This reverts commit adaff46d087799072438dd744b038e6fd50a2d78.
Drop the -O3 checks from default-attributes.hip. I don't know why they
are different on some bots but reverting this is far too disruptive.
Removing it from the codegen pipeline induces a lot of test churn
because llc is no longer optimizing out implicit arguments to kernels.
Mostly mechanical, but there are some creative test updates. I preferred
to take the changes as-is in tests where the ABI isn't relevant. In
cases where it's more relevant, or the optimize out logic was too
ingrained in the test, I pre-run the optimization. Some cases manually
add attributes to disable inputs.
PR #66334 tried to renumber slot indexes before register allocation, but
the numbering was still affected by list entries for instructions which
had been erased. Fix this to make the register allocator's live range
length heuristics even less dependent on the history of how instructions
have been added to and removed from SlotIndexes's maps.
RegAllocGreedy uses SlotIndexes::getApproxInstrDistance to approximate
the length of a live range for its heuristics. Renumbering all slot
indexes with the default instruction distance ensures that this estimate
will be as accurate as possible, and will not depend on the history of
how instructions have been added to and removed from SlotIndexes's maps.
This also means that enabling -early-live-intervals, which runs the
SlotIndexes analysis earlier, will not cause large amounts of churn due
to different register allocator decisions.
Inspired by some of the cases from D145468
Let SimplifyDemandedBits handle the narrowing of lshr to half-width if we don't require the upper bits, the narrowed shift is profitable and the zext/trunc are free.
A future patch will propose the equivalent shl narrowing combine.
Differential Revision: https://reviews.llvm.org/D146121
Documentation for TargetLowering::getShiftAmountTy says that LegalTypes
should generally be true during type legalization, so this patch does
that.
On AMDGPU the effect is that we use i32 (a sane type) instead of i64
(pointer sized type) for more shift amounts, which in turn allows more
formation of rotates and funnel shifts pre-legalization.
Differential Revision: https://reviews.llvm.org/D154960
Currently, a node and its users are added back to the worklist in reverse topological order after it is combined. This diff changes that order to be topological. This is part of a larger migration to get the DAGCombiner to process nodes in topological order.
Reviewed By: RKSimon
Differential Revision: https://reviews.llvm.org/D127115
Currently, a node and its users are added back to the worklist in reverse topological order after it is combined. This diff changes that order to be topological. This is part of a larger migration to get the DAGCombiner to process nodes in topological order.
Reviewed By: RKSimon
Differential Revision: https://reviews.llvm.org/D127115
Occupancy is expressed as waves per SIMD. This means that we need to
take into account the number of SIMDs per "CU" or, to be more precise,
the number of SIMDs over which a workgroup may be distributed.
getOccupancyWithLocalMemSize was wrong because it didn't take SIMDs
into account at all.
At the same time, we need to take into account that WGP mode offers
access to a larger total amount of LDS, since this can affect how
non-power-of-two LDS allocations are rounded. To make this work
consistently, we distinguish between (available) local memory size and
addressable local memory size (which is always limited by 64kB on
gfx10+, even with WGP mode).
This change results in a massive amount of test churn. A lot of it is
caused by the fact that the default work group size is 1024, which means
that (due to rounding effects) the default occupancy on older hardware
is 8 instead of 10, which affects scheduling via register pressure
estimates. I've adjusted most tests by just running the UTC tools, but
in some cases I manually changed the work group size to 32 or 64 to make
sure that work group size chunkiness has no effect.
Differential Revision: https://reviews.llvm.org/D139468
This was disabled to prevent regressions, which appear to be just occurring on AMDGPU (at least in our current lit tests), which I've addressed by adding AMDGPUTargetLowering::isDesirableToCommuteWithShift overrides.
Fixes#57872
Differential Revision: https://reviews.llvm.org/D136042
If we can not prove that f16 operands of a buildvector are canonicalized, then we can not lower into a V_PACK. In this scenario, we would previously lower into some combination of and(sdwa), shr, or. This patch allows for matching into V_PERM instead.
