The /m modifier interleaves source lines with the disassembly of recorded
instructions. This calls disasm.c's gdb_disassembly once for each recorded
instruction to be printed.
This doesn't really work because gdb_disassembly may choose not to print
anything in some situations. And if it does print something, the output
interferes with btrace_insn_history's output around it.
It further results in a separate asm_insns list for each instruction in MI.
Even though there is no MI support for target record, yet, we fix this obvious
issue.
Change record instruction-history /m to use the new gdb_pretty_print_insn
function for printing a single instruction and interleave source lines as
appropriate.
We cannot reuse the new disasm.c do_mixed_source_and_assembly function without
significant changes to it.
gdb/
* record-btrace.c (struct btrace_line_range): New.
(btrace_mk_line_range, btrace_line_range_add)
(btrace_line_range_is_empty, btrace_line_range_contains_range)
(btrace_find_line_range, btrace_print_lines): New.
(btrace_insn_history): Add source interleaving algorithm.
The "record instruction-history" command prints for each instruction in
addition to the instruction's disassembly:
- the instruction number in the recorded execution trace
- a '?' before the instruction if it was executed speculatively
To allow the "record instruction-history" command to use GDB's disassembly
infrastructure, we extend gdb_pretty_print_insn to optionally print those
additional fields and export the function.
Add a new struct disasm_insn to add additional fields describing the
to-be-disassembled instruction. The additional fields are:
number an optional instruction number, zero if omitted.
is_speculative a predicate saying whether the instruction was
executed speculatively.
If non-zero, the instruction number is printed first. It will also appear
as a new optional field "insn-number" in MI. The field will be present if
insn_num is non-zero.
If is_speculative is set, speculative execution will be indicated by a "?"
following the new instruction number field. Unless the PC is omitted, it
will overwrite the first byte of the PC prefix. It will appear as a new
optional field "is-speculative" in MI. The field will contain "?" and will
be present if is_speculative is set.
The speculative execution indication is guarded by a new flag
DISASSEMBLY_SPECULATION.
Replace the PC parameter of gdb_pretty_print_insn with a pointer to the above
struct. GDB's "disassemble" command does not use the new fields.
gdb/
* disasm.h (DISASSEMBLY_SPECULATION): New.
(struct disasm_insn): New.
(gdb_pretty_print_insn): New.
* disasm.c (gdb_pretty_print_insn): Replace parameter PC with INSN.
Update users. Print instruction number and indicate speculative
execution, if requested.
Split disasm.c's dump_insn into two parts:
- print a single instruction
- loop over the specified address range
The first part will be refined in subsequent patches so it can be reused.
gdb/
* disasm.c (dump_insns): Split into this and ...
(gdb_pretty_print_insn): ... this.
This actually fixes the build in C:
/home/simark/src/binutils-gdb/gdb/xtensa-linux-nat.c💯1: error: no previous prototype for ‘supply_gregset_reg’ [-Werror=missing-prototypes]
supply_gregset_reg (struct regcache *regcache,
^
/home/simark/src/binutils-gdb/gdb/xtensa-linux-nat.c:257:1: error: no previous prototype for ‘xtensa_linux_fetch_inferior_registers’ [-Werror=missing-prototypes]
xtensa_linux_fetch_inferior_registers (struct target_ops *ops,
^
/home/simark/src/binutils-gdb/gdb/xtensa-linux-nat.c:272:1: error: no previous prototype for ‘xtensa_linux_store_inferior_registers’ [-Werror=missing-prototypes]
xtensa_linux_store_inferior_registers (struct target_ops *ops,
^
cc1: all warnings being treated as errors
These functions are local to this file, so they should be static.
gdb/ChangeLog:
* xtensa-linux-nat.c (supply_gregset_reg): Make static.
(xtensa_linux_fetch_inferior_registers): Likewise.
(xtensa_linux_store_inferior_registers): Likewise.
Fixes
/home/simark/src/binutils-gdb/gdb/gdbserver/linux-mips-low.c: In function ‘void mips_add_watchpoint(arch_process_info*, CORE_ADDR, int, int)’:
/home/simark/src/binutils-gdb/gdb/gdbserver/linux-mips-low.c:368:19: error: invalid conversion from ‘int’ to ‘target_hw_bp_type’ [-fpermissive]
new_watch->type = watch_type;
^
gdb/gdbserver/ChangeLog:
* linux-mips-low.c (mips_linux_new_thread): Change type of
watch_type to enum target_hw_bp_type.
