Alejandro Piñeiro [Fri, 22 Oct 2021 08:54:24 +0000 (10:54 +0200)]
v3d: handle new TEXTURE_SHADER_STATE v71 YCbCr fields
There are some new fields for YCbCr with pointers for the various
planes in multi-planar formats. These need to match the base address
pointer in the texture state, or the hardware will assume this is a
multi-planar texture.
Notice we don't use an address type for these fields in the XML
description. This is because the addresses are 64-bit aligned (even
though the PRM doesn't say it) which means the 6 LSB bits are
implicitly 0, but the fields are encoded before the 6th bit of their
starting byte, so we can't use the usual trick we do with address
types where the first 6 bits in the byte are implicitly overwritten by
other fields and we have to encode this manually as a uint field. This
would mean that if we had an actual BO we would also need to add it
manually to the job's list, but since we don't have one, we don't have
to do anything about it.
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Alejandro Piñeiro [Thu, 21 Oct 2021 21:21:02 +0000 (23:21 +0200)]
v3d: handle new texture state transfer functions in v71
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Alejandro Piñeiro [Thu, 21 Oct 2021 11:46:11 +0000 (13:46 +0200)]
v3d/uniforms: update VIEWPORT_X/Y_SCALE uniforms for v71
As the packet CLIPPER_XY scaling, this needs to be computed on 1/64ths
of pixel, instead of 1/256ths of pixels.
As this is the usual values that we get from macros, we add manually a
v42 and v71 macro, and define a new helper to get those.
Those granularity values are the same for Vulkan and OpenGL, so
perhaps we should move them to a common place.
As with v3dv, V3D_X macro name is somewhat confusing. It is
specifically created to ask for define values that depends on the
version. But I also felt that V3D_DEFINE_X was too long.
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Alejandro Piñeiro [Thu, 21 Oct 2021 11:37:46 +0000 (13:37 +0200)]
v3d: default vertex attributes values are not needed for v71
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Alejandro Piñeiro [Thu, 21 Oct 2021 11:19:49 +0000 (13:19 +0200)]
v3d: no specific separate_segments flag for V3D 7.1
On V3D 7.1 there is not a flag on the Shader State Record to specify
if we are using shared or separate segments. This is done by setting
the vpm input size to 0 (so we need to ensure that the output would be
the max needed for input/output).
We were already doing the latter on the prog_data_vs, so we just need
to use those values, instead of assigning default values.
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Alejandro Piñeiro [Thu, 21 Oct 2021 11:14:32 +0000 (13:14 +0200)]
v3d: emit CLIPPER_XY_SCALING for v71
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Alejandro Piñeiro [Thu, 21 Oct 2021 11:09:03 +0000 (13:09 +0200)]
v3d: just don't fill up early-z fields for CFG_BITS for v71
v71 doesn't include early_z_enable/early_z_updates_enable. They are
configured with packet 121.
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Alejandro Piñeiro [Wed, 20 Oct 2021 23:47:29 +0000 (01:47 +0200)]
v3d: emit CLEAR_RENDER_TARGETS for v71
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Alejandro Piñeiro [Wed, 20 Oct 2021 13:12:15 +0000 (15:12 +0200)]
v3d: TILE_RENDERING_MODE_CFG_RENDER_TARGET_PART1
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Alejandro Piñeiro [Wed, 20 Oct 2021 12:31:10 +0000 (14:31 +0200)]
v3d: emit TILE_BINNING_MODE_CFG and TILE_RENDERING_MODE_CFG_COMMON for v71
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Alejandro Piñeiro [Tue, 23 May 2023 21:32:37 +0000 (23:32 +0200)]
v3d: add v71 hw generation
Starting point for v71 version inclusion:
* Adds as one of the versions to be compiled on meson
* Updated the v3d_X and v3dX macros to include version 71
* Update the code enough to get it building when using v71.
Any real v71 support will be implemented on following commits.
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Tue, 14 Feb 2023 09:09:53 +0000 (10:09 +0100)]
v3dv: fix depth clipping then Z scale is too small in V3D 7.x
When the Z scale is too small guardband clipping may not clip
correctly, so disable it, which is a new option in V3D 7.x.
