Thanks Rodolph,
I was actually looked at it a little bit and fixed all the compilation
problem. However, it failed during runtime. Then I tried to see if X86 was
successful, and it failed as well.
I was under the impression that X86 was all good. Here's how it failed on
tip. (I have not tried branches)
gdb --args ./d8 --gdbjit foo.js
(gdb) b Builtins_StringPrototypeReplace
Breakpoint 2 at 0x5555567b7bc4
(gdb) r
The program being debugged has been started already.
Start it from the beginning? (y or n) y
Starting program: /work/sirish.pande/v8/v8/out/x64.master.debug.gdbjit/d8
--gdbjit foo.js
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".
[New Thread 0x7ffff662d700 (LWP 23951)]
[New Thread 0x7ffff5e2c700 (LWP 23952)]
[New Thread 0x7ffff562b700 (LWP 23953)]
[New Thread 0x7ffff4e2a700 (LWP 23954)]
[New Thread 0x7ffff4629700 (LWP 23955)]
[New Thread 0x7ffff3e28700 (LWP 23956)]
[New Thread 0x7ffff3627700 (LWP 23957)]
[New Thread 0x7ffff2e26700 (LWP 23958)]
#
# Fatal error in ../../src/isolate.cc, line 198
# Check failed: d.Hash() == d.CreateHash() (18072318529312869522 vs.
8652045362104995461).
#
#
#
#FailureMessage Object: 0x7fffffffc1f0
==== C stack trace ===============================
/work/sirish.pande/v8/v8/out/x64.master.debug.gdbjit/d8(+0x161bcf3)
[0x555556b6fcf3]
/work/sirish.pande/v8/v8/out/x64.master.debug.gdbjit/d8(+0x161b2bb)
[0x555556b6f2bb]
/work/sirish.pande/v8/v8/out/x64.master.debug.gdbjit/d8(+0x1613678)
[0x555556b67678]
/work/sirish.pande/v8/v8/out/x64.master.debug.gdbjit/d8(+0xa36e12)
[0x555555f8ae12]
/work/sirish.pande/v8/v8/out/x64.master.debug.gdbjit/d8(+0xa36c54)
[0x555555f8ac54]
/work/sirish.pande/v8/v8/out/x64.master.debug.gdbjit/d8(+0xa35dd8)
[0x555555f89dd8]
/work/sirish.pande/v8/v8/out/x64.master.debug.gdbjit/d8(+0x3faef1)
[0x55555594eef1]
/work/sirish.pande/v8/v8/out/x64.master.debug.gdbjit/d8(+0x3a9214)
[0x5555558fd214]
/lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0xe7) [0x7ffff7028b97]
/work/sirish.pande/v8/v8/out/x64.master.debug.gdbjit/d8(_start+0x2a)
[0x5555558eb02a]
Thread 1 "d8" received signal SIGILL, Illegal instruction.
v8::base::OS::Abort () at ../../src/base/platform/platform-posix.cc:401
401 V8_IMMEDIATE_CRASH();
(gdb) i b
Num Type Disp Enb Address What
1 breakpoint keep y 0x0000555555f8c3e7 in
v8::internal::Isolate::Init(v8::internal::StartupDeserializer*) at
../../src/isolate.cc:3180
2 breakpoint keep y 0x00005555567b7bc4
<Builtins_StringPrototypeReplace+4>
Should we open bug for this one?
Sirish
On Thursday, December 13, 2018 at 11:13:39 AM UTC-6, Rodolph Perfetta wrote:
>
> You are right the arm/arm64 gdb-jit support is not complete, so it is more
> a feature request. I am happy to help with review if you submit a patch.
>
> Cheers,
> Rodolph.
>
> On Wed, 12 Dec 2018 at 21:36, <[email protected] <javascript:>> wrote:
>
>> Thanks Rodolph. When I try to build with v8_enable_gdbjit=true,
>> unfortunately, I run into compilation errors of gdb-jit for ARM64. Snippet
>> below shows the way try to build. It shows that I am using 7.1.1 - however,
>> it's the same problem with tip/master as well. Should I open a bug for this
>> one?
