Alyssa Rosenzweig [Wed, 17 Jul 2019 23:30:09 +0000 (16:30 -0700)]
panfrost: Make scissor test more robust
See v3d implementation.
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Alyssa Rosenzweig [Wed, 17 Jul 2019 23:19:45 +0000 (16:19 -0700)]
panfrost: Use correct NO_DITHER field on MFBD
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Alyssa Rosenzweig [Wed, 17 Jul 2019 22:49:42 +0000 (15:49 -0700)]
panfrost: Implement Z32F(_S8) support
Z32F uses a dediacted float path. Z32F_S8 uses separate stencil planes
in the hardware, lowered via u_transfer_helper.
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Alyssa Rosenzweig [Wed, 17 Jul 2019 22:43:24 +0000 (15:43 -0700)]
panfrost/decode: Don't disassemble NULL shaders
It is legal to load a shader from a NULL address, particularly when the
TILER job is used strictly for effects on the Z/S buffer with 0x0 color
mask. Don't crash the decoder in this case.
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Alyssa Rosenzweig [Wed, 17 Jul 2019 22:42:48 +0000 (15:42 -0700)]
panfrost: Copy stencil front to back if back disabled
When backside stenciling is disabled, backfacing primitives just do the
same thing as frontfacing primitives.
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Jan Zielinski [Wed, 17 Jul 2019 15:22:16 +0000 (17:22 +0200)]
swr/rast: Refactor memory API between rasterizer core and swr
This commit cleans up API between the core of the rasterizer and swr.
Some formatting changes are also done.
Reviewed-by: Alok Hota <alok.hota@intel.com>
Andreas Baierl [Fri, 31 May 2019 07:54:27 +0000 (09:54 +0200)]
lima/ppir: Add gl_PointCoord handling
Treat gl_PointCoord as a system value and
add the necessary bits for correct codegen.
Signed-off-by: Andreas Baierl <ichgeh@imkreisrum.de>
Reviewed-by: Qiang Yu <yuq825@gmail.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
Andreas Baierl [Tue, 4 Jun 2019 11:25:28 +0000 (13:25 +0200)]
gallium: Add PIPE_CAP_TGSI_FS_POINT_IS_SYSVAL
This adds an option to treat gl_PointCoord as a system value.
Signed-off-by: Andreas Baierl <ichgeh@imkreisrum.de>
Reviewed-by: Qiang Yu <yuq825@gmail.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
Andreas Baierl [Tue, 11 Jun 2019 12:59:11 +0000 (14:59 +0200)]
nir/tgsi: Extend tgsi_to_nir.c to support gl_PointCoord as a system value.
Signed-off-by: Andreas Baierl <ichgeh@imkreisrum.de>
Reviewed-by: Eric Anholt <eric@anholt.net>
Andreas Baierl [Tue, 4 Jun 2019 11:24:53 +0000 (13:24 +0200)]
nir: Add gl_PointCoord system value
gl_PointCoord handling needs some special bits set in lima/ppir code
generation. Treating gl_PointCoord as a system value makes it easier
to distinguish from a regular varying.
Signed-off-by: Andreas Baierl <ichgeh@imkreisrum.de>
Reviewed-by: Qiang Yu <yuq825@gmail.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
Andreas Baierl [Tue, 4 Jun 2019 11:23:44 +0000 (13:23 +0200)]
glsl: Optionally declare gl_PointCoord as a system value
Signed-off-by: Andreas Baierl <ichgeh@imkreisrum.de>
Reviewed-by: Qiang Yu <yuq825@gmail.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
Connor Abbott [Sat, 11 May 2019 16:43:30 +0000 (18:43 +0200)]
lima/gp: Fix problem with complex moves
When writing the scheduler, we forgot that you can't read the complex
unit in certain sources because it gets overwritten to 0 or 1. Fixing
this turned out to be possible without giving up and reducing
GPIR_VALUE_REG_NUM to 10, although it was difficult in a way I didn't
expect. There can be at most 4 next-max nodes that can't have moves
scheduled in the complex slot, so it actually isn't a problem for
getting the number of next-max nodes at 5 or lower. However, it is a
problem for stores. If a given node is a next-max node whose move cannot
go in the complex slot *and* is used by a store that we decide to
schedule, we have to reserve one of the non-complex slots for a move
instead of all the slots, or we can wind up in a situation where only
the complex slot is free and we fail the move. This means that we have
to add another term to the reservation logic, for stores whose children
cannot be in the complex slot.
Acked-by: Qiang Yu <yuq825@gmail.com>
Connor Abbott [Thu, 11 Jan 2018 23:35:58 +0000 (18:35 -0500)]
lima/gpir: Rework the scheduler
Now, we do scheduling at the same time as value register allocation. The
ready list now acts similarly to the array of registers in
value_regalloc, keeping us from running out of slots. Before this, the
value register allocator wasn't aware of the scheduling constraints of
the actual machine, which meant that it sometimes chose the wrong false
dependencies to insert. Now, we assign value registers at the same time
as we actually schedule instructions, making its choices reflect reality
much better. It was also conservative in some cases where the new scheme
doesn't have to be. For example, in something like:
1 = ld_att
2 = ld_uni
3 = add 1, 2
It's possible that one of 1 and 2 can't be scheduled in the same
instruction as 3, meaning that a move needs to be inserted, so the value
register allocator needs to assume that this sequence requires two
registers. But when actually scheduling, we could discover that 1, 2,
and 3 can all be scheduled together, so that they only require one
register. The new scheduler speculatively inserts the instruction under
consideration, as well as all of its child load instructions, and then
counts the number of live value registers after all is said and done.
This lets us be more aggressive with scheduling when we're close to the
limit.
