binutils-gdb/gdb/gdbserver/nto-low.c

991 lines
24 KiB
C
Raw Normal View History

/* QNX Neutrino specific low level interface, for the remote server
for GDB.
Copyright (C) 2009-2015 Free Software Foundation, Inc.
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/>. */
#include "server.h"
#include "gdbthread.h"
#include "nto-low.h"
#include "hostio.h"
#include <limits.h>
#include <fcntl.h>
#include <spawn.h>
#include <sys/procfs.h>
#include <sys/auxv.h>
#include <sys/iomgr.h>
#include <sys/neutrino.h>
extern int using_threads;
int using_threads = 1;
[GDBserver] Multi-process + multi-arch This patch makes GDBserver support multi-process + biarch. Currently, if you're debugging more than one process at once with a single gdbserver (in extended-remote mode), then all processes must have the same architecture (e.g., 64-bit vs 32-bit). Otherwise, you see this: Added inferior 2 [Switching to inferior 2 [<null>] (<noexec>)] Reading symbols from /home/pedro/gdb/tests/main32...done. Temporary breakpoint 2 at 0x4004cf: main. (2 locations) Starting program: /home/pedro/gdb/tests/main32 warning: Selected architecture i386 is not compatible with reported target architecture i386:x86-64 warning: Architecture rejected target-supplied description Remote 'g' packet reply is too long: 000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000090cfffff0000000000000000000000000000000000000000000000000000000000020000000000000000000000000000000000000000000000000000000000000000000000000000b042f7460000000000020000230000002b0000002b0000002b000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000007f03000000000000ffff0000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000801f00003b0000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000 ... etc, etc ... Even though the process was running a 32-bit program, GDBserver sent back to GDB a register set in 64-bit layout. A patch (http://sourceware.org/ml/gdb-patches/2012-11/msg00228.html) a while ago made GDB track a target_gdbarch per inferior, and as consequence, fetch a target description per-inferior. This patch is the GDBserver counterpart, that makes GDBserver keep track of each process'es XML target description and register layout. So in the example above, GDBserver will send the correct register set in 32-bit layout to GDB. A new "struct target_desc" object (tdesc for short) is added, that holds the target description and register layout information about each process. Each `struct process_info' holds a pointer to a target description. The regcache also gains a pointer to a target description, mainly for convenience, and parallel with GDB (and possible future support for programs that flip processor modes). The low target's arch_setup routines are responsible for setting the process'es correct tdesc. This isn't that much different to how things were done before, except that instead of detecting the inferior process'es architecture and calling the corresponding init_registers_FOO routine, which would change the regcache layout globals and recreate the threads' regcaches, the regcache.c globals are gone, and the init_registers_$BAR routines now each initialize a separate global struct target_desc object (one for each arch variant GDBserver supports), and so all the init_registers_$BAR routines that are built into GDBserver are called early at GDBserver startup time (similarly to how GDB handles its built-in target descriptions), and then the arch_setup routine is responsible for making process_info->tdesc point to one of these target description globals. The regcache module is all parameterized to get the regcache's layout from the tdesc object instead of the old register_bytes, etc. globals. The threads' regcaches are now created lazily. The old scheme where we created each of them when we added a new thread doesn't work anymore, because we add the main thread/lwp before we see it stop for the first time, and it is only when we see the thread stop for the first time that we have a chance of determining the inferior's architecture (through the_low_target.arch_setup). Therefore when we add the main thread we don't know which architecture/tdesc its regcache should have. This patch makes the gdb.multi/multi-arch.exp test now pass against (extended-remote) GDBserver. It currently fails, without this patch. The IPA also uses the regcache, so it gains a new global struct target_desc pointer, which points at the description of the process it is loaded in. Re. the linux-low.c & friends changes. Since the register map etc. may differ between processes (64-bit vs 32-bit) etc., the linux_target_ops num_regs, regmap and regset_bitmap data fields are no longer sufficient. A new method is added in their place that returns a pointer to a new struct that includes all info linux-low.c needs to access registers of the current inferior. The patch/discussion that originally introduced linux-low.c:disabled_regsets mentions that the disabled_regsets set may be different per mode (in a biarch setup), and indeed that is cleared whenever we start a new (first) inferior, so that global is moved as well behind the new `struct regs_info'. On the x86 side: I simply replaced the i387-fp.c:num_xmm_registers global with a check for 64-bit or 32-bit process, which is equivalent to how the global was set. This avoided coming up with some more general mechanism that would work for all targets that use this module (GNU/Linux, Windows, etc.). Tested: GNU/Linux IA64 GNU/Linux MIPS64 GNU/Linux PowerPC (Fedora 16) GNU/Linux s390x (Fedora 16) GNU/Linux sparc64 (Debian) GNU/Linux x86_64, -m64 and -m32 (Fedora 17) Cross built, and smoke tested: i686-w64-mingw32, under Wine. GNU/Linux TI C6x, by Yao Qi. Cross built but otherwise not tested: aarch64-linux-gnu arm-linux-gnu m68k-linux nios2-linux-gnu sh-linux-gnu spu tilegx-unknown-linux-gnu Completely untested: GNU/Linux Blackfin GNU/Linux CRIS GNU/Linux CRISv32 GNU/Linux TI Xtensa GNU/Linux M32R LynxOS QNX NTO gdb/gdbserver/ 2013-06-07 Pedro Alves <palves@redhat.com> * Makefile.in (OBS): Add tdesc.o. (IPA_OBJS): Add tdesc-ipa.o. (tdesc-ipa.o): New rule. * ax.c (gdb_eval_agent_expr): Adjust register_size call to new interface. * linux-low.c (new_inferior): Delete. (disabled_regsets, num_regsets): Delete. (linux_add_process): Adjust to set the new per-process new_inferior flag. (linux_detach_one_lwp): Adjust to call regcache_invalidate_thread. (linux_wait_for_lwp): Adjust. Only call arch_setup if the event was a stop. When calling arch_setup, switch the current inferior to the thread that got an event. (linux_resume_one_lwp): Adjust to call regcache_invalidate_thread. (regsets_fetch_inferior_registers) (regsets_store_inferior_registers): New regsets_info parameter. Adjust to use it. (linux_register_in_regsets): New regs_info parameter. Adjust to use it. (register_addr, fetch_register, store_register): New usrregs_info parameter. Adjust to use it. (usr_fetch_inferior_registers, usr_store_inferior_registers): New parameter regs_info. Adjust to use it. (linux_fetch_registers): Get the current inferior's regs_info, and adjust to use it. (linux_store_registers): Ditto. [HAVE_LINUX_REGSETS] (initialize_regsets_info): New. (initialize_low): Don't initialize the target_regsets here. Call initialize_low_arch. * linux-low.h (target_regsets): Delete declaration. (struct regsets_info): New. (struct usrregs_info): New. (struct regs_info): New. (struct process_info_private) <new_inferior>: New field. (struct linux_target_ops): Delete the num_regs, regmap, and regset_bitmap fields. New field regs_info. [HAVE_LINUX_REGSETS] (initialize_regsets_info): Declare. * i387-fp.c (num_xmm_registers): Delete. (i387_cache_to_fsave, i387_fsave_to_cache): Adjust find_regno calls to new interface. (i387_cache_to_fxsave, i387_cache_to_xsave, i387_fxsave_to_cache) (i387_xsave_to_cache): Adjust find_regno calls to new interface. Infer the number of xmm registers from the regcache's target description. * i387-fp.h (num_xmm_registers): Delete. * inferiors.c (add_thread): Don't install the thread's regcache here. * proc-service.c (gregset_info): Fetch the current inferior's regs_info. Adjust to use it. * regcache.c: Include tdesc.h. (register_bytes, reg_defs, num_registers) (gdbserver_expedite_regs): Delete. (get_thread_regcache): If the thread doesn't have a regcache yet, create one, instead of aborting gdbserver. (regcache_invalidate_one): Rename to ... (regcache_invalidate_thread): ... this. (regcache_invalidate_one): New. (regcache_invalidate): Only invalidate registers of the current process. (init_register_cache): Add target_desc parameter, and use it. (new_register_cache): Ditto. Assert the target description has a non zero registers_size. (regcache_cpy): Add assertions. Adjust. (realloc_register_cache, set_register_cache): Delete. (registers_to_string, registers_from_string): Adjust. (find_register_by_name, find_regno, find_register_by_number) (register_cache_size): Add target_desc parameter, and use it. (free_register_cache_thread, free_register_cache_thread_one) (regcache_release, register_cache_size): New. (register_size): Add target_desc parameter, and use it. (register_data, supply_register, supply_register_zeroed) (supply_regblock, supply_register_by_name, collect_register) (collect_register_as_string, collect_register_by_name): Adjust. * regcache.h (struct target_desc): Forward declare. (struct regcache) <tdesc>: New field. (init_register_cache, new_register_cache): Add target_desc parameter. (regcache_invalidate_thread): Declare. (regcache_invalidate_one): Delete declaration. (regcache_release): Declare. (find_register_by_number, register_cache_size, register_size) (find_regno): Add target_desc parameter. (gdbserver_expedite_regs, gdbserver_xmltarget): Delete declarations. * remote-utils.c: Include tdesc.h. (outreg, prepare_resume_reply): Adjust. * server.c: Include tdesc.h. (gdbserver_xmltarget): Delete declaration. (get_features_xml, process_serial_event): Adjust. * server.h [IN_PROCESS_AGENT] (struct target_desc): Forward declare. (struct process_info) <tdesc>: New field. (ipa_tdesc): Declare. * tdesc.c: New file. * tdesc.h: New file. * tracepoint.c: Include tdesc.h. [IN_PROCESS_AGENT] (ipa_tdesc): Define. (get_context_regcache): Adjust to pass ipa_tdesc down. (do_action_at_tracepoint): Adjust to get the register cache size from the context regcache's description. (traceframe_walk_blocks): Adjust to get the register cache size from the current trace frame's description. (traceframe_get_pc): Adjust to get current trace frame's description and pass it down. (gdb_collect): Adjust to get the register cache size from the IPA's description. * linux-amd64-ipa.c (tdesc_amd64_linux): Declare. (gdbserver_xmltarget): Delete. (initialize_low_tracepoint): Set the ipa's target description. * linux-i386-ipa.c (tdesc_i386_linux): Declare. (initialize_low_tracepoint): Set the ipa's target description. * linux-x86-low.c: Include tdesc.h. [__x86_64__] (is_64bit_tdesc): New. (ps_get_thread_area, x86_get_thread_area): Use it. (i386_cannot_store_register): Rename to ... (x86_cannot_store_register): ... this. Use is_64bit_tdesc. (i386_cannot_fetch_register): Rename to ... (x86_cannot_fetch_register): ... this. Use is_64bit_tdesc. (x86_fill_gregset, x86_store_gregset): Adjust register_size calls to new interface. (target_regsets): Rename to ... (x86_regsets): ... this. (x86_get_pc, x86_set_pc): Adjust register_size calls to new interface. (x86_siginfo_fixup): Use is_64bit_tdesc. [__x86_64__] (tdesc_amd64_linux, tdesc_amd64_avx_linux) (tdesc_x32_avx_linux, tdesc_x32_linux) (tdesc_i386_linux, tdesc_i386_mmx_linux, tdesc_i386_avx_linux): Declare. (x86_linux_update_xmltarget): Delete. (I386_LINUX_XSAVE_XCR0_OFFSET): Define. (have_ptrace_getfpxregs, have_ptrace_getregset): New. (AMD64_LINUX_USER64_CS): New. (x86_linux_read_description): New, based on x86_linux_update_xmltarget. (same_process_callback): New. (x86_arch_setup_process_callback): New. (x86_linux_update_xmltarget): New. (x86_regsets_info): New. (amd64_linux_regs_info): New. (i386_linux_usrregs_info): New. (i386_linux_regs_info): New. (x86_linux_regs_info): New. (x86_arch_setup): Reimplement. (x86_install_fast_tracepoint_jump_pad): Use is_64bit_tdesc. (x86_emit_ops): Ditto. (the_low_target): Adjust. Install x86_linux_regs_info, x86_cannot_fetch_register, and x86_cannot_store_register. (initialize_low_arch): New. * linux-ia64-low.c (tdesc_ia64): Declare. (ia64_fetch_register): Adjust. (ia64_usrregs_info, regs_info): New globals. (ia64_regs_info): New function. (the_low_target): Adjust. (initialize_low_arch): New function. * linux-sparc-low.c (tdesc_sparc64): Declare. (sparc_fill_gregset_to_stack, sparc_store_gregset_from_stack): Adjust. (sparc_arch_setup): New function. (sparc_regsets_info, sparc_usrregs_info, regs_info): New globals. (the_low_target): Adjust. (initialize_low_arch): New function. * linux-ppc-low.c (tdesc_powerpc_32l, tdesc_powerpc_altivec32l) (tdesc_powerpc_cell32l, tdesc_powerpc_vsx32l) (tdesc_powerpc_isa205_32l, tdesc_powerpc_isa205_altivec32l) (tdesc_powerpc_isa205_vsx32l, tdesc_powerpc_e500l) (tdesc_powerpc_64l, tdesc_powerpc_altivec64l) (tdesc_powerpc_cell64l, tdesc_powerpc_vsx64l) (tdesc_powerpc_isa205_64l, tdesc_powerpc_isa205_altivec64l) (tdesc_powerpc_isa205_vsx64l): Declare. (ppc_cannot_store_register, ppc_collect_ptrace_register) (ppc_supply_ptrace_register, parse_spufs_run, ppc_get_pc) (ppc_set_pc, ppc_get_hwcap): Adjust. (ppc_usrregs_info): Forward declare. (!__powerpc64__) ppc_regmap_adjusted: New global. (ppc_arch_setup): Adjust to the current process'es target description. (ppc_fill_vsxregset, ppc_store_vsxregset, ppc_fill_vrregset) (ppc_store_vrregset, ppc_fill_evrregset, ppc_store_evrregse) (ppc_store_evrregset): Adjust. (target_regsets): Rename to ... (ppc_regsets): ... this, and make static. (ppc_usrregs_info, ppc_regsets_info, regs_info): New globals. (ppc_regs_info): New function. (the_low_target): Adjust. (initialize_low_arch): New function. * linux-s390-low.c (tdesc_s390_linux32, tdesc_s390_linux32v1) (tdesc_s390_linux32v2, tdesc_s390_linux64, tdesc_s390_linux64v1) (tdesc_s390_linux64v2, tdesc_s390x_linux64, tdesc_s390x_linux64v1) (tdesc_s390x_linux64v2): Declare. (s390_collect_ptrace_register, s390_supply_ptrace_register) (s390_fill_gregset, s390_store_last_break): Adjust. (target_regsets): Rename to ... (s390_regsets): ... this, and make static. (s390_get_pc, s390_set_pc): Adjust. (s390_get_hwcap): New target_desc parameter, and use it. [__s390x__] (have_hwcap_s390_high_gprs): New global. (s390_arch_setup): Adjust to set the current process'es target description. Don't adjust the regmap. (s390_usrregs_info, s390_regsets_info, regs_info): New globals. [__s390x__] (s390_usrregs_info_3264, s390_regsets_info_3264) (regs_info_3264): New globals. (s390_regs_info): New function. (the_low_target): Adjust. (initialize_low_arch): New function. * linux-mips-low.c (tdesc_mips_linux, tdesc_mips_dsp_linux) (tdesc_mips64_linux, tdesc_mips64_dsp_linux): Declare. [__mips64] (init_registers_mips_linux) (init_registers_mips_dsp_linux): Delete defines. [__mips64] (tdesc_mips_linux, tdesc_mips_dsp_linux): New defines. (have_dsp): New global. (mips_read_description): New, based on mips_arch_setup. (mips_arch_setup): Reimplement. (get_usrregs_info): New function. (mips_cannot_fetch_register, mips_cannot_store_register) (mips_get_pc, mips_set_pc, mips_fill_gregset, mips_store_gregset) (mips_fill_fpregset, mips_store_fpregset): Adjust. (target_regsets): Rename to ... (mips_regsets): ... this, and make static. (mips_regsets_info, mips_dsp_usrregs_info, mips_usrregs_info) (dsp_regs_info, regs_info): New globals. (mips_regs_info): New function. (the_low_target): Adjust. (initialize_low_arch): New function. * linux-arm-low.c (tdesc_arm, tdesc_arm_with_iwmmxt) (tdesc_arm_with_vfpv2, tdesc_arm_with_vfpv3, tdesc_arm_with_neon): Declare. (arm_fill_vfpregset, arm_store_vfpregset): Adjust. (arm_read_description): New, with bits factored from arm_arch_setup. (arm_arch_setup): Reimplement. (target_regsets): Rename to ... (arm_regsets): ... this, and make static. (arm_regsets_info, arm_usrregs_info, regs_info): New globals. (arm_regs_info): New function. (the_low_target): Adjust. (initialize_low_arch): New function. * linux-m68k-low.c (tdesc_m68k): Declare. (target_regsets): Rename to ... (m68k_regsets): ... this, and make static. (m68k_regsets_info, m68k_usrregs_info, regs_info): New globals. (m68k_regs_info): New function. (m68k_arch_setup): New function. (the_low_target): Adjust. (initialize_low_arch): New function. * linux-sh-low.c (tdesc_sharch): Declare. (target_regsets): Rename to ... (sh_regsets): ... this, and make static. (sh_regsets_info, sh_usrregs_info, regs_info): New globals. (sh_regs_info, sh_arch_setup): New functions. (the_low_target): Adjust. (initialize_low_arch): New function. * linux-bfin-low.c (tdesc_bfin): Declare. (bfin_arch_setup): New function. (bfin_usrregs_info, regs_info): New globals. (bfin_regs_info): New function. (the_low_target): Adjust. (initialize_low_arch): New function. * linux-cris-low.c (tdesc_cris): Declare. (cris_arch_setup): New function. (cris_usrregs_info, regs_info): New globals. (cris_regs_info): New function. (the_low_target): Adjust. (initialize_low_arch): New function. * linux-cris-low.c (tdesc_crisv32): Declare. (cris_arch_setup): New function. (cris_regsets_info, cris_usrregs_info, regs_info): New globals. (cris_regs_info): New function. (the_low_target): Adjust. (initialize_low_arch): New function. * linux-m32r-low.c (tdesc_m32r): Declare. (m32r_arch_setup): New function. (m32r_usrregs_info, regs_info): New globals. (m32r_regs_info): Adjust. (initialize_low_arch): New function. * linux-tic6x-low.c (tdesc_tic6x_c64xp_linux) (tdesc_tic6x_c64x_linux, tdesc_tic6x_c62x_linux): Declare. (tic6x_usrregs_info): Forward declare. (tic6x_read_description): New function, based on ... (tic6x_arch_setup): ... this. Reimplement. (target_regsets): Rename to ... (tic6x_regsets): ... this, and make static. (tic6x_regsets_info, tic6x_usrregs_info, regs_info): New globals. (tic6x_regs_info): New function. (the_low_target): Adjust. (initialize_low_arch): New function. * linux-xtensa-low.c (tdesc_xtensa): Declare. (xtensa_fill_gregset, xtensa_store_gregset): Adjust. (target_regsets): Rename to ... (xtensa_regsets): ... this, and make static. (xtensa_regsets_info, xtensa_usrregs_info, regs_info): New globals. (xtensa_arch_setup, xtensa_regs_info): New functions. (the_low_target): Adjust. (initialize_low_arch): New function. * linux-nios2-low.c (tdesc_nios2_linux): Declare. (nios2_arch_setup): Set the current process'es tdesc. (target_regsets): Rename to ... (nios2_regsets): ... this. (nios2_regsets_info, nios2_usrregs_info, regs_info): New globals. (nios2_regs_info): New function. (the_low_target): Adjust. (initialize_low_arch): New function. * linux-aarch64-low.c (tdesc_aarch64): Declare. (aarch64_arch_setup): Set the current process'es tdesc. (target_regsets): Rename to ... (aarch64_regsets): ... this. (aarch64_regsets_info, aarch64_usrregs_info, regs_info): New globals. (aarch64_regs_info): New function. (the_low_target): Adjust. (initialize_low_arch): New function. * linux-tile-low.c (tdesc_tilegx, tdesc_tilegx32): Declare globals. (target_regsets): Rename to ... (tile_regsets): ... this. (tile_regsets_info, tile_usrregs_info, regs_info): New globals. (tile_regs_info): New function. (tile_arch_setup): Set the current process'es tdesc. (the_low_target): Adjust. (initialize_low_arch): New function. * spu-low.c (tdesc_spu): Declare. (spu_create_inferior, spu_attach): Set the new process'es tdesc. * win32-arm-low.c (tdesc_arm): Declare. (arm_arch_setup): New function. (the_low_target): Install arm_arch_setup instead of init_registers_arm. * win32-i386-low.c (tdesc_i386, tdesc_amd64): Declare. (init_windows_x86): Rename to ... (i386_arch_setup): ... this. Set `win32_tdesc'. (the_low_target): Adjust. * win32-low.c (win32_tdesc): New global. (child_add_thread): Don't create the thread cache here. (do_initial_child_stuff): Set the new process'es tdesc. * win32-low.h (struct target_desc): Forward declare. (win32_tdesc): Declare. * lynx-i386-low.c (tdesc_i386): Declare global. (lynx_i386_arch_setup): Set `lynx_tdesc'. * lynx-low.c (lynx_tdesc): New global. (lynx_add_process): Set the new process'es tdesc. * lynx-low.h (struct target_desc): Forward declare. (lynx_tdesc): Declare global. * lynx-ppc-low.c (tdesc_powerpc_32): Declare global. (lynx_ppc_arch_setup): Set `lynx_tdesc'. * nto-low.c (nto_tdesc): New global. (do_attach): Set the new process'es tdesc. * nto-low.h (struct target_desc): Forward declare. (nto_tdesc): Declare. * nto-x86-low.c (tdesc_i386): Declare. (nto_x86_arch_setup): Set `nto_tdesc'. gdb/ 2013-06-07 Pedro Alves <palves@redhat.com> * regformats/regdat.sh: Output #include tdesc.h. Make globals static. Output a global target description pointer. (init_registers_${name}): Adjust to initialize a target description structure.
