similar to: [LLVMdev] RFC: Enable Lit testing on Windows buildbot

Hello, I am going to add this step in the nearest time. Thanks Galina 2011/4/08 NAKAMURA Takumi <geek4civic at gmail.com>: >Hello, > >Shall we enable "make check" on msys-mingw hosts? >(I suggest llvm-gcc-native-mingw32-win7 could do) > >(it is llvm+clang's "make check-all") > >Testing Time: 146.64s >Expected Passes : 8297

2015-05-03 21:02 GMT+09:00 Yaron Keren <yaron.keren at gmail.com>: > I now tried the extra tests and get the same failures: > > Clang Tools :: clang-tidy/clang-tidy-diff.cpp > Clang Tools :: clang-tidy/file-filter.cpp > > Takumi, I see you added requires:shell in r211831 and r220837. > > When using MSYS requires:shell is true but test fail, is it OK to >

Chen, see http://llvm.org/docs/GettingStarted.html#pf_12 ...Takumi ps. Excuse me, PE+ (aka pep) means "Executable file format for WIndows x64". 2011/5/16 陈晓宇 <xychen0921 at gmail.com>: > The stack trace: > > Starting program: > C:\MinGW\msys\1.0\home\xchen\llvm-obj\lib\Target\CellSPU/../.. > /../Debug/bin/tblgen.exe -I

On Sat, Jun 16, 2012 at 3:29 AM, Mikael Lyngvig <mikael at lyngvig.org> wrote: > Hi Takumi, > > THANK YOU for your excellent review! It makes all the work worthwhile in > itself. > > It is getting "late" here (it is early morning) so I am about to call it a > day. But I will contemplate and incorporate your suggestions tomorrow, and > send you a full

2012/6/16 Justin Holewinski <justin.holewinski at gmail.com>: > Under Python installation, you mention that Python is only needed for the > test suite.  This is not true, as the normal LLVM build still requires > Python. That's right. Now, python>=2.4 is essential. (In contrast, perl is not required, IIRC) > I would seriously consider splitting this up into

Good evening, Yuan Excuse me, I gave up Debug build earlier due to my poor hosts. I will improve able to build Debug ;) You may build with --enable-optimized=yes (or, make ENABLE_OPTIMIZED=1) btw, DLL builder has implemented since Aug. It might be easier to port DLL stuff to 2.6 :) Do you try? ...Takumi 2010年9月2日 17:16:25 UTC+9 yuan zheng <tsinghuayuan86 at gmail.com>: > Hello,

Hi Takumi, THANK YOU for your excellent review! It makes all the work worthwhile in itself. It is getting "late" here (it is early morning) so I am about to call it a day. But I will contemplate and incorporate your suggestions tomorrow, and send you a full reply. I think I originally wrote a document where I addressed both Visual Studio and MinGW, but then decided that there were

Now I can't make CMake use "MSYS Makefiles". I get an error about GNU v4.7.0 being broken as it can't build a simple test program. Using "MinGW Makefiles", everything works out of the box. And I am trying to give the user the most Windowsy experience (to avoid Cygwin, GnuWin32, and MSYS, if I can). So I think I'll stick to my current angle - which is to let the

Good morning David, I won't accept your patches for now. It breaks lots of things in current environments, autoconf - msys, cmake - msys, cmake -mingw and cmake - msvs. - You should never forget other environments. (eg., MSYS and CMD.EXE) You should implement detecting MKS if you tweak MKS-specific stuff. - Bash should be assumed to be mature. I suppose MKS bash might be

Hello, NAKAMURA Takumi As you said, I check out the head from svn trunk. I build the source code as the following steps: $./configure --enable-shared $ make After 1 hour, the building procedure is stopped and appear the following error message: c:/strawberry/c/bin/../lib/gcc/i686-w64-mingw32/4.4.3/../../../../i686-w64-mingw 32/bin/ld.exe:

As for Ninja, I'd like to wait until it is officially part of the released CMake. As I understand it, Windows support has been checked into the CMake repository but no offical release supporting Ninja has been made yet. I just retrieved Ninja from Github and this time it built without problems. 2012/6/16 NAKAMURA Takumi <geek4civic at gmail.com> > Mikael, > > Thanks for

Oh, didn't realize that there already is a semi-official build of CMake with Ninja support. I am trying it out right now and it seems very promising. So I'll change the document to use Ninja instead of "MinGW Makefiles". No need to waste energy on MinGW and MSYS Makefiles when Ninja outdoes them both. I thought people would have to build CMake themselves, something I'd

Morning, Will! 2012/7/12 Will Schmidt <will_schmidt at vnet.ibm.com>: > llvm]$ gmake check > llvm[0]: Running test suite > gmake[1]: Entering directory `/home/willschm/llvm/test' > Making LLVM 'lit.site.cfg' file... > sed: file lit.tmp line 8: unknown option to `s' > gmake[1]: *** [lit.site.cfg] Error 1 > > The relevant lines in test/Makefile: >

2010/9/2 NAKAMURA Takumi <geek4civic at gmail.com>: > Good evening, Yuan > > Excuse me, I gave up Debug build earlier due to my poor hosts. > I will improve able to build Debug ;) > > You may build with --enable-optimized=yes > (or, make ENABLE_OPTIMIZED=1) > > btw, DLL builder has implemented since Aug. > It might be easier to port DLL stuff to 2.6 :) Do you

Mikael, Thanks for working on. My gross comments... - Canonical URL of LLVM project is; http://llvm.org/ , not www.llvm.org. - "test-suite and cygwin" could be split out. It would be special thing. - Memory/storage requirements would be moved to "A.B Debugging". Less memory would be enough without debug build. - Python (x64 binary) is available, too. I am using. Python

As for memory/storage, I think you want these presented as early as possible so people don't go setting up a box only to find out, half a week later, that they can never do what they intended to do (build Clang in debug mode). The reason I use 32-bit Python is that it is what you need, if you want to make a Windows buildbot slave because the Win32 bindings are needed and I don't think

2010/11/8 Brice Lin <brice.lin at gmail.com>: > $ ~/llvm/Debug+Asserts/bin/llvm-lit ~/llvm/test/Integer/BitCast.ll > llvm-lit: lit.cfg:103: fatal: No site specific configuration available! Have you ever run "make check" and passed? ...Takumi

Good midnight, Kevin. cygming/shared might have been broken since r112976, IIRC. (ToT, too) A trivial patch; --- a/Makefile.rules +++ b/Makefile.rules @@ -942,6 +942,11 @@ ifdef EXPORTED_SYMBOL_FILE # First, set up the native export file, which may differ from the source # export file. +# The option --version-script is not effective on GNU ld win32. +ifneq (,$(filter $(HOST_OS),Cygwin

Ah, ...no. please don't hesitate to write logs ;) It must be the subject in the buildbot, not in developer guys. Galina, I'll write up later, how to tweak this issue. It is not resolved in released versions, though. The easiest way is to tweak svnpoller. 2013/11/16 Daniel Sanders <Daniel.Sanders at imgtec.com>: > That would explain it. The commit messages were quite long

That would explain it. The commit messages were quite long (especially the second one) and a commit with a smaller message has shown up. Thanks for letting me know. I'll try to keep them a bit more reasonable in future. > -----Original Message----- > From: NAKAMURA Takumi [mailto:geek4civic at gmail.com] > Sent: 15 November 2013 16:28 > To: Daniel Sanders > Cc: LLVM Developers