2018-02-26 15:52
cfbolz changed the topic of #pypy to: PyPy, the flexible snake (IRC logs:
https://botbot.me/freenode/pypy/ ) | use cffi for calling C | the secret reason for us trying to get PyPy users: to test the JIT well enough that we're somewhat confident about it
00:02
lritter_ has joined #pypy
00:06
lritter has quit [Ping timeout: 264 seconds]
00:21
tbodt has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
00:48
asmeurer_ has quit [Quit: asmeurer_]
00:49
asmeurer_ has joined #pypy
00:49
jcea has quit [Remote host closed the connection]
00:53
marself has quit [Ping timeout: 276 seconds]
01:27
exarkun has quit [Read error: Connection reset by peer]
01:27
exarkun has joined #pypy
01:39
marky1991 has quit [Read error: Connection reset by peer]
01:57
asmeurer_ has quit [Quit: asmeurer_]
01:59
asmeurer_ has joined #pypy
02:05
oberstet2 has joined #pypy
02:05
oberstet has quit [Ping timeout: 248 seconds]
02:12
Hasimir has quit [Read error: Connection reset by peer]
02:18
Hasimir has joined #pypy
03:31
carljm has joined #pypy
04:01
lritter_ has quit [Ping timeout: 248 seconds]
04:36
Garen has quit [Read error: Connection reset by peer]
04:36
Garen has joined #pypy
05:14
tonyseek has joined #pypy
05:27
energizer has quit [Ping timeout: 265 seconds]
06:27
MerlinTheWise has joined #pypy
06:30
MerlinTheWizard has quit [Ping timeout: 255 seconds]
06:46
asmeurer_ has quit [Quit: asmeurer_]
06:50
asmeurer_ has joined #pypy
07:04
oberstet2 has quit [Ping timeout: 260 seconds]
07:11
energizer has joined #pypy
07:12
energizer has quit [Remote host closed the connection]
07:12
energizer has joined #pypy
07:17
oberstet2 has joined #pypy
07:17
antocuni has joined #pypy
07:23
oberstet3 has joined #pypy
07:24
oberstet2 has quit [Ping timeout: 255 seconds]
07:25
energizer has quit [Remote host closed the connection]
07:29
lazka has joined #pypy
07:43
antocuni has quit [Ping timeout: 260 seconds]
07:46
energizer has joined #pypy
07:49
tayfun26 has joined #pypy
07:52
mcyprian has joined #pypy
08:00
mcyprian has quit [Ping timeout: 248 seconds]
08:00
mcyprian has joined #pypy
08:05
mcyprian has quit [Ping timeout: 264 seconds]
08:05
tonyseek has quit [Quit: tonyseek]
08:11
tonyseek has joined #pypy
08:19
bendlas has quit [Ping timeout: 245 seconds]
08:21
lazka has quit [Quit: Leaving]
08:22
mcyprian has joined #pypy
08:24
bendlas has joined #pypy
08:26
lazka has joined #pypy
08:40
mcyprian has quit [Quit: Leaving.]
08:41
mcyprian has joined #pypy
08:41
mcyprian has quit [Client Quit]
09:11
antocuni has joined #pypy
09:19
tonyseek has quit [Remote host closed the connection]
09:20
tonyseek has joined #pypy
09:24
asmeurer_ has quit [Quit: asmeurer_]
09:28
the_drow has joined #pypy
09:31
energizer has quit [Ping timeout: 240 seconds]
09:36
<
the_drow >
This is similar to the way RapidJSON skips whitespaces in a string.
09:37
<
the_drow >
I've also opened an issue about integrating the code into 'str'.strip() etc.
09:38
<
cfbolz >
the_drow: that is not my question though. My question is whether real world json files have enough whitespace for this to pay off
09:42
<
fijal >
most people who go fancy on JSON soon realize that there are much better protocols out there typically
09:47
<
the_drow >
fijal, browsers communicate via json if we like it or not
09:47
<
the_drow >
But I agree that other serialization protocols are much better
09:47
<
fijal >
I'm not debating your work
09:47
<
fijal >
I'm saying "why we haven't seen that much work in the area"
09:47
<
fijal >
sorry if it came out wrong
09:47
<
the_drow >
yeh I get it
09:47
<
cfbolz >
the_drow: you still didn't answer my question
09:48
<
the_drow >
cfbolz, I think that if the fastest parser in the world includes those optimizations they are not there for no reason
09:48
<
cfbolz >
You'd be surprised
09:48
<
cfbolz >
That's argument by authority
09:48
<
the_drow >
A pretty printted JSON contains at least one whitespace (tab) per nesting
09:49
<
the_drow >
Sometimes spaces are used instead of tabs
09:49
<
the_drow >
Clients are of a variety degree of quality and some do not send compact representations
09:50
<
the_drow >
cfbolz, Whitespace skipping using SIMD can be a 1% performance boost or 10%, depending on the input
09:54
mcyprian has joined #pypy
09:54
<
cfbolz >
Sounds cool, I suppose. Somebody needs to go and implement it
09:55
<
the_drow >
Right now I'm stuck with a translation error I can't figure out
09:56
<
the_drow >
I'm also wondering if we should do it with rffi + C or use rffi to expose interstices to RPython and code this in RPython somehow
09:56
<
the_drow >
The latter is more complex but will allow us to use SIMD more easily
09:56
<
the_drow >
In any case a SIMD module is required to detect at boot time which instructions are available to us
10:01
<
cfbolz >
You'd like a platform independent simd DSL compiler somehow. I wonder whether one exists anywhere
10:01
asmeurer_ has joined #pypy
10:06
asmeurer_ has quit [Ping timeout: 240 seconds]
10:10
mcyprian has quit [Quit: Leaving.]
