State of Clang in MSYS2

classic Classic list List threaded Threaded
9 messages Options
Reply | Threaded
Open this post in threaded view
|

State of Clang in MSYS2

Eric Fiselier via cfe-dev
Hello,

I'm involved with maintenance of Clang and related LLVM packages for MSYS2.
It was based on Martell's patches (with his permission for MSYS2) from his git repository and MSYS2 own fixes so number of patches might be overwhelming: https://github.com/Alexpux/MINGW-packages/tree/master/mingw-w64-clang
Maybe some of these patches could make it's way to the upstream.

Tools state at version 3.9.1 :
Clang:        mostly works (there are issues found)
LLMV:        mostly works (not tested well enough)
LLD:           creates broken executable
LLDB:         runs but not further tested since lldb-server is not build-able for mingw-w64
libc++:        need further testing
libc++abi:    need further testing
compiler-rt:  need further testing


Any help will be welcome.


make check seems to hang during LLVM tests but here is make check-clang:
Failing Tests (18):
    Clang :: Driver/mingw.cpp
    Clang :: Modules/dependency-dump.m
    Clang :: Modules/dependency-gen.modulemap
    Clang :: PCH/chain-categories.m
    Clang :: PCH/chain-categories2.m
    Clang :: PCH/chain-class-extension.m
    Clang :: PCH/chain-conversion-lookup.cpp
    Clang :: PCH/chain-cxx.cpp
    Clang :: PCH/chain-default-argument-instantiation.cpp
    Clang :: PCH/chain-empty-initial-namespace.cpp
    Clang :: PCH/chain-friend-instantiation.cpp
    Clang :: PCH/chain-implicit-definition.cpp
    Clang :: PCH/chain-late-anonymous-namespace.cpp
    Clang :: PCH/chain-openmp-threadprivate.cpp
    Clang :: PCH/chain-pending-instantiations.cpp
    Clang :: PCH/chain-staticvar-instantiation.cpp
    Clang :: PCH/cxx-chain-function-template.cpp
    Clang :: PCH/replaced-decl.m

  Expected Passes    : 9349
  Expected Failures  : 16
  Unsupported Tests  : 227
  Unexpected Failures: 18

Full output is available here: https://gist.github.com/mati865/b4a2c96f82b22ac41eca32714092db1d

Is there any interest from You in providing at lest minimal support to it?

Cheers,
Mateusz Mikuła


_______________________________________________
cfe-dev mailing list
[hidden email]
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-dev
Reply | Threaded
Open this post in threaded view
|

Re: State of Clang in MSYS2

Eric Fiselier via cfe-dev
I would volunteer to try to get these patches in, but I don't have the bandwidth for it. I was hoping someone else would speak up, but I guess not. =/

On Sat, Jan 21, 2017 at 5:50 AM, Mateusz via cfe-dev <[hidden email]> wrote:
Hello,

I'm involved with maintenance of Clang and related LLVM packages for MSYS2.
It was based on Martell's patches (with his permission for MSYS2) from his git repository and MSYS2 own fixes so number of patches might be overwhelming: https://github.com/Alexpux/MINGW-packages/tree/master/mingw-w64-clang
Maybe some of these patches could make it's way to the upstream.

Tools state at version 3.9.1 :
Clang:        mostly works (there are issues found)
LLMV:        mostly works (not tested well enough)
LLD:           creates broken executable
LLDB:         runs but not further tested since lldb-server is not build-able for mingw-w64
libc++:        need further testing
libc++abi:    need further testing
compiler-rt:  need further testing


Any help will be welcome.


make check seems to hang during LLVM tests but here is make check-clang:
Failing Tests (18):
    Clang :: Driver/mingw.cpp
    Clang :: Modules/dependency-dump.m
    Clang :: Modules/dependency-gen.modulemap
    Clang :: PCH/chain-categories.m
    Clang :: PCH/chain-categories2.m
    Clang :: PCH/chain-class-extension.m
    Clang :: PCH/chain-conversion-lookup.cpp
    Clang :: PCH/chain-cxx.cpp
    Clang :: PCH/chain-default-argument-instantiation.cpp
    Clang :: PCH/chain-empty-initial-namespace.cpp
    Clang :: PCH/chain-friend-instantiation.cpp
    Clang :: PCH/chain-implicit-definition.cpp
    Clang :: PCH/chain-late-anonymous-namespace.cpp
    Clang :: PCH/chain-openmp-threadprivate.cpp
    Clang :: PCH/chain-pending-instantiations.cpp
    Clang :: PCH/chain-staticvar-instantiation.cpp
    Clang :: PCH/cxx-chain-function-template.cpp
    Clang :: PCH/replaced-decl.m

  Expected Passes    : 9349
  Expected Failures  : 16
  Unsupported Tests  : 227
  Unexpected Failures: 18

Full output is available here: https://gist.github.com/mati865/b4a2c96f82b22ac41eca32714092db1d

Is there any interest from You in providing at lest minimal support to it?

