fdo-mirrors/mesa
1
0
Fork 0
mirror of https://gitlab.freedesktop.org/mesa/mesa.git synced 2026-01-25 14:30:22 +01:00
Code Issues Projects Releases Packages Wiki Activity Actions 1
184774 commits 66 branches 918 tags 1,006 MiB
Commit graph

184774 commits

This branch
This branch
All branches
Author SHA1 Message Date
Faith Ekstrand
498afcdb88 HACK: Only emit OpBar in compute shaders 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:14 +00:00
Faith Ekstrand
611fe837bb nak: Fix lowering for patch_vertices_in 
We have to grab the 2nd byte of the sysval for some reason.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:14 +00:00
Faith Ekstrand
efbbc3ed93 nak: Implement load_tess_coord 
We could lower this to load_per_vertex_output in NIR but then it
confuses all sorts of NIR passes which assume load_*output only
happens in control shaders.  We could also add a magic NIR intrinsic
but it's probably easier to just special-case this one.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:14 +00:00
Faith Ekstrand
b5cb508523 nak,nvk: Plumb through tessellation info 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:14 +00:00
Faith Ekstrand
bd90031b5b nak/nir: Lower load_primitive_id 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:14 +00:00
Faith Ekstrand
c2768736c9 nak: Implement more attribute I/O 
This commits implement load/store_per_vertex* and load_output which are
required for tessellation shaders.  Because things are getting a bit
complicated, it's easier to combine all the attribute load/store ops in
a single case in the match.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:14 +00:00
Faith Ekstrand
091410c708 nak: Add lowering for per-vertex I/O 
NVIDIA hardware doesn't take a vertex index for per-vertex I/O.
Instead, it takes an offset into the primitive.  This has to be fetched
using a combination of SR_INVOCATION_INFO and the ISBERD instruction.
To keep things simple and allow for maximum CSE, we do the lowering in
NIR and patch the load/store_per_vertex_input/output intrinsic.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:14 +00:00
Faith Ekstrand
aa76c86042 nak: Handle location_frac for FS outputs in nak_from_nir.rs 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:14 +00:00
Faith Ekstrand
1251bb7588 nak: Set per patch attribute count both places in the SPH 
This is what codegen does and it seems to be required.  Maybe it's an
SPH version thing?

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:14 +00:00
Faith Ekstrand
f097303a99 nak: Rework OpALd and OpASt a bit 
Replace the out_load flag with an output flag which we now require to be
set for OpASt.  Also, improve printing so we add .O and .P qualifiers as
needed.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:14 +00:00
Faith Ekstrand
f5805dadab nak: NVIDIA calls them tessellation init shaders 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:13 +00:00
Faith Ekstrand
dfbc03fa88 spirv: Fix locations for per-patch varyings 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:13 +00:00
Faith Ekstrand
c6c2c1b2b2 nak/nir: Lower tessellation and geometry I/O 
We also need to call nir_lower_indirect_derefs.  Even though we're not
lowering anything away, we need to call it so that it will lower clip
distance arrays.  The pass always lowers compact array derfs.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:13 +00:00
Faith Ekstrand
93db2c1254 nak: Plumb clip/cull enables through nak 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:13 +00:00
Faith Ekstrand
d4d7f26151 nak: Add enums for sysvals and attributes 
It's much nicer to have a canonical place for all these things to live
which gives them actual names.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:13 +00:00
Faith Ekstrand
a4c1ec3892 nak: Implement From for SrcRef for more types 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:13 +00:00
Faith Ekstrand
0d725cccbd nak: Implement NIR control barriers 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:13 +00:00
Faith Ekstrand
5e6dd0412b nak/builder: Return the instruction from push_*() 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:13 +00:00
Faith Ekstrand
9b6a4aab8d nak: Add barrier instructions 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:13 +00:00
Faith Ekstrand
2514058c82 nak: Allow encoding Dst::None 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:13 +00:00
Faith Ekstrand
cc6f26b005 nak: Break indirect offset encoding into a helper 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:13 +00:00
Faith Ekstrand
2a61e59bba nak: Add OpNop which can have a label 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:13 +00:00
Faith Ekstrand
e3fa6f3557 nak: Add a Label struct for branch targets 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:13 +00:00
Faith Ekstrand
f2e07cbab9 nvk: Plumb through XFB info from NAK 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:13 +00:00
Faith Ekstrand
30f01c47c2 nak: Translate XFB info 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:13 +00:00
Faith Ekstrand
62ac73d5e3 nouveau/parser: Dump SET_STREAM_OUT_CONTROL_* properly 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:13 +00:00
Faith Ekstrand
c2c010055a nak: Handle non-constant I/O offsets 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:13 +00:00
Faith Ekstrand
827dba398d nak: Don't allocate bitsets in liveness data-flow 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:13 +00:00
Faith Ekstrand
04658a2a11 nak/bitset: Add a helper for modifying in-place 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:13 +00:00
Faith Ekstrand
19e0c52837 nak: Run simple liveness data-flow bottom-up 
Rookie mistake... The liveness algorithm propagates information from
later blocks to earlier blocks so if you run bottom-up it's exactly two
passes when there aren't loops.  If you run top-down, it's quadratic in
the number of blocks.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:13 +00:00
Faith Ekstrand
046a2d5004 nvk: Set clip/cull_enable for NAK shaders 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:13 +00:00
Faith Ekstrand
a4ff5a9fc6 nak: Simplify I/O gathering 
This adds mark_attrs_(read|written)() helpers to VtgIoInfo which take a
range of attribute addresses and mark the range as-needed.  It adds a
similar mark_attr_read() helper to FragmentIoInfo which only marks a
single address and takes a PixelImap.  This gets us down to only needing
to duplicate the address range if ladder twice.  For VTG I/O, having it
take ranges will be more ergonamic when it comes time to handle non-
constant I/O offsets.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:13 +00:00
Faith Ekstrand
32910d3016 nak: rustfmt fixes 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:13 +00:00
Faith Ekstrand
40ec7efa07 nak: Move NIR enum translation out of nak_sph.rs 
Better to keep all the NIR stuff together.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:13 +00:00
Mary Guillemard
145213fd2c nak: Restructure ShaderInfo 
Signed-off-by: Mary Guillemard <mary.guillemard@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:13 +00:00
Mary Guillemard
267a4f07d0 nak: Remove encode_hdr_for_nir 
Now unused as nak_sph::encode_header handles it.

