00:56
Mortir has quit [Ping timeout: 240 seconds]
02:35
lritter has joined #pypy
02:37
lritter has quit [Client Quit]
02:37
lritter has joined #pypy
02:46
jcea has quit [Ping timeout: 260 seconds]
02:49
lritter has quit [Quit: Leaving]
02:50
lritter has joined #pypy
03:30
jacob22_ has quit [Read error: Connection reset by peer]
03:34
jacob22_ has joined #pypy
03:38
oberstet has quit [Quit: Leaving]
04:34
lritter has quit [Ping timeout: 272 seconds]
04:35
lritter has joined #pypy
05:19
lritter has quit [Quit: Leaving]
06:40
_whitelogger has joined #pypy
07:37
_whitelogger has joined #pypy
08:04
<
mattip >
cfbloz: what did you see that looks bad?
08:19
<
cfbolz >
mattip: I just investigated the failing pypyjit tests
08:19
<
cfbolz >
Calls with keywords at the call site are slow
08:19
<
cfbolz >
And I don't know how to fix it
08:34
_whitelogger has joined #pypy
09:24
<
mattip >
ahh, these failures are not new to the branch, they are in py3.7
09:50
<
cfbolz >
Sorry, that was independent of the merge
09:51
<
cfbolz >
But it's a general 3.7 problem
11:38
<
cfbolz >
ah, I had the same problem in a different context already :-/
11:49
oberstet has joined #pypy
12:39
Mortir has joined #pypy
14:15
jacob22_ has quit [Quit: Konversation terminated!]
14:17
jacob22 has joined #pypy
14:27
<
arigato >
_opcode.stack_effect() can probably be implemented from pypy.interpreter.astcompiler.assemble._opcode_stack_effect()
14:27
<
arigato >
though as usual it would take quite some tweaks
14:28
<
arigato >
for now it's fine (I think?) if we ignore the problem, at least now it's a new failing test (just one)
15:41
<
mattip >
arigato: thanks for the tip, I added it to the 3.7 wiki page
15:57
<
arigato >
(fwiw it's also missing from py3.6)
16:04
lritter has joined #pypy
16:06
<
mattip >
if no-one has complained since the first release of 3.6 in Feb 2019, I don't think it is all that important
16:06
<
mattip >
personally I don't want to keep supporting 3.6, as I wrote to pypy-dev, except for obvious bugs and security updates
16:10
<
arigato >
yes, agreed
16:10
<
arigato >
we should have only one 3.x version maintained at a time, and I'm +1 to switch to the next 3.x as soon as the branch works well enough
17:16
dnshane has joined #pypy
18:25
<
antocuni >
+1 for supporting only 1 version
19:42
otisolsen70 has joined #pypy
22:21
<
mattip >
arigato: there is a new failing test on py3.7 windows own tests after py3.7-rsre
22:21
<
mattip >
test: pypy/module/_sre/test/test_app_sre AppTestGetlower/()/test_getlower_locale
22:22
Mortir has quit [Quit: leaving]