incomplete.
Update the names, decoder, and comments to reflect this new
understanding.
Signed-off-by: Alyssa Rosenzweig
---
.../drivers/panfrost/include/panfrost-job.h | 40 --
src/gallium/drivers/panfrost/pan_context.c| 52 ++-
.../drivers/panfrost/pandecod
e'll use these sizes to statically
allocate the polygon list on a per-framebuffer basis.
Alyssa Rosenzweig (11):
panfrost: Integrate kernel names for tiler FBD
panfrost: Add notes about the tiler allocations
panfrost: Rename tiler fields per tiler research
panfrost: Document tile size
Signed-off-by: Alyssa Rosenzweig
---
src/gallium/drivers/panfrost/pan_context.c | 5 +
1 file changed, 5 insertions(+)
diff --git a/src/gallium/drivers/panfrost/pan_context.c
b/src/gallium/drivers/panfrost/pan_context.c
index a30b9e29701..d1e5b4ce647 100644
--- a/src/gallium/drivers
Signed-off-by: Alyssa Rosenzweig
---
src/gallium/drivers/panfrost/pan_context.c | 21 -
1 file changed, 16 insertions(+), 5 deletions(-)
diff --git a/src/gallium/drivers/panfrost/pan_context.c
b/src/gallium/drivers/panfrost/pan_context.c
index 0363591f79f..ecb68c990a0
Quite a bit of refactoring in the main driver will be necessary to make
use of this effectively, so the implementation is incomplete.
Signed-off-by: Alyssa Rosenzweig
---
src/gallium/drivers/panfrost/pan_tiler.c | 21 +
1 file changed, 21 insertions(+)
diff --git a/src
As per the notes at the beginning of pan_tiler.c, we implement a routine
to calculate the size of the polygon list header given the framebuffer
dimensions and the provided hierarchy mask.
Signed-off-by: Alyssa Rosenzweig
---
src/gallium/drivers/panfrost/meson.build | 3 +-
src/gallium/drivers
This is a bit of a hack, but it gets the point across.
Signed-off-by: Alyssa Rosenzweig
---
src/gallium/drivers/panfrost/pan_context.c | 7 ++-
src/gallium/drivers/panfrost/pan_tiler.c | 14 ++
2 files changed, 20 insertions(+), 1 deletion(-)
diff --git a/src/gallium/drivers
Just for readability.
Signed-off-by: Alyssa Rosenzweig
---
src/gallium/drivers/panfrost/pan_context.c | 14 +++---
src/gallium/drivers/panfrost/pan_context.h | 3 +--
src/gallium/drivers/panfrost/pan_drm.c | 2 +-
3 files changed, 9 insertions(+), 10 deletions(-)
diff --git a/src
This series includes misc. fixes to improve robustness for for more
complex apps.
Alyssa Rosenzweig (6):
panfrost: Identify and decode mfbd_flags
panfrost: Disable the tiler for clear-only jobs
panfrost: Improve viewport (clipping) robustness
panfrost: Flush scanout too
panfrost: Remove
On more complex apps (possibly using desktop GL specific extensions?),
our viewport code was getting wacky results for unclear reasons. Let's
be a little less wacky.
Signed-off-by: Alyssa Rosenzweig
---
src/gallium/drivers/panfrost/pan_context.c | 41 ++
1 file change
To do so, we route some basic information through to the FBD creation
routines (currently just a binary toggle of "has draws?"). Eventually,
more refactoring will enable dynamic hierarchy mask selection, but right
now we do the most basic.
Signed-off-by: Alyssa Rosenzweig
---
s
In some cases, Gallium can give us bad info about the texture count,
counting some NULL textures. We pass Gallium's info to the hardware
blindly, which can confuse the hardware in edge cases. This patch
adjusts accordingly.
---
src/gallium/drivers/panfrost/pan_context.c | 47 +-
This worked around a bug in ld versions of Panfrost. Nowadays, its
presence is, at best, *creating* bugs. Let's wack it.
Signed-off-by: Alyssa Rosenzweig
---
src/gallium/drivers/panfrost/pan_context.c | 4
1 file changed, 4 deletions(-)
diff --git a/src/gallium/drivers/pan
Previously known as the unk3 field.
