Commit Graph

948 Commits

Author SHA1 Message Date
Cameron Zwarich
4071a71112 Do some peephole optimizations to remove pointless VMOVs from Neon to integer
registers that arise from argument shuffling with the soft float ABI. These
instructions are particularly slow on Cortex A8. This fixes one half of
<rdar://problem/8674845>.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@128759 91177308-0d34-0410-b5e6-96231b3b80d8
2011-04-02 02:40:43 +00:00
Jim Grosbach
9a3507f091 LDRD/STRD instructions should print both Rt and Rt2 in the asm string.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@128736 91177308-0d34-0410-b5e6-96231b3b80d8
2011-04-01 20:26:57 +00:00
Evan Cheng
5b76c63f83 Add test case.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@128707 91177308-0d34-0410-b5e6-96231b3b80d8
2011-04-01 06:27:25 +00:00
Evan Cheng
df269b9129 FileCheck'ify test.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@128706 91177308-0d34-0410-b5e6-96231b3b80d8
2011-04-01 03:36:33 +00:00
Jakob Stoklund Olesen
ca6fd009ad Fix ARM tests to be register allocator independent.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@128680 91177308-0d34-0410-b5e6-96231b3b80d8
2011-03-31 22:14:03 +00:00
Evan Cheng
463d358f1d Distribute (A + B) * C to (A * C) + (B * C) to make use of NEON multiplier
accumulator forwarding:
vadd d3, d0, d1
vmul d3, d3, d2
=>
vmul d3, d0, d2
vmla d3, d1, d2


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@128665 91177308-0d34-0410-b5e6-96231b3b80d8
2011-03-31 19:38:48 +00:00
Jakob Stoklund Olesen
312babc93f Pick a conservative register class when creating a small live range for remat.
The rematerialized instruction may require a more constrained register class
than the register being spilled. In the test case, the spilled register has been
inflated to the DPR register class, but we are rematerializing a load of the
ssub_0 sub-register which only exists for DPR_VFP2 registers.

The register class is reinflated after spilling, so the conservative choice is
only temporary.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@128610 91177308-0d34-0410-b5e6-96231b3b80d8
2011-03-31 03:54:44 +00:00
Cameron Zwarich
c0e6d780cd Add a ARM-specific SD node for VBSL so that forms with a constant first operand
can be recognized. This fixes <rdar://problem/9183078>.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@128584 91177308-0d34-0410-b5e6-96231b3b80d8
2011-03-30 23:01:21 +00:00
Evan Cheng
92e3916c3b Add intrinsics @llvm.arm.neon.vmulls and @llvm.arm.neon.vmullu.* back. Frontends
was lowering them to sext / uxt + mul instructions. Unfortunately the
optimization passes may hoist the extensions out of the loop and separate them.
When that happens, the long multiplication instructions can be broken into
several scalar instructions, causing significant performance issue.

Note the vmla and vmls intrinsics are not added back. Frontend will codegen them
as intrinsics vmull* + add / sub. Also note the isel optimizations for catching
mul + sext / zext are not changed either.

First part of rdar://8832507, rdar://9203134


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@128502 91177308-0d34-0410-b5e6-96231b3b80d8
2011-03-29 23:06:19 +00:00
Cameron Zwarich
3007d3331b Add Neon SINT_TO_FP and UINT_TO_FP lowering from v4i16 to v4f32. Fixes
<rdar://problem/8875309> and <rdar://problem/9057191>.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@128492 91177308-0d34-0410-b5e6-96231b3b80d8
2011-03-29 21:41:55 +00:00
Evan Cheng
78fe9ababe Optimizing (zext A + zext B) * C, to (VMULL A, C) + (VMULL B, C) during
isel lowering to fold the zero-extend's and take advantage of no-stall
back to back vmul + vmla:
 vmull q0, d4, d6
 vmlal q0, d5, d6
is faster than
 vaddl q0, d4, d5
 vmovl q1, d6                                                                                                                                                                             
 vmul  q0, q0, q1

This allows us to vmull + vmlal for:
    f = vmull_u8(   vget_high_u8(s), c);
    f = vmlal_u8(f, vget_low_u8(s),  c);

rdar://9197392


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@128444 91177308-0d34-0410-b5e6-96231b3b80d8
2011-03-29 01:56:09 +00:00
Devang Patel
36dca60f5c Enable GlobalMerge on darwin.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@128183 91177308-0d34-0410-b5e6-96231b3b80d8
2011-03-23 23:34:19 +00:00
Evan Cheng
2c33915628 Cmp peephole optimization isn't always safe for signed arithmetics.
int tries = INT_MAX;    
while (tries > 0) {
      tries--;
}