10:13
antocuni has quit [Ping timeout: 268 seconds]
10:17
cjwelborn has quit [Remote host closed the connection]
10:36
jcea has joined #pypy
10:37
inad922 has joined #pypy
10:37
demonimin has quit [Ping timeout: 268 seconds]
10:51
iblis17_ is now known as iblis17
10:55
tonyseek has quit [Ping timeout: 264 seconds]
11:00
tonyseek has joined #pypy
11:14
cjwelborn has joined #pypy
11:14
cjwelborn has joined #pypy
11:14
cjwelborn has joined #pypy
11:14
cjwelborn has quit [Changing host]
11:48
<
the_drow >
but it requires us to compile code to their IR
11:48
antocuni has joined #pypy
11:49
<
the_drow >
cfbolz, I think that a feature detection module should be enough
11:52
tonyseek has quit [Quit: tonyseek]
11:56
<
antocuni >
the_drow: as usual rpython is a bit obscure
11:57
<
antocuni >
but I think the error is due to the fact that in the self._raise you use %d and you pass "i"
11:57
inad922 has quit [Ping timeout: 265 seconds]
11:57
<
antocuni >
i is returned by pypy__skip_whitespaces_simd, which is declared as "unsigned int"
11:57
<
antocuni >
so it has an rpython type of r_uint
11:59
<
antocuni >
I don't think that oefmt supports any formatting char for passing unsigneds, though
11:59
<
antocuni >
so your best bet is probably to use intmask() to convert the r_uint into a normal int
11:59
<
the_drow >
antocuni, but i wasn't previously unsigned?
12:00
<
the_drow >
without my code changes?
12:00
<
the_drow >
Because it should be size_t...
12:00
<
antocuni >
we can argue that it should be size_t, indeed
12:01
<
antocuni >
but from the rpython point of view, it's a signed int
12:01
<
antocuni >
look at loads()
12:01
<
antocuni >
it calls decode_any(0), where 0 is an "int"
12:01
<
antocuni >
and then the "i" is passed everywhere, so the type propagates as int
12:01
R3d_Sky has joined #pypy
12:02
<
antocuni >
indeed, probably the easiest thing to do is to make your C function returning a long, and then fix the signedness later
12:06
marky1991 has joined #pypy
12:10
simpson has quit [Ping timeout: 255 seconds]
12:14
inad922 has joined #pypy
12:15
marky1991 has quit [Ping timeout: 240 seconds]
12:16
marr has joined #pypy
12:25
Rhy0lite has joined #pypy
12:33
oberstet3 has quit [Quit: Leaving]
12:45
bremner has quit [Quit: Coyote finally caught me]
12:48
bremner has joined #pypy
13:08
exarkun has quit [Read error: Connection reset by peer]
13:09
lazka has quit [Quit: Leaving]
13:10
simpson has joined #pypy
13:17
exarkun has joined #pypy
13:18
inad922 has quit [Ping timeout: 256 seconds]
13:22
dddddd has joined #pypy
13:22
marky1991 has joined #pypy
13:22
marky1991 has quit [Remote host closed the connection]
13:23
marky1991 has joined #pypy
13:31
inad922 has joined #pypy
13:37
realitix has joined #pypy
13:48
lazka has joined #pypy
13:52
the_drow has quit [Ping timeout: 248 seconds]
13:59
tayfun26 has quit [Remote host closed the connection]
13:59
tayfun26 has joined #pypy
14:02
tayfun27 has joined #pypy
14:03
tayfun26 has quit [Ping timeout: 240 seconds]
14:03
tayfun27 is now known as tayfun26
14:12
raynold has quit [Quit: Connection closed for inactivity]
14:54
marky1991 has quit [Ping timeout: 264 seconds]
15:09
danieljabailey has joined #pypy
15:20
inad922 has quit [Ping timeout: 256 seconds]
15:31
tayfun26 has quit [Quit: tayfun26]
15:36
lazka has quit [Quit: Leaving]
16:12
tbodt has joined #pypy
16:13
ronan has quit [Remote host closed the connection]
16:18
realitix has quit [Ping timeout: 248 seconds]
16:26
lritter_ has joined #pypy
16:31
tonyseek has joined #pypy
16:32
ronan has joined #pypy
16:39
tonyseek has quit [Quit: tonyseek]
16:40
lritter_ is now known as lritter
16:40
lritter has quit [Quit: Leaving]
16:40
lritter has joined #pypy
16:43
marky1991 has joined #pypy
16:44
marky1991 has quit [Remote host closed the connection]
16:44
marky1991 has joined #pypy
16:51
Garen has quit [Read error: Connection reset by peer]
16:52
Garen has joined #pypy
17:00
antocuni has quit [Ping timeout: 240 seconds]
17:37
tonyseek has joined #pypy
17:45
jorge has joined #pypy
17:46
jorge is now known as Guest312
17:47
<
kenaan >
rlamy py3.6 d3671e3e862c /pypy/interpreter/pyparser/parsestring.py: fix translation
17:48
Guest312 has quit [Client Quit]
17:55
tonyseek has quit [Quit: tonyseek]
18:03
bjs has quit [Quit: Bye!]
