RPCS3/llvm-mirror
1
0
Fork 0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2025-01-02 09:01:37 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity
36,510 Commits 4 Branches 5 Tags 634 MiB
cd28ef8950
Commit Graph

36510 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Nick Lewycky
cd28ef8950 Be more careful modifying the use_list while also iterating through it. 
llvm-svn: 46417
 2008-01-27 18:35:00 +00:00
Duncan Sands
e77256b325 Revert r46393: readonly/readnone functions are no 
longer allowed to write through byval arguments.

llvm-svn: 46416
 2008-01-27 18:12:58 +00:00
Chris Lattner
2ab1fd3824 Implement some dag combines that allow doing fneg/fabs/fcopysign in integer 
registers if used by a bitconvert or using a bitconvert.  This allows us to
avoid constant pool loads and use cheaper integer instructions when the
values come from or end up in integer regs anyway.  For example, we now 
compile CodeGen/X86/fp-in-intregs.ll to:

_test1:
	movl	$2147483648, %eax
	xorl	4(%esp), %eax
	ret
_test2:
	movl	$1065353216, %eax
	orl	4(%esp), %eax
	andl	$3212836864, %eax
	ret

Instead of:
_test1:
	movss	4(%esp), %xmm0
	xorps	LCPI2_0, %xmm0
	movd	%xmm0, %eax
	ret
_test2:
	movss	4(%esp), %xmm0
	andps	LCPI3_0, %xmm0
	movss	LCPI3_1, %xmm1
	andps	LCPI3_2, %xmm1
	orps	%xmm0, %xmm1
	movd	%xmm1, %eax
	ret

bitconverts can happen due to various calling conventions that require
fp values to passed in integer regs in some cases, e.g. when returning
a complex.

llvm-svn: 46414
 2008-01-27 17:42:27 +00:00
Chris Lattner
39c52e030b add a note 
llvm-svn: 46413
 2008-01-27 07:31:41 +00:00
Chris Lattner
f4bc2c5718 Use fldz and fld1 for long double constants instead of a constant pool load. 
llvm-svn: 46411
 2008-01-27 06:19:31 +00:00
Chris Lattner
9c66a1b961 make isExactlyValue work for long double. 
llvm-svn: 46410
 2008-01-27 06:19:08 +00:00
Bill Wendling
0e2b8c2c45 The CorrelatedExpressions pass is now no more. 
llvm-svn: 46409
 2008-01-27 06:13:32 +00:00
Bill Wendling
629a569ce9 The CorrelatedExpressionElimination pass is known to be buggy. Remove it. 
This fixes PR1769.

llvm-svn: 46408
 2008-01-27 06:11:41 +00:00
Chris Lattner
f93e3d7dd7 For long double constants, print an approximation of their value to the .s file to make it easier to read. 
llvm-svn: 46407
 2008-01-27 06:09:28 +00:00
Chris Lattner
aa553aa0c1 Fold fptrunc(add (fpextend x), (fpextend y)) -> add(x,y), as GCC does. 
llvm-svn: 46406
 2008-01-27 05:29:54 +00:00
Chris Lattner
00183edf55 Add some notes. 
llvm-svn: 46405
 2008-01-26 20:12:07 +00:00
Chris Lattner
6124c0eb2d Remove some code for inferring alignment info from the x86 backend 
now that the dag combiner does it.

llvm-svn: 46404
 2008-01-26 20:07:42 +00:00
Chris Lattner
e66aea6532 New test to verify that "merging 4 loads into a vec load" continues to work and 
continues to infer alignment info.

llvm-svn: 46403
 2008-01-26 20:06:45 +00:00
Chris Lattner
266b518f13 make this 64-bit safe 
llvm-svn: 46402
 2008-01-26 19:50:30 +00:00
Chris Lattner
682346a7b0 Infer alignment of loads and increase their alignment when we can tell they are 
from the stack.  This allows us to compile stack-align.ll to:

_test:
	movsd	LCPI1_0, %xmm0
	movapd	%xmm0, %xmm1
***	andpd	4(%esp), %xmm1
	andpd	_G, %xmm0
	addsd	%xmm1, %xmm0
	movl	20(%esp), %eax
	movsd	%xmm0, (%eax)
	ret

instead of:

