This patch is a follow up from r240560 and is a step further into
mitigating the compile time performance issues in CaptureTracker.
By providing the CaptureTracker with a "cached ordered basic block"
instead of computing it every time, MemDepAnalysis can use this cache
throughout its calls to AA->callCapturesBefore, avoiding to recompute it
for every scanned instruction. In the same testcase used in r240560,
compile time is reduced from 2min to 30s.
This also fixes PR22348.
rdar://problem/19230319
Differential Revision: http://reviews.llvm.org/D11364
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@243750 91177308-0d34-0410-b5e6-96231b3b80d8
Reapply 243271 with more fixes; although we are not handling multiple
sources with coalescable copies, we were not properly skipping this
case.
- Teaches the ValueTracker in the PeepholeOptimizer to look through PHI
instructions.
- Add findNextSourceAndRewritePHI method to lookup into multiple sources
returnted by the ValueTracker and rewrite PHIs with new sources.
With these changes we can find more register sources and rewrite more
copies to allow coaslescing of bitcast instructions. Hence, we eliminate
unnecessary VR64 <-> GR64 copies in x86, but it could be extended to
other archs by marking "isBitcast" on target specific instructions. The
x86 example follows:
A:
psllq %mm1, %mm0
movd %mm0, %r9
jmp C
B:
por %mm1, %mm0
movd %mm0, %r9
jmp C
C:
movd %r9, %mm0
pshufw $238, %mm0, %mm0
Becomes:
A:
psllq %mm1, %mm0
jmp C
B:
por %mm1, %mm0
jmp C
C:
pshufw $238, %mm0, %mm0
Differential Revision: http://reviews.llvm.org/D11197
rdar://problem/20404526
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@243486 91177308-0d34-0410-b5e6-96231b3b80d8
Reapply r242295 with fixes in the implementation.
- Teaches the ValueTracker in the PeepholeOptimizer to look through PHI
instructions.
- Add findNextSourceAndRewritePHI method to lookup into multiple sources
returnted by the ValueTracker and rewrite PHIs with new sources.
With these changes we can find more register sources and rewrite more
copies to allow coaslescing of bitcast instructions. Hence, we eliminate
unnecessary VR64 <-> GR64 copies in x86, but it could be extended to
other archs by marking "isBitcast" on target specific instructions. The
x86 example follows:
A:
psllq %mm1, %mm0
movd %mm0, %r9
jmp C
B:
por %mm1, %mm0
movd %mm0, %r9
jmp C
C:
movd %r9, %mm0
pshufw $238, %mm0, %mm0
Becomes:
A:
psllq %mm1, %mm0
jmp C
B:
por %mm1, %mm0
jmp C
C:
pshufw $238, %mm0, %mm0
Differential Revision: http://reviews.llvm.org/D11197
rdar://problem/20404526
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@243271 91177308-0d34-0410-b5e6-96231b3b80d8
Reapply r242294.
- Create a new CopyRewriter for Uncoalescable copy-like instructions
- Change the ValueTracker to return a ValueTrackerResult
This makes optimizeUncoalescable looks more like optimizeCoalescable and
use the CopyRewritter infrastructure.
This is also the preparation for looking up into PHI nodes in the
ValueTracker.
rdar://problem/20404526
Differential Revision: http://reviews.llvm.org/D11195
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@242940 91177308-0d34-0410-b5e6-96231b3b80d8
Check whether BaseCst is valid before extracting a GlobalValue.
This fixes PR24163.
Patch by David Majnemer.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@242840 91177308-0d34-0410-b5e6-96231b3b80d8
- Teaches the ValueTracker in the PeepholeOptimizer to look through PHI
instructions.
- Add findNextSourceAndRewritePHI method to lookup into multiple sources
returnted by the ValueTracker and rewrite PHIs with new sources.