With the new scheduler, the kmscube vertex shader is now scheduled in 40
instructions, versus 66 before.
Acked-by: Qiang Yu <yuq825@gmail.com>
Connor Abbott [Mon, 22 Apr 2019 19:54:06 +0000 (21:54 +0200)]
lima/gp: Mark more add-only nodes as maybe-two-slot
Reviewed-by: Qiang Yu <yuq825@gmail.com>
Connor Abbott [Tue, 16 Jan 2018 00:38:17 +0000 (19:38 -0500)]
lima/gpir: Fix some bugs in instruction handling
Reviewed-by: Qiang Yu <yuq825@gmail.com>
Connor Abbott [Fri, 3 Nov 2017 21:34:32 +0000 (17:34 -0400)]
lima: Reintroduce the standalone compiler
I used this to test things without needing to have a device handy.
Acked-by: Qiang Yu <yuq825@gmail.com>
Connor Abbott [Mon, 10 Jun 2019 19:59:12 +0000 (21:59 +0200)]
nir/lower_viewport: Check variable mode first
The location is unused for shader_temp and function_temp variables, and
due to the way we nir_lower_io_to_temproraries demotes shader_out
variables to shader_temp variables, it happened to equal
VARYING_SLOT_POS for the gl_Position temporary, which made this pass
fail with the offline compiler due to this coming before vars_to_ssa.
Reviewed-by: Qiang Yu <yuq825@gmail.com>
Samuel Pitoiset [Thu, 18 Jul 2019 08:14:09 +0000 (10:14 +0200)]
radv/gfx10: set BREAK_WAVE_AT_EOI if TES or GS enable the primitive ID
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Samuel Pitoiset [Thu, 18 Jul 2019 08:14:07 +0000 (10:14 +0200)]
radv/gfx10: move emitting VGT_PRIMITIVEID_EN into the NGG path
And do not emit VGT_GS_MODE which is unnecessary on GFX10.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Samuel Pitoiset [Wed, 17 Jul 2019 13:43:53 +0000 (15:43 +0200)]
radv/gfx10: do not always execute a barrier before the second shader
With NGG, empty waves may still be required to export data.
This fixes dEQP-VK.ycbcr.format.*_unorm.geometry_*.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Samuel Pitoiset [Wed, 17 Jul 2019 08:58:17 +0000 (10:58 +0200)]
radv: fix VGT_GS_MODE if VS uses the primitive ID
Found by inspection.
Cc: <mesa-stable@lists.freedesktop.org>
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Iago Toral Quiroga [Tue, 16 Jul 2019 08:58:55 +0000 (10:58 +0200)]
v3d: emit correct lowering for logic operations with MSAA render targets
v2:
- Drop the writemask from the per-sample color intrinsic (Eric)
Reviewed-by: Eric Anholt <eric@anholt.net>
Iago Toral Quiroga [Tue, 16 Jul 2019 09:20:11 +0000 (11:20 +0200)]
v3d: handle nir_intrinsic_store_tlb_sample_color_v3d
v2:
- Move handling of output intrinsics to ntq_emit_intrinsic() (Eric).
Reviewed-by: Eric Anholt <eric@anholt.net>
Iago Toral Quiroga [Tue, 16 Jul 2019 07:44:52 +0000 (09:44 +0200)]
nir: add a V3D-specific intrinsic for per-sample color writes
For per-sample color writes we need the output intrinsic to pack the
sample index, which is not provided with regular store_output intrinsics
unless we figured out a way to encode it into the base or the offset.
v2:
- Drop the writemask (Eric)
Reviewed-by: Eric Anholt <eric@anholt.net>
Iago Toral Quiroga [Tue, 16 Jul 2019 08:48:06 +0000 (10:48 +0200)]
v3d: implement per-sample tlb color writes
Reviewed-by: Eric Anholt <eric@anholt.net>
Iago Toral Quiroga [Tue, 16 Jul 2019 08:28:53 +0000 (10:28 +0200)]
v3d: refactor the tlb color write code
We want to split the tlb specifier setup from the color writes, because when
we implement per-sample color writes we want to do the latter for all the
samples, but the former only once.
Reviewed-by: Eric Anholt <eric@anholt.net>
Iago Toral Quiroga [Tue, 16 Jul 2019 08:14:17 +0000 (10:14 +0200)]
v3d: move tlb color write emission to a helper function
We will soon be adding per-sample color writes which means additional
complexity and more indentation (we will need another loop to emit
the writes for each individual sample), so this will help keeping
things simple and a bit more readable.
Reviewed-by: Eric Anholt <eric@anholt.net>
Iago Toral Quiroga [Tue, 16 Jul 2019 08:58:12 +0000 (10:58 +0200)]
v3d: implement per-sample tlb color reads
Reviewed-by: Eric Anholt <eric@anholt.net>
Lionel Landwerlin [Wed, 17 Jul 2019 22:00:11 +0000 (01:00 +0300)]
anv: fix format mapping for depth/stencil formats
anv_format is supposed to have a pointer back to the associated
VkFormat, we were missed this for depth/stencil formats.
This doesn't fix anything afaict, but will be needed for future
changes.
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Fixes:
465de47bad70 ("anv: associate vulkan formats with aspects")
Acked-by: Jason Ekstrand <jason@jlekstrand.net>
Dave Airlie [Thu, 18 Jul 2019 06:12:53 +0000 (16:12 +1000)]
radv: put back VGT_FLUSH at ring init on gfx10
I can find no evidence that removing this is a good idea.
Fixes:
9b116173b6a ("radv: do not emit VGT_FLUSH on GFX10")
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Gert Wollny [Tue, 16 Jul 2019 06:05:20 +0000 (08:05 +0200)]
softpipe: Clamp border colors when needed
unorm and snorm require that the border color values are clamped, so when
picking the sampler view copy/clamp the border color from the sampler and
use these adjusted values.
