AlexIndustrial/mesa
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
142,981 Commits 1 Branch 0 Tags
c19dcec60474ef4eaf7a46ad11ce29fe2140eac2
Commit Graph

142981 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Iago Toral Quiroga c19dcec604 v3dv: implement interaction of queries with multiview 
When multiview is enabled, queries must begin and end in the
same subpass and N consecutive queries are implicitly used,
where N is the number of views enabled in the subpass.
Implementations decide how results are split across queries.

In our case, only one query is really used, but we still need
to flag all N queries as available by the time we flag the one
we use so that the application doesn't receive unexpected errors
when trying to retrieve values from them.

Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12034>
 2021-07-27 07:31:32 +00:00
Iago Toral Quiroga 374215de1a v3dv: fix query error handling 
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12034>
 2021-07-27 07:31:32 +00:00
Iago Toral Quiroga b1620caaaf v3dv: track first and last subpass that use a view index 
When multiview is enabled, we no longer care about when a particular
attachment is first or last used in a render pass, since not all views
in the attachment will meet that criteria. Instead, we need to track
each individual view (layer) in each attachment and emit our stores,
loads and clears accordingly.

Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12034>
 2021-07-27 07:31:32 +00:00
Iago Toral Quiroga 34724bafa4 v3dv: skip processing tiles for layers that are not in the view mask 
If a multiview subpass doesn't use a particular layer then we can ignore
that layer completely.

Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12034>
 2021-07-27 07:31:31 +00:00
Iago Toral Quiroga 32cb8809aa v3dv: use correct number of layers for multiview 
The Vulkan spec states that when multiview is enabled the number of
layers in the framebuffer is set to one and that each attachment
must then have at least as many layers as referenced by view masks
in the subpasses in which is used.

Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12034>
 2021-07-27 07:31:31 +00:00
Iago Toral Quiroga f46f5eac7c v3dv: don't merge subpasses with different view masks 
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12034>
 2021-07-27 07:31:31 +00:00
Iago Toral Quiroga 5fd55ab16d v3dv: broadcast multiview draw commands 
We implement multiview by replicating draw commands for all enabled
views and setting a command buffer state for the currently active
view we are broadcasting to.

Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12034>
 2021-07-27 07:31:31 +00:00
Iago Toral Quiroga d5acae3206 broadcom/compiler: implement nir_intrinsic_load_view_index 
This is used for multiview's gl_ViewIndex built-in.

Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12034>
 2021-07-27 07:31:31 +00:00
Iago Toral Quiroga 5872c7ca7b v3dv: inject a custom passthrough geometry shader for multiview pipelines 
This allows us to use layered rendering to broadcast draw calls to the
appropriate views (layers).

Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12034>
 2021-07-27 07:31:31 +00:00
Iago Toral Quiroga 9da26af074 v3dv: move all our NIR pre-processing to preprocess_nir 
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12034>
 2021-07-27 07:31:31 +00:00
Iago Toral Quiroga f5e67e1ce7 v3dv: store multiview info in our render pass data 
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12034>
 2021-07-27 07:31:31 +00:00
Iago Toral Quiroga dc86e032cb v3dv: drop unused parameters 
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12034>
 2021-07-27 07:31:31 +00:00
Samuel Pitoiset 72f55cf7c4 radv: implement VK_EXT_shader_atomic_float2 
Some floating atomic instructions are not available on GFX8-9.
No LLVM support.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12060>
 2021-07-27 08:44:36 +02:00
Samuel Pitoiset 6694c37ea0 aco: implement VK_EXT_shader_atomic_float2 
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12060>
 2021-07-27 08:44:31 +02:00
Chia-I Wu 71912de1d2 venus: log more WSI messages 
They can be useful before resorting to gdb or perfetto.

Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
Reviewed-by: Yiwei Zhang <zzyiwei@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12073>
 2021-07-27 06:21:47 +00:00
Eduardo Lima Mitev ee3495e465 turnip: Add support for VK_VALVE_mutable_descriptor_type 
v1.  Hyunjun Ko <zzoon@igalia.com>
- Add to hanlde VK_DESCRIPTOR_POOL_CREATE_HOST_ONLY_BIT_VALVE
- Don't support VK_DESCRIPTOR_TYPE_INPUT_ATTACHMENT and
VK_DESCRIPTOR_TYPE_COMBINED_IMAGE_SAMPLER

v2.  Hyunjun Ko <zzoon@igalia.com>
- Fix some indentations and nitpicks.
- Add the extension to features.txt

v3.  Hyunjun Ko <zzoon@igalia.com>
- Remove unnecessary asserts.

Signed-off-by: Eduardo Lima Mitev <elima@igalia.com>
Signed-off-by: Hyunjun Ko <zzoon@igalia.com>
Reviewed-by: Danylo Piliaiev <dpiliaiev@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9322>
 2021-07-27 04:53:02 +00:00
Chia-I Wu 567bbfd56b venus: suballocate memory in more cases 
When a dedicated allocation is not required, ignore it and suballocate.

