Ehsan Akhgari [Wed, 9 Jul 2014 00:40:50 +0000 (00:40 +0000)]
[ms-coff] Add a test for proper handling of full Windows path names in the .drectve section
Summary: This test ensures that we can correctly specify a full Windows path to the clang ASAN runtime libraries. This is in preparation to fix PR20246.
Reviewers: rafael
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D4427
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212580
91177308-0d34-0410-b5e6-
96231b3b80d8
NAKAMURA Takumi [Tue, 8 Jul 2014 23:48:22 +0000 (23:48 +0000)]
MipsTargetStreamer.h: Avoid "using" to appease msc17.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212577
91177308-0d34-0410-b5e6-
96231b3b80d8
Kevin Enderby [Tue, 8 Jul 2014 23:47:31 +0000 (23:47 +0000)]
Changed the lvm-nm alias "-s" for -print-armap to "-M".
This will allow the "-s" flag to implemented in the future as it
is in darwin’s nm(1) to list symbols only in the specified section.
Given a LGTM by Shankar Easwaran who originally implemented
the support for lvm-nm’s -print-armap and archive map symbols.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212576
91177308-0d34-0410-b5e6-
96231b3b80d8
Jim Grosbach [Tue, 8 Jul 2014 23:28:48 +0000 (23:28 +0000)]
AArch64: Better codegen for loading from __fp16.
Loading will generally extend to an f32 or an 64, so make sure
to match those patterns directly to load into the FPR16 register
class directly rather than going through the integer GPRs.
This also eliminates an extra step in the convert-to-f64 path
which was first converting to f32 and then to f64 from there.
rdar://
17594379
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212573
91177308-0d34-0410-b5e6-
96231b3b80d8
Hal Finkel [Tue, 8 Jul 2014 23:16:49 +0000 (23:16 +0000)]
Improve BasicAA CS-CS queries
BasicAA contains knowledge of certain intrinsics, such as memcpy and memset,
and uses that information to form more-accurate answers to CallSite vs. Loc
ModRef queries. Unfortunately, it did not use this information when answering
CallSite vs. CallSite queries.
Generically, when an intrinsic takes one or more pointers and the intrinsic is
marked only to read/write from its arguments, the offset/size is unknown. As a
result, the generic code that answers CallSite vs. CallSite (and CallSite vs.
Loc) queries in AA uses UnknownSize when forming Locs from an intrinsic's
arguments. While BasicAA's CallSite vs. Loc override could use more-accurate
size information for some intrinsics, it did not do the same for CallSite vs.
CallSite queries.
This change refactors the intrinsic-specific logic in BasicAA into a generic AA
query function: getArgLocation, which is overridden by BasicAA to supply the
intrinsic-specific knowledge, and used by AA's generic implementation. This
allows the intrinsic-specific knowledge to be used by both CallSite vs. Loc and
CallSite vs. CallSite queries, and simplifies the BasicAA implementation.
Currently, only one function, Mac's memset_pattern16, is handled by BasicAA
(all the rest are intrinsics). As a side-effect of this refactoring, BasicAA's
getModRefBehavior override now also returns OnlyAccessesArgumentPointees for
this function (which is an improvement).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212572
91177308-0d34-0410-b5e6-
96231b3b80d8
Tobias Grosser [Tue, 8 Jul 2014 22:51:03 +0000 (22:51 +0000)]
DominanceInfo is strongly preferred over RegionInfo
This is and always was strong community consensus. Make this clear in the header
in case newcomers may not be aware.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212570
91177308-0d34-0410-b5e6-
96231b3b80d8
Kevin Enderby [Tue, 8 Jul 2014 22:10:02 +0000 (22:10 +0000)]
Add support for BSD format Archive map symbols (aka the table of contents
from a __.SYMDEF or "__.SYMDEF SORTED" archive member).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212568
91177308-0d34-0410-b5e6-
96231b3b80d8
Pete Cooper [Tue, 8 Jul 2014 17:06:03 +0000 (17:06 +0000)]
Revert "GlobalDCE: Delete available_externally initializers if it allows removing the value the initializer is referring to."
This reverts commit
5b55a47e94e28fbb56d0cd5d72c3db9105c15b4c.
A test case was found to crash after this was applied. I'll file a bug to track fixing this with the test case needed.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212550
91177308-0d34-0410-b5e6-
96231b3b80d8
Ulrich Weigand [Tue, 8 Jul 2014 16:16:02 +0000 (16:16 +0000)]
[PowerPC] Implement atomic NAND operations as actual NAND
This changes the implementation of atomic NAND operations
from "a & ~b" (compatible with GCC < 4.4) to actual "~(a & b)"
(compatible with GCC >= 4.4).
This is in line with the common-code and ARM back-end change
implemented in r212433.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212547
91177308-0d34-0410-b5e6-
96231b3b80d8
Andrea Di Biagio [Tue, 8 Jul 2014 15:22:29 +0000 (15:22 +0000)]
[DAG] Teach how to combine a pair of shuffles into a single shuffle if the resulting mask is legal.
