binutils-gdb/gdb/gdbserver/mem-break.h

275 lines
8.4 KiB
C
Raw Normal View History

/* Memory breakpoint interfaces for the remote server for GDB.
Copyright (C) 2002-2016 Free Software Foundation, Inc.
Contributed by MontaVista Software.
This file is part of GDB.
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 3 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program. If not, see <http://www.gnu.org/licenses/>. */
#ifndef MEM_BREAK_H
#define MEM_BREAK_H
[GDBserver] Move Z packet defines and type convertion routines to shared code. The Aarch64, MIPS and x86 Linux backends all have Z packet number defines and corresponding protocol number to internal type convertion routines. Factor them all out to gdbserver's core code, so we only have one shared copy. Tested on x86_64 Fedora 20, and also cross built for aarch64-linux-gnu and mips-linux-gnu. gdb/gdbserver/ 2014-05-20 Pedro Alves <palves@redhat.com> * mem-break.h: Include break-common.h. (Z_PACKET_SW_BP, Z_PACKET_HW_BP, Z_PACKET_WRITE_WP) (Z_PACKET_READ_WP, Z_PACKET_ACCESS_WP): New defines. (Z_packet_to_target_hw_bp_type): New declaration. * mem-break.c (Z_packet_to_target_hw_bp_type): New function. * i386-low.c (Z_PACKET_HW_BP, Z_PACKET_WRITE_WP, Z_PACKET_READ_WP) (Z_PACKET_ACCESS_WP): Delete macros. (Z_packet_to_hw_type): Delete function. * i386-low.h: Don't include break-common.h here. (Z_packet_to_hw_type): Delete declaration. * linux-x86-low.c (x86_insert_point, x86_insert_point): Call Z_packet_to_target_hw_bp_type instead of Z_packet_to_hw_type. * win32-i386-low.c (i386_insert_point, i386_remove_point): Call Z_packet_to_target_hw_bp_type instead of Z_packet_to_hw_type. * linux-aarch64-low.c: Don't include break-common.h here. (Z_PACKET_SW_BP, Z_PACKET_HW_BP, Z_PACKET_WRITE_WP) (Z_PACKET_READ_WP, Z_PACKET_ACCESS_WP): Delete macros. (Z_packet_to_target_hw_bp_type): Delete function. * linux-mips-low.c (rsp_bp_type_to_target_hw_bp_type): Delete function. (mips_insert_point, mips_remove_point): Use Z_packet_to_target_hw_bp_type.
2014-05-20 19:24:27 +02:00
#include "break-common.h"
/* Breakpoints are opaque. */
struct breakpoint;
gdb/gdbserver/ 2010-06-01 Pedro Alves <pedro@codesourcery.com> Stan Shebs <stan@codesourcery.com> * Makefile.in (IPA_DEPFILES, extra_libraries): New. (all): Depend on $(extra_libraries). (install-only): Install the IPA. (IPA_OBJS, IPA_LIB): New. (clean): Remove the IPA lib. (IPAGENT_CFLAGS): New. (tracepoint-ipa.o, utils-ipa.o, remote-utils-ipa.o) (regcache-ipa.o, i386-linux-ipa.o, linux-i386-ipa.o) (linux-amd64-ipa.o, amd64-linux-ipa.o): New rules. * linux-amd64-ipa.c, linux-i386-ipa.c: New files. * configure.ac: Check for atomic builtins support in the compiler. (IPA_DEPFILES, extra_libraries): Define. * configure.srv (ipa_obj): Add description. (ipa_i386_linux_regobj, ipa_amd64_linux_regobj): Define. (i[34567]86-*-linux*): Set ipa_obj. (x86_64-*-linux*): Set ipa_obj. * linux-low.c (stabilizing_threads): New. (supports_fast_tracepoints): New. (linux_detach): Stabilize threads before detaching. (handle_tracepoints): Handle internal tracing breakpoints. Assert the lwp is either not stabilizing, or is moving out of a jump pad. (linux_fast_tracepoint_collecting): New. (maybe_move_out_of_jump_pad): New. (enqueue_one_deferred_signal): New. (dequeue_one_deferred_signal): New. (linux_wait_for_event_1): If moving out of a jump pad, defer pending signals to later. (linux_stabilize_threads): New. (linux_wait_1): Check if threads need moving out of jump pads, and do it if so. (stuck_in_jump_pad_callback): New. (move_out_of_jump_pad_callback): New. (lwp_running): New. (linux_resume_one_lwp): Handle moving out of jump pads. (linux_set_resume_request): Dequeue deferred signals. (need_step_over_p): Also step over fast tracepoint jumps. (start_step_over): Also uninsert fast tracepoint jumps. (finish_step_over): Also reinsert fast tracepoint jumps. (linux_install_fast_tracepoint_jump): New. (linux_target_ops): Install linux_stabilize_threads and linux_install_fast_tracepoint_jump_pad. * linux-low.h (linux_target_ops) <get_thread_area, install_fast_tracepoint_jump_pad>: New fields. (struct lwp_info) <collecting_fast_tracepoint, pending_signals_to_report, exit_jump_pad_bkpt>: New fields. (linux_get_thread_area): Declare. * linux-x86-low.c (jump_insn): New. (x86_get_thread_area): New. (append_insns): New. (push_opcode): New. (amd64_install_fast_tracepoint_jump_pad): New. (i386_install_fast_tracepoint_jump_pad): New. (x86_install_fast_tracepoint_jump_pad): New. (the_low_target): Install x86_get_thread_area and x86_install_fast_tracepoint_jump_pad. * mem-break.c (set_raw_breakpoint_at): Use read_inferior_memory. (struct fast_tracepoint_jump): New. (fast_tracepoint_jump_insn): New. (fast_tracepoint_jump_shadow): New. (find_fast_tracepoint_jump_at): New. (fast_tracepoint_jump_here): New. (delete_fast_tracepoint_jump): New. (set_fast_tracepoint_jump): New. (uninsert_fast_tracepoint_jumps_at): New. (reinsert_fast_tracepoint_jumps_at): New. (set_breakpoint_at): Use write_inferior_memory. (uninsert_raw_breakpoint): Use write_inferior_memory. (check_mem_read): Mask out fast tracepoint jumps. (check_mem_write): Mask out fast tracepoint jumps. * mem-break.h (struct fast_tracepoint_jump): Forward declare. (set_fast_tracepoint_jump): Declare. (delete_fast_tracepoint_jump) (fast_tracepoint_jump_here, uninsert_fast_tracepoint_jumps_at) (reinsert_fast_tracepoint_jumps_at): Declare. * regcache.c: Don't compile many functions when building the in-process agent library. (init_register_cache) [IN_PROCESS_AGENT]: Don't allow allocating the register buffer in the heap. (free_register_cache): If the register buffer isn't owned by the regcache, don't free it. (set_register_cache) [IN_PROCESS_AGENT]: Don't re-alocate pre-existing register caches. * remote-utils.c (convert_int_to_ascii): Constify `from' parameter type. (convert_ascii_to_int): : Constify `from' parameter type. (decode_M_packet, decode_X_packet): Replace the `to' parameter by a `to_p' pointer to pointer parameter. If TO_P is NULL, malloc the needed buffer in-place. (relocate_instruction): New. * server.c (handle_query) <qSymbols>: If the target supports tracepoints, give it a chance of looking up symbols. Report support for fast tracepoints. (handle_status): Stabilize threads. (process_serial_event): Adjust. * server.h (struct fast_tracepoint_jump): Forward declare. (struct process_info) <fast_tracepoint_jumps>: New field. (convert_ascii_to_int, convert_int_to_ascii): Adjust. (decode_X_packet, decode_M_packet): Adjust. (relocate_instruction): Declare. (in_process_agent_loaded): Declare. (tracepoint_look_up_symbols): Declare. (struct fast_tpoint_collect_status): Declare. (fast_tracepoint_collecting): Declare. (force_unlock_trace_buffer): Declare. (handle_tracepoint_bkpts): Declare. (initialize_low_tracepoint) (supply_fast_tracepoint_registers) [IN_PROCESS_AGENT]: Declare. * target.h (struct target_ops) <stabilize_threads, install_fast_tracepoint_jump_pad>: New fields. (stabilize_threads, install_fast_tracepoint_jump_pad): New. * tracepoint.c [HAVE_MALLOC_H]: Include malloc.h. [HAVE_STDINT_H]: Include stdint.h. (trace_debug_1): Rename to ... (trace_vdebug): ... this. (trace_debug): Rename to ... (trace_debug_1): ... this. Add `level' parameter. (trace_debug): New. (ATTR_USED, ATTR_NOINLINE): New. (IP_AGENT_EXPORT): New. (gdb_tp_heap_buffer, gdb_jump_pad_buffer, gdb_jump_pad_buffer_end) (collecting, gdb_collect, stop_tracing, flush_trace_buffer) (about_to_request_buffer_space, trace_buffer_is_full) (stopping_tracepoint, expr_eval_result, error_tracepoint) (tracepoints, tracing, trace_buffer_ctrl, trace_buffer_ctrl_curr) (trace_buffer_lo, trace_buffer_hi, traceframe_read_count) (traceframe_write_count, traceframes_created) (trace_state_variables) New renaming defines. (struct ipa_sym_addresses): New. (STRINGIZE_1, STRINGIZE, IPA_SYM): New. (symbol_list): New. (ipa_sym_addrs): New. (all_tracepoint_symbols_looked_up): New. (in_process_agent_loaded): New. (write_e_ipa_not_loaded): New. (maybe_write_ipa_not_loaded): New. (tracepoint_look_up_symbols): New. (debug_threads) [IN_PROCESS_AGENT]: New. (read_inferior_memory) [IN_PROCESS_AGENT]: New. (UNKNOWN_SIDE_EFFECTS): New. (stop_tracing): New. (flush_trace_buffer): New. (stop_tracing_bkpt): New. (flush_trace_buffer_bkpt): New. (read_inferior_integer): New. (read_inferior_uinteger): New. (read_inferior_data_pointer): New. (write_inferior_data_pointer): New. (write_inferior_integer): New. (write_inferior_uinteger): New. (struct collect_static_trace_data_action): Delete. (enum tracepoint_type): New. (struct tracepoint) <type>: New field `type'. <actions_str, step_actions, step_actions_str>: Only include in GDBserver. <orig_size, obj_addr_on_target, adjusted_insn_addr> <adjusted_insn_addr_end, jump_pad, jump_pad_end>: New fields. (tracepoints): Use IP_AGENT_EXPORT. (last_tracepoint): Don't include in the IPA. (stopping_tracepoint): Use IP_AGENT_EXPORT. (trace_buffer_is_full): Use IP_AGENT_EXPORT. (alloced_trace_state_variables): New. (trace_state_variables): Use IP_AGENT_EXPORT. (traceframe_t): Delete unused variable. (circular_trace_buffer): Don't include in the IPA. (trace_buffer_start): Delete. (struct trace_buffer_control): New. (trace_buffer_free): Delete. (struct ipa_trace_buffer_control): New. (GDBSERVER_FLUSH_COUNT_MASK, GDBSERVER_FLUSH_COUNT_MASK_PREV) (GDBSERVER_FLUSH_COUNT_MASK_CURR, GDBSERVER_UPDATED_FLUSH_COUNT_BIT): New. (trace_buffer_ctrl): New. (TRACE_BUFFER_CTRL_CURR): New. (trace_buffer_start, trace_buffer_free, trace_buffer_end_free): Reimplement as macros. (trace_buffer_wrap): Delete. (traceframe_write_count, traceframe_read_count) (traceframes_created, tracing): Use IP_AGENT_EXPORT. (struct tracepoint_hit_ctx) <type>: New field. (struct fast_tracepoint_ctx): New. (memory_barrier): New. (cmpxchg): New. (record_tracepoint_error): Update atomically in the IPA. (clear_inferior_trace_buffer): New. (about_to_request_buffer_space): New. (trace_buffer_alloc): Handle GDBserver and inferior simulatenous updating the same buffer. (add_tracepoint): Default the tracepoint's type to trap tracepoint, and orig_size to -1. (get_trace_state_variable) [IN_PROCESS_AGENT]: Handle allocated internal variables. (create_trace_state_variable): New parameter `gdb'. Handle it. (clear_installed_tracepoints): Clear fast tracepoint jumps. (cmd_qtdp): Handle fast tracepoints. (cmd_qtdv): Adjust. (max_jump_pad_size): New. (gdb_jump_pad_head): New. (get_jump_space_head): New. (claim_jump_space): New. (sort_tracepoints): New. (MAX_JUMP_SIZE): New. (cmd_qtstart): Handle fast tracepoints. Sync tracepoints with the IPA. (stop_tracing) [IN_PROCESS_AGENT]: Don't include the tdisconnected support. Upload fast traceframes, and delete internal IPA breakpoints. (stop_tracing_handler): New. (flush_trace_buffer_handler): New. (cmd_qtstop): Upload fast tracepoints. (response_tracepoint): Handle fast tracepoints. (tracepoint_finished_step): Upload fast traceframes. Set the tracepoint hit context's tracepoint type. (handle_tracepoint_bkpts): New. (tracepoint_was_hit): Set the tracepoint hit context's tracepoint type. Add comment about fast tracepoints. (collect_data_at_tracepoint) [IN_PROCESS_AGENT]: Don't access the non-existing action_str field. (get_context_regcache): Handle fast tracepoints. (do_action_at_tracepoint) [!IN_PROCESS_AGENT]: Don't write the PC to the regcache. (fast_tracepoint_from_jump_pad_address): New. (fast_tracepoint_from_ipa_tpoint_address): New. (collecting_t): New. (force_unlock_trace_buffer): New. (fast_tracepoint_collecting): New. (collecting): New. (gdb_collect): New. (write_inferior_data_ptr): New. (target_tp_heap): New. (target_malloc): New. (download_agent_expr): New. (UALIGN): New. (download_tracepoints): New. (download_trace_state_variables): New. (upload_fast_traceframes): New. (IPA_FIRST_TRACEFRAME): New. (IPA_NEXT_TRACEFRAME_1): New. (IPA_NEXT_TRACEFRAME): New. [IN_PROCESS_AGENT]: Include sys/mman.h and fcntl.h. [IN_PROCESS_AGENT] (gdb_tp_heap_buffer, gdb_jump_pad_buffer) (gdb_jump_pad_buffer_end): New. [IN_PROCESS_AGENT] (initialize_tracepoint_ftlib): New. (initialize_tracepoint): Adjust. [IN_PROCESS_AGENT]: Allocate the IPA heap, and jump pad scratch buffer. Initialize the low module. * utils.c (PREFIX, TOOLNAME): New. (malloc_failure): Use PREFIX. (error): In the IPA, an error causes an exit. (fatal, warning): Use PREFIX. (internal_error): Use TOOLNAME. (NUMCELLS): Increase to 10. * configure, config.in: Regenerate. gdb/ 2010-06-01 Pedro Alves <pedro@codesourcery.com> * NEWS: Mention gdbserver fast tracepoints support. gdb/doc/ 2010-06-01 Pedro Alves <pedro@codesourcery.com> * gdb.texinfo (Set Tracepoints): Mention tracepoints support in gdbserver, and add cross reference. (Tracepoints support in gdbserver): New subsection.
2010-06-01 15:20:52 +02:00
struct fast_tracepoint_jump;
[GDBserver] Make Zx/zx packet handling idempotent. This patch fixes hardware breakpoint regressions exposed by my fix for "PR breakpoints/7143 - Watchpoint does not trigger when first set", at https://sourceware.org/ml/gdb-patches/2014-03/msg00167.html The testsuite caught them on Linux/x86_64, at least. gdb.sum: gdb.sum: FAIL: gdb.base/hbreak2.exp: next over recursive call FAIL: gdb.base/hbreak2.exp: backtrace from factorial(5.1) FAIL: gdb.base/hbreak2.exp: continue until exit at recursive next test gdb.log: (gdb) next Program received signal SIGTRAP, Trace/breakpoint trap. factorial (value=4) at ../../../src/gdb/testsuite/gdb.base/break.c:113 113 if (value > 1) { /* set breakpoint 7 here */ (gdb) FAIL: gdb.base/hbreak2.exp: next over recursive call Actually, that patch just exposed a latent issue to "breakpoints always-inserted off" mode, not really caused it. After that patch, GDB no longer removes breakpoints at each internal event, thus making some scenarios behave like breakpoint always-inserted on. The bug is easy to trigger with always-inserted on. The issue is that since the target-side breakpoint conditions support, if the stub/server supports evaluating breakpoint conditions on the target side, then GDB is sending duplicate Zx packets to the target without removing them before, and GDBserver is not really expecting that for Z packets other than Z0/z0. E.g., with "set breakpoint always-inserted on" and "set debug remote 1": (gdb) b main Sending packet: $m410943,1#ff...Packet received: 48 Breakpoint 4 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z0,410943,1#48...Packet received: OK ^^^^^^^^^^^^ (gdb) b main Note: breakpoint 4 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Breakpoint 5 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z0,410943,1#48...Packet received: OK ^^^^^^^^^^^^ (gdb) b main Note: breakpoints 4 and 5 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Breakpoint 6 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z0,410943,1#48...Packet received: OK ^^^^^^^^^^^^ (gdb) del Delete all breakpoints? (y or n) y Sending packet: $Z0,410943,1#48...Packet received: OK Sending packet: $Z0,410943,1#48...Packet received: OK Sending packet: $z0,410943,1#68...Packet received: OK And for Z1, similarly: (gdb) hbreak main Sending packet: $m410943,1#ff...Packet received: 48 Hardware assisted breakpoint 4 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ Packet Z1 (hardware-breakpoint) is supported (gdb) hbreak main Note: breakpoint 4 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Hardware assisted breakpoint 5 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ (gdb) hbreak main Note: breakpoints 4 and 5 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Hardware assisted breakpoint 6 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ (gdb) del Delete all breakpoints? (y or n) y Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ Sending packet: $z1,410943,1#69...Packet received: OK ^^^^^^^^^^^^ So GDB sent a bunch of Z1 packets, and then when finally removing the breakpoint, only one z1 packet was sent. On the GDBserver side (with monitor set debug-hw-points 1), in the Z1 case, we see: $ ./gdbserver :9999 ./gdbserver Process ./gdbserver created; pid = 8629 Listening on port 9999 Remote debugging from host 127.0.0.1 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=1 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=2 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=3 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=4 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=5 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 remove_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=4 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 That's one insert_watchpoint call for each Z1 packet, and then one remove_watchpoint call for the z1 packet. Notice how ref.count increased for each insert_watchpoint call, and then in the end, after GDB told GDBserver to forget about the hardware breakpoint, GDBserver ends with the the first debug register still with ref.count=4! IOW, the hardware breakpoint is left armed on the target, while on the GDB end it's gone. If the program happens to execute 0x410943 afterwards, then the CPU traps, GDBserver reports the trap to GDB, and GDB not having a breakpoint set at that address anymore, reports to the user a spurious SIGTRAP. This is exactly what is happening in the hbreak2.exp test, though in that case, it's a shared library event that triggers a breakpoint_re_set, when breakpoints are still inserted (because nowadays GDB doesn't remove breakpoints while handling internal events), and that recreates breakpoint locations, which likewise forces breakpoint reinsertion and Zx packet resends... That is a lot of bogus Zx duplication that should possibly be addressed on the GDB side. GDB resends Zx packets because the way to change the target-side condition, is to resend the breakpoint to the server with the new condition. (That's an option in the packet: e.g., "Z1,410943,1;X3,220027" for "hbreak main if 0". The packets in the examples above are shorter because the breakpoints don't have conditions attached). GDB doesn't remove the breakpoint first before reinserting it because that'd be bad for non-stop, as it'd open a window where the inferior could miss the breakpoint. The conditions actually haven't changed between the resends, but GDB isn't smart enough to realize that. (TBC, if the target doesn't support target-side conditions, then GDB doesn't trigger these resends (init_bp_location calls mark_breakpoint_location_modified, and that does nothing if condition evaluation is on the host side. The resends are caused by the 'loc->condition_changed = condition_modified.' line.) But, even if GDB was made smarter, GDBserver should really still handle the resends anyway. So target-side conditions also aren't really to blame. The documentation of the Z/z packets says: "To avoid potential problems with duplicate packets, the operations should be implemented in an idempotent way." As such, we may want to fix GDB, but we should definitely fix GDBserver. The fix is a prerequisite for target-side conditions on hardware breakpoints anyway (and while at it, on watchpoints too). GDBserver indeed already treats duplicate Z0 packets in an idempotent way. mem-break.c has the concept of high-level and low-level breakpoints, somewhat similar to GDB's split of breakpoints vs breakpoint locations, and keeps track of multiple breakpoints referencing the same address/location, for the case of an internal GDBserver breakpoint or a tracepoint being set at the same address as a GDB breakpoint. But, it only allows GDB to ever contribute one reference to a software breakpoint location. IOW, if gdbserver sees a Z0 packet for the same address where it already had a GDB breakpoint set, then GDBserver won't create another high-level GDB breakpoint. However, mem-break.c only tracks GDB Z0 breakpoints. The same logic should apply to all kinds of Zx packets. Currently, gdbserver passes down each duplicate Zx (other than Z0) request directly to the target->insert_point routine. The x86 watchpoint support itself refcounts watchpoint / hw breakpoint requests, to handle overlapping watchpoints, and save debug registers. But that code doesn't (and really shouldn't) handle the duplicate requests, assuming that for each insert there will be a corresponding remove. So the fix is to generalize mem-break.c to track all kinds of Zx breakpoints, and filter out duplicates. As mentioned, this ends up adding support for target-side conditions on hardware breakpoints and watchpoints too (though GDB itself doesn't support the latter yet). Probably the least obvious change in the patch is that it kind of turns the breakpoint insert/remove APIs inside out. Before, the target methods were only called for GDB breakpoints. The internal breakpoint set/delete methods inserted memory breakpoints directly bypassing the insert/remove target methods. That's not good when the target should use a debug API to set software breakpoints, instead of relying on GDBserver patching memory with breakpoint instructions, as is the case of NTO. Now removal/insertion of all kinds of breakpoints/watchpoints, either internal, or from GDB, always go through the target methods. The insert_point/remove_point methods no longer get passed a Z packet type, but an internal/raw breakpoint type. They're also passed a pointer to the raw breakpoint itself (note that's still opaque outside mem-break.c), so that insert_memory_breakpoint / remove_memory_breakpoint have access to the breakpoint's shadow buffer. I first tried passing down a new structure based on GDB's "struct bp_target_info" (actually with that name exactly), but then decided against it as unnecessary complication. As software/memory breakpoints work by poking at memory, when setting a GDB Z0 breakpoint (but not internal breakpoints, as those can assume the conditions are already right), we need to tell the target to prepare to access memory (which on Linux means stop threads). If that operation fails, we need to return error to GDB. Seeing an error, if this is the first breakpoint of that type that GDB tries to insert, GDB would then assume the breakpoint type is supported, but it may actually not be. So we need to check whether the type is supported at all before preparing to access memory. And to solve that, the patch adds a new target->supports_z_point_type method that is called before actually trying to insert the breakpoint. Other than that, hopefully the change is more or less obvious. New test added that exercises the hbreak2.exp regression in a more direct way, without relying on a breakpoint re-set happening before main is reached. Tested by building GDBserver for: aarch64-linux-gnu arm-linux-gnueabihf i686-pc-linux-gnu i686-w64-mingw32 m68k-linux-gnu mips-linux-gnu mips-uclinux nios2-linux-gnu powerpc-linux-gnu sh-linux-gnu tilegx-unknown-linux-gnu x86_64-redhat-linux x86_64-w64-mingw32 And also regression tested on x86_64 Fedora 20. gdb/gdbserver/ 2014-05-20 Pedro Alves <palves@redhat.com> * linux-aarch64-low.c (aarch64_insert_point) (aarch64_remove_point): No longer check whether the type is supported here. Adjust to new interface. (the_low_target): Install aarch64_supports_z_point_type as supports_z_point_type method. * linux-arm-low.c (raw_bkpt_type_to_arm_hwbp_type): New function. (arm_linux_hw_point_initialize): Take an enum raw_bkpt_type instead of a Z packet char. Adjust. (arm_supports_z_point_type): New function. (arm_insert_point, arm_remove_point): Adjust to new interface. (the_low_target): Install arm_supports_z_point_type. * linux-crisv32-low.c (cris_supports_z_point_type): New function. (cris_insert_point, cris_remove_point): Adjust to new interface. Don't check whether the type is supported here. (the_low_target): Install cris_supports_z_point_type. * linux-low.c (linux_supports_z_point_type): New function. (linux_insert_point, linux_remove_point): Adjust to new interface. * linux-low.h (struct linux_target_ops) <insert_point, remove_point>: Take an enum raw_bkpt_type instead of a char. Add raw_breakpoint pointer parameter. <supports_z_point_type>: New method. * linux-mips-low.c (mips_supports_z_point_type): New function. (mips_insert_point, mips_remove_point): Adjust to new interface. Use mips_supports_z_point_type. (the_low_target): Install mips_supports_z_point_type. * linux-ppc-low.c (the_low_target): Install NULL as supports_z_point_type method. * linux-s390-low.c (the_low_target): Install NULL as supports_z_point_type method. * linux-sparc-low.c (the_low_target): Install NULL as supports_z_point_type method. * linux-x86-low.c (x86_supports_z_point_type): New function. (x86_insert_point): Adjust to new insert_point interface. Use insert_memory_breakpoint. Adjust to new i386_low_insert_watchpoint interface. (x86_remove_point): Adjust to remove_point interface. Use remove_memory_breakpoint. Adjust to new i386_low_remove_watchpoint interface. (the_low_target): Install x86_supports_z_point_type. * lynx-low.c (lynx_target_ops): Install NULL as supports_z_point_type callback. * nto-low.c (nto_supports_z_point_type): New. (nto_insert_point, nto_remove_point): Adjust to new interface. (nto_target_ops): Install nto_supports_z_point_type. * mem-break.c: Adjust intro comment. (struct raw_breakpoint) <raw_type, size>: New fields. <inserted>: Update comment. <shlib_disabled>: Delete field. (enum bkpt_type) <gdb_breakpoint>: Delete value. <gdb_breakpoint_Z0, gdb_breakpoint_Z1, gdb_breakpoint_Z2, gdb_breakpoint_Z3, gdb_breakpoint_Z4>: New values. (raw_bkpt_type_to_target_hw_bp_type): New function. (find_enabled_raw_code_breakpoint_at): New function. (find_raw_breakpoint_at): New type and size parameters. Use them. (insert_memory_breakpoint): New function, based off set_raw_breakpoint_at. (remove_memory_breakpoint): New function. (set_raw_breakpoint_at): Reimplement. (set_breakpoint): New, based on set_breakpoint_at. (set_breakpoint_at): Reimplement. (delete_raw_breakpoint): Go through the_target->remove_point instead of assuming memory breakpoints. (find_gdb_breakpoint_at): Delete. (Z_packet_to_bkpt_type, Z_packet_to_raw_bkpt_type): New functions. (find_gdb_breakpoint): New function. (set_gdb_breakpoint_at): Delete. (z_type_supported): New function. (set_gdb_breakpoint_1): New function, loosely based off set_gdb_breakpoint_at. (check_gdb_bp_preconditions, set_gdb_breakpoint): New functions. (delete_gdb_breakpoint_at): Delete. (delete_gdb_breakpoint_1): New function, loosely based off delete_gdb_breakpoint_at. (delete_gdb_breakpoint): New function. (clear_gdb_breakpoint_conditions): Rename to ... (clear_breakpoint_conditions): ... this. Don't handle a NULL breakpoint. (add_condition_to_breakpoint): Make static. (add_breakpoint_condition): Take a struct breakpoint pointer instead of an address. Adjust. (gdb_condition_true_at_breakpoint): Rename to ... (gdb_condition_true_at_breakpoint_z_type): ... this, and add z_type parameter. (gdb_condition_true_at_breakpoint): Reimplement. (add_breakpoint_commands): Take a struct breakpoint pointer instead of an address. Adjust. (gdb_no_commands_at_breakpoint): Rename to ... (gdb_no_commands_at_breakpoint_z_type): ... this. Add z_type parameter. Return true if no breakpoint was found. Change debug output. (gdb_no_commands_at_breakpoint): Reimplement. (run_breakpoint_commands): Rename to ... (run_breakpoint_commands_z_type): ... this. Add z_type parameter, and change return type to boolean. (run_breakpoint_commands): New function. (gdb_breakpoint_here): Also check for Z1 breakpoints. (uninsert_raw_breakpoint): Don't try to reinsert a disabled breakpoint. Go through the_target->remove_point instead of assuming memory breakpoint. (uninsert_breakpoints_at, uninsert_all_breakpoints): Uninsert software and hardware breakpoints. (reinsert_raw_breakpoint): Go through the_target->insert_point instead of assuming memory breakpoint. (reinsert_breakpoints_at, reinsert_all_breakpoints): Reinsert software and hardware breakpoints. (check_breakpoints, breakpoint_here, breakpoint_inserted_here): Check both software and hardware breakpoints. (validate_inserted_breakpoint): Assert the breakpoint is a software breakpoint. Set the inserted flag to -1 instead of setting shlib_disabled. (delete_disabled_breakpoints): Adjust. (validate_breakpoints): Only validate software breakpoints. Adjust to inserted flag change. (check_mem_read, check_mem_write): Skip breakpoint types other than software breakpoints. Adjust to inserted flag change. * mem-break.h (enum raw_bkpt_type): New enum. (raw_breakpoint, struct process_info): Forward declare. (Z_packet_to_target_hw_bp_type): Delete declaration. (raw_bkpt_type_to_target_hw_bp_type, Z_packet_to_raw_bkpt_type) (set_gdb_breakpoint, delete_gdb_breakpoint) (clear_breakpoint_conditions): New declarations. (set_gdb_breakpoint_at, clear_gdb_breakpoint_conditions): Delete. (breakpoint_inserted_here): Update comment. (add_breakpoint_condition, add_breakpoint_commands): Replace address parameter with a breakpoint pointer parameter. (gdb_breakpoint_here): Update comment. (delete_gdb_breakpoint_at): Delete. (insert_memory_breakpoint, remove_memory_breakpoint): Declare. * server.c (process_point_options): Take a struct breakpoint pointer instead of an address. Adjust. (process_serial_event) <Z/z packets>: Use set_gdb_breakpoint and delete_gdb_breakpoint. * spu-low.c (spu_target_ops): Install NULL as supports_z_point_type method. * target.h: Include mem-break.h. (struct target_ops) <prepare_to_access_memory>: Update comment. <supports_z_point_type>: New field. <insert_point, remove_point>: Take an enum raw_bkpt_type argument instead of a char. Also take a raw breakpoint pointer. * win32-arm-low.c (the_low_target): Install NULL as supports_z_point_type. * win32-i386-low.c (i386_supports_z_point_type): New function. (i386_insert_point, i386_remove_point): Adjust to new interface. (the_low_target): Install i386_supports_z_point_type. * win32-low.c (win32_supports_z_point_type): New function. (win32_insert_point, win32_remove_point): Adjust to new interface. (win32_target_ops): Install win32_supports_z_point_type. * win32-low.h (struct win32_target_ops): <supports_z_point_type>: New method. <insert_point, remove_point>: Take an enum raw_bkpt_type argument instead of a char. Also take a raw breakpoint pointer. gdb/testsuite/ 2014-05-20 Pedro Alves <palves@redhat.com> * gdb.base/break-idempotent.c: New file. * gdb.base/break-idempotent.exp: New file.
2014-05-20 19:24:28 +02:00
struct raw_breakpoint;
struct process_info;
[GDBserver] Move Z packet defines and type convertion routines to shared code. The Aarch64, MIPS and x86 Linux backends all have Z packet number defines and corresponding protocol number to internal type convertion routines. Factor them all out to gdbserver's core code, so we only have one shared copy. Tested on x86_64 Fedora 20, and also cross built for aarch64-linux-gnu and mips-linux-gnu. gdb/gdbserver/ 2014-05-20 Pedro Alves <palves@redhat.com> * mem-break.h: Include break-common.h. (Z_PACKET_SW_BP, Z_PACKET_HW_BP, Z_PACKET_WRITE_WP) (Z_PACKET_READ_WP, Z_PACKET_ACCESS_WP): New defines. (Z_packet_to_target_hw_bp_type): New declaration. * mem-break.c (Z_packet_to_target_hw_bp_type): New function. * i386-low.c (Z_PACKET_HW_BP, Z_PACKET_WRITE_WP, Z_PACKET_READ_WP) (Z_PACKET_ACCESS_WP): Delete macros. (Z_packet_to_hw_type): Delete function. * i386-low.h: Don't include break-common.h here. (Z_packet_to_hw_type): Delete declaration. * linux-x86-low.c (x86_insert_point, x86_insert_point): Call Z_packet_to_target_hw_bp_type instead of Z_packet_to_hw_type. * win32-i386-low.c (i386_insert_point, i386_remove_point): Call Z_packet_to_target_hw_bp_type instead of Z_packet_to_hw_type. * linux-aarch64-low.c: Don't include break-common.h here. (Z_PACKET_SW_BP, Z_PACKET_HW_BP, Z_PACKET_WRITE_WP) (Z_PACKET_READ_WP, Z_PACKET_ACCESS_WP): Delete macros. (Z_packet_to_target_hw_bp_type): Delete function. * linux-mips-low.c (rsp_bp_type_to_target_hw_bp_type): Delete function. (mips_insert_point, mips_remove_point): Use Z_packet_to_target_hw_bp_type.
2014-05-20 19:24:27 +02:00
#define Z_PACKET_SW_BP '0'
#define Z_PACKET_HW_BP '1'
#define Z_PACKET_WRITE_WP '2'
#define Z_PACKET_READ_WP '3'
#define Z_PACKET_ACCESS_WP '4'
[GDBserver] Make Zx/zx packet handling idempotent. This patch fixes hardware breakpoint regressions exposed by my fix for "PR breakpoints/7143 - Watchpoint does not trigger when first set", at https://sourceware.org/ml/gdb-patches/2014-03/msg00167.html The testsuite caught them on Linux/x86_64, at least. gdb.sum: gdb.sum: FAIL: gdb.base/hbreak2.exp: next over recursive call FAIL: gdb.base/hbreak2.exp: backtrace from factorial(5.1) FAIL: gdb.base/hbreak2.exp: continue until exit at recursive next test gdb.log: (gdb) next Program received signal SIGTRAP, Trace/breakpoint trap. factorial (value=4) at ../../../src/gdb/testsuite/gdb.base/break.c:113 113 if (value > 1) { /* set breakpoint 7 here */ (gdb) FAIL: gdb.base/hbreak2.exp: next over recursive call Actually, that patch just exposed a latent issue to "breakpoints always-inserted off" mode, not really caused it. After that patch, GDB no longer removes breakpoints at each internal event, thus making some scenarios behave like breakpoint always-inserted on. The bug is easy to trigger with always-inserted on. The issue is that since the target-side breakpoint conditions support, if the stub/server supports evaluating breakpoint conditions on the target side, then GDB is sending duplicate Zx packets to the target without removing them before, and GDBserver is not really expecting that for Z packets other than Z0/z0. E.g., with "set breakpoint always-inserted on" and "set debug remote 1": (gdb) b main Sending packet: $m410943,1#ff...Packet received: 48 Breakpoint 4 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z0,410943,1#48...Packet received: OK ^^^^^^^^^^^^ (gdb) b main Note: breakpoint 4 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Breakpoint 5 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z0,410943,1#48...Packet received: OK ^^^^^^^^^^^^ (gdb) b main Note: breakpoints 4 and 5 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Breakpoint 6 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z0,410943,1#48...Packet received: OK ^^^^^^^^^^^^ (gdb) del Delete all breakpoints? (y or n) y Sending packet: $Z0,410943,1#48...Packet received: OK Sending packet: $Z0,410943,1#48...Packet received: OK Sending packet: $z0,410943,1#68...Packet received: OK And for Z1, similarly: (gdb) hbreak main Sending packet: $m410943,1#ff...Packet received: 48 Hardware assisted breakpoint 4 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ Packet Z1 (hardware-breakpoint) is supported (gdb) hbreak main Note: breakpoint 4 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Hardware assisted breakpoint 5 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ (gdb) hbreak main Note: breakpoints 4 and 5 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Hardware assisted breakpoint 6 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ (gdb) del Delete all breakpoints? (y or n) y Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ Sending packet: $z1,410943,1#69...Packet received: OK ^^^^^^^^^^^^ So GDB sent a bunch of Z1 packets, and then when finally removing the breakpoint, only one z1 packet was sent. On the GDBserver side (with monitor set debug-hw-points 1), in the Z1 case, we see: $ ./gdbserver :9999 ./gdbserver Process ./gdbserver created; pid = 8629 Listening on port 9999 Remote debugging from host 127.0.0.1 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=1 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=2 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=3 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=4 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=5 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 remove_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=4 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 That's one insert_watchpoint call for each Z1 packet, and then one remove_watchpoint call for the z1 packet. Notice how ref.count increased for each insert_watchpoint call, and then in the end, after GDB told GDBserver to forget about the hardware breakpoint, GDBserver ends with the the first debug register still with ref.count=4! IOW, the hardware breakpoint is left armed on the target, while on the GDB end it's gone. If the program happens to execute 0x410943 afterwards, then the CPU traps, GDBserver reports the trap to GDB, and GDB not having a breakpoint set at that address anymore, reports to the user a spurious SIGTRAP. This is exactly what is happening in the hbreak2.exp test, though in that case, it's a shared library event that triggers a breakpoint_re_set, when breakpoints are still inserted (because nowadays GDB doesn't remove breakpoints while handling internal events), and that recreates breakpoint locations, which likewise forces breakpoint reinsertion and Zx packet resends... That is a lot of bogus Zx duplication that should possibly be addressed on the GDB side. GDB resends Zx packets because the way to change the target-side condition, is to resend the breakpoint to the server with the new condition. (That's an option in the packet: e.g., "Z1,410943,1;X3,220027" for "hbreak main if 0". The packets in the examples above are shorter because the breakpoints don't have conditions attached). GDB doesn't remove the breakpoint first before reinserting it because that'd be bad for non-stop, as it'd open a window where the inferior could miss the breakpoint. The conditions actually haven't changed between the resends, but GDB isn't smart enough to realize that. (TBC, if the target doesn't support target-side conditions, then GDB doesn't trigger these resends (init_bp_location calls mark_breakpoint_location_modified, and that does nothing if condition evaluation is on the host side. The resends are caused by the 'loc->condition_changed = condition_modified.' line.) But, even if GDB was made smarter, GDBserver should really still handle the resends anyway. So target-side conditions also aren't really to blame. The documentation of the Z/z packets says: "To avoid potential problems with duplicate packets, the operations should be implemented in an idempotent way." As such, we may want to fix GDB, but we should definitely fix GDBserver. The fix is a prerequisite for target-side conditions on hardware breakpoints anyway (and while at it, on watchpoints too). GDBserver indeed already treats duplicate Z0 packets in an idempotent way. mem-break.c has the concept of high-level and low-level breakpoints, somewhat similar to GDB's split of breakpoints vs breakpoint locations, and keeps track of multiple breakpoints referencing the same address/location, for the case of an internal GDBserver breakpoint or a tracepoint being set at the same address as a GDB breakpoint. But, it only allows GDB to ever contribute one reference to a software breakpoint location. IOW, if gdbserver sees a Z0 packet for the same address where it already had a GDB breakpoint set, then GDBserver won't create another high-level GDB breakpoint. However, mem-break.c only tracks GDB Z0 breakpoints. The same logic should apply to all kinds of Zx packets. Currently, gdbserver passes down each duplicate Zx (other than Z0) request directly to the target->insert_point routine. The x86 watchpoint support itself refcounts watchpoint / hw breakpoint requests, to handle overlapping watchpoints, and save debug registers. But that code doesn't (and really shouldn't) handle the duplicate requests, assuming that for each insert there will be a corresponding remove. So the fix is to generalize mem-break.c to track all kinds of Zx breakpoints, and filter out duplicates. As mentioned, this ends up adding support for target-side conditions on hardware breakpoints and watchpoints too (though GDB itself doesn't support the latter yet). Probably the least obvious change in the patch is that it kind of turns the breakpoint insert/remove APIs inside out. Before, the target methods were only called for GDB breakpoints. The internal breakpoint set/delete methods inserted memory breakpoints directly bypassing the insert/remove target methods. That's not good when the target should use a debug API to set software breakpoints, instead of relying on GDBserver patching memory with breakpoint instructions, as is the case of NTO. Now removal/insertion of all kinds of breakpoints/watchpoints, either internal, or from GDB, always go through the target methods. The insert_point/remove_point methods no longer get passed a Z packet type, but an internal/raw breakpoint type. They're also passed a pointer to the raw breakpoint itself (note that's still opaque outside mem-break.c), so that insert_memory_breakpoint / remove_memory_breakpoint have access to the breakpoint's shadow buffer. I first tried passing down a new structure based on GDB's "struct bp_target_info" (actually with that name exactly), but then decided against it as unnecessary complication. As software/memory breakpoints work by poking at memory, when setting a GDB Z0 breakpoint (but not internal breakpoints, as those can assume the conditions are already right), we need to tell the target to prepare to access memory (which on Linux means stop threads). If that operation fails, we need to return error to GDB. Seeing an error, if this is the first breakpoint of that type that GDB tries to insert, GDB would then assume the breakpoint type is supported, but it may actually not be. So we need to check whether the type is supported at all before preparing to access memory. And to solve that, the patch adds a new target->supports_z_point_type method that is called before actually trying to insert the breakpoint. Other than that, hopefully the change is more or less obvious. New test added that exercises the hbreak2.exp regression in a more direct way, without relying on a breakpoint re-set happening before main is reached. Tested by building GDBserver for: aarch64-linux-gnu arm-linux-gnueabihf i686-pc-linux-gnu i686-w64-mingw32 m68k-linux-gnu mips-linux-gnu mips-uclinux nios2-linux-gnu powerpc-linux-gnu sh-linux-gnu tilegx-unknown-linux-gnu x86_64-redhat-linux x86_64-w64-mingw32 And also regression tested on x86_64 Fedora 20. gdb/gdbserver/ 2014-05-20 Pedro Alves <palves@redhat.com> * linux-aarch64-low.c (aarch64_insert_point) (aarch64_remove_point): No longer check whether the type is supported here. Adjust to new interface. (the_low_target): Install aarch64_supports_z_point_type as supports_z_point_type method. * linux-arm-low.c (raw_bkpt_type_to_arm_hwbp_type): New function. (arm_linux_hw_point_initialize): Take an enum raw_bkpt_type instead of a Z packet char. Adjust. (arm_supports_z_point_type): New function. (arm_insert_point, arm_remove_point): Adjust to new interface. (the_low_target): Install arm_supports_z_point_type. * linux-crisv32-low.c (cris_supports_z_point_type): New function. (cris_insert_point, cris_remove_point): Adjust to new interface. Don't check whether the type is supported here. (the_low_target): Install cris_supports_z_point_type. * linux-low.c (linux_supports_z_point_type): New function. (linux_insert_point, linux_remove_point): Adjust to new interface. * linux-low.h (struct linux_target_ops) <insert_point, remove_point>: Take an enum raw_bkpt_type instead of a char. Add raw_breakpoint pointer parameter. <supports_z_point_type>: New method. * linux-mips-low.c (mips_supports_z_point_type): New function. (mips_insert_point, mips_remove_point): Adjust to new interface. Use mips_supports_z_point_type. (the_low_target): Install mips_supports_z_point_type. * linux-ppc-low.c (the_low_target): Install NULL as supports_z_point_type method. * linux-s390-low.c (the_low_target): Install NULL as supports_z_point_type method. * linux-sparc-low.c (the_low_target): Install NULL as supports_z_point_type method. * linux-x86-low.c (x86_supports_z_point_type): New function. (x86_insert_point): Adjust to new insert_point interface. Use insert_memory_breakpoint. Adjust to new i386_low_insert_watchpoint interface. (x86_remove_point): Adjust to remove_point interface. Use remove_memory_breakpoint. Adjust to new i386_low_remove_watchpoint interface. (the_low_target): Install x86_supports_z_point_type. * lynx-low.c (lynx_target_ops): Install NULL as supports_z_point_type callback. * nto-low.c (nto_supports_z_point_type): New. (nto_insert_point, nto_remove_point): Adjust to new interface. (nto_target_ops): Install nto_supports_z_point_type. * mem-break.c: Adjust intro comment. (struct raw_breakpoint) <raw_type, size>: New fields. <inserted>: Update comment. <shlib_disabled>: Delete field. (enum bkpt_type) <gdb_breakpoint>: Delete value. <gdb_breakpoint_Z0, gdb_breakpoint_Z1, gdb_breakpoint_Z2, gdb_breakpoint_Z3, gdb_breakpoint_Z4>: New values. (raw_bkpt_type_to_target_hw_bp_type): New function. (find_enabled_raw_code_breakpoint_at): New function. (find_raw_breakpoint_at): New type and size parameters. Use them. (insert_memory_breakpoint): New function, based off set_raw_breakpoint_at. (remove_memory_breakpoint): New function. (set_raw_breakpoint_at): Reimplement. (set_breakpoint): New, based on set_breakpoint_at. (set_breakpoint_at): Reimplement. (delete_raw_breakpoint): Go through the_target->remove_point instead of assuming memory breakpoints. (find_gdb_breakpoint_at): Delete. (Z_packet_to_bkpt_type, Z_packet_to_raw_bkpt_type): New functions. (find_gdb_breakpoint): New function. (set_gdb_breakpoint_at): Delete. (z_type_supported): New function. (set_gdb_breakpoint_1): New function, loosely based off set_gdb_breakpoint_at. (check_gdb_bp_preconditions, set_gdb_breakpoint): New functions. (delete_gdb_breakpoint_at): Delete. (delete_gdb_breakpoint_1): New function, loosely based off delete_gdb_breakpoint_at. (delete_gdb_breakpoint): New function. (clear_gdb_breakpoint_conditions): Rename to ... (clear_breakpoint_conditions): ... this. Don't handle a NULL breakpoint. (add_condition_to_breakpoint): Make static. (add_breakpoint_condition): Take a struct breakpoint pointer instead of an address. Adjust. (gdb_condition_true_at_breakpoint): Rename to ... (gdb_condition_true_at_breakpoint_z_type): ... this, and add z_type parameter. (gdb_condition_true_at_breakpoint): Reimplement. (add_breakpoint_commands): Take a struct breakpoint pointer instead of an address. Adjust. (gdb_no_commands_at_breakpoint): Rename to ... (gdb_no_commands_at_breakpoint_z_type): ... this. Add z_type parameter. Return true if no breakpoint was found. Change debug output. (gdb_no_commands_at_breakpoint): Reimplement. (run_breakpoint_commands): Rename to ... (run_breakpoint_commands_z_type): ... this. Add z_type parameter, and change return type to boolean. (run_breakpoint_commands): New function. (gdb_breakpoint_here): Also check for Z1 breakpoints. (uninsert_raw_breakpoint): Don't try to reinsert a disabled breakpoint. Go through the_target->remove_point instead of assuming memory breakpoint. (uninsert_breakpoints_at, uninsert_all_breakpoints): Uninsert software and hardware breakpoints. (reinsert_raw_breakpoint): Go through the_target->insert_point instead of assuming memory breakpoint. (reinsert_breakpoints_at, reinsert_all_breakpoints): Reinsert software and hardware breakpoints. (check_breakpoints, breakpoint_here, breakpoint_inserted_here): Check both software and hardware breakpoints. (validate_inserted_breakpoint): Assert the breakpoint is a software breakpoint. Set the inserted flag to -1 instead of setting shlib_disabled. (delete_disabled_breakpoints): Adjust. (validate_breakpoints): Only validate software breakpoints. Adjust to inserted flag change. (check_mem_read, check_mem_write): Skip breakpoint types other than software breakpoints. Adjust to inserted flag change. * mem-break.h (enum raw_bkpt_type): New enum. (raw_breakpoint, struct process_info): Forward declare. (Z_packet_to_target_hw_bp_type): Delete declaration. (raw_bkpt_type_to_target_hw_bp_type, Z_packet_to_raw_bkpt_type) (set_gdb_breakpoint, delete_gdb_breakpoint) (clear_breakpoint_conditions): New declarations. (set_gdb_breakpoint_at, clear_gdb_breakpoint_conditions): Delete. (breakpoint_inserted_here): Update comment. (add_breakpoint_condition, add_breakpoint_commands): Replace address parameter with a breakpoint pointer parameter. (gdb_breakpoint_here): Update comment. (delete_gdb_breakpoint_at): Delete. (insert_memory_breakpoint, remove_memory_breakpoint): Declare. * server.c (process_point_options): Take a struct breakpoint pointer instead of an address. Adjust. (process_serial_event) <Z/z packets>: Use set_gdb_breakpoint and delete_gdb_breakpoint. * spu-low.c (spu_target_ops): Install NULL as supports_z_point_type method. * target.h: Include mem-break.h. (struct target_ops) <prepare_to_access_memory>: Update comment. <supports_z_point_type>: New field. <insert_point, remove_point>: Take an enum raw_bkpt_type argument instead of a char. Also take a raw breakpoint pointer. * win32-arm-low.c (the_low_target): Install NULL as supports_z_point_type. * win32-i386-low.c (i386_supports_z_point_type): New function. (i386_insert_point, i386_remove_point): Adjust to new interface. (the_low_target): Install i386_supports_z_point_type. * win32-low.c (win32_supports_z_point_type): New function. (win32_insert_point, win32_remove_point): Adjust to new interface. (win32_target_ops): Install win32_supports_z_point_type. * win32-low.h (struct win32_target_ops): <supports_z_point_type>: New method. <insert_point, remove_point>: Take an enum raw_bkpt_type argument instead of a char. Also take a raw breakpoint pointer. gdb/testsuite/ 2014-05-20 Pedro Alves <palves@redhat.com> * gdb.base/break-idempotent.c: New file. * gdb.base/break-idempotent.exp: New file.