This fixes this test in V3D 7.x without needing any workarounds:
dEQP-VK.draw.renderpass.inverted_depth_ranges.nodepthclamp_deltazero
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Wed, 20 Oct 2021 09:22:11 +0000 (11:22 +0200)]
v3dv: make v3dv_viewport_compute_xform depend on the V3D version
For 4.x we have a workaround for too small Z scale values that is not
required for V3D 7.x.
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Alejandro Piñeiro [Wed, 17 Nov 2021 10:33:59 +0000 (11:33 +0100)]
v3dv: add support for TFU jobs in v71
This includes update the simulator.
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Fri, 15 Oct 2021 11:06:31 +0000 (13:06 +0200)]
v3dv: handle Z clipping in v71
Fixes the following tests:
dEQP-VK.clipping.clip_volume.*
dEQP-VK.draw.inverted_depth_ranges.nodepthclamp_* (except deltazero)
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Thu, 7 Oct 2021 10:43:49 +0000 (12:43 +0200)]
v3dv: don't convert floating point border colors in v71
The TMU does this for us now.
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Alejandro Piñeiro [Fri, 1 Oct 2021 13:18:38 +0000 (15:18 +0200)]
v3dv: no specific separate_segments flag for V3D 7.1
On V3D 7.1 there is not a flag on the Shader State Record to specify
if we are using shared or separate segments. This is done by setting
the vpm input size to 0 (so we need to ensure that the output would be
the max needed for input/output).
We were already doing the latter on the prog_data_vs, so we just need
to use those values, instead of assigning default values.
As we are here, we also add some comments on the compiler part.
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Wed, 29 Sep 2021 07:07:28 +0000 (09:07 +0200)]
v3dv: handle RTs with no color targets in v71
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Wed, 29 Sep 2021 06:22:59 +0000 (08:22 +0200)]
v3dv: handle early Z/S clears for v71
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Tue, 28 Sep 2021 06:59:08 +0000 (08:59 +0200)]
broadcom/compiler: update thread end restrictions validation for v71
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Tue, 28 Sep 2021 06:31:04 +0000 (08:31 +0200)]
v3dv: GFX-1461 does not affect V3D 7.x
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Tue, 28 Sep 2021 06:23:48 +0000 (08:23 +0200)]
v3dv: handle render pass global clear for v71
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Tue, 28 Sep 2021 06:14:11 +0000 (08:14 +0200)]
v3dv: implement noop job for v71
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Sun, 24 Oct 2021 23:38:31 +0000 (01:38 +0200)]
v3dv: handle new texture state transfer functions in v71
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Sun, 24 Oct 2021 23:37:12 +0000 (01:37 +0200)]
v3dv: fix up texture shader state for v71
There are some new fields for YCbCr with pointers for the various
planes in multi-planar formats. These need to match the base address
pointer in the texture state, or the hardware will assume this is a
multi-planar texture.
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Wed, 22 Sep 2021 10:04:21 +0000 (12:04 +0200)]
v3dv: setup TLB clear color for meta operations in v71
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Wed, 22 Sep 2021 10:03:58 +0000 (12:03 +0200)]
v3dv: setup render pass color clears for any format bpp in v71
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Alejandro Piñeiro [Wed, 28 Jul 2021 11:45:52 +0000 (13:45 +0200)]
v3dv/pipeline: handle GL_SHADER_STATE_RECORD changed size on v71
It is likely that we would need more changes, as this packet changed,
but this is enough to get basic tests running. Any additional support
will be handled with new commits.
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Alejandro Piñeiro [Wed, 28 Jul 2021 10:05:26 +0000 (12:05 +0200)]
v3dv/pipeline: default vertex attributes values are not needed for v71
There are not part of the shader state record.
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Alejandro Piñeiro [Wed, 28 Jul 2021 10:01:38 +0000 (12:01 +0200)]
v3dv: default vertex attribute values are gen dependant
Content, structure and size would depend on the generation. Even if it
is needed at all.
So let's move it to the v3dvx files.
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Alejandro Piñeiro [Tue, 27 Jul 2021 12:02:30 +0000 (14:02 +0200)]
v3dv/cmd_buffer: just don't fill up early-z fields for CFG_BITS for v71
For v71 early_z_enable/early_z_updates_enable is configured with
packet 121.
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Alejandro Piñeiro [Tue, 14 Sep 2021 08:08:19 +0000 (10:08 +0200)]
v3dv/uniforms: update VIEWPORT_X/Y_SCALE uniforms for v71
As the packet CLIPPER_XY scaling, this needs to be computed on 1/64ths
of pixel, instead of 1/256ths of pixels.
