Commit Graph

38355 Commits

Author SHA1 Message Date
Anton Korobeynikov
dc28bd43d3 Only allow increase of max alignment value
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50168 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-23 18:23:50 +00:00
Anton Korobeynikov
d52bdaff0c Be over-conservative: scan for all used virtual registers and calculate maximal stack alignment in assumption, that there will be spill of vector register.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50167 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-23 18:23:30 +00:00
Anton Korobeynikov
856914fe00 Add X86 Maximal Stack Alignment Calculator Pass before RA
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50166 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-23 18:23:05 +00:00
Anton Korobeynikov
769b481e9f Add facility for pre-RA passes
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50165 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-23 18:22:28 +00:00
Anton Korobeynikov
aadc780a56 Use precomputed value, if any
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50164 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-23 18:21:50 +00:00
Anton Korobeynikov
2c430cbf03 Do proper book-keeping of offsets and prologue/epilogue code for stack realignment
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50163 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-23 18:21:27 +00:00
Anton Korobeynikov
8e91ec5e29 If stack realignment is used - incoming args will use EBP as base register and locals - ESP
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50162 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-23 18:21:02 +00:00
Anton Korobeynikov
b51dce3013 Eastimate required stack alignment early, so we can decide, whether we will need frame pointer or not
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50161 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-23 18:20:17 +00:00
Anton Korobeynikov
874384e20f Cleanup
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50160 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-23 18:19:47 +00:00
Anton Korobeynikov
d1c133a051 Cleanup
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50159 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-23 18:19:23 +00:00
Anton Korobeynikov
82751e3eda Simplify
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50158 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-23 18:18:36 +00:00
Anton Korobeynikov
45709ae09b Make stack alignment options global for all targets
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50157 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-23 18:18:10 +00:00
Anton Korobeynikov
cfcd20e7be Provide option for enabling-disabling stack realignment
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50156 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-23 18:17:11 +00:00
Anton Korobeynikov
35410a493b Disable stack realignment for functions with dynamic-sized alloca's
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50155 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-23 18:16:43 +00:00
Anton Korobeynikov
890fe888be Provide ABI-correct stack alignment
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50154 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-23 18:16:16 +00:00
Anton Korobeynikov
9bbbea568c Provide convenient helpers for some operations
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50153 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-23 18:15:48 +00:00
Anton Korobeynikov
e201190eb7 Whitespace cleanup
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50152 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-23 18:15:11 +00:00
Dan Gohman
55888274f2 Fix some whitespace.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50151 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-23 17:50:15 +00:00
Bill Wendling
480311b548 Use the Create static method for some of these objects. They no longer
have public constructors. This should fix PR2246.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50145 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-23 07:13:17 +00:00
Chris Lattner
4af6ad3789 simplify code for propagation of constant arguments into
callees.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50142 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-23 06:16:27 +00:00
Chris Lattner
e9625c6c88 Fix a number of bugs in ipconstantprop, simplify the code, fit in 80 cols,
fix read after free bug (PR2238).


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50141 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-23 05:59:23 +00:00
Chris Lattner
c6ee00b8ee Rewrite multiple return value handling in SCCP. Before, the -sccp pass
would turn every getresult instruction into undef.  This helps with
rdar://5778210


