RPCS3/llvm-mirror
1
0
Fork 0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2024-10-20 03:23:01 +02:00
Code Issues Projects Releases Wiki Activity
45,241 Commits 4 Branches 5 Tags 704 MiB
47778669f5
Commit Graph

1548 Commits

Author SHA1 Message Date
Evan Cheng
46e903d2f6 On x86, if the only use of a i64 load is a i64 store, generate a pair of double load and store instead. 
llvm-svn: 66776
 2009-03-12 05:59:15 +00:00
Chris Lattner
a5368fd283 add no-unwind, remove duplicate run line. 
llvm-svn: 66775
 2009-03-12 05:56:37 +00:00
Chris Lattner
cb52d81eea add nounwinds 
llvm-svn: 66773
 2009-03-12 05:35:33 +00:00
Dan Gohman
d30e108f0e Revert r66024. The JIT encoding for CALLpcrel32 is wrong -- see PR3773, and the 
assembly text output uses an indirect call ("call *") instead of a direct call.

llvm-svn: 66735
 2009-03-11 23:01:47 +00:00
Rafael Espindola
a8fe373200 optimize i8 and i16 tls values. 
llvm-svn: 66725
 2009-03-11 22:40:04 +00:00
Evan Cheng
042e05cb31 My last coalescer fix introduced a subtler one. It's aborting a commuting optimization too late and left the live intervals to be out of sync with instructions. This fixes 8b10b. 
llvm-svn: 66715
 2009-03-11 22:18:44 +00:00
Mon P Wang
287e422039 For yonah, fix a vector shuffle case for v16i8 where we didn't properly clear some bits. 
llvm-svn: 66684
 2009-03-11 18:47:57 +00:00
Mon P Wang
2867737ad2 Fixed a v8i16 shuffle case that should generate a pshufb instead of a pshuflw/hw. 
llvm-svn: 66645
 2009-03-11 06:35:11 +00:00
Chris Lattner
a0c4a99ec2 reapply my previous patch (r66358) with a tweak to set the 
alignment of the generated constant pool entry to the
desired alignment of a type.  If we don't do this, we end up
trying to do movsd from 4-byte alignment memory.  This fixes
450.soplex and 456.hmmer.

llvm-svn: 66641
 2009-03-11 05:08:08 +00:00
Evan Cheng
264173da40 Two coalescer fixes in one. 
1. Use the same value# to represent unknown values being merged into sub-registers.
2. When coalescer commute an instruction and the destination is a physical register, update its sub-registers by merging in the extended ranges.

llvm-svn: 66610
 2009-03-11 00:03:21 +00:00
Bill Wendling
5af061a657 Readd test, but XFAIL it. 
llvm-svn: 66581
 2009-03-10 21:31:00 +00:00
Evan Cheng
535dbf4ffd Revert 66358 for now. It's breaking povray, 450.soplex, and 456.hmmer on x86 / Darwin. 
llvm-svn: 66574
 2009-03-10 20:47:18 +00:00
Bill Wendling
4ddb112b2f Add radar number. 
llvm-svn: 66534
 2009-03-10 06:53:54 +00:00
Chris Lattner
03060f6d50 wire up support for emitting "special" values from inline asm 
format strings with the standard ${:foo} syntax.

llvm-svn: 66527
 2009-03-10 05:37:13 +00:00
Chris Lattner
93662edfb8 Fix PR3763 by using proper APInt methods instead of uint64_t's. 
llvm-svn: 66434
 2009-03-09 20:22:18 +00:00
Evan Cheng
3c9a084a1b ARM isLegalAddressImmediate should check if type is a simple type now that optimizer can create values of funky scalar types. 
llvm-svn: 66429
 2009-03-09 19:15:00 +00:00
Evan Cheng
8b47b553f9 Yet another case where the spiller marked two uses of the same register on the same instruction as kill. This fixes PR3706. 
llvm-svn: 66428
 2009-03-09 19:00:05 +00:00
Evan Cheng
67ccd79e39 Recognize triplets starting with armv5-, armv6- etc. And set the ARM arch version accordingly. 
llvm-svn: 66365
 2009-03-08 04:02:49 +00:00
Evan Cheng
483ece4d2e If a MI uses the same register more than once, only mark one of them as 'kill'. 
llvm-svn: 66363
 2009-03-08 03:58:35 +00:00
Chris Lattner
578b634f56 implement an optimization to codegen c ? 1.0 : 2.0 as load { 2.0, 1.0 } + c*4. 
For 2009-03-07-FPConstSelect.ll we now produce:

