Commit Graph

11451 Commits

Author SHA1 Message Date
Eli Bendersky a108a65df2 Add an optimization that does CSE in a group of similar GEPs.
This optimization merges the common part of a group of GEPs, so we can compute
each pointer address by adding a simple offset to the common part.

The optimization is currently only enabled for the NVPTX backend, where it has
a large payoff on some benchmarks.

Review: http://reviews.llvm.org/D3462

Patch by Jingyue Wu.

llvm-svn: 207783
2014-05-01 18:38:36 +00:00
Chandler Carruth 18c2fbb143 Revert r205965, which essentially reverts r205018 for the second time.
=[

Turns out that this was the root cause of PR19621. We found a crasher
only recently (likely due to improvements elsewhere in the SLP
vectorizer) but the reduced test case failed all the way back to here.
I've confirmed that reverting this patch both fixes the reduced test
case in PR19621 and the actual source file that led to it, so it seems
to really be rooted here. I've replied to the commit thread with
discussion of my (feeble) attempts to debug this. Didn't make it very
far, so reverting now that we have a good test case so that things can
get back to healthy while the debugging carries on.

llvm-svn: 207746
2014-05-01 11:24:11 +00:00
Gerolf Hoflehner 3282af13d4 Patch for function cloning to inline all blocks whose address is taken
Not all address taken blocks get inlined. The reason is
that a blocks new address is known only when it is cloned. But e.g.
a branch instruction in a different block could need that address earlier
while it gets cloned. The solution is to collect the set of all
blocks that can potentially get inlined and compute a new block address
up front. Then clone and cleanup.

rdar://16427209

llvm-svn: 207713
2014-04-30 22:05:02 +00:00
Yi Jiang e2d5f29c2f Revert r207571 - Add slp vectorization to LTO passes
llvm-svn: 207693
2014-04-30 19:27:24 +00:00
Carlo Kok 307625c974 [IPO/MergeFunctions] changes so it doesn't try to bitcast a struct return type but instead recreates it with insert/extract value.
llvm-svn: 207679
2014-04-30 17:53:04 +00:00
Benjamin Kramer bf2368d94b Add a <tuple> include to more files that aren't getting it transitively on MSVC.
llvm-svn: 207617
2014-04-30 07:21:01 +00:00
NAKAMURA Takumi 99aa6e156a ConstantHoisting.cpp: Add <tuple> for std::tie, since r207593 removed FileSystem.h, it includes <tuple>.
llvm-svn: 207614
2014-04-30 06:44:50 +00:00
Jim Grosbach 708f80f783 Tidy up.
llvm-svn: 207585
2014-04-29 22:41:58 +00:00
Jim Grosbach 4a7d496059 Spelling.
llvm-svn: 207584
2014-04-29 22:41:55 +00:00
Rafael Espindola 85f3610222 Also handle ConstantAggregateZero when optimizing vpermilvar*.
llvm-svn: 207582
2014-04-29 22:20:40 +00:00
Rafael Espindola 152ee213a4 Remove tabs.
Sorry, new machine and I forgot to change the editor setting.

llvm-svn: 207578
2014-04-29 21:02:37 +00:00
Rafael Espindola eb7bdbd0ce Two fixes to the vpermilvar optimization.
The instcomine logic to handle vpermilvar's pd and 256 variants was incorrect.
The _256 variants have indexes into the individual 128 bit lanes and in all
cases it also has to mask out unused bits.

llvm-svn: 207577
2014-04-29 20:41:54 +00:00
Diego Novillo cd64780d18 Fix vectorization remarks.
This patch changes the vectorization remarks to also inform when
vectorization is possible but not beneficial.

Added tests to exercise some loop remarks.

llvm-svn: 207574
2014-04-29 20:06:10 +00:00
Yi Jiang 1a3f18b161 Continue slp vectorization even the BB already has vectorized store radar://16641956
llvm-svn: 207572
2014-04-29 19:37:20 +00:00
Yi Jiang 4e234aa790 Add slp vectorization to LTO passes
llvm-svn: 207571
2014-04-29 19:35:39 +00:00
Adam Nemet deab6f945c Reapply r207271 without the testcase
PR19608 was filed to find a suitable testcase.

llvm-svn: 207569
2014-04-29 18:25:28 +00:00
Diego Novillo 34fc8a7c4c Add optimization remarks to the loop unroller and vectorizer.
Summary:
This calls emitOptimizationRemark from the loop unroller and vectorizer
at the point where they make a positive transformation. For the
vectorizer, it reports vectorization and interleave factors. For the
loop unroller, it reports all the different supported types of
unrolling.

