Commit Graph

7373 Commits

Author SHA1 Message Date
Chris Lattner
f2a97ed13d fix an oversight caught by Frits!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@122204 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-19 23:24:04 +00:00
Chris Lattner
459f4f8c25 tidy up
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@122190 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-19 20:24:28 +00:00
Chris Lattner
74b64611c4 move a transformation to a more logical place, simplifying it.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@122183 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-19 19:43:52 +00:00
Chris Lattner
e5cbdca3bd recognize an unsigned add with overflow idiom into uadd.
This resolves a README entry and technically resolves PR4916,
but we still get poor code for the testcase in that PR because
GVN isn't CSE'ing uadd with add, filed as PR8817.

Previously we got:

_test7:                                 ## @test7
	addq	%rsi, %rdi
	cmpq	%rdi, %rsi
	movl	$42, %eax
	cmovaq	%rsi, %rax
	ret

Now we get:

_test7:                                 ## @test7
	addq	%rsi, %rdi
	movl	$42, %eax
	cmovbq	%rsi, %rax
	ret



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@122182 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-19 19:37:52 +00:00
Chris Lattner
26b482d7a7 optimize uadd(x, cst) into a comparison when the normal
result is dead.  This is required for my next patch to not
regress the testsuite.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@122181 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-19 19:35:32 +00:00
Chris Lattner
0fe80bbbb6 use IC.ReplaceInstUsesWith instead of a raw RAUW so that uses of
the old thing end up on the instcombine worklist.  Not doing this
can cause an extra top-level iteration of instcombine, burning
compile time.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@122179 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-19 18:38:44 +00:00
Chris Lattner
0a62474830 generalize the sadd creation code to not require that the
sadd formed is half the size of the original type. We can
now compile this into a sadd.i8:

unsigned char X(char a, char b) {
  int res = a+b;
  if ((unsigned )(res+128) > 255U)
    abort();
  return res;
}




git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@122178 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-19 18:35:09 +00:00
Chris Lattner
dd7e837374 fix another miscompile in the llvm.sadd formation logic: it wasn't
checking to see if the high bits of the original add result were dead.
Inserting a smaller add and zexting back to that size is not good enough.

This is likely to be the fix for 8816.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@122177 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-19 18:22:06 +00:00
Chris Lattner
368397bb7d fix a bug (possibly 8816) in the sadd forming xform: it isn't
profitable (or safe) to promote code when the add-with-constant
has other uses.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@122175 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-19 17:59:02 +00:00
Chris Lattner
f0f568b49e rework the code added in r122072 to pull it out to its own
helper function, clean up comments, and reduce indentation.
No functionality change.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@122174 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-19 17:52:50 +00:00
Chris Lattner
1dec0d2704 Enhance LICM to promote alias sets whose pointers themselves are stored,
which doesn't affect the memory address being promoted.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@122172 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-19 05:57:25 +00:00
Chris Lattner
1c0af0ed25 fix PR8602, a bug in an assertion: a volatile store *of* a pointer
does not make the alias set for that pointer volatile, just stores
*to* the pointer.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@122171 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-19 05:51:54 +00:00
Chris Lattner
a0d172f7fe revert r122164, I'm going to go with a different approach.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@122168 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-19 04:23:03 +00:00
Chris Lattner
140f4a315b first step to fixing PR8642: don't fold away empty basic blocks
which have trapping constant exprs in them due to PHI nodes.
Eliminating them can cause the constant expr to be evalutated
on new paths if the input edges are critical.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@122164 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-19 03:02:34 +00:00
Chris Lattner
a37029cd24 simplify this a bit.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@122156 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-18 20:22:49 +00:00
Bill Wendling
2d0537cdd8 Whitespace fixes. No functionality change.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@122110 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-17 23:27:41 +00:00
Nate Begeman
9a3dc55202 Add vector versions of some existing scalar transforms to aid codegen in matching psign & pblend operations to the IR produced by clang/gcc for their C idioms.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@122105 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-17 23:12:19 +00:00
Owen Anderson
e63dda51c2 Reapply r121905 (automatic synthesis of @llvm.sadd.with.overflow) with a fix for a bug that manifested itself
on the DragonEgg self-host bot.  Unfortunately, the testcase is pretty messy and doesn't reduce well due to
interactions with other parts of InstCombine.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@122072 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-17 18:08:00 +00:00
Benjamin Kramer
14c0987bd9 SimplifyCFG: Ranges can be larger than 64 bits. Fixes Release-selfhost build.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@122054 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-17 10:48:14 +00:00
Chris Lattner
e27db74a60 improve switch formation to handle small range
comparisons formed by comparisons.  For example,
this:

void foo(unsigned x) {
  if (x == 0 || x == 1 || x == 3 || x == 4 || x == 6) 
    bar();
}

compiles into:

_foo:                                   ## @foo
## BB#0:                                ## %entry
	cmpl	$6, %edi
	ja	LBB0_2
## BB#1:                                ## %entry
	movl	%edi, %eax
	movl	$91, %ecx
	btq	%rax, %rcx
	jb	LBB0_3

instead of:

_foo:                                   ## @foo
## BB#0:                                ## %entry
	cmpl	$2, %edi
	jb	LBB0_4
## BB#1:                                ## %switch.early.test
	cmpl	$6, %edi
	ja	LBB0_3
## BB#2:                                ## %switch.early.test
	movl	%edi, %eax
	movl	$88, %ecx
	btq	%rax, %rcx
	jb	LBB0_4

This catches a bunch of cases in GCC, which look like this:

 %804 = load i32* @which_alternative, align 4, !tbaa !0
 %805 = icmp ult i32 %804, 2
 %806 = icmp eq i32 %804, 3
 %or.cond121 = or i1 %805, %806
 %807 = icmp eq i32 %804, 4
 %or.cond124 = or i1 %or.cond121, %807
 br i1 %or.cond124, label %.thread, label %808

turning this into a range comparison.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@122045 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-17 06:20:15 +00:00
Dan Gohman
c32046e6ea Revert r64460. strtol and friends cannot be marked readonly, even with
a null endptr argument, because they may write to errno.

This fixes a seflhost miscompile observed on Linux targets when TBAA
was enabled.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@122014 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-17 01:09:43 +00:00
Frits van Bommel
01abcf339f Fix a bug in the loop in JumpThreading::ProcessThreadableEdges() where it could falsely produce a MultipleDestSentinel value if the first predecessor ended with an 'indirectbr'. If that happened, it caused an unnecessary FindMostPopularDest() call.
This wasn't a correctness problem, but it broke the fast path for single-predecessor blocks.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@121966 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-16 12:16:00 +00:00
Duncan Sands
ebef48ea4b Speculatively revert commit 121905 since it looks like it might have broken the
dragonegg self-host buildbot.  Original commit message:

Add an InstCombine transform to recognize instances of manual overflow-safe addition
(performing the addition in a wider type and explicitly checking for overflow), and
fold them down to intrinsics.  This currently only supports signed-addition, but could
be generalized if someone works out the magic constant formulas for other operations.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@121965 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-16 09:40:54 +00:00
Dan Gohman
387f28aff4 Make memcpyopt TBAA-aware.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@121944 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-16 02:51:19 +00:00
Dan Gohman
f4177aa019 Preserve TBAA tags when doing load PRE.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@121921 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-15 23:53:55 +00:00
Owen Anderson
12984de314 Add an InstCombine transform to recognize instances of manual overflow-safe addition
(performing the addition in a wider type and explicitly checking for overflow), and
fold them down to intrinsics.  This currently only supports signed-addition, but could
be generalized if someone works out the magic constant formulas for other operations.

Fixes <rdar://problem/8558713>.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@121905 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-15 22:32:38 +00:00
Dan Gohman
5034dd318a Move Value::getUnderlyingObject to be a standalone
function so that it can live in Analysis instead of
VMCore.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@121885 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-15 20:02:24 +00:00
Duncan Sands
fea3b218d6 Move Sub simplifications and additional Add simplifications out of
instcombine and into InstructionSimplify.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@121861 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-15 14:07:39 +00:00
Frits van Bommel
26e097ca4b Teach jump threading to "look through" a select when the branch direction of a terminator depends on it.
When it sees a promising select it now tries to figure out whether the condition of the select is known in any of the predecessors and if so it maps the operands appropriately.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@121859 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-15 09:51:20 +00:00
Chris Lattner
ba3c815570 make qsort predicate more conformant by returning 0 for equal values.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@121838 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-15 04:52:41 +00:00
Owen Anderson
86e8a700f5 Fix PR8790, another instance where unreachable code can cause instruction simplification to fail,
this case involve a select that simplifies to itself.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@121817 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-15 00:55:35 +00:00
Owen Anderson
93f83deba1 Cleanup trailing whitespace.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@121816 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-15 00:52:44 +00:00
Chris Lattner
3aff13b82a - Insert new instructions before DomBlock's terminator,
which is simpler than finding a place to insert in BB.
 - Don't perform the 'if condition hoisting' xform on certain
   i1 PHIs, as it interferes with switch formation.