2014-05-20 19:24:28 +02:00
/* The low level breakpoint types. */
[GDBserver] Move Z packet defines and type convertion routines to shared code. The Aarch64, MIPS and x86 Linux backends all have Z packet number defines and corresponding protocol number to internal type convertion routines. Factor them all out to gdbserver's core code, so we only have one shared copy. Tested on x86_64 Fedora 20, and also cross built for aarch64-linux-gnu and mips-linux-gnu. gdb/gdbserver/ 2014-05-20 Pedro Alves <palves@redhat.com> * mem-break.h: Include break-common.h. (Z_PACKET_SW_BP, Z_PACKET_HW_BP, Z_PACKET_WRITE_WP) (Z_PACKET_READ_WP, Z_PACKET_ACCESS_WP): New defines. (Z_packet_to_target_hw_bp_type): New declaration. * mem-break.c (Z_packet_to_target_hw_bp_type): New function. * i386-low.c (Z_PACKET_HW_BP, Z_PACKET_WRITE_WP, Z_PACKET_READ_WP) (Z_PACKET_ACCESS_WP): Delete macros. (Z_packet_to_hw_type): Delete function. * i386-low.h: Don't include break-common.h here. (Z_packet_to_hw_type): Delete declaration. * linux-x86-low.c (x86_insert_point, x86_insert_point): Call Z_packet_to_target_hw_bp_type instead of Z_packet_to_hw_type. * win32-i386-low.c (i386_insert_point, i386_remove_point): Call Z_packet_to_target_hw_bp_type instead of Z_packet_to_hw_type. * linux-aarch64-low.c: Don't include break-common.h here. (Z_PACKET_SW_BP, Z_PACKET_HW_BP, Z_PACKET_WRITE_WP) (Z_PACKET_READ_WP, Z_PACKET_ACCESS_WP): Delete macros. (Z_packet_to_target_hw_bp_type): Delete function. * linux-mips-low.c (rsp_bp_type_to_target_hw_bp_type): Delete function. (mips_insert_point, mips_remove_point): Use Z_packet_to_target_hw_bp_type.
2014-05-20 19:24:27 +02:00
[GDBserver] Make Zx/zx packet handling idempotent. This patch fixes hardware breakpoint regressions exposed by my fix for "PR breakpoints/7143 - Watchpoint does not trigger when first set", at https://sourceware.org/ml/gdb-patches/2014-03/msg00167.html The testsuite caught them on Linux/x86_64, at least. gdb.sum: gdb.sum: FAIL: gdb.base/hbreak2.exp: next over recursive call FAIL: gdb.base/hbreak2.exp: backtrace from factorial(5.1) FAIL: gdb.base/hbreak2.exp: continue until exit at recursive next test gdb.log: (gdb) next Program received signal SIGTRAP, Trace/breakpoint trap. factorial (value=4) at ../../../src/gdb/testsuite/gdb.base/break.c:113 113 if (value > 1) { /* set breakpoint 7 here */ (gdb) FAIL: gdb.base/hbreak2.exp: next over recursive call Actually, that patch just exposed a latent issue to "breakpoints always-inserted off" mode, not really caused it. After that patch, GDB no longer removes breakpoints at each internal event, thus making some scenarios behave like breakpoint always-inserted on. The bug is easy to trigger with always-inserted on. The issue is that since the target-side breakpoint conditions support, if the stub/server supports evaluating breakpoint conditions on the target side, then GDB is sending duplicate Zx packets to the target without removing them before, and GDBserver is not really expecting that for Z packets other than Z0/z0. E.g., with "set breakpoint always-inserted on" and "set debug remote 1": (gdb) b main Sending packet: $m410943,1#ff...Packet received: 48 Breakpoint 4 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z0,410943,1#48...Packet received: OK ^^^^^^^^^^^^ (gdb) b main Note: breakpoint 4 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Breakpoint 5 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z0,410943,1#48...Packet received: OK ^^^^^^^^^^^^ (gdb) b main Note: breakpoints 4 and 5 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Breakpoint 6 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z0,410943,1#48...Packet received: OK ^^^^^^^^^^^^ (gdb) del Delete all breakpoints? (y or n) y Sending packet: $Z0,410943,1#48...Packet received: OK Sending packet: $Z0,410943,1#48...Packet received: OK Sending packet: $z0,410943,1#68...Packet received: OK And for Z1, similarly: (gdb) hbreak main Sending packet: $m410943,1#ff...Packet received: 48 Hardware assisted breakpoint 4 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ Packet Z1 (hardware-breakpoint) is supported (gdb) hbreak main Note: breakpoint 4 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Hardware assisted breakpoint 5 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ (gdb) hbreak main Note: breakpoints 4 and 5 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Hardware assisted breakpoint 6 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ (gdb) del Delete all breakpoints? (y or n) y Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ Sending packet: $z1,410943,1#69...Packet received: OK ^^^^^^^^^^^^ So GDB sent a bunch of Z1 packets, and then when finally removing the breakpoint, only one z1 packet was sent. On the GDBserver side (with monitor set debug-hw-points 1), in the Z1 case, we see: $ ./gdbserver :9999 ./gdbserver Process ./gdbserver created; pid = 8629 Listening on port 9999 Remote debugging from host 127.0.0.1 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=1 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=2 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=3 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=4 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=5 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 remove_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=4 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 That's one insert_watchpoint call for each Z1 packet, and then one remove_watchpoint call for the z1 packet. Notice how ref.count increased for each insert_watchpoint call, and then in the end, after GDB told GDBserver to forget about the hardware breakpoint, GDBserver ends with the the first debug register still with ref.count=4! IOW, the hardware breakpoint is left armed on the target, while on the GDB end it's gone. If the program happens to execute 0x410943 afterwards, then the CPU traps, GDBserver reports the trap to GDB, and GDB not having a breakpoint set at that address anymore, reports to the user a spurious SIGTRAP. This is exactly what is happening in the hbreak2.exp test, though in that case, it's a shared library event that triggers a breakpoint_re_set, when breakpoints are still inserted (because nowadays GDB doesn't remove breakpoints while handling internal events), and that recreates breakpoint locations, which likewise forces breakpoint reinsertion and Zx packet resends... That is a lot of bogus Zx duplication that should possibly be addressed on the GDB side. GDB resends Zx packets because the way to change the target-side condition, is to resend the breakpoint to the server with the new condition. (That's an option in the packet: e.g., "Z1,410943,1;X3,220027" for "hbreak main if 0". The packets in the examples above are shorter because the breakpoints don't have conditions attached). GDB doesn't remove the breakpoint first before reinserting it because that'd be bad for non-stop, as it'd open a window where the inferior could miss the breakpoint. The conditions actually haven't changed between the resends, but GDB isn't smart enough to realize that. (TBC, if the target doesn't support target-side conditions, then GDB doesn't trigger these resends (init_bp_location calls mark_breakpoint_location_modified, and that does nothing if condition evaluation is on the host side. The resends are caused by the 'loc->condition_changed = condition_modified.' line.) But, even if GDB was made smarter, GDBserver should really still handle the resends anyway. So target-side conditions also aren't really to blame. The documentation of the Z/z packets says: "To avoid potential problems with duplicate packets, the operations should be implemented in an idempotent way." As such, we may want to fix GDB, but we should definitely fix GDBserver. The fix is a prerequisite for target-side conditions on hardware breakpoints anyway (and while at it, on watchpoints too). GDBserver indeed already treats duplicate Z0 packets in an idempotent way. mem-break.c has the concept of high-level and low-level breakpoints, somewhat similar to GDB's split of breakpoints vs breakpoint locations, and keeps track of multiple breakpoints referencing the same address/location, for the case of an internal GDBserver breakpoint or a tracepoint being set at the same address as a GDB breakpoint. But, it only allows GDB to ever contribute one reference to a software breakpoint location. IOW, if gdbserver sees a Z0 packet for the same address where it already had a GDB breakpoint set, then GDBserver won't create another high-level GDB breakpoint. However, mem-break.c only tracks GDB Z0 breakpoints. The same logic should apply to all kinds of Zx packets. Currently, gdbserver passes down each duplicate Zx (other than Z0) request directly to the target->insert_point routine. The x86 watchpoint support itself refcounts watchpoint / hw breakpoint requests, to handle overlapping watchpoints, and save debug registers. But that code doesn't (and really shouldn't) handle the duplicate requests, assuming that for each insert there will be a corresponding remove. So the fix is to generalize mem-break.c to track all kinds of Zx breakpoints, and filter out duplicates. As mentioned, this ends up adding support for target-side conditions on hardware breakpoints and watchpoints too (though GDB itself doesn't support the latter yet). Probably the least obvious change in the patch is that it kind of turns the breakpoint insert/remove APIs inside out. Before, the target methods were only called for GDB breakpoints. The internal breakpoint set/delete methods inserted memory breakpoints directly bypassing the insert/remove target methods. That's not good when the target should use a debug API to set software breakpoints, instead of relying on GDBserver patching memory with breakpoint instructions, as is the case of NTO. Now removal/insertion of all kinds of breakpoints/watchpoints, either internal, or from GDB, always go through the target methods. The insert_point/remove_point methods no longer get passed a Z packet type, but an internal/raw breakpoint type. They're also passed a pointer to the raw breakpoint itself (note that's still opaque outside mem-break.c), so that insert_memory_breakpoint / remove_memory_breakpoint have access to the breakpoint's shadow buffer. I first tried passing down a new structure based on GDB's "struct bp_target_info" (actually with that name exactly), but then decided against it as unnecessary complication. As software/memory breakpoints work by poking at memory, when setting a GDB Z0 breakpoint (but not internal breakpoints, as those can assume the conditions are already right), we need to tell the target to prepare to access memory (which on Linux means stop threads). If that operation fails, we need to return error to GDB. Seeing an error, if this is the first breakpoint of that type that GDB tries to insert, GDB would then assume the breakpoint type is supported, but it may actually not be. So we need to check whether the type is supported at all before preparing to access memory. And to solve that, the patch adds a new target->supports_z_point_type method that is called before actually trying to insert the breakpoint. Other than that, hopefully the change is more or less obvious. New test added that exercises the hbreak2.exp regression in a more direct way, without relying on a breakpoint re-set happening before main is reached. Tested by building GDBserver for: aarch64-linux-gnu arm-linux-gnueabihf i686-pc-linux-gnu i686-w64-mingw32 m68k-linux-gnu mips-linux-gnu mips-uclinux nios2-linux-gnu powerpc-linux-gnu sh-linux-gnu tilegx-unknown-linux-gnu x86_64-redhat-linux x86_64-w64-mingw32 And also regression tested on x86_64 Fedora 20. gdb/gdbserver/ 2014-05-20 Pedro Alves <palves@redhat.com> * linux-aarch64-low.c (aarch64_insert_point) (aarch64_remove_point): No longer check whether the type is supported here. Adjust to new interface. (the_low_target): Install aarch64_supports_z_point_type as supports_z_point_type method. * linux-arm-low.c (raw_bkpt_type_to_arm_hwbp_type): New function. (arm_linux_hw_point_initialize): Take an enum raw_bkpt_type instead of a Z packet char. Adjust. (arm_supports_z_point_type): New function. (arm_insert_point, arm_remove_point): Adjust to new interface. (the_low_target): Install arm_supports_z_point_type. * linux-crisv32-low.c (cris_supports_z_point_type): New function. (cris_insert_point, cris_remove_point): Adjust to new interface. Don't check whether the type is supported here. (the_low_target): Install cris_supports_z_point_type. * linux-low.c (linux_supports_z_point_type): New function. (linux_insert_point, linux_remove_point): Adjust to new interface. * linux-low.h (struct linux_target_ops) <insert_point, remove_point>: Take an enum raw_bkpt_type instead of a char. Add raw_breakpoint pointer parameter. <supports_z_point_type>: New method. * linux-mips-low.c (mips_supports_z_point_type): New function. (mips_insert_point, mips_remove_point): Adjust to new interface. Use mips_supports_z_point_type. (the_low_target): Install mips_supports_z_point_type. * linux-ppc-low.c (the_low_target): Install NULL as supports_z_point_type method. * linux-s390-low.c (the_low_target): Install NULL as supports_z_point_type method. * linux-sparc-low.c (the_low_target): Install NULL as supports_z_point_type method. * linux-x86-low.c (x86_supports_z_point_type): New function. (x86_insert_point): Adjust to new insert_point interface. Use insert_memory_breakpoint. Adjust to new i386_low_insert_watchpoint interface. (x86_remove_point): Adjust to remove_point interface. Use remove_memory_breakpoint. Adjust to new i386_low_remove_watchpoint interface. (the_low_target): Install x86_supports_z_point_type. * lynx-low.c (lynx_target_ops): Install NULL as supports_z_point_type callback. * nto-low.c (nto_supports_z_point_type): New. (nto_insert_point, nto_remove_point): Adjust to new interface. (nto_target_ops): Install nto_supports_z_point_type. * mem-break.c: Adjust intro comment. (struct raw_breakpoint) <raw_type, size>: New fields. <inserted>: Update comment. <shlib_disabled>: Delete field. (enum bkpt_type) <gdb_breakpoint>: Delete value. <gdb_breakpoint_Z0, gdb_breakpoint_Z1, gdb_breakpoint_Z2, gdb_breakpoint_Z3, gdb_breakpoint_Z4>: New values. (raw_bkpt_type_to_target_hw_bp_type): New function. (find_enabled_raw_code_breakpoint_at): New function. (find_raw_breakpoint_at): New type and size parameters. Use them. (insert_memory_breakpoint): New function, based off set_raw_breakpoint_at. (remove_memory_breakpoint): New function. (set_raw_breakpoint_at): Reimplement. (set_breakpoint): New, based on set_breakpoint_at. (set_breakpoint_at): Reimplement. (delete_raw_breakpoint): Go through the_target->remove_point instead of assuming memory breakpoints. (find_gdb_breakpoint_at): Delete. (Z_packet_to_bkpt_type, Z_packet_to_raw_bkpt_type): New functions. (find_gdb_breakpoint): New function. (set_gdb_breakpoint_at): Delete. (z_type_supported): New function. (set_gdb_breakpoint_1): New function, loosely based off set_gdb_breakpoint_at. (check_gdb_bp_preconditions, set_gdb_breakpoint): New functions. (delete_gdb_breakpoint_at): Delete. (delete_gdb_breakpoint_1): New function, loosely based off delete_gdb_breakpoint_at. (delete_gdb_breakpoint): New function. (clear_gdb_breakpoint_conditions): Rename to ... (clear_breakpoint_conditions): ... this. Don't handle a NULL breakpoint. (add_condition_to_breakpoint): Make static. (add_breakpoint_condition): Take a struct breakpoint pointer instead of an address. Adjust. (gdb_condition_true_at_breakpoint): Rename to ... (gdb_condition_true_at_breakpoint_z_type): ... this, and add z_type parameter. (gdb_condition_true_at_breakpoint): Reimplement. (add_breakpoint_commands): Take a struct breakpoint pointer instead of an address. Adjust. (gdb_no_commands_at_breakpoint): Rename to ... (gdb_no_commands_at_breakpoint_z_type): ... this. Add z_type parameter. Return true if no breakpoint was found. Change debug output. (gdb_no_commands_at_breakpoint): Reimplement. (run_breakpoint_commands): Rename to ... (run_breakpoint_commands_z_type): ... this. Add z_type parameter, and change return type to boolean. (run_breakpoint_commands): New function. (gdb_breakpoint_here): Also check for Z1 breakpoints. (uninsert_raw_breakpoint): Don't try to reinsert a disabled breakpoint. Go through the_target->remove_point instead of assuming memory breakpoint. (uninsert_breakpoints_at, uninsert_all_breakpoints): Uninsert software and hardware breakpoints. (reinsert_raw_breakpoint): Go through the_target->insert_point instead of assuming memory breakpoint. (reinsert_breakpoints_at, reinsert_all_breakpoints): Reinsert software and hardware breakpoints. (check_breakpoints, breakpoint_here, breakpoint_inserted_here): Check both software and hardware breakpoints. (validate_inserted_breakpoint): Assert the breakpoint is a software breakpoint. Set the inserted flag to -1 instead of setting shlib_disabled. (delete_disabled_breakpoints): Adjust. (validate_breakpoints): Only validate software breakpoints. Adjust to inserted flag change. (check_mem_read, check_mem_write): Skip breakpoint types other than software breakpoints. Adjust to inserted flag change. * mem-break.h (enum raw_bkpt_type): New enum. (raw_breakpoint, struct process_info): Forward declare. (Z_packet_to_target_hw_bp_type): Delete declaration. (raw_bkpt_type_to_target_hw_bp_type, Z_packet_to_raw_bkpt_type) (set_gdb_breakpoint, delete_gdb_breakpoint) (clear_breakpoint_conditions): New declarations. (set_gdb_breakpoint_at, clear_gdb_breakpoint_conditions): Delete. (breakpoint_inserted_here): Update comment. (add_breakpoint_condition, add_breakpoint_commands): Replace address parameter with a breakpoint pointer parameter. (gdb_breakpoint_here): Update comment. (delete_gdb_breakpoint_at): Delete. (insert_memory_breakpoint, remove_memory_breakpoint): Declare. * server.c (process_point_options): Take a struct breakpoint pointer instead of an address. Adjust. (process_serial_event) <Z/z packets>: Use set_gdb_breakpoint and delete_gdb_breakpoint. * spu-low.c (spu_target_ops): Install NULL as supports_z_point_type method. * target.h: Include mem-break.h. (struct target_ops) <prepare_to_access_memory>: Update comment. <supports_z_point_type>: New field. <insert_point, remove_point>: Take an enum raw_bkpt_type argument instead of a char. Also take a raw breakpoint pointer. * win32-arm-low.c (the_low_target): Install NULL as supports_z_point_type. * win32-i386-low.c (i386_supports_z_point_type): New function. (i386_insert_point, i386_remove_point): Adjust to new interface. (the_low_target): Install i386_supports_z_point_type. * win32-low.c (win32_supports_z_point_type): New function. (win32_insert_point, win32_remove_point): Adjust to new interface. (win32_target_ops): Install win32_supports_z_point_type. * win32-low.h (struct win32_target_ops): <supports_z_point_type>: New method. <insert_point, remove_point>: Take an enum raw_bkpt_type argument instead of a char. Also take a raw breakpoint pointer. gdb/testsuite/ 2014-05-20 Pedro Alves <palves@redhat.com> * gdb.base/break-idempotent.c: New file. * gdb.base/break-idempotent.exp: New file.
2014-05-20 19:24:28 +02:00
enum raw_bkpt_type
{
/* Software/memory breakpoint. */
raw_bkpt_type_sw,
[GDBserver] Move Z packet defines and type convertion routines to shared code. The Aarch64, MIPS and x86 Linux backends all have Z packet number defines and corresponding protocol number to internal type convertion routines. Factor them all out to gdbserver's core code, so we only have one shared copy. Tested on x86_64 Fedora 20, and also cross built for aarch64-linux-gnu and mips-linux-gnu. gdb/gdbserver/ 2014-05-20 Pedro Alves <palves@redhat.com> * mem-break.h: Include break-common.h. (Z_PACKET_SW_BP, Z_PACKET_HW_BP, Z_PACKET_WRITE_WP) (Z_PACKET_READ_WP, Z_PACKET_ACCESS_WP): New defines. (Z_packet_to_target_hw_bp_type): New declaration. * mem-break.c (Z_packet_to_target_hw_bp_type): New function. * i386-low.c (Z_PACKET_HW_BP, Z_PACKET_WRITE_WP, Z_PACKET_READ_WP) (Z_PACKET_ACCESS_WP): Delete macros. (Z_packet_to_hw_type): Delete function. * i386-low.h: Don't include break-common.h here. (Z_packet_to_hw_type): Delete declaration. * linux-x86-low.c (x86_insert_point, x86_insert_point): Call Z_packet_to_target_hw_bp_type instead of Z_packet_to_hw_type. * win32-i386-low.c (i386_insert_point, i386_remove_point): Call Z_packet_to_target_hw_bp_type instead of Z_packet_to_hw_type. * linux-aarch64-low.c: Don't include break-common.h here. (Z_PACKET_SW_BP, Z_PACKET_HW_BP, Z_PACKET_WRITE_WP) (Z_PACKET_READ_WP, Z_PACKET_ACCESS_WP): Delete macros. (Z_packet_to_target_hw_bp_type): Delete function. * linux-mips-low.c (rsp_bp_type_to_target_hw_bp_type): Delete function. (mips_insert_point, mips_remove_point): Use Z_packet_to_target_hw_bp_type.
2014-05-20 19:24:27 +02:00
[GDBserver] Make Zx/zx packet handling idempotent. This patch fixes hardware breakpoint regressions exposed by my fix for "PR breakpoints/7143 - Watchpoint does not trigger when first set", at https://sourceware.org/ml/gdb-patches/2014-03/msg00167.html The testsuite caught them on Linux/x86_64, at least. gdb.sum: gdb.sum: FAIL: gdb.base/hbreak2.exp: next over recursive call FAIL: gdb.base/hbreak2.exp: backtrace from factorial(5.1) FAIL: gdb.base/hbreak2.exp: continue until exit at recursive next test gdb.log: (gdb) next Program received signal SIGTRAP, Trace/breakpoint trap. factorial (value=4) at ../../../src/gdb/testsuite/gdb.base/break.c:113 113 if (value > 1) { /* set breakpoint 7 here */ (gdb) FAIL: gdb.base/hbreak2.exp: next over recursive call Actually, that patch just exposed a latent issue to "breakpoints always-inserted off" mode, not really caused it. After that patch, GDB no longer removes breakpoints at each internal event, thus making some scenarios behave like breakpoint always-inserted on. The bug is easy to trigger with always-inserted on. The issue is that since the target-side breakpoint conditions support, if the stub/server supports evaluating breakpoint conditions on the target side, then GDB is sending duplicate Zx packets to the target without removing them before, and GDBserver is not really expecting that for Z packets other than Z0/z0. E.g., with "set breakpoint always-inserted on" and "set debug remote 1": (gdb) b main Sending packet: $m410943,1#ff...Packet received: 48 Breakpoint 4 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z0,410943,1#48...Packet received: OK ^^^^^^^^^^^^ (gdb) b main Note: breakpoint 4 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Breakpoint 5 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z0,410943,1#48...Packet received: OK ^^^^^^^^^^^^ (gdb) b main Note: breakpoints 4 and 5 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Breakpoint 6 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z0,410943,1#48...Packet received: OK ^^^^^^^^^^^^ (gdb) del Delete all breakpoints? (y or n) y Sending packet: $Z0,410943,1#48...Packet received: OK Sending packet: $Z0,410943,1#48...Packet received: OK Sending packet: $z0,410943,1#68...Packet received: OK And for Z1, similarly: (gdb) hbreak main Sending packet: $m410943,1#ff...Packet received: 48 Hardware assisted breakpoint 4 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ Packet Z1 (hardware-breakpoint) is supported (gdb) hbreak main Note: breakpoint 4 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Hardware assisted breakpoint 5 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ (gdb) hbreak main Note: breakpoints 4 and 5 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Hardware assisted breakpoint 6 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ (gdb) del Delete all breakpoints? (y or n) y Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ Sending packet: $z1,410943,1#69...Packet received: OK ^^^^^^^^^^^^ So GDB sent a bunch of Z1 packets, and then when finally removing the breakpoint, only one z1 packet was sent. On the GDBserver side (with monitor set debug-hw-points 1), in the Z1 case, we see: $ ./gdbserver :9999 ./gdbserver Process ./gdbserver created; pid = 8629 Listening on port 9999 Remote debugging from host 127.0.0.1 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=1 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=2 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=3 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=4 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=5 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 remove_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=4 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 That's one insert_watchpoint call for each Z1 packet, and then one remove_watchpoint call for the z1 packet. Notice how ref.count increased for each insert_watchpoint call, and then in the end, after GDB told GDBserver to forget about the hardware breakpoint, GDBserver ends with the the first debug register still with ref.count=4! IOW, the hardware breakpoint is left armed on the target, while on the GDB end it's gone. If the program happens to execute 0x410943 afterwards, then the CPU traps, GDBserver reports the trap to GDB, and GDB not having a breakpoint set at that address anymore, reports to the user a spurious SIGTRAP. This is exactly what is happening in the hbreak2.exp test, though in that case, it's a shared library event that triggers a breakpoint_re_set, when breakpoints are still inserted (because nowadays GDB doesn't remove breakpoints while handling internal events), and that recreates breakpoint locations, which likewise forces breakpoint reinsertion and Zx packet resends... That is a lot of bogus Zx duplication that should possibly be addressed on the GDB side. GDB resends Zx packets because the way to change the target-side condition, is to resend the breakpoint to the server with the new condition. (That's an option in the packet: e.g., "Z1,410943,1;X3,220027" for "hbreak main if 0". The packets in the examples above are shorter because the breakpoints don't have conditions attached). GDB doesn't remove the breakpoint first before reinserting it because that'd be bad for non-stop, as it'd open a window where the inferior could miss the breakpoint. The conditions actually haven't changed between the resends, but GDB isn't smart enough to realize that. (TBC, if the target doesn't support target-side conditions, then GDB doesn't trigger these resends (init_bp_location calls mark_breakpoint_location_modified, and that does nothing if condition evaluation is on the host side. The resends are caused by the 'loc->condition_changed = condition_modified.' line.) But, even if GDB was made smarter, GDBserver should really still handle the resends anyway. So target-side conditions also aren't really to blame. The documentation of the Z/z packets says: "To avoid potential problems with duplicate packets, the operations should be implemented in an idempotent way." As such, we may want to fix GDB, but we should definitely fix GDBserver. The fix is a prerequisite for target-side conditions on hardware breakpoints anyway (and while at it, on watchpoints too). GDBserver indeed already treats duplicate Z0 packets in an idempotent way. mem-break.c has the concept of high-level and low-level breakpoints, somewhat similar to GDB's split of breakpoints vs breakpoint locations, and keeps track of multiple breakpoints referencing the same address/location, for the case of an internal GDBserver breakpoint or a tracepoint being set at the same address as a GDB breakpoint. But, it only allows GDB to ever contribute one reference to a software breakpoint location. IOW, if gdbserver sees a Z0 packet for the same address where it already had a GDB breakpoint set, then GDBserver won't create another high-level GDB breakpoint. However, mem-break.c only tracks GDB Z0 breakpoints. The same logic should apply to all kinds of Zx packets. Currently, gdbserver passes down each duplicate Zx (other than Z0) request directly to the target->insert_point routine. The x86 watchpoint support itself refcounts watchpoint / hw breakpoint requests, to handle overlapping watchpoints, and save debug registers. But that code doesn't (and really shouldn't) handle the duplicate requests, assuming that for each insert there will be a corresponding remove. So the fix is to generalize mem-break.c to track all kinds of Zx breakpoints, and filter out duplicates. As mentioned, this ends up adding support for target-side conditions on hardware breakpoints and watchpoints too (though GDB itself doesn't support the latter yet). Probably the least obvious change in the patch is that it kind of turns the breakpoint insert/remove APIs inside out. Before, the target methods were only called for GDB breakpoints. The internal breakpoint set/delete methods inserted memory breakpoints directly bypassing the insert/remove target methods. That's not good when the target should use a debug API to set software breakpoints, instead of relying on GDBserver patching memory with breakpoint instructions, as is the case of NTO. Now removal/insertion of all kinds of breakpoints/watchpoints, either internal, or from GDB, always go through the target methods. The insert_point/remove_point methods no longer get passed a Z packet type, but an internal/raw breakpoint type. They're also passed a pointer to the raw breakpoint itself (note that's still opaque outside mem-break.c), so that insert_memory_breakpoint / remove_memory_breakpoint have access to the breakpoint's shadow buffer. I first tried passing down a new structure based on GDB's "struct bp_target_info" (actually with that name exactly), but then decided against it as unnecessary complication. As software/memory breakpoints work by poking at memory, when setting a GDB Z0 breakpoint (but not internal breakpoints, as those can assume the conditions are already right), we need to tell the target to prepare to access memory (which on Linux means stop threads). If that operation fails, we need to return error to GDB. Seeing an error, if this is the first breakpoint of that type that GDB tries to insert, GDB would then assume the breakpoint type is supported, but it may actually not be. So we need to check whether the type is supported at all before preparing to access memory. And to solve that, the patch adds a new target->supports_z_point_type method that is called before actually trying to insert the breakpoint. Other than that, hopefully the change is more or less obvious. New test added that exercises the hbreak2.exp regression in a more direct way, without relying on a breakpoint re-set happening before main is reached. Tested by building GDBserver for: aarch64-linux-gnu arm-linux-gnueabihf i686-pc-linux-gnu i686-w64-mingw32 m68k-linux-gnu mips-linux-gnu mips-uclinux nios2-linux-gnu powerpc-linux-gnu sh-linux-gnu tilegx-unknown-linux-gnu x86_64-redhat-linux x86_64-w64-mingw32 And also regression tested on x86_64 Fedora 20. gdb/gdbserver/ 2014-05-20 Pedro Alves <palves@redhat.com> * linux-aarch64-low.c (aarch64_insert_point) (aarch64_remove_point): No longer check whether the type is supported here. Adjust to new interface. (the_low_target): Install aarch64_supports_z_point_type as supports_z_point_type method. * linux-arm-low.c (raw_bkpt_type_to_arm_hwbp_type): New function. (arm_linux_hw_point_initialize): Take an enum raw_bkpt_type instead of a Z packet char. Adjust. (arm_supports_z_point_type): New function. (arm_insert_point, arm_remove_point): Adjust to new interface. (the_low_target): Install arm_supports_z_point_type. * linux-crisv32-low.c (cris_supports_z_point_type): New function. (cris_insert_point, cris_remove_point): Adjust to new interface. Don't check whether the type is supported here. (the_low_target): Install cris_supports_z_point_type. * linux-low.c (linux_supports_z_point_type): New function. (linux_insert_point, linux_remove_point): Adjust to new interface. * linux-low.h (struct linux_target_ops) <insert_point, remove_point>: Take an enum raw_bkpt_type instead of a char. Add raw_breakpoint pointer parameter. <supports_z_point_type>: New method. * linux-mips-low.c (mips_supports_z_point_type): New function. (mips_insert_point, mips_remove_point): Adjust to new interface. Use mips_supports_z_point_type. (the_low_target): Install mips_supports_z_point_type. * linux-ppc-low.c (the_low_target): Install NULL as supports_z_point_type method. * linux-s390-low.c (the_low_target): Install NULL as supports_z_point_type method. * linux-sparc-low.c (the_low_target): Install NULL as supports_z_point_type method. * linux-x86-low.c (x86_supports_z_point_type): New function. (x86_insert_point): Adjust to new insert_point interface. Use insert_memory_breakpoint. Adjust to new i386_low_insert_watchpoint interface. (x86_remove_point): Adjust to remove_point interface. Use remove_memory_breakpoint. Adjust to new i386_low_remove_watchpoint interface. (the_low_target): Install x86_supports_z_point_type. * lynx-low.c (lynx_target_ops): Install NULL as supports_z_point_type callback. * nto-low.c (nto_supports_z_point_type): New. (nto_insert_point, nto_remove_point): Adjust to new interface. (nto_target_ops): Install nto_supports_z_point_type. * mem-break.c: Adjust intro comment. (struct raw_breakpoint) <raw_type, size>: New fields. <inserted>: Update comment. <shlib_disabled>: Delete field. (enum bkpt_type) <gdb_breakpoint>: Delete value. <gdb_breakpoint_Z0, gdb_breakpoint_Z1, gdb_breakpoint_Z2, gdb_breakpoint_Z3, gdb_breakpoint_Z4>: New values. (raw_bkpt_type_to_target_hw_bp_type): New function. (find_enabled_raw_code_breakpoint_at): New function. (find_raw_breakpoint_at): New type and size parameters. Use them. (insert_memory_breakpoint): New function, based off set_raw_breakpoint_at. (remove_memory_breakpoint): New function. (set_raw_breakpoint_at): Reimplement. (set_breakpoint): New, based on set_breakpoint_at. (set_breakpoint_at): Reimplement. (delete_raw_breakpoint): Go through the_target->remove_point instead of assuming memory breakpoints. (find_gdb_breakpoint_at): Delete. (Z_packet_to_bkpt_type, Z_packet_to_raw_bkpt_type): New functions. (find_gdb_breakpoint): New function. (set_gdb_breakpoint_at): Delete. (z_type_supported): New function. (set_gdb_breakpoint_1): New function, loosely based off set_gdb_breakpoint_at. (check_gdb_bp_preconditions, set_gdb_breakpoint): New functions. (delete_gdb_breakpoint_at): Delete. (delete_gdb_breakpoint_1): New function, loosely based off delete_gdb_breakpoint_at. (delete_gdb_breakpoint): New function. (clear_gdb_breakpoint_conditions): Rename to ... (clear_breakpoint_conditions): ... this. Don't handle a NULL breakpoint. (add_condition_to_breakpoint): Make static. (add_breakpoint_condition): Take a struct breakpoint pointer instead of an address. Adjust. (gdb_condition_true_at_breakpoint): Rename to ... (gdb_condition_true_at_breakpoint_z_type): ... this, and add z_type parameter. (gdb_condition_true_at_breakpoint): Reimplement. (add_breakpoint_commands): Take a struct breakpoint pointer instead of an address. Adjust. (gdb_no_commands_at_breakpoint): Rename to ... (gdb_no_commands_at_breakpoint_z_type): ... this. Add z_type parameter. Return true if no breakpoint was found. Change debug output. (gdb_no_commands_at_breakpoint): Reimplement. (run_breakpoint_commands): Rename to ... (run_breakpoint_commands_z_type): ... this. Add z_type parameter, and change return type to boolean. (run_breakpoint_commands): New function. (gdb_breakpoint_here): Also check for Z1 breakpoints. (uninsert_raw_breakpoint): Don't try to reinsert a disabled breakpoint. Go through the_target->remove_point instead of assuming memory breakpoint. (uninsert_breakpoints_at, uninsert_all_breakpoints): Uninsert software and hardware breakpoints. (reinsert_raw_breakpoint): Go through the_target->insert_point instead of assuming memory breakpoint. (reinsert_breakpoints_at, reinsert_all_breakpoints): Reinsert software and hardware breakpoints. (check_breakpoints, breakpoint_here, breakpoint_inserted_here): Check both software and hardware breakpoints. (validate_inserted_breakpoint): Assert the breakpoint is a software breakpoint. Set the inserted flag to -1 instead of setting shlib_disabled. (delete_disabled_breakpoints): Adjust. (validate_breakpoints): Only validate software breakpoints. Adjust to inserted flag change. (check_mem_read, check_mem_write): Skip breakpoint types other than software breakpoints. Adjust to inserted flag change. * mem-break.h (enum raw_bkpt_type): New enum. (raw_breakpoint, struct process_info): Forward declare. (Z_packet_to_target_hw_bp_type): Delete declaration. (raw_bkpt_type_to_target_hw_bp_type, Z_packet_to_raw_bkpt_type) (set_gdb_breakpoint, delete_gdb_breakpoint) (clear_breakpoint_conditions): New declarations. (set_gdb_breakpoint_at, clear_gdb_breakpoint_conditions): Delete. (breakpoint_inserted_here): Update comment. (add_breakpoint_condition, add_breakpoint_commands): Replace address parameter with a breakpoint pointer parameter. (gdb_breakpoint_here): Update comment. (delete_gdb_breakpoint_at): Delete. (insert_memory_breakpoint, remove_memory_breakpoint): Declare. * server.c (process_point_options): Take a struct breakpoint pointer instead of an address. Adjust. (process_serial_event) <Z/z packets>: Use set_gdb_breakpoint and delete_gdb_breakpoint. * spu-low.c (spu_target_ops): Install NULL as supports_z_point_type method. * target.h: Include mem-break.h. (struct target_ops) <prepare_to_access_memory>: Update comment. <supports_z_point_type>: New field. <insert_point, remove_point>: Take an enum raw_bkpt_type argument instead of a char. Also take a raw breakpoint pointer. * win32-arm-low.c (the_low_target): Install NULL as supports_z_point_type. * win32-i386-low.c (i386_supports_z_point_type): New function. (i386_insert_point, i386_remove_point): Adjust to new interface. (the_low_target): Install i386_supports_z_point_type. * win32-low.c (win32_supports_z_point_type): New function. (win32_insert_point, win32_remove_point): Adjust to new interface. (win32_target_ops): Install win32_supports_z_point_type. * win32-low.h (struct win32_target_ops): <supports_z_point_type>: New method. <insert_point, remove_point>: Take an enum raw_bkpt_type argument instead of a char. Also take a raw breakpoint pointer. gdb/testsuite/ 2014-05-20 Pedro Alves <palves@redhat.com> * gdb.base/break-idempotent.c: New file. * gdb.base/break-idempotent.exp: New file.
2014-05-20 19:24:28 +02:00
/* Hardware-assisted breakpoint. */
raw_bkpt_type_hw,
Teach linux gdbserver to step-over-breakpoints. * linux-low.c (can_hardware_single_step): New. (supports_breakpoints): New. (handle_extended_wait): If stopping threads, read the stop pc of the new cloned LWP. (get_pc): New. (get_stop_pc): Add `lwp' parameter. Handle it. Bail out if the low target doesn't support retrieving the PC. (add_lwp): Set last_resume_kind to resume_continue. (linux_attach_lwp_1): Adjust comments. Always set stop_expected. (linux_attach): Don't clear stop_expected. Set the lwp's last_resume_kind to resume_stop. (linux_detach_one_lwp): Don't check for removed breakpoints. (check_removed_breakpoint): Delete. (status_pending_p): Rename to ... (status_pending_p_callback): ... this. Don't check for removed breakpoints. Don't consider threads that are stopped from GDB's perspective. (linux_wait_for_lwp): Always read the stop_pc here. (cancel_breakpoint): New. (step_over_bkpt): New global. (linux_wait_for_event_1): Implement stepping over breakpoints. (gdb_wants_lwp_stopped): New. (gdb_wants_all_stopped): New. (linux_wait_1): Tag threads as gdb-wants-stopped. Cancel finished single-step traps here. Store the thread's last reported target wait status. (send_sigstop): Don't clear stop_expected. Always set it, instead. (mark_lwp_dead): Remove reference to pending_is_breakpoint. (cancel_finished_single_step): New. (cancel_finished_single_steps): New. (wait_for_sigstop): Don't cancel finished single-step traps here. (linux_resume_one_lwp): Don't check for removed breakpoints. Don't set `step' on non-hardware step archs. (linux_set_resume_request): Ignore resume_stop requests if already stopping or stopped. Set the lwp's last_resume_kind. (resume_status_pending_p): Don't check for removed breakpoints. (need_step_over_p): New. (start_step_over): New. (finish_step_over): New. (linux_resume_one_thread): Always queue a sigstop for resume_stop requests. Clear the thread's last reported target waitstatus. Don't use the `suspended' flag. Don't consider pending breakpoints. (linux_resume): Start a step-over if necessary. (proceed_one_lwp): New. (proceed_all_lwps): New. (unstop_all_lwps): New. * linux-low.h (struct lwp_info): Rewrite comment for the `suspended' flag. Add the `stop_pc' field. Delete the `pending_stop_pc' field. Tweak the `stepping' flag's comment. Add `'last_resume_kind' and `need_step_over' fields. * inferiors.c (struct thread_info): Delete, moved elsewhere. * mem-break.c (struct breakpoint): Delete `reinserting' flag. Delete `breakpoint_to_reinsert' field. New flag `inserted'. (set_raw_breakpoint_at): New. (set_breakpoint_at): Rewrite to use it. (reinsert_breakpoint_handler): Delete. (set_reinsert_breakpoint): New. (reinsert_breakpoint_by_bp): Delete. (delete_reinsert_breakpoints): New. (uninsert_breakpoint): Rewrite. (uninsert_breakpoints_at): New. (reinsert_breakpoint): Rewrite. (reinsert_breakpoints_at): New. (check_breakpoints): Rewrite. (breakpoint_here): New. (breakpoint_inserted_here): New. (check_mem_read): Adjust. * mem-break.h (breakpoints_supported, breakpoint_here) (breakpoint_inserted_here, set_reinsert_breakpoint): Declare. (reinsert_breakpoint_by_bp): Delete declaration. (delete_reinsert_breakpoints): Declare. (reinsert_breakpoint): Delete declaration. (reinsert_breakpoints_at): Declare. (uninsert_breakpoint): Delete declaration. (uninsert_breakpoints_at): Declare. (check_breakpoints): Adjust prototype. * server.h: Adjust include order. (struct thread_info): Declare here. Add a `last_status' field.
2010-03-24 01:05:03 +01:00
[GDBserver] Make Zx/zx packet handling idempotent. This patch fixes hardware breakpoint regressions exposed by my fix for "PR breakpoints/7143 - Watchpoint does not trigger when first set", at https://sourceware.org/ml/gdb-patches/2014-03/msg00167.html The testsuite caught them on Linux/x86_64, at least. gdb.sum: gdb.sum: FAIL: gdb.base/hbreak2.exp: next over recursive call FAIL: gdb.base/hbreak2.exp: backtrace from factorial(5.1) FAIL: gdb.base/hbreak2.exp: continue until exit at recursive next test gdb.log: (gdb) next Program received signal SIGTRAP, Trace/breakpoint trap. factorial (value=4) at ../../../src/gdb/testsuite/gdb.base/break.c:113 113 if (value > 1) { /* set breakpoint 7 here */ (gdb) FAIL: gdb.base/hbreak2.exp: next over recursive call Actually, that patch just exposed a latent issue to "breakpoints always-inserted off" mode, not really caused it. After that patch, GDB no longer removes breakpoints at each internal event, thus making some scenarios behave like breakpoint always-inserted on. The bug is easy to trigger with always-inserted on. The issue is that since the target-side breakpoint conditions support, if the stub/server supports evaluating breakpoint conditions on the target side, then GDB is sending duplicate Zx packets to the target without removing them before, and GDBserver is not really expecting that for Z packets other than Z0/z0. E.g., with "set breakpoint always-inserted on" and "set debug remote 1": (gdb) b main Sending packet: $m410943,1#ff...Packet received: 48 Breakpoint 4 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z0,410943,1#48...Packet received: OK ^^^^^^^^^^^^ (gdb) b main Note: breakpoint 4 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Breakpoint 5 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z0,410943,1#48...Packet received: OK ^^^^^^^^^^^^ (gdb) b main Note: breakpoints 4 and 5 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Breakpoint 6 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z0,410943,1#48...Packet received: OK ^^^^^^^^^^^^ (gdb) del Delete all breakpoints? (y or n) y Sending packet: $Z0,410943,1#48...Packet received: OK Sending packet: $Z0,410943,1#48...Packet received: OK Sending packet: $z0,410943,1#68...Packet received: OK And for Z1, similarly: (gdb) hbreak main Sending packet: $m410943,1#ff...Packet received: 48 Hardware assisted breakpoint 4 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ Packet Z1 (hardware-breakpoint) is supported (gdb) hbreak main Note: breakpoint 4 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Hardware assisted breakpoint 5 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ (gdb) hbreak main Note: breakpoints 4 and 5 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Hardware assisted breakpoint 6 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ (gdb) del Delete all breakpoints? (y or n) y Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ Sending packet: $z1,410943,1#69...Packet received: OK ^^^^^^^^^^^^ So GDB sent a bunch of Z1 packets, and then when finally removing the breakpoint, only one z1 packet was sent. On the GDBserver side (with monitor set debug-hw-points 1), in the Z1 case, we see: $ ./gdbserver :9999 ./gdbserver Process ./gdbserver created; pid = 8629 Listening on port 9999 Remote debugging from host 127.0.0.1 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=1 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=2 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=3 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=4 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=5 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 remove_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=4 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 That's one insert_watchpoint call for each Z1 packet, and then one remove_watchpoint call for the z1 packet. Notice how ref.count increased for each insert_watchpoint call, and then in the end, after GDB told GDBserver to forget about the hardware breakpoint, GDBserver ends with the the first debug register still with ref.count=4! IOW, the hardware breakpoint is left armed on the target, while on the GDB end it's gone. If the program happens to execute 0x410943 afterwards, then the CPU traps, GDBserver reports the trap to GDB, and GDB not having a breakpoint set at that address anymore, reports to the user a spurious SIGTRAP. This is exactly what is happening in the hbreak2.exp test, though in that case, it's a shared library event that triggers a breakpoint_re_set, when breakpoints are still inserted (because nowadays GDB doesn't remove breakpoints while handling internal events), and that recreates breakpoint locations, which likewise forces breakpoint reinsertion and Zx packet resends... That is a lot of bogus Zx duplication that should possibly be addressed on the GDB side. GDB resends Zx packets because the way to change the target-side condition, is to resend the breakpoint to the server with the new condition. (That's an option in the packet: e.g., "Z1,410943,1;X3,220027" for "hbreak main if 0". The packets in the examples above are shorter because the breakpoints don't have conditions attached). GDB doesn't remove the breakpoint first before reinserting it because that'd be bad for non-stop, as it'd open a window where the inferior could miss the breakpoint. The conditions actually haven't changed between the resends, but GDB isn't smart enough to realize that. (TBC, if the target doesn't support target-side conditions, then GDB doesn't trigger these resends (init_bp_location calls mark_breakpoint_location_modified, and that does nothing if condition evaluation is on the host side. The resends are caused by the 'loc->condition_changed = condition_modified.' line.) But, even if GDB was made smarter, GDBserver should really still handle the resends anyway. So target-side conditions also aren't really to blame. The documentation of the Z/z packets says: "To avoid potential problems with duplicate packets, the operations should be implemented in an idempotent way." As such, we may want to fix GDB, but we should definitely fix GDBserver. The fix is a prerequisite for target-side conditions on hardware breakpoints anyway (and while at it, on watchpoints too). GDBserver indeed already treats duplicate Z0 packets in an idempotent way. mem-break.c has the concept of high-level and low-level breakpoints, somewhat similar to GDB's split of breakpoints vs breakpoint locations, and keeps track of multiple breakpoints referencing the same address/location, for the case of an internal GDBserver breakpoint or a tracepoint being set at the same address as a GDB breakpoint. But, it only allows GDB to ever contribute one reference to a software breakpoint location. IOW, if gdbserver sees a Z0 packet for the same address where it already had a GDB breakpoint set, then GDBserver won't create another high-level GDB breakpoint. However, mem-break.c only tracks GDB Z0 breakpoints. The same logic should apply to all kinds of Zx packets. Currently, gdbserver passes down each duplicate Zx (other than Z0) request directly to the target->insert_point routine. The x86 watchpoint support itself refcounts watchpoint / hw breakpoint requests, to handle overlapping watchpoints, and save debug registers. But that code doesn't (and really shouldn't) handle the duplicate requests, assuming that for each insert there will be a corresponding remove. So the fix is to generalize mem-break.c to track all kinds of Zx breakpoints, and filter out duplicates. As mentioned, this ends up adding support for target-side conditions on hardware breakpoints and watchpoints too (though GDB itself doesn't support the latter yet). Probably the least obvious change in the patch is that it kind of turns the breakpoint insert/remove APIs inside out. Before, the target methods were only called for GDB breakpoints. The internal breakpoint set/delete methods inserted memory breakpoints directly bypassing the insert/remove target methods. That's not good when the target should use a debug API to set software breakpoints, instead of relying on GDBserver patching memory with breakpoint instructions, as is the case of NTO. Now removal/insertion of all kinds of breakpoints/watchpoints, either internal, or from GDB, always go through the target methods. The insert_point/remove_point methods no longer get passed a Z packet type, but an internal/raw breakpoint type. They're also passed a pointer to the raw breakpoint itself (note that's still opaque outside mem-break.c), so that insert_memory_breakpoint / remove_memory_breakpoint have access to the breakpoint's shadow buffer. I first tried passing down a new structure based on GDB's "struct bp_target_info" (actually with that name exactly), but then decided against it as unnecessary complication. As software/memory breakpoints work by poking at memory, when setting a GDB Z0 breakpoint (but not internal breakpoints, as those can assume the conditions are already right), we need to tell the target to prepare to access memory (which on Linux means stop threads). If that operation fails, we need to return error to GDB. Seeing an error, if this is the first breakpoint of that type that GDB tries to insert, GDB would then assume the breakpoint type is supported, but it may actually not be. So we need to check whether the type is supported at all before preparing to access memory. And to solve that, the patch adds a new target->supports_z_point_type method that is called before actually trying to insert the breakpoint. Other than that, hopefully the change is more or less obvious. New test added that exercises the hbreak2.exp regression in a more direct way, without relying on a breakpoint re-set happening before main is reached. Tested by building GDBserver for: aarch64-linux-gnu arm-linux-gnueabihf i686-pc-linux-gnu i686-w64-mingw32 m68k-linux-gnu mips-linux-gnu mips-uclinux nios2-linux-gnu powerpc-linux-gnu sh-linux-gnu tilegx-unknown-linux-gnu x86_64-redhat-linux x86_64-w64-mingw32 And also regression tested on x86_64 Fedora 20. gdb/gdbserver/ 2014-05-20 Pedro Alves <palves@redhat.com> * linux-aarch64-low.c (aarch64_insert_point) (aarch64_remove_point): No longer check whether the type is supported here. Adjust to new interface. (the_low_target): Install aarch64_supports_z_point_type as supports_z_point_type method. * linux-arm-low.c (raw_bkpt_type_to_arm_hwbp_type): New function. (arm_linux_hw_point_initialize): Take an enum raw_bkpt_type instead of a Z packet char. Adjust. (arm_supports_z_point_type): New function. (arm_insert_point, arm_remove_point): Adjust to new interface. (the_low_target): Install arm_supports_z_point_type. * linux-crisv32-low.c (cris_supports_z_point_type): New function. (cris_insert_point, cris_remove_point): Adjust to new interface. Don't check whether the type is supported here. (the_low_target): Install cris_supports_z_point_type. * linux-low.c (linux_supports_z_point_type): New function. (linux_insert_point, linux_remove_point): Adjust to new interface. * linux-low.h (struct linux_target_ops) <insert_point, remove_point>: Take an enum raw_bkpt_type instead of a char. Add raw_breakpoint pointer parameter. <supports_z_point_type>: New method. * linux-mips-low.c (mips_supports_z_point_type): New function. (mips_insert_point, mips_remove_point): Adjust to new interface. Use mips_supports_z_point_type. (the_low_target): Install mips_supports_z_point_type. * linux-ppc-low.c (the_low_target): Install NULL as supports_z_point_type method. * linux-s390-low.c (the_low_target): Install NULL as supports_z_point_type method. * linux-sparc-low.c (the_low_target): Install NULL as supports_z_point_type method. * linux-x86-low.c (x86_supports_z_point_type): New function. (x86_insert_point): Adjust to new insert_point interface. Use insert_memory_breakpoint. Adjust to new i386_low_insert_watchpoint interface. (x86_remove_point): Adjust to remove_point interface. Use remove_memory_breakpoint. Adjust to new i386_low_remove_watchpoint interface. (the_low_target): Install x86_supports_z_point_type. * lynx-low.c (lynx_target_ops): Install NULL as supports_z_point_type callback. * nto-low.c (nto_supports_z_point_type): New. (nto_insert_point, nto_remove_point): Adjust to new interface. (nto_target_ops): Install nto_supports_z_point_type. * mem-break.c: Adjust intro comment. (struct raw_breakpoint) <raw_type, size>: New fields. <inserted>: Update comment. <shlib_disabled>: Delete field. (enum bkpt_type) <gdb_breakpoint>: Delete value. <gdb_breakpoint_Z0, gdb_breakpoint_Z1, gdb_breakpoint_Z2, gdb_breakpoint_Z3, gdb_breakpoint_Z4>: New values. (raw_bkpt_type_to_target_hw_bp_type): New function. (find_enabled_raw_code_breakpoint_at): New function. (find_raw_breakpoint_at): New type and size parameters. Use them. (insert_memory_breakpoint): New function, based off set_raw_breakpoint_at. (remove_memory_breakpoint): New function. (set_raw_breakpoint_at): Reimplement. (set_breakpoint): New, based on set_breakpoint_at. (set_breakpoint_at): Reimplement. (delete_raw_breakpoint): Go through the_target->remove_point instead of assuming memory breakpoints. (find_gdb_breakpoint_at): Delete. (Z_packet_to_bkpt_type, Z_packet_to_raw_bkpt_type): New functions. (find_gdb_breakpoint): New function. (set_gdb_breakpoint_at): Delete. (z_type_supported): New function. (set_gdb_breakpoint_1): New function, loosely based off set_gdb_breakpoint_at. (check_gdb_bp_preconditions, set_gdb_breakpoint): New functions. (delete_gdb_breakpoint_at): Delete. (delete_gdb_breakpoint_1): New function, loosely based off delete_gdb_breakpoint_at. (delete_gdb_breakpoint): New function. (clear_gdb_breakpoint_conditions): Rename to ... (clear_breakpoint_conditions): ... this. Don't handle a NULL breakpoint. (add_condition_to_breakpoint): Make static. (add_breakpoint_condition): Take a struct breakpoint pointer instead of an address. Adjust. (gdb_condition_true_at_breakpoint): Rename to ... (gdb_condition_true_at_breakpoint_z_type): ... this, and add z_type parameter. (gdb_condition_true_at_breakpoint): Reimplement. (add_breakpoint_commands): Take a struct breakpoint pointer instead of an address. Adjust. (gdb_no_commands_at_breakpoint): Rename to ... (gdb_no_commands_at_breakpoint_z_type): ... this. Add z_type parameter. Return true if no breakpoint was found. Change debug output. (gdb_no_commands_at_breakpoint): Reimplement. (run_breakpoint_commands): Rename to ... (run_breakpoint_commands_z_type): ... this. Add z_type parameter, and change return type to boolean. (run_breakpoint_commands): New function. (gdb_breakpoint_here): Also check for Z1 breakpoints. (uninsert_raw_breakpoint): Don't try to reinsert a disabled breakpoint. Go through the_target->remove_point instead of assuming memory breakpoint. (uninsert_breakpoints_at, uninsert_all_breakpoints): Uninsert software and hardware breakpoints. (reinsert_raw_breakpoint): Go through the_target->insert_point instead of assuming memory breakpoint. (reinsert_breakpoints_at, reinsert_all_breakpoints): Reinsert software and hardware breakpoints. (check_breakpoints, breakpoint_here, breakpoint_inserted_here): Check both software and hardware breakpoints. (validate_inserted_breakpoint): Assert the breakpoint is a software breakpoint. Set the inserted flag to -1 instead of setting shlib_disabled. (delete_disabled_breakpoints): Adjust. (validate_breakpoints): Only validate software breakpoints. Adjust to inserted flag change. (check_mem_read, check_mem_write): Skip breakpoint types other than software breakpoints. Adjust to inserted flag change. * mem-break.h (enum raw_bkpt_type): New enum. (raw_breakpoint, struct process_info): Forward declare. (Z_packet_to_target_hw_bp_type): Delete declaration. (raw_bkpt_type_to_target_hw_bp_type, Z_packet_to_raw_bkpt_type) (set_gdb_breakpoint, delete_gdb_breakpoint) (clear_breakpoint_conditions): New declarations. (set_gdb_breakpoint_at, clear_gdb_breakpoint_conditions): Delete. (breakpoint_inserted_here): Update comment. (add_breakpoint_condition, add_breakpoint_commands): Replace address parameter with a breakpoint pointer parameter. (gdb_breakpoint_here): Update comment. (delete_gdb_breakpoint_at): Delete. (insert_memory_breakpoint, remove_memory_breakpoint): Declare. * server.c (process_point_options): Take a struct breakpoint pointer instead of an address. Adjust. (process_serial_event) <Z/z packets>: Use set_gdb_breakpoint and delete_gdb_breakpoint. * spu-low.c (spu_target_ops): Install NULL as supports_z_point_type method. * target.h: Include mem-break.h. (struct target_ops) <prepare_to_access_memory>: Update comment. <supports_z_point_type>: New field. <insert_point, remove_point>: Take an enum raw_bkpt_type argument instead of a char. Also take a raw breakpoint pointer. * win32-arm-low.c (the_low_target): Install NULL as supports_z_point_type. * win32-i386-low.c (i386_supports_z_point_type): New function. (i386_insert_point, i386_remove_point): Adjust to new interface. (the_low_target): Install i386_supports_z_point_type. * win32-low.c (win32_supports_z_point_type): New function. (win32_insert_point, win32_remove_point): Adjust to new interface. (win32_target_ops): Install win32_supports_z_point_type. * win32-low.h (struct win32_target_ops): <supports_z_point_type>: New method. <insert_point, remove_point>: Take an enum raw_bkpt_type argument instead of a char. Also take a raw breakpoint pointer. gdb/testsuite/ 2014-05-20 Pedro Alves <palves@redhat.com> * gdb.base/break-idempotent.c: New file. * gdb.base/break-idempotent.exp: New file.
