00:08
speeder39_ has joined #pypy
00:17
lritter has quit [Quit: Leaving]
00:54
jcea has quit [Quit: jcea]
02:26
jcea has joined #pypy
02:38
speeder39_ has quit [Quit: Connection closed for inactivity]
02:47
jcea has quit [Quit: jcea]
03:33
dansan has quit [Ping timeout: 258 seconds]
04:48
jacob22 has quit [Read error: Connection reset by peer]
04:51
jacob22 has joined #pypy
05:41
_whitelogger has joined #pypy
08:17
Alex_Gaynor has quit [Read error: Connection reset by peer]
08:17
Olorin has quit [Read error: Connection reset by peer]
08:17
ctismer has quit [Read error: Connection reset by peer]
08:17
fijal has quit [Read error: Connection reset by peer]
08:17
Guest68750 has quit [Read error: Connection reset by peer]
08:18
ctismer has joined #pypy
08:19
Guest68750 has joined #pypy
08:19
<
mattip >
about 26691d5fd920: post_include_bits will only be used with separate_module_sources,
08:20
<
mattip >
when using seperate_module_files the post_include_bits are thrown away
08:20
Alex_Gaynor has joined #pypy
08:20
<
mattip >
well, since no C source is generated they are ignored
08:22
fijal has joined #pypy
08:23
Olorin has joined #pypy
08:24
wallet42_ has quit [Ping timeout: 244 seconds]
08:25
altendky has quit [Read error: Connection reset by peer]
08:32
altendky has joined #pypy
08:32
wallet42_ has joined #pypy
08:59
jacob22 has quit [Read error: Connection reset by peer]
09:01
otisolsen70 has joined #pypy
09:03
jacob22 has joined #pypy
09:04
<
arigato >
mattip: RPYTHON_JITLOG appears not used at all. I guess it's a leftover
09:05
<
arigato >
but note that "if not we_are_translated():" at global module level will always pass
09:05
<
arigato >
i.e. it's not testing anything, written like that
09:05
<
mattip >
hmm, that's less good
09:06
<
arigato >
also, "#define JITLOG xxx" has no effect on a string "JITLOG" used in C
09:07
<
arigato >
you need to write in C something like #ifdef JITLOG ... #else ... #endif
09:08
<
arigato >
I suggest that you edit rpython.translator.goal.translate
09:08
<
arigato >
so that this file replaces things around to give the end result ``getenv("JITLOG")`` in C
09:09
<
arigato >
and by default this is ``getenv("JITLOG_FORTESTING")``
09:10
<
arigato >
you can do that in a way similar to RPY_REVERSE_DEBUGGER in translator/c/genc.py
09:10
<
arigato >
which is an env var set explicitly there
09:11
<
mattip >
arigato: feel like taking it over?
09:14
<
arigato >
ah I see you did the macro thing
09:14
<
arigato >
with Py_STR
09:15
<
arigato >
and also, I was confused
09:15
<
arigato >
and still am
09:15
<
arigato >
there is PYPYLOG and also JITLOG
09:16
<
arigato >
two different things I guess
09:16
<
arigato >
I thought this was about PYPYLOG, but probably they both have the same problem
09:19
<
arigato >
does it make sense to apply the same solution to both? i.e. replace PYPYLOG with PYPYLOG_FORTESTS?
09:32
<
arigato >
mattip: done for JITLOG
09:32
<
arigato >
PYPYLOG would be similar, if it is needed
09:33
<
mattip >
thanks again. Does PYPYLOG have tests that are failing?
09:34
<
arigato >
I don't know the status
09:35
<
arigato >
right, but which revisions are run with pypy instead of cpython here?
09:35
<
mattip >
the last three own tests
09:36
<
arigato >
no need to fix a non-problem, I guess. let's only fix PYPYLOG if we see it's an issue somewhere
09:36
<
mattip >
right, that is why I am doing this very slowly
09:37
<
mattip >
in the first link, I don't understand the pypy.interpreter.test.test_app_main failures
09:37
<
mattip >
they passs locally for me
09:38
<
mattip >
I seem to remember this "TypeError: __init__() takes exactly 1 argument (2 given)" from another issue,
09:39
<
mattip >
but why is it showing up in the tests?