This re-fixes "example 7", without breaking the world hopefully.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@121764 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-14 08:46:09 +00:00
Chris Lattner
60d410d7bb fix two significant issues with FoldTwoEntryPHINode:
first, it can kick in on blocks whose conditions have been
folded to a constant, even though one of the edges will be
trivially folded.

second, it doesn't clean up the "if diamond" that it just 
eliminated away.  This is a problem because other simplifycfg
xforms kick in depending on the order of block visitation,
causing pointless work.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@121762 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-14 08:01:53 +00:00
Chris Lattner
071edc81f2 remove the instsimplify logic I added in r121754. It is apparently
breaking the selfhost builds, though I can't fathom how.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@121761 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-14 07:53:03 +00:00
Chris Lattner
44da7ca421 clean up logic, convert std::set to SmallPtrSet, handle the case
when all 2-entry phis are simplified away.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@121760 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-14 07:41:39 +00:00
Chris Lattner
e0b18e5912 tidy up a bit, move DEBUG down to when we commit to doing the transform so we
don't print it unless the xform happens.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@121758 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-14 07:23:10 +00:00
Chris Lattner
07ff3539f5 use SimplifyInstruction instead of reimplementing part of it.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@121757 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-14 07:20:29 +00:00
Chris Lattner
995ba1bd49 simplify GetIfCondition by using getSinglePredecessor.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@121756 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-14 07:15:21 +00:00
Chris Lattner
6de0a28dfb use AddPredecessorToBlock in 3 places instead of a manual loop.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@121755 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-14 07:09:42 +00:00
Chris Lattner
73c50a68a7 make FoldTwoEntryPHINode use instsimplify a bit, make
GetIfCondition faster by avoiding pred_iterator.  No
really interesting change.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@121754 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-14 07:00:00 +00:00
Chris Lattner
2112bbc42d remove the dead (and terrible) llvm::RemoveSuccessor function.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@121753 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-14 06:51:55 +00:00
Chris Lattner
302ba6fc1c improve DEBUG's a bit, switch to eraseFromParent() to simplify
code a bit, switch from constant folding to instsimplify.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@121751 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-14 06:17:25 +00:00
Chris Lattner
117f8cffc5 reapply my recent change that disables a piece of the switch formation
work, but fixes 400.perlbmk.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@121749 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-14 05:57:30 +00:00
Owen Anderson
2d9220e8f5 Fix recent buildbot breakage by pulling SimplifyCFG back to its state as of r121694, the most recent state
where I'm confident there were no crashes or miscompilations.  XFAIL the test added since then for now.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@121733 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-13 23:49:28 +00:00
Chris Lattner
f9a1b2a4cf temporarily disable part of my previous patch, which causes an iterator invalidation issue, causing a crash on some versions of perlbmk.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@121728 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-13 23:02:19 +00:00
Chris Lattner
94c58a0906 add some DEBUG's.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@121711 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-13 19:55:30 +00:00
Benjamin Kramer
cf8b3257c0 Fix sort predicate. qsort(3)'s predicate semantics differ from std::sort's. Fixes PR 8780.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@121705 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-13 18:20:38 +00:00
Chris Lattner
a9f6bbea62 reinstate my patch: the miscompile was caused by an inverted branch in the
'and' case.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@121695 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-13 08:12:19 +00:00
Chris Lattner
92407e5895 Completely disable the optimization I added in r121680 until
I can track down a miscompile.  This should bring the buildbots
back to life


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@121693 91177308-0d34-0410-b5e6-96231b3b80d8
2010-12-13 07:41:29 +00:00