This patch teaches how to fold a shuffle according to rule:
shuffle (shuffle (x, undef, M0), undef, M1) -> shuffle(x, undef, M2)
We do this only if the resulting mask M2 is legal; this is to avoid introducing
illegal shuffles that are potentially expanded into a sub-optimal sequence
of target specific dag nodes.
This patch has the advantage of being target independent, since it works on ISD
nodes. Therefore, all targets (not only x86) can take advantage of this rule.
The idea behind this patch is that most shuffle pairs can be safely combined
before we run the legalizer on vector operations. This allows us to
combine/simplify dag nodes earlier in the process and not only immediately
before instruction selection stage.
That said. This patch is not meant to replace any existing target specific
combine rules; backends might still introduce new shuffles during legalization
stage. Also, this rule is very simple and avoids to aggressively optimize
shuffles.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212539
91177308-0d34-0410-b5e6-
96231b3b80d8
Benjamin Kramer [Tue, 8 Jul 2014 14:55:06 +0000 (14:55 +0000)]
Fix some Twine locals.
Two of those are use after frees. Found by clang-tidy, fixed by me.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212537
91177308-0d34-0410-b5e6-
96231b3b80d8
Timur Iskhodzhanov [Tue, 8 Jul 2014 13:18:58 +0000 (13:18 +0000)]
[ASan/Win] Don't instrument private COMDAT globals until PR20244 is properly fixed
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212530
91177308-0d34-0410-b5e6-
96231b3b80d8
Daniel Sanders [Tue, 8 Jul 2014 13:13:42 +0000 (13:13 +0000)]
[mips] Fixed struct/class mismatch introduced in r212522.
Clang emits a warning about this.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212528
91177308-0d34-0410-b5e6-
96231b3b80d8
Daniel Sanders [Tue, 8 Jul 2014 10:35:52 +0000 (10:35 +0000)]
Fix r212522 - [mips] Improve encapsulation of the .MIPS.abiflags implementation and limit scope of related enums
Added two lines that should have been in r212522.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212523
91177308-0d34-0410-b5e6-
96231b3b80d8
Daniel Sanders [Tue, 8 Jul 2014 10:11:38 +0000 (10:11 +0000)]
[mips] Improve encapsulation of the .MIPS.abiflags implementation and limit scope of related enums
Summary:
Follow on to r212519 to improve the encapsulation and limit the scope of the enums.
Also merged two very similar parser functions, fixed a bug where ASE's
were not being reported, and marked CPR1's as being 128-bit when MSA is
enabled.
Differential Revision: http://reviews.llvm.org/D4384
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212522
91177308-0d34-0410-b5e6-
96231b3b80d8
Renato Golin [Tue, 8 Jul 2014 10:06:16 +0000 (10:06 +0000)]
Revert "Refactor ARM subarchitecture parsing"
This reverts commit
7b4a6882467e7fef4516a0cbc418cbfce0fc6f6d.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212521
91177308-0d34-0410-b5e6-
96231b3b80d8
Arnaud A. de Grandmaison [Tue, 8 Jul 2014 09:53:04 +0000 (09:53 +0000)]
Truncate the immediate in logical operation to the register width
And continue to produce an error if the 32 most significant bits are not all ones or zeros.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212520
91177308-0d34-0410-b5e6-
96231b3b80d8
Vladimir Medic [Tue, 8 Jul 2014 08:59:22 +0000 (08:59 +0000)]
Mips.abiflags is a new implicitly generated section that will be present on all new modules. The section contains a versioned data structure which represents essentially information to allow a program loader to determine the requirements of the application. This patch implements mips.abiflags section and provides test cases for it.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212519
91177308-0d34-0410-b5e6-
96231b3b80d8
Chandler Carruth [Tue, 8 Jul 2014 08:45:38 +0000 (08:45 +0000)]
[x86,SDAG] Sink the logic for folding shuffles of splats more
aggressively from the x86 shuffle lowering to the generic SDAG vector
shuffle formation code.
This code already tried to fold away shuffles of splats! It just had
lots of bugs and couldn't handle the case my new x86 shuffle lowering
needed.
First, it failed to correctly compute whether N2 was undef because it
pre-computed this, then did transformations which could *make* N2 undef,
then failed to ever re-consider the precomputed state.
Second, it didn't look through bitcasts at all, even in the safe cases
where they are just element-type bitcasts with no change to the number
of elements.
Third, it didn't handle all-zero bit casts nicely the way my code in the
x86 side of things did, which is essential to getting good zext-shuffle
lowerings.
But all of these are generic. I just ported the code down to this layer
and fixed the surrounding bugs. Tests exercising this in the x86 backend
still pass and some silly code in widen_cast-6.ll gets better. I updated
that test to be a bit more precise but it's still pretty unclear what
the value of the test is in this day and age.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212517
91177308-0d34-0410-b5e6-
96231b3b80d8
Chandler Carruth [Tue, 8 Jul 2014 07:44:15 +0000 (07:44 +0000)]
[SDAG] Actually check for a non-constant splat and clarify comments
around the handling of UNDEF lanes in boolean vector content analysis.