As this is the usual values that we get from macros, we add manually a
v42 and v71 macro, and define a new helper (V3DV_X) to get the value
for the current hw version.
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Alejandro Piñeiro [Sun, 19 Sep 2021 21:37:32 +0000 (23:37 +0200)]
v3dv/cmd_buffer: emit CLIPPER_XY_SCALING for v71
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Alejandro Piñeiro [Mon, 26 Jul 2021 13:08:11 +0000 (15:08 +0200)]
v3dvx/cmd_buffer: emit CLEAR_RENDER_TARGETS for v71
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Alejandro Piñeiro [Thu, 22 Jul 2021 12:26:13 +0000 (14:26 +0200)]
v3dv/cmd_buffer: emit TILE_RENDERING_MODE_CFG_RENDER_TARGET_PART1 for v71
Signed-off-by: Alejandro Piñeiro <apinheiro@igalia.com>
Signed-off-by: Iago Toral Quiroga <itoral@igalia.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Alejandro Piñeiro [Tue, 20 Jul 2021 12:00:44 +0000 (14:00 +0200)]
v3dv: emit TILE_BINNING_MODE_CFG and TILE_RENDERING_MODE_CFG_COMMON for v71
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Wed, 10 Nov 2021 06:54:35 +0000 (07:54 +0100)]
v3dv/device: handle new rpi5 device (bcm2712)
This includes both master and primary devices.
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Wed, 10 Nov 2021 09:06:50 +0000 (10:06 +0100)]
v3dv: expose V3D revision number in device name
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Alejandro Piñeiro [Tue, 29 Jun 2021 09:59:53 +0000 (11:59 +0200)]
v3dv/meson: add v71 hw generation
Starting point for v71 version inclusion.
This just adds it as one of the versions to be compiled (on meson),
updates the v3dX/v3dv_X macros, and update the code enough to get it
compiling when building using the two versions. For any packet not
available on v71 we just provide a generic asserted placeholder of
generation not supported.
Any real v71 support will be implemented on following commits.
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Mon, 15 May 2023 08:02:10 +0000 (10:02 +0200)]
broadcom/compiler: only assign rf0 as last resort in V3D 7.x
So we can use it for ldunif(a) and avoid generating ldunif(a)rf which
can't be paired with conditional instructions.
shader-db (pi5):
total instructions in shared programs:
11357802 ->
11338883 (-0.17%)
instructions in affected programs: 7117889 -> 7098970 (-0.27%)
helped: 24264
HURT: 17574
Instructions are helped.
total uniforms in shared programs: 3857808 -> 3857815 (<.01%)
uniforms in affected programs: 92 -> 99 (7.61%)
helped: 0
HURT: 1
total max-temps in shared programs: 2230904 -> 2230199 (-0.03%)
max-temps in affected programs: 52309 -> 51604 (-1.35%)
helped: 1219
HURT: 725
Max-temps are helped.
total sfu-stalls in shared programs: 15021 -> 15236 (1.43%)
sfu-stalls in affected programs: 6848 -> 7063 (3.14%)
helped: 1866
HURT: 1704
Inconclusive result
total inst-and-stalls in shared programs:
11372823 ->
11354119 (-0.16%)
inst-and-stalls in affected programs: 7149177 -> 7130473 (-0.26%)
helped: 24315
HURT: 17561
Inst-and-stalls are helped.
total nops in shared programs: 273624 -> 273711 (0.03%)
nops in affected programs: 31562 -> 31649 (0.28%)
helped: 1619
HURT: 1854
Inconclusive result (value mean confidence interval includes 0).