2013-06-07 12:46:59 +02:00
const struct target_desc *nto_tdesc;
static void
nto_trace (const char *fmt, ...)
{
va_list arg_list;
if (debug_threads == 0)
return;
fprintf (stderr, "nto:");
va_start (arg_list, fmt);
vfprintf (stderr, fmt, arg_list);
va_end (arg_list);
}
#define TRACE nto_trace
/* Structure holding neutrino specific information about
inferior. */
struct nto_inferior
{
char nto_procfs_path[PATH_MAX];
int ctl_fd;
pid_t pid;
int exit_signo; /* For tracking exit status. */
};
static struct nto_inferior nto_inferior;
static void
init_nto_inferior (struct nto_inferior *nto_inferior)
{
memset (nto_inferior, 0, sizeof (struct nto_inferior));
nto_inferior->ctl_fd = -1;
nto_inferior->pid = -1;
}
static void
do_detach (void)
{
if (nto_inferior.ctl_fd != -1)
{
nto_trace ("Closing fd\n");
close (nto_inferior.ctl_fd);
init_nto_inferior (&nto_inferior);
}
}
/* Set current thread. Return 1 on success, 0 otherwise. */
static int
nto_set_thread (ptid_t ptid)
{
int res = 0;
TRACE ("%s pid: %d tid: %ld\n", __func__, ptid_get_pid (ptid),
ptid_get_lwp (ptid));
if (nto_inferior.ctl_fd != -1
&& !ptid_equal (ptid, null_ptid)
&& !ptid_equal (ptid, minus_one_ptid))
{
pthread_t tid = ptid_get_lwp (ptid);
if (EOK == devctl (nto_inferior.ctl_fd, DCMD_PROC_CURTHREAD, &tid,
sizeof (tid), 0))
res = 1;
else
TRACE ("%s: Error: failed to set current thread\n", __func__);
}
return res;
}
/* This function will determine all alive threads. Note that we do not list
dead but unjoined threads even though they are still in the process' thread
list.
NTO_INFERIOR must not be NULL. */
static void
nto_find_new_threads (struct nto_inferior *nto_inferior)
{
pthread_t tid;
TRACE ("%s pid:%d\n", __func__, nto_inferior->pid);
if (nto_inferior->ctl_fd == -1)
return;
for (tid = 1;; ++tid)
{
procfs_status status;
ptid_t ptid;
int err;
status.tid = tid;
err = devctl (nto_inferior->ctl_fd, DCMD_PROC_TIDSTATUS, &status,
sizeof (status), 0);
if (err != EOK || status.tid == 0)
break;
/* All threads in between are gone. */
while (tid != status.tid || status.state == STATE_DEAD)
{
struct thread_info *ti;
ptid = ptid_build (nto_inferior->pid, tid, 0);
ti = find_thread_ptid (ptid);
if (ti != NULL)
{
TRACE ("Removing thread %d\n", tid);
remove_thread (ti);
}
if (tid == status.tid)
break;
++tid;
}
if (status.state != STATE_DEAD)
{
TRACE ("Adding thread %d\n", tid);
ptid = ptid_build (nto_inferior->pid, tid, 0);
if (!find_thread_ptid (ptid))
add_thread (ptid, NULL);
}
}
}
/* Given pid, open procfs path. */
static pid_t
do_attach (pid_t pid)
{
procfs_status status;
struct sigevent event;
if (nto_inferior.ctl_fd != -1)
{
close (nto_inferior.ctl_fd);
init_nto_inferior (&nto_inferior);
}
xsnprintf (nto_inferior.nto_procfs_path, PATH_MAX - 1, "/proc/%d/as", pid);
nto_inferior.ctl_fd = open (nto_inferior.nto_procfs_path, O_RDWR);
if (nto_inferior.ctl_fd == -1)
{
TRACE ("Failed to open %s\n", nto_inferior.nto_procfs_path);
init_nto_inferior (&nto_inferior);
return -1;
}
if (devctl (nto_inferior.ctl_fd, DCMD_PROC_STOP, &status, sizeof (status), 0)
!= EOK)
{
do_detach ();
return -1;
}
nto_inferior.pid = pid;
/* Define a sigevent for process stopped notification. */
event.sigev_notify = SIGEV_SIGNAL_THREAD;
event.sigev_signo = SIGUSR1;
event.sigev_code = 0;
event.sigev_value.sival_ptr = NULL;
event.sigev_priority = -1;
devctl (nto_inferior.ctl_fd, DCMD_PROC_EVENT, &event, sizeof (event), 0);
if (devctl (nto_inferior.ctl_fd, DCMD_PROC_STATUS, &status, sizeof (status),
0) == EOK
&& (status.flags & _DEBUG_FLAG_STOPPED))
{
ptid_t ptid;
[GDBserver] Multi-process + multi-arch This patch makes GDBserver support multi-process + biarch. Currently, if you're debugging more than one process at once with a single gdbserver (in extended-remote mode), then all processes must have the same architecture (e.g., 64-bit vs 32-bit). Otherwise, you see this: Added inferior 2 [Switching to inferior 2 [<null>] (<noexec>)] Reading symbols from /home/pedro/gdb/tests/main32...done. Temporary breakpoint 2 at 0x4004cf: main. (2 locations) Starting program: /home/pedro/gdb/tests/main32 warning: Selected architecture i386 is not compatible with reported target architecture i386:x86-64 warning: Architecture rejected target-supplied description Remote 'g' packet reply is too long: 000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000090cfffff0000000000000000000000000000000000000000000000000000000000020000000000000000000000000000000000000000000000000000000000000000000000000000b042f7460000000000020000230000002b0000002b0000002b000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000007f03000000000000ffff0000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000801f00003b0000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000 ... etc, etc ... Even though the process was running a 32-bit program, GDBserver sent back to GDB a register set in 64-bit layout. A patch (http://sourceware.org/ml/gdb-patches/2012-11/msg00228.html) a while ago made GDB track a target_gdbarch per inferior, and as consequence, fetch a target description per-inferior. This patch is the GDBserver counterpart, that makes GDBserver keep track of each process'es XML target description and register layout. So in the example above, GDBserver will send the correct register set in 32-bit layout to GDB. A new "struct target_desc" object (tdesc for short) is added, that holds the target description and register layout information about each process. Each `struct process_info' holds a pointer to a target description. The regcache also gains a pointer to a target description, mainly for convenience, and parallel with GDB (and possible future support for programs that flip processor modes). The low target's arch_setup routines are responsible for setting the process'es correct tdesc. This isn't that much different to how things were done before, except that instead of detecting the inferior process'es architecture and calling the corresponding init_registers_FOO routine, which would change the regcache layout globals and recreate the threads' regcaches, the regcache.c globals are gone, and the init_registers_$BAR routines now each initialize a separate global struct target_desc object (one for each arch variant GDBserver supports), and so all the init_registers_$BAR routines that are built into GDBserver are called early at GDBserver startup time (similarly to how GDB handles its built-in target descriptions), and then the arch_setup routine is responsible for making process_info->tdesc point to one of these target description globals. The regcache module is all parameterized to get the regcache's layout from the tdesc object instead of the old register_bytes, etc. globals. The threads' regcaches are now created lazily. The old scheme where we created each of them when we added a new thread doesn't work anymore, because we add the main thread/lwp before we see it stop for the first time, and it is only when we see the thread stop for the first time that we have a chance of determining the inferior's architecture (through the_low_target.arch_setup). Therefore when we add the main thread we don't know which architecture/tdesc its regcache should have. This patch makes the gdb.multi/multi-arch.exp test now pass against (extended-remote) GDBserver. It currently fails, without this patch. The IPA also uses the regcache, so it gains a new global struct target_desc pointer, which points at the description of the process it is loaded in. Re. the linux-low.c & friends changes. Since the register map etc. may differ between processes (64-bit vs 32-bit) etc., the linux_target_ops num_regs, regmap and regset_bitmap data fields are no longer sufficient. A new method is added in their place that returns a pointer to a new struct that includes all info linux-low.c needs to access registers of the current inferior. The patch/discussion that originally introduced linux-low.c:disabled_regsets mentions that the disabled_regsets set may be different per mode (in a biarch setup), and indeed that is cleared whenever we start a new (first) inferior, so that global is moved as well behind the new `struct regs_info'. On the x86 side: I simply replaced the i387-fp.c:num_xmm_registers global with a check for 64-bit or 32-bit process, which is equivalent to how the global was set. This avoided coming up with some more general mechanism that would work for all targets that use this module (GNU/Linux, Windows, etc.). Tested: GNU/Linux IA64 GNU/Linux MIPS64 GNU/Linux PowerPC (Fedora 16) GNU/Linux s390x (Fedora 16) GNU/Linux sparc64 (Debian) GNU/Linux x86_64, -m64 and -m32 (Fedora 17) Cross built, and smoke tested: i686-w64-mingw32, under Wine. GNU/Linux TI C6x, by Yao Qi. Cross built but otherwise not tested: aarch64-linux-gnu arm-linux-gnu m68k-linux nios2-linux-gnu sh-linux-gnu spu tilegx-unknown-linux-gnu Completely untested: GNU/Linux Blackfin GNU/Linux CRIS GNU/Linux CRISv32 GNU/Linux TI Xtensa GNU/Linux M32R LynxOS QNX NTO gdb/gdbserver/ 2013-06-07 Pedro Alves <palves@redhat.com> * Makefile.in (OBS): Add tdesc.o. (IPA_OBJS): Add tdesc-ipa.o. (tdesc-ipa.o): New rule. * ax.c (gdb_eval_agent_expr): Adjust register_size call to new interface. * linux-low.c (new_inferior): Delete. (disabled_regsets, num_regsets): Delete. (linux_add_process): Adjust to set the new per-process new_inferior flag. (linux_detach_one_lwp): Adjust to call regcache_invalidate_thread. (linux_wait_for_lwp): Adjust. Only call arch_setup if the event was a stop. When calling arch_setup, switch the current inferior to the thread that got an event. (linux_resume_one_lwp): Adjust to call regcache_invalidate_thread. (regsets_fetch_inferior_registers) (regsets_store_inferior_registers): New regsets_info parameter. Adjust to use it. (linux_register_in_regsets): New regs_info parameter. Adjust to use it. (register_addr, fetch_register, store_register): New usrregs_info parameter. Adjust to use it. (usr_fetch_inferior_registers, usr_store_inferior_registers): New parameter regs_info. Adjust to use it. (linux_fetch_registers): Get the current inferior's regs_info, and adjust to use it. (linux_store_registers): Ditto. [HAVE_LINUX_REGSETS] (initialize_regsets_info): New. (initialize_low): Don't initialize the target_regsets here. Call initialize_low_arch. * linux-low.h (target_regsets): Delete declaration. (struct regsets_info): New. (struct usrregs_info): New. (struct regs_info): New. (struct process_info_private) <new_inferior>: New field. (struct linux_target_ops): Delete the num_regs, regmap, and regset_bitmap fields. New field regs_info. [HAVE_LINUX_REGSETS] (initialize_regsets_info): Declare. * i387-fp.c (num_xmm_registers): Delete. (i387_cache_to_fsave, i387_fsave_to_cache): Adjust find_regno calls to new interface. (i387_cache_to_fxsave, i387_cache_to_xsave, i387_fxsave_to_cache) (i387_xsave_to_cache): Adjust find_regno calls to new interface. Infer the number of xmm registers from the regcache's target description. * i387-fp.h (num_xmm_registers): Delete. * inferiors.c (add_thread): Don't install the thread's regcache here. * proc-service.c (gregset_info): Fetch the current inferior's regs_info. Adjust to use it. * regcache.c: Include tdesc.h. (register_bytes, reg_defs, num_registers) (gdbserver_expedite_regs): Delete. (get_thread_regcache): If the thread doesn't have a regcache yet, create one, instead of aborting gdbserver. (regcache_invalidate_one): Rename to ... (regcache_invalidate_thread): ... this. (regcache_invalidate_one): New. (regcache_invalidate): Only invalidate registers of the current process. (init_register_cache): Add target_desc parameter, and use it. (new_register_cache): Ditto. Assert the target description has a non zero registers_size. (regcache_cpy): Add assertions. Adjust. (realloc_register_cache, set_register_cache): Delete. (registers_to_string, registers_from_string): Adjust. (find_register_by_name, find_regno, find_register_by_number) (register_cache_size): Add target_desc parameter, and use it. (free_register_cache_thread, free_register_cache_thread_one) (regcache_release, register_cache_size): New. (register_size): Add target_desc parameter, and use it. (register_data, supply_register, supply_register_zeroed) (supply_regblock, supply_register_by_name, collect_register) (collect_register_as_string, collect_register_by_name): Adjust. * regcache.h (struct target_desc): Forward declare. (struct regcache) <tdesc>: New field. (init_register_cache, new_register_cache): Add target_desc parameter. (regcache_invalidate_thread): Declare. (regcache_invalidate_one): Delete declaration. (regcache_release): Declare. (find_register_by_number, register_cache_size, register_size) (find_regno): Add target_desc parameter. (gdbserver_expedite_regs, gdbserver_xmltarget): Delete declarations. * remote-utils.c: Include tdesc.h. (outreg, prepare_resume_reply): Adjust. * server.c: Include tdesc.h. (gdbserver_xmltarget): Delete declaration. (get_features_xml, process_serial_event): Adjust. * server.h [IN_PROCESS_AGENT] (struct target_desc): Forward declare. (struct process_info) <tdesc>: New field. (ipa_tdesc): Declare. * tdesc.c: New file. * tdesc.h: New file. * tracepoint.c: Include tdesc.h. [IN_PROCESS_AGENT] (ipa_tdesc): Define. (get_context_regcache): Adjust to pass ipa_tdesc down. (do_action_at_tracepoint): Adjust to get the register cache size from the context regcache's description. (traceframe_walk_blocks): Adjust to get the register cache size from the current trace frame's description. (traceframe_get_pc): Adjust to get current trace frame's description and pass it down. (gdb_collect): Adjust to get the register cache size from the IPA's description. * linux-amd64-ipa.c (tdesc_amd64_linux): Declare. (gdbserver_xmltarget): Delete. (initialize_low_tracepoint): Set the ipa's target description. * linux-i386-ipa.c (tdesc_i386_linux): Declare. (initialize_low_tracepoint): Set the ipa's target description. * linux-x86-low.c: Include tdesc.h. [__x86_64__] (is_64bit_tdesc): New. (ps_get_thread_area, x86_get_thread_area): Use it. (i386_cannot_store_register): Rename to ... (x86_cannot_store_register): ... this. Use is_64bit_tdesc. (i386_cannot_fetch_register): Rename to ... (x86_cannot_fetch_register): ... this. Use is_64bit_tdesc. (x86_fill_gregset, x86_store_gregset): Adjust register_size calls to new interface. (target_regsets): Rename to ... (x86_regsets): ... this. (x86_get_pc, x86_set_pc): Adjust register_size calls to new interface. (x86_siginfo_fixup): Use is_64bit_tdesc. [__x86_64__] (tdesc_amd64_linux, tdesc_amd64_avx_linux) (tdesc_x32_avx_linux, tdesc_x32_linux) (tdesc_i386_linux, tdesc_i386_mmx_linux, tdesc_i386_avx_linux): Declare. (x86_linux_update_xmltarget): Delete. (I386_LINUX_XSAVE_XCR0_OFFSET): Define. (have_ptrace_getfpxregs, have_ptrace_getregset): New. (AMD64_LINUX_USER64_CS): New. (x86_linux_read_description): New, based on x86_linux_update_xmltarget. (same_process_callback): New. (x86_arch_setup_process_callback): New. (x86_linux_update_xmltarget): New. (x86_regsets_info): New. (amd64_linux_regs_info): New. (i386_linux_usrregs_info): New. (i386_linux_regs_info): New. (x86_linux_regs_info): New. (x86_arch_setup): Reimplement. (x86_install_fast_tracepoint_jump_pad): Use is_64bit_tdesc. (x86_emit_ops): Ditto. (the_low_target): Adjust. Install x86_linux_regs_info, x86_cannot_fetch_register, and x86_cannot_store_register. (initialize_low_arch): New. * linux-ia64-low.c (tdesc_ia64): Declare. (ia64_fetch_register): Adjust. (ia64_usrregs_info, regs_info): New globals. (ia64_regs_info): New function. (the_low_target): Adjust. (initialize_low_arch): New function. * linux-sparc-low.c (tdesc_sparc64): Declare. (sparc_fill_gregset_to_stack, sparc_store_gregset_from_stack): Adjust. (sparc_arch_setup): New function. (sparc_regsets_info, sparc_usrregs_info, regs_info): New globals. (the_low_target): Adjust. (initialize_low_arch): New function. * linux-ppc-low.c (tdesc_powerpc_32l, tdesc_powerpc_altivec32l) (tdesc_powerpc_cell32l, tdesc_powerpc_vsx32l) (tdesc_powerpc_isa205_32l, tdesc_powerpc_isa205_altivec32l) (tdesc_powerpc_isa205_vsx32l, tdesc_powerpc_e500l) (tdesc_powerpc_64l, tdesc_powerpc_altivec64l) (tdesc_powerpc_cell64l, tdesc_powerpc_vsx64l) (tdesc_powerpc_isa205_64l, tdesc_powerpc_isa205_altivec64l) (tdesc_powerpc_isa205_vsx64l): Declare. (ppc_cannot_store_register, ppc_collect_ptrace_register) (ppc_supply_ptrace_register, parse_spufs_run, ppc_get_pc) (ppc_set_pc, ppc_get_hwcap): Adjust. (ppc_usrregs_info): Forward declare. (!