2014-05-20 19:24:28 +02:00
/* Hardware-assisted write watchpoint. */
raw_bkpt_type_write_wp,
Teach linux gdbserver to step-over-breakpoints. * linux-low.c (can_hardware_single_step): New. (supports_breakpoints): New. (handle_extended_wait): If stopping threads, read the stop pc of the new cloned LWP. (get_pc): New. (get_stop_pc): Add `lwp' parameter. Handle it. Bail out if the low target doesn't support retrieving the PC. (add_lwp): Set last_resume_kind to resume_continue. (linux_attach_lwp_1): Adjust comments. Always set stop_expected. (linux_attach): Don't clear stop_expected. Set the lwp's last_resume_kind to resume_stop. (linux_detach_one_lwp): Don't check for removed breakpoints. (check_removed_breakpoint): Delete. (status_pending_p): Rename to ... (status_pending_p_callback): ... this. Don't check for removed breakpoints. Don't consider threads that are stopped from GDB's perspective. (linux_wait_for_lwp): Always read the stop_pc here. (cancel_breakpoint): New. (step_over_bkpt): New global. (linux_wait_for_event_1): Implement stepping over breakpoints. (gdb_wants_lwp_stopped): New. (gdb_wants_all_stopped): New. (linux_wait_1): Tag threads as gdb-wants-stopped. Cancel finished single-step traps here. Store the thread's last reported target wait status. (send_sigstop): Don't clear stop_expected. Always set it, instead. (mark_lwp_dead): Remove reference to pending_is_breakpoint. (cancel_finished_single_step): New. (cancel_finished_single_steps): New. (wait_for_sigstop): Don't cancel finished single-step traps here. (linux_resume_one_lwp): Don't check for removed breakpoints. Don't set `step' on non-hardware step archs. (linux_set_resume_request): Ignore resume_stop requests if already stopping or stopped. Set the lwp's last_resume_kind. (resume_status_pending_p): Don't check for removed breakpoints. (need_step_over_p): New. (start_step_over): New. (finish_step_over): New. (linux_resume_one_thread): Always queue a sigstop for resume_stop requests. Clear the thread's last reported target waitstatus. Don't use the `suspended' flag. Don't consider pending breakpoints. (linux_resume): Start a step-over if necessary. (proceed_one_lwp): New. (proceed_all_lwps): New. (unstop_all_lwps): New. * linux-low.h (struct lwp_info): Rewrite comment for the `suspended' flag. Add the `stop_pc' field. Delete the `pending_stop_pc' field. Tweak the `stepping' flag's comment. Add `'last_resume_kind' and `need_step_over' fields. * inferiors.c (struct thread_info): Delete, moved elsewhere. * mem-break.c (struct breakpoint): Delete `reinserting' flag. Delete `breakpoint_to_reinsert' field. New flag `inserted'. (set_raw_breakpoint_at): New. (set_breakpoint_at): Rewrite to use it. (reinsert_breakpoint_handler): Delete. (set_reinsert_breakpoint): New. (reinsert_breakpoint_by_bp): Delete. (delete_reinsert_breakpoints): New. (uninsert_breakpoint): Rewrite. (uninsert_breakpoints_at): New. (reinsert_breakpoint): Rewrite. (reinsert_breakpoints_at): New. (check_breakpoints): Rewrite. (breakpoint_here): New. (breakpoint_inserted_here): New. (check_mem_read): Adjust. * mem-break.h (breakpoints_supported, breakpoint_here) (breakpoint_inserted_here, set_reinsert_breakpoint): Declare. (reinsert_breakpoint_by_bp): Delete declaration. (delete_reinsert_breakpoints): Declare. (reinsert_breakpoint): Delete declaration. (reinsert_breakpoints_at): Declare. (uninsert_breakpoint): Delete declaration. (uninsert_breakpoints_at): Declare. (check_breakpoints): Adjust prototype. * server.h: Adjust include order. (struct thread_info): Declare here. Add a `last_status' field.
2010-03-24 01:05:03 +01:00
[GDBserver] Make Zx/zx packet handling idempotent. This patch fixes hardware breakpoint regressions exposed by my fix for "PR breakpoints/7143 - Watchpoint does not trigger when first set", at https://sourceware.org/ml/gdb-patches/2014-03/msg00167.html The testsuite caught them on Linux/x86_64, at least. gdb.sum: gdb.sum: FAIL: gdb.base/hbreak2.exp: next over recursive call FAIL: gdb.base/hbreak2.exp: backtrace from factorial(5.1) FAIL: gdb.base/hbreak2.exp: continue until exit at recursive next test gdb.log: (gdb) next Program received signal SIGTRAP, Trace/breakpoint trap. factorial (value=4) at ../../../src/gdb/testsuite/gdb.base/break.c:113 113 if (value > 1) { /* set breakpoint 7 here */ (gdb) FAIL: gdb.base/hbreak2.exp: next over recursive call Actually, that patch just exposed a latent issue to "breakpoints always-inserted off" mode, not really caused it. After that patch, GDB no longer removes breakpoints at each internal event, thus making some scenarios behave like breakpoint always-inserted on. The bug is easy to trigger with always-inserted on. The issue is that since the target-side breakpoint conditions support, if the stub/server supports evaluating breakpoint conditions on the target side, then GDB is sending duplicate Zx packets to the target without removing them before, and GDBserver is not really expecting that for Z packets other than Z0/z0. E.g., with "set breakpoint always-inserted on" and "set debug remote 1": (gdb) b main Sending packet: $m410943,1#ff...Packet received: 48 Breakpoint 4 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z0,410943,1#48...Packet received: OK ^^^^^^^^^^^^ (gdb) b main Note: breakpoint 4 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Breakpoint 5 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z0,410943,1#48...Packet received: OK ^^^^^^^^^^^^ (gdb) b main Note: breakpoints 4 and 5 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Breakpoint 6 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z0,410943,1#48...Packet received: OK ^^^^^^^^^^^^ (gdb) del Delete all breakpoints? (y or n) y Sending packet: $Z0,410943,1#48...Packet received: OK Sending packet: $Z0,410943,1#48...Packet received: OK Sending packet: $z0,410943,1#68...Packet received: OK And for Z1, similarly: (gdb) hbreak main Sending packet: $m410943,1#ff...Packet received: 48 Hardware assisted breakpoint 4 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ Packet Z1 (hardware-breakpoint) is supported (gdb) hbreak main Note: breakpoint 4 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Hardware assisted breakpoint 5 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ (gdb) hbreak main Note: breakpoints 4 and 5 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Hardware assisted breakpoint 6 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ (gdb) del Delete all breakpoints? (y or n) y Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ Sending packet: $z1,410943,1#69...Packet received: OK ^^^^^^^^^^^^ So GDB sent a bunch of Z1 packets, and then when finally removing the breakpoint, only one z1 packet was sent. On the GDBserver side (with monitor set debug-hw-points 1), in the Z1 case, we see: $ ./gdbserver :9999 ./gdbserver Process ./gdbserver created; pid = 8629 Listening on port 9999 Remote debugging from host 127.0.0.1 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=1 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=2 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=3 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=4 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=5 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 remove_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=4 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 That's one insert_watchpoint call for each Z1 packet, and then one remove_watchpoint call for the z1 packet. Notice how ref.count increased for each insert_watchpoint call, and then in the end, after GDB told GDBserver to forget about the hardware breakpoint, GDBserver ends with the the first debug register still with ref.count=4! IOW, the hardware breakpoint is left armed on the target, while on the GDB end it's gone. If the program happens to execute 0x410943 afterwards, then the CPU traps, GDBserver reports the trap to GDB, and GDB not having a breakpoint set at that address anymore, reports to the user a spurious SIGTRAP. This is exactly what is happening in the hbreak2.exp test, though in that case, it's a shared library event that triggers a breakpoint_re_set, when breakpoints are still inserted (because nowadays GDB doesn't remove breakpoints while handling internal events), and that recreates breakpoint locations, which likewise forces breakpoint reinsertion and Zx packet resends... That is a lot of bogus Zx duplication that should possibly be addressed on the GDB side. GDB resends Zx packets because the way to change the target-side condition, is to resend the breakpoint to the server with the new condition. (That's an option in the packet: e.g., "Z1,410943,1;X3,220027" for "hbreak main if 0". The packets in the examples above are shorter because the breakpoints don't have conditions attached). GDB doesn't remove the breakpoint first before reinserting it because that'd be bad for non-stop, as it'd open a window where the inferior could miss the breakpoint. The conditions actually haven't changed between the resends, but GDB isn't smart enough to realize that. (TBC, if the target doesn't support target-side conditions, then GDB doesn't trigger these resends (init_bp_location calls mark_breakpoint_location_modified, and that does nothing if condition evaluation is on the host side. The resends are caused by the 'loc->condition_changed = condition_modified.' line.) But, even if GDB was made smarter, GDBserver should really still handle the resends anyway. So target-side conditions also aren't really to blame. The documentation of the Z/z packets says: "To avoid potential problems with duplicate packets, the operations should be implemented in an idempotent way." As such, we may want to fix GDB, but we should definitely fix GDBserver. The fix is a prerequisite for target-side conditions on hardware breakpoints anyway (and while at it, on watchpoints too). GDBserver indeed already treats duplicate Z0 packets in an idempotent way. mem-break.c has the concept of high-level and low-level breakpoints, somewhat similar to GDB's split of breakpoints vs breakpoint locations, and keeps track of multiple breakpoints referencing the same address/location, for the case of an internal GDBserver breakpoint or a tracepoint being set at the same address as a GDB breakpoint. But, it only allows GDB to ever contribute one reference to a software breakpoint location. IOW, if gdbserver sees a Z0 packet for the same address where it already had a GDB breakpoint set, then GDBserver won't create another high-level GDB breakpoint. However, mem-break.c only tracks GDB Z0 breakpoints. The same logic should apply to all kinds of Zx packets. Currently, gdbserver passes down each duplicate Zx (other than Z0) request directly to the target->insert_point routine. The x86 watchpoint support itself refcounts watchpoint / hw breakpoint requests, to handle overlapping watchpoints, and save debug registers. But that code doesn't (and really shouldn't) handle the duplicate requests, assuming that for each insert there will be a corresponding remove. So the fix is to generalize mem-break.c to track all kinds of Zx breakpoints, and filter out duplicates. As mentioned, this ends up adding support for target-side conditions on hardware breakpoints and watchpoints too (though GDB itself doesn't support the latter yet). Probably the least obvious change in the patch is that it kind of turns the breakpoint insert/remove APIs inside out. Before, the target methods were only called for GDB breakpoints. The internal breakpoint set/delete methods inserted memory breakpoints directly bypassing the insert/remove target methods. That's not good when the target should use a debug API to set software breakpoints, instead of relying on GDBserver patching memory with breakpoint instructions, as is the case of NTO. Now removal/insertion of all kinds of breakpoints/watchpoints, either internal, or from GDB, always go through the target methods. The insert_point/remove_point methods no longer get passed a Z packet type, but an internal/raw breakpoint type. They're also passed a pointer to the raw breakpoint itself (note that's still opaque outside mem-break.c), so that insert_memory_breakpoint / remove_memory_breakpoint have access to the breakpoint's shadow buffer. I first tried passing down a new structure based on GDB's "struct bp_target_info" (actually with that name exactly), but then decided against it as unnecessary complication. As software/memory breakpoints work by poking at memory, when setting a GDB Z0 breakpoint (but not internal breakpoints, as those can assume the conditions are already right), we need to tell the target to prepare to access memory (which on Linux means stop threads). If that operation fails, we need to return error to GDB. Seeing an error, if this is the first breakpoint of that type that GDB tries to insert, GDB would then assume the breakpoint type is supported, but it may actually not be. So we need to check whether the type is supported at all before preparing to access memory. And to solve that, the patch adds a new target->supports_z_point_type method that is called before actually trying to insert the breakpoint. Other than that, hopefully the change is more or less obvious. New test added that exercises the hbreak2.exp regression in a more direct way, without relying on a breakpoint re-set happening before main is reached. Tested by building GDBserver for: aarch64-linux-gnu arm-linux-gnueabihf i686-pc-linux-gnu i686-w64-mingw32 m68k-linux-gnu mips-linux-gnu mips-uclinux nios2-linux-gnu powerpc-linux-gnu sh-linux-gnu tilegx-unknown-linux-gnu x86_64-redhat-linux x86_64-w64-mingw32 And also regression tested on x86_64 Fedora 20. gdb/gdbserver/ 2014-05-20 Pedro Alves <palves@redhat.com> * linux-aarch64-low.c (aarch64_insert_point) (aarch64_remove_point): No longer check whether the type is supported here. Adjust to new interface. (the_low_target): Install aarch64_supports_z_point_type as supports_z_point_type method. * linux-arm-low.c (raw_bkpt_type_to_arm_hwbp_type): New function. (arm_linux_hw_point_initialize): Take an enum raw_bkpt_type instead of a Z packet char. Adjust. (arm_supports_z_point_type): New function. (arm_insert_point, arm_remove_point): Adjust to new interface. (the_low_target): Install arm_supports_z_point_type. * linux-crisv32-low.c (cris_supports_z_point_type): New function. (cris_insert_point, cris_remove_point): Adjust to new interface. Don't check whether the type is supported here. (the_low_target): Install cris_supports_z_point_type. * linux-low.c (linux_supports_z_point_type): New function. (linux_insert_point, linux_remove_point): Adjust to new interface. * linux-low.h (struct linux_target_ops) <insert_point, remove_point>: Take an enum raw_bkpt_type instead of a char. Add raw_breakpoint pointer parameter. <supports_z_point_type>: New method. * linux-mips-low.c (mips_supports_z_point_type): New function. (mips_insert_point, mips_remove_point): Adjust to new interface. Use mips_supports_z_point_type. (the_low_target): Install mips_supports_z_point_type. * linux-ppc-low.c (the_low_target): Install NULL as supports_z_point_type method. * linux-s390-low.c (the_low_target): Install NULL as supports_z_point_type method. * linux-sparc-low.c (the_low_target): Install NULL as supports_z_point_type method. * linux-x86-low.c (x86_supports_z_point_type): New function. (x86_insert_point): Adjust to new insert_point interface. Use insert_memory_breakpoint. Adjust to new i386_low_insert_watchpoint interface. (x86_remove_point): Adjust to remove_point interface. Use remove_memory_breakpoint. Adjust to new i386_low_remove_watchpoint interface. (the_low_target): Install x86_supports_z_point_type. * lynx-low.c (lynx_target_ops): Install NULL as supports_z_point_type callback. * nto-low.c (nto_supports_z_point_type): New. (nto_insert_point, nto_remove_point): Adjust to new interface. (nto_target_ops): Install nto_supports_z_point_type. * mem-break.c: Adjust intro comment. (struct raw_breakpoint) <raw_type, size>: New fields. <inserted>: Update comment. <shlib_disabled>: Delete field. (enum bkpt_type) <gdb_breakpoint>: Delete value. <gdb_breakpoint_Z0, gdb_breakpoint_Z1, gdb_breakpoint_Z2, gdb_breakpoint_Z3, gdb_breakpoint_Z4>: New values. (raw_bkpt_type_to_target_hw_bp_type): New function. (find_enabled_raw_code_breakpoint_at): New function. (find_raw_breakpoint_at): New type and size parameters. Use them. (insert_memory_breakpoint): New function, based off set_raw_breakpoint_at. (remove_memory_breakpoint): New function. (set_raw_breakpoint_at): Reimplement. (set_breakpoint): New, based on set_breakpoint_at. (set_breakpoint_at): Reimplement. (delete_raw_breakpoint): Go through the_target->remove_point instead of assuming memory breakpoints. (find_gdb_breakpoint_at): Delete. (Z_packet_to_bkpt_type, Z_packet_to_raw_bkpt_type): New functions. (find_gdb_breakpoint): New function. (set_gdb_breakpoint_at): Delete. (z_type_supported): New function. (set_gdb_breakpoint_1): New function, loosely based off set_gdb_breakpoint_at. (check_gdb_bp_preconditions, set_gdb_breakpoint): New functions. (delete_gdb_breakpoint_at): Delete. (delete_gdb_breakpoint_1): New function, loosely based off delete_gdb_breakpoint_at. (delete_gdb_breakpoint): New function. (clear_gdb_breakpoint_conditions): Rename to ... (clear_breakpoint_conditions): ... this. Don't handle a NULL breakpoint. (add_condition_to_breakpoint): Make static. (add_breakpoint_condition): Take a struct breakpoint pointer instead of an address. Adjust. (gdb_condition_true_at_breakpoint): Rename to ... (gdb_condition_true_at_breakpoint_z_type): ... this, and add z_type parameter. (gdb_condition_true_at_breakpoint): Reimplement. (add_breakpoint_commands): Take a struct breakpoint pointer instead of an address. Adjust. (gdb_no_commands_at_breakpoint): Rename to ... (gdb_no_commands_at_breakpoint_z_type): ... this. Add z_type parameter. Return true if no breakpoint was found. Change debug output. (gdb_no_commands_at_breakpoint): Reimplement. (run_breakpoint_commands): Rename to ... (run_breakpoint_commands_z_type): ... this. Add z_type parameter, and change return type to boolean. (run_breakpoint_commands): New function. (gdb_breakpoint_here): Also check for Z1 breakpoints. (uninsert_raw_breakpoint): Don't try to reinsert a disabled breakpoint. Go through the_target->remove_point instead of assuming memory breakpoint. (uninsert_breakpoints_at, uninsert_all_breakpoints): Uninsert software and hardware breakpoints. (reinsert_raw_breakpoint): Go through the_target->insert_point instead of assuming memory breakpoint. (reinsert_breakpoints_at, reinsert_all_breakpoints): Reinsert software and hardware breakpoints. (check_breakpoints, breakpoint_here, breakpoint_inserted_here): Check both software and hardware breakpoints. (validate_inserted_breakpoint): Assert the breakpoint is a software breakpoint. Set the inserted flag to -1 instead of setting shlib_disabled. (delete_disabled_breakpoints): Adjust. (validate_breakpoints): Only validate software breakpoints. Adjust to inserted flag change. (check_mem_read, check_mem_write): Skip breakpoint types other than software breakpoints. Adjust to inserted flag change. * mem-break.h (enum raw_bkpt_type): New enum. (raw_breakpoint, struct process_info): Forward declare. (Z_packet_to_target_hw_bp_type): Delete declaration. (raw_bkpt_type_to_target_hw_bp_type, Z_packet_to_raw_bkpt_type) (set_gdb_breakpoint, delete_gdb_breakpoint) (clear_breakpoint_conditions): New declarations. (set_gdb_breakpoint_at, clear_gdb_breakpoint_conditions): Delete. (breakpoint_inserted_here): Update comment. (add_breakpoint_condition, add_breakpoint_commands): Replace address parameter with a breakpoint pointer parameter. (gdb_breakpoint_here): Update comment. (delete_gdb_breakpoint_at): Delete. (insert_memory_breakpoint, remove_memory_breakpoint): Declare. * server.c (process_point_options): Take a struct breakpoint pointer instead of an address. Adjust. (process_serial_event) <Z/z packets>: Use set_gdb_breakpoint and delete_gdb_breakpoint. * spu-low.c (spu_target_ops): Install NULL as supports_z_point_type method. * target.h: Include mem-break.h. (struct target_ops) <prepare_to_access_memory>: Update comment. <supports_z_point_type>: New field. <insert_point, remove_point>: Take an enum raw_bkpt_type argument instead of a char. Also take a raw breakpoint pointer. * win32-arm-low.c (the_low_target): Install NULL as supports_z_point_type. * win32-i386-low.c (i386_supports_z_point_type): New function. (i386_insert_point, i386_remove_point): Adjust to new interface. (the_low_target): Install i386_supports_z_point_type. * win32-low.c (win32_supports_z_point_type): New function. (win32_insert_point, win32_remove_point): Adjust to new interface. (win32_target_ops): Install win32_supports_z_point_type. * win32-low.h (struct win32_target_ops): <supports_z_point_type>: New method. <insert_point, remove_point>: Take an enum raw_bkpt_type argument instead of a char. Also take a raw breakpoint pointer. gdb/testsuite/ 2014-05-20 Pedro Alves <palves@redhat.com> * gdb.base/break-idempotent.c: New file. * gdb.base/break-idempotent.exp: New file.
2014-05-20 19:24:28 +02:00
/* Hardware-assisted read watchpoint. */
raw_bkpt_type_read_wp,
/* Hardware-assisted access watchpoint. */
raw_bkpt_type_access_wp
};
/* Map the protocol breakpoint/watchpoint type Z_TYPE to the internal
raw breakpoint type. */
enum raw_bkpt_type Z_packet_to_raw_bkpt_type (char z_type);
/* Map a raw breakpoint type to an enum target_hw_bp_type. */
enum target_hw_bp_type raw_bkpt_type_to_target_hw_bp_type
(enum raw_bkpt_type raw_type);
Support breakpoint kinds for software breakpoints in GDBServer. There's two ways to set breakpoints in GDBServer. - GDBServer setting its own breakpoints, through API set_breakpoint_at. - GDBServer setting breakpoints according to the information in Z packets, through API set_gdb_breakpoint. Before this patch the breakpoint kinds were a concept unique to GDB and Z packets, as GDBServer never had to set different kinds of breakpoint on its own. This patch teaches GDBServer to handle breakpoint kinds for its own breakpoints. It generalizes the breakpoint kind as per Z packets to represent different kinds of breakpoints directly set by GDBServer also. GDBServer now querys breakpoint_kind_from_pc to know what breakpoint kind to set on its own. As the kind is now a differentiating factor equivalent to size for the breakpoint struct and that it's size can be queried using sw_breakpoint_from_kind, the size field has been replaced with the kind field. All references to size are now replaced by kind or a call to bp_size that wraps sw_breakpoing_from_kind and returns the size of the breakpoint in memory. To fetch the software breakpoint data bp_opcode is called and wraps the sw_breakpoint_from_kind call. No regressions on Ubuntu 14.04 on ARMv7 and x86. With gdbserver-{native,extended} / { -marm -mthumb } gdb/gdbserver/ChangeLog: * linux-low.c (initialize_low): Ajdust for breakpoint global variables removal. * mem-break.c : Remove breakpoint_data/breakpoint_len global variables. (struct raw_breakpoint) <size>: Remove. (struct raw_breakpoint) <kind>: Add. (bp_size): New function. (bp_opcode): Likewise. (find_raw_breakpoint_at): Adjust for kind. (insert_memory_breakpoint): Adjust for kind call bp_size,bp_opcode. (remove_memory_breakpoint): Adjust for kind call bp_size. (set_raw_breakpoint_at): Adjust for kind. (set_breakpoint): Likewise. (set_breakpoint_at): Call breakpoint_kind_from_pc. (delete_raw_breakpoint): Adjust for kind. (delete_breakpoint): Likewise. (find_gdb_breakpoint): Likewise. (set_gdb_breakpoint_1): Likewise. (set_gdb_breakpoint): Likewise. (delete_gdb_breakpoint_1): Likewise. (delete_gdb_breakpoint): Likewise. (uninsert_raw_breakpoint): Likewise. (reinsert_raw_breakpoint): Likewise. (set_breakpoint_data): Remove. (validate_inserted_breakpoint): Adjust for kind call bp_size,bp_opcode. (check_mem_read): Adjust for kind call bp_size. (check_mem_write): Adjust for kind call bp_size,bp_opcode. (clone_one_breakpoint): Adjust for kind. * mem-break.h (set_gdb_breakpoint): Likewise. (delete_gdb_breakpoint): Likewise. * server.c (process_serial_event): Likewise.
2015-10-21 17:13:40 +02:00
/* Create a new GDB breakpoint of type Z_TYPE at ADDR with kind KIND.
[GDBserver] Make Zx/zx packet handling idempotent. This patch fixes hardware breakpoint regressions exposed by my fix for "PR breakpoints/7143 - Watchpoint does not trigger when first set", at https://sourceware.org/ml/gdb-patches/2014-03/msg00167.html The testsuite caught them on Linux/x86_64, at least. gdb.sum: gdb.sum: FAIL: gdb.base/hbreak2.exp: next over recursive call FAIL: gdb.base/hbreak2.exp: backtrace from factorial(5.1) FAIL: gdb.base/hbreak2.exp: continue until exit at recursive next test gdb.log: (gdb) next Program received signal SIGTRAP, Trace/breakpoint trap. factorial (value=4) at ../../../src/gdb/testsuite/gdb.base/break.c:113 113 if (value > 1) { /* set breakpoint 7 here */ (gdb) FAIL: gdb.base/hbreak2.exp: next over recursive call Actually, that patch just exposed a latent issue to "breakpoints always-inserted off" mode, not really caused it. After that patch, GDB no longer removes breakpoints at each internal event, thus making some scenarios behave like breakpoint always-inserted on. The bug is easy to trigger with always-inserted on. The issue is that since the target-side breakpoint conditions support, if the stub/server supports evaluating breakpoint conditions on the target side, then GDB is sending duplicate Zx packets to the target without removing them before, and GDBserver is not really expecting that for Z packets other than Z0/z0. E.g., with "set breakpoint always-inserted on" and "set debug remote 1": (gdb) b main Sending packet: $m410943,1#ff...Packet received: 48 Breakpoint 4 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z0,410943,1#48...Packet received: OK ^^^^^^^^^^^^ (gdb) b main Note: breakpoint 4 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Breakpoint 5 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z0,410943,1#48...Packet received: OK ^^^^^^^^^^^^ (gdb) b main Note: breakpoints 4 and 5 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Breakpoint 6 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z0,410943,1#48...Packet received: OK ^^^^^^^^^^^^ (gdb) del Delete all breakpoints? (y or n) y Sending packet: $Z0,410943,1#48...Packet received: OK Sending packet: $Z0,410943,1#48...Packet received: OK Sending packet: $z0,410943,1#68...Packet received: OK And for Z1, similarly: (gdb) hbreak main Sending packet: $m410943,1#ff...Packet received: 48 Hardware assisted breakpoint 4 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ Packet Z1 (hardware-breakpoint) is supported (gdb) hbreak main Note: breakpoint 4 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Hardware assisted breakpoint 5 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ (gdb) hbreak main Note: breakpoints 4 and 5 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Hardware assisted breakpoint 6 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ (gdb) del Delete all breakpoints? (y or n) y Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ Sending packet: $z1,410943,1#69...Packet received: OK ^^^^^^^^^^^^ So GDB sent a bunch of Z1 packets, and then when finally removing the breakpoint, only one z1 packet was sent. On the GDBserver side (with monitor set debug-hw-points 1), in the Z1 case, we see: $ ./gdbserver :9999 ./gdbserver Process ./gdbserver created; pid = 8629 Listening on port 9999 Remote debugging from host 127.0.0.1 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=1 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=2 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=3 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=4 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=5 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 remove_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=4 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 That's one insert_watchpoint call for each Z1 packet, and then one remove_watchpoint call for the z1 packet. Notice how ref.count increased for each insert_watchpoint call, and then in the end, after GDB told GDBserver to forget about the hardware breakpoint, GDBserver ends with the the first debug register still with ref.count=4! IOW, the hardware breakpoint is left armed on the target, while on the GDB end it's gone. If the program happens to execute 0x410943 afterwards, then the CPU traps, GDBserver reports the trap to GDB, and GDB not having a breakpoint set at that address anymore, reports to the user a spurious SIGTRAP. This is exactly what is happening in the hbreak2.exp test, though in that case, it's a shared library event that triggers a breakpoint_re_set, when breakpoints are still inserted (because nowadays GDB doesn't remove breakpoints while handling internal events), and that recreates breakpoint locations, which likewise forces breakpoint reinsertion and Zx packet resends... That is a lot of bogus Zx duplication that should possibly be addressed on the GDB side. GDB resends Zx packets because the way to change the target-side condition, is to resend the breakpoint to the server with the new condition. (That's an option in the packet: e.g., "Z1,410943,1;X3,220027" for "hbreak main if 0". The packets in the examples above are shorter because the breakpoints don't have conditions attached). GDB doesn't remove the breakpoint first before reinserting it because that'd be bad for non-stop, as it'd open a window where the inferior could miss the breakpoint. The conditions actually haven't changed between the resends, but GDB isn't smart enough to realize that. (TBC, if the target doesn't support target-side conditions, then GDB doesn't trigger these resends (init_bp_location calls mark_breakpoint_location_modified, and that does nothing if condition evaluation is on the host side. The resends are caused by the 'loc->condition_changed = condition_modified.' line.) But, even if GDB was made smarter, GDBserver should really still handle the resends anyway. So target-side conditions also aren't really to blame. The documentation of the Z/z packets says: "To avoid potential problems with duplicate packets, the operations should be implemented in an idempotent way." As such, we may want to fix GDB, but we should definitely fix GDBserver. The fix is a prerequisite for target-side conditions on hardware breakpoints anyway (and while at it, on watchpoints too). GDBserver indeed already treats duplicate Z0 packets in an idempotent way. mem-break.c has the concept of high-level and low-level breakpoints, somewhat similar to GDB's split of breakpoints vs breakpoint locations, and keeps track of multiple breakpoints referencing the same address/location, for the case of an internal GDBserver breakpoint or a tracepoint being set at the same address as a GDB breakpoint. But, it only allows GDB to ever contribute one reference to a software breakpoint location. IOW, if gdbserver sees a Z0 packet for the same address where it already had a GDB breakpoint set, then GDBserver won't create another high-level GDB breakpoint. However, mem-break.c only tracks GDB Z0 breakpoints. The same logic should apply to all kinds of Zx packets. Currently, gdbserver passes down each duplicate Zx (other than Z0) request directly to the target->insert_point routine. The x86 watchpoint support itself refcounts watchpoint / hw breakpoint requests, to handle overlapping watchpoints, and save debug registers. But that code doesn't (and really shouldn't) handle the duplicate requests, assuming that for each insert there will be a corresponding remove. So the fix is to generalize mem-break.c to track all kinds of Zx breakpoints, and filter out duplicates. As mentioned, this ends up adding support for target-side conditions on hardware breakpoints and watchpoints too (though GDB itself doesn't support the latter yet). Probably the least obvious change in the patch is that it kind of turns the breakpoint insert/remove APIs inside out. Before, the target methods were only called for GDB breakpoints. The internal breakpoint set/delete methods inserted memory breakpoints directly bypassing the insert/remove target methods. That's not good when the target should use a debug API to set software breakpoints, instead of relying on GDBserver patching memory with breakpoint instructions, as is the case of NTO. Now removal/insertion of all kinds of breakpoints/watchpoints, either internal, or from GDB, always go through the target methods. The insert_point/remove_point methods no longer get passed a Z packet type, but an internal/raw breakpoint type. They're also passed a pointer to the raw breakpoint itself (note that's still opaque outside mem-break.c), so that insert_memory_breakpoint / remove_memory_breakpoint have access to the breakpoint's shadow buffer. I first tried passing down a new structure based on GDB's "struct bp_target_info" (actually with that name exactly), but then decided against it as unnecessary complication. As software/memory breakpoints work by poking at memory, when setting a GDB Z0 breakpoint (but not internal breakpoints, as those can assume the conditions are already right), we need to tell the target to prepare to access memory (which on Linux means stop threads). If that operation fails, we need to return error to GDB. Seeing an error, if this is the first breakpoint of that type that GDB tries to insert, GDB would then assume the breakpoint type is supported, but it may actually not be. So we need to check whether the type is supported at all before preparing to access memory. And to solve that, the patch adds a new target->supports_z_point_type method that is called before actually trying to insert the breakpoint. Other than that, hopefully the change is more or less obvious. New test added that exercises the hbreak2.exp regression in a more direct way, without relying on a breakpoint re-set happening before main is reached. Tested by building GDBserver for: aarch64-linux-gnu arm-linux-gnueabihf i686-pc-linux-gnu i686-w64-mingw32 m68k-linux-gnu mips-linux-gnu mips-uclinux nios2-linux-gnu powerpc-linux-gnu sh-linux-gnu tilegx-unknown-linux-gnu x86_64-redhat-linux x86_64-w64-mingw32 And also regression tested on x86_64 Fedora 20. gdb/gdbserver/ 2014-05-20 Pedro Alves <palves@redhat.com> * linux-aarch64-low.c (aarch64_insert_point) (aarch64_remove_point): No longer check whether the type is supported here. Adjust to new interface. (the_low_target): Install aarch64_supports_z_point_type as supports_z_point_type method. * linux-arm-low.c (raw_bkpt_type_to_arm_hwbp_type): New function. (arm_linux_hw_point_initialize): Take an enum raw_bkpt_type instead of a Z packet char. Adjust. (arm_supports_z_point_type): New function. (arm_insert_point, arm_remove_point): Adjust to new interface. (the_low_target): Install arm_supports_z_point_type. * linux-crisv32-low.c (cris_supports_z_point_type): New function. (cris_insert_point, cris_remove_point): Adjust to new interface. Don't check whether the type is supported here. (the_low_target): Install cris_supports_z_point_type. * linux-low.c (linux_supports_z_point_type): New function. (linux_insert_point, linux_remove_point): Adjust to new interface. * linux-low.h (struct linux_target_ops) <insert_point, remove_point>: Take an enum raw_bkpt_type instead of a char. Add raw_breakpoint pointer parameter. <supports_z_point_type>: New method. * linux-mips-low.c (mips_supports_z_point_type): New function. (mips_insert_point, mips_remove_point): Adjust to new interface. Use mips_supports_z_point_type. (the_low_target): Install mips_supports_z_point_type. * linux-ppc-low.c (the_low_target): Install NULL as supports_z_point_type method. * linux-s390-low.c (the_low_target): Install NULL as supports_z_point_type method. * linux-sparc-low.c (the_low_target): Install NULL as supports_z_point_type method. * linux-x86-low.c (x86_supports_z_point_type): New function. (x86_insert_point): Adjust to new insert_point interface. Use insert_memory_breakpoint. Adjust to new i386_low_insert_watchpoint interface. (x86_remove_point): Adjust to remove_point interface. Use remove_memory_breakpoint. Adjust to new i386_low_remove_watchpoint interface. (the_low_target): Install x86_supports_z_point_type. * lynx-low.c (lynx_target_ops): Install NULL as supports_z_point_type callback. * nto-low.c (nto_supports_z_point_type): New. (nto_insert_point, nto_remove_point): Adjust to new interface. (nto_target_ops): Install nto_supports_z_point_type. * mem-break.c: Adjust intro comment. (struct raw_breakpoint) <raw_type, size>: New fields. <inserted>: Update comment. <shlib_disabled>: Delete field. (enum bkpt_type) <gdb_breakpoint>: Delete value. <gdb_breakpoint_Z0, gdb_breakpoint_Z1, gdb_breakpoint_Z2, gdb_breakpoint_Z3, gdb_breakpoint_Z4>: New values. (raw_bkpt_type_to_target_hw_bp_type): New function. (find_enabled_raw_code_breakpoint_at): New function. (find_raw_breakpoint_at): New type and size parameters. Use them. (insert_memory_breakpoint): New function, based off set_raw_breakpoint_at. (remove_memory_breakpoint): New function. (set_raw_breakpoint_at): Reimplement. (set_breakpoint): New, based on set_breakpoint_at. (set_breakpoint_at): Reimplement. (delete_raw_breakpoint): Go through the_target->remove_point instead of assuming memory breakpoints. (find_gdb_breakpoint_at): Delete. (Z_packet_to_bkpt_type, Z_packet_to_raw_bkpt_type): New functions. (find_gdb_breakpoint): New function. (set_gdb_breakpoint_at): Delete. (z_type_supported): New function. (set_gdb_breakpoint_1): New function, loosely based off set_gdb_breakpoint_at. (check_gdb_bp_preconditions, set_gdb_breakpoint): New functions. (delete_gdb_breakpoint_at): Delete. (delete_gdb_breakpoint_1): New function, loosely based off delete_gdb_breakpoint_at. (delete_gdb_breakpoint): New function. (clear_gdb_breakpoint_conditions): Rename to ... (clear_breakpoint_conditions): ... this. Don't handle a NULL breakpoint. (add_condition_to_breakpoint): Make static. (add_breakpoint_condition): Take a struct breakpoint pointer instead of an address. Adjust. (gdb_condition_true_at_breakpoint): Rename to ... (gdb_condition_true_at_breakpoint_z_type): ... this, and add z_type parameter. (gdb_condition_true_at_breakpoint): Reimplement. (add_breakpoint_commands): Take a struct breakpoint pointer instead of an address. Adjust. (gdb_no_commands_at_breakpoint): Rename to ... (gdb_no_commands_at_breakpoint_z_type): ... this. Add z_type parameter. Return true if no breakpoint was found. Change debug output. (gdb_no_commands_at_breakpoint): Reimplement. (run_breakpoint_commands): Rename to ... (run_breakpoint_commands_z_type): ... this. Add z_type parameter, and change return type to boolean. (run_breakpoint_commands): New function. (gdb_breakpoint_here): Also check for Z1 breakpoints. (uninsert_raw_breakpoint): Don't try to reinsert a disabled breakpoint. Go through the_target->remove_point instead of assuming memory breakpoint. (uninsert_breakpoints_at, uninsert_all_breakpoints): Uninsert software and hardware breakpoints. (reinsert_raw_breakpoint): Go through the_target->insert_point instead of assuming memory breakpoint. (reinsert_breakpoints_at, reinsert_all_breakpoints): Reinsert software and hardware breakpoints. (check_breakpoints, breakpoint_here, breakpoint_inserted_here): Check both software and hardware breakpoints. (validate_inserted_breakpoint): Assert the breakpoint is a software breakpoint. Set the inserted flag to -1 instead of setting shlib_disabled. (delete_disabled_breakpoints): Adjust. (validate_breakpoints): Only validate software breakpoints. Adjust to inserted flag change. (check_mem_read, check_mem_write): Skip breakpoint types other than software breakpoints. Adjust to inserted flag change. * mem-break.h (enum raw_bkpt_type): New enum. (raw_breakpoint, struct process_info): Forward declare. (Z_packet_to_target_hw_bp_type): Delete declaration. (raw_bkpt_type_to_target_hw_bp_type, Z_packet_to_raw_bkpt_type) (set_gdb_breakpoint, delete_gdb_breakpoint) (clear_breakpoint_conditions): New declarations. (set_gdb_breakpoint_at, clear_gdb_breakpoint_conditions): Delete. (breakpoint_inserted_here): Update comment. (add_breakpoint_condition, add_breakpoint_commands): Replace address parameter with a breakpoint pointer parameter. (gdb_breakpoint_here): Update comment. (delete_gdb_breakpoint_at): Delete. (insert_memory_breakpoint, remove_memory_breakpoint): Declare. * server.c (process_point_options): Take a struct breakpoint pointer instead of an address. Adjust. (process_serial_event) <Z/z packets>: Use set_gdb_breakpoint and delete_gdb_breakpoint. * spu-low.c (spu_target_ops): Install NULL as supports_z_point_type method. * target.h: Include mem-break.h. (struct target_ops) <prepare_to_access_memory>: Update comment. <supports_z_point_type>: New field. <insert_point, remove_point>: Take an enum raw_bkpt_type argument instead of a char. Also take a raw breakpoint pointer. * win32-arm-low.c (the_low_target): Install NULL as supports_z_point_type. * win32-i386-low.c (i386_supports_z_point_type): New function. (i386_insert_point, i386_remove_point): Adjust to new interface. (the_low_target): Install i386_supports_z_point_type. * win32-low.c (win32_supports_z_point_type): New function. (win32_insert_point, win32_remove_point): Adjust to new interface. (win32_target_ops): Install win32_supports_z_point_type. * win32-low.h (struct win32_target_ops): <supports_z_point_type>: New method. <insert_point, remove_point>: Take an enum raw_bkpt_type argument instead of a char. Also take a raw breakpoint pointer. gdb/testsuite/ 2014-05-20 Pedro Alves <palves@redhat.com> * gdb.base/break-idempotent.c: New file. * gdb.base/break-idempotent.exp: New file.
2014-05-20 19:24:28 +02:00
Returns a pointer to the newly created breakpoint on success. On
failure returns NULL and sets *ERR to either -1 for error, or 1 if
Z_TYPE breakpoints are not supported on this target. */
Support breakpoint kinds for software breakpoints in GDBServer. There's two ways to set breakpoints in GDBServer. - GDBServer setting its own breakpoints, through API set_breakpoint_at. - GDBServer setting breakpoints according to the information in Z packets, through API set_gdb_breakpoint. Before this patch the breakpoint kinds were a concept unique to GDB and Z packets, as GDBServer never had to set different kinds of breakpoint on its own. This patch teaches GDBServer to handle breakpoint kinds for its own breakpoints. It generalizes the breakpoint kind as per Z packets to represent different kinds of breakpoints directly set by GDBServer also. GDBServer now querys breakpoint_kind_from_pc to know what breakpoint kind to set on its own. As the kind is now a differentiating factor equivalent to size for the breakpoint struct and that it's size can be queried using sw_breakpoint_from_kind, the size field has been replaced with the kind field. All references to size are now replaced by kind or a call to bp_size that wraps sw_breakpoing_from_kind and returns the size of the breakpoint in memory. To fetch the software breakpoint data bp_opcode is called and wraps the sw_breakpoint_from_kind call. No regressions on Ubuntu 14.04 on ARMv7 and x86. With gdbserver-{native,extended} / { -marm -mthumb } gdb/gdbserver/ChangeLog: * linux-low.c (initialize_low): Ajdust for breakpoint global variables removal. * mem-break.c : Remove breakpoint_data/breakpoint_len global variables. (struct raw_breakpoint) <size>: Remove. (struct raw_breakpoint) <kind>: Add. (bp_size): New function. (bp_opcode): Likewise. (find_raw_breakpoint_at): Adjust for kind. (insert_memory_breakpoint): Adjust for kind call bp_size,bp_opcode. (remove_memory_breakpoint): Adjust for kind call bp_size. (set_raw_breakpoint_at): Adjust for kind. (set_breakpoint): Likewise. (set_breakpoint_at): Call breakpoint_kind_from_pc. (delete_raw_breakpoint): Adjust for kind. (delete_breakpoint): Likewise. (find_gdb_breakpoint): Likewise. (set_gdb_breakpoint_1): Likewise. (set_gdb_breakpoint): Likewise. (delete_gdb_breakpoint_1): Likewise. (delete_gdb_breakpoint): Likewise. (uninsert_raw_breakpoint): Likewise. (reinsert_raw_breakpoint): Likewise. (set_breakpoint_data): Remove. (validate_inserted_breakpoint): Adjust for kind call bp_size,bp_opcode. (check_mem_read): Adjust for kind call bp_size. (check_mem_write): Adjust for kind call bp_size,bp_opcode. (clone_one_breakpoint): Adjust for kind. * mem-break.h (set_gdb_breakpoint): Likewise. (delete_gdb_breakpoint): Likewise. * server.c (process_serial_event): Likewise.
