Commit Graph

143 Commits

Author SHA1 Message Date
Bill Wendling
de2b151dbf Consider this code snippet:
float t1(int argc) {
  return (argc == 1123) ? 1.234f : 2.38213f;
}

We would generate truly awful code on ARM (those with a weak stomach should look
away):

_t1:
  movw   r1, #1123
  movs   r2, #1
  movs   r3, #0
  cmp    r0, r1
  mov.w  r0, #0
  it     eq
  moveq  r0, r2
  movs   r1, #4
  cmp    r0, #0
  it     ne
  movne  r3, r1
  adr    r0, #LCPI1_0
  ldr    r0, [r0, r3]
  bx     lr

The problem was that legalization was creating a cascade of SELECT_CC nodes, for
for the comparison of "argc == 1123" which was fed into a SELECT node for the ?:
statement which was itself converted to a SELECT_CC node. This is because the
ARM back-end doesn't have custom lowering for SELECT nodes, so it used the
default "Expand".

I added a fairly simple "LowerSELECT" to the ARM back-end. It takes care of this
testcase, but can obviously be expanded to include more cases.

Now we generate this, which looks optimal to me:

_t1:
  movw   r1, #1123
  movs   r2, #0
  cmp    r0, r1
  adr    r0, #LCPI0_0
  it     eq
  moveq  r2, #4
  ldr    r0, [r0, r2]
  bx     lr
  .align  2
LCPI0_0:
  .long   1075344593  @ float 2.382130e+00
  .long   1067316150  @ float 1.234000e+00



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@110799 91177308-0d34-0410-b5e6-96231b3b80d8
2010-08-11 08:43:16 +00:00
Nate Begeman
d1fb583128 Add support for getting & setting the FPSCR application register on ARM when VFP is enabled.
Add support for using the FPSCR in conjunction with the vcvtr instruction, for controlling fp to int rounding.
Add support for the FLT_ROUNDS_ node now that the FPSCR is exposed.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@110152 91177308-0d34-0410-b5e6-96231b3b80d8
2010-08-03 21:31:55 +00:00
Jim Grosbach
f0d7e36691 Remove dead prototype
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@109691 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-28 23:16:12 +00:00
Anton Korobeynikov
cec36f4c11 Hook in GlobalMerge pass
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@109359 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-24 21:52:08 +00:00
Evan Cheng
3144687df7 - Allow target to specify when is register pressure "too high". In most cases,
it's too late to start backing off aggressive latency scheduling when most
  of the registers are in use so the threshold should be a bit tighter.
- Correctly handle live out's and extract_subreg etc.
- Enable register pressure aware scheduling by default for hybrid scheduler.
  For ARM, this is almost always a win on # of instructions. It's runtime
  neutral for most of the tests. But for some kernels with high register
  pressure it can be a huge win. e.g. 464.h264ref reduced number of spills by
  54 and sped up by 20%.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@109279 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-23 22:39:59 +00:00
Eric Christopher
ab695889c6 Baby steps towards ARM fast-isel.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@109047 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-21 22:26:11 +00:00
Evan Cheng
4f6b4674be Teach bottom up pre-ra scheduler to track register pressure. Work in progress.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108991 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-21 06:09:07 +00:00
Evan Cheng
d70f57b254 ARM has to provide its own TargetLowering::findRepresentativeClass because its scalar floating point registers alias its vector registers.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108761 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-19 22:15:08 +00:00
Jim Grosbach
e1102caf86 Since ARM emits inline jump tables as part of the ConstantIsland pass,
it should set the jump table encloding the EK_Inline. This prevents
a second, unused, copy of the table from being emitted after the function
body. PR6581.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108730 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-19 17:20:38 +00:00
Jim Grosbach
350afb16ec revert so I can get the right PR# in the log message.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108727 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-19 17:19:40 +00:00
Jim Grosbach
0bb9895a78 Since ARM emits inline jump tables as part of the ConstantIsland pass,
it should set the jump table encloding the EK_Inline. This prevents
a second, unused, copy of the table from being emitted after the function
body. PR7499.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108722 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-19 17:18:28 +00:00
Jim Grosbach
469bbdb597 Add basic support to code-gen the ARM/Thumb2 bit-field insert (BFI) instruction
and a combine pattern to use it for setting a bit-field to a constant
value. More to come for non-constant stores.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108570 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-16 23:05:05 +00:00
Bob Wilson
7e3f0d2690 Add support for NEON VMVN immediate instructions.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108324 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-14 06:31:50 +00:00
Bob Wilson
cba270d042 Use a target-specific VMOVIMM DAG node instead of BUILD_VECTOR to represent
NEON VMOV-immediate instructions.  This simplifies some things.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108275 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-13 21:16:48 +00:00
Evan Cheng
218977b53e Extend the r107852 optimization which turns some fp compare to code sequence using only i32 operations. It now optimize some f64 compares when fp compare is exceptionally slow (e.g. cortex-a8). It also catches comparison against 0.0.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108258 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-13 19:27:42 +00:00
Evan Cheng
515fe3a588 Optimize some vfp comparisons to integer ones. This patch implements the simplest case when the following conditions are met:
1. The arguments are f32.
2. The arguments are loads and they have no uses other than the comparison.
3. The comparison code is EQ or NE.

