Commit Graph

57358 Commits

Author SHA1 Message Date
Dan Gohman
6020d85c41 Fix a comment.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96178 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-14 18:51:20 +00:00
Dan Gohman
ce947366ec When complicated expressions are broken down into subexpressions
with multiplication by constants distributed through, occasionally
those subexpressions can include both x and -x. For now, if this
condition is discovered within LSR, just prune such cases away,
as they won't be profitable. This fixes a "zero allocated in a
base register" assertion failure.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96177 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-14 18:50:49 +00:00
Sanjiv Gupta
44de6823fc fixes to pagesel/banksel inserter.
1. restore these across direct/indirect calls.
2. restore pagesel for any macros with gotos.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96175 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-14 18:27:42 +00:00
Anton Korobeynikov
77709ac7db Forgot to commit the header
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96174 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-14 18:25:41 +00:00
Chris Lattner
782ab584e5 follow-on to PR6280
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96172 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-14 18:20:09 +00:00
Anton Korobeynikov
245ba9698e Drop winmcasminfo and use normal AT&T COFF for all windows targets.
Otherwise AT&T asm printer is used with non-compatible MCAsmInfo and
there is no way to override this behaviour.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96165 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-14 15:19:54 +00:00
Johnny Chen
2faf3919d4 Try to factorize the specification of saturating add/subtract operations a bit,
as suggested by Bob Wilson.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96153 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-14 06:32:20 +00:00
Dan Gohman
590bfe8641 Actually, this code doesn't have to be quite so conservative in
the no-TLI case. But it should still default to declining the
transformation.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96152 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-14 03:21:49 +00:00
Dan Gohman
1d826a76f5 In rememberInstruction, if the value being remembered is the
current insertion point, advance the current insertion point.
This avoids a use-before-def situation in a testcase extracted
from clang which is difficult to reduce to a reasonable-sized
regression test.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96151 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-14 03:12:47 +00:00
Dan Gohman
0402577f38 Simplify this code; no need for a custom subclass if it doesn't need
to override anything from the parent class.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96150 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-14 02:48:58 +00:00
Dan Gohman
9269926bfb Remove a 'protected' keyword, now that SCEVExpander is no longer
intended to be subclassed.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96149 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-14 02:47:26 +00:00
Dan Gohman
2763dfdc70 Don't attempt aggressive post-inc uses if TargetLowering is not available,
because profitability can't be sufficiently approximated.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96148 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-14 02:45:21 +00:00
Daniel Dunbar
ca1c816f25 2.7: Note that DataTypes.h moved.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96143 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-14 01:47:19 +00:00
John McCall
6d09ab8ba2 Make LSR not crash if invoked without target lowering info, e.g. if invoked
from opt.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96135 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-13 23:40:16 +00:00
Eric Christopher
77ffe3b31e Fix a problem where we had bitcasted operands that gave us
odd offsets since the bitcasted pointer size and the offset pointer
size are going to be different types for the GEP vs base object.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96134 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-13 23:38:01 +00:00
Mikhail Glushenkov
18c67b994c Forward parameter options as '-option=parameter'.
Some tools do not like the '-option parameter' form. Should this be
configurable?

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96130 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-13 22:37:28 +00:00
Mikhail Glushenkov
49e496b1cc Support some more Darwin-only options.
We really need a conditional compilation mechanism...

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96129 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-13 22:37:13 +00:00
Mikhail Glushenkov
b88c6f6cdc Support -mfix-and-continue properly.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96128 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-13 22:37:00 +00:00
Mikhail Glushenkov
2ac7ab129d Revert r94752, turns out we don't need to touch these options.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96127 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-13 22:36:43 +00:00
Daniel Dunbar
2df89fa407 Trim trailing spaces (aka, trigger rebuild).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96126 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-13 22:23:47 +00:00
Chris Lattner
a0cdf17f37 pull a bunch of huge inline methods in the PatternCodeEmitter
class out of line.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96113 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-13 20:06:50 +00:00
Chris Lattner
1cea10a663 teach the encoder to handle pseudo instructions like FP_REG_KILL,
encoding them into nothing.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96110 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-13 19:16:53 +00:00
Chris Lattner
c09687bb7e remove dead code.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96109 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-13 19:07:06 +00:00
Daniel Dunbar
591047f714 MCAssembler: Fix pcrel relocations. Oh and,
--
ddunbar@ozzy:tmp$ clang -m32 -integrated-as hello.c && ./a.out
hello world!
--