The code before my changes here also failed to check for non-constant
splats in a buildvector. I have no idea how to trigger this, I just
spotted by inspection when trying to understand the code. It seems
extremely unlikely to be worth the trouble to teach the only caller of
this code (DAG combining setcc patterns) how to cleverly handle undef
lanes, so I've just commented more thoroughly that we're giving up
there.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212515
91177308-0d34-0410-b5e6-
96231b3b80d8
Chandler Carruth [Tue, 8 Jul 2014 07:19:55 +0000 (07:19 +0000)]
[SDAG] Build up a more rich set of APIs for querying build-vector SDAG
nodes about whether they are splats. This is factored out and improved
from r212324 which got reverted as it was far too aggressive. The new
API should help more conservatively handle buildvectors that are
a mixture of splatted and undef values.
No functionality change at this point. The hope is to slowly
re-introduce the undef-tolerant optimization of splats, but each time
being forced to make a concious decision about how to handle the undefs
in a way that doesn't lead to contradicting assumptions about the
collapsed value.
Hal has pointed out in discussions that this may not end up being the
desired API and instead it may be more convenient to get a mask of the
undef elements or something similar. I'm starting simple and will expand
the API as I adapt actual callers and see exactly what they need.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212514
91177308-0d34-0410-b5e6-
96231b3b80d8
Alexey Samsonov [Tue, 8 Jul 2014 00:50:49 +0000 (00:50 +0000)]
[ASan] Completely remove sanitizer blacklist file from instrumentation pass.
All blacklisting logic is now moved to the frontend (Clang).
If a function (or source file it is in) is blacklisted, it doesn't
get sanitize_address attribute and is therefore not instrumented.
If a global variable (or source file it is in) is blacklisted, it is
reported to be blacklisted by the entry in llvm.asan.globals metadata,
and is not modified by the instrumentation.
The latter may lead to certain false positives - not all the globals
created by Clang are described in llvm.asan.globals metadata (e.g,
RTTI descriptors are not), so we may start reporting errors on them
even if "module" they appear in is blacklisted. We assume it's fine
to take such risk:
1) errors on these globals are rare and usually indicate wild memory access
2) we can lazily add descriptors for these globals into llvm.asan.globals
lazily.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212505
91177308-0d34-0410-b5e6-
96231b3b80d8
Adam Nemet [Tue, 8 Jul 2014 00:22:32 +0000 (00:22 +0000)]
[X86] AVX512: Only allow k1-k7 as predicates to vpcmp*
As destination k0 is allowed but not as predicate/writemask.
I also modified the test to allow checking of error messages by the assembler.
I applied a similar approach to the test ret.s in the same directory.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212504
91177308-0d34-0410-b5e6-
96231b3b80d8
Alexey Samsonov [Tue, 8 Jul 2014 00:03:11 +0000 (00:03 +0000)]
Kill unnecessary include
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212503
91177308-0d34-0410-b5e6-
96231b3b80d8
Andrea Di Biagio [Mon, 7 Jul 2014 23:25:23 +0000 (23:25 +0000)]
[x86] Fix assertion failure caused by a wrong combine of PSHUFD nodes with different types.
When combining a sequence of two PSHUFD dag nodes into a single PSHUFD,
make sure that we assign the correct type to the resulting PSHUFD.
X86ISD::PSHUFD dag nodes can be either MVT::v4i32 or MVT::v4f32.
Before this change, an assertion failure was triggered in method
'DAGCombinerInfo::CombineTo' when trying to combine the shuffles from the test
below into a single PSHUFD.
define <4 x float> @test1(<4 x float> %V) {
%1 = shufflevector <4 x float> %V, <4 x float> undef, <4 x i32> <i32 3, i32 0, i32 2, i32 1>
%2 = shufflevector <4 x float> %1, <4 x float> undef, <4 x i32> <i32 3, i32 0, i32 2, i32 1>
ret <4 x float> %2
}
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212498
91177308-0d34-0410-b5e6-
96231b3b80d8
Sanjay Patel [Mon, 7 Jul 2014 22:13:58 +0000 (22:13 +0000)]
fixed some typos
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212495
91177308-0d34-0410-b5e6-
96231b3b80d8
Juergen Ributzka [Mon, 7 Jul 2014 21:52:21 +0000 (21:52 +0000)]
[FastISel][X86] Fix smul.with.overflow.i8 lowering.
Add custom lowering code for signed multiply instruction selection, because the
default FastISel instruction selection for ISD::MUL will use unsigned multiply
for the i8 type and signed multiply for all other types. This would set the
incorrect flags for the overflow check.
This fixes <rdar://problem/
17549300>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212493
91177308-0d34-0410-b5e6-
96231b3b80d8
Louis Gerbarg [Mon, 7 Jul 2014 21:37:51 +0000 (21:37 +0000)]
Allow AArch64FastISel to degrade graceully in the presence of an MVT::i128
Currently AArch64FastISel crashes if it tries to extend an integer into an
MVT::i128. This can happen by creating 128 bit integers like so:
typedef unsigned int uint128_t __attribute__((mode(TI)));
typedef int sint128_t __attribute__((mode(TI)));
This patch makes EmitIntExt check for their presence and then falls back to
SelectionDAG.
