2d6a0a8620
gallivm: add subgroup ballot support
...
Reviewed-by: Roland Scheidegger <sroland@vmware.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9645 >
2021-04-07 19:56:17 +00:00
7b3073ad44
gallivm: add subgroup reduction + in/ex scan support
...
Reviewed-by: Roland Scheidegger <sroland@vmware.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9645 >
2021-04-07 19:56:17 +00:00
2e1266d1fe
gallivm: add subgroup elect intrinsic support.
...
Reviewed-by: Roland Scheidegger <sroland@vmware.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9645 >
2021-04-07 19:56:17 +00:00
52415cd39e
gallivm: add subgroup system values support
...
This just adds support for retrieving subgroup system values.
Reviewed-by: Roland Scheidegger <sroland@vmware.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9645 >
2021-04-07 19:56:17 +00:00
1b8145eee9
gallivm: move get_flt_bld to header.
...
Rewrite a little as we have to add 16-bit support later in life
Reviewed-by: Roland Scheidegger <sroland@vmware.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9645 >
2021-04-07 19:56:17 +00:00
2e23108274
gallivm: add subgroup vote 64-bit and feq support.
...
These are both required for vulkan subgroup support.
Reviewed-by: Roland Scheidegger <sroland@vmware.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9645 >
2021-04-07 19:56:17 +00:00
8a18ed018b
lavapipe: fix missing protected memory properties
...
This is needed for vk 1.1
Reviewed-by: Roland Scheidegger <sroland@vmware.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9645 >
2021-04-07 19:56:17 +00:00
e47b742810
lavapipe: add vulkan 1.1 properties/features apis
...
Still have to add subgroups before advertising 1.1
Reviewed-by: Roland Scheidegger <sroland@vmware.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9645 >
2021-04-07 19:56:17 +00:00
35c02f79c9
lavapipe: add some asserts for blit region extents
...
this enforces spec for 3d vs array texture blits, more to come
Reviewed-by: Dave Airlie <airlied@redhat.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10005 >
2021-04-07 19:45:26 +00:00
be477f39aa
ci: update zink/virgl results for depth clamping fixes
...
Reviewed-by: Roland Scheidegger <sroland@vmware.com >
Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10068 >
2021-04-08 05:19:00 +10:00
60bdf61d68
llvmpipe: always take depth clamping from state tracker
...
Don't be smarter than state tracker here, of d3d10 wants to do
something the state tracker should hard code that. Since lavapipe
wants to use clip_halfz and depth clipping independently.
This fixes some issues blitting Z that zink was seeing
Reviewed-by: Roland Scheidegger <sroland@vmware.com >
Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10068 >
2021-04-08 05:18:16 +10:00
f96f016c22
llvmpipe: when depth clamp is disable clamp to 0.0/1.0
...
When depth clamp is disabled the viewport values aren't meaningful,
however the value is about to be converted to a unorm so needs
to still be clamped to 0/1.
This might not be the best place for this, maybe it should be in
the write swizzled code.
Reviewed-by: Roland Scheidegger <sroland@vmware.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10068 >
2021-04-08 05:17:53 +10:00
da166f648f
radv: enable DCC for concurrent images on GFX10
...
The driver now supports DCC stores on GFX10 that means that we can
keep DCC compressed on all layouts/queues.
This should help games which use aync compute or which declare
all images as concurrent like Youngblood.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com >
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9850 >
2021-04-07 18:33:34 +00:00
37d48de83e
radv: simplify a check when enabling DCC for concurrent images
...
If the image is marked as concurrent, each element of
pQueueFamilyIndices must be unique, so queue_family_mask is
different than 1.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com >
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9850 >
2021-04-07 18:33:34 +00:00
05ecb6780c
zink: emit WorkgroupSize when not using ExecutionModeLocalSize
...
the system_values_read bit might not be set if the value isn't read or is
DCE, but it still needs to be emitted
probably.
fixes #4591
Reviewed-by: Witold Baryluk <witold.baryluk@gmail.com >
Tested-by: Witold Baryluk <witold.baryluk@gmail.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10079 >
2021-04-07 17:55:53 +00:00
636170c888
docs: update calendar and link releases notes for 21.0.2
...
