This patch fixes PR24546, which demonstrates a segfault during the VSX
swap removal pass. The problem is that debug value instructions were
not excluded from the list of instructions to be analyzed for webs of
related computation. I've added the test case from the PR as a crash
test in test/CodeGen/PowerPC.
llvm-svn: 245862
The FP16_TO_FP node only uses the bottom 16 bits of its input, so the
following pattern can be optimised by removing the AND:
(FP16_TO_FP (AND op, 0xffff)) -> (FP16_TO_FP op)
This is a common pattern for ARM targets when functions have __fp16
arguments, as they are passed as floats (so that they get passed in the
correct registers), but then bitcast and truncated to ignore the top 16
bits.
llvm-svn: 245832
Not only do we not need to do anything to read correct values from the
object files, but the current logic actually wrongly applies twice the
section base address when there is no LoadedObjectInfo passed to the
DWARFContext creation (as the added test shows).
Simply do not apply any relocations on the mach-o debug info if there is
no load offset to apply.
llvm-svn: 245807
This patch adds all the refactored tests in new files, the old
tests will be removed by a followup commit.
Thanks to D. Blaikie for all the feedback.
llvm-svn: 245803
Summary:
WinEHPrepare is going to require that cleanuppad and catchpad produce values
of token type which are consumed by any cleanupret or catchret exiting the
pad. This change updates the signatures of those operators to require/enforce
that the type produced by the pads is token type and that the rets have an
appropriate argument.
The catchpad argument of a `CatchReturnInst` must be a `CatchPadInst` (and
similarly for `CleanupReturnInst`/`CleanupPadInst`). To accommodate that
restriction, this change adds a notion of an operator constraint to both
LLParser and BitcodeReader, allowing appropriate sentinels to be constructed
for forward references and appropriate error messages to be emitted for
illegal inputs.
Also add a verifier rule (noted in LangRef) that a catchpad with a catchpad
predecessor must have no other predecessors; this ensures that WinEHPrepare
will see the expected linear relationship between sibling catches on the
same try.
Lastly, remove some superfluous/vestigial casts from instruction operand
setters operating on BasicBlocks.
Reviewers: rnk, majnemer
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D12108
llvm-svn: 245797
Some debug info was drastically out of date, from the days where we used
to emit a list of length one (with a single null entry) rather than an
empty list (or, more recently, no list at all) for list fields that have
no elements.
llvm-svn: 245796
There was already a good error path for this. Added a test for it & made
a minor code change to ensure the error path was actually reached,
rather than crashing before we got that far.
llvm-svn: 245795
Summary:
__shared__ variable may now emit undef value as initializer, do not
throw error on that.
Test Plan: test/CodeGen/NVPTX/global-addrspace.ll
Patch by Xuetian Weng
Reviewers: jholewinski, tra, jingyue
Subscribers: llvm-commits, jholewinski
Differential Revision: http://reviews.llvm.org/D12242
llvm-svn: 245785
Summary:
Merge functions previously relied on unsigned comparisons of pointer values to
order functions. This caused observable non-determinism in the compiler for
large bitcode programs. Basically, opt -mergefuncs program.bc | md5sum produces
different hashes when run repeatedly on the same machine. Differing output was
observed on three large bitcodes, but it was less frequent on the smallest file.
It is possible that this only manifests on the large inputs, hence remaining
undetected until now.
This patch fixes this by removing (almost, see below) all places where
comparisons between pointers are used to order functions. Most of these changes
are local, but the comparison of global values requires assigning an identifier
to each local in the order it is visited. This is very similar to the way the
comparison function identifies Value*'s defined within a function. Because the
order of visiting the functions and their subparts is deterministic, the
identifiers assigned to the globals will be as well, and the order of functions
will be deterministic.
With these changes, there is no more observed non-determinism. There is also
only minor slowdowns (negligible to 4%) compared to the baseline, which is
likely a result of the fact that global comparisons involve hash lookups and not
just pointer comparisons.