With these changes we can find more register sources and rewrite more
copies to allow coaslescing of bitcast instructions. Hence, we eliminate
unnecessary VR64 <-> GR64 copies in x86, but it could be extended to
other archs by marking "isBitcast" on target specific instructions. The
x86 example follows:
A:
psllq %mm1, %mm0
movd %mm0, %r9
jmp C
B:
por %mm1, %mm0
movd %mm0, %r9
jmp C
C:
movd %r9, %mm0
pshufw $238, %mm0, %mm0
Becomes:
A:
psllq %mm1, %mm0
jmp C
B:
por %mm1, %mm0
jmp C
C:
pshufw $238, %mm0, %mm0
Differential Revision: http://reviews.llvm.org/D11197
rdar://problem/20404526
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@242295 91177308-0d34-0410-b5e6-96231b3b80d8
- Create a new CopyRewriter for Uncoalescable copy-like instructions
- Change the ValueTracker to return a ValueTrackerResult
This makes optimizeUncoalescable looks more like optimizeCoalescable and
use the CopyRewritter infrastructure.
This is also the preparation for looking up into PHI nodes in the
ValueTracker.
Differential Revision: http://reviews.llvm.org/D11195
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@242294 91177308-0d34-0410-b5e6-96231b3b80d8
MOVSDto64rr and MOV64toSDrr are defined to convert between FR64 (%xmm)
<-> GR64 registers, not VR64 (%mm) <-> GR64. This is wrong.
I found this by inspection and could not find a suitable testcase for it
since (1) we don't handle MMX bitcasts in Peephole optimizer as to
generate COPYs that (2) could be expanded back to the appropriate x86
instruction in ExpandPostRA.
Switch to use the appropriate instructions: MMX_MOVD64from64rr and
MMX_MOVD64to64rr here.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@242191 91177308-0d34-0410-b5e6-96231b3b80d8
CaptureTracking becomes very expensive in large basic blocks while
calling PointerMayBeCaptured. PointerMayBeCaptured scans the BB the
number of times equal to the number of uses of 'BeforeHere', which is
currently capped at 20 and bails out with Tracker->tooManyUses().
The bottleneck here is the number of calls to PointerMayBeCaptured * the
basic block scan. In a testcase with a 82k instruction BB,
PointerMayBeCaptured is called 130k times, leading to 'shouldExplore'
taking 527k runs, this currently takes ~12min.
To fix this we locally (within PointerMayBeCaptured) number the
instructions in the basic block using a DenseMap to cache instruction
positions/numbers. We build the cache incrementally every time we need
to scan an unexplored part of the BB, improving compile time to only
take ~2min.
This triggers in the flow: DeadStoreElimination -> MepDepAnalysis ->
CaptureTracking.
Side note: after multiple runs in the test-suite I've seen no
performance nor compile time regressions, but could note a couple of
compile time improvements:
Performance Improvements - Compile Time Delta Previous Current StdDev
SingleSource/Benchmarks/Misc-C++/bigfib -4.48% 0.8547 0.8164 0.0022
MultiSource/Benchmarks/TSVC/LoopRerolling-dbl/LoopRerolling-dbl -1.47% 1.3912 1.3707 0.0056
Differential Revision: http://reviews.llvm.org/D7010
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@240560 91177308-0d34-0410-b5e6-96231b3b80d8
Follow up from r231505.
Fix the non-determinism by using a MapVector and reintroduce the AArch64
testcase. Defer deleting the got candidates up to the end and remove
them in a bulk, avoiding linear time removal of each element.
Thanks to Renato Golin for trying it out on other platforms.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@231830 91177308-0d34-0410-b5e6-96231b3b80d8
Follow up from r231497. Using XFAIL would still trigger fail on some
buildbots. Will re-introduce it as soon as I have a fix.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@231505 91177308-0d34-0410-b5e6-96231b3b80d8
The checking for extgotequiv and localgotequiv rely on the emission
order, which is not guaranteed because we use DenseMap to hold the GOT
equivalents. XFAIL this now until I get time to use MapVector and test
out the solution. In the meantime, appease buildbots.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@231497 91177308-0d34-0410-b5e6-96231b3b80d8
Add MachO 32-bit (i.e. arm and x86) support for replacing global GOT equivalent
symbol accesses. Unlike 64-bit targets, there's no GOTPCREL relocation, and
access through a non_lazy_symbol_pointers section is used instead.
