Splitting critical edges when one of the source edges is an indirectbr
is hard in general (because it requires changing the memory the indirectbr
reads). But if a block only has a single indirectbr predecessor (which is
the common case), we can simulate splitting that edge by splitting
the destination block, and retargeting the *direct* branches.
This is motivated by the use of computed gotos in python 2.7: PyEval_EvalFrame()
ends up using an indirect branch with ~100 successors, and passing a constant to
each of those. Since MachineSink can't break indirect critical edges on demand
(and doing this in MIR doesn't look feasible), this causes us to emit about ~100
defs of registers containing constants, which we in the predecessor block, where
only one of those constants is used in each successor. So, at each computed goto,
we needlessly spill about a 100 constants to stack. The end result is that a
clang-compiled python interpreter can be about ~2.5x slower on a simple python
reduction loop than a gcc-compiled interpreter.
Differential Revision: https://reviews.llvm.org/D29916
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@296149 91177308-0d34-0410-b5e6-96231b3b80d8
The current pattern for extract bits in range is typically:
Mask.lshr(BitOffset).trunc(SubSizeInBits);
Which can be particularly slow for large APInts (MaskSizeInBits > 64) as they require the allocation of memory for the temporary variable.
This is another of the compile time issues identified in PR32037 (see also D30265).
This patch adds the APInt::extractBits() helper method which avoids the temporary memory allocation.
Differential Revision: https://reviews.llvm.org/D30336
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@296147 91177308-0d34-0410-b5e6-96231b3b80d8
This patch merges the existing floating-point induction variable widening code
into the integer induction variable widening code, creating a single set of
functions for both kinds of inductions. The primary motivation for doing this
is to enable vector phi node creation for floating-point induction variables.
Differential Revision: https://reviews.llvm.org/D30211
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@296145 91177308-0d34-0410-b5e6-96231b3b80d8
Provide a 64-bit pattern to use SUBFIC for subtracting from a 16-bit immediate.
The corresponding pattern already exists for 32-bit integers.
Committing on behalf of Hiroshi Inoue.
Differential Revision: https://reviews.llvm.org/D29387
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@296144 91177308-0d34-0410-b5e6-96231b3b80d8
The current pattern for extract bits in range is typically:
Mask.lshr(BitOffset).trunc(SubSizeInBits);
Which can be particularly slow for large APInts (MaskSizeInBits > 64) as they require the allocation of memory for the temporary variable.
This is another of the compile time issues identified in PR32037 (see also D30265).
This patch adds the APInt::extractBits() helper method which avoids the temporary memory allocation.
Differential Revision: https://reviews.llvm.org/D30336
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@296141 91177308-0d34-0410-b5e6-96231b3b80d8
The motivation for filling out these select-of-constants cases goes back to D24480,
where we discussed removing an IR fold from add(zext) --> select. And that goes back to:
https://reviews.llvm.org/rL75531https://reviews.llvm.org/rL159230
The idea is that we should always canonicalize patterns like this to a select-of-constants
in IR because that's the smallest IR and the best for value tracking. Note that we currently
do the opposite in some cases (like the cases in *this* patch). Ie, the proposed folds in
this patch already exist in InstCombine today:
https://github.com/llvm-mirror/llvm/blob/master/lib/Transforms/InstCombine/InstCombineSelect.cpp#L1151
As this patch shows, most targets generate better machine code for simple ext/add/not ops
rather than a select of constants. So the follow-up steps to make this less of a patchwork
of special-case folds and missing IR canonicalization:
1. Have DAGCombiner convert any select of constants into ext/add/not ops.
2 Have InstCombine canonicalize in the other direction (create more selects).
Differential Revision: https://reviews.llvm.org/D30180
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@296137 91177308-0d34-0410-b5e6-96231b3b80d8
This time with the missing files.
Similar to PR/25526, fast-regalloc introduces spills at the end of basic
blocks. When this occurs in between an ll and sc, the store can cause the
atomic sequence to fail.
