#
✕ nick:i509vcb
#asahi-gpu · 2023-09-19 20:08
i509vcbs/pipeline/shader
#asahi-gpu · 2023-09-19 20:07
i509vcbI'm not 100% familar with layered rendering in vk, how would you typically indicate that to the pipeline?
#asahi-gpu · 2023-09-19 19:49
i509vcb(alyssa said that was something still wip)
#asahi-gpu · 2023-09-19 19:47
i509vcbStill patiently waiting for collabora to fix image copies in vk meta of course
#asahi-gpu · 2023-09-19 18:25
i509vcbella-0: I think part of implementing CmdBlit with vk_meta requires layered rendering, so I guess I'll probably get a lot of that done
#asahi-gpu · 2023-09-19 15:51
i509vcbYeah I suspected that
#asahi-gpu · 2023-09-19 15:06
i509vcbI guess I'll try to see if the layered rendering stuff gives me working blits
#asahi-gpu · 2023-09-19 15:02
i509vcbneat, I guess I'm wondering what mistake I was making during rebase as to make every commit be "click incoming changes" in editor
#asahi-gpu · 2023-09-18 20:06
i509vcbblits I am currently working on
#asahi-gpu · 2023-09-18 20:06
i509vcbimage copies are a bit dependent on something else in mesa
#asahi-gpu · 2023-09-18 20:05
i509vcbRunning the dEQP-vk is probably a good sign of what needs work as well
#asahi-gpu · 2023-09-18 20:04
i509vcbCmdDrawIndirect, CmdDrawIndexedIndirect are probably two easy things, you just need to adjust the stuff for Draw and DrawIndexed, look at Dispatch and DispatchIndirect for reference. There are also extensions that can be implemented
#asahi-gpu · 2023-09-18 20:04
i509vcboh okay you came back
#asahi-gpu · 2023-09-18 20:01
i509vcb(And quit...)
#asahi-gpu · 2023-09-18 19:52
i509vcbUhh let me recall what is still low hanging fruit
#asahi-gpu · 2023-09-18 19:51
i509vcbthere is definitely some stuff that can be done
#asahi-gpu · 2023-09-18 19:20
i509vcbAlthough 0-3 feels a bit meaningless, is it expected to map to VkQueueGlobalPriorityKHR values or something else?
#asahi-gpu · 2023-09-18 19:06
i509vcbdrm_asahi_queue_create:: priority seems to cover that?
#asahi-gpu · 2023-09-18 16:12
i509vcbSimilar question for VK_KHR_global_priority as well
#asahi-gpu · 2023-09-18 16:11
i509vcbWondering about EGL_IMG_context_priority, I'm guess the kernel uabi would need to allow specifying a priority for the drm device?
#asahi-gpu · 2023-09-17 18:36
i509vcbwhere my current branch is effectively agxv/main
#asahi-gpu · 2023-09-17 18:36
i509vcbgit merge-base --fork-point upstream/main
#asahi-gpu · 2023-09-17 18:36
i509vcbthat commit hash came fro
#asahi-gpu · 2023-09-17 18:36
i509vcbgit rebase --onto upstream/asahi/main 7bdc188a9453fc2a513c2ffd386c858aa958e839
#asahi-gpu · 2023-09-17 18:36
i509vcbjannau: I tried the following and I still seem to get a lot of "choose the incoming change"
#asahi-gpu · 2023-09-16 04:25
i509vcbwhich agxv really isn't supposed to be touching
#asahi-gpu · 2023-09-16 04:24
i509vcbI feel I am doing something wrong if the vast majority of the conflicts are occuring in the gallium driver
#asahi-gpu · 2023-09-16 03:48
i509vcb2. rebase --onto asahi/main
#asahi-gpu · 2023-09-16 03:48
i509vcb1. Find the point at which asahi/main and agxv/main forked
#asahi-gpu · 2023-09-16 03:48
i509vcbI thought it would be something like this:
#asahi-gpu · 2023-09-16 03:48
i509vcbI'm not sure if I am doing a really naive rebase
#asahi-gpu · 2023-09-15 23:55
i509vcbAlthough I imagine there is a better way than mass cherry-picking the entire branch from where it diverges
#asahi-gpu · 2023-09-15 23:54
i509vcbI can give it a try
#asahi-gpu · 2023-09-15 20:07
i509vcbWell if you consider a tile to be a workgroup of sorts?
#asahi-gpu · 2023-09-15 20:04
i509vcbalyssa: the sr2 response is interesting, so the workgroup index in agx is not strictly a compute thing?
#asahi-gpu · 2023-09-13 19:12
i509vcbHope you feel better soon :)
#asahi-gpu · 2023-09-13 18:48
i509vcbalyssa: do you plan to rebase agxv onto asahi/main eventually? Ideally so I can try to test the layer id stuff
#asahi-gpu · 2023-09-13 01:45
i509vcbhopefully G16 is just on time for agxv being done or beyond done
#asahi-gpu · 2023-09-13 01:44
i509vcbalyssa: something something 10x-30x or something like that more iphones sold per year than macs
#asahi-gpu · 2023-09-08 23:17
i509vcbI'll peek there I guess
#asahi-gpu · 2023-09-08 23:13
i509vcb(trying to get blits to at least try to run)
#asahi-gpu · 2023-09-08 23:13
i509vcbalyssa: any pointers for implementing load_layer_id on agxv? I noticed a few drivers involve creating an nir_variable but I suspect this might not be right for agxv
#asahi-gpu · 2023-09-08 13:51
i509vcbNo idea, I think it's either max memory allocation size or UINT64_MAX
#asahi-gpu · 2023-09-08 13:50
i509vcbOkay then I guess I just ignore the above
#asahi-gpu · 2023-09-08 13:50
i509vcbmaxBufferSize from https://registry.khronos.org/vulkan/specs/1.3-extensions/man/html/VkPhysicalDeviceMaintenance4PropertiesKHR.html
#asahi-gpu · 2023-09-08 13:44
i509vcblina: Is the mentioned theoretical 4GiB maximum buffer size related to the usage of the buffer or is it more of a strict limit? (Want to understand this before blindly updating the limit in agxv)
#asahi-gpu · 2023-09-07 02:44
i509vcbOn that topic what is PBE?
#asahi-gpu · 2023-09-07 02:40
i509vcbhmm that wants load_layer_id, random crapshoot at the cmdbuf that appears to be something related to AGX_PBE?
#asahi-gpu · 2023-09-07 02:03
i509vcb(I'm guessing yes because I saw the blit stuff already is there)
#asahi-gpu · 2023-09-07 01:54
i509vcbalyssa: was the plan to implement vkCmdBlitImage with vk meta like image and buffer copies?

plus ›