The one caveat so far is that programs containing BlockAddress constants can
still be non-deterministic. It is not clear what the right solution is here. In
particular, even if the global numbers are used to order by function, we still
need a way to order the BasicBlock*'s. Unfortunately, we cannot just bail out
and fail to order the functions or consider them equal, because we require a
total order over functions. Note that programs with BlockAddress constants are
relatively rare, so the impact of leaving this in is minor as long as this pass
is opt-in.
Author: jrkoenig
Reviewers: nlewycky, jfb, dschuff
Subscribers: jevinskie, llvm-commits, chapuni
Differential revision: http://reviews.llvm.org/D12168
llvm-svn: 245762
SCEV expansion can invalidate previously expanded values. For example
in SCEVExpander::ReuseOrCreateCast, if we already have the requested
cast value but it's not at the desired location, a new cast is inserted
and the old cast will be invalidated.
Therefore, when expanding the bounds for the pointers, a later entry can
invalidate the IR value for an earlier one. The fix is to store a value
handle rather than the value itself.
The newly added test has a more detailed description of how the bug
triggers.
This bug can have a negative but potentially highly variable performance
impact in Loop Distribution. Because one of the bound values was
invalidated and is an undef expression now, InstCombine is free to
transform the array overlap check:
Start0 <= End1 && Start1 <= End0
into:
Start0 <= End1
So depending on the runtime location of the arrays, we would detect a
conflict and fall back on the original loop of the versioned loop.
Also tested compile time with SPEC2006 LTO bc files.
llvm-svn: 245760
We can wait on either VM, EXP or LGKM.
The waits are independent.
Without this patch, a wait inserted because of one of them
would also wait for all the previous others.
This patch makes s_wait only wait for the ones we need for the next
instruction.
Here's an example of subtle perf reduction this patch solves:
This is without the patch:
buffer_load_format_xyzw v[8:11], v0, s[44:47], 0 idxen
buffer_load_format_xyzw v[12:15], v0, s[48:51], 0 idxen
s_load_dwordx4 s[44:47], s[8:9], 0xc
s_waitcnt lgkmcnt(0)
buffer_load_format_xyzw v[16:19], v0, s[52:55], 0 idxen
s_load_dwordx4 s[48:51], s[8:9], 0x10
s_waitcnt vmcnt(1)
buffer_load_format_xyzw v[20:23], v0, s[44:47], 0 idxen
The s_waitcnt vmcnt(1) is useless.
The reason it is added is because the last
buffer_load_format_xyzw needs s[44:47], which was issued
by the first s_load_dwordx4. It waits for all VM
before that call to have finished.
Internally after every instruction, 3 counters (for VM, EXP and LGTM)
are updated after every instruction. For example buffer_load_format_xyzw
will
increase the VM counter, and s_load_dwordx4 the LGKM one.
Without the patch, for every defined register,
the current 3 counters are stored, and are used to know
how long to wait when an instruction needs the register.
Because of that, the s[44:47] counter includes that to use the register
you need to wait for the previous buffer_load_format_xyzw.
Instead this patch stores only the counters that matter for the
register,
and puts zero for the other ones, since we don't need any wait for them.
Patch by: Axel Davy
Differential Revision: http://reviews.llvm.org/D11883
llvm-svn: 245755
The original checkin was buggy, this change has a fix.
Original commit message:
[InstCombine] Transform A & (L - 1) u< L --> L != 0
Summary:
This transform is never a pessimization at the IR level (since it
replaces an `icmp` with another), and has potentiall payoffs:
1. It may make the `icmp` fold away or become loop invariant.
2. It may make the `A & (L - 1)` computation dead.
This shows up in Java, in range checks generated by array accesses of
the form `a[i & (a.length - 1)]`.
Reviewers: reames, majnemer
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D12210
llvm-svn: 245753
When PPCVSXFMAMutate would look at the input addend register, it would get its
input value number. This would fail, however, if the register was undef,
causing a segfault. Don't segfault (just skip such FMA instructions).
Fixes the test case from PR24542 (although that may have been over-reduced).
llvm-svn: 245741
See discussion in D12154 ( http://reviews.llvm.org/D12154 ), AMD Software
Optimization Guides for 10H/12H/15H/16H, and Agner Fog's experimental data.
llvm-svn: 245733
This will confirm that the patch in D12154 is actually NFC.
