Artur Pilipenko [Mon, 17 Aug 2015 15:54:26 +0000 (15:54 +0000)]
Take alignment into account in isSafeToSpeculativelyExecute and isSafeToLoadUnconditionally.
Reviewed By: hfinkel, sanjoy, MatzeB
Differential Revision: http://reviews.llvm.org/D9791
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245223
91177308-0d34-0410-b5e6-
96231b3b80d8
Benjamin Kramer [Mon, 17 Aug 2015 14:35:25 +0000 (14:35 +0000)]
Extend MCAsmLexer so that it can peek forward several tokens
This commit adds a virtual `peekTokens()` function to `MCAsmLexer`
which can peek forward an arbitrary number of tokens.
It also makes the `peekTok()` method call `peekTokens()` method, but
only requesting one token.
The idea is to better support targets which more more ambiguous
assembly syntaxes.
Patch by Dylan McKay!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245221
91177308-0d34-0410-b5e6-
96231b3b80d8
Aaron Ballman [Mon, 17 Aug 2015 14:25:57 +0000 (14:25 +0000)]
Correcting a -Woverflow warning where 0xFFFF was overflowing an implicit constant conversion.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245220
91177308-0d34-0410-b5e6-
96231b3b80d8
Joseph Tremoulet [Mon, 17 Aug 2015 13:51:37 +0000 (13:51 +0000)]
[WinEHPrepare] Fix catchret successor phi demotion
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
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245218
91177308-0d34-0410-b5e6-
96231b3b80d8
Tobias Grosser [Mon, 17 Aug 2015 10:58:03 +0000 (10:58 +0000)]
Revert "Disable targetdatalayoutcheck"
I committed by accident a local hack that should not have made it upstream.
Sorry for the noise.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245212
91177308-0d34-0410-b5e6-
96231b3b80d8
Tobias Grosser [Mon, 17 Aug 2015 10:56:35 +0000 (10:56 +0000)]
Disable targetdatalayoutcheck
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245210
91177308-0d34-0410-b5e6-
96231b3b80d8
Daniel Sanders [Mon, 17 Aug 2015 10:11:55 +0000 (10:11 +0000)]
[mips] [IAS] Add support for the DLA pseudo-instruction and fix problems with DLI
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
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245208
91177308-0d34-0410-b5e6-
96231b3b80d8
Michael Kuperstein [Mon, 17 Aug 2015 10:06:08 +0000 (10:06 +0000)]
[GMR] isNonEscapingGlobalNoAlias() should look through Bitcasts/GEPs when looking at loads.
This fixes yet another case from PR24288.
Differential Revision: http://reviews.llvm.org/D12064
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245207
91177308-0d34-0410-b5e6-
96231b3b80d8
James Molloy [Mon, 17 Aug 2015 07:13:20 +0000 (07:13 +0000)]
Remove hand-rolled matching for fmin and fmax.
SDAGBuilder now does this all for us.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245198
91177308-0d34-0410-b5e6-
96231b3b80d8
James Molloy [Mon, 17 Aug 2015 07:13:15 +0000 (07:13 +0000)]
Rip out hand-rolled matching code for VMIN, VMAX, VMINNM and VMAXNM
This is no longer needed - SDAGBuilder will do this for us.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245197
91177308-0d34-0410-b5e6-
96231b3b80d8
James Molloy [Mon, 17 Aug 2015 07:13:10 +0000 (07:13 +0000)]
Generate FMINNAN/FMINNUM/FMAXNAN/FMAXNUM from SDAGBuilder.
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.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245196
91177308-0d34-0410-b5e6-
96231b3b80d8
Karthik Bhat [Mon, 17 Aug 2015 05:51:39 +0000 (05:51 +0000)]
Fix PR24469 resulting from r245025 and re-enable dead store elimination across basicblocks.
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.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245195
91177308-0d34-0410-b5e6-
96231b3b80d8
David Majnemer [Mon, 17 Aug 2015 03:11:26 +0000 (03:11 +0000)]
Revert "[InstCombinePHI] Partial simplification of identity operations."
This reverts commit r244887, it caused PR24470.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245194
91177308-0d34-0410-b5e6-
96231b3b80d8
Chandler Carruth [Mon, 17 Aug 2015 02:08:17 +0000 (02:08 +0000)]
[PM] Port ScalarEvolution to the new pass manager.
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
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245193
91177308-0d34-0410-b5e6-
96231b3b80d8
Chandler Carruth [Sun, 16 Aug 2015 23:17:27 +0000 (23:17 +0000)]
[ADT] Teach FoldingSet to be movable.
