This patch change aarch prologue analyzer using code cache, in order
to improve the performance of remote debugging.
gdb.perf/skip-prologue.exp (measured by wall-time) is improved when
the program is compiled without debug information.
Original Patched Original Patched
without dbg without dbg with dbg with dbg
/ 11.1635239124 9.99472999573 9.65339517593 9.66648793221
-fstack-protector-all 11.2560930252 9.338118 9.63896489143 9.59474396706
gdb:
2016-12-9 Yao Qi <yao.qi@linaro.org>
* aarch64-tdep.c (instruction_reader::read): Call
read_code_unsigned_integer instead of
read_memory_unsigned_integer.
This patch teaches GDB AArch64 backend to recognize STR instructions
in prologue, like 'str x19, [sp, #-48]!' or 'str w0, [sp, #44]'.
The unit test is added too.
gdb:
2016-12-02 Yao Qi <yao.qi@linaro.org>
Pedro Alves <palves@redhat.com>
* aarch64-tdep.c (aarch64_analyze_prologue): Recognize STR
instruction.
(aarch64_analyze_prologue_test): More tests.
We don't have an effective way to test prologue analyzer which is
highly dependent on instruction patterns in prologue generated by
compiler. GDB prologue analyzer may not handle the new sequences
generated by new compiler, or may still handle some sequences that
generated by very old compilers which are no longer used. The
former is a functionality issue, while the latter is a maintenance
issue.
The input and output of prologue analyzer is quite clear, so it
fits for unit test. The input is series of instructions, and the
output are 1) where prologue end, 2) where registers are saved.
In aarch64, they are represented in 'struct aarch64_prologue_cache'.
This patch refactors aarch64_analyze_prologue so it can read
instructions from either real target or test harness. In unit
test aarch64_analyze_prologue_test, aarch64_analyze_prologue gets
instructions we prepared in the test, as the input of prologue
analyzer. Then, we checked various fields in
'struct aarch64_prologue_cache'.
gdb:
2016-12-02 Yao Qi <yao.qi@linaro.org>
Pedro Alves <palves@redhat.com>
* aarch64-tdep.c: Include "selftest.h".
(abstract_instruction_reader): New class.
(instruction_reader): New class.
(aarch64_analyze_prologue): Add new parameter reader. Call
reader.read instead of read_memory_unsigned_integer.
[GDB_SELF_TEST] (instruction_reader_test): New class.
(aarch64_analyze_prologue_test): New function.
(_initialize_aarch64_tdep) [GDB_SELF_TEST]: Register
selftests::aarch64_analyze_prologue_test.
* trad-frame.c (trad_frame_cache_zalloc):
(trad_frame_alloc_saved_regs): Add a new function.
* trad-frame.h (trad_frame_alloc_saved_regs): Declare.
Use regcache in software_single_step.
gdb:
2016-11-22 Yao Qi <yao.qi@linaro.org>
* aarch64-tdep.c (aarch64_software_single_step): Call
get_regcache_arch instead of get_frame_arch. Call
regcache_read_pc instead of get_frame_pc.
Both of them are used in conversion. We can remove them since the
conversion is done.
There are many architectures only have one breakpoint instruction,
so their gdbarch methods breakpoint_kind_from_pc and
sw_breakpoint_from_kind look very similar. Instead of macro, we
use template "template <size_t, const gdb_byte *> struct bp_manipulation"
for these architectures. In order to use template, I also change
breakpoint instruction of type "static const gdb_byte[]" to
"constexpr gdb_byte[]", and rename them to ARCH_break_insn.
gdb:
2016-11-03 Yao Qi <yao.qi@linaro.org>
Pedro Alves <palves@redhat.com>
* aarch64-tdep.c (aarch64_default_breakpoint): Change it to
constexpr. Don't use GDBARCH_BREAKPOINT_MANIPULATION.
(aarch64_gdbarch_init): Don't use
SET_GDBARCH_BREAKPOINT_MANIPULATION.
* alpha-tdep.c (break_insn): Rename to alpha_break_insn.
Don't use GDBARCH_BREAKPOINT_MANIPULATION.
(alpha_gdbarch_init): Don't use
SET_GDBARCH_BREAKPOINT_MANIPULATION.
* arc-tdep.c (arc_gdbarch_init): Don't use
SET_GDBARCH_BREAKPOINT_MANIPULATION.
* arch-utils.h (GDBARCH_BREAKPOINT_MANIPULATION): Remove.
(struct bp_manipulation): New.
(SET_GDBARCH_BREAKPOINT_MANIPULATION): Remove.
(struct bp_manipulation_endian): New.
(BP_MANIPULATION): New.
(BP_MANIPULATION_ENDIAN): New.
* arm-tdep.c (arm_gdbarch_init): Don't use
SET_GDBARCH_BREAKPOINT_MANIPULATION.
* avr-tdep.c (avr_break_insn): Change it constexpr.
(avr_gdbarch_init): Don't use
SET_GDBARCH_BREAKPOINT_MANIPULATION.
* bfin-tdep.c (bfin_gdbarch_init): Likewise.
* cris-tdep.c (cris_gdbarch_init): Likewise.
* frv-tdep.c (breakpoint): Rename it to frv_break_insn, and
change its type to constexpr. Don't use
GDBARCH_BREAKPOINT_MANIPULATION.
(frv_gdbarch_init): Don't use
SET_GDBARCH_BREAKPOINT_MANIPULATION.
* ft32-tdep.c (breakpoint): Rename it to ft32_break_insn and
change its type to constexpr. Don't use
GDBARCH_BREAKPOINT_MANIPULATION.
(ft32_gdbarch_init): Don't use
SET_GDBARCH_BREAKPOINT_MANIPULATION.
* h8300-tdep.c (breakpoint): Rename it to h8300_break_insn.
Don't use GDBARCH_BREAKPOINT_MANIPULATION.
(h8300_gdbarch_init): Don't use
SET_GDBARCH_BREAKPOINT_MANIPULATION.