-- before
_extgotequiv:
.long _extfoo
_delta:
.long _extgotequiv-_delta
-- after
_delta:
.long L_extfoo$non_lazy_ptr-_delta
.section __IMPORT,__pointers,non_lazy_symbol_pointers
L_extfoo$non_lazy_ptr:
.indirect_symbol _extfoo
.long 0
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@231475 91177308-0d34-0410-b5e6-96231b3b80d8
Follow up r230264 and add ARM64 support for replacing global GOT
equivalent symbol accesses by references to the GOT entry for the final
symbol instead, example:
-- before
.globl _foo
_foo:
.long 42
.globl _gotequivalent
_gotequivalent:
.quad _foo
.globl _delta
_delta:
.long _gotequivalent-_delta
-- after
.globl _foo
_foo:
.long 42
.globl _delta
Ltmp3:
.long _foo@GOT-Ltmp3
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@231474 91177308-0d34-0410-b5e6-96231b3b80d8
Turns out that after the past MMX commits, we don't need to rely on this
flag to get better codegen for MMX. Also update the tests to become
triple neutral.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@230637 91177308-0d34-0410-b5e6-96231b3b80d8
Reapply r230248.
Teach the peephole optimizer to work with MMX instructions by adding
entries into the foldable tables. This covers folding opportunities not
handled during isel.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@230499 91177308-0d34-0410-b5e6-96231b3b80d8
MMX_MOVD64rm zero-extends i32 load results into i64 registers.
The peephole optimizer will try to fold it in other MMX foldable
instructions, the wrong thing to do, since there's no MMX memory
instruction that loads from i32 and does implict zero extension.
Remove 'canFoldAsLoad' from MOVD64rm in order to prevent such folding.
The current MMX tests already test this, but since there are no MMX
instructions in the foldable tables yet, this did not trigger. This
commit prepares the addition of those instructions.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@230498 91177308-0d34-0410-b5e6-96231b3b80d8
Front-ends could use global unnamed_addr to hold pointers to other
symbols, like @gotequivalent below:
@foo = global i32 42
@gotequivalent = private unnamed_addr constant i32* @foo
@delta = global i32 trunc (i64 sub (i64 ptrtoint (i32** @gotequivalent to i64),
i64 ptrtoint (i32* @delta to i64))
to i32)
The global @delta holds a data "PC"-relative offset to @gotequivalent,
an unnamed pointer to @foo. The darwin/x86-64 assembly output for this follows:
.globl _foo
_foo:
.long 42
.globl _gotequivalent
_gotequivalent:
.quad _foo
.globl _delta
_delta:
.long _gotequivalent-_delta
Since unnamed_addr indicates that the address is not significant, only
the content, we can optimize the case above by replacing pc-relative
accesses to "GOT equivalent" globals, by a PC relative access to the GOT
entry of the final symbol instead. Therefore, "delta" can contain a pc
relative relocation to foo's GOT entry and we avoid the emission of
"gotequivalent", yielding the assembly code below:
.globl _foo
_foo:
.long 42
.globl _delta
_delta:
.long _foo@GOTPCREL+4
There are a couple of advantages of doing this: (1) Front-ends that need
to emit a great deal of data to store pointers to external symbols could
save space by not emitting such "got equivalent" globals and (2) IR
constructs combined with this opt opens a way to represent GOT pcrel
relocations by using the LLVM IR, which is something we previously had
no way to express.
Differential Revision: http://reviews.llvm.org/D6922
rdar://problem/18534217
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@230264 91177308-0d34-0410-b5e6-96231b3b80d8
This test failed in several buildbots, a bit unclear how that happen
since this was the previous behavior before r230248.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@230258 91177308-0d34-0410-b5e6-96231b3b80d8
Teach the peephole optimizer to work with MMX instructions by adding
entries into the foldable tables. This covers folding opportunities not
handled during isel.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@230226 91177308-0d34-0410-b5e6-96231b3b80d8
Add tests to cover the RR form of the pslli, psrli and psrai intrinsics.
