01:00 <bbot2> Started: http://buildbot.pypy.org/builders/pypy-c-jit-linux-aarch64/builds/1052 [py3.7]

01:00 <bbot2> Started: http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-32/builds/6436 [py3.7]

01:00 <bbot2> Started: http://buildbot.pypy.org/builders/pypy-c-jit-linux-s390x/builds/1549 [py3.7]

01:00 <bbot2> Started: http://buildbot.pypy.org/builders/pypy-c-jit-macosx-x86-64/builds/5630 [py3.7]

01:00 <bbot2> Started: http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-64/builds/7759 [py3.7]

01:00 <bbot2> Started: http://buildbot.pypy.org/builders/pypy-c-jit-win-x86-64/builds/306 [py3.7]

01:00 <bbot2> Started: http://buildbot.pypy.org/builders/own-win-x86-64/builds/360 [py3.7]

01:00 <bbot2> Started: http://buildbot.pypy.org/builders/own-linux-x86-64/builds/8765 [py3.7]

01:00 <bbot2> Started: http://buildbot.pypy.org/builders/own-linux-aarch64/builds/957 [py3.7]

01:00 <bbot2> Started: http://buildbot.pypy.org/builders/own-linux-x86-32/builds/7582 [py3.7]

01:59 <bbot2> Failure: http://buildbot.pypy.org/builders/own-linux-x86-64/builds/8765 [py3.7]

02:01 <bbot2> Failure: http://buildbot.pypy.org/builders/own-linux-x86-32/builds/7582 [py3.7]

02:09 <bbot2> Failure: http://buildbot.pypy.org/builders/pypy-c-jit-linux-s390x/builds/1549 [py3.7]

02:14 <bbot2> Failure: http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-64/builds/7759 [py3.7]

02:24 <bbot2> Failure: http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-32/builds/6436 [py3.7]

02:41 <bbot2> Failure: http://buildbot.pypy.org/builders/pypy-c-jit-win-x86-64/builds/306 [py3.7]

02:47 <kor1> the faq mentions that cython type annotations could be used for optimizations but the exact benefits are unclear (as of 2015). are things more clear now?

02:54 <bbot2> Failure: http://buildbot.pypy.org/builders/pypy-c-jit-macosx-x86-64/builds/5630 [py3.7]

03:51 <bbot2> Failure: http://buildbot.pypy.org/builders/own-win-x86-64/builds/360 [py3.7]

03:55 <bbot2> Failure: http://buildbot.pypy.org/builders/pypy-c-jit-linux-aarch64/builds/1052 [py3.7]

06:22 <bbot2> Failure: http://buildbot.pypy.org/builders/own-linux-aarch64/builds/957 [py3.7]

09:00 <bbot2> Started: http://buildbot.pypy.org/builders/jit-benchmark-linux-x86-64/builds/3331 [py3.7]

10:29 <bbot2> Success: http://buildbot.pypy.org/builders/jit-benchmark-linux-x86-64/builds/3331 [py3.7]

11:16 <smarr> when using `jit.promote` on a variable that is updated frequently, let's say I have an array I use as a stack, and have a stack pointer that is fixed for each specific bytecode executing in a sequence, do I need to call promote after every modification? Or perhaps always at the beginning of the bytecode loop, or just once at the beginning of a function?

11:17 <smarr> I guess, if I need to call it after each modification, would I gain anything from promoting also `stack_ptr + 1` and `stack_ptr + 2` to extract arguments of a function call from the stack?

14:05 <cfbolz> smarr: at some place at the beginning of the loop should be enough

15:09 <smarr> cfbolz: thanks, indeed, results are basically identical

15:16 <cfbolz> smarr: did you fix the recursive call problem?

15:36 <smarr> cfbolz: no, that's in the AST interpreter, haven't yet understood why the tracing is different