It will also confirm that the proposed changes for the AMD chips
are behaving as expected.
llvm-svn: 245704
This is intended to improve code generation for GEPs, as the index value is
shifted by the element size and in GEPs of multi-dimensional arrays the index
of higher dimensions is multiplied by the lower dimension size.
Differential Revision: http://reviews.llvm.org/D12197
llvm-svn: 245689
Note: I do not implement a base pointer, so it's still impossible to
have dynamic realignment AND dynamic alloca in the same function.
This also moves the code for determining the frame index reference
into getFrameIndexReference, where it belongs, instead of inline in
eliminateFrameIndex.
[Begin long-winded screed]
Now, stack realignment for Sparc is actually a silly thing to support,
because the Sparc ABI has no need for it -- unlike the situation on
x86, the stack is ALWAYS aligned to the required alignment for the CPU
instructions: 8 bytes on sparcv8, and 16 bytes on sparcv9.
However, LLVM unfortunately implements user-specified overalignment
using stack realignment support, so for now, I'm going to go along
with that tradition. GCC instead treats objects which have alignment
specification greater than the maximum CPU-required alignment for the
target as a separate block of stack memory, with their own virtual
base pointer (which gets aligned). Doing it that way avoids needing to
implement per-target support for stack realignment, except for the
targets which *actually* have an ABI-specified stack alignment which
is too small for the CPU's requirements.
Further unfortunately in LLVM, the default canRealignStack for all
targets effectively returns true, despite that implementing that is
something a target needs to do specifically. So, the previous behavior
on Sparc was to silently ignore the user's specified stack
alignment. Ugh.
Yet MORE unfortunate, if a target actually does return false from
canRealignStack, that also causes the user-specified alignment to be
*silently ignored*, rather than emitting an error.
(I started looking into fixing that last, but it broke a bunch of
tests, because LLVM actually *depends* on having it silently ignored:
some architectures (e.g. non-linux i386) have smaller stack alignment
than spilled-register alignment. But, the fact that a register needs
spilling is not known until within the register allocator. And by that
point, the decision to not reserve the frame pointer has been frozen
in place. And without a frame pointer, stack realignment is not
possible. So, canRealignStack() returns false, and
needsStackRealignment() then returns false, assuming everyone can just
go on their merry way assuming the alignment requirements were
probably just suggestions after-all. Sigh...)
Differential Revision: http://reviews.llvm.org/D12208
llvm-svn: 245668
The module splitter splits a module into linkable partitions. It will
be used to implement parallel LTO code generation.
This initial version of the splitter does not attempt to deal with the
somewhat subtle symbol visibility issues around module splitting. These
will be dealt with in a future change.
Differential Revision: http://reviews.llvm.org/D12132
llvm-svn: 245662
When producing conditional compare sequences for or operations we need
to negate the operands and the finally tested flags. The thing is if we negate
the finally tested flags this equals a logical negation of all previously
emitted expressions. There was a case missing where we have to order OR
expressions so they get emitted first.
This fixes http://llvm.org/PR24459
llvm-svn: 245641
Create CMP;CCMP sequences from and/or trees does not gain us anything if
the and/or tree is materialized to a GP register anyway. While most of
the code already checked for hasOneUse() there was one important case
missing.
llvm-svn: 245640
Summary:
This transform is never a pessimization at the IR level (since it
replaces an `icmp` with another), and has potentiall payoffs:
1. It may make the `icmp` fold away or become loop invariant.
2. It may make the `A & (L - 1)` computation dead.
This shows up in Java, in range checks generated by array accesses of
the form `a[i & (a.length - 1)]`.
Reviewers: reames, majnemer
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D12210
llvm-svn: 245635
Fixes PR23464: one way to use the broadcast intrinsics is:
_mm256_broadcastw_epi16(_mm_cvtsi32_si128(*(int*)src));
We don't currently fold this, but now that we use native IR for
the intrinsics (r245605), we can look through one bitcast to find
the broadcast scalar.
Differential Revision: http://reviews.llvm.org/D10557
llvm-svn: 245613
Summary:
Add an LSR test that exercises isTruncateFree. Without this change, LSR creates
another indvar representing the truncated value.
