저에게 항상 밝은 미소로 하루종일 저와같은 어려운 분들에게 항상 작은 힘을 주고 있는 분들에게 저가 고마운 마음으로 올해가 끝나기전에
새해 인사를 하고자 합니다... 감사합니다.

2016. 12. 5. 오후 8:38에 "ranjit… via monorail" <
monorail+v2.3493406661@chromium.org>님이 작성:

Unable to reproduce the issue on Linux 14.04 chrome version 56.0.2924.18 - chrome://inspect/#devices is displayed fine.

Could you please upgrade to latest dev and see if issue still exists.

Yes.

An addition observation. If chrome://inspect is the only tab, it works fine, until a new tab is opened, upon which it takes down the browser immediately.

I was mistaken. It is less unstable when the only tab open, but eventually exhibits the same problem.

If you can reproduce this on an official build can you post the crash ID from chrome://crashes here? That will let me look up the stack trace. I haven't been able to reproduce this on HEAD.

​It's not an official build, but now that I've found the instructions, I'll
produce a stack trace.​

OK, in a build with debug symbols, this message is printed in console once per second. May or may not be related.

[18269:18339:1209/165346.707656:ERROR:socket_posix.cc(82)] CreatePlatformSocket() returned an error, errno=97: Address family not supported by protocol

And the stack trace from gdb.

Program received signal SIGSEGV, Segmentation fault.
0x0000555556d152fd in Invoke<base::WeakPtr<content::(anonymous namespace)::WebContentsCaptureMachine> const&> (
    method=&virtual table offset 93825019270908, receiver_ptr=...) at ../../base/bind_internal.h:214
214	../../base/bind_internal.h: No such file or directory.

(gdb) bt
#0  0x0000555556d152fd in Invoke<base::WeakPtr<content::(anonymous namespace)::WebContentsCaptureMachine> const&> (
    method=&virtual table offset 93825019270908, receiver_ptr=...) at ../../base/bind_internal.h:214
#1  MakeItSo<void (content::(anonymous namespace)::WebContentsCaptureMachine::* const&)(), base::WeakPtr<content::(anonymous namespace)::WebContentsCaptureMachine> const&> (functor=<optimized out>, weak_ptr=...)
    at ../../base/bind_internal.h:305
#2  RunImpl<void (content::(anonymous namespace)::WebContentsCaptureMachine::* const&)(), std::tuple<base::WeakPtr<content::(anonymous namespace)::WebContentsCaptureMachine> > const&, 0> (functor=<optimized out>, bound=...)
    at ../../base/bind_internal.h:361
#3  base::internal::Invoker<base::internal::BindState<void (content::(anonymous namespace)::WebContentsCaptureMachine::*)(), base::WeakPtr<content::(anonymous namespace)::WebContentsCaptureMachine> >, void ()>::Run(base::internal::BindStateBase*) (base=0x1d7bb330c500) at ../../base/bind_internal.h:339
#4  0x00005555573bc4ce in Run (this=<optimized out>) at ../../base/callback.h:47
#5  base::debug::TaskAnnotator::RunTask (this=<optimized out>, queue_function=0x55555a1203e3 "MessageLoop::PostTask", 
    pending_task=0x7fffffffd930) at ../../base/debug/task_annotator.cc:52
#6  0x000055555735a71c in base::MessageLoop::RunTask (this=0x1d7bb290f380, pending_task=0x7fffffffd930)
    at ../../base/message_loop/message_loop.cc:413
#7  0x000055555735a9c8 in base::MessageLoop::DeferOrRunPendingTask (this=0x1d7bb290f380, pending_task=...)
    at ../../base/message_loop/message_loop.cc:422
#8  0x000055555735ae8e in base::MessageLoop::DoDelayedWork (this=0x1d7bb290f380, 
    next_delayed_work_time=<optimized out>) at ../../base/message_loop/message_loop.cc:554
#9  0x000055555735c2c0 in HandleDispatch (this=<optimized out>) at ../../base/message_loop/message_pump_glib.cc:278
#10 base::(anonymous namespace)::WorkSourceDispatch (source=<optimized out>, unused_func=<optimized out>, 
    unused_data=<optimized out>) at ../../base/message_loop/message_pump_glib.cc:109
#11 0x00007ffff6b3ae04 in g_main_context_dispatch () from /lib/x86_64-linux-gnu/libglib-2.0.so.0
#12 0x00007ffff6b3b048 in ?? () from /lib/x86_64-linux-gnu/libglib-2.0.so.0
#13 0x00007ffff6b3b0ec in g_main_context_iteration () from /lib/x86_64-linux-gnu/libglib-2.0.so.0
#14 0x000055555735c186 in base::MessagePumpGlib::Run (this=0x1d7bb298e680, delegate=<optimized out>)
    at ../../base/message_loop/message_pump_glib.cc:309
#15 0x0000555557375951 in base::RunLoop::Run (this=0x7fffffffdbe8) at ../../base/run_loop.cc:35
#16 0x0000555557289e81 in ChromeBrowserMainParts::MainMessageLoopRun (this=0x1d7bb2852b40, result_code=0x1d7bb284ba98)
    at ../../chrome/browser/chrome_browser_main.cc:2010
#17 0x0000555556be6854 in content::BrowserMainLoop::RunMainMessageLoopParts (this=0x1d7bb284ba80)
    at ../../content/browser/browser_main_loop.cc:984
#18 0x0000555556be919f in content::BrowserMainRunnerImpl::Run (this=0x1d7bb283bf80)
    at ../../content/browser/browser_main_runner.cc:141
#19 0x0000555556be2f36 in content::BrowserMain (parameters=...) at ../../content/browser/browser_main.cc:46
#20 0x00005555570f3d10 in content::ContentMainRunnerImpl::Run (this=0x1d7bb28382d0)
    at ../../content/app/content_main_runner.cc:774
#21 0x00005555570f2e94 in content::ContentMain (params=...) at ../../content/app/content_main.cc:20
#22 0x00005555560d1a37 in ChromeMain (argc=5, argv=0x7fffffffde88) at ../../chrome/app/chrome_main.cc:108
#23 0x00007ffff1626f45 in __libc_start_main () from /lib/x86_64-linux-gnu/libc.so.6
#24 0x00005555560d1925 in _start ()

