AlexIndustrial/mesa
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
137,931 Commits 1 Branch 0 Tags
fcd01ad4448ccd7bd587a0531e8a968e7eee6029
Commit Graph

127229 Commits

Author SHA1 Message Date
Marek Olšák fcd01ad444 radeonsi: add si_can_fast_clear_depth/stencil helpers 
for later use

Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10003>
 2021-04-13 03:17:42 +00:00
Marek Olšák 4dd8d58ad5 radeonsi: clean up some mess around htile_stencil_disabled 
Set the final value in si_texture_create_object, so that other places
don't have to derive it redundantly.

The only thing to remember is that HTILE stencil can be enabled when
stencil is not present, and it can be disabled when stencil is present
due to various workarounds.

Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10003>
 2021-04-13 03:17:42 +00:00
Marek Olšák bcd1a69f79 radeonsi: parallelize Z/S conversion into TC-compatible with fast color clears 
It's not really a fast clear, but it's the next logical step towards doing
HTILE clears here.

Acked-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10003>
 2021-04-13 03:17:42 +00:00
Marek Olšák fb72d41b18 radeonsi: implement Z/S fast clear for non-zero mipmap levels 
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10003>
 2021-04-13 03:17:42 +00:00
Marek Olšák 6434b0b652 radeonsi: implement per-level DCC and CMASK fast clears for gfx10+ 
Fast clears are only used for level 0. This enables clearing level 0
of CMASK and DCC on gfx10+ when there are multiple mipmap levels.
vi_dcc_clear_level can also clear any level now.

Mipmapped array textures are still cleared slowly.

Acked-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10003>
 2021-04-13 03:17:42 +00:00
Marek Olšák 059f042fb1 radeonsi: allow trivial DCC clears for shared textures with DCC constant encode 
This relaxes the existing restriction.

Acked-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10003>
 2021-04-13 03:17:42 +00:00
Marek Olšák c00b314ae4 radeonsi: restructure DCC disablement into a switch 
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10003>
 2021-04-13 03:17:42 +00:00
Marek Olšák fa43d61953 radeonsi: don't cache FMASK transactions from CB in L2 
FMASK is usually pretty large. It's better to leave the cache to shaders.
FMASK stores are still cached, but they can be evicted sooner, which is
the same as other color stores. Only DCC, HTILE, and CMASK are cached.

I haven't benchmarked this, but it seems like the right thing to do.
This only affected APUs.

Acked-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10003>
 2021-04-13 03:17:42 +00:00
Nanley Chery bc1f715076 iris: Support I915_FORMAT_MOD_Y_TILED_GEN12_RC_CCS_CC 
With a resolution of 1600x1200, I measured FPS increases in:

* glxgears 18.04% +/- 0.65% (n=691)
* Nexuiz    3.58% +/- 0.09% (n=553)

compared to the master branch at commit
3f614c6f7c.

Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9230>
 2021-04-13 02:53:11 +00:00
Nanley Chery 4785aad6a0 iris: Support RC_CCS_CC modifier in plane queries 
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9230>
 2021-04-13 02:53:11 +00:00
Nanley Chery f02951d727 iris: Support clear color plane imports for RC_CCS_CC 
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9230>
 2021-04-13 02:53:11 +00:00
Nanley Chery 11a0584c48 isl: Describe I915_FORMAT_MOD_Y_TILED_GEN12_RC_CCS_CC 
v2. Simplify the diff. (Jason)

Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9230>
 2021-04-13 02:53:11 +00:00
Nanley Chery 0092219cfe iris: Set BO maps to NULL in bo_free 
bo_free is called on external BOs when there are no objects left which
reference them. The function unmaps the address range associated with
any maps which occured. However, if the BO is busy (not idle), it
doesn't mark the pointer to the start address as invalid. This can lead
to a segfault later on.

At the end of bo_free, these BOs are still present in the handle hash
table. If such a BO is reused (i.e., when a DMABUF with the same handle
is reimported) and the driver attempts to get another mapping, the
bufmgr will incorrectly assume that the map pointer is still valid and
reuse it. This leads to a segfault. Set the pointer to NULL to mark it
as invalid.

Enables iris to run and pass the piglit test,
ext_image_dma_buf_import-reimport-bug.

