From empirical tests (on a660) R8G8 with UBWC enabled requires 256b
alignment, otherwise there would be a GPU fault during blits.
Set alignment to 4096 for all UBWC images since that's what blob does
and this area is heavily undertested.
Fixes GPU fault in Borderlands 3 running through DXVK.
cc: mesa-stable
Signed-off-by: Danylo Piliaiev <dpiliaiev@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19298>
This involves computing the significand with a 64-bit precision type,
and implementing the normalization and packing manually instead of
relying on u2f32, since the significand can no longer be represented
as a 32-bit integer. This fixes 64-bit integer to 64-bit float
conversions on devices that support 64-bit float natively but lack
64-bit integer support, like Intel MTL hardware.
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com> (v1)
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19128>
The existing code for this appears to work okay for conversions
involving 64-bit floats, relax the assert and enable the lowering
path. This fixes 64-bit float to 64-bit integer integer conversions
on devices that have native support for 64-bit floats but lack 64-bit
integer support, like Intel MTL hardware.
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19128>
Bifrost onwards handle this in hardware, and the Midgard lowering isn't
too terrible. Enable the format, otherwise desktop GL apps such as
Hacknet try to render to the format and get an incomplete framebuffer.
Cc stable because apparently we've been advertising this format
unintentionally as a result of some other interaction? Unclear how
Hacknet is hitting this, maybe it's an app bug. Shrug, it's not a big
deal regardless.
Additionally, we need to restrict texturing from 32-bit normalized due
to a restriction added with the v7 pixel format fiasco. That means
restricting rendering to 32-bit normalized on v7 onwards.
Closes: #7251
Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Tested-by: Dang Huynh <danct12@disroot.org>
Cc: mesa-stable
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19358>
Trace-based testing has not worked for Panfrost. It was a neat
experiment, and I'm glad we tried it, but the results have been mostly
negative for the driver. Disable the trace-based tests.
For testing that specific API features work correctly, we run the
conformance tests (dEQP), which are thorough for OpenGL ES. For big GL
features, we run Piglit, and if there are big GL features that we are
not testing adequately, we should extend Piglit for these. For
fine-grained driver correctness, we are already covered.
Where trace-based testing can fit in is as a smoke test, ensuring that
the overall rendering of complex scenes does not regress. In principle,
that's a lovely idea, but the current implementation has not worked out
for Panfrost thus far. The crux of the issue is that the trace based
tests are based on checksums, not fuzzy-compared reference images. That
requires updating checksums any time rendering changes. However, a
rendering change to a trace is NOT a regression. The behaviour of OpenGL
is specified very loosely. For a given trace, there are many different
valid checksums. That means that correct changes to core code frequently
fail CI after running through the rest of CI, only because a checksum
changed in a still correct way. That's a pain to deal with, exacerbated
by rebase pains, and provides negative value to the project. Some recent
examples of this I've hit in the past two weeks alone:
panfrost: Enable rendering to 16-bit and 32-bit
4b49241f7d ("panfrost: Use proper formats for pntc varying")
ac2964dfbd ("nir: Be smarter fusing ffma")
The last example were virgl traces, but were especially bad: due to a
rebase fail, I had to update traces /twice/, wasting two full runs of
pre-merge CI across *all* hardware. This was extremely wasteful.
The value of trace-based testing is as a smoke test to check that traces
still render correctly. That is useful, but it turns out that checksums
are the wrong way to go about it. A better implementation would be
storing only a single reference image from a software rasterizer per
trace. No driver-specific references would be stored. That reference
image must never change, provided the trace never changes. CI would then
check rendered results against that image with tolerant fuzzy
comparisons. That tolerance matches with the fuzzy comparison that the
human eye would do when investigating a checksum change anyway. Yes, the
image comparison JavaScript will now report that
0 pixels changed within the tolerance, but there's nothing a human eye
can do with that information other than an error prone copypaste of new
checksums back in the yaml file and kicking it back to CI, itself a
waste of time.
Finally, in the time we've had trace-based testing alongside the
conformance tests, I cannot remember a single actual regression in one
of my commits the trace jobs have identified that the conformance tests
have not also identified. By contrast, the conformance test coverage has
prevented the merge of a number of actual regressions, with very few
flakes or xfail changes, and I am grateful we have that coverage. That
means the value added from the trace jobs is close to zero, while the
above checksum issues means that the cost is tremendous, even ignoring
the physical cost of the extra CI jobs.