Change-Id: Ifa4a74fdb81ef44f22ba490c7fdf81ec8aebc945
This patch allows SimplifyDemandedBits to call SimplifyMultipleUseDemandedBits in cases where the ISD::SRL source operand has other uses, enabling us to peek through the shifted value if we don't demand all the bits/elts.
This is another step towards removing SelectionDAG::GetDemandedBits and just using TargetLowering::SimplifyMultipleUseDemandedBits.
There a few cases where we end up with extra register moves which I think we can accept in exchange for the increased ILP.
Differential Revision: https://reviews.llvm.org/D77804
Similar to the existing (shl (srl x, c1), c2) fold
Part of the work to fix the regressions in D77804
Differential Revision: https://reviews.llvm.org/D125836
Fold immediates regardless of how many uses they have. This is expected
to increase overall code size, but decrease register usage.
Differential Revision: https://reviews.llvm.org/D114644
Previously SIFoldOperands::foldInstOperand would only fold a
non-inlinable immediate into a single user, so as not to increase code
size by adding the same 32-bit literal operand to many instructions.
This patch removes that restriction, so that a non-inlinable immediate
will be folded into any number of users. The rationale is:
- It reduces the number of registers used for holding constant values,
which might increase occupancy. (On the other hand, many of these
registers are SGPRs which no longer affect occupancy on GFX10+.)
- It reduces ALU stalls between the instruction that loads a constant
into a register, and the instruction that uses it.
- The above benefits are expected to outweigh any increase in code size.
Differential Revision: https://reviews.llvm.org/D114643
This change enables divergence-driven instruction selection for the build_vector DAG nodes.
It also enables packed i16 instructions for GFX9.
Reviewed By: rampitec
Differential Revision: https://reviews.llvm.org/D116187
Using a BufferSize of one for memory ProcResources will result in better
ILP since it more accurately models the dependencies between memory ops
and their consumers on an in-order processor. After this change, the
scheduler will treat the data edges from loads as blocking so that
stalls are guaranteed when waiting for data to be retreaved from memory.
Since we don't actually track waitcnt here, this should do a better job
at modeling their behavior.
Practically, this means that the scheduler will trigger the 'STALL'
heuristic more often.
This type of change needs to be evaluated experimentally. Preliminary
results are positive.
Fixes: SWDEV-282962
Reviewed By: rampitec
Differential Revision: https://reviews.llvm.org/D114777
Change VOP_PAT_GEN to default to not generating an instruction selection
pattern for the VOP2 (e32) form of an instruction, only for the VOP3
(e64) form. This allows SIFoldOperands maximum freedom to fold copies
into the operands of an instruction, before SIShrinkInstructions tries
to shrink it back to the smaller encoding.
This affects the following VOP2 instructions:
v_min_i32
v_max_i32
v_min_u32
v_max_u32
v_and_b32
v_or_b32
v_xor_b32
v_lshr_b32
v_ashr_i32
v_lshl_b32
A further cleanup could simplify or remove VOP_PAT_GEN, since its
optional second argument is never used.
Differential Revision: https://reviews.llvm.org/D114252
Use GCNHazardRecognizer in postra sched.
Updated tests for the new schedules.
Reviewed By: arsenm
Differential Revision: https://reviews.llvm.org/D109536
Change-Id: Ia86ba2ae168f12fb34b4d8efdab491f84d936cde
the compilation time and there is no case for which we see any improvement in
performance. This patch removes this pass and its associated test cases from
the tree.
Differential Revision: https://reviews.llvm.org/D101313
Change-Id: I0599169a7609c19a887f8d847a71e664030cc141
Change some test cases to use divergent addresses for vector loads,
which should be the common case in real world code. Using uniform
addresses causes poor instruction selection for the surrounding
code which has to be fixed up post-register-allocation, and this causes
a lot of testsuite churn for a forthcoming patch to stop selecting
24-bit vector multiply instructions for uniform multiplies.
This shows up some problems in the idot tests where we fail to select
v_dot instructions because the patterns only match MUL_[UI]24 ISD nodes,
but the DAG contains i16 mul nodes instead.