Signed-off-by: Alyssa Rosenzweig
---
.../drivers/panfrost/include/panfrost-job.h | 4 ++--
src/gallium/drivers/panfrost/pan_mfbd.c | 12 ++--
src/gallium/drivers/panfrost/pandecode/decode.c | 16 ++--
3 files changed, 22
In a poorly coded app, the framebuffer can be partially drawn, an FBO
switched, switch back to the framebuffer and keep drawing, etc.
Reordering would fix this, but for now we need to just be careful about
flushing scanout too.
Signed-off-by: Alyssa Rosenzweig
---
src/gallium/drivers/panfrost
> Except it would have to be util_is_power_of_two_or_zero because
> util_is_power_of_two(0x + 1) is false. :)
Corner cases, corner cases!
> Is there actually a 2/2 for this? We usually wouldn't land something
> like this without a caller.
The 1/2 on there was accidentally, sorry. The us
Honestly, maybe I should just retract the patch. What's the lingo for
that? :P
signature.asc
Description: PGP signature
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/mesa-dev
Signed-off-by: Alyssa Rosenzweig
---
src/gallium/drivers/panfrost/midgard/midgard_compile.c | 10 ++
1 file changed, 10 insertions(+)
diff --git a/src/gallium/drivers/panfrost/midgard/midgard_compile.c
b/src/gallium/drivers/panfrost/midgard/midgard_compile.c
index 7d7bda6ee12
Exactly what it says on the tin. Decode them and implement them.
Alyssa Rosenzweig (8):
panfrost/midgard: Add swizzle_of/mask_of helpers
panfrost/midgard: Fix 3D texture masks/swizzles
panfrost: Specify 3D in texture descriptor
panfrost: Implement 3D texture resource management
panfrost
Signed-off-by: Alyssa Rosenzweig
---
src/gallium/drivers/panfrost/pan_context.c | 11 +++
1 file changed, 11 insertions(+)
diff --git a/src/gallium/drivers/panfrost/pan_context.c
b/src/gallium/drivers/panfrost/pan_context.c
index 4541b84754c..ec0e4ef7876 100644
--- a/src/gallium
Signed-off-by: Alyssa Rosenzweig
---
src/gallium/drivers/panfrost/pan_resource.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/src/gallium/drivers/panfrost/pan_resource.c
b/src/gallium/drivers/panfrost/pan_resource.c
index 7c0d54a1f9f..81a74735592 100644
--- a/src
Signed-off-by: Alyssa Rosenzweig
---
src/gallium/drivers/panfrost/include/panfrost-job.h | 3 +--
src/gallium/drivers/panfrost/pandecode/decode.c | 6 --
2 files changed, 5 insertions(+), 4 deletions(-)
diff --git a/src/gallium/drivers/panfrost/include/panfrost-job.h
b/src/gallium
Signed-off-by: Alyssa Rosenzweig
---
.../drivers/panfrost/midgard/midgard_compile.c| 11 ---
1 file changed, 8 insertions(+), 3 deletions(-)
diff --git a/src/gallium/drivers/panfrost/midgard/midgard_compile.c
b/src/gallium/drivers/panfrost/midgard/midgard_compile.c
index
Passes dEQP-GLES3.functional.texture.format.unsized.*3d*
Signed-off-by: Alyssa Rosenzweig
---
src/gallium/drivers/panfrost/pan_resource.c | 47 ++---
1 file changed, 42 insertions(+), 5 deletions(-)
diff --git a/src/gallium/drivers/panfrost/pan_resource.c
b/src/gallium/drivers
Signed-off-by: Alyssa Rosenzweig
---
src/gallium/drivers/panfrost/include/panfrost-job.h | 3 +++
src/gallium/drivers/panfrost/pan_context.c | 2 +-
2 files changed, 4 insertions(+), 1 deletion(-)
diff --git a/src/gallium/drivers/panfrost/include/panfrost-job.h
b/src/gallium/drivers
These make manipulating vectors in the Midgard compiler easier.
Signed-off-by: Alyssa Rosenzweig
---
.../panfrost/midgard/midgard_compile.c| 36 +++
1 file changed, 30 insertions(+), 6 deletions(-)
diff --git a/src/gallium/drivers/panfrost/midgard/midgard_compile.c
b
Tag to the 3D texture series. With this it passes 100% of
dEQP-GLES3.functional.texture.format.unsized.*
I plan to squash this in but don't want to resend the whole series just
for this.