The check should be:
        subs    r4, #1
        cmp     r4, #0
        bgt     LBB0_1

The subs can set the overflow V bit when r4 is INT_MAX+1 (which loop
canonicalization apparently does in this case). cmp #0 would have cleared
it while not changing the N and Z bits. Since BGT is dependent on the V
bit, i.e. (N == V) && !Z, it is not safe to eliminate the cmp #0.

rdar://9172742


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@128179 91177308-0d34-0410-b5e6-96231b3b80d8
2011-03-23 22:52:04 +00:00
Rafael Espindola
7c18fa87a4 Write the section table and the section data in the same order that
gun as does. This makes it a lot easier to compare the output of both
as the addresses are now a lot closer.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@127972 91177308-0d34-0410-b5e6-96231b3b80d8
2011-03-20 18:44:20 +00:00
Evan Cheng
3f30af3f45 Match a few more obvious patterns to revsh. rdar://9147637.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@127913 91177308-0d34-0410-b5e6-96231b3b80d8
2011-03-18 21:52:42 +00:00
Daniel Dunbar
83320a023c Revert r127757, "Patch to a fix dwarf relocation problem on ARM. One-line fix
plus the test where it used to break.", which broke Clang self-host of a
Debug+Asserts compiler, on OS X.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@127763 91177308-0d34-0410-b5e6-96231b3b80d8
2011-03-16 22:16:39 +00:00
Renato Golin
2589d0aedf Patch to a fix dwarf relocation problem on ARM. One-line fix plus the test where it used to break.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@127757 91177308-0d34-0410-b5e6-96231b3b80d8
2011-03-16 21:05:52 +00:00
Bill Wendling
a24cb40be2 Some minor cleanups based on feedback.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@127694 91177308-0d34-0410-b5e6-96231b3b80d8
2011-03-15 20:47:26 +00:00
Evan Cheng
0191952296 Do not form thumb2 ldrd / strd if the offset is by multiple of 4. rdar://9133587
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@127683 91177308-0d34-0410-b5e6-96231b3b80d8
2011-03-15 18:41:52 +00:00
Evan Cheng
d158fba3e4 Add a peephole optimization to optimize pairs of bitcasts. e.g.
v2 = bitcast v1
...
v3 = bitcast v2
...
   = v3
=>
v2 = bitcast v1
...
   = v1
if v1 and v3 are of in the same register class.

bitcast between i32 and fp (and others) are often not nops since they
are in different register classes. These bitcast instructions are often
left because they are in different basic blocks and cannot be
eliminated by dag combine.

rdar://9104514


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@127668 91177308-0d34-0410-b5e6-96231b3b80d8
2011-03-15 05:13:13 +00:00
Bill Wendling
49c69bb70f Testcase for r127630.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@127648 91177308-0d34-0410-b5e6-96231b3b80d8
2011-03-15 01:49:08 +00:00
Jim Grosbach
5edf24efac Clean up ARM tail calls a bit. They're pseudo-instructions for normal branches.
Also more cleanly separate the ARM vs. Thumb functionality. Previously, the
encoding would be incorrect for some Thumb instructions (the indirect calls).

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@127637 91177308-0d34-0410-b5e6-96231b3b80d8
2011-03-15 00:30:40 +00:00
Bill Wendling
69a05a7b92 Generate a VTBL instruction instead of a series of loads and stores when we
can. As Nate pointed out, VTBL isn't super performant, but it *has* to be better
than this:

_shuf:
@ BB#0:       @ %entry
  push        {r4, r7, lr}
  add         r7, sp, #4
  sub         sp, #12
  mov         r4, sp
  bic         r4, r4, #7
  mov         sp, r4
  mov         r2, sp
  vmov        d16, r0, r1
  orr         r0, r2, #6
  orr         r3, r2, #7
  vst1.8      {d16[0]}, [r3]
  vst1.8      {d16[5]}, [r0]
  subs        r4, r7, #4
  orr         r0, r2, #5
  vst1.8      {d16[4]}, [r0]
  orr         r0, r2, #4
  vst1.8      {d16[4]}, [r0]
  orr         r0, r2, #3
  vst1.8      {d16[0]}, [r0]
  orr         r0, r2, #2
  vst1.8      {d16[2]}, [r0]
  orr         r0, r2, #1
  vst1.8      {d16[1]}, [r0]
  vst1.8      {d16[3]}, [r2]
  vldr.64     d16, [sp]
  vmov        r0, r1, d16
  mov         sp, r4
  pop         {r4, r7, pc}