This patch fixes the issue by introducing more pseudos to represent atomic
operations and moving their lowering to after the expansion of postRA
pseudos.
This resolves PR/32020.
Thanks to James Cowgill for reporting the issue!
Reviewers: slthakur
Differential Revision: https://reviews.llvm.org/D30257
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@296134 91177308-0d34-0410-b5e6-96231b3b80d8
Similar to PR/25526, fast-regalloc introduces spills at the end of basic
blocks. When this occurs in between an ll and sc, the store can cause the
atomic sequence to fail.
This patch fixes the issue by introducing more pseudos to represent atomic
operations and moving their lowering to after the expansion of postRA
pseudos.
This resolves PR/32020.
Thanks to James Cowgill for reporting the issue!
Reviewers: slthakur
Differential Revision: https://reviews.llvm.org/D30257
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@296132 91177308-0d34-0410-b5e6-96231b3b80d8
Summary:
This isn't testable for AArch64 by itself so this patch also adds
support for constant immediates in the pattern and physical
register uses in the result.
The new IntOperandMatcher matches the constant in patterns such as
'(set $rd:GPR32, (G_XOR $rs:GPR32, -1))'. It's always safe to fold
immediates into an instruction so this is the first rule that will match
across multiple BB's.
The Renderer hierarchy is responsible for adding operands to the result
instruction. Renderers can copy operands (CopyRenderer) or add physical
registers (in particular %wzr and %xzr) to the result instruction
in any order (OperandMatchers now import the operand names from
SelectionDAG to allow renderers to access any operand). This allows us to
emit the result instruction for:
%1 = G_XOR %0, -1 --> %1 = ORNWrr %wzr, %0
%1 = G_XOR -1, %0 --> %1 = ORNWrr %wzr, %0
although the latter is untested since the matcher/importer has not been
taught about commutativity yet.
Added BuildMIAction which can build new instructions and mutate them where
possible. W.r.t the mutation aspect, MatchActions are now told the name of
an instruction they can recycle and BuildMIAction will emit mutation code
when the renderers are appropriate. They are appropriate when all operands
are rendered using CopyRenderer and the indices are the same as the matcher.
This currently assumes that all operands have at least one matcher.
Finally, this change also fixes a crash in
AArch64InstructionSelector::select() caused by an immediate operand
passing isImm() rather than isCImm(). This was uncovered by the other
changes and was detected by existing tests.
Depends on D29711
Reviewers: t.p.northover, ab, qcolombet, rovka, aditya_nandakumar, javed.absar
Reviewed By: rovka
Subscribers: aemerson, dberris, kristof.beyls, llvm-commits
Differential Revision: https://reviews.llvm.org/D29712
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@296131 91177308-0d34-0410-b5e6-96231b3b80d8
Noticed while profiling PR32037, the target shuffle ops were being stored in SmallVector<*,8> types but the combiner could store as many as 16 ops at maximum depth (2 per depth).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@296130 91177308-0d34-0410-b5e6-96231b3b80d8
This one seems more obvious than D30270 that it can't make improvements because an extension always needs
all of the incoming bits. There's one specific transform in SimplifyDemandedInstructionBits of converting
a sext to a zext when the sign-bit is known zero, but that is handled explicitly in visitSext() with
ComputeSignBit().
Like D30270, there are no IR differences (other than instruction names) for the case in PR32037:
https://bugs.llvm.org//show_bug.cgi?id=32037
...and no regression test differences.
Zext/sext are a smaller part of the profile, but this still appears to shave off another 0.5% or so from
'opt -O2'.
Differential Revision: https://reviews.llvm.org/D30280
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@296129 91177308-0d34-0410-b5e6-96231b3b80d8
Previously LLVM was assuming 32-bit signed immediates which results in and with
a bitmask that has bit 31 set to incorrectly include bits 63-32 in the result.
After applying this patch I can now compile all of the FreeBSD mips assembly
code with clang.
