| Commit message (Collapse) | Author | Age | Files | Lines | 
| | 
| 
| 
| 
| 
| 
| 
| 
|  | 
file.
Changing code that is covered by these tests is just too hard to debug
currently, and now it will be clear the nature of the changes.
llvm-svn: 216643
 | 
| | 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
|  | 
The included test case would fail, because the MI PHI node would have two
operands from the same predecessor.
This problem occurs when a switch instruction couldn't be selected. This happens
always, because there is no default switch support for FastISel to begin with.
The problem was that FastISel would first add the operand to the PHI nodes and
then fall-back to SelectionDAG, which would then in turn add the same operands
to the PHI nodes again.
This fix removes these duplicate PHI node operands by reseting the
PHINodesToUpdate to its original state before FastISel tried to select the
instruction.
This fixes <rdar://problem/18155224>.
llvm-svn: 216640
 | 
| | 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
|  | 
Currently instructions are folded very aggressively for AArch64 into the memory
operation, which can lead to the use of killed operands:
  %vreg1<def> = ADDXri %vreg0<kill>, 2
  %vreg2<def> = LDRBBui %vreg0, 2
  ... = ... %vreg1 ...
This usually happens when the result is also used by another non-memory
instruction in the same basic block, or any instruction in another basic block.
This fix teaches hasTrivialKill to not only check the LLVM IR that the value has
a single use, but also to check if the register that represents that value has
already been used. This can happen when the instruction with the use was folded
into another instruction (in this particular case a load instruction).
This fixes rdar://problem/18142857.
llvm-svn: 216634
 | 
| | 
| 
| 
| 
| 
| 
| 
|  | 
the memory operation."
Quentin pointed out that this is not the correct approach and there is a better and easier solution.
llvm-svn: 216632
 | 
| | 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
|  | 
operation.
Currently instructions are folded very aggressively into the memory operation,
which can lead to the use of killed operands:
  %vreg1<def> = ADDXri %vreg0<kill>, 2
  %vreg2<def> = LDRBBui %vreg0, 2
  ... = ... %vreg1 ...
This usually happens when the result is also used by another non-memory
instruction in the same basic block, or any instruction in another basic block.
If the computed address is used by only memory operations in the same basic
block, then it is safe to fold them. This is because all memory operations will
fold the address computation and the original computation will never be emitted.
This fixes rdar://problem/18142857.
llvm-svn: 216629
 | 
| | 
| 
| 
| 
| 
| 
| 
| 
|  | 
When the address comes directly from a shift instruction then the address
computation cannot be folded into the memory instruction, because the zero
register is not available as a base register. Simplify addess needs to emit the
shift instruction and use the result as base register.
llvm-svn: 216621
 | 
| | 
| 
| 
| 
| 
| 
| 
| 
|  | 
Use the zero register directly when possible to avoid an unnecessary register
copy and a wasted register at -O0. This also uses integer stores to store a
positive floating-point zero. This saves us from materializing the positive zero
in a register and then storing it.
llvm-svn: 216617
 | 
| | 
| 
| 
| 
| 
| 
| 
|  | 
This teaches the AArch64 backend to deal with the operations required
to deal with the operations on v4f16 and v8f16 which are exposed by
NEON intrinsics, plus the add, sub, mul and div operations.
llvm-svn: 216555
 | 
| | 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
|  | 
we stopped efficiently lowering sextload using the SSE41 instructions
for that operation.
This is a consequence of a bad predicate I used thinking of the memory
access needs. The code actually handles the cases where the predicate
doesn't apply, and handles them much better. =] Simple fix and a test
case added. Fixes PR20767.
llvm-svn: 216538
 | 
| | 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
|  | 
This combine is essentially combining target-specific nodes back into target
independent nodes that it "knows" will be combined yet again by a target
independent DAG combine into a different set of target-independent nodes that
are legal (not custom though!) and thus "ok". This seems... deeply flawed. The
crux of the problem is that we don't combine un-legalized shuffles that are
introduced by legalizing other operations, and thus we don't see a very
profitable combine opportunity. So the backend just forces the input to that
combine to re-appear.