>>
>> cc01:v8:[7.1.1]>gn gen --args="target_cpu=\"arm64\"
>> is_component_build=false v8_enable_gdbjit=true" out/7.1.1.debug.gdbjit
>> Done. Made 159 targets from 80 files in 346ms
>> cc01:v8:[7.1.1]>autoninja -C out/7.1.1.debug.gdbjit d8
>> ninja -C out/7.1.1.debug.gdbjit d8 -l 128
>> ninja: Entering directory `out/7.1.1.debug.gdbjit'
>> [1340/2173] CXX obj/v8_base/gdb-jit.o
>> FAILED: obj/v8_base/gdb-jit.o
>> ../../third_party/llvm-build/Release+Asserts/bin/clang++ -MMD -MF
>> obj/v8_base/gdb-jit.o.d -DV8_DEPRECATION_WARNINGS -DUSE_UDEV -DUSE_AURA=1
>> -DUSE_GLIB=1 -DUSE_NSS_CERTS=1 -DUSE_X11=1 -DNO_TCMALLOC
>> -DFULL_SAFE_BROWSING -DSAFE_BROWSING_CSD -DSAFE_BROWSING_DB_LOCAL
>> -DCHROMIUM_BUILD -DFIELDTRIAL_TESTING_ENABLED -D_FILE_OFFSET_BITS=64
>> -D_LARGEFILE_SOURCE -D_LARGEFILE64_SOURCE -D_GNU_SOURCE
>> -DCR_CLANG_REVISION=\"338452-1\" -D__STDC_CONSTANT_MACROS
>> -D__STDC_FORMAT_MACROS -D_LIBCPP_DISABLE_VISIBILITY_ANNOTATIONS
>> -D_LIBCXXABI_DISABLE_VISIBILITY_ANNOTATIONS -DCR_LIBCXX_REVISION=332543
>> -DCR_LIBCXXABI_REVISION=331450
>> -DCR_SYSROOT_HASH=c68b2062879db201c4047d03c016227d3d1e7b35 -D_DEBUG
>> -DDYNAMIC_ANNOTATIONS_ENABLED=1 -DWTF_USE_DYNAMIC_ANNOTATIONS=1
>> -DENABLE_DISASSEMBLER -DV8_TYPED_ARRAY_MAX_SIZE_IN_HEAP=64
>> -DENABLE_GDB_JIT_INTERFACE -DENABLE_MINOR_MC -DOBJECT_PRINT -DVERIFY_HEAP
>> -DV8_TRACE_MAPS -DV8_ENABLE_ALLOCATION_TIMEOUT -DV8_ENABLE_FORCE_SLOW_PATH
>> -DV8_ENABLE_CHECKS -DV8_DEPRECATION_WARNINGS
>> -DV8_IMMINENT_DEPRECATION_WARNINGS -DV8_INTL_SUPPORT
>> -DENABLE_HANDLE_ZAPPING -DV8_USE_SNAPSHOT -DV8_USE_EXTERNAL_STARTUP_DATA
>> -DV8_CONCURRENT_MARKING -DV8_CHECK_MICROTASKS_SCOPES_CONSISTENCY
>> -DV8_EMBEDDED_BUILTINS -DV8_TARGET_ARCH_ARM64 -DDEBUG
>> -DU_USING_ICU_NAMESPACE=0 -DU_ENABLE_DYLOAD=0 -DU_STATIC_IMPLEMENTATION
>> -DICU_UTIL_DATA_IMPL=ICU_UTIL_DATA_FILE -DUCHAR_TYPE=uint16_t -I../.. -Igen
>> -I../.. -Igen -I../../third_party/icu/source/common
>> -I../../third_party/icu/source/i18n -I../../include -fno-strict-aliasing
>> --param=ssp-buffer-size=4 -fstack-protector -Wno-builtin-macro-redefined
>> -D__DATE__= -D__TIME__= -D__TIMESTAMP__= -funwind-tables -fPIC -pthread
>> -fcolor-diagnostics -fmerge-all-constants -no-canonical-prefixes
>> -fcomplete-member-pointers --target=aarch64-linux-gnu -Wall -Werror -Wextra
>> -Wimplicit-fallthrough -Wthread-safety -Wno-missing-field-initializers
>> -Wno-unused-parameter -Wno-c++11-narrowing -Wno-covered-switch-default
>> -Wno-unneeded-internal-declaration -Wno-undefined-var-template
>> -Wno-nonportable-include-path -Wno-user-defined-warnings
>> -Wno-unused-lambda-capture -Wno-null-pointer-arithmetic
>> -Wno-enum-compare-switch -Wno-ignored-pragma-optimize
>> -fno-omit-frame-pointer -g2 -fvisibility=hidden -Wheader-hygiene
>> -Wstring-conversion -Wtautological-overlap-compare
>> -Wmissing-field-initializers -Winconsistent-missing-override
>> -Wunreachable-code -Wshorten-64-to-32 -O3 -fno-ident -fdata-sections
>> -ffunction-sections -Wno-undefined-bool-conversion
>> -Wno-tautological-undefined-compare -std=c++14 -fno-exceptions -fno-rtti
>> -nostdinc++ -isystem../../buildtools/third_party/libc++/trunk/include
>> -isystem../../buildtools/third_party/libc++abi/trunk/include
>> --sysroot=../../build/linux/debian_sid_arm64-sysroot
>> -fvisibility-inlines-hidden -c ../../src/gdb-jit.cc -o obj/v8_base/gdb-jit.o
>> ../../src/gdb-jit.cc:668:2: error: Unsupported target architecture.
>> #error Unsupported target architecture.
>> ^
>> ../../src/gdb-jit.cc:697:2: error: Unsupported target architecture.
>> #error Unsupported target architecture.