2015-10-21 17:13:40 +02:00
struct breakpoint *set_gdb_breakpoint (char z_type, CORE_ADDR addr, int kind,
[GDBserver] Make Zx/zx packet handling idempotent. This patch fixes hardware breakpoint regressions exposed by my fix for "PR breakpoints/7143 - Watchpoint does not trigger when first set", at https://sourceware.org/ml/gdb-patches/2014-03/msg00167.html The testsuite caught them on Linux/x86_64, at least. gdb.sum: gdb.sum: FAIL: gdb.base/hbreak2.exp: next over recursive call FAIL: gdb.base/hbreak2.exp: backtrace from factorial(5.1) FAIL: gdb.base/hbreak2.exp: continue until exit at recursive next test gdb.log: (gdb) next Program received signal SIGTRAP, Trace/breakpoint trap. factorial (value=4) at ../../../src/gdb/testsuite/gdb.base/break.c:113 113 if (value > 1) { /* set breakpoint 7 here */ (gdb) FAIL: gdb.base/hbreak2.exp: next over recursive call Actually, that patch just exposed a latent issue to "breakpoints always-inserted off" mode, not really caused it. After that patch, GDB no longer removes breakpoints at each internal event, thus making some scenarios behave like breakpoint always-inserted on. The bug is easy to trigger with always-inserted on. The issue is that since the target-side breakpoint conditions support, if the stub/server supports evaluating breakpoint conditions on the target side, then GDB is sending duplicate Zx packets to the target without removing them before, and GDBserver is not really expecting that for Z packets other than Z0/z0. E.g., with "set breakpoint always-inserted on" and "set debug remote 1": (gdb) b main Sending packet: $m410943,1#ff...Packet received: 48 Breakpoint 4 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z0,410943,1#48...Packet received: OK ^^^^^^^^^^^^ (gdb) b main Note: breakpoint 4 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Breakpoint 5 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z0,410943,1#48...Packet received: OK ^^^^^^^^^^^^ (gdb) b main Note: breakpoints 4 and 5 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Breakpoint 6 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z0,410943,1#48...Packet received: OK ^^^^^^^^^^^^ (gdb) del Delete all breakpoints? (y or n) y Sending packet: $Z0,410943,1#48...Packet received: OK Sending packet: $Z0,410943,1#48...Packet received: OK Sending packet: $z0,410943,1#68...Packet received: OK And for Z1, similarly: (gdb) hbreak main Sending packet: $m410943,1#ff...Packet received: 48 Hardware assisted breakpoint 4 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ Packet Z1 (hardware-breakpoint) is supported (gdb) hbreak main Note: breakpoint 4 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Hardware assisted breakpoint 5 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ (gdb) hbreak main Note: breakpoints 4 and 5 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Hardware assisted breakpoint 6 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ (gdb) del Delete all breakpoints? (y or n) y Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ Sending packet: $z1,410943,1#69...Packet received: OK ^^^^^^^^^^^^ So GDB sent a bunch of Z1 packets, and then when finally removing the breakpoint, only one z1 packet was sent. On the GDBserver side (with monitor set debug-hw-points 1), in the Z1 case, we see: $ ./gdbserver :9999 ./gdbserver Process ./gdbserver created; pid = 8629 Listening on port 9999 Remote debugging from host 127.0.0.1 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=1 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=2 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=3 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=4 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=5 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 remove_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=4 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 That's one insert_watchpoint call for each Z1 packet, and then one remove_watchpoint call for the z1 packet. Notice how ref.count increased for each insert_watchpoint call, and then in the end, after GDB told GDBserver to forget about the hardware breakpoint, GDBserver ends with the the first debug register still with ref.count=4! IOW, the hardware breakpoint is left armed on the target, while on the GDB end it's gone. If the program happens to execute 0x410943 afterwards, then the CPU traps, GDBserver reports the trap to GDB, and GDB not having a breakpoint set at that address anymore, reports to the user a spurious SIGTRAP. This is exactly what is happening in the hbreak2.exp test, though in that case, it's a shared library event that triggers a breakpoint_re_set, when breakpoints are still inserted (because nowadays GDB doesn't remove breakpoints while handling internal events), and that recreates breakpoint locations, which likewise forces breakpoint reinsertion and Zx packet resends... That is a lot of bogus Zx duplication that should possibly be addressed on the GDB side. GDB resends Zx packets because the way to change the target-side condition, is to resend the breakpoint to the server with the new condition. (That's an option in the packet: e.g., "Z1,410943,1;X3,220027" for "hbreak main if 0". The packets in the examples above are shorter because the breakpoints don't have conditions attached). GDB doesn't remove the breakpoint first before reinserting it because that'd be bad for non-stop, as it'd open a window where the inferior could miss the breakpoint. The conditions actually haven't changed between the resends, but GDB isn't smart enough to realize that. (TBC, if the target doesn't support target-side conditions, then GDB doesn't trigger these resends (init_bp_location calls mark_breakpoint_location_modified, and that does nothing if condition evaluation is on the host side. The resends are caused by the 'loc->condition_changed = condition_modified.' line.) But, even if GDB was made smarter, GDBserver should really still handle the resends anyway. So target-side conditions also aren't really to blame. The documentation of the Z/z packets says: "To avoid potential problems with duplicate packets, the operations should be implemented in an idempotent way." As such, we may want to fix GDB, but we should definitely fix GDBserver. The fix is a prerequisite for target-side conditions on hardware breakpoints anyway (and while at it, on watchpoints too). GDBserver indeed already treats duplicate Z0 packets in an idempotent way. mem-break.c has the concept of high-level and low-level breakpoints, somewhat similar to GDB's split of breakpoints vs breakpoint locations, and keeps track of multiple breakpoints referencing the same address/location, for the case of an internal GDBserver breakpoint or a tracepoint being set at the same address as a GDB breakpoint. But, it only allows GDB to ever contribute one reference to a software breakpoint location. IOW, if gdbserver sees a Z0 packet for the same address where it already had a GDB breakpoint set, then GDBserver won't create another high-level GDB breakpoint. However, mem-break.c only tracks GDB Z0 breakpoints. The same logic should apply to all kinds of Zx packets. Currently, gdbserver passes down each duplicate Zx (other than Z0) request directly to the target->insert_point routine. The x86 watchpoint support itself refcounts watchpoint / hw breakpoint requests, to handle overlapping watchpoints, and save debug registers. But that code doesn't (and really shouldn't) handle the duplicate requests, assuming that for each insert there will be a corresponding remove. So the fix is to generalize mem-break.c to track all kinds of Zx breakpoints, and filter out duplicates. As mentioned, this ends up adding support for target-side conditions on hardware breakpoints and watchpoints too (though GDB itself doesn't support the latter yet). Probably the least obvious change in the patch is that it kind of turns the breakpoint insert/remove APIs inside out. Before, the target methods were only called for GDB breakpoints. The internal breakpoint set/delete methods inserted memory breakpoints directly bypassing the insert/remove target methods. That's not good when the target should use a debug API to set software breakpoints, instead of relying on GDBserver patching memory with breakpoint instructions, as is the case of NTO. Now removal/insertion of all kinds of breakpoints/watchpoints, either internal, or from GDB, always go through the target methods. The insert_point/remove_point methods no longer get passed a Z packet type, but an internal/raw breakpoint type. They're also passed a pointer to the raw breakpoint itself (note that's still opaque outside mem-break.c), so that insert_memory_breakpoint / remove_memory_breakpoint have access to the breakpoint's shadow buffer. I first tried passing down a new structure based on GDB's "struct bp_target_info" (actually with that name exactly), but then decided against it as unnecessary complication. As software/memory breakpoints work by poking at memory, when setting a GDB Z0 breakpoint (but not internal breakpoints, as those can assume the conditions are already right), we need to tell the target to prepare to access memory (which on Linux means stop threads). If that operation fails, we need to return error to GDB. Seeing an error, if this is the first breakpoint of that type that GDB tries to insert, GDB would then assume the breakpoint type is supported, but it may actually not be. So we need to check whether the type is supported at all before preparing to access memory. And to solve that, the patch adds a new target->supports_z_point_type method that is called before actually trying to insert the breakpoint. Other than that, hopefully the change is more or less obvious. New test added that exercises the hbreak2.exp regression in a more direct way, without relying on a breakpoint re-set happening before main is reached. Tested by building GDBserver for: aarch64-linux-gnu arm-linux-gnueabihf i686-pc-linux-gnu i686-w64-mingw32 m68k-linux-gnu mips-linux-gnu mips-uclinux nios2-linux-gnu powerpc-linux-gnu sh-linux-gnu tilegx-unknown-linux-gnu x86_64-redhat-linux x86_64-w64-mingw32 And also regression tested on x86_64 Fedora 20. gdb/gdbserver/ 2014-05-20 Pedro Alves <palves@redhat.com> * linux-aarch64-low.c (aarch64_insert_point) (aarch64_remove_point): No longer check whether the type is supported here. Adjust to new interface. (the_low_target): Install aarch64_supports_z_point_type as supports_z_point_type method. * linux-arm-low.c (raw_bkpt_type_to_arm_hwbp_type): New function. (arm_linux_hw_point_initialize): Take an enum raw_bkpt_type instead of a Z packet char. Adjust. (arm_supports_z_point_type): New function. (arm_insert_point, arm_remove_point): Adjust to new interface. (the_low_target): Install arm_supports_z_point_type. * linux-crisv32-low.c (cris_supports_z_point_type): New function. (cris_insert_point, cris_remove_point): Adjust to new interface. Don't check whether the type is supported here. (the_low_target): Install cris_supports_z_point_type. * linux-low.c (linux_supports_z_point_type): New function. (linux_insert_point, linux_remove_point): Adjust to new interface. * linux-low.h (struct linux_target_ops) <insert_point, remove_point>: Take an enum raw_bkpt_type instead of a char. Add raw_breakpoint pointer parameter. <supports_z_point_type>: New method. * linux-mips-low.c (mips_supports_z_point_type): New function. (mips_insert_point, mips_remove_point): Adjust to new interface. Use mips_supports_z_point_type. (the_low_target): Install mips_supports_z_point_type. * linux-ppc-low.c (the_low_target): Install NULL as supports_z_point_type method. * linux-s390-low.c (the_low_target): Install NULL as supports_z_point_type method. * linux-sparc-low.c (the_low_target): Install NULL as supports_z_point_type method. * linux-x86-low.c (x86_supports_z_point_type): New function. (x86_insert_point): Adjust to new insert_point interface. Use insert_memory_breakpoint. Adjust to new i386_low_insert_watchpoint interface. (x86_remove_point): Adjust to remove_point interface. Use remove_memory_breakpoint. Adjust to new i386_low_remove_watchpoint interface. (the_low_target): Install x86_supports_z_point_type. * lynx-low.c (lynx_target_ops): Install NULL as supports_z_point_type callback. * nto-low.c (nto_supports_z_point_type): New. (nto_insert_point, nto_remove_point): Adjust to new interface. (nto_target_ops): Install nto_supports_z_point_type. * mem-break.c: Adjust intro comment. (struct raw_breakpoint) <raw_type, size>: New fields. <inserted>: Update comment. <shlib_disabled>: Delete field. (enum bkpt_type) <gdb_breakpoint>: Delete value. <gdb_breakpoint_Z0, gdb_breakpoint_Z1, gdb_breakpoint_Z2, gdb_breakpoint_Z3, gdb_breakpoint_Z4>: New values. (raw_bkpt_type_to_target_hw_bp_type): New function. (find_enabled_raw_code_breakpoint_at): New function. (find_raw_breakpoint_at): New type and size parameters. Use them. (insert_memory_breakpoint): New function, based off set_raw_breakpoint_at. (remove_memory_breakpoint): New function. (set_raw_breakpoint_at): Reimplement. (set_breakpoint): New, based on set_breakpoint_at. (set_breakpoint_at): Reimplement. (delete_raw_breakpoint): Go through the_target->remove_point instead of assuming memory breakpoints. (find_gdb_breakpoint_at): Delete. (Z_packet_to_bkpt_type, Z_packet_to_raw_bkpt_type): New functions. (find_gdb_breakpoint): New function. (set_gdb_breakpoint_at): Delete. (z_type_supported): New function. (set_gdb_breakpoint_1): New function, loosely based off set_gdb_breakpoint_at. (check_gdb_bp_preconditions, set_gdb_breakpoint): New functions. (delete_gdb_breakpoint_at): Delete. (delete_gdb_breakpoint_1): New function, loosely based off delete_gdb_breakpoint_at. (delete_gdb_breakpoint): New function. (clear_gdb_breakpoint_conditions): Rename to ... (clear_breakpoint_conditions): ... this. Don't handle a NULL breakpoint. (add_condition_to_breakpoint): Make static. (add_breakpoint_condition): Take a struct breakpoint pointer instead of an address. Adjust. (gdb_condition_true_at_breakpoint): Rename to ... (gdb_condition_true_at_breakpoint_z_type): ... this, and add z_type parameter. (gdb_condition_true_at_breakpoint): Reimplement. (add_breakpoint_commands): Take a struct breakpoint pointer instead of an address. Adjust. (gdb_no_commands_at_breakpoint): Rename to ... (gdb_no_commands_at_breakpoint_z_type): ... this. Add z_type parameter. Return true if no breakpoint was found. Change debug output. (gdb_no_commands_at_breakpoint): Reimplement. (run_breakpoint_commands): Rename to ... (run_breakpoint_commands_z_type): ... this. Add z_type parameter, and change return type to boolean. (run_breakpoint_commands): New function. (gdb_breakpoint_here): Also check for Z1 breakpoints. (uninsert_raw_breakpoint): Don't try to reinsert a disabled breakpoint. Go through the_target->remove_point instead of assuming memory breakpoint. (uninsert_breakpoints_at, uninsert_all_breakpoints): Uninsert software and hardware breakpoints. (reinsert_raw_breakpoint): Go through the_target->insert_point instead of assuming memory breakpoint. (reinsert_breakpoints_at, reinsert_all_breakpoints): Reinsert software and hardware breakpoints. (check_breakpoints, breakpoint_here, breakpoint_inserted_here): Check both software and hardware breakpoints. (validate_inserted_breakpoint): Assert the breakpoint is a software breakpoint. Set the inserted flag to -1 instead of setting shlib_disabled. (delete_disabled_breakpoints): Adjust. (validate_breakpoints): Only validate software breakpoints. Adjust to inserted flag change. (check_mem_read, check_mem_write): Skip breakpoint types other than software breakpoints. Adjust to inserted flag change. * mem-break.h (enum raw_bkpt_type): New enum. (raw_breakpoint, struct process_info): Forward declare. (Z_packet_to_target_hw_bp_type): Delete declaration. (raw_bkpt_type_to_target_hw_bp_type, Z_packet_to_raw_bkpt_type) (set_gdb_breakpoint, delete_gdb_breakpoint) (clear_breakpoint_conditions): New declarations. (set_gdb_breakpoint_at, clear_gdb_breakpoint_conditions): Delete. (breakpoint_inserted_here): Update comment. (add_breakpoint_condition, add_breakpoint_commands): Replace address parameter with a breakpoint pointer parameter. (gdb_breakpoint_here): Update comment. (delete_gdb_breakpoint_at): Delete. (insert_memory_breakpoint, remove_memory_breakpoint): Declare. * server.c (process_point_options): Take a struct breakpoint pointer instead of an address. Adjust. (process_serial_event) <Z/z packets>: Use set_gdb_breakpoint and delete_gdb_breakpoint. * spu-low.c (spu_target_ops): Install NULL as supports_z_point_type method. * target.h: Include mem-break.h. (struct target_ops) <prepare_to_access_memory>: Update comment. <supports_z_point_type>: New field. <insert_point, remove_point>: Take an enum raw_bkpt_type argument instead of a char. Also take a raw breakpoint pointer. * win32-arm-low.c (the_low_target): Install NULL as supports_z_point_type. * win32-i386-low.c (i386_supports_z_point_type): New function. (i386_insert_point, i386_remove_point): Adjust to new interface. (the_low_target): Install i386_supports_z_point_type. * win32-low.c (win32_supports_z_point_type): New function. (win32_insert_point, win32_remove_point): Adjust to new interface. (win32_target_ops): Install win32_supports_z_point_type. * win32-low.h (struct win32_target_ops): <supports_z_point_type>: New method. <insert_point, remove_point>: Take an enum raw_bkpt_type argument instead of a char. Also take a raw breakpoint pointer. gdb/testsuite/ 2014-05-20 Pedro Alves <palves@redhat.com> * gdb.base/break-idempotent.c: New file. * gdb.base/break-idempotent.exp: New file.
2014-05-20 19:24:28 +02:00
int *err);
Support breakpoint kinds for software breakpoints in GDBServer. There's two ways to set breakpoints in GDBServer. - GDBServer setting its own breakpoints, through API set_breakpoint_at. - GDBServer setting breakpoints according to the information in Z packets, through API set_gdb_breakpoint. Before this patch the breakpoint kinds were a concept unique to GDB and Z packets, as GDBServer never had to set different kinds of breakpoint on its own. This patch teaches GDBServer to handle breakpoint kinds for its own breakpoints. It generalizes the breakpoint kind as per Z packets to represent different kinds of breakpoints directly set by GDBServer also. GDBServer now querys breakpoint_kind_from_pc to know what breakpoint kind to set on its own. As the kind is now a differentiating factor equivalent to size for the breakpoint struct and that it's size can be queried using sw_breakpoint_from_kind, the size field has been replaced with the kind field. All references to size are now replaced by kind or a call to bp_size that wraps sw_breakpoing_from_kind and returns the size of the breakpoint in memory. To fetch the software breakpoint data bp_opcode is called and wraps the sw_breakpoint_from_kind call. No regressions on Ubuntu 14.04 on ARMv7 and x86. With gdbserver-{native,extended} / { -marm -mthumb } gdb/gdbserver/ChangeLog: * linux-low.c (initialize_low): Ajdust for breakpoint global variables removal. * mem-break.c : Remove breakpoint_data/breakpoint_len global variables. (struct raw_breakpoint) <size>: Remove. (struct raw_breakpoint) <kind>: Add. (bp_size): New function. (bp_opcode): Likewise. (find_raw_breakpoint_at): Adjust for kind. (insert_memory_breakpoint): Adjust for kind call bp_size,bp_opcode. (remove_memory_breakpoint): Adjust for kind call bp_size. (set_raw_breakpoint_at): Adjust for kind. (set_breakpoint): Likewise. (set_breakpoint_at): Call breakpoint_kind_from_pc. (delete_raw_breakpoint): Adjust for kind. (delete_breakpoint): Likewise. (find_gdb_breakpoint): Likewise. (set_gdb_breakpoint_1): Likewise. (set_gdb_breakpoint): Likewise. (delete_gdb_breakpoint_1): Likewise. (delete_gdb_breakpoint): Likewise. (uninsert_raw_breakpoint): Likewise. (reinsert_raw_breakpoint): Likewise. (set_breakpoint_data): Remove. (validate_inserted_breakpoint): Adjust for kind call bp_size,bp_opcode. (check_mem_read): Adjust for kind call bp_size. (check_mem_write): Adjust for kind call bp_size,bp_opcode. (clone_one_breakpoint): Adjust for kind. * mem-break.h (set_gdb_breakpoint): Likewise. (delete_gdb_breakpoint): Likewise. * server.c (process_serial_event): Likewise.
2015-10-21 17:13:40 +02:00
/* Delete a GDB breakpoint of type Z_TYPE and kind KIND previously
[GDBserver] Make Zx/zx packet handling idempotent. This patch fixes hardware breakpoint regressions exposed by my fix for "PR breakpoints/7143 - Watchpoint does not trigger when first set", at https://sourceware.org/ml/gdb-patches/2014-03/msg00167.html The testsuite caught them on Linux/x86_64, at least. gdb.sum: gdb.sum: FAIL: gdb.base/hbreak2.exp: next over recursive call FAIL: gdb.base/hbreak2.exp: backtrace from factorial(5.1) FAIL: gdb.base/hbreak2.exp: continue until exit at recursive next test gdb.log: (gdb) next Program received signal SIGTRAP, Trace/breakpoint trap. factorial (value=4) at ../../../src/gdb/testsuite/gdb.base/break.c:113 113 if (value > 1) { /* set breakpoint 7 here */ (gdb) FAIL: gdb.base/hbreak2.exp: next over recursive call Actually, that patch just exposed a latent issue to "breakpoints always-inserted off" mode, not really caused it. After that patch, GDB no longer removes breakpoints at each internal event, thus making some scenarios behave like breakpoint always-inserted on. The bug is easy to trigger with always-inserted on. The issue is that since the target-side breakpoint conditions support, if the stub/server supports evaluating breakpoint conditions on the target side, then GDB is sending duplicate Zx packets to the target without removing them before, and GDBserver is not really expecting that for Z packets other than Z0/z0. E.g., with "set breakpoint always-inserted on" and "set debug remote 1": (gdb) b main Sending packet: $m410943,1#ff...Packet received: 48 Breakpoint 4 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z0,410943,1#48...Packet received: OK ^^^^^^^^^^^^ (gdb) b main Note: breakpoint 4 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Breakpoint 5 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z0,410943,1#48...Packet received: OK ^^^^^^^^^^^^ (gdb) b main Note: breakpoints 4 and 5 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Breakpoint 6 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z0,410943,1#48...Packet received: OK ^^^^^^^^^^^^ (gdb) del Delete all breakpoints? (y or n) y Sending packet: $Z0,410943,1#48...Packet received: OK Sending packet: $Z0,410943,1#48...Packet received: OK Sending packet: $z0,410943,1#68...Packet received: OK And for Z1, similarly: (gdb) hbreak main Sending packet: $m410943,1#ff...Packet received: 48 Hardware assisted breakpoint 4 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ Packet Z1 (hardware-breakpoint) is supported (gdb) hbreak main Note: breakpoint 4 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Hardware assisted breakpoint 5 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ (gdb) hbreak main Note: breakpoints 4 and 5 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Hardware assisted breakpoint 6 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ (gdb) del Delete all breakpoints? (y or n) y Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ Sending packet: $z1,410943,1#69...Packet received: OK ^^^^^^^^^^^^ So GDB sent a bunch of Z1 packets, and then when finally removing the breakpoint, only one z1 packet was sent. On the GDBserver side (with monitor set debug-hw-points 1), in the Z1 case, we see: $ ./gdbserver :9999 ./gdbserver Process ./gdbserver created; pid = 8629 Listening on port 9999 Remote debugging from host 127.0.0.1 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=1 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=2 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=3 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=4 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=5 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 remove_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=4 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 That's one insert_watchpoint call for each Z1 packet, and then one remove_watchpoint call for the z1 packet. Notice how ref.count increased for each insert_watchpoint call, and then in the end, after GDB told GDBserver to forget about the hardware breakpoint, GDBserver ends with the the first debug register still with ref.count=4! IOW, the hardware breakpoint is left armed on the target, while on the GDB end it's gone. If the program happens to execute 0x410943 afterwards, then the CPU traps, GDBserver reports the trap to GDB, and GDB not having a breakpoint set at that address anymore, reports to the user a spurious SIGTRAP. This is exactly what is happening in the hbreak2.exp test, though in that case, it's a shared library event that triggers a breakpoint_re_set, when breakpoints are still inserted (because nowadays GDB doesn't remove breakpoints while handling internal events), and that recreates breakpoint locations, which likewise forces breakpoint reinsertion and Zx packet resends... That is a lot of bogus Zx duplication that should possibly be addressed on the GDB side. GDB resends Zx packets because the way to change the target-side condition, is to resend the breakpoint to the server with the new condition. (That's an option in the packet: e.g., "Z1,410943,1;X3,220027" for "hbreak main if 0". The packets in the examples above are shorter because the breakpoints don't have conditions attached). GDB doesn't remove the breakpoint first before reinserting it because that'd be bad for non-stop, as it'd open a window where the inferior could miss the breakpoint. The conditions actually haven't changed between the resends, but GDB isn't smart enough to realize that. (TBC, if the target doesn't support target-side conditions, then GDB doesn't trigger these resends (init_bp_location calls mark_breakpoint_location_modified, and that does nothing if condition evaluation is on the host side. The resends are caused by the 'loc->condition_changed = condition_modified.' line.) But, even if GDB was made smarter, GDBserver should really still handle the resends anyway. So target-side conditions also aren't really to blame. The documentation of the Z/z packets says: "To avoid potential problems with duplicate packets, the operations should be implemented in an idempotent way." As such, we may want to fix GDB, but we should definitely fix GDBserver. The fix is a prerequisite for target-side conditions on hardware breakpoints anyway (and while at it, on watchpoints too). GDBserver indeed already treats duplicate Z0 packets in an idempotent way. mem-break.c has the concept of high-level and low-level breakpoints, somewhat similar to GDB's split of breakpoints vs breakpoint locations, and keeps track of multiple breakpoints referencing the same address/location, for the case of an internal GDBserver breakpoint or a tracepoint being set at the same address as a GDB breakpoint. But, it only allows GDB to ever contribute one reference to a software breakpoint location. IOW, if gdbserver sees a Z0 packet for the same address where it already had a GDB breakpoint set, then GDBserver won't create another high-level GDB breakpoint. However, mem-break.c only tracks GDB Z0 breakpoints. The same logic should apply to all kinds of Zx packets. Currently, gdbserver passes down each duplicate Zx (other than Z0) request directly to the target->insert_point routine. The x86 watchpoint support itself refcounts watchpoint / hw breakpoint requests, to handle overlapping watchpoints, and save debug registers. But that code doesn't (and really shouldn't) handle the duplicate requests, assuming that for each insert there will be a corresponding remove. So the fix is to generalize mem-break.c to track all kinds of Zx breakpoints, and filter out duplicates. As mentioned, this ends up adding support for target-side conditions on hardware breakpoints and watchpoints too (though GDB itself doesn't support the latter yet). Probably the least obvious change in the patch is that it kind of turns the breakpoint insert/remove APIs inside out. Before, the target methods were only called for GDB breakpoints. The internal breakpoint set/delete methods inserted memory breakpoints directly bypassing the insert/remove target methods. That's not good when the target should use a debug API to set software breakpoints, instead of relying on GDBserver patching memory with breakpoint instructions, as is the case of NTO. Now removal/insertion of all kinds of breakpoints/watchpoints, either internal, or from GDB, always go through the target methods. The insert_point/remove_point methods no longer get passed a Z packet type, but an internal/raw breakpoint type. They're also passed a pointer to the raw breakpoint itself (note that's still opaque outside mem-break.c), so that insert_memory_breakpoint / remove_memory_breakpoint have access to the breakpoint's shadow buffer. I first tried passing down a new structure based on GDB's "struct bp_target_info" (actually with that name exactly), but then decided against it as unnecessary complication. As software/memory breakpoints work by poking at memory, when setting a GDB Z0 breakpoint (but not internal breakpoints, as those can assume the conditions are already right), we need to tell the target to prepare to access memory (which on Linux means stop threads). If that operation fails, we need to return error to GDB. Seeing an error, if this is the first breakpoint of that type that GDB tries to insert, GDB would then assume the breakpoint type is supported, but it may actually not be. So we need to check whether the type is supported at all before preparing to access memory. And to solve that, the patch adds a new target->supports_z_point_type method that is called before actually trying to insert the breakpoint. Other than that, hopefully the change is more or less obvious. New test added that exercises the hbreak2.exp regression in a more direct way, without relying on a breakpoint re-set happening before main is reached. Tested by building GDBserver for: aarch64-linux-gnu arm-linux-gnueabihf i686-pc-linux-gnu i686-w64-mingw32 m68k-linux-gnu mips-linux-gnu mips-uclinux nios2-linux-gnu powerpc-linux-gnu sh-linux-gnu tilegx-unknown-linux-gnu x86_64-redhat-linux x86_64-w64-mingw32 And also regression tested on x86_64 Fedora 20. gdb/gdbserver/ 2014-05-20 Pedro Alves <palves@redhat.com> * linux-aarch64-low.c (aarch64_insert_point) (aarch64_remove_point): No longer check whether the type is supported here. Adjust to new interface. (the_low_target): Install aarch64_supports_z_point_type as supports_z_point_type method. * linux-arm-low.c (raw_bkpt_type_to_arm_hwbp_type): New function. (arm_linux_hw_point_initialize): Take an enum raw_bkpt_type instead of a Z packet char. Adjust. (arm_supports_z_point_type): New function. (arm_insert_point, arm_remove_point): Adjust to new interface. (the_low_target): Install arm_supports_z_point_type. * linux-crisv32-low.c (cris_supports_z_point_type): New function. (cris_insert_point, cris_remove_point): Adjust to new interface. Don't check whether the type is supported here. (the_low_target): Install cris_supports_z_point_type. * linux-low.c (linux_supports_z_point_type): New function. (linux_insert_point, linux_remove_point): Adjust to new interface. * linux-low.h (struct linux_target_ops) <insert_point, remove_point>: Take an enum raw_bkpt_type instead of a char. Add raw_breakpoint pointer parameter. <supports_z_point_type>: New method. * linux-mips-low.c (mips_supports_z_point_type): New function. (mips_insert_point, mips_remove_point): Adjust to new interface. Use mips_supports_z_point_type. (the_low_target): Install mips_supports_z_point_type. * linux-ppc-low.c (the_low_target): Install NULL as supports_z_point_type method. * linux-s390-low.c (the_low_target): Install NULL as supports_z_point_type method. * linux-sparc-low.c (the_low_target): Install NULL as supports_z_point_type method. * linux-x86-low.c (x86_supports_z_point_type): New function. (x86_insert_point): Adjust to new insert_point interface. Use insert_memory_breakpoint. Adjust to new i386_low_insert_watchpoint interface. (x86_remove_point): Adjust to remove_point interface. Use remove_memory_breakpoint. Adjust to new i386_low_remove_watchpoint interface. (the_low_target): Install x86_supports_z_point_type. * lynx-low.c (lynx_target_ops): Install NULL as supports_z_point_type callback. * nto-low.c (nto_supports_z_point_type): New. (nto_insert_point, nto_remove_point): Adjust to new interface. (nto_target_ops): Install nto_supports_z_point_type. * mem-break.c: Adjust intro comment. (struct raw_breakpoint) <raw_type, size>: New fields. <inserted>: Update comment. <shlib_disabled>: Delete field. (enum bkpt_type) <gdb_breakpoint>: Delete value. <gdb_breakpoint_Z0, gdb_breakpoint_Z1, gdb_breakpoint_Z2, gdb_breakpoint_Z3, gdb_breakpoint_Z4>: New values. (raw_bkpt_type_to_target_hw_bp_type): New function. (find_enabled_raw_code_breakpoint_at): New function. (find_raw_breakpoint_at): New type and size parameters. Use them. (insert_memory_breakpoint): New function, based off set_raw_breakpoint_at. (remove_memory_breakpoint): New function. (set_raw_breakpoint_at): Reimplement. (set_breakpoint): New, based on set_breakpoint_at. (set_breakpoint_at): Reimplement. (delete_raw_breakpoint): Go through the_target->remove_point instead of assuming memory breakpoints. (find_gdb_breakpoint_at): Delete. (Z_packet_to_bkpt_type, Z_packet_to_raw_bkpt_type): New functions. (find_gdb_breakpoint): New function. (set_gdb_breakpoint_at): Delete. (z_type_supported): New function. (set_gdb_breakpoint_1): New function, loosely based off set_gdb_breakpoint_at. (check_gdb_bp_preconditions, set_gdb_breakpoint): New functions. (delete_gdb_breakpoint_at): Delete. (delete_gdb_breakpoint_1): New function, loosely based off delete_gdb_breakpoint_at. (delete_gdb_breakpoint): New function. (clear_gdb_breakpoint_conditions): Rename to ... (clear_breakpoint_conditions): ... this. Don't handle a NULL breakpoint. (add_condition_to_breakpoint): Make static. (add_breakpoint_condition): Take a struct breakpoint pointer instead of an address. Adjust. (gdb_condition_true_at_breakpoint): Rename to ... (gdb_condition_true_at_breakpoint_z_type): ... this, and add z_type parameter. (gdb_condition_true_at_breakpoint): Reimplement. (add_breakpoint_commands): Take a struct breakpoint pointer instead of an address. Adjust. (gdb_no_commands_at_breakpoint): Rename to ... (gdb_no_commands_at_breakpoint_z_type): ... this. Add z_type parameter. Return true if no breakpoint was found. Change debug output. (gdb_no_commands_at_breakpoint): Reimplement. (run_breakpoint_commands): Rename to ... (run_breakpoint_commands_z_type): ... this. Add z_type parameter, and change return type to boolean. (run_breakpoint_commands): New function. (gdb_breakpoint_here): Also check for Z1 breakpoints. (uninsert_raw_breakpoint): Don't try to reinsert a disabled breakpoint. Go through the_target->remove_point instead of assuming memory breakpoint. (uninsert_breakpoints_at, uninsert_all_breakpoints): Uninsert software and hardware breakpoints. (reinsert_raw_breakpoint): Go through the_target->insert_point instead of assuming memory breakpoint. (reinsert_breakpoints_at, reinsert_all_breakpoints): Reinsert software and hardware breakpoints. (check_breakpoints, breakpoint_here, breakpoint_inserted_here): Check both software and hardware breakpoints. (validate_inserted_breakpoint): Assert the breakpoint is a software breakpoint. Set the inserted flag to -1 instead of setting shlib_disabled. (delete_disabled_breakpoints): Adjust. (validate_breakpoints): Only validate software breakpoints. Adjust to inserted flag change. (check_mem_read, check_mem_write): Skip breakpoint types other than software breakpoints. Adjust to inserted flag change. * mem-break.h (enum raw_bkpt_type): New enum. (raw_breakpoint, struct process_info): Forward declare. (Z_packet_to_target_hw_bp_type): Delete declaration. (raw_bkpt_type_to_target_hw_bp_type, Z_packet_to_raw_bkpt_type) (set_gdb_breakpoint, delete_gdb_breakpoint) (clear_breakpoint_conditions): New declarations. (set_gdb_breakpoint_at, clear_gdb_breakpoint_conditions): Delete. (breakpoint_inserted_here): Update comment. (add_breakpoint_condition, add_breakpoint_commands): Replace address parameter with a breakpoint pointer parameter. (gdb_breakpoint_here): Update comment. (delete_gdb_breakpoint_at): Delete. (insert_memory_breakpoint, remove_memory_breakpoint): Declare. * server.c (process_point_options): Take a struct breakpoint pointer instead of an address. Adjust. (process_serial_event) <Z/z packets>: Use set_gdb_breakpoint and delete_gdb_breakpoint. * spu-low.c (spu_target_ops): Install NULL as supports_z_point_type method. * target.h: Include mem-break.h. (struct target_ops) <prepare_to_access_memory>: Update comment. <supports_z_point_type>: New field. <insert_point, remove_point>: Take an enum raw_bkpt_type argument instead of a char. Also take a raw breakpoint pointer. * win32-arm-low.c (the_low_target): Install NULL as supports_z_point_type. * win32-i386-low.c (i386_supports_z_point_type): New function. (i386_insert_point, i386_remove_point): Adjust to new interface. (the_low_target): Install i386_supports_z_point_type. * win32-low.c (win32_supports_z_point_type): New function. (win32_insert_point, win32_remove_point): Adjust to new interface. (win32_target_ops): Install win32_supports_z_point_type. * win32-low.h (struct win32_target_ops): <supports_z_point_type>: New method. <insert_point, remove_point>: Take an enum raw_bkpt_type argument instead of a char. Also take a raw breakpoint pointer. gdb/testsuite/ 2014-05-20 Pedro Alves <palves@redhat.com> * gdb.base/break-idempotent.c: New file. * gdb.base/break-idempotent.exp: New file.
2014-05-20 19:24:28 +02:00
inserted at ADDR with set_gdb_breakpoint_at. Returns 0 on success,
-1 on error, and 1 if Z_TYPE breakpoints are not supported on this
target. */
Support breakpoint kinds for software breakpoints in GDBServer. There's two ways to set breakpoints in GDBServer. - GDBServer setting its own breakpoints, through API set_breakpoint_at. - GDBServer setting breakpoints according to the information in Z packets, through API set_gdb_breakpoint. Before this patch the breakpoint kinds were a concept unique to GDB and Z packets, as GDBServer never had to set different kinds of breakpoint on its own. This patch teaches GDBServer to handle breakpoint kinds for its own breakpoints. It generalizes the breakpoint kind as per Z packets to represent different kinds of breakpoints directly set by GDBServer also. GDBServer now querys breakpoint_kind_from_pc to know what breakpoint kind to set on its own. As the kind is now a differentiating factor equivalent to size for the breakpoint struct and that it's size can be queried using sw_breakpoint_from_kind, the size field has been replaced with the kind field. All references to size are now replaced by kind or a call to bp_size that wraps sw_breakpoing_from_kind and returns the size of the breakpoint in memory. To fetch the software breakpoint data bp_opcode is called and wraps the sw_breakpoint_from_kind call. No regressions on Ubuntu 14.04 on ARMv7 and x86. With gdbserver-{native,extended} / { -marm -mthumb } gdb/gdbserver/ChangeLog: * linux-low.c (initialize_low): Ajdust for breakpoint global variables removal. * mem-break.c : Remove breakpoint_data/breakpoint_len global variables. (struct raw_breakpoint) <size>: Remove. (struct raw_breakpoint) <kind>: Add. (bp_size): New function. (bp_opcode): Likewise. (find_raw_breakpoint_at): Adjust for kind. (insert_memory_breakpoint): Adjust for kind call bp_size,bp_opcode. (remove_memory_breakpoint): Adjust for kind call bp_size. (set_raw_breakpoint_at): Adjust for kind. (set_breakpoint): Likewise. (set_breakpoint_at): Call breakpoint_kind_from_pc. (delete_raw_breakpoint): Adjust for kind. (delete_breakpoint): Likewise. (find_gdb_breakpoint): Likewise. (set_gdb_breakpoint_1): Likewise. (set_gdb_breakpoint): Likewise. (delete_gdb_breakpoint_1): Likewise. (delete_gdb_breakpoint): Likewise. (uninsert_raw_breakpoint): Likewise. (reinsert_raw_breakpoint): Likewise. (set_breakpoint_data): Remove. (validate_inserted_breakpoint): Adjust for kind call bp_size,bp_opcode. (check_mem_read): Adjust for kind call bp_size. (check_mem_write): Adjust for kind call bp_size,bp_opcode. (clone_one_breakpoint): Adjust for kind. * mem-break.h (set_gdb_breakpoint): Likewise. (delete_gdb_breakpoint): Likewise. * server.c (process_serial_event): Likewise.
2015-10-21 17:13:40 +02:00
int delete_gdb_breakpoint (char z_type, CORE_ADDR addr, int kind);
[GDBserver] Make Zx/zx packet handling idempotent. This patch fixes hardware breakpoint regressions exposed by my fix for "PR breakpoints/7143 - Watchpoint does not trigger when first set", at https://sourceware.org/ml/gdb-patches/2014-03/msg00167.html The testsuite caught them on Linux/x86_64, at least. gdb.sum: gdb.sum: FAIL: gdb.base/hbreak2.exp: next over recursive call FAIL: gdb.base/hbreak2.exp: backtrace from factorial(5.1) FAIL: gdb.base/hbreak2.exp: continue until exit at recursive next test gdb.log: (gdb) next Program received signal SIGTRAP, Trace/breakpoint trap. factorial (value=4) at ../../../src/gdb/testsuite/gdb.base/break.c:113 113 if (value > 1) { /* set breakpoint 7 here */ (gdb) FAIL: gdb.base/hbreak2.exp: next over recursive call Actually, that patch just exposed a latent issue to "breakpoints always-inserted off" mode, not really caused it. After that patch, GDB no longer removes breakpoints at each internal event, thus making some scenarios behave like breakpoint always-inserted on. The bug is easy to trigger with always-inserted on. The issue is that since the target-side breakpoint conditions support, if the stub/server supports evaluating breakpoint conditions on the target side, then GDB is sending duplicate Zx packets to the target without removing them before, and GDBserver is not really expecting that for Z packets other than Z0/z0. E.g., with "set breakpoint always-inserted on" and "set debug remote 1": (gdb) b main Sending packet: $m410943,1#ff...Packet received: 48 Breakpoint 4 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z0,410943,1#48...Packet received: OK ^^^^^^^^^^^^ (gdb) b main Note: breakpoint 4 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Breakpoint 5 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z0,410943,1#48...Packet received: OK ^^^^^^^^^^^^ (gdb) b main Note: breakpoints 4 and 5 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Breakpoint 6 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z0,410943,1#48...Packet received: OK ^^^^^^^^^^^^ (gdb) del Delete all breakpoints? (y or n) y Sending packet: $Z0,410943,1#48...Packet received: OK Sending packet: $Z0,410943,1#48...Packet received: OK Sending packet: $z0,410943,1#68...Packet received: OK And for Z1, similarly: (gdb) hbreak main Sending packet: $m410943,1#ff...Packet received: 48 Hardware assisted breakpoint 4 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ Packet Z1 (hardware-breakpoint) is supported (gdb) hbreak main Note: breakpoint 4 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Hardware assisted breakpoint 5 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ (gdb) hbreak main Note: breakpoints 4 and 5 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Hardware assisted breakpoint 6 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ (gdb) del Delete all breakpoints? (y or n) y Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ Sending packet: $z1,410943,1#69...Packet received: OK ^^^^^^^^^^^^ So GDB sent a bunch of Z1 packets, and then when finally removing the breakpoint, only one z1 packet was sent. On the GDBserver side (with monitor set debug-hw-points 1), in the Z1 case, we see: $ ./gdbserver :9999 ./gdbserver Process ./gdbserver created; pid = 8629 Listening on port 9999 Remote debugging from host 127.0.0.1 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=1 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=2 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=3 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=4 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=5 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 remove_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=4 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 That's one insert_watchpoint call for each Z1 packet, and then one remove_watchpoint call for the z1 packet. Notice how ref.count increased for each insert_watchpoint call, and then in the end, after GDB told GDBserver to forget about the hardware breakpoint, GDBserver ends with the the first debug register still with ref.count=4! IOW, the hardware breakpoint is left armed on the target, while on the GDB end it's gone. If the program happens to execute 0x410943 afterwards, then the CPU traps, GDBserver reports the trap to GDB, and GDB not having a breakpoint set at that address anymore, reports to the user a spurious SIGTRAP. This is exactly what is happening in the hbreak2.exp test, though in that case, it's a shared library event that triggers a breakpoint_re_set, when breakpoints are still inserted (because nowadays GDB doesn't remove breakpoints while handling internal events), and that recreates breakpoint locations, which likewise forces breakpoint reinsertion and Zx packet resends... That is a lot of bogus Zx duplication that should possibly be addressed on the GDB side. GDB resends Zx packets because the way to change the target-side condition, is to resend the breakpoint to the server with the new condition. (That's an option in the packet: e.g., "Z1,410943,1;X3,220027" for "hbreak main if 0". The packets in the examples above are shorter because the breakpoints don't have conditions attached). GDB doesn't remove the breakpoint first before reinserting it because that'd be bad for non-stop, as it'd open a window where the inferior could miss the breakpoint. The conditions actually haven't changed between the resends, but GDB isn't smart enough to realize that. (TBC, if the target doesn't support target-side conditions, then GDB doesn't trigger these resends (init_bp_location calls mark_breakpoint_location_modified, and that does nothing if condition evaluation is on the host side. The resends are caused by the 'loc->condition_changed = condition_modified.' line.) But, even if GDB was made smarter, GDBserver should really still handle the resends anyway. So target-side conditions also aren't really to blame. The documentation of the Z/z packets says: "To avoid potential problems with duplicate packets, the operations should be implemented in an idempotent way." As such, we may want to fix GDB, but we should definitely fix GDBserver. The fix is a prerequisite for target-side conditions on hardware breakpoints anyway (and while at it, on watchpoints too). GDBserver indeed already treats duplicate Z0 packets in an idempotent way. mem-break.c has the concept of high-level and low-level breakpoints, somewhat similar to GDB's split of breakpoints vs breakpoint locations, and keeps track of multiple breakpoints referencing the same address/location, for the case of an internal GDBserver breakpoint or a tracepoint being set at the same address as a GDB breakpoint. But, it only allows GDB to ever contribute one reference to a software breakpoint location. IOW, if gdbserver sees a Z0 packet for the same address where it already had a GDB breakpoint set, then GDBserver won't create another high-level GDB breakpoint. However, mem-break.c only tracks GDB Z0 breakpoints. The same logic should apply to all kinds of Zx packets. Currently, gdbserver passes down each duplicate Zx (other than Z0) request directly to the target->insert_point routine. The x86 watchpoint support itself refcounts watchpoint / hw breakpoint requests, to handle overlapping watchpoints, and save debug registers. But that code doesn't (and really shouldn't) handle the duplicate requests, assuming that for each insert there will be a corresponding remove. So the fix is to generalize mem-break.c to track all kinds of Zx breakpoints, and filter out duplicates. As mentioned, this ends up adding support for target-side conditions on hardware breakpoints and watchpoints too (though GDB itself doesn't support the latter yet). Probably the least obvious change in the patch is that it kind of turns the breakpoint insert/remove APIs inside out. Before, the target methods were only called for GDB breakpoints. The internal breakpoint set/delete methods inserted memory breakpoints directly bypassing the insert/remove target methods. That's not good when the target should use a debug API to set software breakpoints, instead of relying on GDBserver patching memory with breakpoint instructions, as is the case of NTO. Now removal/insertion of all kinds of breakpoints/watchpoints, either internal, or from GDB, always go through the target methods. The insert_point/remove_point methods no longer get passed a Z packet type, but an internal/raw breakpoint type. They're also passed a pointer to the raw breakpoint itself (note that's still opaque outside mem-break.c), so that insert_memory_breakpoint / remove_memory_breakpoint have access to the breakpoint's shadow buffer. I first tried passing down a new structure based on GDB's "struct bp_target_info" (actually with that name exactly), but then decided against it as unnecessary complication. As software/memory breakpoints work by poking at memory, when setting a GDB Z0 breakpoint (but not internal breakpoints, as those can assume the conditions are already right), we need to tell the target to prepare to access memory (which on Linux means stop threads). If that operation fails, we need to return error to GDB. Seeing an error, if this is the first breakpoint of that type that GDB tries to insert, GDB would then assume the breakpoint type is supported, but it may actually not be. So we need to check whether the type is supported at all before preparing to access memory. And to solve that, the patch adds a new target->supports_z_point_type method that is called before actually trying to insert the breakpoint. Other than that, hopefully the change is more or less obvious. New test added that exercises the hbreak2.exp regression in a more direct way, without relying on a breakpoint re-set happening before main is reached. Tested by building GDBserver for: aarch64-linux-gnu arm-linux-gnueabihf i686-pc-linux-gnu i686-w64-mingw32 m68k-linux-gnu mips-linux-gnu mips-uclinux nios2-linux-gnu powerpc-linux-gnu sh-linux-gnu tilegx-unknown-linux-gnu x86_64-redhat-linux x86_64-w64-mingw32 And also regression tested on x86_64 Fedora 20. gdb/gdbserver/ 2014-05-20 Pedro Alves <palves@redhat.com> * linux-aarch64-low.c (aarch64_insert_point) (aarch64_remove_point): No longer check whether the type is supported here. Adjust to new interface. (the_low_target): Install aarch64_supports_z_point_type as supports_z_point_type method. * linux-arm-low.c (raw_bkpt_type_to_arm_hwbp_type): New function. (arm_linux_hw_point_initialize): Take an enum raw_bkpt_type instead of a Z packet char. Adjust. (arm_supports_z_point_type): New function. (arm_insert_point, arm_remove_point): Adjust to new interface. (the_low_target): Install arm_supports_z_point_type. * linux-crisv32-low.c (cris_supports_z_point_type): New function. (cris_insert_point, cris_remove_point): Adjust to new interface. Don't check whether the type is supported here. (the_low_target): Install cris_supports_z_point_type. * linux-low.c (linux_supports_z_point_type): New function. (linux_insert_point, linux_remove_point): Adjust to new interface. * linux-low.h (struct linux_target_ops) <insert_point, remove_point>: Take an enum raw_bkpt_type instead of a char. Add raw_breakpoint pointer parameter. <supports_z_point_type>: New method. * linux-mips-low.c (mips_supports_z_point_type): New function. (mips_insert_point, mips_remove_point): Adjust to new interface. Use mips_supports_z_point_type. (the_low_target): Install mips_supports_z_point_type. * linux-ppc-low.c (the_low_target): Install NULL as supports_z_point_type method. * linux-s390-low.c (the_low_target): Install NULL as supports_z_point_type method. * linux-sparc-low.c (the_low_target): Install NULL as supports_z_point_type method. * linux-x86-low.c (x86_supports_z_point_type): New function. (x86_insert_point): Adjust to new insert_point interface. Use insert_memory_breakpoint. Adjust to new i386_low_insert_watchpoint interface. (x86_remove_point): Adjust to remove_point interface. Use remove_memory_breakpoint. Adjust to new i386_low_remove_watchpoint interface. (the_low_target): Install x86_supports_z_point_type. * lynx-low.c (lynx_target_ops): Install NULL as supports_z_point_type callback. * nto-low.c (nto_supports_z_point_type): New. (nto_insert_point, nto_remove_point): Adjust to new interface. (nto_target_ops): Install nto_supports_z_point_type. * mem-break.c: Adjust intro comment. (struct raw_breakpoint) <raw_type, size>: New fields. <inserted>: Update comment. <shlib_disabled>: Delete field. (enum bkpt_type) <gdb_breakpoint>: Delete value. <gdb_breakpoint_Z0, gdb_breakpoint_Z1, gdb_breakpoint_Z2, gdb_breakpoint_Z3, gdb_breakpoint_Z4>: New values. (raw_bkpt_type_to_target_hw_bp_type): New function. (find_enabled_raw_code_breakpoint_at): New function. (find_raw_breakpoint_at): New type and size parameters. Use them. (insert_memory_breakpoint): New function, based off set_raw_breakpoint_at. (remove_memory_breakpoint): New function. (set_raw_breakpoint_at): Reimplement. (set_breakpoint): New, based on set_breakpoint_at. (set_breakpoint_at): Reimplement. (delete_raw_breakpoint): Go through the_target->remove_point instead of assuming memory breakpoints. (find_gdb_breakpoint_at): Delete. (Z_packet_to_bkpt_type, Z_packet_to_raw_bkpt_type): New functions. (find_gdb_breakpoint): New function. (set_gdb_breakpoint_at): Delete. (z_type_supported): New function. (set_gdb_breakpoint_1): New function, loosely based off set_gdb_breakpoint_at. (check_gdb_bp_preconditions, set_gdb_breakpoint): New functions. (delete_gdb_breakpoint_at): Delete. (delete_gdb_breakpoint_1): New function, loosely based off delete_gdb_breakpoint_at. (delete_gdb_breakpoint): New function. (clear_gdb_breakpoint_conditions): Rename to ... (clear_breakpoint_conditions): ... this. Don't handle a NULL breakpoint. (add_condition_to_breakpoint): Make static. (add_breakpoint_condition): Take a struct breakpoint pointer instead of an address. Adjust. (gdb_condition_true_at_breakpoint): Rename to ... (gdb_condition_true_at_breakpoint_z_type): ... this, and add z_type parameter. (gdb_condition_true_at_breakpoint): Reimplement. (add_breakpoint_commands): Take a struct breakpoint pointer instead of an address. Adjust. (gdb_no_commands_at_breakpoint): Rename to ... (gdb_no_commands_at_breakpoint_z_type): ... this. Add z_type parameter. Return true if no breakpoint was found. Change debug output. (gdb_no_commands_at_breakpoint): Reimplement. (run_breakpoint_commands): Rename to ... (run_breakpoint_commands_z_type): ... this. Add z_type parameter, and change return type to boolean. (run_breakpoint_commands): New function. (gdb_breakpoint_here): Also check for Z1 breakpoints. (uninsert_raw_breakpoint): Don't try to reinsert a disabled breakpoint. Go through the_target->remove_point instead of assuming memory breakpoint. (uninsert_breakpoints_at, uninsert_all_breakpoints): Uninsert software and hardware breakpoints. (reinsert_raw_breakpoint): Go through the_target->insert_point instead of assuming memory breakpoint. (reinsert_breakpoints_at, reinsert_all_breakpoints): Reinsert software and hardware breakpoints. (check_breakpoints, breakpoint_here, breakpoint_inserted_here): Check both software and hardware breakpoints. (validate_inserted_breakpoint): Assert the breakpoint is a software breakpoint. Set the inserted flag to -1 instead of setting shlib_disabled. (delete_disabled_breakpoints): Adjust. (validate_breakpoints): Only validate software breakpoints. Adjust to inserted flag change. (check_mem_read, check_mem_write): Skip breakpoint types other than software breakpoints. Adjust to inserted flag change. * mem-break.h (enum raw_bkpt_type): New enum. (raw_breakpoint, struct process_info): Forward declare. (Z_packet_to_target_hw_bp_type): Delete declaration. (raw_bkpt_type_to_target_hw_bp_type, Z_packet_to_raw_bkpt_type) (set_gdb_breakpoint, delete_gdb_breakpoint) (clear_breakpoint_conditions): New declarations. (set_gdb_breakpoint_at, clear_gdb_breakpoint_conditions): Delete. (breakpoint_inserted_here): Update comment. (add_breakpoint_condition, add_breakpoint_commands): Replace address parameter with a breakpoint pointer parameter. (gdb_breakpoint_here): Update comment. (delete_gdb_breakpoint_at): Delete. (insert_memory_breakpoint, remove_memory_breakpoint): Declare. * server.c (process_point_options): Take a struct breakpoint pointer instead of an address. Adjust. (process_serial_event) <Z/z packets>: Use set_gdb_breakpoint and delete_gdb_breakpoint. * spu-low.c (spu_target_ops): Install NULL as supports_z_point_type method. * target.h: Include mem-break.h. (struct target_ops) <prepare_to_access_memory>: Update comment. <supports_z_point_type>: New field. <insert_point, remove_point>: Take an enum raw_bkpt_type argument instead of a char. Also take a raw breakpoint pointer. * win32-arm-low.c (the_low_target): Install NULL as supports_z_point_type. * win32-i386-low.c (i386_supports_z_point_type): New function. (i386_insert_point, i386_remove_point): Adjust to new interface. (the_low_target): Install i386_supports_z_point_type. * win32-low.c (win32_supports_z_point_type): New function. (win32_insert_point, win32_remove_point): Adjust to new interface. (win32_target_ops): Install win32_supports_z_point_type. * win32-low.h (struct win32_target_ops): <supports_z_point_type>: New method. <insert_point, remove_point>: Take an enum raw_bkpt_type argument instead of a char. Also take a raw breakpoint pointer. gdb/testsuite/ 2014-05-20 Pedro Alves <palves@redhat.com> * gdb.base/break-idempotent.c: New file. * gdb.base/break-idempotent.exp: New file.