Fixes:
dEQP-GLES31.functional.texture.border_clamp.range_clamp.linear_compressed_color
dEQP-GLES31.functional.texture.border_clamp.range_clamp.linear_snorm_color
dEQP-GLES31.functional.texture.border_clamp.range_clamp.linear_srgb_color
dEQP-GLES31.functional.texture.border_clamp.range_clamp.linear_unorm_color
dEQP-GLES31.functional.texture.border_clamp.range_clamp.nearest_compressed_color
dEQP-GLES31.functional.texture.border_clamp.range_clamp.nearest_snorm_color
dEQP-GLES31.functional.texture.border_clamp.range_clamp.nearest_srgb_color
dEQP-GLES31.functional.texture.border_clamp.range_clamp.nearest_unorm_color
dEQP-GLES31.functional.texture.border_clamp.range_clamp.nearest_unorm_depth
dEQP-GLES31.functional.texture.border_clamp.range_clamp.nearest_unorm_depth_uint_stencil_sample_depth
Signed-off-by: Gert Wollny <gert.wollny@collabora.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Gert Wollny [Mon, 15 Jul 2019 10:51:06 +0000 (12:51 +0200)]
softpipe: set a lower minimum clamp value for texture coordinate border clamp
The value of -0.5f is not small enough to produce negative coordinates,
so lower the minimum clamp value to -1.0f. This fixes a number of tests
from
dEQP-GLES31.functional.texture.border_clamp.*
Signed-off-by: Gert Wollny <gert.wollny@collabora.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Gert Wollny [Mon, 15 Jul 2019 11:46:53 +0000 (13:46 +0200)]
softpipe: Correct repeat-mirror evaluation
when mirroring the texture corrdinates the indices must be mirrored as
well and the half pixel shift must be applied in reverse.
Fixes a number of tests from:
dEQP-GLES31.functional.texture.gather.offset.*
dEQP-GLES31.functional.texture.gather.offsets.*
Signed-off-by: Gert Wollny <gert.wollny@collabora.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Gert Wollny [Fri, 24 May 2019 10:08:26 +0000 (12:08 +0200)]
softpipe: Also mark textures as dirty when updating the framebuffer state
At this point all the draw caches are flushed to the old attached textures,
so the read caches of these textures will need to be updated too.
Fixes:
dEQP-GLES3.functional.fbo.color.repeated_clear.sample.tex2d.*
Signed-off-by: Gert Wollny <gert.wollny@collabora.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Jonathan Marek [Fri, 12 Jul 2019 23:26:32 +0000 (19:26 -0400)]
etnaviv: set DITHER_MODE
This fixes a rendering glitch observed in SDL testscale test, where alpha
blending samples with value (1.0, 1.0, 1.0, 0.0) whitens the target instead
of having no effect.
Signed-off-by: Jonathan Marek <jonathan@marek.ca>
Reviewed-by: Christian Gmeiner <christian.gmeiner@gmail.com>
Jonathan Marek [Fri, 12 Jul 2019 23:25:33 +0000 (19:25 -0400)]
etnaviv: update headers from rnndb
Update to etna_viv commit a16a418.
Signed-off-by: Jonathan Marek <jonathan@marek.ca>
Reviewed-by: Christian Gmeiner <christian.gmeiner@gmail.com>
Jonathan Marek [Wed, 3 Jul 2019 18:06:17 +0000 (14:06 -0400)]
etnaviv: fix blend color on newer GPUs
Newer GPUs use the half float ALPHA_COLOR_EXT register.
Signed-off-by: Jonathan Marek <jonathan@marek.ca>
Reviewed-by: Christian Gmeiner <christian.gmeiner@gmail.com>
Jonathan Marek [Wed, 3 Jul 2019 18:04:20 +0000 (14:04 -0400)]
etnaviv: fix alpha blending cases
We need to check rgb_func/alpha_func when determining if blend or separate
alpha is required.
Signed-off-by: Jonathan Marek <jonathan@marek.ca>
Reviewed-by: Christian Gmeiner <christian.gmeiner@gmail.com>
Jonathan Marek [Wed, 3 Jul 2019 18:01:33 +0000 (14:01 -0400)]
etnaviv: fix polygon offset
Dividing the fui result by 65535 is obviously wrong, and from testing, on
GC7000L at least there is no division by 65535.
Fixes dEQP-GLES2.functional.polygon_offset.fixed16_displacement_with_units
Signed-off-by: Jonathan Marek <jonathan@marek.ca>
Reviewed-by: Christian Gmeiner <christian.gmeiner@gmail.com>
Timothy Arceri [Wed, 17 Jul 2019 04:20:55 +0000 (14:20 +1000)]
radv: dont store disasm string unless keep_shader_info flag set
This fixes the memory use regression from bug 111107.
Fixes:
726a31df705 ("radv: Add the concept of radv shader binaries.")
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=111107
Dave Airlie [Tue, 16 Jul 2019 05:21:56 +0000 (15:21 +1000)]
radv/gfx10: set the pgm rsrc3/4 regs using index sh reg set
This is ported from AMDVLK, it's probably not requires unless
we want to use "real time queues", but it might be nice to just have
in place.
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Dave Airlie [Wed, 17 Jul 2019 04:55:52 +0000 (14:55 +1000)]
radv: use correct register setter for ngg hw addr
this shouldn't matter, but it's good to be correct.
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Eric Anholt [Wed, 17 Jul 2019 19:56:12 +0000 (12:56 -0700)]
freedreno/a6xx: Drop the WFI in the program update stateobj.