_test:
	movsd	LCPI1_0, %xmm0
**	movsd	4(%esp), %xmm1
**	andpd	%xmm0, %xmm1
	andpd	_G, %xmm0
	addsd	%xmm1, %xmm0
	movl	20(%esp), %eax
	movsd	%xmm0, (%eax)
	ret

llvm-svn: 46401
 2008-01-26 19:45:50 +00:00
Chris Lattner
f0c3240135 remove a useless xfailed test. 
llvm-svn: 46400
 2008-01-26 19:35:46 +00:00
Duncan Sands
9fae964ef7 Invert this test, because it is wrong if we allow 
readonly functions to use byval parameters as local
storage (how much do we want this?).

llvm-svn: 46399
 2008-01-26 12:33:01 +00:00
Bill Wendling
7b83688c73 If there's no instructions being emitted on X86 for a function, emit a 
nop. Emit the nop directly for PPC.

llvm-svn: 46398
 2008-01-26 09:03:52 +00:00
Bill Wendling
26fb9335f5 Need to convert to LLVM code and not C. 
llvm-svn: 46397
 2008-01-26 06:56:08 +00:00
Bill Wendling
3e622b88b6 Rename the .c to .ll 
llvm-svn: 46396
 2008-01-26 06:53:40 +00:00
Bill Wendling
7151e8d92c Move testcase to the code gen directory. 
llvm-svn: 46395
 2008-01-26 06:53:06 +00:00
Bill Wendling
1c92468074 If there are no machine instructions emitted for a function, then insert 
a "nop" instruction so that we don't have the function's label associated
with something that it's not supposed to be associated with.

llvm-svn: 46394
 2008-01-26 06:51:24 +00:00
Duncan Sands
792234c366 Create an explicit copy for byval parameters even 
when inlining a readonly function.

llvm-svn: 46393
 2008-01-26 06:41:49 +00:00
Bill Wendling
1e56a2ffb6 If we have a function like this: 
void bork() {
  int *address = 0;
  *address = 0;
}

It's compiled into LLVM code that looks like this:

define void @bork() noreturn nounwind  {
entry:
        unreachable
}

This is bad on some platforms (like PPC) because it will generate the label for
the function but no body. The label could end up being associated with some
non-code related stuff, like a section. This places a "trap" instruction if the
SimplifyCFG pass removed all code from the function leaving only one
"unreachable" instruction.

llvm-svn: 46387
 2008-01-26 01:43:44 +00:00
Devang Patel
c40820e322 Add another testcase. 
llvm-svn: 46385
 2008-01-26 01:21:48 +00:00
Chris Lattner
53a98f46fd Fix some bugs in SimplifyNodeWithTwoResults where it would call deletenode to 
delete a node even if it was not dead in some cases.  Instead, just add it to
the worklist.  Also, make sure to use the CombineTo methods, as it was doing
things that were unsafe: the top level combine loop could touch dangling memory.

This fixes CodeGen/Generic/2008-01-25-dag-combine-mul.ll

llvm-svn: 46384
 2008-01-26 01:09:19 +00:00
Chris Lattner
8d8b31d5f4 don't bother making x&-1 only to simplify it in dag combine. This commonly occurs expanding i64 ops. 
llvm-svn: 46383
 2008-01-26 01:05:42 +00:00
Evan Cheng
e62e9a8d96 New test case. 
llvm-svn: 46382
 2008-01-26 00:35:43 +00:00
Chris Lattner
6b56f4031f reduce indentation 
llvm-svn: 46377
 2008-01-25 23:34:24 +00:00
Chris Lattner
c2df169459 add a testcase for a bug Duncan pointed out. 
llvm-svn: 46372
 2008-01-25 22:36:24 +00:00
Duncan Sands
10316ffbb2 Do this more neatly. 
llvm-svn: 46369
 2008-01-25 22:06:51 +00:00
Chris Lattner
f2a0e6c176 Reorder a field to reduce the size of StackObject. Note that this 
may require a clean rebuild on leopard. :(

llvm-svn: 46360
 2008-01-25 17:49:41 +00:00
Duncan Sands
ced29554f7 Test for PR1942. 
llvm-svn: 46357
 2008-01-25 17:36:44 +00:00
Chris Lattner
8c621a474e fix long lines. 
llvm-svn: 46355
 2008-01-25 17:24:52 +00:00
Chris Lattner
00ead854ef JITEmitter.cpp was trying to sync the icache for function stubs, but 
was actually passing a completely incorrect size to sys_icache_invalidate.
Instead of having the JITEmitter do this (which doesn't have the correct 
size), just make the target sync its own stubs.