Reviewers: jholewinski, eliben
Subscribers: jholewinski, llvm-commits
Differential Revision: http://reviews.llvm.org/D12058
llvm-svn: 245611
Since r245605, the clang headers don't use these anymore.
r245165 updated some of the tests already; update the others, add
an autoupgrade, remove the intrinsics, and cleanup the definitions.
Differential Revision: http://reviews.llvm.org/D10555
llvm-svn: 245606
Instruction::dropUnknownMetadata(KnownSet) is supposed to preserve all
metadata in KnownSet, but the condition for DebugLocs was inverted.
Most users of dropUnknownMetadata() actually worked around this by not
adding LLVMContext::MD_dbg to their list of KnowIDs.
This is now made explicit.
llvm-svn: 245589
Caught by the famous "DebugLoc describes the currect SubProgram" assertion.
When GVN is removing a nonlocal load it updates the debug location of the
SSA value it replaced the load with with the one of the load. In the
testcase this actually overwrites a valid debug location with an empty one.
In reality GVN has to make an arbitrary choice between two equally valid
debug locations. This patch changes to behavior to only update the
location if the value doesn't already have a debug location.
llvm-svn: 245588
Summary: We know that -x & 1 is equivalent to x & 1, avoid using negation for testing if a negative integer is even or odd.
Reviewers: majnemer
Subscribers: junbuml, mssimpso, gberry, mcrosier, llvm-commits
Differential Revision: http://reviews.llvm.org/D12156
llvm-svn: 245569
COMISD should receive QWORD because it is defined as
(V)COMISD xmm1, xmm2/m64
COMISS should receive DWORD because it is defined as
(V)COMISS xmm1, xmm2/m32
Differential Revision: http://reviews.llvm.org/D11712
llvm-svn: 245551
Usually DSE is not supposed to remove lifetime intrinsics, but it's
actually ok to remove them for dead objects in terminating blocks,
because they convey no extra information there. Until we hit a lifetime
start that cannot be removed, that is. Because from that point on the
lifetime intrinsics become interesting again, e.g. for stack coloring.
Reviewers: reames
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D11710
llvm-svn: 245542
XVCMPEQDP is used for VSX v2f64 equality comparisons, but the value type needs
to be v2i64 (as that's the corresponding SETCC type).
Fixes PR24225.
llvm-svn: 245535
This DAGCombine was creating custom SDAG nodes with an illegal ppc_fp128
operand type because it was triggering on f64/f32 int2fp(fp2int(ppc_fp128 x)),
but shouldn't (it should only apply to f32/f64 types). The result was a crash.
llvm-svn: 245530
This commit modifies the serialization syntax so that the global IR values in
machine memory operands use the global value '@<name>' syntax instead of the
current '%ir.<name>' syntax.
The unnamed global IR values are handled by this commit as well, as the
existing global value parsing method can parse the unnamed globals already.
llvm-svn: 245527
The global IR values in machine memory operands should use the global value
'@<name>' syntax instead of the current '%ir.<name>' syntax.
However, the global value call entry pseudo source values use the global value
syntax already. Therefore, the syntax for the call entry pseudo source values
has to be changed so that the global values and call entry global value PSVs
can be parsed without ambiguities.
llvm-svn: 245526
We still need to add constant folding of vector comparisons to fold the tests for targets that don't support the respective min/max nodes
I needed to update 2011-12-06-AVXVectorExtractCombine to load a vector instead of using a constant vector to prevent it folding
Differential Revision: http://reviews.llvm.org/D12118
llvm-svn: 245503
We are already falling back to SelectionDAG when encountering an shift with UB.
This adds the same checks for shifts with UB that get folded into arithmetic or
logical operations.
This fixes rdar://problem/22345295.
llvm-svn: 245499
We don't do a great job with >= 0 comparisons against zero when the
result is used as an i8.
Given something like:
void f(long long LL, bool *B) {
*B = LL >= 0;
}
We used to generate:
shrq $63, %rdi
xorb $1, %dil
movb %dil, (%rsi)
Now we generate:
testq %rdi, %rdi
setns (%rsi)
Differential Revision: http://reviews.llvm.org/D12136
llvm-svn: 245498
Previously WebAssembly's datalayout string had -v128:8:128. This had been an
attempt to declare a certain level of support for unaligned SIMD accesses.