Differential Revision: https://reviews.llvm.org/D97062
Support for XNACK and SRAMECC is not static on some GPUs. We must be able
to differentiate between different scenarios for these dynamic subtarget
features.
The possible settings are:
- Unsupported: The GPU has no support for XNACK/SRAMECC.
- Any: Preference is unspecified. Use conservative settings that can run anywhere.
- Off: Request support for XNACK/SRAMECC Off
- On: Request support for XNACK/SRAMECC On
GCNSubtarget will track the four options based on the following criteria. If
the subtarget does not support XNACK/SRAMECC we say the setting is
"Unsupported". If no subtarget features for XNACK/SRAMECC are requested we
must support "Any" mode. If the subtarget features XNACK/SRAMECC exist in the
feature string when initializing the subtarget, the settings are "On/Off".
The defaults are updated to be conservatively correct, meaning if no setting
for XNACK or SRAMECC is explicitly requested, defaults will be used which
generate code that can be run anywhere. This corresponds to the "Any" setting.
Differential Revision: https://reviews.llvm.org/D85882
tryLatency compares two sched candidates. For the top zone it prefers
the one with lesser depth, but only if that depth is greater than the
total latency of the instructions we've already scheduled -- otherwise
its latency would be hidden and there would be no stall.
Unfortunately it only tests the depth of one of the candidates. This can
lead to situations where the TopDepthReduce heuristic does not kick in,
but a lower priority heuristic chooses the other candidate, whose depth
*is* greater than the already scheduled latency, which causes a stall.
The fix is to apply the heuristic if the depth of *either* candidate is
greater than the already scheduled latency.
All this also applies to the BotHeightReduce heuristic in the bottom
zone.
Differential Revision: https://reviews.llvm.org/D72392
Enable clausing of memory loads on gfx10 by adding a new pass to insert
the s_clause instructions that mark the start of each hard clause.
Differential Revision: https://reviews.llvm.org/D79792
The two code paths have the same goal, legalizing a load of a non-byte-sized vector by loading the "flattened" representation in memory, slicing off each single element and then building a vector out of those pieces.
The technique employed by `ExpandLoad` is slightly more convoluted and produces slightly better codegen on ARM, AMDGPU and x86 but suffers from some bugs (D78480) and is wrong for BE machines.
Differential Revision: https://reviews.llvm.org/D79096
Add the scratch wave offset to the scratch buffer descriptor (SRSrc) in
the entry function prologue. This allows us to removes the scratch wave
offset register from the calling convention ABI.
As part of this change, allow the use of an inline constant zero for the
SOffset of MUBUF instructions accessing the stack in entry functions
when a frame pointer is not requested/required. Entry functions with
calls still need to set up the calling convention ABI stack pointer
register, and reference it in order to address arguments of called
functions. The ABI stack pointer register remains unswizzled, but is now
wave-relative instead of queue-relative.
Non-entry functions also use an inline constant zero SOffset for
wave-relative scratch access, but continue to use the stack and frame
pointers as before. When the stack or frame pointer is converted to a
swizzled offset it is now scaled directly, as the scratch wave offset no
longer needs to be subtracted first.
Update llvm/docs/AMDGPUUsage.rst to reflect these changes to the calling
convention.
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D75138
Summary:
pickNodeBidirectional tried to compare the best top candidate and the
best bottom candidate by examining TopCand.Reason and BotCand.Reason.
This is unsound because, after calling pickNodeFromQueue, Cand.Reason
does not reflect the most important reason why Cand was chosen. Rather
it reflects the most recent reason why it beat some other potential
candidate, which could have been for some low priority tie breaker
reason.
I have seen this cause problems where TopCand is a good candidate, but
because TopCand.Reason is ORDER (which is very low priority) it is
repeatedly ignored in favour of a mediocre BotCand. This is not how
bidirectional scheduling is supposed to work.
To fix this I changed the code to always compare TopCand and BotCand
directly, like the generic implementation of pickNodeBidirectional does.
