#pypy on 2021-08-12 — irc logs at libera.irclog.whitequark.org

09:56 <yuiza> cfbolz: sorry to bother you again. Who manages the runtime stack? It seems framestack in pyjitpl.py handles, but don't know that for sure.

09:59 <cfbolz> yuiza: ah, I was more wondering whether you cannot do it purely in the interpreter

10:02 <yuiza> cfbolz: i see. doing in the interpreter is simpler.

10:02 <cfbolz> yuiza: but in general, I am a bit worried about this problem

10:02 <cfbolz> yuiza: if you have more complex state

10:07 <yuiza> cfbolz: me too. It might be that doing it in RPython level is better to resolve more complex state, if we can know which stack should be saved

10:08 <cfbolz> yuiza: sure, but there are things you cannot roll back! eg if one of the explored paths has an actual print in it

10:08 <cfbolz> but I am happy to leave this problem for a while longer ;-)

10:11 <yuiza> cfbolz: yes, side effects are problematic. I agree with your suggestion :) > but I am happy to leave this problem for a while longer ;-)

10:12 gef has quit [Ping timeout: 268 seconds]

10:13 gef_ has quit [Ping timeout: 256 seconds]

10:13 Julian has quit [Quit: leaving]

10:15 yuiza_ has joined #pypy

10:30 <bbot2> Success: http://buildbot.pypy.org/builders/jit-benchmark-linux-x86-64/builds/3357 [py3.7]

10:38 lessshaste has joined #pypy

10:39 lesshaste has quit [Read error: Connection reset by peer]

10:40 gef has joined #pypy

10:42 gef_ has joined #pypy

12:00 lessshaste has quit [Ping timeout: 268 seconds]

12:01 leshaste has joined #pypy

12:04 Atque has joined #pypy

12:45 yuiza_ has quit [Quit: Connection closed for inactivity]

12:49 Atque has quit [Quit: ...]

12:56 lritter has joined #pypy

13:36 Julian has joined #pypy

14:15 Julian has quit [Quit: leaving]

14:36 gef has quit [Ping timeout: 248 seconds]

14:37 yuiza_ has joined #pypy

14:37 gef_ has quit [Ping timeout: 268 seconds]

14:38 gef has joined #pypy

14:41 gef_ has joined #pypy

15:04 Joannah has joined #pypy

15:27 <Joannah> I may be mistaken but IIRC, I must have bumped into a PyPy branch when the project was still on bitbucket where cfbolz was experimenting with allocation sites. Do we have such branch living somewhere? or am mistaking it with artifacts from the storage strategies paper?

15:42 gef has quit [Ping timeout: 272 seconds]

15:42 gef_ has quit [Ping timeout: 245 seconds]

15:44 leshaste has quit [Remote host closed the connection]

15:44 lesshaste has joined #pypy

15:45 mathieu_ has quit [Ping timeout: 245 seconds]

15:46 yuiza has quit [Quit: Leaving.]

15:46 yuiza has joined #pypy

15:47 yuiza has left #pypy [#pypy]

15:47 yuiza has joined #pypy

15:48 yuiza has quit [Client Quit]

15:48 stkrdknmibalz has quit [Quit: WeeChat 3.0.1]

15:49 Joannah has quit [Quit: Client closed]

15:53 Dejan has quit [Quit: Leaving]

15:54 gef has joined #pypy

15:57 <cfbolz> Joannah: sorry, I don't remember that at all :-(

15:57 <cfbolz> Not even doing that work

16:00 gef has quit [Ping timeout: 256 seconds]

16:01 yuiza has joined #pypy

16:04 <yuiza> cfbolz: I naively ran pystone on py3.7-hack-measure-bytecode-dispatch and confirmed that the threaded code version is a little bit slow :( I continue to take this benchmark more precisely

16:08 gef has joined #pypy

16:08 gef_ has joined #pypy

16:13 gef has quit [Ping timeout: 248 seconds]

16:13 gef_ has quit [Ping timeout: 256 seconds]

16:22 gef has joined #pypy

16:25 mathieu_ has joined #pypy

16:28 gef has quit [Ping timeout: 245 seconds]

16:33 gef has joined #pypy

16:33 gef_ has joined #pypy

17:02 Joannah has joined #pypy

17:12 <Joannah> cfbolz , my bad I must be remembering wrongly

17:14 lritter has quit [Ping timeout: 256 seconds]