* hppa-tdep.c (breakpoint): Rename it to h8300_break_insn.
Don't use GDBARCH_BREAKPOINT_MANIPULATION.
(hppa_gdbarch_init): Don't use
SET_GDBARCH_BREAKPOINT_MANIPULATION.
* i386-tdep.c (break_insn): Rename it to i386_break_insn.
Don't use GDBARCH_BREAKPOINT_MANIPULATION.
(i386_gdbarch_init): Don't use
SET_GDBARCH_BREAKPOINT_MANIPULATION.
* iq2000-tdep.c (iq2000_gdbarch_init): Don't use
SET_GDBARCH_BREAKPOINT_MANIPULATION.
* lm32-tdep.c (breakpoint): Rename it to lm32_break_insn and
change its type to constexpr. Don't use
GDBARCH_BREAKPOINT_MANIPULATION.
(lm32_gdbarch_init): Don't use
SET_GDBARCH_BREAKPOINT_MANIPULATION.
* m32c-tdep.c (break_insn): Rename it to m32c_break_insn and change
its type to constexpr. Don't use GDBARCH_BREAKPOINT_MANIPULATION.
(m32c_gdbarch_init): Don't use
SET_GDBARCH_BREAKPOINT_MANIPULATION.
* m32r-tdep.c (m32r_gdbarch_init): Likewise.
* m68hc11-tdep.c (breakpoint): Rename it to m68hc11_break_insn and
change its type to constexpr. Don't use GDBARCH_BREAKPOINT_MANIPULATION.
(m68hc11_gdbarch_init): Don't use SET_GDBARCH_BREAKPOINT_MANIPULATION.
* m68k-tdep.c (break_insn): Rename it to m68k_break_insn and change
its type to constexpr. Don't use GDBARCH_BREAKPOINT_MANIPULATION.
(m68k_gdbarch_init): Don't use
SET_GDBARCH_BREAKPOINT_MANIPULATION.
* m88k-tdep.c (break_insn): Rename it to m88k_break_insn and change
its type to constexpr. Don't use GDBARCH_BREAKPOINT_MANIPULATION.
(m88k_gdbarch_init): Don't use
SET_GDBARCH_BREAKPOINT_MANIPULATION.
* mep-tdep.c (breakpoint): Rename it to mep_break_insn and change
its type to constexpr. Don't use GDBARCH_BREAKPOINT_MANIPULATION.
(mep_gdbarch_init): Don't use
SET_GDBARCH_BREAKPOINT_MANIPULATION.
* microblaze-tdep.c (break_insn): Rename it to
microblaze_break_insn and change its type to constexpr. Don't use
GDBARCH_BREAKPOINT_MANIPULATION.
(microblaze_gdbarch_init): Don't use
SET_GDBARCH_BREAKPOINT_MANIPULATION.
* mips-tdep.c (mips_gdbarch_init): Likewise.
* mn10300-tdep.c (breakpoint): Rename it to mn10300_break_insn and
change its type to constexpr. Don't use
GDBARCH_BREAKPOINT_MANIPULATION.
(mn10300_gdbarch_init): Don't use
SET_GDBARCH_BREAKPOINT_MANIPULATION.
* moxie-tdep.c (breakpoint): Rename it to moxie_break_insn and
change its type to constexpr. Don't use
GDBARCH_BREAKPOINT_MANIPULATION.
(moxie_gdbarch_init): Don't use
SET_GDBARCH_BREAKPOINT_MANIPULATION.
* msp430-tdep.c (breakpoint): Rename it to msp430_break_insn
and change its type to constexpr. Don't use
GDBARCH_BREAKPOINT_MANIPULATION.
(msp430_gdbarch_init): Don't use
SET_GDBARCH_BREAKPOINT_MANIPULATION.
* mt-tdep.c (mt_gdbarch_init): Likewise.
* nds32-tdep.c (break_insn): Rename it to nds32_break_insn
and change its type to constexpr. Don't use
GDBARCH_BREAKPOINT_MANIPULATION.
(nds32_gdbarch_init): Don't use
SET_GDBARCH_BREAKPOINT_MANIPULATION.
* nios2-tdep.c (nios2_gdbarch_init): Likewise.
* rl78-tdep.c (breakpoint): Rename it to rl78_break_ins
and change its type to rl78_break_insn. Don't use
GDBARCH_BREAKPOINT_MANIPULATION.
(rl78_gdbarch_init): Don't use
SET_GDBARCH_BREAKPOINT_MANIPULATION.
* rs6000-tdep.c (big_breakpoint): Change its type to
constexpr.
(little_breakpoint): Likewise.
Don't use GDBARCH_BREAKPOINT_MANIPULATION_ENDIAN.
(rs6000_gdbarch_init): Don't use
SET_GDBARCH_BREAKPOINT_MANIPULATION.
* rx-tdep.c (breakpoint): Rename it to rx_break_insn and
change its type to constexpr. Don't use
GDBARCH_BREAKPOINT_MANIPULATION.
(rx_gdbarch_init): Don't use
SET_GDBARCH_BREAKPOINT_MANIPULATION.
* s390-linux-tdep.c (breakpoint): Rename it to s390_break_insn
and change its type to constexpr. Don't use
GDBARCH_BREAKPOINT_MANIPULATION
(s390_gdbarch_init): Don't use
SET_GDBARCH_BREAKPOINT_MANIPULATION.
* score-tdep.c (score_gdbarch_init): Likewise.
* sh-tdep.c (sh_gdbarch_init): Likewise.
* sh64-tdep.c (sh64_gdbarch_init): Likewise.
* sparc-tdep.c (break_insn): Rename it to sparc_break_insn
and change its type to constexpr. Don't use
GDBARCH_BREAKPOINT_MANIPULATION.
(sparc32_gdbarch_init): Don't use
SET_GDBARCH_BREAKPOINT_MANIPULATION.