2014-05-20 19:24:28 +02:00
/* Returns TRUE if there's a software or hardware (code) breakpoint at
ADDR in our tables, inserted, or not. */
Teach linux gdbserver to step-over-breakpoints. * linux-low.c (can_hardware_single_step): New. (supports_breakpoints): New. (handle_extended_wait): If stopping threads, read the stop pc of the new cloned LWP. (get_pc): New. (get_stop_pc): Add `lwp' parameter. Handle it. Bail out if the low target doesn't support retrieving the PC. (add_lwp): Set last_resume_kind to resume_continue. (linux_attach_lwp_1): Adjust comments. Always set stop_expected. (linux_attach): Don't clear stop_expected. Set the lwp's last_resume_kind to resume_stop. (linux_detach_one_lwp): Don't check for removed breakpoints. (check_removed_breakpoint): Delete. (status_pending_p): Rename to ... (status_pending_p_callback): ... this. Don't check for removed breakpoints. Don't consider threads that are stopped from GDB's perspective. (linux_wait_for_lwp): Always read the stop_pc here. (cancel_breakpoint): New. (step_over_bkpt): New global. (linux_wait_for_event_1): Implement stepping over breakpoints. (gdb_wants_lwp_stopped): New. (gdb_wants_all_stopped): New. (linux_wait_1): Tag threads as gdb-wants-stopped. Cancel finished single-step traps here. Store the thread's last reported target wait status. (send_sigstop): Don't clear stop_expected. Always set it, instead. (mark_lwp_dead): Remove reference to pending_is_breakpoint. (cancel_finished_single_step): New. (cancel_finished_single_steps): New. (wait_for_sigstop): Don't cancel finished single-step traps here. (linux_resume_one_lwp): Don't check for removed breakpoints. Don't set `step' on non-hardware step archs. (linux_set_resume_request): Ignore resume_stop requests if already stopping or stopped. Set the lwp's last_resume_kind. (resume_status_pending_p): Don't check for removed breakpoints. (need_step_over_p): New. (start_step_over): New. (finish_step_over): New. (linux_resume_one_thread): Always queue a sigstop for resume_stop requests. Clear the thread's last reported target waitstatus. Don't use the `suspended' flag. Don't consider pending breakpoints. (linux_resume): Start a step-over if necessary. (proceed_one_lwp): New. (proceed_all_lwps): New. (unstop_all_lwps): New. * linux-low.h (struct lwp_info): Rewrite comment for the `suspended' flag. Add the `stop_pc' field. Delete the `pending_stop_pc' field. Tweak the `stepping' flag's comment. Add `'last_resume_kind' and `need_step_over' fields. * inferiors.c (struct thread_info): Delete, moved elsewhere. * mem-break.c (struct breakpoint): Delete `reinserting' flag. Delete `breakpoint_to_reinsert' field. New flag `inserted'. (set_raw_breakpoint_at): New. (set_breakpoint_at): Rewrite to use it. (reinsert_breakpoint_handler): Delete. (set_reinsert_breakpoint): New. (reinsert_breakpoint_by_bp): Delete. (delete_reinsert_breakpoints): New. (uninsert_breakpoint): Rewrite. (uninsert_breakpoints_at): New. (reinsert_breakpoint): Rewrite. (reinsert_breakpoints_at): New. (check_breakpoints): Rewrite. (breakpoint_here): New. (breakpoint_inserted_here): New. (check_mem_read): Adjust. * mem-break.h (breakpoints_supported, breakpoint_here) (breakpoint_inserted_here, set_reinsert_breakpoint): Declare. (reinsert_breakpoint_by_bp): Delete declaration. (delete_reinsert_breakpoints): Declare. (reinsert_breakpoint): Delete declaration. (reinsert_breakpoints_at): Declare. (uninsert_breakpoint): Delete declaration. (uninsert_breakpoints_at): Declare. (check_breakpoints): Adjust prototype. * server.h: Adjust include order. (struct thread_info): Declare here. Add a `last_status' field.
2010-03-24 01:05:03 +01:00
int breakpoint_here (CORE_ADDR addr);
[GDBserver] Make Zx/zx packet handling idempotent. This patch fixes hardware breakpoint regressions exposed by my fix for "PR breakpoints/7143 - Watchpoint does not trigger when first set", at https://sourceware.org/ml/gdb-patches/2014-03/msg00167.html The testsuite caught them on Linux/x86_64, at least. gdb.sum: gdb.sum: FAIL: gdb.base/hbreak2.exp: next over recursive call FAIL: gdb.base/hbreak2.exp: backtrace from factorial(5.1) FAIL: gdb.base/hbreak2.exp: continue until exit at recursive next test gdb.log: (gdb) next Program received signal SIGTRAP, Trace/breakpoint trap. factorial (value=4) at ../../../src/gdb/testsuite/gdb.base/break.c:113 113 if (value > 1) { /* set breakpoint 7 here */ (gdb) FAIL: gdb.base/hbreak2.exp: next over recursive call Actually, that patch just exposed a latent issue to "breakpoints always-inserted off" mode, not really caused it. After that patch, GDB no longer removes breakpoints at each internal event, thus making some scenarios behave like breakpoint always-inserted on. The bug is easy to trigger with always-inserted on. The issue is that since the target-side breakpoint conditions support, if the stub/server supports evaluating breakpoint conditions on the target side, then GDB is sending duplicate Zx packets to the target without removing them before, and GDBserver is not really expecting that for Z packets other than Z0/z0. E.g., with "set breakpoint always-inserted on" and "set debug remote 1": (gdb) b main Sending packet: $m410943,1#ff...Packet received: 48 Breakpoint 4 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z0,410943,1#48...Packet received: OK ^^^^^^^^^^^^ (gdb) b main Note: breakpoint 4 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Breakpoint 5 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z0,410943,1#48...Packet received: OK ^^^^^^^^^^^^ (gdb) b main Note: breakpoints 4 and 5 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Breakpoint 6 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z0,410943,1#48...Packet received: OK ^^^^^^^^^^^^ (gdb) del Delete all breakpoints? (y or n) y Sending packet: $Z0,410943,1#48...Packet received: OK Sending packet: $Z0,410943,1#48...Packet received: OK Sending packet: $z0,410943,1#68...Packet received: OK And for Z1, similarly: (gdb) hbreak main Sending packet: $m410943,1#ff...Packet received: 48 Hardware assisted breakpoint 4 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ Packet Z1 (hardware-breakpoint) is supported (gdb) hbreak main Note: breakpoint 4 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Hardware assisted breakpoint 5 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ (gdb) hbreak main Note: breakpoints 4 and 5 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Hardware assisted breakpoint 6 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ (gdb) del Delete all breakpoints? (y or n) y Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ Sending packet: $z1,410943,1#69...Packet received: OK ^^^^^^^^^^^^ So GDB sent a bunch of Z1 packets, and then when finally removing the breakpoint, only one z1 packet was sent. On the GDBserver side (with monitor set debug-hw-points 1), in the Z1 case, we see: $ ./gdbserver :9999 ./gdbserver Process ./gdbserver created; pid = 8629 Listening on port 9999 Remote debugging from host 127.0.0.1 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=1 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=2 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=3 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=4 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=5 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 remove_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=4 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 That's one insert_watchpoint call for each Z1 packet, and then one remove_watchpoint call for the z1 packet. Notice how ref.count increased for each insert_watchpoint call, and then in the end, after GDB told GDBserver to forget about the hardware breakpoint, GDBserver ends with the the first debug register still with ref.count=4! IOW, the hardware breakpoint is left armed on the target, while on the GDB end it's gone. If the program happens to execute 0x410943 afterwards, then the CPU traps, GDBserver reports the trap to GDB, and GDB not having a breakpoint set at that address anymore, reports to the user a spurious SIGTRAP. This is exactly what is happening in the hbreak2.exp test, though in that case, it's a shared library event that triggers a breakpoint_re_set, when breakpoints are still inserted (because nowadays GDB doesn't remove breakpoints while handling internal events), and that recreates breakpoint locations, which likewise forces breakpoint reinsertion and Zx packet resends... That is a lot of bogus Zx duplication that should possibly be addressed on the GDB side. GDB resends Zx packets because the way to change the target-side condition, is to resend the breakpoint to the server with the new condition. (That's an option in the packet: e.g., "Z1,410943,1;X3,220027" for "hbreak main if 0". The packets in the examples above are shorter because the breakpoints don't have conditions attached). GDB doesn't remove the breakpoint first before reinserting it because that'd be bad for non-stop, as it'd open a window where the inferior could miss the breakpoint. The conditions actually haven't changed between the resends, but GDB isn't smart enough to realize that. (TBC, if the target doesn't support target-side conditions, then GDB doesn't trigger these resends (init_bp_location calls mark_breakpoint_location_modified, and that does nothing if condition evaluation is on the host side. The resends are caused by the 'loc->condition_changed = condition_modified.' line.) But, even if GDB was made smarter, GDBserver should really still handle the resends anyway. So target-side conditions also aren't really to blame. The documentation of the Z/z packets says: "To avoid potential problems with duplicate packets, the operations should be implemented in an idempotent way." As such, we may want to fix GDB, but we should definitely fix GDBserver. The fix is a prerequisite for target-side conditions on hardware breakpoints anyway (and while at it, on watchpoints too). GDBserver indeed already treats duplicate Z0 packets in an idempotent way. mem-break.c has the concept of high-level and low-level breakpoints, somewhat similar to GDB's split of breakpoints vs breakpoint locations, and keeps track of multiple breakpoints referencing the same address/location, for the case of an internal GDBserver breakpoint or a tracepoint being set at the same address as a GDB breakpoint. But, it only allows GDB to ever contribute one reference to a software breakpoint location. IOW, if gdbserver sees a Z0 packet for the same address where it already had a GDB breakpoint set, then GDBserver won't create another high-level GDB breakpoint. However, mem-break.c only tracks GDB Z0 breakpoints. The same logic should apply to all kinds of Zx packets. Currently, gdbserver passes down each duplicate Zx (other than Z0) request directly to the target->insert_point routine. The x86 watchpoint support itself refcounts watchpoint / hw breakpoint requests, to handle overlapping watchpoints, and save debug registers. But that code doesn't (and really shouldn't) handle the duplicate requests, assuming that for each insert there will be a corresponding remove. So the fix is to generalize mem-break.c to track all kinds of Zx breakpoints, and filter out duplicates. As mentioned, this ends up adding support for target-side conditions on hardware breakpoints and watchpoints too (though GDB itself doesn't support the latter yet). Probably the least obvious change in the patch is that it kind of turns the breakpoint insert/remove APIs inside out. Before, the target methods were only called for GDB breakpoints. The internal breakpoint set/delete methods inserted memory breakpoints directly bypassing the insert/remove target methods. That's not good when the target should use a debug API to set software breakpoints, instead of relying on GDBserver patching memory with breakpoint instructions, as is the case of NTO. Now removal/insertion of all kinds of breakpoints/watchpoints, either internal, or from GDB, always go through the target methods. The insert_point/remove_point methods no longer get passed a Z packet type, but an internal/raw breakpoint type. They're also passed a pointer to the raw breakpoint itself (note that's still opaque outside mem-break.c), so that insert_memory_breakpoint / remove_memory_breakpoint have access to the breakpoint's shadow buffer. I first tried passing down a new structure based on GDB's "struct bp_target_info" (actually with that name exactly), but then decided against it as unnecessary complication. As software/memory breakpoints work by poking at memory, when setting a GDB Z0 breakpoint (but not internal breakpoints, as those can assume the conditions are already right), we need to tell the target to prepare to access memory (which on Linux means stop threads). If that operation fails, we need to return error to GDB. Seeing an error, if this is the first breakpoint of that type that GDB tries to insert, GDB would then assume the breakpoint type is supported, but it may actually not be. So we need to check whether the type is supported at all before preparing to access memory. And to solve that, the patch adds a new target->supports_z_point_type method that is called before actually trying to insert the breakpoint. Other than that, hopefully the change is more or less obvious. New test added that exercises the hbreak2.exp regression in a more direct way, without relying on a breakpoint re-set happening before main is reached. Tested by building GDBserver for: aarch64-linux-gnu arm-linux-gnueabihf i686-pc-linux-gnu i686-w64-mingw32 m68k-linux-gnu mips-linux-gnu mips-uclinux nios2-linux-gnu powerpc-linux-gnu sh-linux-gnu tilegx-unknown-linux-gnu x86_64-redhat-linux x86_64-w64-mingw32 And also regression tested on x86_64 Fedora 20. gdb/gdbserver/ 2014-05-20 Pedro Alves <palves@redhat.com> * linux-aarch64-low.c (aarch64_insert_point) (aarch64_remove_point): No longer check whether the type is supported here. Adjust to new interface. (the_low_target): Install aarch64_supports_z_point_type as supports_z_point_type method. * linux-arm-low.c (raw_bkpt_type_to_arm_hwbp_type): New function. (arm_linux_hw_point_initialize): Take an enum raw_bkpt_type instead of a Z packet char. Adjust. (arm_supports_z_point_type): New function. (arm_insert_point, arm_remove_point): Adjust to new interface. (the_low_target): Install arm_supports_z_point_type. * linux-crisv32-low.c (cris_supports_z_point_type): New function. (cris_insert_point, cris_remove_point): Adjust to new interface. Don't check whether the type is supported here. (the_low_target): Install cris_supports_z_point_type. * linux-low.c (linux_supports_z_point_type): New function. (linux_insert_point, linux_remove_point): Adjust to new interface. * linux-low.h (struct linux_target_ops) <insert_point, remove_point>: Take an enum raw_bkpt_type instead of a char. Add raw_breakpoint pointer parameter. <supports_z_point_type>: New method. * linux-mips-low.c (mips_supports_z_point_type): New function. (mips_insert_point, mips_remove_point): Adjust to new interface. Use mips_supports_z_point_type. (the_low_target): Install mips_supports_z_point_type. * linux-ppc-low.c (the_low_target): Install NULL as supports_z_point_type method. * linux-s390-low.c (the_low_target): Install NULL as supports_z_point_type method. * linux-sparc-low.c (the_low_target): Install NULL as supports_z_point_type method. * linux-x86-low.c (x86_supports_z_point_type): New function. (x86_insert_point): Adjust to new insert_point interface. Use insert_memory_breakpoint. Adjust to new i386_low_insert_watchpoint interface. (x86_remove_point): Adjust to remove_point interface. Use remove_memory_breakpoint. Adjust to new i386_low_remove_watchpoint interface. (the_low_target): Install x86_supports_z_point_type. * lynx-low.c (lynx_target_ops): Install NULL as supports_z_point_type callback. * nto-low.c (nto_supports_z_point_type): New. (nto_insert_point, nto_remove_point): Adjust to new interface. (nto_target_ops): Install nto_supports_z_point_type. * mem-break.c: Adjust intro comment. (struct raw_breakpoint) <raw_type, size>: New fields. <inserted>: Update comment. <shlib_disabled>: Delete field. (enum bkpt_type) <gdb_breakpoint>: Delete value. <gdb_breakpoint_Z0, gdb_breakpoint_Z1, gdb_breakpoint_Z2, gdb_breakpoint_Z3, gdb_breakpoint_Z4>: New values. (raw_bkpt_type_to_target_hw_bp_type): New function. (find_enabled_raw_code_breakpoint_at): New function. (find_raw_breakpoint_at): New type and size parameters. Use them. (insert_memory_breakpoint): New function, based off set_raw_breakpoint_at. (remove_memory_breakpoint): New function. (set_raw_breakpoint_at): Reimplement. (set_breakpoint): New, based on set_breakpoint_at. (set_breakpoint_at): Reimplement. (delete_raw_breakpoint): Go through the_target->remove_point instead of assuming memory breakpoints. (find_gdb_breakpoint_at): Delete. (Z_packet_to_bkpt_type, Z_packet_to_raw_bkpt_type): New functions. (find_gdb_breakpoint): New function. (set_gdb_breakpoint_at): Delete. (z_type_supported): New function. (set_gdb_breakpoint_1): New function, loosely based off set_gdb_breakpoint_at. (check_gdb_bp_preconditions, set_gdb_breakpoint): New functions. (delete_gdb_breakpoint_at): Delete. (delete_gdb_breakpoint_1): New function, loosely based off delete_gdb_breakpoint_at. (delete_gdb_breakpoint): New function. (clear_gdb_breakpoint_conditions): Rename to ... (clear_breakpoint_conditions): ... this. Don't handle a NULL breakpoint. (add_condition_to_breakpoint): Make static. (add_breakpoint_condition): Take a struct breakpoint pointer instead of an address. Adjust. (gdb_condition_true_at_breakpoint): Rename to ... (gdb_condition_true_at_breakpoint_z_type): ... this, and add z_type parameter. (gdb_condition_true_at_breakpoint): Reimplement. (add_breakpoint_commands): Take a struct breakpoint pointer instead of an address. Adjust. (gdb_no_commands_at_breakpoint): Rename to ... (gdb_no_commands_at_breakpoint_z_type): ... this. Add z_type parameter. Return true if no breakpoint was found. Change debug output. (gdb_no_commands_at_breakpoint): Reimplement. (run_breakpoint_commands): Rename to ... (run_breakpoint_commands_z_type): ... this. Add z_type parameter, and change return type to boolean. (run_breakpoint_commands): New function. (gdb_breakpoint_here): Also check for Z1 breakpoints. (uninsert_raw_breakpoint): Don't try to reinsert a disabled breakpoint. Go through the_target->remove_point instead of assuming memory breakpoint. (uninsert_breakpoints_at, uninsert_all_breakpoints): Uninsert software and hardware breakpoints. (reinsert_raw_breakpoint): Go through the_target->insert_point instead of assuming memory breakpoint. (reinsert_breakpoints_at, reinsert_all_breakpoints): Reinsert software and hardware breakpoints. (check_breakpoints, breakpoint_here, breakpoint_inserted_here): Check both software and hardware breakpoints. (validate_inserted_breakpoint): Assert the breakpoint is a software breakpoint. Set the inserted flag to -1 instead of setting shlib_disabled. (delete_disabled_breakpoints): Adjust. (validate_breakpoints): Only validate software breakpoints. Adjust to inserted flag change. (check_mem_read, check_mem_write): Skip breakpoint types other than software breakpoints. Adjust to inserted flag change. * mem-break.h (enum raw_bkpt_type): New enum. (raw_breakpoint, struct process_info): Forward declare. (Z_packet_to_target_hw_bp_type): Delete declaration. (raw_bkpt_type_to_target_hw_bp_type, Z_packet_to_raw_bkpt_type) (set_gdb_breakpoint, delete_gdb_breakpoint) (clear_breakpoint_conditions): New declarations. (set_gdb_breakpoint_at, clear_gdb_breakpoint_conditions): Delete. (breakpoint_inserted_here): Update comment. (add_breakpoint_condition, add_breakpoint_commands): Replace address parameter with a breakpoint pointer parameter. (gdb_breakpoint_here): Update comment. (delete_gdb_breakpoint_at): Delete. (insert_memory_breakpoint, remove_memory_breakpoint): Declare. * server.c (process_point_options): Take a struct breakpoint pointer instead of an address. Adjust. (process_serial_event) <Z/z packets>: Use set_gdb_breakpoint and delete_gdb_breakpoint. * spu-low.c (spu_target_ops): Install NULL as supports_z_point_type method. * target.h: Include mem-break.h. (struct target_ops) <prepare_to_access_memory>: Update comment. <supports_z_point_type>: New field. <insert_point, remove_point>: Take an enum raw_bkpt_type argument instead of a char. Also take a raw breakpoint pointer. * win32-arm-low.c (the_low_target): Install NULL as supports_z_point_type. * win32-i386-low.c (i386_supports_z_point_type): New function. (i386_insert_point, i386_remove_point): Adjust to new interface. (the_low_target): Install i386_supports_z_point_type. * win32-low.c (win32_supports_z_point_type): New function. (win32_insert_point, win32_remove_point): Adjust to new interface. (win32_target_ops): Install win32_supports_z_point_type. * win32-low.h (struct win32_target_ops): <supports_z_point_type>: New method. <insert_point, remove_point>: Take an enum raw_bkpt_type argument instead of a char. Also take a raw breakpoint pointer. gdb/testsuite/ 2014-05-20 Pedro Alves <palves@redhat.com> * gdb.base/break-idempotent.c: New file. * gdb.base/break-idempotent.exp: New file.
2014-05-20 19:24:28 +02:00
/* Returns TRUE if there's any inserted software or hardware (code)
breakpoint set at ADDR. */
Teach linux gdbserver to step-over-breakpoints. * linux-low.c (can_hardware_single_step): New. (supports_breakpoints): New. (handle_extended_wait): If stopping threads, read the stop pc of the new cloned LWP. (get_pc): New. (get_stop_pc): Add `lwp' parameter. Handle it. Bail out if the low target doesn't support retrieving the PC. (add_lwp): Set last_resume_kind to resume_continue. (linux_attach_lwp_1): Adjust comments. Always set stop_expected. (linux_attach): Don't clear stop_expected. Set the lwp's last_resume_kind to resume_stop. (linux_detach_one_lwp): Don't check for removed breakpoints. (check_removed_breakpoint): Delete. (status_pending_p): Rename to ... (status_pending_p_callback): ... this. Don't check for removed breakpoints. Don't consider threads that are stopped from GDB's perspective. (linux_wait_for_lwp): Always read the stop_pc here. (cancel_breakpoint): New. (step_over_bkpt): New global. (linux_wait_for_event_1): Implement stepping over breakpoints. (gdb_wants_lwp_stopped): New. (gdb_wants_all_stopped): New. (linux_wait_1): Tag threads as gdb-wants-stopped. Cancel finished single-step traps here. Store the thread's last reported target wait status. (send_sigstop): Don't clear stop_expected. Always set it, instead. (mark_lwp_dead): Remove reference to pending_is_breakpoint. (cancel_finished_single_step): New. (cancel_finished_single_steps): New. (wait_for_sigstop): Don't cancel finished single-step traps here. (linux_resume_one_lwp): Don't check for removed breakpoints. Don't set `step' on non-hardware step archs. (linux_set_resume_request): Ignore resume_stop requests if already stopping or stopped. Set the lwp's last_resume_kind. (resume_status_pending_p): Don't check for removed breakpoints. (need_step_over_p): New. (start_step_over): New. (finish_step_over): New. (linux_resume_one_thread): Always queue a sigstop for resume_stop requests. Clear the thread's last reported target waitstatus. Don't use the `suspended' flag. Don't consider pending breakpoints. (linux_resume): Start a step-over if necessary. (proceed_one_lwp): New. (proceed_all_lwps): New. (unstop_all_lwps): New. * linux-low.h (struct lwp_info): Rewrite comment for the `suspended' flag. Add the `stop_pc' field. Delete the `pending_stop_pc' field. Tweak the `stepping' flag's comment. Add `'last_resume_kind' and `need_step_over' fields. * inferiors.c (struct thread_info): Delete, moved elsewhere. * mem-break.c (struct breakpoint): Delete `reinserting' flag. Delete `breakpoint_to_reinsert' field. New flag `inserted'. (set_raw_breakpoint_at): New. (set_breakpoint_at): Rewrite to use it. (reinsert_breakpoint_handler): Delete. (set_reinsert_breakpoint): New. (reinsert_breakpoint_by_bp): Delete. (delete_reinsert_breakpoints): New. (uninsert_breakpoint): Rewrite. (uninsert_breakpoints_at): New. (reinsert_breakpoint): Rewrite. (reinsert_breakpoints_at): New. (check_breakpoints): Rewrite. (breakpoint_here): New. (breakpoint_inserted_here): New. (check_mem_read): Adjust. * mem-break.h (breakpoints_supported, breakpoint_here) (breakpoint_inserted_here, set_reinsert_breakpoint): Declare. (reinsert_breakpoint_by_bp): Delete declaration. (delete_reinsert_breakpoints): Declare. (reinsert_breakpoint): Delete declaration. (reinsert_breakpoints_at): Declare. (uninsert_breakpoint): Delete declaration. (uninsert_breakpoints_at): Declare. (check_breakpoints): Adjust prototype. * server.h: Adjust include order. (struct thread_info): Declare here. Add a `last_status' field.
2010-03-24 01:05:03 +01:00
int breakpoint_inserted_here (CORE_ADDR addr);
[gdbserver] linux-low.c: better starvation avoidance, handle non-stop mode too This patch applies the same starvation avoidance improvements of the previous patch to the Linux gdbserver side. Without this, the test added by the following commit (gdb.threads/non-stop-fair-events.exp) always fails with time outs. gdb/gdbserver/ 2015-01-09 Pedro Alves <palves@redhat.com> * linux-low.c (step_over_bkpt): Move higher up in the file. (handle_extended_wait): Don't store the stop_pc here. (get_stop_pc): Adjust comments and rename to ... (check_stopped_by_breakpoint): ... this. Record whether the LWP stopped for a software breakpoint or hardware breakpoint. (thread_still_has_status_pending_p): New function. (status_pending_p_callback): Use thread_still_has_status_pending_p. If the event is no longer interesting, resume the LWP. (handle_tracepoints): Add assert. (maybe_move_out_of_jump_pad): Remove cancel_breakpoints call. (wstatus_maybe_breakpoint): New function. (cancel_breakpoint): Delete function. (check_stopped_by_watchpoint): New function, factored out from linux_low_filter_event. (lp_status_maybe_breakpoint): Delete function. (linux_low_filter_event): Remove filter_ptid argument. Leave thread group exits pending here. Store the LWP's stop PC. Always leave events pending. (linux_wait_for_event_filtered): Pull all events out of the kernel, and leave them all pending. (count_events_callback, select_event_lwp_callback): Consider all events. (cancel_breakpoints_callback, linux_cancel_breakpoints): Delete. (select_event_lwp): Only give preference to the stepping LWP in all-stop mode. Adjust comments. (ignore_event): New function. (linux_wait_1): Delete 'retry' label. Use ignore_event. Remove references to cancel_breakpoints. Adjust to renames. Also give equal priority to all LWPs that have had events in non-stop mode. If reporting a software breakpoint event, unadjust the LWP's PC. (linux_wait): If linux_wait_1 returned an ignored event, retry. (stuck_in_jump_pad_callback, move_out_of_jump_pad_callback): Adjust. (linux_resume_one_lwp): Store the LWP's PC. Adjust. (resume_status_pending_p): Use thread_still_has_status_pending_p. (linux_stopped_by_watchpoint): Adjust. (linux_target_ops): Remove reference to linux_cancel_breakpoints. * linux-low.h (enum lwp_stop_reason): New. (struct lwp_info) <stop_pc>: Adjust comment. <stopped_by_watchpoint>: Delete field. <stop_reason>: New field. * linux-x86-low.c (x86_linux_prepare_to_resume): Adjust. * mem-break.c (software_breakpoint_inserted_here) (hardware_breakpoint_inserted_here): New function. * mem-break.h (software_breakpoint_inserted_here) (hardware_breakpoint_inserted_here): Declare. * target.h (struct target_ops) <cancel_breakpoints>: Remove field. (cancel_breakpoints): Delete. * tracepoint.c (clear_installed_tracepoints, stop_tracing) (upload_fast_traceframes): Remove references to cancel_breakpoints.
2014-12-29 20:41:07 +01:00
/* Returns TRUE if there's any inserted software breakpoint at
ADDR. */
int software_breakpoint_inserted_here (CORE_ADDR addr);
/* Returns TRUE if there's any inserted hardware (code) breakpoint at
ADDR. */
int hardware_breakpoint_inserted_here (CORE_ADDR addr);
Fix instruction skipping when using software single step in GDBServer Without this patch, when doing a software single step, with for example a conditional breakpoint, gdbserver would wrongly avance the pc of breakpoint_len and skips an instruction. This is due to gdbserver assuming that it's hardware single stepping. When it resumes from the breakpoint address it expects the trap to be caused by ptrace and if it's rather caused by a software breakpoint it assumes this is a permanent breakpoint and that it needs to skip over it. However when software single stepping, this breakpoint is legitimate as it's the reinsert breakpoint gdbserver has put in place to break at the next instruction. Thus gdbserver wrongly advances the pc and skips an instruction. This patch fixes this behavior so that gdbserver checks if it is a reinsert breakpoint from software single stepping. If it is it won't advance the pc. And if there's no reinsert breakpoint there we assume then that it's a permanent breakpoint and advance the pc. Here's a commented log of what would happen before and after the fix on gdbserver : /* Here there is a conditional breakpoint at 0x10428 that needs to be stepped over. */ Need step over [LWP 11204]? yes, found breakpoint at 0x10428 ... /* e7f001f0 is a breakpoint instruction on arm Here gdbserver writes the software breakpoint we would like to hit */ Writing e7f001f0 to 0x0001042c in process 11204 ... Resuming lwp 11220 (continue, signal 0, stop not expected) pending reinsert at 0x10428 stop pc is 00010428 continue from pc 0x10428 ... /* Here gdbserver hit the software breakpoint that was in place for the step over */ stop pc is 0001042c pc is 0x1042c step-over for LWP 11220.11220 executed software breakpoint Finished step over. Could not find fast tracepoint jump at 0x10428 in list (reinserting). /* Here gdbserver writes back the original instruction */ Writing e50b3008 to 0x0001042c in process 11220 Step-over finished. Need step over [LWP 11220]? No /* Here because gdbserver assumes this is a permenant breakpoint it advances the pc of breakpoint_len, in this case 4 bytes, so we have just skipped the instruction that was written back here : Writing e50b3008 to 0x0001042c in process 11220 */ stop pc is 00010430 pc is 0x10430 Need step over [LWP 11220]? No, no breakpoint found at 0x10430 Proceeding, no step-over needed proceed_one_lwp: lwp 11220 stop pc is 00010430 This patch fixes this situation and we get the right behavior : Writing e50b3008 to 0x0001042c in process 11245 Hit a gdbserver breakpoint. Hit a gdbserver breakpoint. Step-over finished. proceeding all threads. Need step over [LWP 11245]? No stop pc is 0001042c pc is 0x1042c Need step over [LWP 11245]? No, no breakpoint found at 0x1042c Proceeding, no step-over needed proceed_one_lwp: lwp 11245 stop pc is 0001042c pc is 0x1042c Resuming lwp 11245 (continue, signal 0, stop not expected) stop pc is 0001042c continue from pc 0x1042c It also works if the value at 0x0001042c is a permanent breakpoint. If so gdbserver will finish the step over, remove the reinserted breakpoint, resume at that location and on the next SIGTRAP gdbserver will trigger the advance PC condition as reinsert_breakpoint_inserted_here will be false. I also tested this against bp-permanent.exp on arm (with a work in progress software single step patchset) without any regressions. It's also tested against x86 bp-permanent.exp without any regression. So both software and hardware single step are tested. No regressions on Ubuntu 14.04 on ARMv7 and x86. With gdbserver-{native,extended} / { -marm -mthumb } gdb/gdbserver/ChangeLog: * linux-low.c (linux_wait_1): Fix pc advance condition. * mem-break.c (reinsert_breakpoint_inserted_here): New function. * mem-break.h (reinsert_breakpoint_inserted_here): New declaration.
2015-11-30 21:16:22 +01:00
/* Returns TRUE if there's any reinsert breakpoint at ADDR. */
int reinsert_breakpoint_inserted_here (CORE_ADDR addr);
/* Clear all breakpoint conditions and commands associated with a
breakpoint. */
void clear_breakpoint_conditions_and_commands (struct breakpoint *bp);
Conditional Z1 breakpoint hangs GDBserver. While trying to fix hbreak2.exp against GDBserver I noticed this... (gdb) hbreak main if 1 Sending packet: $m400580,40#2e...Packet received: e8d2ffffff5dc3554889e54883ec10c745fc00000000eb0eb800000000e8c1ffffff8345fc01817dfce70300007ee9b800000000c9c3662e0f1f840000000000 Sending packet: $m40058f,1#31...Packet received: c7 Hardware assisted breakpoint 1 at 0x40058f: file ../../../src/gdb/testsuite/gdb.base/break-idempotent.c, line 46. Sending packet: $Z1,40058f,1;X3,220127#9b... *hangs forever* The issue is that nothing advances the packet pointer if add_breakpoint_condition either fails to parse the agent expression, or fails to find the breakpoint, resulting in an infinite loop in process_point_options. The latter case should really be fixed by GDBserver tracking GDB Z1 breakpoints in its breakpoint structures like Z0 breakpoints are, but the latter case still needs handling. add_breakpoint_commands has the same issue, though at present I don't know any way to trigger it other than sending a manually cooked packet. Unbelievably, it doesn't look like we have any test that tries setting a conditional hardware breakpoint. Looking at cond-eval-mode.exp, it looks like the file was meant to actually test something, but it's mostly empty today. This patch adds tests that tries all sorts of conditional breakpoints and watchpoints. The test hangs/fails without the GDBserver fix. Tested on x86_64 Fedora 17. gdb/gdbserver/ 2014-04-10 Pedro Alves <palves@redhat.com> * mem-break.c (add_breakpoint_condition, add_breakpoint_commands): Check if the condition or command is NULL before checking if the breakpoint is known. On success, return true. * mem-break.h (add_breakpoint_condition): Document return. (add_breakpoint_commands): Add describing comment. * server.c (skip_to_semicolon): New function. (process_point_options): Use it. gdb/testsuite/ 2014-04-10 Pedro Alves <palves@redhat.com> * gdb.base/cond-eval-mode.c: New file. * gdb.base/cond-eval-mode.exp: Use standard_testfile. Adjust prepare_for_testing to build the new file. Check result of runto_main. (test_break, test_watch): New procedures. (top level): Use them.
2014-04-10 20:22:23 +02:00
/* Set target-side condition CONDITION to the breakpoint at ADDR.
Returns false on failure. On success, advances CONDITION pointer
past the condition and returns true. */
[GDBserver] Make Zx/zx packet handling idempotent. This patch fixes hardware breakpoint regressions exposed by my fix for "PR breakpoints/7143 - Watchpoint does not trigger when first set", at https://sourceware.org/ml/gdb-patches/2014-03/msg00167.html The testsuite caught them on Linux/x86_64, at least. gdb.sum: gdb.sum: FAIL: gdb.base/hbreak2.exp: next over recursive call FAIL: gdb.base/hbreak2.exp: backtrace from factorial(5.1) FAIL: gdb.base/hbreak2.exp: continue until exit at recursive next test gdb.log: (gdb) next Program received signal SIGTRAP, Trace/breakpoint trap. factorial (value=4) at ../../../src/gdb/testsuite/gdb.base/break.c:113 113 if (value > 1) { /* set breakpoint 7 here */ (gdb) FAIL: gdb.base/hbreak2.exp: next over recursive call Actually, that patch just exposed a latent issue to "breakpoints always-inserted off" mode, not really caused it. After that patch, GDB no longer removes breakpoints at each internal event, thus making some scenarios behave like breakpoint always-inserted on. The bug is easy to trigger with always-inserted on. The issue is that since the target-side breakpoint conditions support, if the stub/server supports evaluating breakpoint conditions on the target side, then GDB is sending duplicate Zx packets to the target without removing them before, and GDBserver is not really expecting that for Z packets other than Z0/z0. E.g., with "set breakpoint always-inserted on" and "set debug remote 1": (gdb) b main Sending packet: $m410943,1#ff...Packet received: 48 Breakpoint 4 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z0,410943,1#48...Packet received: OK ^^^^^^^^^^^^ (gdb) b main Note: breakpoint 4 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Breakpoint 5 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z0,410943,1#48...Packet received: OK ^^^^^^^^^^^^ (gdb) b main Note: breakpoints 4 and 5 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Breakpoint 6 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z0,410943,1#48...Packet received: OK ^^^^^^^^^^^^ (gdb) del Delete all breakpoints? (y or n) y Sending packet: $Z0,410943,1#48...Packet received: OK Sending packet: $Z0,410943,1#48...Packet received: OK Sending packet: $z0,410943,1#68...Packet received: OK And for Z1, similarly: (gdb) hbreak main Sending packet: $m410943,1#ff...Packet received: 48 Hardware assisted breakpoint 4 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ Packet Z1 (hardware-breakpoint) is supported (gdb) hbreak main Note: breakpoint 4 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Hardware assisted breakpoint 5 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ (gdb) hbreak main Note: breakpoints 4 and 5 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Hardware assisted breakpoint 6 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ (gdb) del Delete all breakpoints? (y or n) y Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ Sending packet: $z1,410943,1#69...Packet received: OK ^^^^^^^^^^^^ So GDB sent a bunch of Z1 packets, and then when finally removing the breakpoint, only one z1 packet was sent. On the GDBserver side (with monitor set debug-hw-points 1), in the Z1 case, we see: $ ./gdbserver :9999 ./gdbserver Process ./gdbserver created; pid = 8629 Listening on port 9999 Remote debugging from host 127.0.0.1 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=1 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=2 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=3 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=4 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=5 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 remove_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=4 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 That's one insert_watchpoint call for each Z1 packet, and then one remove_watchpoint call for the z1 packet. Notice how ref.count increased for each insert_watchpoint call, and then in the end, after GDB told GDBserver to forget about the hardware breakpoint, GDBserver ends with the the first debug register still with ref.count=4! IOW, the hardware breakpoint is left armed on the target, while on the GDB end it's gone. If the program happens to execute 0x410943 afterwards, then the CPU traps, GDBserver reports the trap to GDB, and GDB not having a breakpoint set at that address anymore, reports to the user a spurious SIGTRAP. This is exactly what is happening in the hbreak2.exp test, though in that case, it's a shared library event that triggers a breakpoint_re_set, when breakpoints are still inserted (because nowadays GDB doesn't remove breakpoints while handling internal events), and that recreates breakpoint locations, which likewise forces breakpoint reinsertion and Zx packet resends... That is a lot of bogus Zx duplication that should possibly be addressed on the GDB side. GDB resends Zx packets because the way to change the target-side condition, is to resend the breakpoint to the server with the new condition. (That's an option in the packet: e.g., "Z1,410943,1;X3,220027" for "hbreak main if 0". The packets in the examples above are shorter because the breakpoints don't have conditions attached). GDB doesn't remove the breakpoint first before reinserting it because that'd be bad for non-stop, as it'd open a window where the inferior could miss the breakpoint. The conditions actually haven't changed between the resends, but GDB isn't smart enough to realize that. (TBC, if the target doesn't support target-side conditions, then GDB doesn't trigger these resends (init_bp_location calls mark_breakpoint_location_modified, and that does nothing if condition evaluation is on the host side. The resends are caused by the 'loc->condition_changed = condition_modified.' line.) But, even if GDB was made smarter, GDBserver should really still handle the resends anyway. So target-side conditions also aren't really to blame. The documentation of the Z/z packets says: "To avoid potential problems with duplicate packets, the operations should be implemented in an idempotent way." As such, we may want to fix GDB, but we should definitely fix GDBserver. The fix is a prerequisite for target-side conditions on hardware breakpoints anyway (and while at it, on watchpoints too). GDBserver indeed already treats duplicate Z0 packets in an idempotent way. mem-break.c has the concept of high-level and low-level breakpoints, somewhat similar to GDB's split of breakpoints vs breakpoint locations, and keeps track of multiple breakpoints referencing the same address/location, for the case of an internal GDBserver breakpoint or a tracepoint being set at the same address as a GDB breakpoint. But, it only allows GDB to ever contribute one reference to a software breakpoint location. IOW, if gdbserver sees a Z0 packet for the same address where it already had a GDB breakpoint set, then GDBserver won't create another high-level GDB breakpoint. However, mem-break.c only tracks GDB Z0 breakpoints. The same logic should apply to all kinds of Zx packets. Currently, gdbserver passes down each duplicate Zx (other than Z0) request directly to the target->insert_point routine. The x86 watchpoint support itself refcounts watchpoint / hw breakpoint requests, to handle overlapping watchpoints, and save debug registers. But that code doesn't (and really shouldn't) handle the duplicate requests, assuming that for each insert there will be a corresponding remove. So the fix is to generalize mem-break.c to track all kinds of Zx breakpoints, and filter out duplicates. As mentioned, this ends up adding support for target-side conditions on hardware breakpoints and watchpoints too (though GDB itself doesn't support the latter yet). Probably the least obvious change in the patch is that it kind of turns the breakpoint insert/remove APIs inside out. Before, the target methods were only called for GDB breakpoints. The internal breakpoint set/delete methods inserted memory breakpoints directly bypassing the insert/remove target methods. That's not good when the target should use a debug API to set software breakpoints, instead of relying on GDBserver patching memory with breakpoint instructions, as is the case of NTO. Now removal/insertion of all kinds of breakpoints/watchpoints, either internal, or from GDB, always go through the target methods. The insert_point/remove_point methods no longer get passed a Z packet type, but an internal/raw breakpoint type. They're also passed a pointer to the raw breakpoint itself (note that's still opaque outside mem-break.c), so that insert_memory_breakpoint / remove_memory_breakpoint have access to the breakpoint's shadow buffer. I first tried passing down a new structure based on GDB's "struct bp_target_info" (actually with that name exactly), but then decided against it as unnecessary complication. As software/memory breakpoints work by poking at memory, when setting a GDB Z0 breakpoint (but not internal breakpoints, as those can assume the conditions are already right), we need to tell the target to prepare to access memory (which on Linux means stop threads). If that operation fails, we need to return error to GDB. Seeing an error, if this is the first breakpoint of that type that GDB tries to insert, GDB would then assume the breakpoint type is supported, but it may actually not be. So we need to check whether the type is supported at all before preparing to access memory. And to solve that, the patch adds a new target->supports_z_point_type method that is called before actually trying to insert the breakpoint. Other than that, hopefully the change is more or less obvious. New test added that exercises the hbreak2.exp regression in a more direct way, without relying on a breakpoint re-set happening before main is reached. Tested by building GDBserver for: aarch64-linux-gnu arm-linux-gnueabihf i686-pc-linux-gnu i686-w64-mingw32 m68k-linux-gnu mips-linux-gnu mips-uclinux nios2-linux-gnu powerpc-linux-gnu sh-linux-gnu tilegx-unknown-linux-gnu x86_64-redhat-linux x86_64-w64-mingw32 And also regression tested on x86_64 Fedora 20. gdb/gdbserver/ 2014-05-20 Pedro Alves <palves@redhat.com> * linux-aarch64-low.c (aarch64_insert_point) (aarch64_remove_point): No longer check whether the type is supported here. Adjust to new interface. (the_low_target): Install aarch64_supports_z_point_type as supports_z_point_type method. * linux-arm-low.c (raw_bkpt_type_to_arm_hwbp_type): New function. (arm_linux_hw_point_initialize): Take an enum raw_bkpt_type instead of a Z packet char. Adjust. (arm_supports_z_point_type): New function. (arm_insert_point, arm_remove_point): Adjust to new interface. (the_low_target): Install arm_supports_z_point_type. * linux-crisv32-low.c (cris_supports_z_point_type): New function. (cris_insert_point, cris_remove_point): Adjust to new interface. Don't check whether the type is supported here. (the_low_target): Install cris_supports_z_point_type. * linux-low.c (linux_supports_z_point_type): New function. (linux_insert_point, linux_remove_point): Adjust to new interface. * linux-low.h (struct linux_target_ops) <insert_point, remove_point>: Take an enum raw_bkpt_type instead of a char. Add raw_breakpoint pointer parameter. <supports_z_point_type>: New method. * linux-mips-low.c (mips_supports_z_point_type): New function. (mips_insert_point, mips_remove_point): Adjust to new interface. Use mips_supports_z_point_type. (the_low_target): Install mips_supports_z_point_type. * linux-ppc-low.c (the_low_target): Install NULL as supports_z_point_type method. * linux-s390-low.c (the_low_target): Install NULL as supports_z_point_type method. * linux-sparc-low.c (the_low_target): Install NULL as supports_z_point_type method. * linux-x86-low.c (x86_supports_z_point_type): New function. (x86_insert_point): Adjust to new insert_point interface. Use insert_memory_breakpoint. Adjust to new i386_low_insert_watchpoint interface. (x86_remove_point): Adjust to remove_point interface. Use remove_memory_breakpoint. Adjust to new i386_low_remove_watchpoint interface. (the_low_target): Install x86_supports_z_point_type. * lynx-low.c (lynx_target_ops): Install NULL as supports_z_point_type callback. * nto-low.c (nto_supports_z_point_type): New. (nto_insert_point, nto_remove_point): Adjust to new interface. (nto_target_ops): Install nto_supports_z_point_type. * mem-break.c: Adjust intro comment. (struct raw_breakpoint) <raw_type, size>: New fields. <inserted>: Update comment. <shlib_disabled>: Delete field. (enum bkpt_type) <gdb_breakpoint>: Delete value. <gdb_breakpoint_Z0, gdb_breakpoint_Z1, gdb_breakpoint_Z2, gdb_breakpoint_Z3, gdb_breakpoint_Z4>: New values. (raw_bkpt_type_to_target_hw_bp_type): New function. (find_enabled_raw_code_breakpoint_at): New function. (find_raw_breakpoint_at): New type and size parameters. Use them. (insert_memory_breakpoint): New function, based off set_raw_breakpoint_at. (remove_memory_breakpoint): New function. (set_raw_breakpoint_at): Reimplement. (set_breakpoint): New, based on set_breakpoint_at. (set_breakpoint_at): Reimplement. (delete_raw_breakpoint): Go through the_target->remove_point instead of assuming memory breakpoints. (find_gdb_breakpoint_at): Delete. (Z_packet_to_bkpt_type, Z_packet_to_raw_bkpt_type): New functions. (find_gdb_breakpoint): New function. (set_gdb_breakpoint_at): Delete. (z_type_supported): New function. (set_gdb_breakpoint_1): New function, loosely based off set_gdb_breakpoint_at. (check_gdb_bp_preconditions, set_gdb_breakpoint): New functions. (delete_gdb_breakpoint_at): Delete. (delete_gdb_breakpoint_1): New function, loosely based off delete_gdb_breakpoint_at. (delete_gdb_breakpoint): New function. (clear_gdb_breakpoint_conditions): Rename to ... (clear_breakpoint_conditions): ... this. Don't handle a NULL breakpoint. (add_condition_to_breakpoint): Make static. (add_breakpoint_condition): Take a struct breakpoint pointer instead of an address. Adjust. (gdb_condition_true_at_breakpoint): Rename to ... (gdb_condition_true_at_breakpoint_z_type): ... this, and add z_type parameter. (gdb_condition_true_at_breakpoint): Reimplement. (add_breakpoint_commands): Take a struct breakpoint pointer instead of an address. Adjust. (gdb_no_commands_at_breakpoint): Rename to ... (gdb_no_commands_at_breakpoint_z_type): ... this. Add z_type parameter. Return true if no breakpoint was found. Change debug output. (gdb_no_commands_at_breakpoint): Reimplement. (run_breakpoint_commands): Rename to ... (run_breakpoint_commands_z_type): ... this. Add z_type parameter, and change return type to boolean. (run_breakpoint_commands): New function. (gdb_breakpoint_here): Also check for Z1 breakpoints. (uninsert_raw_breakpoint): Don't try to reinsert a disabled breakpoint. Go through the_target->remove_point instead of assuming memory breakpoint. (uninsert_breakpoints_at, uninsert_all_breakpoints): Uninsert software and hardware breakpoints. (reinsert_raw_breakpoint): Go through the_target->insert_point instead of assuming memory breakpoint. (reinsert_breakpoints_at, reinsert_all_breakpoints): Reinsert software and hardware breakpoints. (check_breakpoints, breakpoint_here, breakpoint_inserted_here): Check both software and hardware breakpoints. (validate_inserted_breakpoint): Assert the breakpoint is a software breakpoint. Set the inserted flag to -1 instead of setting shlib_disabled. (delete_disabled_breakpoints): Adjust. (validate_breakpoints): Only validate software breakpoints. Adjust to inserted flag change. (check_mem_read, check_mem_write): Skip breakpoint types other than software breakpoints. Adjust to inserted flag change. * mem-break.h (enum raw_bkpt_type): New enum. (raw_breakpoint, struct process_info): Forward declare. (Z_packet_to_target_hw_bp_type): Delete declaration. (raw_bkpt_type_to_target_hw_bp_type, Z_packet_to_raw_bkpt_type) (set_gdb_breakpoint, delete_gdb_breakpoint) (clear_breakpoint_conditions): New declarations. (set_gdb_breakpoint_at, clear_gdb_breakpoint_conditions): Delete. (breakpoint_inserted_here): Update comment. (add_breakpoint_condition, add_breakpoint_commands): Replace address parameter with a breakpoint pointer parameter. (gdb_breakpoint_here): Update comment. (delete_gdb_breakpoint_at): Delete. (insert_memory_breakpoint, remove_memory_breakpoint): Declare. * server.c (process_point_options): Take a struct breakpoint pointer instead of an address. Adjust. (process_serial_event) <Z/z packets>: Use set_gdb_breakpoint and delete_gdb_breakpoint. * spu-low.c (spu_target_ops): Install NULL as supports_z_point_type method. * target.h: Include mem-break.h. (struct target_ops) <prepare_to_access_memory>: Update comment. <supports_z_point_type>: New field. <insert_point, remove_point>: Take an enum raw_bkpt_type argument instead of a char. Also take a raw breakpoint pointer. * win32-arm-low.c (the_low_target): Install NULL as supports_z_point_type. * win32-i386-low.c (i386_supports_z_point_type): New function. (i386_insert_point, i386_remove_point): Adjust to new interface. (the_low_target): Install i386_supports_z_point_type. * win32-low.c (win32_supports_z_point_type): New function. (win32_insert_point, win32_remove_point): Adjust to new interface. (win32_target_ops): Install win32_supports_z_point_type. * win32-low.h (struct win32_target_ops): <supports_z_point_type>: New method. <insert_point, remove_point>: Take an enum raw_bkpt_type argument instead of a char. Also take a raw breakpoint pointer. gdb/testsuite/ 2014-05-20 Pedro Alves <palves@redhat.com> * gdb.base/break-idempotent.c: New file. * gdb.base/break-idempotent.exp: New file.