Cheers,
Mateusz Mikuła


_______________________________________________
cfe-dev mailing list
[hidden email]
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-dev



_______________________________________________
cfe-dev mailing list
[hidden email]
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-dev
Reply | Threaded
Open this post in threaded view
|

Re: State of Clang in MSYS2

Eric Fiselier via cfe-dev
I briefly looked over them and they seem to be just a huge pile of
different unrelated changes, e.g. there is lldb patch, patch for
libc++ to allow compilation with gcc, some reverts of reverts, some
backports and so on.

Will it be possible to clean them up, rebase to 4.0 and submit to
Phabricator? Ideally as "one patch per topic".

On Wed, Feb 1, 2017 at 10:25 PM, Reid Kleckner via cfe-dev
<[hidden email]> wrote:

> I would volunteer to try to get these patches in, but I don't have the
> bandwidth for it. I was hoping someone else would speak up, but I guess not.
> =/
>
> On Sat, Jan 21, 2017 at 5:50 AM, Mateusz via cfe-dev
> <[hidden email]> wrote:
>>
>> Hello,
>>
>> I'm involved with maintenance of Clang and related LLVM packages for
>> MSYS2.
>> It was based on Martell's patches (with his permission for MSYS2) from his
>> git repository and MSYS2 own fixes so number of patches might be
>> overwhelming:
>> https://github.com/Alexpux/MINGW-packages/tree/master/mingw-w64-clang
>> Maybe some of these patches could make it's way to the upstream.
>>
>> Tools state at version 3.9.1 :
>> Clang:        mostly works (there are issues found)
>> LLMV:        mostly works (not tested well enough)
>> LLD:           creates broken executable
>> LLDB:         runs but not further tested since lldb-server is not
>> build-able for mingw-w64
>> libc++:        need further testing
>> libc++abi:    need further testing
>> compiler-rt:  need further testing
>>
>> Open clang issues (at MSYS2):
>> https://github.com/Alexpux/MINGW-packages/issues/1677
>> https://github.com/Alexpux/MINGW-packages/issues/1678 (not that much Clang
>> issue but rather mingw-64 with GCC)
>> https://github.com/Alexpux/MINGW-packages/issues/1834
>>
>> Any help will be welcome.
>>
>>
>> make check seems to hang during LLVM tests but here is make check-clang:
>> Failing Tests (18):
>>     Clang :: Driver/mingw.cpp
>>     Clang :: Modules/dependency-dump.m
>>     Clang :: Modules/dependency-gen.modulemap
>>     Clang :: PCH/chain-categories.m
>>     Clang :: PCH/chain-categories2.m
>>     Clang :: PCH/chain-class-extension.m
>>     Clang :: PCH/chain-conversion-lookup.cpp
>>     Clang :: PCH/chain-cxx.cpp
>>     Clang :: PCH/chain-default-argument-instantiation.cpp
>>     Clang :: PCH/chain-empty-initial-namespace.cpp
>>     Clang :: PCH/chain-friend-instantiation.cpp
>>     Clang :: PCH/chain-implicit-definition.cpp
>>     Clang :: PCH/chain-late-anonymous-namespace.cpp
>>     Clang :: PCH/chain-openmp-threadprivate.cpp
>>     Clang :: PCH/chain-pending-instantiations.cpp
>>     Clang :: PCH/chain-staticvar-instantiation.cpp
>>     Clang :: PCH/cxx-chain-function-template.cpp
>>     Clang :: PCH/replaced-decl.m
>>
>>   Expected Passes    : 9349
>>   Expected Failures  : 16
>>   Unsupported Tests  : 227
>>   Unexpected Failures: 18
>>
>> Full output is available here:
>> https://gist.github.com/mati865/b4a2c96f82b22ac41eca32714092db1d
>>
>> Is there any interest from You in providing at lest minimal support to it?
>>
>> Cheers,
>> Mateusz Mikuła
>>
>>
>> _______________________________________________
>> cfe-dev mailing list
>> [hidden email]
>> http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-dev
>>
>
>
> _______________________________________________
> cfe-dev mailing list
> [hidden email]
> http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-dev
>



--
With best regards, Anton Korobeynikov
Department of Statistical Modelling, Saint Petersburg State University
_______________________________________________
cfe-dev mailing list
[hidden email]
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-dev
Reply | Threaded
Open this post in threaded view
|

Re: State of Clang in MSYS2

Eric Fiselier via cfe-dev
LLVM patches start with 00xx and Clang patches start with 01xx.