_f:
	xorl	%eax, %eax
	testl	%edi, %edi
	movl	$4, %ecx
	cmovne	%rax, %rcx
	leaq	LCPI1_0(%rip), %rax
	movss	(%rcx,%rax), %xmm0
	ret

previously we produced:

_f:
	subl	$4, %esp
	cmpl	$0, 8(%esp)
	movss	LCPI1_0, %xmm0
	je	LBB1_2	## entry
LBB1_1:	## entry
	movss	LCPI1_1, %xmm0
LBB1_2:	## entry
	movss	%xmm0, (%esp)
	flds	(%esp)
	addl	$4, %esp
	ret

on PPC the code also improves to:

_f:
	cntlzw r2, r3
	srwi r2, r2, 5
	li r3, lo16(LCPI1_0)
	slwi r2, r2, 2
	addis r3, r3, ha16(LCPI1_0)
	lfsx f1, r3, r2
	blr 

from:

_f:
	li r2, lo16(LCPI1_1)
	cmplwi cr0, r3, 0
	addis r2, r2, ha16(LCPI1_1)
	beq cr0, LBB1_2	; entry
LBB1_1:	; entry
	li r2, lo16(LCPI1_0)
	addis r2, r2, ha16(LCPI1_0)
LBB1_2:	; entry
	lfs f1, 0(r2)
	blr 

This also improves the existing pic-cpool case from:

foo:
	subl	$12, %esp
	call	.Lllvm$1.$piclabel
.Lllvm$1.$piclabel:
	popl	%eax
	addl	$_GLOBAL_OFFSET_TABLE_ + [.-.Lllvm$1.$piclabel], %eax
	cmpl	$0, 16(%esp)
	movsd	.LCPI1_0@GOTOFF(%eax), %xmm0
	je	.LBB1_2	# entry
.LBB1_1:	# entry
	movsd	.LCPI1_1@GOTOFF(%eax), %xmm0
.LBB1_2:	# entry
	movsd	%xmm0, (%esp)
	fldl	(%esp)
	addl	$12, %esp
	ret

to:

foo:
	call	.Lllvm$1.$piclabel
.Lllvm$1.$piclabel:
	popl	%eax
	addl	$_GLOBAL_OFFSET_TABLE_ + [.-.Lllvm$1.$piclabel], %eax
	xorl	%ecx, %ecx
	cmpl	$0, 4(%esp)
	movl	$8, %edx
	cmovne	%ecx, %edx
	fldl	.LCPI1_0@GOTOFF(%eax,%edx)
	ret

This triggers a few dozen times in spec FP 2000.

llvm-svn: 66358
 2009-03-08 01:51:30 +00:00
Dan Gohman
b9c32f1aca Arithmetic instructions don't set EFLAGS bits OF and CF bits 
the same say the "test" instruction does in overflow cases,
so eliminating the test is only safe when those bits aren't
needed, as is the case for COND_E and COND_NE, or if it
can be proven that no overflow will occur. For now, just
restrict the optimization to COND_E and COND_NE and don't
do any overflow analysis.

llvm-svn: 66318
 2009-03-07 01:58:32 +00:00
Dan Gohman
42049dafd7 Fix ScheduleDAGRRList::CopyAndMoveSuccessors' handling of nodes 
with multiple chain operands. This can occur when the scheduler
has added chain operands to a node that already has a chain
operand, in order to handle physical register dependencies.

This fixes an llvm-gcc bootstrap failure on x86-64 introduced
in r66058.

llvm-svn: 66240
 2009-03-06 02:23:01 +00:00
Dan Gohman
f6f684b206 Fix the "test" optimization to recognize "dec" as an add of 
negative one, as subtracts of immediates are canonicalized
to adds.

llvm-svn: 66180
 2009-03-05 19:32:48 +00:00
Dan Gohman
5e88bf4c00 Make this test more thorough. Not only should there be no %esi, 
there should be no spilling of anything.