Tests included.
rdar://
17516686
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212492
91177308-0d34-0410-b5e6-
96231b3b80d8
Sanjay Patel [Mon, 7 Jul 2014 21:19:00 +0000 (21:19 +0000)]
Fix for PR17073 ( llvm.org/pr17073 ), simplifycfg illegally hoists an operation in a phi node that can trap.
This patch adds to an existing loop over phi nodes in SimplifyCondBranchToCondBranch() to check for trapping ops and bails out of the optimization if we find one of those.
The test cases verify that trapping ops are not hoisted and non-trapping ops are still optimized as expected.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212490
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Mon, 7 Jul 2014 20:34:51 +0000 (20:34 +0000)]
Use raw_fd_ostream instead of std::ofstream.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212483
91177308-0d34-0410-b5e6-
96231b3b80d8
Renato Golin [Mon, 7 Jul 2014 20:01:11 +0000 (20:01 +0000)]
Refactor ARM subarchitecture parsing
According to a FIXME in ARMMCTargetDesc.cpp the ARM version parsing should be
in the Triple helper class.
Patch by: Gabor Ballabas
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212479
91177308-0d34-0410-b5e6-
96231b3b80d8
Ulrich Weigand [Mon, 7 Jul 2014 19:41:54 +0000 (19:41 +0000)]
[PowerPC] Fix testcase regression
Use -mcpu to avoid different codegen depending on host platform.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212478
91177308-0d34-0410-b5e6-
96231b3b80d8
Ulrich Weigand [Mon, 7 Jul 2014 19:39:44 +0000 (19:39 +0000)]
[PowerPC] Fix no-assert build
r212476 caused a compile failure (unused variable) in a non-assertion
build ...
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212477
91177308-0d34-0410-b5e6-
96231b3b80d8
Ulrich Weigand [Mon, 7 Jul 2014 19:26:41 +0000 (19:26 +0000)]
[PowerPC] Fix "byval align" arguments
Arguments passed as "byval align" should get the specified alignment
in the parameter save area. There was some code in PPCISelLowering.cpp
that attempted to implement this, but this didn't work correctly:
while code did update the ArgOffset value, it neglected to update
the PtrOff value (which was already computed from the old ArgOffset),
and it also neglected to update GPR_idx -- fields skipped due to
alignment in the save area must likewise be skipped in GPRs.
This patch fixes and simplifies this logic by:
- handling argument offset alignment right at the beginning
of argument processing, using a new helper routine
CalculateStackSlotAlignment (this avoids having to update
PtrOff and other derived values later on)
- not tracking GPR_idx separately, but always computing the
correct GPR_idx for each argument *from* its ArgOffset
- removing some redundant computation in LowerFormalArguments:
MinReservedArea must equal ArgOffset after argument processing,
so there's no use in computing it twice.
[This doesn't change the behavior of the current clang front-end,
since that never creates "byval align" arguments at the moment.
This will change with a follow-on patch, however.]
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212476
91177308-0d34-0410-b5e6-
96231b3b80d8
Chandler Carruth [Mon, 7 Jul 2014 19:03:32 +0000 (19:03 +0000)]
[x86] Revert r212324 which was too aggressive w.r.t. allowing undef
lanes in vector splats.
The core problem here is that undef lanes can't *unilaterally* be
considered to contribute to splats. Their handling needs to be more
cautious. There is also a reported failure of the nightly testers
(thanks Tobias!) that may well stem from the same core issue. I'm going
to fix this theoretical issue, factor the APIs a bit better, and then
verify that I don't see anything bad with Tobias's reduction from the
test suite before recommitting.
Original commit message for r212324:
[x86] Generalize BuildVectorSDNode::getConstantSplatValue to work for
any constant, constant FP, or undef splat and to tolerate any undef
lanes in a splat, then replace all uses of isSplatVector in X86's
lowering with it.
This fixes issues where undef lanes in an otherwise splat vector would
prevent the splat logic from firing. It is a touch more awkward to use
this interface, but it is much more accurate. Suggestions for better
interface structuring welcome.
With this fix, the code generated with the widening legalization
strategy for widen_cast-4.ll is *dramatically* improved as the special
lowering strategies for a v16i8 SRA kick in even though the high lanes
are undef.
We also get a slightly different choice for broadcasting an aligned
memory location, and use vpshufd instead of vbroadcastss. This looks
like a minor win for pipelining and domain crossing, but a minor loss
for the number of micro-ops. I suspect its a wash, but folks can
easily tweak the lowering if they want.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212475
91177308-0d34-0410-b5e6-
96231b3b80d8
Matt Arsenault [Mon, 7 Jul 2014 18:34:45 +0000 (18:34 +0000)]
R600: Fix mishandling of load / store chains.
Fixes various bugs with reordering loads and stores.
