This is the common case in the BE when we serialize condition and then
rematerialize it. Use either original or inverted condition.
Differential Revision: https://reviews.llvm.org/D48246
llvm-svn: 334882
Try to access pieces 4 bytes at a time. This helps
various hasOneUse extract_vector_elt combines, such
as load width reductions.
Avoids test regressions in a future commit.
llvm-svn: 334836
Summary: The same pattern as D48010, but this one is IR-canonical as of D47428.
Reviewers: nhaehnle, bogner, tstellar, arsenm
Reviewed By: arsenm
Subscribers: arsenm, kzhuravl, wdng, yaxunl, dstuttard, tpr, t-tye, llvm-commits
Tags: #amdgpu
Differential Revision: https://reviews.llvm.org/D48012
llvm-svn: 334817
Summary:
As a followup for D48007.
Since we already handle `x << (bitwidth - y) >> (bitwidth - y)` pattern,
which does not have ub for both the edge cases (`y == 0`, `y == bitwidth`),
i think also handling a pattern that is ub for `y == bitwidth` should be fine.
Reviewers: nhaehnle, bogner, tstellar, arsenm
Reviewed By: arsenm
Subscribers: arsenm, kzhuravl, wdng, yaxunl, dstuttard, tpr, t-tye, llvm-commits
Tags: #amdgpu
Differential Revision: https://reviews.llvm.org/D48010
llvm-svn: 334816
Summary:
D47980 will canonicalize the `x << (32 - y) >> (32 - y)`,
which is the pattern the AMDGPU expects to `x & (-1 >> (32 - y))`,
which is not recognized by AMDGPU.
Thus, it needs to be recognized, too.
Reviewers: nhaehnle, bogner, tstellar, arsenm
Reviewed By: arsenm
Subscribers: arsenm, kzhuravl, wdng, yaxunl, dstuttard, tpr, t-tye, llvm-commits
Tags: #amdgpu
Differential Revision: https://reviews.llvm.org/D48007
llvm-svn: 334815
Currently the handle type is a global pointer which holds 8 bytes.
We need a larger type which hold 16 bytes, therefore change it
to [i64 x 2].
Differential Revision: https://reviews.llvm.org/D48094
llvm-svn: 334625
Summary:
The code that handles ISD:Register and ISD::CopyFromReg assumes
the target is amdgcn, so this is broken on r600. We don't
need this analysis on r600 anyway so we can safely move
it to SITargetLowering.
Reviewers: alex-t, arsenm, nhaehnle
Reviewed By: arsenm
Subscribers: msearles, kzhuravl, wdng, yaxunl, dstuttard, tpr, t-tye, llvm-commits
Differential Revision: https://reviews.llvm.org/D46298
llvm-svn: 334607
- Do not emit following assembler directives:
- .hsa_code_object_version
- .hsa_code_object_isa
- .amd_amdgpu_isa
- .amd_amdgpu_hsa_metadata
- .amd_amdgpu_pal_metadata
- Do not emit .note entries
- Cleanup and bring in sync kernel descriptor header file
- Emit kernel descriptor into .rodata with appropriate relocations and
alignments
llvm-svn: 334519
The use iterator, used within findMaskOperands(), can return anything which is
not a def. isUse() requires a register, so check isReg() before calling isUse().
Differential Revision: https://reviews.llvm.org/D48047
llvm-svn: 334459
Rational: if there is indirect access that is usually an issue
because load is not ready by the use. However, if use is inside a
loop and load is outside that is potentially an issue for a first
iteration only.
Differential Revision: https://reviews.llvm.org/D47740
llvm-svn: 334420
AMDGPU inline assembler support i16, half and i128 typed variables in constraints, but they were reported as error.
Needed to fix https://github.com/RadeonOpenCompute/ROCm/issues/341,
e.g. to be able to load with global_load_dwordx4 to a 128bit integer variable
Differential Revision: https://reviews.llvm.org/D44920
llvm-svn: 334301
- Make code easier to maintain.
- Avoid generating waitcnts for VMEM if the address sppace does not involve VMEM.
- Add support to generate waitcnts for LDS and GDS memory.
Differential Revision: https://reviews.llvm.org/D47504
llvm-svn: 334241
This has two main components. First, widen
widen short constant loads in DAG when they have
the correct alignment. This is already done a bit in
AMDGPUCodeGenPrepare, since that has access to
DivergenceAnalysis. This can't help kernarg loads
created in the DAG. Start to use DAG divergence analysis
to help this case.