Rob Clark thinks this was likely a workaround for our const buffer
update bugs, and now that it's passing tests, we should be able to
drop it.
renderdoc-traces results:
traces/android/clashofclans.rdc: +6.1% +/- 1.1%
traces/android/candycrush.rdc: +5.2% +/- 1.6%
Reviewed-by: Rob Clark <robdclark@gmail.com>
Eric Anholt [Wed, 19 Jun 2019 18:36:29 +0000 (11:36 -0700)]
freedreno/a6xx: Drop the WFI in constant uploads.
Now that the bin vs render constlen is fixed, we can skip these waits.
Improves webgl aquarium performance at 10k fish from 27fps to 33.
Some highlights from renderdoc-traces:
traces/android/minecraft.rdc: +17.1% +/- 3.4%
traces/glmark2/ideas-speed=duration.rdc: +11.6% +/- 2.4%
traces/android/candycrush.rdc: +5.4% +/- 1.1%
traces/android/clashofclans.rdc: +4.4% +/- 1.3%
Reviewed-by: Rob Clark <robdclark@gmail.com>
Eric Anholt [Wed, 26 Jun 2019 22:35:44 +0000 (15:35 -0700)]
freedreno: Assert that we don't exceed constlen.
We actually could go up to vs->constlen in the binning shader on a6xx,
but for sanity let's make sure that we're always under constlen. This
would have caught the bug fixed in
572c76fd8826 ("freedreno: Clamp UBO
uploads to the constlen decided by the shader.")
Reviewed-by: Rob Clark <robdclark@gmail.com>
Eric Anholt [Wed, 26 Jun 2019 23:47:10 +0000 (16:47 -0700)]
freedreno: Fix more constlen overflows.
Fixes constlen overflow in
dEQP-GLES31.functional.shaders.builtin_var.compute.num_work_groups and
dEQP-GLES31.functional.image_load_store.buffer.image_size.readonly_32
and probably others.
Reviewed-by: Rob Clark <robdclark@gmail.com>
Eric Anholt [Wed, 19 Jun 2019 19:00:57 +0000 (12:00 -0700)]
freedreno: Drop stale comment about skipping uploads.
We already skip the upload if it's unused, due to the constlen >
offset check.
Reviewed-by: Rob Clark <robdclark@gmail.com>
Lepton Wu [Wed, 17 Jul 2019 17:02:20 +0000 (10:02 -0700)]
virgl: Set meta data for textures from handle.
The set of meta data was removed by commit 8083464. It broke lots of
dEQP tests when running with pbuffer surface type.
Fixes:
80834640137 ("virgl: remove dead code")
Signed-off-by: Lepton Wu <lepton@chromium.org>
Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Reviewed-by: Chia-I Wu <olvaffe@gmail.com>
Bas Nieuwenhuizen [Wed, 17 Jul 2019 00:58:59 +0000 (02:58 +0200)]
radv: Only save the descriptor set if we have one.
After reset, if valid does not contain the relevant bit the descriptor
can be != NULL but still not be valid.
CC: <mesa-stable@lists.freedesktop.org>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Lionel Landwerlin [Wed, 17 Jul 2019 05:46:53 +0000 (08:46 +0300)]
anv: report timestampComputeAndGraphics true
Spec says :
"timestampComputeAndGraphics specifies support for timestamps on all
graphics and compute queues. If this limit is set to VK_TRUE, all
queues that advertise the VK_QUEUE_GRAPHICS_BIT or
VK_QUEUE_COMPUTE_BIT in the VkQueueFamilyProperties::queueFlags
support VkQueueFamilyProperties::timestampValidBits of at least 36."
On gen7+ this should be true (we only have 32bits of timestamp on
gen6 and below).
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Fixes:
802f00219addb3 ("anv/device: Update features and limits")
Reported-by: Timothy Strelchun <timothy.strelchun@intel.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Rafael Antognolli [Tue, 25 Jun 2019 21:14:22 +0000 (14:14 -0700)]
iris: Enable fast clears on other miplevels and layers than 0.
Until now we only supported fast clear colors on the first miplevel and
layer. The main reason for it is that we can't have different fast clear
values at different levels/layers, since the surface state only supports
one clear value.
We can, however, enable it if we make sure we only use the same value
for all levels/layers, and if one of them changes, we resolve all the
others. We already do that for depth fast clears so hopefully it will be
fine for color fast clears too.
v2: Add check for partial clear too (Ken).
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Rafael Antognolli [Wed, 3 Jul 2019 16:37:47 +0000 (09:37 -0700)]
iris: Allow resolving clear color of CCS_D surfaces.
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Kenneth Graunke [Wed, 17 Jul 2019 20:43:04 +0000 (13:43 -0700)]
iris: Make iris_has_color_unresolved non-static
We want to use this in the transfer code and possibly for fast clears.
Andreas Bergmeier [Sun, 14 Jul 2019 20:23:15 +0000 (22:23 +0200)]
broadcom: Move v3d_get_device_info to common
In common we can use implementation for Vulkan.
Caio Marcelo de Oliveira Filho [Wed, 5 Jun 2019 20:08:56 +0000 (13:08 -0700)]
nir/large_constants: Use dominance information to find more constants
Relax the restriction that all the writes need to be in the first
block: now accept variables that have all the writes in the same
block, and all the reads are dominated by that block.
This let the pass identify large constants that are local to a helper
function. The writes will be at the place that the function is
inlined, possibly not in the first block (but still all in the same
block).