* spu-tdep.c (breakpoint): Rename it to spu_break_insn and change
its type to constexpr. Don't use
GDBARCH_BREAKPOINT_MANIPULATION.
(spu_gdbarch_init): Don't use
SET_GDBARCH_BREAKPOINT_MANIPULATION.
* tic6x-tdep.c (tic6x_gdbarch_init): Likewise.
* tilegx-tdep.c (breakpoint): Rename it to tilegx_break_insn
and change its type to constexpr. Don't use
GDBARCH_BREAKPOINT_MANIPULATION.
(tilegx_gdbarch_init): Don't use
SET_GDBARCH_BREAKPOINT_MANIPULATION.
* v850-tdep.c (v850_gdbarch_init): Likewise.
* vax-tdep.c (break_insn): Rename it to vax_break_insn and
change its type to constexpr.
Don't use GDBARCH_BREAKPOINT_MANIPULATION.
(vax_gdbarch_init): Don't use
SET_GDBARCH_BREAKPOINT_MANIPULATION.
* xstormy16-tdep.c (breakpoint): Rename it to
xstormy16_break_insn and change its type to constexpr.
Don't use GDBARCH_BREAKPOINT_MANIPULATION.
(xstormy16_gdbarch_init): Don't use
SET_GDBARCH_BREAKPOINT_MANIPULATION.
* xtensa-tdep.c (xtensa_gdbarch_init): Likewise.
We don't track FP registers in aarch64 prologue analyzer, so this causes
an internal error when FP registers are saved by "stp" instruction in
prologue (stp d8, d9, [sp,#128]),
tbreak _Unwind_RaiseException^M
aarch64-tdep.c:335: internal-error: CORE_ADDR aarch64_analyze_prologue(gdbarch*, CORE_ADDR, CORE_ADDR, aarch64_prologue_cache*): Assertion `inst.operands[0].type == AARCH64_OPND_Rt' failed.^M
A problem internal to GDB has been detected,
This patch teaches GDB to track FP registers (D registers) in prologue
analyzer.
gdb:
2016-10-12 Yao Qi <yao.qi@linaro.org>
PR tdep/20682
* aarch64-tdep.c: Replace 32 with AARCH64_D_REGISTER_COUNT.
(aarch64_analyze_prologue): Extend array 'regs' for D registers.
Assert that operand 0 and 1 can be X or D registers. Update
register number for D registers. Update registers in frame
cache.
* aarch64-tdep.h (AARCH64_D_REGISTER_COUNT): New macro.
Nowadays, we only match pre-indexed STP in prologue. Due to the change
in gcc, https://gcc.gnu.org/ml/gcc-patches/2016-07/msg01933.html, it
may generate "STP with base register" in prologue, which GDB doesn't
handle. That is to say, previously GCC generates prologue like this,
sub sp, sp, #490
stp x29, x30, [sp, #-96]!
mov x29, sp
with the gcc patch above, GCC generates prologue like like this,
sub sp, sp, #4f0
stp x29, x30, [sp]
mov x29, sp
This patch is to teach GDB to recognize this instruction in prologue
analysis.
gdb:
2016-08-19 Yao Qi <yao.qi@linaro.org>
* aarch64-tdep.c (aarch64_analyze_prologue): Handle register
based STP instruction.
In AArch64 displaced stepping and fast tracepoint, GDB/GDBserver needs
to check whether the offset can fit in the range. We are using int32_t
for offset, it is sufficient to get an offset from an instruction, but
it is not enough to get an offset from two addresses. For example,
we have a BL in shared lib which is at 0x0000002000040774, and the
scratch pad for displaced stepping is at 0x400698. The offset can't
fit in 28 bit imm. However, since we are using int32_t for offset, GDB
thinks the offset can fit it, and generate the B instruction with wrong
offset.
It fixes the following fail,
-FAIL: gdb.base/dso2dso.exp: next over call to sub2
gdb:
2016-06-28 Yao Qi <yao.qi@linaro.org>
* aarch64-tdep.c (aarch64_displaced_step_b): Use int64_t for
variable new_offset.
gdb/gdbserver:
2016-06-28 Yao Qi <yao.qi@linaro.org>
* linux-aarch64-low.c (aarch64_ftrace_insn_reloc_b): Use int64_t
for variable new_offset.
(aarch64_ftrace_insn_reloc_b_cond): Likewise.
(aarch64_ftrace_insn_reloc_cb): Likewise.
(aarch64_ftrace_insn_reloc_tb): Likewise.
(aarch64_install_fast_tracepoint_jump_pad): Likewise. Use
PRIx64 instead of PRIx32.
Fixes, with x86_64-apple-darwin15-gcc (gcc 5.3.0):
.../src/gdb/aarch64-tdep.c: In function 'aarch64_record_load_store':
.../src/gdb/aarch64-tdep.c:3479:67: error: passing argument 3 of 'regcache_raw_read_unsigned' from incompatible pointer type [-Werror=incompatible-pointer-types]
bits (aarch64_insn_r->aarch64_insn, 16, 20), ®_rm_val);
^
In file included from .../src/gdb/regcache.h:23:0,
from .../src/gdb/gdbarch.h:69,
from .../src/gdb/defs.h:620,
from .../src/gdb/aarch64-tdep.c:21:
.../src/gdb/common/common-regcache.h:60:29: note: expected 'ULONGEST * {aka long unsigned int *}' but argument is of type 'uint64_t * {aka long long unsigned int *}'
extern enum register_status regcache_raw_read_unsigned
^
gdb/ChangeLog:
2016-04-20 Pedro Alves <palves@redhat.com>
* aarch64-tdep.c (aarch64_record_load_store): Change type of
'reg_rm_val' local to ULONGEST.
core_addr_to_string_nz returns string which has "0x" prefix, so don't
need to print "0x" again. This patch is to remove the "0x".
gdb:
2016-02-15 Yao Qi <yao.qi@linaro.org>
* aarch64-tdep.c (aarch64_analyze_prologue): Remove "0x".