e.g.
        vldr.32 s0, [r1]
        vldr.32 s1, [r0]
        vcmpe.f32       s1, s0
        vmrs    apsr_nzcv, fpscr
	beq     LBB0_2
=>
        ldr     r1, [r1]
        ldr     r0, [r0]
        cmp     r0, r1
        beq     LBB0_2

More complicated cases will be implemented in subsequent patches.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@107852 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-08 02:08:50 +00:00
Dan Gohman
c9403659a9 Split the SDValue out of OutputArg so that SelectionDAG-independent
code can do calling-convention queries. This obviates OutputArgReg.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@107786 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-07 15:54:55 +00:00
Dale Johannesen
1784d160e4 The hasMemory argument is irrelevant to how the argument
for an "i" constraint should get lowered; PR 6309.  While
this argument was passed around a lot, this is the only
place it was used, so it goes away from a lot of other
places.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106893 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-25 21:55:36 +00:00
Bob Wilson
827b2106fe Add basic support for NEON modified immediates besides VMOV.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106030 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-15 19:05:35 +00:00
Bob Wilson
d3c4284849 Rename functions referring to VMOV immediates to refer to NEON "modified
immediate" operands.  These functions have so far only been used for VMOV
but they also apply to other NEON instructions with modified immediate
operands.  No functional changes.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@105969 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-14 22:19:57 +00:00
Bob Wilson
40cbe7d5d4 For NEON vectors with 32- or 64-bit elements, select BUILD_VECTORs and
VECTOR_SHUFFLEs to REG_SEQUENCE instructions.  The standard ISD::BUILD_VECTOR
node corresponds closely to REG_SEQUENCE but I couldn't use it here because
its operands do not get legalized.  That is pretty awful, but I guess it
makes sense for other targets.  Instead, I have added an ARM-specific version
of BUILD_VECTOR that will have its operands properly legalized.
This fixes the rest of Radar 7872877.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@105439 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-04 00:04:02 +00:00
Dale Johannesen
51e28e6348 Early implementation of tail call for ARM.
A temporary flag -arm-tail-calls defaults to off,
so there is no functional change by default.
Intrepid users may try this; simple cases work
but there are bugs.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@105413 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-03 21:09:53 +00:00
Jim Grosbach
18f30e6f5e Clean up 80 column violations. No functional change.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@105350 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-02 21:53:11 +00:00
Jim Grosbach
a65850230a back out 104862/104869. Can reuse stacksave after all. Very cool.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104897 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-27 23:11:57 +00:00
Jim Grosbach
ad9aaf038e add ISD::STACKADDR to get the current stack pointer. Will be used by sjlj EH
to update the jmpbuf in the presence of VLAs.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104862 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-27 18:23:48 +00:00
Jim Grosbach
23ff7cff52 Adjust eh.sjlj.setjmp to properly have a chain and to have an opcode entry in
ISD::. No functional change.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104734 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-26 20:22:18 +00:00
Evan Cheng
2457f2c661 Implement @llvm.returnaddress. rdar://8015977.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104421 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-22 01:47:14 +00:00
Jim Grosbach
5eb1951539 Implement eh.sjlj.longjmp for ARM. Clean up the intrinsic a bit.
Followups: docs patch for the builtin and eh.sjlj.setjmp cleanup to match
longjmp.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104419 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-22 01:06:18 +00:00
Evan Cheng
1cc3984148 Allow targets more controls on what nodes are scheduled by reg pressure, what for latency in hybrid mode.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@104293 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-20 23:26:43 +00:00
Evan Cheng
06b666c705 Allow TargetLowering::getRegClassFor() to be called on illegal types. Also
allow target to override it in order to map register classes to illegal
but synthesizable types. e.g. v4i64, v8i64 for ARM / NEON.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@103854 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-15 02:18:07 +00:00
Dan Gohman
ff7a562751 Implement a bunch more TargetSelectionDAGInfo infrastructure.
Move EmitTargetCodeForMemcpy, EmitTargetCodeForMemset, and
EmitTargetCodeForMemmove out of TargetLowering and into
SelectionDAGInfo to exercise this.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@103481 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-11 17:31:57 +00:00
Dan Gohman
419e4f9263 Remove the TargetLowering::getSubtarget() virtual function, which
was unused. TargetMachine::getSubtarget() is used instead.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@103474 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-11 16:21:03 +00:00
Dan Gohman
af1d8ca44a Get rid of the EdgeMapping map. Instead, just check for BasicBlock
changes before doing phi lowering for switches.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@102809 91177308-0d34-0410-b5e6-96231b3b80d8
2010-05-01 00:01:06 +00:00
Dan Gohman
d858e90f03 Use const qualifiers with TargetLowering. This eliminates several
const_casts, and it reinforces the design of the Target classes being
immutable.