__powerpc64__) ppc_regmap_adjusted: New global. (ppc_arch_setup): Adjust to the current process'es target description. (ppc_fill_vsxregset, ppc_store_vsxregset, ppc_fill_vrregset) (ppc_store_vrregset, ppc_fill_evrregset, ppc_store_evrregse) (ppc_store_evrregset): Adjust. (target_regsets): Rename to ... (ppc_regsets): ... this, and make static. (ppc_usrregs_info, ppc_regsets_info, regs_info): New globals. (ppc_regs_info): New function. (the_low_target): Adjust. (initialize_low_arch): New function. * linux-s390-low.c (tdesc_s390_linux32, tdesc_s390_linux32v1) (tdesc_s390_linux32v2, tdesc_s390_linux64, tdesc_s390_linux64v1) (tdesc_s390_linux64v2, tdesc_s390x_linux64, tdesc_s390x_linux64v1) (tdesc_s390x_linux64v2): Declare. (s390_collect_ptrace_register, s390_supply_ptrace_register) (s390_fill_gregset, s390_store_last_break): Adjust. (target_regsets): Rename to ... (s390_regsets): ... this, and make static. (s390_get_pc, s390_set_pc): Adjust. (s390_get_hwcap): New target_desc parameter, and use it. [__s390x__] (have_hwcap_s390_high_gprs): New global. (s390_arch_setup): Adjust to set the current process'es target description. Don't adjust the regmap. (s390_usrregs_info, s390_regsets_info, regs_info): New globals. [__s390x__] (s390_usrregs_info_3264, s390_regsets_info_3264) (regs_info_3264): New globals. (s390_regs_info): New function. (the_low_target): Adjust. (initialize_low_arch): New function. * linux-mips-low.c (tdesc_mips_linux, tdesc_mips_dsp_linux) (tdesc_mips64_linux, tdesc_mips64_dsp_linux): Declare. [__mips64] (init_registers_mips_linux) (init_registers_mips_dsp_linux): Delete defines. [__mips64] (tdesc_mips_linux, tdesc_mips_dsp_linux): New defines. (have_dsp): New global. (mips_read_description): New, based on mips_arch_setup. (mips_arch_setup): Reimplement. (get_usrregs_info): New function. (mips_cannot_fetch_register, mips_cannot_store_register) (mips_get_pc, mips_set_pc, mips_fill_gregset, mips_store_gregset) (mips_fill_fpregset, mips_store_fpregset): Adjust. (target_regsets): Rename to ... (mips_regsets): ... this, and make static. (mips_regsets_info, mips_dsp_usrregs_info, mips_usrregs_info) (dsp_regs_info, regs_info): New globals. (mips_regs_info): New function. (the_low_target): Adjust. (initialize_low_arch): New function. * linux-arm-low.c (tdesc_arm, tdesc_arm_with_iwmmxt) (tdesc_arm_with_vfpv2, tdesc_arm_with_vfpv3, tdesc_arm_with_neon): Declare. (arm_fill_vfpregset, arm_store_vfpregset): Adjust. (arm_read_description): New, with bits factored from arm_arch_setup. (arm_arch_setup): Reimplement. (target_regsets): Rename to ... (arm_regsets): ... this, and make static. (arm_regsets_info, arm_usrregs_info, regs_info): New globals. (arm_regs_info): New function. (the_low_target): Adjust. (initialize_low_arch): New function. * linux-m68k-low.c (tdesc_m68k): Declare. (target_regsets): Rename to ... (m68k_regsets): ... this, and make static. (m68k_regsets_info, m68k_usrregs_info, regs_info): New globals. (m68k_regs_info): New function. (m68k_arch_setup): New function. (the_low_target): Adjust. (initialize_low_arch): New function. * linux-sh-low.c (tdesc_sharch): Declare. (target_regsets): Rename to ... (sh_regsets): ... this, and make static. (sh_regsets_info, sh_usrregs_info, regs_info): New globals. (sh_regs_info, sh_arch_setup): New functions. (the_low_target): Adjust. (initialize_low_arch): New function. * linux-bfin-low.c (tdesc_bfin): Declare. (bfin_arch_setup): New function. (bfin_usrregs_info, regs_info): New globals. (bfin_regs_info): New function. (the_low_target): Adjust. (initialize_low_arch): New function. * linux-cris-low.c (tdesc_cris): Declare. (cris_arch_setup): New function. (cris_usrregs_info, regs_info): New globals. (cris_regs_info): New function. (the_low_target): Adjust. (initialize_low_arch): New function. * linux-cris-low.c (tdesc_crisv32): Declare. (cris_arch_setup): New function. (cris_regsets_info, cris_usrregs_info, regs_info): New globals. (cris_regs_info): New function. (the_low_target): Adjust. (initialize_low_arch): New function. * linux-m32r-low.c (tdesc_m32r): Declare. (m32r_arch_setup): New function. (m32r_usrregs_info, regs_info): New globals. (m32r_regs_info): Adjust. (initialize_low_arch): New function. * linux-tic6x-low.c (tdesc_tic6x_c64xp_linux) (tdesc_tic6x_c64x_linux, tdesc_tic6x_c62x_linux): Declare. (tic6x_usrregs_info): Forward declare. (tic6x_read_description): New function, based on ... (tic6x_arch_setup): ... this. Reimplement. (target_regsets): Rename to ... (tic6x_regsets): ... this, and make static. (tic6x_regsets_info, tic6x_usrregs_info, regs_info): New globals. (tic6x_regs_info): New function. (the_low_target): Adjust. (initialize_low_arch): New function. * linux-xtensa-low.c (tdesc_xtensa): Declare. (xtensa_fill_gregset, xtensa_store_gregset): Adjust. (target_regsets): Rename to ... (xtensa_regsets): ... this, and make static. (xtensa_regsets_info, xtensa_usrregs_info, regs_info): New globals. (xtensa_arch_setup, xtensa_regs_info): New functions. (the_low_target): Adjust. (initialize_low_arch): New function. * linux-nios2-low.c (tdesc_nios2_linux): Declare. (nios2_arch_setup): Set the current process'es tdesc. (target_regsets): Rename to ... (nios2_regsets): ... this. (nios2_regsets_info, nios2_usrregs_info, regs_info): New globals. (nios2_regs_info): New function. (the_low_target): Adjust. (initialize_low_arch): New function. * linux-aarch64-low.c (tdesc_aarch64): Declare. (aarch64_arch_setup): Set the current process'es tdesc. (target_regsets): Rename to ... (aarch64_regsets): ... this. (aarch64_regsets_info, aarch64_usrregs_info, regs_info): New globals. (aarch64_regs_info): New function. (the_low_target): Adjust. (initialize_low_arch): New function. * linux-tile-low.c (tdesc_tilegx, tdesc_tilegx32): Declare globals. (target_regsets): Rename to ... (tile_regsets): ... this. (tile_regsets_info, tile_usrregs_info, regs_info): New globals. (tile_regs_info): New function. (tile_arch_setup): Set the current process'es tdesc. (the_low_target): Adjust. (initialize_low_arch): New function. * spu-low.c (tdesc_spu): Declare. (spu_create_inferior, spu_attach): Set the new process'es tdesc. * win32-arm-low.c (tdesc_arm): Declare. (arm_arch_setup): New function. (the_low_target): Install arm_arch_setup instead of init_registers_arm. * win32-i386-low.c (tdesc_i386, tdesc_amd64): Declare. (init_windows_x86): Rename to ... (i386_arch_setup): ... this. Set `win32_tdesc'. (the_low_target): Adjust. * win32-low.c (win32_tdesc): New global. (child_add_thread): Don't create the thread cache here. (do_initial_child_stuff): Set the new process'es tdesc. * win32-low.h (struct target_desc): Forward declare. (win32_tdesc): Declare. * lynx-i386-low.c (tdesc_i386): Declare global. (lynx_i386_arch_setup): Set `lynx_tdesc'. * lynx-low.c (lynx_tdesc): New global. (lynx_add_process): Set the new process'es tdesc. * lynx-low.h (struct target_desc): Forward declare. (lynx_tdesc): Declare global. * lynx-ppc-low.c (tdesc_powerpc_32): Declare global. (lynx_ppc_arch_setup): Set `lynx_tdesc'. * nto-low.c (nto_tdesc): New global. (do_attach): Set the new process'es tdesc. * nto-low.h (struct target_desc): Forward declare. (nto_tdesc): Declare. * nto-x86-low.c (tdesc_i386): Declare. (nto_x86_arch_setup): Set `nto_tdesc'. gdb/ 2013-06-07 Pedro Alves <palves@redhat.com> * regformats/regdat.sh: Output #include tdesc.h. Make globals static. Output a global target description pointer. (init_registers_${name}): Adjust to initialize a target description structure.
2013-06-07 12:46:59 +02:00
struct process_info *proc;
kill (pid, SIGCONT);
ptid = ptid_build (status.pid, status.tid, 0);
the_low_target.arch_setup ();
[GDBserver] Multi-process + multi-arch This patch makes GDBserver support multi-process + biarch. Currently, if you're debugging more than one process at once with a single gdbserver (in extended-remote mode), then all processes must have the same architecture (e.g., 64-bit vs 32-bit). Otherwise, you see this: Added inferior 2 [Switching to inferior 2 [<null>] (<noexec>)] Reading symbols from /home/pedro/gdb/tests/main32...done. Temporary breakpoint 2 at 0x4004cf: main. (2 locations) Starting program: /home/pedro/gdb/tests/main32 warning: Selected architecture i386 is not compatible with reported target architecture i386:x86-64 warning: Architecture rejected target-supplied description Remote 'g' packet reply is too long: 000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000090cfffff0000000000000000000000000000000000000000000000000000000000020000000000000000000000000000000000000000000000000000000000000000000000000000b042f7460000000000020000230000002b0000002b0000002b000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000007f03000000000000ffff0000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000801f00003b0000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000 ... etc, etc ... Even though the process was running a 32-bit program, GDBserver sent back to GDB a register set in 64-bit layout. A patch (http://sourceware.org/ml/gdb-patches/2012-11/msg00228.html) a while ago made GDB track a target_gdbarch per inferior, and as consequence, fetch a target description per-inferior. This patch is the GDBserver counterpart, that makes GDBserver keep track of each process'es XML target description and register layout. So in the example above, GDBserver will send the correct register set in 32-bit layout to GDB. A new "struct target_desc" object (tdesc for short) is added, that holds the target description and register layout information about each process. Each `struct process_info' holds a pointer to a target description. The regcache also gains a pointer to a target description, mainly for convenience, and parallel with GDB (and possible future support for programs that flip processor modes). The low target's arch_setup routines are responsible for setting the process'es correct tdesc. This isn't that much different to how things were done before, except that instead of detecting the inferior process'es architecture and calling the corresponding init_registers_FOO routine, which would change the regcache layout globals and recreate the threads' regcaches, the regcache.c globals are gone, and the init_registers_$BAR routines now each initialize a separate global struct target_desc object (one for each arch variant GDBserver supports), and so all the init_registers_$BAR routines that are built into GDBserver are called early at GDBserver startup time (similarly to how GDB handles its built-in target descriptions), and then the arch_setup routine is responsible for making process_info->tdesc point to one of these target description globals. The regcache module is all parameterized to get the regcache's layout from the tdesc object instead of the old register_bytes, etc. globals. The threads' regcaches are now created lazily. The old scheme where we created each of them when we added a new thread doesn't work anymore, because we add the main thread/lwp before we see it stop for the first time, and it is only when we see the thread stop for the first time that we have a chance of determining the inferior's architecture (through the_low_target.arch_setup). Therefore when we add the main thread we don't know which architecture/tdesc its regcache should have. This patch makes the gdb.multi/multi-arch.exp test now pass against (extended-remote) GDBserver. It currently fails, without this patch. The IPA also uses the regcache, so it gains a new global struct target_desc pointer, which points at the description of the process it is loaded in. Re. the linux-low.c & friends changes. Since the register map etc. may differ between processes (64-bit vs 32-bit) etc., the linux_target_ops num_regs, regmap and regset_bitmap data fields are no longer sufficient. A new method is added in their place that returns a pointer to a new struct that includes all info linux-low.c needs to access registers of the current inferior. The patch/discussion that originally introduced linux-low.c:disabled_regsets mentions that the disabled_regsets set may be different per mode (in a biarch setup), and indeed that is cleared whenever we start a new (first) inferior, so that global is moved as well behind the new `struct regs_info'. On the x86 side: I simply replaced the i387-fp.c:num_xmm_registers global with a check for 64-bit or 32-bit process, which is equivalent to how the global was set. This avoided coming up with some more general mechanism that would work for all targets that use this module (GNU/Linux, Windows, etc.). Tested: GNU/Linux IA64 GNU/Linux MIPS64 GNU/Linux PowerPC (Fedora 16) GNU/Linux s390x (Fedora 16) GNU/Linux sparc64 (Debian) GNU/Linux x86_64, -m64 and -m32 (Fedora 17) Cross built, and smoke tested: i686-w64-mingw32, under Wine. GNU/Linux TI C6x, by Yao Qi. Cross built but otherwise not tested: aarch64-linux-gnu arm-linux-gnu m68k-linux nios2-linux-gnu sh-linux-gnu spu tilegx-unknown-linux-gnu Completely untested: GNU/Linux Blackfin GNU/Linux CRIS GNU/Linux CRISv32 GNU/Linux TI Xtensa GNU/Linux M32R LynxOS QNX NTO gdb/gdbserver/ 2013-06-07 Pedro Alves <palves@redhat.com> * Makefile.in (OBS): Add tdesc.o. (IPA_OBJS): Add tdesc-ipa.o. (tdesc-ipa.o): New rule. * ax.c (gdb_eval_agent_expr): Adjust register_size call to new interface. * linux-low.c (new_inferior): Delete. (disabled_regsets, num_regsets): Delete. (linux_add_process): Adjust to set the new per-process new_inferior flag. (linux_detach_one_lwp): Adjust to call regcache_invalidate_thread. (linux_wait_for_lwp): Adjust. Only call arch_setup if the event was a stop. When calling arch_setup, switch the current inferior to the thread that got an event. (linux_resume_one_lwp): Adjust to call regcache_invalidate_thread. (regsets_fetch_inferior_registers) (regsets_store_inferior_registers): New regsets_info parameter. Adjust to use it. (linux_register_in_regsets): New regs_info parameter. Adjust to use it. (register_addr, fetch_register, store_register): New usrregs_info parameter. Adjust to use it. (usr_fetch_inferior_registers, usr_store_inferior_registers): New parameter regs_info. Adjust to use it. (linux_fetch_registers): Get the current inferior's regs_info, and adjust to use it. (linux_store_registers): Ditto. [HAVE_LINUX_REGSETS] (initialize_regsets_info): New. (initialize_low): Don't initialize the target_regsets here. Call initialize_low_arch. * linux-low.h (target_regsets): Delete declaration. (struct regsets_info): New. (struct usrregs_info): New. (struct regs_info): New. (struct process_info_private) <new_inferior>: New field. (struct linux_target_ops): Delete the num_regs, regmap, and regset_bitmap fields. New field regs_info. [HAVE_LINUX_REGSETS] (initialize_regsets_info): Declare. * i387-fp.c (num_xmm_registers): Delete. (i387_cache_to_fsave, i387_fsave_to_cache): Adjust find_regno calls to new interface. (i387_cache_to_fxsave, i387_cache_to_xsave, i387_fxsave_to_cache) (i387_xsave_to_cache): Adjust find_regno calls to new interface. Infer the number of xmm registers from the regcache's target description. * i387-fp.h (num_xmm_registers): Delete. * inferiors.c (add_thread): Don't install the thread's regcache here. * proc-service.c (gregset_info): Fetch the current inferior's regs_info. Adjust to use it. * regcache.c: Include tdesc.h. (register_bytes, reg_defs, num_registers) (gdbserver_expedite_regs): Delete. (get_thread_regcache): If the thread doesn't have a regcache yet, create one, instead of aborting gdbserver. (regcache_invalidate_one): Rename to ... (regcache_invalidate_thread): ... this. (regcache_invalidate_one): New. (regcache_invalidate): Only invalidate registers of the current process. (init_register_cache): Add target_desc parameter, and use it. (new_register_cache): Ditto. Assert the target description has a non zero registers_size. (regcache_cpy): Add assertions. Adjust. (realloc_register_cache, set_register_cache): Delete. (registers_to_string, registers_from_string): Adjust. (find_register_by_name, find_regno, find_register_by_number) (register_cache_size): Add target_desc