Yaron Keren via llvm-dev
2017-Apr-21 15:45 UTC
[llvm-dev] r300964 and the llvm-clang-x86_64-expensive-checks-win buildbot
clean buld forced the regeneration of X86GenGlobalISel.inc & AArch64GenGlobalISel.inc, which may have not happened otherwise. בתאריך יום ו׳, 21 באפר׳ 2017 ב-18:29 מאת Simon Pilgrim via llvm-dev < llvm-dev at lists.llvm.org>:> For reasons I’m not sure about it needs a clean build every so often - > I’ve forced a build at > http://lab.llvm.org:8011/builders/llvm-clang-x86_64-expensive-checks-win/builds/1341 > and its gone green again. > > On 21 Apr 2017, at 15:17, Daniel Sanders via llvm-dev < > llvm-dev at lists.llvm.org> wrote: > > Hi, > > One of my commits earlier today (r300964) is causing > llvm-clang-x86_64-expensive-checks-win to fail with a compilation error ( > http://lab.llvm.org:8011/builders/llvm-clang-x86_64-expensive-checks-win/builds/1331) > and I'm having trouble figuring out why. I had seen something similar > before on MSVC so I tried fixing it in r300966 but with no luck. I then > managed to reproduce it using an online MSVC compiler and fixed that issue > in r300970 but still no luck. I'm going to revert the commit to make the > bot green while I figure it out but I don't have a MSVC environment at the > moment so I was wondering if anyone who does can reproduce the problem and > tell me what it is that MSVC doesn't like. > > Thanks > _______________________________________________ > LLVM Developers mailing list > llvm-dev at lists.llvm.org > http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev > > > > _______________________________________________ > LLVM Developers mailing list > llvm-dev at lists.llvm.org > http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev >-------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20170421/3148b072/attachment.html>
Daniel Sanders via llvm-dev
2017-Apr-21 15:52 UTC
[llvm-dev] r300964 and the llvm-clang-x86_64-expensive-checks-win buildbot
> On 21 Apr 2017, at 16:45, Yaron Keren <yaron.keren at gmail.com> wrote: > > clean buld forced the regeneration of X86GenGlobalISel.inc & AArch64GenGlobalISel.inc, which may have not happened otherwise.Thanks, I think that's the problem. I can see the tablegen commands in #1341 but they aren't in #1331 or #1340. I'll re-land this commit and we'll see if a clean build makes it work properly.> בתאריך יום ו׳, 21 באפר׳ 2017 ב-18:29 מאת Simon Pilgrim via llvm-dev <llvm-dev at lists.llvm.org <mailto:llvm-dev at lists.llvm.org>>: > For reasons I’m not sure about it needs a clean build every so often - I’ve forced a build at http://lab.llvm.org:8011/builders/llvm-clang-x86_64-expensive-checks-win/builds/1341 <http://lab.llvm.org:8011/builders/llvm-clang-x86_64-expensive-checks-win/builds/1341> and its gone green again.Thanks. My revert landed in http://lab.llvm.org:8011/builders/llvm-clang-x86_64-expensive-checks-win/builds/1340 <http://lab.llvm.org:8011/builders/llvm-clang-x86_64-expensive-checks-win/builds/1340> but it continued to fail so I was very confused. Is forcing a clean build something I can do from the buildbot interface? I've got a few patches to commit that are likely to trigger this problem.>> On 21 Apr 2017, at 15:17, Daniel Sanders via llvm-dev <llvm-dev at lists.llvm.org <mailto:llvm-dev at lists.llvm.org>> wrote: >> >> Hi, >> >> One of my commits earlier today (r300964) is causing llvm-clang-x86_64-expensive-checks-win to fail with a compilation error (http://lab.llvm.org:8011/builders/llvm-clang-x86_64-expensive-checks-win/builds/1331 <http://lab.llvm.org:8011/builders/llvm-clang-x86_64-expensive-checks-win/builds/1331>) and I'm having trouble figuring out why. I had seen something similar before on MSVC so I tried fixing it in r300966 but with no luck. I then managed to reproduce it using an online MSVC compiler and fixed that issue in r300970 but still no luck. I'm going to revert the commit to make the bot green while I figure it out but I don't have a MSVC