However, for this to work, the conditions detected to re-form the unlegalized
nodes must be *exactly* right. Previously, failing this would have caused poor
code (if you're lucky) or a crasher when we failed to select instructions.
After r215611 we would fall back into the legalizer. In some cases, this just
"fixed" the crasher by produces bad code. But in the test case added it caused
the legalizer and the dag combiner to iterate forever.
The fix is to make the alignment checking in the x86 side of things match the
alignment checking in the generic DAG combine exactly. This isn't really a
satisfying or principled fix, but it at least make the code work as intended.
It also highlights that it would be nice to detect the availability of under
aligned loads for a given type rather than bailing on this optimization. I've
left a FIXME to document this.
Original commit message for r215611 which covers the rest of the chang:
  [SDAG] Fix a case where we would iteratively legalize a node during
  combining by replacing it with something else but not re-process the
  node afterward to remove it.
  In a truly remarkable stroke of bad luck, this would (in the test case
  attached) end up getting some other node combined into it without ever
  getting re-processed. By adding it back on to the worklist, in addition
  to deleting the dead nodes more quickly we also ensure that if it
  *stops* being dead for any reason it makes it back through the
  legalizer. Without this, the test case will end up failing during
  instruction selection due to an and node with a type we don't have an
  instruction pattern for.
It took many million runs of the shuffle fuzz tester to find this.
llvm-svn: 216537
 | 
| | 
| 
| 
|  | 
llvm-svn: 216530
 | 
| | 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
|  | 
When a shift with extension or an add with shift and extension cannot be folded
into the memory operation, then the address calculation has to be materialized
separately. While doing so the code forgot to consider a possible sign-/zero-
extension. This fix folds now also the sign-/zero-extension into the add or
shift instruction which is used to materialize the address.
This fixes rdar://problem/18141718.
llvm-svn: 216511
 | 
| | 
| 
| 
|  | 
llvm-svn: 216510
 | 
| | 
| 
| 
|  | 
llvm-svn: 216451
 | 
| | 
| 
| 
| 
| 
| 
|  | 
Patterns for lowering libm calls to VCVT{A,N,P,M} are also included.
Phabricator Revision: http://reviews.llvm.org/D5033
llvm-svn: 216388
 | 
| | 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
|  | 
Adds code generation support for dcbtst (data cache prefetch for write) and
icbt (instruction cache prefetch for read - Book E cores only).
We still end up with a 'cannot select' error for the non-supported prefetch
intrinsic forms. This will be fixed in a later commit.
Fixes PR20692.
llvm-svn: 216339
 | 
| | 
| 
| 
| 
| 
| 
|  | 
This reverts commit r215862 due to nightly failures.  Will work on getting a
reduced test case, but I wanted to get our bots green in the meantime.
llvm-svn: 216325
 | 
| | 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
|  | 
these DAG combines.
The DAG auto-CSE thing is truly terrible. Due to it, when RAUW-ing
a node with its operand, you can cause its uses to CSE to itself, which
then causes their uses to become your uses which causes them to be
picked up by the RAUW. For nodes that are determined to be "no-ops",
this is "fine". But if the RAUW is one of several steps to enact
a transformation, this causes the DAG to really silently eat an discard
nodes that you would never expect. It took days for me to actually
pinpoint a test case triggering this and a really frustrating amount of
time to even comprehend the bug because I never even thought about the
ability of RAUW to iteratively consume nodes due to CSE-ing them into
itself.
To fix this, we have to build up a brand-new chain of operations any
time we are combining across (potentially) intervening nodes. But once
the logic is added to do this, another issue surfaces: CombineTo eagerly
deletes the one node combined, *but no others*. This is... really
frustrating. If deleting it makes its operands become dead, those
operand nodes often won't go onto the worklist in the
order you would want -- they're already on it and not near the top. That
means things higher on the worklist will get combined prior to these
dead nodes being GCed out of the worklist, and if the chain is long, the
immediate users won't be enough to re-detect where the root of the chain
is that became single-use again after deleting the dead nodes. The
better way to do this is to never immediately delete nodes, and instead
to just enqueue them so we can recursively delete them. The
combined-from node is typically not on the worklist anyways by virtue of
having been popped off.... But that in turn breaks other tests that
*require* CombineTo to delete unused nodes. :: sigh ::
Fortunately, there is a better way. This whole routine should have been
returning the replacement rather than using CombineTo which is quite
hacky. Switch to that, and all the pieces fall together.