>> ^
>> ../../src/gdb-jit.cc:670:27: error: use of undeclared identifier 'ident'
>> memcpy(header->ident, ident, 16);
>> ^
>> ../../src/gdb-jit.cc:838:27: error: use of undeclared identifier
>> 'SerializedLayout'
>> void Write(Writer::Slot<SerializedLayout> s, ELFStringTable* t) const {
>> ^
>> ../../src/gdb-jit.cc:912:49: error: no member named 'SerializedLayout' in
>> 'v8::internal::GDBJITInterface::ELFSymbol'
>> Writer::Slot<ELFSymbol::SerializedLayout> dst,
>> ~~~~~~~~~~~^
>> ../../src/gdb-jit.cc:870:29: error: no member named 'SerializedLayout' in
>> 'v8::internal::GDBJITInterface::ELFSymbol'
>> Writer::Slot<ELFSymbol::SerializedLayout> symbols =
>> ~~~~~~~~~~~^
>> ../../src/gdb-jit.cc:870:47: error: C++ requires a type specifier for all
>> declarations
>> Writer::Slot<ELFSymbol::SerializedLayout> symbols =
>> ^
>> ../../src/gdb-jit.cc:871:39: error: no member named 'SerializedLayout' in
>> 'v8::internal::GDBJITInterface::ELFSymbol'
>> w->CreateSlotsHere<ELFSymbol::SerializedLayout>(
>> ~~~~~~~~~~~^
>> ../../src/gdb-jit.cc:871:56: error: expected unqualified-id
>> w->CreateSlotsHere<ELFSymbol::SerializedLayout>(
>> ^
>> ../../src/gdb-jit.cc:880:34: error: no member named 'SerializedLayout' in
>> 'v8::internal::GDBJITInterface::ELFSymbol'
>> symbols.at(0).set(ELFSymbol::SerializedLayout(0,
>> ~~~~~~~~~~~^
>> ../../src/gdb-jit.cc:907:44: error: no member named 'SerializedLayout' in
>> 'v8::internal::GDBJITInterface::ELFSymbol'
>> header->entry_size = sizeof(ELFSymbol::SerializedLayout);
>> ~~~~~~~~~~~^
>> ../../src/gdb-jit.cc:1160:2: error: Unsupported target architecture.
>> #error Unsupported target architecture.
>> ^
>> 12 errors generated.
>> [1469/2173] CXX obj/v8_base/code-stub-assembler.o
>> ninja: build stopped: subcommand failed.
>>
>>
>> On Wednesday, December 12, 2018 at 12:08:13 PM UTC-6, Rodolph Perfetta
>> wrote:
>>>
>>> Just checking, when building v8, you did enable gdbjit?
>>> v8_enable_gdbjit=true in gn args. The blog you pointed to is a bit out of
>>> date.
>>>
>>>
>>>
>>> On Wed, 12 Dec 2018 at 07:49, Jakob Gruber <[email protected]> wrote:
>>>
>>>>
>>>>
>>>> On Tue, Dec 11, 2018 at 11:45 PM <[email protected]> wrote:
>>>>
>>>>> Hi,
>>>>>
>>>>> In some of the profiles that I have been looking at CheckObjectType
>>>>> comes out are pretty hot. I noticed that this is called multiple times
>>>>> from
>>>>> various builtins or other jitted code. I wanted to know where in jitted
>>>>> code or builtins is this call coming from.
>>>>>
>>>>
>>>> This doesn't answer your main question, but CheckObjectType calls only
>>>> occur in debug mode and are emitted when creating TNodes (typed nodes),
>>>> see
>>>> here
>>>> <https://cs.chromium.org/chromium/src/v8/src/compiler/code-assembler.h?l=703&rcl=72448bedcf497b26a742fe96bb1d387d7421d10d>
>>>> .
>>>>
>>>>
>>>>>
>>>>> Unfortunately, I can not step into jitted code while running d8 with
>>>>> gdb. I found the following link which talks about gdb-jit
>>>>> https://v8.dev/docs/gdb-jit. It mostly talks about ia32.
>>>>>
>>>>> I was wondering if anyone has looked at running GDB on d8 on ARM64? Is
>>>>> it possible to break in the native code at all in ARM64?
>>>>>
>>>>> --
>>>> --
>>>> v8-dev mailing list
>>>> [email protected]
>>>> http://groups.google.com/group/v8-dev
>>>> ---
>>>> You received this message because you are subscribed to the Google
>>>> Groups "v8-dev" group.
>>>> To unsubscribe from this group and stop receiving emails from it, send
>>>> an email to [email protected].
>>>> For more options, visit https://groups.google.com/d/optout.
>>>>
>>> --
>> --
>> v8-dev mailing list
>> [email protected] <javascript:>
>> http://groups.google.com/group/v8-dev
>> ---
>> You received this message because you are subscribed to the Google Groups
>> "v8-dev" group.
>> To unsubscribe from this group and stop receiving emails from it, send an
>> email to [email protected] <javascript:>.
>> For more options, visit https://groups.google.com/d/optout.
>>
>
--
--
v8-dev mailing list
[email protected]
http://groups.google.com/group/v8-dev
---
You received this message because you are subscribed to the Google Groups
"v8-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
For more options, visit https://groups.google.com/d/optout.
