http://bugs.winehq.org/show_bug.cgi?id=15272
Summary: Pipes (PeekNamedPipe, NtQueryInformationFile) crash Wine Product: Wine Version: 1.0.0 Platform: Other URL: http://www.inf.unisi.ch/phd/bonzini/test-poll.tar.gz OS/Version: other Status: UNCONFIRMED Severity: normal Priority: P2 Component: kernel32 AssignedTo: wine-bugs@winehq.org ReportedBy: bonzini@gnu.org
The program in the URL, which tests a native Windows emulation of poll(2) similar to Cygwin's, exposes several bugs in Wine:
1) if compiled with -DINTERACTIVE, Wine crashes on test_pipe if test_tty is enabled
2) if not compiled with -DINTERACTIVE, Wine crashes on test_pipe if it is not placed first.
3) test_pipe crashes Wineserver (there is a comment in the function -- if you comment everything after the function, it works).
It should be compiled using MinGW tools (./configure && make), and run with "wine gltests/test-poll.exe".
The TTY test fails, but as I couldn't test it under Windows, I don't know if it is a bug. The important part is the crash on pipes, anyway! On a brighter note, socket tests pass with flashing lights.
http://bugs.winehq.org/show_bug.cgi?id=15272
Paolo Bonzini bonzini@gnu.org changed:
What |Removed |Added ---------------------------------------------------------------------------- Keywords| |download, source, testcase
http://bugs.winehq.org/show_bug.cgi?id=15272
Paolo Bonzini bonzini@gnu.org changed:
What |Removed |Added ---------------------------------------------------------------------------- OS/Version|other |Mac OS X 10.5 Platform|Other |PC
http://bugs.winehq.org/show_bug.cgi?id=15272
Dan Kegel dank@kegel.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |dank@kegel.com Status|UNCONFIRMED |NEW Component|kernel32 |wineserver Ever Confirmed|0 |1 OS/Version|Mac OS X 10.5 |All
--- Comment #1 from Dan Kegel dank@kegel.com 2008-09-13 10:09:42 --- The wineserver crash seems more serious, setting component to wineserver. Reproduced with very recent wine.
Program received signal SIGSEGV, Segmentation fault. default_fd_signaled (obj=0x822c238, thread=0x8224060) at fd.c:1737 1737 int ret = fd->signaled; (gdb) bt #0 default_fd_signaled (obj=0x822c238, thread=0x8224060) at fd.c:1737 #1 0x0806f0f2 in check_wait (thread=0x8224060) at thread.c:547 #2 0x0807110f in req_select (req=0x8224148, reply=0xbf83962c) at thread.c:672 #3 0x0806b838 in call_req_handler (thread=<value optimized out>) at request.c:288 #4 0x0806b947 in read_request (thread=0x8224060) at request.c:343 #5 0x08056634 in main_loop () at fd.c:455 #6 0x0805a9c5 in main (argc=4, argv=0xbf839dd4) at main.c:147 (gdb) print fd $1 = (struct fd *) 0x0
To build, I had to jank it around a bit: CC=i586-mingw32msvc-gcc AR=i586-mingw32msvc-ar RANLIB=i586-mingw32msvc-ranlib ./configure
then edit gllib/Makefile to fix the broken AR and RANLIB definitions, which ignored the override.
Then it built.
http://bugs.winehq.org/show_bug.cgi?id=15272
Vitaliy Margolen vitaliy@kievinfo.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Severity|normal |major OS/Version|All |other
http://bugs.winehq.org/show_bug.cgi?id=15272
--- Comment #2 from Dan Kegel dank@kegel.com 2008-09-13 12:14:07 --- Configuring like
./configure CC=i586-mingw32msvc-gcc AR=i586-mingw32msvc-ar RANLIB=i586-mingw32msvc-ranlib
seemed slightly better, I only had to fix the AR override.
http://bugs.winehq.org/show_bug.cgi?id=15272
--- Comment #3 from Paolo Bonzini bonzini@gnu.org 2008-09-14 09:30:51 --- Just ./configure --host=i586-mingw32msvc worked here.
http://bugs.winehq.org/show_bug.cgi?id=15272
--- Comment #4 from Vincent Povirk madewokherd@gmail.com 2009-06-08 15:51:59 --- Created an attachment (id=21658) --> (http://bugs.winehq.org/attachment.cgi?id=21658) simpler test case
It seems the crash is triggered by waiting on a named pipe client that has been disconnected. Here it is as a patch to the Wine test suite.
http://bugs.winehq.org/show_bug.cgi?id=15272
Dan Kegel dank@kegel.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Target Milestone|--- |1.2.0
--- Comment #5 from Dan Kegel dank@kegel.com 2009-06-08 15:56:45 --- easy test case, crucial component -> nominating for 1.2
http://bugs.winehq.org/show_bug.cgi?id=15272
--- Comment #6 from Alexandre Julliard julliard@winehq.org 2009-06-08 16:08:22 --- Created an attachment (id=21660) --> (http://bugs.winehq.org/attachment.cgi?id=21660) Check for valid fd in default_fd_signaled
Something like this should fix it.
http://bugs.winehq.org/show_bug.cgi?id=15272
David Laštovička david@lastovicka.cz changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |david@lastovicka.cz
http://bugs.winehq.org/show_bug.cgi?id=15272
Alexandre Julliard julliard@winehq.org changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |FIXED
--- Comment #7 from Alexandre Julliard julliard@winehq.org 2009-10-28 06:37:59 --- Fixed by 7a344c14af0d47c4479198bdfe4c8f1923ec5ca1.
http://bugs.winehq.org/show_bug.cgi?id=15272
Alexandre Julliard julliard@winehq.org changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |CLOSED
--- Comment #8 from Alexandre Julliard julliard@winehq.org 2009-11-13 12:43:11 --- Closing bugs fixed in 1.1.33.
-
wine-bugs＠winehq.org