I suspect the same kind of miscompile is possible in the half-shuffle
folding code, and potentially the recursive folding code. I'll be
switching those over to a pattern more like this one for safety's sake
even though I don't immediately have any test cases for them. Note that
the only way I got a test case for this instance was with *heavily* DAG
combined 256-bit shuffle sequences generated by my fuzzer. ;]
llvm-svn: 216319
 | 
| | 
| 
| 
| 
| 
|  | 
reduced testcase in that bug.
llvm-svn: 216307
 | 
| | 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
|  | 
There's no need to do this if the user doesn't call va_start. In the
future, we're going to have thunks that forward these register
parameters with musttail calls, and they won't need these spills for
handling va_start.
Most of the test suite changes are adding va_start calls to existing
tests to keep things working.
llvm-svn: 216294
 | 
| | 
| 
| 
|  | 
llvm-svn: 216279
 | 
| | 
| 
| 
|  | 
llvm-svn: 216278
 | 
| | 
| 
| 
| 
| 
| 
| 
| 
|  | 
instruction from ARMInstrInfo to ARMBaseInstrInfo.
That way, thumb mode can also benefit from the advanced copy optimization.
<rdar://problem/12702965>
llvm-svn: 216274
 | 
| | 
| 
| 
| 
| 
| 
| 
|  | 
calling convention if FP is 64-bit and +nooddspreg is used.
Differential Revision: http://reviews.llvm.org/D4981.diff
llvm-svn: 216262
 | 
| | 
| 
| 
| 
| 
| 
| 
|  | 
This adds the missing variable shift support for value type i8, i16, and i32.
This fixes <rdar://problem/18095685>.
llvm-svn: 216242
 | 
| | 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
|  | 
happy.
This is mostly achieved by providing the correct register class manually,
because getRegClassFor always returns the GPR*AllRegClass for MVT::i32 and
MVT::i64.
Also cleanup the code to use the FastEmitInst_* method whenever possible. This
makes sure that the operands' register class is properly constrained. For all
the remaining cases this adds the missing constrainOperandRegClass calls for
each operand.
llvm-svn: 216225
 | 
| | 
| 
| 
|  | 
llvm-svn: 216220
 | 
| | 
| 
| 
| 
| 
|  | 
This fixes a crash in an ocl conformance test.
llvm-svn: 216219
 | 
| | 
| 
| 
| 
| 
| 
| 
| 
|  | 
The AdvSIMD pass may produce copies that are not coalescer-friendly. The
peephole optimizer knows how to fix that as demonstrated in the test case.
<rdar://problem/12702965>
llvm-svn: 216200
 | 
| | 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
|  | 
There are two add-immediate instructions in Thumb1: tADDi8 and tADDi3. Only
the latter supports using different source and destination registers, so
whenever we materialize a new base register (at a certain offset) we'd do
so by moving the base register value to the new register and then adding in
place. This patch changes the code to use a single tADDi3 if the offset is
small enough to fit in 3 bits.
Differential Revision: http://reviews.llvm.org/D5006
llvm-svn: 216193
 | 
| | 
| 
| 
| 
| 
|  | 
These tests and many more are already covered by fast-isel-addressing-modes.ll.
llvm-svn: 216186
 | 
| | 
| 
| 
| 
| 
| 
| 
|  | 
systems
http://reviews.llvm.org/D4984
llvm-svn: 216182
 | 
| | 
| 
| 
| 
| 
| 
| 
|  | 
with many arguments.
PR20677
llvm-svn: 216175
 | 
| | 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
|  | 
The FPv4-SP floating-point unit is generally referred to as
single-precision only, but it does have double-precision registers and
load, store and GPR<->DPR move instructions which operate on them.
This patch enables the use of these registers, the main advantage of
which is that we now comply with the AAPCS-VFP calling convention.
