vectorizing loops with memory accesses to non-zero address spaces. It
simply dropped the AS info. Fixes PR16306.
Merged from r184103
Author: Pekka Jaaskelainen <pekka.jaaskelainen@tut.fi>
Date: Mon Jun 17 18:49:06 2013 +0000
git-svn-id: https://llvm.org/svn/llvm-project/llvm/branches/release_33@185869 91177308-0d34-0410-b5e6-96231b3b80d8
- Also split out a section on build portability and updated the list of features and improvements.
--- This now reflects the up-coming merge of patches from trunk into the release_33 branch.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/branches/release_33@183554 91177308-0d34-0410-b5e6-96231b3b80d8
------------------------------------------------------------------------
r183060 | atrick | 2013-05-31 16:34:46 -0700 (Fri, 31 May 2013) | 11 lines
Prevent loop-unroll from making assumptions about undefined behavior.
Fixes rdar:14036816, PR16130.
There is an opportunity to compute precise trip counts for 'or'
expressions and multi-exit loops.
rdar:14038809: Optimize trip count computation for multi-exit loops.
To do this we need to record the fact that ExitLimit assumes NSW. When
it does not we can safely assume that the loop trip count is the
minimum ExitLimt across all subexpressions and loop exits.
------------------------------------------------------------------------
git-svn-id: https://llvm.org/svn/llvm-project/llvm/branches/release_33@183246 91177308-0d34-0410-b5e6-96231b3b80d8
------------------------------------------------------------------------
r182989 | atrick | 2013-05-30 23:43:25 -0700 (Thu, 30 May 2013) | 13 lines
Fix ScalarEvolution::ComputeExitLimitFromCond for 'or' conditions.
Fixes PR16130 - clang produces incorrect code with loop/expression at -O2.
This is a 2+ year old bug that's now holding up the release. It's a
case where we knowingly made aggressive assumptions about undefined
behavior. These assumptions are wrong when SCEV is computing a
subexpression that does not directly control the branch. With this
fix, we avoid making assumptions in those cases but still optimize the
common case. SCEV's trip count computation for exits controlled by
'or' expressions is now analagous to the trip count computation for
loops with multiple exits. I had already fixed the multiple exit case
to be conservative.
------------------------------------------------------------------------
git-svn-id: https://llvm.org/svn/llvm-project/llvm/branches/release_33@183245 91177308-0d34-0410-b5e6-96231b3b80d8
------------------------------------------------------------------------
r183035 | arnolds | 2013-05-31 12:53:50 -0700 (Fri, 31 May 2013) | 7 lines
LoopVectorize: PHIs with only outside users should prevent vectorization
We check that instructions in the loop don't have outside users (except if
they are reduction values). Unfortunately, we skipped this check for
if-convertable PHIs.
Fixes PR16184.
------------------------------------------------------------------------
git-svn-id: https://llvm.org/svn/llvm-project/llvm/branches/release_33@183189 91177308-0d34-0410-b5e6-96231b3b80d8
Fixes rdar:14036816, PR16130.
There is an opportunity to compute precise trip counts for 'or'
expressions and multi-exit loops.
rdar:14038809: Optimize trip count computation for multi-exit loops.
To do this we need to record the fact that ExitLimit assumes NSW. When
it does not we can safely assume that the loop trip count is the
minimum ExitLimt across all subexpressions and loop exits.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@183060 91177308-0d34-0410-b5e6-96231b3b80d8
git-svn-id: https://llvm.org/svn/llvm-project/llvm/branches/release_33@183066 91177308-0d34-0410-b5e6-96231b3b80d8
Fixes PR16130 - clang produces incorrect code with loop/expression at -O2.
This is a 2+ year old bug that's now holding up the release. It's a
case where we knowingly made aggressive assumptions about undefined
behavior. These assumptions are wrong when SCEV is computing a
subexpression that does not directly control the branch. With this
fix, we avoid making assumptions in those cases but still optimize the
common case. SCEV's trip count computation for exits controlled by
'or' expressions is now analagous to the trip count computation for
loops with multiple exits. I had already fixed the multiple exit case
to be conservative.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@182989 91177308-0d34-0410-b5e6-96231b3b80d8
git-svn-id: https://llvm.org/svn/llvm-project/llvm/branches/release_33@183013 91177308-0d34-0410-b5e6-96231b3b80d8
------------------------------------------------------------------------
r182585 | aaronballman | 2013-05-23 07:55:00 -0700 (Thu, 23 May 2013) | 1 line
Setting the default value (fixes CRT assertions about uninitialized variable use when doing debug MSVC builds), and fixing coding style.