2014-05-20 19:24:28 +02:00
int add_breakpoint_condition (struct breakpoint *bp, char **condition);
Conditional Z1 breakpoint hangs GDBserver. While trying to fix hbreak2.exp against GDBserver I noticed this... (gdb) hbreak main if 1 Sending packet: $m400580,40#2e...Packet received: e8d2ffffff5dc3554889e54883ec10c745fc00000000eb0eb800000000e8c1ffffff8345fc01817dfce70300007ee9b800000000c9c3662e0f1f840000000000 Sending packet: $m40058f,1#31...Packet received: c7 Hardware assisted breakpoint 1 at 0x40058f: file ../../../src/gdb/testsuite/gdb.base/break-idempotent.c, line 46. Sending packet: $Z1,40058f,1;X3,220127#9b... *hangs forever* The issue is that nothing advances the packet pointer if add_breakpoint_condition either fails to parse the agent expression, or fails to find the breakpoint, resulting in an infinite loop in process_point_options. The latter case should really be fixed by GDBserver tracking GDB Z1 breakpoints in its breakpoint structures like Z0 breakpoints are, but the latter case still needs handling. add_breakpoint_commands has the same issue, though at present I don't know any way to trigger it other than sending a manually cooked packet. Unbelievably, it doesn't look like we have any test that tries setting a conditional hardware breakpoint. Looking at cond-eval-mode.exp, it looks like the file was meant to actually test something, but it's mostly empty today. This patch adds tests that tries all sorts of conditional breakpoints and watchpoints. The test hangs/fails without the GDBserver fix. Tested on x86_64 Fedora 17. gdb/gdbserver/ 2014-04-10 Pedro Alves <palves@redhat.com> * mem-break.c (add_breakpoint_condition, add_breakpoint_commands): Check if the condition or command is NULL before checking if the breakpoint is known. On success, return true. * mem-break.h (add_breakpoint_condition): Document return. (add_breakpoint_commands): Add describing comment. * server.c (skip_to_semicolon): New function. (process_point_options): Use it. gdb/testsuite/ 2014-04-10 Pedro Alves <palves@redhat.com> * gdb.base/cond-eval-mode.c: New file. * gdb.base/cond-eval-mode.exp: Use standard_testfile. Adjust prepare_for_testing to build the new file. Check result of runto_main. (test_break, test_watch): New procedures. (top level): Use them.
2014-04-10 20:22:23 +02:00
/* Set target-side commands COMMANDS to the breakpoint at ADDR.
Returns false on failure. On success, advances COMMANDS past the
commands and returns true. If PERSIST, the commands should run
even while GDB is disconnected. */
[GDBserver] Make Zx/zx packet handling idempotent. This patch fixes hardware breakpoint regressions exposed by my fix for "PR breakpoints/7143 - Watchpoint does not trigger when first set", at https://sourceware.org/ml/gdb-patches/2014-03/msg00167.html The testsuite caught them on Linux/x86_64, at least. gdb.sum: gdb.sum: FAIL: gdb.base/hbreak2.exp: next over recursive call FAIL: gdb.base/hbreak2.exp: backtrace from factorial(5.1) FAIL: gdb.base/hbreak2.exp: continue until exit at recursive next test gdb.log: (gdb) next Program received signal SIGTRAP, Trace/breakpoint trap. factorial (value=4) at ../../../src/gdb/testsuite/gdb.base/break.c:113 113 if (value > 1) { /* set breakpoint 7 here */ (gdb) FAIL: gdb.base/hbreak2.exp: next over recursive call Actually, that patch just exposed a latent issue to "breakpoints always-inserted off" mode, not really caused it. After that patch, GDB no longer removes breakpoints at each internal event, thus making some scenarios behave like breakpoint always-inserted on. The bug is easy to trigger with always-inserted on. The issue is that since the target-side breakpoint conditions support, if the stub/server supports evaluating breakpoint conditions on the target side, then GDB is sending duplicate Zx packets to the target without removing them before, and GDBserver is not really expecting that for Z packets other than Z0/z0. E.g., with "set breakpoint always-inserted on" and "set debug remote 1": (gdb) b main Sending packet: $m410943,1#ff...Packet received: 48 Breakpoint 4 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z0,410943,1#48...Packet received: OK ^^^^^^^^^^^^ (gdb) b main Note: breakpoint 4 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Breakpoint 5 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z0,410943,1#48...Packet received: OK ^^^^^^^^^^^^ (gdb) b main Note: breakpoints 4 and 5 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Breakpoint 6 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z0,410943,1#48...Packet received: OK ^^^^^^^^^^^^ (gdb) del Delete all breakpoints? (y or n) y Sending packet: $Z0,410943,1#48...Packet received: OK Sending packet: $Z0,410943,1#48...Packet received: OK Sending packet: $z0,410943,1#68...Packet received: OK And for Z1, similarly: (gdb) hbreak main Sending packet: $m410943,1#ff...Packet received: 48 Hardware assisted breakpoint 4 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ Packet Z1 (hardware-breakpoint) is supported (gdb) hbreak main Note: breakpoint 4 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Hardware assisted breakpoint 5 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ (gdb) hbreak main Note: breakpoints 4 and 5 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Hardware assisted breakpoint 6 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ (gdb) del Delete all breakpoints? (y or n) y Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ Sending packet: $z1,410943,1#69...Packet received: OK ^^^^^^^^^^^^ So GDB sent a bunch of Z1 packets, and then when finally removing the breakpoint, only one z1 packet was sent. On the GDBserver side (with monitor set debug-hw-points 1), in the Z1 case, we see: $ ./gdbserver :9999 ./gdbserver Process ./gdbserver created; pid = 8629 Listening on port 9999 Remote debugging from host 127.0.0.1 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=1 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=2 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=3 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=4 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=5 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 remove_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=4 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 That's one insert_watchpoint call for each Z1 packet, and then one remove_watchpoint call for the z1 packet. Notice how ref.count increased for each insert_watchpoint call, and then in the end, after GDB told GDBserver to forget about the hardware breakpoint, GDBserver ends with the the first debug register still with ref.count=4! IOW, the hardware breakpoint is left armed on the target, while on the GDB end it's gone. If the program happens to execute 0x410943 afterwards, then the CPU traps, GDBserver reports the trap to GDB, and GDB not having a breakpoint set at that address anymore, reports to the user a spurious SIGTRAP. This is exactly what is happening in the hbreak2.exp test, though in that case, it's a shared library event that triggers a breakpoint_re_set, when breakpoints are still inserted (because nowadays GDB doesn't remove breakpoints while handling internal events), and that recreates breakpoint locations, which likewise forces breakpoint reinsertion and Zx packet resends... That is a lot of bogus Zx duplication that should possibly be addressed on the GDB side. GDB resends Zx packets because the way to change the target-side condition, is to resend the breakpoint to the server with the new condition. (That's an option in the packet: e.g., "Z1,410943,1;X3,220027" for "hbreak main if 0". The packets in the examples above are shorter because the breakpoints don't have conditions attached). GDB doesn't remove the breakpoint first before reinserting it because that'd be bad for non-stop, as it'd open a window where the inferior could miss the breakpoint. The conditions actually haven't changed between the resends, but GDB isn't smart enough to realize that. (TBC, if the target doesn't support target-side conditions, then GDB doesn't trigger these resends (init_bp_location calls mark_breakpoint_location_modified, and that does nothing if condition evaluation is on the host side. The resends are caused by the 'loc->condition_changed = condition_modified.' line.) But, even if GDB was made smarter, GDBserver should really still handle the resends anyway. So target-side conditions also aren't really to blame. The documentation of the Z/z packets says: "To avoid potential problems with duplicate packets, the operations should be implemented in an idempotent way." As such, we may want to fix GDB, but we should definitely fix GDBserver. The fix is a prerequisite for target-side conditions on hardware breakpoints anyway (and while at it, on watchpoints too). GDBserver indeed already treats duplicate Z0 packets in an idempotent way. mem-break.c has the concept of high-level and low-level breakpoints, somewhat similar to GDB's split of breakpoints vs breakpoint locations, and keeps track of multiple breakpoints referencing the same address/location, for the case of an internal GDBserver breakpoint or a tracepoint being set at the same address as a GDB breakpoint. But, it only allows GDB to ever contribute one reference to a software breakpoint location. IOW, if gdbserver sees a Z0 packet for the same address where it already had a GDB breakpoint set, then GDBserver won't create another high-level GDB breakpoint. However, mem-break.c only tracks GDB Z0 breakpoints. The same logic should apply to all kinds of Zx packets. Currently, gdbserver passes down each duplicate Zx (other than Z0) request directly to the target->insert_point routine. The x86 watchpoint support itself refcounts watchpoint / hw breakpoint requests, to handle overlapping watchpoints, and save debug registers. But that code doesn't (and really shouldn't) handle the duplicate requests, assuming that for each insert there will be a corresponding remove. So the fix is to generalize mem-break.c to track all kinds of Zx breakpoints, and filter out duplicates. As mentioned, this ends up adding support for target-side conditions on hardware breakpoints and watchpoints too (though GDB itself doesn't support the latter yet). Probably the least obvious change in the patch is that it kind of turns the breakpoint insert/remove APIs inside out. Before, the target methods were only called for GDB breakpoints. The internal breakpoint set/delete methods inserted memory breakpoints directly bypassing the insert/remove target methods. That's not good when the target should use a debug API to set software breakpoints, instead of relying on GDBserver patching memory with breakpoint instructions, as is the case of NTO. Now removal/insertion of all kinds of breakpoints/watchpoints, either internal, or from GDB, always go through the target methods. The insert_point/remove_point methods no longer get passed a Z packet type, but an internal/raw breakpoint type. They're also passed a pointer to the raw breakpoint itself (note that's still opaque outside mem-break.c), so that insert_memory_breakpoint / remove_memory_breakpoint have access to the breakpoint's shadow buffer. I first tried passing down a new structure based on GDB's "struct bp_target_info" (actually with that name exactly), but then decided against it as unnecessary complication. As software/memory breakpoints work by poking at memory, when setting a GDB Z0 breakpoint (but not internal breakpoints, as those can assume the conditions are already right), we need to tell the target to prepare to access memory (which on Linux means stop threads). If that operation fails, we need to return error to GDB. Seeing an error, if this is the first breakpoint of that type that GDB tries to insert, GDB would then assume the breakpoint type is supported, but it may actually not be. So we need to check whether the type is supported at all before preparing to access memory. And to solve that, the patch adds a new target->supports_z_point_type method that is called before actually trying to insert the breakpoint. Other than that, hopefully the change is more or less obvious. New test added that exercises the hbreak2.exp regression in a more direct way, without relying on a breakpoint re-set happening before main is reached. Tested by building GDBserver for: aarch64-linux-gnu arm-linux-gnueabihf i686-pc-linux-gnu i686-w64-mingw32 m68k-linux-gnu mips-linux-gnu mips-uclinux nios2-linux-gnu powerpc-linux-gnu sh-linux-gnu tilegx-unknown-linux-gnu x86_64-redhat-linux x86_64-w64-mingw32 And also regression tested on x86_64 Fedora 20. gdb/gdbserver/ 2014-05-20 Pedro Alves <palves@redhat.com> * linux-aarch64-low.c (aarch64_insert_point) (aarch64_remove_point): No longer check whether the type is supported here. Adjust to new interface. (the_low_target): Install aarch64_supports_z_point_type as supports_z_point_type method. * linux-arm-low.c (raw_bkpt_type_to_arm_hwbp_type): New function. (arm_linux_hw_point_initialize): Take an enum raw_bkpt_type instead of a Z packet char. Adjust. (arm_supports_z_point_type): New function. (arm_insert_point, arm_remove_point): Adjust to new interface. (the_low_target): Install arm_supports_z_point_type. * linux-crisv32-low.c (cris_supports_z_point_type): New function. (cris_insert_point, cris_remove_point): Adjust to new interface. Don't check whether the type is supported here. (the_low_target): Install cris_supports_z_point_type. * linux-low.c (linux_supports_z_point_type): New function. (linux_insert_point, linux_remove_point): Adjust to new interface. * linux-low.h (struct linux_target_ops) <insert_point, remove_point>: Take an enum raw_bkpt_type instead of a char. Add raw_breakpoint pointer parameter. <supports_z_point_type>: New method. * linux-mips-low.c (mips_supports_z_point_type): New function. (mips_insert_point, mips_remove_point): Adjust to new interface. Use mips_supports_z_point_type. (the_low_target): Install mips_supports_z_point_type. * linux-ppc-low.c (the_low_target): Install NULL as supports_z_point_type method. * linux-s390-low.c (the_low_target): Install NULL as supports_z_point_type method. * linux-sparc-low.c (the_low_target): Install NULL as supports_z_point_type method. * linux-x86-low.c (x86_supports_z_point_type): New function. (x86_insert_point): Adjust to new insert_point interface. Use insert_memory_breakpoint. Adjust to new i386_low_insert_watchpoint interface. (x86_remove_point): Adjust to remove_point interface. Use remove_memory_breakpoint. Adjust to new i386_low_remove_watchpoint interface. (the_low_target): Install x86_supports_z_point_type. * lynx-low.c (lynx_target_ops): Install NULL as supports_z_point_type callback. * nto-low.c (nto_supports_z_point_type): New. (nto_insert_point, nto_remove_point): Adjust to new interface. (nto_target_ops): Install nto_supports_z_point_type. * mem-break.c: Adjust intro comment. (struct raw_breakpoint) <raw_type, size>: New fields. <inserted>: Update comment. <shlib_disabled>: Delete field. (enum bkpt_type) <gdb_breakpoint>: Delete value. <gdb_breakpoint_Z0, gdb_breakpoint_Z1, gdb_breakpoint_Z2, gdb_breakpoint_Z3, gdb_breakpoint_Z4>: New values. (raw_bkpt_type_to_target_hw_bp_type): New function. (find_enabled_raw_code_breakpoint_at): New function. (find_raw_breakpoint_at): New type and size parameters. Use them. (insert_memory_breakpoint): New function, based off set_raw_breakpoint_at. (remove_memory_breakpoint): New function. (set_raw_breakpoint_at): Reimplement. (set_breakpoint): New, based on set_breakpoint_at. (set_breakpoint_at): Reimplement. (delete_raw_breakpoint): Go through the_target->remove_point instead of assuming memory breakpoints. (find_gdb_breakpoint_at): Delete. (Z_packet_to_bkpt_type, Z_packet_to_raw_bkpt_type): New functions. (find_gdb_breakpoint): New function. (set_gdb_breakpoint_at): Delete. (z_type_supported): New function. (set_gdb_breakpoint_1): New function, loosely based off set_gdb_breakpoint_at. (check_gdb_bp_preconditions, set_gdb_breakpoint): New functions. (delete_gdb_breakpoint_at): Delete. (delete_gdb_breakpoint_1): New function, loosely based off delete_gdb_breakpoint_at. (delete_gdb_breakpoint): New function. (clear_gdb_breakpoint_conditions): Rename to ... (clear_breakpoint_conditions): ... this. Don't handle a NULL breakpoint. (add_condition_to_breakpoint): Make static. (add_breakpoint_condition): Take a struct breakpoint pointer instead of an address. Adjust. (gdb_condition_true_at_breakpoint): Rename to ... (gdb_condition_true_at_breakpoint_z_type): ... this, and add z_type parameter. (gdb_condition_true_at_breakpoint): Reimplement. (add_breakpoint_commands): Take a struct breakpoint pointer instead of an address. Adjust. (gdb_no_commands_at_breakpoint): Rename to ... (gdb_no_commands_at_breakpoint_z_type): ... this. Add z_type parameter. Return true if no breakpoint was found. Change debug output. (gdb_no_commands_at_breakpoint): Reimplement. (run_breakpoint_commands): Rename to ... (run_breakpoint_commands_z_type): ... this. Add z_type parameter, and change return type to boolean. (run_breakpoint_commands): New function. (gdb_breakpoint_here): Also check for Z1 breakpoints. (uninsert_raw_breakpoint): Don't try to reinsert a disabled breakpoint. Go through the_target->remove_point instead of assuming memory breakpoint. (uninsert_breakpoints_at, uninsert_all_breakpoints): Uninsert software and hardware breakpoints. (reinsert_raw_breakpoint): Go through the_target->insert_point instead of assuming memory breakpoint. (reinsert_breakpoints_at, reinsert_all_breakpoints): Reinsert software and hardware breakpoints. (check_breakpoints, breakpoint_here, breakpoint_inserted_here): Check both software and hardware breakpoints. (validate_inserted_breakpoint): Assert the breakpoint is a software breakpoint. Set the inserted flag to -1 instead of setting shlib_disabled. (delete_disabled_breakpoints): Adjust. (validate_breakpoints): Only validate software breakpoints. Adjust to inserted flag change. (check_mem_read, check_mem_write): Skip breakpoint types other than software breakpoints. Adjust to inserted flag change. * mem-break.h (enum raw_bkpt_type): New enum. (raw_breakpoint, struct process_info): Forward declare. (Z_packet_to_target_hw_bp_type): Delete declaration. (raw_bkpt_type_to_target_hw_bp_type, Z_packet_to_raw_bkpt_type) (set_gdb_breakpoint, delete_gdb_breakpoint) (clear_breakpoint_conditions): New declarations. (set_gdb_breakpoint_at, clear_gdb_breakpoint_conditions): Delete. (breakpoint_inserted_here): Update comment. (add_breakpoint_condition, add_breakpoint_commands): Replace address parameter with a breakpoint pointer parameter. (gdb_breakpoint_here): Update comment. (delete_gdb_breakpoint_at): Delete. (insert_memory_breakpoint, remove_memory_breakpoint): Declare. * server.c (process_point_options): Take a struct breakpoint pointer instead of an address. Adjust. (process_serial_event) <Z/z packets>: Use set_gdb_breakpoint and delete_gdb_breakpoint. * spu-low.c (spu_target_ops): Install NULL as supports_z_point_type method. * target.h: Include mem-break.h. (struct target_ops) <prepare_to_access_memory>: Update comment. <supports_z_point_type>: New field. <insert_point, remove_point>: Take an enum raw_bkpt_type argument instead of a char. Also take a raw breakpoint pointer. * win32-arm-low.c (the_low_target): Install NULL as supports_z_point_type. * win32-i386-low.c (i386_supports_z_point_type): New function. (i386_insert_point, i386_remove_point): Adjust to new interface. (the_low_target): Install i386_supports_z_point_type. * win32-low.c (win32_supports_z_point_type): New function. (win32_insert_point, win32_remove_point): Adjust to new interface. (win32_target_ops): Install win32_supports_z_point_type. * win32-low.h (struct win32_target_ops): <supports_z_point_type>: New method. <insert_point, remove_point>: Take an enum raw_bkpt_type argument instead of a char. Also take a raw breakpoint pointer. gdb/testsuite/ 2014-05-20 Pedro Alves <palves@redhat.com> * gdb.base/break-idempotent.c: New file. * gdb.base/break-idempotent.exp: New file.
2014-05-20 19:24:28 +02:00
int add_breakpoint_commands (struct breakpoint *bp, char **commands,
int persist);
Add target-side support for dynamic printf. * NEWS: Mention the additional style. * breakpoint.h (struct bp_target_info): New fields tcommands, persist. (struct bp_location): New field cmd_bytecode. * breakpoint.c: Include format.h. (disconnected_dprintf): New global. (parse_cmd_to_aexpr): New function. (build_target_command_list): New function. (insert_bp_location): Call it. (remove_breakpoints_pid): Skip dprintf breakpoints. (print_one_breakpoint_location): Ditto. (dprintf_style_agent): New global. (dprintf_style_enums): Add dprintf_style_agent. (update_dprintf_command_list): Add agent case. (agent_printf_command): New function. (_initialize_breakpoint): Add new commands. * common/ax.def (printf): New bytecode. * ax.h (ax_string): Declare. * ax-gdb.h (gen_printf): Declare. * ax-gdb.c: Include cli-utils.h, format.h. (gen_printf): New function. (maint_agent_print_command): New function. (_initialize_ax_gdb): Add maint agent-printf command. * ax-general.c (ax_string): New function. (ax_print): Add printf disassembly. * Makefile.in (SFILES): Add format.c (COMMON_OBS): Add format.o. * common/format.h: New file. * common/format.c: New file. * printcmd.c: Include format.h. (ui_printf): Call parse_format_string. * remote.c (remote_state): New field breakpoint_commands. (PACKET_BreakpointCommands): New enum. (remote_breakpoint_commands_feature): New function. (remote_protocol_features): Add new BreakpointCommands entry. (remote_can_run_breakpoint_commands): New function. (remote_add_target_side_commands): New function. (remote_insert_breakpoint): Call it. (remote_insert_hw_breakpoint): Ditto. (_initialize_remote): Add new packet configuration for target-side breakpoint commands. * target.h (struct target_ops): New field to_can_run_breakpoint_commands. (target_can_run_breakpoint_commands): New macro. * target.c (update_current_target): Handle to_can_run_breakpoint_commands. [gdbserver] * Makefile.in (WARN_CFLAGS_NO_FORMAT): Define. (ax.o): Add it to build rule. (ax-ipa.o): Ditto. (OBS): Add format.o. (IPA_OBS): Add format.o. * server.c (handle_query): Claim support for breakpoint commands. (process_point_options): Add command case. (process_serial_event): Leave running if there are printfs in effect. * mem-break.h (any_persistent_commands): Declare. (add_breakpoint_commands): Declare. (gdb_no_commands_at_breakpoint): Declare. (run_breakpoint_commands): Declare. * mem-break.c (struct point_command_list): New struct. (struct breakpoint): New field command_list. (any_persistent_commands): New function. (add_commands_to_breakpoint): New function. (add_breakpoint_commands): New function. (gdb_no_commands_at_breakpoint): New function. (run_breakpoint_commands): New function. * linux-low.c (linux_wait_1): Test for and run breakpoint commands locally. * ax.c: Include format.h. (ax_printf): New function. (gdb_eval_agent_expr): Add printf opcode. [doc] * gdb.texinfo (Dynamic Printf): Mention agent style and disconnected dprintf. (Maintenance Commands): Describe maint agent-printf. (General Query Packets): Mention BreakpointCommands feature. (Packets): Document commands extension to Z0 packet. * agentexpr.texi (Bytecode Descriptions): Document printf bytecode. [testsuite] * gdb.base/dprintf.exp: Add agent style tests.
2012-07-02 17:29:39 +02:00
int any_persistent_commands (void);
/* Evaluation condition (if any) at breakpoint BP. Return 1 if
true and 0 otherwise. */
int gdb_condition_true_at_breakpoint (CORE_ADDR where);
Add target-side support for dynamic printf. * NEWS: Mention the additional style. * breakpoint.h (struct bp_target_info): New fields tcommands, persist. (struct bp_location): New field cmd_bytecode. * breakpoint.c: Include format.h. (disconnected_dprintf): New global. (parse_cmd_to_aexpr): New function. (build_target_command_list): New function. (insert_bp_location): Call it. (remove_breakpoints_pid): Skip dprintf breakpoints. (print_one_breakpoint_location): Ditto. (dprintf_style_agent): New global. (dprintf_style_enums): Add dprintf_style_agent. (update_dprintf_command_list): Add agent case. (agent_printf_command): New function. (_initialize_breakpoint): Add new commands. * common/ax.def (printf): New bytecode. * ax.h (ax_string): Declare. * ax-gdb.h (gen_printf): Declare. * ax-gdb.c: Include cli-utils.h, format.h. (gen_printf): New function. (maint_agent_print_command): New function. (_initialize_ax_gdb): Add maint agent-printf command. * ax-general.c (ax_string): New function. (ax_print): Add printf disassembly. * Makefile.in (SFILES): Add format.c (COMMON_OBS): Add format.o. * common/format.h: New file. * common/format.c: New file. * printcmd.c: Include format.h. (ui_printf): Call parse_format_string. * remote.c (remote_state): New field breakpoint_commands. (PACKET_BreakpointCommands): New enum. (remote_breakpoint_commands_feature): New function. (remote_protocol_features): Add new BreakpointCommands entry. (remote_can_run_breakpoint_commands): New function. (remote_add_target_side_commands): New function. (remote_insert_breakpoint): Call it. (remote_insert_hw_breakpoint): Ditto. (_initialize_remote): Add new packet configuration for target-side breakpoint commands. * target.h (struct target_ops): New field to_can_run_breakpoint_commands. (target_can_run_breakpoint_commands): New macro. * target.c (update_current_target): Handle to_can_run_breakpoint_commands. [gdbserver] * Makefile.in (WARN_CFLAGS_NO_FORMAT): Define. (ax.o): Add it to build rule. (ax-ipa.o): Ditto. (OBS): Add format.o. (IPA_OBS): Add format.o. * server.c (handle_query): Claim support for breakpoint commands. (process_point_options): Add command case. (process_serial_event): Leave running if there are printfs in effect. * mem-break.h (any_persistent_commands): Declare. (add_breakpoint_commands): Declare. (gdb_no_commands_at_breakpoint): Declare. (run_breakpoint_commands): Declare. * mem-break.c (struct point_command_list): New struct. (struct breakpoint): New field command_list. (any_persistent_commands): New function. (add_commands_to_breakpoint): New function. (add_breakpoint_commands): New function. (gdb_no_commands_at_breakpoint): New function. (run_breakpoint_commands): New function. * linux-low.c (linux_wait_1): Test for and run breakpoint commands locally. * ax.c: Include format.h. (ax_printf): New function. (gdb_eval_agent_expr): Add printf opcode. [doc] * gdb.texinfo (Dynamic Printf): Mention agent style and disconnected dprintf. (Maintenance Commands): Describe maint agent-printf. (General Query Packets): Mention BreakpointCommands feature. (Packets): Document commands extension to Z0 packet. * agentexpr.texi (Bytecode Descriptions): Document printf bytecode. [testsuite] * gdb.base/dprintf.exp: Add agent style tests.
2012-07-02 17:29:39 +02:00
int gdb_no_commands_at_breakpoint (CORE_ADDR where);
void run_breakpoint_commands (CORE_ADDR where);
[GDBserver] Make Zx/zx packet handling idempotent. This patch fixes hardware breakpoint regressions exposed by my fix for "PR breakpoints/7143 - Watchpoint does not trigger when first set", at https://sourceware.org/ml/gdb-patches/2014-03/msg00167.html The testsuite caught them on Linux/x86_64, at least. gdb.sum: gdb.sum: FAIL: gdb.base/hbreak2.exp: next over recursive call FAIL: gdb.base/hbreak2.exp: backtrace from factorial(5.1) FAIL: gdb.base/hbreak2.exp: continue until exit at recursive next test gdb.log: (gdb) next Program received signal SIGTRAP, Trace/breakpoint trap. factorial (value=4) at ../../../src/gdb/testsuite/gdb.base/break.c:113 113 if (value > 1) { /* set breakpoint 7 here */ (gdb) FAIL: gdb.base/hbreak2.exp: next over recursive call Actually, that patch just exposed a latent issue to "breakpoints always-inserted off" mode, not really caused it. After that patch, GDB no longer removes breakpoints at each internal event, thus making some scenarios behave like breakpoint always-inserted on. The bug is easy to trigger with always-inserted on. The issue is that since the target-side breakpoint conditions support, if the stub/server supports evaluating breakpoint conditions on the target side, then GDB is sending duplicate Zx packets to the target without removing them before, and GDBserver is not really expecting that for Z packets other than Z0/z0. E.g., with "set breakpoint always-inserted on" and "set debug remote 1": (gdb) b main Sending packet: $m410943,1#ff...Packet received: 48 Breakpoint 4 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z0,410943,1#48...Packet received: OK ^^^^^^^^^^^^ (gdb) b main Note: breakpoint 4 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Breakpoint 5 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z0,410943,1#48...Packet received: OK ^^^^^^^^^^^^ (gdb) b main Note: breakpoints 4 and 5 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Breakpoint 6 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z0,410943,1#48...Packet received: OK ^^^^^^^^^^^^ (gdb) del Delete all breakpoints? (y or n) y Sending packet: $Z0,410943,1#48...Packet received: OK Sending packet: $Z0,410943,1#48...Packet received: OK Sending packet: $z0,410943,1#68...Packet received: OK And for Z1, similarly: (gdb) hbreak main Sending packet: $m410943,1#ff...Packet received: 48 Hardware assisted breakpoint 4 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ Packet Z1 (hardware-breakpoint) is supported (gdb) hbreak main Note: breakpoint 4 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Hardware assisted breakpoint 5 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ (gdb) hbreak main Note: breakpoints 4 and 5 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Hardware assisted breakpoint 6 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ (gdb) del Delete all breakpoints? (y or n) y Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ Sending packet: $z1,410943,1#69...Packet received: OK ^^^^^^^^^^^^ So GDB sent a bunch of Z1 packets, and then when finally removing the breakpoint, only one z1 packet was sent. On the GDBserver side (with monitor set debug-hw-points 1), in the Z1 case, we see: $ ./gdbserver :9999 ./gdbserver Process ./gdbserver created; pid = 8629 Listening on port 9999 Remote debugging from host 127.0.0.1 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=1 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=2 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=3 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=4 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=5 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 remove_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=4 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 That's one insert_watchpoint call for each Z1 packet, and then one remove_watchpoint call for the z1 packet. Notice how ref.count increased for each insert_watchpoint call, and then in the end, after GDB told GDBserver to forget about the hardware breakpoint, GDBserver ends with the the first debug register still with ref.count=4! IOW, the hardware breakpoint is left armed on the target, while on the GDB end it's gone. If the program happens to execute 0x410943 afterwards, then the CPU traps, GDBserver reports the trap to GDB, and GDB not having a breakpoint set at that address anymore, reports to the user a spurious SIGTRAP. This is exactly what is happening in the hbreak2.exp test, though in that case, it's a shared library event that triggers a breakpoint_re_set, when breakpoints are still inserted (because nowadays GDB doesn't remove breakpoints while handling internal events), and that recreates breakpoint locations, which likewise forces breakpoint reinsertion and Zx packet resends... That is a lot of bogus Zx duplication that should possibly be addressed on the GDB side. GDB resends Zx packets because the way to change the target-side condition, is to resend the breakpoint to the server with the new condition. (That's an option in the packet: e.g., "Z1,410943,1;X3,220027" for "hbreak main if 0". The packets in the examples above are shorter because the breakpoints don't have conditions attached). GDB doesn't remove the breakpoint first before reinserting it because that'd be bad for non-stop, as it'd open a window where the inferior could miss the breakpoint. The conditions actually haven't changed between the resends, but GDB isn't smart enough to realize that. (TBC, if the target doesn't support target-side conditions, then GDB doesn't trigger these resends (init_bp_location calls mark_breakpoint_location_modified, and that does nothing if condition evaluation is on the host side. The resends are caused by the 'loc->condition_changed = condition_modified.' line.) But, even if GDB was made smarter, GDBserver should really still handle the resends anyway. So target-side conditions also aren't really to blame. The documentation of the Z/z packets says: "To avoid potential problems with duplicate packets, the operations should be implemented in an idempotent way." As such, we may want to fix GDB, but we should definitely fix GDBserver. The fix is a prerequisite for target-side conditions on hardware breakpoints anyway (and while at it, on watchpoints too). GDBserver indeed already treats duplicate Z0 packets in an idempotent way. mem-break.c has the concept of high-level and low-level breakpoints, somewhat similar to GDB's split of breakpoints vs breakpoint locations, and keeps track of multiple breakpoints referencing the same address/location, for the case of an internal GDBserver breakpoint or a tracepoint being set at the same address as a GDB breakpoint. But, it only allows GDB to ever contribute one reference to a software breakpoint location. IOW, if gdbserver sees a Z0 packet for the same address where it already had a GDB breakpoint set, then GDBserver won't create another high-level GDB breakpoint. However, mem-break.c only tracks GDB Z0 breakpoints. The same logic should apply to all kinds of Zx packets. Currently, gdbserver passes down each duplicate Zx (other than Z0) request directly to the target->insert_point routine. The x86 watchpoint support itself refcounts watchpoint / hw breakpoint requests, to handle overlapping watchpoints, and save debug registers. But that code doesn't (and really shouldn't) handle the duplicate requests, assuming that for each insert there will be a corresponding remove. So the fix is to generalize mem-break.c to track all kinds of Zx breakpoints, and filter out duplicates. As mentioned, this ends up adding support for target-side conditions on hardware breakpoints and watchpoints too (though GDB itself doesn't support the latter yet). Probably the least obvious change in the patch is that it kind of turns the breakpoint insert/remove APIs inside out. Before, the target methods were only called for GDB breakpoints. The internal breakpoint set/delete methods inserted memory breakpoints directly bypassing the insert/remove target methods. That's not good when the target should use a debug API to set software breakpoints, instead of relying on GDBserver patching memory with breakpoint instructions, as is the case of NTO. Now removal/insertion of all kinds of breakpoints/watchpoints, either internal, or from GDB, always go through the target methods. The insert_point/remove_point methods no longer get passed a Z packet type, but an internal/raw breakpoint type. They're also passed a pointer to the raw breakpoint itself (note that's still opaque outside mem-break.c), so that insert_memory_breakpoint / remove_memory_breakpoint have access to the breakpoint's shadow buffer. I first tried passing down a new structure based on GDB's "struct bp_target_info" (actually with that name exactly), but then decided against it as unnecessary complication. As software/memory breakpoints work by poking at memory, when setting a GDB Z0 breakpoint (but not internal breakpoints, as those can assume the conditions are already right), we need to tell the target to prepare to access memory (which on Linux means stop threads). If that operation fails, we need to return error to GDB. Seeing an error, if this is the first breakpoint of that type that GDB tries to insert, GDB would then assume the breakpoint type is supported, but it may actually not be. So we need to check whether the type is supported at all before preparing to access memory. And to solve that, the patch adds a new target->supports_z_point_type method that is called before actually trying to insert the breakpoint. Other than that, hopefully the change is more or less obvious. New test added that exercises the hbreak2.exp regression in a more direct way, without relying on a breakpoint re-set happening before main is reached. Tested by building GDBserver for: aarch64-linux-gnu arm-linux-gnueabihf i686-pc-linux-gnu i686-w64-mingw32 m68k-linux-gnu mips-linux-gnu mips-uclinux nios2-linux-gnu powerpc-linux-gnu sh-linux-gnu tilegx-unknown-linux-gnu x86_64-redhat-linux x86_64-w64-mingw32 And also regression tested on x86_64 Fedora 20. gdb/gdbserver/ 2014-05-20 Pedro Alves <palves@redhat.com> * linux-aarch64-low.c (aarch64_insert_point) (aarch64_remove_point): No longer check whether the type is supported here. Adjust to new interface. (the_low_target): Install aarch64_supports_z_point_type as supports_z_point_type method. * linux-arm-low.c (raw_bkpt_type_to_arm_hwbp_type): New function. (arm_linux_hw_point_initialize): Take an enum raw_bkpt_type instead of a Z packet char. Adjust. (arm_supports_z_point_type): New function. (arm_insert_point, arm_remove_point): Adjust to new interface. (the_low_target): Install arm_supports_z_point_type. * linux-crisv32-low.c (cris_supports_z_point_type): New function. (cris_insert_point, cris_remove_point): Adjust to new interface. Don't check whether the type is supported here. (the_low_target): Install cris_supports_z_point_type. * linux-low.c (linux_supports_z_point_type): New function. (linux_insert_point, linux_remove_point): Adjust to new interface. * linux-low.h (struct linux_target_ops) <insert_point, remove_point>: Take an enum raw_bkpt_type instead of a char. Add raw_breakpoint pointer parameter. <supports_z_point_type>: New method. * linux-mips-low.c (mips_supports_z_point_type): New function. (mips_insert_point, mips_remove_point): Adjust to new interface. Use mips_supports_z_point_type. (the_low_target): Install mips_supports_z_point_type. * linux-ppc-low.c (the_low_target): Install NULL as supports_z_point_type method. * linux-s390-low.c (the_low_target): Install NULL as supports_z_point_type method. * linux-sparc-low.c (the_low_target): Install NULL as supports_z_point_type method. * linux-x86-low.c (x86_supports_z_point_type): New function. (x86_insert_point): Adjust to new insert_point interface. Use insert_memory_breakpoint. Adjust to new i386_low_insert_watchpoint interface. (x86_remove_point): Adjust to remove_point interface. Use remove_memory_breakpoint. Adjust to new i386_low_remove_watchpoint interface. (the_low_target): Install x86_supports_z_point_type. * lynx-low.c (lynx_target_ops): Install NULL as supports_z_point_type callback. * nto-low.c (nto_supports_z_point_type): New. (nto_insert_point, nto_remove_point): Adjust to new interface. (nto_target_ops): Install nto_supports_z_point_type. * mem-break.c: Adjust intro comment. (struct raw_breakpoint) <raw_type, size>: New fields. <inserted>: Update comment. <shlib_disabled>: Delete field. (enum bkpt_type) <gdb_breakpoint>: Delete value. <gdb_breakpoint_Z0, gdb_breakpoint_Z1, gdb_breakpoint_Z2, gdb_breakpoint_Z3, gdb_breakpoint_Z4>: New values. (raw_bkpt_type_to_target_hw_bp_type): New function. (find_enabled_raw_code_breakpoint_at): New function. (find_raw_breakpoint_at): New type and size parameters. Use them. (insert_memory_breakpoint): New function, based off set_raw_breakpoint_at. (remove_memory_breakpoint): New function. (set_raw_breakpoint_at): Reimplement. (set_breakpoint): New, based on set_breakpoint_at. (set_breakpoint_at): Reimplement. (delete_raw_breakpoint): Go through the_target->remove_point instead of assuming memory breakpoints. (find_gdb_breakpoint_at): Delete. (Z_packet_to_bkpt_type, Z_packet_to_raw_bkpt_type): New functions. (find_gdb_breakpoint): New function. (set_gdb_breakpoint_at): Delete. (z_type_supported): New function. (set_gdb_breakpoint_1): New function, loosely based off set_gdb_breakpoint_at. (check_gdb_bp_preconditions, set_gdb_breakpoint): New functions. (delete_gdb_breakpoint_at): Delete. (delete_gdb_breakpoint_1): New function, loosely based off delete_gdb_breakpoint_at. (delete_gdb_breakpoint): New function. (clear_gdb_breakpoint_conditions): Rename to ... (clear_breakpoint_conditions): ... this. Don't handle a NULL breakpoint. (add_condition_to_breakpoint): Make static. (add_breakpoint_condition): Take a struct breakpoint pointer instead of an address. Adjust. (gdb_condition_true_at_breakpoint): Rename to ... (gdb_condition_true_at_breakpoint_z_type): ... this, and add z_type parameter. (gdb_condition_true_at_breakpoint): Reimplement. (add_breakpoint_commands): Take a struct breakpoint pointer instead of an address. Adjust. (gdb_no_commands_at_breakpoint): Rename to ... (gdb_no_commands_at_breakpoint_z_type): ... this. Add z_type parameter. Return true if no breakpoint was found. Change debug output. (gdb_no_commands_at_breakpoint): Reimplement. (run_breakpoint_commands): Rename to ... (run_breakpoint_commands_z_type): ... this. Add z_type parameter, and change return type to boolean. (run_breakpoint_commands): New function. (gdb_breakpoint_here): Also check for Z1 breakpoints. (uninsert_raw_breakpoint): Don't try to reinsert a disabled breakpoint. Go through the_target->remove_point instead of assuming memory breakpoint. (uninsert_breakpoints_at, uninsert_all_breakpoints): Uninsert software and hardware breakpoints. (reinsert_raw_breakpoint): Go through the_target->insert_point instead of assuming memory breakpoint. (reinsert_breakpoints_at, reinsert_all_breakpoints): Reinsert software and hardware breakpoints. (check_breakpoints, breakpoint_here, breakpoint_inserted_here): Check both software and hardware breakpoints. (validate_inserted_breakpoint): Assert the breakpoint is a software breakpoint. Set the inserted flag to -1 instead of setting shlib_disabled. (delete_disabled_breakpoints): Adjust. (validate_breakpoints): Only validate software breakpoints. Adjust to inserted flag change. (check_mem_read, check_mem_write): Skip breakpoint types other than software breakpoints. Adjust to inserted flag change. * mem-break.h (enum raw_bkpt_type): New enum. (raw_breakpoint, struct process_info): Forward declare. (Z_packet_to_target_hw_bp_type): Delete declaration. (raw_bkpt_type_to_target_hw_bp_type, Z_packet_to_raw_bkpt_type) (set_gdb_breakpoint, delete_gdb_breakpoint) (clear_breakpoint_conditions): New declarations. (set_gdb_breakpoint_at, clear_gdb_breakpoint_conditions): Delete. (breakpoint_inserted_here): Update comment. (add_breakpoint_condition, add_breakpoint_commands): Replace address parameter with a breakpoint pointer parameter. (gdb_breakpoint_here): Update comment. (delete_gdb_breakpoint_at): Delete. (insert_memory_breakpoint, remove_memory_breakpoint): Declare. * server.c (process_point_options): Take a struct breakpoint pointer instead of an address. Adjust. (process_serial_event) <Z/z packets>: Use set_gdb_breakpoint and delete_gdb_breakpoint. * spu-low.c (spu_target_ops): Install NULL as supports_z_point_type method. * target.h: Include mem-break.h. (struct target_ops) <prepare_to_access_memory>: Update comment. <supports_z_point_type>: New field. <insert_point, remove_point>: Take an enum raw_bkpt_type argument instead of a char. Also take a raw breakpoint pointer. * win32-arm-low.c (the_low_target): Install NULL as supports_z_point_type. * win32-i386-low.c (i386_supports_z_point_type): New function. (i386_insert_point, i386_remove_point): Adjust to new interface. (the_low_target): Install i386_supports_z_point_type. * win32-low.c (win32_supports_z_point_type): New function. (win32_insert_point, win32_remove_point): Adjust to new interface. (win32_target_ops): Install win32_supports_z_point_type. * win32-low.h (struct win32_target_ops): <supports_z_point_type>: New method. <insert_point, remove_point>: Take an enum raw_bkpt_type argument instead of a char. Also take a raw breakpoint pointer. gdb/testsuite/ 2014-05-20 Pedro Alves <palves@redhat.com> * gdb.base/break-idempotent.c: New file. * gdb.base/break-idempotent.exp: New file.
2014-05-20 19:24:28 +02:00
/* Returns TRUE if there's a GDB breakpoint (Z0 or Z1) set at
WHERE. */
* linux-low.c (linux_wait_1): Avoid setting need_step_over is there's a GDB breakpoint at stop_pc. Always report a trap to GDB if we could tell there's a GDB breakpoint at stop_pc. (need_step_over_p): Don't do a step over if we find a GDB breakpoint at the resume PC. * mem-break.c (struct raw_breakpoint): New. (enum bkpt_type): New type `gdb_breakpoint'. (struct breakpoint): Delete the `PC', `old_data' and `inserted' fields. New field `raw'. (find_raw_breakpoint_at): New. (set_raw_breakpoint_at): Handle refcounting. Create a raw breakpoint instead. (set_breakpoint_at): Adjust. (delete_raw_breakpoint): New. (release_breakpoint): New. (delete_breakpoint): Rename to... (delete_breakpoint_1): ... this. Add proc parameter. Use release_breakpoint. Return ENOENT. (delete_breakpoint): Reimplement. (find_breakpoint_at): Delete. (find_gdb_breakpoint_at): New. (delete_breakpoint_at): Delete. (set_gdb_breakpoint_at): New. (delete_gdb_breakpoint_at): New. (gdb_breakpoint_here): New. (set_reinsert_breakpoint): Use release_breakpoint. (uninsert_breakpoint): Rename to ... (uninsert_raw_breakpoint): ... this. (uninsert_breakpoints_at): Adjust to handle raw breakpoints. (reinsert_raw_breakpoint): Change parameter type to raw_breakpoint. (reinsert_breakpoints_at): Adjust to handle raw breakpoints instead. (check_breakpoints): Adjust. Use release_breakpoint. (breakpoint_here): Rewrite using find_raw_breakpoint_at. (breakpoint_inserted_here): Ditto. (check_mem_read): Adjust to iterate over raw breakpoints instead. Don't trust the breakpoint's shadow if it is not inserted. (check_mem_write): Adjust to iterate over raw breakpoints instead. (delete_all_breakpoints): Adjust. (free_all_breakpoints): Mark all breakpoints as uninserted, and use delete_breakpoint_1. * mem-break.h (breakpoints_supported): Delete declaration. (set_gdb_breakpoint_at): Declare. (gdb_breakpoint_here): Declare. (delete_breakpoint_at): Delete. (delete_gdb_breakpoint_at): Declare. * server.h (struct raw_breakpoint): Forward declare. (struct process_info): New field `raw_breakpoints'. * linux-x86-low.c (x86_insert_point, x86_remote_point): Handle Z0 breakpoints.
