antocuni changed the topic of #pypy to: PyPy, the flexible snake (IRC logs: https://botbot.me/freenode/pypy/ ) | use cffi for calling C | "PyPy: the Gradual Reduction of Magic (tm)"
<Joghurt>
How can I correctly define a highlevel entrypoint in RPython that takes an int[] correctly? The following crashes with incminimark gc, but not with boehm gc: https://pastebin.com/xCXnTWfX
ArneBab has quit [Remote host closed the connection]
tbodt has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
marky1991 has joined #pypy
<Duality>
LarstiQ: good one, uh somewhere in my program it crashes :D
marky1991 has quit [Remote host closed the connection]
marky1991 has joined #pypy
forgottenone has quit [Quit: Konversation terminated!]
nunatak has joined #pypy
marky1991 has quit [Ping timeout: 260 seconds]
ArneBab has joined #pypy
ArneBab has left #pypy [#pypy]
altendky has joined #pypy
Duality has quit [Remote host closed the connection]
tbodt has joined #pypy
<kenaan_>
arigo cffi/cffi 6d56592a5c9b /: Issue #357: fix the out-of-line ABI mode when we see structs containing anonymous unions or vice-versa
tbodt has quit [Client Quit]
<kenaan_>
arigo cffi/cffi acf2cf3a6959 /testing/cffi1/test_re_python.py: test for the recursion in _struct_collecttype(): fails if the recursion is replaced with a simple _do_collect_type()
demonimin has quit [Remote host closed the connection]
<blachance>
random question about greens: if a jitdriver's greens is [], will there be any profiling to see if something gets hot?
demonimin has joined #pypy
<simpson>
I'm not sure what happens without greens.
tbodt has joined #pypy
<mjacob>
blachance: what do you mean by "any profiling"?
<cfbolz>
blachance: it will just create a single trace
<blachance>
by profiling I mean counting to determine whether or not to make a trace; my current (potentially flawed) understanding is that there's a counter for each combination of greens, and when that counter reaches a threshold a trace gets taken
<cfbolz>
yes
<blachance>
cfbolz: a single trace for the entire run of the program?
<cfbolz>
so if there is no green, you will get a single counter
<cfbolz>
yes
<blachance>
interesting; and action increments that counter single counter?
<cfbolz>
not for the entire program
<blachance>
* and what action
<cfbolz>
for one iteration of the loop that the can_enter_jit is in
nakka has joined #pypy
<blachance>
thanks, that makes sense now---a trace starts at some combination of greens and ends at the same one, and if all combinations are the same then it'll only trace one iteration