------------------------------------------------------------------------
git-svn-id: https://llvm.org/svn/llvm-project/llvm/branches/release_33@182830 91177308-0d34-0410-b5e6-96231b3b80d8
------------------------------------------------------------------------
r182298 | jholewinski | 2013-05-20 09:42:18 -0700 (Mon, 20 May 2013) | 1 line
[NVPTX] Fix mis-use of CurrentFnSym in NVPTXAsmPrinter. This was causing a symbol name error in the output PTX.
------------------------------------------------------------------------
git-svn-id: https://llvm.org/svn/llvm-project/llvm/branches/release_33@182828 91177308-0d34-0410-b5e6-96231b3b80d8
------------------------------------------------------------------------
r182254 | jholewinski | 2013-05-20 05:13:32 -0700 (Mon, 20 May 2013) | 12 lines
[NVPTX] Add GenericToNVVM IR converter to better handle idiomatic LLVM IR inputs
This converter currently only handles global variables in address space 0. For
these variables, they are promoted to address space 1 (global memory), and all
uses are updated to point to the result of a cvta.global instruction on the new
variable.
The motivation for this is address space 0 global variables are illegal since we
cannot declare variables in the generic address space. Instead, we place the
variables in address space 1 and explicitly convert the pointer to address
space 0. This is primarily intended to help new users who expect to be able to
place global variables in the default address space.
------------------------------------------------------------------------
git-svn-id: https://llvm.org/svn/llvm-project/llvm/branches/release_33@182826 91177308-0d34-0410-b5e6-96231b3b80d8
------------------------------------------------------------------------
r182253 | jholewinski | 2013-05-20 05:13:28 -0700 (Mon, 20 May 2013) | 1 line
[NVPTX] Fix i1 kernel parameters and global variables. ABI rules say we need to use .u8 for i1 parameters for kernels.
------------------------------------------------------------------------
git-svn-id: https://llvm.org/svn/llvm-project/llvm/branches/release_33@182825 91177308-0d34-0410-b5e6-96231b3b80d8
------------------------------------------------------------------------
r182485 | arnolds | 2013-05-22 09:54:56 -0700 (Wed, 22 May 2013) | 7 lines
LoopVectorize: Make Value pointers that could be RAUW'ed a VH
The Value pointers we store in the induction variable list can be RAUW'ed by a
call to SCEVExpander::expandCodeFor, use a TrackingVH instead. Do the same thing
in some other places where we store pointers that could potentially be RAUW'ed.
Fixes PR16073.
------------------------------------------------------------------------
git-svn-id: https://llvm.org/svn/llvm-project/llvm/branches/release_33@182492 91177308-0d34-0410-b5e6-96231b3b80d8
------------------------------------------------------------------------
r182486 | d0k | 2013-05-22 10:01:12 -0700 (Wed, 22 May 2013) | 3 lines
X86: When expanding PCMPGTQ to PCMPGTD we always want to compare the lower halves as unsigned.
Take #2 on fixing PR15977.
------------------------------------------------------------------------
git-svn-id: https://llvm.org/svn/llvm-project/llvm/branches/release_33@182489 91177308-0d34-0410-b5e6-96231b3b80d8
------------------------------------------------------------------------
r181366 | enderby | 2013-05-07 14:40:58 -0700 (Tue, 07 May 2013) | 6 lines
Fix a bug in the MC asm parser evaluating expressions. It was treating:
A = 9
B = 3 * A - 2 * A + 1 as B = 3 * A - (2 * A + 1)
rdar://13816516
------------------------------------------------------------------------
git-svn-id: https://llvm.org/svn/llvm-project/llvm/branches/release_33@182438 91177308-0d34-0410-b5e6-96231b3b80d8
------------------------------------------------------------------------
r182385 | hfinkel | 2013-05-21 07:21:09 -0700 (Tue, 21 May 2013) | 9 lines
Fix PPC branch selection for counter-based branches
Although I had added some support for the BDZ/BDNZ branches into the selector
(in r158204), I had not correctly adjusted the condition at the top of the
loop. As a result, these branches were still essentially unsupported.