This partially reverts r209650, which added some AAPCS-VFP support,
but did not handle return values or alignment of double arguments in
registers.
This patch also adds tests for Thumb2 code generation for
floating-point instructions and intrinsics, which previously only
existed for ARM.
llvm-svn: 216172
 | 
| | 
| 
| 
|  | 
llvm-svn: 216164
 | 
| | 
| 
| 
|  | 
llvm-svn: 216162
 | 
| | 
| 
| 
| 
| 
|  | 
Builder and type".
llvm-svn: 216147
 | 
| | 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
|  | 
advanced copy optimization.
This is the final step patch toward transforming:
udiv    r0, r0, r2
udiv    r1, r1, r3
vmov.32 d16[0], r0
vmov.32 d16[1], r1
vmov    r0, r1, d16
bx      lr
into:
udiv    r0, r0, r2
udiv    r1, r1, r3
bx      lr
Indeed, thanks to this patch, this optimization is able to look through
vmov.32 d16[0], r0
vmov.32 d16[1], r1
and is able to rewrite the following sequence:
vmov.32 d16[0], r0
vmov.32 d16[1], r1
vmov    r0, r1, d16
into simple generic GPR copies that the coalescer managed to remove.
<rdar://problem/12702965>
llvm-svn: 216144
 | 
| | 
| 
| 
| 
| 
| 
| 
| 
| 
| 
|  | 
On pre-v6 hardware, 'MOV lo, lo' gives undefined results, so such copies need to
be avoided. This patch trades simplicity for implementation time at the expense
of performance... As they say: correctness first, then performance.
See http://lists.cs.uiuc.edu/pipermail/llvmdev/2014-August/075998.html for a few
ideas on how to make this better.
llvm-svn: 216138
 | 
| | 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
|  | 
Fix for PR20648 - http://llvm.org/bugs/show_bug.cgi?id=20648
This patch checks the operands of a vselect to see if all values are constants.
If yes, bail out of any further attempts to create a blend or shuffle because
SelectionDAGLegalize knows how to turn this kind of vselect into a single load.
This already happens for machines without SSE4.1, so the added checks just send
more targets down that path.
Differential Revision: http://reviews.llvm.org/D4934
llvm-svn: 216121
 | 
| | 
| 
| 
|  | 
llvm-svn: 216120
 | 
| | 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
|  | 
The goal of the patch is to implement section 3.2.3 of the AMD64 ABI
correctly.  The controlling sentence is, "The size of each argument gets
rounded up to eightbytes.  Therefore the stack will always be eightbyte
aligned." The equivalent sentence in the i386 ABI page 37 says, "At all
times, the stack pointer should point to a word-aligned area."  For both
architectures, the stack pointer is not being rounded up to the nearest
eightbyte or word between the last normal argument and the first
variadic argument.
Patch by Thomas Jablin!
llvm-svn: 216119
 | 
| | 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
|  | 
Summary: This fixes http://llvm.org/bugs/show_bug.cgi?id=19530.
The problem is that X86ISelLowering erroneously thought the third call
was eligible for tail call elimination.
It would have been if it's return value was actually the one returned
by the calling function, but here that is not the case and
additional values are being returned.
Test Plan: Test case from the original bug report is included.
Reviewers: rafael
Reviewed By: rafael
Subscribers: rafael, llvm-commits
Differential Revision: http://reviews.llvm.org/D4968
llvm-svn: 216117
 | 
| | 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
|  | 
(PR20308)
In PR20308 ( http://llvm.org/bugs/show_bug.cgi?id=20308 ), the critical-anti-dependency breaker
caused a miscompile because it broke a WAR hazard using a register that it thinks is available
based on info from a kill inst. Until PR18663 is solved, we shouldn't use any def/use info from
a kill because they are really just nops.
This patch adds guard checks for kills around calls to ScanInstruction() where the DefIndices
array is set. For good measure, add an assert in ScanInstruction() so we don't hit this bug again.
The test case is a reduced version of the code from the bug report.
Differential Revision: http://reviews.llvm.org/D4977
llvm-svn: 216114
 | 
| | 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
|  | 
the isRegSequence property.
