llvm-svn: 246108

By Greg Bedwell!

Differential Revision: http://reviews.llvm.org/D12368

llvm-svn: 246090

By Brenden Blanco!

llvm-svn: 246072

Patch by Greg Bedwell!

Differential Revision: http://reviews.llvm.org/D12367

llvm-svn: 246071

By Tobias Grosser!

llvm-svn: 246053

llvm-svn: 246036

SCEVExpander, which we are using during code generation, only allows
instructions as insert locations, but breaks in case BasicBlock->end() iterators
are passed to it due to it trying to obtain the basic block in which code should
be generated by calling Instruction->getParent(), which is not defined for
->end() iterators.

This change adds an assert to Polly that ensures we only pass valid instructions
to SCEVExpander and it fixes one case, where we used IRBuilder->SetInsertBlock()
to set an ->end() insert location which was later passed to SCEVExpander.

In general, Polly is always trying to build up the CFG first, before we actually
insert instructions into the CFG sceleton. As a result, each basic block should
already have at least one branch instruction before we start adding code. Hence,
always requiring the IRBuilder insert location to be set to a real instruction
should always be possible.

Thanks Utpal Bora <cs14mtech11017@iith.ac.in> for his help with test case
reduction.

This is a backport from r243830 as it was committed on trunk.

llvm-svn: 246029

llvm-svn: 246013

llvm-svn: 245973

llvm-svn: 245972

llvm-svn: 245969

llvm-svn: 245968

By Mehdi Amini.

llvm-svn: 245948

------------------------------------------------------------------------
r245927 | slthakur | 2015-08-25 02:52:59 -0700 (Tue, 25 Aug 2015) | 9 lines

Fix build on mips

Setting and getting register values as bytes instead of depending on the 128 bit integer support in register value.
This patch will fix the build failure in the release branch.

Reviewers: tberghammer, clayborg, hans
Subscribers: bhushan, nitesh.jain, jaydeep, lldb-commits
Differential: http://reviews.llvm.org/D12275

------------------------------------------------------------------------

llvm-svn: 245947

------------------------------------------------------------------------
r245902 | hans | 2015-08-24 16:34:28 -0700 (Mon, 24 Aug 2015) | 3 lines

Revert r245355 "Release script: correctly symlink clang-tools-extra into the build (PR22765)"

This worked with the CMake build but broke the Autoconf one.
------------------------------------------------------------------------

llvm-svn: 245903

llvm-svn: 245857

Update CREDITS.TXT with Clang OpenMP implementation + test suite contributors from AMD, Argonne National Lab., IBM, Intel, Texas Instruments, University of Houston and many others. 

llvm-svn: 245845

------------------------------------------------------------------------

llvm-svn: 245623

------------------------------------------------------------------------
r245560 | ogoffart | 2015-08-20 06:11:14 -0700 (Thu, 20 Aug 2015) | 5 lines

Fix crash with two typos in the arguments of a function

The problem is that the arguments are of TheCall are reset later
to the ones in Args, making TypoExpr put back. Some TypoExpr that have
already  been diagnosed and will assert later in Sema::getTypoExprState
------------------------------------------------------------------------

llvm-svn: 245615

[ARM] Don't try and custom lower a vNi64 SETCC.

It won't go well. We've already marked 64-bit SETCCs as non-Custom, but it's
just possible that a SETCC has a legal result type but an illegal operand
type. If this happens, bail out before we create unselectable nodes.

Fixes PR24292. I tried to create a testcase but in 99% of cases we can't
trigger this - not surprising that this bug has been latent since 2009.

llvm-svn: 245578

------------------------------------------------------------------------
r245535 | hfinkel | 2015-08-19 20:02:02 -0700 (Wed, 19 Aug 2015) | 6 lines

[PowerPC] Fix value type on XVCMPEQDP for v2f64 comparisons

XVCMPEQDP is used for VSX v2f64 equality comparisons, but the value type needs
to be v2i64 (as that's the corresponding SETCC type).

Fixes PR24225.
------------------------------------------------------------------------

llvm-svn: 245574

------------------------------------------------------------------------
r245530 | hfinkel | 2015-08-19 18:18:20 -0700 (Wed, 19 Aug 2015) | 5 lines

[PowerPC] Fix the int2fp(fp2int(x)) DAGCombine to ignore ppc_fp128

This DAGCombine was creating custom SDAG nodes with an illegal ppc_fp128
operand type because it was triggering on f64/f32 int2fp(fp2int(ppc_fp128 x)),
but shouldn't (it should only apply to f32/f64 types). The result was a crash.
------------------------------------------------------------------------

llvm-svn: 245573

------------------------------------------------------------------------
r245365 | majnemer | 2015-08-18 15:07:25 -0700 (Tue, 18 Aug 2015) | 4 lines

