<javierm>
tzimmermann: yeah, I consider that but didn't want to change the drm_copy_field() function signature and this really shouldn't happen if the drivers are not buggy
<tzimmermann>
ok
<tzimmermann>
feel free to add it to drm-misc-fixes
<javierm>
tzimmermann: will do, thanks again
<javierm>
tzimmermann: actually, I prefer to just put it in drm-misc-next
<javierm>
there's no rush on pushing these really
<tzimmermann>
ok, your choice
<javierm>
tzimmermann: ok
<tzimmermann>
javierm, BTW do you have comments on sravn's via patches?
<pq>
dcz_, oh, sorry, I don't speak glium.
<javierm>
tzimmermann: I don't. I think is the correct approach
<tzimmermann>
ok. i think he'll merge it today or tomorrow
<dcz_>
my FBO actually works :/ so I don't expect any problems there (I did use that as a reference)
<pq>
dcz_, you were confused about glBindTexture before, could that be the problem? Or are you using something else like bindless textures or whatever if you're on GL4?
<dcz_>
I'm binding textures every time before they enter the shader, and I'm using GL2 only
<pq>
GL2, not GLES2?
<dcz_>
I didn't see a reason to use GLES2
<pq>
okay, just confirming - I haven't touched GL2 in... a decade?
<dcz_>
I think switching to GLES2 wouldn't be hard, considering that I haven't seen any differences so far
<pq>
there are quite many steps to getting a texture into shader
danvet has joined #dri-devel
<pq>
glActiveTexture, glBindTexture, setting the sampler uniform to the active texture index...
<pq>
dcz_, did you try with a simple texture defined with glTexImage2D instead? With just some hardcoded pixel data, to see if anything comes through?
mszyprow has joined #dri-devel
<dcz_>
I'm not sure how to fill a GL texture with data
<pq>
you literally call glTexImage2D(), that's it
<dcz_>
won't that be black?
<pq>
no, when you feed in pixel data that is not black
<pq>
you pass a pointer to pixel data into that function
<dcz_>
oh, there's a data pointer
<dcz_>
won't hurt to try, thanks
<pq>
yes, that's the main purpose of that function
<pq>
whatever texture name you have bound to the GL_TEXTURE_2D slot, glTexImage2D will drop the old texture data and replace it with whatever you give in as parameters
linearcannon has quit [Ping timeout: 480 seconds]
toolchains has quit [Ping timeout: 480 seconds]
toolchains has joined #dri-devel
vliaskov has joined #dri-devel
lynxeye has joined #dri-devel
kts has quit [Ping timeout: 480 seconds]
mvlad has joined #dri-devel
fahien has quit [Quit: fahien]
bmodem has joined #dri-devel
heat has joined #dri-devel
pochu has joined #dri-devel
bmodem has quit []
bmodem has joined #dri-devel
pcercuei has joined #dri-devel
saurabhg has joined #dri-devel
bmodem has quit []
ahajda_ has joined #dri-devel
toolchains has quit [Ping timeout: 480 seconds]
rkanwal has joined #dri-devel
toolchains has joined #dri-devel
fahien has joined #dri-devel
danvet has quit [Ping timeout: 480 seconds]
pochu has quit [Remote host closed the connection]
itoral has quit [Remote host closed the connection]
pochu has joined #dri-devel
toolchains has quit [Ping timeout: 480 seconds]
mclasen has joined #dri-devel
toolchains has joined #dri-devel
toolchains has quit [Ping timeout: 480 seconds]
kts has joined #dri-devel
alatiera has quit [Ping timeout: 480 seconds]
Duke`` has quit [Ping timeout: 480 seconds]
toolchains has joined #dri-devel
simon-perretta-img has joined #dri-devel
simon-perretta-img__ has quit [Ping timeout: 480 seconds]
Duke`` has joined #dri-devel
adarshgm has quit [Ping timeout: 480 seconds]
kts has quit [Read error: Connection reset by peer]
<eric_engestrom>
lynxeye: networking issues like these are usually transient; I would just retry, and if it happens again you can ask on #freedesktop
kts has joined #dri-devel
<eric_engestrom>
I'm more worried about all the other failures in that pipeline honestly, they're all caused by the artifact storage system (is it minio?) crashing ("500 Internal Server Error")
<eric_engestrom>
(I did a retry of the job you linked and it instantly passed)
<eric_engestrom>
I also retried one of the jobs failing with the artifact upload and it also passed; I think you should just re-assign it to marge
ppascher has quit [Quit: Gateway shutdown]
toolchains has joined #dri-devel
<lynxeye>
eric_engestrom: Okay, I already did try to reassign to marge once and it failed with the same error. Just wanted to avoid an endless retry cycle if this is something that could be fixed somehow.