I cannot reliably reproduce this now by just waiting on chrome://inspect. The bt is from opening chrome://inspect, and then opening a new tab, which works reliably.

The same bt, but with glib debug symbols, which I don't think matter.

(gdb) bt
#0  0x0000555556d152fd in Invoke<base::WeakPtr<content::(anonymous namespace)::WebContentsCaptureMachine> const&> (
    method=&virtual table offset 93825019270908, receiver_ptr=...) at ../../base/bind_internal.h:214
#1  MakeItSo<void (content::(anonymous namespace)::WebContentsCaptureMachine::* const&)(), base::WeakPtr<content::(anonymous namespace)::WebContentsCaptureMachine> const&> (functor=<optimized out>, weak_ptr=...)
    at ../../base/bind_internal.h:305
#2  RunImpl<void (content::(anonymous namespace)::WebContentsCaptureMachine::* const&)(), std::tuple<base::WeakPtr<content::(anonymous namespace)::WebContentsCaptureMachine> > const&, 0> (functor=<optimized out>, bound=...)
    at ../../base/bind_internal.h:361
#3  base::internal::Invoker<base::internal::BindState<void (content::(anonymous namespace)::WebContentsCaptureMachine::*)(), base::WeakPtr<content::(anonymous namespace)::WebContentsCaptureMachine> >, void ()>::Run(base::internal::BindStateBase*) (base=0xf6cdeb171c0) at ../../base/bind_internal.h:339
#4  0x00005555573bc4ce in Run (this=<optimized out>) at ../../base/callback.h:47
#5  base::debug::TaskAnnotator::RunTask (this=<optimized out>, queue_function=0x55555a1203e3 "MessageLoop::PostTask", 
    pending_task=0x7fffffffd930) at ../../base/debug/task_annotator.cc:52
#6  0x000055555735a71c in base::MessageLoop::RunTask (this=0xf6cdded5380, pending_task=0x7fffffffd930)
    at ../../base/message_loop/message_loop.cc:413
#7  0x000055555735a9c8 in base::MessageLoop::DeferOrRunPendingTask (this=0xf6cdded5380, pending_task=...)
    at ../../base/message_loop/message_loop.cc:422
#8  0x000055555735ae8e in base::MessageLoop::DoDelayedWork (this=0xf6cdded5380, next_delayed_work_time=<optimized out>)
    at ../../base/message_loop/message_loop.cc:554
#9  0x000055555735c2c0 in HandleDispatch (this=<optimized out>) at ../../base/message_loop/message_pump_glib.cc:278
#10 base::(anonymous namespace)::WorkSourceDispatch (source=<optimized out>, unused_func=<optimized out>, 
    unused_data=<optimized out>) at ../../base/message_loop/message_pump_glib.cc:109
#11 0x00007ffff6b3ae04 in g_main_dispatch (context=0xf6cdddfacc0) at /build/buildd/glib2.0-2.40.2/./glib/gmain.c:3064
#12 g_main_context_dispatch (context=context@entry=0xf6cdddfacc0) at /build/buildd/glib2.0-2.40.2/./glib/gmain.c:3663
#13 0x00007ffff6b3b048 in g_main_context_iterate (context=context@entry=0xf6cdddfacc0, block=block@entry=1, 
    dispatch=dispatch@entry=1, self=<optimized out>) at /build/buildd/glib2.0-2.40.2/./glib/gmain.c:3734
#14 0x00007ffff6b3b0ec in g_main_context_iteration (context=0xf6cdddfacc0, may_block=1)
    at /build/buildd/glib2.0-2.40.2/./glib/gmain.c:3795
#15 0x000055555735c186 in base::MessagePumpGlib::Run (this=0xf6cddf53600, delegate=<optimized out>)
    at ../../base/message_loop/message_pump_glib.cc:309
#16 0x0000555557375951 in base::RunLoop::Run (this=0x7fffffffdbe8) at ../../base/run_loop.cc:35
#17 0x0000555557289e81 in ChromeBrowserMainParts::MainMessageLoopRun (this=0xf6cdde17b40, result_code=0xf6cdde10a98)
    at ../../chrome/browser/chrome_browser_main.cc:2010
#18 0x0000555556be6854 in content::BrowserMainLoop::RunMainMessageLoopParts (this=0xf6cdde10a80)
    at ../../content/browser/browser_main_loop.cc:984
#19 0x0000555556be919f in content::BrowserMainRunnerImpl::Run (this=0xf6cdde00f80)
    at ../../content/browser/browser_main_runner.cc:141
#20 0x0000555556be2f36 in content::BrowserMain (parameters=...) at ../../content/browser/browser_main.cc:46
#21 0x00005555570f3d10 in content::ContentMainRunnerImpl::Run (this=0xf6cdddfd2d0)
    at ../../content/app/content_main_runner.cc:774
#22 0x00005555570f2e94 in content::ContentMain (params=...) at ../../content/app/content_main.cc:20
#23 0x00005555560d1a37 in ChromeMain (argc=5, argv=0x7fffffffde88) at ../../chrome/app/chrome_main.cc:108
#24 0x00007ffff1626f45 in __libc_start_main () from /lib/x86_64-linux-gnu/libc.so.6
#25 0x00005555560d1925 in _start ()

I cannot reproduce this with Debug builds, nor with Chromium snapshots from around the same revision.

OK, this is more likely to be a clang bug, than a Chrome bug. I can only reproduce it with a non-default release build configuration that I've been using.

optimize_for_size = true # false on Linux by default

Further, compiler/BUILD.gn is patched.

-        "-march=x86-64",
+        "-march=haswell",

With either missing, it doesn't reproduce.

It appears this commit has caused it.

https://chromium.googlesource.com/chromium/src/+/99de02ba952b0a69291f81c5b8ca14d81cc1f74f

I've previously build M53 successfully with the above configuration, and the mentioned commit was introduced in 54.0.2786.0 so it adds up. I've let the author know, in case.

제주특별자치도 서귀포시 남원읍 태수로 26번길14


2016. 12. 10. 오전 8:17에 "pdk… via monorail" <
monorail+v2.2825603754@chromium.org>님이 작성:

I would try "disas" to see how the broken pointer to member function
(broken vtable offset) has produced...

56.0.2924.28

Program received signal SIGSEGV, Segmentation fault.
0x0000555556cd8bad in Invoke<base::WeakPtr<content::(anonymous namespace)::WebContentsCaptureMachine> const&> (
    method=&virtual table offset 93825019022562, receiver_ptr=...) at ../../base/bind_internal.h:214
214	    return (receiver.*method)(std::forward<RunArgs>(args)...);

