v2: use abort(), describe which LLVM version is affected
Cc: 12.0 <mesa-stable@lists.freedesktop.org>
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Not sure if this is the right way to do it, but it seems to work.
v2: make it a no-op on LLVM <= 3.5
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Fix compilation on Cygwin, since 50b22354, by adding c99_alloca.h include,
which should know how to portably make the alloc() prototype available.
Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Setup for pixel shader push constants is the same as for other
stages. Note that on gen8+ the if-else branches were identical
and the generation check for packet size redundant.
Signed-off-by: Topi Pohjolainen <topi.pohjolainen@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
This is partial revert of commit cc2d0e64.
It looks that even though blorp disables a stage the corresponding
3DSTATE_CONSTANT_XS packet is needed to be programmed. Hardware
seems to try to fetch the constants even for disabled stages.
Therefore care needs to be taken that the constant buffer is
set up properly. Blorp will continue to trash it into non-existing
such as before.
It is possible that this could be omitted on SKL where the
constant buffer is considered when the corresponding binding table
settings are changed. Bspec:
"The 3DSTATE_CONSTANT_* command is not committed to the shader
unit until the corresponding (same shader)
3DSTATE_BINDING_TABLE_POINTER_* command is parsed."
However, as CONSTANT_XS packet itself does not seem to stall on its
own, it is safer to emit the packets for SKL also.
Possible alternative to blorp trashing could have been to setup
defaults in the beginning of each batch buffer. However, hardware
doesn't seem to tolerate these packets being programmed multiple
times per primitive. Bspec for IVB:
"It is invalid to execute this command more than once between
3D_PRIMITIVE commands."
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=96878
Signed-off-by: Topi Pohjolainen <topi.pohjolainen@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Coverity's analysis is too weak to understand that
r600_init_flushed_depth(_, _, NULL) only returns true when
flushed_depth_texture was assigned a non-NULL value.
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
So that we can have gen7 split large writes produced by this lowering pass.
Signed-off-by: Samuel Iglesias Gonsálvez <siglesias@igalia.com>
Reviewed-by: Francisco Jerez <currojerez@riseup.net>
So far we only used instructions with this size in situations where we
did not operate per-channel and we wanted to ignore the execution mask,
but gen7 fp64 will need to emit code with a width of 4 that needs
normal execution masking.
v2:
- Modify the assert instead of deleting it (Curro)
Reviewed-by: Francisco Jerez <currojerez@riseup.net>
In fp64 we can produce code like this:
mov(16) vgrf2<2>:UD, vgrf3<2>:UD
That our simd lowering pass would typically split in instructions with a
width of 8, writing to two consecutive registers each. Unfortunately, gen7
hardware has a bug affecting execution masking and as a result, the
second GRF register write won't work properly. Curro verified this:
"The problem is that pre-Gen8 EUs are hardwired to use the QtrCtrl+1
(where QtrCtrl is the 8-bit quarter of the execution mask signals
specified in the instruction control fields) for the second
compressed half of any single-precision instruction (for
double-precision instructions it's hardwired to use NibCtrl+1,
at least on HSW), which means that the EU will apply the wrong
execution controls for the second sequential GRF write if the number
of channels per GRF is not exactly eight in single-precision mode (or
four in double-float mode)."
In practice, this means that we cannot write more than one
consecutive GRF in a single instruction if the number of channels
per GRF is not exactly eight in single-precision mode (or four
in double-float mode).
This patch makes our SIMD lowering pass split this kind of instructions
so that the split versions only write to a single register. In the
example above this means that we split the write in 4 instructions, each
one writing 4 UD elements (width = 4) to a single register.
v2 (Curro):
- Make explicit that the thing about hardwiring NibCtrl+1 for the second
compressed half is known to happen in Haswell and the issue with IVB
might not be exactly the same.
- Assign max_width instead of returning early so that we can handle
multiple restrictions affecting to the same instruction.
- Avoid division by 0 if the instruction does not write any registers.
- Ignore instructions what have WE_all set.