This fixes PR16086. Unfortunately, any test case would be very large (because
it would need to force the loop backedge to exceed the range of the 16-bit
immediate).
------------------------------------------------------------------------
git-svn-id: https://llvm.org/svn/llvm-project/llvm/branches/release_33@182431 91177308-0d34-0410-b5e6-96231b3b80d8
------------------------------------------------------------------------
r182387 | jholewinski | 2013-05-21 07:37:16 -0700 (Tue, 21 May 2013) | 7 lines
Drop @llvm.annotation and @llvm.ptr.annotation intrinsics during codegen.
The intrinsic calls are dropped, but the annotated value is propagated.
Fixes PR 15253
Original patch by Zeng Bin!
------------------------------------------------------------------------
git-svn-id: https://llvm.org/svn/llvm-project/llvm/branches/release_33@182417 91177308-0d34-0410-b5e6-96231b3b80d8
------------------------------------------------------------------------
r182364 | d0k | 2013-05-21 02:58:54 -0700 (Tue, 21 May 2013) | 4 lines
X86: When emulating unsigned PCMPGTQ with PCMPGTD, fix the sign bit for the smaller type.
Otherwise we'll get a mix of signed and unsigned compares.
Fixes PR15977.
------------------------------------------------------------------------
git-svn-id: https://llvm.org/svn/llvm-project/llvm/branches/release_33@182413 91177308-0d34-0410-b5e6-96231b3b80d8
------------------------------------------------------------------------
r182344 | mren | 2013-05-20 17:57:22 -0700 (Mon, 20 May 2013) | 7 lines
Dwarf: use a single line table to generate assembly when .loc is used.
This is to fix PR15408 where an undefined symbol Lline_table_start1 is used.
Since we do not generate the debug_line section when .loc is used,
Lline_table_start1 is not emitted and we can't refer to it when calculating
at_stmt_list for a compile unit.
------------------------------------------------------------------------
git-svn-id: https://llvm.org/svn/llvm-project/llvm/branches/release_33@182346 91177308-0d34-0410-b5e6-96231b3b80d8
------------------------------------------------------------------------
r181864 | chapuni | 2013-05-14 19:16:23 -0700 (Tue, 14 May 2013) | 10 lines
ELFRelocationEntry::operator<(): Try to stabilize the order. r_offset was insufficient to sort Relocs.
It should fix llvm/test/CodeGen/ARM/ehabi-mc-compact-pr*.ll on some hosts.
RELOCATION RECORDS FOR [.ARM.exidx]:
0 R_ARM_PREL31 .text
0 R_ARM_NONE __aeabi_unwind_cpp_pr0
FIXME: I am not sure of the directions of extra comparators, in Type and Index.
For now, they are different from the direction in r_offset.
------------------------------------------------------------------------
git-svn-id: https://llvm.org/svn/llvm-project/llvm/branches/release_33@182149 91177308-0d34-0410-b5e6-96231b3b80d8
------------------------------------------------------------------------
r181678 | rafael | 2013-05-12 17:18:24 -0700 (Sun, 12 May 2013) | 1 line
XFAIL this test for mingw too.
------------------------------------------------------------------------
git-svn-id: https://llvm.org/svn/llvm-project/llvm/branches/release_33@182146 91177308-0d34-0410-b5e6-96231b3b80d8
------------------------------------------------------------------------
r181600 | aaronballman | 2013-05-10 07:42:16 -0700 (Fri, 10 May 2013) | 1 line
XFAILing this test on Win32 to unbreak the build bots.
------------------------------------------------------------------------
git-svn-id: https://llvm.org/svn/llvm-project/llvm/branches/release_33@182145 91177308-0d34-0410-b5e6-96231b3b80d8
------------------------------------------------------------------------
r181529 | void | 2013-05-09 11:21:45 -0700 (Thu, 09 May 2013) | 8 lines
Simplify the code a bit.
The compact unwind registers were defined in two different
places. It's better just to place them in the function that uses them
and specify that this is a 64-bit or 32-bit machine.
No functionality change.
------------------------------------------------------------------------
git-svn-id: https://llvm.org/svn/llvm-project/llvm/branches/release_33@181956 91177308-0d34-0410-b5e6-96231b3b80d8
------------------------------------------------------------------------
r181540 | void | 2013-05-09 13:10:38 -0700 (Thu, 09 May 2013) | 11 lines
Generate a compact unwind encoding in the face of a stack alignment push.
