Da_Coynul has quit [Quit: My MacBook Air has gone to sleep. ZZZzzz…]
jrmuizel has quit [Remote host closed the connection]
_whitelogger has joined #lima
adjtm_ has joined #lima
adjtm has quit [Ping timeout: 258 seconds]
dddddd has quit [Remote host closed the connection]
_whitelogger has joined #lima
hoijui has joined #lima
_whitelogger has joined #lima
_whitelogger has joined #lima
_whitelogger has joined #lima
_whitelogger has joined #lima
sphalerite has joined #lima
dddddd has joined #lima
_whitelogger has joined #lima
_whitelogger has joined #lima
Da_Coynul has joined #lima
<enunes>
anarsoul: my only question on that MR is does const lowering really need to be moved to a separate pass?
<MoeIcenowy>
Oh, android build of lima is broken again
<enunes>
recently while working on another MR I also had a patch doing const loewring after everything else, but then realized that it wasn't really necessary
Da_Coynul has quit [Quit: My MacBook Air has gone to sleep. ZZZzzz…]
megi has quit [Ping timeout: 248 seconds]
jrmuizel has joined #lima
Da_Coynul has joined #lima
jrmuizel has quit [Remote host closed the connection]
Da_Coynul has quit [Quit: My MacBook Air has gone to sleep. ZZZzzz…]
jrmuizel has joined #lima
yuq825 has quit [Remote host closed the connection]
Da_Coynul has joined #lima
jrmuizel has quit [Remote host closed the connection]
Da_Coynul has quit [Quit: My MacBook Air has gone to sleep. ZZZzzz…]
jrmuizel has joined #lima
jrmuizel has quit [Remote host closed the connection]
Da_Coynul has joined #lima
jrmuizel has joined #lima
niceplaces has joined #lima
jrmuizel has quit [Remote host closed the connection]
jrmuizel has joined #lima
<anarsoul>
enunes: yes, since sin/cos and branch lowering creates consts
<anarsoul>
so we need to make sure that lowering is done at the point when no new const nodes are created
<enunes>
anarsoul: well but then they just become lowered normally as they are added to the end of the list and the normal lowering loop will go through them, no?
<anarsoul>
enunes: yes if we're using pred/succ lists in lowering
<anarsoul>
but I'm trying to get away from it
<anarsoul>
so lowering won't use successors/predecessors lists anymore
<enunes>
I see, it is because of the same change then, makes sense
jrmuizel has quit [Ping timeout: 245 seconds]
<enunes>
anarsoul: would it make sense to try to duplicate consts in nir or in a lima specific nir pass?