In the next commit, the loads are going to be folded and the
instructions use the RM form.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@230224 91177308-0d34-0410-b5e6-96231b3b80d8
Implement a BITCAST dag combine to transform i32->mmx conversion patterns
into a X86 specific node (MMX_MOVW2D) and guarantee that moves between
i32 and x86mmx are better handled, i.e., don't use store-load to do the
conversion..
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@228293 91177308-0d34-0410-b5e6-96231b3b80d8
Avoid regression in previously supported MMX code by adding different
combinations of tests which exercise MMX bitcasts. Small improvements
to these patterns should come next.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@228292 91177308-0d34-0410-b5e6-96231b3b80d8
Improve EXTRACT_VECTOR_ELT DAG combine to catch conversion patterns
between x86mmx and i32 with more layers of indirection.
Before:
movq2dq %mm0, %xmm0
movd %xmm0, %eax
After:
movd %mm0, %eax
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@227969 91177308-0d34-0410-b5e6-96231b3b80d8
LLVM ToT produces poor MMX code compared to 3.5. However, part of the previous
functionality can be achieved by using -x86-experimental-vector-widening-legalization.
Add tests to be sure we don't regress again.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@227869 91177308-0d34-0410-b5e6-96231b3b80d8
- Rename mmx-builtins to mmx-intrinsics to match other intrinsic test naming.
- Remove tests that duplicate functionality from mmx-intrinsics.ll.
- Move arith related tests to mmx-arith.ll.
- MMX related shuffle goes to vector-shuffle-mmx.ll.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@227130 91177308-0d34-0410-b5e6-96231b3b80d8
Handle the poor codegen for i64/x86xmm->v2i64 (%mm -> %xmm) moves. Instead of
using stack store/load pair to do the job, use scalar_to_vector directly, which
in the MMX case can use movq2dq. This was the current behavior prior to
improvements for vector legalization of extloads in r213897.
This commit fixes the regression and as a side-effect also remove some
unnecessary shuffles.
In the new attached testcase, we go from:
pshufw $-18, (%rdi), %mm0
movq %mm0, -8(%rsp)
movq -8(%rsp), %xmm0
pshufd $-44, %xmm0, %xmm0
movd %xmm0, %eax
...
To:
pshufw $-18, (%rdi), %mm0
movq2dq %mm0, %xmm0
movd %xmm0, %eax
...
Differential Revision: http://reviews.llvm.org/D7126
rdar://problem/19413324
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@226953 91177308-0d34-0410-b5e6-96231b3b80d8
Take two disjoint Loops L1 and L2.
LoopSimplify fails to simplify some loops (e.g. when indirect branches
are involved). In such situations, it can happen that an exit for L1 is
the header of L2. Thus, when we create PHIs in one of such exits we are
also inserting PHIs in L2 header.
This could break LCSSA form for L2 because these inserted PHIs can also
have uses in L2 exits, which are never handled in the current
implementation. Provide a fix for this corner case and test that we
don't assert/crash on that.
Differential Revision: http://reviews.llvm.org/D6624
rdar://problem/19166231
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@224740 91177308-0d34-0410-b5e6-96231b3b80d8
Currently, when ctpop is supported for scalar types, the expansion of
@llvm.ctpop.vXiY uses vector element extractions, insertions and individual
calls to @llvm.ctpop.iY. When not, expansion with bit-math operations is used
for the scalar calls.
Local haswell measurements show that we can improve vector @llvm.ctpop.vXiY
expansion in some cases by using a using a vector parallel bit twiddling
approach, based on:
v = v - ((v >> 1) & 0x55555555);
v = (v & 0x33333333) + ((v >> 2) & 0x33333333);
v = ((v + (v >> 4) & 0xF0F0F0F)
v = v + (v >> 8)
v = v + (v >> 16)
v = v & 0x0000003F
(from http://graphics.stanford.edu/~seander/bithacks.html#CountBitsSetParallel)
When scalar ctpop isn't supported, the approach above performs better for
v2i64, v4i32, v4i64 and v8i32 (see numbers below). And even when scalar ctpop
is supported, this approach performs ~2x better for v8i32.