llvm-svn: 66179
 2009-03-05 19:31:32 +00:00
Evan Cheng
24c138a1cd Do not split edges to EH landing pads. It will cause code size explosion. 
llvm-svn: 66140
 2009-03-05 06:31:26 +00:00
Dan Gohman
31fb085c2e Re-apply 66008, now that the unfoldMemoryOperand bug is fixed. 
llvm-svn: 66058
 2009-03-04 19:44:21 +00:00
Owen Anderson
fb7f64ea0d Add a restore folder, which shaves a dozen or so machineinstrs off oggenc. Update a testcase to check this. 
llvm-svn: 66029
 2009-03-04 08:52:31 +00:00
Evan Cheng
7d9019d0f3 Fix PR3666: isel calls to constant addresses. 
llvm-svn: 66024
 2009-03-04 06:48:53 +00:00
Eli Friedman
1bed40b86a PR3686: make the legalizer handle bitcast from i80 to x86 long double. 
llvm-svn: 66021
 2009-03-04 06:23:34 +00:00
Dan Gohman
6831e2c2a6 Revert r66004 for now; it's causing a variety of test failures. 
llvm-svn: 66008
 2009-03-04 03:54:19 +00:00
Evan Cheng
32eef2f73f Rename test. 
llvm-svn: 66006
 2009-03-04 02:47:25 +00:00
Dan Gohman
c6c669cc1e Teach the x86 backend to eliminate "test" instructions by using the EFLAGS 
result from add, sub, inc, and dec instructions in simple cases.

llvm-svn: 66004
 2009-03-04 02:33:24 +00:00
Evan Cheng
db402a7a49 Fix PR3701. 1. X86 target renamed eflags register to flags. This matches what llvm-gcc generates so codegen knows flags register is being clobbered by inline asm. 2. BURR scheduler should also check if inline asm nodes can clobber "live" physical registers. Previously it was only checking target nodes with implicit defs. 
llvm-svn: 65996
 2009-03-04 01:41:49 +00:00
Bill Wendling
93eeea0493 The DAG combiner was performing a BT combine. The BT combine had a value of -1, 
so it changed it into a 31 via the TLO.ShrinkDemandedConstant() call. Then it
would go through the DAG combiner again. This time it had a value of 31, which
was turned into a -1 by TLI.SimplifyDemandedBits(). This would ping pong
forever.

Teach the TLO.ShrinkDemandedConstant() call not to lower a value if the demanded
value is an XOR of all ones.

llvm-svn: 65985
 2009-03-04 00:18:06 +00:00
Nate Begeman
6b41d33726 Fix a problem with DAGCombine on 64b targets where folding 
extracts + build_vector into a shuffle would fail, because the
type of the new build_vector would not be legal.  Try harder to
create a legal build_vector type.  Note: this will be totally 
irrelevant once vector_shuffle no longer takes a build_vector for
shuffle mask.

New:
_foo:
	xorps	%xmm0, %xmm0
	xorps	%xmm1, %xmm1
	subps	%xmm1, %xmm1
	mulps	%xmm0, %xmm1
	addps	%xmm0, %xmm1
	movaps	%xmm1, 0

Old:
_foo:
	xorps	%xmm0, %xmm0
	movss	%xmm0, %xmm1
	xorps	%xmm2, %xmm2
	unpcklps	%xmm1, %xmm2
	pshufd	$80, %xmm1, %xmm1
	unpcklps	%xmm1, %xmm2
	pslldq	$16, %xmm2
	pshufd	$57, %xmm2, %xmm1
	subps	%xmm0, %xmm1
	mulps	%xmm0, %xmm1
	addps	%xmm0, %xmm1
	movaps	%xmm1, 0

llvm-svn: 65791
 2009-03-01 23:44:07 +00:00
Evan Cheng
276f9b02c5 Minor optimization: 
Look for situations like this:                                                                                                                                                              
%reg1024<def> = MOV r1                                                                                                                                                                      
%reg1025<def> = MOV r0                                                                                                                                                                      
%reg1026<def> = ADD %reg1024, %reg1025                                                                                                                                                      
r0            = MOV %reg1026                                                                                                                                                                
Commute the ADD to hopefully eliminate an otherwise unavoidable copy.

llvm-svn: 65752
 2009-03-01 02:03:43 +00:00
Evan Cheng
9c3ce7905e Last commit accidentially deleted this code. 
llvm-svn: 65679
 2009-02-28 06:02:14 +00:00
Rafael Espindola
880e63bf01 Refactor TLS code and add some tests. The tests and expected results are: 
pic |  declaration | linkage  | visibility |