Trivial casts for C++.
Fixes things like
In file included from /home/simark/src/binutils-gdb/gdb/gdbserver/../common/common-defs.h:39:0,
from /home/simark/src/binutils-gdb/gdb/gdbserver/server.h:22,
from /home/simark/src/binutils-gdb/gdb/gdbserver/linux-arm-low.c:19:
/home/simark/src/binutils-gdb/gdb/gdbserver/linux-arm-low.c: In function ‘int arm_get_hwcap(long unsigned int*)’:
/home/simark/src/binutils-gdb/gdb/gdbserver/../../include/libiberty.h:711:38: error: invalid conversion from ‘void*’ to ‘unsigned char*’ [-fpermissive]
# define alloca(x) __builtin_alloca(x)
^
/home/simark/src/binutils-gdb/gdb/gdbserver/linux-arm-low.c:807:25: note: in expansion of macro ‘alloca’
unsigned char *data = alloca (8);
^
gdb/gdbserver/ChangeLog:
* linux-aarch32-low.c (arm_fill_gregset): Add cast.
(arm_store_gregset): Likewise.
* linux-arm-low.c (arm_get_hwcap): Likewise.
(arm_read_description): Likewise.
Mostly some casts from "generic arg" void* to the actual type.
There are two (enum gdb_signal) casts. I tried to see if it would have
been better to change the type of sigrc, but it has a double role, as an
enum and as an integer, so I left it as is.
gdb/ChangeLog:
* remote-sim.c (check_for_duplicate_sim_descriptor): Add casts.
(get_sim_inferior_data): Likewise.
(sim_inferior_data_cleanup): Likewise.
(gdbsim_close_inferior): Likewise.
(gdbsim_resume_inferior): Likewise.
(gdbsim_wait): Likewise.
(simulator_command): Likewise.
(sim_command_completer): Likewise.
g++ doesn't like that we forward-declare a variable that is initialized
later in the file. It's easy enough to re-order things to fix it.
Fixes
/home/simark/src/binutils-gdb/gdb/gdbserver/linux-ppc-low.c:663:28: error: redefinition of ‘usrregs_info ppc_usrregs_info’
static struct usrregs_info ppc_usrregs_info =
^
/home/simark/src/binutils-gdb/gdb/gdbserver/linux-ppc-low.c:381:28: note: ‘usrregs_info ppc_usrregs_info’ previously declared here
static struct usrregs_info ppc_usrregs_info;
^
gdb/gdbserver/ChangeLog:
* linux-ppc-low.c (ppc_usrregs_info): Remove
forward-declaration.
(ppc_arch_setup): Move lower in file.
readelf ought to notify when a symbol wrongly has both a version
definition and a needed version. This patch does that, and removes
the heuristic that only defined symbols in SHT_NOBITS sections have
verneed entries.
* readelf (process_version_sections): Check DT_VERNEED and
DT_VERDEF for all symbols. Report "*both*" should a symbol
have both a verneed and verdef.
(get_symbol_version_string): Reduce indentation by early
exits. Don't use SHT_NOBITS heuristic to detect case where a
defined symbol has a verneed entry.
continue_to_breakpoint always continues to the next breakpoint, not to the
one named in parameter. This rendered the tests effectively useless, since
marker2 was never reached.
gdb/testsuite/ChangeLog:
* gdb.reverse/fstatat-reverse.exp: Set breakpoint on marker1 after
reaching marker2.
* gdb.reverse/getresuid-reverse.exp: Likewise.
* gdb.reverse/pipe-reverse.exp: Likewise.
* gdb.reverse/readv-reverse.exp: Likewise.
* gdb.reverse/recvmsg-reverse.exp: Likewise.
* gdb.reverse/time-reverse.exp: Likewise.
* gdb.reverse/waitpid-reverse.exp: Likewise and add KFAILs.
opcode * rx.h (enum RX_Operand_Type): Add RX_Operand_Zero_Indirect.
opcodes * rx-decode.opc (rx_disp): If the displacement is zero, set the
type to RX_Operand_Zero_Indirect.
* rx-decode.c: Regenerate.