Results for vkpipeline-db in SKL:
total instructions in shared programs: 3624891 -> 3623145 (-0.05%)
instructions in affected programs: 79416 -> 77670 (-2.20%)
helped: 16
HURT: 0
total cycles in shared programs:
1458149667 ->
1458147273 (<.01%)
cycles in affected programs:
30154164 ->
30151770 (<.01%)
helped: 14
HURT: 2
total loops in shared programs: 2437 -> 2437 (0.00%)
loops in affected programs: 0 -> 0
helped: 0
HURT: 0
total spills in shared programs: 8813 -> 8745 (-0.77%)
spills in affected programs: 2894 -> 2826 (-2.35%)
helped: 8
HURT: 0
total fills in shared programs: 23470 -> 23392 (-0.33%)
fills in affected programs: 12248 -> 12170 (-0.64%)
helped: 6
HURT: 2
LOST: 0
GAINED: 0
Results for shader-db in SKL with Iris:
total instructions in shared programs:
15379442 ->
15379392 (<.01%)
instructions in affected programs: 837 -> 787 (-5.97%)
helped: 2
HURT: 2
helped stats (abs) min: 27 max: 27 x̄: 27.00 x̃: 27
helped stats (rel) min: 10.47% max: 10.67% x̄: 10.57% x̃: 10.57%
HURT stats (abs) min: 2 max: 2 x̄: 2.00 x̃: 2
HURT stats (rel) min: 1.23% max: 1.23% x̄: 1.23% x̃: 1.23%
95% mean confidence interval for instructions value: -39.14 14.14
95% mean confidence interval for instructions %-change: -15.51% 6.17%
Inconclusive result (value mean confidence interval includes 0).
total loops in shared programs: 4880 -> 4880 (0.00%)
loops in affected programs: 0 -> 0
helped: 0
HURT: 0
total cycles in shared programs:
370677237 ->
370676567 (<.01%)
cycles in affected programs: 17852 -> 17182 (-3.75%)
helped: 2
HURT: 1
helped stats (abs) min: 338 max: 356 x̄: 347.00 x̃: 347
helped stats (rel) min: 13.98% max: 14.64% x̄: 14.31% x̃: 14.31%
HURT stats (abs) min: 24 max: 24 x̄: 24.00 x̃: 24
HURT stats (rel) min: 0.18% max: 0.18% x̄: 0.18% x̃: 0.18%
total spills in shared programs: 11772 -> 11772 (0.00%)
spills in affected programs: 0 -> 0
helped: 0
HURT: 0
total fills in shared programs: 24948 -> 24948 (0.00%)
fills in affected programs: 0 -> 0
helped: 0
HURT: 0
LOST: 0
GAINED: 0
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Jason Ekstrand [Sat, 13 Jul 2019 23:35:20 +0000 (18:35 -0500)]
intel/fs: Use a strided MOV instead of a conversion for load_* destinations
In many cases, the compiler can just copy-prop the strided MOV whereas
the conversion is a bit trickier. This cuts 5% of the instructions off
of one particular Vulkan CTS test which does lots of load_ssbo.
Reviewed-by: Matt Turner <mattst88@gmail.com>
Jason Ekstrand [Sat, 13 Jul 2019 04:26:59 +0000 (23:26 -0500)]
nir/algebraic: Optimize comparisons and up-casts
These seem like obvious enough optimizations in the world of multiple
integer bit sizes. The only known thing which hits these at the moment
is some Vulkan CTS tests for 16-bit SSBO values which like to up-cast
and check for equality. However, it's something that's bound to come up
as we start seeing more integers in shaders.
The optimizations of comparisons of casted values with constants are
something which we would ideally do with range analysis. However,
lacking that, we can do it in opt_algebraic as long as one side is a
constant.
In dEQP-VK.ssbo.phys.layout.random.16bit.scalar.13, this commit, along
with the previous commit, reduce the number of instructions emitted on
Skylake from 55328 to 44546, a reduction of 20%.
Acked-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Jason Ekstrand [Sat, 13 Jul 2019 04:26:48 +0000 (23:26 -0500)]
nir/algebraic: Optimize comparing unpacked values
We could, in theory, add the same optimization for 64-bit unpack
operations but that's likely to fight with 64-bit integer lowering on
platforms which require it so it will require more infrastructure before
that will be a good idea.
Reviewed-by: Matt Turner <mattst88@gmail.com>
Jason Ekstrand [Sat, 13 Jul 2019 15:57:24 +0000 (10:57 -0500)]
nir/algebraic: Print out the list of transforms in the C file
This helps greatly when debugging algebraic transform generators because
you can now actually see the output and verify that your transforms are
getting generated.
Acked-by: Matt Turner <mattst88@gmail.com>
Jason Ekstrand [Sat, 13 Jul 2019 17:06:00 +0000 (12:06 -0500)]
intel/fs: Properly stride NULL replacement regs in DCE
This fixes some validation errors generated by certain D->W conversions
but is likely not a full solution. Calculating an actual register
stride is a far more complex problem in general and should probably be
handled by the brw_fs_generator.
Reviewed-by: Matt Turner <mattst88@gmail.com>
Eric Anholt [Tue, 16 Jul 2019 19:38:10 +0000 (12:38 -0700)]
nir: Fix nir_lower_alu_to_scalar's instr filtering.
It was checking if the dest or src[0] SSA values were vectors, rather than
whether the ALU op was using the source as a vector resulting in a
nir_fdot4 making it through to vc4 and v3d:
vec1 32 ssa_6 = fdot4 ssa_4.xxxx, ssa_5
Fixes:
c1cffa4249ca ("nir/alu_to_scalar: Use the new NIR lowering framework")
v2: Use Jason's recommendation to look at input_sizes.
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Alyssa Rosenzweig [Tue, 16 Jul 2019 18:36:11 +0000 (11:36 -0700)]
panfrost: Merge varyings_mem into transient buffers
Theoretically we would like these split since varyings can have
specially optimized flags (no map, coherent local). For now, since
neither of these flags is particularly meaningful right now, merge them
together instead of special casing varyings_mem.
Saves upwards of 64MB of RAM per context.
