<chewitt>
^ was trying to turn up brightness (above the 2 key) in the dark .. je m'excuse
empty_string has quit [Ping timeout: 272 seconds]
empty_string has joined #panfrost
tgall_foo has quit [Quit: My iMac has gone to sleep. ZZZzzz…]
vstehle has joined #panfrost
camus has joined #panfrost
kaspter has quit [Read error: Connection reset by peer]
camus is now known as kaspter
chewitt has quit [Read error: Connection reset by peer]
chewitt has joined #panfrost
_whitelogger has joined #panfrost
davidlt has joined #panfrost
_whitelogger has joined #panfrost
_whitelogger has joined #panfrost
chewitt has quit [Quit: Adios!]
camus has joined #panfrost
kaspter has quit [Ping timeout: 265 seconds]
camus is now known as kaspter
<icecream95>
There are still some artefacts (certain colours are wrongly displayed as #00ff00 or #0000ff) after restricting gamma to 8-bit
davidlt has quit [Ping timeout: 264 seconds]
<HdkR>
icecream95: clamping versus scaling the range or something?
icecream95 has quit [Ping timeout: 260 seconds]
raster has joined #panfrost
chewitt has joined #panfrost
stikonas has joined #panfrost
stikonas has quit [Remote host closed the connection]
robmur01_ is now known as robmur01
kaspter has quit [Ping timeout: 260 seconds]
stikonas has joined #panfrost
kaspter has joined #panfrost
davidlt has joined #panfrost
<urjaman>
fatal: packfile .git/objects/pack/pack-6fecbea4a6af03c1ad53b637a3a12493edd4160d.pack cannot be mapped: Cannot allocate memory
<urjaman>
.... sigh. (the kernel git is getting too big for 32bit systems, i feel)
<HdkR>
:D
<HdkR>
Time to checkout shallow copies?
<urjaman>
i'm gonna check if i was actually running out of RAM+swap (my swap on the C201 is tiny), first but like ... wouldnt be surprised if it was out of VM space :P
<urjaman>
those are quite close anyways for this device :P (4GB of RAM, 512M of swap, 32bit...)
<urjaman>
oh yeah for this op (mapping a file) it'd be VM space ... hmm
* urjaman
is tired
<urjaman>
and yeah there isnt a problem if you dont need to repack anything (the line after that is, "fatal: index-pack failed"), but i was trying to fetch the upstream stable git repo onto my git that had only my branch from github (i messed up my old kernel git somehow, so i was trying to get a fresh setup)
<urjaman>
i guess i can do the git work on ... one of my x64 things, it's just fetching, rebasing, pushing
<HdkR>
What a fun failure case
camus has joined #panfrost
kaspter has quit [Ping timeout: 240 seconds]
camus is now known as kaspter
tgall_foo has joined #panfrost
<urjaman>
yeah it's a 3.0G packfile ... not exactly surprising one can't mmap that (atleast while doing anything else at all :P)
<HdkR>
Wow
alpernebbi has joined #panfrost
kaspter has quit [Quit: kaspter]
mixfix41 has quit [Quit: leaving]
stikonas has quit [Remote host closed the connection]
stikonas has joined #panfrost
kaspter has joined #panfrost
leah has quit [Quit: WeeChat 2.8]
chewitt has quit [Read error: Connection reset by peer]
chewitt has joined #panfrost
<urjaman>
I stuffed the git repo to an x64 machine (my PCEngines APU1 lol, i bet the C201 is faster...), limited the pack size to 512M (git has a knob for that but the default is unlimited... uhh), did that fetch (just because), repacked the repo, and moved it back to the C201... lets see what breaks next :P
unoccupied has joined #panfrost
gcl has quit [Ping timeout: 246 seconds]
gcl has joined #panfrost
endrift has quit [Ping timeout: 264 seconds]
yann has quit [Ping timeout: 246 seconds]
yann has joined #panfrost
alpernebbi has quit [Quit: alpernebbi]
stikonas has quit [Remote host closed the connection]
<macc24>
icecream95: i redone cadmium kernel building stuff to pull kernel from kernel.org and then patch it to be duet-compatible, already fixed ;)
<alyssa>
how many hundreds of commits until i realize merging my compiler changes is a good idea..
<urjaman>
hmm..?
<alyssa>
urjaman: have been working furiously in a private branch but have been too distracted by $LIFE (thanks 2020) to get my patches cleaned up and upstreamed
<alyssa>
I started work on the real scheduler today
<alyssa>
won't be done by christmas alas but hey, I can get a _start_ on Santa Claus
<urjaman>
:)
<urjaman>
It's actually somewhat common for me to be building a kernel on christmas
<alyssa>
lol
<alyssa>
also crap 34 open panfrost MRs
<icecream95>
alyssa: And I have one more ready to send, fixing a bug with SuperTxuKart
<alyssa>
🎉
<macc24>
🎉
<alyssa>
icecream95: I see you poking at OpenCL too apparently? :p
<icecream95>
alyssa: I got it to the point of calling launch_grid, but then I get page faults at address 0, even with a shader that doesn't actually do anything
<alyssa>
Ruh roh.
<alyssa>
At some point (~Nov 2019) I had hello world OpenCL working on t860
<alyssa>
but I got too frustrated with llvm versioning and.. ;p
tomboy64 has quit [Remote host closed the connection]
tomboy64 has joined #panfrost
Ashleee has quit [Quit: Leaving... bbl]
* alyssa
is in git rebase hell
listen_ers has joined #panfrost
leah has quit [Quit: WeeChat 2.8]
<alyssa>
but hey, we'll have shiny compiler after
macc24 has quit [Ping timeout: 246 seconds]
listen_ers has quit [Ping timeout: 245 seconds]
<alyssa>
all the fixups are gone!
<alyssa>
...still need to fix regressions => more fixups :<
<HdkR>
woo fixes
macc24 has joined #panfrost
<macc24>
did i miss anything?
<HdkR>
Not much
<HdkR>
Mostly alyssa being awesome as usual
cwabbott has quit [Ping timeout: 264 seconds]
<icecream95>
Hmm, so instancing with 17 vertices is more broken than I thought...
kinkinkijkin has quit [Ping timeout: 272 seconds]
narmstrong has quit [Ping timeout: 264 seconds]
steev has quit [Ping timeout: 272 seconds]
<alyssa>
wasn't me
<alyssa>
icecream95: Instancing code is super special cased and also done from r/e fwiw
<alyssa>
although, computational details aside, the hardware just wants something in the form (2m + 1) * 2^n