ChanServ changed the topic of #lima to: Development channel for open source lima driver for ARM Mali4** GPUs - Kernel has landed in mainline, userspace driver is part of mesa - Logs at https://people.freedesktop.org/~cbrill/dri-log/index.php?channel=lima and https://freenode.irclog.whitequark.org/lima - Contact ARM for binary driver support!
rellla has quit [Ping timeout: 256 seconds]
warpme_ has quit [Quit: Connection closed for inactivity]
rellla has joined #lima
buzzmarshall has quit [Remote host closed the connection]
cp- has quit [Quit: Disappeared in a puff of smoke]
cp- has joined #lima
cp- has quit [Client Quit]
cp- has joined #lima
cp- has quit [Client Quit]
cp- has joined #lima
dddddd has quit [Remote host closed the connection]
Barada has joined #lima
Elpaulo has joined #lima
<anarsoul> enunes: rellla: any comments on https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4055 ?
<MoeIcenowy> anarsoul: does this break usual applications?
<anarsoul> nope
<anarsoul> why should it?
<rellla> anarsoul: lgtm, the only difference is, that we do the check in *gpir_instr_create already, isn't it?
warpme_ has joined #lima
yann has quit [Ping timeout: 258 seconds]
ChanServ has quit [*.net *.split]
rellla has quit [*.net *.split]
narmstrong has quit [*.net *.split]
romangg has quit [*.net *.split]
paulk-leonov has quit [*.net *.split]
arti has quit [*.net *.split]
egbert has quit [*.net *.split]
plaes has quit [*.net *.split]
ente has quit [*.net *.split]
hell__ has quit [*.net *.split]
griffinp has quit [*.net *.split]
bhoj has quit [*.net *.split]
forkbomb has quit [*.net *.split]
eightdot has quit [*.net *.split]
xexaxo1 has quit [*.net *.split]
mariogrip has quit [*.net *.split]
marex-cloud has quit [*.net *.split]
Barada has quit [*.net *.split]
_whitelogger has joined #lima
yann has joined #lima
UnivrslSuprBox has joined #lima
anarsoul|c has joined #lima
wiewo has joined #lima
jernej has joined #lima
deesix has joined #lima
bshah has joined #lima
<rellla> glad to get it reviewed. i think this could be useful ...
kaspter has quit [Quit: kaspter]
kaspter has joined #lima
kaspter has quit [Quit: kaspter]
gcl_ has joined #lima
gcl has quit [Ping timeout: 268 seconds]
buzzmarshall has joined #lima
Barada has quit [Quit: Barada]
dddddd has joined #lima
danqo has joined #lima
yann has quit [Ping timeout: 256 seconds]
rak-zero has joined #lima
<anarsoul> op18.v1 $0.x $2.zzzz is interesting :)
<anarsoul> that's some alu op we don't know yet
buzzmarshall has quit [Remote host closed the connection]
yann has joined #lima
rak-zero has quit [Quit: https://quassel-irc.org - Chat comfortably. Anywhere.]
<rellla> anarsoul: op18.v1 does the following:
<anarsoul> dramatic pause? :)
<rellla> op18.v1 $0.x $2.xyyy adds the first two components of $2 and subtracts the value in .s0 from it. or kind of it
<rellla> (mobil phone :p)
<anarsoul> how do you know that?
<rellla> so op18.v1 $1.x $2.zyyy is equivalent with
<rellla> $1 = $2.z + $2.y - $0.w
<rellla> by reading the above example you linked?
<rellla> compare the disassembled code with the original shader
<anarsoul> hold on
<anarsoul> @ 20: op18.v1 $0.x $2.zzzz doesn't have anything in s0
<rellla> then it's the last sth. value
<rellla> it's $1.x which was last used in 11: mul.s0
<rellla> and op18 in 16: uses $0.w from the mul.s0 before
<anarsoul> we haven't seen such kind of deps between instructions before :\
<rellla> it's the only operation, which makes it equivalent to the shader...
<rellla> maybe we should grep for other op18's...
<anarsoul> or any other op
<rellla> ok, there are more tests with op18. i will look into them tomorrow...
warpme_ has quit [Quit: Connection closed for inactivity]
buzzmarshall has joined #lima