Subscribers: llvm-commits

Differential Revision: http://reviews.llvm.org/D3456

llvm-svn: 207528
2014-04-29 14:27:31 +00:00
Zinovy Nis 487268574a [BUG] Fix -Wunused-variable warning in Release mode. Thnx to Kostya Serebryany for pointing.
llvm-svn: 207516
2014-04-29 09:45:08 +00:00
Kostya Serebryany dc8e551d84 fix -Wunused-variable warning in Release mode
llvm-svn: 207514
2014-04-29 09:33:02 +00:00
Zinovy Nis d373fec199 [OPENMP][LV][D3423] Respect Hints.Force meta-data for loops in LoopVectorizer
llvm-svn: 207512
2014-04-29 08:55:11 +00:00
Michael Zolotukhin a93fec040a Fix a typo in comment
llvm-svn: 207499
2014-04-29 07:35:33 +00:00
Chandler Carruth c71b2c3c7f Revert r207271 for now. This commit introduced a test case that ran
clang directly from the LLVM test suite! That doesn't work. I've
followed up on the review thread to try and get a viable solution sorted
out, but trying to get the tree clean here.

llvm-svn: 207462
2014-04-28 23:07:49 +00:00
Hans Wennborg e36e116826 InstCombine: don't drop 'inalloca' in PromoteCastOfAllocation (PR19569)
llvm-svn: 207426
2014-04-28 17:40:03 +00:00
Chandler Carruth 5bdf72cef6 Fix rampant quadratic behavior in UpdatePHINodes. The operation of
mapping from a basic block to an incoming value, either for removal or
just lookup, is linear in the number of predecessors, and we were doing
this for every entry in the 'Preds' list which is in many cases almost
all of them!

Unfortunately, the fixes are quite ugly. PHI nodes just don't make this
operation easy. The efficient way to fix this is to have a clever
'remove_if' operation on PHI nodes that lets us do a single pass over
all the incoming values of the original PHI node, extracting the ones we
care about. Then we could quickly construct the new phi node from this
list. This would remove the remaining underlying quadratic movement of
unrelated incoming values and the need for silly backwards looping to
"minimize" how often we hit the quadratic case.

This is the last obvious fix for PR19499. It shaves another 20% off the
compile time for me, and while UpdatePHINodes remains in the profile,
most of the time is now stemming from the well known inefficiencies of
LVI and jump threading.

llvm-svn: 207409
2014-04-28 10:37:30 +00:00
Craig Topper e73658ddbb [C++] Use 'nullptr'.
llvm-svn: 207394
2014-04-28 04:05:08 +00:00
Gerolf Hoflehner af7a87d2e3 RecursivelyDeleteTriviallyDeadInstructions() could remove
more than 1 instruction. The caller need to be aware of this
and adjust instruction iterators accordingly.

rdar://16679376

Repaired r207302.

llvm-svn: 207309
2014-04-26 05:58:11 +00:00
Gerolf Hoflehner 1da7cbd584 Restore CloneFunction.cpp which got accidently
overwritten by previous backout of r207303

llvm-svn: 207308
2014-04-26 05:43:41 +00:00
Gerolf Hoflehner c46e9b0423 Revert commit r207302 since build failures
have been reported.

llvm-svn: 207303
2014-04-26 02:03:17 +00:00
Gerolf Hoflehner 34210108b3 RecursivelyDeleteTriviallyDeadInstructions() could remove
more than 1 instruction. The caller need to be aware of this
and adjust instruction iterators accordingly.

rdar://16679376

llvm-svn: 207302
2014-04-26 01:19:16 +00:00
Andrea Di Biagio 8cc9059ce8 [InstCombine][X86] Teach how to fold calls to SSE2/AVX2 packed logical shift
right intrinsics.

A packed logical shift right with a shift count bigger than or equal to the
element size always produces a zero vector. In all other cases, it can be
safely replaced by a 'lshr' instruction.

llvm-svn: 207299
2014-04-26 01:03:22 +00:00
Adrian Prantl 232897feaa Unbreak the gdb buildbot by not lowering dbg.declare intrinsics for arrays.
llvm-svn: 207284
2014-04-25 23:00:25 +00:00
Adam Nemet 03d91c51e4 [LoopStrengthReduce] Don't trim formula that uses a subset of required registers
Consider this use from the new testcase:

  LSR Use: Kind=ICmpZero, Offsets={0}, widest fixup type: i32
    reg({1000,+,-1}<nw><%for.body>)
    -3003 + reg({3,+,3}<nw><%for.body>)
    -1001 + reg({1,+,1}<nuw><nsw><%for.body>)
    -1000 + reg({0,+,1}<nw><%for.body>)
    -3000 + reg({0,+,3}<nuw><%for.body>)
    reg({-1000,+,1}<nw><%for.body>)
    reg({-3000,+,3}<nsw><%for.body>)

This is the last use we consider for a solution in SolveRecurse, so CurRegs is
a large set.  (CurRegs is the set of registers that are needed by the
previously visited uses in the in-progress solution.)