2010-04-01 16:25:34 +02:00
int gdb_breakpoint_here (CORE_ADDR where);
/* Create a new breakpoint at WHERE, and call HANDLER when
it is hit. HANDLER should return 1 if the breakpoint
should be deleted, 0 otherwise. */
struct breakpoint *set_breakpoint_at (CORE_ADDR where,
int (*handler) (CORE_ADDR));
gdb/gdbserver/ * server.h (LONGEST): New. (struct thread_info) <while_stepping>: New field. (unpack_varlen_hex, xrealloc, pulongest, plongest, phex_nz): Declare. (initialize_tracepoint, handle_tracepoint_general_set) (handle_tracepoint_query, tracepoint_finished_step) (tracepoint_was_hit, release_while_stepping_state_list): (current_traceframe): Declare. * server.c (handle_general_set): Handle tracepoint packets. (read_memory): New. (write_memory): New. (handle_search_memory_1): Use read_memory. (handle_query): Report support for conditional tracepoints, trace state variables, and tracepoint sources. Handle tracepoint queries. (main): Initialize the tracepoints module. (process_serial_event): Handle traceframe reads/writes. * linux-low.c (handle_tracepoints): New. (linux_wait_1): Call it. (linux_resume_one_lwp): Handle while-stepping. (linux_supports_tracepoints, linux_read_pc, linux_write_pc): New. (linux_target_ops): Install them. * linux-low.h (struct linux_target_ops) <supports_tracepoints>: New field. * linux-x86-low.c (x86_supports_tracepoints): New. (the_low_target). Install it. * mem-break.h (delete_breakpoint): Declare. * mem-break.c (delete_breakpoint): Make external. * target.h (struct target_ops): Add `supports_tracepoints', `read_pc', and `write_pc' fields. (target_supports_tracepoints): Define. * utils.c (xrealloc, decimal2str, pulongest, plongest, thirty_two) (phex_nz): New. * regcache.h (struct regcache) <registers_owned>: New field. (init_register_cache, regcache_cpy): Declare. (regcache_read_pc, regcache_write_pc): Declare. (register_cache_size): Declare. (supply_regblock): Declare. * regcache.c (init_register_cache): New. (new_register_cache): Use it. (regcache_cpy): New. (register_cache_size): New. (supply_regblock): New. (regcache_read_pc, regcache_write_pc): New. * tracepoint.c: New. * Makefile.in (OBS): Add tracepoint.o. (tracepoint.o): New rule. gdb/ * regformats/regdat.sh: Include server.h. Don't include regcache.h.
2010-04-09 05:40:00 +02:00
/* Delete a breakpoint. */
int delete_breakpoint (struct breakpoint *bkpt);
Teach linux gdbserver to step-over-breakpoints. * linux-low.c (can_hardware_single_step): New. (supports_breakpoints): New. (handle_extended_wait): If stopping threads, read the stop pc of the new cloned LWP. (get_pc): New. (get_stop_pc): Add `lwp' parameter. Handle it. Bail out if the low target doesn't support retrieving the PC. (add_lwp): Set last_resume_kind to resume_continue. (linux_attach_lwp_1): Adjust comments. Always set stop_expected. (linux_attach): Don't clear stop_expected. Set the lwp's last_resume_kind to resume_stop. (linux_detach_one_lwp): Don't check for removed breakpoints. (check_removed_breakpoint): Delete. (status_pending_p): Rename to ... (status_pending_p_callback): ... this. Don't check for removed breakpoints. Don't consider threads that are stopped from GDB's perspective. (linux_wait_for_lwp): Always read the stop_pc here. (cancel_breakpoint): New. (step_over_bkpt): New global. (linux_wait_for_event_1): Implement stepping over breakpoints. (gdb_wants_lwp_stopped): New. (gdb_wants_all_stopped): New. (linux_wait_1): Tag threads as gdb-wants-stopped. Cancel finished single-step traps here. Store the thread's last reported target wait status. (send_sigstop): Don't clear stop_expected. Always set it, instead. (mark_lwp_dead): Remove reference to pending_is_breakpoint. (cancel_finished_single_step): New. (cancel_finished_single_steps): New. (wait_for_sigstop): Don't cancel finished single-step traps here. (linux_resume_one_lwp): Don't check for removed breakpoints. Don't set `step' on non-hardware step archs. (linux_set_resume_request): Ignore resume_stop requests if already stopping or stopped. Set the lwp's last_resume_kind. (resume_status_pending_p): Don't check for removed breakpoints. (need_step_over_p): New. (start_step_over): New. (finish_step_over): New. (linux_resume_one_thread): Always queue a sigstop for resume_stop requests. Clear the thread's last reported target waitstatus. Don't use the `suspended' flag. Don't consider pending breakpoints. (linux_resume): Start a step-over if necessary. (proceed_one_lwp): New. (proceed_all_lwps): New. (unstop_all_lwps): New. * linux-low.h (struct lwp_info): Rewrite comment for the `suspended' flag. Add the `stop_pc' field. Delete the `pending_stop_pc' field. Tweak the `stepping' flag's comment. Add `'last_resume_kind' and `need_step_over' fields. * inferiors.c (struct thread_info): Delete, moved elsewhere. * mem-break.c (struct breakpoint): Delete `reinserting' flag. Delete `breakpoint_to_reinsert' field. New flag `inserted'. (set_raw_breakpoint_at): New. (set_breakpoint_at): Rewrite to use it. (reinsert_breakpoint_handler): Delete. (set_reinsert_breakpoint): New. (reinsert_breakpoint_by_bp): Delete. (delete_reinsert_breakpoints): New. (uninsert_breakpoint): Rewrite. (uninsert_breakpoints_at): New. (reinsert_breakpoint): Rewrite. (reinsert_breakpoints_at): New. (check_breakpoints): Rewrite. (breakpoint_here): New. (breakpoint_inserted_here): New. (check_mem_read): Adjust. * mem-break.h (breakpoints_supported, breakpoint_here) (breakpoint_inserted_here, set_reinsert_breakpoint): Declare. (reinsert_breakpoint_by_bp): Delete declaration. (delete_reinsert_breakpoints): Declare. (reinsert_breakpoint): Delete declaration. (reinsert_breakpoints_at): Declare. (uninsert_breakpoint): Delete declaration. (uninsert_breakpoints_at): Declare. (check_breakpoints): Adjust prototype. * server.h: Adjust include order. (struct thread_info): Declare here. Add a `last_status' field.
2010-03-24 01:05:03 +01:00
/* Set a reinsert breakpoint at STOP_AT. */
void set_reinsert_breakpoint (CORE_ADDR stop_at);
/* Delete all reinsert breakpoints. */
Teach linux gdbserver to step-over-breakpoints. * linux-low.c (can_hardware_single_step): New. (supports_breakpoints): New. (handle_extended_wait): If stopping threads, read the stop pc of the new cloned LWP. (get_pc): New. (get_stop_pc): Add `lwp' parameter. Handle it. Bail out if the low target doesn't support retrieving the PC. (add_lwp): Set last_resume_kind to resume_continue. (linux_attach_lwp_1): Adjust comments. Always set stop_expected. (linux_attach): Don't clear stop_expected. Set the lwp's last_resume_kind to resume_stop. (linux_detach_one_lwp): Don't check for removed breakpoints. (check_removed_breakpoint): Delete. (status_pending_p): Rename to ... (status_pending_p_callback): ... this. Don't check for removed breakpoints. Don't consider threads that are stopped from GDB's perspective. (linux_wait_for_lwp): Always read the stop_pc here. (cancel_breakpoint): New. (step_over_bkpt): New global. (linux_wait_for_event_1): Implement stepping over breakpoints. (gdb_wants_lwp_stopped): New. (gdb_wants_all_stopped): New. (linux_wait_1): Tag threads as gdb-wants-stopped. Cancel finished single-step traps here. Store the thread's last reported target wait status. (send_sigstop): Don't clear stop_expected. Always set it, instead. (mark_lwp_dead): Remove reference to pending_is_breakpoint. (cancel_finished_single_step): New. (cancel_finished_single_steps): New. (wait_for_sigstop): Don't cancel finished single-step traps here. (linux_resume_one_lwp): Don't check for removed breakpoints. Don't set `step' on non-hardware step archs. (linux_set_resume_request): Ignore resume_stop requests if already stopping or stopped. Set the lwp's last_resume_kind. (resume_status_pending_p): Don't check for removed breakpoints. (need_step_over_p): New. (start_step_over): New. (finish_step_over): New. (linux_resume_one_thread): Always queue a sigstop for resume_stop requests. Clear the thread's last reported target waitstatus. Don't use the `suspended' flag. Don't consider pending breakpoints. (linux_resume): Start a step-over if necessary. (proceed_one_lwp): New. (proceed_all_lwps): New. (unstop_all_lwps): New. * linux-low.h (struct lwp_info): Rewrite comment for the `suspended' flag. Add the `stop_pc' field. Delete the `pending_stop_pc' field. Tweak the `stepping' flag's comment. Add `'last_resume_kind' and `need_step_over' fields. * inferiors.c (struct thread_info): Delete, moved elsewhere. * mem-break.c (struct breakpoint): Delete `reinserting' flag. Delete `breakpoint_to_reinsert' field. New flag `inserted'. (set_raw_breakpoint_at): New. (set_breakpoint_at): Rewrite to use it. (reinsert_breakpoint_handler): Delete. (set_reinsert_breakpoint): New. (reinsert_breakpoint_by_bp): Delete. (delete_reinsert_breakpoints): New. (uninsert_breakpoint): Rewrite. (uninsert_breakpoints_at): New. (reinsert_breakpoint): Rewrite. (reinsert_breakpoints_at): New. (check_breakpoints): Rewrite. (breakpoint_here): New. (breakpoint_inserted_here): New. (check_mem_read): Adjust. * mem-break.h (breakpoints_supported, breakpoint_here) (breakpoint_inserted_here, set_reinsert_breakpoint): Declare. (reinsert_breakpoint_by_bp): Delete declaration. (delete_reinsert_breakpoints): Declare. (reinsert_breakpoint): Delete declaration. (reinsert_breakpoints_at): Declare. (uninsert_breakpoint): Delete declaration. (uninsert_breakpoints_at): Declare. (check_breakpoints): Adjust prototype. * server.h: Adjust include order. (struct thread_info): Declare here. Add a `last_status' field.
2010-03-24 01:05:03 +01:00
void delete_reinsert_breakpoints (void);
Handle reinsert breakpoints for vforked child When a thread is doing step-over with reinsert breakpoint, and the instruction executed is a syscall doing vfork, both parent and child share the memory, so the reinsert breakpoint in the space is visible to both of them. Also, removing the reinsert breakpoints from the child will effectively remove them from the parent. We should carefully manipulate reinsert breakpoints for both processes. What we are doing here is that - uninsert reinsert breakpoints from the parent before cloning the breakpoint list. We use "uninsert" instead of "remove", because we need to "reinsert" them back after vfork is done. In fact, "uninsert" removes them from both child and parent process space. - reinsert breakpoints in parent process are still copied to child's breakpoint list, - remove them from child's breakpoint list as what we did for fork, at this point, reinsert breakpoints are removed from the child and the parent, but they are still tracked by the parent's breakpoint list, - once vfork is done, "reinsert" them back to the parent, gdb/gdbserver: 2016-06-17 Yao Qi <yao.qi@linaro.org> * linux-low.c (handle_extended_wait): Call uninsert_reinsert_breakpoints for the parent process. Remove reinsert breakpoints from the child process. Reinsert them to the parent process when vfork is done. * mem-break.c (uninsert_reinsert_breakpoints): New function. (reinsert_reinsert_breakpoints): New function. * mem-break.h (uninsert_reinsert_breakpoints): Declare (reinsert_reinsert_breakpoints): Declare.
2016-06-17 11:25:13 +02:00
/* Reinsert all reinsert breakpoints of the current process. */
void reinsert_reinsert_breakpoints (void);
/* Uninsert all reinsert breakpoints of the current process. This
still leaves the reinsert breakpoints in the table. */
void uninsert_reinsert_breakpoints (void);
Teach linux gdbserver to step-over-breakpoints. * linux-low.c (can_hardware_single_step): New. (supports_breakpoints): New. (handle_extended_wait): If stopping threads, read the stop pc of the new cloned LWP. (get_pc): New. (get_stop_pc): Add `lwp' parameter. Handle it. Bail out if the low target doesn't support retrieving the PC. (add_lwp): Set last_resume_kind to resume_continue. (linux_attach_lwp_1): Adjust comments. Always set stop_expected. (linux_attach): Don't clear stop_expected. Set the lwp's last_resume_kind to resume_stop. (linux_detach_one_lwp): Don't check for removed breakpoints. (check_removed_breakpoint): Delete. (status_pending_p): Rename to ... (status_pending_p_callback): ... this. Don't check for removed breakpoints. Don't consider threads that are stopped from GDB's perspective. (linux_wait_for_lwp): Always read the stop_pc here. (cancel_breakpoint): New. (step_over_bkpt): New global. (linux_wait_for_event_1): Implement stepping over breakpoints. (gdb_wants_lwp_stopped): New. (gdb_wants_all_stopped): New. (linux_wait_1): Tag threads as gdb-wants-stopped. Cancel finished single-step traps here. Store the thread's last reported target wait status. (send_sigstop): Don't clear stop_expected. Always set it, instead. (mark_lwp_dead): Remove reference to pending_is_breakpoint. (cancel_finished_single_step): New. (cancel_finished_single_steps): New. (wait_for_sigstop): Don't cancel finished single-step traps here. (linux_resume_one_lwp): Don't check for removed breakpoints. Don't set `step' on non-hardware step archs. (linux_set_resume_request): Ignore resume_stop requests if already stopping or stopped. Set the lwp's last_resume_kind. (resume_status_pending_p): Don't check for removed breakpoints. (need_step_over_p): New. (start_step_over): New. (finish_step_over): New. (linux_resume_one_thread): Always queue a sigstop for resume_stop requests. Clear the thread's last reported target waitstatus. Don't use the `suspended' flag. Don't consider pending breakpoints. (linux_resume): Start a step-over if necessary. (proceed_one_lwp): New. (proceed_all_lwps): New. (unstop_all_lwps): New. * linux-low.h (struct lwp_info): Rewrite comment for the `suspended' flag. Add the `stop_pc' field. Delete the `pending_stop_pc' field. Tweak the `stepping' flag's comment. Add `'last_resume_kind' and `need_step_over' fields. * inferiors.c (struct thread_info): Delete, moved elsewhere. * mem-break.c (struct breakpoint): Delete `reinserting' flag. Delete `breakpoint_to_reinsert' field. New flag `inserted'. (set_raw_breakpoint_at): New. (set_breakpoint_at): Rewrite to use it. (reinsert_breakpoint_handler): Delete. (set_reinsert_breakpoint): New. (reinsert_breakpoint_by_bp): Delete. (delete_reinsert_breakpoints): New. (uninsert_breakpoint): Rewrite. (uninsert_breakpoints_at): New. (reinsert_breakpoint): Rewrite. (reinsert_breakpoints_at): New. (check_breakpoints): Rewrite. (breakpoint_here): New. (breakpoint_inserted_here): New. (check_mem_read): Adjust. * mem-break.h (breakpoints_supported, breakpoint_here) (breakpoint_inserted_here, set_reinsert_breakpoint): Declare. (reinsert_breakpoint_by_bp): Delete declaration. (delete_reinsert_breakpoints): Declare. (reinsert_breakpoint): Delete declaration. (reinsert_breakpoints_at): Declare. (uninsert_breakpoint): Delete declaration. (uninsert_breakpoints_at): Declare. (check_breakpoints): Adjust prototype. * server.h: Adjust include order. (struct thread_info): Declare here. Add a `last_status' field.
2010-03-24 01:05:03 +01:00
/* Reinsert breakpoints at WHERE (and change their status to
inserted). */
Teach linux gdbserver to step-over-breakpoints. * linux-low.c (can_hardware_single_step): New. (supports_breakpoints): New. (handle_extended_wait): If stopping threads, read the stop pc of the new cloned LWP. (get_pc): New. (get_stop_pc): Add `lwp' parameter. Handle it. Bail out if the low target doesn't support retrieving the PC. (add_lwp): Set last_resume_kind to resume_continue. (linux_attach_lwp_1): Adjust comments. Always set stop_expected. (linux_attach): Don't clear stop_expected. Set the lwp's last_resume_kind to resume_stop. (linux_detach_one_lwp): Don't check for removed breakpoints. (check_removed_breakpoint): Delete. (status_pending_p): Rename to ... (status_pending_p_callback): ... this. Don't check for removed breakpoints. Don't consider threads that are stopped from GDB's perspective. (linux_wait_for_lwp): Always read the stop_pc here. (cancel_breakpoint): New. (step_over_bkpt): New global. (linux_wait_for_event_1): Implement stepping over breakpoints. (gdb_wants_lwp_stopped): New. (gdb_wants_all_stopped): New. (linux_wait_1): Tag threads as gdb-wants-stopped. Cancel finished single-step traps here. Store the thread's last reported target wait status. (send_sigstop): Don't clear stop_expected. Always set it, instead. (mark_lwp_dead): Remove reference to pending_is_breakpoint. (cancel_finished_single_step): New. (cancel_finished_single_steps): New. (wait_for_sigstop): Don't cancel finished single-step traps here. (linux_resume_one_lwp): Don't check for removed breakpoints. Don't set `step' on non-hardware step archs. (linux_set_resume_request): Ignore resume_stop requests if already stopping or stopped. Set the lwp's last_resume_kind. (resume_status_pending_p): Don't check for removed breakpoints. (need_step_over_p): New. (start_step_over): New. (finish_step_over): New. (linux_resume_one_thread): Always queue a sigstop for resume_stop requests. Clear the thread's last reported target waitstatus. Don't use the `suspended' flag. Don't consider pending breakpoints. (linux_resume): Start a step-over if necessary. (proceed_one_lwp): New. (proceed_all_lwps): New. (unstop_all_lwps): New. * linux-low.h (struct lwp_info): Rewrite comment for the `suspended' flag. Add the `stop_pc' field. Delete the `pending_stop_pc' field. Tweak the `stepping' flag's comment. Add `'last_resume_kind' and `need_step_over' fields. * inferiors.c (struct thread_info): Delete, moved elsewhere. * mem-break.c (struct breakpoint): Delete `reinserting' flag. Delete `breakpoint_to_reinsert' field. New flag `inserted'. (set_raw_breakpoint_at): New. (set_breakpoint_at): Rewrite to use it. (reinsert_breakpoint_handler): Delete. (set_reinsert_breakpoint): New. (reinsert_breakpoint_by_bp): Delete. (delete_reinsert_breakpoints): New. (uninsert_breakpoint): Rewrite. (uninsert_breakpoints_at): New. (reinsert_breakpoint): Rewrite. (reinsert_breakpoints_at): New. (check_breakpoints): Rewrite. (breakpoint_here): New. (breakpoint_inserted_here): New. (check_mem_read): Adjust. * mem-break.h (breakpoints_supported, breakpoint_here) (breakpoint_inserted_here, set_reinsert_breakpoint): Declare. (reinsert_breakpoint_by_bp): Delete declaration. (delete_reinsert_breakpoints): Declare. (reinsert_breakpoint): Delete declaration. (reinsert_breakpoints_at): Declare. (uninsert_breakpoint): Delete declaration. (uninsert_breakpoints_at): Declare. (check_breakpoints): Adjust prototype. * server.h: Adjust include order. (struct thread_info): Declare here. Add a `last_status' field.
2010-03-24 01:05:03 +01:00
void reinsert_breakpoints_at (CORE_ADDR where);
Switch to current thread in finish_step_over This patch adds some sanity check that reinsert breakpoints must be there when doing step-over on software single step target. The check triggers an assert when running forking-threads-plus-breakpoint.exp on arm-linux target, gdb/gdbserver/linux-low.c:4714: A problem internal to GDBserver has been detected.^M int finish_step_over(lwp_info*): Assertion `has_reinsert_breakpoints ()' failed. the error happens when GDBserver has already resumed a thread of process A for step-over (and wait for it hitting reinsert breakpoint), but receives detach request for process B from GDB, which is shown in the backtrace below, (gdb) bt #2 0x000228aa in finish_step_over (lwp=0x12bbd98) at /home/yao/SourceCode/gnu/gdb/git/gdb/gdbserver/linux-low.c:4703 #3 0x00025a50 in finish_step_over (lwp=0x12bbd98) at /home/yao/SourceCode/gnu/gdb/git/gdb/gdbserver/linux-low.c:4749 #4 complete_ongoing_step_over () at /home/yao/SourceCode/gnu/gdb/git/gdb/gdbserver/linux-low.c:4760 #5 linux_detach (pid=25228) at /home/yao/SourceCode/gnu/gdb/git/gdb/gdbserver/linux-low.c:1503 #6 0x00012bae in process_serial_event () at /home/yao/SourceCode/gnu/gdb/git/gdb/gdbserver/server.c:3974 #7 handle_serial_event (err=<optimized out>, client_data=<optimized out>) at /home/yao/SourceCode/gnu/gdb/git/gdb/gdbserver/server.c:4347 #8 0x00016d68 in handle_file_event (event_file_desc=<optimized out>) at /home/yao/SourceCode/gnu/gdb/git/gdb/gdbserver/event-loop.c:429 #9 0x000173ea in process_event () at /home/yao/SourceCode/gnu/gdb/git/gdb/gdbserver/event-loop.c:184 #10 start_event_loop () at /home/yao/SourceCode/gnu/gdb/git/gdb/gdbserver/event-loop.c:547 #11 0x0000aa2c in captured_main (argv=<optimized out>, argc=<optimized out>) at /home/yao/SourceCode/gnu/gdb/git/gdb/gdbserver/server.c:3719 #12 main (argc=<optimized out>, argv=<optimized out>) at /home/yao/SourceCode/gnu/gdb/git/gdb/gdbserver/server.c:3804 the sanity check tries to find the reinsert breakpoint from process B, but nothing is found. It is wrong, we need to search in process A, since we started step-over of a thread of process A. (gdb) p lwp->thread->entry.id $3 = {pid = 25120, lwp = 25131, tid = 0} (gdb) p current_thread->entry.id $4 = {pid = 25228, lwp = 25228, tid = 0} This patch switched current_thread to the thread we are doing step-over in finish_step_over. gdb/gdbserver: 2016-06-17 Yao Qi <yao.qi@linaro.org> * linux-low.c (maybe_hw_step): New function. (linux_resume_one_lwp_throw): Call maybe_hw_step. (finish_step_over): Switch current_thread to lwp temporarily, and assert has_reinsert_breakpoints returns true. (proceed_one_lwp): Call maybe_hw_step. * mem-break.c (has_reinsert_breakpoints): New function. * mem-break.h (has_reinsert_breakpoints): Declare.
2016-06-17 11:25:12 +02:00
/* Process PROC has reinsert breakpoints or not. */
int has_reinsert_breakpoints (struct process_info *proc);
Teach linux gdbserver to step-over-breakpoints. * linux-low.c (can_hardware_single_step): New. (supports_breakpoints): New. (handle_extended_wait): If stopping threads, read the stop pc of the new cloned LWP. (get_pc): New. (get_stop_pc): Add `lwp' parameter. Handle it. Bail out if the low target doesn't support retrieving the PC. (add_lwp): Set last_resume_kind to resume_continue. (linux_attach_lwp_1): Adjust comments. Always set stop_expected. (linux_attach): Don't clear stop_expected. Set the lwp's last_resume_kind to resume_stop. (linux_detach_one_lwp): Don't check for removed breakpoints. (check_removed_breakpoint): Delete. (status_pending_p): Rename to ... (status_pending_p_callback): ... this. Don't check for removed breakpoints. Don't consider threads that are stopped from GDB's perspective. (linux_wait_for_lwp): Always read the stop_pc here. (cancel_breakpoint): New. (step_over_bkpt): New global. (linux_wait_for_event_1): Implement stepping over breakpoints. (gdb_wants_lwp_stopped): New. (gdb_wants_all_stopped): New. (linux_wait_1): Tag threads as gdb-wants-stopped. Cancel finished single-step traps here. Store the thread's last reported target wait status. (send_sigstop): Don't clear stop_expected. Always set it, instead. (mark_lwp_dead): Remove reference to pending_is_breakpoint. (cancel_finished_single_step): New. (cancel_finished_single_steps): New. (wait_for_sigstop): Don't cancel finished single-step traps here. (linux_resume_one_lwp): Don't check for removed breakpoints. Don't set `step' on non-hardware step archs. (linux_set_resume_request): Ignore resume_stop requests if already stopping or stopped. Set the lwp's last_resume_kind. (resume_status_pending_p): Don't check for removed breakpoints. (need_step_over_p): New. (start_step_over): New. (finish_step_over): New. (linux_resume_one_thread): Always queue a sigstop for resume_stop requests. Clear the thread's last reported target waitstatus. Don't use the `suspended' flag. Don't consider pending breakpoints. (linux_resume): Start a step-over if necessary. (proceed_one_lwp): New. (proceed_all_lwps): New. (unstop_all_lwps): New. * linux-low.h (struct lwp_info): Rewrite comment for the `suspended' flag. Add the `stop_pc' field. Delete the `pending_stop_pc' field. Tweak the `stepping' flag's comment. Add `'last_resume_kind' and `need_step_over' fields. * inferiors.c (struct thread_info): Delete, moved elsewhere. * mem-break.c (struct breakpoint): Delete `reinserting' flag. Delete `breakpoint_to_reinsert' field. New flag `inserted'. (set_raw_breakpoint_at): New. (set_breakpoint_at): Rewrite to use it. (reinsert_breakpoint_handler): Delete. (set_reinsert_breakpoint): New. (reinsert_breakpoint_by_bp): Delete. (delete_reinsert_breakpoints): New. (uninsert_breakpoint): Rewrite. (uninsert_breakpoints_at): New. (reinsert_breakpoint): Rewrite. (reinsert_breakpoints_at): New. (check_breakpoints): Rewrite. (breakpoint_here): New. (breakpoint_inserted_here): New. (check_mem_read): Adjust. * mem-break.h (breakpoints_supported, breakpoint_here) (breakpoint_inserted_here, set_reinsert_breakpoint): Declare. (reinsert_breakpoint_by_bp): Delete declaration. (delete_reinsert_breakpoints): Declare. (reinsert_breakpoint): Delete declaration. (reinsert_breakpoints_at): Declare. (uninsert_breakpoint): Delete declaration. (uninsert_breakpoints_at): Declare. (check_breakpoints): Adjust prototype. * server.h: Adjust include order. (struct thread_info): Declare here. Add a `last_status' field.
2010-03-24 01:05:03 +01:00
/* Uninsert breakpoints at WHERE (and change their status to
uninserted). This still leaves the breakpoints in the table. */
Teach linux gdbserver to step-over-breakpoints. * linux-low.c (can_hardware_single_step): New. (supports_breakpoints): New. (handle_extended_wait): If stopping threads, read the stop pc of the new cloned LWP. (get_pc): New. (get_stop_pc): Add `lwp' parameter. Handle it. Bail out if the low target doesn't support retrieving the PC. (add_lwp): Set last_resume_kind to resume_continue. (linux_attach_lwp_1): Adjust comments. Always set stop_expected. (linux_attach): Don't clear stop_expected. Set the lwp's last_resume_kind to resume_stop. (linux_detach_one_lwp): Don't check for removed breakpoints. (check_removed_breakpoint): Delete. (status_pending_p): Rename to ... (status_pending_p_callback): ... this. Don't check for removed breakpoints. Don't consider threads that are stopped from GDB's perspective. (linux_wait_for_lwp): Always read the stop_pc here. (cancel_breakpoint): New. (step_over_bkpt): New global. (linux_wait_for_event_1): Implement stepping over breakpoints. (gdb_wants_lwp_stopped): New. (gdb_wants_all_stopped): New. (linux_wait_1): Tag threads as gdb-wants-stopped. Cancel finished single-step traps here. Store the thread's last reported target wait status. (send_sigstop): Don't clear stop_expected. Always set it, instead. (mark_lwp_dead): Remove reference to pending_is_breakpoint. (cancel_finished_single_step): New. (cancel_finished_single_steps): New. (wait_for_sigstop): Don't cancel finished single-step traps here. (linux_resume_one_lwp): Don't check for removed breakpoints. Don't set `step' on non-hardware step archs. (linux_set_resume_request): Ignore resume_stop requests if already stopping or stopped. Set the lwp's last_resume_kind. (resume_status_pending_p): Don't check for removed breakpoints. (need_step_over_p): New. (start_step_over): New. (finish_step_over): New. (linux_resume_one_thread): Always queue a sigstop for resume_stop requests. Clear the thread's last reported target waitstatus. Don't use the `suspended' flag. Don't consider pending breakpoints. (linux_resume): Start a step-over if necessary. (proceed_one_lwp): New. (proceed_all_lwps): New. (unstop_all_lwps): New. * linux-low.h (struct lwp_info): Rewrite comment for the `suspended' flag. Add the `stop_pc' field. Delete the `pending_stop_pc' field. Tweak the `stepping' flag's comment. Add `'last_resume_kind' and `need_step_over' fields. * inferiors.c (struct thread_info): Delete, moved elsewhere. * mem-break.c (struct breakpoint): Delete `reinserting' flag. Delete `breakpoint_to_reinsert' field. New flag `inserted'. (set_raw_breakpoint_at): New. (set_breakpoint_at): Rewrite to use it. (reinsert_breakpoint_handler): Delete. (set_reinsert_breakpoint): New. (reinsert_breakpoint_by_bp): Delete. (delete_reinsert_breakpoints): New. (uninsert_breakpoint): Rewrite. (uninsert_breakpoints_at): New. (reinsert_breakpoint): Rewrite. (reinsert_breakpoints_at): New. (check_breakpoints): Rewrite. (breakpoint_here): New. (breakpoint_inserted_here): New. (check_mem_read): Adjust. * mem-break.h (breakpoints_supported, breakpoint_here) (breakpoint_inserted_here, set_reinsert_breakpoint): Declare. (reinsert_breakpoint_by_bp): Delete declaration. (delete_reinsert_breakpoints): Declare. (reinsert_breakpoint): Delete declaration. (reinsert_breakpoints_at): Declare. (uninsert_breakpoint): Delete declaration. (uninsert_breakpoints_at): Declare. (check_breakpoints): Adjust prototype. * server.h: Adjust include order. (struct thread_info): Declare here. Add a `last_status' field.
2010-03-24 01:05:03 +01:00
void uninsert_breakpoints_at (CORE_ADDR where);
Static tracepoints support, and UST integration. gdb/gdbserver/ * configure.ac: Handle --with-ust. substitute ustlibs and ustinc. * mem-break.c (uninsert_all_breakpoints) (reinsert_all_breakpoints): New. * mem-break.h (reinsert_all_breakpoints, uninsert_all_breakpoints): * tracepoint.c (ust_loaded, helper_thread_id, cmd_buf): New. (gdb_agent_ust_loaded, helper_thread_id) (gdb_agent_helper_thread_id): New macros. (struct ipa_sym_addresses): Add addr_ust_loaded, addr_helper_thread_id, addr_cmd_buf. (symbol_list): Add ust_loaded, helper_thread_id, cmd_buf. (in_process_agent_loaded_ust): New. (write_e_ust_not_loaded): New. (maybe_write_ipa_ust_not_loaded): New. (struct collect_static_trace_data_action): New. (enum tracepoint_type) <static_tracepoint>: New. (struct tracepoint) <handle>: Mention static tracepoints. (struct static_tracepoint_ctx): New. (CMD_BUF_SIZE): New. (add_tracepoint_action): Handle static tracepoint actions. (unprobe_marker_at): New. (clear_installed_tracepoints): Handle static tracepoints. (cmd_qtdp): Handle static tracepoints. (probe_marker_at): New. (cmd_qtstart): Handle static tracepoints. (response_tracepoint): Handle static tracepoints. (cmd_qtfstm, cmd_qtsstm, cmd_qtstmat): New. (handle_tracepoint_query): Handle qTfSTM, qTsSTM and qTSTMat. (get_context_regcache): Handle static tracepoints. (do_action_at_tracepoint): Handle static tracepoint actions. (traceframe_find_block_type): Handle static trace data blocks. (traceframe_read_sdata): New. (download_tracepoints): Download static tracepoint actions. [HAVE_UST] Include ust/ust.h, dlfcn.h, sys/socket.h, and sys/un.h. (GDB_PROBE_NAME): New. (ust_ops): New. (GET_UST_SYM): New. (USTF): New. (dlsym_ust): New. (ust_marker_to_static_tracepoint): New. (gdb_probe): New. (collect_ust_data_at_tracepoint): New. (gdb_ust_probe): New. (UNIX_PATH_MAX, SOCK_DIR): New. (gdb_ust_connect_sync_socket): New. (resume_thread, stop_thread): New. (run_inferior_command): New. (init_named_socket): New. (gdb_ust_socket_init): New. (cstr_to_hexstr): New. (next_st): New. (first_marker, next_marker): New. (response_ust_marker): New. (cmd_qtfstm, cmd_qtsstm): New. (unprobe_marker_at, probe_marker_at): New. (cmd_qtstmat, gdb_ust_thread): New. (gdb_ust_init): New. (initialize_tracepoint_ftlib): Call gdb_ust_init. * linux-amd64-ipa.c [HAVE_UST]: Include ust/processor.h (ST_REGENTRY): New. (x86_64_st_collect_regmap): New. (X86_64_NUM_ST_COLLECT_GREGS): New. (AMD64_RIP_REGNUM): New. (supply_static_tracepoint_registers): New. * linux-i386-ipa.c [HAVE_UST]: Include ust/processor.h (ST_REGENTRY): New. (i386_st_collect_regmap): New. (i386_NUM_ST_COLLECT_GREGS): New. (supply_static_tracepoint_registers): New. * server.c (handle_query): Handle qXfer:statictrace:read. <qSupported>: Report support for StaticTracepoints, and qXfer:statictrace:read features. * server.h (traceframe_read_sdata) (supply_static_tracepoint_registers): Declare. * remote-utils.c (convert_int_to_ascii, hexchars, ishex, tohex) (unpack_varlen_hex): Include in IPA build. * Makefile.in (ustlibs, ustinc): New. (IPA_OBJS): Add remote-utils-ipa.o. ($(IPA_LIB)): Link -ldl and -lpthread. (UST_CFLAGS): New. (IPAGENT_CFLAGS): Add UST_CFLAGS. * config.in, configure: Regenerate. gdb/ * NEWS: Mention new support for static tracepoints. (New packets): Mention qTfSTM, qTsSTM, qTSTMat and qXfer:statictrace:read. (New features in the GDB remote stub, GDBserver): Mention static tracepoints support using an UST based backend. (New commands): Mention "info static-tracepoint-markers" and "strace". * breakpoint.c (is_marker_spec): New. (is_tracepoint): Handle static tracepoints. (validate_commands_for_breakpoint): Static tracepoints can't do while-stepping. (static_tracepoints_here): New. (bpstat_what): Handle static tracepoints. (print_one_breakpoint_location, allocate_bp_location, mention): Ditto. (create_breakpoint_sal): Ditto. (decode_static_tracepoint_spec): New. (create_breakpoint): Replace `hardwareflag', and `traceflag' with `type_wanted'. Adjust. Handle static tracepoint marker locations. (break_command_1): Adjust. (update_static_tracepoint): New. (update_breakpoint_locations): Handle static tracepoints. (breakpoint_re_set_one): Handle static tracepoint marker locations. (disable_command, enable_command): Handle static tracepoints. (trace_command, ftrace_command): Adjust. (strace_command): New. (create_tracepoint_from_upload): Adjust. (save_breakpoints): Handle static tracepoints. (_initialize_breakpoint): Install the "strace" command. * breakpoint.h (enum bptype): New bp_static_tracepoint type. (struct breakpoint): New fields static_trace_marker_id and static_trace_marker_id_idx. (breakpoints_here_p): Declare. (create_breakpoint): Adjust. (static_tracepoints_here): Declare. * remote.c (struct remote_state) <static_tracepoints>: New field. (PACKET_qXfer_statictrace_read, PACKET_StaticTracepoints): New. (remote_static_tracepoint_marker_at): New. (remote_static_tracepoint_markers_by_strid): New. (remote_static_tracepoint_feature): New. (remote_disconnected_tracing_feature): Handle "StaticTracepoints". (remote_xfer_partial): Handle TARGET_OBJECT_STATIC_TRACE_DATA. (remote_supports_static_tracepoints): New. (remote_download_tracepoint): Download static tracepoints. (init_remote_ops): Install remote_static_tracepoint_marker_at and remote_static_tracepoint_markers_by_strid. (_initialize_remote): Install set|show remote static-tracepoints, and set|show remote read-sdata-object commands. * target.c (update_current_target): Inherit and default to_static_tracepoint_marker_at, and to_static_tracepoint_markers_by_strid. * target.h (static_tracepoint_marker): Forward declare. (enum target_object): New object TARGET_OBJECT_STATIC_TRACE_DATA. (static_tracepoint_marker_p): New typedef. (DEF_VEC_P(static_tracepoint_marker_p)): New VEC type. (struct target_ops): New fields to_static_tracepoint_marker_at and to_static_tracepoint_markers_by_strid. (target_static_tracepoint_marker_at) (target_static_tracepoint_markers_by_strid): New. * tracepoint.c: Include source.h. (validate_actionline): Handle $_sdata. (struct collection_list): New field strace_data. (add_static_trace_data): New. (clear_collection_list): Clear strace_data. (stringify_collection_list): Account for a possible static trace data collection. (encode_actions_1): Encode an $_sdata collection. (parse_tracepoint_definition): Handle static tracepoints. (parse_static_tracepoint_marker_definition): New. (release_static_tracepoint_marker): New. (print_one_static_tracepoint_marker): New. (info_static_tracepoint_markers_command): New. (sdata_make_value): New. (_initialize_tracepoint): Create the $_sdata convenience variable. Add the "info static-tracepoint-markers" command. Mention $_sdata in the "collect" command's help output. * tracepoint.h (struct static_tracepoint_marker): New. (parse_static_tracepoint_marker_definition) (release_static_tracepoint_marker): Declare. * mi/mi-cmd-break.c (mi_cmd_break_insert): Adjust. * python/py-breakpoint.c (bppy_new): Adjust. doc/ * gdb.texinfo (Convenience Variables): Document $_sdata. (Commands to Set Tracepoints): Describe static tracepoints. Add `Listing Static Tracepoint Markers' menu entry. Document "strace". (Tracepoint Action Lists): Document collecting $_sdata. (Listing Static Tracepoint Markers): New subsection. (Tracepoints support in gdbserver): Mention static tracepoints. (remote packets, enabling and disabling): Mention read-sdata-object. (General Query Packets) <qSupported>: Document qXfer:sdata:read and StaticTracepoint. Mention qTfSTM, qTsSTM and qTSTMat as tracepoint packets. Document qXfer:sdata:read. (Tracepoint packets): Document qTfSTM, qTsSTM and qTSTMat.
2010-07-01 12:36:12 +02:00
/* Reinsert all breakpoints of the current process (and change their
status to inserted). */
void reinsert_all_breakpoints (void);
/* Uninsert all breakpoints of the current process (and change their
status to uninserted). This still leaves the breakpoints in the
table. */
void uninsert_all_breakpoints (void);
/* See if any breakpoint claims ownership of STOP_PC. Call the handler for
the breakpoint, if found. */
Teach linux gdbserver to step-over-breakpoints. * linux-low.c (can_hardware_single_step): New. (supports_breakpoints): New. (handle_extended_wait): If stopping threads, read the stop pc of the new cloned LWP. (get_pc): New. (get_stop_pc): Add `lwp' parameter. Handle it. Bail out if the low target doesn't support retrieving the PC. (add_lwp): Set last_resume_kind to resume_continue. (linux_attach_lwp_1): Adjust comments. Always set stop_expected. (linux_attach): Don't clear stop_expected. Set the lwp's last_resume_kind to resume_stop. (linux_detach_one_lwp): Don't check for removed breakpoints. (check_removed_breakpoint): Delete. (status_pending_p): Rename to ... (status_pending_p_callback): ... this. Don't check for removed breakpoints. Don't consider threads that are stopped from GDB's perspective. (linux_wait_for_lwp): Always read the stop_pc here. (cancel_breakpoint): New. (step_over_bkpt): New global. (linux_wait_for_event_1): Implement stepping over breakpoints. (gdb_wants_lwp_stopped): New. (gdb_wants_all_stopped): New. (linux_wait_1): Tag threads as gdb-wants-stopped. Cancel finished single-step traps here. Store the thread's last reported target wait status. (send_sigstop): Don't clear stop_expected. Always set it, instead. (mark_lwp_dead): Remove reference to pending_is_breakpoint. (cancel_finished_single_step): New. (cancel_finished_single_steps): New. (wait_for_sigstop): Don't cancel finished single-step traps here. (linux_resume_one_lwp): Don't check for removed breakpoints. Don't set `step' on non-hardware step archs. (linux_set_resume_request): Ignore resume_stop requests if already stopping or stopped. Set the lwp's last_resume_kind. (resume_status_pending_p): Don't check for removed breakpoints. (need_step_over_p): New. (start_step_over): New. (finish_step_over): New. (linux_resume_one_thread): Always queue a sigstop for resume_stop requests. Clear the thread's last reported target waitstatus. Don't use the `suspended' flag. Don't consider pending breakpoints. (linux_resume): Start a step-over if necessary. (proceed_one_lwp): New. (proceed_all_lwps): New. (unstop_all_lwps): New. * linux-low.h (struct lwp_info): Rewrite comment for the `suspended' flag. Add the `stop_pc' field. Delete the `pending_stop_pc' field. Tweak the `stepping' flag's comment. Add `'last_resume_kind' and `need_step_over' fields. * inferiors.c (struct thread_info): Delete, moved elsewhere. * mem-break.c (struct breakpoint): Delete `reinserting' flag. Delete `breakpoint_to_reinsert' field. New flag `inserted'. (set_raw_breakpoint_at): New. (set_breakpoint_at): Rewrite to use it. (reinsert_breakpoint_handler): Delete. (set_reinsert_breakpoint): New. (reinsert_breakpoint_by_bp): Delete. (delete_reinsert_breakpoints): New. (uninsert_breakpoint): Rewrite. (uninsert_breakpoints_at): New. (reinsert_breakpoint): Rewrite. (reinsert_breakpoints_at): New. (check_breakpoints): Rewrite. (breakpoint_here): New. (breakpoint_inserted_here): New. (check_mem_read): Adjust. * mem-break.h (breakpoints_supported, breakpoint_here) (breakpoint_inserted_here, set_reinsert_breakpoint): Declare. (reinsert_breakpoint_by_bp): Delete declaration. (delete_reinsert_breakpoints): Declare. (reinsert_breakpoint): Delete declaration. (reinsert_breakpoints_at): Declare. (uninsert_breakpoint): Delete declaration. (uninsert_breakpoints_at): Declare. (check_breakpoints): Adjust prototype. * server.h: Adjust include order. (struct thread_info): Declare here. Add a `last_status' field.
2010-03-24 01:05:03 +01:00
void check_breakpoints (CORE_ADDR stop_pc);
/* See if any breakpoints shadow the target memory area from MEM_ADDR
to MEM_ADDR + MEM_LEN. Update the data already read from the target
(in BUF) if necessary. */
* acconfig.h: Remove. * configure.ac: Add a test for socklen_t. Use three-argument AC_DEFINE throughout. * config.in: Regenerated using autoheader 2.59. * configure: Regenerated. * gdbreplay.c (socklen_t): Provide a default. (remote_open): Use socklen_t. * remote-utils.c (socklen_t): Provide a default. (remote_open): Use socklen_t. (convert_int_to_ascii, convert_ascii_to_int, decode_M_packet): Use unsigned char. * i387-fp.c (struct i387_fsave, struct i387_fxsave): Use unsigned char for buffers. * linux-low.c (linux_read_memory, linux_write_memory) (linux_read_auxv): Likewise. * mem-break.c (breakpoint_data, set_breakpoint_data, check_mem_read) (check_mem_write): Likewise. * mem-break.h (set_breakpoint_data, check_mem_read, check_mem_write): Likewise. * regcache.c (struct inferior_rgcache_data, registers_to_string) (registers_from_string, register_data): Likewise. * server.c (handle_query, main): Likewise. * server.h (convert_ascii_to_int, convert_int_to_ascii) (decode_M_packet): Likewise. * target.c (read_inferior_memory, write_inferior_memory): Likewise. * target.h (struct target_ops): Update read_memory, write_memory, and read_auxv. (read_inferior_memory, write_inferior_memory): Update. * linux-low.h (struct linux_target_ops): Change type of breakpoint to unsigned char *. * linux-arm-low.c, linux-cris-low.c, linux-crisv32-low.c, linux-i386-low.c, linux-m32r-low.c, linux-m68k-low.c, linux-mips-low.c, linux-ppc-low.c, linux-ppc64-low.c, linux-s390-low.c, linux-sh-low.c: Update for changes in read_inferior_memory and the_low_target->breakpoint.