Two small changes so everything builds with latest GCC and its
-Wmisleading-indentation.
In the aarch64-tdep.c case, the two misindented lines should actually be
part of the for loop. It looks like the indentation is all done using
spaces in that file though... I fixed it (changed for tabs + spaces) for
the lines I touched.
In the xcoffread.c case, we can simply remove the braces and fix the
indentation.
gdb/ChangeLog:
* aarch64-tdep.c (aarch64_record_asimd_load_store): Add braces
to for include additional lines.
* xcoffread.c (scan_xcoff_symtab): Remove unnecessary braces.
Hi,
AddressSanitizer reports an error like this,
(gdb) PASS: gdb.base/call-ar-st.exp: continue to tbreak9
print print_long_arg_list(a, b, c, d, e, f, *struct1, *struct2, *struct3, *struct4, *flags, *flags_combo, *three_char, *five_char, *int_char_combo, *d1, *d2, *d3, *f1, *f2, *f3)
=================================================================
==6236==ERROR: AddressSanitizer: heap-buffer-overflow on address 0x60200008eb50 at pc 0x89e432 bp 0x7fffa3df9080 sp 0x7fffa3df9078
READ of size 5 at 0x60200008eb50 thread T0
#0 0x89e431 in memory_xfer_partial gdb/target.c:1264
#1 0x89e6c7 in target_xfer_partial gdb/target.c:1320
#2 0x89f267 in target_write_partial gdb/target.c:1595^M
#3 0x8a014b in target_write_with_progress gdb/target.c:1889^M
#4 0x8a0262 in target_write gdb/target.c:1914^M
#5 0x89ee59 in target_write_memory gdb/target.c:1492^M
#6 0x9a1c74 in write_memory gdb/corefile.c:393^M
#7 0x467ea5 in aarch64_push_dummy_call gdb/aarch64-tdep.c:1388
The problem is that an instance of stack_item_t is created to adjust
stack for alignment, the item.len is correct, but item.data is buf,
which is wrong, because item.len can be greater than the length of
buf. This patch sets item.data to NULL, and only update sp (no
inferior memory writes on stack for this item).
gdb:
2015-12-17 Yao Qi <yao.qi@linaro.org>
* aarch64-tdep.c (struct stack_item_t): Update comments.
(pass_on_stack): Set item.data to NULL.
(aarch64_push_dummy_call): Call write_memory if si->data
isn't NULL.
AArch64 AAPCS defined HFA (homogeneous floating-point aggregate)
and HVF (homogeneous short vector aggregate), bug GDB only handles the
former. In the AAPCS doc, both types are treated exactly the same
in terms of alignment and passing locations (on registers or stack).
This patch is to extend is_hfa to handle both HFA and HVA.
gdb:
2015-11-27 Yao Qi <yao.qi@linaro.org>
* aarch64-tdep.c (is_hfa): Rename to ...
(is_hfa_or_hva): ... this. Handle vector type. All callers
updated.
(aarch64_extract_return_value): Update debugging message.
(aarch64_store_return_value): Likewise.
(aarch64_return_in_memory): Update comments.
As defined in AArch64 AAPCS, short vectors are passed through V
registers, and its maximum alignment is 16-byte. This patch is
to reflect these rules in GDB. This patch fixes some fails in
gdb.base/gnu_vector.exp.
gdb:
2015-11-27 Yao Qi <yao.qi@linaro.org>
* aarch64-tdep.c (aarch64_type_align): For vector type, return
its length, but with the maximum of 16 bytes.
(is_hfa): Return zero for vector type.
(aarch64_push_dummy_call): Handle short vectors.
(aarch64_extract_return_value): Likewise.
(aarch64_store_return_value): Likewise.
Hi,
I build GDB with -fsanitize=address, and run testsuite. In
gdb.base/callfuncs.exp, I see the following error,
p t_float_values(0.0,0.0)
=================================================================
==8088==ERROR: AddressSanitizer: heap-buffer-overflow on address 0x6020000cb650 at pc 0x6e195c bp 0x7fff164f9770 sp 0x7fff164f9768
READ of size 16 at 0x6020000cb650 thread T0^
#0 0x6e195b in regcache_raw_write /home/yao/SourceCode/gnu/gdb/git/gdb/regcache.c:912
#1 0x6e1e52 in regcache_cooked_write /home/yao/SourceCode/gnu/gdb/git/gdb/regcache.c:945
#2 0x466d69 in pass_in_v /home/yao/SourceCode/gnu/gdb/git/gdb/aarch64-tdep.c:1101
#3 0x467512 in pass_in_v_or_stack /home/yao/SourceCode/gnu/gdb/git/gdb/aarch64-tdep.c:1196
#4 0x467d7d in aarch64_push_dummy_call /home/yao/SourceCode/gnu/gdb/git/gdb/aarch64-tdep.c:1335
The code in pass_in_v read contents from V registers (128 bit), but the
data passed through V registers can be less than 128 bit. In this case,
float is passed. So writing V registers contents into contents buff
will cause overflow. In this patch, we add an array reg[V_REGISTER_SIZE],
which is to hold the contents from V registers, and then copy useful
bits to buf.
gdb:
2015-11-18 Yao Qi <yao.qi@linaro.org>
* aarch64-tdep.c (pass_in_v): Add argument len. Add local array
reg. Callers updated.