ReqRegs is {
  {3,+,3}<nw><%for.body>,
  {1,+,1}<nuw><nsw><%for.body>
}

This is the intersection of the regs used by any of the formulas for the
current use and CurRegs.

Now, the code requires a formula to contain *all* these regs (the comment is
simply wrong), otherwise the formula is immediately disqualified.  Obviously,
no formula for this use contains two regs so they will all get disqualified.

The fix modifies the check to allow the formula in this case.  The idea is
that neither of these formulae is introducing any new registers which is the
point of this early pruning as far as I understand.

In terms of set arithmetic, we now allow formulas whose used regs are a subset
of the required regs not just the other way around.

There are few more loops in the test-suite that are now successfully LSRed.  I
have benchmarked those and found very minimal change.

Fixes <rdar://problem/13965777>

llvm-svn: 207271
2014-04-25 21:02:21 +00:00
Adrian Prantl 32da88923a This reapplies r207235 with an additional bugfixes caught by the msan
buildbot - do not insert debug intrinsics before phi nodes.

Debug info for optimized code: Support variables that are on the stack and
described by DBG_VALUEs during their lifetime.

Previously, when a variable was at a FrameIndex for any part of its
lifetime, this would shadow all other DBG_VALUEs and only a single
fbreg location would be emitted, which in fact is only valid for a small
range and not the entire lexical scope of the variable. The included
dbg-value-const-byref testcase demonstrates this.

This patch fixes this by
Local
- emitting dbg.value intrinsics for allocas that are passed by reference
- dropping all dbg.declares (they are now fully lowered to dbg.values)
SelectionDAG
- renamed constructors for SDDbgValue for better readability.
- fix UserValue::match() to handle indirect values correctly
- not inserting an MMI table entries for dbg.values that describe allocas.
- lowering dbg.values that describe allocas into *indirect* DBG_VALUEs.
CodeGenPrepare
- leaving dbg.values for an alloca were they are (see comment)
Other
- regenerated/updated instcombine.ll testcase and included source

rdar://problem/16679879
http://reviews.llvm.org/D3374

llvm-svn: 207269
2014-04-25 20:49:25 +00:00
Duncan P. N. Exon Smith d2b2facb07 SCC: Change clients to use const, NFC
It's fishy to be changing the `std::vector<>` owned by the iterator, and
no one actual does it, so I'm going to remove the ability in a
subsequent commit.  First, update the users.

<rdar://problem/14292693>

llvm-svn: 207252
2014-04-25 18:24:50 +00:00
Adrian Prantl d2d9b76e48 Revert "This reapplies r207130 with an additional testcase+and a missing check for"
This reverts commit 207235 to investigate msan buildbot breakage.

llvm-svn: 207250
2014-04-25 18:18:09 +00:00
Manman Ren 3c44067a30 [inline cold threshold] Command line argument for inline threshold will
override the default cold threshold.

When we use command line argument to set the inline threshold, the default
cold threshold will not be used. This is in line with how we use
OptSizeThreshold. When we want a higher threshold for all functions, we
do not have to set both inline threshold and cold threshold.

llvm-svn: 207245
2014-04-25 17:34:55 +00:00
Adrian Prantl 0840a22452 Reapply r207135 without modifications.
Debug info: Let dbg.values inserted by LowerDbgDeclare inherit the location
of the dbg.value. This gets rid of tons of redundant variable DIEs in
subscopes.

rdar://problem/14874886, rdar://problem/16679936

llvm-svn: 207236
2014-04-25 17:01:04 +00:00
Adrian Prantl f5834a4b49 This reapplies r207130 with an additional testcase+and a missing check for
AllocaInst that was missing in one location.
Debug info for optimized code: Support variables that are on the stack and
described by DBG_VALUEs during their lifetime.

Previously, when a variable was at a FrameIndex for any part of its
lifetime, this would shadow all other DBG_VALUEs and only a single
fbreg location would be emitted, which in fact is only valid for a small
range and not the entire lexical scope of the variable. The included
dbg-value-const-byref testcase demonstrates this.