We generate a `push' of a random register (%rax) if the stack needs to be
aligned by the size of that register. However, this could mess up compact unwind
generation. In particular, we want to still generate compact unwind in the
presence of this monstrosity.
Check if the push of of the %rax/%eax register. If it is and it's marked with
the `FrameSetup' flag, then we can generate a compact unwind encoding for the
function only if the push is the last FrameSetup instruction.
------------------------------------------------------------------------
git-svn-id: https://llvm.org/svn/llvm-project/llvm/branches/release_33@181955 91177308-0d34-0410-b5e6-96231b3b80d8
------------------------------------------------------------------------
r181580 | tstellar | 2013-05-09 19:09:45 -0700 (Thu, 09 May 2013) | 10 lines
R600: Remove AMDILPeeopholeOptimizer and replace optimizations with tablegen patterns
The BFE optimization was the only one we were actually using, and it was
emitting an intrinsic that we don't support.
https://bugs.freedesktop.org/show_bug.cgi?id=64201
Reviewed-by: Christian König <christian.koenig@amd.com>
NOTE: This is a candidate for the 3.3 branch.
------------------------------------------------------------------------
git-svn-id: https://llvm.org/svn/llvm-project/llvm/branches/release_33@181954 91177308-0d34-0410-b5e6-96231b3b80d8
------------------------------------------------------------------------
r181579 | tstellar | 2013-05-09 19:09:39 -0700 (Thu, 09 May 2013) | 8 lines
R600: Expand SUB for v2i32/v4i32
Patch by: Aaron Watry
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
Signed-off-by: Aaron Watry <awatry@gmail.com>
NOTE: This is a candidate for the 3.3 branch.
------------------------------------------------------------------------
git-svn-id: https://llvm.org/svn/llvm-project/llvm/branches/release_33@181953 91177308-0d34-0410-b5e6-96231b3b80d8
------------------------------------------------------------------------
r181578 | tstellar | 2013-05-09 19:09:34 -0700 (Thu, 09 May 2013) | 10 lines
R600: Expand MUL for v4i32/v2i32
Fixes piglit test for OpenCL builtin mul24, and allows mad24 to run.
Patch by: Aaron Watry
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
Signed-off-by: Aaron Watry <awatry@gmail.com>
NOTE: This is a candidate for the 3.3 branch.
------------------------------------------------------------------------
git-svn-id: https://llvm.org/svn/llvm-project/llvm/branches/release_33@181952 91177308-0d34-0410-b5e6-96231b3b80d8
------------------------------------------------------------------------
r181792 | tstellar | 2013-05-14 07:42:56 -0700 (Tue, 14 May 2013) | 8 lines
R600/SI: Add processor type for Hainan asic
Patch by: Alex Deucher
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
NOTE: This is a candidate for the 3.3 branch.
------------------------------------------------------------------------
git-svn-id: https://llvm.org/svn/llvm-project/llvm/branches/release_33@181949 91177308-0d34-0410-b5e6-96231b3b80d8
------------------------------------------------------------------------
r181842 | arnolds | 2013-05-14 15:33:24 -0700 (Tue, 14 May 2013) | 14 lines
ARM ISel: Don't create illegal types during LowerMUL
The transformation happening here is that we want to turn a
"mul(ext(X), ext(X))" into a "vmull(X, X)", stripping off the extension. We have
to make sure that X still has a valid vector type - possibly recreate an
extension to a smaller type. In case of a extload of a memory type smaller than
64 bit we used create a ext(load()). The problem with doing this - instead of
recreating an extload - is that an illegal type is exposed.
This patch fixes this by creating extloads instead of ext(load()) sequences.
Fixes PR15970.
radar://13871383
------------------------------------------------------------------------
git-svn-id: https://llvm.org/svn/llvm-project/llvm/branches/release_33@181946 91177308-0d34-0410-b5e6-96231b3b80d8
------------------------------------------------------------------------
r181524 | rafael | 2013-05-09 10:22:59 -0700 (Thu, 09 May 2013) | 4 lines
Don't replace an alias in llvm.used with its target.
When we replace an internal alias with its target, be careful not to
replace the entry in llvm.used (and llvm.compiler_used).
------------------------------------------------------------------------
git-svn-id: https://llvm.org/svn/llvm-project/llvm/branches/release_33@181909 91177308-0d34-0410-b5e6-96231b3b80d8