However, clang makes its own determinations for SIMD alignment that are
independent of the datalayout string, so this wasn't actually meaningful.
llvm-svn: 245494
Check to see if this is a CONCAT_VECTORS of a bunch of EXTRACT_SUBVECTOR operations. If so, and if the EXTRACT_SUBVECTOR vector inputs come from at most two distinct vectors the same size as the result, attempt to turn this into a legal shuffle.
Differential Revision: http://reviews.llvm.org/D12125
llvm-svn: 245490
This commit serializes the machine instruction's register operand ties.
The ties are printed out only when the instructon has register ties that are
different from the ties that are specified in the instruction's description.
llvm-svn: 245482
This revision has introduced an issue that only affects bootstrapped compiler
when it is printing the ASM. I am working on resolving the issue, but in the
meantime, I'm disabling the legalization of scalar_to_vector operation for v2i64
and the associated testing until I can get this fixed.
llvm-svn: 245481
The defined registers are already serialized - they are represented by placing
them before the '=' in a machine instruction. However, certain instructions like
INLINEASM can have defined register operands after the '=', so this commit
introduces the 'def' register flag for such operands.
llvm-svn: 245480
Reintroduce r245442. Remove an overly conservative assertion introduced
in r245442. We could replace the assertion to use `shareSameRegisterFile`
instead, but in that point in `insertPHI` we already lost the original
Def subreg to check against. So drop the assertion completely.
Original commit message:
- Teaches the ValueTracker in the PeepholeOptimizer to look through PHI
instructions.
- Add findNextSourceAndRewritePHI method to lookup into multiple sources
returnted by the ValueTracker and rewrite PHIs with new sources.
With these changes we can find more register sources and rewrite more
copies to allow coaslescing of bitcast instructions. Hence, we eliminate
unnecessary VR64 <-> GR64 copies in x86, but it could be extended to
other archs by marking "isBitcast" on target specific instructions. The
x86 example follows:
A:
psllq %mm1, %mm0
movd %mm0, %r9
jmp C
B:
por %mm1, %mm0
movd %mm0, %r9
jmp C
C:
movd %r9, %mm0
pshufw $238, %mm0, %mm0
Becomes:
A:
psllq %mm1, %mm0
jmp C
B:
por %mm1, %mm0
jmp C
C:
pshufw $238, %mm0, %mm0
Differential Revision: http://reviews.llvm.org/D11197
rdar://problem/20404526
llvm-svn: 245479
Since r244955, we try to use the short-form ErrorInfo when both
tries failed, and the long-form match failed on a suffix operand.
However, this means we sometimes mix ErrorInfo and MatchResult
(one manifestation of this being PR24498). Instead, restore both.
llvm-svn: 245469
This patch updates the X86 lowering so that the Exception Pointer and Selector
are 64-bit wide only if Subtarget.isTarget64BitLP64.
Patch by João Porto
Reviewers: dschuff, rnk
Differential Revision: http://reviews.llvm.org/D12111
llvm-svn: 245454
Reapply r243486.
- Teaches the ValueTracker in the PeepholeOptimizer to look through PHI
instructions.
- Add findNextSourceAndRewritePHI method to lookup into multiple sources
returnted by the ValueTracker and rewrite PHIs with new sources.
With these changes we can find more register sources and rewrite more
copies to allow coaslescing of bitcast instructions. Hence, we eliminate
unnecessary VR64 <-> GR64 copies in x86, but it could be extended to
other archs by marking "isBitcast" on target specific instructions. The
x86 example follows:
A:
psllq %mm1, %mm0
movd %mm0, %r9
jmp C
B:
por %mm1, %mm0
movd %mm0, %r9
jmp C
C:
movd %r9, %mm0
pshufw $238, %mm0, %mm0
Becomes:
A:
psllq %mm1, %mm0
jmp C
B:
por %mm1, %mm0
jmp C
C:
pshufw $238, %mm0, %mm0
Differential Revision: http://reviews.llvm.org/D11197
rdar://problem/20404526
llvm-svn: 245442
Summary:
The mid-end was generating vector smin/smax/umin/umax nodes, but
we were using vbsl to generatate the code. This adds the vmin/vmax
patterns and a test to check that we are now generating vmin/vmax
instructions.