This is a very minimal move support - it leaves the moved-from object in
a zombie state that is only valid for destruction and move assignment.
This seems fine to me, and leaving it in the default constructed state
would require adding more state to the object and potentially allocating
memory (!!!) and so seems like a Bad Idea.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245192
91177308-0d34-0410-b5e6-
96231b3b80d8
Craig Topper [Sun, 16 Aug 2015 21:27:10 +0000 (21:27 +0000)]
[TableGen] Use range-based for loop.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245191
91177308-0d34-0410-b5e6-
96231b3b80d8
Craig Topper [Sun, 16 Aug 2015 21:27:08 +0000 (21:27 +0000)]
[TableGen] Move the ConversionRow vector into the ConversionTable instead of copying.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245190
91177308-0d34-0410-b5e6-
96231b3b80d8
Benjamin Kramer [Sun, 16 Aug 2015 21:16:37 +0000 (21:16 +0000)]
[SimplifyLibCalls] Drop default template args. No functional change.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245189
91177308-0d34-0410-b5e6-
96231b3b80d8
Benjamin Kramer [Sun, 16 Aug 2015 21:16:26 +0000 (21:16 +0000)]
[IR] Simplify code. No functionality change.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245188
91177308-0d34-0410-b5e6-
96231b3b80d8
Sanjay Patel [Sun, 16 Aug 2015 20:18:19 +0000 (20:18 +0000)]
transform fmin/fmax calls when possible (PR24314)
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
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245187
91177308-0d34-0410-b5e6-
96231b3b80d8
Sanjoy Das [Sun, 16 Aug 2015 18:22:46 +0000 (18:22 +0000)]
[LSR][NFC] Don’t duplicate entity name at the beginning of the comment.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245183
91177308-0d34-0410-b5e6-
96231b3b80d8
Sanjoy Das [Sun, 16 Aug 2015 18:22:43 +0000 (18:22 +0000)]
[LSR][NFC] Use camelCase for method names in Formula and RegUseTracker.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245182
91177308-0d34-0410-b5e6-
96231b3b80d8
Sanjay Patel [Sun, 16 Aug 2015 17:54:28 +0000 (17:54 +0000)]
use SDValue bool operator; NFCI
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245181
91177308-0d34-0410-b5e6-
96231b3b80d8
Yaron Keren [Sun, 16 Aug 2015 07:55:08 +0000 (07:55 +0000)]
Add missing include guard.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245173
91177308-0d34-0410-b5e6-
96231b3b80d8
David Majnemer [Sun, 16 Aug 2015 07:11:59 +0000 (07:11 +0000)]
Revert "Add support for cross block dse. This patch enables dead stroe elimination across basicblocks."
This reverts commit r245025, it caused PR24469.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245172
91177308-0d34-0410-b5e6-
96231b3b80d8
David Majnemer [Sun, 16 Aug 2015 07:09:17 +0000 (07:09 +0000)]
[InstCombine] Replace an and+icmp with a trunc+icmp
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.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245171
91177308-0d34-0410-b5e6-
96231b3b80d8
Chandler Carruth [Sun, 16 Aug 2015 06:35:19 +0000 (06:35 +0000)]
Revert r244127: [PM] Remove a failed attempt to port the CallGraph
analysis ...
It turns out that we *do* need the old CallGraph ported to the new pass
manager. There are times where this model of a call graph is really
superior to the one provided by the LazyCallGraph. For example,
GlobalsModRef very specifically needs the model provided by CallGraph.
While here, I've tried to make the move semantics actually work. =]
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245170
91177308-0d34-0410-b5e6-
96231b3b80d8
David Majnemer [Sun, 16 Aug 2015 04:52:11 +0000 (04:52 +0000)]
[X86] Widen the 'AND' mask if doing so shrinks the encoding size
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
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245169
91177308-0d34-0410-b5e6-
96231b3b80d8
NAKAMURA Takumi [Sun, 16 Aug 2015 02:41:23 +0000 (02:41 +0000)]
MergeFunc: Quick fix for r245140, Ignore second, aka Function*, in sorting.
Don't assume second would be ordered in the module.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245168
91177308-0d34-0410-b5e6-
96231b3b80d8
Yaron Keren [Sat, 15 Aug 2015 19:06:14 +0000 (19:06 +0000)]
Try to appease VS 2015 warnings from reviews.llvm.org/D11890
ByteSize and BitSize should not be size_t but unsigned, considering
1) They are at most 2^16 and 2^19, respectively.
