|
From: Robert W. <rj...@du...> - 2006-01-17 02:45:27
|
As of 2 minutes ago, the entire test suite is failing on my amd64 box, all like this: vex: priv/main/vex_main.c:279 (LibVEX_Translate): Assertion `vta->dispatch = !=3D ((void*)0)' failed. vex storage: T total 0 bytes allocated valgrind: the 'impossible' happened: LibVEX called failure_exit(). at 0x........: report_and_quit (m_libcassert.c:136) by 0x........: panic (m_libcassert.c:209) by 0x........: vgPlain_core_panic_at (m_libcassert.c:214) by 0x........: vgPlain_core_panic (m_libcassert.c:219) by 0x........: failure_exit (m_translate.c:470) by 0x........: vex_assert_fail (vex_util.c:218) by 0x........: LibVEX_Translate (vex_main.c:279) by 0x........: vgPlain_translate (m_translate.c:996) by 0x........: handle_tt_miss (scheduler.c:692) by 0x........: vgPlain_scheduler (scheduler.c:864) by 0x........: thread_wrapper (syswrap-linux.c:86) by 0x........: run_a_thread_NORETURN (syswrap-linux.c:119) Looks like some VEX stuff changed earlier. Any thoughts? Regards, Robert. --=20 Robert Walsh Amalgamated Durables, Inc. - "We don't make the things you buy." Email: rj...@du... |
|
From: Julian S. <js...@ac...> - 2006-01-17 03:04:18
|
Even with a clean build, Robin? I did check in a load of stuff just now, but revs 1540/5541 appear to work OK on my amd64 box. J p.s. Shouldn't it be "Holy smoking test suite!" ? On Tuesday 17 January 2006 02:45, Robert Walsh wrote: > As of 2 minutes ago, the entire test suite is failing on my amd64 box, > all like this: > > vex: priv/main/vex_main.c:279 (LibVEX_Translate): Assertion `vta->dispatch > != ((void*)0)' failed. vex storage: T total 0 bytes allocated |
|
From: Robert W. <rj...@du...> - 2006-01-17 03:11:17
|
> Even with a clean build, Robin? I did check in a load of=20 > stuff just now, but revs 1540/5541 appear to work OK on my > amd64 box. Yup - even with a clean build. Details are: Athlon64 with 1GB memory running stock FC4 w/ glibc-2.3.5-10.3. It looks like every single test fails. > p.s. Shouldn't it be "Holy smoking test suite!" ? I thought "flaming" would give a better idea of the severity of the situation :-) Regards, Robert. --=20 Robert Walsh Amalgamated Durables, Inc. - "We don't make the things you buy." Email: rj...@du... |
|
From: Julian S. <js...@ac...> - 2006-01-17 04:25:09
|
On Tuesday 17 January 2006 03:11, Robert Walsh wrote: > > Even with a clean build, Robin? I did check in a load of > > stuff just now, but revs 1540/5541 appear to work OK on my > > amd64 box. > > Yup - even with a clean build. Details are: Athlon64 with 1GB memory > running stock FC4 w/ glibc-2.3.5-10.3. It looks like every single test > fails. Hmm. Looking at the nightly builds, there's clearly something up, but I can't reproduce it. A fresh checkout of 1540/5541 builds and tests ok for me on SuSE 9.2 (amd64) and SuSE 10.0 (ppc32). Might this be to do with svn up-ing an existing tree vs using a new checkout? J |
|
From: Robert W. <rj...@du...> - 2006-01-18 17:43:58
|
On Tue, 2006-01-17 at 04:25 +0000, Julian Seward wrote: > On Tuesday 17 January 2006 03:11, Robert Walsh wrote: > > > Even with a clean build, Robin? I did check in a load of > > > stuff just now, but revs 1540/5541 appear to work OK on my > > > amd64 box. > > > > Yup - even with a clean build. Details are: Athlon64 with 1GB memory > > running stock FC4 w/ glibc-2.3.5-10.3. It looks like every single test > > fails. >=20 > Hmm. Looking at the nightly builds, there's clearly something up, but > I can't reproduce it. A fresh checkout of 1540/5541 builds and tests > ok for me on SuSE 9.2 (amd64) and SuSE 10.0 (ppc32). Might this be > to do with svn up-ing an existing tree vs using a new checkout? Well, that wasn't the problem, but it did go away after later check-ins, so I think everything is OK now. Regards, Robert. --=20 Robert Walsh Amalgamated Durables, Inc. - "We don't make the things you buy." Email: rj...@du... |