Signal SIGPIPE received after m_arbitraryHeaders.IsEmpty() is TRUE

UNCONFIRMED
Unassigned

Status

Thunderbird
General
--
critical
UNCONFIRMED
a year ago
4 hours ago

People

(Reporter: Jim Miller, Unassigned, NeedInfo)

Tracking

({crash})

45 Branch
crash

Firefox Tracking Flags

(Not tracked)

Details

Attachments

(1 attachment)

(Reporter)

Description

a year ago
Created attachment 8851610 [details]
tb-crash-032517.rtf

User Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_11_6) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/56.0.2924.87 Safari/537.36

Steps to reproduce:

I've been having repeated crashes (multiple times a day) with TB 45.8.0 on my Mac OS 10.11.6 machine. I finally took the time to compile TB and enable debugging so I could run it productively in lldb.  I've also tried on TB 45.2 and get crashes there as well just not as often.

I think the crash is trigged when a message arrives and Enigmail tries to decrypt but I can't be certain.  See attached debug.

The crash is always a SIGPIPE related error.

Process 44934 stopped
thread #1: tid = 0x14748a9, 0x00007fff9bdfbd36 libdispatch.dylib_dispatch_runloop_root_queue_get_port_4CF, queue = 'com.apple.main-thread', stop reason = signal SIGPIPE frame #0: 0x00007fff9bdfbd36 libdispatch.dylib_dispatch_runloop_root_queue_get_port_4CF
libdispatch.dylib`_dispatch_runloop_root_queue_get_port_4CF:
-> 0x7fff9bdfbd36 <+0>: movq (%rdi), %rax
0x7fff9bdfbd39 <+3>: cmpq -0x1fc65cd8(%rip), %rax ; (void )0x00007fff7c1a42c0: _dispatch_queue_runloop_vtable
0x7fff9bdfbd40 <+10>: jne 0x7fff9bdfbd46 ; <+16>
0x7fff9bdfbd42 <+12>: movl 0x20(%rdi), %eax


Actual results:

App crashes periodically.


Expected results:

App should not crash.

Updated

a year ago
Severity: normal → critical
Component: Untriaged → General
Keywords: crash
We'd need a stacktrace with symbols.
Have you determined the cause to be enigmail?
Flags: needinfo?(jmiller)
(Reporter)

Comment 2

a year ago
I can't be 100% certain that the problem is enigmail but it seems to crash after mimeDecrypt.jsm is run.

Regarding the stack trace with symbols.  I'm not super familiar with lldb but in the attachment I just ran a "bt" to produce the backtrace with each frame.  It looks like symbols were included but I'm not sure if you want something else?  I'll send another crash when it dies again shortly.

Jim
Flags: needinfo?(jmiller)

Comment 3

4 hours ago
Jiim, do you still see this issue when using newest versions?
Flags: needinfo?(jmiller)
You need to log in before you can comment on or make changes to this bug.