09:40
<
mattip >
is some state cached somewhere that should be cleared?
09:41
<
arigato >
I agree that it's very strange
09:41
<
arigato >
I can do type(sys.flags)([0]*16) in pypy on linux
09:43
<
arigato >
works on windows too
09:47
<
mattip >
adding config.translation.rpython_translate may be helpful in other places in this transition too
09:49
<
arigato >
the error doesn't make sense, but maybe would if we're accidentally calling CPython here and it's a slightly older version of CPython 2.7.x where the error message for calling type(sys.flags)() was different than what it is now
09:49
<
arigato >
or maybe it's an old version of pypy2, for all I know
10:17
otisolsen70_ has joined #pypy
10:21
otisolsen70 has quit [Ping timeout: 240 seconds]
10:25
Taggnostr has quit [Ping timeout: 240 seconds]
10:27
Taggnostr has joined #pypy
10:28
otisolsen70__ has joined #pypy
10:31
otisolsen70_ has quit [Ping timeout: 240 seconds]
11:00
dansan has joined #pypy
11:03
otisolsen70_ has joined #pypy
11:07
otisolsen70__ has quit [Ping timeout: 258 seconds]
11:30
lritter has joined #pypy
11:49
otisolsen70_ has quit [Quit: Leaving]
13:53
_whitelogger has joined #pypy
14:16
todda7 has joined #pypy
14:40
<
mattip >
it seems sys.flags is being translated from site.Flags, but where does site.Flags come from?
14:50
<
mattip >
ahh, it is something to do with the virtualenv python
14:50
<
mattip >
which is pypy, but not quite correctly
14:51
<
mattip >
(own tests run in a virtualenv, see the setup steps before the actual tests)
14:52
jcea has joined #pypy
15:09
<
mattip >
running in a pypy virtualenv fails the same on linux
15:21
<
mattip >
so the problem is that app_main.py has its own we_are_translated(), and it is too blunt a tool:
15:21
<
mattip >
it looks for sys.pypy_translation_info, which any pypy will have
15:24
<
arigato >
time to add 'sys.pypy_untranslated' maybe, only on untranslated pypy's?
15:32
<
mattip >
I think the fix for this one is easier: in test_app_main, in "def check", I can add a def we_are_translated(): return False
15:33
<
mattip >
it seems to work on cpython2, pypy, pypy-in-a-venv
15:34
<
mattip >
well, and then assign that function to app_main.we_are_translated
15:34
<
mattip >
but if you think it will be simple to add a sys.pypy_untranslated, that would be great
16:32
todda7 has quit [Ping timeout: 258 seconds]
20:03
speeder39_ has joined #pypy
20:35
epony has quit [Ping timeout: 258 seconds]
20:46
epsilonKNOT has joined #pypy
21:41
altendky has quit [Ping timeout: 240 seconds]
21:43
wallet42_ has quit [Ping timeout: 240 seconds]
21:43
altendky has joined #pypy
21:44
altendky has quit [Excess Flood]
21:45
wallet42_ has joined #pypy
21:46
altendky has joined #pypy
22:05
epony has joined #pypy
22:13
speeder39_ has quit [Quit: Connection closed for inactivity]
22:29
epsilonKNOT has joined #pypy
22:56
RemoteFox has joined #pypy
22:56
RemoteFox has left #pypy [#pypy]
23:07
igitoor has quit [Ping timeout: 260 seconds]
23:08
igitoor has joined #pypy
23:12
epsilonKNOT_ has joined #pypy
23:13
epsilonKNOT has quit [Ping timeout: 256 seconds]
23:15
igitoor has joined #pypy
23:15
igitoor has quit [Changing host]
23:29
lritter has quit [Quit: Leaving]