Fixes dEQP-VK.api.invariance.random.

Cc: 21.2 mesa-stable
Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
Reviewed-by: Yiwei Zhang <zzyiwei@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12072>
 2021-07-27 03:53:52 +00:00
Chia-I Wu f06b0df741 venus: clean up vn_AllocateMemory 
Mainly to add vn_device_memory_should_suballocate.

Cc: 21.2 mesa-stable
Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
Reviewed-by: Yiwei Zhang <zzyiwei@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12072>
 2021-07-27 03:53:52 +00:00
Roland Scheidegger cac5711d43 llvmpipe: fix nir dot products (fsum op) 
When the dot product uses a source which can be optimized to a scalar,
after a bunch of nir optimization steps the source to fsum will be a scalar
with a x replicate swizzle. Hence nir_src_num_components is just 1 and the
fsum was just a no-op which is not correct. Arguably this could be optimized
a bit better, but just determine the number of addends by using nir_op_infos
instead (the operand fetch was fixed already by 39a938ecf4 doing the same).

Fixes: 4eb0475b5a ("gallivm/nir: add fsum support")
Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12066>
 2021-07-27 00:50:55 +00:00
Mike Blumenkrantz 6691a48bcc zink: EXT_vertex_input_dynamic_state 
this eliminates vertex attributes from the pipeline state, massively
deduplicating the number of pipelines needed

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12009>
 2021-07-27 00:36:27 +00:00
Mike Blumenkrantz 907e4a0f02 zink: improve oom flushing 
flush on 30k works and also check oom_flush

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12010>
 2021-07-27 00:28:02 +00:00
Mike Blumenkrantz 9460c350b1 zink: merge draw_count and compute_count, move to batch struct 
one fewer deref and now it's a single counter

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12010>
 2021-07-27 00:28:02 +00:00
Mike Blumenkrantz 48473ec967 zink: more explicitly check shader stages during compile 
this is a bit more obvious

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12013>
 2021-07-27 00:14:44 +00:00
Mike Blumenkrantz 7665e656c1 zink: remove no longer used internal resource function 
Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12008>
 2021-07-26 19:52:49 -04:00
Mike Blumenkrantz 3137795e38 zink: replace some direct batch_usage calls with resource abstractions 
Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12008>
 2021-07-26 19:52:47 -04:00
Mike Blumenkrantz 12417c030b zink: use new resource batch usage utils for is_resource_busy 
Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12008>
 2021-07-26 19:52:00 -04:00
Mike Blumenkrantz 6bda555cc0 zink: simplify some dumb code in invalidate_buffer 
Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12008>
 2021-07-26 19:51:53 -04:00
Mike Blumenkrantz 79ccc17da5 zink: use resource batch usage helpers in invalidate_buffer() 
Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12008>
 2021-07-26 19:48:07 -04:00
Mike Blumenkrantz e5d6ce6e5e zink: collapse a conditional in zink_batch_resource_usage_set() 
Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12008>
 2021-07-26 19:48:07 -04:00
Mike Blumenkrantz 13d732ec5b zink: add some resource util functions for batch usage 
Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12008>
 2021-07-26 19:48:07 -04:00
Mike Blumenkrantz 0a2b03376d zink: force batch completion check on query result 
non-timeline drivers have no screen-based method of "checking" batch
completion, so the context method has to be used here to avoid an infinite
loop

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12008>
 2021-07-26 19:48:07 -04:00
Mike Blumenkrantz 5bab4cc546 zink: add screen function for checking usage completion 
Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12008>
 2021-07-26 19:48:07 -04:00
Rob Clark 53c39f710f freedreno/a6xx: Add missing PC_CCU_INVALIDATE_x 
The kernel normally inserts these between submits, but when we merge
submits in userspace we need to add them ourselves.

Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/5114
Fixes: cccdc513e3 ("freedreno/drm/sp: Implement deferred submit merging")
Signed-off-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12049>
 2021-07-26 21:11:35 +00:00
Mike Blumenkrantz 1dc7b9b08b zink: use array size in spirv bo length calculations 
I don't know why I was dropping this, but doing so breaks drivers that have
optimization passes based on the lengths of these variables

Fixes: c1cdf30a11 ("zink: apply Delete All The Code methodology to the ubo/ssbo variables")

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12070>
 2021-07-26 20:59:07 +00:00
Roman Stratiienko ed98e54f19 kmsro: Add 'kirin' driver support 
Kirin is used by hisilicon SOCs (Hikey boards, etc.)
Driver is available in the mainline kernel [1]

[1]: https://elixir.bootlin.com/linux/v5.13/source/drivers/gpu/drm/hisilicon/kirin/kirin_drm_ade.c#L928

Signed-off-by: Roman Stratiienko <r.stratiienko@gmail.com>
Reviewed-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11755>
 2021-07-26 20:39:17 +00:00
Dave Airlie 5bf6ec31cc crocus/gen4: restrict memcpy mapping to gen5 
This is due to gen4 + 4.5 having some rather strange swizzling
that we can't actually detect properly in userspace

Fixes: f3630548f1 ("crocus: initial gallium driver for Intel gfx 4-7")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12058>
 2021-07-26 20:28:37 +00:00
Thong Thai 856e2b7529 frontends/va: change to per-layer rate control 
Allows for each layer in a Scalable Video Coding (SVC) video to have its
own rate control setting, when encoding H.264.