- Use the instruction execution type size instead of the dst type size.
v3 (Curro):
- Move the implementation down so it is not placed in the middle of another
workaround.
- Declare channels_per_grf as const.
- Don't break the loop early if we find a BAD_FILE source.
- Fix the number of channels that the hardware shifts for the second half
of a compressed instruction to be 8 in single precision and 4 in double
precision.
Reviewed-by: Francisco Jerez <currojerez@riseup.net>
Gen7 hardware does not support double immediates so these need
to be moved in 32-bit chunks to a regular vgrf instead. Instead
of doing this every time we need to create a DF immediate,
create a helper function that does the right thing depending
on the hardware generation.
v2:
- Define setup_imm_df() as an independent function (Curro)
- Create a specific builder to get rid of some instruction field
assignments (Curro).
v3:
- Get devinfo from builder (Kenneth)
Signed-off-by: Samuel Iglesias Gonsálvez <siglesias@igalia.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Previously, there were occasionally NIR registers in our programs, but
they were always actually used SSA-only. Now that we're trying to support
control flow, we need to actually conditionally move to registers based on
whether channels are active or not.
We're already checking that branch instructions are within the
contents of the shader and the proper PROG_END sequence is present.
The other thing we need in the presence of branching is to verify that
the shader doesn't overflow past the end of the uniforms stream.
To do that, we require that at the start of any basic block reading
uniforms have the following instructions:
load_imm temp, <offset within uniform stream>
add unif_addr, temp, unif
The instructions are generated by userspace, and the kernel verifies
that the load_imm is of the expected offset, and that the add adds it
to a uniform. We track which uniform in the stream that is, and at
draw call time fix up the uniform stream to have the address of the
start of the shader's uniforms for that draw call.
Signed-off-by: Eric Anholt <eric@anholt.net>
This isn't used yet, it's just a first step toward loop validation.
During the main parsing of instructions, we need to know when we hit a new
basic block so that we can reset validated state.
Basically we just treat each block independently. The only inter-block
scheduling I can think of that would be be interesting would be to move
texture result collection to after a short loop/if block that doesn't do
texturing. However, the kernel disallows that as part of its security
validation.
We still decide which uniform to lower based on how many
instructions-that-need-lowering use that uniform, but now we emit a new
temporary uniform load in each of the basic blocks containing an
instruction being lowered.
This commit is best reviewed with diff -b.
We need to apply the peephole pass to each of the blocks in the program.
We don't do dataflow analysis for SF across blocks, but we also don't
generate code that would need us to do so.
The optimization passes and scheduling aren't actually ready for multiple
blocks with control flow yet (as seen by the "cur_block" references in
them instead of iterating over blocks), but this creates the structures
necessary for converting them.
We have the prior list_foreach() all over the code, but I need to move
where instructions live as part of adding support for control flow. Start
by just converting to a helper iterator macro. (The simpler
"qir_for_each_inst()" will be used for the for-each-inst-in-a-block
iterator macro later)
This avoids a bunch of code gen regressions when enabling loops in vc4.
Prior to that, the GLSL that would have generated these optimizable phi
nodes was being lowered to csels between either (undef, a) or (a, a), and
those were being dealt with by nir_opt_undef and nir_opt_algebraic.
The allocation has succeeded by that point, so it needs to be freed.
CovID: 1358929
Signed-off-by: Eric Engestrom <eric@engestrom.ch>
Reviewed-by: Eric Anholt <eric@anholt.net>
This was appearing in vc4 VS/CS in mupen64, due to vertex attrib lowering
producing some constants that were getting compared.
total instructions in shared programs: 112276 -> 112198 (-0.07%)
instructions in affected programs: 2239 -> 2161 (-3.48%)
total estimated cycles in shared programs: 283102 -> 283038 (-0.02%)
estimated cycles in affected programs: 2365 -> 2301 (-2.71%)
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Small api cleanup. Make all api functions call GetContext instead
of locally casting handle. Makes debugging easier by providing a
single point to track context changes.
Signed-off-by: Tim Rowley <timothy.o.rowley@intel.com>