Reviewers: rengolin, jmolloy
Subscribers: aemerson, rengolin, llvm-commits
Differential Revision: http://reviews.llvm.org/D12105
llvm-svn: 245439
There are some cases where the mul sequence is smaller, but for the most part,
using a div is preferable. This does not apply to vectors, since x86 doesn't
have vector idiv, and a vector mul/shifts sequence ought to be smaller than a
scalarized division.
Differential Revision: http://reviews.llvm.org/D12082
llvm-svn: 245431
Fix how DependenceAnalysis calls delinearization, mirroring what is done in
Delinearization.cpp (mostly by making sure to call getSCEVAtScope before
delinearizing, and by removing the unnecessary 'Pairs == 1' check).
Patch by Vaivaswatha Nagaraj!
llvm-svn: 245408
Here we make ScalarEvolution::isKnownPredicate, indirectly, a little smarter.
Given some relational comparison operator OP, and two AddRec SCEVs, {I,+,S} OP
{J,+,T}, we can reduce this to the comparison I OP J when S == T, both AddRecs
are for the same loop, and both are known not to wrap.
As it turns out, because of the way that backedge-guard expressions can be
leveraged when computing known predicates, this allows indvars to simplify the
if-statement comparison in this loop:
void foo (int *a, int *b, int n) {
for (int i = 0; i < n; ++i) {
if (i > n)
a[i] = b[i] + 1;
}
}
which, somewhat surprisingly, we were not previously optimizing away.
llvm-svn: 245400
This commit adds support for bit mask target flag serialization to the MIR
printer and the MIR parser. It also adds support for the machine operand's
target flag serialization to the AArch64 target.
Reviewers: Duncan P. N. Exon Smith
llvm-svn: 245383
To properly handle this, define the *a instructions as separate
instruction classes by refactoring the LoadA and StoreA multiclasses.
Move the instruction tests into the sparcv9 file to test the difference.
llvm-svn: 245360
The current code normalizes select(C0, x, select(C1, x, y)) towards
select(C0|C1, x, y) if the targets prefers that form. This patch adds an
additional rule that if the select(C1, x, y) part already exists in the
function then we want to normalize into the other direction because the
effects of reusing the existing value are bigger than transforming into
the target preferred form.
This addresses regressions following r238793, see also:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20150727/290272.html
Differential Revision: http://reviews.llvm.org/D11616
llvm-svn: 245350
State numbers are calculated by performing a walk from the innermost
funclet to the outermost funclet. Rudimentary support for the new EH
constructs has been added to the assembly printer, just enough to test
the new machinery.
Differential Revision: http://reviews.llvm.org/D12098
llvm-svn: 245331
Summary: This is the correct way to handle JAL instructions when PIC is enabled.
Patch by Toma Tabacu
Reviewers: seanbruno, tomatabacu
Subscribers: brooks, seanbruno, emaste, llvm-commits
Differential Revision: http://reviews.llvm.org/D6231
llvm-svn: 245305
This is (almost) everything under MC/MachO/ARM. There are still some
cases missing, because llvm-readobj doesn't (yet) support some features,
that macho-dump provides. I plan to reduce the gap between them shortly.
llvm-svn: 245302
After hitting @llvm.assume(X) we can:
- propagate equality that X == true
- if X is icmp/fcmp (with eq operation), and one of operand
is constant we can change all variables with constants in the same BasicBlock
http://reviews.llvm.org/D11918
llvm-svn: 245265
It is possible to be in a situation where more than one funclet token is
a valid SSA value. If we see a terminator which exits a funclet which
doesn't use the funclet's token, replace it with unreachable.
Differential Revision: http://reviews.llvm.org/D12074
llvm-svn: 245238
Summary:
Increase the estimated costs for insert/extract element operations on
AArch64. This is motivated by results from benchmarking interleaved
accesses.