This patch fixes this by
Local
- emitting dbg.value intrinsics for allocas that are passed by reference
- dropping all dbg.declares (they are now fully lowered to dbg.values)
SelectionDAG
- renamed constructors for SDDbgValue for better readability.
- fix UserValue::match() to handle indirect values correctly
- not inserting an MMI table entries for dbg.values that describe allocas.
- lowering dbg.values that describe allocas into *indirect* DBG_VALUEs.
CodeGenPrepare
- leaving dbg.values for an alloca were they are (see comment)
Other
- regenerated/updated instcombine.ll testcase and included source

rdar://problem/16679879
http://reviews.llvm.org/D3374

llvm-svn: 207235
2014-04-25 17:01:00 +00:00
Craig Topper f40110f4d8 [C++] Use 'nullptr'. Transforms edition.
llvm-svn: 207196
2014-04-25 05:29:35 +00:00
Karthik Bhat 6a48f7d66e Allow vectorization of bit intrinsics in BB Vectorizer.
This patch adds support for vectorization of  bit intrinsics such as bswap,ctpop,ctlz,cttz.

llvm-svn: 207174
2014-04-25 03:33:48 +00:00
Adrian Prantl 6e5de2ea06 Revert "This reapplies r207130 with an additional testcase+and a missing check for"
Typo in testcase.

llvm-svn: 207166
2014-04-25 00:42:50 +00:00
Adrian Prantl 3512190ab3 This reapplies r207130 with an additional testcase+and a missing check for
AllocaInst that was missing in one location.
Debug info for optimized code: Support variables that are on the stack and
described by DBG_VALUEs during their lifetime.

Previously, when a variable was at a FrameIndex for any part of its
lifetime, this would shadow all other DBG_VALUEs and only a single
fbreg location would be emitted, which in fact is only valid for a small
range and not the entire lexical scope of the variable. The included
dbg-value-const-byref testcase demonstrates this.

This patch fixes this by
Local
- emitting dbg.value intrinsics for allocas that are passed by reference
- dropping all dbg.declares (they are now fully lowered to dbg.values)
SelectionDAG
- renamed constructors for SDDbgValue for better readability.
- fix UserValue::match() to handle indirect values correctly
- not inserting an MMI table entries for dbg.values that describe allocas.
- lowering dbg.values that describe allocas into *indirect* DBG_VALUEs.
CodeGenPrepare
- leaving dbg.values for an alloca were they are (see comment)
Other
- regenerated/updated instcombine.ll testcase and included source

rdar://problem/16679879
http://reviews.llvm.org/D3374

llvm-svn: 207165
2014-04-25 00:38:40 +00:00
Adrian Prantl ff4282a204 Revert "Debug info for optimized code: Support variables that are on the stack and"
This reverts commit 207130 for buildbot breakage.

llvm-svn: 207162
2014-04-25 00:04:49 +00:00
Adrian Prantl 5ad11841f7 Revert "Debug info: Let dbg.values inserted by LowerDbgDeclare inherit the location"
This reverts commit 207130 for buildbot breakage.

llvm-svn: 207159
2014-04-24 23:53:29 +00:00
Adrian Prantl f4a701092e Debug info: Let dbg.values inserted by LowerDbgDeclare inherit the location
of the dbg.value. This gets rid of tons of redundant variable DIEs in
subscopes.

rdar://problem/14874886, rdar://problem/16679936

llvm-svn: 207135
2014-04-24 18:44:15 +00:00
Adrian Prantl f4223918de Debug info for optimized code: Support variables that are on the stack and
described by DBG_VALUEs during their lifetime.

Previously, when a variable was at a FrameIndex for any part of its
lifetime, this would shadow all other DBG_VALUEs and only a single
fbreg location would be emitted, which in fact is only valid for a small
range and not the entire lexical scope of the variable. The included
dbg-value-const-byref testcase demonstrates this.

This patch fixes this by
Local
- emitting dbg.value intrinsics for allocas that are passed by reference
- dropping all dbg.declares (they are now fully lowered to dbg.values)
SelectionDAG
- renamed constructors for SDDbgValue for better readability.
- fix UserValue::match() to handle indirect values correctly
- not inserting an MMI table entries for dbg.values that describe allocas.
- lowering dbg.values that describe allocas into *indirect* DBG_VALUEs.
CodeGenPrepare
- leaving dbg.values for an alloca were they are (see comment)
Other
- regenerated/updated instcombine-intrinsics testcase and included source


rdar://problem/16679879
http://reviews.llvm.org/D3374

llvm-svn: 207130
2014-04-24 17:41:45 +00:00
Karthik Bhat 81e6bf0a41 Allow vectorization of few missed llvm intrinsic calls in BBVectorizor by handling them in isVectorizableIntrinsic function.
llvm-svn: 207085
2014-04-24 07:29:55 +00:00
Michael J. Spencer dee4b2c379 [InstCombine][x86] Constant fold psll intrinsics.
This excludes avx512 as I don't have hardware to verify. It excludes _dq
variants because they are represented in the IR as <{2,4} x i64> when it's
actually a byte shift of the entire i{128,265}.