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Tue, 2 May 2023 08:17:47 +0000 (10:17 +0200)]
broadcom/compiler: don't assign registers to unused nodes/temps
In programs with a lot of unused temps, if we don't do this, we may
end up recycling previously used rfs more often, which can be
detrimental to instruction pairing.
total instructions in shared programs:
11464335 ->
11444136 (-0.18%)
instructions in affected programs: 8976743 -> 8956544 (-0.23%)
helped: 33196
HURT: 33778
Inconclusive result
total max-temps in shared programs: 2230150 -> 2229445 (-0.03%)
max-temps in affected programs: 86413 -> 85708 (-0.82%)
helped: 2217
HURT: 1523
Max-temps are helped.
total sfu-stalls in shared programs: 18077 -> 17104 (-5.38%)
sfu-stalls in affected programs: 8669 -> 7696 (-11.22%)
helped: 2657
HURT: 2182
Sfu-stalls are helped.
total inst-and-stalls in shared programs:
11482412 ->
11461240 (-0.18%)
inst-and-stalls in affected programs: 8995697 -> 8974525 (-0.24%)
helped: 33319
HURT: 33708
Inconclusive result
total nops in shared programs: 298140 -> 296185 (-0.66%)
nops in affected programs: 52805 -> 50850 (-3.70%)
helped: 3797
HURT: 2662
Inconclusive result
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Tue, 2 May 2023 08:12:37 +0000 (10:12 +0200)]
broadcom/compiler: improve allocation for final program instructions
The last 3 instructions can't use specific registers so flag all the
nodes for temps used in the last program instructions and try to
avoid assigning any of these. This may help us avoid injecting nops
for the last thread switch instruction.
Because regisster allocation needs to happen before QPU scheduling
and instruction merging we can't tell exactly what the last 3
instructions will be, so we do this for a few more instructions than
just 3.
We only do this for fragment shaders because other shader stages
always end with VPM store instructions that take an small immediate
and therefore will never allow us to merge the final thread switch
earlier, so limiting allocation for these shaders will never improve
anything and might instead be detrimental.
total instructions in shared programs:
11471389 ->
11464335 (-0.06%)
instructions in affected programs: 582908 -> 575854 (-1.21%)
helped: 4669
HURT: 578
Instructions are helped.
total max-temps in shared programs: 2230497 -> 2230150 (-0.02%)
max-temps in affected programs: 5662 -> 5315 (-6.13%)
helped: 344
HURT: 44
Max-temps are helped.
total sfu-stalls in shared programs: 18068 -> 18077 (0.05%)
sfu-stalls in affected programs: 264 -> 273 (3.41%)
helped: 37
HURT: 48
Inconclusive result (value mean confidence interval includes 0).
total inst-and-stalls in shared programs:
11489457 ->
11482412 (-0.06%)
inst-and-stalls in affected programs: 585180 -> 578135 (-1.20%)
helped: 4659
HURT: 588
Inst-and-stalls are helped.
total nops in shared programs: 301738 -> 298140 (-1.19%)
nops in affected programs: 14680 -> 11082 (-24.51%)
helped: 3252
HURT: 108
Nops are helped.
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Tue, 18 Apr 2023 06:50:13 +0000 (08:50 +0200)]
broadcom/compiler: don't allocate spill base to rf0 in V3D 7.x
Otherwise it can be stomped by instructions doing implicit rf0 writes.
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Alejandro Piñeiro [Fri, 26 Nov 2021 00:24:12 +0000 (01:24 +0100)]
broadcom/qpu: new packing/conversion v71 instructions
This commits adds the qpu definitions for several new v71
instructions.
Packing:
* vpack does a 2x32 to 2x16 bit integer pack
* v8pack: Pack 2 x 2x16 bit integers into 4x8 bits
* v10pack packs parts of 2 2x16 bit integer into r10g10b10a2.
* v11fpack packs parts of 2 2x16 bit float into r11g11b10 rounding
to nearest
Conversion to unorm/snorm:
* vftounorm8/vftosnorm8: converts from 2x16-bit floating point
to 2x8 bit unorm/snorm.
* ftounorm16/ftosnorm16: converts floating point to 16-bit
unorm/snorm
* vftounorm10lo: Convert 2x16-bit floating point to 2x10-bit unorm
* vftounorm10hi: Convert 2x16-bit floating point to one 2-bit and one 10-bit unorm
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Tue, 9 Nov 2021 10:34:59 +0000 (11:34 +0100)]
broadcom/compiler: fix up copy propagation for v71
Update rules for unsafe copy propagations to match v7.x.