Hi,
I build GDB with -fsanitize=address, and run testsuite. In
gdb.base/callfuncs.exp, I see the following error,
p/c fun1()
=================================================================^M
==9601==ERROR: AddressSanitizer: stack-buffer-overflow on address 0x7fffee858530 at pc 0x6df079 bp 0x7fffee8583a0 sp 0x7fffee858398
WRITE of size 16 at 0x7fffee858530 thread T0
#0 0x6df078 in regcache_raw_read /home/yao/SourceCode/gnu/gdb/git/gdb/regcache.c:673
#1 0x6dfe1e in regcache_cooked_read /home/yao/SourceCode/gnu/gdb/git/gdb/regcache.c:751
#2 0x4696a3 in aarch64_extract_return_value /home/yao/SourceCode/gnu/gdb/git/gdb/aarch64-tdep.c:1708
#3 0x46ae57 in aarch64_return_value /home/yao/SourceCode/gnu/gdb/git/gdb/aarch64-tdep.c:1918
We are extracting return value from V registers (128 bit), but only
allocate X_REGISTER_SIZE-byte array, which isn't sufficient. This
patch changes the array to V_REGISTER_SIZE.
gdb:
2015-11-16 Yao Qi <yao.qi@linaro.org>
* aarch64-tdep.c (aarch64_extract_return_value): Change array
buf's length to V_REGISTER_SIZE.
This patch changes the last argument of functions pass_in_x_or_stack,
pass_in_v_or_stack, pass_on_stack, and pass_in_x to type value *.
gdb:
2015-11-16 Yao Qi <yao.qi@linaro.org>
* aarch64-tdep.c (pass_in_x_or_stack): Change argument type
from bfd_byte * to value *. Caller updated.
(pass_in_x): Likewise.
(pass_in_v_or_stack): Likewise.
(pass_on_stack): Likewise.
Both aarch64_push_dummy_call and bfin_push_dummy_call only use args[i]
contents but then never write to them, so that we can use
value_contents instead.
gdb:
2015-11-16 Yao Qi <yao.qi@linaro.org>
* aarch64-tdep.c (aarch64_push_dummy_call): Call value_contents instead
of value_contents_writeable.
* bfin-tdep.c (bfin_push_dummy_call): Likewise.
gdb:
2015-11-05 Yao Qi <yao.qi@linaro.org>
* aarch64-tdep.c (aarch64_displaced_step_copy_insn): Call
aarch64_decode_insn and decode instruction by aarch64_inst.
This patch convert aarch64_analyze_prologue to using aarch64_decode_insn
to decode instructions. After this change, aarch64_analyze_prologue
looks much simple, and some aarch64_decode_* functions are removed
accordingly.
gdb:
2015-11-05 Yao Qi <yao.qi@linaro.org>
* aarch64-tdep.c (extract_signed_bitfield): Remove.
(decode_masked_match): Remove.
(aarch64_decode_add_sub_imm): Remove.
(aarch64_decode_br): Remove.
(aarch64_decode_eret): Remove.
(aarch64_decode_movz): Remove.
(aarch64_decode_orr_shifted_register_x): Remove.
(aarch64_decode_ret): Remove.
(aarch64_decode_stp_offset): Remove.
(aarch64_decode_stur): Remove.
(aarch64_analyze_prologue): Call aarch64_decode_insn
and use aarch64_inst to decode instructions.
This patch combines both aarch64_decode_stp_offset_wb and
aarch64_decode_stp_offset together.
gdb:
2015-11-05 Yao Qi <yao.qi@linaro.org>
* aarch64-tdep.c (aarch64_decode_stp_offset): New argument
wback.
(aarch64_decode_stp_offset_wb): Removed.
(aarch64_analyze_prologue): Don't use
aarch64_decode_stp_offset_wb.
Nowadays aarch64_decode_insn is a public interface used by both
opcodes and gdb. However, its behaviour relies on a global variable
no_aliases, which isn't a good practise. On the other hand, In default,
no_aliases is zero, but in GDB, we do want no alias when decoding
instructions for prologue analysis (patches to be posted), so that we
can handle both instructions "add" and "mov" (an alias of "add") as
"add". The code in GDB can be simplified.
This patch adds a new argument in aarch64_decode_insn, and pass no_aliases
to it. In GDB side, always pass 1 to it.
include/opcode:
2015-10-28 Yao Qi <yao.qi@linaro.org>
* aarch64.h (aarch64_decode_insn): Update declaration.
opcodes:
2015-10-28 Yao Qi <yao.qi@linaro.org>
* aarch64-dis.c (aarch64_decode_insn): Add one argument
noaliases_p. Update comments. Pass noaliases_p rather than
no_aliases to aarch64_opcode_decode.
(print_insn_aarch64_word): Pass no_aliases to
aarch64_decode_insn.
gdb:
2015-10-28 Yao Qi <yao.qi@linaro.org>
* aarch64-tdep.c (aarch64_software_single_step): Pass 1 to
aarch64_decode_insn.
This patch initialize dsd.insn_count, otherwise, it triggers the assert
below on testings we did recently.
gdb:
2015-10-23 Yao Qi <yao.qi@linaro.org>
* aarch64-tdep.c (aarch64_displaced_step_copy_insn): Set
dsd.insn_count to zero.
As emit_insn becomes extern, the prefix "aarch64_" is needed. This
patch renames emit_insn to aarch64_emit_insn.
gdb:
2015-10-12 Yao Qi <yao.qi@linaro.org>
* arch/aarch64-insn.c (emit_insn): Rename to ...
(aarch64_emit_insn): ... it. All callers updated.
gdb/gdbserver:
2015-10-12 Yao Qi <yao.qi@linaro.org>
* linux-aarch64-low.c: Update all callers of function renaming
from emit_insn to aarch64_emit_insn.
This patch is to support displaced stepping in aarch64-linux. A
visitor is implemented for displaced stepping, and used to record
information to fixup pc after displaced stepping if needed. Some
emit_* functions are converted to macros, and moved to
arch/aarch64-insn.{c,h} so that they can be shared.
gdb:
2015-10-12 Yao Qi <yao.qi@linaro.org>
* aarch64-linux-tdep.c: Include arch-utils.h.
(aarch64_linux_init_abi): Call set_gdbarch_max_insn_length,
set_gdbarch_displaced_step_copy_insn,
set_gdbarch_displaced_step_fixup,
set_gdbarch_displaced_step_free_closure,
set_gdbarch_displaced_step_location,
and set_gdbarch_displaced_step_hw_singlestep.