!pic |  declaration | external | default    | tls1.ll     tls2.ll     | local exec
 pic |  declaration | external | default    | tls1-pic.ll tls2-pic.ll | general dynamic
!pic | !declaration | external | default    | tls3.ll     tls4.ll     | initial exec
 pic | !declaration | external | default    | tls3-pic.ll tls4-pic.ll | general dynamic

!pic |  declaration | external | hidden     | tls7.ll     tls8.ll     | local exec
 pic |  declaration | external | hidden     | X                       | local dynamic
!pic | !declaration | external | hidden     | tls9.ll     tls10.ll    | local exec
 pic | !declaration | external | hidden     | X                       | local dynamic

!pic |  declaration | internal | default    | tls5.ll     tls6.ll     | local exec
 pic |  declaration | internal | default    | X                       | local dynamic

The ones marked with an X have not been implemented since local dynamic is not implemented.

llvm-svn: 65632
 2009-02-27 13:37:18 +00:00
Evan Cheng
257e065df6 Make sure this test passes on linux-ppc. 
llvm-svn: 65600
 2009-02-27 00:51:50 +00:00
Evan Cheng
0ca6e3dba5 MachineLICM CSE should match destination register classes; avoid hoisting implicit_def's. 
llvm-svn: 65592
 2009-02-27 00:02:22 +00:00
Evan Cheng
4014a9a5b8 ADDS{D|S}rr_Int and MULS{D|S}rr_Int are not commutable. The users of these intrinsics expect the high bits will not be modified. 
llvm-svn: 65499
 2009-02-26 03:12:02 +00:00
Evan Cheng
86fc9440db The last commit was overly conservative. It's ok to reuse value that's already marked livein. 
llvm-svn: 65498
 2009-02-26 03:02:21 +00:00
Evan Cheng
ec34226c2b Revert BuildVectorSDNode related patches: 65426, 65427, and 65296. 
llvm-svn: 65482
 2009-02-25 22:49:59 +00:00
Dan Gohman
3766eeea36 Fast-isel can't do TLS yet, so it should fall back to SDISel 
if it sees TLS addresses.

llvm-svn: 65341
 2009-02-23 22:03:08 +00:00
Dan Gohman
c1229a3a86 Use the -stack-alignment option instead of using a target triple 
for avoiding dynamic stack realignment.

llvm-svn: 65319
 2009-02-23 16:34:46 +00:00
Evan Cheng
dd139e795c Only v1i16 (i.e. _m64) is returned via RAX / RDX. 
llvm-svn: 65313
 2009-02-23 09:03:22 +00:00
Nate Begeman
7d9e99560f Make this test use darwin targe triple, to avoid stack traffic on linux. 
llvm-svn: 65312
 2009-02-23 09:03:06 +00:00
Nate Begeman
e0093d2501 Generate better code for v8i16 shuffles on SSE2 
Generate better code for v16i8 shuffles on SSE2 (avoids stack)
Generate pshufb for v8i16 and v16i8 shuffles on SSSE3 where it is fewer uops.
Document the shuffle matching logic and add some FIXMEs for later further
  cleanups.
New tests that test the above.

Examples:

New:
_shuf2:
	pextrw	$7, %xmm0, %eax
	punpcklqdq	%xmm1, %xmm0
	pshuflw	$128, %xmm0, %xmm0
	pinsrw	$2, %eax, %xmm0

Old:
_shuf2:
	pextrw	$2, %xmm0, %eax
	pextrw	$7, %xmm0, %ecx
	pinsrw	$2, %ecx, %xmm0
	pinsrw	$3, %eax, %xmm0
	movd	%xmm1, %eax
	pinsrw	$4, %eax, %xmm0
	ret

=========

New:
_shuf4:
	punpcklqdq	%xmm1, %xmm0
	pshufb	LCPI1_0, %xmm0

Old:
_shuf4:
	pextrw	$3, %xmm0, %eax
	movsd	%xmm1, %xmm0
	pextrw	$3, %xmm1, %ecx
	pinsrw	$4, %ecx, %xmm0
	pinsrw	$5, %eax, %xmm0