Add missing costs for zext/sext/trunc instructions and some integer to
floating point conversions. These costs were previously calculated
by scalarizing these operation and were affected by the cost increase of
the insert/extract element operations.
Reviewers: rengolin
Subscribers: mcrosier, aemerson, rengolin, llvm-commits
Differential Revision: http://reviews.llvm.org/D11939
llvm-svn: 245226
Summary:
This change limits the minimum cost of an insert/extract
element operation to 2 in cases where this would result
in mixing of NEON and VFP code.
Reviewers: rengolin
Subscribers: mssimpso, aemerson, llvm-commits, rengolin
Differential Revision: http://reviews.llvm.org/D12030
llvm-svn: 245225
Summary:
When demoting an SSA value that has a use on a phi and one of the phi's
predecessors terminates with catchret, the edge needs to be split and the
load inserted in the new block, else we'll still have a cross-funclet SSA
value.
Add a test for this, and for the similar case where a def to be spilled is
on and invoke and a critical edge, which was already implemented but
missing a test.
Reviewers: majnemer
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D12065
llvm-svn: 245218
Summary: It is the same as LA, except that it can also load 64-bit addresses and it only works on 64-bit MIPS architectures.
Reviewers: tomatabacu, seanbruno, vkalintiris
Subscribers: brooks, seanbruno, emaste, llvm-commits
Differential Revision: http://reviews.llvm.org/D9524
llvm-svn: 245208
These only get generated if the target supports them. If one of the variants is not legal and the other is, and it is safe to do so, the other variant will be emitted.
For example on AArch32 (V8), we have scalar fminnm but not fmin.
Fix up a couple of tests while we're here - one now produces better code, and the other was just plain wrong to start with.
llvm-svn: 245196
PR24469 resulted because DeleteDeadInstruction in handleNonLocalStoreDeletion was
deleting the next basic block iterator. Fixed the same by resetting the basic block iterator
post call to DeleteDeadInstruction.
llvm-svn: 245195
This change makes ScalarEvolution a stand-alone object and just produces
one from a pass as needed. Making this work well requires making the
object movable, using references instead of overwritten pointers in
a number of places, and other refactorings.
I've also wired it up to the new pass manager and added a RUN line to
a test to exercise it under the new pass manager. This includes basic
printing support much like with other analyses.
But there is a big and somewhat scary change here. Prior to this patch
ScalarEvolution was never *actually* invalidated!!! Re-running the pass
just re-wired up the various other analyses and didn't remove any of the
existing entries in the SCEV caches or clear out anything at all. This
might seem OK as everything in SCEV that can uses ValueHandles to track
updates to the values that serve as SCEV keys. However, this still means
that as we ran SCEV over each function in the module, we kept
accumulating more and more SCEVs into the cache. At the end, we would
have a SCEV cache with every value that we ever needed a SCEV for in the
entire module!!! Yowzers. The releaseMemory routine would dump all of
this, but that isn't realy called during normal runs of the pipeline as
far as I can see.
To make matters worse, there *is* actually a key that we don't update
with value handles -- there is a map keyed off of Loop*s. Because
LoopInfo *does* release its memory from run to run, it is entirely
possible to run SCEV over one function, then over another function, and
then lookup a Loop* from the second function but find an entry inserted
for the first function! Ouch.
To make matters still worse, there are plenty of updates that *don't*
trip a value handle. It seems incredibly unlikely that today GVN or
another pass that invalidates SCEV can update values in *just* such
a way that a subsequent run of SCEV will incorrectly find lookups in
a cache, but it is theoretically possible and would be a nightmare to
debug.
With this refactoring, I've fixed all this by actually destroying and
recreating the ScalarEvolution object from run to run. Technically, this
could increase the amount of malloc traffic we see, but then again it is
also technically correct. ;] I don't actually think we're suffering from
tons of malloc traffic from SCEV because if we were, the fact that we
never clear the memory would seem more likely to have come up as an
actual problem before now. So, I've made the simple fix here. If in fact
there are serious issues with too much allocation and deallocation,
I can work on a clever fix that preserves the allocations (while
clearing the data) between each run, but I'd prefer to do that kind of
optimization with a test case / benchmark that shows why we need such
cleverness (and that can test that we actually make it faster). It's
possible that this will make some things faster by making the SCEV
caches have higher locality (due to being significantly smaller) so
until there is a clear benchmark, I think the simple change is best.