parameter, and use it. (free_register_cache_thread, free_register_cache_thread_one) (regcache_release, register_cache_size): New. (register_size): Add target_desc parameter, and use it. (register_data, supply_register, supply_register_zeroed) (supply_regblock, supply_register_by_name, collect_register) (collect_register_as_string, collect_register_by_name): Adjust. * regcache.h (struct target_desc): Forward declare. (struct regcache) <tdesc>: New field. (init_register_cache, new_register_cache): Add target_desc parameter. (regcache_invalidate_thread): Declare. (regcache_invalidate_one): Delete declaration. (regcache_release): Declare. (find_register_by_number, register_cache_size, register_size) (find_regno): Add target_desc parameter. (gdbserver_expedite_regs, gdbserver_xmltarget): Delete declarations. * remote-utils.c: Include tdesc.h. (outreg, prepare_resume_reply): Adjust. * server.c: Include tdesc.h. (gdbserver_xmltarget): Delete declaration. (get_features_xml, process_serial_event): Adjust. * server.h [IN_PROCESS_AGENT] (struct target_desc): Forward declare. (struct process_info) <tdesc>: New field. (ipa_tdesc): Declare. * tdesc.c: New file. * tdesc.h: New file. * tracepoint.c: Include tdesc.h. [IN_PROCESS_AGENT] (ipa_tdesc): Define. (get_context_regcache): Adjust to pass ipa_tdesc down. (do_action_at_tracepoint): Adjust to get the register cache size from the context regcache's description. (traceframe_walk_blocks): Adjust to get the register cache size from the current trace frame's description. (traceframe_get_pc): Adjust to get current trace frame's description and pass it down. (gdb_collect): Adjust to get the register cache size from the IPA's description. * linux-amd64-ipa.c (tdesc_amd64_linux): Declare. (gdbserver_xmltarget): Delete. (initialize_low_tracepoint): Set the ipa's target description. * linux-i386-ipa.c (tdesc_i386_linux): Declare. (initialize_low_tracepoint): Set the ipa's target description. * linux-x86-low.c: Include tdesc.h. [__x86_64__] (is_64bit_tdesc): New. (ps_get_thread_area, x86_get_thread_area): Use it. (i386_cannot_store_register): Rename to ... (x86_cannot_store_register): ... this. Use is_64bit_tdesc. (i386_cannot_fetch_register): Rename to ... (x86_cannot_fetch_register): ... this. Use is_64bit_tdesc. (x86_fill_gregset, x86_store_gregset): Adjust register_size calls to new interface. (target_regsets): Rename to ... (x86_regsets): ... this. (x86_get_pc, x86_set_pc): Adjust register_size calls to new interface. (x86_siginfo_fixup): Use is_64bit_tdesc. [__x86_64__] (tdesc_amd64_linux, tdesc_amd64_avx_linux) (tdesc_x32_avx_linux, tdesc_x32_linux) (tdesc_i386_linux, tdesc_i386_mmx_linux, tdesc_i386_avx_linux): Declare. (x86_linux_update_xmltarget): Delete. (I386_LINUX_XSAVE_XCR0_OFFSET): Define. (have_ptrace_getfpxregs, have_ptrace_getregset): New. (AMD64_LINUX_USER64_CS): New. (x86_linux_read_description): New, based on x86_linux_update_xmltarget. (same_process_callback): New. (x86_arch_setup_process_callback): New. (x86_linux_update_xmltarget): New. (x86_regsets_info): New. (amd64_linux_regs_info): New. (i386_linux_usrregs_info): New. (i386_linux_regs_info): New. (x86_linux_regs_info): New. (x86_arch_setup): Reimplement. (x86_install_fast_tracepoint_jump_pad): Use is_64bit_tdesc. (x86_emit_ops): Ditto. (the_low_target): Adjust. Install x86_linux_regs_info, x86_cannot_fetch_register, and x86_cannot_store_register. (initialize_low_arch): New. * linux-ia64-low.c (tdesc_ia64): Declare. (ia64_fetch_register): Adjust. (ia64_usrregs_info, regs_info): New globals. (ia64_regs_info): New function. (the_low_target): Adjust. (initialize_low_arch): New function. * linux-sparc-low.c (tdesc_sparc64): Declare. (sparc_fill_gregset_to_stack, sparc_store_gregset_from_stack): Adjust. (sparc_arch_setup): New function. (sparc_regsets_info, sparc_usrregs_info, regs_info): New globals. (the_low_target): Adjust. (initialize_low_arch): New function. * linux-ppc-low.c (tdesc_powerpc_32l, tdesc_powerpc_altivec32l) (tdesc_powerpc_cell32l, tdesc_powerpc_vsx32l) (tdesc_powerpc_isa205_32l, tdesc_powerpc_isa205_altivec32l) (tdesc_powerpc_isa205_vsx32l, tdesc_powerpc_e500l) (tdesc_powerpc_64l, tdesc_powerpc_altivec64l) (tdesc_powerpc_cell64l, tdesc_powerpc_vsx64l) (tdesc_powerpc_isa205_64l, tdesc_powerpc_isa205_altivec64l) (tdesc_powerpc_isa205_vsx64l): Declare. (ppc_cannot_store_register, ppc_collect_ptrace_register) (ppc_supply_ptrace_register, parse_spufs_run, ppc_get_pc) (ppc_set_pc, ppc_get_hwcap): Adjust. (ppc_usrregs_info): Forward declare. (!__powerpc64__) ppc_regmap_adjusted: New global. (ppc_arch_setup): Adjust to the current process'es target description. (ppc_fill_vsxregset, ppc_store_vsxregset, ppc_fill_vrregset) (ppc_store_vrregset, ppc_fill_evrregset, ppc_store_evrregse) (ppc_store_evrregset): Adjust. (target_regsets): Rename to ... (ppc_regsets): ... this, and make static. (ppc_usrregs_info, ppc_regsets_info, regs_info): New globals. (ppc_regs_info): New function. (the_low_target): Adjust. (initialize_low_arch): New function. * linux-s390-low.c (tdesc_s390_linux32, tdesc_s390_linux32v1) (tdesc_s390_linux32v2, tdesc_s390_linux64, tdesc_s390_linux64v1) (tdesc_s390_linux64v2, tdesc_s390x_linux64, tdesc_s390x_linux64v1) (tdesc_s390x_linux64v2): Declare. (s390_collect_ptrace_register, s390_supply_ptrace_register) (s390_fill_gregset, s390_store_last_break): Adjust. (target_regsets): Rename to ... (s390_regsets): ... this, and make static. (s390_get_pc, s390_set_pc): Adjust. (s390_get_hwcap): New target_desc parameter, and use it. [__s390x__] (have_hwcap_s390_high_gprs): New global. (s390_arch_setup): Adjust to set the current process'es target description. Don't adjust the regmap. (s390_usrregs_info, s390_regsets_info, regs_info): New globals. [__s390x__] (s390_usrregs_info_3264, s390_regsets_info_3264) (regs_info_3264): New globals. (s390_regs_info): New function. (the_low_target): Adjust. (initialize_low_arch): New function. * linux-mips-low.c (tdesc_mips_linux, tdesc_mips_dsp_linux) (tdesc_mips64_linux, tdesc_mips64_dsp_linux): Declare. [__mips64] (init_registers_mips_linux) (init_registers_mips_dsp_linux): Delete defines. [__mips64] (tdesc_mips_linux, tdesc_mips_dsp_linux): New defines. (have_dsp): New global. (mips_read_description): New, based on mips_arch_setup. (mips_arch_setup): Reimplement. (get_usrregs_info): New function. (mips_cannot_fetch_register, mips_cannot_store_register) (mips_get_pc, mips_set_pc, mips_fill_gregset, mips_store_gregset) (mips_fill_fpregset, mips_store_fpregset): Adjust. (target_regsets): Rename to ... (mips_regsets): ... this, and make static. (mips_regsets_info, mips_dsp_usrregs_info, mips_usrregs_info) (dsp_regs_info, regs_info): New globals. (mips_regs_info): New function. (the_low_target): Adjust. (initialize_low_arch): New function. * linux-arm-low.c (tdesc_arm, tdesc_arm_with_iwmmxt) (tdesc_arm_with_vfpv2, tdesc_arm_with_vfpv3, tdesc_arm_with_neon): Declare. (arm_fill_vfpregset, arm_store_vfpregset): Adjust. (arm_read_description): New, with bits factored from arm_arch_setup. (arm_arch_setup): Reimplement. (target_regsets): Rename to ... (arm_regsets): ... this, and make static. (arm_regsets_info, arm_usrregs_info, regs_info): New globals. (arm_regs_info): New function. (the_low_target): Adjust. (initialize_low_arch): New function. * linux-m68k-low.c (tdesc_m68k): Declare. (target_regsets): Rename to ... (m68k_regsets): ... this, and make static. (m68k_regsets_info, m68k_usrregs_info, regs_info): New globals. (m68k_regs_info): New function. (m68k_arch_setup): New function. (the_low_target): Adjust. (initialize_low_arch): New function. * linux-sh-low.c (tdesc_sharch): Declare. (target_regsets): Rename to ... (sh_regsets): ... this, and make static. (sh_regsets_info, sh_usrregs_info, regs_info): New globals. (sh_regs_info, sh_arch_setup): New functions. (the_low_target): Adjust. (initialize_low_arch): New function. * linux-bfin-low.c (tdesc_bfin): Declare. (bfin_arch_setup): New function. (bfin_usrregs_info, regs_info): New globals. (bfin_regs_info): New function. (the_low_target): Adjust. (initialize_low_arch): New function. * linux-cris-low.c (tdesc_cris): Declare. (cris_arch_setup): New function. (cris_usrregs_info, regs_info): New globals. (cris_regs_info): New function. (the_low_target): Adjust. (initialize_low_arch): New function. * linux-cris-low.c (tdesc_crisv32): Declare. (cris_arch_setup): New function. (cris_regsets_info, cris_usrregs_info, regs_info): New globals. (cris_regs_info): New function. (the_low_target): Adjust. (initialize_low_arch): New function. * linux-m32r-low.c (tdesc_m32r): Declare. (m32r_arch_setup): New function. (m32r_usrregs_info, regs_info): New globals. (m32r_regs_info): Adjust. (initialize_low_arch): New function. * linux-tic6x-low.c (tdesc_tic6x_c64xp_linux) (tdesc_tic6x_c64x_linux, tdesc_tic6x_c62x_linux): Declare. (tic6x_usrregs_info): Forward declare. (tic6x_read_description): New function, based on ... (tic6x_arch_setup): ... this. Reimplement. (target_regsets): Rename to ... (tic6x_regsets): ... this, and make static. (tic6x_regsets_info, tic6x_usrregs_info, regs_info): New globals. (tic6x_regs_info): New function. (the_low_target): Adjust. (initialize_low_arch): New function. * linux-xtensa-low.c (tdesc_xtensa): Declare. (xtensa_fill_gregset, xtensa_store_gregset): Adjust. (target_regsets): Rename to ... (xtensa_regsets): ... this, and make static. (xtensa_regsets_info, xtensa_usrregs_info, regs_info): New globals. (xtensa_arch_setup, xtensa_regs_info): New functions. (the_low_target): Adjust. (initialize_low_arch): New function. * linux-nios2-low.c (tdesc_nios2_linux): Declare. (nios2_arch_setup): Set the current process'es tdesc. (target_regsets): Rename to ... (nios2_regsets): ... this. (nios2_regsets_info, nios2_usrregs_info, regs_info): New globals. (nios2_regs_info): New function. (the_low_target): Adjust. (initialize_low_arch): New function. * linux-aarch64-low.c (tdesc_aarch64): Declare. (aarch64_arch_setup): Set the current process'es tdesc. (target_regsets): Rename to ... (aarch64_regsets): ... this. (aarch64_regsets_info, aarch64_usrregs_info, regs_info): New globals. (aarch64_regs_info): New function. (the_low_target): Adjust. (initialize_low_arch): New function. * linux-tile-low.c (tdesc_tilegx, tdesc_tilegx32): Declare globals. (target_regsets): Rename to ... (tile_regsets): ... this. (tile_regsets_info, tile_usrregs_info, regs_info): New globals. (tile_regs_info): New function. (tile_arch_setup): Set the current process'es tdesc. (the_low_target): Adjust. (initialize_low_arch): New function. * spu-low.c (tdesc_spu): Declare. (spu_create_inferior, spu_attach): Set the new process'es tdesc. * win32-arm-low.c (tdesc_arm): Declare. (arm_arch_setup): New function. (the_low_target): Install arm_arch_setup instead of init_registers_arm. * win32-i386-low.c (tdesc_i386, tdesc_amd64): Declare. (init_windows_x86): Rename to ... (i386_arch_setup): ... this. Set `win32_tdesc'. (the_low_target): Adjust. * win32-low.c (win32_tdesc): New global. (child_add_thread): Don't create the thread cache here. (do_initial_child_stuff): Set the new process'es tdesc. * win32-low.h (struct target_desc): Forward declare. (win32_tdesc): Declare. * lynx-i386-low.c (tdesc_i386): Declare global. (lynx_i386_arch_setup): Set `lynx_tdesc'. * lynx-low.c (lynx_tdesc): New global. (lynx_add_process): Set the new process'es tdesc. * lynx-low.h (struct target_desc): Forward declare. (lynx_tdesc): Declare global. * lynx-ppc-low.c (tdesc_powerpc_32): Declare global. (lynx_ppc_arch_setup): Set `lynx_tdesc'. * nto-low.c (nto_tdesc): New global. (do_attach): Set the new process'es tdesc. * nto-low.h (struct target_desc): Forward declare. (nto_tdesc): Declare. * nto-x86-low.c (tdesc_i386): Declare. (nto_x86_arch_setup): Set `nto_tdesc'. gdb/ 2013-06-07 Pedro Alves <palves@redhat.com> * regformats/regdat.sh: Output #include tdesc.h. Make globals static. Output a global target description pointer. (init_registers_${name}): Adjust to initialize a target description structure.
2013-06-07 12:46:59 +02:00
proc = add_process (status.pid, 1);
proc->tdesc = nto_tdesc;
TRACE ("Adding thread: pid=%d tid=%ld\n", status.pid,
ptid_get_lwp (ptid));
nto_find_new_threads (&nto_inferior);
}
else
{
do_detach ();
return -1;
}
return pid;
}
/* Read or write LEN bytes from/to inferior's MEMADDR memory address
into gdbservers's MYADDR buffer. Return number of bytes actually
transfered. */
static int
nto_xfer_memory (off_t memaddr, unsigned char *myaddr, int len,
int dowrite)
{
int nbytes = 0;
if (lseek (nto_inferior.ctl_fd, memaddr, SEEK_SET) == memaddr)
{
if (dowrite)
nbytes = write (nto_inferior.ctl_fd, myaddr, len);
else
nbytes = read (nto_inferior.ctl_fd, myaddr, len);
if (nbytes < 0)
nbytes = 0;
}
if (nbytes == 0)
{
int e = errno;
TRACE ("Error in %s : errno=%d (%s)\n", __func__, e, strerror (e));
}
return nbytes;
}
/* Insert or remove breakpoint or watchpoint at address ADDR.
TYPE can be one of Neutrino breakpoint types. SIZE must be 0 for
inserting the point, -1 for removing it.
Return 0 on success, 1 otherwise. */
static int
nto_breakpoint (CORE_ADDR addr, int type, int size)
{
procfs_break brk;
brk.type = type;
brk.addr = addr;
brk.size = size;
if (devctl (nto_inferior.ctl_fd, DCMD_PROC_BREAK, &brk, sizeof (brk), 0)
!= EOK)
return 1;
return 0;
}
/* Read auxiliary vector from inferior's initial stack into gdbserver's
MYADDR buffer, up to LEN bytes.
Return number of bytes read. */
static int
nto_read_auxv_from_initial_stack (CORE_ADDR initial_stack,
unsigned char *myaddr,
unsigned int len)
{
int data_ofs = 0;
int anint;
unsigned int len_read = 0;
/* Skip over argc, argv and envp... Comment from ldd.c:
The startup frame is set-up so that we have:
auxv
NULL
...
envp2
envp1 <----- void *frame + (argc + 2) * sizeof(char *)
NULL
...
argv2
argv1
argc <------ void * frame
On entry to ldd, frame gives the address of argc on the stack. */
if (nto_xfer_memory (initial_stack, (unsigned char *)&anint,
sizeof (anint), 0) != sizeof (anint))
return 0;
/* Size of pointer is assumed to be 4 bytes (32 bit arch. ) */
data_ofs += (anint + 2) * sizeof (void *); /* + 2 comes from argc itself and
NULL terminating pointer in
argv. */
/* Now loop over env table: */
while (nto_xfer_memory (initial_stack + data_ofs,
(unsigned char *)&anint, sizeof (anint), 0)
== sizeof (anint))
{
data_ofs += sizeof (anint);
if (anint == 0)
break;
}
initial_stack += data_ofs;
memset (myaddr, 0, len);
while (len_read <= len - sizeof (auxv_t))
{
auxv_t *auxv = (auxv_t *)myaddr;
/* Search backwards until we have read AT_PHDR (num. 3),
AT_PHENT (num 4), AT_PHNUM (num 5) */
if (nto_xfer_memory (initial_stack, (unsigned char *)auxv,
sizeof (auxv_t), 0) == sizeof (auxv_t))
{
if (auxv->a_type != AT_NULL)
{
auxv++;
len_read += sizeof (auxv_t);
}
if (auxv->a_type == AT_PHNUM) /* That's all we need. */
break;
initial_stack += sizeof (auxv_t);
}
else
break;
}
TRACE ("auxv: len_read: %d\n", len_read);
return len_read;
}
/* Start inferior specified by PROGRAM passing arguments ALLARGS. */
static int
nto_create_inferior (char *program, char **allargs)
{
struct inheritance inherit;
pid_t pid;
sigset_t set;
TRACE ("%s %s\n", __func__, program);
/* Clear any pending SIGUSR1's but keep the behavior the same. */
signal (SIGUSR1, signal (SIGUSR1, SIG_IGN));
sigemptyset (&set);
sigaddset (&set, SIGUSR1);
sigprocmask (SIG_UNBLOCK, &set, NULL);
memset (&inherit, 0, sizeof (inherit));
inherit.flags |= SPAWN_SETGROUP | SPAWN_HOLD;
inherit.pgroup = SPAWN_NEWPGROUP;
pid = spawnp (program, 0, NULL, &inherit, allargs, 0);
sigprocmask (SIG_BLOCK, &set, NULL);
if (pid == -1)
return -1;
if (do_attach (pid) != pid)
return -1;
return pid;
}
/* Attach to process PID. */
static int
nto_attach (unsigned long pid)
{
TRACE ("%s %ld\n", __func__, pid);
if (do_attach (pid) != pid)
error ("Unable to attach to %ld\n", pid);
return 0;
}
/* Send signal to process PID. */
static int
nto_kill (int pid)
{
TRACE ("%s %d\n", __func__, pid);
kill (pid, SIGKILL);
do_detach ();
return 0;
}
/* Detach from process PID. */
static int
nto_detach (int pid)
{
TRACE ("%s %d\n", __func__, pid);
do_detach ();
return 0;
}
static void
nto_mourn (struct process_info *process)
{
remove_process (process);
}
/* Check if the given thread is alive.