* aarch64-tdep.c (struct displaced_step_closure): New.
(struct aarch64_displaced_step_data): New.
(aarch64_displaced_step_b): New function.
(aarch64_displaced_step_b_cond): Likewise.
(aarch64_register): Likewise.
(aarch64_displaced_step_cb): Likewise.
(aarch64_displaced_step_tb): Likewise.
(aarch64_displaced_step_adr): Likewise.
(aarch64_displaced_step_ldr_literal): Likewise.
(aarch64_displaced_step_others): Likewise.
(aarch64_displaced_step_copy_insn): Likewise.
(aarch64_displaced_step_fixup): Likewise.
(aarch64_displaced_step_hw_singlestep): Likewise.
* aarch64-tdep.h (DISPLACED_MODIFIED_INSNS): New macro.
(aarch64_displaced_step_copy_insn): Declare.
(aarch64_displaced_step_fixup): Declare.
(aarch64_displaced_step_hw_singlestep): Declare.
* arch/aarch64-insn.c (emit_insn): Moved from
gdbserver/linux-aarch64-low.c.
(emit_load_store): Likewise.
* arch/aarch64-insn.h (enum aarch64_opcodes): Moved from
gdbserver/linux-aarch64-low.c.
(struct aarch64_register): Likewise.
(struct aarch64_memory_operand): Likewise.
(ENCODE): Likewise.
(can_encode_int32): New macro.
(emit_b, emit_bcond, emit_cb, emit_ldr, emit_ldrsw): Likewise.
(emit_tb, emit_nop): Likewise.
(emit_insn): Declare.
(emit_load_store): Declare.
gdb/gdbserver:
2015-10-12 Yao Qi <yao.qi@linaro.org>
* linux-aarch64-low.c (enum aarch64_opcodes): Move to
arch/aarch64-insn.h.
(struct aarch64_memory_operand): Likewise.
(ENCODE): Likewise.
(emit_insn): Move to arch/aarch64-insn.c.
(emit_b, emit_bcond, emit_cb, emit_tb): Remove.
(emit_load_store): Move to arch/aarch64-insn.c.
(emit_ldr, emit_ldrb, emit_ldrsw, emit_nop): Remove.
(can_encode_int32): Remove.
There are a bunch of places where a void* is implicitely casted into a
gdb_byte*. The auto-insert-casts script added explicit casts at those
places. However, in many cases, it makes more sense to just change the
void* to a gdb_byte*.
gdb/ChangeLog:
* aarch64-tdep.c (stack_item_t): Change type of data to gdb_byte*.
* arm-tdep.c (struct stack_item): Likewise.
(push_stack_item): Add gdb_byte* cast.
* avr-tdep.c (struct stack_item): Change type of data to gdb_byte*.
(push_stack_item): Add gdb_byte* cast.
* cli/cli-dump.c (dump_memory_to_file): Change type of buf to gdb_byte*
and add cast.
* cris-tdep.c (struct stack_item): Change type of data to gdb_byte*.
(push_stack_item): Add gdb_byte* cast.
* gcore.c (gcore_copy_callback): Change type of memhunk to gdb_byte* and
add cast.
* gdbtypes.h (print_scalar_formatted): Change type of first parameter to
gdb_byte*.
* h8300-tdep.c (h8300_extract_return_value): Change type of valbuf to
gdb_byte* and remove unnecessary cast.
(h8300h_extract_return_value): Likewise.
(h8300_store_return_value): Change type of valbuf to gdb_byte*.
(h8300h_store_return_value): Likewise.
* iq2000-tdep.c (iq2000_extract_return_value): Change type of valbuf to
gdb_byte* and remove unnecessary cast.
* jit.c (jit_reader_try_read_symtab): Change type of gdb_mem to gdb_byte*
and add cast.
* m32r-tdep.c (m32r_store_return_value): Change type of valbuf to
gdb_byte* and remove unnecessary cast.
(m32r_extract_return_value): Change type of dst to gdb_byte* and remove
valbuf.
* mep-tdep.c (mep_pseudo_cr32_read): Change type of buf to gdb_byte*.
(mep_pseudo_cr64_read): Likewise.
(mep_pseudo_csr_write): Likewise.
(mep_pseudo_cr32_write): Likewise.
(mep_pseudo_cr64_write): Likewise.
* mi/mi-main.c (mi_cmd_data_write_memory): Change type of buffer to
gdb_byte* and add cast.
* moxie-tdep.c (moxie_store_return_value): Change type of valbuf to
gdb_byte* and remove unnecessary cast.
(moxie_extract_return_value): Change type of dst to gdb_byte* and remove
valbuf.
* p-valprint.c (print_scalar_formatted): Change type of valaddr to
gdb_byte*.
* printcmd.c (void): Likewise.
* python/py-inferior.c (infpy_read_memory): Change type of buffer to
gdb_byte* and add cast.
(infpy_write_memory): Likewise.
(infpy_search_memory): Likewise.
* regcache.c (regcache_raw_write_signed): Change type of buf to gdb_byte*
and add cast.
(regcache_raw_write_unsigned): Likewise.
(regcache_cooked_write_signed): Likewise.
(regcache_cooked_write_unsigned): Likewise.
* sh64-tdep.c (h64_extract_return_value): Change type of valbuf to
gdb_byte*.
In this patch, we start to use aarch64_decode_insn to decode instructions
in aarch64_software_single_step.
gdb:
2015-10-07 Yao Qi <yao.qi@linaro.org>
* aarch64-tdep.c: Include opcode/aarch64.h.
(submask): Move it above.
(bit): Likewise.
(bits): Likewise.
(aarch64_software_single_step): Call aarch64_decode_insn.
Decode instruction by aarch64_inst instead of using
aarch64_decode_bcond and decode_masked_match.
We will need to decode both ADR and ADRP instructions in GDBserver.