Signed-off-by: Alyssa Rosenzweig
---
src/gallium/drivers/panfrost/pan_resource.c | 2 ++
1 file chang
Reviewed-by, seems like a good idea to have the full set :)
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/mesa-dev
I'm not sure if there is a more idiomatic "NIR" way to express the body
of nir_lower_txs_lod, so let's wait to hear some feedback. In principle,
I think this is a great idea :)
signature.asc
Description: PGP signature
___
mesa-dev mailing list
mesa-dev@
This conflicts with texture bias/LOD work (in a functional sense, not
a git sense). We can probably reuse the switch from the previous
function rather than duplicating the op list :)
signature.asc
Description: PGP signature
___
mesa-dev mailing list
mes
Reviewed-by: Alyssa Rosenzweig , thank
you!
signature.asc
Description: PGP signature
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/mesa-dev
> +/*
> + * Note: the first call is here to lower RECT and TXP, the second one
> + * to lower the TXS(lod) instructions generated by the RECT lowering
> + * done in the first pass.
> + *
> + * FIXME: we should probably have a
> + *
> +
> No problem, I'll rebase once this work has landed.
I thought it has?
> Hm, the list of allowed ops is likely to differ (doesn't make sense to
> allow txs instructions when building a midgard tex instruction) so I'm
> not sure it's a good idea to share this switch statement.
Well, I was maybe m
Reviewed-by, thank you!
signature.asc
Description: PGP signature
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/mesa-dev
> Note that I don't have write permissions to the mesa tree, so I'll need
> someone to apply the patch once it's considered ready to be merged.
I can apply v2 once it has the necessary R-b's (Jason's will do, of
course :) )
signature.asc
Description: PGP signature
___
From: Connor Abbott
This effectively does the opposite of nir_lower_alus_to_scalar, trying
to combine per-component ALU operations with the same sources but
different swizzles into one larger ALU operation. It uses a similar
model as CSE, where we do a depth-first approach and keep around a hash
Enable nir_opt_vectorize.
Signed-off-by: Alyssa Rosenzweig
---
src/gallium/drivers/panfrost/midgard/midgard_compile.c | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/src/gallium/drivers/panfrost/midgard/midgard_compile.c
b/src/gallium/drivers/panfrost/midgard
Reviewed-by: Alyssa Rosenzweig , thank
you!
signature.asc
Description: PGP signature
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/mesa-dev
Not sure what would have fixed any of these but sure, A-b, I won't say
no to fixed tests ^_^
What about the fragment_ops.scissor.* tests?
signature.asc
Description: PGP signature
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
https://lists.fr
> Don't know, what about them? They seem to pass.
There were a few that were failing before and have now been fixe-- oh
wait, I remember to update the expectations myself. Never mind! :)
signature.asc
Description: PGP signature
___
mesa-dev mailing lis
Woah :)
I'm okay with this if CI is. I recall having issues when I tried this
but maybe I'm just bad at memory.
Actually, maybe I don't recall issues. Did I even try this?
Like I said. Bad at memory.
R-b, thank you ^^
signature.asc
Description: PGP signature
__
2.x, which matters for glamor.
Alyssa Rosenzweig (9):
panfrost: Refactor texture targets
panfrost: Decode rendering block type
panfrost: Implement tiled rendering
panfrost: Add sRGB render target flag
panfrost: Implement sRGB texturing
panfrost: Specify sRGB in the render target
panfro
otice the sRGB enable bit, which we document and
decode as well here so we don't forget about it.
Signed-off-by: Alyssa Rosenzweig
---
.../drivers/panfrost/include/panfrost-job.h | 23 +-
.../drivers/panfrost/midgard/disassemble.c| 12
.../drivers/panfrost/midgard/mi
We already can sample from Mali's linear/tiled encoding (the one from
Utgard -- AFBC is mostly unrelated); let's be able to render to it as
well.
Signed-off-by: Alyssa Rosenzweig
---
src/gallium/drivers/panfrost/pan_mfbd.c | 4
1 file changed, 4 insertions(+)
diff --git a/s
Signed-off-by: Alyssa Rosenzweig
---
src/gallium/drivers/panfrost/include/panfrost-job.h | 1 +
src/gallium/drivers/panfrost/pandecode/decode.c | 1 +
2 files changed, 2 insertions(+)
diff --git a/src/gallium/drivers/panfrost/include/panfrost-job.h
b/src/gallium/drivers/panfrost/include
The performance impact is slightly mitigated by tiling the render
target, but it's undeniably still slow compared to AFBC. Unfortunately,
it doesn't look like AFBC and sRGB play nice...