Signed-off-by: Thong Thai <thong.thai@amd.com>
Reviewed-by: Boyuan Zhang <Boyuan.Zhang@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11850>
 2021-07-26 15:06:22 -04:00
Thong Thai c2695823fd frontends/omx: change rate ctrl struct to array 
Signed-off-by: Thong Thai <thong.thai@amd.com>
Reviewed-by: Boyuan Zhang <Boyuan.Zhang@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11850>
 2021-07-26 15:06:17 -04:00
Thong Thai 687bbe2270 radeon/vcn/enc: change to per-temporal layer rate control 
Signed-off-by: Thong Thai <thong.thai@amd.com>
Reviewed-by: Boyuan Zhang <Boyuan.Zhang@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11850>
 2021-07-26 15:06:13 -04:00
Thong Thai 65ad80f1d8 radeon/vce: change rate ctrl struct to array 
Signed-off-by: Thong Thai <thong.thai@amd.com>
Reviewed-by: Boyuan Zhang <Boyuan.Zhang@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11850>
 2021-07-26 15:06:09 -04:00
Thong Thai 21da3be7a2 r600: change rate ctrl struct to array 
Signed-off-by: Thong Thai <thong.thai@amd.com>
Reviewed-by: Boyuan Zhang <Boyuan.Zhang@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11850>
 2021-07-26 15:06:05 -04:00
Thong Thai 1e0d512595 gallium: change rate ctrl struct to array 
Signed-off-by: Thong Thai <thong.thai@amd.com>
Reviewed-by: Boyuan Zhang <Boyuan.Zhang@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11850>
 2021-07-26 15:06:00 -04:00
Thong Thai ec53e519c3 frontends/va: handle h264 num_temporal_layers for SVC encoding 
Allows for the number of temporal layers to be specified when encoding
Scalable Video Coding (SVC) H.264 videos.

Signed-off-by: Thong Thai <thong.thai@amd.com>
Reviewed-by: Boyuan Zhang <Boyuan.Zhang@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11850>
 2021-07-26 15:05:56 -04:00
Thong Thai 8236320a0b radeonsi: enable H.264 temporal encoding support for VCN 
Signed-off-by: Thong Thai <thong.thai@amd.com>
Reviewed-by: Boyuan Zhang <Boyuan.Zhang@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11850>
 2021-07-26 15:05:48 -04:00
Thong Thai 51935d594e radeon/vcn/enc: H.264 SVC encode 
Implement H.264 temporal, Scalable Video Coding (SVC) for VCN devices by
sending the required parameters to the firmware, and creating H.264 NALU
prefix and SEI scalability_info headers.

Signed-off-by: Thong Thai <thong.thai@amd.com>
Reviewed-by: Boyuan Zhang <Boyuan.Zhang@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11850>
 2021-07-26 15:05:43 -04:00
Thong Thai 41f4b69354 gallium: update h264 struct to track temporal layers 
Signed-off-by: Thong Thai <thong.thai@amd.com>
Reviewed-by: Boyuan Zhang <Boyuan.Zhang@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11850>
 2021-07-26 15:05:39 -04:00
Thong Thai 96c581e897 frontends/va: check number of temporal layers supported by encoder 
Checks the encoder to determine the number of temporal layers supported,
and returns max_num_temporal_layers_minus1, along with setting the
corresponding control flag if multiple layers are supported.

Signed-off-by: Thong Thai <thong.thai@amd.com>
Reviewed-by: Boyuan Zhang <Boyuan.Zhang@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11850>
 2021-07-26 15:05:33 -04:00
Thong Thai a23beb9c43 gallium: add temporal layers cap enum 
Determine the number of temporal layers the encoder supports. Used for
encoding Scalable Video Coding (SVC) videos.

Signed-off-by: Thong Thai <thong.thai@amd.com>
Reviewed-by: Boyuan Zhang <Boyuan.Zhang@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11850>
 2021-07-26 15:05:26 -04:00
Danylo Piliaiev b45cddda18 tu: handle half-reg fs outputs 
This would allow to enable translation of RelaxedPrecision spirv
variable decorator into mediump which for us means fp16.

Signed-off-by: Danylo Piliaiev <dpiliaiev@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12067>
 2021-07-26 16:25:14 +00:00
Karol Herbst 1387d1d411 nv30: fix emulated vertex index buffers 
We ended up applying the offset twice. Quite embarrassing.

This fixes a bunch of vertex shader related issues like the gnome desktop
is less broken and probably a lot of other applications.

Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/5061
Cc: mesa-stable
Signed-off-by: Karol Herbst <kherbst@redhat.com>
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12054>
 2021-07-26 15:39:25 +00:00