Signed-off-by: Mary Guillemard <mary.guillemard@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:13 +00:00
Mary Guillemard
a926c67335 nak: Collect information to create SPH 
This should cover everything except tesselation.

Signed-off-by: Mary Guillemard <mary.guillemard@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:13 +00:00
Mary Guillemard
8d07d033ae nak: Add a complete wrapper around SPH 
Signed-off-by: Mary Guillemard <mary.guillemard@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:13 +00:00
Faith Ekstrand
97c249a353 nak: Fix printing of OpASt 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:13 +00:00
Faith Ekstrand
a7eb42119a nak: Take component into account in store_output 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:13 +00:00
Faith Ekstrand
330f3e6fa8 nvk: Lower more draw system values 
In codegen, these are handled by the backend by passing in cbuf numbers
and offsets.  NAK expects us to lower these in the driver.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:13 +00:00
Faith Ekstrand
263dfd4838 nvk: Add a helper for lowering system values to root table loads 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:13 +00:00
Mary Guillemard
1a2bed819d nak: Fix ALD/AST encoding for vtx and offset 
The bit offsets were inverted compared to gv100 codegen.

Signed-off-by: Mary Guillemard <mary.guillemard@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:13 +00:00
Faith Ekstrand
66c9c40f68 nak: Handle modifiers in dedup_srcs() in opt_lop() 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:13 +00:00
Faith Ekstrand
1b6962eb7e nak: Provide more detail when printing IR after passes 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:13 +00:00
Faith Ekstrand
5654366a55 nak: Fold source modifiers in legalize 
We were smashing them to None later but didn't actually take them into
account in fold_lop_src() to ensure the smash was valid.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:13 +00:00
Faith Ekstrand
6da06bf7d7 nak/nir: Return one sample for gl_SampleMaskIn[0] when sample shading 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:13 +00:00
Faith Ekstrand
0e393f0d58 nak/nir: Interpolate gl_PointCoord 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:13 +00:00
Faith Ekstrand
062735c1cc nak/nir: Add a load_frag_w helper 
Also, rework things a bit.  The one functional change is that we now put
1/w in gl_FragCoord.w which should fix a CTS test.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:13 +00:00
Faith Ekstrand
8ec753749a nak/nir: Fix sample vs. pixel input interpolation 
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24998>
 2023-11-14 00:48:13 +00:00