fahien has quit [Ping timeout: 480 seconds]
<eric_engestrom>
I retried the failing jobs without re-running the whole pipeline, and they all passed
<eric_engestrom>
I think it's a good reflex to ask instead of just retrying in a loop, but in this case I think you should just retry once more ^^
saurabhg has quit [Ping timeout: 480 seconds]
ppascher has joined #dri-devel
rkanwal has quit [Read error: Connection reset by peer]
rkanwal has joined #dri-devel
toolchains has quit [Ping timeout: 480 seconds]
icecream95 has quit [Ping timeout: 480 seconds]
JoniSt has joined #dri-devel
smpl has joined #dri-devel
<javierm>
tzimmermann: when you re-spin your "fbdev: Maintain device ownership with aperture helpers" series, feel free to also include my r-b to patch 02/11
<javierm>
tzimmermann: I think we are on the same page on that one too
<tzimmermann>
javierm, thanks. i'm working on it right now
<javierm>
tzimmermann: awesome
toolchains has joined #dri-devel
pochu has quit [Quit: leaving]
toolchains has quit [Ping timeout: 480 seconds]
fahien has joined #dri-devel
Jeremy_Rand_Talos has quit [Remote host closed the connection]
jewins has joined #dri-devel
MajorBiscuit has joined #dri-devel
toolchains has joined #dri-devel
srslypascal has quit [Quit: Leaving]
<dcz_>
okay, I have the proximate cause: my texture cannot be u8 because gl_ext_texture_integer is missing
<dcz_>
I got there by trying to make one in glium and tracing where it blew up
<pq>
hmm? oh, a Rust problem, ok. I can't help with those.
<dcz_>
nope, not a Rust problem. It's a GPU problem where it doesn't accept u8 textures
<dcz_>
now I have to find a way to feed it one regardless
<pq>
err
<dcz_>
glium blowing up is actually a feature, because I'm not half as good at error checking :P
<pq>
why would you try integer textures?
<dcz_>
because my Bayer data has 8 bits per pixel
<pq>
GL_RGBA, GL_UNSIGNED_BYTE should be good enough
<pq>
...which is 8-bit data, but not "integer" even though the raw data is uint8_8
<pq>
*uint8_t
<pq>
or just GL_RED to use one channel instead of gour
<pq>
*four
<dcz_>
thanks, I'll follow those leads
<pq>
the usual texture formats always imply conversion from whatever raw data format you have to floating-point 0.0 - 1.0
<pq>
I suppose integer textures are something else
Company has joined #dri-devel
kts has quit [Ping timeout: 480 seconds]
<JoniSt>
pq: The GL wiki says that unsigned GL_RED_INTEGER must be supported by all GL implementations, isn't that exactly what you need?
nicholasbishop has joined #dri-devel
<pq>
JoniSt, in GL 2.0?
toolchains has quit [Ping timeout: 480 seconds]
<dcz_>
it might, now I have to find out where to plug it
<pq>
JoniSt, or in GL ES 2.0?
<pq>
FWIW, I have never used that format, and I work with 8-bit data all the time
<JoniSt>
Regular GL 2.0 if the Wiki is anything to go by
<pq>
but why do you even want INTEGER?
nchery has joined #dri-devel
<JoniSt>
Hmm, GL_R8UI with usampler2d should work
kts has joined #dri-devel
<JoniSt>
Produces an uvec4 color=(red, 0, 0, 1)
<pq>
yeah, but what you get from using integer formats intead of the normal formats?
<lynxeye>
pq: slower execution and rounding issues when implementing the debayer filter
<lynxeye>
JoniSt: Seriously, you want a regular R8 texture for that and operate on fp in the shader.
<pq>
slower execution? because one has to only convert to float instead of convert to float and scale?
camus has quit []
saurabhg has joined #dri-devel
<lynxeye>
pq: First you likely don't have a integer RT when your don't have integer textures, so there would likely be some conversion needed. Also there are some braindamaged embedded GPUs that can do int math only at a lower rate than fp.
YuGiOhJCJ has quit [Quit: YuGiOhJCJ]
sdutt has joined #dri-devel
<pq>
lynxeye, oooh! I understood you exactly the opposite of what you said. :-)
<pq>
right, we agree then
toolchains has joined #dri-devel
<dcz_>
lynxeye: what's "integer RT"?
<JoniSt>
dcz_: Integer render target.
