#
#asahi-gpu · 2023-12-24 17:20
chaos_princessam creating the bo in the same way as geometry count buffer is done. Any ideas on what i might be doing wrong?
#asahi-gpu · 2023-12-24 17:20
chaos_princessim trying to do some stuff in mesa, and added a field to `struct agx_geometry_params`, GLOBAL(uchar), and trying to read from it's contents in shader code. However instead of loading data from that buffer, the shader appears to dereference it one extra time, and fault (filled the buffer with NANs, the fault address looks very much like a nan bit pattern.) the nir seems sane, and i
#asahi-gpu · 2023-12-21 18:26
millenialhackerpk
#asahi-gpu · 2023-12-18 17:31
FerryPieHey, I was wondering if it's a good idea to use syzkaller to fuzz the asahi drivers ? Does anybody have any experience in something related to this ?
#asahi-gpu · 2023-12-17 22:26
ChaosPrincesslina, alyssa: is there a way to find out what specific instruction the gpu faulted on?
#asahi-gpu · 2023-12-17 18:20
kaazooSomebody was so kind to document the steps to build qemu with ratagaba: https://linaro.atlassian.net/wiki/spaces/ORKO/pages/28985622530/Building+QEMU+with+virtio-gpu+and+rutabaga_gfx
#asahi-gpu · 2023-12-17 18:19
kaazooHi, did anybody try out virtio-gpu rutabaga with qemu to have fast OpenGL in guest VMs? I know slp was working on a microVM approach. But what about generic VMs? Using virglrender exposes hardware OpenGL support to the VM, but it's quite slow when comparing glmark results. (~ 80 points in VM vs. ~6000 points in host)
#asahi-gpu · 2023-12-04 10:39
rosefromthedeadoh neat, ty
#asahi-gpu · 2023-12-04 00:47
alyssahttps://vulkan.gpuinfo.org/displaydevicelimit.php?name=maxDescriptorSetStorageBuffers&platform=linux
#asahi-gpu · 2023-12-04 00:45
alyssaso whatever we set it to will be arbitrary
#asahi-gpu · 2023-12-04 00:45
alyssafor maxPerStageDescriptorStorageBuffers, there is no hw limit
#asahi-gpu · 2023-12-04 00:44
alyssarosefromthedead: agxv isn't running real sutff yet so.. "they weren't"
#asahi-gpu · 2023-12-02 15:27
rosefromthedeadi have a wgpu program wanting 8 maxPerStageDescriptorStorageBuffers but agxv only advertises 4. the metal backend for wgpu hardcodes 31 for this limit, so is 4 a driver limitation or a guess?
#asahi-gpu · 2023-12-02 15:26
rosefromthedeadhow were the limits in agxv_limits.h determined?
#asahi-gpu · 2023-11-26 22:16
ella-0i509vcb: https://gitlab.freedesktop.org/Ella-0/mesa/-/blob/agxv/meta/randomcrapfori509vcbgoodluck/src/asahi/vulkan/agxv_format.c?ref_type=heads#L214
#asahi-gpu · 2023-11-26 22:14
ella-0my vk meta code should already deal with that
#asahi-gpu · 2023-11-26 20:12
i509vcb(Ella did give me the vk-meta code)
#asahi-gpu · 2023-11-26 20:08
i509vcbIf I disable all the compressed formats then its Pass: 8170, Fail: 708, Crash: 2, Flake: 1 for the copy_and_blit.image_to_image tests
#asahi-gpu · 2023-11-26 20:05
i509vcbwhen a meta shader is built agx_nir_lower_tilebuffer seems to have problems when emitting code with a compressed format (like PIPE_FORMAT_ETC2_R11_SNORM for where agx_tilebuffer_physical_format returns PIPE_FORMAT_NONE). I'm guessing that these types of formats need some sort of emulation for reads and writes?
#asahi-gpu · 2023-11-24 13:49
alyssaignore it
#asahi-gpu · 2023-11-24 13:49
alyssathe intrinsic is an etnaviv weirdness that i was opposed to adding because it's unnecessary.
#asahi-gpu · 2023-11-24 13:49
alyssaas for txs, it's a texture op
#asahi-gpu · 2023-11-24 13:49
alyssa#2 has slight performance implications but meh
#asahi-gpu · 2023-11-24 13:48
alyssa2. just make textures bindless always, and lower texture derefs to bindless textures directly and don't bother with the texture state registers
#asahi-gpu · 2023-11-24 13:47
alyssa1. what we do in the gl driver. agx_nir_lower_bindings.c has the logic
#asahi-gpu · 2023-11-24 13:47
alyssai509vcb: two options
#asahi-gpu · 2023-11-24 02:51
i509vcbSeems like 2 ways to do the same thing?