Signed-off-by: Alyssa Rosenzweig
---
src/gallium/drivers/panfrost/pan_afbc.c | 7 +++
1 file
Signed-off-by: Alyssa Rosenzweig
---
src/gallium/drivers/panfrost/pan_context.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/gallium/drivers/panfrost/pan_context.c
b/src/gallium/drivers/panfrost/pan_context.c
index 4d9705671bd..5dac0820920 100644
--- a/src/gallium
Now that sRGB formats are supported for both rendering and sampling,
advertise support.
Signed-off-by: Alyssa Rosenzweig
---
src/gallium/drivers/panfrost/pan_screen.c | 4
1 file changed, 4 deletions(-)
diff --git a/src/gallium/drivers/panfrost/pan_screen.c
b/src/gallium/drivers/panfrost
Signed-off-by: Alyssa Rosenzweig
---
src/gallium/drivers/panfrost/pan_mfbd.c | 5 -
1 file changed, 4 insertions(+), 1 deletion(-)
diff --git a/src/gallium/drivers/panfrost/pan_mfbd.c
b/src/gallium/drivers/panfrost/pan_mfbd.c
index f9a69e47e45..c082d0b91a3 100644
--- a/src/gallium/drivers
A mode for rendering tiled/uncompressed was noticed, so we reshuffle the
MFBD render target definitions to explicitly include block type.
Signed-off-by: Alyssa Rosenzweig
---
.../drivers/panfrost/include/panfrost-job.h | 14 +---
src/gallium/drivers/panfrost/pan_mfbd.c | 8
For fixed-function, we have hardware to handle sRGB so we just set a
flag. For blend shaders, it's rather more involved; this is currently
unimplemented. Assert it out for now; we don't need it quite yet.
Signed-off-by: Alyssa Rosenzweig
---
.../drivers/panfrost/include/panf
Midgard texture ops taking bias/LOD(/offset) have both immediate and
register modes, depending on whether the argument is constant or not.
Rather than always using register mode and emitting dumb modes all over
the place, support packing constant bias/LOD.
Alyssa Rosenzweig (3):
panfrost
Signed-off-by: Alyssa Rosenzweig
---
.../panfrost/midgard/midgard_compile.c| 34 +++
1 file changed, 34 insertions(+)
diff --git a/src/gallium/drivers/panfrost/midgard/midgard_compile.c
b/src/gallium/drivers/panfrost/midgard/midgard_compile.c
index 1cc551b603c
Signed-off-by: Alyssa Rosenzweig
---
src/gallium/drivers/panfrost/midgard/disassemble.c | 7 ---
src/gallium/drivers/panfrost/midgard/midgard.h | 2 +-
2 files changed, 5 insertions(+), 4 deletions(-)
diff --git a/src/gallium/drivers/panfrost/midgard/disassemble.c
b/src/gallium/drivers
Signed-off-by: Alyssa Rosenzweig
---
.../panfrost/midgard/midgard_compile.c| 97 +--
1 file changed, 48 insertions(+), 49 deletions(-)
diff --git a/src/gallium/drivers/panfrost/midgard/midgard_compile.c
b/src/gallium/drivers/panfrost/midgard/midgard_compile.c
index
Looking at the scissor, we can discard some tiles. We specifially don't
care about the scissor on the wallpaper, since that's a no-op if the
entire tile is culled.
Signed-off-by: Alyssa Rosenzweig
---
src/gallium/drivers/panfrost/pan_context.c | 9 +
src/gallium/driver
> Actually, I think this is the commit that introduced these regressions:
>
> panfrost: Figure out job requirements in pan_job.c
...Hmm.
signature.asc
Description: PGP signature
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
https://lists.fr
This series includes a number of assorted improvements to the decode
infrastructure to make compute shaders easier to read.
Alyssa Rosenzweig (10):
panfrost/decode: Identify "compute FBD"
panfrost/decode: Expand pointers in compute descriptor
panfrost/decode: Show int uniforms
Just as an aid.
Signed-off-by: Alyssa Rosenzweig
---
.../drivers/panfrost/include/panfrost-job.h | 11 ++-
.../drivers/panfrost/pandecode/decode.c | 30 +++
2 files changed, 34 insertions(+), 7 deletions(-)
diff --git a/src/gallium/drivers/panfrost/include/panfrost
In OpenGL, uniforms generally represent fp32 vec4s (at least in highp
mode). In OpenCL, they represent vec2s of 64-bit pointers.