* rx-dis (print_insn): Handle RX_Operand_Zero_Indirect.
gas * config/rx-parse.y: Allow zero value for 5-bit displacements.
tests * gas/rx/mov.sm: Add tests for zero offset indirect moves.
* gas/rx/mov.d: Update expected output.
Assuming displaced stepping is enabled, and a breakpoint is set in the
memory region of the scratch pad, things break. One of two cases can
happen:
#1 - The breakpoint wasn't inserted yet (all threads were stopped), so
after setting up the displaced stepping scratch pad with the
adjusted copy of the instruction we're trying to single-step, we
insert the breakpoint, which corrupts the scratch pad, and the
inferior executes the wrong instruction. (Example below.)
This is clearly unacceptable.
#2 - The breakpoint was already inserted, so setting up the displaced
stepping scratch pad overwrites the breakpoint. This is OK in
the sense that we already assume that no thread is going to
executes the code in the scratch pad range (after initial
startup) anyway.
This commit addresses both cases by simply punting on displaced
stepping if we have a breakpoint in the scratch pad range.
The #1 case above explains a few regressions exposed by the AS/NS
series on x86:
Running ./gdb.dwarf2/callframecfa.exp ...
FAIL: gdb.dwarf2/callframecfa.exp: set display for call-frame-cfa
FAIL: gdb.dwarf2/callframecfa.exp: step 1 for call-frame-cfa
FAIL: gdb.dwarf2/callframecfa.exp: step 2 for call-frame-cfa
FAIL: gdb.dwarf2/callframecfa.exp: step 3 for call-frame-cfa
FAIL: gdb.dwarf2/callframecfa.exp: step 4 for call-frame-cfa
Running ./gdb.dwarf2/typeddwarf.exp ...
FAIL: gdb.dwarf2/typeddwarf.exp: continue to breakpoint: continue to typeddwarf.c:53
FAIL: gdb.dwarf2/typeddwarf.exp: check value of x at typeddwarf.c:53
FAIL: gdb.dwarf2/typeddwarf.exp: check value of y at typeddwarf.c:53
FAIL: gdb.dwarf2/typeddwarf.exp: check value of z at typeddwarf.c:53
FAIL: gdb.dwarf2/typeddwarf.exp: continue to breakpoint: continue to typeddwarf.c:73
FAIL: gdb.dwarf2/typeddwarf.exp: check value of w at typeddwarf.c:73
FAIL: gdb.dwarf2/typeddwarf.exp: check value of x at typeddwarf.c:73
FAIL: gdb.dwarf2/typeddwarf.exp: check value of y at typeddwarf.c:73
FAIL: gdb.dwarf2/typeddwarf.exp: check value of z at typeddwarf.c:73
Enabling "maint set target-non-stop on" implies displaced stepping
enabled as well, and it's the latter that's to blame here. We can see
the same failures with "maint set target-non-stop off + set displaced
on".
Diffing (good/bad) gdb.log for callframecfa.exp shows:
@@ -99,29 +99,29 @@ Breakpoint 2 at 0x80481b0: file q.c, lin
continue
Continuing.
-Breakpoint 2, func (arg=77) at q.c:2
+Breakpoint 2, func (arg=52301) at q.c:2
2 in q.c
(gdb) PASS: gdb.dwarf2/callframecfa.exp: continue to breakpoint: continue to breakpoint for call-frame-cfa
display arg
-1: arg = 77
-(gdb) PASS: gdb.dwarf2/callframecfa.exp: set display for call-frame-cfa
+1: arg = 52301
+(gdb) FAIL: gdb.dwarf2/callframecfa.exp: set display for call-frame-cfa
The problem is here, when setting up the func call:
Breakpoint 1, main (argc=-13345, argv=0x0) at q.c:7
7 in q.c
(gdb) disassemble
Dump of assembler code for function main:
0x080481bb <+0>: push %ebp
0x080481bc <+1>: mov %esp,%ebp
0x080481be <+3>: sub $0x4,%esp
=> 0x080481c1 <+6>: movl $0x4d,(%esp)
0x080481c8 <+13>: call 0x80481b0 <func>
0x080481cd <+18>: leave
0x080481ce <+19>: ret
End of assembler dump.