2) BitSize is an argument to Type::getIntNTy which takes unsigned.
Also, use the correct utostr instead itostr and cache the string result.
Thanks to James Touton for reporting this!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245167
91177308-0d34-0410-b5e6-
96231b3b80d8
Sanjay Patel [Sat, 15 Aug 2015 17:01:54 +0000 (17:01 +0000)]
[x86] enable machine combiner reassociations for scalar single-precision minimums
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245166
91177308-0d34-0410-b5e6-
96231b3b80d8
Simon Pilgrim [Sat, 15 Aug 2015 16:54:18 +0000 (16:54 +0000)]
Updated broadcast stack folding test to avoid use of broadcast intrinsics.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245165
91177308-0d34-0410-b5e6-
96231b3b80d8
Sanjay Patel [Sat, 15 Aug 2015 16:53:08 +0000 (16:53 +0000)]
fix typos; NFC
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245164
91177308-0d34-0410-b5e6-
96231b3b80d8
Sanjay Patel [Sat, 15 Aug 2015 16:49:50 +0000 (16:49 +0000)]
add test case to show current codegen
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245163
91177308-0d34-0410-b5e6-
96231b3b80d8
Yaron Keren [Sat, 15 Aug 2015 14:54:43 +0000 (14:54 +0000)]
Silence VS2015 warning.
Patch by James Touton!
http://reviews.llvm.org/D11890
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245161
91177308-0d34-0410-b5e6-
96231b3b80d8
Simon Pilgrim [Sat, 15 Aug 2015 13:27:30 +0000 (13:27 +0000)]
[DAGCombiner] Attempt to mask vectors before zero extension instead of after.
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
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245160
91177308-0d34-0410-b5e6-
96231b3b80d8
Chandler Carruth [Sat, 15 Aug 2015 09:22:21 +0000 (09:22 +0000)]
[PM/AA] Delete the LibCallAliasAnalysis and all the associated
infrastructure.
This AA was never used in tree. It's infrastructure also completely
overlaps that of TargetLibraryInfo which is used heavily by BasicAA to
achieve similar goals to those stated for this analysis.
As has come up in several discussions, the use case here is still really
important, but this code isn't helping move toward that use case. Any
progress on better supporting rich AA information for runtime library
environments would likely be better off starting from scratch or
starting from TargetLibraryInfo than from this base.
Differential Revision: http://reviews.llvm.org/D12028
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245155
91177308-0d34-0410-b5e6-
96231b3b80d8
David Majnemer [Sat, 15 Aug 2015 03:19:29 +0000 (03:19 +0000)]
[IR] Update CreateCatchRet to take a return value
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245152
91177308-0d34-0410-b5e6-
96231b3b80d8
Matt Arsenault [Sat, 15 Aug 2015 02:58:49 +0000 (02:58 +0000)]
AMDGPU/SI: Only look at live out SGPR defs
When trying to fix SGPR live ranges, skip defs that are
killed in the same block as the def. I don't think
we need to worry about these cases as long as the
live ranges of the SGPRs in dominating blocks are
correct.
This reduces the number of elements the second
loop over the function needs to look at, and makes
it generally easier to understand. The second loop
also only considers if the live range is live
in to a block, which logically means it
must have been live out from another.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245150
91177308-0d34-0410-b5e6-
96231b3b80d8
David Majnemer [Sat, 15 Aug 2015 02:46:08 +0000 (02:46 +0000)]
[IR] Give catchret an optional 'return value' operand
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.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245149
91177308-0d34-0410-b5e6-
96231b3b80d8
James Y Knight [Sat, 15 Aug 2015 02:32:35 +0000 (02:32 +0000)]
Remove redundant TargetFrameLowering::getFrameIndexOffset virtual
function.
This was the same as getFrameIndexReference, but without the FrameReg
output.
Differential Revision: http://reviews.llvm.org/D12042
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245148
91177308-0d34-0410-b5e6-
96231b3b80d8
JF Bastien [Sat, 15 Aug 2015 01:23:28 +0000 (01:23 +0000)]
[WebAssembly] Add Relooper
This is just an initial checkin of an implementation of the Relooper algorithm, in preparation for WebAssembly codegen to utilize. It doesn't do anything yet by itself.
The Relooper algorithm takes an arbitrary control flow graph and generates structured control flow from that, utilizing a helper variable when necessary to handle irreducibility. The WebAssembly backend will be able to use this in order to generate an AST for its binary format.