Scalarized vector loads weren't collecting the chains
at all.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212473
91177308-0d34-0410-b5e6-
96231b3b80d8
Matt Arsenault [Mon, 7 Jul 2014 18:34:42 +0000 (18:34 +0000)]
Fix typo, weird indentation
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212472
91177308-0d34-0410-b5e6-
96231b3b80d8
Tim Northover [Mon, 7 Jul 2014 15:26:53 +0000 (15:26 +0000)]
[testing]: lld generally lives in tools/, so fix llvm-lit.
Otherwise we can't run individual tests directly ("llvm-lit /path/to/test")
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212461
91177308-0d34-0410-b5e6-
96231b3b80d8
Benjamin Kramer [Mon, 7 Jul 2014 14:47:51 +0000 (14:47 +0000)]
Make helper functions static.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212460
91177308-0d34-0410-b5e6-
96231b3b80d8
Tim Northover [Mon, 7 Jul 2014 14:06:42 +0000 (14:06 +0000)]
X86: revert unintentional change to X86FastISel.
This crept in with r212443.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212459
91177308-0d34-0410-b5e6-
96231b3b80d8
Evgeniy Stepanov [Mon, 7 Jul 2014 13:57:37 +0000 (13:57 +0000)]
[asan] Generate asm instrumentation in MC.
Generate entire ASan asm instrumentation in MC without
relying on runtime helper functions.
Patch by Yuri Gorshenin.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212455
91177308-0d34-0410-b5e6-
96231b3b80d8
Evgeniy Stepanov [Mon, 7 Jul 2014 13:28:31 +0000 (13:28 +0000)]
[msan] Fix handling of phi in blacklisted functions.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212454
91177308-0d34-0410-b5e6-
96231b3b80d8
Benjamin Kramer [Mon, 7 Jul 2014 11:01:16 +0000 (11:01 +0000)]
InstCombine: Simplify code, no functionality change.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212449
91177308-0d34-0410-b5e6-
96231b3b80d8
Chandler Carruth [Mon, 7 Jul 2014 09:06:58 +0000 (09:06 +0000)]
[x86] Teach the new vector shuffle lowering code to handle what is
essentially a DAG combine that never gets a chance to run.
We might typically expect DAG combining to remove shuffles-of-splats and
other similar patterns, but we don't get a chance to run the DAG
combiner when we recursively form sub-shuffles during the lowering of
a shuffle. So instead hand-roll a really important combine directly into
the lowering code to detect shuffles-of-splats, especially shuffles of
an all-zero splat which needn't even have the same element width, etc.
This lets the new vector shuffle lowering handle shuffles which
implement things like zero-extension really nicely. This will become
even more important when I wire the legalization of zero-extension to
vector shuffles with the new widening legalization strategy.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212444
91177308-0d34-0410-b5e6-
96231b3b80d8
Tim Northover [Mon, 7 Jul 2014 09:06:35 +0000 (09:06 +0000)]
CodeGen: it turns out that NAND is not the same thing as BIC. At all.
We've been performing the wrong operation on ARM for "atomicrmw nand" for
years, since "a NAND b" is "~(a & b)" rather than ARM's very tempting "a & ~b".
This bled over into the generic expansion pass.
So I assume no-one has ever actually tried to do an atomic nand in the real
world. Oh well.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212443
91177308-0d34-0410-b5e6-
96231b3b80d8
Saleem Abdulrasool [Mon, 7 Jul 2014 05:18:35 +0000 (05:18 +0000)]
ARM: properly lower dllimport'ed global values
This completes the handling for DLL import storage symbols when lowering
instructions. A DLL import storage symbol must have an additional load
performed prior to use. This is applicable to variables and functions.
This is particularly important for non-function symbols as it is possible to
handle function references by emitting a thunk which performs the translation
from the unprefixed __imp_ symbol to the proper symbol (although, this is a
non-optimal lowering). For a variable symbol, no such thunk can be
accommodated.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212431
91177308-0d34-0410-b5e6-
96231b3b80d8
Saleem Abdulrasool [Mon, 7 Jul 2014 05:18:30 +0000 (05:18 +0000)]
ARM: correctly mangle dllimport symbols
Add support for tracking DLLImport storage class information on a per symbol
basis in the ARM instruction selection. Use that information to correctly
mangle the symbol (dllimport symbols are referenced via *__imp_<name>).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212430
91177308-0d34-0410-b5e6-
96231b3b80d8
Saleem Abdulrasool [Mon, 7 Jul 2014 05:18:22 +0000 (05:18 +0000)]
ARM: unify symbol name retrieval
Ensure that all paths that retrieve the symbol name go through GetARMGVSymbol
rather than getSymbol. This is desirable so that any global symbol mangling can
be centralised to this function. The motivation for this is handling of symbols
that are marked as having dll import dll storage. Such a symbol requires an
extra load that is currently handled in the backend and a __imp_ prefix on the
symbol name.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212429
91177308-0d34-0410-b5e6-
96231b3b80d8
Kevin Qin [Mon, 7 Jul 2014 02:45:40 +0000 (02:45 +0000)]
[AArch64] Normalize all constants to build a vector.
