00:32
Rhy0lite has joined #pypy
00:47
oberstet has quit [Remote host closed the connection]
01:38
Ninpo has quit [Ping timeout: 264 seconds]
01:39
jcea has quit [Quit: jcea]
01:45
Rhy0lite has quit [Quit: Leaving]
01:51
Ninpo has joined #pypy
02:26
_whitelogger has joined #pypy
03:11
_whitelogger has joined #pypy
04:11
_whitelogger has joined #pypy
04:46
marky1991_2 has joined #pypy
04:49
marky1991 has quit [Ping timeout: 260 seconds]
06:05
<
mattip >
on windows 32-bit, the tool.jitlogparser.test.test_parser tests fail since the test uses 64-bit semantics,
06:06
<
mattip >
so the raw_asm has "['ffffffff <.data>:\r\n', 'ffff:\t64-bit address is disabled\r\n']"
06:06
<
mattip >
and the asm is an empty list
06:14
forgottenone has joined #pypy
07:18
<
arigo >
mattip: because there is no "nm.exe" on Windows. you should just skip these failures
07:22
<
arigo >
or, I guess, for the last of the 3 failures, because it finds some "objdump.exe" which doesn't support disassembling 64-bit code
07:23
<
arigo >
this was never really meant to run on windows, I guess it would work if you install some specific versions of unix-y tools but there is no point
07:42
otisolsen70_ has joined #pypy
07:44
otisolsen70_ has quit [Remote host closed the connection]
07:45
otisolsen70_ has joined #pypy
07:46
otisolsen70 has quit [Ping timeout: 240 seconds]
07:46
otisolsen70_ has quit [Remote host closed the connection]
08:11
otisolsen70 has joined #pypy
08:20
_whitelogger has joined #pypy
08:48
otisolsen70 has quit [Ping timeout: 256 seconds]
09:20
oberstet has joined #pypy
09:39
<
mattip >
arigo: I think the streamio changes in 87e22968f9b4 were exactly the kind of thing you were warning against
10:21
<
arigo >
mattip: well, yes
10:21
<
arigo >
you'll end up replacing all of os.xxx with rposix.xxx, over time
10:21
<
arigo >
(which may not be a bad thing)
10:49
<
mattip >
yay for cffi arm64 wheels
11:13
pmp-p has quit [Ping timeout: 256 seconds]
11:15
pmp-p has joined #pypy
11:26
_whitelogger has joined #pypy
11:59
otisolsen70 has joined #pypy
12:02
lritter has joined #pypy
12:27
otisolsen70 has quit [Ping timeout: 265 seconds]
12:37
otisolsen70 has joined #pypy
12:37
otisolsen70 has quit [Remote host closed the connection]
13:12
marky1991_2 has quit [Remote host closed the connection]
13:13
marky1991_2 has joined #pypy
13:15
<
arigo >
mmap module: why yield integers when you can yield bytes? who cares about consistency
13:58
marky1991_2 has quit [Remote host closed the connection]
13:59
marky1991_2 has joined #pypy
14:10
marky1991 has joined #pypy
14:14
marky1991_2 has quit [Ping timeout: 264 seconds]
14:25
jcea has joined #pypy
14:26
<
cfbolz >
arigo: cool, are you actively working on supporting win64?
14:30
forgottenone has quit [Ping timeout: 246 seconds]
14:59
<
arigo >
cfbolz: yes
15:00
<
arigo >
cfbolz: well no, overall :-)
15:00
<
arigo >
only in the JIT, which involves mainly only fixing the calling convension
15:09
<
cfbolz >
arigo: ah, cool
15:11
<
arigo >
see a recent issue where a contributor reached already relatively far for win64 support
16:02
<
arigo >
cfbolz: btw, do you remember test_lshift in x86/test/test_regalloc.py which you wrote (I guess) for reghint.py?
16:03
<
arigo >
this test as written cannot pass on win64: the first argument in passed in ecx, and ecx is also where the int_lshift count must be
16:03
<
arigo >
so we get one extra move (correctly detected by the logic of the test)
16:19
<
cfbolz >
No, I fear I have no memory of that test
16:33
<
arigo >
the test works again if I pass it as 2nd argument, i.e. != ecx, so I think I must not be completely wrong
17:18
oberstet has quit [Remote host closed the connection]
17:30
oberstet has joined #pypy
19:14
marky1991_2 has joined #pypy
19:16
marky1991 has quit [Ping timeout: 256 seconds]
19:20
marky1991_2 has quit [Ping timeout: 240 seconds]
19:33
oberstet has quit [Remote host closed the connection]
23:20
lritter has quit [Ping timeout: 264 seconds]