(gdb) disas
Dump of assembler code for function base::internal::Invoker<base::internal::BindState<void (content::(anonymous namespace)::WebContentsCaptureMachine::*)(), base::WeakPtr<content::(anonymous namespace)::WebContentsCaptureMachine> >, void ()>::Run(base::internal::BindStateBase*):
   0x0000555556cd8b64 <+0>:	push   %r15
   0x0000555556cd8b66 <+2>:	push   %r14
   0x0000555556cd8b68 <+4>:	push   %r12
   0x0000555556cd8b6a <+6>:	push   %rbx
   0x0000555556cd8b6b <+7>:	push   %rax
   0x0000555556cd8b6c <+8>:	mov    %rdi,%r14
   0x0000555556cd8b6f <+11>:	lea    0x30(%r14),%rbx
   0x0000555556cd8b73 <+15>:	mov    %rbx,%rdi
   0x0000555556cd8b76 <+18>:	callq  0x55555731bf00 <base::internal::WeakReference::is_valid() const>
   0x0000555556cd8b7b <+23>:	test   %al,%al
   0x0000555556cd8b7d <+25>:	je     0x555556cd8bbf <base::internal::Invoker<base::internal::BindState<void (content::(anonymous namespace)::WebContentsCaptureMachine::*)(), base::WeakPtr<content::(anonymous namespace)::WebContentsCaptureMachine> >, void ()>::Run(base::internal::BindStateBase*)+91>
   0x0000555556cd8b7f <+27>:	cmpq   $0x0,0x8(%rbx)
   0x0000555556cd8b84 <+32>:	je     0x555556cd8bbf <base::internal::Invoker<base::internal::BindState<void (content::(anonymous namespace)::WebContentsCaptureMachine::*)(), base::WeakPtr<content::(anonymous namespace)::WebContentsCaptureMachine> >, void ()>::Run(base::internal::BindStateBase*)+91>
   0x0000555556cd8b86 <+34>:	mov    0x20(%r14),%r12
   0x0000555556cd8b8a <+38>:	mov    0x28(%r14),%r15
   0x0000555556cd8b8e <+42>:	mov    %rbx,%rdi
   0x0000555556cd8b91 <+45>:	callq  0x55555731bf00 <base::internal::WeakReference::is_valid() const>
   0x0000555556cd8b96 <+50>:	mov    %r12,%rcx
   0x0000555556cd8b99 <+53>:	xor    %edi,%edi
   0x0000555556cd8b9b <+55>:	test   %al,%al
   0x0000555556cd8b9d <+57>:	cmovne 0x38(%r14),%rdi
   0x0000555556cd8ba2 <+62>:	add    %r15,%rdi
   0x0000555556cd8ba5 <+65>:	test   $0x1,%cl
   0x0000555556cd8ba8 <+68>:	je     0x555556cd8bb2 <base::internal::Invoker<base::internal::BindState<void (content::(anonymous namespace)::WebContentsCaptureMachine::*)(), base::WeakPtr<content::(anonymous namespace)::WebContentsCaptureMachine> >, void ()>::Run(base::internal::BindStateBase*)+78>
   0x0000555556cd8baa <+70>:	mov    (%rdi),%rax
=> 0x0000555556cd8bad <+73>:	mov    -0x1(%rax,%rcx,1),%rcx
   0x0000555556cd8bb2 <+78>:	add    $0x8,%rsp
   0x0000555556cd8bb6 <+82>:	pop    %rbx
   0x0000555556cd8bb7 <+83>:	pop    %r12
   0x0000555556cd8bb9 <+85>:	pop    %r14
   0x0000555556cd8bbb <+87>:	pop    %r15
   0x0000555556cd8bbd <+89>:	jmpq   *%rcx
   0x0000555556cd8bbf <+91>:	add    $0x8,%rsp
   0x0000555556cd8bc3 <+95>:	pop    %rbx
   0x0000555556cd8bc4 <+96>:	pop    %r12
   0x0000555556cd8bc6 <+98>:	pop    %r14
   0x0000555556cd8bc8 <+100>:	pop    %r15
   0x0000555556cd8bca <+102>:	retq   
End of assembler dump.

Doesn't seem to be an issue caused by DevTools

pdknsk@gmail.com， Can you provide the crash ID? it is in chrome://crashes/

Thank you for providing more feedback. Adding requester "tkonchada@chromium.org" for another review and adding "Needs-Review" label for tracking.

For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot

Unable to reproduce the issue on windows 7, Linux Ubuntu 14.04 using chrome version 56.0.2924.76.

pdknsk@Could you please confirm are you still facing this issue on latest chrome version 56.0.2924.76.Please provide us the crash id from chrome://crashes to triage the issue further.

Thanks,

no response for 2 weeks. close this bug.