The value of constant operands will be truncated to fit element width.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212428
91177308-0d34-0410-b5e6-
96231b3b80d8
Sanjay Patel [Sun, 6 Jul 2014 23:24:53 +0000 (23:24 +0000)]
fixed typos in comments
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212424
91177308-0d34-0410-b5e6-
96231b3b80d8
Sanjay Patel [Sun, 6 Jul 2014 23:10:24 +0000 (23:10 +0000)]
fixed some typos in comments
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212423
91177308-0d34-0410-b5e6-
96231b3b80d8
Saleem Abdulrasool [Sun, 6 Jul 2014 22:13:26 +0000 (22:13 +0000)]
AArch64: whitespace cleanup
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212420
91177308-0d34-0410-b5e6-
96231b3b80d8
Aaron Ballman [Sun, 6 Jul 2014 20:20:02 +0000 (20:20 +0000)]
These should be EXPECT_TRUE, not EXPECT_FALSE. Amends r212415.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212419
91177308-0d34-0410-b5e6-
96231b3b80d8
Aaron Ballman [Sun, 6 Jul 2014 19:34:52 +0000 (19:34 +0000)]
Fixing compile errors related to changes with MemoryBuffer::getFile.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212415
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Sun, 6 Jul 2014 17:43:13 +0000 (17:43 +0000)]
Update the MemoryBuffer API to use ErrorOr.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212405
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Sun, 6 Jul 2014 14:31:22 +0000 (14:31 +0000)]
Declare variable on first use.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212403
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Sun, 6 Jul 2014 14:24:03 +0000 (14:24 +0000)]
This only needs a StringRef.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212402
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Sun, 6 Jul 2014 14:17:29 +0000 (14:17 +0000)]
This only needs a StringRef.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212401
91177308-0d34-0410-b5e6-
96231b3b80d8
Alp Toker [Sun, 6 Jul 2014 10:54:41 +0000 (10:54 +0000)]
Fix the MSVC build following r212382
Looks like the casts are needed there after all.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212399
91177308-0d34-0410-b5e6-
96231b3b80d8
Alp Toker [Sun, 6 Jul 2014 10:33:31 +0000 (10:33 +0000)]
SourceMgr: make valid buffer IDs start from one
Use 0 for the invalid buffer instead of -1/~0 and switch to unsigned
representation to enable more idiomatic usage.
Also introduce a trivial SourceMgr::getMainFileID() instead of hard-coding 0/1
to identify the main file.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212398
91177308-0d34-0410-b5e6-
96231b3b80d8
Alp Toker [Sun, 6 Jul 2014 10:32:55 +0000 (10:32 +0000)]
Don't use StringRef iterator functions for data access
And also remove some redundant casts from r212371.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212397
91177308-0d34-0410-b5e6-
96231b3b80d8
Alp Toker [Sat, 5 Jul 2014 22:20:59 +0000 (22:20 +0000)]
Remove IntrusiveRefCntPtr::getPtr() function
It was deprecated in r212366 and all uses have been switched to get().
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212382
91177308-0d34-0410-b5e6-
96231b3b80d8
Matt Arsenault [Sat, 5 Jul 2014 21:16:43 +0000 (21:16 +0000)]
Use cast<> instead of dyn_cast + assert
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212380
91177308-0d34-0410-b5e6-
96231b3b80d8
Matt Arsenault [Sat, 5 Jul 2014 21:16:40 +0000 (21:16 +0000)]
Fix grammar
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212379
91177308-0d34-0410-b5e6-
96231b3b80d8
Saleem Abdulrasool [Sat, 5 Jul 2014 20:09:24 +0000 (20:09 +0000)]
ARM: mark matching ARM intrinsics as MSBuiltin
A number of the ARM intrinsics are aliased with alternative names in MSVC
compatibility mode. This change indicates those intrinsics to permit tablegen
to construct an appropriate list of MSBuiltins. With the corresponding change
in clang, these intrinsics can then be mapped from the frontend.
The tests to validate the intrinsics are aliased correctly will be added with
the corresponding clang change.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212377
91177308-0d34-0410-b5e6-
96231b3b80d8
Ehsan Akhgari [Sat, 5 Jul 2014 19:46:10 +0000 (19:46 +0000)]
Revert r212375 because of test failures
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212376
91177308-0d34-0410-b5e6-
96231b3b80d8
Ehsan Akhgari [Sat, 5 Jul 2014 19:40:35 +0000 (19:40 +0000)]
Add a test case for the tilde operator in Microsoft inline assembly
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212375
91177308-0d34-0410-b5e6-
96231b3b80d8
Simon Atanasyan [Sat, 5 Jul 2014 19:28:49 +0000 (19:28 +0000)]
[llvm-readobj] Fix output of MIPS GOT without local and global entries.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212374
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Sat, 5 Jul 2014 11:38:52 +0000 (11:38 +0000)]
This only needs a StringRef. No functionality change.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212371
91177308-0d34-0410-b5e6-
96231b3b80d8
David Majnemer [Sat, 5 Jul 2014 06:12:30 +0000 (06:12 +0000)]
ADT: Add a drop_back() helper to ArrayRef
The slice(N, M) interface is powerful but not concise when wanting to
drop a few elements off of an ArrayRef, fix this by adding a drop_back
method.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212370
91177308-0d34-0410-b5e6-
96231b3b80d8
Alp Toker [Sat, 5 Jul 2014 03:03:21 +0000 (03:03 +0000)]
Deprecate IntrusiveRefCntPtr::getPtr() in favour of get()
This better aligns with other LLVM-specific and C++ standard library smart
pointer types.