Author: azakai
Reviewers: jfb, sunfish
Subscribers: jevinskie, arsenm, jroelofs, llvm-commits
Differential revision: http://reviews.llvm.org/D11691
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245142
91177308-0d34-0410-b5e6-
96231b3b80d8
JF Bastien [Sat, 15 Aug 2015 01:18:18 +0000 (01:18 +0000)]
Accelerate MergeFunctions with hashing
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
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245140
91177308-0d34-0410-b5e6-
96231b3b80d8
Alex Lorenz [Sat, 15 Aug 2015 01:06:06 +0000 (01:06 +0000)]
MIRLangRef: Describe the syntax that is used to represent machine basic blocks.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245138
91177308-0d34-0410-b5e6-
96231b3b80d8
Matt Arsenault [Sat, 15 Aug 2015 00:53:06 +0000 (00:53 +0000)]
LoopStrengthReduce: Try to pass address space to isLegalAddressingMode
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.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245137
91177308-0d34-0410-b5e6-
96231b3b80d8
Matt Arsenault [Sat, 15 Aug 2015 00:12:39 +0000 (00:12 +0000)]
AMDGPU/SI: Fix printing useless info with amdhsa
The comments at the bottom would all report 0 if
amdhsa was used.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245135
91177308-0d34-0410-b5e6-
96231b3b80d8
Matt Arsenault [Sat, 15 Aug 2015 00:12:37 +0000 (00:12 +0000)]
AMDGPU/SI: Update LiveVariables
This is simple but won't work if/when this pass
is moved to be post-SSA.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245134
91177308-0d34-0410-b5e6-
96231b3b80d8
Matt Arsenault [Sat, 15 Aug 2015 00:12:35 +0000 (00:12 +0000)]
AMDGPU/SI: Update LiveIntervals during SIFixSGPRLiveRanges
Does not mark SlotIndexes as reserved, although I think
that might be OK.
LiveVariables still need to be handled.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245133
91177308-0d34-0410-b5e6-
96231b3b80d8
Matt Arsenault [Sat, 15 Aug 2015 00:12:32 +0000 (00:12 +0000)]
AMDGPU: Remove unnecessary assert
These shouldn't ever be null. The number of successors
was already asserted to be 2.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245132
91177308-0d34-0410-b5e6-
96231b3b80d8
Matt Arsenault [Sat, 15 Aug 2015 00:12:30 +0000 (00:12 +0000)]
AMDGPU/SI: Make comments more precise.
True branch instructions do behave as expected with liveness.
Avoid the phrasing "branch decision is based on a value in an SGPR"
because this could be misleading. A VALU compare instruction's
result is still based on an SGPR, even though that condition
may be divergent.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245131
91177308-0d34-0410-b5e6-
96231b3b80d8
Sanjay Patel [Fri, 14 Aug 2015 23:03:01 +0000 (23:03 +0000)]
make current codegen visible in the checks, so we can decide if it's right
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245120
91177308-0d34-0410-b5e6-
96231b3b80d8
Nick Lewycky [Fri, 14 Aug 2015 22:46:49 +0000 (22:46 +0000)]
Fix a crash where a utility function wasn't aware of fcmp vectors and created a value with the wrong type. Fixes PR24458!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245119
91177308-0d34-0410-b5e6-
96231b3b80d8
Bjarke Hammersholt Roune [Fri, 14 Aug 2015 22:45:26 +0000 (22:45 +0000)]
[SCEV] Apply NSW and NUW flags via poison value analysis for sub, mul and shl
Summary:
http://reviews.llvm.org/D11212 made Scalar Evolution able to propagate NSW and NUW flags from instructions to SCEVs for add instructions. This patch expands that to sub, mul and shl instructions.
This change makes LSR able to generate pointer induction variables for loops like these, where the index is 32 bit and the pointer is 64 bit:
for (int i = 0; i < numIterations; ++i)
sum += ptr[i - offset];
for (int i = 0; i < numIterations; ++i)
sum += ptr[i * stride];
for (int i = 0; i < numIterations; ++i)
sum += ptr[3 * (i << 7)];
Reviewers: atrick, sanjoy
Subscribers: sanjoy, majnemer, hfinkel, llvm-commits, meheff, jingyue, eliben
Differential Revision: http://reviews.llvm.org/D11860
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245118
91177308-0d34-0410-b5e6-
96231b3b80d8
Pat Gavlin [Fri, 14 Aug 2015 22:41:43 +0000 (22:41 +0000)]
Add a target environment for CoreCLR.