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50140 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-23 05:38:20 +00:00
Chris Lattner
a925a14698 regenerate
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50139 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-23 05:37:08 +00:00
Chris Lattner
f7dedf2d85 Validate that the result of a function type is valid using shared
logic with vmcore.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50138 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-23 05:36:58 +00:00
Chris Lattner
d52b62ae05 Enforce that multiple return values have to have at least one result.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50137 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-23 05:36:34 +00:00
Chris Lattner
69e6317083 remove this testcase. It isn't testing loop rotate, it is testing all
of -std-compile-opts and is now failing because other passes are generating
IR that looks different to input of loop rotate.  Devang, please 
introduce a testcase that only runs loop rotate.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50136 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-23 05:36:04 +00:00
Chris Lattner
7e4687a9f2 returning an empty multiple return list is not valid.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50135 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-23 05:29:14 +00:00
Chris Lattner
f4cde4ec81 fix description of 'ret' to be more correct with multiple return
values.  Clarify that it is impossible to 'multiply return' a 
struct with zero elements.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50131 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-23 04:59:35 +00:00
Chris Lattner
1c406d77a6 Allow undef as an operand to getresult.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50130 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-23 04:06:52 +00:00
Chris Lattner
42369b7e4b Verify that the operand of a getresult instruction is a
call/invoke or undef.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50129 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-23 04:06:15 +00:00
Chris Lattner
525af20937 make this test more interesting.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50128 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-23 03:49:32 +00:00
Chris Lattner
ed98501ab7 distill down the essense of this test.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50125 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-23 03:03:42 +00:00
Dale Johannesen
02dc1a7f1f new test
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50123 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-23 01:22:22 +00:00
Dale Johannesen
63e7eb4e12 Do not change the type of a ByVal argument to a
type of a different size.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50121 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-23 01:03:05 +00:00
Evan Cheng
f30752cbcc Don't do: "(X & 4) >> 1 == 2 --> (X & 4) == 4" if there are more than one uses of the shift result.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50118 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-23 00:38:06 +00:00
Chris Lattner
a5ddb59a13 Start doing the significantly useful part of jump threading: handle cases
where a comparison has a phi input and that phi is a constant.  For example,
stuff like:

  Threading edge through bool from 'bb2149' to 'bb2231' with cost: 1, across block:
bb2237:		; preds = %bb2231, %bb2149
	%tmp2328.rle = phi i32 [ %tmp2232, %bb2231 ], [ %tmp2232439, %bb2149 ]		; <i32> [#uses=2]
	%done.0 = phi i32 [ %done.2, %bb2231 ], [ 0, %bb2149 ]		; <i32> [#uses=1]
	%tmp2239 = icmp eq i32 %done.0, 0		; <i1> [#uses=1]
	br i1 %tmp2239, label %bb2231, label %bb2327

or

bb38.i298:		; preds = %bb33.i295, %bb1693
	%tmp39.i296.rle = phi %struct.ibox* [ null, %bb1693 ], [ %tmp39.i296.rle1109, %bb33.i295 ]		; <%struct.ibox*> [#uses=2]
	%minspan.1.i291.reg2mem.1 = phi i32 [ 32000, %bb1693 ], [ %minspan.0.i288, %bb33.i295 ]		; <i32> [#uses=1]
	%tmp40.i297 = icmp eq %struct.ibox* %tmp39.i296.rle, null		; <i1> [#uses=1]
	br i1 %tmp40.i297, label %implfeeds.exit311, label %bb43.i301

This triggers thousands of times in spec.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50110 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-22 21:40:39 +00:00
Chris Lattner
ae65b3c791 Dig through multiple levels of AND to thread jumps if needed.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50106 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-22 20:46:09 +00:00
Owen Anderson
d2368dc344 Move MemCpyOpt after GVN.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50097 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-22 07:12:26 +00:00
Chris Lattner
6bf77500c6 Teach jump threading to thread through blocks like:
br (and X, phi(Y, Z, false)), label L1, label L2

This triggers once on 252.eon and 6 times on 176.gcc.  Blocks 
in question often look like this:

bb262:		; preds = %bb261, %bb248
	%iftmp.251.0 = phi i1 [ true, %bb261 ], [ false, %bb248 ]		; <i1> [#uses=4]
	%tmp270 = icmp eq %struct.rtx_def* %tmp.0.i, null		; <i1> [#uses=1]
	%bothcond = or i1 %iftmp.251.0, %tmp270		; <i1> [#uses=1]
	br i1 %bothcond, label %bb288, label %bb273

In this case, it is clear that it doesn't matter if tmp.0.i is null when coming from bb261.  When coming from bb248, it is all that matters.


