|
From: Julian S. <js...@ac...> - 2008-06-02 10:00:02
|
I have put a release candidate for 3.3.1 at http://www.valgrind.org/downloads/valgrind-3.3.1.RC1.tar.bz2 (md5 = 7b8325af2156c679d825d15eb38a0bd4) Please download and try it on platforms that are important to you, and let me know of any critical breakage. If no critical problems are reported, I plan to make make the final 3.3.1 release on Wednesday 4 June. For your convenience, the 3.3.1 release notes are attached below. The main important change is support for glibc-2.8 systems. In particular this tarball works well on Fedora 9, openSUSE 11, and Ubuntu 8.04. J Release 3.3.1 (4 June 2008) ~~~~~~~~~~~~~~~~~~~~~~~~~~~ 3.3.1 fixes a bunch of bugs in 3.3.0, adds support for glibc-2.8 based systems (openSUSE 11, Fedora Core 9), improves the existing glibc-2.7 support, and adds support for the SSSE3 (Core 2) instruction set. 3.3.1 will likely be the last release that supports some very old systems. In particular, the next major release, 3.4.0, will drop support for the old LinuxThreads threading library, and for gcc versions prior to 3.0. The fixed bugs are as follows. Note that "n-i-bz" stands for "not in bugzilla" -- that is, a bug that was reported to us but never got a bugzilla entry. We encourage you to file bugs in bugzilla (http://bugs.kde.org/enter_valgrind_bug.cgi) rather than mailing the developers (or mailing lists) directly -- bugs that are not entered into bugzilla tend to get forgotten about or ignored. n-i-bz Massif segfaults at exit n-i-bz Memcheck asserts on Altivec code n-i-bz fix sizeof bug in Helgrind n-i-bz check fd on sys_llseek n-i-bz update syscall lists to kernel 2.6.23.1 n-i-bz support sys_sync_file_range n-i-bz handle sys_sysinfo, sys_getresuid, sys_getresgid on ppc64-linux n-i-bz intercept memcpy in 64-bit ld.so's n-i-bz Fix wrappers for sys_{futimesat,utimensat} n-i-bz Minor false-error avoidance fixes for Memcheck n-i-bz libmpiwrap.c: add a wrapper for MPI_Waitany n-i-bz helgrind support for glibc-2.8 n-i-bz partial fix for mc_leakcheck.c:698 assert: 'lc_shadows[i]->data + lc_shadows[i] ... n-i-bz Massif/Cachegrind output corruption when programs fork n-i-bz register allocator fix: handle spill stores correctly n-i-bz add support for PA6T PowerPC CPUs 126389 vex x86->IR: 0xF 0xAE (FXRSTOR) 158525 ==126389 152818 vex x86->IR: 0xF3 0xAC (repz lodsb) 153196 vex x86->IR: 0xF2 0xA6 (repnz cmpsb) 155011 vex x86->IR: 0xCF (iret) 155091 Warning [...] unhandled DW_OP_ opcode 0x23 156960 ==155901 155528 support Core2/SSSE3 insns on x86/amd64 155929 ms_print fails on massif outputs containing long lines 157665 valgrind fails on shmdt(0) after shmat to 0 157748 support x86 PUSHFW/POPFW 158212 helgrind: handle pthread_rwlock_try{rd,wr}lock. 158425 sys_poll incorrectly emulated when RES==0 158744 vex amd64->IR: 0xF0 0x41 0xF 0xC0 (xaddb) 160907 Support for a couple of recent Linux syscalls 161285 Patch -- support for eventfd() syscall 161378 illegal opcode in debug libm (FUCOMPP) 160136 ==161378 161487 number of suppressions files is limited to 10 162386 ms_print typo in milliseconds time unit for massif 161036 exp-drd: client allocated memory was never freed 162663 signalfd_wrapper fails on 64bit linux (3.3.1.RC1: 2 June 2008, vex r1854, valgrind r8169). |
|
From: Ashley P. <api...@co...> - 2008-06-02 10:10:54
Attachments:
xml-qual.patch
|
On Mon, 2008-06-02 at 11:53 +0200, Julian Seward wrote: > I have put a release candidate for 3.3.1 at > > http://www.valgrind.org/downloads/valgrind-3.3.1.RC1.tar.bz2 > (md5 = 7b8325af2156c679d825d15eb38a0bd4) > > Please download and try it on platforms that are important to you, and > let me know of any critical breakage. If no critical problems are > reported, I plan to make make the final 3.3.1 release on Wednesday 4 June. Could you apply this patch to re-enable the log file qualifier in the xml output. Also did the VPATH build issue get fixed? It's not hard to do it but it means adding and removing files so it's not something I can send as a patch. Ashley, |
|
From: Bart V. A. <bar...@gm...> - 2008-06-02 18:12:24
|
On Mon, Jun 2, 2008 at 11:53 AM, Julian Seward <js...@ac...> wrote: > > I have put a release candidate for 3.3.1 at > http://www.valgrind.org/downloads/valgrind-3.3.1.RC1.tar.bz2 > (md5 = 7b8325af2156c679d825d15eb38a0bd4) By this time I have made a minor exp-drd documentation fix. I'm not expecting that you rebuild the RC tarball because of this, but in case the tarball has to be rebuilt, it would be nice if that fix could be included. Bart. |
|
From: Rich C. <Ric...@me...> - 2008-06-03 16:55:43
|
Compiled, built, and tested:
Slackware 12.1 i686 smp
== 302 tests, 21 stderr failures, 4 stdout failures, 0 post failures ==
openSUSE 10.2 x86_64 smp
== 336 tests, 59 stderr failures, 50 stdout failures, 0 post failures ==
No problems for the simple test program I ran.
