[cfe-dev] [3.8 Release] RC1 has been tagged

[Dimitry Andric via cfe-dev]

Unfortunately I'm having lots of trouble with rc1 at this point:
* libcxxabi can't build, because it requires unwind.h, which we do not yet have on FreeBSD 10.x (Ed Maste is working on it for 11.x, but that is not ready for general consumption).
* The test-release.sh script has no option to disable only libcxxabi, you can only disable libcxx, libcxxabi and libunwind together (maybe this can be improved)
* Last time I hand-built libcxx, it still had a lot of test failures in the locale parts, but I haven't had time to investigate.
* OpenMP does not support i386-freebsd, so I have to disable it there
* Last but not least: the host compiler on FreeBSD 10.x is clang 3.4.1 (the last version that can build without C++11 support), and it crashes with a segfault during building of CGBlocks.cpp.  I'll need to find some way to work around this failure, since we cannot upgrade the compiler easily on FreeBSD 10.x.

I also had to hack the test-release.sh script to fix a number of problems that I encountered during the 3.7.1 release, but haven't gotten to upstreaming them.  E.g. the way the source code is checked out with symlinks all over the place does not work, and I need to add a custom patch to clang-tools-extra to make the tests succeed, because there is a race condition in the Makefile.

-Dimitry

> On 20 Jan 2016, at 11:30, Daniel Sanders <Daniel.Sanders at imgtec.com> wrote:
> 
> This isn't rc1 but the tip of the release branch had some X86 test failures on my most recent build:
> Failing Tests (24):
>    libc++ :: std/input.output/file.streams/fstreams/filebuf.virtuals/overflow.pass.cpp
>    libc++ :: std/input.output/file.streams/fstreams/filebuf.virtuals/underflow.pass.cpp
>    libc++ :: std/input.output/iostream.format/ext.manip/get_time.pass.cpp
>    libc++ :: std/input.output/iostream.format/ext.manip/put_time.pass.cpp
>    libc++ :: std/input.output/iostreams.base/ios.base/ios.base.callback/register_callback.pass.cpp
>    libc++ :: std/input.output/iostreams.base/ios.base/ios.base.locales/imbue.pass.cpp
>    libc++ :: std/input.output/iostreams.base/ios/basic.ios.members/imbue.pass.cpp
>    libc++ :: std/input.output/stream.buffers/streambuf/streambuf.cons/copy.pass.cpp
>    libc++ :: std/input.output/stream.buffers/streambuf/streambuf.cons/default.pass.cpp
>    libc++ :: std/input.output/stream.buffers/streambuf/streambuf.protected/streambuf.assign/assign.pass.cpp
>    libc++ :: std/input.output/stream.buffers/streambuf/streambuf.protected/streambuf.assign/swap.pass.cpp
>    libc++ :: std/localization/locale.categories/category.collate/locale.collate.byname/hash.pass.cpp
>    libc++ :: std/localization/locale.categories/category.collate/locale.collate.byname/types.pass.cpp
>    libc++ :: std/localization/locale.categories/category.ctype/locale.codecvt.byname/ctor_wchar_t.pass.cpp
>    libc++ :: std/localization/locale.categories/category.ctype/locale.ctype.byname/types.pass.cpp
>    libc++ :: std/localization/locales/locale/locale.cons/default.pass.cpp
>    libc++ :: std/localization/locales/locale/locale.members/name.pass.cpp
>    libc++ :: std/localization/locales/locale/locale.operators/eq.pass.cpp
>    libc++ :: std/localization/locales/locale/locale.statics/global.pass.cpp
>    libc++ :: std/re/re.regex/re.regex.locale/imbue.pass.cpp
>    libc++ :: std/re/re.traits/default.pass.cpp
>    libc++ :: std/re/re.traits/getloc.pass.cpp
>    libc++ :: std/re/re.traits/imbue.pass.cpp
>    libomp :: barrier/omp_barrier.c
> 
> Big-endian mips gave this during phase 3:
> 	CMake Error at cmake/modules/HandleLLVMOptions.cmake:43 (message):
> 	  Host Clang must be able to find libstdc++4.7 or newer!
> It's possible that this is a machine config issue. We moved offices over the weekend (just across the street) and my normal machine somehow lost the ability to boot in the process. I'm borrowing a replacement at the moment.
> 
> Little-endian mips is just about to finish Phase 2 so I'll know if it's a more general problem soon.
> 
>> -----Original Message-----
>> From: hwennborg at google.com [mailto:hwennborg at google.com] On Behalf
>> Of Hans Wennborg
>> Sent: 19 January 2016 23:56
>> To: Ben Pope; Dimitry Andric; Daniel Sanders; Nikola Smiljanić; Brian Cain;
>> Renato Golin; Sylvestre Ledru
>> Cc: cfe-dev; lldb-dev at lists.llvm.org; openmp-dev at lists.llvm.org; llvm-dev
>> Subject: Re: [3.8 Release] RC1 has been tagged
>> 
>> (cc'ing non-legacy llvm-dev this time; apologies if you get this
>> twice. Please don't reply-all to the first one.)
>> 
>> On Tue, Jan 19, 2016 at 3:47 PM, Hans Wennborg <hans at chromium.org>
>> wrote:
>>> Dear testers,
>>> 
>>> Start your engines; 3.8.0-rc1 was just tagged from the 3.8 branch at
>>> r258223. (It took a little longer than I'd planned, sorry about that.)
>>> 
>>> There are still a bunch of open merge requests and bug reports, but I
>>> wanted to get the tag in so we can see what the build and test status
>>> are on the various platforms.
>>> 
>>> I verified that it currently builds and tests cleanly for me on x86_64
>>> Linux, Mac OS X* and Windows.
>>> 
>>> Please build, test, and upload binaries to the sftp. Let me know if how it
>> goes.
>>> 
>>> Thanks,
>>> Hans
>>> 
>>> 
>>> [*] For Mac, I had to set CFLAGS="-isysroot `xcrun -show-sdk-path`"
>>> CXXFLAGS="-isysroot `xcrun -show-sdk-path`" for the build to work,
>>> otherwise stage-2 Clang couldn't find the SDK. I don't remember if I
>>> had to do this last time; maybe some upgrade changed something.

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 194 bytes
Desc: Message signed with OpenPGP using GPGMail
URL: <http://lists.llvm.org/pipermail/cfe-dev/attachments/20160120/e720be85/attachment.sig>