18:06
energizer has joined #pypy
18:26
asmeurer has joined #pypy
18:31
lazka has joined #pypy
18:38
marky1991 has quit [Ping timeout: 240 seconds]
18:41
marky1991 has joined #pypy
18:42
nunatak has joined #pypy
18:54
lazka has quit [Quit: Leaving]
19:05
tbodt has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
19:06
marky1991 has quit [Ping timeout: 260 seconds]
19:07
kipras is now known as kipras`away
19:10
kipras`away is now known as kipras
19:28
bjs has joined #pypy
19:32
Rhy0lite has quit [Quit: Leaving]
19:32
tbodt has joined #pypy
19:37
tbodt has quit [Ping timeout: 256 seconds]
19:48
mattip has quit [Ping timeout: 240 seconds]
19:53
mattip has joined #pypy
20:04
JStoker has quit [Remote host closed the connection]
20:07
tbodt has joined #pypy
20:20
JStoker has joined #pypy
20:57
tbodt has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
20:59
tbodt has joined #pypy
21:04
marky1991 has joined #pypy
21:08
nunatak has quit [Remote host closed the connection]
21:10
marky1991 has quit [Remote host closed the connection]
21:11
marky1991 has joined #pypy
21:12
marky1991 has quit [Remote host closed the connection]
21:12
marky1991 has joined #pypy
21:22
Guest95978 has quit [Remote host closed the connection]
21:23
marvin has joined #pypy
21:33
lritter has quit [Remote host closed the connection]
21:44
antocuni has joined #pypy
21:52
tbodt has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
21:54
tbodt has joined #pypy
21:55
marvin has quit [Remote host closed the connection]
21:55
marvin has joined #pypy
21:57
marvin has quit [Remote host closed the connection]
21:58
marvin has joined #pypy
21:58
marvin has quit [Remote host closed the connection]
21:58
marky1991 has quit [Remote host closed the connection]
21:58
marvin has joined #pypy
21:58
marky1991 has joined #pypy
22:00
lazka has joined #pypy
22:01
lazka has quit [Client Quit]
22:04
tbodt has quit [Read error: Connection reset by peer]
22:04
lritter has joined #pypy
22:05
marvin has quit [Remote host closed the connection]
22:05
marvin has joined #pypy
22:07
tbodt has joined #pypy
22:16
tbodt has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
22:18
tbodt has joined #pypy
22:27
demonimin has joined #pypy
22:29
marr has quit [Ping timeout: 248 seconds]
22:32
energizer has quit [Remote host closed the connection]
22:40
ssbr has quit [Remote host closed the connection]
22:41
ssbr has joined #pypy
22:57
antocuni has quit [Ping timeout: 264 seconds]
23:13
exarkun has quit [Ping timeout: 248 seconds]
23:15
Eyess has joined #pypy
23:16
SunDwarf has quit [*.net *.split]
23:16
ronny has quit [*.net *.split]
23:16
mvantellingen has quit [*.net *.split]
23:16
Ganwell has quit [*.net *.split]
23:16
infernix has quit [*.net *.split]
23:16
kanaka has quit [*.net *.split]
23:18
exarkun has joined #pypy
23:19
pepesza- has quit [Ping timeout: 260 seconds]
23:21
pepesza has joined #pypy
23:23
kanaka has joined #pypy
23:23
ronny has joined #pypy
23:24
Ganwell has joined #pypy
23:31
infernix has joined #pypy
23:46
marky1991 has quit [Read error: Connection reset by peer]