One of my test programs SIGSEGV, but I think that has more to do with the
flakiness in the program than with anythin in V.
R.
On Mon, 2 Jun 2008 11:53:36 +0200
Julian Seward <js...@ac...> wrote:
>
> I have put a release candidate for 3.3.1 at
>
> http://www.valgrind.org/downloads/valgrind-3.3.1.RC1.tar.bz2
> (md5 = 7b8325af2156c679d825d15eb38a0bd4)
>
> Please download and try it on platforms that are important to you, and
> let me know of any critical breakage. If no critical problems are
> reported, I plan to make make the final 3.3.1 release on Wednesday 4 June.
>
> For your convenience, the 3.3.1 release notes are attached below. The
> main important change is support for glibc-2.8 systems. In particular
> this tarball works well on Fedora 9, openSUSE 11, and Ubuntu 8.04.
>
> J
>
>
> Release 3.3.1 (4 June 2008)
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~
> 3.3.1 fixes a bunch of bugs in 3.3.0, adds support for glibc-2.8 based
> systems (openSUSE 11, Fedora Core 9), improves the existing glibc-2.7
> support, and adds support for the SSSE3 (Core 2) instruction set.
>
> 3.3.1 will likely be the last release that supports some very old
> systems. In particular, the next major release, 3.4.0, will drop
> support for the old LinuxThreads threading library, and for gcc
> versions prior to 3.0.
>
> The fixed bugs are as follows. Note that "n-i-bz" stands for "not in
> bugzilla" -- that is, a bug that was reported to us but never got a
> bugzilla entry. We encourage you to file bugs in bugzilla
> (http://bugs.kde.org/enter_valgrind_bug.cgi) rather than mailing the
> developers (or mailing lists) directly -- bugs that are not entered
> into bugzilla tend to get forgotten about or ignored.
>
> n-i-bz Massif segfaults at exit
> n-i-bz Memcheck asserts on Altivec code
> n-i-bz fix sizeof bug in Helgrind
> n-i-bz check fd on sys_llseek
> n-i-bz update syscall lists to kernel 2.6.23.1
> n-i-bz support sys_sync_file_range
> n-i-bz handle sys_sysinfo, sys_getresuid, sys_getresgid on ppc64-linux
> n-i-bz intercept memcpy in 64-bit ld.so's
> n-i-bz Fix wrappers for sys_{futimesat,utimensat}
> n-i-bz Minor false-error avoidance fixes for Memcheck
> n-i-bz libmpiwrap.c: add a wrapper for MPI_Waitany
> n-i-bz helgrind support for glibc-2.8
> n-i-bz partial fix for mc_leakcheck.c:698 assert:
> 'lc_shadows[i]->data + lc_shadows[i] ...
> n-i-bz Massif/Cachegrind output corruption when programs fork
> n-i-bz register allocator fix: handle spill stores correctly
> n-i-bz add support for PA6T PowerPC CPUs
> 126389 vex x86->IR: 0xF 0xAE (FXRSTOR)
> 158525 ==126389
> 152818 vex x86->IR: 0xF3 0xAC (repz lodsb)
> 153196 vex x86->IR: 0xF2 0xA6 (repnz cmpsb)
> 155011 vex x86->IR: 0xCF (iret)
> 155091 Warning [...] unhandled DW_OP_ opcode 0x23
> 156960 ==155901
> 155528 support Core2/SSSE3 insns on x86/amd64
> 155929 ms_print fails on massif outputs containing long lines
> 157665 valgrind fails on shmdt(0) after shmat to 0
> 157748 support x86 PUSHFW/POPFW
> 158212 helgrind: handle pthread_rwlock_try{rd,wr}lock.