Return 1 if alive, 0 otherwise. */
static int
nto_thread_alive (ptid_t ptid)
{
int res;
TRACE ("%s pid:%d tid:%d\n", __func__, ptid_get_pid (ptid),
ptid_get_lwp (ptid));
if (SignalKill (0, ptid_get_pid (ptid), ptid_get_lwp (ptid),
0, 0, 0) == -1)
res = 0;
else
res = 1;
TRACE ("%s: %s\n", __func__, res ? "yes" : "no");
return res;
}
/* Resume inferior's execution. */
static void
nto_resume (struct thread_resume *resume_info, size_t n)
{
/* We can only work in all-stop mode. */
procfs_status status;
procfs_run run;
int err;
TRACE ("%s\n", __func__);
/* Workaround for aliasing rules violation. */
sigset_t *run_fault = (sigset_t *) (void *) &run.fault;
nto_set_thread (resume_info->thread);
run.flags = _DEBUG_RUN_FAULT | _DEBUG_RUN_TRACE;
if (resume_info->kind == resume_step)
run.flags |= _DEBUG_RUN_STEP;
run.flags |= _DEBUG_RUN_ARM;
sigemptyset (run_fault);
sigaddset (run_fault, FLTBPT);
sigaddset (run_fault, FLTTRACE);
sigaddset (run_fault, FLTILL);
sigaddset (run_fault, FLTPRIV);
sigaddset (run_fault, FLTBOUNDS);
sigaddset (run_fault, FLTIOVF);
sigaddset (run_fault, FLTIZDIV);
sigaddset (run_fault, FLTFPE);
sigaddset (run_fault, FLTPAGE);
sigaddset (run_fault, FLTSTACK);
sigaddset (run_fault, FLTACCESS);
sigemptyset (&run.trace);
if (resume_info->sig)
{
int signal_to_pass;
devctl (nto_inferior.ctl_fd, DCMD_PROC_STATUS, &status, sizeof (status),
0);
signal_to_pass = resume_info->sig;
if (status.why & (_DEBUG_WHY_SIGNALLED | _DEBUG_WHY_FAULTED))
{
if (signal_to_pass != status.info.si_signo)
{
kill (status.pid, signal_to_pass);
run.flags |= _DEBUG_RUN_CLRFLT | _DEBUG_RUN_CLRSIG;
}
else /* Let it kill the program without telling us. */
sigdelset (&run.trace, signal_to_pass);
}
}
else
run.flags |= _DEBUG_RUN_CLRSIG | _DEBUG_RUN_CLRFLT;
sigfillset (&run.trace);
regcache_invalidate ();
err = devctl (nto_inferior.ctl_fd, DCMD_PROC_RUN, &run, sizeof (run), 0);
if (err != EOK)
TRACE ("Error: %d \"%s\"\n", err, strerror (err));
}
/* Wait for inferior's event.
Return ptid of thread that caused the event. */
static ptid_t
nto_wait (ptid_t ptid,
struct target_waitstatus *ourstatus, int target_options)
{
sigset_t set;
siginfo_t info;
procfs_status status;
const int trace_mask = (_DEBUG_FLAG_TRACE_EXEC | _DEBUG_FLAG_TRACE_RD
| _DEBUG_FLAG_TRACE_WR | _DEBUG_FLAG_TRACE_MODIFY);
TRACE ("%s\n", __func__);
ourstatus->kind = TARGET_WAITKIND_SPURIOUS;
sigemptyset (&set);
sigaddset (&set, SIGUSR1);
devctl (nto_inferior.ctl_fd, DCMD_PROC_STATUS, &status, sizeof (status), 0);
while (!(status.flags & _DEBUG_FLAG_ISTOP))
{
sigwaitinfo (&set, &info);
devctl (nto_inferior.ctl_fd, DCMD_PROC_STATUS, &status, sizeof (status),
0);
}
nto_find_new_threads (&nto_inferior);
if (status.flags & _DEBUG_FLAG_SSTEP)
{
TRACE ("SSTEP\n");
ourstatus->kind = TARGET_WAITKIND_STOPPED;
gdb/ 2012-05-24 Pedro Alves <palves@redhat.com> PR gdb/7205 Replace TARGET_SIGNAL_ with GDB_SIGNAL_ throughout. gdb/gdbserver/ 2012-05-24 Pedro Alves <palves@redhat.com> PR gdb/7205 Replace TARGET_SIGNAL_ with GDB_SIGNAL_ throughout. include/gdb/ 2012-05-24 Pedro Alves <palves@redhat.com> PR gdb/7205 * gdb/signals.def: Replace TARGET_SIGNAL_ with GDB_SIGNAL_ throughout. sim/arm/ 2012-05-24 Pedro Alves <palves@redhat.com> PR gdb/7205 Replace TARGET_SIGNAL_ with GDB_SIGNAL_ throughout. sim/avr/ 2012-05-24 Pedro Alves <palves@redhat.com> PR gdb/7205 Replace TARGET_SIGNAL_ with GDB_SIGNAL_ throughout. sim/common/ 2012-05-24 Pedro Alves <palves@redhat.com> PR gdb/7205 Replace TARGET_SIGNAL_ with GDB_SIGNAL_ throughout. sim/cr16/ 2012-05-24 Pedro Alves <palves@redhat.com> PR gdb/7205 Replace TARGET_SIGNAL_ with GDB_SIGNAL_ throughout. sim/d10v/ 2012-05-24 Pedro Alves <palves@redhat.com> PR gdb/7205 Replace TARGET_SIGNAL_ with GDB_SIGNAL_ throughout. sim/erc32/ 2012-05-24 Pedro Alves <palves@redhat.com> PR gdb/7205 Replace TARGET_SIGNAL_ with GDB_SIGNAL_ throughout. sim/m32c/ 2012-05-24 Pedro Alves <palves@redhat.com> PR gdb/7205 Replace TARGET_SIGNAL_ with GDB_SIGNAL_ throughout. sim/ppc/ 2012-05-24 Pedro Alves <palves@redhat.com> PR gdb/7205 Replace TARGET_SIGNAL_ with GDB_SIGNAL_ throughout. sim/rl78/ 2012-05-24 Pedro Alves <palves@redhat.com> PR gdb/7205 Replace TARGET_SIGNAL_ with GDB_SIGNAL_ throughout. sim/rx/ 2012-05-24 Pedro Alves <palves@redhat.com> PR gdb/7205 Replace TARGET_SIGNAL_ with GDB_SIGNAL_ throughout.
2012-05-24 18:51:47 +02:00
ourstatus->value.sig = GDB_SIGNAL_TRAP;
}
/* Was it a breakpoint? */
else if (status.flags & trace_mask)
{
TRACE ("STOPPED\n");
ourstatus->kind = TARGET_WAITKIND_STOPPED;
gdb/ 2012-05-24 Pedro Alves <palves@redhat.com> PR gdb/7205 Replace TARGET_SIGNAL_ with GDB_SIGNAL_ throughout. gdb/gdbserver/ 2012-05-24 Pedro Alves <palves@redhat.com> PR gdb/7205 Replace TARGET_SIGNAL_ with GDB_SIGNAL_ throughout. include/gdb/ 2012-05-24 Pedro Alves <palves@redhat.com> PR gdb/7205 * gdb/signals.def: Replace TARGET_SIGNAL_ with GDB_SIGNAL_ throughout. sim/arm/ 2012-05-24 Pedro Alves <palves@redhat.com> PR gdb/7205 Replace TARGET_SIGNAL_ with GDB_SIGNAL_ throughout. sim/avr/ 2012-05-24 Pedro Alves <palves@redhat.com> PR gdb/7205 Replace TARGET_SIGNAL_ with GDB_SIGNAL_ throughout. sim/common/ 2012-05-24 Pedro Alves <palves@redhat.com> PR gdb/7205 Replace TARGET_SIGNAL_ with GDB_SIGNAL_ throughout. sim/cr16/ 2012-05-24 Pedro Alves <palves@redhat.com> PR gdb/7205 Replace TARGET_SIGNAL_ with GDB_SIGNAL_ throughout. sim/d10v/ 2012-05-24 Pedro Alves <palves@redhat.com> PR gdb/7205 Replace TARGET_SIGNAL_ with GDB_SIGNAL_ throughout. sim/erc32/ 2012-05-24 Pedro Alves <palves@redhat.com> PR gdb/7205 Replace TARGET_SIGNAL_ with GDB_SIGNAL_ throughout. sim/m32c/ 2012-05-24 Pedro Alves <palves@redhat.com> PR gdb/7205 Replace TARGET_SIGNAL_ with GDB_SIGNAL_ throughout. sim/ppc/ 2012-05-24 Pedro Alves <palves@redhat.com> PR gdb/7205 Replace TARGET_SIGNAL_ with GDB_SIGNAL_ throughout. sim/rl78/ 2012-05-24 Pedro Alves <palves@redhat.com> PR gdb/7205 Replace TARGET_SIGNAL_ with GDB_SIGNAL_ throughout. sim/rx/ 2012-05-24 Pedro Alves <palves@redhat.com> PR gdb/7205 Replace TARGET_SIGNAL_ with GDB_SIGNAL_ throughout.
2012-05-24 18:51:47 +02:00
ourstatus->value.sig = GDB_SIGNAL_TRAP;
}
else if (status.flags & _DEBUG_FLAG_ISTOP)
{
TRACE ("ISTOP\n");
switch (status.why)
{
case _DEBUG_WHY_SIGNALLED:
TRACE (" SIGNALLED\n");
ourstatus->kind = TARGET_WAITKIND_STOPPED;
ourstatus->value.sig =
gdb_signal_from_host (status.info.si_signo);
nto_inferior.exit_signo = ourstatus->value.sig;
break;
case _DEBUG_WHY_FAULTED:
TRACE (" FAULTED\n");
ourstatus->kind = TARGET_WAITKIND_STOPPED;
if (status.info.si_signo == SIGTRAP)
{
ourstatus->value.sig = 0;
nto_inferior.exit_signo = 0;
}
else
{
ourstatus->value.sig =
gdb_signal_from_host (status.info.si_signo);
nto_inferior.exit_signo = ourstatus->value.sig;
}
break;
case _DEBUG_WHY_TERMINATED:
{
int waitval = 0;
TRACE (" TERMINATED\n");
waitpid (ptid_get_pid (ptid), &waitval, WNOHANG);
if (nto_inferior.exit_signo)
{
/* Abnormal death. */
ourstatus->kind = TARGET_WAITKIND_SIGNALLED;
ourstatus->value.sig = nto_inferior.exit_signo;
}
else
{
/* Normal death. */
ourstatus->kind = TARGET_WAITKIND_EXITED;
ourstatus->value.integer = WEXITSTATUS (waitval);
}
nto_inferior.exit_signo = 0;
break;
}
case _DEBUG_WHY_REQUESTED:
TRACE ("REQUESTED\n");
/* We are assuming a requested stop is due to a SIGINT. */
ourstatus->kind = TARGET_WAITKIND_STOPPED;
gdb/ 2012-05-24 Pedro Alves <palves@redhat.com> PR gdb/7205 Replace TARGET_SIGNAL_ with GDB_SIGNAL_ throughout. gdb/gdbserver/ 2012-05-24 Pedro Alves <palves@redhat.com> PR gdb/7205 Replace TARGET_SIGNAL_ with GDB_SIGNAL_ throughout. include/gdb/ 2012-05-24 Pedro Alves <palves@redhat.com> PR gdb/7205 * gdb/signals.def: Replace TARGET_SIGNAL_ with GDB_SIGNAL_ throughout. sim/arm/ 2012-05-24 Pedro Alves <palves@redhat.com> PR gdb/7205 Replace TARGET_SIGNAL_ with GDB_SIGNAL_ throughout. sim/avr/ 2012-05-24 Pedro Alves <palves@redhat.com> PR gdb/7205 Replace TARGET_SIGNAL_ with GDB_SIGNAL_ throughout. sim/common/ 2012-05-24 Pedro Alves <palves@redhat.com> PR gdb/7205 Replace TARGET_SIGNAL_ with GDB_SIGNAL_ throughout. sim/cr16/ 2012-05-24 Pedro Alves <palves@redhat.com> PR gdb/7205 Replace TARGET_SIGNAL_ with GDB_SIGNAL_ throughout. sim/d10v/ 2012-05-24 Pedro Alves <palves@redhat.com> PR gdb/7205 Replace TARGET_SIGNAL_ with GDB_SIGNAL_ throughout. sim/erc32/ 2012-05-24 Pedro Alves <palves@redhat.com> PR gdb/7205 Replace TARGET_SIGNAL_ with GDB_SIGNAL_ throughout. sim/m32c/ 2012-05-24 Pedro Alves <palves@redhat.com> PR gdb/7205 Replace TARGET_SIGNAL_ with GDB_SIGNAL_ throughout. sim/ppc/ 2012-05-24 Pedro Alves <palves@redhat.com> PR gdb/7205 Replace TARGET_SIGNAL_ with GDB_SIGNAL_ throughout. sim/rl78/ 2012-05-24 Pedro Alves <palves@redhat.com> PR gdb/7205 Replace TARGET_SIGNAL_ with GDB_SIGNAL_ throughout. sim/rx/ 2012-05-24 Pedro Alves <palves@redhat.com> PR gdb/7205 Replace TARGET_SIGNAL_ with GDB_SIGNAL_ throughout.
2012-05-24 18:51:47 +02:00
ourstatus->value.sig = GDB_SIGNAL_INT;
nto_inferior.exit_signo = 0;
break;
}
}
return ptid_build (status.pid, status.tid, 0);
}
/* Fetch inferior's registers for currently selected thread (CURRENT_INFERIOR).
If REGNO is -1, fetch all registers, or REGNO register only otherwise. */
static void
* regcache.h (struct thread_info): Forward declare. (struct regcache): New. (new_register_cache): Adjust prototype. (get_thread_regcache): Declare. (free_register_cache): Adjust prototype. (registers_to_string, registers_from_string): Ditto. (supply_register, supply_register_by_name, collect_register) (collect_register_as_string, collect_register_by_name): Ditto. * regcache.c (struct inferior_regcache_data): Delete. (get_regcache): Rename to ... (get_thread_regcache): ... this. Adjust. Switch inferior before fetching registers. (regcache_invalidate_one): Adjust. (regcache_invalidate): Fix prototype. (new_register_cache): Return the new register cache. (free_register_cache): Change prototype. (realloc_register_cache): Adjust. (registers_to_string): Change prototype to take a regcache. Adjust. (registers_from_string): Ditto. (register_data): Ditto. (supply_register): Ditto. (supply_register_by_name): Ditto. (collect_register): Ditto. (collect_register_as_string): Ditto. (collect_register_by_name): Ditto. * server.c (process_serial_event): Adjust. * linux-low.h (regset_fill_func, regset_store_func): Change prototype. (get_pc, set_pc, collect_ptrace_register, supply_ptrace_register): Change prototype. * linux-low.c (get_stop_pc): Adjust. (check_removed_breakpoint): Adjust. (linux_wait_for_event): Adjust. (linux_resume_one_lwp): Adjust. (fetch_register): Add regcache parameter. Adjust. (usr_store_inferior_registers): Ditto. (regsets_fetch_inferior_registers): Ditto. (regsets_store_inferior_registers): Ditto. (linux_fetch_registers, linux_store_registers): Ditto. * i387-fp.c (i387_cache_to_fsave): Change prototype to take a regcache. Adjust. (i387_fsave_to_cache, i387_cache_to_fxsave, i387_fxsave_to_cache): Ditto. * i387-fp.h (i387_cache_to_fsave, i387_fsave_to_cache): Change prototype to take a regcache. (i387_cache_to_fxsave, i387_fxsave_to_cache): Ditto. * remote-utils.c (convert_ascii_to_int, outreg) (prepare_resume_reply): Change prototype to take a regcache. Adjust. * target.h (struct target_ops) <fetch_registers, store_registers>: Change prototype to take a regcache. (fetch_inferior_registers, store_inferior_registers): Change prototype to take a regcache. Adjust. * proc-service.c (ps_lgetregs): Adjust. * linux-x86-low.c (x86_fill_gregset, x86_store_gregset) (x86_fill_fpregset, x86_store_fpregset, x86_fill_fpxregset) (x86_store_fpxregset, x86_get_pc, x86_set_pc): Change prototype to take a regcache. Adjust. * linux-arm-low.c (arm_fill_gregset, arm_store_gregset) (arm_fill_wmmxregset, arm_store_wmmxregset, arm_fill_vfpregset) (arm_store_vfpregset, arm_get_pc, arm_set_pc): (arm_breakpoint_at): Change prototype to take a regcache. Adjust. * linux-cris-low.c (cris_get_pc, cris_set_pc) (cris_cannot_fetch_register): (cris_breakpoint_at): Change prototype to take a regcache. Adjust. * linux-crisv32-low.c (cris_get_pc, cris_set_pc, cris_reinsert_addr, cris_write_data_breakpoint): Change prototype to take a regcache. Adjust. (cris_breakpoint_at, cris_insert_point, cris_remove_point): Adjust. * linux-m32r-low.c (m32r_get_pc, m32r_set_pc): Change prototype to take a regcache. Adjust. * linux-m68k-low.c (m68k_fill_gregset, m68k_store_gregset) (m68k_fill_fpregset, m68k_store_fpregset, m68k_get_pc, (m68k_set_pc): Change prototype to take a regcache. Adjust. * linux-mips-low.c (mips_get_pc): (mips_set_pc): Change prototype to take a regcache. Adjust. (mips_reinsert_addr): Adjust. (mips_collect_register): Change prototype to take a regcache. Adjust. (mips_supply_register): (mips_collect_register_32bit, mips_supply_register_32bit) (mips_fill_gregset, mips_store_gregset, mips_fill_fpregset) (mips_store_fpregset): Ditto. * linux-ppc-low.c (ppc_supply_ptrace_register, ppc_supply_ptrace_register): Ditto. (parse_spufs_run): Adjust. (ppc_get_pc, ppc_set_pc, ppc_fill_gregset, ppc_fill_vsxregset) (ppc_store_vsxregset, ppc_fill_vrregset, ppc_store_vrregset) (ppc_fill_evrregset, ppc_store_evrregset): Change prototype to take a regcache. Adjust. * linux-s390-low.c (s390_collect_ptrace_register) (s390_supply_ptrace_register, s390_fill_gregset, s390_get_pc) (s390_set_pc): Change prototype to take a regcache. Adjust. (s390_arch_setup): Adjust. * linux-sh-low.c (sh_get_pc, sh_breakpoint_at) (sh_fill_gregset): Change prototype to take a regcache. Adjust. * linux-sparc-low.c (sparc_fill_gregset_to_stack) (sparc_fill_gregset, sparc_store_gregset_from_stack) (sparc_store_gregset, sparc_get_pc): Change prototype to take a regcache. Adjust. (sparc_breakpoint_at): Adjust. * linux-xtensa-low.c (xtensa_fill_gregset): (xtensa_store_gregset): (xtensa_fill_xtregset, xtensa_store_xtregset, xtensa_get_pc) (xtensa_set_pc): Change prototype to take a regcache. Adjust. * nto-low.c (nto_fetch_registers, nto_store_registers): Change prototype to take a regcache. Adjust. * win32-arm-low.c (arm_fetch_inferior_register) (arm_store_inferior_register): Change prototype to take a regcache. Adjust. * win32-i386-low.c (i386_fetch_inferior_register) (i386_store_inferior_register): Change prototype to take a regcache. Adjust. * win32-low.c (child_fetch_inferior_registers) (child_store_inferior_registers): Change prototype to take a regcache. Adjust. (win32_wait): Adjust. (win32_fetch_inferior_registers): Change prototype to take a regcache. Adjust. (win32_store_inferior_registers): Adjust. * win32-low.h (struct win32_target_ops) <fetch_inferior_register, store_inferior_register>: Change prototype to take a regcache.