In particular there are at least a few uses of intrusive refcounting in the
frontend where it's worth investigating std::shared_ptr as a more appropriate
alternative.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212366
91177308-0d34-0410-b5e6-
96231b3b80d8
David Majnemer [Sat, 5 Jul 2014 00:39:52 +0000 (00:39 +0000)]
MC: make MCSymbolData::dump work on const objects
This just lets us dump a const MCSymbolData object, no functionality
changed.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212365
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Sat, 5 Jul 2014 00:39:08 +0000 (00:39 +0000)]
Make a helper function static. No functionality change.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212364
91177308-0d34-0410-b5e6-
96231b3b80d8
David Majnemer [Fri, 4 Jul 2014 23:20:46 +0000 (23:20 +0000)]
MC: Correct comment in ExportSymbol
No functionality changed, just make it so that the code _could_ be
uncommented.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212363
91177308-0d34-0410-b5e6-
96231b3b80d8
David Majnemer [Fri, 4 Jul 2014 23:15:28 +0000 (23:15 +0000)]
MC: Cleanup COFFAsmParser::ParseSectionFlags
Switch a normal for-loop to a range-based for. No functionality changed.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212362
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Fri, 4 Jul 2014 22:44:18 +0000 (22:44 +0000)]
Make RecordStreamer.h private.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212361
91177308-0d34-0410-b5e6-
96231b3b80d8
David Majnemer [Fri, 4 Jul 2014 22:05:26 +0000 (22:05 +0000)]
IR: Fold away compares between GV GEPs and GVs
A GEP of a non-weak global variable will not be equivalent to another
non-weak global variable or a GEP of such a variable.
Differential Revision: http://reviews.llvm.org/D4238
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212360
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Fri, 4 Jul 2014 20:05:56 +0000 (20:05 +0000)]
Fix a bug in the conversion to ErrorOr.
The regular end of the bitcode parsing is in the BitstreamEntry::EndBlock
case.
Should fix the LTO bootstrap on OS X (this function is only used by ld64).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212357
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Fri, 4 Jul 2014 20:02:42 +0000 (20:02 +0000)]
Revert "Convert a few std::strings to StringRef."
This reverts commit r212342.
We can get a StringRef into the current Record, but not one in the bitcode
itself since the string is compressed in it.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212356
91177308-0d34-0410-b5e6-
96231b3b80d8
Sanjay Patel [Fri, 4 Jul 2014 19:40:43 +0000 (19:40 +0000)]
fixed typos
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212355
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Fri, 4 Jul 2014 19:31:27 +0000 (19:31 +0000)]
Ignore llvm specific symbols in the LTOModule.
These are the llvm.* globals and functions.
I don't think it is possible to test this directly since llvm-lto is not
a full linker and will not report duplicated symbols, but this fixes
bootstrap with gold and lto enabled.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212354
91177308-0d34-0410-b5e6-
96231b3b80d8
Ehsan Akhgari [Fri, 4 Jul 2014 19:13:05 +0000 (19:13 +0000)]
Add support for parsing the not operator in Microsoft inline assembly
This fixes http://llvm.org/PR20202
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212352
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Fri, 4 Jul 2014 19:08:22 +0000 (19:08 +0000)]
Ignore llvm.* globals.
It is not clear if llvm.global_ctors should or should not be in llvm.metadata,
but in practice it is not and we need to ignore it for LTO.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212351
91177308-0d34-0410-b5e6-
96231b3b80d8
Saleem Abdulrasool [Fri, 4 Jul 2014 18:42:25 +0000 (18:42 +0000)]
TableGen: introduce support for MSBuiltin
Add MSBuiltin which is similar in vein to GCCBuiltin. This allows for adding
intrinsics for Microsoft compatibility to individual instructions. This is
needed to permit the creation of ARM specific MSVC extensions.
This is not currently in use, and requires an associated change in clang to
enable use of the intrinsics defined by this new class. This merely sets the
LLVM portion of the infrastructure in place to permit the use of this
functionality. A separate set of changes will enable the new intrinsics.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212350
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Fri, 4 Jul 2014 18:40:36 +0000 (18:40 +0000)]
Implement LTOModule on top of IRObjectFile.
IRObjectFile provides all the logic for producing mangled names and getting
symbols from inline assembly.
LTOModule then adds logic for linking specific tasks, like constructing
llvm.compiler_user or extracting linker options from the bitcode.