> 158425 sys_poll incorrectly emulated when RES==0
> 158744 vex amd64->IR: 0xF0 0x41 0xF 0xC0 (xaddb)
> 160907 Support for a couple of recent Linux syscalls
> 161285 Patch -- support for eventfd() syscall
> 161378 illegal opcode in debug libm (FUCOMPP)
> 160136 ==161378
> 161487 number of suppressions files is limited to 10
> 162386 ms_print typo in milliseconds time unit for massif
> 161036 exp-drd: client allocated memory was never freed
> 162663 signalfd_wrapper fails on 64bit linux
>
> (3.3.1.RC1: 2 June 2008, vex r1854, valgrind r8169).
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft
> Defy all challenges. Microsoft(R) Visual Studio 2008.
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> _______________________________________________
> Valgrind-developers mailing list
> Val...@li...
> https://lists.sourceforge.net/lists/listinfo/valgrind-developers
>
--
Rich Coe ric...@me...
Virtual Principle Engineer General Electric Healthcare Technologies
Global Software Platforms, Computer Technology Team
|
|
From: Julian S. <js...@ac...> - 2008-06-04 06:54:13
|
On Monday 02 June 2008 12:10, Ashley Pittman wrote: > On Mon, 2008-06-02 at 11:53 +0200, Julian Seward wrote: > > I have put a release candidate for 3.3.1 at > > > > http://www.valgrind.org/downloads/valgrind-3.3.1.RC1.tar.bz2 > > (md5 = 7b8325af2156c679d825d15eb38a0bd4) > > > > Please download and try it on platforms that are important to you, and > > let me know of any critical breakage. If no critical problems are > > reported, I plan to make make the final 3.3.1 release on Wednesday 4 > > June. > > Could you apply this patch to re-enable the log file qualifier in the > xml output. Applied to the trunk. Ta. I'd prefer not to change the visible behaviour of the stable branch at this point. > Also did the VPATH build issue get fixed? No .. there's a patch floating around somewhere, isn't there, that you sent before? Not sure where though. Any ideas? J |
|
From: Ashley P. <api...@co...> - 2008-06-05 07:39:16
Attachments:
vpath-omega.patch
|
On Wed, 2008-06-04 at 08:47 +0200, Julian Seward wrote: > On Monday 02 June 2008 12:10, Ashley Pittman wrote: > > On Mon, 2008-06-02 at 11:53 +0200, Julian Seward wrote: > > > I have put a release candidate for 3.3.1 at > > > > > > http://www.valgrind.org/downloads/valgrind-3.3.1.RC1.tar.bz2 > > > (md5 = 7b8325af2156c679d825d15eb38a0bd4) > > > > > > Please download and try it on platforms that are important to you, and > > > let me know of any critical breakage. If no critical problems are > > > reported, I plan to make make the final 3.3.1 release on Wednesday 4 > > > June. > > > > Could you apply this patch to re-enable the log file qualifier in the > > xml output. > > Applied to the trunk. Ta. I'd prefer not to change the visible > behaviour of the stable branch at this point. So this will in the 3.3.1 release? That's all I care about, my users are waiting on the malinfo stuff so will be on that anyway. > > Also did the VPATH build issue get fixed? > > No .. there's a patch floating around somewhere, isn't there, that > you sent before? Not sure where though. Any ideas? I've got this but I don't think it's a full patch, I seem to remember I also had to remove VG_(start_debugger) from pub_core_debugger.h and add it somewhere else, at this point I decided that the entire header file probably needed re-naming as pub_tool_debugger.h hence my comment about not being able to make a patch that encompasses the changes. You'll need to do a svn move and lots of build build changes to do this properly so perhaps it's better to leave that header file empty and create a new one with the declaration in it. Let me know if you want me to look at this deeper, I'm off to work now but will be able to take a proper look this evening if you require. Would it be prudent to adjust one of the nightly builds to use VPATH? It's not the first time it's broken and it often takes a while to spot that's anythings wrong. Ashley, |
|
From: Rich C. <Ric...@me...> - 2008-06-05 19:26:48
|
I'm supposing this patch will fix bug 162753.