Here, x86_64 implies -march=corei7-avx without ctpop and x86_64h includes ctpop
support with -march=core-avx2.
== [x86_64h - new]
v8i32: 0.661685
v4i32: 0.514678
v4i64: 0.652009
v2i64: 0.324289
== [x86_64h - old]
v8i32: 1.29578
v4i32: 0.528807
v4i64: 0.65981
v2i64: 0.330707
== [x86_64 - new]
v8i32: 1.003
v4i32: 0.656273
v4i64: 1.11711
v2i64: 0.754064
== [x86_64 - old]
v8i32: 2.34886
v4i32: 1.72053
v4i64: 1.41086
v2i64: 1.0244
More work for other vector types will come next.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@224725 91177308-0d34-0410-b5e6-96231b3b80d8
The visitSwitchInst generates SUB constant expressions to recompute the
switch condition. When truncating the condition to a smaller type, SUB
expressions should use the previous type (before trunc) for both
operands. Also, fix code to also return the modified switch when only
the truncation is performed.
This fixes an assertion crash.
Differential Revision: http://reviews.llvm.org/D6644
rdar://problem/19191835
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@224588 91177308-0d34-0410-b5e6-96231b3b80d8
Reverts commit r224574 to appease buildbots:
The visitSwitchInst generates SUB constant expressions to recompute the
switch condition. When truncating the condition to a smaller type, SUB
expressions should use the previous type (before trunc) for both
operands. This fixes an assertion crash.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@224576 91177308-0d34-0410-b5e6-96231b3b80d8
The visitSwitchInst generates SUB constant expressions to recompute the
switch condition. When truncating the condition to a smaller type, SUB
expressions should use the previous type (before trunc) for both
operands. This fixes an assertion crash.
Differential Revision: http://reviews.llvm.org/D6644
rdar://problem/19191835
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@224574 91177308-0d34-0410-b5e6-96231b3b80d8
Fix a compact unwind encoding logic bug which would try to encode
more callee saved registers than it should, leading to early bail out
in the encoding logic and abusive use of DWARF frame mode unnecessarily.
Also remove no-compact-unwind.ll which was testing the wrong thing
based on this bug and move it to valid 'compact unwind' tests. Added
other few more tests too.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223676 91177308-0d34-0410-b5e6-96231b3b80d8
Offset == 0 is a valid offset for kernel code model according to the
x86_64 System V ABI. Found by inspection, no testcase.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223383 91177308-0d34-0410-b5e6-96231b3b80d8
Follow up from r222926. Also handle multiple destinations from merged
cases on multiple and subsequent phi instructions.
rdar://problem/19106978
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223135 91177308-0d34-0410-b5e6-96231b3b80d8
Load instructions are inserted into loop preheaders when sinking stores
and later removed if not used by the SSA updater. Avoid sinking if the
loop has no preheader and avoid crashes. This fixes one more side effect
of not handling indirectbr instructions properly on LoopSimplify.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@223119 91177308-0d34-0410-b5e6-96231b3b80d8
Loop simplify skips exit-block insertion when exits contain indirectbr
instructions. This leads to an assertion in LICM when trying to sink
stores out of non-dedicated loop exits containing indirectbr
instructions. This patch fix this issue by re-checking for dedicated
exits in LICM prior to store sink attempts.
Differential Revision: http://reviews.llvm.org/D6414
rdar://problem/18943047
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@222927 91177308-0d34-0410-b5e6-96231b3b80d8
Switch cases statements with sequential values that branch to the same
destination BB may often be handled together in a single new source BB.
In this scenario we need to remove remaining incoming values from PHI
instructions in the destination BB, as to match the number of source
branches.