This issue also affects the nor, slt and sltu macros and I will fix those in a
separate review.
Patch By: Alexander Richardson
Commit message reformatted by sdardis.
Reviewers: atanasyan, theraven, sdardis
Differential Revision: https://reviews.llvm.org/D30298
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@296125 91177308-0d34-0410-b5e6-96231b3b80d8
Make the MIPS disassembler consistent with the other targets in returning
a Size of zero when the input buffer cannot contain an instruction due
to it's size. Previously it reported the minimum instruction size when
it failed due to the buffer not being big enough for an instruction
causing llvm-objdump to crash when disassembling all sections.
Reviewers: slthakur
Differential Revision: https://reviews.llvm.org/D29984
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@296105 91177308-0d34-0410-b5e6-96231b3b80d8
The current pattern for setting bits in range is typically:
Mask |= APInt::getBitsSet(MaskSizeInBits, LoPos, HiPos);
Which can be particularly slow for large APInts (MaskSizeInBits > 64) as they require the allocation memory for the temporary variable.
This is one of the key compile time issues identified in PR32037.
This patch adds the APInt::setBits() helper method which avoids the temporary memory allocation completely, this first implementation uses setBit() internally instead but already significantly reduces the regression in PR32037 (~10% drop). Additional optimization may be possible.
I investigated whether there is need for APInt::clearBits() and APInt::flipBits() equivalents but haven't seen these patterns to be particularly common, but reusing the code would be trivial.
Differential Revision: https://reviews.llvm.org/D30265
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@296102 91177308-0d34-0410-b5e6-96231b3b80d8
The Fuchsia ABI defines slots from the thread pointer where the
stack-guard value for stack-protector, and the unsafe stack pointer
for safe-stack, are stored. This parallels the Android ABI support.
Patch by Roland McGrath
Differential Revision: https://reviews.llvm.org/D30237
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@296081 91177308-0d34-0410-b5e6-96231b3b80d8
LoopUnswitch/simplify-with-nonvalness.ll is the test case for this.
The LIC has 2 users and deleting the 1st user when it can be simplified
invalidated the iterator for the 2nd user.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@296069 91177308-0d34-0410-b5e6-96231b3b80d8
Splitting critical edges when one of the source edges is an indirectbr
is hard in general (because it requires changing the memory the indirectbr
reads). But if a block only has a single indirectbr predecessor (which is
the common case), we can simulate splitting that edge by splitting
the destination block, and retargeting the *direct* branches.
This is motivated by the use of computed gotos in python 2.7: PyEval_EvalFrame()
ends up using an indirect branch with ~100 successors, and passing a constant to
each of those. Since MachineSink can't break indirect critical edges on demand
(and doing this in MIR doesn't look feasible), this causes us to emit about ~100
defs of registers containing constants, which we in the predecessor block, where
only one of those constants is used in each successor. So, at each computed goto,
we needlessly spill about a 100 constants to stack. The end result is that a
clang-compiled python interpreter can be about ~2.5x slower on a simple python
reduction loop than a gcc-compiled interpreter.
Differential Revision: https://reviews.llvm.org/D29916
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@296060 91177308-0d34-0410-b5e6-96231b3b80d8
Justin added support for DISubprogram locs in r295531 and r296052.
Use that instead of no-loc for constants and arguments.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@296058 91177308-0d34-0410-b5e6-96231b3b80d8
Add an optimization remark to asm-printer that summarizes the number
of instructions emitted per function.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@296053 91177308-0d34-0410-b5e6-96231b3b80d8
Before this, MSan poisoned exactly one element of any array alloca,
even if the number of elements was zero.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@296050 91177308-0d34-0410-b5e6-96231b3b80d8
This allows the ability to call IPDBSession::getGlobalScope with a NativeSession and
to then query it for some basic fields from the PDB's InfoStream.
Note that the symbols now have non-const references back to the Session so that
NativeRawSymbol can access the PDBFile through the Session.