The second part is to avoid kernel argument lowering
breaking the alignment of short vector elements because
calling convention lowering wants to split everything
into legal register types.
When loading a split type, load the nearest 4-byte aligned
segment and shift to get the desired bits. This extra
load of the earlier argument piece ends up merging,
and the bit extract hopefully folds out.
There are a number of improvements and regressions with
this, but I think as-is this is a better compromise between
several of the worst parts of SelectionDAG.
Particularly when i16 is legal, this produces worse code
for i8 and i16 element vector kernel arguments. This is
partially due to the very weak load merging the DAG does.
It only looks for fairly specific combines between pairs
of loads which no longer appear. In particular this
causes v4i16 loads to be split into 2 components when
previously the two halves were merged.
Worse, because of the newly introduced shifts, there
is a lot more unnecessary vector packing and unpacking code
emitted. At least some of this is due to reporting
false for isTypeDesirableForOp for i16 as a workaround for
the lack of divergence information in the DAG. The cases
where this happens it doesn't actually matter, but the
relevant code in SimplifyDemandedBits doens't have the context
to know to ignore this.
The use of the scalar cache is probably more important
than the mess of mostly scalar instructions doing this packing
and unpacking. Future work can fix this, possibly by making better
use of the new DAG divergence information for controlling promotion
decisions, or adding another version of shift + trunc + shift
combines that doesn't only know about the used types.
llvm-svn: 334180
When denormals are supported we are producing a full division for
1.0f / x. That still can be replaced by the faster version:
bool c = fabs(x) > 0x1.0p+96f;
float s = c ? 0x1.0p-32f : 1.0f;
x *= s;
return s * v_rcp_f32(x)
in case if requested accuracy is 2.5ulp or less. The same version
is used if denormals are not supported for non 1.0 numerators, where
just v_rcp_f32 is then used for 1.0 numerator.
The optimization of 1/x is extended to the case -1/x, which is the
same except for the resulting sign bit.
OpenCL conformance passed with both enabled and disabled denorms.
Differential Revision: https://reviews.llvm.org/D47805
llvm-svn: 334142
Fixes terrible code on targets without f16 support. The
legalization creates a mess that is difficult to recover
from. Also should avoid randomly breaking these tests
multiple times in sequence in future commits.
Some regressions in cases where it happens to be better
to pull the source modifier after the conversion.
llvm-svn: 334132
On targets like Arm some relaxations may only be performed when certain
architectural features are available. As functions can be compiled with
differing levels of architectural support we must make a judgement on
whether we can relax based on the MCSubtargetInfo for the function. This
change passes through the MCSubtargetInfo for the function to
fixupNeedsRelaxation so that the decision on whether to relax can be made
per function. In this patch, only the ARM backend makes use of this
information. We must also pass the MCSubtargetInfo to applyFixup because
some fixups skip error checking on the assumption that relaxation has
occurred, to prevent code-generation errors applyFixup must see the same
MCSubtargetInfo as fixupNeedsRelaxation.
Differential Revision: https://reviews.llvm.org/D44928
llvm-svn: 334078
Preserves the low bound of the !range. I don't think
it's legal to do anything with the top half since it's
theoretically reading garbage.
llvm-svn: 334045
Review feedback from r328165. Split out just the one function from the
file that's used by Analysis. (As chandlerc pointed out, the original
change only moved the header and not the implementation anyway - which
was fine for the one function that was used (since it's a
template/inlined in the header) but not in general)
llvm-svn: 333954
On GFX9 and earlier, flat memory ops may decrement VMCNT out-of-order as well as LGKMCNT out-of-order.
Differential Revision: https://reviews.llvm.org/D46616
llvm-svn: 333926
Summary:
The new rules are straightforward. The main rules to keep in mind
are:
1. NAME is an implicit template argument of class and multiclass,
and will be substituted by the name of the instantiating def/defm.
2. The name of a def/defm in a multiclass must contain a reference
to NAME. If such a reference is not present, it is automatically
prepended.
And for some additional subtleties, consider these:
3. defm with no name generates a unique name but has no special
behavior otherwise.
