Search
Wednesday, 12th of February 2020, 18:14:35 UTC
18:36:55
stassats
for some reason the closed-over reference ends up in a different component
18:37:05
stassats
even though the function is eventually-dead
18:37:48
stassats
does FIND-INITIAL-DFO know something before everything else?
18:40:13
stassats
checking for node-to-be-deleted-p appears to be enough, but then, do we even want the AVER?
18:43:30
stassats
find-initial-dfo doesn't appear to work all that well since i made it prefer not to split functions (not that it behaved well before that)
19:36:38
stassats
electing to do nothing
20:04:23
pfdietz
Well that's one way to solve a failed AVER :)
2:53:04
|3b|
hmm, with SPEED 3, FILL on a (simple-array ub32 1) with ub32 fill value notes about boxing the value passed to ub32-bash-fill
2:57:17
|3b|
ACTION wonders how many elements it takes for writing 2 words at a time to outweigh that
3:43:41
|3b|
what does "note: couldn't inline expand because expansion calls this LET-converted local function:" mean?
4:17:40
karlosz
|3b|: can you show the code?
4:18:48
|3b|
for the inlining note? possibly, need to unbreak it first
4:21:12
|3b|
ah, still have an old copy from before i broke it, let me try to reduce it a bit
4:27:20
|3b|
still messy, but compiles without other code https://gist.github.com/3b/0de02be5a32ccdf0d4cc0eacaa420a3c
4:29:49
|3b|
updated gist with simpler version
4:35:39
|3b|
even simpler version https://gist.github.com/3b/0de02be5a32ccdf0d4cc0eacaa420a3c#gistcomment-3175366
4:36:49
|3b|
seems to need 2 calls through funcall
6:13:31
|3b|
hmm, optimizing by adding NOTINLINE for local functions :p
6:14:21
|3b|
with them inlined, it splits the termination test and the body of the loop into opposite ends of the functions, which seems to be slightly slower
Thursday, 13th of February 2020, 6:14:35 UTC