18fa17cf3d
Fix the folding of floating-point math library calls, like sin(infinity),
...
so that it detects errors on platforms where libm doesn't set errno.
It's still subject to host libm details though.
llvm-svn: 114148
2010-09-17 01:38:06 +00:00
fe12f23024
Add a reduced testcase for the infinite loop fixed in r113763.
...
llvm-svn: 113770
2010-09-13 18:28:40 +00:00
c237a849e3
Re-apply r113679, which was reverted in r113720, which added a paid of new instcombine transforms
...
to expose greater opportunities for store narrowing in codegen. This patch fixes a potential
infinite loop in instcombine caused by one of the introduced transforms being overly aggressive.
llvm-svn: 113763
2010-09-13 17:59:27 +00:00
26abd3e0c2
Revert 113679, it was causing an infinite loop in a testcase that I've sent
...
on to Owen.
llvm-svn: 113720
2010-09-12 06:09:23 +00:00
70f4524427
Invert and-of-or into or-of-and when doing so would allow us to clear bits of the and's mask.
...
This can result in increased opportunities for store narrowing in code generation. Update a number of
tests for this change. This fixes <rdar://problem/8285027>.
Additionally, because this inverts the order of ors and ands, some patterns for optimizing or-of-and-of-or
no longer fire in instances where they did originally. Add a simple transform which recaptures most of these
opportunities: if we have an or-of-constant-or and have failed to fold away the inner or, commute the order
of the two ors, to give the non-constant or a chance for simplification instead.
llvm-svn: 113679
2010-09-11 05:48:06 +00:00
8c35fb0739
Teach InstructionSimplify to fold (A & B) & A -> A & B and (A | B) | A -> A | B.
...
Reassociate does this but it doesn't catch all cases (e.g. if the operands are i1).
llvm-svn: 113651
2010-09-10 22:39:55 +00:00
3fe002dfb5
Generalize instcombine's support for combining multiple bit checks into a single test. Patch by Dirk Steinke!
...
llvm-svn: 113423
2010-09-08 22:16:17 +00:00
6e27b3e004
Fix a serious performance regression introduced by r108687 on linux:
...
turning (fptrunc (sqrt (fpext x))) -> (sqrtf x) is great, but we have
to delete the original sqrt as well. Not doing so causes us to do
two sqrt's when building with -fmath-errno (the default on linux).
llvm-svn: 113260
2010-09-07 20:01:38 +00:00
29570bd695
rename test.
...
llvm-svn: 113257
2010-09-07 19:57:06 +00:00
064cb4c807
Add a test for PR4413, which was apparently fixed at some point in the past.
...
llvm-svn: 112987
2010-09-03 18:33:08 +00:00
65fb25a257
more test cleanup
...
llvm-svn: 112892
2010-09-02 22:38:56 +00:00
376597c13e
Remove r111665, which implemented store-narrowing in InstCombine. Chris discovered a miscompilation in it, and it's not easily
...
fixable at the optimizer level. I'll investigate reimplementing it in DAGCombine.
llvm-svn: 112575
2010-08-31 04:41:06 +00:00
d0214f3efe
handle the constant case of vector insertion. For something
...
like this:
struct S { float A, B, C, D; };
struct S g;
struct S bar() {
struct S A = g;
++A.B;
A.A = 42;
return A;
}
we now generate:
_bar: ## @bar
## BB#0: ## %entry
movq _g@GOTPCREL(%rip), %rax
movss 12(%rax), %xmm0
pshufd $16, %xmm0, %xmm0
movss 4(%rax), %xmm2
movss 8(%rax), %xmm1
pshufd $16, %xmm1, %xmm1
unpcklps %xmm0, %xmm1
addss LCPI1_0(%rip), %xmm2
pshufd $16, %xmm2, %xmm2
movss LCPI1_1(%rip), %xmm0
pshufd $16, %xmm0, %xmm0
unpcklps %xmm2, %xmm0
ret
instead of:
_bar: ## @bar
## BB#0: ## %entry
movq _g@GOTPCREL(%rip), %rax
movss 12(%rax), %xmm0
pshufd $16, %xmm0, %xmm0
movss 4(%rax), %xmm2
movss 8(%rax), %xmm1
pshufd $16, %xmm1, %xmm1
unpcklps %xmm0, %xmm1
addss LCPI1_0(%rip), %xmm2
movd %xmm2, %eax
shlq $32, %rax
addq $1109917696, %rax ## imm = 0x42280000
movd %rax, %xmm0
ret
llvm-svn: 112345
2010-08-28 01:50:57 +00:00
dd6601048e
optimize bitcasts from large integers to vector into vector
...
element insertion from the pieces that feed into the vector.