#asahi-gpu · 2023-11-24 02:51
i509vcbAlso there seems to be a texture size intrinsic and texoo_txs?
#asahi-gpu · 2023-11-24 02:50
i509vcbI guess while you are here, I'm struggling to understand how I'd change a binded access (if that is even the right term) to a bindless for the bindless image stuff the compiler wants
#asahi-gpu · 2023-11-24 02:48
i509vcbYeah I mentioned it yesterday to ella and she said she has patched
#asahi-gpu · 2023-11-24 02:47
alyssaella-0: has vkmeta patches i think
#asahi-gpu · 2023-11-24 02:47
alyssai509vcb: that assert is because format mutation isn't hnadled correctly
#asahi-gpu · 2023-11-23 17:18
i509vcbodd... Commenting out some stuff in agxv while trying to do the proper bindless image lowering I can crash my entire session when running vkcube
#asahi-gpu · 2023-11-21 23:20
i509vcbabout 75% of the draw tests in the dEQP that are supported crash in vk_meta on that one assert
#asahi-gpu · 2023-11-21 23:19
i509vcbregarding the comment in vk_meta, what specifically was broken? https://gitlab.freedesktop.org/asahi/mesa/-/blob/agxv/main/src/vulkan/runtime/vk_meta_copy.c?ref_type=heads#L406
#asahi-gpu · 2023-11-21 03:40
i509vcbVK_QUEUE_GLOBAL_PRIORITY_REALTIME_KHR does exist which is technically higher than all values, which could make sense if everything else is shifted down by 1
#asahi-gpu · 2023-11-21 03:39
i509vcbGallium and agxv currently use priority = 1
#asahi-gpu · 2023-11-21 03:38
i509vcbFor drm_asahi_queue_create::priority: 0 = EGL_CONTEXT_PRIORITY_HIGH_IMG/VK_QUEUE_GLOBAL_PRIORITY_HIGH_KHR, 1 = EGL_CONTEXT_PRIORITY_MEDIUM_IMG/VK_QUEUE_GLOBAL_PRIORITY_MEDIUM_KHR, 2 = EGL_CONTEXT_PRIORITY_LOW_IMG/VK_QUEUE_GLOBAL_PRIORITY_LOW_KHR, 3 = ????
#asahi-gpu · 2023-11-21 02:49
i509vcblina: timelines appear to work: https://gitlab.freedesktop.org/asahi/mesa/-/merge_requests/219
#asahi-gpu · 2023-11-12 07:45
linaIf you need to consider the "nothing submitted on this sync" case as unsignaled, you need to track that in userspace (that rule is for sync objects from the point of view of the kernel interface).
#asahi-gpu · 2023-11-12 07:44
linai509vcb: Sync objects are always either signaled or will be signaled by an ongoing GPU submission. It's not legal for a sync object to be non-signaled for an indefinite amount of time, which is why a "fresh" sync object is always signaled on creation. It becomes unsignaled when you use it to submit work, until that work completes.
#asahi-gpu · 2023-11-12 06:21
i509vcbsorry for the annoyance
#asahi-gpu · 2023-11-12 06:21
i509vcbHmm more discussion, it seems to be a misunderstanding of EGL
#asahi-gpu · 2023-11-12 02:19
Lestat9Become a ghost person, get in a situation where the mind gives the perceived death signal, watch a scary movie alone in the dark,when it does just ignore it or pay no attention to it. -Copyright: TXu 2-037-947- Rest the scientists- ghostpeople@gmx.com
#asahi-gpu · 2023-11-12 02:15
Lestat9Become a ghost person, get in a situation where the mind gives the perceived death signal, watch a scary movie alone in the dark,when it does just ignore it or pay no attention to it. -Copyright: TXu 2-037-947- Rest the scientists- ghostpeople@gmx.com
#asahi-gpu · 2023-11-12 02:11
Lestat9Become a ghost person, get in a situation where the mind gives the perceived death signal, watch a scary movie alone in the dark,when it does just ignore it or pay no attention to it. -Copyright: TXu 2-037-947- Rest the scientists- ghostpeople@gmx.com
#asahi-gpu · 2023-11-12 02:07
Lestat9Become a ghost person, get in a situation where the mind gives the perceived death signal, watch a scary movie alone in the dark,when it does just ignore it or pay no attention to it. -Copyright: TXu 2-037-947- Rest the scientists- ghostpeople@gmx.com
#asahi-gpu · 2023-11-12 00:18
j`eyhttps://docs.kernel.org/gpu/drm-mm.html#host-side-wait-on-syncobjs might help?
#asahi-gpu · 2023-11-12 00:12
i509vcbI have no idea what semantics that should have
#asahi-gpu · 2023-11-12 00:12
i509vcbWould be great if I could find documentation for drmSyncobjWait somewhere

plus ›