SelectionDAGISel::IsLegalToFold is now a static member function, because
PIC16 uses it in an unconventional way. There is more room for API
cleanup here.

And PIC16's AsmPrinter no longer uses TargetLowering.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@101635 91177308-0d34-0410-b5e6-96231b3b80d8
2010-04-17 15:26:15 +00:00
Dan Gohman
1e93df6f0b Move per-function state out of TargetLowering subclasses and into
MachineFunctionInfo subclasses.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@101634 91177308-0d34-0410-b5e6-96231b3b80d8
2010-04-17 14:41:14 +00:00
Mon P Wang
20adc9dc46 Reapply address space patch after fixing an issue in MemCopyOptimizer.
Added support for address spaces and added a isVolatile field to memcpy, memmove, and memset,
e.g., llvm.memcpy.i32(i8*, i8*, i32, i32) -> llvm.memcpy.p0i8.p0i8.i32(i8*, i8*, i32, i32, i1)


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@100304 91177308-0d34-0410-b5e6-96231b3b80d8
2010-04-04 03:10:48 +00:00
Mon P Wang
e754d3fb85 Revert r100191 since it breaks objc in clang
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@100199 91177308-0d34-0410-b5e6-96231b3b80d8
2010-04-02 18:43:02 +00:00
Mon P Wang
e33c848fa4 Reapply address space patch after fixing an issue in MemCopyOptimizer.
Added support for address spaces and added a isVolatile field to memcpy, memmove, and memset,
e.g., llvm.memcpy.i32(i8*, i8*, i32, i32) -> llvm.memcpy.p0i8.p0i8.i32(i8*, i8*, i32, i32, i1)


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@100191 91177308-0d34-0410-b5e6-96231b3b80d8
2010-04-02 18:04:15 +00:00
Bob Wilson
100f090add Revert Mon Ping's change 99928, since it broke all the llvm-gcc buildbots.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99948 91177308-0d34-0410-b5e6-96231b3b80d8
2010-03-30 22:27:04 +00:00
Mon P Wang
808bab0169 Added support for address spaces and added a isVolatile field to memcpy, memmove, and memset,
e.g., llvm.memcpy.i32(i8*, i8*, i32, i32) -> llvm.memcpy.p0i8.p0i8.i32(i8*, i8*, i32, i32, i1)
A update of langref will occur in a subsequent checkin.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99928 91177308-0d34-0410-b5e6-96231b3b80d8
2010-03-30 20:55:56 +00:00
Bob Wilson
76a312b7d1 Revert this change, since it was causing ARM performance regressions.
--- Reverse-merging r98889 into '.':
U    lib/Target/ARM/ARMInstrNEON.td
U    lib/Target/ARM/ARMISelLowering.h
U    lib/Target/ARM/ARMInstrInfo.td
U    lib/Target/ARM/ARMInstrVFP.td
U    lib/Target/ARM/ARMISelLowering.cpp
U    lib/Target/ARM/ARMInstrFormats.td


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99010 91177308-0d34-0410-b5e6-96231b3b80d8
2010-03-19 22:51:32 +00:00
Anton Korobeynikov
341ab138fb Get rid of target-specific fp <-> int nodes when still I'm here.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@98889 91177308-0d34-0410-b5e6-96231b3b80d8
2010-03-18 22:35:45 +00:00
Anton Korobeynikov
f0d500768d Get rid of target-specific nodes for fp16 <-> fp32 conversion.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@98888 91177308-0d34-0410-b5e6-96231b3b80d8
2010-03-18 22:35:37 +00:00
Anton Korobeynikov
bec3dd25c7 Add codegen support for FP16 on ARM
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@98502 91177308-0d34-0410-b5e6-96231b3b80d8
2010-03-14 18:42:31 +00:00
Bob Wilson
9f6c4c141f Use NEON vmin/vmax instructions for floating-point selects.
Radar 7461718.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96572 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-18 06:05:53 +00:00
Jim Grosbach
a87ded2695 tighten up eh.setjmp sequence a bit.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@95603 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-08 23:22:00 +00:00
Evan Cheng
022d9e1cef Revert 95130.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@95160 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-02 23:55:14 +00:00
Evan Cheng
942619695f Pass callsite return type to TargetLowering::LowerCall and use that to check sibcall eligibility.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@95130 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-02 21:29:10 +00:00
Evan Cheng
0c439eb2c8 Eliminate target hook IsEligibleForTailCallOptimization.
Target independent isel should always pass along the "tail call" property. Change
target hook LowerCall's parameter "isTailCall" into a refernce. If the target
decides it's impossible to honor the tail call request, it should set isTailCall
to false to make target independent isel happy.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@94626 91177308-0d34-0410-b5e6-96231b3b80d8
2010-01-27 00:07:07 +00:00
Jim Grosbach
3482c8003a Patch by David Conrad:
"On ARMv6T2 this turns cttz into rbit, clz instead of the 4 instruction
 sequence it is now."




git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@93758 91177308-0d34-0410-b5e6-96231b3b80d8
2010-01-18 19:58:49 +00:00