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Lionel Landwerlin [Mon, 15 Jul 2019 06:01:20 +0000 (09:01 +0300)]
vulkan/wsi: update swapchain status on vkQueuePresent
With the following chain of events :
vkQueuePresent()
<- Surface resize
vkQueuePresent()
We should be able to report SUBOPTIMAL or OUT_OF_DATE on the second
vkQueuePresent() call. Currently we only look at X11 events in the
vkAcquireNextImage() path so we're not able to report this.
This change checks the queue of events and process any available ones
to update the swapchain status.
v2: Be consistent about reporting the current error state of the
swapchain (Jason)
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=111097
Cc: <mesa-stable@lists.freedesktop.org>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Samuel Pitoiset [Wed, 17 Jul 2019 12:19:10 +0000 (14:19 +0200)]
radv: add an option for disabling NGG on GFX10
Will be useful for testing the legacy path.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Erik Faye-Lund [Tue, 7 May 2019 10:13:43 +0000 (12:13 +0200)]
softpipe: pass stream-out targets to draw-module early
This is essensially a port of
ed53e61bec9 from LLVMpipe to softpipe,
as it makes things a bit simpler and more performant.
Signed-off-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Reviewed-By: Gert Wollny <gert.wollny@collabora.com>
Alejandro Piñeiro [Sat, 14 Oct 2017 08:52:51 +0000 (10:52 +0200)]
spirv_extensions: i965: initialize SPIR-V extensions
v2: Rebase update after changes on previous patches.
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
Alejandro Piñeiro [Fri, 13 Oct 2017 18:43:01 +0000 (20:43 +0200)]
spirv_extensions: add spirv_supported_extensions on gl_constants
We can use it to get real values for ARB_spirv_extensions methods.
Signed-off-by: Alejandro Piñeiro <apinheiro@igalia.com>
Signed-off-by: Arcady Goldmints-Orlov <agoldmints@igalia.com>
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
Alejandro Piñeiro [Wed, 19 Jun 2019 18:58:30 +0000 (13:58 -0500)]
spirv_extensions: define spirv_extensions_supported
Add a struct to maintain which SPIR-V extensions are supported, and an
utility method to initialize it based on
nir_spirv_supported_capabilities.
v2:
* Fixing code style (Ian Romanick)
* Adding a prefix (spirv) to fill_supported_spirv_extensions (Ian Romanick)
v3: rebase update (nir_spirv_supported_extensions renamed)
v4: include AMD_gcn_shader support
v5: move spirv_fill_supported_spirv_extensions to
src/mesa/main/spirv_extensions.c
Signed-off-by: Alejandro Piñeiro <apinheiro@igalia.com>
Signed-off-by: Arcady Goldmints-Orlov <agoldmints@igalia.com>
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
Alejandro Piñeiro [Fri, 13 Oct 2017 14:17:51 +0000 (16:17 +0200)]
spirv_extensions: add list of extensions and to_string method
Ideally this should be generated somehow. One option would be gather
all the extension dependencies listed on the core grammar, but there
would be the possibility of not including some of the extensions.
Note that spirv-tools is doing it just slightly better, as it has a
hardcoded list of extensions manually took from the registry, that
they parse to get the enum and the to_string method (see
generate_grammar_tables.py).
v2:
* Use a macro to improve readability. (Tapani Pälli)
* Add unreachable on the switch, no default (Eric Engestrom)
* No typedef enum (Ian Romanick)
* Sort extensions names (Ian Romanick)
* Don't add extensions unlikely to be supported by Mesa at any point
(Ian Romanick)
v3: rebase update
v4: Include AMD_gcn_shader
v5: move spirv_extensions_to_string to src/mesa/main/spirv_extensions.c
Signed-off-by: Alejandro Piñeiro <apinheiro@igalia.com>
Signed-off-by: Arcady Goldmints-Orlov <agoldmints@igalia.com>
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
Alejandro Piñeiro [Tue, 10 Oct 2017 14:09:25 +0000 (16:09 +0200)]
spirv_extensions: add GL_ARB_spirv_extensions boilerplate
v2:
* Mention extension gap at gl_API.xml (Emil Velikov)
* Bail with INVALID_ENUM if extension not available on getStringi (Emil Velikov)
* Use EXTRA_EXT macro when defining the extension at
get.c/get_hash_params.py (Emil Velikov)
* Rename source files (spirvextensions.[ch] -> spirv_extensions.[ch]) (Ian)
v3:
* Fix GL_PROGRAM_BINARY_FORMATS glGet query, broken by error on a
previous rebase
v4:
* Fix rebase conflicts on getstring.c after
GL_SHADING_LANGUAGE_VERSION query was added
v5:
* Remove src/mapi/glapi/gen/Makefile.am as it no longer exists in
master
Signed-off-by: Alejandro Piñeiro <apinheiro@igalia.com>
Signed-off-by: Arcady Goldmints-Orlov <agoldmints@igalia.com>
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
Samuel Pitoiset [Tue, 16 Jul 2019 15:11:50 +0000 (17:11 +0200)]
radv/gfx10: implement VK_EXT_post_depth_coverage
I did implement this extension a while ago but it didn't work
on pre GFX10 for some reasons. Now all CTS pass.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Samuel Pitoiset [Tue, 16 Jul 2019 15:35:00 +0000 (17:35 +0200)]
radv/gfx10: disable the TC compat zrange workaround
Unnecessary.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Samuel Pitoiset [Tue, 16 Jul 2019 14:39:16 +0000 (16:39 +0200)]
radv/gfx10: fallback to the legacy path if tess and extreme geometry
This is unsupported and hangs.