This is a follow-up of r215394 and r215404, which respectively introduces the
isRegSequence property and uses it for ARM.
Thanks to the property introduced by the previous commits, this patch is able
to optimize the following sequence:
vmov	d0, r2, r3
vmov	d1, r0, r1
vmov	r0, s0
vmov	r1, s2
udiv	r0, r1, r0
vmov	r1, s1
vmov	r2, s3
udiv	r1, r2, r1
vmov.32	d16[0], r0
vmov.32	d16[1], r1
vmov	r0, r1, d16
bx	lr
into:
udiv	r0, r0, r2
udiv	r1, r1, r3
vmov.32	d16[0], r0
vmov.32	d16[1], r1
vmov	r0, r1, d16
bx	lr
This patch refactors how the copy optimizations are done in the peephole
optimizer. Prior to this patch, we had one copy-related optimization that
replaced a copy or bitcast by a generic, more suitable (in terms of register
file), copy.
With this patch, the peephole optimizer features two copy-related optimizations:
1. One for rewriting generic copies to generic copies:
PeepholeOptimizer::optimizeCoalescableCopy.
2. One for replacing non-generic copies with generic copies:
PeepholeOptimizer::optimizeUncoalescableCopy.
The goals of these two optimizations are slightly different: one rewrite the
operand of the instruction (#1), the other kills off the non-generic instruction
and replace it by a (sequence of) generic instruction(s).
Both optimizations rely on the ValueTracker introduced in r212100.
The ValueTracker has been refactored to use the information from the
TargetInstrInfo for non-generic instruction. As part of the refactoring, we
switched the tracking from the index of the definition to the actual register
(virtual or physical). This one change is to provide better consistency with
register related APIs and to ease the use of the TargetInstrInfo.
Moreover, this patch introduces a new helper class CopyRewriter used to ease the
rewriting of generic copies (i.e., #1).
Finally, this patch adds a dead code elimination pass right after the peephole
optimizer to get rid of dead code that may appear after rewriting.
This is related to <rdar://problem/12702965>.
Review: http://reviews.llvm.org/D4874
llvm-svn: 216088
 | 
| | 
| 
| 
| 
| 
| 
| 
| 
| 
|  | 
This fixes a bug I introduced in a previous commit (r216033). Sign-/Zero-
extension from i1 cannot be folded into the ADDS/SUBS instructions. Instead both
operands have to be sign-/zero-extended with separate instructions.
Related to <rdar://problem/17913111>.
llvm-svn: 216073
 | 
| | 
| 
| 
| 
| 
| 
| 
|  | 
legalization stage. With those two optimizations, fewer signed/zero extension
instructions can be inserted, and then we can expose more opportunities to
Machine CSE pass in back-end.
llvm-svn: 216066
 | 
| | 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
|  | 
Summary:
Fixes http://llvm.org/bugs/show_bug.cgi?id=20016 reproducible on new
lea-5.ll case.
Also use RSP/RBP for x32 lea to save 1 byte used for 0x67 prefix in
ESP/EBP case.
Test Plan: lea tests modified to include x32/nacl and new test added
Reviewers: nadav, dschuff, t.p.northover
Subscribers: llvm-commits, zinovy.nis
Differential Revision: http://reviews.llvm.org/D4929
llvm-svn: 216065
 | 
| | 
| 
| 
| 
| 
| 
| 
| 
| 
| 
| 
|  | 
LLVM generates illegal `rbit r0, #352` instruction for rbit intrinsic.
According to ARM ARM, rbit only takes register as argument, not immediate.
The correct instruction should be rbit <Rd>, <Rm>.
The bug was originally introduced in r211057.
Differential Revision: http://reviews.llvm.org/D4980
llvm-svn: 216064
 | 
| | 
| 
| 
| 
| 
| 
| 
| 
| 
|  | 
Use FMOVWSr/FMOVXDr instead of FMOVSr/FMOVDr, which have the proper register
class to be used with the zero register. This makes the MachineInstruction
verifier happy again.
This is related to <rdar://problem/18027157>.
llvm-svn: 216040
 |