Bug#626021: linux-image-2.6.38-2-686: icebp (opcode 0xf1) no longer causing a SIGTRAP, breaks Wine
On Tue, May 10, 2011 at 07:04:15PM -0500, Austin English wrote:
> On Tue, May 10, 2011 at 19:00, Frederic Weisbecker <fweisbec@gmail.com> wrote:
> > On Tue, May 10, 2011 at 06:16:12PM -0500, Austin English wrote:
> >> On Tue, May 10, 2011 at 18:08, Frederic Weisbecker <fweisbec@gmail.com> wrote:
> >> > On Tue, May 10, 2011 at 06:01:28PM -0500, Austin English wrote:
> >> >> On Tue, May 10, 2011 at 17:55, Frederic Weisbecker <fweisbec@gmail.com> wrote:
> >> >> > I'm trying to launch this wine test from dlls/ntdll/tests/
> >> >> > with ../../../tools/runtest -q -P wine -M ntdll.dll -T ../../.. -p
> >> >> > and a git clone of the wine repo from today.
> >> >> >
> >> >> > And I get this result: "shift: 89: can't shift that many", which I have
> >> >> > no clue what it's about.
> >> >> >
> >> >> > So I don't know how to reproduce. Any idea on how I should launch
> >> >> > that wine self test, or which mistake I may have made?
> >> >>
> >> >> do:
> >> >> $ ./configure && make
> >> >>
> >> >> as usual, then:
> >> >> $ cd dlls/ntdll/tests
> >> >> $ make exception.ok
> >> >>
> >> >> to repeat the test after success, 'rm exception.ok' or 'make testclean'.
> >> >
> >> > Cool, so I can reproduce:
> >> >
> >> > exception.c:1463: Tests skipped: child process not mapped at same address (0x68758fe4/0x68753fe4)
> >> >
> >> > Do you know a kernel version that works with this test?
> >>
> >> Works on my gentoo machine, which has 2.6.37. Not sure what custom
> >> patches they're using though, off hand.
> >> You can see test results at http://test.winehq.org/data/. Using
> >> results from a few days ago:
> >> http://test.winehq.org/data/f555c2d8572b83595e5a9742ec4444c3ade17d3d/index_Wine.html
> >>
> >> I see it passes on another debian testing machine, though amd64:
> >> fg-deb64-win32 /
> >> http://test.winehq.org/data/6a3255bbbde0bdec3407ea89ef640b66dc45773c/wine_fg-deb64-win32/version.html
> >> / 2.6.38-2-amd64 /
> >> http://test.winehq.org/data/6a3255bbbde0bdec3407ea89ef640b66dc45773c/wine_fg-deb64-win32/ntdll:exception.html
> >>
> >> I don't know of a version in debian 32-bit that passes. If there are
> >> some old .deb's you'd like me to test, let me know.
> >
> > Hmm, doesn't work for me with amd64 on 2.6.37 (my previous test was 2.6.38.5 on amd64):
> >
> > exception.c:1463: Tests skipped: child process not mapped at same address (0x68758fe4/0x68753fe4)
> >
> > So may be it depends on the config. Or on something else than the kernel.
> >
> > Lemme check with an oldy like 2.6.32
>
> Fails for me with
> http://snapshot.debian.org/archive/debian/20110505T002605Z/pool/main/l/linux-2.6/linux-image-2.6.32-5-686_2.6.32-34_i386.deb
Same for me, amd64.
It seems not adapted to 64 bits in fact.
The part that complains is in an "#ifdef __i386__", and there
is an alternate __x86_64__ section. It seems the section that has been
compiled is the i386 even if I'm running amd64, which may be the
cause of the issue.
So the problem appears to be more in the test implementation of wine.
Reply to: