The logic in SIInsertWaits::getHwCounts() only really made sense for SMRD
instructions, and trying to shoehorn it into handling DS_WRITE_B32 caused
it to corrupt the encoding of that by clobbering the first operand with
the second one.
Undo that damage and only apply the SMRD logic to that.
Fixes some derivates related piglit regressions with radeonsi.
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@188558 91177308-0d34-0410-b5e6-96231b3b80d8
The SIInsertWaits pass was overwriting the first operand (gds bit) of
DS_WRITE_B32 with the second operand (value to write). This meant that
any time the value to write was stored in an odd number VGPR, the gds
bit would be set causing the instruction to write to GDS instead of LDS.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@188522 91177308-0d34-0410-b5e6-96231b3b80d8
Now that compute support is better on SI, we can't continue using v16i8
for descriptors since this is also a legal type in OpenCL.
This patch fixes numerous hangs with the piglit OpenCL test and since
we now use a target specific DAG node for LOAD_CONSTANT with the
correct MemOperandFlags, this should also fix:
https://bugs.freedesktop.org/show_bug.cgi?id=66805
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@188429 91177308-0d34-0410-b5e6-96231b3b80d8
Using REG_SEQUENCE for BUILD_VECTOR rather than a series of INSERT_SUBREG
instructions should make it easier for the register allocator to coalasce
unnecessary copies.
v2:
- Use an SGPR register class if all the operands of BUILD_VECTOR are
SGPRs.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@188427 91177308-0d34-0410-b5e6-96231b3b80d8
The instruction selector will now try to infer the destination register
so it can decided whether to use V_MOV_B32 or S_MOV_B32 when copying
immediates.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@188426 91177308-0d34-0410-b5e6-96231b3b80d8
The previous code declared the operand as unknown:$vaddr, which made
it possible for scalar registers to be used instead of vector registers.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@188425 91177308-0d34-0410-b5e6-96231b3b80d8
This fixes the F2U opcode for the Mesa driver.
Patch by: Marek Olšák
Signed-off-by: Marek Olšák <marek.olsak@amd.com>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@188418 91177308-0d34-0410-b5e6-96231b3b80d8
R600 doesn't need to do any scheduling on the SelectionDAG now that it
has a very good MachineScheduler. Also, using the VLIW SelectionDAG
scheduler was having a major impact on compile times. For example with
the phatk kernel here are the LLVM IR to machine code compile times:
With Sched::VLIW
Total Compile Time: 1.4890 Seconds (User + System)
SelectionDAG Instruction Scheduling: 1.1670 Seconds (User + System)
With Sched::Source
Total Compile Time: 0.3330 Seconds (User + System)
SelectionDAG Instruction Scheduling: 0.0070 Seconds (User + System)
The code ouput was identical with both schedulers. This may not be true
for all programs, but it gives me confidence that there won't be much
reduction, if any, in code quality by using Sched::Source.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@188215 91177308-0d34-0410-b5e6-96231b3b80d8
Since the VSrc_* register classes contain both VGPRs and SGPRs, copies
that used be emitted by isel like this:
SGPR = COPY VGPR
Will now be emitted like this:
VSrC = COPY VGPR
This patch also adds a pass that tries to identify and fix situations where
a VGPR to SGPR copy may occur. Hopefully, these changes will make it
impossible for the compiler to generate illegal VGPR to SGPR copies.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@187831 91177308-0d34-0410-b5e6-96231b3b80d8
Without explicit dependencies, both per-file action and in-CommonTableGen action could run in parallel.
It races to emit *.inc files simultaneously.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@187780 91177308-0d34-0410-b5e6-96231b3b80d8
We use MVT::i32 for the vector index type, because we use 32-bit
operations to caculate offsets when dynamically indexing vectors.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@187749 91177308-0d34-0410-b5e6-96231b3b80d8
* Added R600_Reg64 class
* Added T#Index#.XY registers definition
* Added v2i32 register reads from parameter and global space
* Added f32 and i32 elements extraction from v2f32 and v2i32
* Added v2i32 -> v2f32 conversions
Tom Stellard:
- Mark vec2 operations as expand. The addition of a vec2 register
class made them all legal.
Patch by: Dmitry Cherkassov
Signed-off-by: Dmitry Cherkassov <dcherkassov@gmail.com>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@187582 91177308-0d34-0410-b5e6-96231b3b80d8
If we merge vector when a vector is used, it will generate an artificial
antidependency that can prevent 2 tex/vtx instructions to use the same
clause and thus generate extra clauses that reduce performance.
There is no test case as such situation is really hard to predict.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@187516 91177308-0d34-0410-b5e6-96231b3b80d8
There are a lot of restrictions on instruction groups that contain
LDS instructions, so for now we will be conservative and not packetize
anything else with them.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@187513 91177308-0d34-0410-b5e6-96231b3b80d8
We were using two instructions for similar purpose : break and
predicated break. Only predicated_break was emitted and it was
lowered at R600ControlFlowFinalizer to JUMP;CF_BREAK;POP.
This commit simplify the situation by making AMDILCFGStructurizer
emit IF_PREDICATE;BREAK;ENDIF; instead of predicated_break (which
is now removed).
There is no functionality change.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@187510 91177308-0d34-0410-b5e6-96231b3b80d8
build_vector is lowered to REG_SEQUENCE, which is something the register
allocator does a good job at optimizing.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@187397 91177308-0d34-0410-b5e6-96231b3b80d8
Merge consecutive if-regions if they contain identical statements.
Both transformations reduce number of branches. The transformation
is guarded by a target-hook, and is currently enabled only for +R600,
but the correctness has been tested on X86 target using a variety of
CPU benchmarks.
Patch by: Mei Ye
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@187278 91177308-0d34-0410-b5e6-96231b3b80d8
This commit also implements these functions for R600 and removes a test
case that was relying on the buggy behavior.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@187007 91177308-0d34-0410-b5e6-96231b3b80d8
These are really the same address space in hardware. The only
difference is that CONSTANT_ADDRESS uses a special cache for faster
access. When we are unable to use the constant kcache for some reason
(e.g. smaller types or lack of indirect addressing) then the instruction
selector must use GLOBAL_ADDRESS loads instead.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@187006 91177308-0d34-0410-b5e6-96231b3b80d8
This increases the number of opportunites we have for folding. With the
previous implementation we were unable to fold into any instructions
other than the first when multiple instructions were selected from a
single SDNode.
Reviewed-by: Vincent Lejeune <vljn at ovi.com>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@186919 91177308-0d34-0410-b5e6-96231b3b80d8
A side-effect of this is that now the compiler expects kernel arguments
to be 4-byte aligned.
Reviewed-by: Vincent Lejeune <vljn at ovi.com>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@186916 91177308-0d34-0410-b5e6-96231b3b80d8
Enough for the radeonsi driver to use it for calculating derivatives.
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@186012 91177308-0d34-0410-b5e6-96231b3b80d8
Test is not included as it is several 1000 lines long.
To test this functionnality, a test case must generate at least 2 ALU clauses,
where an ALU clause is ~110 instructions long.
NOTE: This is a candidate for the stable branch.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@185943 91177308-0d34-0410-b5e6-96231b3b80d8
This is dead code since PIC16 was removed in 2010. The result was an odd mix,
where some parts would carefully pass it along and others would assert it was
zero (most of the object streamer for example).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@185436 91177308-0d34-0410-b5e6-96231b3b80d8
Patch By: Alex Deucher
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@185209 91177308-0d34-0410-b5e6-96231b3b80d8
By default, we expand these operations for both EG and SI. Move the
duplicated code into a common space for now. If the targets ever actually
implement these operations as instructions, we can override that in the relevant
target.
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@184848 91177308-0d34-0410-b5e6-96231b3b80d8
Add test cases for both vector sizes on SI and also add v2i32 test for EG.
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@184846 91177308-0d34-0410-b5e6-96231b3b80d8
Also add lit test for both cases on SI, and v2i32 for evergreen.
Note: I followed the guidance of the v4i32 EG check... UREM produces really
complex code, so let's just check that the instruction was lowered
successfully.
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@184844 91177308-0d34-0410-b5e6-96231b3b80d8
Also add lit test for both cases on SI, and v2i32 for evergreen.
Note: I followed the guidance of the v4i32 EG check... UDIV produces really
complex code, so let's just check that the instruction was lowered
successfully.
