Once LCRA has run, we have a map from IR indices to byte offsets into
the register file, so we need to "install" these results, rewriting the
IR to use native registers and fixing up writemasks/swizzles to
substitute vectorization for adjacent registers (for LCRA, we're
modeling in terms of real vectors).
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Tested-by: Marge Bot <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4158>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4158>
We model the machine as vector (with restrictions) to natively handle
mixed types and I/O and other goodies. We use LCRA for the heavylifting.
This commit adds only the modeling to feed into LCRA and spit LCRA
solutions back; next commit will integrate it with the IR.
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4158>
We want types to be consistent throughout the IR so we don't have to
make exceptions to parse things out. These cases just got missed.
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4158>
We have native FMA which works for graphics usage (unlike Midgard where
it's really reserved for compute for various reasons), let's use it.
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4150>
We can move e v e n more code to be shared and let bi_block inherit from
pan_block, which will allow us to use the shared data flow analysis.
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4150>
Now that it's all abstracted nicely with an implementation shared with
Midgard, this is pretty easy to get.
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4150>
Only on supported GPUs at the moment; for older Bifrost that don't
support these, I'm not sure yet where the right place to do the lowering
is. NIR algebraic rules would be "nice" but probably impractical -- but
it wouldn't be hard to do it directly in BIR (as a lowering pass or
alternative implementation).
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4139>
These are all regular ALU ops found in GLES2 which makes them
particularly nice targets at the moment. Just translate straight to our
IR.
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4139>
Soon we'll have a NIR support to handle this the Right Way along with
pos and sat_signed support, but we'll always need the fallback anyway.
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4139>
This will allow us to optimize out the constant moves (although that
will require a DCE pass which has yet to be written).
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4139>
We do the bare minimum translation, just enough for fmov/fadd/fmul right
now with no modifiers / inlined constants / etc. The rest is to come!
But hopefully I got bitsize handling right this time around.
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4139>
I feel so dirty. But this will let the IR be a lot more flexible seeing
as we really are vector in a certain sense (I/O, small types)
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4139>
We'd really rather not emit extracts. We are approaching on a vector IR
anyway which is annoying but really necessary to handle I/O and fp16
correctly. So let's just go all the way and deal with swizzles and masks
within reason; it'll still be somewhat saner in the long-term.
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4139>
Do the absolute simplest possible thing -- create a clause for every
instruction, and just pick whichever slot we can, nopping the other,
copying whatever constant we have whether it's used or not.
To be clear - this is not to be used in a production compiler. But this
lets actual bundles and clauses show up in the BIR, which unblocks work
on final code generation and packing (which can happen more or less in
parallel to NIR->BIR, optimization, register allocation, and writing an
actual scheduling).
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4097>
In the laziest possible way... We can just emit worst case moves which
DCE will eat for breakfast anyway, and inline constants on instructions
where that is supported directly. This approach eliminates a lot of
nasty corner cases that Midgard's crazy cache scheme has hit, at the
expense of slightly more work for DCE (but it's only a single iteration
of an O(N) pass that has to run anyway..)
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4097>
Not very useful without also handling breaks and continues, of course.
We use the strategy from v3d (vir_to_nir) instead of Midgard's, since
the latter is mildly insane. I mean, it passes deqp but...
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4097>
Copypaste from Midgard with some cleanups. That seems to be a trend
these days. Hopefully boilerplate will come to a close soon.
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4097>
Pretty much a copypaste from Midgard except where architectural
decisions diverge around vectorization. On that note, we will need our
own ALU scalarization pass at some point (or rather we'll need to extend
nir_lower_alu_scalar) to allow partial lowering for 8/16-bit ops. I.e.
we'll approximately need to lower
vec4 16 ssa_2 = fadd ssa_0, ssa_1
to
vec2 16 ssa_2 = fadd ssa_0.xy, ssa_1.xy
vec2 16 ssa_3 = fadd ssa_0.zw, ssa_1.zw
vec4 16 ssa_4 = vec4 ssa_2.x, ssa_2.y, ssa_3.x, ssa_4
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4097>
Modeled after the Midgard system. Already we know of two
compiler-visible differences between G52 and G71, so let's keep track so
we can eventually port the compiler to other Bifrost systems.
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4061>
We're making some pretty dramatic design pivots so this early on it'll
be easier to start from scratch, I think.
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4061>
Currently the schedule_program implementation being used is picked
at compile time, which on the Android platform means that the
bifrost compiler & scheduler is used for all targets, including
midgard based hardware.
This commit disambiguates between the two schedule_program functions.
Signed-off-by: Robert Foss <robert.foss@collabora.com>
Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Set of opcodes doesn't have enough flexibility in certain cases. E.g.
Utgard PP has vector conditional select operation, but condition is always
scalar. Lowering all the vector selects to scalar increases instruction
number, so we need a way to filter only those ops that can't be handled
in hardware.
Reviewed-by: Qiang Yu <yuq825@gmail.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Signed-off-by: Vasily Khoruzhick <anarsoul@gmail.com>
Panfrost is the only user of the macro; we are better off expanding than
having random stuff in nir.h.
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>