This also excludes _dq_bs as they aren't at all supported by the backend.
There are also no corresponding instructions in the ISA. I have no idea why
they exist...

llvm-svn: 207058
2014-04-24 00:58:18 +00:00
Filipe Cabecinhas 1a80595a2b Optimize some special cases for SSE4a insertqi
Summary:
Since the upper 64 bits of the destination register are undefined when
performing this operation, we can substitute it and let the optimizer
figure out that only a copy is needed.

Also added range merging, if an instruction copies a range that can be
merged with a previous copied range.

Added test cases for both optimizations.

Reviewers: grosbach, nadav

CC: llvm-commits

Differential Revision: http://reviews.llvm.org/D3357

llvm-svn: 207055
2014-04-24 00:38:14 +00:00
Matt Arsenault 60728177fb Handle addrspacecast when looking at memcpys from globals
llvm-svn: 207054
2014-04-24 00:01:09 +00:00
Matt Arsenault be55888849 Remove more default address space argument usage.
These places are inconsequential in practice.

llvm-svn: 207021
2014-04-23 20:58:57 +00:00
Matt Arsenault fcd7401bbf Don't use default address space arguments in GlobalOpt
llvm-svn: 207019
2014-04-23 20:36:10 +00:00
Alexander Potapenko a51e483846 [ASan] Move the shadow range on 32-bit iOS (and iOS Simulator)
to 0x40000000-0x60000000 to avoid address space clash with system libraries.
The solution has been proposed by tahabekireren@gmail.com in https://code.google.com/p/address-sanitizer/issues/detail?id=210
This is also known to fix some Chromium iOS tests.

llvm-svn: 207002
2014-04-23 17:14:45 +00:00
Matt Arsenault 6b4bed4b83 Remove dead code in instcombine.
Don't replace shifts greater than the type with the maximum shift.

This isn't hit anywhere in the tests, and somewhere else is replacing
these with undef.

llvm-svn: 207000
2014-04-23 16:48:40 +00:00
Evgeniy Stepanov 119cb2eed5 Fix handling of missing DataLayout in sanitizers.
Pass::doInitialization is supposed to return False when it did not
change the program, not when a fatal error occurs.

llvm-svn: 206975
2014-04-23 12:51:32 +00:00
Alexander Musman f0785f4db4 [LV] Statistics numbers for LoopVectorize introduced: a number of analyzed loops & a number of vectorized loops.
Use -stats to see how many loops were analyzed for possible vectorization and how many of them were actually vectorized.
Patch by Zinovy Nis

Differential Revision: http://reviews.llvm.org/D3438

llvm-svn: 206956
2014-04-23 08:40:37 +00:00
Juergen Ributzka 575bcb770a [Constant Hoisting] Materialize the constant before the cloned cast instruction.
In the case where the constant comes from a cloned cast instruction, the
materialization code has to go before the cloned cast instruction.

This commit fixes the method that finds the materialization insertion point
by making it aware of this case.

This fixes <rdar://problem/15532441>

llvm-svn: 206913
2014-04-22 18:06:58 +00:00
Juergen Ributzka a1444b39fb [Constant Hoisting] Print the instructions in the correct order for debugging. No functional change.
llvm-svn: 206912
2014-04-22 18:06:51 +00:00
Kostya Serebryany c9a2c17ad3 [asan] Support outline instrumentation for wide types and delete dead code, patch by Yuri Gribov
llvm-svn: 206883
2014-04-22 11:19:45 +00:00
Chandler Carruth 964daaaf19 [Modules] Fix potential ODR violations by sinking the DEBUG_TYPE
definition below all of the header #include lines, lib/Transforms/...
edition.

This one is tricky for two reasons. We again have a couple of passes
that define something else before the includes as well. I've sunk their
name macros with the DEBUG_TYPE.