Differential Revision: http://reviews.llvm.org/D6415
rdar://problem/19040894
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@222926 91177308-0d34-0410-b5e6-96231b3b80d8
ConstantFolding crashes when trying to InstSimplify the following load:
@a = private unnamed_addr constant %mst {
i8* inttoptr (i64 -1 to i8*),
i8* inttoptr (i64 -1 to i8*)
}, align 8
%x = load <2 x i8*>* bitcast (%mst* @a to <2 x i8*>*), align 8
This patch fix this by adding support to this type of folding:
%x = load <2 x i8*>* bitcast (%mst* @a to <2 x i8*>*), align 8
==> gets folded to:
%x = <2 x i8*> <i8* inttoptr (i64 -1 to i8*), i8* inttoptr (i64 -1 to i8*)>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220380 91177308-0d34-0410-b5e6-96231b3b80d8
- Problem
One program takes ~3min to compile under -O2. This happens after a certain
function A is inlined ~700 times in a function B, inserting thousands of new
BBs. This leads to 80% of the compilation time spent in
GVN::processNonLocalLoad and
MemoryDependenceAnalysis::getNonLocalPointerDependency, while searching for
nonlocal information for basic blocks.
Usually, to avoid spending a long time to process nonlocal loads, GVN bails out
if it gets more than 100 deps as a result from
MD->getNonLocalPointerDependency. However this only happens *after* all
nonlocal information for BBs have been computed, which is the bottleneck in
this scenario. For instance, there are 8280 times where
getNonLocalPointerDependency returns deps with more than 100 bbs and from
those, 600 times it returns more than 1000 blocks.
- Solution
Bail out early during the nonlocal info computation whenever we reach a
specified threshold. This patch proposes a 100 BBs threshold, it also
reduces the compile time from 3min to 23s.
- Testing
The test-suite presented no compile nor execution time regressions.
Some numbers from my machine (x86_64 darwin):
- 17s under -Oz (which avoids inlining).
- 1.3s under -O1.
- 2m51s under -O2 ToT
*** 23s under -O2 w/ Result.size() > 100
- 1m54s under -O2 w/ Result.size() > 500
With NumResultsLimit = 100, GVN yields the same outcome as in the
unlimited 3min version.
http://reviews.llvm.org/D5532
rdar://problem/18188041
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@218792 91177308-0d34-0410-b5e6-96231b3b80d8
Machine Sink uses loop depth information to select between successors BBs to
sink machine instructions into, where BBs within smaller loop depths are
preferable. This patch adds support for choosing between successors by using
profile information from BlockFrequencyInfo instead, whenever the information
is available.
Tested it under SPEC2006 train (average of 30 runs for each program); ~1.5%
execution speedup in average on x86-64 darwin.
<rdar://problem/18021659>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@218472 91177308-0d34-0410-b5e6-96231b3b80d8
GlobalDCE deletes global vars and updates their initializers to nullptr
while leaving underlying constants to be cleaned up later by its uses.
The clean up may never happen, fix this by forcing it every time it's
safe to destroy constants.
Final patch by Rafael Espindola
http://reviews.llvm.org/D4931
<rdar://problem/17523868>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@216390 91177308-0d34-0410-b5e6-96231b3b80d8
Implement encoder methods getJumpTargetOpValue and getBranchTargetOpValue
for jmptarget and brtarget Mips tablegen operand types in the code emitter
for old-style JIT. Rename the pc relative relocation for branches - new
name is Mips::reloc_mips_pc16.
Patch by Sasa Stankovic
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@147382 91177308-0d34-0410-b5e6-96231b3b80d8
tablegen patterns for scalar FMA4 operations and intrinsic. Also
add tests for vfmaddsd.
Patch by Jan Sjodin
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145133 91177308-0d34-0410-b5e6-96231b3b80d8
"With this patch we can now generate runnable Mips code through LLVM
direct object emission. We have run numerous simple programs, both C
and C++ and with -O0 and -O3 from the output. The code is not production
ready, but quite useful for experimentation." Patch and message by
Jack Carter
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144414 91177308-0d34-0410-b5e6-96231b3b80d8
implements unaligned loads and stores with assembler macro-instructions
ulw, usw, ulh, ulhu, ush, and this patch emits corresponding instructions
instead of these macros. Since each unaligned load/store is expanded
into two corresponding loads/stores where offset for second load/store is
modified by +3 (for words) or +1 (for halfwords).