(gdb) disassemble /r
Dump of assembler code for function main:
0x080481bb <+0>: 55 push %ebp
0x080481bc <+1>: 89 e5 mov %esp,%ebp
0x080481be <+3>: 83 ec 04 sub $0x4,%esp
=> 0x080481c1 <+6>: c7 04 24 4d 00 00 00 movl $0x4d,(%esp)
0x080481c8 <+13>: e8 e3 ff ff ff call 0x80481b0 <func>
0x080481cd <+18>: c9 leave
0x080481ce <+19>: c3 ret
End of assembler dump.
Note the breakpoint at main is set at 0x080481c1. Right at the
instruction that sets up func's argument. Executing that instruction
should write 0x4d to the address pointed at by $esp. However, if we
stepi, the program manages to write 52301/0xcc4d there instead (0xcc
is int3, the x86 breakpoint instruction), because the breakpoint
address is 4 bytes inside the scratch pad location, which is
0x080481bd:
(gdb) p 0x080481c1 - 0x080481bd
$1 = 4
IOW, instead of executing:
"c7 04 24 4d 00 00 00" [ movl $0x4d,(%esp) ]
the inferior executes:
"c7 04 24 4d cc 00 00" [ movl $0xcc4d,(%esp) ]
gdb/ChangeLog:
2015-10-30 Pedro Alves <palves@redhat.com>
* breakpoint.c (breakpoint_in_range_p)
(breakpoint_location_address_range_overlap): New functions.
* breakpoint.h (breakpoint_in_range_p): New declaration.
* infrun.c (displaced_step_prepare_throw): If there's a breakpoint
in the scratch pad range, don't displaced step.
Nowadays, test gdb.threads/wp-replication.exp uses a while loop to
repeatedly insert HW watchpoint, resume and check no error message
coming out, in order to count HW watchpoints There are some
drawbacks in this way,
- the loop could be endless. I think this is use to making trouble
to S/390, since we had such comment
# Some targets (like S/390) behave as though supporting
# unlimited hardware watchpoints. In this case we just take a
# safe exit out of the loop.
I hit this today too because a GDB internal error is triggered
on "continue" in the loop, and $done is 0 invariantly, so the loop
can't end.
- the code counting hardware watchpoint is too complicated. We can
use "set breakpoint always-inserted on" to get the result of inserting
HW watchpoint without resuming the inferior. In this way,
watch_count_done and empty_cycle in c file is no longer needed.
In this patch, I change to use "set breakpoint always-inserted on" trick,
and only iterate $NR_THREADS times, to count the HW watchpoint. In this
way, the loop can't be endless, and GDB doesn't need to resume the inferior.
gdb/testsuite:
2015-10-30 Yao Qi <yao.qi@linaro.org>
* gdb.threads/wp-replication.c (watch_count_done): Remove.
(empty_cycle): Remove.
(main): Don't call empty_cycle. Don't use watch_count_done.
* gdb.threads/wp-replication.exp: Don't set breakpoint on
empty_cycle. Rewrite the code counting HW watchpoints.
60 bytes is the size of glibc's struct termios, the one used by kernel is
36 bytes long.
gdb/ChangeLog:
* aarch64-linux-tdep.c (aarch64_linux_init_abi): Fix size_termios.
* amd64-linux-tdep.c (amd64_linux_init_abi): Fix size_termios.
(amd64_x32_linux_init_abi): Fix size_termios.
We have to use extract_unsigned_integer to read paramaters structure - target
pointers can have different endianness and size.
gdb/ChangeLog:
* linux-record.c (record_linux_system_call): Fix old_select.
Memory size for getgroups16 needs to be multiplied by entry count, and only
needs recording if the pointer is non-NULL. setgroups16, on the other hand,
doesn't write to user memory and doesn't need special handling at all.
gdb/ChangeLog:
* linux-record.c (record_linux_system_call): Fix [gs]etgroups16.
The code failed to account for padding between the int and subsequent
pointer present on 64-bit architectures.
gdb/ChangeLog:
* linux-record.c (record_linux_msghdr): Fix msg_namelen handling.
getdents buffer size is given in bytes, not dirent entries (which have
variable size anyway). We don't need size_dirent and size_dirent64 for
this reason.
readdir, on the other hand, needs size of old_linux_dirent, which is
a somewhat different structure. Accordingly, rename size_dirent
to size_old_dirent.
gdb/ChangeLog:
* aarch64-linux-tdep.c (aarch64_linux_init_abi): Remove
size_dirent{,64}, add size_old_dirent.