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10078 >
2021-04-07 17:51:27 +00:00
6f4e17a4bb
relnotes: Add sha256sum for 21.0.2
...
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10078 >
2021-04-07 17:51:27 +00:00
8368598c44
docs: add release notes for 21.0.2
...
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10078 >
2021-04-07 17:51:27 +00:00
4e5ea37612
ci: Use lavacli from master
...
This allows to fix some credential leaks and provides compatibility with
the latest LAVA version.
Avoids failures when the LAVA instance returns Error 500
Signed-off-by: Corentin Noël <corentin.noel@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9851 >
2021-04-07 17:26:29 +00:00
d4db68b338
radv: enable DCC stores with MSAA 4x/8x on GFX10+
...
Should work now.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com >
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10052 >
2021-04-07 17:13:51 +00:00
1286e73c2c
nir/lower_idiv: Add 8-bit and 16-bit lowering path
...
Roundtrip to a larger float and divide there. The extra details for
mod/rem are handled directly in integer space to simplify verification
of rounding details. The one issue is that the mantissa might be
rounded down which will cause issues; adding 1 unconditionally (proposed
by Jonathan Marek) fixes this. The lowerings here were tested
exhaustively on all pairs of 16-bit integers.
v2: Update idiv lowering per Rhys Perry's comment.
v3: Rewrite lowerings.
v4: Remove useless ftrunc, fix 8-bit issue, simplify code.
v5: Remove useless ffloor
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com >
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net >
Tested-by: Danylo Piliaiev <dpiliaiev@igalia.com >
Tested-by: Jason Ekstrand <jason@jlekstrand.net >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/8339 >
2021-04-07 15:48:15 +00:00
e91dec1327
nir/lower_idiv: Factor out numer/denom load
...
No need to duplicate across paths.
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com >
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/8339 >
2021-04-07 15:48:15 +00:00
7b0eb4aa00
nir/lower_idiv: Convert to lower_instructions
...
Helps deduplicate some code between the two lowering paths. In
particular, it ports the missing 32-bit? check to the precise pass. This
does not change anything immediately: drivers depending on this to lower
16-bit did not work before due to type mismatches and will not work now
since it'll refuse to lower. But that means sub-32-bit idiv can be
lowered more efficiently in an algebraic pass.
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com >
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/8339 >
2021-04-07 15:48:15 +00:00
e4da24bd24
nir: Add {i2f, u2f, f2i, f2u} helpers
...
Convenient for bitsize independent lowerings, will be used in the idiv
lowering.
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com >
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/8339 >
2021-04-07 15:48:15 +00:00
6b19711645
nir: Add nir_type_convert
...
Generalizes nir_convert_to_bit_size, which we implement as a
special-case.
v2: Take a sized dest type but allow unsized or sized source to address
Jason's feedback. Shorten name.
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com >
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/8339 >
2021-04-07 15:48:15 +00:00
f280367a27
radv: perform MSAA color decompression for storage images with DCC
...
For MSAA storage images with DCC, we also need to perform a MSAA
color decompression.
Fixes dEQP-VK.pipeline.multisample.storage_image.* if DCC stores
is enabled.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com >
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9854 >
2021-04-07 15:35:07 +00:00
835c5b7ebf
aco: fix integer tg4 workaround with unnormalized coordinates
...
Same as LLVM from 2abf62d348pendingchaos02@gmail.com >
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10053 >
2021-04-07 15:21:51 +00:00
5a70c4d4a0
ir3: Don't copy propagate arrays in ir3_cp
...
We don't check whether there's an intervening write in this pass, which
makes it incorrect. ir3_cp_postsched does check correctly, but we were
accidentally doing it here anyway for some sources.
While we're here, delete some code that was only used in the array case.
Fixes: f370e954robdclark@gmail.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10076 >
2021-04-07 14:35:13 +00:00
1ad5ee5a04
ir3/cp_postsched: Set address of uses for relative mov's
...
Fixes: 680ca5brobdclark@gmail.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10076 >
2021-04-07 14:35:13 +00:00
dcc26a3945
ir3: Fix valid flags for STIB
...
Disallow immediates for the source. This was hidden by the fact that we
didn't copy-propagate trivial collect instructions.