Patch by Petar Jovanovic and Sasa Stankovic.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@144081 91177308-0d34-0410-b5e6-96231b3b80d8
compile and use a bunch of stuff using o32 abi). Also the rt-rk.com team
claims that the JIT support they contributed, is complete for the mips
"static" relocation model.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@142950 91177308-0d34-0410-b5e6-96231b3b80d8
-Fix binary codes and rename operands in .td files so that automatically
generated function MipsCodeEmitter::getBinaryCodeForInstr gives correct
encoding for instructions.
-Define new class FMem for instructions that access memory.
-Define new class FFRGPR for instructions that move data between GPR and
FPU general and control registers.
-Define custom encoder methods for memory operands, and also for size
operands of ext and ins instructions.
-Only static relocation model is currently implemented.
Patch by Sasa Stankovic
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@142378 91177308-0d34-0410-b5e6-96231b3b80d8
compiled on mips32r1 processors because it uses synci and rdhwr instructions
which are supported only on mips32r2, so I replaced this function with the
call to function cacheflush which works for both mips32r1 and mips32r2.
Patch by Sasa Stankovic
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@141564 91177308-0d34-0410-b5e6-96231b3b80d8
are declared with load patterns. This fix the crash in PR10941. No testcases,
since a fold is triggered and then converted back to the register form
afterwards.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@139953 91177308-0d34-0410-b5e6-96231b3b80d8
This PR basically reports a problem where a crash in generated code
happened due to %rbp being clobbered:
pushq %rbp
movq %rsp, %rbp
....
vmovmskps %ymm12, %ebp
....
movq %rbp, %rsp
popq %rbp
ret
Since Eric's r123367 commit, the default stack alignment for x86 32-bit
has changed to be 16-bytes. Since then, the MaxStackAlignmentHeuristicPass
hasn't been really used, but with AVX it becomes useful again, since per
ABI compliance we don't always align the stack to 256-bit, but only when
there are 256-bit incoming arguments.
ReserveFP was only used by this pass, but there's no RA target hook that
uses getReserveFP() to check for the presence of FP (since nothing was
triggering the pass to run, the uses of getReserveFP() were removed
through time without being noticed). Change this pass to use
setForceFramePointer, which is properly called by MachineFunction
hasFP method.
The testcase is very big and dependent on RA, not sure if it's worth
adding to test/CodeGen/X86.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@139939 91177308-0d34-0410-b5e6-96231b3b80d8
take into consideration the presence of AVX. This change, together with
the SSEDomainFix enabled for AVX, makes AVX codegen to always (hopefully)
emit the same code as SSE for 128-bit vector ops. I don't
have a testcase for this, but AVX now beats SSE in performance for
128-bit ops in the majority of programas in the llvm testsuite
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@139817 91177308-0d34-0410-b5e6-96231b3b80d8
- Add TSFlags for the instruction formats. The idea here is to use
as much encoding as possible from getBinaryCodeForInstr, and having
TSFLags formats for that would make it easier to encode most part
of the instructions (since Mips encodings are pretty straightforward)
- Improve the mips mechanism for compilation callback
- Add Mips specific code for invalidating the instruction cache
- Next patch will address wrong tablegen encoding
Commit msg added by my own but the patch is from Sasa Stankovic.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@139688 91177308-0d34-0410-b5e6-96231b3b80d8
alignment check for 256-bit classes more strict. There're no testcases
but we catch more folding cases for AVX while running single and multi
sources in the llvm testsuite.
Since some 128-bit AVX instructions have different number of operands
than their SSE counterparts, they are placed in different tables.
256-bit AVX instructions should also be added in the table soon. And
there a few more 128-bit versions to handled, which should come in
the following commits.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@139687 91177308-0d34-0410-b5e6-96231b3b80d8
more strict about the alignment checking. This was found by inspection
and I don't have any testcases so far, although the llvm testsuite runs
without any problem.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@139625 91177308-0d34-0410-b5e6-96231b3b80d8
Bruno Cardoso Lopes