This is the mail archive of the cygwin-developers@cygwin.com mailing list for the Cygwin project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: 20011101 snapshot broken (also 20011103)


On Mon, Nov 05, 2001 at 11:32:36AM +0900, Kazuhiro Fujieda wrote:
>>>> On Sun, 4 Nov 2001 20:37:18 -0500
>>>> Christopher Faylor <cgf@redhat.com> said:
>
>> Oddly enough, both of the above work fine for me.
>> 
>> If you have a clue about what is now incorrect in the handling of
>> unreliable pipes, please share it.
>
>I'm sorry I don't have any clue at present. I tried to get
>the strace log but failed by STATUS_ACCESS_VIOLATION: 
>
>$ strace -o /tmp/sh.log sh -c '`echo tako`'
> 329833 [main] SH 360109 handle_exceptions: Exception: STATUS_ACCESS_VIOLATION
> 332560 [main] SH 360109 open_stackdumpfile: Dumping stack trace to SH.EXE.stackdump
>$
>
>I attached /tmp/sh.log below. I also failed the JIT debugging
>because gdb also died by STATUS_ACCESS_VIOLATION (NT4sp6a) or
>can't attach the process (on WinME).

Here's how to debug strace problems:

Set a breakpoint somewhere in the beginning of dll_crt0_1.
Then:

(gdb) set strace.active=1
(gdb) continue

This will pinpoint the place where an invalid strace is dying.

I just did this and traced the SEGV to a bad debug_printf call in
dtable::build_fhandler.

cgf


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]