This handles a pattern that occurs frequently due to code
generated for the x86-64 abi. We now compile something like
this:
struct S { float A, B, C, D; };
struct S g;
struct S bar() {
struct S A = g;
++A.A;
++A.C;
return A;
}
into all nice vector operations:
_bar: ## @bar
## BB#0: ## %entry
movq _g@GOTPCREL(%rip), %rax
movss LCPI1_0(%rip), %xmm1
movss (%rax), %xmm0
addss %xmm1, %xmm0
pshufd $16, %xmm0, %xmm0
movss 4(%rax), %xmm2
movss 12(%rax), %xmm3
pshufd $16, %xmm2, %xmm2
unpcklps %xmm2, %xmm0
addss 8(%rax), %xmm1
pshufd $16, %xmm1, %xmm1
pshufd $16, %xmm3, %xmm2
unpcklps %xmm2, %xmm1
ret
instead of icky integer operations:
_bar: ## @bar
movq _g@GOTPCREL(%rip), %rax
movss LCPI1_0(%rip), %xmm1
movss (%rax), %xmm0
addss %xmm1, %xmm0
movd %xmm0, %ecx
movl 4(%rax), %edx
movl 12(%rax), %esi
shlq $32, %rdx
addq %rcx, %rdx
movd %rdx, %xmm0
addss 8(%rax), %xmm1
movd %xmm1, %eax
shlq $32, %rsi
addq %rax, %rsi
movd %rsi, %xmm1
ret
This resolves rdar://8360454
llvm-svn: 112343
2010-08-28 01:20:38 +00:00
954e9557e3
tidy up test.
...
llvm-svn: 112321
2010-08-27 23:15:21 +00:00
6c1395f62a
Enhance the shift propagator to handle the case when you have:
...
A = shl x, 42
...
B = lshr ..., 38
which can be transformed into:
A = shl x, 4
...
iff we can prove that the would-be-shifted-in bits
are already zero. This eliminates two shifts in the testcase
and allows eliminate of the whole i128 chain in the real example.
llvm-svn: 112314
2010-08-27 22:53:44 +00:00
18d7fc8fc6
Implement a pretty general logical shift propagation
...
framework, which is good at ripping through bitfield
operations. This generalize a bunch of the existing
xforms that instcombine does, such as
(x << c) >> c -> and
to handle intermediate logical nodes. This is useful for
ripping up the "promote to large integer" code produced by
SRoA.
llvm-svn: 112304
2010-08-27 22:24:38 +00:00
606b76eba6
merge and filecheckize test
...
llvm-svn: 112289
2010-08-27 20:44:45 +00:00
c665156e9f
merge two tests
...
llvm-svn: 112288
2010-08-27 20:42:10 +00:00
7398434675
teach the truncation optimization that an entire chain of
...
computation can be truncated if it is fed by a sext/zext that doesn't
have to be exactly equal to the truncation result type.
llvm-svn: 112285
2010-08-27 20:32:06 +00:00
90cd746e63
Add an instcombine to clean up a common pattern produced
...
by the SRoA "promote to large integer" code, eliminating
some type conversions like this:
%94 = zext i16 %93 to i32 ; <i32> [#uses=2]
%96 = lshr i32 %94, 8 ; <i32> [#uses=1]
%101 = trunc i32 %96 to i8 ; <i8> [#uses=1]
This also unblocks other xforms from happening, now clang is able to compile:
struct S { float A, B, C, D; };
float foo(struct S A) { return A.A + A.B+A.C+A.D; }
into:
_foo: ## @foo
## BB#0: ## %entry
pshufd $1, %xmm0, %xmm2
addss %xmm0, %xmm2
movdqa %xmm1, %xmm3
addss %xmm2, %xmm3
pshufd $1, %xmm1, %xmm0
addss %xmm3, %xmm0
ret
on x86-64, instead of:
_foo: ## @foo
## BB#0: ## %entry
movd %xmm0, %rax
shrq $32, %rax
movd %eax, %xmm2
addss %xmm0, %xmm2
movapd %xmm1, %xmm3
addss %xmm2, %xmm3
movd %xmm1, %rax
shrq $32, %rax
movd %eax, %xmm0
addss %xmm3, %xmm0
ret
This seems pretty close to optimal to me, at least without
using horizontal adds. This also triggers in lots of other
code, including SPEC.
llvm-svn: 112278
2010-08-27 18:31:05 +00:00
c188b96bbe
filecheckize
...
llvm-svn: 112235
2010-08-26 22:23:39 +00:00
387d6bcdcb
rename test.
...
llvm-svn: 112234
2010-08-26 22:20:47 +00:00
bfd2228182
optimize "integer extraction out of the middle of a vector" as produced
...
by SRoA. This is part of rdar://7892780, but needs another xform to
expose this.
llvm-svn: 112232
2010-08-26 22:14:59 +00:00
d4ebd6df5a
optimize bitcast(trunc(bitcast(x))) where the result is a float and 'x'
...
