03:36 <GitHub131> [artiq] sbourdeauducq commented on issue #790: Funded by Oxford. https://github.com/m-labs/artiq/issues/790#issuecomment-333351018

03:49 <sb0> _florent_, have you been able to run artiq on the sayma amc+rtm system?

03:50 <sb0> whitequark, so what is wrong exactly with copying the numba code? it doesn't seem particularly awful to me, and that would solve the problem quickly.

03:51 <sb0> well I haven't looked at the floating point version...

04:52 <sb0> ah I had missed the sawg changes when merging sinara... sorry about that

04:54 <GitHub144> [artiq] sbourdeauducq pushed 1 new commit to master: https://github.com/m-labs/artiq/commit/a87031f8c3561924d4a31bc9e759d0fa53ae6fff

04:54 <GitHub144> artiq/master a87031f Sebastien Bourdeauducq: RELEASING: reminder to close GH milestones

05:19 <bb-m-labs_> build #808 of artiq-board is complete: Success [build successful] Build details are at http://buildbot.m-labs.hk/builders/artiq-board/builds/808

05:30 <bb-m-labs_> build #579 of artiq-win64-test is complete: Success [build successful] Build details are at http://buildbot.m-labs.hk/builders/artiq-win64-test/builds/579

05:30 <bb-m-labs_> build #1698 of artiq is complete: Success [build successful] Build details are at http://buildbot.m-labs.hk/builders/artiq/builds/1698

05:38 <sb0> hmm github doesn't have a way of denying the deletion of branches via the footgun that is git push --delete ...

06:15 <sb0> it would be nice if push --delete could be completely blocked, and branches could only be deleted via the github interface that would warn you if they are not fully merged

06:19 <sb0> anyway, short of having it blocked, we can simply refrain from using it

08:58 <larsc> sb0: if the branch is marked as protected it can't be deleted

08:58 <sb0> yes, but I'd want to automatically protect all branches

09:30 <rjo> sb0: np. i wasn't sure whether i even pushed that.

11:05 <_florent_> sb0: haven't tested artiq on sayma yet, I'll work on it next week

14:15 <whitequark> sb0: ok, I finally figured out how to do it nicely

14:15 <whitequark> which is actually what numba does, it just wasn't obvious

14:15 <whitequark> they emit an intrinsic as LLVM IR

14:16 <whitequark> so this avoids code duplication (it's emitted no more than once) but also gives LLVM visibility into it

14:18 <GitHub26> [artiq] whitequark deleted now-as-ssa at 1b9236f: https://github.com/m-labs/artiq/commit/1b9236f

18:33 <GitHub151> [artiq] whitequark pushed 1 new commit to master: https://github.com/m-labs/artiq/commit/62f2693e36cc30c4088c9e76c2ecb52fe0d38f06

18:33 <GitHub151> artiq/master 62f2693 whitequark: compiler: correct semantics of integer % operator (#830).

18:43 <whitequark> rjo: sb0: what do we do with "signed zeroes"?

18:44 <whitequark> i currently do not make any effort to support signed zeroes and nans anywhere in artiq python

18:44 <whitequark> so in modulo code as well

18:58 <bb-m-labs_> build #809 of artiq-board is complete: Success [build successful] Build details are at http://buildbot.m-labs.hk/builders/artiq-board/builds/809

19:09 <GitHub169> [artiq] whitequark pushed 2 new commits to master: https://github.com/m-labs/artiq/compare/62f2693e36cc...843786a0919e

19:09 <GitHub169> artiq/master 843786a whitequark: compiler: minor intrinsic refactoring.

19:09 <GitHub169> artiq/master 491c7ef whitequark: compiler: correct semantics of floating point % operator (fix #830).

19:09 <GitHub8> [artiq] whitequark closed issue #830: Modulo arithmetic with negative numbers on the kernel https://github.com/m-labs/artiq/issues/830

19:09 <bb-m-labs_> build #580 of artiq-win64-test is complete: Success [build successful] Build details are at http://buildbot.m-labs.hk/builders/artiq-win64-test/builds/580

19:09 <bb-m-labs_> build #1699 of artiq is complete: Success [build successful] Build details are at http://buildbot.m-labs.hk/builders/artiq/builds/1699

19:33 <bb-m-labs_> build #810 of artiq-board is complete: Success [build successful] Build details are at http://buildbot.m-labs.hk/builders/artiq-board/builds/810

19:44 <bb-m-labs_> build #581 of artiq-win64-test is complete: Success [build successful] Build details are at http://buildbot.m-labs.hk/builders/artiq-win64-test/builds/581

19:44 <bb-m-labs_> build #1700 of artiq is complete: Success [build successful] Build details are at http://buildbot.m-labs.hk/builders/artiq/builds/1700

19:44 <GitHub33> [buildbot-config] whitequark pushed 1 new commit to master: https://git.io/vdnWA

19:44 <GitHub33> buildbot-config/master a9c9650 whitequark: Set meaningful context for GitHub status reporter.

20:26 <GitHub29> [buildbot-config] whitequark pushed 1 new commit to master: https://git.io/vdn8y

20:26 <GitHub29> buildbot-config/master f70218a whitequark: Revert "Set meaningful context for GitHub status reporter."...

20:29 <GitHub130> [artiq] whitequark pushed 1 new commit to master: https://github.com/m-labs/artiq/commit/f86744c65cc3e84605d7fa25eae815fc0f9d523c

20:29 <GitHub130> artiq/master f86744c whitequark: compiler: implement ~ operator (fix #836).

20:29 <GitHub179> [artiq] whitequark closed issue #836: Compiler crash using unary operator https://github.com/m-labs/artiq/issues/836

20:32 <whitequark> rjo: can you please put your access token for github back into buildbot's secrets.json file?

20:32 <whitequark> i wanted to make github status reporting more clear but broke it instead

20:53 <bb-m-labs> build #811 of artiq-board is complete: Success [build successful] Build details are at http://buildbot.m-labs.hk/builders/artiq-board/builds/811

21:07 <rjo> whitequark: status_api_token?

21:07 <whitequark> rjo: yes

21:08 <rjo> whitequark: done. had to regenerate it but afaict that's just fine.

21:08 <whitequark> I made an OAuth app so it would appear under m-labs' account but that didn't work

21:08 <whitequark> not sure why and I spent too much time on this silliness already

21:08 <rjo> whitequark: note that i deleted what was in there before.

21:08 <whitequark> yeah sure

21:08 <rjo> ok.