Reviewed-by: Eric Anholt <eric@anholt.net >
Reviewed-by: Rob Clark <robdclark@gmail.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10076 >
2021-04-07 14:35:13 +00:00
94beaa1d92
ir3/legalize: Fix last input (ss) insertion
...
If there was a mix of ldlv and bary.f and we inserted an (ss) *after*
the last input which was a bary.f, then last_input_needs_ss would get
unset, even though it shouldn't. For figuring out whether we need the
(ss), we need to know whether there are any pending ldlv's when
last_input gets executed, not at the end of the block, which means that
the existing code's strategy of inserting it after the whole block has
been processed won't work. Rework it to do the last_input processing in
the main loop instead.
Reviewed-by: Rob Clark <robdclark@gmail.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10076 >
2021-04-07 14:35:13 +00:00
35ffe4fec1
freedreno/a3xx: Fix SP_FS_CTRL_REG1_INITIALOUTSTANDING
...
Unfortunately this didn't fix anything, but I thought I might as well
include it.
Reviewed-by: Eric Anholt <eric@anholt.net >
Reviewed-by: Rob Clark <robdclark@gmail.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10076 >
2021-04-07 14:35:13 +00:00
8cd7950014
util/bitset: Avoid out-of-bounds reads
...
I missed a corner case here: when the next range ends right at the end
of the bitset, we need to return immediately to avoid trying to search
after the bitset. And when finding the next end, we similarly need to
bail if the range is size 1 at the very end of the range. In practice
this probably would'nt have been noticed, because it would break out of
the loop anyway, but I happened to be running something using this under
Valgrind and it complained.
Reviewed-by: Eric Anholt <eric@anholt.net >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10076 >
2021-04-07 14:35:13 +00:00
387189a955
intel: avoid dumping null cs sampler/binding states
...
this crashes the driver
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com >
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10077 >
2021-04-07 14:24:08 +00:00
7ebb215495
nir_to_tgsi: run constant folding after nir_opt_algebraic
...
Otherwise, an infinite loop can occur.
The first nir_opt_algebraic in the loop can optimize:
op(bcsel(a, #b, c), #d) to bcsel(a, op(b, d), op(c, d))
which (because op(b, d) is not constant folded), will be reverted by the
second nir_opt_algebraic in the loop.
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com >
Reviewed-by: Neha Bhende <bhenden@vmware.com >
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10071 >
2021-04-07 13:55:31 +00:00
26c3a509de
intel: Drop gen_device_info::has_resource_streamer
...
We've never done anything with that HW and it was removed in ICL.
There's no point in carrying around the useless field.
Acked-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com >
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10069 >
2021-04-07 13:33:18 +00:00
8ec530d982
wsi/x11: Wait for fences with IMMEDIATE on Xwayland
...
The commit below was already meant to do this, but accidentally missed
this part.
Fixes stutter when the frame-rate drops below the refresh rate.
Fixes: e8f50bd600ajax@redhat.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10026 >
2021-04-07 13:20:20 +00:00
519eb735a3
turnip: implement variableMultisampleRate
...
If subpass doesn't have depth/color attachments - samples count is
devised from VkPipelineMultisampleStateCreateInfo::rasterizationSamples.
Without variableMultisampleRate enabled all pipelines in such subpass
should have the same samples count; variableMultisampleRate allows
to have pipelines with different number of samples in one subpass,
given that it doesn't have depth/color attachments.
Blob doesn't have it enabled but there is no known reason for this.
Passes:
dEQP-VK.pipeline.multisample.variable_rate.*
Fixes test:
dEQP-VK.pipeline.framebuffer_attachment.no_attachments_ms
Signed-off-by: Danylo Piliaiev <dpiliaiev@igalia.com >
Reviewed-by: Hyunjun Ko <zzoon@igalia.com >
Reviewed-by: Eric Anholt <eric@anholt.net >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9556 >
2021-04-07 12:04:45 +00:00
f31a84233b
ci: Don't retry failed test runs
...
After a recent LAVA update, failures in running a test would cause the
whole test to be retried. We only want for the boot to be retried (in
case of infrastructuer errors), so set the nr of retries for tests to 1.