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@184843 91177308-0d34-0410-b5e6-96231b3b80d8
Also add lit test for both cases on SI, and v2i32 for evergreen.
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@184842 91177308-0d34-0410-b5e6-96231b3b80d8
Also add lit test for both cases on SI, and v2i32 for evergreen.
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@184841 91177308-0d34-0410-b5e6-96231b3b80d8
Also add lit test for both cases on SI, and v2i32 for evergreen.
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@184840 91177308-0d34-0410-b5e6-96231b3b80d8
Also add lit test for both cases on SI, and v2i32 for evergreen.
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@184839 91177308-0d34-0410-b5e6-96231b3b80d8
Also add lit test for both cases on SI, and v2i32 for evergreen.
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@184838 91177308-0d34-0410-b5e6-96231b3b80d8
Also add lit test for both cases on SI, and v2i32 for evergreen.
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@184837 91177308-0d34-0410-b5e6-96231b3b80d8
In reality, some unaligned memory accesses are legal for 32-bit types and
smaller too, but it all depends on the address space. Allowing
unaligned loads/stores for > 32-bit types is mainly to prevent the
legalizer from splitting one load into multiple loads of smaller types.
https://bugs.freedesktop.org/show_bug.cgi?id=65873
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@184822 91177308-0d34-0410-b5e6-96231b3b80d8
This should only make a difference in programs that use a lot of the
vector ALU instructions like BFI_INT and BIT_ALIGN. There is a slight
improvement in the phatk bitcoin mining kernel with this patch on
Evergreen (vector size == 1):
Before:
1173 Instruction Groups / 9520 dwords
After:
1167 Instruction Groups / 9510 dwords
Reviewed-by: Reviewed-by: Vincent Lejeune<vljn at ovi.com>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@184819 91177308-0d34-0410-b5e6-96231b3b80d8
Also add a v2i32 test to the existing v4i32 test.
Patch by: Aaron Watry
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
Signed-off-by: Aaron Watry<awatry@gmail.com>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@184482 91177308-0d34-0410-b5e6-96231b3b80d8
Also add SI tests to existing file and a v2i32 test for both
R600 and SI.
Patch by: Aaron Watry
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
Signed-off-by: Aaron Watry <awatry@gmail.com>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@184481 91177308-0d34-0410-b5e6-96231b3b80d8
The custom lowering causes llc to crash with a segfault.
Ideally, the custom lowering can be fixed, but this allows
programs which load/store v2i32 to work without crashing.
Patch by: Aaron Watry
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
Signed-off-by: Aaron Watry<awatry@gmail.com>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@184480 91177308-0d34-0410-b5e6-96231b3b80d8
Also add a seperate vector lit test file, since r600 doesn't seem to handle
v2i32 load/store yet, but we can test both for SI.
Patch by: Aaron Watry
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
Signed-off-by: Aaron Watry <awatry@gmail.com>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@184021 91177308-0d34-0410-b5e6-96231b3b80d8
We were using RAT_INST_STORE_RAW, which seemed to work, but the docs
say this instruction doesn't exist for Cayman, so it's probably safer
to use a documented instruction instead.
Reviewed-by: Vincent Lejeune<vljn at ovi.com>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@184015 91177308-0d34-0410-b5e6-96231b3b80d8
This is using a hint from AMD APP OpenCL Programming Guide with
empirically tweaked parameters.
I used Unigine Heaven 3.0 to determine best parameters on my system
(i7 2600/Radeon 6950/Kernel 3.9.4) the benchmark :
it went from 38.8 average fps to 39.6, which is ~3% gain.
(Lightmark 2008.2 gain is much more marginal: from 537 to 539)
There is no lit test provided as the parameter were determined
empirically and it it would be nearly impossiblet to find a test
program that check for optimal behavior.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@183593 91177308-0d34-0410-b5e6-96231b3b80d8
We weren't computing structure size correctly and we were relying on
the original alloca instruction to compute the offset, which isn't
always reliable.
Reviewed-by: Vincent Lejeune <vljn@ovi.com>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@183568 91177308-0d34-0410-b5e6-96231b3b80d8
This should simplify the subtarget definitions and make it easier to
add new ones.
Reviewed-by: Vincent Lejeune <vljn@ovi.com>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@183566 91177308-0d34-0410-b5e6-96231b3b80d8