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Mon, 29 Nov 2021 12:23:11 +0000 (13:23 +0100)]
broadcom/compiler: lift restriction on vpmwt in last instruction for V3D 7.x
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Thu, 25 Nov 2021 12:00:34 +0000 (13:00 +0100)]
broadcom/compiler: validate restrictions after TLB Z write
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Fri, 26 Nov 2021 09:37:05 +0000 (10:37 +0100)]
broadcom/compiler: start allocating from RF 4 in V7.x
In V3D 4.x we start at RF3 so that we allocate RF0-2 only if there
aren't any other RFs available. This is useful with small shaders to
ensure that our TLB writes don't use these registers because these are
the last instructions we emit in fragment shaders and the last
instructions in a program can't write to these registers, so if we do,
we need to emit NOPs.
In V3D 7.x the registers affected by this restriction are RF2-3, so we
choose to start at RF4.
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Thu, 25 Nov 2021 07:31:02 +0000 (08:31 +0100)]
broadcom/compiler: lift restriction for branch + msfign after setmsf for v7.x
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Tue, 23 Nov 2021 09:04:49 +0000 (10:04 +0100)]
broadcom/compiler: update ldvary thread switch delay slot restriction for v7.x
In V3D 7.x we don't have accumulators which would not survive a thread
switch, so the only restriction is that ldvary can't be placed in the
second delay slot of a thread switch.
shader-db results for UnrealEngine4 shaders:
total instructions in shared programs: 446458 -> 446401 (-0.01%)
instructions in affected programs: 13492 -> 13435 (-0.42%)
helped: 58
HURT: 3
Instructions are helped.
total nops in shared programs: 19571 -> 19541 (-0.15%)
nops in affected programs: 161 -> 131 (-18.63%)
helped: 30
HURT: 0
Nops are helped.
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Mon, 22 Nov 2021 11:56:03 +0000 (12:56 +0100)]
broadcom/compiler: update thread end restrictions for v7.x
In 4.x it is not allowed to write to the register file in the last 3
instructions, but in 7.x we only have this restriction in the thread
end instruction itself, and only if the write comes from the ALU
ports.
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Wed, 3 Nov 2021 09:34:19 +0000 (10:34 +0100)]
broadcom/compiler: implement small immediates for v71
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Mon, 25 Oct 2021 07:38:57 +0000 (09:38 +0200)]
broadcom/compiler: convert mul to add when needed to allow merge
V3D 7.x added 'mov' opcodes to the ADD alu, so now it is possible to
move these to the ADD alu to facilitate merging them with other MUL
instructions.
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Fri, 29 Oct 2021 11:00:56 +0000 (13:00 +0200)]
broadcom/compiler: don't assign rf0 to temps that conflict with ldvary
ldvary writes to rf0 implicitly, so we don't want to allocate rf0 to
any temps that are live across ldvary's rf0 live ranges.
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Thu, 28 Oct 2021 12:13:29 +0000 (14:13 +0200)]
broadcom/compiler: try to use ldunif(a) instead of ldunif(a)rf in v71
The rf variants need to encode the destination in the cond bits, which
prevents these to be merged with any other instruction that need them.
In 4.x, ldunif(a) write to r5 which is a special register that only
ldunif(a) and ldvary can write so we have a special register class for
it and only allow it for them. Then when we need to choose a register
for a node, if this register is available we always use it.
In 7.x these instructions write to rf0, which can be used by any
instruction, so instead of restricting rf0, we track the temps that
are used as ldunif(a) destinations and use that information to favor
rf0 for them.
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Wed, 27 Oct 2021 09:35:12 +0000 (11:35 +0200)]
broadcom/compiler: enable ldvary pipelining on v71
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Wed, 6 Oct 2021 11:58:27 +0000 (13:58 +0200)]
broadcom/compiler: handle rf0 flops storage restriction in v71
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Tue, 26 Oct 2021 06:37:54 +0000 (08:37 +0200)]
broadcom/qpu: add packing for fmov on ADD alu
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Tue, 26 Oct 2021 09:43:02 +0000 (11:43 +0200)]
broadcom/compiler: update peripheral access restrictions for v71
In V3D 4.x only a couple of simultaneous accesses where allowed, but
V3D 7.x is a bit more flexible, so rather than trying to check for all
the allowed combinations it is easier to check if we are one of the
disallows.