Signed-off-by: Tomeu Vizoso <tomeu.vizoso@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9950 >
2021-04-07 13:36:14 +02:00
4297fe1d9d
ci: Set more reasonable timeouts for LAVA jobs
...
The test timeout shouldn't be bigger than the whole job's timeout.
Also, reduce the timeouts to a more sane level and remove timeouts that
don't bring value.
Signed-off-by: Tomeu Vizoso <tomeu.vizoso@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9950 >
2021-04-07 13:36:14 +02:00
93ec399b28
ci: Use a single template for LAVA jobs
...
It's a pain to keep both templates in sync and there aren't really that
many differences.
Signed-off-by: Tomeu Vizoso <tomeu.vizoso@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9950 >
2021-04-07 13:34:14 +02:00
8926c4a313
aux/trace: add a set_inlinable_constants hook
...
Reviewed-by: Christian Gmeiner <christian.gmeiner@gmail.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10057 >
2021-04-07 10:30:00 +00:00
eddbbd8b68
v3d: use uint type in _gen_unpack_uint
...
Use a unsigned int type in the loop to avoid unintended sign extensions.
Fixes CID#1414500 (Unintended sign extension [SIGN_EXTENSION]).
Reviewed-by: Jose Maria Casanova Crespo <jmcasanova@igalia.com >
Signed-off-by: Juan A. Suarez Romero <jasuarez@igalia.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10060 >
2021-04-07 09:39:42 +00:00
d760a995e3
panfrost: Hook up indirect draw support
...
Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com >
Reviewed-by: Alyssa Rosenzweig <alyssa@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/8700 >
2021-04-07 08:27:58 +00:00
54526d8eca
panfrost: Prepare things for indirect draws
...
Several things need to be tweaked to re-use existing helpers for
indirect draws:
* Indirect draws should always be considered as instanced draws since
we don't know in advance how many instances will be requested. For
each vertex attribute buffer entry we store the element divisor which
will be extracted by the compute shader and transformed into a HW
divisor
* PRIMITIVE.index_count should be initialized to its default value
(one, or zero after the minus(1) modification) waiting for the
compute shader to patch it
Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com >
Reviewed-by: Alyssa Rosenzweig <alyssa@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/8700 >
2021-04-07 08:27:58 +00:00
2e6d94c198
panfrost: Add helpers to support indirect draws
...
Indirect draws are implemented with compute jobs patching the
vertex/tiler jobs. Provide helpers to do that.
Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com >
Reviewed-by: Alyssa Rosenzweig <alyssa@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/8700 >
2021-04-07 08:27:58 +00:00
9a08c9097f
panfrost: Stop including pan_device.h from pan_bo.h
...
We want to define structures containing pan_pool objects in pan_device.h
but it is prevented by the
pan_device.h -> pan_pool.h -> pan_bo.h
^____________________________|
loop.
Break this loop by not including pan_device.h from pan_bo.h.
Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com >
Reviewed-by: Alyssa Rosenzweig <alyssa@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/8700 >
2021-04-07 08:27:58 +00:00
363c1ef0c0
gallium/u_threaded: split draws that don't fit in a batch
...
Multidraws with large draw count will try to call tc_add_slot_based_call
with a payload larger than TC_CALLS_PER_BATCH.
This commit split these draws in multiple commands to avoid this issue.
Cc: mesa-stable
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/4534
Reviewed-by: Marek Olšák <marek.olsak@amd.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9920 >
2021-04-07 08:00:36 +00:00
e25daef16f
panfrost: Skip disabled RTs when selecting a RT for transaction elimination
...
If the render target is disabled, the CRC won't be updated, and the
slice state should stay unchanged.
Fixes dEQP-GLES2.functional.fbo.render.shared_colorbuffer.tex2d_rgba_stencil_index8
and dEQP-GLES2.functional.fbo.render.shared_colorbuffer.rbo_rgba4_depth_component16.
Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com >
Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9907 >
2021-04-07 09:44:50 +02:00
13ff5756ae
panfrost: Don't reserve space in the color buffer for disabled RTs
...
Render targets that have no clear or draws are disabled and thus don't
use the color buffer. Don't reserve space for those.
Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com >
Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9907 >
2021-04-07 09:44:37 +02:00
Dave Airlie