This fixes GPU hangs with
dEQP-VK.tessellation.geometry_interaction.limits.output_required_*.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Samuel Pitoiset [Tue, 16 Jul 2019 14:39:15 +0000 (16:39 +0200)]
radv/gfx10: always build the GS copy shader but uses it on-demand
It should be possible to build it on-demand too but it requires
more work. On GFX10, the GS copy shader is required when tess
is enabled with extreme geometry.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Gert Wollny [Thu, 2 May 2019 13:54:55 +0000 (15:54 +0200)]
softpipe: Remove unused static function
Thanks to Eric Engestrom for pointing out that there was something wrong
with that function.
Fixes:
724a73509e1bc1ce3abf9500e457bb2911b642db
softpipe: Prepare handling explicit gradients
Signed-off-by: Gert Wollny <gert.wollny@collabora.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
Caio Marcelo de Oliveira Filho [Wed, 10 Jul 2019 16:00:47 +0000 (09:00 -0700)]
spirv: Bail when we see CounterBuffer decoration
This decoration can be ignored, so we can just skip the next steps.
Otherwise we'd have to also handle it in apply_var_decoration.
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Kenneth Graunke [Mon, 15 Jul 2019 18:51:14 +0000 (11:51 -0700)]
iris: Drop copy and pasted iris_timebase_scale
Lionel moved brw_timebase_scale to gen_device_info_timebase_scale a few
months ago, so we should just use that, and not our own copy in iris.
Jason Ekstrand [Fri, 12 Jul 2019 16:01:40 +0000 (11:01 -0500)]
nir/regs_to_ssa: Handle regs in phi sources properly
Sources of phi instructions act as if they occur at the very end of the
predecessor block not the block in which the phi lives. In order to
handle them correctly, we have to skip phi sources on the normal
instruction walk and handle them as a separate walk over the successor
phis. While registers in phi instructions is a bit of an oddity it can
happen when we temporarily go out-of-SSA for control-flow manipulations.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=111075
Cc: mesa-stable@lists.freedesktop.org
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
Jason Ekstrand [Wed, 10 Jul 2019 22:55:44 +0000 (17:55 -0500)]
spirv: Add a warning for ArrayStride on arrays of blocks
It's disallowed according to the SPIR-V spec or at least I think that's
what the spec says. It's in a section explicitly about explicit layout
of things in the StorageBuffer, Uniform, and PushConstant storage
classes so it's not 100% clear that it applies with other storage
classes. However, it seems like it should apply in general and
violating it can trigger (fairly harmless) asserts in NIR.
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
Caio Marcelo de Oliveira Filho [Fri, 12 Jul 2019 18:33:38 +0000 (11:33 -0700)]
anv: Increase state allocation size limit to 2MB
When running on ICL the
dEQP-VK.ssbo.phys.layout.random.16bit.scalar.13 needs more than 1M for
the shader, so bump it.
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Yevhenii Kolesnikov [Tue, 9 Jul 2019 12:11:15 +0000 (15:11 +0300)]
meta: leaking of BO with DrawPixels
ctx->Unpack.BufferObj wasn't unreferenced.
Fixes:
d492e7b0171 (meta: Fix invalid PBO access from DrawPixels when
trying to just alloc.)
CC: Eric Anholt <eric@anholt.net>
Signed-off-by: Yevhenii Kolesnikov <yevhenii.kolesnikov@globallogic.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
Eric Anholt [Tue, 2 Jul 2019 00:20:55 +0000 (17:20 -0700)]
swrast: Move _mesa_format_pack_colormask() to the only caller.
This avoids needing format_pack to have access to the GLenum return
functions for mesa_format. It seems like an odd function and unlikely
to be reused.
Reviewed-by: Thomas Helland <thomashelland90@gmail.com>
Reviewed-by: Kristian H. Kristensen <hoegsberg@google.com>
Eric Anholt [Mon, 1 Jul 2019 22:43:19 +0000 (15:43 -0700)]
mesa: Give _mesa_format_get_color_encoding a clearer name.
It only returned one of two values.
Reviewed-by: Thomas Helland <thomashelland90@gmail.com>
Reviewed-by: Kristian H. Kristensen <hoegsberg@google.com>
Eric Anholt [Mon, 1 Jul 2019 22:38:56 +0000 (15:38 -0700)]
mesa: Drop redundant checks for sRGB before sRGB to linear conversion.
_mesa_get_srgb_format_linear() just returns the original format if it
wasn't sRGB.
Reviewed-by: Thomas Helland <thomashelland90@gmail.com>
Reviewed-by: Kristian H. Kristensen <hoegsberg@google.com>
Eric Anholt [Mon, 1 Jul 2019 23:45:32 +0000 (16:45 -0700)]
mesa: Fold _mesa_unpack_depth_stencil_row() into its only caller.
This was the last bit of gl.h usage in format packing.
Reviewed-by: Thomas Helland <thomashelland90@gmail.com>
Reviewed-by: Kristian H. Kristensen <hoegsberg@google.com>
Eric Anholt [Mon, 1 Jul 2019 23:27:24 +0000 (16:27 -0700)]
mesa: Convert format_pack/unpack off of GL types.
Reviewed-by: Rob Clark <robdclark@gmail.com>
Reviewed-by: Kristian H. Kristensen <hoegsberg@google.com>
Eric Anholt [Mon, 1 Jul 2019 23:34:23 +0000 (16:34 -0700)]
mesa: Port format_pack/unpack off of _mesa_problem().
unreachable() should be plenty of debug for these.
Reviewed-by: Thomas Helland <thomashelland90@gmail.com>
Reviewed-by: Kristian H. Kristensen <hoegsberg@google.com>
Eric Anholt [Mon, 1 Jul 2019 22:23:19 +0000 (15:23 -0700)]
mesa: Mostly switch Mesa format info off of GL types other than GLenum.