4. def with no name generates an anonymous record, whose name is
unique but undefined. In particular, the name won't contain a
reference to NAME.
Keeping rules 1&2 in mind should allow a predictable behavior of
name resolution that is simple to follow.
The old "rules" were rather surprising: sometimes (but not always),
NAME would correspond to the name of the toplevel defm. They were
also plain bonkers when you pushed them to their limits, as the old
version of the TableGen test case shows.
Having NAME correspond to the name of the toplevel defm introduces
"spooky action at a distance" and breaks composability:
refactoring the upper layers of a hierarchy of nested multiclass
instantiations can cause unexpected breakage by changing the value
of NAME at a lower level of the hierarchy. The new rules don't
suffer from this problem.
Some existing .td files have to be adjusted because they ended up
depending on the details of the old implementation.
Change-Id: I694095231565b30f563e6fd0417b41ee01a12589
Reviewers: tra, simon_tatham, craig.topper, MartinO, arsenm, javed.absar
Subscribers: wdng, llvm-commits
Differential Revision: https://reviews.llvm.org/D47430
llvm-svn: 333900
Summary:
They've been deprecated in favor of UADDO/ADDCARRY or USUBO/SUBCARRY for a while.
Target that uses these opcodes are changed in order to ensure their behavior doesn't change.
Reviewers: efriedma, craig.topper, dblaikie, bkramer
Subscribers: jholewinski, arsenm, jyknight, sdardis, nemanjai, nhaehnle, kbarton, fedor.sergeev, asb, rbar, johnrusso, simoncook, jordy.potman.lists, apazos, sabuasal, niosHD, jrtc27, zzheng, edward-jones, mgrang, atanasyan, llvm-commits
Differential Revision: https://reviews.llvm.org/D47422
llvm-svn: 333748
Memory clauses are formed into bundles in presence of xnack.
Their source operands are marked as early-clobber.
This allows to allocate distinct source and destination registers
within a clause and prevent breaking the clause with s_nop in the
hazard recognizer.
Clauses are undone before post-RA scheduler to allow some rescheduling,
which will not break the clause since artificial edges are created in
the dag to keep memory operations together. Yet this allows a better
ILP in some cases.
Differential Revision: https://reviews.llvm.org/D47511
llvm-svn: 333691
Keep track of achieved occupancy in SIMachineFunctionInfo.
At the moment we have a lot of duplicated or even missed code to
query and maintain occupancy info. Record it in the MFI and
query in a single call. Interfaces:
- getOccupancy() - returns current recorded achieved occupancy.
- getMinAllowedOccupancy() - returns lesser of the achieved occupancy
and the lowest occupancy we are ready to tolerate. For example if
a kernel is memory bound we are ready to tolerate 4 waves.
- limitOccupancy() - record occupancy level if we have to lower it.
- increaseOccupancy() - record occupancy if scheduler managed to
increase the occupancy.
MFI takes care of integrating different checks affecting occupancy,
including LDS use and waves-per-eu attribute. Note that scheduler
starts with not yet known register pressure, so has to record either
limit or increase in occupancy after it is done. Later passes can
just query a resulting value.
New interface is used in the active scheduler and NFC wrt its work.
Changes are also made to experimental schedulers to use it and record
an occupancy after they are done. Before the change waves-per-eu was
ignored by experimental schedulers and tolerance window for memory
bound kernels was not used.
Differential Revision: https://reviews.llvm.org/D47509
llvm-svn: 333629
v2: use "ensureAlignment"
make functions cache line aligned
Fixes GPU hangs since r333219:
"AMDGPU: Split R600 AsmPrinter code into its own class"
Differential Revision: https://reviews.llvm.org/D47516
llvm-svn: 333622
This was just emitting loads with the ABI alignment
for the raw type. The true alignment is often better,
especially when an illegal vector type was scalarized.
The better alignment allows using a scalar load
more often.
llvm-svn: 333558
In terms of waitcnt insertion/if necessary, the waitcnt pass forces convergence
for a loop. Previously, that kicked if greater than 2 passes over a loop, which
doesn't account for loop with many bottom blocks. So, increase the threshold to
(n+1), where n is the number of bottom blocks. This gives the pass an
opportunity to consider the contribution of each bottom block, to the overall
loop, before the forced convergence potentially kicks in.
Differential Revision: https://reviews.llvm.org/D47488
llvm-svn: 333556