Signed-off-by: Alyssa Rosenzweig
---
.../drivers/panfrost/pandecode/decode.c | 25 ++-
1 file changed, 24 insertions(+), 1 deletion(-)
diff
Looking at internal evidence (later fields including a literal other
compute job inception-style, seeming memory corruption, no clear
function, and the field after this being a pointer to *itself*), it
looks like this is really a much smaller descriptor.
Signed-off-by: Alyssa Rosenzweig
---
src
Float is ambiguous.
Signed-off-by: Alyssa Rosenzweig
---
src/gallium/drivers/panfrost/pandecode/decode.c | 12
1 file changed, 8 insertions(+), 4 deletions(-)
diff --git a/src/gallium/drivers/panfrost/pandecode/decode.c
b/src/gallium/drivers/panfrost/pandecode/decode.c
index
This is all zero for anything but fragment shaders.
Signed-off-by: Alyssa Rosenzweig
---
.../drivers/panfrost/pandecode/decode.c | 72 ---
1 file changed, 48 insertions(+), 24 deletions(-)
diff --git a/src/gallium/drivers/panfrost/pandecode/decode.c
b/src/gallium/drivers
There is fundamentally not a framebuffer associated with a compute job.
Allocate a new structure for it so we don't mess up graphics when
decoding.
Signed-off-by: Alyssa Rosenzweig
---
.../drivers/panfrost/include/panfrost-job.h | 7 ++
.../drivers/panfrost/pandecode/decode.c
Signed-off-by: Alyssa Rosenzweig
---
src/gallium/drivers/panfrost/pandecode/decode.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/gallium/drivers/panfrost/pandecode/decode.c
b/src/gallium/drivers/panfrost/pandecode/decode.c
index 76baa06dda2..2c74d807d63 100644
--- a
Signed-off-by: Alyssa Rosenzweig
---
.../drivers/panfrost/pandecode/decode.c | 22 ++-
1 file changed, 16 insertions(+), 6 deletions(-)
diff --git a/src/gallium/drivers/panfrost/pandecode/decode.c
b/src/gallium/drivers/panfrost/pandecode/decode.c
index 9b04646f43f
These do more harm than good at this point.
Signed-off-by: Alyssa Rosenzweig
---
.../drivers/panfrost/pandecode/decode.c | 20 ---
1 file changed, 20 deletions(-)
diff --git a/src/gallium/drivers/panfrost/pandecode/decode.c
b/src/gallium/drivers/panfrost/pandecode
Signed-off-by: Alyssa Rosenzweig
---
src/gallium/drivers/panfrost/pandecode/decode.c | 4
1 file changed, 4 deletions(-)
diff --git a/src/gallium/drivers/panfrost/pandecode/decode.c
b/src/gallium/drivers/panfrost/pandecode/decode.c
index ce948f9ce02..5bc6dd84c60 100644
--- a/src/gallium
NAK. There's a patch for this I pushed to your repo which should fix
these regressions (they were legitimate bugs that had nothing to do with
autovectorization and would have cropped up anyhow).
signature.asc
Description: PGP signature
___
mesa-dev mail
How about a partial revert: call panfrost_job_set_requirements() in
emit_for_draw(), so the flow is the same, but we still have the code
logically with the other panfrost_job code?
signature.asc
Description: PGP signature
___
mesa-dev mailing list
mesa-
> dEQP-GLES2.functional.shaders.matrix.mul.dynamic_highp_mat4_vec4_vertex
No clue what's up here, but I suppose A-b for the moment :(
signature.asc
Description: PGP signature
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
https://lists.freede
Oh, this looks lovely! Thank you!
One concern I have is that we're leaking work_vec* for
REGS_VARIANTS_COUNT-1 variants, since we reinitialize those classes for
each RA variant (I believe that is correct behaviour) but only store the
last one (I believe that is incorrect).
Perhaps we should make
Makes sense, R-b :)
signature.asc
Description: PGP signature
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/mesa-dev
R-b, good one :)
signature.asc
Description: PGP signature
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/mesa-dev
R-b, though let's wait for Rohan :)
signature.asc
Description: PGP signature
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/mesa-dev
Fixes regression in shaders using ball/etc.
v2: Don't also regress every other swizzle in Equestria.