First patch assigned to Reid: https://reviews.llvm.org/D29464
More patches will come soon

Patches are here: https://github.com/mati865/MINGW-packages/tree/clang-svn/mingw-w64-clang-svn


make check-clang:
Failing Tests (21):
Clang :: Driver/mingw.cpp
Clang :: Index/pr20320.cpp
Clang :: Modules/dependency-dump-dependent-module.m
Clang :: Modules/dependency-dump.m
Clang :: Modules/dependency-gen.modulemap
Clang :: PCH/chain-categories.m
Clang :: PCH/chain-categories2.m
Clang :: PCH/chain-class-extension.m
Clang :: PCH/chain-conversion-lookup.cpp
Clang :: PCH/chain-cxx.cpp
Clang :: PCH/chain-default-argument-instantiation.cpp
Clang :: PCH/chain-empty-initial-namespace.cpp
Clang :: PCH/chain-friend-instantiation.cpp
Clang :: PCH/chain-implicit-definition.cpp
Clang :: PCH/chain-late-anonymous-namespace.cpp
Clang :: PCH/chain-openmp-threadprivate.cpp
Clang :: PCH/chain-pending-instantiations.cpp
Clang :: PCH/chain-staticvar-instantiation.cpp
Clang :: PCH/cxx-chain-function-template.cpp
Clang :: PCH/debug-info-pch-path.c
Clang :: PCH/replaced-decl.m
Expected Passes : 10094
Expected Failures : 16
Unsupported Tests : 250
Unexpected Failures: 21

https://gist.github.com/mati865/3a350709f853e4a8cb1fc40980a4f5b0

2017-02-02 1:14 GMT+01:00 Anton Korobeynikov <[hidden email]>:
I briefly looked over them and they seem to be just a huge pile of
different unrelated changes, e.g. there is lldb patch, patch for
libc++ to allow compilation with gcc, some reverts of reverts, some
backports and so on.

Will it be possible to clean them up, rebase to 4.0 and submit to
Phabricator? Ideally as "one patch per topic".

On Wed, Feb 1, 2017 at 10:25 PM, Reid Kleckner via cfe-dev
<[hidden email]> wrote:
> I would volunteer to try to get these patches in, but I don't have the
> bandwidth for it. I was hoping someone else would speak up, but I guess not.
> =/
>
> On Sat, Jan 21, 2017 at 5:50 AM, Mateusz via cfe-dev
> <[hidden email]> wrote:
>>
>> Hello,
>>
>> I'm involved with maintenance of Clang and related LLVM packages for
>> MSYS2.
>> It was based on Martell's patches (with his permission for MSYS2) from his
>> git repository and MSYS2 own fixes so number of patches might be
>> overwhelming:
>> https://github.com/Alexpux/MINGW-packages/tree/master/mingw-w64-clang
>> Maybe some of these patches could make it's way to the upstream.
>>
>> Tools state at version 3.9.1 :
>> Clang:        mostly works (there are issues found)
>> LLMV:        mostly works (not tested well enough)
>> LLD:           creates broken executable
>> LLDB:         runs but not further tested since lldb-server is not
>> build-able for mingw-w64
>> libc++:        need further testing
>> libc++abi:    need further testing
>> compiler-rt:  need further testing
>>
>> Open clang issues (at MSYS2):
>> https://github.com/Alexpux/MINGW-packages/issues/1677
>> https://github.com/Alexpux/MINGW-packages/issues/1678 (not that much Clang
>> issue but rather mingw-64 with GCC)
>> https://github.com/Alexpux/MINGW-packages/issues/1834
>>
>> Any help will be welcome.
>>
>>
>> make check seems to hang during LLVM tests but here is make check-clang:
>> Failing Tests (18):
>>     Clang :: Driver/mingw.cpp
>>     Clang :: Modules/dependency-dump.m
>>     Clang :: Modules/dependency-gen.modulemap
>>     Clang :: PCH/chain-categories.m
>>     Clang :: PCH/chain-categories2.m
>>     Clang :: PCH/chain-class-extension.m
>>     Clang :: PCH/chain-conversion-lookup.cpp
>>     Clang :: PCH/chain-cxx.cpp
>>     Clang :: PCH/chain-default-argument-instantiation.cpp
>>     Clang :: PCH/chain-empty-initial-namespace.cpp
>>     Clang :: PCH/chain-friend-instantiation.cpp
>>     Clang :: PCH/chain-implicit-definition.cpp
>>     Clang :: PCH/chain-late-anonymous-namespace.cpp
>>     Clang :: PCH/chain-openmp-threadprivate.cpp
>>     Clang :: PCH/chain-pending-instantiations.cpp
>>     Clang :: PCH/chain-staticvar-instantiation.cpp
>>     Clang :: PCH/cxx-chain-function-template.cpp
>>     Clang :: PCH/replaced-decl.m
>>
>>   Expected Passes    : 9349
>>   Expected Failures  : 16
>>   Unsupported Tests  : 227
>>   Unexpected Failures: 18
>>
>> Full output is available here:
>> https://gist.github.com/mati865/b4a2c96f82b22ac41eca32714092db1d
>>
>> Is there any interest from You in providing at lest minimal support to it?
>>
>> Cheers,
>> Mateusz Mikuła
>>
>>
>> _______________________________________________
>> cfe-dev mailing list
>> [hidden email]
>> http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-dev
>>
>
>
> _______________________________________________
> cfe-dev mailing list
> [hidden email]
> http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-dev
>



