ChanServ changed the topic of #dri-devel to: <ajax> nothing involved with X should ever be unable to find a bar
Company has joined #dri-devel
nchery has quit [Quit: Leaving]
March-123 has quit [Ping timeout: 480 seconds]
hch12907 has joined #dri-devel
gawin has quit [Quit: Konversation terminated!]
ahajda has quit [Read error: Connection reset by peer]
hch12907_ has joined #dri-devel
<airlied> jenatali: hey my spirv llvm translator fix for the entrypoints finally got merged!
hch12907 has quit [Ping timeout: 480 seconds]
<mattst88> wow, originally opened in August
<daniels> :o
<mattst88> contributing to Khronos projects through the public github is... not easy
<airlied> yeah esp when you don't want to get up at 2am to harass ppl on phone calls
ybogdano has joined #dri-devel
alatiera2 has joined #dri-devel
pnowack has quit [Quit: pnowack]
<airlied> I also forgot about it for a few weeks :-P
<airlied> and of course someone else merged stuff that broke it
<jenatali> airlied: I did see that one, about time!
pcercuei has quit [Quit: dodo]
<jenatali> That means we won't have to apply our workaround patch when we jump to LLVM 14
danvet has quit [Ping timeout: 480 seconds]
ybogdano has quit [Ping timeout: 480 seconds]
rasterman has quit [Quit: Gettin' stinky!]
iive has quit []
<alyssa> airlied: with that translator we can build a gpu, nice!
<airlied> alyssa: seems like the perfect basis for a gpu :-P
<airlied> whether you want and llvm or spir-v gpu is the question we yet to answer
<imirkin> the real question is why doesn't the GPU just consume SPIR-V as its ISA... that would make it so much easier to write a backend
<imirkin> so much waste.
<airlied> imirkin: PTX in firmware is the way forward :-P
<imirkin> :)
<imirkin> airlied: llvm *does* have a PTX backend iirc
ybogdano has joined #dri-devel
<imirkin> no clue what quality stuff it produces
* airlied really wants a PTX parser
<airlied> well not really wants, but would like to see one
<imirkin> i've never looked at how it's encoded
<imirkin> nvdisasm can decode it
<imirkin> or you mean a parser for the "assembly language"?
<airlied> oh there is a PTX frontend project for llvm from nvidia, or at least slides about it :-P
<alyssa> imirkin: You, you make an excellent case
<imirkin> airlied: there's a NV-PTX backend in llvm, that's all i know
<imirkin> airlied: i somewhat assume it produces PTX "binaries", which you can then feed to CUDA somehow
<imirkin> ah, didn't know about the existence of that
<airlied> not sure it does beyond the slides :-P
<alyssa> if i maintained llvm i would nak that since the code consuming ptx is proprietary...
tursulin has quit [Read error: Connection reset by peer]
* alyssa grumpy
<alyssa> we can't all be airlied and danvet...
<airlied> we should propose some stuff for jekstand to get NAK practice in :P
<alyssa> airlied: hmmmmm I can start writing a driver for a Mali that doesn't exist, that seems NAK worthy :-p
<airlied> alyssa: just port llvmpipe to an imaginary risc-v
kevinx has joined #dri-devel
ybogdano has quit [Ping timeout: 480 seconds]
LexSfX has quit []
LexSfX has joined #dri-devel
<alyssa> airlied: why risc-v? i was thinking of adding fp64 extensions to 6502 and running llvmpipe on that...
<jekstrand> Why do I need to start NAKing more things?
<jekstrand> I've thought about writing a NIR -> PTX back-end...
* jekstrand runs
* zmike would rather have nir -> smoked pork shoulder
<zmike> work on things that matter, jason.
co1umbarius has joined #dri-devel
columbarius has quit [Ping timeout: 480 seconds]
bbrezillon has quit [Read error: Connection reset by peer]
mripard has quit [Read error: Connection reset by peer]
<jekstrand> zmike: Like what? Reviewing your patches? :-P
mripard has joined #dri-devel
bbrezillon has joined #dri-devel
<zmike> I'm not sure how that's going to get us artisanally cooked meats?
<airlied> need some sponsorship deal with a proscuito maker
<jekstrand> zmike: I didn't go to Weta just to keep reviewing Zink patches. :-P
<zmike> "keep" implies it's something you've been doing all along though, and everyone here knows that you've really been using that time to work on IVI systems for Big Automotive
<airlied> I've definitely been using IVI time for zink review
<jekstrand> Given the choices...
ngcortes has quit [Remote host closed the connection]
LexSfX has quit []
LexSfX has joined #dri-devel