Signed-off-by: Alyssa Rosenzweig
---
.../panfrost/midgard/midgard_compile.c| 55 +--
1 file changed, 39 insertions(+), 16 deletions(-)
diff --git a/src/gallium/dr
Looking at the scissor, we can discard some tiles. We specifially don't
care about the scissor on the wallpaper, since that's a no-op if the
entire tile is culled.
Signed-off-by: Alyssa Rosenzweig
---
src/gallium/drivers/panfrost/pan_context.c | 9 +
src/gallium/driver
> Isn't enough with ctx->blitter->running ?
What if we're legitimately blitting something?
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/mesa-dev
st).
Incidentally, this moves quite a bit of state from the driver context to
the batch, which helps with Rohan's refactor to eventually permit
pipelining across framebuffers (one important outstanding optimization
for FBO-heavy workloads).
Signed-off-by: Alyssa Rosenzweig
---
src/g
A-b
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/mesa-dev
operations (like wallpaper
blits) when reading from an uninitialized buffer; since it's
uninitialized, the results of these operations are undefined, and it's
legal to take the fast path ^_^
Signed-off-by: Alyssa Rosenzweig
---
src/gallium/drivers/panfrost/pan_context.c | 9 +
s
Just a fun optimization, shaves some cycles off glmark shaders.
Alyssa Rosenzweig (3):
panfrost/midgard: Share swizzle/mask code
panfrost/midgard: Share swizzle compose
panfrost/midgard: Merge embedded constants
.../drivers/panfrost/midgard/helpers.h| 60
Signed-off-by: Alyssa Rosenzweig
---
.../drivers/panfrost/midgard/helpers.h| 29 +++
.../drivers/panfrost/midgard/midgard_ra.c | 9 +-
2 files changed, 30 insertions(+), 8 deletions(-)
diff --git a/src/gallium/drivers/panfrost/midgard/helpers.h
b/src/gallium
Now that scoreboarding is working, this is trivial.
Alyssa Rosenzweig (2):
panfrost: Implement rasterizer discard
panfrost: Conditionally submit fragment job
src/gallium/drivers/panfrost/pan_context.c | 14 --
src/gallium/drivers/panfrost/pan_drm.c | 5 -
2 files
Signed-off-by: Alyssa Rosenzweig
---
.../drivers/panfrost/midgard/helpers.h| 31 +++
.../panfrost/midgard/midgard_compile.c| 29 -
2 files changed, 31 insertions(+), 29 deletions(-)
diff --git a/src/gallium/drivers/panfrost/midgard/helpers.h
b
If there are no tiling jobs and no clears, there is no need to submit a
fragment job (relevant for transform feedback).
Signed-off-by: Alyssa Rosenzweig
---
src/gallium/drivers/panfrost/pan_drm.c | 5 -
1 file changed, 4 insertions(+), 1 deletion(-)
diff --git a/src/gallium/drivers
fuse into a single (pipelined!) bundle:
vmul.fmul [temp], [a], r26.xyyz
vadd.fadd [out], [temp], r26.zwwy
fconstants 1.5, 0.5, 1.0, 2.3
Signed-off-by: Alyssa Rosenzweig
---
.../panfrost/midgard/midgard_schedule.c | 85 ++-
1 file changed, 66 insertions(+), 19 dele
D'aww, look cute that is now that scoreboarding is setup.
Signed-off-by: Alyssa Rosenzweig
---
src/gallium/drivers/panfrost/pan_context.c | 14 --
1 file changed, 12 insertions(+), 2 deletions(-)
diff --git a/src/gallium/drivers/panfrost/pan_context.c
b/src/gallium/dr
> Very nice, this will fix a few tests.
Ah, of course :)
> Good riddance!
To be fair, a lot of that just got moved in with panfrost_job.. :)
> Can we remove now the has_draws arg?
Maybe? The has_draws arg controls a lot of things beyond just, you
know, having draws. The whole API there nee
R-b, that's long overdue seeing as the DRM driver has a totally
different (better? :) ) interface anyway (Thank you!)
signature.asc
Description: PGP signature
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailma
Oh, nice! I'm going to give this a tentative Reviewed-by. The changes
themselves look good (and there shouldn't be functional changes). I do
want this to happen; I just need to give some thought to ensuring that
yes, this is really what we want to do.
I'm having troubles thinking of when we might
901 - 991 of 991 matches
Mail list logo