The "illegal" testcase in vext.ll is no longer illegal.
<rdar://problem/9078775>


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@127630 91177308-0d34-0410-b5e6-96231b3b80d8
2011-03-14 23:02:38 +00:00
Eric Christopher
29cfa21326 Fix this test up a bit.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@127621 91177308-0d34-0410-b5e6-96231b3b80d8
2011-03-14 21:05:21 +00:00
Evan Cheng
5ae1da96f2 Minor optimization. sign-ext/anyext of undef is still undef.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@127598 91177308-0d34-0410-b5e6-96231b3b80d8
2011-03-14 18:15:55 +00:00
Eric Christopher
8b3000b08e Saving files before committing is overrated.
Add a RUN line to this test.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@127520 91177308-0d34-0410-b5e6-96231b3b80d8
2011-03-12 01:36:23 +00:00
Eric Christopher
af3dce5149 Sometimes isPredicable lies to us and tells us we don't need the operands.
Go ahead and add them on when we might want to use them and let
later passes remove them.

Fixes rdar://9118569


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@127518 91177308-0d34-0410-b5e6-96231b3b80d8
2011-03-12 01:09:29 +00:00
Jim Grosbach
dd11988c99 Properly pseudo-ize the ARM LDMIA_RET instruction. This has the nice side-
effect that we get proper instruction printing using the "pop" mnemonic for it.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@127502 91177308-0d34-0410-b5e6-96231b3b80d8
2011-03-11 22:51:41 +00:00
Cameron Zwarich
899eaa3569 Roll r127459 back in:
Optimize trivial branches in CodeGenPrepare, which often get created from the
lowering of objectsize intrinsics. Unfortunately, a number of tests were relying
on llc not optimizing trivial branches, so I had to add an option to allow them
to continue to test what they originally tested.

This fixes <rdar://problem/8785296> and <rdar://problem/9112893>.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@127498 91177308-0d34-0410-b5e6-96231b3b80d8
2011-03-11 21:52:04 +00:00
Daniel Dunbar
950d3db5f4 Revert r127459, "Optimize trivial branches in CodeGenPrepare, which often get
created from the", it broke some GCC test suite tests.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@127477 91177308-0d34-0410-b5e6-96231b3b80d8
2011-03-11 19:30:30 +00:00
Cameron Zwarich
592ca3fda9 Optimize trivial branches in CodeGenPrepare, which often get created from the
lowering of objectsize intrinsics. Unfortunately, a number of tests were relying
on llc not optimizing trivial branches, so I had to add an option to allow them
to continue to test what they originally tested.

This fixes <rdar://problem/8785296> and <rdar://problem/9112893>.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@127459 91177308-0d34-0410-b5e6-96231b3b80d8
2011-03-11 04:54:27 +00:00
Evan Cheng
9ef82ce4fe Avoid replacing the value of a directly stored load with the stored value if the load is indexed. rdar://9117613.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@127440 91177308-0d34-0410-b5e6-96231b3b80d8
2011-03-11 00:48:56 +00:00
Jim Grosbach
d4a16ad85d Properly pseudo-ize MOVCCr and MOVCCs.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@127434 91177308-0d34-0410-b5e6-96231b3b80d8
2011-03-10 23:56:09 +00:00
Bob Wilson
79f56c9618 Fix a compiler crash where a Glue value had multiple uses. Radar 9049552.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@127198 91177308-0d34-0410-b5e6-96231b3b80d8
2011-03-08 01:17:20 +00:00
Joerg Sonnenberger
89e0f386f3 Be nice to Xcore and the XMOS assembler and avoid quoting section names
that contain only letters, digits and the characters "_" and ".".


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@127028 91177308-0d34-0410-b5e6-96231b3b80d8
2011-03-04 20:03:14 +00:00
Devang Patel
c8888c7d80 XFAIL for all. These tests are darwin specific anyway.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@127022 91177308-0d34-0410-b5e6-96231b3b80d8
2011-03-04 19:38:10 +00:00
Devang Patel
53dc40a45f Disable ARMGlobalMerge on darwin. The debugger is not yet able to extract individual variable's info from merged global.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@127019 91177308-0d34-0410-b5e6-96231b3b80d8
2011-03-04 19:11:05 +00:00
Joerg Sonnenberger
ea83b13350 Bug#9033: For the ELF assembler output, always quote the section name.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@126963 91177308-0d34-0410-b5e6-96231b3b80d8
2011-03-03 22:31:08 +00:00
Cameron Zwarich
56e3793acf Eliminate the unused CodeGenPrepare option to split critical edges.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@126825 91177308-0d34-0410-b5e6-96231b3b80d8
2011-03-02 03:31:46 +00:00
Bill Wendling
a656b63ee4 Narrow right shifts need to encode their immediates differently from a normal
shift.