Although targeting CoreCLR is similar to targeting MSVC, there are
certain important differences that the backend must be aware of
(e.g. differences in stack probes, EH, and library calls).
Differential Revision: http://reviews.llvm.org/D11012
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245115
91177308-0d34-0410-b5e6-
96231b3b80d8
Sanjay Patel [Fri, 14 Aug 2015 22:10:59 +0000 (22:10 +0000)]
make current codegen visible in the checks, so we can decide if it's right
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245108
91177308-0d34-0410-b5e6-
96231b3b80d8
Ahmed Bougacha [Fri, 14 Aug 2015 22:06:05 +0000 (22:06 +0000)]
[AArch64] Fix FMLS scalar-indexed-from-2s-after-neg patterns.
We canonicalize V64 vectors to V128 through insert_subvector: the other
FMLA/FMLS/FMUL/FMULX patterns match that already, but this one doesn't,
so we'd fail to match fmls and generate fneg+fmla instead.
The vector equivalents are already tested and functional.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245107
91177308-0d34-0410-b5e6-
96231b3b80d8
Evgeniy Stepanov [Fri, 14 Aug 2015 22:03:50 +0000 (22:03 +0000)]
[msan] Fix handling of musttail calls.
MSan instrumentation for return values of musttail calls is not
allowed by the IR constraints, and not needed at the same time.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245106
91177308-0d34-0410-b5e6-
96231b3b80d8
Alexei Starovoitov [Fri, 14 Aug 2015 22:00:45 +0000 (22:00 +0000)]
[bpf] add documentation and instruction set description
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245105
91177308-0d34-0410-b5e6-
96231b3b80d8
Alex Lorenz [Fri, 14 Aug 2015 21:55:58 +0000 (21:55 +0000)]
MIR Serialization: Serialize the '.cfi_same_value' CFI directive.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245103
91177308-0d34-0410-b5e6-
96231b3b80d8
Alex Lorenz [Fri, 14 Aug 2015 21:14:50 +0000 (21:14 +0000)]
MIR Serialization: Serialize the external symbol call entry pseudo source
values.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245098
91177308-0d34-0410-b5e6-
96231b3b80d8
Alex Lorenz [Fri, 14 Aug 2015 21:08:30 +0000 (21:08 +0000)]
MIR Serialization: Serialize the global value call entry pseudo source values.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245097
91177308-0d34-0410-b5e6-
96231b3b80d8
Michael Kruse [Fri, 14 Aug 2015 20:20:00 +0000 (20:20 +0000)]
[RegionInfo] Remove unused and broken function splitBlock
Summary:
It always makes NewBB the entry of the region instead of OldBB. This breaks if there are edges from inside the region to OldBB. OldBB is moved out of the region and hence there are exiting edges to OldBB and the region's exit block, contradicting the single-exit condition for regions.
The only use from Polly is going to be removed, hence I propose to remove the function completely.
Reviewers: grosser
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D11873
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245092
91177308-0d34-0410-b5e6-
96231b3b80d8
Tom Stellard [Fri, 14 Aug 2015 19:46:05 +0000 (19:46 +0000)]
AMDGPU/SI: Add missing spill class
The compiler was failing to spill for some shaders.
Patch By: Axel Davy
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245087
91177308-0d34-0410-b5e6-
96231b3b80d8
Renato Golin [Fri, 14 Aug 2015 19:35:47 +0000 (19:35 +0000)]
Revert "[ARM] Fix MachO CPU Subtype selection"
This reverts commit r245081, as it breaks many builds.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245086
91177308-0d34-0410-b5e6-
96231b3b80d8
Alex Lorenz [Fri, 14 Aug 2015 19:07:07 +0000 (19:07 +0000)]
MIR Serialization: Serialize the 'internal' register operand flag.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245085
91177308-0d34-0410-b5e6-
96231b3b80d8
Alex Lorenz [Fri, 14 Aug 2015 18:57:24 +0000 (18:57 +0000)]
MIR Serialization: Serialize the bundled machine instructions.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245082
91177308-0d34-0410-b5e6-
96231b3b80d8
Vedant Kumar [Fri, 14 Aug 2015 18:36:47 +0000 (18:36 +0000)]
[ARM] Fix MachO CPU Subtype selection
This patch makes the Darwin ARM backend take advantage of TargetParser. It
also teaches TargetParser about ARMV7K for the first time. This makes target
triple parsing more consistent across llvm.