2005-06-13 03:59:22 +02:00
void check_mem_read (CORE_ADDR mem_addr, unsigned char *buf, int mem_len);
/* See if any breakpoints shadow the target memory area from MEM_ADDR
to MEM_ADDR + MEM_LEN. Update the data to be written to the target
(in BUF, a copy of MYADDR on entry) if necessary, as well as the
original data for any breakpoints. */
void check_mem_write (CORE_ADDR mem_addr,
unsigned char *buf, const unsigned char *myaddr, int mem_len);
/* Delete all breakpoints. */
void delete_all_breakpoints (void);
/* Clear the "inserted" flag in all breakpoints of PROC. */
void mark_breakpoints_out (struct process_info *proc);
2009-04-01 Pedro Alves <pedro@codesourcery.com> Implement the multiprocess extensions, and add linux multiprocess support. * server.h (ULONGEST): Declare. (struct ptid, ptid_t): New. (minus_one_ptid, null_ptid): Declare. (ptid_build, pid_to_ptid, ptid_get_pid, ptid_get_lwp) (ptid_get_tid, ptid_equal, ptid_is_pid): Declare. (struct inferior_list_entry): Change `id' type from unsigned from to ptid_t. (struct sym_cache, struct breakpoint, struct process_info_private): Forward declare. (struct process_info): Declare. (current_process): Declare. (all_processes): Declare. (initialize_inferiors): Declare. (add_thread): Adjust to use ptid_t. (thread_id_to_gdb_id, thread_to_gdb_id, gdb_id_to_thread_id): Ditto. (add_process, remove_process, find_thread_pid): Declare. (find_inferior_id): Adjust to use ptid_t. (cont_thread, general_thread, step_thread): Change type to ptid_t. (multi_process): Declare. (push_event): Adjust to use ptid_t. (read_ptid, write_ptid): Declare. (prepare_resume_reply): Adjust to use ptid_t. (clear_symbol_cache): Declare. * inferiors.c (all_processes): New. (null_ptid, minus_one_ptid): New. (ptid_build, pid_to_ptid, ptid_get_pid, ptid_get_lwp) (ptid_get_tid, ptid_equal, ptid_is_pid): New. (add_thread): Change unsigned long to ptid. Remove gdb_id parameter. Adjust. (thread_id_to_gdb_id, thread_to_gdb_id): Change unsigned long to ptid. (gdb_id_to_thread): Rename to ... (find_thread_pid): ... this. Change unsigned long to ptid. (gdb_id_to_thread_id, find_inferior_id): Change unsigned long to ptid. (loaded_dll, pull_pid_from_list): Adjust. (add_process, remove_process, find_process_pid) (get_thread_process, current_process, initialize_inferiors): New. * target.h (struct thread_resume) <thread>: Change type to ptid_t. (struct target_waitstatus) <related_pid>: Ditto. (struct target_ops) <kill, detach>: Add `pid' argument. Change return type to int. (struct target_ops) <join>: Add `pid' argument. (struct target_ops) <thread_alive>: Change pid's type to ptid_t. (struct target_ops) <wait>: Add `ptid' field. Change return type to ptid. (kill_inferior, detach_inferior, join_inferior): Add `pid' argument. (mywait): Add `ptid' argument. Change return type to ptid_t. (target_pid_to_str): Declare. * target.c (set_desired_inferior): Adjust to use ptids. (mywait): Add new `ptid' argument. Adjust. (target_pid_to_str): New. * mem-break.h (free_all_breakpoints): Declare. * mem-break.c (breakpoints): Delelete. (set_breakpoint_at, delete_breakpoint, find_breakpoint_at) (check_mem_read, check_mem_write, delete_all_breakpoints): Adjust to use per-process breakpoint list. (free_all_breakpoints): New. * remote-utils.c (struct sym_cache) <name>: Drop `const'. (symbol_cache, all_symbols_looked_up): Delete. (hexchars): New. (ishex, unpack_varlen_hex, write_ptid, hex_or_minus_one, read_ptid): New. (prepare_resume_reply): Change ptid argument's type from unsigned long to ptid_t. Adjust. Implement W;process and X;process. (free_sym_cache, clear_symbol_cache): New. (look_up_one_symbol): Adjust to per-process symbol cache. * * server.c (cont_thread, general_thread, step_thread): Change type to ptid_t. (attached): Delete. (multi_process): New. (last_ptid): Change type to ptid_t. (struct vstop_notif) <ptid>: Change type to ptid_t. (queue_stop_reply, push_event): Change `ptid' argument's type to ptid_t. (discard_queued_stop_replies): Add `pid' argument. (start_inferior): Adjust to use ptids. Adjust to mywait interface changes. Don't reference the `attached' global. (attach_inferior): Adjust to mywait interface changes. (handle_query): Adjust to use ptids. Parse GDB's qSupported features. Handle and report "multiprocess+". Handle "qAttached:PID". (handle_v_cont): Adjust to use ptids. Adjust to mywait interface changes. (handle_v_kill): New. (handle_v_stopped): Adjust to use target_pid_to_str. (handle_v_requests): Allow multiple attaches and runs when multiprocess extensions are in effect. Handle "vKill". (myresume): Adjust to use ptids. (queue_stop_reply_callback): Add `arg' parameter. Handle it. (handle_status): Adjust to discard_queued_stop_replies interface change. (first_thread_of, kill_inferior_callback) (detach_or_kill_inferior_callback, join_inferiors_callback): New. (main): Call initialize_inferiors. Adjust to use ptids, killing and detaching from all inferiors. Handle multiprocess packet variants. * linux-low.h: Include gdb_proc_service.h. (struct process_info_private): New. (struct linux_target_ops) <pid_of>: Use ptid_get_pid. <lwpid_of>: Use ptid_get_lwp. (get_lwp_thread): Adjust. (struct lwp_info): Add `dead' member. (find_lwp_pid): Declare. * linux-low.c (thread_db_active): Delete. (new_inferior): Adjust comment. (inferior_pid): Delete. (linux_add_process): New. (handle_extended_wait): Adjust. (add_lwp): Change unsigned long to ptid. (linux_create_inferior): Add process to processes table. Adjust to use ptids. Don't set new_inferior here. (linux_attach_lwp): Rename to ... (linux_attach_lwp_1): ... this. Add `initial' argument. Handle it. Adjust to use ptids. (linux_attach_lwp): New. (linux_attach): Add process to processes table. Don't set new_inferior here. (struct counter): New. (second_thread_of_pid_p, last_thread_of_process_p): New. (linux_kill_one_lwp): Add `args' parameter. Handle it. Adjust to multiple processes. (linux_kill): Add `pid' argument. Handle it. Adjust to multiple processes. Remove process from process table. (linux_detach_one_lwp): Add `args' parameter. Handle it. Adjust to multiple processes. (any_thread_of): New. (linux_detach): Add `pid' argument, and handle it. Remove process from processes table. (linux_join): Add `pid' argument. Handle it. (linux_thread_alive): Change unsighed long argument to ptid_t. Consider dead lwps as not being alive. (status_pending_p): Rename `dummy' argument to `arg'. Filter out threads we're not interested in. (same_lwp, find_lwp_pid): New. (linux_wait_for_lwp): Change `pid' argument's type from int to ptid_t. Adjust. (linux_wait_for_event): Rename to ... (linux_wait_for_event_1): ... this. Change `pid' argument's type from int to ptid_t. Adjust. (linux_wait_for_event): New. (linux_wait_1): Add `ptid' argument. Change return type to ptid_t. Adjust. Use last_thread_of_process_p. Remove processes that exit from the process table. (linux_wait): Add `ptid' argument. Change return type to ptid_t. Adjust. (mark_lwp_dead): New. (wait_for_sigstop): Adjust to use ptids. If a process exits while stopping all threads, mark its main lwp as dead. (linux_set_resume_request, linux_resume_one_thread): Adjust to use ptids. (fetch_register, usr_store_inferior_registers) (regsets_fetch_inferior_registers) (regsets_store_inferior_registers, linux_read_memory) (linux_write_memory): Inline `inferior_pid'. (linux_look_up_symbols): Adjust to use per-process `thread_db_active'. (linux_request_interrupt): Adjust to use ptids. (linux_read_auxv): Inline `inferior_pid'. (initialize_low): Don't reference thread_db_active. * gdb_proc_service.h (struct ps_prochandle) <pid>: Remove. * proc-service.c (ps_lgetregs): Use find_lwp_pid. (ps_getpid): Return the pid of the current inferior. * thread-db.c (proc_handle, thread_agent): Delete. (thread_db_create_event, thread_db_enable_reporting): Adjust to per-process data. (find_one_thread): Change argument type to ptid_t. Adjust to per-process data. (maybe_attach_thread): Adjust to per-process data and ptids. (thread_db_find_new_threads): Ditto. (thread_db_init): Ditto. * spu-low.c (spu_create_inferior, spu_attach): Add process to processes table. Adjust to use ptids. (spu_kill, spu_detach): Adjust interface. Remove process from processes table. (spu_join, spu_thread_alive): Adjust interface. (spu_wait): Adjust interface. Remove process from processes table. Adjust to use ptids. * win32-low.c (current_inferior_tid): Delete. (current_inferior_ptid): New. (debug_event_ptid): New. (thread_rec): Take a ptid. Adjust. (child_add_thread): Add `pid' argument. Adjust to use ptids. (child_delete_thread): Ditto. (do_initial_child_stuff): Add `attached' argument. Add process to processes table. (child_fetch_inferior_registers, child_store_inferior_registers): Adjust. (win32_create_inferior): Pass 0 to do_initial_child_stuff. (win32_attach): Pass 1 to do_initial_child_stuff. (win32_kill): Adjust interface. Remove process from processes table. (win32_detach): Ditto. (win32_join): Adjust interface. (win32_thread_alive): Take a ptid. (win32_resume): Adjust to use ptids. (get_child_debug_event): Ditto. (win32_wait): Adjust interface. Remove exiting process from processes table.
2009-04-02 00:50:24 +02:00
/* Delete all breakpoints, but do not try to un-insert them from the
inferior. */
void free_all_breakpoints (struct process_info *proc);
/* Check if breakpoints still seem to be inserted in the inferior. */
void validate_breakpoints (void);
gdb/gdbserver/ 2010-06-01 Pedro Alves <pedro@codesourcery.com> Stan Shebs <stan@codesourcery.com> * Makefile.in (IPA_DEPFILES, extra_libraries): New. (all): Depend on $(extra_libraries). (install-only): Install the IPA. (IPA_OBJS, IPA_LIB): New. (clean): Remove the IPA lib. (IPAGENT_CFLAGS): New. (tracepoint-ipa.o, utils-ipa.o, remote-utils-ipa.o) (regcache-ipa.o, i386-linux-ipa.o, linux-i386-ipa.o) (linux-amd64-ipa.o, amd64-linux-ipa.o): New rules. * linux-amd64-ipa.c, linux-i386-ipa.c: New files. * configure.ac: Check for atomic builtins support in the compiler. (IPA_DEPFILES, extra_libraries): Define. * configure.srv (ipa_obj): Add description. (ipa_i386_linux_regobj, ipa_amd64_linux_regobj): Define. (i[34567]86-*-linux*): Set ipa_obj. (x86_64-*-linux*): Set ipa_obj. * linux-low.c (stabilizing_threads): New. (supports_fast_tracepoints): New. (linux_detach): Stabilize threads before detaching. (handle_tracepoints): Handle internal tracing breakpoints. Assert the lwp is either not stabilizing, or is moving out of a jump pad. (linux_fast_tracepoint_collecting): New. (maybe_move_out_of_jump_pad): New. (enqueue_one_deferred_signal): New. (dequeue_one_deferred_signal): New. (linux_wait_for_event_1): If moving out of a jump pad, defer pending signals to later. (linux_stabilize_threads): New. (linux_wait_1): Check if threads need moving out of jump pads, and do it if so. (stuck_in_jump_pad_callback): New. (move_out_of_jump_pad_callback): New. (lwp_running): New. (linux_resume_one_lwp): Handle moving out of jump pads. (linux_set_resume_request): Dequeue deferred signals. (need_step_over_p): Also step over fast tracepoint jumps. (start_step_over): Also uninsert fast tracepoint jumps. (finish_step_over): Also reinsert fast tracepoint jumps. (linux_install_fast_tracepoint_jump): New. (linux_target_ops): Install linux_stabilize_threads and linux_install_fast_tracepoint_jump_pad. * linux-low.h (linux_target_ops) <get_thread_area, install_fast_tracepoint_jump_pad>: New fields. (struct lwp_info) <collecting_fast_tracepoint, pending_signals_to_report, exit_jump_pad_bkpt>: New fields. (linux_get_thread_area): Declare. * linux-x86-low.c (jump_insn): New. (x86_get_thread_area): New. (append_insns): New. (push_opcode): New. (amd64_install_fast_tracepoint_jump_pad): New. (i386_install_fast_tracepoint_jump_pad): New. (x86_install_fast_tracepoint_jump_pad): New. (the_low_target): Install x86_get_thread_area and x86_install_fast_tracepoint_jump_pad. * mem-break.c (set_raw_breakpoint_at): Use read_inferior_memory. (struct fast_tracepoint_jump): New. (fast_tracepoint_jump_insn): New. (fast_tracepoint_jump_shadow): New. (find_fast_tracepoint_jump_at): New. (fast_tracepoint_jump_here): New. (delete_fast_tracepoint_jump): New. (set_fast_tracepoint_jump): New. (uninsert_fast_tracepoint_jumps_at): New. (reinsert_fast_tracepoint_jumps_at): New. (set_breakpoint_at): Use write_inferior_memory. (uninsert_raw_breakpoint): Use write_inferior_memory. (check_mem_read): Mask out fast tracepoint jumps. (check_mem_write): Mask out fast tracepoint jumps. * mem-break.h (struct fast_tracepoint_jump): Forward declare. (set_fast_tracepoint_jump): Declare. (delete_fast_tracepoint_jump) (fast_tracepoint_jump_here, uninsert_fast_tracepoint_jumps_at) (reinsert_fast_tracepoint_jumps_at): Declare. * regcache.c: Don't compile many functions when building the in-process agent library. (init_register_cache) [IN_PROCESS_AGENT]: Don't allow allocating the register buffer in the heap. (free_register_cache): If the register buffer isn't owned by the regcache, don't free it. (set_register_cache) [IN_PROCESS_AGENT]: Don't re-alocate pre-existing register caches. * remote-utils.c (convert_int_to_ascii): Constify `from' parameter type. (convert_ascii_to_int): : Constify `from' parameter type. (decode_M_packet, decode_X_packet): Replace the `to' parameter by a `to_p' pointer to pointer parameter. If TO_P is NULL, malloc the needed buffer in-place. (relocate_instruction): New. * server.c (handle_query) <qSymbols>: If the target supports tracepoints, give it a chance of looking up symbols. Report support for fast tracepoints. (handle_status): Stabilize threads. (process_serial_event): Adjust. * server.h (struct fast_tracepoint_jump): Forward declare. (struct process_info) <fast_tracepoint_jumps>: New field. (convert_ascii_to_int, convert_int_to_ascii): Adjust. (decode_X_packet, decode_M_packet): Adjust. (relocate_instruction): Declare. (in_process_agent_loaded): Declare. (tracepoint_look_up_symbols): Declare. (struct fast_tpoint_collect_status): Declare. (fast_tracepoint_collecting): Declare. (force_unlock_trace_buffer): Declare. (handle_tracepoint_bkpts): Declare. (initialize_low_tracepoint) (supply_fast_tracepoint_registers) [IN_PROCESS_AGENT]: Declare. * target.h (struct target_ops) <stabilize_threads, install_fast_tracepoint_jump_pad>: New fields. (stabilize_threads, install_fast_tracepoint_jump_pad): New. * tracepoint.c [HAVE_MALLOC_H]: Include malloc.h. [HAVE_STDINT_H]: Include stdint.h. (trace_debug_1): Rename to ... (trace_vdebug): ... this. (trace_debug): Rename to ... (trace_debug_1): ... this. Add `level' parameter. (trace_debug): New. (ATTR_USED, ATTR_NOINLINE): New. (IP_AGENT_EXPORT): New. (gdb_tp_heap_buffer, gdb_jump_pad_buffer, gdb_jump_pad_buffer_end) (collecting, gdb_collect, stop_tracing, flush_trace_buffer) (about_to_request_buffer_space, trace_buffer_is_full) (stopping_tracepoint, expr_eval_result, error_tracepoint) (tracepoints, tracing, trace_buffer_ctrl, trace_buffer_ctrl_curr) (trace_buffer_lo, trace_buffer_hi, traceframe_read_count) (traceframe_write_count, traceframes_created) (trace_state_variables) New renaming defines. (struct ipa_sym_addresses): New. (STRINGIZE_1, STRINGIZE, IPA_SYM): New. (symbol_list): New. (ipa_sym_addrs): New. (all_tracepoint_symbols_looked_up): New. (in_process_agent_loaded): New. (write_e_ipa_not_loaded): New. (maybe_write_ipa_not_loaded): New. (tracepoint_look_up_symbols): New. (debug_threads) [IN_PROCESS_AGENT]: New. (read_inferior_memory) [IN_PROCESS_AGENT]: New. (UNKNOWN_SIDE_EFFECTS): New. (stop_tracing): New. (flush_trace_buffer): New. (stop_tracing_bkpt): New. (flush_trace_buffer_bkpt): New. (read_inferior_integer): New. (read_inferior_uinteger): New. (read_inferior_data_pointer): New. (write_inferior_data_pointer): New. (write_inferior_integer): New. (write_inferior_uinteger): New. (struct collect_static_trace_data_action): Delete. (enum tracepoint_type): New. (struct tracepoint) <type>: New field `type'. <actions_str, step_actions, step_actions_str>: Only include in GDBserver. <orig_size, obj_addr_on_target, adjusted_insn_addr> <adjusted_insn_addr_end, jump_pad, jump_pad_end>: New fields. (tracepoints): Use IP_AGENT_EXPORT. (last_tracepoint): Don't include in the IPA. (stopping_tracepoint): Use IP_AGENT_EXPORT. (trace_buffer_is_full): Use IP_AGENT_EXPORT. (alloced_trace_state_variables): New. (trace_state_variables): Use IP_AGENT_EXPORT. (traceframe_t): Delete unused variable. (circular_trace_buffer): Don't include in the IPA. (trace_buffer_start): Delete. (struct trace_buffer_control): New. (trace_buffer_free): Delete. (struct ipa_trace_buffer_control): New. (GDBSERVER_FLUSH_COUNT_MASK, GDBSERVER_FLUSH_COUNT_MASK_PREV) (GDBSERVER_FLUSH_COUNT_MASK_CURR, GDBSERVER_UPDATED_FLUSH_COUNT_BIT): New. (trace_buffer_ctrl): New. (TRACE_BUFFER_CTRL_CURR): New. (trace_buffer_start, trace_buffer_free, trace_buffer_end_free): Reimplement as macros. (trace_buffer_wrap): Delete. (traceframe_write_count, traceframe_read_count) (traceframes_created, tracing): Use IP_AGENT_EXPORT. (struct tracepoint_hit_ctx) <type>: New field. (struct fast_tracepoint_ctx): New. (memory_barrier): New. (cmpxchg): New. (record_tracepoint_error): Update atomically in the IPA. (clear_inferior_trace_buffer): New. (about_to_request_buffer_space): New. (trace_buffer_alloc): Handle GDBserver and inferior simulatenous updating the same buffer. (add_tracepoint): Default the tracepoint's type to trap tracepoint, and orig_size to -1. (get_trace_state_variable) [IN_PROCESS_AGENT]: Handle allocated internal variables. (create_trace_state_variable): New parameter `gdb'. Handle it. (clear_installed_tracepoints): Clear fast tracepoint jumps. (cmd_qtdp): Handle fast tracepoints. (cmd_qtdv): Adjust. (max_jump_pad_size): New. (gdb_jump_pad_head): New. (get_jump_space_head): New. (claim_jump_space): New. (sort_tracepoints): New. (MAX_JUMP_SIZE): New. (cmd_qtstart): Handle fast tracepoints. Sync tracepoints with the IPA. (stop_tracing) [IN_PROCESS_AGENT]: Don't include the tdisconnected support. Upload fast traceframes, and delete internal IPA breakpoints. (stop_tracing_handler): New. (flush_trace_buffer_handler): New. (cmd_qtstop): Upload fast tracepoints. (response_tracepoint): Handle fast tracepoints. (tracepoint_finished_step): Upload fast traceframes. Set the tracepoint hit context's tracepoint type. (handle_tracepoint_bkpts): New. (tracepoint_was_hit): Set the tracepoint hit context's tracepoint type. Add comment about fast tracepoints. (collect_data_at_tracepoint) [IN_PROCESS_AGENT]: Don't access the non-existing action_str field. (get_context_regcache): Handle fast tracepoints. (do_action_at_tracepoint) [!IN_PROCESS_AGENT]: Don't write the PC to the regcache. (fast_tracepoint_from_jump_pad_address): New. (fast_tracepoint_from_ipa_tpoint_address): New. (collecting_t): New. (force_unlock_trace_buffer): New. (fast_tracepoint_collecting): New. (collecting): New. (gdb_collect): New. (write_inferior_data_ptr): New. (target_tp_heap): New. (target_malloc): New. (download_agent_expr): New. (UALIGN): New. (download_tracepoints): New. (download_trace_state_variables): New. (upload_fast_traceframes): New. (IPA_FIRST_TRACEFRAME): New. (IPA_NEXT_TRACEFRAME_1): New. (IPA_NEXT_TRACEFRAME): New. [IN_PROCESS_AGENT]: Include sys/mman.h and fcntl.h. [IN_PROCESS_AGENT] (gdb_tp_heap_buffer, gdb_jump_pad_buffer) (gdb_jump_pad_buffer_end): New. [IN_PROCESS_AGENT] (initialize_tracepoint_ftlib): New. (initialize_tracepoint): Adjust. [IN_PROCESS_AGENT]: Allocate the IPA heap, and jump pad scratch buffer. Initialize the low module. * utils.c (PREFIX, TOOLNAME): New. (malloc_failure): Use PREFIX. (error): In the IPA, an error causes an exit. (fatal, warning): Use PREFIX. (internal_error): Use TOOLNAME. (NUMCELLS): Increase to 10. * configure, config.in: Regenerate. gdb/ 2010-06-01 Pedro Alves <pedro@codesourcery.com> * NEWS: Mention gdbserver fast tracepoints support. gdb/doc/ 2010-06-01 Pedro Alves <pedro@codesourcery.com> * gdb.texinfo (Set Tracepoints): Mention tracepoints support in gdbserver, and add cross reference. (Tracepoints support in gdbserver): New subsection.
2010-06-01 15:20:52 +02:00
/* Insert a fast tracepoint jump at WHERE, using instruction INSN, of
LENGTH bytes. */
struct fast_tracepoint_jump *set_fast_tracepoint_jump (CORE_ADDR where,
unsigned char *insn,
ULONGEST length);
/* Increment reference counter of JP. */
void inc_ref_fast_tracepoint_jump (struct fast_tracepoint_jump *jp);
gdb/gdbserver/ 2010-06-01 Pedro Alves <pedro@codesourcery.com> Stan Shebs <stan@codesourcery.com> * Makefile.in (IPA_DEPFILES, extra_libraries): New. (all): Depend on $(extra_libraries). (install-only): Install the IPA. (IPA_OBJS, IPA_LIB): New. (clean): Remove the IPA lib. (IPAGENT_CFLAGS): New. (tracepoint-ipa.o, utils-ipa.o, remote-utils-ipa.o) (regcache-ipa.o, i386-linux-ipa.o, linux-i386-ipa.o) (linux-amd64-ipa.o, amd64-linux-ipa.o): New rules. * linux-amd64-ipa.c, linux-i386-ipa.c: New files. * configure.ac: Check for atomic builtins support in the compiler. (IPA_DEPFILES, extra_libraries): Define. * configure.srv (ipa_obj): Add description. (ipa_i386_linux_regobj, ipa_amd64_linux_regobj): Define. (i[34567]86-*-linux*): Set ipa_obj. (x86_64-*-linux*): Set ipa_obj. * linux-low.c (stabilizing_threads): New. (supports_fast_tracepoints): New. (linux_detach): Stabilize threads before detaching. (handle_tracepoints): Handle internal tracing breakpoints. Assert the lwp is either not stabilizing, or is moving out of a jump pad. (linux_fast_tracepoint_collecting): New. (maybe_move_out_of_jump_pad): New. (enqueue_one_deferred_signal): New. (dequeue_one_deferred_signal): New. (linux_wait_for_event_1): If moving out of a jump pad, defer pending signals to later. (linux_stabilize_threads): New. (linux_wait_1): Check if threads need moving out of jump pads, and do it if so. (stuck_in_jump_pad_callback): New. (move_out_of_jump_pad_callback): New. (lwp_running): New. (linux_resume_one_lwp): Handle moving out of jump pads. (linux_set_resume_request): Dequeue deferred signals. (need_step_over_p): Also step over fast tracepoint jumps. (start_step_over): Also uninsert fast tracepoint jumps. (finish_step_over): Also reinsert fast tracepoint jumps. (linux_install_fast_tracepoint_jump): New. (linux_target_ops): Install linux_stabilize_threads and linux_install_fast_tracepoint_jump_pad. * linux-low.h (linux_target_ops) <get_thread_area, install_fast_tracepoint_jump_pad>: New fields. (struct lwp_info) <collecting_fast_tracepoint, pending_signals_to_report, exit_jump_pad_bkpt>: New fields. (linux_get_thread_area): Declare. * linux-x86-low.c (jump_insn): New. (x86_get_thread_area): New. (append_insns): New. (push_opcode): New. (amd64_install_fast_tracepoint_jump_pad): New. (i386_install_fast_tracepoint_jump_pad): New. (x86_install_fast_tracepoint_jump_pad): New. (the_low_target): Install x86_get_thread_area and x86_install_fast_tracepoint_jump_pad. * mem-break.c (set_raw_breakpoint_at): Use read_inferior_memory. (struct fast_tracepoint_jump): New. (fast_tracepoint_jump_insn): New. (fast_tracepoint_jump_shadow): New. (find_fast_tracepoint_jump_at): New. (fast_tracepoint_jump_here): New. (delete_fast_tracepoint_jump): New. (set_fast_tracepoint_jump): New. (uninsert_fast_tracepoint_jumps_at): New. (reinsert_fast_tracepoint_jumps_at): New. (set_breakpoint_at): Use write_inferior_memory. (uninsert_raw_breakpoint): Use write_inferior_memory. (check_mem_read): Mask out fast tracepoint jumps. (check_mem_write): Mask out fast tracepoint jumps. * mem-break.h (struct fast_tracepoint_jump): Forward declare. (set_fast_tracepoint_jump): Declare. (delete_fast_tracepoint_jump) (fast_tracepoint_jump_here, uninsert_fast_tracepoint_jumps_at) (reinsert_fast_tracepoint_jumps_at): Declare. * regcache.c: Don't compile many functions when building the in-process agent library. (init_register_cache) [IN_PROCESS_AGENT]: Don't allow allocating the register buffer in the heap. (free_register_cache): If the register buffer isn't owned by the regcache, don't free it. (set_register_cache) [IN_PROCESS_AGENT]: Don't re-alocate pre-existing register caches. * remote-utils.c (convert_int_to_ascii): Constify `from' parameter type. (convert_ascii_to_int): : Constify `from' parameter type. (decode_M_packet, decode_X_packet): Replace the `to' parameter by a `to_p' pointer to pointer parameter. If TO_P is NULL, malloc the needed buffer in-place. (relocate_instruction): New. * server.c (handle_query) <qSymbols>: If the target supports tracepoints, give it a chance of looking up symbols. Report support for fast tracepoints. (handle_status): Stabilize threads. (process_serial_event): Adjust. * server.h (struct fast_tracepoint_jump): Forward declare. (struct process_info) <fast_tracepoint_jumps>: New field. (convert_ascii_to_int, convert_int_to_ascii): Adjust. (decode_X_packet, decode_M_packet): Adjust. (relocate_instruction): Declare. (in_process_agent_loaded): Declare. (tracepoint_look_up_symbols): Declare. (struct fast_tpoint_collect_status): Declare. (fast_tracepoint_collecting): Declare. (force_unlock_trace_buffer): Declare. (handle_tracepoint_bkpts): Declare. (initialize_low_tracepoint) (supply_fast_tracepoint_registers) [IN_PROCESS_AGENT]: Declare. * target.h (struct target_ops) <stabilize_threads, install_fast_tracepoint_jump_pad>: New fields. (stabilize_threads, install_fast_tracepoint_jump_pad): New. * tracepoint.c [HAVE_MALLOC_H]: Include malloc.h. [HAVE_STDINT_H]: Include stdint.h. (trace_debug_1): Rename to ... (trace_vdebug): ... this. (trace_debug): Rename to ... (trace_debug_1): ... this. Add `level' parameter. (trace_debug): New. (ATTR_USED, ATTR_NOINLINE): New. (IP_AGENT_EXPORT): New. (gdb_tp_heap_buffer, gdb_jump_pad_buffer, gdb_jump_pad_buffer_end) (collecting, gdb_collect, stop_tracing, flush_trace_buffer) (about_to_request_buffer_space, trace_buffer_is_full) (stopping_tracepoint, expr_eval_result, error_tracepoint) (tracepoints, tracing, trace_buffer_ctrl, trace_buffer_ctrl_curr) (trace_buffer_lo, trace_buffer_hi, traceframe_read_count) (traceframe_write_count, traceframes_created) (trace_state_variables) New renaming defines. (struct ipa_sym_addresses): New. (STRINGIZE_1, STRINGIZE, IPA_SYM): New. (symbol_list): New. (ipa_sym_addrs): New. (all_tracepoint_symbols_looked_up): New. (in_process_agent_loaded): New. (write_e_ipa_not_loaded): New. (maybe_write_ipa_not_loaded): New. (tracepoint_look_up_symbols): New. (debug_threads) [IN_PROCESS_AGENT]: New. (read_inferior_memory) [IN_PROCESS_AGENT]: New. (UNKNOWN_SIDE_EFFECTS): New. (stop_tracing): New. (flush_trace_buffer): New. (stop_tracing_bkpt): New. (flush_trace_buffer_bkpt): New. (read_inferior_integer): New. (read_inferior_uinteger): New. (read_inferior_data_pointer): New. (write_inferior_data_pointer): New. (write_inferior_integer): New. (write_inferior_uinteger): New. (struct collect_static_trace_data_action): Delete. (enum tracepoint_type): New. (struct tracepoint) <type>: New field `type'. <actions_str, step_actions, step_actions_str>: Only include in GDBserver. <orig_size, obj_addr_on_target, adjusted_insn_addr> <adjusted_insn_addr_end, jump_pad, jump_pad_end>: New fields. (tracepoints): Use IP_AGENT_EXPORT. (last_tracepoint): Don't include in the IPA. (stopping_tracepoint): Use IP_AGENT_EXPORT. (trace_buffer_is_full): Use IP_AGENT_EXPORT. (alloced_trace_state_variables): New. (trace_state_variables): Use IP_AGENT_EXPORT. (traceframe_t): Delete unused variable. (circular_trace_buffer): Don't include in the IPA. (trace_buffer_start): Delete. (struct trace_buffer_control): New. (trace_buffer_free): Delete. (struct ipa_trace_buffer_control): New. (GDBSERVER_FLUSH_COUNT_MASK, GDBSERVER_FLUSH_COUNT_MASK_PREV) (GDBSERVER_FLUSH_COUNT_MASK_CURR, GDBSERVER_UPDATED_FLUSH_COUNT_BIT): New. (trace_buffer_ctrl): New. (TRACE_BUFFER_CTRL_CURR): New. (trace_buffer_start, trace_buffer_free, trace_buffer_end_free): Reimplement as macros. (trace_buffer_wrap): Delete. (traceframe_write_count, traceframe_read_count) (traceframes_created, tracing): Use IP_AGENT_EXPORT. (struct tracepoint_hit_ctx) <type>: New field. (struct fast_tracepoint_ctx): New. (memory_barrier): New. (cmpxchg): New. (record_tracepoint_error): Update atomically in the IPA. (clear_inferior_trace_buffer): New. (about_to_request_buffer_space): New. (trace_buffer_alloc): Handle GDBserver and inferior simulatenous updating the same buffer. (add_tracepoint): Default the tracepoint's type to trap tracepoint, and orig_size to -1. (get_trace_state_variable) [IN_PROCESS_AGENT]: Handle allocated internal variables. (create_trace_state_variable): New parameter `gdb'. Handle it. (clear_installed_tracepoints): Clear fast tracepoint jumps. (cmd_qtdp): Handle fast tracepoints. (cmd_qtdv): Adjust. (max_jump_pad_size): New. (gdb_jump_pad_head): New. (get_jump_space_head): New. (claim_jump_space): New. (sort_tracepoints): New. (MAX_JUMP_SIZE): New. (cmd_qtstart): Handle fast tracepoints. Sync tracepoints with the IPA. (stop_tracing) [IN_PROCESS_AGENT]: Don't include the tdisconnected support. Upload fast traceframes, and delete internal IPA breakpoints. (stop_tracing_handler): New. (flush_trace_buffer_handler): New. (cmd_qtstop): Upload fast tracepoints. (response_tracepoint): Handle fast tracepoints. (tracepoint_finished_step): Upload fast traceframes. Set the tracepoint hit context's tracepoint type. (handle_tracepoint_bkpts): New. (tracepoint_was_hit): Set the tracepoint hit context's tracepoint type. Add comment about fast tracepoints. (collect_data_at_tracepoint) [IN_PROCESS_AGENT]: Don't access the non-existing action_str field. (get_context_regcache): Handle fast tracepoints. (do_action_at_tracepoint) [!IN_PROCESS_AGENT]: Don't write the PC to the regcache. (fast_tracepoint_from_jump_pad_address): New. (fast_tracepoint_from_ipa_tpoint_address): New. (collecting_t): New. (force_unlock_trace_buffer): New. (fast_tracepoint_collecting): New. (collecting): New. (gdb_collect): New. (write_inferior_data_ptr): New. (target_tp_heap): New. (target_malloc): New. (download_agent_expr): New. (UALIGN): New. (download_tracepoints): New. (download_trace_state_variables): New. (upload_fast_traceframes): New. (IPA_FIRST_TRACEFRAME): New. (IPA_NEXT_TRACEFRAME_1): New. (IPA_NEXT_TRACEFRAME): New. [IN_PROCESS_AGENT]: Include sys/mman.h and fcntl.h. [IN_PROCESS_AGENT] (gdb_tp_heap_buffer, gdb_jump_pad_buffer) (gdb_jump_pad_buffer_end): New. [IN_PROCESS_AGENT] (initialize_tracepoint_ftlib): New. (initialize_tracepoint): Adjust. [IN_PROCESS_AGENT]: Allocate the IPA heap, and jump pad scratch buffer. Initialize the low module. * utils.c (PREFIX, TOOLNAME): New. (malloc_failure): Use PREFIX. (error): In the IPA, an error causes an exit. (fatal, warning): Use PREFIX. (internal_error): Use TOOLNAME. (NUMCELLS): Increase to 10. * configure, config.in: Regenerate. gdb/ 2010-06-01 Pedro Alves <pedro@codesourcery.com> * NEWS: Mention gdbserver fast tracepoints support. gdb/doc/ 2010-06-01 Pedro Alves <pedro@codesourcery.com> * gdb.texinfo (Set Tracepoints): Mention tracepoints support in gdbserver, and add cross reference. (Tracepoints support in gdbserver): New subsection.
2010-06-01 15:20:52 +02:00
/* Delete fast tracepoint jump TODEL from our tables, and uninsert if
from memory. */
int delete_fast_tracepoint_jump (struct fast_tracepoint_jump *todel);
/* Returns true if there's fast tracepoint jump set at WHERE. */
int fast_tracepoint_jump_here (CORE_ADDR);
/* Uninsert fast tracepoint jumps at WHERE (and change their status to
uninserted). This still leaves the tracepoints in the table. */
void uninsert_fast_tracepoint_jumps_at (CORE_ADDR pc);
/* Reinsert fast tracepoint jumps at WHERE (and change their status to
inserted). */
void reinsert_fast_tracepoint_jumps_at (CORE_ADDR where);
[GDBserver] Make Zx/zx packet handling idempotent. This patch fixes hardware breakpoint regressions exposed by my fix for "PR breakpoints/7143 - Watchpoint does not trigger when first set", at https://sourceware.org/ml/gdb-patches/2014-03/msg00167.html The testsuite caught them on Linux/x86_64, at least. gdb.sum: gdb.sum: FAIL: gdb.base/hbreak2.exp: next over recursive call FAIL: gdb.base/hbreak2.exp: backtrace from factorial(5.1) FAIL: gdb.base/hbreak2.exp: continue until exit at recursive next test gdb.log: (gdb) next Program received signal SIGTRAP, Trace/breakpoint trap. factorial (value=4) at ../../../src/gdb/testsuite/gdb.base/break.c:113 113 if (value > 1) { /* set breakpoint 7 here */ (gdb) FAIL: gdb.base/hbreak2.exp: next over recursive call Actually, that patch just exposed a latent issue to "breakpoints always-inserted off" mode, not really caused it. After that patch, GDB no longer removes breakpoints at each internal event, thus making some scenarios behave like breakpoint always-inserted on. The bug is easy to trigger with always-inserted on. The issue is that since the target-side breakpoint conditions support, if the stub/server supports evaluating breakpoint conditions on the target side, then GDB is sending duplicate Zx packets to the target without removing them before, and GDBserver is not really expecting that for Z packets other than Z0/z0. E.g., with "set breakpoint always-inserted on" and "set debug remote 1": (gdb) b main Sending packet: $m410943,1#ff...Packet received: 48 Breakpoint 4 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z0,410943,1#48...Packet received: OK ^^^^^^^^^^^^ (gdb) b main Note: breakpoint 4 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Breakpoint 5 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z0,410943,1#48...Packet received: OK ^^^^^^^^^^^^ (gdb) b main Note: breakpoints 4 and 5 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Breakpoint 6 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z0,410943,1#48...Packet received: OK ^^^^^^^^^^^^ (gdb) del Delete all breakpoints? (y or n) y Sending packet: $Z0,410943,1#48...Packet received: OK Sending packet: $Z0,410943,1#48...Packet received: OK Sending packet: $z0,410943,1#68...Packet received: OK And for Z1, similarly: (gdb) hbreak main Sending packet: $m410943,1#ff...Packet received: 48 Hardware assisted breakpoint 4 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ Packet Z1 (hardware-breakpoint) is supported (gdb) hbreak main Note: breakpoint 4 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Hardware assisted breakpoint 5 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ (gdb) hbreak main Note: breakpoints 4 and 5 also set at pc 0x410943. Sending packet: $m410943,1#ff...Packet received: 48 Hardware assisted breakpoint 6 at 0x410943: file ../../../src/gdb/gdbserver/server.c, line 3028. Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ (gdb) del Delete all breakpoints? (y or n) y Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ Sending packet: $Z1,410943,1#49...Packet received: OK ^^^^^^^^^^^^ Sending packet: $z1,410943,1#69...Packet received: OK ^^^^^^^^^^^^ So GDB sent a bunch of Z1 packets, and then when finally removing the breakpoint, only one z1 packet was sent. On the GDBserver side (with monitor set debug-hw-points 1), in the Z1 case, we see: $ ./gdbserver :9999 ./gdbserver Process ./gdbserver created; pid = 8629 Listening on port 9999 Remote debugging from host 127.0.0.1 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=1 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=2 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=3 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=4 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 insert_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=5 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 remove_watchpoint (addr=410943, len=1, type=instruction-execute): CONTROL (DR7): 00000101 STATUS (DR6): 00000000 DR0: addr=0x410943, ref.count=4 DR1: addr=0x0, ref.count=0 DR2: addr=0x0, ref.count=0 DR3: addr=0x0, ref.count=0 That's one insert_watchpoint call for each Z1 packet, and then one remove_watchpoint call for the z1 packet. Notice how ref.count increased for each insert_watchpoint call, and then in the end, after GDB told GDBserver to forget about the hardware breakpoint, GDBserver ends with the the first debug register still with ref.count=4! IOW, the hardware breakpoint is left armed on the target, while on the GDB end it's gone. If the program happens to execute 0x410943 afterwards, then the CPU traps, GDBserver reports the trap to GDB, and GDB not having a breakpoint set at that address anymore, reports to the user a spurious SIGTRAP. This is exactly what is happening in the hbreak2.exp test, though in that case, it's a shared library event that triggers a breakpoint_re_set, when breakpoints are still inserted (because nowadays GDB doesn't remove breakpoints while handling internal events), and that recreates breakpoint locations, which likewise forces breakpoint reinsertion and Zx packet resends... That is a lot of bogus Zx duplication that should possibly be addressed on the GDB side. GDB resends Zx packets because the way to change the target-side condition, is to resend the breakpoint to the server with the new condition. (That's an option in the packet: e.g., "Z1,410943,1;X3,220027" for "hbreak main if 0". The packets in the examples above are shorter because the breakpoints don't have conditions attached). GDB doesn't remove the breakpoint first before reinserting it because that'd be bad for non-stop, as it'd open a window where the inferior could miss the breakpoint. The conditions actually haven't changed between the resends, but GDB isn't smart enough to realize that. (TBC, if the target doesn't support target-side conditions, then GDB doesn't trigger these resends (init_bp_location calls mark_breakpoint_location_modified, and that does nothing if condition evaluation is on the host side. The resends are caused by the 'loc->condition_changed = condition_modified.' line.) But, even if GDB was made smarter, GDBserver should really still handle the resends anyway. So target-side conditions also aren't really to blame. The documentation of the Z/z packets says: "To avoid potential problems with duplicate packets, the operations should be implemented in an idempotent way." As such, we may want to fix GDB, but we should definitely fix GDBserver. The fix is a prerequisite for target-side conditions on hardware breakpoints anyway (and while at it, on watchpoints too). GDBserver indeed already treats duplicate Z0 packets in an idempotent way. mem-break.c has the concept of high-level and low-level breakpoints, somewhat similar to GDB's split of breakpoints vs breakpoint locations, and keeps track of multiple breakpoints referencing the same address/location, for the case of an internal GDBserver breakpoint or a tracepoint being set at the same address as a GDB breakpoint. But, it only allows GDB to ever contribute one reference to a software breakpoint location. IOW, if gdbserver sees a Z0 packet for the same address where it already had a GDB breakpoint set, then GDBserver won't create another high-level GDB breakpoint. However, mem-break.c only tracks GDB Z0 breakpoints. The same logic should apply to all kinds of Zx packets. Currently, gdbserver passes down each duplicate Zx (other than Z0) request directly to the target->insert_point routine. The x86 watchpoint support itself refcounts watchpoint / hw breakpoint requests, to handle overlapping watchpoints, and save debug registers. But that code doesn't (and really shouldn't) handle the duplicate requests, assuming that for each insert there will be a corresponding remove. So the fix is to generalize mem-break.c to track all kinds of Zx breakpoints, and filter out duplicates. As mentioned, this ends up adding support for target-side conditions on hardware breakpoints and watchpoints too (though GDB itself doesn't support the latter yet). Probably the least obvious change in the patch is that it kind of turns the breakpoint insert/remove APIs inside out. Before, the target methods were only called for GDB breakpoints. The internal breakpoint set/delete methods inserted memory breakpoints directly bypassing the insert/remove target methods. That's not good when the target should use a debug API to set software breakpoints, instead of relying on GDBserver patching memory with breakpoint instructions, as is the case of NTO. Now removal/insertion of all kinds of breakpoints/watchpoints, either internal, or from GDB, always go through the target methods. The insert_point/remove_point methods no longer get passed a Z packet type, but an internal/raw breakpoint type. They're also passed a pointer to the raw breakpoint itself (note that's still opaque outside mem-break.c), so that insert_memory_breakpoint / remove_memory_breakpoint have access to the breakpoint's shadow buffer. I first tried passing down a new structure based on GDB's "struct bp_target_info" (actually with that name exactly), but then decided against it as unnecessary complication. As software/memory breakpoints work by poking at memory, when setting a GDB Z0 breakpoint (but not internal breakpoints, as those can assume the conditions are already right), we need to tell the target to prepare to access memory (which on Linux means stop threads). If that operation fails, we need to return error to GDB. Seeing an error, if this is the first breakpoint of that type that GDB tries to insert, GDB would then assume the breakpoint type is supported, but it may actually not be. So we need to check whether the type is supported at all before preparing to access memory. And to solve that, the patch adds a new target->supports_z_point_type method that is called before actually trying to insert the breakpoint. Other than that, hopefully the change is more or less obvious. New test added that exercises the hbreak2.exp regression in a more direct way, without relying on a breakpoint re-set happening before main is reached. Tested by building GDBserver for: aarch64-linux-gnu arm-linux-gnueabihf i686-pc-linux-gnu i686-w64-mingw32 m68k-linux-gnu mips-linux-gnu mips-uclinux nios2-linux-gnu powerpc-linux-gnu sh-linux-gnu tilegx-unknown-linux-gnu x86_64-redhat-linux x86_64-w64-mingw32 And also regression tested on x86_64 Fedora 20. gdb/gdbserver/ 2014-05-20 Pedro Alves <palves@redhat.com> * linux-aarch64-low.c (aarch64_insert_point) (aarch64_remove_point): No longer check whether the type is supported here. Adjust to new interface. (the_low_target): Install aarch64_supports_z_point_type as supports_z_point_type method. * linux-arm-low.c (raw_bkpt_type_to_arm_hwbp_type): New function. (arm_linux_hw_point_initialize): Take an enum raw_bkpt_type instead of a Z packet char. Adjust. (arm_supports_z_point_type): New function. (arm_insert_point, arm_remove_point): Adjust to new interface. (the_low_target): Install arm_supports_z_point_type. * linux-crisv32-low.c (cris_supports_z_point_type): New function. (cris_insert_point, cris_remove_point): Adjust to new interface. Don't check whether the type is supported here. (the_low_target): Install cris_supports_z_point_type. * linux-low.c (linux_supports_z_point_type): New function. (linux_insert_point, linux_remove_point): Adjust to new interface. * linux-low.h (struct linux_target_ops) <insert_point, remove_point>: Take an enum raw_bkpt_type instead of a char. Add raw_breakpoint pointer parameter. <supports_z_point_type>: New method. * linux-mips-low.c (mips_supports_z_point_type): New function. (mips_insert_point, mips_remove_point): Adjust to new interface. Use mips_supports_z_point_type. (the_low_target): Install mips_supports_z_point_type. * linux-ppc-low.c (the_low_target): Install NULL as supports_z_point_type method. * linux-s390-low.c (the_low_target): Install NULL as supports_z_point_type method. * linux-sparc-low.c (the_low_target): Install NULL as supports_z_point_type method. * linux-x86-low.c (x86_supports_z_point_type): New function. (x86_insert_point): Adjust to new insert_point interface. Use insert_memory_breakpoint. Adjust to new i386_low_insert_watchpoint interface. (x86_remove_point): Adjust to remove_point interface. Use remove_memory_breakpoint. Adjust to new i386_low_remove_watchpoint interface. (the_low_target): Install x86_supports_z_point_type. * lynx-low.c (lynx_target_ops): Install NULL as supports_z_point_type callback. * nto-low.c (nto_supports_z_point_type): New. (nto_insert_point, nto_remove_point): Adjust to new interface. (nto_target_ops): Install nto_supports_z_point_type. * mem-break.c: Adjust intro comment. (struct raw_breakpoint) <raw_type, size>: New fields. <inserted>: Update comment. <shlib_disabled>: Delete field. (enum bkpt_type) <gdb_breakpoint>: Delete value. <gdb_breakpoint_Z0, gdb_breakpoint_Z1, gdb_breakpoint_Z2, gdb_breakpoint_Z3, gdb_breakpoint_Z4>: New values. (raw_bkpt_type_to_target_hw_bp_type): New function. (find_enabled_raw_code_breakpoint_at): New function. (find_raw_breakpoint_at): New type and size parameters. Use them. (insert_memory_breakpoint): New function, based off set_raw_breakpoint_at. (remove_memory_breakpoint): New function. (set_raw_breakpoint_at): Reimplement. (set_breakpoint): New, based on set_breakpoint_at. (set_breakpoint_at): Reimplement. (delete_raw_breakpoint): Go through the_target->remove_point instead of assuming memory breakpoints. (find_gdb_breakpoint_at): Delete. (Z_packet_to_bkpt_type, Z_packet_to_raw_bkpt_type): New functions. (find_gdb_breakpoint): New function. (set_gdb_breakpoint_at): Delete. (z_type_supported): New function. (set_gdb_breakpoint_1): New function, loosely based off set_gdb_breakpoint_at. (check_gdb_bp_preconditions, set_gdb_breakpoint): New functions. (delete_gdb_breakpoint_at): Delete. (delete_gdb_breakpoint_1): New function, loosely based off delete_gdb_breakpoint_at. (delete_gdb_breakpoint): New function. (clear_gdb_breakpoint_conditions): Rename to ... (clear_breakpoint_conditions): ... this. Don't handle a NULL breakpoint. (add_condition_to_breakpoint): Make static. (add_breakpoint_condition): Take a struct breakpoint pointer instead of an address. Adjust. (gdb_condition_true_at_breakpoint): Rename to ... (gdb_condition_true_at_breakpoint_z_type): ... this, and add z_type parameter. (gdb_condition_true_at_breakpoint): Reimplement. (add_breakpoint_commands): Take a struct breakpoint pointer instead of an address. Adjust. (gdb_no_commands_at_breakpoint): Rename to ... (gdb_no_commands_at_breakpoint_z_type): ... this. Add z_type parameter. Return true if no breakpoint was found. Change debug output. (gdb_no_commands_at_breakpoint): Reimplement. (run_breakpoint_commands): Rename to ... (run_breakpoint_commands_z_type): ... this. Add z_type parameter, and change return type to boolean. (run_breakpoint_commands): New function. (gdb_breakpoint_here): Also check for Z1 breakpoints. (uninsert_raw_breakpoint): Don't try to reinsert a disabled breakpoint. Go through the_target->remove_point instead of assuming memory breakpoint. (uninsert_breakpoints_at, uninsert_all_breakpoints): Uninsert software and hardware breakpoints. (reinsert_raw_breakpoint): Go through the_target->insert_point instead of assuming memory breakpoint. (reinsert_breakpoints_at, reinsert_all_breakpoints): Reinsert software and hardware breakpoints. (check_breakpoints, breakpoint_here, breakpoint_inserted_here): Check both software and hardware breakpoints. (validate_inserted_breakpoint): Assert the breakpoint is a software breakpoint. Set the inserted flag to -1 instead of setting shlib_disabled. (delete_disabled_breakpoints): Adjust. (validate_breakpoints): Only validate software breakpoints. Adjust to inserted flag change. (check_mem_read, check_mem_write): Skip breakpoint types other than software breakpoints. Adjust to inserted flag change. * mem-break.h (enum raw_bkpt_type): New enum. (raw_breakpoint, struct process_info): Forward declare. (Z_packet_to_target_hw_bp_type): Delete declaration. (raw_bkpt_type_to_target_hw_bp_type, Z_packet_to_raw_bkpt_type) (set_gdb_breakpoint, delete_gdb_breakpoint) (clear_breakpoint_conditions): New declarations. (set_gdb_breakpoint_at, clear_gdb_breakpoint_conditions): Delete. (breakpoint_inserted_here): Update comment. (add_breakpoint_condition, add_breakpoint_commands): Replace address parameter with a breakpoint pointer parameter. (gdb_breakpoint_here): Update comment. (delete_gdb_breakpoint_at): Delete. (insert_memory_breakpoint, remove_memory_breakpoint): Declare. * server.c (process_point_options): Take a struct breakpoint pointer instead of an address. Adjust. (process_serial_event) <Z/z packets>: Use set_gdb_breakpoint and delete_gdb_breakpoint. * spu-low.c (spu_target_ops): Install NULL as supports_z_point_type method. * target.h: Include mem-break.h. (struct target_ops) <prepare_to_access_memory>: Update comment. <supports_z_point_type>: New field. <insert_point, remove_point>: Take an enum raw_bkpt_type argument instead of a char. Also take a raw breakpoint pointer. * win32-arm-low.c (the_low_target): Install NULL as supports_z_point_type. * win32-i386-low.c (i386_supports_z_point_type): New function. (i386_insert_point, i386_remove_point): Adjust to new interface. (the_low_target): Install i386_supports_z_point_type. * win32-low.c (win32_supports_z_point_type): New function. (win32_insert_point, win32_remove_point): Adjust to new interface. (win32_target_ops): Install win32_supports_z_point_type. * win32-low.h (struct win32_target_ops): <supports_z_point_type>: New method. <insert_point, remove_point>: Take an enum raw_bkpt_type argument instead of a char. Also take a raw breakpoint pointer. gdb/testsuite/ 2014-05-20 Pedro Alves <palves@redhat.com> * gdb.base/break-idempotent.c: New file. * gdb.base/break-idempotent.exp: New file.
2014-05-20 19:24:28 +02:00
/* Insert a memory breakpoint. */
int insert_memory_breakpoint (struct raw_breakpoint *bp);
/* Remove a previously inserted memory breakpoint. */
int remove_memory_breakpoint (struct raw_breakpoint *bp);
/* Create a new breakpoint list NEW_BKPT_LIST that is a copy of SRC. */
void clone_all_breakpoints (struct breakpoint **new_bkpt_list,
struct raw_breakpoint **new_raw_bkpt_list,
const struct breakpoint *src);
#endif /* MEM_BREAK_H */