#
✕ nick:i509vcb
#asahi-gpu · 2024-02-15 02:02
i509vcbmismatched ppp size?
#asahi-gpu · 2024-02-15 01:52
i509vcbvkcube seems to die in the compiler
#asahi-gpu · 2024-02-15 01:50
i509vcblet me smoke test vkcube quickly
#asahi-gpu · 2024-02-15 01:50
i509vcbrebase is done
#asahi-gpu · 2024-02-15 01:50
i509vcbalyssa: https://gitlab.freedesktop.org/i509VCB/mesa/-/commits/rebase-10010
#asahi-gpu · 2024-02-15 01:30
i509vcbOkay if you are fine with the squashing
#asahi-gpu · 2024-02-15 00:53
i509vcbI assume it's not VR related?
#asahi-gpu · 2024-02-15 00:52
i509vcbhmm what is TEXTURE_TYPE_XR? Not related to the rebase, just don't know
#asahi-gpu · 2024-02-15 00:16
i509vcbalthough that's quite involved
#asahi-gpu · 2024-02-15 00:16
i509vcbI think I got the rebase done, just going through each commit and fixing compiler errors I see along the way
#asahi-gpu · 2024-02-14 22:09
i509vcblittle confused why some intel compiler stuff is being built with just asahi and swrast enabled for builds but not my problem
#asahi-gpu · 2024-02-14 20:12
i509vcbAlthough we arent't exactly doing great so far given there is no nice render pass stuff for mesa to do for us
#asahi-gpu · 2024-02-14 20:11
i509vcbday 1 ESO oof
#asahi-gpu · 2024-02-14 16:00
i509vcb(well 2d fully passed)
#asahi-gpu · 2024-02-14 16:00
i509vcbI'm tempted to just ship what's working so far with 2d and 1d copies and fix 3d copies later
#asahi-gpu · 2024-02-14 15:54
i509vcbWell vulkan needs more work anyways so I guess I'm still dying in 3d copy land
#asahi-gpu · 2024-02-14 15:54
i509vcbOh damn gl 4.6 snuck up on me lol
#asahi-gpu · 2024-02-11 21:10
i509vcbthat's probably more optimal using the ANE which is being implemented elsewhere
#asahi-gpu · 2024-02-11 21:09
i509vcbLLaMA models are probably not going to use OpenGL or Vulkan compute however
#asahi-gpu · 2024-02-11 19:59
i509vcbVulkan is a whole other story since the rest of the driver needs to be implemented still, but I did get compute working in some of the dEQP tests there
#asahi-gpu · 2024-02-11 19:58
i509vcbOpenCL no idea
#asahi-gpu · 2024-02-11 19:58
i509vcbAnd GL_ARB_compute_shader is advertised on the OpenGL side
#asahi-gpu · 2024-02-11 19:57
i509vcbI have ES 3.1 on M2 so if you just need GLES compute that would work fine
#asahi-gpu · 2024-01-21 01:15
i509vcbErr more of an imply
#asahi-gpu · 2024-01-21 01:13
i509vcbVUID-VkCopyImageInfo2-dstImage-04444 seems to state that I should be using the depth and dstOffset for layerCount and baseArrayLayer in the 2D Array view of the destination 3d view
#asahi-gpu · 2024-01-20 23:53
i509vcbFrom what I can tell, image views add the result of ail_get_layer_offset_B to the base address of the image to get a view that starts at some layer
#asahi-gpu · 2024-01-20 23:46
i509vcbalthough it's all layer 0 writes still
#asahi-gpu · 2024-01-20 23:46
i509vcbI do tell the fragment shader to always have a layered tib layout and lower with layer_id_sr
#asahi-gpu · 2024-01-20 23:45
i509vcb4. This sounds like forcing 2D array for everything? I've tried this a few times but it ends in breaking everything
#asahi-gpu · 2024-01-20 23:45
i509vcb2. Seems like vk-meta does do this and creates 2d views of the 3d image for single layer copies, although why these still fail is still in the air
#asahi-gpu · 2024-01-20 19:25
i509vcb(Although the problem with forcing the tib layout to always be layered is that everything kind of breaks lol)
#asahi-gpu · 2024-01-20 19:25
i509vcbbut assuming always layered means that the issue of Arrays not being allowed to be linear becomes a problem
#asahi-gpu · 2024-01-20 19:24
i509vcbI did notice the EOT shader was never initialized as layered for some reason, I guess that checking if the layerCount > 1 is pretty much wrong
#asahi-gpu · 2024-01-20 19:23
i509vcbIf the texture layout is linear, Depth is ignored? (And Depth(Linear) is used instead)?
#asahi-gpu · 2024-01-09 03:26
i509vcbarray copies start at .0032, up to ..0047 https://usercontent.irccloud-cdn.com/file/W8l9QzTm/dEQP-VK.api.copy_and_blit.core.image_to_image.3d_images.2d_to_3d_whole.zip
#asahi-gpu · 2024-01-09 03:18
i509vcbthis shouldn't be possible?
#asahi-gpu · 2024-01-09 03:17
i509vcb ...
#asahi-gpu · 2024-01-09 03:17
i509vcb Depth: 256
#asahi-gpu · 2024-01-09 03:17
i509vcbTexture
#asahi-gpu · 2024-01-09 03:17
i509vcb Dimension: 2D
#asahi-gpu · 2024-01-09 03:12
i509vcbUsing the dstSubresource just like the src image view in vk-meta didn't seem to change anything either: https://gitlab.freedesktop.org/i509VCB/mesa/-/commit/85f680952dc24685d39c049cf5ba968643a5a0ee#ccd82ce6d5431b9371915af7969698022e62444e_312_308
#asahi-gpu · 2024-01-09 03:10
i509vcbOn possible issues, the PBE and TEXTURE descriptors look just like gallium's
#asahi-gpu · 2024-01-09 03:10
i509vcbdEQP-VK.api.copy_and_blit.core.image_to_image.3d_images.2d_to_3d_whole does a per layer copy (only copying one layer at a time) but seems to just only ever write on the first layer?
#asahi-gpu · 2024-01-08 23:51
i509vcbthe gpu channel is generally pretty quiet until I stumble around the vulkan driver
#asahi-gpu · 2024-01-08 23:50
i509vcbit's probably not going to be here (the display and gpu hardware are seperate units), better luck probably in #asahi or #asahi-dev, but I imagine those questions appear every day, so probably just need to wait
#asahi-gpu · 2024-01-08 23:47
i509vcbqvy: depends what you mean by active
#asahi-gpu · 2024-01-08 23:46
i509vcbs/khronos/lunarg
#asahi-gpu · 2024-01-08 23:45
i509vcbsadly the khronos debug layers aren't packaged on fedora (api trace), so debugging is a little more painful
#asahi-gpu · 2024-01-08 23:44
i509vcbthis is certainly odd
#asahi-gpu · 2024-01-08 23:43
i509vcbalthough some of the 3d blits started working when I did turn on layered rendering

plus ›