Differential Revision: http://reviews.llvm.org/D11996
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245081
91177308-0d34-0410-b5e6-
96231b3b80d8
Sanjay Patel [Fri, 14 Aug 2015 17:53:40 +0000 (17:53 +0000)]
[x86] fix allowsMisalignedMemoryAccess() implementation
This patch fixes the x86 implementation of allowsMisalignedMemoryAccess() to correctly
return the 'Fast' output parameter for 32-byte accesses. To test that, an existing load
merging optimization is changed to use the TLI hook. This exposes a shortcoming in the
current logic and results in the regression test update. Changing other direct users of
the isUnalignedMem32Slow() x86 CPU attribute would be a follow-on patch.
Without the fix in allowsMisalignedMemoryAccesses(), we will infinite loop when targeting
SandyBridge because LowerINSERT_SUBVECTOR() creates 32-byte loads from two 16-byte loads
while PerformLOADCombine() splits them back into 16-byte loads.
Differential Revision: http://reviews.llvm.org/D10662
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245075
91177308-0d34-0410-b5e6-
96231b3b80d8
Vedant Kumar [Fri, 14 Aug 2015 17:42:50 +0000 (17:42 +0000)]
[test] Testing write access to llvm
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245074
91177308-0d34-0410-b5e6-
96231b3b80d8
Justin Bogner [Fri, 14 Aug 2015 17:03:45 +0000 (17:03 +0000)]
[sancov] Fix an unused variable warning introduced in r245067
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245072
91177308-0d34-0410-b5e6-
96231b3b80d8
Kit Barton [Fri, 14 Aug 2015 16:54:32 +0000 (16:54 +0000)]
Reverting patch r244235.
This patch will be redone in a different way. See
http://lists.llvm.org/pipermail/llvm-commits/Week-of-Mon-
20150810/292978.html
for more details.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245071
91177308-0d34-0410-b5e6-
96231b3b80d8
Reid Kleckner [Fri, 14 Aug 2015 16:48:34 +0000 (16:48 +0000)]
[cmake] Start adding support for LLVM_USE_SANITIZER=Address on Windows
Pass "-fsanitize=address" to the compiler and "-debug" to the linker.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245070
91177308-0d34-0410-b5e6-
96231b3b80d8
Reid Kleckner [Fri, 14 Aug 2015 16:45:42 +0000 (16:45 +0000)]
[sancov] Leave llvm.localescape in the entry block
Summary: Similar to the change we applied to ASan. The same test case works.
Reviewers: samsonov
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D11961
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245067
91177308-0d34-0410-b5e6-
96231b3b80d8
Chad Rosier [Fri, 14 Aug 2015 16:34:15 +0000 (16:34 +0000)]
Cleanup test whitespace or lack thereof. NFC.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245065
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Bieneman [Fri, 14 Aug 2015 16:20:31 +0000 (16:20 +0000)]
[CMake] Fix PR14200, llvm-config output misses -fno-rtti
This change adds RTTI and Exception flags to llvm-config's cxxflags. This solution is a minimal patch to solve the issue, and is recommended for the 3.7 release branch. Tom Stellard's outstanding work is the longer term solution.
Patch By: David Wiberg
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245064
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Fri, 14 Aug 2015 15:48:41 +0000 (15:48 +0000)]
Revert "Centralize the information about which object format we are using."
This reverts commit r245047.
It was failing on the darwin bots. The problem was that when running
./bin/llc -march=msp430
llc gets to
if (TheTriple.getTriple().empty())
TheTriple.setTriple(sys::getDefaultTargetTriple());
Which means that we go with an arch of msp430 but a triple of
x86_64-apple-darwin14.4.0 which fails badly.
That code has to be updated to select a triple based on the value of
march, but that is not a trivial fix.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245062
91177308-0d34-0410-b5e6-
96231b3b80d8
Davide Italiano [Fri, 14 Aug 2015 15:16:37 +0000 (15:16 +0000)]
Convert tests under MC/ELF from macho-dump to llvm-readobj.
Yet another step towards deprecating macho-dump.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245059
91177308-0d34-0410-b5e6-
96231b3b80d8
Sanjay Patel [Fri, 14 Aug 2015 15:11:42 +0000 (15:11 +0000)]
don't repeaat function names in comments; NFC
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245058
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Fri, 14 Aug 2015 13:31:17 +0000 (13:31 +0000)]
Centralize the information about which object format we are using.