I'm considering moving most of this code to src/util/, and I want that
code to not expose GL types in its interfaces.
Reviewed-by: Rob Clark <robdclark@gmail.com>
Reviewed-by: Kristian H. Kristensen <hoegsberg@google.com>
Eric Anholt [Mon, 1 Jul 2019 21:29:45 +0000 (14:29 -0700)]
mesa: Rename gl_pack typedefs to mesa_pack.
These are packing mesa formats, not a GL format/type.
Reviewed-by: Thomas Helland <thomashelland90@gmail.com>
Reviewed-by: Kristian H. Kristensen <hoegsberg@google.com>
Eric Anholt [Mon, 1 Jul 2019 21:07:30 +0000 (14:07 -0700)]
mesa: Rename gl_format_info to mesa_format_info.
It's about MESA_FORMATs, after all.
Reviewed-by: Thomas Helland <thomashelland90@gmail.com>
Reviewed-by: Kristian H. Kristensen <hoegsberg@google.com>
Eric Anholt [Mon, 1 Jul 2019 20:06:09 +0000 (13:06 -0700)]
state_tracker: Move the format test out to be an actual unit test.
We want errors in the table to show up as unit test failures in MRs.
Also keeps unit test code out of the built drivers.
Reviewed-by: Thomas Helland <thomashelland90@gmail.com>
Reviewed-by: Kristian H. Kristensen <hoegsberg@google.com>
Eric Anholt [Mon, 1 Jul 2019 19:38:41 +0000 (12:38 -0700)]
u_format: Remove pointless comments.
Reviewed-by: Thomas Helland <thomashelland90@gmail.com>
Reviewed-by: Kristian H. Kristensen <hoegsberg@google.com>
Eric Anholt [Thu, 27 Jun 2019 22:53:53 +0000 (15:53 -0700)]
src/util: Switch _mesa_half_to_float() to u_half.h's version.
The two implementations differ across the entire input range only in
that u_half.h preserves mantissa bits for NaNs. The u_half.h version
shaves 15% off of the text size of half_float.o.
Reviewed-by: Thomas Helland <thomashelland90@gmail.com>
Reviewed-by: Kristian H. Kristensen <hoegsberg@google.com>
Eric Anholt [Thu, 27 Jun 2019 22:46:28 +0000 (15:46 -0700)]
u_half_test: Turn it into an actual unit test.
You could break the test and meson test wouldn't complain, since we
returned success either way.
Reviewed-by: Thomas Helland <thomashelland90@gmail.com>
Reviewed-by: Kristian H. Kristensen <hoegsberg@google.com>
Mauro Rossi [Sun, 7 Jul 2019 23:17:22 +0000 (01:17 +0200)]
android: radv/gfx10: generate gfx10_format_table.h
This patch adds the missing building rules for Android,
to avoid following building errors:
In file included from external/mesa/src/amd/vulkan/radv_debug.c:35:
In file included from external/mesa/src/amd/vulkan/radv_debug.h:27:
external/mesa/src/amd/vulkan/radv_private.h:95:10:
fatal error: 'gfx10_format_table.h' file not found
^~~~~~~~~~~~~~~~~~~~~~
1 error generated.
In file included from external/mesa/src/amd/vulkan/radv_android.c:31:
external/mesa/src/amd/vulkan/radv_private.h:95:10:
fatal error: 'gfx10_format_table.h' file not found
^~~~~~~~~~~~~~~~~~~~~~
1 error generated.
Fixes:
3dc5ec5d16 ("radv/gfx10: generate gfx10_format_table.h")
Signed-off-by: Mauro Rossi <issor.oruam@gmail.com>
Acked-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Rob Clark [Mon, 15 Jul 2019 16:05:55 +0000 (09:05 -0700)]
mesa/st: add sampler uniforms
Add sampler uniforms for the UV plane(s), so driver can count the
uniforms and get the correct sampler count.
Fixes lowered YUV on a6xx which actually wants to know # of samplers.
Signed-off-by: Rob Clark <robdclark@chromium.org>
Reviewed-by: Kristian H. Kristensen <hoegsberg@google.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
Rob Clark [Fri, 12 Jul 2019 16:17:30 +0000 (09:17 -0700)]
egl/android: handle multi-fd native windows
We can hit multi-fd EGL_NATIVE_BUFFER_ANDROID case when the native
android buffer is YUV. So we need to handle that.
Currently this went unnoticed because, even though we have two or
three fd's for YUV native android buffers, they all reference the
same backing buffer. But we really shouldn't rely on that.
Signed-off-by: Rob Clark <robdclark@chromium.org>
Reviewed-by: Eric Engestrom <eric@engestrom.ch>
Reviewed-by: Kristian H. Kristensen <hoegsberg@google.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
Jason Ekstrand [Thu, 11 Jul 2019 21:59:31 +0000 (16:59 -0500)]
st,i965: Stop looping on 64-bit lowering
Now that the 64-bit lowering passes do a complete lowering in one go, we
don't need to loop anymore. We do, however, have to ensure that int64
lowering happens after double lowering because double lowering can
produce int64 ops.
Reviewed-by: Eric Anholt <eric@anholt.net>
Jason Ekstrand [Thu, 11 Jul 2019 22:06:31 +0000 (17:06 -0500)]
nir/lower_doubles: Handle fdiv and fsub directly
Reviewed-by: Eric Anholt <eric@anholt.net>
Jason Ekstrand [Thu, 11 Jul 2019 21:54:01 +0000 (16:54 -0500)]
nir/lower_doubles: Use the new NIR lowering framework
One advantage of this is that we no longer need to run in a loop because
the new framework handles lowering instructions added by lowering.
Reviewed-by: Eric Anholt <eric@anholt.net>