Differential Revision: https://reviews.llvm.org/D30314
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@296049 91177308-0d34-0410-b5e6-96231b3b80d8
We were stopping the translation of the parent block when the
translation of an instruction failed, but we were still trying to
translate the other blocks of the parent function.
Don't do that.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@296047 91177308-0d34-0410-b5e6-96231b3b80d8
This is the compromise between having a per-function IRTranslator
and manually managing the per-function state.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@296046 91177308-0d34-0410-b5e6-96231b3b80d8
Summary: In case we do not know what the condition is in an unswitched loop, but we know its definitely NOT a known constant. We can perform simplifcations based on this information.
Reviewers: sanjoy, hfinkel, chenli, efriedma
Reviewed By: efriedma
Subscribers: david2050, llvm-commits, mzolotukhin
Differential Revision: https://reviews.llvm.org/D28968
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@296041 91177308-0d34-0410-b5e6-96231b3b80d8
These are only used when emitting remarks without ORE directly using the free
functions emitOptimizationRemark*.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@296037 91177308-0d34-0410-b5e6-96231b3b80d8
Summary:
The helper will be used in a later change. This change itself is NFC
since the only user of this new function is its unit test.
Reviewers: majnemer, efriedma
Reviewed By: efriedma
Subscribers: efriedma, mcrosier, llvm-commits
Differential Revision: https://reviews.llvm.org/D30184
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@296035 91177308-0d34-0410-b5e6-96231b3b80d8
FastISel wasn't checking the isFPOnlySP subtarget feature before emitting
double-precision operations, so it got completely invalid CodeGen for doubles
on Cortex-M4F.
The normal ISel testing wasn't spectacular either so I added a second RUN line
to improve that while I was in the area.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@296031 91177308-0d34-0410-b5e6-96231b3b80d8
While not CVP's fault, this caused miscompiles (PR31181). Reverting
until those are resolved.
(This also reverts the follow-ups r288154 and r288161 which removed the
flag.)
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@296030 91177308-0d34-0410-b5e6-96231b3b80d8
Summary: SamplePGO uses branch_weight annotation to represent callsite hotness. When ICP promotes an indirect call to direct call, we need to make sure the direct call is annotated with branch_weight in SamplePGO mode, so that downstream function inliner can use hot callsite heuristic.
Reviewers: davidxl, eraman, xur
Reviewed By: davidxl, xur
Subscribers: mehdi_amini, llvm-commits
Differential Revision: https://reviews.llvm.org/D30282
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@296028 91177308-0d34-0410-b5e6-96231b3b80d8
In the bit tracker, references to other bit values in which the register
is 0 are prohibited. This means that generating self-referential register
cells like { w:32 [0-15]:s[0-15] [16-31]:s[15] } is impossible. In order
to get a self-referential cell, it had to be stored into a map and then
reloaded from it. To avoid this step, add a function that will set the
register to a given value without going through the map.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@296025 91177308-0d34-0410-b5e6-96231b3b80d8
Clang issues warning about hidden overload. That was intended, so
add "using AMDGPUGenRegisterInfo::getRegUnitWeight;" to mute it.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@296021 91177308-0d34-0410-b5e6-96231b3b80d8
Last use was killed in my previous patch. The preferred way is now to
construct the remark, pipe things to it and pass it to ORE.emit.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@296019 91177308-0d34-0410-b5e6-96231b3b80d8
The need for this removed when I converted everything to use the opt-remark
classes directly with the streaming interface.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@296017 91177308-0d34-0410-b5e6-96231b3b80d8
The need for this removed when I converted everything to use the opt-remark
classes directly with the streaming interface.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@296016 91177308-0d34-0410-b5e6-96231b3b80d8
Having more fine-grained information on the specific construct that
caused us to fallback is valuable for large-scale data collection.
We still have the fallback warning, that's also used for FastISel.
We still need to remove the fallback warning, and teach FastISel to also
emit remarks (it currently has a combination of the warning, stats, and
debug prints: the remarks could unify all three).