--
With best regards, Anton Korobeynikov
Department of Statistical Modelling, Saint Petersburg State University


_______________________________________________
cfe-dev mailing list
[hidden email]
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-dev
Reply | Threaded
Open this post in threaded view
|

Re: State of Clang in MSYS2

Eric Fiselier via cfe-dev
In reply to this post by Eric Fiselier via cfe-dev

I don't have rights to commit, should I ask someone to do it with: https://reviews.llvm.org/D29464, https://reviews.llvm.org/D29772, https://reviews.llvm.org/D29952?

Also while browsing bugzilla I came into https://bugs.llvm.org//show_bug.cgi?id=30685.
I have developed workaround for MinGW https://github.com/mati865/MINGW-packages/blob/clang-svn/mingw-w64-clang-svn/0107-enable-__float128-for-X86-MinGW.patch
It tricks libstc++ to think Clang supports _float128 but __FLOAT128__ isn't defined so you can compile code with <type_traits> included but not with _float128 actually being used. I think it's better that way since using _float128 with mingw will seg fault Clang.


W dniu 01.02.2017 o 23:25, Reid Kleckner pisze:
I would volunteer to try to get these patches in, but I don't have the bandwidth for it. I was hoping someone else would speak up, but I guess not. =/

On Sat, Jan 21, 2017 at 5:50 AM, Mateusz via cfe-dev <[hidden email]> wrote:
Hello,

I'm involved with maintenance of Clang and related LLVM packages for MSYS2.
It was based on Martell's patches (with his permission for MSYS2) from his git repository and MSYS2 own fixes so number of patches might be overwhelming: https://github.com/Alexpux/MINGW-packages/tree/master/mingw-w64-clang
Maybe some of these patches could make it's way to the upstream.

Tools state at version 3.9.1 :
Clang:        mostly works (there are issues found)
LLMV:        mostly works (not tested well enough)
LLD:           creates broken executable
LLDB:         runs but not further tested since lldb-server is not build-able for mingw-w64
libc++:        need further testing
libc++abi:    need further testing
compiler-rt:  need further testing


Any help will be welcome.


make check seems to hang during LLVM tests but here is make check-clang:
Failing Tests (18):
    Clang :: Driver/mingw.cpp
    Clang :: Modules/dependency-dump.m
    Clang :: Modules/dependency-gen.modulemap
    Clang :: PCH/chain-categories.m
    Clang :: PCH/chain-categories2.m
    Clang :: PCH/chain-class-extension.m
    Clang :: PCH/chain-conversion-lookup.cpp
    Clang :: PCH/chain-cxx.cpp
    Clang :: PCH/chain-default-argument-instantiation.cpp
    Clang :: PCH/chain-empty-initial-namespace.cpp
    Clang :: PCH/chain-friend-instantiation.cpp
    Clang :: PCH/chain-implicit-definition.cpp
    Clang :: PCH/chain-late-anonymous-namespace.cpp
    Clang :: PCH/chain-openmp-threadprivate.cpp
    Clang :: PCH/chain-pending-instantiations.cpp
    Clang :: PCH/chain-staticvar-instantiation.cpp
    Clang :: PCH/cxx-chain-function-template.cpp
    Clang :: PCH/replaced-decl.m

  Expected Passes    : 9349
  Expected Failures  : 16
  Unsupported Tests  : 227
  Unexpected Failures: 18

Full output is available here: https://gist.github.com/mati865/b4a2c96f82b22ac41eca32714092db1d

Is there any interest from You in providing at lest minimal support to it?