   16-bit: imm6<5:3> = '001', 8 - <imm> is encded in imm6<2:0>
   32-bit: imm6<5:4> = '01',16 - <imm> is encded in imm6<3:0>
   64-bit: imm6<5> = '1', 32 - <imm> is encded in imm6<4:0>


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@126723 91177308-0d34-0410-b5e6-96231b3b80d8
2011-03-01 01:00:59 +00:00
Jakob Stoklund Olesen
bafd516d3e Fix typo introduced by r126661: "Fix a typo which ..."
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@126666 91177308-0d34-0410-b5e6-96231b3b80d8
2011-02-28 19:18:59 +00:00
Evan Cheng
c24ab5c654 Fix a typo which cause dag combine crash. rdar://9059537.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@126661 91177308-0d34-0410-b5e6-96231b3b80d8
2011-02-28 18:45:27 +00:00
Bob Wilson
da52506792 Add patterns to use post-increment addressing for Neon VST1-lane instructions.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@126477 91177308-0d34-0410-b5e6-96231b3b80d8
2011-02-25 06:42:42 +00:00
Devang Patel
3fe3424a21 Move arch specific tests in arch specific directories.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@126401 91177308-0d34-0410-b5e6-96231b3b80d8
2011-02-24 19:06:27 +00:00
Devang Patel
1d6b2c74c0 Check only relevant strings in output to increase stability of the tests.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@126338 91177308-0d34-0410-b5e6-96231b3b80d8
2011-02-23 22:35:57 +00:00
Evan Cheng
e573fb3255 More fcopysign correctness and performance fix.
The previous codegen for the slow path (when values are in VFP / NEON
registers) was incorrect if the source is NaN.

The new codegen uses NEON vbsl instruction to copy the sign bit. e.g.
        vmov.i32        d1, #0x80000000
        vbsl    d1, d2, d0
If NEON is not available, it uses integer instructions to copy the sign bit.
rdar://9034702


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@126295 91177308-0d34-0410-b5e6-96231b3b80d8
2011-02-23 02:24:55 +00:00
Evan Cheng
af05c69ba0 available_externally (hidden or not) GVs are always accessed via stubs. rdar://9027648.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@126191 91177308-0d34-0410-b5e6-96231b3b80d8
2011-02-22 06:58:34 +00:00
Bob Wilson
56519aac04 PR9139: Specify ARM/Darwin triple for vector-DAGCombine.ll test.
The i64_buildvector test in this file relies on the alignment of i64 and
f64 types being the same, which is true for Darwin but not AAPCS.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@125525 91177308-0d34-0410-b5e6-96231b3b80d8
2011-02-14 22:12:50 +00:00
Nate Begeman
7973f350b7 Implement sdiv & udiv for <4 x i16> and <8 x i8> NEON vector types.
This avoids moving each element to the integer register file and calling __divsi3 etc. on it.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@125402 91177308-0d34-0410-b5e6-96231b3b80d8
2011-02-11 20:53:29 +00:00
Evan Cheng
c143dd4f63 Fix buggy fcopysign lowering.
This
define float @foo(float %x, float %y) nounwind readnone {
entry:
  %0 = tail call float @copysignf(float %x, float %y) nounwind readnone
  ret float %0
}

Was compiled to:
    vmov     s0, r1
    bic      r0, r0, #-2147483648
    vmov     s1, r0
    vcmpe.f32    s0, #0
    vmrs         apsr_nzcv, fpscr
    it           lt
    vneglt.f32   s1, s1
    vmov         r0, s1
    bx           lr

This fails to copy the sign of -0.0f because it's lost during the float to int
conversion. Also, it's sub-optimal when the inputs are in GPR registers.

Now it uses integer and + or operations when it's profitable. And it's correct!
    lsrs    r1, r1, #31
    bfi     r0, r1, #31, #1
    bx      lr
rdar://8984306


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@125357 91177308-0d34-0410-b5e6-96231b3b80d8
2011-02-11 02:28:55 +00:00