This patch makes common code handle both cases, even if GDB only needs
to decode the ADRP instruction.
gdb/ChangeLog:
* aarch64-tdep.c (aarch64_analyze_prologue): New is_adrp
variable. Call aarch64_decode_adr instead of
aarch64_decode_adrp.
* arch/aarch64-insn.h (aarch64_decode_adrp): Delete.
(aarch64_decode_adr): New function declaration.
* arch/aarch64-insn.c (aarch64_decode_adrp): Delete.
(aarch64_decode_adr): New function, factored out from
aarch64_decode_adrp to decode both adr and adrp instructions.
This patch moves the following functions into the arch/ common
directory, in new files arch/aarch64-insn.{h,c}. They are prefixed with
'aarch64_':
- aarch64_decode_adrp
- aarch64_decode_b
- aarch64_decode_cb
- aarch64_decode_tb
We will need them to implement fast tracepoints in GDBserver.
For consistency, this patch also adds the 'aarch64_' prefix to static
decoding functions that do not need to be shared right now.
V2:
make sure the formatting issues propagated
fix `gdbserver/configure.srv'.
gdb/ChangeLog:
* Makefile.in (ALL_64_TARGET_OBS): Add aarch64-insn.o.
(HFILES_NO_SRCDIR): Add arch/aarch64-insn.h.
(aarch64-insn.o): New rule.
* configure.tgt (aarch64*-*-elf): Add aarch64-insn.o.
(aarch64*-*-linux*): Likewise.
* arch/aarch64-insn.c: New file.
* arch/aarch64-insn.h: New file.
* aarch64-tdep.c: Include arch/aarch64-insn.h.
(aarch64_debug): Move to arch/aarch64-insn.c. Declare in
arch/aarch64-insn.h.
(decode_add_sub_imm): Rename to ...
(aarch64_decode_add_sub_imm): ... this.
(decode_adrp): Rename to ...
(aarch64_decode_adrp): ... this. Move to arch/aarch64-insn.c.
Declare in arch/aarch64-insn.h.
(decode_b): Rename to ...
(aarch64_decode_b): ... this. Move to arch/aarch64-insn.c.
Declare in arch/aarch64-insn.h.
(decode_bcond): Rename to ...
(aarch64_decode_bcond): ... this. Move to arch/aarch64-insn.c.
Declare in arch/aarch64-insn.h.
(decode_br): Rename to ...
(aarch64_decode_br): ... this.
(decode_cb): Rename to ...
(aarch64_decode_cb): ... this. Move to arch/aarch64-insn.c.
Declare in arch/aarch64-insn.h.
(decode_eret): Rename to ...
(aarch64_decode_eret): ... this.
(decode_movz): Rename to ...
(aarch64_decode_movz): ... this.
(decode_orr_shifted_register_x): Rename to ...
(aarch64_decode_orr_shifted_register_x): ... this.
(decode_ret): Rename to ...
(aarch64_decode_ret): ... this.
(decode_stp_offset): Rename to ...
(aarch64_decode_stp_offset): ... this.
(decode_stp_offset_wb): Rename to ...
(aarch64_decode_stp_offset_wb): ... this.
(decode_stur): Rename to ...
(aarch64_decode_stur): ... this.
(decode_tb): Rename to ...
(aarch64_decode_tb): ... this. Move to arch/aarch64-insn.c.
Declare in arch/aarch64-insn.h.
(aarch64_analyze_prologue): Adjust calls to renamed functions.
gdb/gdbserver/ChangeLog:
* Makefile.in (aarch64-insn.o): New rule.
* configure.srv (aarch64*-*-linux*): Add aarch64-insn.o.
gdb/ChangeLog:
* aarch64-tdep.c (decode_cb): Move up comment describing the
encoding.
(decode_tb): Fix a typo in comment above the function. Move up
comment describing the encoding.
The encoding of the b.cond instruction is described in the architecture
reference manual as:
b.cond 0101 0100 iiii iiii iiii iiii iii0 cccc
So the mask should be 0xff000010.
gdb/ChangeLog:
* aarch64-tdep.c (decode_bcond): Fix incorrect mask.
This patch cleans up the decoding functions using booleans when they can
decode two instructions. The boolean argument is used to know which of
the two instructions was decoded.
The instructions affected are BR/BLR, B/BL, CBZ/CBNZ and TBZ/TBNZ.
These arguments would be named after a named bit in the instruction
encoding, this patch renames them to 'is_XXX'. Furthermore, the
'unsigned' type would be used to describe a boolean while
aarch64_decode_cb would use 'int' (see the 'is64' argument). This patch
makes all booleans be 'int' and decoded bitfields be 'unsigned'.
gdb/ChangeLog:
* aarch64-tdep.c (decode_b): Rename link argument to is_bl.
Change its type to int *.
(decode_br): Rename link argument to is_blr. Change its type to
int *.
(decode_cb): Rename op argument to is_cbnz. Change its type to
int *.
(decode_tb): Rename op argument to is_tbnz. Change its type to
int *. Set is_tbnz to either 1 or 0.
(aarch64_analyze_prologue): Change type of is_link to int. Add
new variables is_cbnz and is_tbnz. Adjust call to
aarch64_decode_cb and aarch64_decode_tb.
I noticed two failure in gdb.trace/mi-trace-frame-collected.exp:
FAIL: gdb.trace/mi-trace-frame-collected.exp: live:
-trace-frame-collected (register)
FAIL: gdb.trace/mi-trace-frame-collected.exp: tfile:
-trace-frame-collected (register)
In these cases, we are not collecting registers so the MI command
-trace-frame-collected should only give us the value of the PC.
However, it also gives us all of the single precision pseudo registers,
initialized with 0x0.
We can reproduce this error by simply issuing the
'maint print cooked-register' when no inferior is connected:
~~~
...