Shader-db (pi5):
total instructions in shared programs:
11338883 ->
11307386 (-0.28%)
instructions in affected programs: 2727201 -> 2695704 (-1.15%)
helped: 12555
HURT: 289
Instructions are helped.
total max-temps in shared programs: 2230199 -> 2229260 (-0.04%)
max-temps in affected programs: 20508 -> 19569 (-4.58%)
helped: 608
HURT: 4
Max-temps are helped.
total sfu-stalls in shared programs: 15236 -> 15293 (0.37%)
sfu-stalls in affected programs: 148 -> 205 (38.51%)
helped: 38
HURT: 64
Inconclusive result (%-change mean confidence interval includes 0).
total inst-and-stalls in shared programs:
11354119 ->
11322679 (-0.28%)
inst-and-stalls in affected programs: 2732262 -> 2700822 (-1.15%)
helped: 12550
HURT: 304
Inst-and-stalls are helped.
total nops in shared programs: 273711 -> 274095 (0.14%)
nops in affected programs: 9626 -> 10010 (3.99%)
helped: 186
HURT: 397
Nops are HURT.
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Alejandro Piñeiro [Tue, 19 Oct 2021 21:52:30 +0000 (23:52 +0200)]
broadcom/compiler: update payload registers handling when computing live intervals
As for v71 the payload registers are not the same. Specifically now
rf3 is used as payload register, so this is needed to avoid rf3 being
selected as a instruction dst by the register allocator, overwriting
the payload value that could be still used.
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Alejandro Piñeiro [Tue, 19 Oct 2021 09:51:32 +0000 (11:51 +0200)]
broadcom/compiler: update ldunif/ldvary comment for v71
For v42 and below ldunif/ldvary write both on r5, but with a different
delay, so we need to take that into account when scheduling both.
For v71 the register used is rf0, but the behaviour is the same. So
the scheduling code can be the same, but the comment needs update.
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Alejandro Piñeiro [Tue, 19 Oct 2021 09:16:43 +0000 (11:16 +0200)]
broadcom/compiler: update one TMUWT restriction for v71
TMUWT not allowed in the final instruction restriction doesn't apply
for v71.
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Thu, 14 Oct 2021 12:16:40 +0000 (14:16 +0200)]
broadcom/compiler: v71 isn't affected by double-rounding of viewport X,Y coords
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Fri, 8 Oct 2021 13:10:24 +0000 (15:10 +0200)]
broadcom/compiler: generalize check for shaders using pixel center W
V3D 4.x has pixel center W in rf0 and V3D 7.x has it in rf3. We already
account for this when we setup the c->payload_w, so use that.
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Wed, 6 Oct 2021 10:01:10 +0000 (12:01 +0200)]
broadcom/qpu: fail packing on unhandled mul pack/unpack
We are doing this for the ADD alu already and it may be helpful to
identify cases where we have QPU code with pack/unpack modifiers on
MUL opcodes that we then are not packing into the actual QPU
instructions.
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Wed, 6 Oct 2021 07:27:43 +0000 (09:27 +0200)]
broadcom/qpu: add MOV integer packing/unpacking variants
These are new in v71 and cover MOV on both the ADD and the MUL alus.
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Thu, 30 Sep 2021 11:22:48 +0000 (13:22 +0200)]
broadcom/compiler: allow instruction merges in v71
In v3d 4.x there were restrictions based on the number of raddrs used
by the combined instructions, but we don't have these restrictions in
v3d 7.x.
It should be noted that while there are no restrictions on the number
of raddrs addressed, a QPU instruction can only address a single small
immediate, so we should be careful about that when we add support for
small immediates.
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Fri, 22 Oct 2021 11:39:48 +0000 (13:39 +0200)]
broadcom/compiler: don't schedule rf0 writes right after ldvary
ldvary writes rf0 implicitly on the next cycle so they would clash.
This case is not handled correctly by our normal dependency tracking,
which doesn't know anything about delayed writes from instructions
and thinks the rf0 write happens on the same cycle ldvary is emitted.
Fixes (v71):
dEQP-VK.glsl.conversions.matrix_to_matrix.mat2x3_to_mat4x2_fragment
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Tue, 28 Sep 2021 11:37:28 +0000 (13:37 +0200)]
broadcom/compiler: CS payload registers have changed in v71
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Wed, 29 Sep 2021 10:14:04 +0000 (12:14 +0200)]
broadcom/compiler: don't assign rf0 to temps across implicit rf0 writes
In platforms that don't have accumulators and have implicit writes to
the register file we need to be careful and avoid assigning a physical
register to a temp that lives across an implicit write to that same
physical register.