This removes some uncommented AMDGPU-specific logic; if this logic turns
out to be important then perhaps it could be moved into an override of
tryCandidate instead.
Graphics shader benchmarking on gfx10 shows a lot more positive than
negative effects from this change.
Reviewers: arsenm, tstellar, rampitec, kzhuravl, vpykhtin, dstuttard, tpr, atrick, MatzeB
Subscribers: jvesely, wdng, nhaehnle, yaxunl, t-tye, hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D68338
Similar to what we already do with SimplifyDemandedVectorElts, call SimplifyDemandedBits across all the extracted elements of the source vector, treating it as single use.
There's a minor regression in store-weird-sizes.ll which will be addressed in an upcoming SimplifyDemandedBits patch.
We are relying on atrificial DAG edges inserted by the
MemOpClusterMutation to keep loads and stores together in the
post-RA scheduler. This does not work all the time since it
allows to schedule a completely independent instruction in the
middle of the cluster.
Removed the DAG mutation and added pass to bundle already
clustered instructions. These bundles are unpacked before the
memory legalizer because it does not work with bundles but also
because it allows to insert waitcounts in the middle of a store
cluster.
Removing artificial edges also allows a more relaxed scheduling.
Differential Revision: https://reviews.llvm.org/D72737
Summary:
The typo has been present since memOpsHaveSameBasePtr was introduced in
r313208.
It caused SIInstrInfo::shouldClusterMemOps to cluster more mem ops than
it was supposed to.
Subscribers: arsenm, kzhuravl, jvesely, wdng, nhaehnle, yaxunl, dstuttard, tpr, t-tye, hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D71616
There were 2 problems here. First, these patterns were duplicated to
handle the inverted shift operands instead of using the commuted
PatFrags.
Second, the point of the zext folding patterns don't apply to the
non-0ing high subtargets. They should be skipped instead of inserting
the extension. The zeroing high code would be emitted when necessary
anyway. This was also emitting unnecessary zexts in cases where the
high bits were undefined.
llvm-svn: 374092
If anything called the recursive isKnownNeverNaN/computeKnownBits/ComputeNumSignBits/SimplifyDemandedBits/SimplifyMultipleUseDemandedBits with an incorrect depth then we could continue to recurse if we'd already exceeded the depth limit.
This replaces the limit check (Depth == 6) with a (Depth >= 6) to make sure that we don't circumvent it.
This causes a couple of regressions as a mixture of calls (SimplifyMultipleUseDemandedBits + combineX86ShufflesRecursively) were calling with depths that were already over the limit. I've fixed SimplifyMultipleUseDemandedBits to not do this. combineX86ShufflesRecursively is trickier as we get a lot of regressions if we reduce its own limit from 8 to 6 (it also starts at Depth == 1 instead of Depth == 0 like the others....) - I'll see what I can do in future patches.
llvm-svn: 367171
This patch introduces the DAG version of SimplifyMultipleUseDemandedBits, which attempts to peek through ops (mainly and/or/xor so far) that don't contribute to the demandedbits/elts of a node - which means we can do this even in cases where we have multiple uses of an op, which normally requires us to demanded all bits/elts. The intention is to remove a similar instruction - SelectionDAG::GetDemandedBits - once SimplifyMultipleUseDemandedBits has matured.
The InstCombine version of SimplifyMultipleUseDemandedBits can constant fold which I haven't added here yet, and so far I've only wired this up to some basic binops (and/or/xor/add/sub/mul) to demonstrate its use.
We do see a couple of regressions that need to be addressed:
AMDGPU unsigned dot product codegen retains an AND mask (for ZERO_EXTEND) that it previously removed (but otherwise the dotproduct codegen is a lot better).
X86/AVX2 has poor handling of vector ANY_EXTEND/ANY_EXTEND_VECTOR_INREG - it prematurely gets converted to ZERO_EXTEND_VECTOR_INREG.
The code owners have confirmed its ok for these cases to fixed up in future patches.
Differential Revision: https://reviews.llvm.org/D63281
llvm-svn: 366799