Cc: mesa-stable
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9230>
 2021-04-13 02:53:11 +00:00
Mike Blumenkrantz 5a61a4dbfa aux/trace: do deep dumps of fb state for triggered traces 
having the full surface info available here is very useful

Acked-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10093>
 2021-04-13 02:04:22 +00:00
Mike Blumenkrantz 5ac3bb806b aux/trace: dump current fb state on trigger-mode draw if it hasn't been seen yet 
this info is important to have for a given frame, but it requires that the base
structs be copied and stored to the trace context for later use

Acked-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10093>
 2021-04-13 02:04:22 +00:00
Mike Blumenkrantz 5ecffaa7a6 aux/trace: enhance trigger mode to dump context states during bind 
when dumping a single frame, the creation info for these states hasn't been
dumped yet, so always dump it during bind so it's visible

Acked-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10093>
 2021-04-13 02:04:22 +00:00
Chia-I Wu 51216d656d venus: check vn_renderer_info::vk_xml_version 
When crosvm does not support venus, it still advertises
VIRGL_RENDERER_CAPSET_VENUS but provides no or zeroed capset data.
vk_xml_version will be zero.

It is a good idea to verify vk_xml_version anyway.

v2: print required version suggested by Ryan

Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
Reviewed-by: Yiwei Zhang <zzyiwei@chromium.org>
Reviewed-by: Ryan Neph <ryanneph@google.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10192>
 2021-04-13 01:03:52 +00:00
James Park f3bd5f9590 amd: Hide drm_fourcc.h on Windows 
Declare missing definitions instead.

Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9708>
 2021-04-13 00:24:02 +00:00
James Park 7dfc9e4431 amd: Hide amdgpu_drm.h on Windows 
Declare missing definitions instead.

Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9708>
 2021-04-13 00:24:02 +00:00
Icecream95 821d68652d panfrost: Add fast path for graphics work group computation 
Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10158>
 2021-04-12 23:53:41 +00:00
Icecream95 c8620005d2 panfrost: Align BO size to 4096 bytes 
Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10158>
 2021-04-12 23:53:41 +00:00
Icecream95 1b41707bbc panfrost: Only add resource checksum BOs to the batch once 
Previously panfrost_batch_add_bo was called MAX_MIP_LEVELS times on
the same batch.

Fixes: cbf68b21fb ("panfrost: Move checksum_bo to panfrost_resource")

Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10158>
 2021-04-12 23:53:41 +00:00
James Park 1aeebac4e6 ac/rgp: BSD elf library compatibility 
Allow compilation on Windows using modified BSD elf library.

Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9210>
 2021-04-12 22:50:52 +00:00
Thong Thai e62c7e7c6c radeon: Add cropping to encoded H.265 when padding is used 
Because the VCN encoder needs the surface to be memory aligned, the
resolution of the image passed to the encoder might be larger and have
extra padding added - this change crops the resulting output to
compensate for the extra padding that might have been added.

Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/4559
Signed-off-by: Thong Thai <thong.thai@amd.com>
Reviewed-by: Leo Liu <leo.liu@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10137>
 2021-04-12 22:42:53 +00:00
Vasily Khoruzhick 77fdabdfa5 lima: limit number of draws per job 
Otherwise we may hit tile heap size limit if an app issues too many
draws per job.

Reviewed-by: Erico Nunes <nunes.erico@gmail.com>
Signed-off-by: Vasily Khoruzhick <anarsoul@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10121>
 2021-04-12 22:34:12 +00:00
Vasily Khoruzhick c8e80c24d4 lima: use passed surface to get mipmap level for reload, not cbuf 
lima_pack_reload_plbu_cmd() can be used for either cbuf or zsbuf, so
use passed surface other using cbuf unconditionally.

Fixes: b8c31ac06d ("lima: fix glCopyTexSubImage2D")
Reviewed-by: Erico Nunes <nunes.erico@gmail.com>
Signed-off-by: Vasily Khoruzhick <anarsoul@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10121>
 2021-04-12 22:34:12 +00:00
Erico Nunes b071d98c04 lima: increase epsilon for depthrange near == far 
When the application sets depthrange near == far, lima inserts an
'epsilon' difference of decreasing 'near' by 1 unit, to prevent
application bugs.
For some specific values of depthrange this does not seem to be
sufficient, presumably due to precision limitations.
Increase the difference by also incrementing the 'far' distance in
this case, to further avoid bugs.

Signed-off-by: Erico Nunes <nunes.erico@gmail.com>
Reviewed-by: Vasily Khoruzhick <anarsoul@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9933>
 2021-04-12 21:08:35 +00:00
Marek Olšák faf10bd49d ac/surface: use named "color and "zs" structures in unions 
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10083>
 2021-04-12 20:53:45 +00:00
Marek Olšák 468836317b ac/surface: unify htile_* and dcc_* fields as meta_* fields 
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10083>
 2021-04-12 20:53:45 +00:00
Marek Olšák 34fd07aa56 ac/surface: unify htile_levels and dcc_levels as meta_levels 
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10083>
 2021-04-12 20:53:45 +00:00
Marek Olšák b16c7b706f ac/surface: pack radeon_surf better 
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10083>
 2021-04-12 20:53:45 +00:00
Marek Olšák 988f148db3 ac/surface: overlap color and Z/S fields using a union in gfx9_surf_layout 
to save space

Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10083>
 2021-04-12 20:53:45 +00:00
Marek Olšák ab00557685 ac/surface: pack alignments by storing log2 in radeon_surf 
Only one bit is set in alignments, so store the bit offset (log2) and
change the type from uint32_t to uint8_t.

Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10083>
 2021-04-12 20:53:45 +00:00
Marek Olšák cb016bb600 ac/surface: pack radeon_surf::num_htile_levels better 
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10083>
 2021-04-12 20:53:45 +00:00
Marek Olšák 8837d1d833 ac/surface: pack gfx9_surf_layout:resource_type better to save 8 bytes 
Yes, this saves 8 bytes. See pahole for yourself.

Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10083>
 2021-04-12 20:53:45 +00:00
Marek Olšák 4bd5f62966 ac/surface: pack gfx9_surf_meta_flags better 
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10083>
 2021-04-12 20:53:45 +00:00
Marek Olšák 3f074d2f45 ac/surface: inline and reorder gfx9_surf_flags for better packing 
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10083>
 2021-04-12 20:53:45 +00:00
Marek Olšák 48dbdc62bf ac/surface: change legacy_surf_level::offset to 32-bit offset_256B shifted by 8 
Images are always aligned to 256B (enforced by register and descriptor
fields) and limited to 40-bit addresses. This saves some space.

Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10083>
 2021-04-12 20:53:45 +00:00
Marek Olšák bbda20bf29 ac/surface: overlap color and Z/S fields using a union in legacy_surf_layout 
to save space

Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10083>
 2021-04-12 20:53:45 +00:00
Marek Olšák 2fd8018845 ac/surface: split dcc level info from surface_info to save space 
stencil level info doesn't have DCC

Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10083>
 2021-04-12 20:53:45 +00:00
Rhys Perry e9dc3df868 nir/loop_unroll: fix is_indirect_load() with load_global 
load_global only has one source.

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Fixes: dfe429eb41 ("nir/loop_unroll: unroll more aggressively if it can improve load scheduling")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10186>
 2021-04-12 20:28:57 +00:00
Jesse Natalie 3b44d481e4 driconf: Remove default values from string driconf entries 
None of the call sites pass a string here, which produces warnings
for MSVC, for not passing an argument to a macro which requires it.

Looks like GCC/clang stringize an unpassed argument to ""

Acked-by: Eric Engestrom <eric@engestrom.ch>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10157>
 2021-04-12 19:12:40 +00:00
Jesse Natalie 8cc3100226 microsoft/clc: Update unit test to always use COMMON state for buffers 
New Windows versions have new debug validation warning that for resources that
aren't actually UPLOAD/READBACK (which these aren't, thanks to the
GetCustomHeapProperties call), initial state that's not COMMON doesn't actually
do anything, which causes these tests to all fail because they verify that
they don't produce debug layer messages.

Reviewed-By: Bill Kristiansen <billkris@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10190>
 2021-04-12 18:50:20 +00:00
Rhys Perry 0f2bf55c7e nir/lcssa: fix nondeterminism in predecessor iteration 
set_foreach()'s order on a list of nir_block * isn't deterministic, so we
need to sort the predecessor list.

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/3364>
 2021-04-12 18:17:19 +00:00
Rhys Perry 7050896be0 nir: add nir_block_get_predecessors_sorted() helper 
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/3364>
 2021-04-12 18:17:19 +00:00
Jason Ekstrand 21582016c5 anv: Make memory type and queue family pointers const 
These point to data inside the instance so no one should be modifying
them.

Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10184>
 2021-04-12 18:09:27 +00:00
Leo Liu 544044b29e radeonsi: replace the hard coded video decode and encode caps 
For the caps of max_width, max_height and max_level with those
that are available via querying the kernel.

Signed-off-by: Leo Liu <leo.liu@amd.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10095>
 2021-04-12 17:33:32 +00:00
Leo Liu 75a725e4c5 ac: add function for querying video capabilities 
It will be used to query caps of decode and encode
for hardware AMDGPU supports

Signed-off-by: Leo Liu <leo.liu@amd.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10095>
 2021-04-12 17:33:32 +00:00
Ilia Mirkin 1b06189f4a nv50: enable ARB_framebuffer_no_attachments 
This is a required part of ES3.1 and fairly trivial to enable. In
practice it's almost entirely useless since the only information one
would be able to obtain are queries.

Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Reviewed-by: Karol Herbst <kherbst@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10162>
 2021-04-12 16:45:40 +00:00
Ilia Mirkin 97d61ae771 nv50/ir: fake SV_THREAD_KILL support 
If we say that there are no helper invocations, it's suboptimal but
allows ES 3.1 to chug along.

Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Reviewed-by: Karol Herbst <kherbst@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10162>
 2021-04-12 16:45:40 +00:00