(gdb) maint print cooked-register
Name Nr Rel Offset Size Type Cooked value
x0 0 0 0 8 long <unavailable>
x1 1 1 8 8 long <unavailable>
...
d30 130 62 1540 8 *1 <unavailable>
d31 131 63 1548 8 *1 <unavailable>
s0 132 64 1556 4 *1 0x00000000
s1 133 65 1560 4 *1 0x00000000
s2 134 66 1564 4 *1 0x00000000
...
s28 160 92 1668 4 *1 0x00000000
s29 161 93 1672 4 *1 0x00000000
s30 162 94 1676 4 *1 0x00000000
s31 163 95 1680 4 *1 0x00000000
h0 164 96 1684 2 *1 <unavailable>
h1 165 97 1686 2 *1 <unavailable>
h2 166 98 1688 2 *1 <unavailable>
...
~~~
It turns out GDB does not check if S registers are valid before returning
a value for them. It should return <unavailable> in this case.
gdb/ChangeLog:
* aarch64-tdep.c (aarch64_pseudo_read_value): Mark S register as
unavailable if invalid.
This patch implements the 'collect $_ret' command to collect the return
address of a function in a tracepoint. It marks the LR register for
collection.
gdb/ChangeLog:
* aarch64-tdep.c: Add ax.h and ax-gdb.h includes.
(aarch64_gen_return_address): New function.
(aarch64_gdbarch_init): Hook it.
The stub unwinder is used on AArch64 if the target's memory is not
readable at the current PC. For example, the user could try to call at
an invalid address such as 0x0, as covered in the gdb.base/signull.exp
test case. Many GDB ports use a similar unwinder to handle this case
too.
If we purposely kill the inferior before examining the trace then we get
the following issue:
~~~
...
(gdb) trace f
Tracepoint 3 at 0x7fb7fc28c0
(gdb) tstart
(gdb) continue
...
(gdb) tstop
(gdb) tsave /tmp/trace
(gdb) kill
...
(gdb) target tfile /tmp/trace
...
(gdb) tfind
Register 31 is not available.
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Found trace frame 0, tracepoint 3
#-1 0x0000007fb7fc28c0 in f () ...
^^^
~~~
This patch teaches the stub unwinder to report to the core frame code
with UNWIND_UNAVAILABLE when either the stack pointer of the return
address are unavailable to read from the target.
gdb/ChangeLog:
* aarch64-tdep.c (aarch64_make_stub_cache): Set available_p and
swallow NOT_AVAILABLE_ERROR.
(aarch64_stub_this_id): Call frame_id_build_unavailable_stack if
available_p is not set.
(aarch64_stub_frame_unwind_stop_reason): New function.
(aarch64_stub_unwind): Install it.
Without debugging information, we have the following issue when
examining a trace buffer:
~~~
...
(gdb) trace f
Tracepoint 3 at 0x7fb7fc28c0
(gdb) tstart
(gdb) continue
...
(gdb) tstop
(gdb) tfind start
Register 31 is not available.
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Found trace frame 0, tracepoint 3
#-1 0x0000007fb7fc28c0 in f () ...
^^^
~~~
The reason for this is that the target's stack pointer is unavailable
when examining the trace buffer. What we are seeing is due to the
'tfind' command creating a sentinel frame and unwinding it. If an
exception is thrown, we are left with the sentinel frame being displayed
at level #-1. The exception is thrown when the prologue unwinder tries
to read the stack pointer to construct an ID for the frame.
This patch fixes this by making the prologue unwinder catch
NOT_AVAILABLE_ERROR exceptions when either registers or memory is
unreadable and report back to the frame core code with
UNWIND_UNAVAILABLE.
gdb/ChangeLog:
* aarch64-tdep.c (aarch64_prologue_cache) <available_p>: New
field.
(aarch64_make_prologue_cache_1): New function, factored out from
aarch64_make_prologue_cache. Do not allocate cache. Set
available_p.
(aarch64_make_prologue_cache): Reimplement wrapping
aarch64_make_prologue_cache_1, and swallowing
NOT_AVAILABLE_ERROR.
(aarch64_prologue_frame_unwind_stop_reason): New function.
Return UNWIND_UNAVAILABLE if available_p is not set.
(aarch64_prologue_unwind): Install it.
(aarch64_prologue_this_id): Move prev_pc and prev_sp limit
checks into aarch64_prologue_frame_unwind_stop_reason. Call
frame_id_build_unavailable_stack if available_p is not set.
This patch moves the address of the start of a function (func) and the
address from which it was called (prev_pc) into aarch64_prologue_cache.
The idea is to keep accesses to the inferior's registers into
aarch64_make_prologue_cache and aarch64_make_stub_cache.
gdb/ChangeLog:
* aarch64-tdep.c (aarch64_prologue_cache) <func, prev_pc>: New
fields.
(aarch64_scan_prologue): Set prev_pc.
(aarch64_make_prologue_cache): Set func.
(aarch64_make_stub_cache): Set prev_pc.
(aarch64_prologue_this_id): Remove local variables id, pc and
func. Read prev_pc and func from cache.
(aarch64_stub_this_id): Read prev_pc from cache.
We would previously have to make sure the frame cache was not already
created before calling aarch64_make_stub_cache. This patch makes this
function check it so the caller does not need to do so.
gdb/ChangeLog:
* aarch64-tdep.c (aarch64_make_stub_cache): Update comment. New
argument this_cache. Remove unused local variables reg and
unwound_fp. Return early if this_cache is already set. Set
this_cache.
(aarch64_stub_this_id): Update call to aarch64_make_stub_cache.
We would previously have to make sure the frame cache was not already
created before calling aarch64_make_prologue_cache. This patch makes
this function check it so that the caller does not need to do so.
gdb/ChangeLog:
* aarch64-tdep.c (aarch64_make_prologue_cache): Update comment.
New argument this_cache. Return early if this_cache is already
set. Set this_cache.
(aarch64_prologue_this_id): Update call to
aarch64_make_prologue_cache.
(aarch64_prologue_prev_register): Likewise.
(aarch64_normal_frame_base): Likewise.