http://bugs.kde.org/show_bug.cgi?id=162753
Ashley Pittman <api...@co...> wrote:
>
> On Wed, 2008-06-04 at 08:47 +0200, Julian Seward wrote:
> > On Monday 02 June 2008 12:10, Ashley Pittman wrote:
> > > On Mon, 2008-06-02 at 11:53 +0200, Julian Seward wrote:
> > > > I have put a release candidate for 3.3.1 at
> > > >
> > > > http://www.valgrind.org/downloads/valgrind-3.3.1.RC1.tar.bz2
> > > > (md5 = 7b8325af2156c679d825d15eb38a0bd4)
> > > >
> > > > Please download and try it on platforms that are important to you, and
> > > > let me know of any critical breakage. If no critical problems are
> > > > reported, I plan to make make the final 3.3.1 release on Wednesday 4
> > > > June.
> > >
> > > Could you apply this patch to re-enable the log file qualifier in the
> > > xml output.
> >
> > Applied to the trunk. Ta. I'd prefer not to change the visible
> > behaviour of the stable branch at this point.
>
> So this will in the 3.3.1 release? That's all I care about, my users
> are waiting on the malinfo stuff so will be on that anyway.
>
> > > Also did the VPATH build issue get fixed?
> >
> > No .. there's a patch floating around somewhere, isn't there, that
> > you sent before? Not sure where though. Any ideas?
>
> I've got this but I don't think it's a full patch, I seem to remember I
> also had to remove VG_(start_debugger) from pub_core_debugger.h and add
> it somewhere else, at this point I decided that the entire header file
> probably needed re-naming as pub_tool_debugger.h hence my comment about
> not being able to make a patch that encompasses the changes. You'll
> need to do a svn move and lots of build build changes to do this
> properly so perhaps it's better to leave that header file empty and
> create a new one with the declaration in it.
>
> Let me know if you want me to look at this deeper, I'm off to work now
> but will be able to take a proper look this evening if you require.
>
> Would it be prudent to adjust one of the nightly builds to use VPATH?
> It's not the first time it's broken and it often takes a while to spot
> that's anythings wrong.
>
> Ashley,
>
--
Rich Coe ric...@me...
Virtual Principle Engineer General Electric Healthcare Technologies
Global Software Platforms, Computer Technology Team
|
|
From: Ashley P. <api...@co...> - 2008-06-05 21:58:53
|
When I've had time to finish it properly yes it will. Ashley, On Thu, 2008-06-05 at 14:26 -0500, Rich Coe wrote: > I'm supposing this patch will fix bug 162753. > http://bugs.kde.org/show_bug.cgi?id=162753 > > Ashley Pittman <api...@co...> wrote: > > > > On Wed, 2008-06-04 at 08:47 +0200, Julian Seward wrote: > > > On Monday 02 June 2008 12:10, Ashley Pittman wrote: > > > > On Mon, 2008-06-02 at 11:53 +0200, Julian Seward wrote: > > > > > I have put a release candidate for 3.3.1 at > > > > > > > > > > http://www.valgrind.org/downloads/valgrind-3.3.1.RC1.tar.bz2 > > > > > (md5 = 7b8325af2156c679d825d15eb38a0bd4) > > > > > > > > > > Please download and try it on platforms that are important to you, and > > > > > let me know of any critical breakage. If no critical problems are > > > > > reported, I plan to make make the final 3.3.1 release on Wednesday 4 > > > > > June. > > > > > > > > Could you apply this patch to re-enable the log file qualifier in the > > > > xml output. > > > > > > Applied to the trunk. Ta. I'd prefer not to change the visible > > > behaviour of the stable branch at this point. > > > > So this will in the 3.3.1 release? That's all I care about, my users > > are waiting on the malinfo stuff so will be on that anyway. > > > > > > Also did the VPATH build issue get fixed? > > > > > > No .. there's a patch floating around somewhere, isn't there, that > > > you sent before? Not sure where though. Any ideas? > > > > I've got this but I don't think it's a full patch, I seem to remember I > > also had to remove VG_(start_debugger) from pub_core_debugger.h and add > > it somewhere else, at this point I decided that the entire header file > > probably needed re-naming as pub_tool_debugger.h hence my comment about > > not being able to make a patch that encompasses the changes. You'll > > need to do a svn move and lots of build build changes to do this > > properly so perhaps it's better to leave that header file empty and > > create a new one with the declaration in it. > > > > Let me know if you want me to look at this deeper, I'm off to work now > > but will be able to take a proper look this evening if you require. > > > > Would it be prudent to adjust one of the nightly builds to use VPATH? > > It's not the first time it's broken and it often takes a while to spot > > that's anythings wrong. > > > > Ashley, > > > > |