[LLVMbugs] [Bug 18379] New: [TSan RTL] hang in symbolizer code

[bugzilla-daemon at llvm.org]

http://llvm.org/bugs/show_bug.cgi?id=18379

            Bug ID: 18379
           Summary: [TSan RTL] hang in symbolizer code
           Product: compiler-rt
           Version: unspecified
          Hardware: PC
                OS: Linux
            Status: NEW
          Severity: normal
          Priority: P
         Component: compiler-rt
          Assignee: unassignedbugs at nondot.org
          Reporter: todd at lipcon.org
                CC: llvmbugs at cs.uiuc.edu
    Classification: Unclassified

Running a program that reports a lot of TSAN errors, I ended up with one of my
threads stuck in the following stack:

(gdb) bt
#0  0x00007fab3cb07319 in syscall () from /lib64/libc.so.6
#1  0x00007fab3febaa61 in __sanitizer::internal_read(int, void*, unsigned long)
()
#2  0x00007fab3fef3fe8 in __tsan::SymbolizeCodeAddr2Line(unsigned long) ()
#3  0x00007fab3feefbbd in __tsan::SymbolizeStack(__tsan::StackTrace const&) ()
#4  0x00007fab3fef1aa2 in __tsan::PrintCurrentStack(__tsan::ThreadState*,
unsigned long) ()
#5  0x00007fab3feeee88 in __tsan::MutexUnlock(__tsan::ThreadState*, unsigned
long, unsigned long) ()

If I look for hung addr2line processes, I see:

(gdb) bt
#0  0x0000003c982db630 in __read_nocancel () from /lib64/libc.so.6
#1  0x0000003c98271fa8 in _IO_new_file_underflow () from /lib64/libc.so.6
#2  0x0000003c98273aae in _IO_default_uflow_internal () from /lib64/libc.so.6
#3  0x0000003c9826818a in _IO_getline_info_internal () from /lib64/libc.so.6
#4  0x0000003c98266fe9 in fgets () from /lib64/libc.so.6
#5  0x00000000004020e9 in ?? ()
#6  0x0000003c9821ecdd in __libc_start_main () from /lib64/libc.so.6

which implies that, for some reason, the Addr2Line code is sending some kind of
partial input into the addr2line binary, and then expecting a response.
Addr2line is still expecting the input, and we end up hanging forever.

This is with llvm/clang 3.3.

-- 
You are receiving this mail because:
You are on the CC list for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-bugs/attachments/20140104/d5e49f5f/attachment.html>