2010-01-20 23:55:38 +01:00
nto_fetch_registers (struct regcache *regcache, int regno)
{
int regsize;
procfs_greg greg;
ptid_t ptid;
TRACE ("%s (regno=%d)\n", __func__, regno);
if (regno >= the_low_target.num_regs)
return;
Rename current_inferior as current_thread in gdbserver GDB has a function named "current_inferior" and gdbserver has a global variable named "current_inferior", but the two are not equivalent; indeed, gdbserver does not have any real equivalent of what GDB calls an inferior. What gdbserver's "current_inferior" is actually pointing to is a structure describing the current thread. This commit renames current_inferior as current_thread in gdbserver to clarify this. It also renames the function "set_desired_inferior" to "set_desired_thread" and renames various local variables from foo_inferior to foo_thread. gdb/gdbserver/ChangeLog: * inferiors.h (current_inferior): Renamed as... (current_thread): New variable. All uses updated. * linux-low.c (get_pc): Renamed saved_inferior as saved_thread. (maybe_move_out_of_jump_pad): Likewise. (cancel_breakpoint): Likewise. (linux_low_filter_event): Likewise. (wait_for_sigstop): Likewise. (linux_resume_one_lwp): Likewise. (need_step_over_p): Likewise. (start_step_over): Likewise. (linux_stabilize_threads): Renamed save_inferior as saved_thread. * linux-x86-low.c (x86_linux_update_xmltarget): Likewise. * proc-service.c (ps_lgetregs): Renamed reg_inferior as reg_thread and save_inferior as saved_thread. * regcache.c (get_thread_regcache): Renamed saved_inferior as saved_thread. (regcache_invalidate_thread): Likewise. * remote-utils.c (prepare_resume_reply): Likewise. * thread-db.c (thread_db_get_tls_address): Likewise. (disable_thread_event_reporting): Likewise. (remove_thread_event_breakpoints): Likewise. * tracepoint.c (gdb_agent_about_to_close): Renamed save_inferior as saved_thread. * target.h (set_desired_inferior): Renamed as... (set_desired_thread): New declaration. All uses updated. * server.c (myresume): Updated comment to reference thread instead of inferior. (handle_serial_event): Likewise. (handle_target_event): Likewise.
2014-09-10 11:37:11 +02:00
if (current_thread == NULL)
{
Rename current_inferior as current_thread in gdbserver GDB has a function named "current_inferior" and gdbserver has a global variable named "current_inferior", but the two are not equivalent; indeed, gdbserver does not have any real equivalent of what GDB calls an inferior. What gdbserver's "current_inferior" is actually pointing to is a structure describing the current thread. This commit renames current_inferior as current_thread in gdbserver to clarify this. It also renames the function "set_desired_inferior" to "set_desired_thread" and renames various local variables from foo_inferior to foo_thread. gdb/gdbserver/ChangeLog: * inferiors.h (current_inferior): Renamed as... (current_thread): New variable. All uses updated. * linux-low.c (get_pc): Renamed saved_inferior as saved_thread. (maybe_move_out_of_jump_pad): Likewise. (cancel_breakpoint): Likewise. (linux_low_filter_event): Likewise. (wait_for_sigstop): Likewise. (linux_resume_one_lwp): Likewise. (need_step_over_p): Likewise. (start_step_over): Likewise. (linux_stabilize_threads): Renamed save_inferior as saved_thread. * linux-x86-low.c (x86_linux_update_xmltarget): Likewise. * proc-service.c (ps_lgetregs): Renamed reg_inferior as reg_thread and save_inferior as saved_thread. * regcache.c (get_thread_regcache): Renamed saved_inferior as saved_thread. (regcache_invalidate_thread): Likewise. * remote-utils.c (prepare_resume_reply): Likewise. * thread-db.c (thread_db_get_tls_address): Likewise. (disable_thread_event_reporting): Likewise. (remove_thread_event_breakpoints): Likewise. * tracepoint.c (gdb_agent_about_to_close): Renamed save_inferior as saved_thread. * target.h (set_desired_inferior): Renamed as... (set_desired_thread): New declaration. All uses updated. * server.c (myresume): Updated comment to reference thread instead of inferior. (handle_serial_event): Likewise. (handle_target_event): Likewise.
2014-09-10 11:37:11 +02:00
TRACE ("current_thread is NULL\n");
return;
}
Rename current_inferior as current_thread in gdbserver GDB has a function named "current_inferior" and gdbserver has a global variable named "current_inferior", but the two are not equivalent; indeed, gdbserver does not have any real equivalent of what GDB calls an inferior. What gdbserver's "current_inferior" is actually pointing to is a structure describing the current thread. This commit renames current_inferior as current_thread in gdbserver to clarify this. It also renames the function "set_desired_inferior" to "set_desired_thread" and renames various local variables from foo_inferior to foo_thread. gdb/gdbserver/ChangeLog: * inferiors.h (current_inferior): Renamed as... (current_thread): New variable. All uses updated. * linux-low.c (get_pc): Renamed saved_inferior as saved_thread. (maybe_move_out_of_jump_pad): Likewise. (cancel_breakpoint): Likewise. (linux_low_filter_event): Likewise. (wait_for_sigstop): Likewise. (linux_resume_one_lwp): Likewise. (need_step_over_p): Likewise. (start_step_over): Likewise. (linux_stabilize_threads): Renamed save_inferior as saved_thread. * linux-x86-low.c (x86_linux_update_xmltarget): Likewise. * proc-service.c (ps_lgetregs): Renamed reg_inferior as reg_thread and save_inferior as saved_thread. * regcache.c (get_thread_regcache): Renamed saved_inferior as saved_thread. (regcache_invalidate_thread): Likewise. * remote-utils.c (prepare_resume_reply): Likewise. * thread-db.c (thread_db_get_tls_address): Likewise. (disable_thread_event_reporting): Likewise. (remove_thread_event_breakpoints): Likewise. * tracepoint.c (gdb_agent_about_to_close): Renamed save_inferior as saved_thread. * target.h (set_desired_inferior): Renamed as... (set_desired_thread): New declaration. All uses updated. * server.c (myresume): Updated comment to reference thread instead of inferior. (handle_serial_event): Likewise. (handle_target_event): Likewise.
2014-09-10 11:37:11 +02:00
ptid = thread_to_gdb_id (current_thread);
if (!nto_set_thread (ptid))
return;
if (devctl (nto_inferior.ctl_fd, DCMD_PROC_GETGREG, &greg, sizeof (greg),
&regsize) == EOK)
{
if (regno == -1) /* All registers. */
{
for (regno = 0; regno != the_low_target.num_regs; ++regno)
{
const unsigned int registeroffset
= the_low_target.register_offset (regno);
supply_register (regcache, regno,
((char *)&greg) + registeroffset);
}
}
else
{
const unsigned int registeroffset
= the_low_target.register_offset (regno);
if (registeroffset == -1)
return;
* regcache.h (struct thread_info): Forward declare. (struct regcache): New. (new_register_cache): Adjust prototype. (get_thread_regcache): Declare. (free_register_cache): Adjust prototype. (registers_to_string, registers_from_string): Ditto. (supply_register, supply_register_by_name, collect_register) (collect_register_as_string, collect_register_by_name): Ditto. * regcache.c (struct inferior_regcache_data): Delete. (get_regcache): Rename to ... (get_thread_regcache): ... this. Adjust. Switch inferior before fetching registers. (regcache_invalidate_one): Adjust. (regcache_invalidate): Fix prototype. (new_register_cache): Return the new register cache. (free_register_cache): Change prototype. (realloc_register_cache): Adjust. (registers_to_string): Change prototype to take a regcache. Adjust. (registers_from_string): Ditto. (register_data): Ditto. (supply_register): Ditto. (supply_register_by_name): Ditto. (collect_register): Ditto. (collect_register_as_string): Ditto. (collect_register_by_name): Ditto. * server.c (process_serial_event): Adjust. * linux-low.h (regset_fill_func, regset_store_func): Change prototype. (get_pc, set_pc, collect_ptrace_register, supply_ptrace_register): Change prototype. * linux-low.c (get_stop_pc): Adjust. (check_removed_breakpoint): Adjust. (linux_wait_for_event): Adjust. (linux_resume_one_lwp): Adjust. (fetch_register): Add regcache parameter. Adjust. (usr_store_inferior_registers): Ditto. (regsets_fetch_inferior_registers): Ditto. (regsets_store_inferior_registers): Ditto. (linux_fetch_registers, linux_store_registers): Ditto. * i387-fp.c (i387_cache_to_fsave): Change prototype to take a regcache. Adjust. (i387_fsave_to_cache, i387_cache_to_fxsave, i387_fxsave_to_cache): Ditto. * i387-fp.h (i387_cache_to_fsave, i387_fsave_to_cache): Change prototype to take a regcache. (i387_cache_to_fxsave, i387_fxsave_to_cache): Ditto. * remote-utils.c (convert_ascii_to_int, outreg) (prepare_resume_reply): Change prototype to take a regcache. Adjust. * target.h (struct target_ops) <fetch_registers, store_registers>: Change prototype to take a regcache. (fetch_inferior_registers, store_inferior_registers): Change prototype to take a regcache. Adjust. * proc-service.c (ps_lgetregs): Adjust. * linux-x86-low.c (x86_fill_gregset, x86_store_gregset) (x86_fill_fpregset, x86_store_fpregset, x86_fill_fpxregset) (x86_store_fpxregset, x86_get_pc, x86_set_pc): Change prototype to take a regcache. Adjust. * linux-arm-low.c (arm_fill_gregset, arm_store_gregset) (arm_fill_wmmxregset, arm_store_wmmxregset, arm_fill_vfpregset) (arm_store_vfpregset, arm_get_pc, arm_set_pc): (arm_breakpoint_at): Change prototype to take a regcache. Adjust. * linux-cris-low.c (cris_get_pc, cris_set_pc) (cris_cannot_fetch_register): (cris_breakpoint_at): Change prototype to take a regcache. Adjust. * linux-crisv32-low.c (cris_get_pc, cris_set_pc, cris_reinsert_addr, cris_write_data_breakpoint): Change prototype to take a regcache. Adjust. (cris_breakpoint_at, cris_insert_point, cris_remove_point): Adjust. * linux-m32r-low.c (m32r_get_pc, m32r_set_pc): Change prototype to take a regcache. Adjust. * linux-m68k-low.c (m68k_fill_gregset, m68k_store_gregset) (m68k_fill_fpregset, m68k_store_fpregset, m68k_get_pc, (m68k_set_pc): Change prototype to take a regcache. Adjust. * linux-mips-low.c (mips_get_pc): (mips_set_pc): Change prototype to take a regcache. Adjust. (mips_reinsert_addr): Adjust. (mips_collect_register): Change prototype to take a regcache. Adjust. (mips_supply_register): (mips_collect_register_32bit, mips_supply_register_32bit) (mips_fill_gregset, mips_store_gregset, mips_fill_fpregset) (mips_store_fpregset): Ditto. * linux-ppc-low.c (ppc_supply_ptrace_register, ppc_supply_ptrace_register): Ditto. (parse_spufs_run): Adjust. (ppc_get_pc, ppc_set_pc, ppc_fill_gregset, ppc_fill_vsxregset) (ppc_store_vsxregset, ppc_fill_vrregset, ppc_store_vrregset) (ppc_fill_evrregset, ppc_store_evrregset): Change prototype to take a regcache. Adjust. * linux-s390-low.c (s390_collect_ptrace_register) (s390_supply_ptrace_register, s390_fill_gregset, s390_get_pc) (s390_set_pc): Change prototype to take a regcache. Adjust. (s390_arch_setup): Adjust. * linux-sh-low.c (sh_get_pc, sh_breakpoint_at) (sh_fill_gregset): Change prototype to take a regcache. Adjust. * linux-sparc-low.c (sparc_fill_gregset_to_stack) (sparc_fill_gregset, sparc_store_gregset_from_stack) (sparc_store_gregset, sparc_get_pc): Change prototype to take a regcache. Adjust. (sparc_breakpoint_at): Adjust. * linux-xtensa-low.c (xtensa_fill_gregset): (xtensa_store_gregset): (xtensa_fill_xtregset, xtensa_store_xtregset, xtensa_get_pc) (xtensa_set_pc): Change prototype to take a regcache. Adjust. * nto-low.c (nto_fetch_registers, nto_store_registers): Change prototype to take a regcache. Adjust. * win32-arm-low.c (arm_fetch_inferior_register) (arm_store_inferior_register): Change prototype to take a regcache. Adjust. * win32-i386-low.c (i386_fetch_inferior_register) (i386_store_inferior_register): Change prototype to take a regcache. Adjust. * win32-low.c (child_fetch_inferior_registers) (child_store_inferior_registers): Change prototype to take a regcache. Adjust. (win32_wait): Adjust. (win32_fetch_inferior_registers): Change prototype to take a regcache. Adjust. (win32_store_inferior_registers): Adjust. * win32-low.h (struct win32_target_ops) <fetch_inferior_register, store_inferior_register>: Change prototype to take a regcache.
2010-01-20 23:55:38 +01:00
supply_register (regcache, regno, ((char *)&greg) + registeroffset);
}
}
else
TRACE ("ERROR reading registers from inferior.\n");
}
/* Store registers for currently selected thread (CURRENT_INFERIOR).
We always store all registers, regardless of REGNO. */
static void
* regcache.h (struct thread_info): Forward declare. (struct regcache): New. (new_register_cache): Adjust prototype. (get_thread_regcache): Declare. (free_register_cache): Adjust prototype. (registers_to_string, registers_from_string): Ditto. (supply_register, supply_register_by_name, collect_register) (collect_register_as_string, collect_register_by_name): Ditto. * regcache.c (struct inferior_regcache_data): Delete. (get_regcache): Rename to ... (get_thread_regcache): ... this. Adjust. Switch inferior before fetching registers. (regcache_invalidate_one): Adjust. (regcache_invalidate): Fix prototype. (new_register_cache): Return the new register cache. (free_register_cache): Change prototype. (realloc_register_cache): Adjust. (registers_to_string): Change prototype to take a regcache. Adjust. (registers_from_string): Ditto. (register_data): Ditto. (supply_register): Ditto. (supply_register_by_name): Ditto. (collect_register): Ditto. (collect_register_as_string): Ditto. (collect_register_by_name): Ditto. * server.c (process_serial_event): Adjust. * linux-low.h (regset_fill_func, regset_store_func): Change prototype. (get_pc, set_pc, collect_ptrace_register, supply_ptrace_register): Change prototype. * linux-low.c (get_stop_pc): Adjust. (check_removed_breakpoint): Adjust. (linux_wait_for_event): Adjust. (linux_resume_one_lwp): Adjust. (fetch_register): Add regcache parameter. Adjust. (usr_store_inferior_registers): Ditto. (regsets_fetch_inferior_registers): Ditto. (regsets_store_inferior_registers): Ditto. (linux_fetch_registers, linux_store_registers): Ditto. * i387-fp.c (i387_cache_to_fsave): Change prototype to take a regcache. Adjust. (i387_fsave_to_cache, i387_cache_to_fxsave, i387_fxsave_to_cache): Ditto. * i387-fp.h (i387_cache_to_fsave, i387_fsave_to_cache): Change prototype to take a regcache. (i387_cache_to_fxsave, i387_fxsave_to_cache): Ditto. * remote-utils.c (convert_ascii_to_int, outreg) (prepare_resume_reply): Change prototype to take a regcache. Adjust. * target.h (struct target_ops) <fetch_registers, store_registers>: Change prototype to take a regcache. (fetch_inferior_registers, store_inferior_registers): Change prototype to take a regcache. Adjust. * proc-service.c (ps_lgetregs): Adjust. * linux-x86-low.c (x86_fill_gregset, x86_store_gregset) (x86_fill_fpregset, x86_store_fpregset, x86_fill_fpxregset) (x86_store_fpxregset, x86_get_pc, x86_set_pc): Change prototype to take a regcache. Adjust. * linux-arm-low.c (arm_fill_gregset, arm_store_gregset) (arm_fill_wmmxregset, arm_store_wmmxregset, arm_fill_vfpregset) (arm_store_vfpregset, arm_get_pc, arm_set_pc): (arm_breakpoint_at): Change prototype to take a regcache. Adjust. * linux-cris-low.c (cris_get_pc, cris_set_pc) (cris_cannot_fetch_register): (cris_breakpoint_at): Change prototype to take a regcache. Adjust. * linux-crisv32-low.c (cris_get_pc, cris_set_pc, cris_reinsert_addr, cris_write_data_breakpoint): Change prototype to take a regcache. Adjust. (cris_breakpoint_at, cris_insert_point, cris_remove_point): Adjust. * linux-m32r-low.c (m32r_get_pc, m32r_set_pc): Change prototype to take a regcache. Adjust. * linux-m68k-low.c (m68k_fill_gregset, m68k_store_gregset) (m68k_fill_fpregset, m68k_store_fpregset, m68k_get_pc, (m68k_set_pc): Change prototype to take a regcache. Adjust. * linux-mips-low.c (mips_get_pc): (mips_set_pc): Change prototype to take a regcache. Adjust. (mips_reinsert_addr): Adjust. (mips_collect_register): Change prototype to take a regcache. Adjust. (mips_supply_register): (mips_collect_register_32bit, mips_supply_register_32bit) (mips_fill_gregset, mips_store_gregset, mips_fill_fpregset) (mips_store_fpregset): Ditto. * linux-ppc-low.c (ppc_supply_ptrace_register, ppc_supply_ptrace_register): Ditto. (parse_spufs_run): Adjust. (ppc_get_pc, ppc_set_pc, ppc_fill_gregset, ppc_fill_vsxregset) (ppc_store_vsxregset, ppc_fill_vrregset, ppc_store_vrregset) (ppc_fill_evrregset, ppc_store_evrregset): Change prototype to take a regcache. Adjust. * linux-s390-low.c (s390_collect_ptrace_register) (s390_supply_ptrace_register, s390_fill_gregset, s390_get_pc) (s390_set_pc): Change prototype to take a regcache. Adjust. (s390_arch_setup): Adjust. * linux-sh-low.c (sh_get_pc, sh_breakpoint_at) (sh_fill_gregset): Change prototype to take a regcache. Adjust. * linux-sparc-low.c (sparc_fill_gregset_to_stack) (sparc_fill_gregset, sparc_store_gregset_from_stack) (sparc_store_gregset, sparc_get_pc): Change prototype to take a regcache. Adjust. (sparc_breakpoint_at): Adjust. * linux-xtensa-low.c (xtensa_fill_gregset): (xtensa_store_gregset): (xtensa_fill_xtregset, xtensa_store_xtregset, xtensa_get_pc) (xtensa_set_pc): Change prototype to take a regcache. Adjust. * nto-low.c (nto_fetch_registers, nto_store_registers): Change prototype to take a regcache. Adjust. * win32-arm-low.c (arm_fetch_inferior_register) (arm_store_inferior_register): Change prototype to take a regcache. Adjust. * win32-i386-low.c (i386_fetch_inferior_register) (i386_store_inferior_register): Change prototype to take a regcache. Adjust. * win32-low.c (child_fetch_inferior_registers) (child_store_inferior_registers): Change prototype to take a regcache. Adjust. (win32_wait): Adjust. (win32_fetch_inferior_registers): Change prototype to take a regcache. Adjust. (win32_store_inferior_registers): Adjust. * win32-low.h (struct win32_target_ops) <fetch_inferior_register, store_inferior_register>: Change prototype to take a regcache.