* amd64-linux-tdep.c (amd64_linux_init_abi): Remove size_dirent{,64},
add size_old_dirent.
(amd64_x32_linux_init_abi): Remove size_dirent{,64}, add
size_old_dirent.
* arm-linux-tdep.c (arm_linux_init_abi): Remove size_dirent{,64},
add size_old_dirent.
* i386-linux-tdep.c (i386_linux_init_abi): Remove size_dirent{,64},
add size_old_dirent.
* linux-record.c (record_linux_system_call): Fix handling of readdir
and getdents{,64}.
* linux-record.h (struct linux_record_tdep): Remove size_dirent{,64},
add size_old_dirent.
* ppc-linux-tdep.c (ppc_init_linux_record_tdep): Remove
size_dirent{,64}, add size_old_dirent.
i386 and arm wrongly set them to 2, when it should be 4. size_[ug]id is used
by getgroups32 etc syscalls, while size_old_[ug]id is used for getgroups16
and friends.
gdb/ChangeLog:
* arm-linux-tdep.c (arm_linux_init_abi): Fix size_[ug]id.
* i386-linux-tdep.c (i386_linux_init_abi): Fix size_[ug]id.
gdb/testsuite/ChangeLog:
* gdb.reverse/fstatat-reverse.c: New test.
* gdb.reverse/fstatat-reverse.exp: New file.
* gdb.reverse/getresuid-reverse.c: New test.
* gdb.reverse/getresuid-reverse.exp: New file.
* gdb.reverse/pipe-reverse.c: New test.
* gdb.reverse/pipe-reverse.exp: New file.
* gdb.reverse/readv-reverse.c: New test.
* gdb.reverse/readv-reverse.exp: New file.
* gdb.reverse/recvmsg-reverse.c: New test.
* gdb.reverse/recvmsg-reverse.exp: New file.
* gdb.reverse/time-reverse.c: New test.
* gdb.reverse/time-reverse.exp: New file.
* gdb.reverse/waitpid-reverse.c: New test.
* gdb.reverse/waitpid-reverse.exp: New file.
Fixes on i386:
../../../binutils-gdb/gdb/gdbserver/proc-service.c: In function ps_pdread:
../../../binutils-gdb/gdb/gdbserver/proc-service.c:83:25: error: cast from pointer to integer of different size [-Werror=pointer-to-int-cast]
../../../binutils-gdb/gdb/gdbserver/proc-service.c: In function ps_pdwrite:
../../../binutils-gdb/gdb/gdbserver/proc-service.c:93:30: error: cast from pointer to integer of different size [-Werror=pointer-to-int-cast]
I could have kept both casts:
(CORE_ADDR) (uintptr_t) addr
but it's cleaner this way. The uintptr_t implicitely gets promoted to a
CORE_ADDR, which is at least as long as uintptr_t.
gdb/gdbserver/ChangeLog:
* proc-service.c (ps_pdread): Change CORE_ADDR cast to uintptr_t.
(ps_pdwrite): Likewise.
gdb/gdbserver/ChangeLog:
* linux-arm-low.c (arm_new_thread): Move pointer dereference
to after assert checks.
Signed-off-by: Henrik Wallin <henrik.wallin@windriver.com>
The casts are required because ps_pd{read,write} must respect a fixed
interface.
gdb/gdbserver/ChangeLog:
* proc-service.c (ps_pdread): Add/adjust casts.
(ps_pdwrite): Add/adjust casts.
This fixes this error in C++ mode:
/home/pedro/gdb/mygit/cxx-convertion/src/gdb/mdebugread.c:654:11: error: invalid conversion from ‘int’ to ‘address_class’ [-fpermissive]
theclass = mdebug_register_index;
^
The "theclass" local is of type enum address_class, however, what it
really holds is an address class index. Class index values by design
match the address class values up until LOC_FINAL_VALUE, but extend
beyond that, so it's not really right to store an address class index
in an enum address_class.
The fix is really the same making the 'theclass' local be of type int,
but while we're at it, we get rid of the goto, and thus the local
becomes the 'aclass_index' parameter in the new add_data_symbol
function.
gdb/ChangeLog:
2015-10-29 Pedro Alves <palves@redhat.com>
* mdebugread.c (add_data_symbol): New function, factored out from
...
(parse_symbol): ... here. Delete 'theclass' local.