[InstSimplify] Don't assume getAggregateElement will succeed

It isn't always possible to get a value from getAggregateElement.
This fixes PR24488.
------------------------------------------------------------------------

------------------------------------------------------------------------
r245369 | majnemer | 2015-08-18 15:18:22 -0700 (Tue, 18 Aug 2015) | 3 lines

[InstSimplify] Remove unused variable

No functionality change is intended.
------------------------------------------------------------------------

llvm-svn: 245572

I have underestimated the importance of this patch, and
James has got a fix for it in the making. Sorry for the noise.

llvm-svn: 245570

This reverts commit r229099 in branch 37 only, because it caused PR24292.
I'll continue investigating and will fix on trunk, but being an optimization
change, we can let the rest of the release go without this one.

llvm-svn: 245568

llvm-svn: 245534

------------------------------------------------------------------------
r245395 | qcolombet | 2015-08-18 17:08:26 -0700 (Tue, 18 Aug 2015) | 3 lines

[BasicAA] Add a test for PR24468 to be sure we won't regress
when we finally get the GEP aliasing right.

------------------------------------------------------------------------

llvm-svn: 245478

------------------------------------------------------------------------
r245394 | qcolombet | 2015-08-18 17:07:20 -0700 (Tue, 18 Aug 2015) | 3 lines

[BasicAA] Revert r221876 because it can produce incorrect aliasing
information: see PR24468.

------------------------------------------------------------------------

llvm-svn: 245476

------------------------------------------------------------------------
r244448 | fcormack | 2015-08-10 07:48:47 -0700 (Mon, 10 Aug 2015) | 12 lines

Prevent the scalarizer from caching incorrect entries

The scalarizer can cache incorrect entries when walking up a chain of
insertelement instructions. This occurs when it encounters more than one
instruction that it is not actively searching for, as it unconditionally caches
every element it finds. The fix is to only cache the first element that it
isn't searching for so we don't overwrite correct entries.

Reviewers: hfinkel

Differential Revision: http://reviews.llvm.org/D11559

------------------------------------------------------------------------

llvm-svn: 245457

------------------------------------------------------------------------
r245084 | martell | 2015-08-14 12:05:56 -0700 (Fri, 14 Aug 2015) | 13 lines

WindowsX86: long double is x87DoubleExtended on mingw

    Summary:
    long double on x86 mingw is 80bits and is aligned to 16bytes

	Fixes:
    https://llvm.org/bugs/show_bug.cgi?id=24398

    Reviewers: rnk

    Subscribers: cfe-commits

    Differential Revision: http://reviews.llvm.org/D12037
------------------------------------------------------------------------

llvm-svn: 245456

------------------------------------------------------------------------
r244902 | martell | 2015-08-13 08:41:04 -0700 (Thu, 13 Aug 2015) | 12 lines

Driver: Fix include directories when not using libgcc under mingw

Summary:
When we want to use mingw-w64 and clang with compiler-rt we should not
need to have libgcc installed. This fixes finding includes when libgcc
is not installed

Reviewers: yaron.keren

Subscribers: cfe-commits

Differential Revision: http://reviews.llvm.org/D11808
------------------------------------------------------------------------

llvm-svn: 245393

It avoids doing key function optimizations when the key function is
dllimported.

Fixed PR24409.

llvm-svn: 245388

It's actually probably a no-op, since the 3.7 branch will not generate
available externally vtables.

llvm-svn: 245387

------------------------------------------------------------------------
r245119 | nicholas | 2015-08-14 15:46:49 -0700 (Fri, 14 Aug 2015) | 2 lines

Fix a crash where a utility function wasn't aware of fcmp vectors and created a value with the wrong type. Fixes PR24458!

------------------------------------------------------------------------

llvm-svn: 245386

------------------------------------------------------------------------
r245355 | hans | 2015-08-18 14:10:17 -0700 (Tue, 18 Aug 2015) | 1 line

Release script: correctly symlink clang-tools-extra into the build (PR22765)
------------------------------------------------------------------------

llvm-svn: 245356

llvm-svn: 245347

llvm-svn: 245328

------------------------------------------------------------------------
r245259 | hans | 2015-08-17 16:38:56 -0700 (Mon, 17 Aug 2015) | 3 lines

Doxygen: add build option to use svg instead of png files for graphs

Differential Revision: http://reviews.llvm.org/D11994
------------------------------------------------------------------------

llvm-svn: 245313

------------------------------------------------------------------------
r245256 | hans | 2015-08-17 16:24:17 -0700 (Mon, 17 Aug 2015) | 3 lines

Doxygen: add build option to use svg instead of png files for graphs

Differential Revision: http://reviews.llvm.org/D11994
------------------------------------------------------------------------

llvm-svn: 245312

llvm-svn: 245308