Cheers,
Mateusz Mikuła


_______________________________________________
cfe-dev mailing list
[hidden email]
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-dev




_______________________________________________
cfe-dev mailing list
[hidden email]
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-dev

signature.asc (836 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: State of Clang in MSYS2

Eric Fiselier via cfe-dev

On Mar 5, 2017, at 11:25 AM, Mateusz Mikuła via cfe-dev <[hidden email]> wrote:

I don't have rights to commit, should I ask someone to do it with: https://reviews.llvm.org/D29464, https://reviews.llvm.org/D29772, https://reviews.llvm.org/D29952?

Just mention it in the review when it is accepter and the reviewer usually will commit it for you right after.

— 
Mehdi


Also while browsing bugzilla I came into https://bugs.llvm.org//show_bug.cgi?id=30685.
I have developed workaround for MinGW https://github.com/mati865/MINGW-packages/blob/clang-svn/mingw-w64-clang-svn/0107-enable-__float128-for-X86-MinGW.patch
It tricks libstc++ to think Clang supports _float128 but __FLOAT128__ isn't defined so you can compile code with <type_traits> included but not with _float128 actually being used. I think it's better that way since using _float128 with mingw will seg fault Clang.


W dniu 01.02.2017 o 23:25, Reid Kleckner pisze:
I would volunteer to try to get these patches in, but I don't have the bandwidth for it. I was hoping someone else would speak up, but I guess not. =/

On Sat, Jan 21, 2017 at 5:50 AM, Mateusz via cfe-dev <[hidden email]> wrote:
Hello,

I'm involved with maintenance of Clang and related LLVM packages for MSYS2.
It was based on Martell's patches (with his permission for MSYS2) from his git repository and MSYS2 own fixes so number of patches might be overwhelming: https://github.com/Alexpux/MINGW-packages/tree/master/mingw-w64-clang
Maybe some of these patches could make it's way to the upstream.

Tools state at version 3.9.1 :
Clang:        mostly works (there are issues found)
LLMV:        mostly works (not tested well enough)
LLD:           creates broken executable
LLDB:         runs but not further tested since lldb-server is not build-able for mingw-w64
libc++:        need further testing
libc++abi:    need further testing
compiler-rt:  need further testing


Any help will be welcome.


make check seems to hang during LLVM tests but here is make check-clang:
Failing Tests (18):
    Clang :: Driver/mingw.cpp
    Clang :: Modules/dependency-dump.m
    Clang :: Modules/dependency-gen.modulemap
    Clang :: PCH/chain-categories.m
    Clang :: PCH/chain-categories2.m
    Clang :: PCH/chain-class-extension.m
    Clang :: PCH/chain-conversion-lookup.cpp
    Clang :: PCH/chain-cxx.cpp
    Clang :: PCH/chain-default-argument-instantiation.cpp
    Clang :: PCH/chain-empty-initial-namespace.cpp
    Clang :: PCH/chain-friend-instantiation.cpp
    Clang :: PCH/chain-implicit-definition.cpp
    Clang :: PCH/chain-late-anonymous-namespace.cpp
    Clang :: PCH/chain-openmp-threadprivate.cpp
    Clang :: PCH/chain-pending-instantiations.cpp
    Clang :: PCH/chain-staticvar-instantiation.cpp
    Clang :: PCH/cxx-chain-function-template.cpp
    Clang :: PCH/replaced-decl.m

  Expected Passes    : 9349
  Expected Failures  : 16
  Unsupported Tests  : 227
  Unexpected Failures: 18

Full output is available here: https://gist.github.com/mati865/b4a2c96f82b22ac41eca32714092db1d

Is there any interest from You in providing at lest minimal support to it?

Cheers,
Mateusz Mikuła


_______________________________________________
cfe-dev mailing list
[hidden email]
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-dev



_______________________________________________
cfe-dev mailing list
[hidden email]
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-dev


_______________________________________________
cfe-dev mailing list
[hidden email]
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-dev
Reply | Threaded
Open this post in threaded view
|

Re: State of Clang in MSYS2

Eric Fiselier via cfe-dev
In reply to this post by Eric Fiselier via cfe-dev
I committed the two obvious ones. D29952 contains some discussion, so
I left it as-is.