Also, InstCombine contains headers that need DEBUG_TYPE, so now those
headers #define and #undef DEBUG_TYPE around their code, leaving them
well formed modular headers. Fixing these headers was a large motivation
for all of these changes, as "leaky" macros of this form are hard on the
modules implementation.

llvm-svn: 206844
2014-04-22 02:55:47 +00:00
Chandler Carruth e96dd8975f [Modules] Make Support/Debug.h modular. This requires it to not change
behavior based on other files defining DEBUG_TYPE, which means it cannot
define DEBUG_TYPE at all. This is actually better IMO as it forces folks
to define relevant DEBUG_TYPEs for their files. However, it requires all
files that currently use DEBUG(...) to define a DEBUG_TYPE if they don't
already. I've updated all such files in LLVM and will do the same for
other upstream projects.

This still leaves one important change in how LLVM uses the DEBUG_TYPE
macro going forward: we need to only define the macro *after* header
files have been #include-ed. Previously, this wasn't possible because
Debug.h required the macro to be pre-defined. This commit removes that.
By defining DEBUG_TYPE after the includes two things are fixed:

- Header files that need to provide a DEBUG_TYPE for some inline code
  can do so by defining the macro before their inline code and undef-ing
  it afterward so the macro does not escape.

- We no longer have rampant ODR violations due to including headers with
  different DEBUG_TYPE definitions. This may be mostly an academic
  violation today, but with modules these types of violations are easy
  to check for and potentially very relevant.

Where necessary to suppor headers with DEBUG_TYPE, I have moved the
definitions below the includes in this commit. I plan to move the rest
of the DEBUG_TYPE macros in LLVM in subsequent commits; this one is big
enough.

The comments in Debug.h, which were hilariously out of date already,
have been updated to reflect the recommended practice going forward.

llvm-svn: 206822
2014-04-21 22:55:11 +00:00
Rafael Espindola bad3f77703 Simplify a vpermil* with constant mask.
With a constant mask a vpermil* is just a shufflevector. This patch implements
that simplification. This allows us to produce denser code. It should also
allow more folding down the line.

llvm-svn: 206801
2014-04-21 22:06:04 +00:00
David Blaikie bc44220eb8 Use unique_ptr to handle GlobalOpt's Evaluator members
llvm-svn: 206790
2014-04-21 20:49:36 +00:00
Reid Kleckner 9b2cc647eb Fix PR7272 in -tailcallelim instead of the inliner
The -tailcallelim pass should be checking if byval or inalloca args can
be captured before marking calls as tail calls.  This was the real root
cause of PR7272.

With a better fix in place, revert the inliner change from r105255.  The
test case it introduced still passes and has been moved to
test/Transforms/Inline/byval-tail-call.ll.

Reviewers: chandlerc

Differential Revision: http://reviews.llvm.org/D3403

llvm-svn: 206789
2014-04-21 20:48:47 +00:00
David Blaikie eb038915ab Simplify expression that was explicitly naming an operator overload in a call.
llvm-svn: 206788
2014-04-21 20:43:51 +00:00
David Blaikie 229de508bb Use unique_ptr to handle ownership of GCOVFunctions in GCOVProfiler.
llvm-svn: 206786
2014-04-21 20:41:55 +00:00
Chandler Carruth 5f1f26e891 [Modules] Sink all the DEBUG_TYPE defines for InstCombine out of the
header files and into the cpp files.

These files will require more touches as the header files actually use
DEBUG(). Eventually, I'll have to introduce a matched #define and #undef
of DEBUG_TYPE for the header files, but that comes as step N of many to
clean all of this up.

llvm-svn: 206777
2014-04-21 19:51:41 +00:00
Evgeniy Stepanov 3939f54820 [msan] Enable out-of-line instrumentation for large functions by default.
llvm-svn: 206759
2014-04-21 15:04:05 +00:00
Kostya Serebryany 35e538312a [asan] add a run-time flag detect_container_overflow=true/false
llvm-svn: 206756
2014-04-21 14:35:00 +00:00
Kostya Serebryany 94c81ca8e1 [asan] instead of inserting inline instrumentation around memset/memcpy/memmove, replace the intrinsic with __asan_memset/etc. This makes the memset/etc handling more complete and consistent with what we do in msan. It may slowdown some cases (when the intrinsic was actually inlined) and speedup other cases (when it was not inlined)
llvm-svn: 206746
2014-04-21 11:50:42 +00:00
Kostya Serebryany 94f57d199a [asan] temporary disable generating __asan_loadN/__asan_storeN
llvm-svn: 206741
2014-04-21 10:28:13 +00:00
Kostya Serebryany 86332c010f [asan] insert __asan_loadN/__asan_storeN as out-lined asan checks, llvm part
llvm-svn: 206734
2014-04-21 07:10:43 +00:00
Alp Toker 9844434151 Remove some empty statements
Cleanup only.

llvm-svn: 206710
2014-04-19 23:56:35 +00:00
Nick Lewycky 05e0f1ccff Check whether functions have any lines associated before emitting coverage info for them. This isn't just a size/time saving, gcov may crash on these.
llvm-svn: 206671
2014-04-18 23:32:28 +00:00
Evgeniy Stepanov 65120ec8c6 [msan] Add -msan-instrumentation-with-call-threshold.
This flag replaces inline instrumentation for checks and origin stores with
calls into MSan runtime library. This is a workaround for PR17409.