llvm-svn: 46354
 2008-01-25 16:41:09 +00:00
Duncan Sands
b9f1e3df90 Add more assertions to catch accesses outside of 
arrays.  Also, as a convenience, don't barf, just
return false, if someone calls isTruncStoreLegal
or isLoadXLegal with an extended type for the in
memory type.

llvm-svn: 46352
 2008-01-25 10:20:53 +00:00
Owen Anderson
a4ff15c69f DeadStoreElimination can treat byval parameters as if there were alloca's for the purpose of removing end-of-function stores. 
llvm-svn: 46351
 2008-01-25 10:10:33 +00:00
Chris Lattner
93f4f5970c move this field back. Moving the field causes miscompilations (!) of voronoi and others. 
llvm-svn: 46350
 2008-01-25 07:29:34 +00:00
Chris Lattner
59dc439330 Add skeletal code to increase the alignment of loads and stores when 
we can infer it.  This will eventually help stuff, though it doesn't
do much right now because all fixed FI's have an alignment of 1.

llvm-svn: 46349
 2008-01-25 07:20:16 +00:00
Chris Lattner
1ce075bcaa move MachineFrameInfo::CreateFixedObject out of line, give MachineFrameInfo 
a reference to TargetFrameInfo.  Rearrange order of fields in StackObject to
save a word.

llvm-svn: 46348
 2008-01-25 07:19:06 +00:00
Chris Lattner
74d4792eb2 include alignment and volatility information in -view-*-dags output 
llvm-svn: 46347
 2008-01-25 06:40:45 +00:00
Chris Lattner
72795598b9 add a fixme. 
llvm-svn: 46346
 2008-01-25 06:20:20 +00:00
Chris Lattner
726a4e45e5 optimize fxor like for 
llvm-svn: 46345
 2008-01-25 06:14:17 +00:00
Chris Lattner
79076fdf2a Add target-specific dag combines for FAND(x,0) and FOR(x,0). This allows 
us to compile:

double test(double X) {
  return copysign(0.0, X);
}

into:

_test:
	andpd	LCPI1_0(%rip), %xmm0
	ret

instead of:
_test:
	pxor	%xmm1, %xmm1
	andpd	LCPI1_0(%rip), %xmm1
	movapd	%xmm0, %xmm2
	andpd	LCPI1_1(%rip), %xmm2
	movapd	%xmm1, %xmm0
	orpd	%xmm2, %xmm0
	ret

llvm-svn: 46344
 2008-01-25 05:46:26 +00:00
Anton Korobeynikov
37309ed741 Provide correct DWARF register numbering for debug information emission on x86-32/Darwin. 
This should fix bunch of issues.

llvm-svn: 46337
 2008-01-25 00:34:13 +00:00
Devang Patel
587591ba1b New test. 
llvm-svn: 46333
 2008-01-24 23:55:34 +00:00
Chris Lattner
6dfec0fb0e Don't dump the function! 
llvm-svn: 46320
 2008-01-24 19:28:11 +00:00
Chris Lattner
2c3646ea4a getUnderlyingObject can return null, handle this. 
llvm-svn: 46318
 2008-01-24 19:07:10 +00:00
Ted Kremenek
274df5b7db Added additional overloaded operators for APSInt to match the operators of 
APInt.

While some operators were already specifically overloaded for APSInt, others
resulted in using the overloaded operator methods in APInt, which would result
in the signedness bit being lost.

Modified the APSInt(APInt&) constructor to be "explicit" and to take an
extra (optional) flag to indicate the signedness.  Making the ctor explicit
will catch any implicit conversations between APSInt -> APInt -> APSInt that
results in the signedness flag being lost.

llvm-svn: 46316
 2008-01-24 18:59:52 +00:00
Chris Lattner
cd5013eb2f Teach basicaa that 'byval' arguments define a new memory location that 
can't be aliased to other known objects.  This allows us to know that byval 
pointer args don't alias globals, etc.

llvm-svn: 46315
 2008-01-24 18:00:32 +00:00