is a vector to be a vector element extraction. This allows clang to
compile:
struct S { float A, B, C, D; };
float foo(struct S A) { return A.A + A.B+A.C+A.D; }
into:
_foo: ## @foo
## BB#0: ## %entry
movd %xmm0, %rax
shrq $32, %rax
movd %eax, %xmm2
addss %xmm0, %xmm2
movapd %xmm1, %xmm3
addss %xmm2, %xmm3
movd %xmm1, %rax
shrq $32, %rax
movd %eax, %xmm0
addss %xmm3, %xmm0
ret
instead of:
_foo: ## @foo
## BB#0: ## %entry
movd %xmm0, %rax
movd %eax, %xmm0
shrq $32, %rax
movd %eax, %xmm2
addss %xmm0, %xmm2
movd %xmm1, %rax
movd %eax, %xmm1
addss %xmm2, %xmm1
shrq $32, %rax
movd %eax, %xmm0
addss %xmm1, %xmm0
ret
... eliminating half of the horribleness.
llvm-svn: 112227
2010-08-26 21:55:42 +00:00
3c19d3d5c3
filecheckize
...
llvm-svn: 112225
2010-08-26 21:51:41 +00:00
7717c616bd
rename test
...
llvm-svn: 112224
2010-08-26 21:50:56 +00:00
84c29a096b
Re-apply r111568 with a fix for the clang self-host.
...
llvm-svn: 111665
2010-08-20 18:24:43 +00:00
3323651ec7
Previous revert failed to remove this file.
...
llvm-svn: 111582
2010-08-19 23:45:15 +00:00
43057cd56a
Revert r111568 to unbreak clang self-host.
...
llvm-svn: 111571
2010-08-19 23:25:16 +00:00
bb723b228a
When a set of bitmask operations, typically from a bitfield initialization, only modifies the low bytes of a value,
...
we can narrow the store to only over-write the affected bytes.
llvm-svn: 111568
2010-08-19 22:15:40 +00:00
51edc7b7e1
Temporarily revert r110987 as it's causing some miscompares in
...
vector heavy code. I'll re-enable when we've tracked down the problem.
llvm-svn: 111318
2010-08-17 22:55:27 +00:00
2a0ca3e937
Reapply this transformation now that it is passing the external test which it previously failed.
...
llvm-svn: 110987
2010-08-13 00:17:53 +00:00
ac40d49c70
Temporarily revert 110737 and 110734, they were causing failures
...
in an external testsuite.
llvm-svn: 110905
2010-08-12 07:01:22 +00:00
3ec892c167
Add test for recent instcombine vector shuffle enhancement
...
llvm-svn: 110737
2010-08-10 21:58:00 +00:00
f99e7e6643
PR7853: fix a silly mistake introduced in r101899, and add a test to make sure
...
it doesn't regress again.
llvm-svn: 110597
2010-08-09 20:49:43 +00:00
3619660529
Make instcombine set explicit alignments on load or store
...
instructions with alignment 0, so that subsequent passes don't
need to bother checking the TargetData ABI size manually.
llvm-svn: 110128
2010-08-03 18:20:32 +00:00
8f306a779b
Re-apply the infamous r108614, with a fix pointed out by Dirk Steinke.
...
llvm-svn: 110036
2010-08-02 09:32:13 +00:00
0b636a24c7
Speculatively revert r108614, "Another attempt at getting the clang self-host to
...
like my instcombine patch.", in an attempt to fix Clang i386 bootstrap.
- Also PR7719.
llvm-svn: 109953
2010-07-31 19:51:11 +00:00
3ccd81864f
Testcase for r108687.
...
llvm-svn: 108689
2010-07-19 08:14:26 +00:00
7d2818b073
Another attempt at getting the clang self-host to like my instcombine patch.
...
llvm-svn: 108614
2010-07-17 06:56:35 +00:00
15a81cddb4
Also revert 108422, it's causing some test failures.
...
Working on testcases for Owen.
llvm-svn: 108494
2010-07-16 01:36:12 +00:00
7151dfd48a
Reapply r108378, with bugfixes, testcase, and improved comment formatting.
...
This now passes LIT, nighty test, and llvm-gcc bootstrap on my machine.
llvm-svn: 108422
2010-07-15 15:00:23 +00:00
ec0e7b1643
revert r108320, I see the failures now...
...
llvm-svn: 108322
2010-07-14 06:16:35 +00:00
658680b2f5
reapply benjamin's instcombine patch, I don't see anything wrong with it and can't repro any problems with a manual self-host.
...
llvm-svn: 108320
2010-07-14 05:59:13 +00:00
8f36402ac2
Nope, still breaks the release selfhost bots :(
...
llvm-svn: 108153
2010-07-12 16:38:48 +00:00
07b695e052
Reapply the "or" half of r108136, which seems to be less problematic.
...
llvm-svn: 108152
2010-07-12 16:15:48 +00:00
c719e8ae9e
Revert r108141 again, sigh.
...
llvm-svn: 108148
2010-07-12 14:42:04 +00:00
f578c36035
Reapply 108136 with an ugly pasto fixed.
...
llvm-svn: 108141
2010-07-12 13:44:00 +00:00
9675e759cf
Revert r108136 until I figure out why it broke selfhost.
...
llvm-svn: 108139
2010-07-12 12:35:49 +00:00
Dan Gohman