For now, we have the case of implicit writes to rf0 from various
signals, but it should be easy to extend this to include additional
registers if needed.
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Wed, 29 Sep 2021 10:10:31 +0000 (12:10 +0200)]
broadcom/compiler: only handle accumulator classes if present
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Wed, 29 Sep 2021 10:03:50 +0000 (12:03 +0200)]
broadcom/compiler: rename vir_writes_rX to vir_writes_rX_implicitly
Since that represents more accurately what they check..
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Wed, 29 Sep 2021 09:54:18 +0000 (11:54 +0200)]
broadcom/compiler: make vir_write_rX return false on platforms without accums
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Alejandro Piñeiro [Mon, 27 Sep 2021 23:17:08 +0000 (01:17 +0200)]
broadcom/qpu: implement switch rules for fmin/fmax fadd/faddnf for v71
They use the same opcodes, and switch between one and the other based
on raddr.
Note that the rule includes also if small_imm_a/b are used. That is
still not in place so that part is hardcoded. Would be updated later
when small immediates support for v71 gets implemented.
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Mon, 4 Oct 2021 11:07:35 +0000 (13:07 +0200)]
broadcom/qpu: fix packing/unpacking of fmov variants for v71
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Mon, 27 Sep 2021 11:26:04 +0000 (13:26 +0200)]
broadcom/qpu: add new ADD opcodes for FMOV/MOV in v71
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Mon, 27 Sep 2021 09:49:24 +0000 (11:49 +0200)]
broadcom/compiler: prevent rf2-3 usage in thread end delay slots for v71
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Wed, 6 Oct 2021 11:58:00 +0000 (13:58 +0200)]
broadcom/compiler: add a v3d71_qpu_writes_waddr_explicitly helper
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Thu, 23 Sep 2021 09:44:59 +0000 (11:44 +0200)]
broadcom/compiler: implement read stall check for v71
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Thu, 23 Sep 2021 09:19:58 +0000 (11:19 +0200)]
broadcom/compiler: implement "reads/writes too soon" checks for v71
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Alejandro Piñeiro [Wed, 15 Sep 2021 22:49:25 +0000 (00:49 +0200)]
broadcom/compiler: update register classes to not include accumulators on v71
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Alejandro Piñeiro [Wed, 15 Sep 2021 09:12:59 +0000 (11:12 +0200)]
broadcom/qpu_schedule: update write deps for v71
We just need to add a write dep if rf0 is written implicitly.
Note that we don't need to check if we have accumulators when checking
for r3/r4/r5, as v3d_qpu_writes_rX would return false for hw version
that doesn't have accumulators.
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Alejandro Piñeiro [Tue, 14 Sep 2021 23:14:15 +0000 (01:14 +0200)]
broadcom/compiler: payload_w is loaded on rf3 for v71
And in general rf0 is now used for other needs.
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Alejandro Piñeiro [Tue, 14 Sep 2021 08:42:55 +0000 (10:42 +0200)]
broadcom/compiler: add support for varyings on nir to vir generation for v71
Needs update as v71 doesn't have accumulators anymore, and ldvary uses
now rf0 to return the value.
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Alejandro Piñeiro [Wed, 15 Sep 2021 08:55:49 +0000 (10:55 +0200)]
broadcom/qpu: return false on qpu_writes_accumulatorXX helpers for v71
As for v71 doesn't have accumulators (devinfo->has_accumulators set to
false), those methods would always return false.
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Alejandro Piñeiro [Thu, 9 Sep 2021 23:20:44 +0000 (01:20 +0200)]
broadcom/qpu: update disasm_raddr for v71
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Alejandro Piñeiro [Thu, 9 Sep 2021 21:59:28 +0000 (23:59 +0200)]
broadcom/qpu_schedule: add process_raddr_deps
On v71 we don't have muxes, but more raddr. Adding a equivalent add
deps function.
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Alejandro Piñeiro [Wed, 8 Sep 2021 23:18:54 +0000 (01:18 +0200)]
broadcom/compiler: update vir_to_qpu::set_src for v71
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Alejandro Piñeiro [Wed, 8 Sep 2021 22:28:53 +0000 (00:28 +0200)]
broadcom/vir: implement is_no_op_mov for v71
Did some refactoring/splitting.