========

New:
_shuf1:
	pushl	%ebx
	pushl	%edi
	pushl	%esi
	pextrw	$1, %xmm0, %eax
	rolw	$8, %ax
	movd	%xmm0, %ecx
	rolw	$8, %cx
	pextrw	$5, %xmm0, %edx
	pextrw	$4, %xmm0, %esi
	pextrw	$3, %xmm0, %edi
	pextrw	$2, %xmm0, %ebx
	movaps	%xmm0, %xmm1
	pinsrw	$0, %ecx, %xmm1
	pinsrw	$1, %eax, %xmm1
	rolw	$8, %bx
	pinsrw	$2, %ebx, %xmm1
	rolw	$8, %di
	pinsrw	$3, %edi, %xmm1
	rolw	$8, %si
	pinsrw	$4, %esi, %xmm1
	rolw	$8, %dx
	pinsrw	$5, %edx, %xmm1
	pextrw	$7, %xmm0, %eax
	rolw	$8, %ax
	movaps	%xmm1, %xmm0
	pinsrw	$7, %eax, %xmm0
	popl	%esi
	popl	%edi
	popl	%ebx
	ret

Old:
_shuf1:
	subl	$252, %esp
	movaps	%xmm0, (%esp)
	movaps	%xmm0, 16(%esp)
	movaps	%xmm0, 32(%esp)
	movaps	%xmm0, 48(%esp)
	movaps	%xmm0, 64(%esp)
	movaps	%xmm0, 80(%esp)
	movaps	%xmm0, 96(%esp)
	movaps	%xmm0, 224(%esp)
	movaps	%xmm0, 208(%esp)
	movaps	%xmm0, 192(%esp)
	movaps	%xmm0, 176(%esp)
	movaps	%xmm0, 160(%esp)
	movaps	%xmm0, 144(%esp)
	movaps	%xmm0, 128(%esp)
	movaps	%xmm0, 112(%esp)
	movzbl	14(%esp), %eax
	movd	%eax, %xmm1
	movzbl	22(%esp), %eax
	movd	%eax, %xmm2
	punpcklbw	%xmm1, %xmm2
	movzbl	42(%esp), %eax
	movd	%eax, %xmm1
	movzbl	50(%esp), %eax
	movd	%eax, %xmm3
	punpcklbw	%xmm1, %xmm3
	punpcklbw	%xmm2, %xmm3
	movzbl	77(%esp), %eax
	movd	%eax, %xmm1
	movzbl	84(%esp), %eax
	movd	%eax, %xmm2
	punpcklbw	%xmm1, %xmm2
	movzbl	104(%esp), %eax
	movd	%eax, %xmm1
	punpcklbw	%xmm1, %xmm0
	punpcklbw	%xmm2, %xmm0
	movaps	%xmm0, %xmm1
	punpcklbw	%xmm3, %xmm1
	movzbl	127(%esp), %eax
	movd	%eax, %xmm0
	movzbl	135(%esp), %eax
	movd	%eax, %xmm2
	punpcklbw	%xmm0, %xmm2
	movzbl	155(%esp), %eax
	movd	%eax, %xmm0
	movzbl	163(%esp), %eax
	movd	%eax, %xmm3
	punpcklbw	%xmm0, %xmm3
	punpcklbw	%xmm2, %xmm3
	movzbl	188(%esp), %eax
	movd	%eax, %xmm0
	movzbl	197(%esp), %eax
	movd	%eax, %xmm2
	punpcklbw	%xmm0, %xmm2
	movzbl	217(%esp), %eax
	movd	%eax, %xmm4
	movzbl	225(%esp), %eax
	movd	%eax, %xmm0
	punpcklbw	%xmm4, %xmm0
	punpcklbw	%xmm2, %xmm0
	punpcklbw	%xmm3, %xmm0
	punpcklbw	%xmm1, %xmm0
	addl	$252, %esp
	ret

llvm-svn: 65311
 2009-02-23 08:49:38 +00:00
Scott Michel
3f8637305f Introduce the BuildVectorSDNode class that encapsulates the ISD::BUILD_VECTOR 
instruction. The class also consolidates the code for detecting constant
splats that's shared across PowerPC and the CellSPU backends (and might be
useful for other backends.) Also introduces SelectionDAG::getBUID_VECTOR() for
generating new BUILD_VECTOR nodes.

llvm-svn: 65296
 2009-02-22 23:36:09 +00:00
Richard Pennington
72dc7c621e bug 3610: Test case. 
llvm-svn: 65287
 2009-02-22 15:54:44 +00:00