If you work on trace-based testing and would like to understand how it
could adapted to be useful for Panfrost, see my recommendations above.
If you work on CI in general and would like to improve Panfrost's CI
coverage, what we need right now is not trace-based testing, it's
GLES3.1 conformance runs on MediaTek MT8192 or MT8195. That hardware is
already in the Collabora LAVA lab, but it's not being used for Mesa CI
as the required kernel patches haven't made their way to mainline yet
and nobody has cherry-picked them to the gfx-ci kernel. If you are a
Collaboran and interested in improving Panfrost CI, please ping
AngeloGioacchino for information on which specific patches need to be
backported or cherry-picked to our gfx-ci kernel. Thank you.
Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Acked-by: Jason Ekstrand <jason.ekstrand@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19358>
On macOS, we don't have DRM or any real WSI, so Asahi has to pretend to be a
software rasterizer to load. On Linux, we do have DRM and proper WSI, so we
don't want that. For faking Asahi devices on Linux, we should use drm-shim
instead. This makes sure we don't accidentally load Asahi on non-M1 Linux.
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Acked-by: Eric Engestrom <eric@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/15940>
Copy-paste a pile of winsys code from panfrost and find-and-replace the name to
asahi. This should contain all the glue code needed for asahi+kmsro.
The kernel driver is under way (led by Asahi Lina, not me), but it's not
wred up here. My goal was rather to run shader-db, which expects a
render node, which means drm-shim, which means DRM loader support. With
this patch and a trivial drm-shim, shader-db runs.
In general I am reticent to touch UABI related code when the UABI hasn't been
finalized upstream, or started design at all, hence the RFC. Realistically this
patch assumes the following about the future UABI:
0. It will be a DRM driver. This is nonnegotiable.
1. The render node will be named "asahi". The other reasonable name would be
"apple", which I'm using for the display controller (not yet upstream, but
getting close).
2. Display and rendering will be split in the kernel, requiring kmsro in
userspace, as agreed in past discussions.
The 3D accelerator (AGX) and the display controller (DCP) are completely
orthogonal blocks with separate lineages. True, Apple A14 (~= M1) has AGX and
DCP together, and it seems like all the chips that will get upstream support
will have this for the forseeable future. Nevertheless, it's a historical
coincidence. Apple A12 had an AGX block with a pre-DCP Apple display
controller, which would use a completely different display driver. Older SoCs
had a PowerVR block with an Apple shader core, with a pre-DCP Apple display
controller. Even older SoCs had a pure PowerVR block (+ Apple display).
The AGX and DCP kernel drivers are not expected to share any nontrivial code.
We don't gain anything by bundling them together. Likewise, the many
codec blocks are completely orthogonal. This is all standard practice
for Arm SoCs.
It is true that AGX has never been used with a non-Apple display
controller; it is highly unlikely this would change (either by AGX
licensing out or something like Mali-DP getting licensed in). But
an extra kmsro user doesn't actually add more complexity to Mesa, so
shrug.
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: Eric Engestrom <eric@igalia.com> [meson, ack on gallium]
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/15940>
After a great deal of spec lawyering in #dri-devel, I am convinced this
is probably okay for the same reasons as v3d and freedreno. The batch
reordering and flush deferral optimizations are seemingly still ok. The
requirement that writes are visible "immediately" in the spec actually
means "in the subsequent [OpenGL] command" for the CPU -> GPU direction,
which avoids pitfalls where PERSISTENT|COHERENT could be used as a
"doorbell". With that understanding, the extension doesn't actually
require anything special for tilers other than coherency at GPU submit
boundaries, which is true for any driver that does not use a sync ioctl.
After this commit, the remaining drivers that don't set the CAP are
d3d12, softpipe, etnaviv, and i915g. I am unsure about d3d12, but the
latter 3 could probbaly enable it trivially for the same reason.
v2: Don't use copy_resource path for persistent mappings (Emma). Emma
explained on GitLab:
I don't think you should have the copy_resource path taken for
PIPE_RESOURCE_FLAG_MAP_PERSISTENT BOs. Imagine the user has a
general-purpose BO they're streaming stuff into and doing draws that
they keep persistently mapped until wrapping. They call some GL
function on the same buffer that does a fallback write map on the BO
(u_default_buffer_subdata, util_resource_copy_region, whatever) -- the
buffer is in use, copy triggers, allocates a new BO. Whoops, the user's
pointer for streaming writes is now freed.
Closes: #7570
Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Reviewed-by: Emma Anholt <emma@anholt.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19351>