2010-01-20 23:55:38 +01:00
nto_store_registers (struct regcache *regcache, int regno)
{
procfs_greg greg;
int err;
ptid_t ptid;
TRACE ("%s (regno:%d)\n", __func__, regno);
Rename current_inferior as current_thread in gdbserver GDB has a function named "current_inferior" and gdbserver has a global variable named "current_inferior", but the two are not equivalent; indeed, gdbserver does not have any real equivalent of what GDB calls an inferior. What gdbserver's "current_inferior" is actually pointing to is a structure describing the current thread. This commit renames current_inferior as current_thread in gdbserver to clarify this. It also renames the function "set_desired_inferior" to "set_desired_thread" and renames various local variables from foo_inferior to foo_thread. gdb/gdbserver/ChangeLog: * inferiors.h (current_inferior): Renamed as... (current_thread): New variable. All uses updated. * linux-low.c (get_pc): Renamed saved_inferior as saved_thread. (maybe_move_out_of_jump_pad): Likewise. (cancel_breakpoint): Likewise. (linux_low_filter_event): Likewise. (wait_for_sigstop): Likewise. (linux_resume_one_lwp): Likewise. (need_step_over_p): Likewise. (start_step_over): Likewise. (linux_stabilize_threads): Renamed save_inferior as saved_thread. * linux-x86-low.c (x86_linux_update_xmltarget): Likewise. * proc-service.c (ps_lgetregs): Renamed reg_inferior as reg_thread and save_inferior as saved_thread. * regcache.c (get_thread_regcache): Renamed saved_inferior as saved_thread. (regcache_invalidate_thread): Likewise. * remote-utils.c (prepare_resume_reply): Likewise. * thread-db.c (thread_db_get_tls_address): Likewise. (disable_thread_event_reporting): Likewise. (remove_thread_event_breakpoints): Likewise. * tracepoint.c (gdb_agent_about_to_close): Renamed save_inferior as saved_thread. * target.h (set_desired_inferior): Renamed as... (set_desired_thread): New declaration. All uses updated. * server.c (myresume): Updated comment to reference thread instead of inferior. (handle_serial_event): Likewise. (handle_target_event): Likewise.
2014-09-10 11:37:11 +02:00
if (current_thread == NULL)
{
Rename current_inferior as current_thread in gdbserver GDB has a function named "current_inferior" and gdbserver has a global variable named "current_inferior", but the two are not equivalent; indeed, gdbserver does not have any real equivalent of what GDB calls an inferior. What gdbserver's "current_inferior" is actually pointing to is a structure describing the current thread. This commit renames current_inferior as current_thread in gdbserver to clarify this. It also renames the function "set_desired_inferior" to "set_desired_thread" and renames various local variables from foo_inferior to foo_thread. gdb/gdbserver/ChangeLog: * inferiors.h (current_inferior): Renamed as... (current_thread): New variable. All uses updated. * linux-low.c (get_pc): Renamed saved_inferior as saved_thread. (maybe_move_out_of_jump_pad): Likewise. (cancel_breakpoint): Likewise. (linux_low_filter_event): Likewise. (wait_for_sigstop): Likewise. (linux_resume_one_lwp): Likewise. (need_step_over_p): Likewise. (start_step_over): Likewise. (linux_stabilize_threads): Renamed save_inferior as saved_thread. * linux-x86-low.c (x86_linux_update_xmltarget): Likewise. * proc-service.c (ps_lgetregs): Renamed reg_inferior as reg_thread and save_inferior as saved_thread. * regcache.c (get_thread_regcache): Renamed saved_inferior as saved_thread. (regcache_invalidate_thread): Likewise. * remote-utils.c (prepare_resume_reply): Likewise. * thread-db.c (thread_db_get_tls_address): Likewise. (disable_thread_event_reporting): Likewise. (remove_thread_event_breakpoints): Likewise. * tracepoint.c (gdb_agent_about_to_close): Renamed save_inferior as saved_thread. * target.h (set_desired_inferior): Renamed as... (set_desired_thread): New declaration. All uses updated. * server.c (myresume): Updated comment to reference thread instead of inferior. (handle_serial_event): Likewise. (handle_target_event): Likewise.
2014-09-10 11:37:11 +02:00
TRACE ("current_thread is NULL\n");
return;
}
Rename current_inferior as current_thread in gdbserver GDB has a function named "current_inferior" and gdbserver has a global variable named "current_inferior", but the two are not equivalent; indeed, gdbserver does not have any real equivalent of what GDB calls an inferior. What gdbserver's "current_inferior" is actually pointing to is a structure describing the current thread. This commit renames current_inferior as current_thread in gdbserver to clarify this. It also renames the function "set_desired_inferior" to "set_desired_thread" and renames various local variables from foo_inferior to foo_thread. gdb/gdbserver/ChangeLog: * inferiors.h (current_inferior): Renamed as... (current_thread): New variable. All uses updated. * linux-low.c (get_pc): Renamed saved_inferior as saved_thread. (maybe_move_out_of_jump_pad): Likewise. (cancel_breakpoint): Likewise. (linux_low_filter_event): Likewise. (wait_for_sigstop): Likewise. (linux_resume_one_lwp): Likewise. (need_step_over_p): Likewise. (start_step_over): Likewise. (linux_stabilize_threads): Renamed save_inferior as saved_thread. * linux-x86-low.c (x86_linux_update_xmltarget): Likewise. * proc-service.c (ps_lgetregs): Renamed reg_inferior as reg_thread and save_inferior as saved_thread. * regcache.c (get_thread_regcache): Renamed saved_inferior as saved_thread. (regcache_invalidate_thread): Likewise. * remote-utils.c (prepare_resume_reply): Likewise. * thread-db.c (thread_db_get_tls_address): Likewise. (disable_thread_event_reporting): Likewise. (remove_thread_event_breakpoints): Likewise. * tracepoint.c (gdb_agent_about_to_close): Renamed save_inferior as saved_thread. * target.h (set_desired_inferior): Renamed as... (set_desired_thread): New declaration. All uses updated. * server.c (myresume): Updated comment to reference thread instead of inferior. (handle_serial_event): Likewise. (handle_target_event): Likewise.
2014-09-10 11:37:11 +02:00
ptid = thread_to_gdb_id (current_thread);
if (!nto_set_thread (ptid))
return;
memset (&greg, 0, sizeof (greg));
for (regno = 0; regno != the_low_target.num_regs; ++regno)
{
const unsigned int regoffset
= the_low_target.register_offset (regno);
* regcache.h (struct thread_info): Forward declare. (struct regcache): New. (new_register_cache): Adjust prototype. (get_thread_regcache): Declare. (free_register_cache): Adjust prototype. (registers_to_string, registers_from_string): Ditto. (supply_register, supply_register_by_name, collect_register) (collect_register_as_string, collect_register_by_name): Ditto. * regcache.c (struct inferior_regcache_data): Delete. (get_regcache): Rename to ... (get_thread_regcache): ... this. Adjust. Switch inferior before fetching registers. (regcache_invalidate_one): Adjust. (regcache_invalidate): Fix prototype. (new_register_cache): Return the new register cache. (free_register_cache): Change prototype. (realloc_register_cache): Adjust. (registers_to_string): Change prototype to take a regcache. Adjust. (registers_from_string): Ditto. (register_data): Ditto. (supply_register): Ditto. (supply_register_by_name): Ditto. (collect_register): Ditto. (collect_register_as_string): Ditto. (collect_register_by_name): Ditto. * server.c (process_serial_event): Adjust. * linux-low.h (regset_fill_func, regset_store_func): Change prototype. (get_pc, set_pc, collect_ptrace_register, supply_ptrace_register): Change prototype. * linux-low.c (get_stop_pc): Adjust. (check_removed_breakpoint): Adjust. (linux_wait_for_event): Adjust. (linux_resume_one_lwp): Adjust. (fetch_register): Add regcache parameter. Adjust. (usr_store_inferior_registers): Ditto. (regsets_fetch_inferior_registers): Ditto. (regsets_store_inferior_registers): Ditto. (linux_fetch_registers, linux_store_registers): Ditto. * i387-fp.c (i387_cache_to_fsave): Change prototype to take a regcache. Adjust. (i387_fsave_to_cache, i387_cache_to_fxsave, i387_fxsave_to_cache): Ditto. * i387-fp.h (i387_cache_to_fsave, i387_fsave_to_cache): Change prototype to take a regcache. (i387_cache_to_fxsave, i387_fxsave_to_cache): Ditto. * remote-utils.c (convert_ascii_to_int, outreg) (prepare_resume_reply): Change prototype to take a regcache. Adjust. * target.h (struct target_ops) <fetch_registers, store_registers>: Change prototype to take a regcache. (fetch_inferior_registers, store_inferior_registers): Change prototype to take a regcache. Adjust. * proc-service.c (ps_lgetregs): Adjust. * linux-x86-low.c (x86_fill_gregset, x86_store_gregset) (x86_fill_fpregset, x86_store_fpregset, x86_fill_fpxregset) (x86_store_fpxregset, x86_get_pc, x86_set_pc): Change prototype to take a regcache. Adjust. * linux-arm-low.c (arm_fill_gregset, arm_store_gregset) (arm_fill_wmmxregset, arm_store_wmmxregset, arm_fill_vfpregset) (arm_store_vfpregset, arm_get_pc, arm_set_pc): (arm_breakpoint_at): Change prototype to take a regcache. Adjust. * linux-cris-low.c (cris_get_pc, cris_set_pc) (cris_cannot_fetch_register): (cris_breakpoint_at): Change prototype to take a regcache. Adjust. * linux-crisv32-low.c (cris_get_pc, cris_set_pc, cris_reinsert_addr, cris_write_data_breakpoint): Change prototype to take a regcache. Adjust. (cris_breakpoint_at, cris_insert_point, cris_remove_point): Adjust. * linux-m32r-low.c (m32r_get_pc, m32r_set_pc): Change prototype to take a regcache. Adjust. * linux-m68k-low.c (m68k_fill_gregset, m68k_store_gregset) (m68k_fill_fpregset, m68k_store_fpregset, m68k_get_pc, (m68k_set_pc): Change prototype to take a regcache. Adjust. * linux-mips-low.c (mips_get_pc): (mips_set_pc): Change prototype to take a regcache. Adjust. (mips_reinsert_addr): Adjust. (mips_collect_register): Change prototype to take a regcache. Adjust. (mips_supply_register): (mips_collect_register_32bit, mips_supply_register_32bit) (mips_fill_gregset, mips_store_gregset, mips_fill_fpregset) (mips_store_fpregset): Ditto. * linux-ppc-low.c (ppc_supply_ptrace_register, ppc_supply_ptrace_register): Ditto. (parse_spufs_run): Adjust. (ppc_get_pc, ppc_set_pc, ppc_fill_gregset, ppc_fill_vsxregset) (ppc_store_vsxregset, ppc_fill_vrregset, ppc_store_vrregset) (ppc_fill_evrregset, ppc_store_evrregset): Change prototype to take a regcache. Adjust. * linux-s390-low.c (s390_collect_ptrace_register) (s390_supply_ptrace_register, s390_fill_gregset, s390_get_pc) (s390_set_pc): Change prototype to take a regcache. Adjust. (s390_arch_setup): Adjust. * linux-sh-low.c (sh_get_pc, sh_breakpoint_at) (sh_fill_gregset): Change prototype to take a regcache. Adjust. * linux-sparc-low.c (sparc_fill_gregset_to_stack) (sparc_fill_gregset, sparc_store_gregset_from_stack) (sparc_store_gregset, sparc_get_pc): Change prototype to take a regcache. Adjust. (sparc_breakpoint_at): Adjust. * linux-xtensa-low.c (xtensa_fill_gregset): (xtensa_store_gregset): (xtensa_fill_xtregset, xtensa_store_xtregset, xtensa_get_pc) (xtensa_set_pc): Change prototype to take a regcache. Adjust. * nto-low.c (nto_fetch_registers, nto_store_registers): Change prototype to take a regcache. Adjust. * win32-arm-low.c (arm_fetch_inferior_register) (arm_store_inferior_register): Change prototype to take a regcache. Adjust. * win32-i386-low.c (i386_fetch_inferior_register) (i386_store_inferior_register): Change prototype to take a regcache. Adjust. * win32-low.c (child_fetch_inferior_registers) (child_store_inferior_registers): Change prototype to take a regcache. Adjust. (win32_wait): Adjust. (win32_fetch_inferior_registers): Change prototype to take a regcache. Adjust. (win32_store_inferior_registers): Adjust. * win32-low.h (struct win32_target_ops) <fetch_inferior_register, store_inferior_register>: Change prototype to take a regcache.
2010-01-20 23:55:38 +01:00
collect_register (regcache, regno, ((char *)&greg) + regoffset);
}
err = devctl (nto_inferior.ctl_fd, DCMD_PROC_SETGREG, &greg, sizeof (greg),
0);
if (err != EOK)
TRACE ("Error: setting registers.\n");
}
/* Read LEN bytes from inferior's memory address MEMADDR into
gdbserver's MYADDR buffer.
Return 0 on success -1 otherwise. */
static int
nto_read_memory (CORE_ADDR memaddr, unsigned char *myaddr, int len)
{
TRACE ("%s memaddr:0x%08lx, len:%d\n", __func__, memaddr, len);
if (nto_xfer_memory (memaddr, myaddr, len, 0) != len)
{
TRACE ("Failed to read memory\n");
return -1;
}
return 0;
}
/* Write LEN bytes from gdbserver's buffer MYADDR into inferior's
memory at address MEMADDR.
Return 0 on success -1 otherwise. */
static int
nto_write_memory (CORE_ADDR memaddr, const unsigned char *myaddr, int len)
{
int len_written;
TRACE ("%s memaddr: 0x%08llx len: %d\n", __func__, memaddr, len);
if ((len_written = nto_xfer_memory (memaddr, (unsigned char *)myaddr, len,
1))
!= len)
{
TRACE ("Wanted to write: %d but written: %d\n", len, len_written);
return -1;
}
return 0;
}
/* Stop inferior. We always stop all threads. */
static void
nto_request_interrupt (void)
{
TRACE ("%s\n", __func__);
nto_set_thread (ptid_build (nto_inferior.pid, 1, 0));
if (EOK != devctl (nto_inferior.ctl_fd, DCMD_PROC_STOP, NULL, 0, 0))
TRACE ("Error stopping inferior.\n");
}
/* Read auxiliary vector from inferior's memory into gdbserver's buffer
MYADDR. We always read whole auxv.
Return number of bytes stored in MYADDR buffer, 0 if OFFSET > 0
or -1 on error. */
static int
nto_read_auxv (CORE_ADDR offset, unsigned char *myaddr, unsigned int len)
{
int err;
CORE_ADDR initial_stack;
procfs_info procinfo;
TRACE ("%s\n", __func__);
if (offset > 0)
return 0;
err = devctl (nto_inferior.ctl_fd, DCMD_PROC_INFO, &procinfo,
sizeof procinfo, 0);
if (err != EOK)
return -1;
initial_stack = procinfo.initial_stack;
return nto_read_auxv_from_initial_stack (initial_stack, myaddr, len);
}
[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
static int
nto_supports_z_point_type (char z_type)
{
switch (z_type)
{
case Z_PACKET_SW_BP:
case Z_PACKET_HW_BP:
case Z_PACKET_WRITE_WP:
case Z_PACKET_READ_WP:
case Z_PACKET_ACCESS_WP:
return 1;
default:
return 0;
}
}
/* Insert {break/watch}point at address ADDR. SIZE is not used. */
static int
[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
nto_insert_point (enum raw_bkpt_type type, CORE_ADDR addr,
int size, struct raw_breakpoint *bp)
{
int wtype = _DEBUG_BREAK_HW; /* Always request HW. */
TRACE ("%s type:%c addr: 0x%08lx len:%d\n", __func__, (int)type, addr, len);
switch (type)
{
[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
case raw_bkpt_type_sw:
wtype = _DEBUG_BREAK_EXEC;
break;
[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
case raw_bkpt_type_hw:
wtype |= _DEBUG_BREAK_EXEC;
break;
[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
case raw_bkpt_type_write_wp:
wtype |= _DEBUG_BREAK_RW;
break;
[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
case raw_bkpt_type_read_wp:
wtype |= _DEBUG_BREAK_RD;
break;
[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
case raw_bkpt_type_access_wp:
wtype |= _DEBUG_BREAK_RW;
break;
default:
return 1; /* Not supported. */
}
return nto_breakpoint (addr, wtype, 0);
}
[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
/* Remove {break/watch}point at address ADDR. SIZE is not used. */
static int
[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
nto_remove_point (enum raw_bkpt_type type, CORE_ADDR addr,
int size, struct raw_breakpoint *bp)
{
int wtype = _DEBUG_BREAK_HW; /* Always request HW. */
TRACE ("%s type:%c addr: 0x%08lx len:%d\n", __func__, (int)type, addr, len);
switch (type)
{
[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
case raw_bkpt_type_sw:
wtype = _DEBUG_BREAK_EXEC;
break;
[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
case raw_bkpt_type_hw:
wtype |= _DEBUG_BREAK_EXEC;
break;
[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
case raw_bkpt_type_write_wp:
wtype |= _DEBUG_BREAK_RW;
break;
[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
case raw_bkpt_type_read_wp:
wtype |= _DEBUG_BREAK_RD;
break;
[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
case raw_bkpt_type_access_wp:
wtype |= _DEBUG_BREAK_RW;
break;
default:
return 1; /* Not supported. */
}
return nto_breakpoint (addr, wtype, -1);
}
/* Check if the reason of stop for current thread (CURRENT_INFERIOR) is
a watchpoint.