environment at the moment so I was wondering if anyone who does can reproduce the problem and tell me what it is that MSVC doesn't like. >> >> Thanks >> _______________________________________________ >> LLVM Developers mailing list >> llvm-dev at lists.llvm.org <mailto:llvm-dev at lists.llvm.org> >> http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev <http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev> > > > _______________________________________________ > LLVM Developers mailing list > llvm-dev at lists.llvm.org <mailto:llvm-dev at lists.llvm.org> > http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev <http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev>-------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20170421/c32db632/attachment.html>
Daniel Sanders via llvm-dev
2017-Apr-21 16:13 UTC
[llvm-dev] r300964 and the llvm-clang-x86_64-expensive-checks-win buildbot
The re-commit is r300993.> On 21 Apr 2017, at 16:52, Daniel Sanders via llvm-dev <llvm-dev at lists.llvm.org> wrote: > >> >> On 21 Apr 2017, at 16:45, Yaron Keren <yaron.keren at gmail.com <mailto:yaron.keren at gmail.com>> wrote: >> >> clean buld forced the regeneration of X86GenGlobalISel.inc & AArch64GenGlobalISel.inc, which may have not happened otherwise. > > Thanks, I think that's the problem. I can see the tablegen commands in #1341 but they aren't in #1331 or #1340. > > I'll re-land this commit and we'll see if a clean build makes it work properly. > >> בתאריך יום ו׳, 21 באפר׳ 2017 ב-18:29 מאת Simon Pilgrim via llvm-dev <llvm-dev at lists.llvm.org <mailto:llvm-dev at lists.llvm.org>>: >> For reasons I’m not sure about it needs a clean build every so often - I’ve forced a build at http://lab.llvm.org:8011/builders/llvm-clang-x86_64-expensive-checks-win/builds/1341 <http://lab.llvm.org:8011/builders/llvm-clang-x86_64-expensive-checks-win/builds/1341> and its gone green again. > > Thanks. My revert landed in http://lab.llvm.org:8011/builders/llvm-clang-x86_64-expensive-checks-win/builds/1340 <http://lab.llvm.org:8011/builders/llvm-clang-x86_64-expensive-checks-win/builds/1340> but it continued to fail so I was very confused. > > Is forcing a clean build something I can do from the buildbot interface? I've got a few patches to commit that are likely to trigger this problem. > >>> On 21 Apr 2017, at 15:17, Daniel Sanders via llvm-dev <llvm-dev at lists.llvm.org <mailto:llvm-dev at lists.llvm.org>> wrote: >>> >>> Hi, >>> >>> One of my commits earlier today (r300964) is causing llvm-clang-x86_64-expensive-checks-win to fail with a compilation error (http://lab.llvm.org:8011/builders/llvm-clang-x86_64-expensive-checks-win/builds/1331 <http://lab.llvm.org:8011/builders/llvm-clang-x86_64-expensive-checks-win/builds/1331>) and I'm having trouble figuring out why. I had seen something similar before on MSVC so I tried fixing it in r300966 but with no luck. I then managed to reproduce it using an online MSVC compiler and fixed that issue in r300970 but still no luck. I'm going to revert the commit to make the bot green while I figure it out but I don't have a MSVC environment at the moment so I was wondering if anyone who does can reproduce the problem and tell me what it is that MSVC doesn't like. >>> >>> Thanks >>> _______________________________________________ >>> LLVM Developers mailing list >>> llvm-dev at lists.llvm.org <mailto:llvm-dev at lists.llvm.org> >>> http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev <http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev> >> >> >> _______________________________________________ >> LLVM Developers mailing list >> llvm-dev at lists.llvm.org <mailto:llvm-dev at lists.llvm.org> >> http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev <http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev> > > _______________________________________________ > LLVM Developers mailing list > llvm-dev at lists.llvm.org <mailto:llvm-dev at lists.llvm.org> > http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev <http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev>-------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20170421/a77c2fdf/attachment.html>