The abort-on-fallback path could also be better handled using remarks:
one could imagine a "-Rpass-error", analoguous to "-Werror", which would
promote missed/failed remarks to errors. It's not clear whether that
would be useful for other remarks though, so we're not there yet.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@296013 91177308-0d34-0410-b5e6-96231b3b80d8
If a subreg is used in an instruction it counts as a whole superreg
for the purpose of register pressure calculation. This patch corrects
improper register pressure calculation by examining operand's lane mask.
Differential Revision: https://reviews.llvm.org/D29835
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@296009 91177308-0d34-0410-b5e6-96231b3b80d8
This reverts commit r295749 while investigating PR32042.
It looks like this check uncovered a problem in the frontend that
needs to be fixed before the check can be enabled again.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@296005 91177308-0d34-0410-b5e6-96231b3b80d8
In OptimizeAdd, we scan the operand list to see if there are any common factors
between operands that can be factored out to reduce the number of multiplies
(e.g., 'A*A+A*B*C+D' -> 'A*(A+B*C)+D'). For each operand of the operand list, we
only consider unique factors (which is tracked by the Duplicate set). Now if we
find a factor that is a negative constant, we add the negated value as a factor
as well, because we can percolate the negate out. However, we mistakenly don't
add this negated constant to the Duplicates set.
Consider the expression A*2*-2 + B. Obviously, nothing to factor.
For the added value A*2*-2 we over count 2 as a factor without this change,
which causes the assert reported in PR30256. The problem is that this code is
assuming that all the multiply operands of the add are already reassociated.
This change avoids the issue by making OptimizeAdd tolerate multiplies which
haven't been completely optimized; this sort of works, but we're doing wasted
work: we'll end up revisiting the add later anyway.
Another possible approach would be to enforce RPO iteration order more strongly.
If we have RedoInsts, we process them immediately in RPO order, rather than
waiting until we've finished processing the whole function. Intuitively, it
seems like the natural approach: reassociation works on expression trees, so
the optimization only works in one direction. That said, I'm not sure how
practical that is given the current Reassociate; the "optimal" form for an
expression depends on its use list (see all the uses of "user_back()"), so
Reassociate is really an iterative optimization of sorts, so any changes here
would probably get messy.
PR30256
Differential Revision: https://reviews.llvm.org/D30228
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@296003 91177308-0d34-0410-b5e6-96231b3b80d8
Summary: The discriminator has been encoded, and only the base discriminator should be used during profile matching.
Reviewers: dblaikie, davidxl
Reviewed By: dblaikie, davidxl
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D30218
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@295999 91177308-0d34-0410-b5e6-96231b3b80d8
Since LoopInfo is not available in machine passes as universally as in IR
passes, using the same approach for OptimizationRemarkEmitter as we did for IR
will run LoopInfo and DominatorTree unnecessarily. (LoopInfo is not used
lazily by ORE.)
To fix this, I am modifying the approach I took in D29836. LazyMachineBFI now
uses its client passes including MachineBFI itself that are available or
otherwise compute them on the fly.
So for example GreedyRegAlloc, since it's already using MBFI, will reuse that
instance. On the other hand, AsmPrinter in Justin's patch will generate DT,
LI and finally BFI on the fly.
(I am of course wondering now if the simplicity of this approach is even
preferable in IR. I will do some experiments.)
Testing is provided by an updated version of D29837 which requires Justin's
patch to bring ORE to the AsmPrinter.
Differential Revision: https://reviews.llvm.org/D30128
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@295996 91177308-0d34-0410-b5e6-96231b3b80d8
Introduce a common ValueHandler for call returns and formal arguments, and
inherit two different versions for handling the differences (at the moment the
only difference is the way physical registers are marked as used).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@295973 91177308-0d34-0410-b5e6-96231b3b80d8
result
Summary:
If the same value is used several times as an extra value, SLP
vectorizer takes it into account only once instead of actual number of
using.