Differential Revision: http://reviews.llvm.org/D12063
llvm-svn: 245193
If we can ignore NaNs, fmin/fmax libcalls can become compare and select
(this is what we turn std::min / std::max into).
This IR should then be optimized in the backend to whatever is best for
any given target. Eg, x86 can use minss/maxss instructions.
This should solve PR24314:
https://llvm.org/bugs/show_bug.cgi?id=24314
Differential Revision: http://reviews.llvm.org/D11866
llvm-svn: 245187
Bitwise arithmetic can obscure a simple sign-test. If replacing the
mask with a truncate is preferable if the type is legal because it
permits us to rephrase the comparison more explicitly.
llvm-svn: 245171
We can set additional bits in a mask given that we know the other
operand of an AND already has some bits set to zero. This can be more
efficient if doing so allows us to use an instruction which implicitly
sign extends the immediate.
This fixes PR24085.
Differential Revision: http://reviews.llvm.org/D11289
llvm-svn: 245169
For cases where we TRUNCATE and then ZERO_EXTEND to a larger size (often from vector legalization), see if we can mask the source data and then ZERO_EXTEND (instead of after a ANY_EXTEND). This can help avoid having to generate a larger mask, and possibly applying it to several sub-vectors.
(zext (truncate x)) -> (zext (and(x, m))
Includes a minor patch to SystemZ to better recognise 8/16-bit zero extension patterns from RISBG bit-extraction code.
This is the first of a number of minor patches to help improve the conversion of byte masks to clear mask shuffles.
Differential Revision: http://reviews.llvm.org/D11764
llvm-svn: 245160
Some personality routines require funclet exit points to be clearly
marked, this is done by producing a token at the funclet pad and
consuming it at the corresponding ret instruction. CleanupReturnInst
already had a spot for this operand but CatchReturnInst did not.
Other personality routines don't need to use this which is why it has
been made optional.
llvm-svn: 245149
This patch makes the Merge Functions pass faster by calculating and comparing
a hash value which captures the essential structure of a function before
performing a full function comparison.
The hash is calculated by hashing the function signature, then walking the basic
blocks of the function in the same order as the main comparison function. The
opcode of each instruction is hashed in sequence, which means that different
functions according to the existing total order cannot have the same hash, as
the comparison requires the opcodes of the two functions to be the same order.
The hash function is a static member of the FunctionComparator class because it
is tightly coupled to the exact comparison function used. For example, functions
which are equivalent modulo a single variant callsite might be merged by a more
aggressive MergeFunctions, and the hash function would need to be insensitive to
these differences in order to exploit this.
The hashing function uses a utility class which accumulates the values into an
internal state using a standard bit-mixing function. Note that this is a different interface
than a regular hashing routine, because the values to be hashed are scattered
amongst the properties of a llvm::Function, not linear in memory. This scheme is
fast because only one word of state needs to be kept, and the mixing function is
a few instructions.
The main runOnModule function first computes the hash of each function, and only
further processes functions which do not have a unique function hash. The hash
is also used to order the sorted function set. If the hashes differ, their
values are used to order the functions, otherwise the full comparison is done.
Both of these are helpful in speeding up MergeFunctions. Together they result in
speedups of 9% for mysqld (a mostly C application with little redundancy), 46%
for libxul in Firefox, and 117% for Chromium. (These are all LTO builds.) In all
three cases, the new speed of MergeFunctions is about half that of the module
verifier, making it relatively inexpensive even for large LTO builds with
hundreds of thousands of functions. The same functions are merged, so this
change is free performance.
Author: jrkoenig
Reviewers: nlewycky, dschuff, jfb
Subscribers: llvm-commits, aemerson
Differential revision: http://reviews.llvm.org/D11923
llvm-svn: 245140
This seems to only work some of the time. In some situations,
this seems to use a nonsensical type and isn't actually aware of the
memory being accessed. e.g. if branch condition is an icmp of a pointer,
it checks the addressing mode of i1.
llvm-svn: 245137
Dan Gohman