On Sun, Mar 5, 2017 at 10:25 PM, Mateusz Mikuła <[hidden email]> wrote:

> I don't have rights to commit, should I ask someone to do it with:
> https://reviews.llvm.org/D29464, https://reviews.llvm.org/D29772,
> https://reviews.llvm.org/D29952?
>
> Also while browsing bugzilla I came into
> https://bugs.llvm.org//show_bug.cgi?id=30685.
> I have developed workaround for MinGW
> https://github.com/mati865/MINGW-packages/blob/clang-svn/mingw-w64-clang-svn/0107-enable-__float128-for-X86-MinGW.patch
> It tricks libstc++ to think Clang supports _float128 but __FLOAT128__ isn't
> defined so you can compile code with <type_traits> included but not with
> _float128 actually being used. I think it's better that way since using
> _float128 with mingw will seg fault Clang.
>
>
> W dniu 01.02.2017 o 23:25, Reid Kleckner pisze:
>
> I would volunteer to try to get these patches in, but I don't have the
> bandwidth for it. I was hoping someone else would speak up, but I guess not.
> =/
>
> On Sat, Jan 21, 2017 at 5:50 AM, Mateusz via cfe-dev
> <[hidden email]> wrote:
>>
>> Hello,
>>
>> I'm involved with maintenance of Clang and related LLVM packages for
>> MSYS2.
>> It was based on Martell's patches (with his permission for MSYS2) from his
>> git repository and MSYS2 own fixes so number of patches might be
>> overwhelming:
>> https://github.com/Alexpux/MINGW-packages/tree/master/mingw-w64-clang
>> Maybe some of these patches could make it's way to the upstream.
>>
>> Tools state at version 3.9.1 :
>> Clang:        mostly works (there are issues found)
>> LLMV:        mostly works (not tested well enough)
>> LLD:           creates broken executable
>> LLDB:         runs but not further tested since lldb-server is not
>> build-able for mingw-w64
>> libc++:        need further testing
>> libc++abi:    need further testing
>> compiler-rt:  need further testing
>>
>> Open clang issues (at MSYS2):
>> https://github.com/Alexpux/MINGW-packages/issues/1677
>> https://github.com/Alexpux/MINGW-packages/issues/1678 (not that much Clang
>> issue but rather mingw-64 with GCC)
>> https://github.com/Alexpux/MINGW-packages/issues/1834
>>
>> Any help will be welcome.
>>
>>
>> make check seems to hang during LLVM tests but here is make check-clang:
>> Failing Tests (18):
>>     Clang :: Driver/mingw.cpp
>>     Clang :: Modules/dependency-dump.m
>>     Clang :: Modules/dependency-gen.modulemap
>>     Clang :: PCH/chain-categories.m
>>     Clang :: PCH/chain-categories2.m
>>     Clang :: PCH/chain-class-extension.m
>>     Clang :: PCH/chain-conversion-lookup.cpp
>>     Clang :: PCH/chain-cxx.cpp
>>     Clang :: PCH/chain-default-argument-instantiation.cpp
>>     Clang :: PCH/chain-empty-initial-namespace.cpp
>>     Clang :: PCH/chain-friend-instantiation.cpp
>>     Clang :: PCH/chain-implicit-definition.cpp
>>     Clang :: PCH/chain-late-anonymous-namespace.cpp
>>     Clang :: PCH/chain-openmp-threadprivate.cpp
>>     Clang :: PCH/chain-pending-instantiations.cpp
>>     Clang :: PCH/chain-staticvar-instantiation.cpp
>>     Clang :: PCH/cxx-chain-function-template.cpp
>>     Clang :: PCH/replaced-decl.m
>>
>>   Expected Passes    : 9349
>>   Expected Failures  : 16
>>   Unsupported Tests  : 227
>>   Unexpected Failures: 18
>>
>> Full output is available here:
>> https://gist.github.com/mati865/b4a2c96f82b22ac41eca32714092db1d
>>
>> Is there any interest from You in providing at lest minimal support to it?
>>
>> Cheers,
>> Mateusz Mikuła
>>
>>
>> _______________________________________________
>> cfe-dev mailing list
>> [hidden email]
>> http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-dev
>>
>
>



--
With best regards, Anton Korobeynikov
Department of Statistical Modelling, Saint Petersburg State University
_______________________________________________
cfe-dev mailing list
[hidden email]
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-dev
Reply | Threaded
Open this post in threaded view
|

Re: State of Clang in MSYS2

Eric Fiselier via cfe-dev
Thank you Anton.

I was waiting for Reid response about D29952 so it's fine.