Disabled by default.

llvm-svn: 206585
2014-04-18 12:17:20 +00:00
Kostya Serebryany 22e8810838 [asan] one more workaround for PR17409: don't do BB-level coverage instrumentation if there are more than N (=1500) basic blocks. This makes ASanCoverage work on libjpeg_turbo/jchuff.c used by Chrome, which has 1824 BBs
llvm-svn: 206564
2014-04-18 08:02:42 +00:00
Duncan P. N. Exon Smith 49f3ec80c2 PMBuilder: Expose an option to disable tail calls
Adds API to allow frontends to disable tail calls in PassManagerBuilder.

<rdar://problem/16050591>

llvm-svn: 206542
2014-04-18 01:05:15 +00:00
Diego Novillo 0915c047c2 Fix bug 19437 - Only add discriminators for DWARF 4 and above.
Summary:
This prevents the discriminator generation pass from triggering if
the DWARF version being used in the module is prior to 4.

Reviewers: echristo, dblaikie

CC: llvm-commits

Differential Revision: http://reviews.llvm.org/D3413

llvm-svn: 206507
2014-04-17 22:33:50 +00:00
Nuno Lopes 9ced19abe8 remove some dead code
lib/Analysis/IPA/InlineCost.cpp         |   18 ------------------
 lib/Analysis/RegionPass.cpp             |    1 -
 lib/Analysis/TypeBasedAliasAnalysis.cpp |    1 -
 lib/Transforms/Scalar/LoopUnswitch.cpp  |   21 ---------------------
 lib/Transforms/Utils/LCSSA.cpp          |    2 --
 lib/Transforms/Utils/LoopSimplify.cpp   |    6 ------
 utils/TableGen/AsmWriterEmitter.cpp     |   13 -------------
 utils/TableGen/DFAPacketizerEmitter.cpp |    7 -------
 utils/TableGen/IntrinsicEmitter.cpp     |    2 --
 9 files changed, 71 deletions(-)

llvm-svn: 206506
2014-04-17 22:26:44 +00:00
NAKAMURA Takumi cd1fc4bc1b Inliner::OptimizationRemark: Fix crash in clang/test/Frontend/optimization-remark.c on some hosts, including --vg.
DebugLoc in Callsite would not live after Inliner. It should be copied before Inliner.

llvm-svn: 206459
2014-04-17 12:22:14 +00:00
Kostya Serebryany 0c02d26d6b [asan] add two new hidden compile-time flags for asan: asan-instrumentation-with-call-threshold and asan-memory-access-callback-prefix. This is part of the workaround for PR17409 (instrument huge functions with callbacks instead of inlined code). These flags will also help us experiment with kasan (kernel-asan) and clang
llvm-svn: 206383
2014-04-16 12:12:19 +00:00
Julien Lerouge be4fe32eb8 Add lifetime markers for allocas created to hold byval arguments, make them
appear in the InlineFunctionInfo.

llvm-svn: 206308
2014-04-15 18:06:46 +00:00
Julien Lerouge 957e91c4d8 Split byval argument initialization so the memcpy(s) are injected at the
beginning of the first new block after inlining.

llvm-svn: 206307
2014-04-15 18:01:54 +00:00
Duncan P. N. Exon Smith 2b69189c9c LTO: Add more loop simplification passes to LTO
Similar to r202051, add missing loop simplification passes to the LTO
optimization pipeline.

Patch by Rafael Espindola.

llvm-svn: 206306
2014-04-15 17:48:15 +00:00
Duncan P. N. Exon Smith 6ef5f284d6 verify-di: Implement DebugInfoVerifier
Implement DebugInfoVerifier, which steals verification relying on
DebugInfoFinder from Verifier.

  - Adds LegacyDebugInfoVerifierPassPass, a ModulePass which wraps
    DebugInfoVerifier.  Uses -verify-di command-line flag.

  - Change verifyModule() to invoke DebugInfoVerifier as well as
    Verifier.

  - Add a call to createDebugInfoVerifierPass() wherever there was a
    call to createVerifierPass().