Other than some places that were handling unknown as ELF, this should
have no change. The test updates are because we were detecting
arm-coff or x86_64-win64-coff as ELF targets before.
It is not clear if the enum should live on the Triple. At least now it lives
in a single location and should be easier to move somewhere else.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245047
91177308-0d34-0410-b5e6-
96231b3b80d8
James Molloy [Fri, 14 Aug 2015 11:09:09 +0000 (11:09 +0000)]
Separate out BDCE's analysis into a separate DemandedBits analysis.
This allows other areas of the compiler to use BDCE's bit-tracking.
NFCI.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245039
91177308-0d34-0410-b5e6-
96231b3b80d8
Simon Pilgrim [Fri, 14 Aug 2015 11:03:31 +0000 (11:03 +0000)]
Renamed min tests (typo)
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245038
91177308-0d34-0410-b5e6-
96231b3b80d8
James Molloy [Fri, 14 Aug 2015 09:08:50 +0000 (09:08 +0000)]
[AArch64] FMINNAN/FMAXNAN on f16 is not legal.
Spotted by Ahmed - in r244594 I inadvertently marked f16 min/max as legal.
I've reverted it here, and marked min/max on scalar f16's as promote. I've also added a testcase. The test just checks that the compiler doesn't fall over - it doesn't create fmin nodes for f16 yet.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245035
91177308-0d34-0410-b5e6-
96231b3b80d8
Chandler Carruth [Fri, 14 Aug 2015 08:39:32 +0000 (08:39 +0000)]
[PM/AA] Remove two no-op overridden functions that just delegated to the
base class anyways.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245034
91177308-0d34-0410-b5e6-
96231b3b80d8
Adam Nemet [Fri, 14 Aug 2015 06:30:26 +0000 (06:30 +0000)]
[LVer] Remove unused Pass parameter from versionLoop, NFC
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245032
91177308-0d34-0410-b5e6-
96231b3b80d8
Lang Hames [Fri, 14 Aug 2015 06:26:42 +0000 (06:26 +0000)]
[RuntimeDyld] Make sure code-sections aren't under-aligned.
Code-section alignment should be at least as high as the minimum
stub alignment. If the section alignment is lower it can cause
padding to be emitted resulting in alignment errors if the section
is mapped to a higher alignment on the target.
E.g. If a text section with a 4-byte alignment gets 4-bytes of
padding to guarantee 8-byte alignment for stubs but is re-mapped to
an 8-byte alignment on the target, the 4-bytes of padding will push
the stubs to 4-byte alignment causing a crash.
No test case: There is currently no way to control host section
alignment in llvm-rtdyld. This could be made testable by adding
a custom memory manager. I'll look at that in a follow-up patch.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245031
91177308-0d34-0410-b5e6-
96231b3b80d8
David Majnemer [Fri, 14 Aug 2015 05:09:07 +0000 (05:09 +0000)]
[IR] Add token types
This introduces the basic functionality to support "token types".
The motivation stems from the need to perform operations on a Value
whose provenance cannot be obscured.
There are several applications for such a type but my immediate
motivation stems from WinEH. Our personality routine enforces a
single-entry - single-exit regime for cleanups. After several rounds of
optimizations, we may be left with a terminator whose "cleanup-entry
block" is not entirely clear because control flow has merged two
cleanups together. We have experimented with using labels as operands
inside of instructions which are not terminators to indicate where we
came from but found that LLVM does not expect such exotic uses of
BasicBlocks.
Instead, we can use this new type to clearly associate the "entry point"
and "exit point" of our cleanup. This is done by having the cleanuppad
yield a Token and consuming it at the cleanupret.
The token type makes it impossible to obscure or otherwise hide the
Value, making it trivial to track the relationship between the two
points.
What is the burden to the optimizer? Well, it turns out we have already
paid down this cost by accepting that there are certain calls that we
are not permitted to duplicate, optimizations have to watch out for
such instructions anyway. There are additional places in the optimizer
that we will probably have to update but early examination has given me
the impression that this will not be heroic.
Differential Revision: http://reviews.llvm.org/D11861
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245029
91177308-0d34-0410-b5e6-
96231b3b80d8
Karthik Bhat [Fri, 14 Aug 2015 04:17:23 +0000 (04:17 +0000)]
Add support for cross block dse.
This patch enables dead stroe elimination across basicblocks.