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Alejandro Piñeiro [Thu, 16 Sep 2021 23:07:06 +0000 (01:07 +0200)]
broadcom/compiler: don't favor/select accum registers for hw not supporting it
Note that what we do is to just return false on the favor/select accum
methods. We could just avoid to call them, but as the select is called
more than once, it is just easier this way.
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Alejandro Piñeiro [Mon, 23 Aug 2021 00:18:43 +0000 (02:18 +0200)]
broadcom/compiler: phys index depends on hw version
For 7.1 there are not accumulators. So we replace the macro with a
function call.
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Iago Toral Quiroga [Sat, 28 Jan 2023 23:27:11 +0000 (00:27 +0100)]
broadcom/compiler: update node/temp translation for v71
As the offset applied needs to take into account if we have
accumulators or not.
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Alejandro Piñeiro [Sat, 7 Aug 2021 00:20:39 +0000 (02:20 +0200)]
broadcom/qpu: add pack/unpack support for v71
Note that we provide new v71 alu pack/unpack methods. As there are a
lot that it is equivalent, initially we tried to use existing methods
as template and add version checks on the existing methods. At some
early point that become just really unreadable, so it become better to
just provide new methods, even if v42 and v71 methods have a really
similar structure.
Note that we have splitted the op tables, and created a two (add/mul)
for v71. As the description struct include versioning info, we could
have just used one table. But, specially with the add table, there are
a lot of differences with v71. So it is slightly tidier this
way. Also, taking into account that we do a linear search on the
tables, this can be even justified by performance.
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Alejandro Piñeiro [Wed, 15 Sep 2021 08:56:43 +0000 (10:56 +0200)]
broadcom/qpu: add qpu_writes_rf0_implicitly helper
On v71 rf0 replaces r5 as the register that gets updated implicitly
with uniform loads, and gets the C coefficient with ldvary. This
helper return if rf0 gets implicitly updated.
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Alejandro Piñeiro [Thu, 16 Sep 2021 23:04:31 +0000 (01:04 +0200)]
broadcom/commmon: add has_accumulators field on v3d_device_info
Even if we can just check for the version on the code, checking for
this field makes several places more readable. So for example, on the
register allocate code we doesn't assign an accumulator because we
don't have accumulators on that hw, instead of because hw version is a
given one.
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Alejandro Piñeiro [Thu, 12 Aug 2021 00:24:02 +0000 (02:24 +0200)]
broadcom/qpu: defining shift/mask for raddr_c/d
On V3D 7.x it replaces mul_a/b and add_a/b
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Alejandro Piñeiro [Thu, 5 Aug 2021 23:33:32 +0000 (01:33 +0200)]
broadcom/qpu: add raddr on v3d_qpu_input
On V3D 7.x mux are not used, and raddr_a/b/c/d are used instead
This is not perfect, as for v71, the raddr_a/b defined at qpu_instr
became superfluous. But the alternative would be to define two
different structs, or even having them defined based on version
ifdefs, so this is a reasonable compromise.
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Alejandro Piñeiro [Thu, 5 Aug 2021 23:22:31 +0000 (01:22 +0200)]
broadcom/qpu: define v3d_qpu_input, use on v3d_qpu_alu_instr
At this point it just tidy up a little the alu_instr structure.
But also serves to prepare the structure for new changes, as 7.x uses
raddr instead of mux, and it is just easier to add the raddr to the
new input structure.
Signed-off-by: Alejandro Piñeiro <apinheiro@igalia.com>
Signed-off-by: Iago Toral Quiroga <itoral@igalia.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Alejandro Piñeiro [Tue, 3 Aug 2021 23:11:16 +0000 (01:11 +0200)]
broadcom/qpu: add v71 signal map
Compared with v41, the differences are:
* 14, 15, 29 and 30 are now about immediate a, b, c, d respectively
* 23 is now reserved. On v42 this was for rotate signals, that are
gone on v71.
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>
Alejandro Piñeiro [Wed, 4 Aug 2021 22:50:12 +0000 (00:50 +0200)]
broadcom/compiler: add small_imm a/c/d on v3d_qpu_sig
small_imm_a, small_imm_c and small_imm_d added on top of the already
existing small_imm_b, as V3D 7.1 defines 4 small immediates, tied to
the 4 raddr. Note that this is only the definition, and just a inst
validation rule to check that are not used before v71. Any real use is
still pending.
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25450>