The rule of the thumb is that IRObjectFile has the functionality that is
needed by both LTO and llvm-ar.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212349
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Fri, 4 Jul 2014 16:37:02 +0000 (16:37 +0000)]
Avoid mangling names twice. No functionality change.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212348
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Fri, 4 Jul 2014 15:58:00 +0000 (15:58 +0000)]
Mark intrinsic functions as llvm-specific.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212347
91177308-0d34-0410-b5e6-
96231b3b80d8
Daniel Sanders [Fri, 4 Jul 2014 15:21:53 +0000 (15:21 +0000)]
[mips][mips64r6] Set ELF e_flags for MIPS32r6/MIPS64r6. Also do MIPS-I to MIPS-V
Differential Revision: http://reviews.llvm.org/D4386
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212346
91177308-0d34-0410-b5e6-
96231b3b80d8
Daniel Sanders [Fri, 4 Jul 2014 15:16:14 +0000 (15:16 +0000)]
[mips] Add tests for the 'ret', 'call', and 'indirectbr' LLVM IR instruction.
Summary:
The tests in this directory are intended to test a single IR instruction
with as few dependencies on other instructions as possible. The aim is to
be very confident that each LLVM-IR instruction is implemented correctly and
with the optimal sequence of instructions, as well as to make it easy to tell
what is tested, and make it easier to bring up new ISA revisions in the
future. This gives us a good foundation on which to test bigger things.
These particular tests will allow testing that MIPS32r6/MIPS64r6 generate
the correct return instruction for returns, calls, and indirect branches.
This will be a bit tricky since the assembly text is identical but the
instruction is actually different. On MIPS32r6/MIPS64r6 'jr $rs' has been
removed in favour of the equivalent 'jalr $zero, $rs'. 'jr $rs' remains as
an alias for 'jalr $zero, $rs'.
Differential Revision: http://reviews.llvm.org/D4266
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212345
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Fri, 4 Jul 2014 15:03:17 +0000 (15:03 +0000)]
Don't include llvm.metadata variables in archive symbol tables.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212344
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Fri, 4 Jul 2014 14:19:41 +0000 (14:19 +0000)]
Change LTOModule`s getTargetTriple and setTargetTriple to use c++ types.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212343
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Fri, 4 Jul 2014 14:12:46 +0000 (14:12 +0000)]
Convert a few std::strings to StringRef.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212342
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Fri, 4 Jul 2014 13:52:01 +0000 (13:52 +0000)]
Convert these functions to use ErrorOr.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212341
91177308-0d34-0410-b5e6-
96231b3b80d8
Rafael Espindola [Fri, 4 Jul 2014 13:30:13 +0000 (13:30 +0000)]
Remove unused old-style error handling.
If needed, an ErrorOr should be used.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212340
91177308-0d34-0410-b5e6-
96231b3b80d8
Benjamin Kramer [Fri, 4 Jul 2014 12:36:05 +0000 (12:36 +0000)]
GlobalDCE: Delete available_externally initializers if it allows removing the value the initializer is referring to.
This is useful for functions that are not actually available externally but
referenced by a vtable of some kind. Clang emits functions like this for the MS
ABI.
PR20182.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212337
91177308-0d34-0410-b5e6-
96231b3b80d8
NAKAMURA Takumi [Fri, 4 Jul 2014 11:58:03 +0000 (11:58 +0000)]
llvm/test/CodeGen/XCore/dwarf_debug.ll: Fix not to be affected by *-win32.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212335
91177308-0d34-0410-b5e6-
96231b3b80d8
NAKAMURA Takumi [Fri, 4 Jul 2014 11:55:40 +0000 (11:55 +0000)]
llvm/test/CodeGen/X86/vector-gep.ll: Appease to add -mtriple=i686-linux.
This doesn't pass if stack alignment is not 16, like cygming, *bsd.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212334
91177308-0d34-0410-b5e6-
96231b3b80d8
Tim Northover [Fri, 4 Jul 2014 10:58:05 +0000 (10:58 +0000)]
ARM: when falling back to scattered relocs, keep the type.
The linker relies on relocation type info (e.g. is it a branch?) to perform the
correct actions, so we should keep that even when we end up using a scattered
relocation for whatever reason.
rdar://problem/
17553104
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212333
91177308-0d34-0410-b5e6-
96231b3b80d8
Tim Northover [Fri, 4 Jul 2014 10:57:56 +0000 (10:57 +0000)]
llvm-readobj: fix MachO relocatoin printing a bit.
There were two issues here:
1. At the very least, scattered relocations cannot use the same code to
determine the corresponding symbol being referred to. For some reason we
pretend there is no symbol, even when one actually exists in the symtab, so to
match this behaviour getRelocationSymbol should simply return symbols_end for
scattered relocations.
2. Printing "-" when we can't get a symbol (including the scattered case, but
not exclusively), isn't that helpful. In both cases there *is* interesting
information in that field, so we should print it. As hex will do.
Small part of rdar://problem/
17553104
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212332
91177308-0d34-0410-b5e6-
96231b3b80d8
Benjamin Kramer [Fri, 4 Jul 2014 10:22:21 +0000 (10:22 +0000)]
InstCombine: Strength reduce sadd.with.overflow into a regular nsw add if we can prove that it cannot overflow.
PR20194
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212331
91177308-0d34-0410-b5e6-
96231b3b80d8