Galina Kistanova via llvm-dev
2017-Apr-21 17:56 UTC
[llvm-dev] r300964 and the llvm-clang-x86_64-expensive-checks-win buildbot
Hello Daniel,> Is forcing a clean build something I can do from the buildbot interface?I've got a few patches to commit that are likely to trigger this problem. There are 2 possible ways to do so: 1. You can force a clean build from the buildbot WebUI by specifying a property 'clean' with the value 'True' (http://lab.llvm.org:8011/buil ders/llvm-clang-x86_64-expensive-checks-win, "Property 1 Name:" filed is 'clean' and the corresponding "Value:" field is 'True', then click the "Force Build" button.) 2. If you know in advance that your commit requires a clean build, you may add something like "Requires a clean build." somewhere in the commit message. That would trigger clean builds. Not all builders support this, though. The llvm-clang-x86_64-expensive-checks-win buildbot does. I'm in the middle of expanding these features to the rest of the builders. As of the original issue, it seems those inc files are missing dependencies. I'll look into this. Thanks Galina On Fri, Apr 21, 2017 at 8:52 AM, Daniel Sanders via llvm-dev < llvm-dev at lists.llvm.org> wrote:> > On 21 Apr 2017, at 16:45, Yaron Keren <yaron.keren at gmail.com> wrote: > > clean buld forced the regeneration of X86GenGlobalISel.inc & > AArch64GenGlobalISel.inc, which may have not happened otherwise. > > > Thanks, I think that's the problem. I can see the tablegen commands in > #1341 but they aren't in #1331 or #1340. > > I'll re-land this commit and we'll see if a clean build makes it work > properly. > > בתאריך יום ו׳, 21 באפר׳ 2017 ב-18:29 מאת Simon Pilgrim via llvm-dev < > llvm-dev at lists.llvm.org>: > >> For reasons I’m not sure about it needs a clean build every so often - >> I’ve forced a build at http://lab.llvm.org:8011/ >> builders/llvm-clang-x86_64-expensive-checks-win/builds/1341 and its gone >> green again. >> > > Thanks. My revert landed in http://lab.llvm.org:8011/ > builders/llvm-clang-x86_64-expensive-checks-win/builds/1340 but it > continued to fail so I was very confused. > > Is forcing a clean build something I can do from the buildbot interface? > I've got a few patches to commit that are likely to trigger this problem. > > On 21 Apr 2017, at 15:17, Daniel Sanders via llvm-dev < >> llvm-dev at lists.llvm.org> wrote: >> >> Hi, >> >> One of my commits earlier today (r300964) is causing >> llvm-clang-x86_64-expensive-checks-win to fail with a compilation error ( >> http://lab.llvm.org:8011/builders/llvm-clang-x86_64- >> expensive-checks-win/builds/1331) and I'm having trouble figuring out >> why. I had seen something similar before on MSVC so I tried fixing it in >> r300966 but with no luck. I then managed to reproduce it using an online >> MSVC compiler and fixed that issue in r300970 but still no luck. I'm going >> to revert the commit to make the bot green while I figure it out but I >> don't have a MSVC environment at the moment so I was wondering if anyone >> who does can reproduce the problem and tell me what it is that MSVC doesn't >> like. >> >> Thanks >> _______________________________________________ >> LLVM Developers mailing list >> llvm-dev at lists.llvm.org >> http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev >> >> >> >> _______________________________________________ >> LLVM Developers mailing list >> llvm-dev at lists.llvm.org >> http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev >> > > > _______________________________________________ > LLVM Developers mailing list > llvm-dev at lists.llvm.org > http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev > >-------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20170421/3731b970/attachment.html>