Re: [lmi] [lmi-commits] odd/eraseme_error 69d06e9 10/10: Attempt to find a problem

[Vadim Zeitlin]

On Wed,  7 Jul 2021 06:22:15 -0400 (EDT) Greg Chicares 
<gchicares@sbcglobal.net> wrote:

GC> branch: odd/eraseme_error
GC> commit 69d06e9c8f3ff5f6b03b7a538335b7af7bf4a01b
GC> Author: Gregory W. Chicares <gchicares@sbcglobal.net>
GC> Commit: Gregory W. Chicares <gchicares@sbcglobal.net>
GC> 
GC>     Attempt to find a problem
GC>     
GC>     Symptom observed after the immediately preceding commit: 'gui_test.sh'
GC>     freezes with i686-w64-mingw32 and wine-4.0.3, here...
GC>     
GC>         paste_census: started
GC>         time=2067ms (for paste_census)
GC>         paste_census: ok
GC>         pdf_census: started
GC>         ^^^^^^^^^^^^^^^^^^^ freezes
GC>     
GC>     ...which seems to be about where the GUI test moves the mouse pointer.
GC>     Is this a wine defect, an lmi GUI defect, or a defect in 'zero.hpp',
GC>     the only relevant file that had changed since origin/master, where the
GC>     GUI test succeeded?

 Is this hang reproducible? If so, I could try debugging it.

 Generally speaking, the best technique for debugging problems under Wine
remains the one consisting in running gdbserver itself under it, as
described in this message:

        https://lists.nongnu.org/archive/html/lmi/2016-03/msg00023.html

(searching for gdbserver.exe will find it in this rather long -- but what
would you expect considering its author -- message). It's quite simple and
should be perfectly adequate for at least finding where does it hang, as
you should be able to break into gdb (although there are Wine-unrelated
problems with Ctrl-C handling in chroots, so maybe it would be better to
send SIGINT to the process using kill from another shell instead) and use
"bt" to at least see where is it stuck.

 Please let me know if you need my help with this,
VZ

pgpU1UtQuMG4O.pgp
Description: PGP signature