<pq>
render target, I presume, which I suppose refers to writing in integers instead of the usual [0, 1] floating-point
<dcz_>
this really underscores how much better compute shaders would have been for the purpose
toolchains has quit [Ping timeout: 480 seconds]
<pq>
I'm not so sure about that. GL shaders operate naturally in floating-point, but all textures and render targets use some non-floating-point pixel formats, and OpenGL converts automatically between the two.
nchery has quit [Ping timeout: 480 seconds]
<pq>
computing with integers is some new fancy stuff compared to that
<pq>
and hardware makes those conversion very efficient
<ishitatsuyuki>
(I saw you reassign, but just curious)
<daniels>
quite a loud failure I would've thought :P
<daniels>
I don't know why the artifact-upload errors are popping up
<ishitatsuyuki>
I looked at the wrong pipeline probably, I see thanks
mbrost has joined #dri-devel
<sravn>
javierm, tzimmermann: via patches will be merged tomorrow (no time left today). I have a few more patches in the works: s/DRIVER_LEGACY/DRIVER_DRI1 and s/CONFIG_DRM_LEGACY/CONFIG_DRM_DRI1 - will send them to dri-devel when they are rebased upon the via stuff.
srslypascal has joined #dri-devel
<sravn>
javierm: Did you suggest gpu/drm/dri1/ or gpu/dri1 as the new home for dri1 drivers? I did not look at this yet but may try to give it a spin
ahajda_ has quit []
alyssa has quit [Quit: leaving]
toolchains has joined #dri-devel
tzimmermann has quit [Quit: Leaving]
JohnnyonFlame has joined #dri-devel
toolchains has quit [Ping timeout: 480 seconds]
nicholasbishop has quit [Ping timeout: 480 seconds]
<dcz_>
pq: thanks for the help. The problem was some missing calls stemming from having no idea what I was doing
<dcz_>
turns out that that "unsafe" wasn't enforced just for show
simon-perretta-img has quit [Ping timeout: 480 seconds]
toolchains has joined #dri-devel
ybogdano is now known as Guest5102
ybogdano has joined #dri-devel
mbrost has quit [Ping timeout: 480 seconds]
Guest5101 has quit [Ping timeout: 480 seconds]
Guest5102 has quit [Ping timeout: 480 seconds]
egbert has quit [Ping timeout: 480 seconds]
egbert has joined #dri-devel
toolchains has quit [Ping timeout: 480 seconds]
egbert is now known as Guest5104
egbert has joined #dri-devel
Guest5104 has quit [Ping timeout: 480 seconds]
ngcortes has joined #dri-devel
iive has joined #dri-devel
egbert has quit [Ping timeout: 480 seconds]
nchery has quit [Ping timeout: 480 seconds]
nchery has joined #dri-devel
egbert has joined #dri-devel
egbert has quit [Ping timeout: 480 seconds]
egbert has joined #dri-devel
alarumbe has quit [Ping timeout: 480 seconds]
egbert is now known as Guest5106
egbert has joined #dri-devel
Guest5106 has quit [Ping timeout: 480 seconds]
egbert is now known as Guest5107
egbert has joined #dri-devel
kts has quit [Quit: Konversation terminated!]
Guest5107 has quit [Ping timeout: 480 seconds]
egbert is now known as Guest5109
egbert has joined #dri-devel
egbert is now known as Guest5111
Guest5109 has quit [Ping timeout: 480 seconds]
ngcortes has quit [Remote host closed the connection]
Guest5111 has quit [Ping timeout: 480 seconds]
egbert has joined #dri-devel
vliaskov has quit [Remote host closed the connection]
egbert is now known as Guest5114
egbert has joined #dri-devel
egbert is now known as Guest5115
egbert has joined #dri-devel
Guest5114 has quit [Ping timeout: 480 seconds]
Guest5115 has quit [Ping timeout: 480 seconds]
* Lyude
resolving merge conflict on drm-tip…
egbert is now known as Guest5117
<airlied>
Lyude: amd ones?
<airlied>
i thougbt i fixed that yesterday
<Lyude>
airlied: ahhhh, yeah that I'm still hitting it. if you want to try pushing your fixes again I can wait a moment
<Lyude>
*that. I'm still
mszyprow has quit [Ping timeout: 480 seconds]
<airlied>
Lyude: not near my pc yet
Guest5117 has quit [Ping timeout: 480 seconds]
srslypascal has quit [Remote host closed the connection]
srslypascal has joined #dri-devel
<Lyude>
airlied: gotcha, if I come up with a diff can you verify that it looks correct?
<airlied>
Lyude: can do, for vram mgr the resolution should look close to the one in fixes
rkanwal has quit [Ping timeout: 480 seconds]
<Lyude>
hm, looks like it's since been reverted so I assume it's probably supposed to be the opposite now (925b6e59138c ("Revert "drm/amdgpu: add drm buddy support to amdgpu""))
<Lyude>
ah, but then there's an actual fix in drm-misc-next-fixes
<javierm>
sravn: I did suggest drivers/gpu/drm/dri1/ but feel free to put it in other path if you think that makes more sense
<javierm>
sravn: great, if you can take on that it would be great since I'm busy right now with non-drm stuff