For example:
```
int val = 1;
for (int y = 0; y < 8; y++) {
for (int x = 0; x < 8; x++) {
val = val + input[y * 8 + x] + 3;
}
}
```
We have 2 extra rguments: `1` - initial value of horizontal reduction
and `3`, which is added 8*8 times to the reduction. Before the patch we
added `1` to the reduction value and added once `3`, though it must be
added 64 times.
Reviewers: mkuper, mzolotukhin
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D30262
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@295972 91177308-0d34-0410-b5e6-96231b3b80d8
Add support for lowering calls with parameters than can fit into regs. Use the
same ValueHandler that we used for function returns, but rename it to match its
new, extended purpose.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@295971 91177308-0d34-0410-b5e6-96231b3b80d8
This patch adjusts the most relaxed predicate of immediate operands to accept
immediate forms such as ~(0xf0000000|0x000f00000). Previously these forms
would be accepted by GAS and rejected by IAS.
This partially resolves PR/30383.
Thanks to Sean Bruno for reporting the issue!
Reviewers: slthakur, seanbruno
Differential Revision: https://reviews.llvm.org/D29218
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@295965 91177308-0d34-0410-b5e6-96231b3b80d8
The ARMConstantIslandPass didn't have support for handling accesses to
constant island objects through ARM::t2LDRBpci instructions. This adds
support for that.
This fixes PR31997.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@295964 91177308-0d34-0410-b5e6-96231b3b80d8
result
Summary:
If the same value is used several times as an extra value, SLP
vectorizer takes it into account only once instead of actual number of
using.
For example:
```
int val = 1;
for (int y = 0; y < 8; y++) {
for (int x = 0; x < 8; x++) {
val = val + input[y * 8 + x] + 3;
}
}
```
We have 2 extra rguments: `1` - initial value of horizontal reduction
and `3`, which is added 8*8 times to the reduction. Before the patch we
added `1` to the reduction value and added once `3`, though it must be
added 64 times.
Reviewers: mkuper, mzolotukhin
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D30262
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@295956 91177308-0d34-0410-b5e6-96231b3b80d8
result
Summary:
If the same value is used several times as an extra value, SLP
vectorizer takes it into account only once instead of actual number of
using.
For example:
```
int val = 1;
for (int y = 0; y < 8; y++) {
for (int x = 0; x < 8; x++) {
val = val + input[y * 8 + x] + 3;
}
}
```
We have 2 extra rguments: `1` - initial value of horizontal reduction
and `3`, which is added 8*8 times to the reduction. Before the patch we
added `1` to the reduction value and added once `3`, though it must be
added 64 times.
Reviewers: mkuper, mzolotukhin
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D30262
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@295949 91177308-0d34-0410-b5e6-96231b3b80d8
Implement isLegalToVectorizeLoadChain for AMDGPU to avoid
producing private address spaces accesses that will need to be
split up later. This was doing the wrong thing in the case
where the queried chain was an even number of elements.
A possible <4 x i32> store was being split into
store <2 x i32>
store i32
store i32
rather than
store <2 x i32>
store <2 x i32>
when legal.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@295933 91177308-0d34-0410-b5e6-96231b3b80d8
There were some older intrinsics that only existed for less than a month in 2012 that still exist in some out of tree test files that start with this string, but aren't able to be handled by the current upgrade code and fire an assert. Now we'll go back to treating them as not intrinsics at all and just passing them through to output.
Fixes PR32041, sort of.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@295930 91177308-0d34-0410-b5e6-96231b3b80d8
The manual is unclear on the details of this. It's not
clear to me if denormals are not allowed with clamp,
or if that is only omod. Not allowing denorms for
fp16 or fp64 isn't useful so I also question if that
is really a restriction. Same with whether this is valid
without IEEE mode enabled.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@295905 91177308-0d34-0410-b5e6-96231b3b80d8
We need to investigate this, but for now it just causes too much
headache when trying to run these tests.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@295900 91177308-0d34-0410-b5e6-96231b3b80d8