Return 1 if stopped by watchpoint, 0 otherwise. */
static int
nto_stopped_by_watchpoint (void)
{
int ret = 0;
TRACE ("%s\n", __func__);
Rename current_inferior as current_thread in gdbserver GDB has a function named "current_inferior" and gdbserver has a global variable named "current_inferior", but the two are not equivalent; indeed, gdbserver does not have any real equivalent of what GDB calls an inferior. What gdbserver's "current_inferior" is actually pointing to is a structure describing the current thread. This commit renames current_inferior as current_thread in gdbserver to clarify this. It also renames the function "set_desired_inferior" to "set_desired_thread" and renames various local variables from foo_inferior to foo_thread. gdb/gdbserver/ChangeLog: * inferiors.h (current_inferior): Renamed as... (current_thread): New variable. All uses updated. * linux-low.c (get_pc): Renamed saved_inferior as saved_thread. (maybe_move_out_of_jump_pad): Likewise. (cancel_breakpoint): Likewise. (linux_low_filter_event): Likewise. (wait_for_sigstop): Likewise. (linux_resume_one_lwp): Likewise. (need_step_over_p): Likewise. (start_step_over): Likewise. (linux_stabilize_threads): Renamed save_inferior as saved_thread. * linux-x86-low.c (x86_linux_update_xmltarget): Likewise. * proc-service.c (ps_lgetregs): Renamed reg_inferior as reg_thread and save_inferior as saved_thread. * regcache.c (get_thread_regcache): Renamed saved_inferior as saved_thread. (regcache_invalidate_thread): Likewise. * remote-utils.c (prepare_resume_reply): Likewise. * thread-db.c (thread_db_get_tls_address): Likewise. (disable_thread_event_reporting): Likewise. (remove_thread_event_breakpoints): Likewise. * tracepoint.c (gdb_agent_about_to_close): Renamed save_inferior as saved_thread. * target.h (set_desired_inferior): Renamed as... (set_desired_thread): New declaration. All uses updated. * server.c (myresume): Updated comment to reference thread instead of inferior. (handle_serial_event): Likewise. (handle_target_event): Likewise.
2014-09-10 11:37:11 +02:00
if (nto_inferior.ctl_fd != -1 && current_thread != NULL)
{
ptid_t ptid;
Rename current_inferior as current_thread in gdbserver GDB has a function named "current_inferior" and gdbserver has a global variable named "current_inferior", but the two are not equivalent; indeed, gdbserver does not have any real equivalent of what GDB calls an inferior. What gdbserver's "current_inferior" is actually pointing to is a structure describing the current thread. This commit renames current_inferior as current_thread in gdbserver to clarify this. It also renames the function "set_desired_inferior" to "set_desired_thread" and renames various local variables from foo_inferior to foo_thread. gdb/gdbserver/ChangeLog: * inferiors.h (current_inferior): Renamed as... (current_thread): New variable. All uses updated. * linux-low.c (get_pc): Renamed saved_inferior as saved_thread. (maybe_move_out_of_jump_pad): Likewise. (cancel_breakpoint): Likewise. (linux_low_filter_event): Likewise. (wait_for_sigstop): Likewise. (linux_resume_one_lwp): Likewise. (need_step_over_p): Likewise. (start_step_over): Likewise. (linux_stabilize_threads): Renamed save_inferior as saved_thread. * linux-x86-low.c (x86_linux_update_xmltarget): Likewise. * proc-service.c (ps_lgetregs): Renamed reg_inferior as reg_thread and save_inferior as saved_thread. * regcache.c (get_thread_regcache): Renamed saved_inferior as saved_thread. (regcache_invalidate_thread): Likewise. * remote-utils.c (prepare_resume_reply): Likewise. * thread-db.c (thread_db_get_tls_address): Likewise. (disable_thread_event_reporting): Likewise. (remove_thread_event_breakpoints): Likewise. * tracepoint.c (gdb_agent_about_to_close): Renamed save_inferior as saved_thread. * target.h (set_desired_inferior): Renamed as... (set_desired_thread): New declaration. All uses updated. * server.c (myresume): Updated comment to reference thread instead of inferior. (handle_serial_event): Likewise. (handle_target_event): Likewise.
2014-09-10 11:37:11 +02:00
ptid = thread_to_gdb_id (current_thread);
if (nto_set_thread (ptid))
{
const int watchmask = _DEBUG_FLAG_TRACE_RD | _DEBUG_FLAG_TRACE_WR
| _DEBUG_FLAG_TRACE_MODIFY;
procfs_status status;
int err;
err = devctl (nto_inferior.ctl_fd, DCMD_PROC_STATUS, &status,
sizeof (status), 0);
if (err == EOK && (status.flags & watchmask))
ret = 1;
}
}
TRACE ("%s: %s\n", __func__, ret ? "yes" : "no");
return ret;
}
/* Get instruction pointer for CURRENT_INFERIOR thread.
Return inferior's instruction pointer value, or 0 on error. */
static CORE_ADDR
nto_stopped_data_address (void)
{
CORE_ADDR ret = (CORE_ADDR)0;
TRACE ("%s\n", __func__);
Rename current_inferior as current_thread in gdbserver GDB has a function named "current_inferior" and gdbserver has a global variable named "current_inferior", but the two are not equivalent; indeed, gdbserver does not have any real equivalent of what GDB calls an inferior. What gdbserver's "current_inferior" is actually pointing to is a structure describing the current thread. This commit renames current_inferior as current_thread in gdbserver to clarify this. It also renames the function "set_desired_inferior" to "set_desired_thread" and renames various local variables from foo_inferior to foo_thread. gdb/gdbserver/ChangeLog: * inferiors.h (current_inferior): Renamed as... (current_thread): New variable. All uses updated. * linux-low.c (get_pc): Renamed saved_inferior as saved_thread. (maybe_move_out_of_jump_pad): Likewise. (cancel_breakpoint): Likewise. (linux_low_filter_event): Likewise. (wait_for_sigstop): Likewise. (linux_resume_one_lwp): Likewise. (need_step_over_p): Likewise. (start_step_over): Likewise. (linux_stabilize_threads): Renamed save_inferior as saved_thread. * linux-x86-low.c (x86_linux_update_xmltarget): Likewise. * proc-service.c (ps_lgetregs): Renamed reg_inferior as reg_thread and save_inferior as saved_thread. * regcache.c (get_thread_regcache): Renamed saved_inferior as saved_thread. (regcache_invalidate_thread): Likewise. * remote-utils.c (prepare_resume_reply): Likewise. * thread-db.c (thread_db_get_tls_address): Likewise. (disable_thread_event_reporting): Likewise. (remove_thread_event_breakpoints): Likewise. * tracepoint.c (gdb_agent_about_to_close): Renamed save_inferior as saved_thread. * target.h (set_desired_inferior): Renamed as... (set_desired_thread): New declaration. All uses updated. * server.c (myresume): Updated comment to reference thread instead of inferior. (handle_serial_event): Likewise. (handle_target_event): Likewise.
2014-09-10 11:37:11 +02:00
if (nto_inferior.ctl_fd != -1 && current_thread != NULL)
{
ptid_t ptid;
Rename current_inferior as current_thread in gdbserver GDB has a function named "current_inferior" and gdbserver has a global variable named "current_inferior", but the two are not equivalent; indeed, gdbserver does not have any real equivalent of what GDB calls an inferior. What gdbserver's "current_inferior" is actually pointing to is a structure describing the current thread. This commit renames current_inferior as current_thread in gdbserver to clarify this. It also renames the function "set_desired_inferior" to "set_desired_thread" and renames various local variables from foo_inferior to foo_thread. gdb/gdbserver/ChangeLog: * inferiors.h (current_inferior): Renamed as... (current_thread): New variable. All uses updated. * linux-low.c (get_pc): Renamed saved_inferior as saved_thread. (maybe_move_out_of_jump_pad): Likewise. (cancel_breakpoint): Likewise. (linux_low_filter_event): Likewise. (wait_for_sigstop): Likewise. (linux_resume_one_lwp): Likewise. (need_step_over_p): Likewise. (start_step_over): Likewise. (linux_stabilize_threads): Renamed save_inferior as saved_thread. * linux-x86-low.c (x86_linux_update_xmltarget): Likewise. * proc-service.c (ps_lgetregs): Renamed reg_inferior as reg_thread and save_inferior as saved_thread. * regcache.c (get_thread_regcache): Renamed saved_inferior as saved_thread. (regcache_invalidate_thread): Likewise. * remote-utils.c (prepare_resume_reply): Likewise. * thread-db.c (thread_db_get_tls_address): Likewise. (disable_thread_event_reporting): Likewise. (remove_thread_event_breakpoints): Likewise. * tracepoint.c (gdb_agent_about_to_close): Renamed save_inferior as saved_thread. * target.h (set_desired_inferior): Renamed as... (set_desired_thread): New declaration. All uses updated. * server.c (myresume): Updated comment to reference thread instead of inferior. (handle_serial_event): Likewise. (handle_target_event): Likewise.
2014-09-10 11:37:11 +02:00
ptid = thread_to_gdb_id (current_thread);
if (nto_set_thread (ptid))
{
procfs_status status;
if (devctl (nto_inferior.ctl_fd, DCMD_PROC_STATUS, &status,
sizeof (status), 0) == EOK)
ret = status.ip;
}
}
TRACE ("%s: 0x%08lx\n", __func__, ret);
return ret;
}
/* We do not currently support non-stop. */
static int
nto_supports_non_stop (void)
{
TRACE ("%s\n", __func__);
return 0;
}
static struct target_ops nto_target_ops = {
nto_create_inferior,
Initialise target descrption after skipping extra traps for --wrapper Nowadays, when --wrapper is used, GDBserver skips extra traps/stops in the wrapper program, and stops at the first instruction of the program to be debugged. However, GDBserver created target description in the first stop of inferior, and the executable of the inferior is the wrapper program rather than the program to be debugged. In this way, the target description can be wrong if the architectures of wrapper program and program to be debugged are different. This is shown by some fails in gdb.server/wrapper.exp on buildbot. We are testing i686-linux GDB (Fedora-i686) on an x86_64-linux box (fedora-x86-64-4) in buildbot, such configuration causes fails in gdb.server/wrapper.exp like this: spawn /home/gdb-buildbot-2/fedora-x86-64-4/fedora-i686/build/gdb/testsuite/../../gdb/gdbserver/gdbserver --once --wrapper env TEST=1 -- :2346 /home/gdb-buildbot-2/fedora-x86-64-4/fedora-i686/build/gdb/testsuite/outputs/gdb.server/wrapper/wrapper Process /home/gdb-buildbot-2/fedora-x86-64-4/fedora-i686/build/gdb/testsuite/outputs/gdb.server/wrapper/wrapper created; pid = 8795 Can't debug 64-bit process with 32-bit GDBserver Exiting target remote localhost:2346 localhost:2346: Connection timed out. (gdb) FAIL: gdb.server/wrapper.exp: setting breakpoint at marker See https://sourceware.org/ml/gdb-testers/2015-q3/msg01541.html In this case, program to be debugged ("wrapper") is 32-bit but wrapper program ("/usr/bin/env") is 64-bit, so GDBserver gets the 64-bit target description instead of 32-bit. The root cause of this problem is that GDBserver creates target description too early, and the rationale of fix could be creating target description once the GDBserver skips extra traps and inferior stops at the first instruction of the program we want to debug. IOW, when GDBserver skips extra traps, the inferior's tdesc is NULL, and mywait and its callees shouldn't use inferior's tdesc, so in this patch, we skip code that requires register access, see changes in linux_resume_one_lwp_throw and need_step_over_p. In linux_low_filter_event, if target description isn't initialised and GDBserver attached the process, we create target description immediately, because GDBserver don't have to skip extra traps for attach, IOW, it makes no sense to use --attach and --wrapper together. Otherwise, the process is launched by GDBserver, we keep the status pending, and return. After GDBserver skipped extra traps in start_inferior, we call a target_ops hook arch_setup to initialise target description there. gdb/gdbserver: 2015-07-24 Yao Qi <yao.qi@linaro.org> * linux-low.c (linux_arch_setup): New function. (linux_low_filter_event): If proc->tdesc is NULL and proc->attached is true, call the_low_target.arch_setup. Otherwise, keep status pending, and return. (linux_resume_one_lwp_throw): Don't call get_pc if thread->while_stepping isn't NULL. Don't call get_thread_regcache if proc->tdesc is NULL. (need_step_over_p): Return 0 if proc->tdesc is NULL. (linux_target_ops): Install arch_setup. * server.c (start_inferior): Call the_target->arch_setup. * target.h (struct target_ops) <arch_setup>: New field. (target_arch_setup): New marco. * lynx-low.c (lynx_target_ops): Update. * nto-low.c (nto_target_ops): Update. * spu-low.c (spu_target_ops): Update. * win32-low.c (win32_target_ops): Update.
2015-07-24 15:40:34 +02:00
NULL, /* arch_setup */
nto_attach,
nto_kill,
nto_detach,
nto_mourn,
NULL, /* nto_join */
nto_thread_alive,
nto_resume,
nto_wait,
nto_fetch_registers,
nto_store_registers,
NULL, /* prepare_to_access_memory */
NULL, /* done_accessing_memory */
nto_read_memory,
nto_write_memory,
NULL, /* nto_look_up_symbols */
nto_request_interrupt,
nto_read_auxv,
[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
nto_supports_z_point_type,
nto_insert_point,
nto_remove_point,
NULL, /* stopped_by_sw_breakpoint */
NULL, /* supports_stopped_by_sw_breakpoint */
NULL, /* stopped_by_hw_breakpoint */
NULL, /* supports_stopped_by_hw_breakpoint */
[gdbserver] Disable conditional breakpoints on no-hardware-single-step targets GDBserver steps over breakpoint if the condition is false, but if target doesn't support hardware single step, the step over is very simple, if not incorrect, in linux-arm-low.c: /* We only place breakpoints in empty marker functions, and thread locking is outside of the function. So rather than importing software single-step, we can just run until exit. */ static CORE_ADDR arm_reinsert_addr (void) { struct regcache *regcache = get_thread_regcache (current_thread, 1); unsigned long pc; collect_register_by_name (regcache, "lr", &pc); return pc; } and linux-mips-low.c does the same. GDBserver sets a breakpoint at the return address of the current function, resume and wait the program hits the breakpoint in order to achieve "breakpoint step over". What if program hits other user breakponits during this "step over"? It is worse if the arm/thumb interworking is considered. Nowadays, GDBserver arm backend unconditionally inserts arm breakpoint, /* Define an ARM-mode breakpoint; we only set breakpoints in the C library, which is most likely to be ARM. If the kernel supports clone events, we will never insert a breakpoint, so even a Thumb C library will work; so will mixing EABI/non-EABI gdbserver and application. */ (const unsigned char *) &arm_breakpoint, (const unsigned char *) &arm_eabi_breakpoint, note that the comments are no longer valid as C library can be compiled in thumb mode. When GDBserver steps over a breakpoint in arm mode function, which returns to thumb mode, GDBserver will insert arm mode breakpoint by mistake and the program will crash. GDBserver alone is unable to determine the arm/thumb mode given a PC address. See how GDB does it in arm-tdep.c:arm_pc_is_thumb. After thinking about how to teach GDBserver inserting right breakpoint (arm or thumb) for a while, I reconsider it from a different direction that it may be unreasonable to run target-side conditional breakpoint for targets without hardware single step. Pedro also pointed this out here https://sourceware.org/ml/gdb-patches/2015-04/msg00337.html This patch is to add a new target_ops hook supports_conditional_breakpoints, and only reply ";ConditionalBreakpoints+" if it is true. On linux targets, supports_conditional_breakpoints returns true if target has hardware single step, on other targets, (win32, lynx, nto, spu), set it to NULL, because conditional breakpoint is a linux-specific feature. gdb/gdbserver: 2015-05-08 Yao Qi <yao.qi@linaro.org> * linux-low.c (linux_supports_conditional_breakpoints): New function. (linux_target_ops): Install new target method. * lynx-low.c (lynx_target_ops): Install NULL hook for supports_conditional_breakpoints. * nto-low.c (nto_target_ops): Likewise. * spu-low.c (spu_target_ops): Likewise. * win32-low.c (win32_target_ops): Likewise. * server.c (handle_query): Check target_supports_conditional_breakpoints. * target.h (struct target_ops) <supports_conditional_breakpoints>: New field. (target_supports_conditional_breakpoints): New macro.
2015-05-08 13:29:13 +02:00
/* Although nto has hardware single step, still disable this
feature for not, because it is implemented in linux-low.c instead
of in generic code. */
NULL, /* supports_conditional_breakpoints */
nto_stopped_by_watchpoint,
nto_stopped_data_address,
NULL, /* nto_read_offsets */
NULL, /* thread_db_set_tls_address */
NULL,
hostio_last_error_from_errno,
NULL, /* nto_qxfer_osdata */
NULL, /* xfer_siginfo */
nto_supports_non_stop,
NULL, /* async */
NULL /* start_non_stop */
};
/* Global function called by server.c. Initializes QNX Neutrino
gdbserver. */
void
initialize_low (void)
{
sigset_t set;
TRACE ("%s\n", __func__);
set_target_ops (&nto_target_ops);
set_breakpoint_data (the_low_target.breakpoint,
the_low_target.breakpoint_len);
/* We use SIGUSR1 to gain control after we block waiting for a process.
We use sigwaitevent to wait. */
sigemptyset (&set);
sigaddset (&set, SIGUSR1);
sigprocmask (SIG_BLOCK, &set, NULL);
}