Example:
define void @test_02(i32 %N) {
%1 = alloca i32
store i32 %N, i32* %1
store i32 10, i32* @x
%2 = load i32, i32* %1
%3 = icmp ne i32 %2, 0
br i1 %3, label %4, label %5
; <label>:4
store i32 5, i32* @x
br label %7
; <label>:5
%6 = load i32, i32* @x
store i32 %6, i32* @y
br label %7
; <label>:7
store i32 15, i32* @x
ret void
}
In the above example dead store "store i32 5, i32* @x" is now eliminated.
Differential Revision: http://reviews.llvm.org/D11143
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245025
91177308-0d34-0410-b5e6-
96231b3b80d8
Chandler Carruth [Fri, 14 Aug 2015 03:57:00 +0000 (03:57 +0000)]
[PM/AA] Run clang-format over the ObjCARC Alias Analysis code to
normalize its formatting before I make more substantial changes.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245024
91177308-0d34-0410-b5e6-
96231b3b80d8
Chandler Carruth [Fri, 14 Aug 2015 03:55:36 +0000 (03:55 +0000)]
[PM/AA] Don't bother forward declaring Function and Value, just include
their headers.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245023
91177308-0d34-0410-b5e6-
96231b3b80d8
Saleem Abdulrasool [Fri, 14 Aug 2015 03:48:35 +0000 (03:48 +0000)]
PowerPC: remove dead initialization (NFC)
Identified by the clang static analyzer. No functional change intended.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245022
91177308-0d34-0410-b5e6-
96231b3b80d8
Chandler Carruth [Fri, 14 Aug 2015 03:48:20 +0000 (03:48 +0000)]
[PM/AA] Extract the interface for GlobalsModRef into a header along with
its creation function.
This required shifting a bunch of method definitions to be out-of-line
so that we could leave most of the implementation guts in the .cpp file.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245021
91177308-0d34-0410-b5e6-
96231b3b80d8
Chandler Carruth [Fri, 14 Aug 2015 03:33:48 +0000 (03:33 +0000)]
[PM/AA] Hoist the interface to TBAA into a dedicated header along with
its creation function. Update the relevant includes accordingly.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245019
91177308-0d34-0410-b5e6-
96231b3b80d8
Chandler Carruth [Fri, 14 Aug 2015 03:26:15 +0000 (03:26 +0000)]
[PM/AA] Run clang-format over TBAA code to normalize the formatting
before making substantial changes.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245017
91177308-0d34-0410-b5e6-
96231b3b80d8
Chandler Carruth [Fri, 14 Aug 2015 03:16:11 +0000 (03:16 +0000)]
[PM/AA] Remove a stray #include that snuck in via copy/paste when
creating this header.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245016
91177308-0d34-0410-b5e6-
96231b3b80d8
Chandler Carruth [Fri, 14 Aug 2015 03:14:50 +0000 (03:14 +0000)]
[PM/AA] Clean up the SCEV-AA comment formatting and typos.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245015
91177308-0d34-0410-b5e6-
96231b3b80d8
Chandler Carruth [Fri, 14 Aug 2015 03:12:16 +0000 (03:12 +0000)]
[PM/AA] Run clang-format over the SCEV-AA code to normalize the
formatting.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245014
91177308-0d34-0410-b5e6-
96231b3b80d8
Chandler Carruth [Fri, 14 Aug 2015 03:11:16 +0000 (03:11 +0000)]
[PM/AA] Hoist the SCEV-AA interface to its own header and pull the
creation function into that header.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245013
91177308-0d34-0410-b5e6-
96231b3b80d8
Chandler Carruth [Fri, 14 Aug 2015 02:55:50 +0000 (02:55 +0000)]
[PM/AA] Hoist ScopedNoAliasAA's interface into a header and move the
creation function there.
Same basic refactoring as the other alias analyses. Nothing special
required this time around.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245012
91177308-0d34-0410-b5e6-
96231b3b80d8
Chandler Carruth [Fri, 14 Aug 2015 02:50:34 +0000 (02:50 +0000)]
[PM/AA] Hoist the value handle definition for CFLAA into the header to
satisfy libc++'s std::forward_list which requires the value type to be
complete.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245011
91177308-0d34-0410-b5e6-
96231b3b80d8
Chandler Carruth [Fri, 14 Aug 2015 02:46:07 +0000 (02:46 +0000)]
[PM/AA] Run clang-format over the ScopedNoAliasAA pass prior to making
substantial changes to normalize any formatting.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@245010
91177308-0d34-0410-b5e6-
96231b3b80d8