Another random example:

check_asm_operands.exit:		; preds = %check_asm_operands.exit.thr_comm, %bb30.i, %bb12.i, %bb6.i413
	%tmp.0.i420 = phi i1 [ true, %bb6.i413 ], [ true, %bb12.i ], [ true, %bb30.i ], [ false, %check_asm_operands.exit.thr_comm ; <i1> [#uses=1]
	call void @llvm.stackrestore( i8* %savedstack ) nounwind 
	%tmp4389 = icmp eq i32 %added_sets_1.0, 0		; <i1> [#uses=1]
	%tmp4394 = icmp eq i32 %added_sets_2.0, 0		; <i1> [#uses=1]
	%bothcond80 = and i1 %tmp4389, %tmp4394		; <i1> [#uses=1]
	%bothcond81 = and i1 %bothcond80, %tmp.0.i420		; <i1> [#uses=1]
	br i1 %bothcond81, label %bb4398, label %bb4397

Here is the case from 252.eon:

bb290.i.i:		; preds = %bb23.i57.i.i, %bb8.i39.i.i, %bb100.i.i, %bb100.i.i, %bb85.i.i110
	%myEOF.1.i.i = phi i1 [ true, %bb100.i.i ], [ true, %bb100.i.i ], [ true, %bb85.i.i110 ], [ true, %bb8.i39.i.i ], [ false, %bb23.i57.i.i ]		; <i1> [#uses=2]
	%i.4.i.i = phi i32 [ %i.1.i.i, %bb85.i.i110 ], [ %i.0.i.i, %bb100.i.i ], [ %i.0.i.i, %bb100.i.i ], [ %i.3.i.i, %bb8.i39.i.i ], [ %i.3.i.i, %bb23.i57.i.i ]		; <i32> [#uses=3]
	%tmp292.i.i = load i8* %tmp16.i.i100, align 1		; <i8> [#uses=1]
	%tmp293.not.i.i = icmp ne i8 %tmp292.i.i, 0		; <i1> [#uses=1]
	%bothcond.i.i = and i1 %tmp293.not.i.i, %myEOF.1.i.i		; <i1> [#uses=1]
	br i1 %bothcond.i.i, label %bb202.i.i, label %bb301.i.i
  Factoring out 3 common predecessors.

On the path from any blocks other than bb23.i57.i.i, the load and compare 
are dead.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50096 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-22 07:05:46 +00:00
Evan Cheng
c290a5b9ff Added command line option -no-lazy to disable JIT lazy compilation.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50095 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-22 06:51:41 +00:00
Chris Lattner
d38c14e6ed refactor some code, no functionality change.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50094 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-22 06:36:15 +00:00
Chris Lattner
f29984f695 add a basic testcase.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50093 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-22 06:35:14 +00:00
Nick Lewycky
75712b8b3f Whoops! Undo r50087, unbreak the build.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50088 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-22 05:20:06 +00:00
Nick Lewycky
59091c4630 Reverse r47989. Part of removing 'unwinds to' support.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50087 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-22 05:16:51 +00:00
Nick Lewycky
dc2421f786 Start removing 'unwinds to' support from mainline in preparation for 2.3.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50086 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-22 05:16:02 +00:00
Owen Anderson
ecbc963eda Add MemCpyOpt to LTO.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50081 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-22 03:57:36 +00:00
Chris Lattner
9d4b94d3e1 remove dead code.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50080 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-22 03:21:48 +00:00
Chris Lattner
10c0d9179e optimize "p != gep p, ..." better. This allows us to compile
getelementptr-seteq.ll into:

define i1 @test(i64 %X, %S* %P) {
	%C = icmp eq i64 %X, -1		; <i1> [#uses=1]
	ret i1 %C
}

instead of:

define i1 @test(i64 %X, %S* %P) {
	%A.idx.mask = and i64 %X, 4611686018427387903		; <i64> [#uses=1]
	%C = icmp eq i64 %A.idx.mask, 4611686018427387903		; <i1> [#uses=1]
	ret i1 %C
}

And fixes the second half of PR2235.  This speeds up the insertion sort
case by 45%, from 1.12s to 0.77s.  In practice, this will significantly
speed up for loops structured like:

for (double *P = Base + N; P != Base; --P)
  ...

Which happens frequently for C++ iterators.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50079 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-22 02:53:33 +00:00
Chris Lattner
eb411291fa more fallout from Nicholas' asmprinter patch.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50078 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-22 02:45:44 +00:00
Dan Gohman
61a9213440 Implement an x86-64 ABI detail of passing structs by hidden first
argument. The x86-64 ABI requires the incoming value of %rdi to
be copied to %rax on exit from a function that is returning a
large C struct.

Also, add a README-X86-64 entry detailing the missed optimization
opportunity and proposing an alternative approach.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50075 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-21 23:59:07 +00:00