W dniu 06.03.2017 o 10:51, Anton Korobeynikov pisze:

> I committed the two obvious ones. D29952 contains some discussion, so
> I left it as-is.
>
> On Sun, Mar 5, 2017 at 10:25 PM, Mateusz Mikuła <[hidden email]> wrote:
>> I don't have rights to commit, should I ask someone to do it with:
>> https://reviews.llvm.org/D29464, https://reviews.llvm.org/D29772,
>> https://reviews.llvm.org/D29952?
>>
>> Also while browsing bugzilla I came into
>> https://bugs.llvm.org//show_bug.cgi?id=30685.
>> I have developed workaround for MinGW
>> https://github.com/mati865/MINGW-packages/blob/clang-svn/mingw-w64-clang-svn/0107-enable-__float128-for-X86-MinGW.patch
>> It tricks libstc++ to think Clang supports _float128 but __FLOAT128__ isn't
>> defined so you can compile code with <type_traits> included but not with
>> _float128 actually being used. I think it's better that way since using
>> _float128 with mingw will seg fault Clang.
>>
>>
>> W dniu 01.02.2017 o 23:25, Reid Kleckner pisze:
>>
>> I would volunteer to try to get these patches in, but I don't have the
>> bandwidth for it. I was hoping someone else would speak up, but I guess not.
>> =/
>>
>> On Sat, Jan 21, 2017 at 5:50 AM, Mateusz via cfe-dev
>> <[hidden email]> wrote:
>>> Hello,
>>>
>>> I'm involved with maintenance of Clang and related LLVM packages for
>>> MSYS2.
>>> It was based on Martell's patches (with his permission for MSYS2) from his
>>> git repository and MSYS2 own fixes so number of patches might be
>>> overwhelming:
>>> https://github.com/Alexpux/MINGW-packages/tree/master/mingw-w64-clang
>>> Maybe some of these patches could make it's way to the upstream.
>>>
>>> Tools state at version 3.9.1 :
>>> Clang:        mostly works (there are issues found)
>>> LLMV:        mostly works (not tested well enough)
>>> LLD:           creates broken executable
>>> LLDB:         runs but not further tested since lldb-server is not
>>> build-able for mingw-w64
>>> libc++:        need further testing
>>> libc++abi:    need further testing
>>> compiler-rt:  need further testing
>>>
>>> Open clang issues (at MSYS2):
>>> https://github.com/Alexpux/MINGW-packages/issues/1677
>>> https://github.com/Alexpux/MINGW-packages/issues/1678 (not that much Clang
>>> issue but rather mingw-64 with GCC)
>>> https://github.com/Alexpux/MINGW-packages/issues/1834
>>>
>>> Any help will be welcome.
>>>
>>>
>>> make check seems to hang during LLVM tests but here is make check-clang:
>>> Failing Tests (18):
>>>     Clang :: Driver/mingw.cpp
>>>     Clang :: Modules/dependency-dump.m
>>>     Clang :: Modules/dependency-gen.modulemap
>>>     Clang :: PCH/chain-categories.m
>>>     Clang :: PCH/chain-categories2.m
>>>     Clang :: PCH/chain-class-extension.m
>>>     Clang :: PCH/chain-conversion-lookup.cpp
>>>     Clang :: PCH/chain-cxx.cpp
>>>     Clang :: PCH/chain-default-argument-instantiation.cpp
>>>     Clang :: PCH/chain-empty-initial-namespace.cpp
>>>     Clang :: PCH/chain-friend-instantiation.cpp
>>>     Clang :: PCH/chain-implicit-definition.cpp
>>>     Clang :: PCH/chain-late-anonymous-namespace.cpp
>>>     Clang :: PCH/chain-openmp-threadprivate.cpp
>>>     Clang :: PCH/chain-pending-instantiations.cpp
>>>     Clang :: PCH/chain-staticvar-instantiation.cpp
>>>     Clang :: PCH/cxx-chain-function-template.cpp
>>>     Clang :: PCH/replaced-decl.m
>>>
>>>   Expected Passes    : 9349
>>>   Expected Failures  : 16
>>>   Unsupported Tests  : 227
>>>   Unexpected Failures: 18
>>>
>>> Full output is available here:
>>> https://gist.github.com/mati865/b4a2c96f82b22ac41eca32714092db1d
>>>
>>> Is there any interest from You in providing at lest minimal support to it?
>>>
>>> Cheers,
>>> Mateusz Mikuła
>>>
>>>
>>> _______________________________________________
>>> cfe-dev mailing list
>>> [hidden email]
>>> http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-dev
>>>
>>
>
>


_______________________________________________
cfe-dev mailing list
[hidden email]
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-dev

signature.asc (836 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: State of Clang in MSYS2

Eric Fiselier via cfe-dev
In reply to this post by Eric Fiselier via cfe-dev
Looks like there was issue with https://reviews.llvm.org/D29772 and
cfe/trunk/ prefix is missing.

Should I recreate patch and update this diff?