This implementation as a module pass should sidestep efficiency issues,
allowing us to turn debug info verification back on.

<rdar://problem/15500563>

llvm-svn: 206300
2014-04-15 16:27:38 +00:00
Alexey Bataev b97f9e8698 D3348 - [BUG] "Rotate Loop" pass kills "llvm.vectorizer.enable" metadata
llvm-svn: 206266
2014-04-15 09:37:30 +00:00
Matt Arsenault fed3dc8dc6 Revert "Revert r206045, "Fix shift by constants for vector.""
Fix cases where the Value itself is used, and not the constant value.

llvm-svn: 206214
2014-04-14 21:50:37 +00:00
NAKAMURA Takumi 58ad0c87f8 Whitespace.
llvm-svn: 206154
2014-04-14 07:03:13 +00:00
NAKAMURA Takumi 26afa982ec Revert r206045, "Fix shift by constants for vector."
It broke some builders, at least, i686.

llvm-svn: 206153
2014-04-14 07:02:57 +00:00
Serge Pavlov b5f3ddc7a1 Use APInt arithmetic, fixed typo. Thanks to Benjamin Kramer for noticing that.
llvm-svn: 206144
2014-04-14 02:20:19 +00:00
Serge Pavlov 4bb54d51c8 Recognize test for overflow in integer multiplication.
If multiplication involves zero-extended arguments and the result is
compared as in the patterns:

    %mul32 = trunc i64 %mul64 to i32
    %zext = zext i32 %mul32 to i64
    %overflow = icmp ne i64 %mul64, %zext
or
    %overflow = icmp ugt i64 %mul64 , 0xffffffff

then the multiplication may be replaced by call to umul.with.overflow.
This change fixes PR4917 and PR4918.

Differential Revision: http://llvm-reviews.chandlerc.com/D2814

llvm-svn: 206137
2014-04-13 18:23:41 +00:00
Matt Arsenault 173a1e577c Fix shift by constants for vector.
ashr <N x iM>, <N x iM> M -> undef

llvm-svn: 206045
2014-04-11 17:57:53 +00:00
David Blaikie ceec2bdaa5 Implement depth_first and inverse_depth_first range factory functions.
Also updated as many loops as I could find using df_begin/idf_begin -
strangely I found no uses of idf_begin. Is that just used out of tree?

Also a few places couldn't use df_begin because either they used the
member functions of the depth first iterators or had specific ordering
constraints (I added a comment in the latter case).

Based on a patch by Jim Grosbach. (Jim - you just had iterator_range<T>
where you needed iterator_range<idf_iterator<T>>)

llvm-svn: 206016
2014-04-11 01:50:01 +00:00
Arnold Schwaighofer b373e01d87 Reapply "SLPVectorizer: Ignore users that are insertelements we can reschedule them"
This commit reapplies 205018. After 205855 we should correctly vectorize
intrinsics.

llvm-svn: 205965
2014-04-10 13:41:35 +00:00
Alp Toker 16f98b255d Fix some doc and comment typos
llvm-svn: 205899
2014-04-09 14:47:27 +00:00
Arnold Schwaighofer fd0bf5d6e5 SLPVectorizer: Only vectorize intrinsics whose operands are widened equally
The vectorizer only knows how to vectorize intrinics by widening all operands by
the same factor.

Patch by Tyler Nowicki!

llvm-svn: 205855
2014-04-09 14:20:47 +00:00
Diego Novillo a9298b2297 Add support for optimization reports.
Summary:
This patch adds backend support for -Rpass=, which indicates the name
of the optimization pass that should emit remarks stating when it
made a transformation to the code.

Pass names are taken from their DEBUG_NAME definitions.

When emitting an optimization report diagnostic, the lack of debug
information causes the diagnostic to use "<unknown>:0:0" as the
location string.

This is the back end counterpart for

http://llvm-reviews.chandlerc.com/D3226

Reviewers: qcolombet

CC: llvm-commits

Differential Revision: http://llvm-reviews.chandlerc.com/D3227

llvm-svn: 205774
2014-04-08 16:42:34 +00:00
Eric Christopher fcd222aa47 Add NDEBUG markers around debug only function.
llvm-svn: 205706
2014-04-07 12:46:30 +00:00
Eric Christopher 6ab6a0648b Add debug location information to the vectorizer debug statements.
Patch by Zinovy Nis.

llvm-svn: 205705
2014-04-07 12:32:17 +00:00
David Blaikie 6425696818 Fixing typo.
Differential Revision: http://reviews.llvm.org/D3154

llvm-svn: 205674
2014-04-05 20:30:31 +00:00