W dniu 06.03.2017 o 10:51, Anton Korobeynikov pisze:

> I committed the two obvious ones. D29952 contains some discussion, so
> I left it as-is.
>
> On Sun, Mar 5, 2017 at 10:25 PM, Mateusz Mikuła <[hidden email]> wrote:
>> I don't have rights to commit, should I ask someone to do it with:
>> https://reviews.llvm.org/D29464, https://reviews.llvm.org/D29772,
>> https://reviews.llvm.org/D29952?
>>
>> Also while browsing bugzilla I came into
>> https://bugs.llvm.org//show_bug.cgi?id=30685.
>> I have developed workaround for MinGW
>> https://github.com/mati865/MINGW-packages/blob/clang-svn/mingw-w64-clang-svn/0107-enable-__float128-for-X86-MinGW.patch
>> It tricks libstc++ to think Clang supports _float128 but __FLOAT128__ isn't
>> defined so you can compile code with <type_traits> included but not with
>> _float128 actually being used. I think it's better that way since using
>> _float128 with mingw will seg fault Clang.
>>
>>
>> W dniu 01.02.2017 o 23:25, Reid Kleckner pisze:
>>
>> I would volunteer to try to get these patches in, but I don't have the
>> bandwidth for it. I was hoping someone else would speak up, but I guess not.
>> =/
>>
>> On Sat, Jan 21, 2017 at 5:50 AM, Mateusz via cfe-dev
>> <[hidden email]> wrote:
>>> Hello,
>>>
>>> I'm involved with maintenance of Clang and related LLVM packages for
>>> MSYS2.
>>> It was based on Martell's patches (with his permission for MSYS2) from his
>>> git repository and MSYS2 own fixes so number of patches might be
>>> overwhelming:
>>> https://github.com/Alexpux/MINGW-packages/tree/master/mingw-w64-clang
>>> Maybe some of these patches could make it's way to the upstream.
>>>
>>> Tools state at version 3.9.1 :
>>> Clang:        mostly works (there are issues found)
>>> LLMV:        mostly works (not tested well enough)
>>> LLD:           creates broken executable
>>> LLDB:         runs but not further tested since lldb-server is not
>>> build-able for mingw-w64
>>> libc++:        need further testing
>>> libc++abi:    need further testing
>>> compiler-rt:  need further testing
>>>
>>> Open clang issues (at MSYS2):
>>> https://github.com/Alexpux/MINGW-packages/issues/1677
>>> https://github.com/Alexpux/MINGW-packages/issues/1678 (not that much Clang
>>> issue but rather mingw-64 with GCC)
>>> https://github.com/Alexpux/MINGW-packages/issues/1834
>>>
>>> Any help will be welcome.
>>>
>>>
>>> make check seems to hang during LLVM tests but here is make check-clang:
>>> Failing Tests (18):
>>>     Clang :: Driver/mingw.cpp
>>>     Clang :: Modules/dependency-dump.m
>>>     Clang :: Modules/dependency-gen.modulemap
>>>     Clang :: PCH/chain-categories.m
>>>     Clang :: PCH/chain-categories2.m
>>>     Clang :: PCH/chain-class-extension.m
>>>     Clang :: PCH/chain-conversion-lookup.cpp
>>>     Clang :: PCH/chain-cxx.cpp
>>>     Clang :: PCH/chain-default-argument-instantiation.cpp
>>>     Clang :: PCH/chain-empty-initial-namespace.cpp
>>>     Clang :: PCH/chain-friend-instantiation.cpp
>>>     Clang :: PCH/chain-implicit-definition.cpp
>>>     Clang :: PCH/chain-late-anonymous-namespace.cpp
>>>     Clang :: PCH/chain-openmp-threadprivate.cpp
>>>     Clang :: PCH/chain-pending-instantiations.cpp
>>>     Clang :: PCH/chain-staticvar-instantiation.cpp
>>>     Clang :: PCH/cxx-chain-function-template.cpp
>>>     Clang :: PCH/replaced-decl.m
>>>
>>>   Expected Passes    : 9349
>>>   Expected Failures  : 16
>>>   Unsupported Tests  : 227
>>>   Unexpected Failures: 18
>>>
>>> Full output is available here:
>>> https://gist.github.com/mati865/b4a2c96f82b22ac41eca32714092db1d
>>>
>>> Is there any interest from You in providing at lest minimal support to it?
>>>
>>> Cheers,
>>> Mateusz Mikuła
>>>
>>>
>>> _______________________________________________
>>> cfe-dev mailing list
>>> [hidden email]
>>> http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-dev
>>>
>>
>
>


_______________________________________________
cfe-dev mailing list
[hidden email]
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-dev

signature.asc (836 bytes) Download Attachment