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96096 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-13 09:45:59 +00:00
Daniel Dunbar
f6346769b3 MC/Mach-O: Start emitting fixups/relocations for instructions.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96095 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-13 09:29:02 +00:00
Daniel Dunbar
2be2fd0730 MCAssembler: Switch MCAsmFixup to storing MCFixupKind instead of just a size.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96094 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-13 09:28:54 +00:00
Daniel Dunbar
0bcf074867 MCAssembler: Sink fixup list into MCDataFragment.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96093 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-13 09:28:43 +00:00
Daniel Dunbar
a4766d7af9 MCAssembler: Switch MCFillFragment to only taking constant values. Symbolic expressions can always be emitted as data + fixups.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96092 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-13 09:28:32 +00:00
Daniel Dunbar
45f4874611 MC/Mach-O: Implement EmitValue using data fragments + fixups instead of fill fragment.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96091 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-13 09:28:22 +00:00
Daniel Dunbar
3a30b827a5 MCAssembler: Start applying fixups in the data section.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96090 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-13 09:28:15 +00:00
Daniel Dunbar
b7c3a4b195 MCAssembler: Add assorted dump() methods.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96089 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-13 09:28:03 +00:00
Daniel Dunbar
a8dfb79e6f X86: Move extended MCFixupKinds into X86FixupKinds.h
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96088 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-13 09:27:52 +00:00
Chris Lattner
98d5c3141e Split some code out to a helper function (FindReusablePredBB)
and add a doxygen comment.

Cache the phi entry to avoid doing tons of 
PHINode::getBasicBlockIndex calls in the common case.

On my insane testcase from re2c, this speeds up CGP from
617.4s to 7.9s (78x).



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96083 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-13 05:35:08 +00:00
Chris Lattner
6686c6bf68 Speed up codegen prepare from 3.58s to 0.488s.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96081 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-13 05:01:14 +00:00
Chris Lattner
277cccc58f PHINode::getBasicBlockIndex is O(n) in the number of inputs
to a PHI, avoid it in the common case where the BB occurs
in the same index for multiple phis.  This speeds up CGP on
an insane testcase from 8.35 to 3.58s.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96080 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-13 04:24:19 +00:00
Chris Lattner
2f36ea8b74 iterate over preds using PHI information when available instead of
using pred_begin/end.  It is much faster.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96079 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-13 04:15:26 +00:00
Chris Lattner
3f65b5e733 speed up CGP a bit by scanning predecessors through phi operands
instead of with pred_begin/end.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96078 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-13 04:04:42 +00:00
Chris Lattner
b779033a23 add encoder support and tests for rdtscp
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96076 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-13 03:42:24 +00:00
Johnny Chen
a1e7621510 Add SETEND and BXJ instructions for disassembly only.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96075 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-13 02:51:09 +00:00
Sean Callanan
cebe955779 Added the rdtscp instruction to the x86 instruction
tables.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96073 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-13 02:06:11 +00:00
Jakob Stoklund Olesen
d135f14f61 Fix PR6283.
When coalescing with a physreg, remember to add imp-def and imp-kill when
dealing with sub-registers.

Also fix a related bug in VirtRegRewriter where substitutePhysReg may
reallocate the operand list on an instruction and invalidate the reg_iterator.
This can happen when a register is mentioned twice on the same instruction.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96072 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-13 02:06:10 +00:00
Dan Gohman
9214b82c54 Fix a pruning heuristic which implicitly assumed that SmallPtrSet is
deterministically sorted.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96071 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-13 02:06:02 +00:00
Evan Cheng
7545f49a5e Teach MachineFrameInfo to track maximum alignment while stack objects are being
created. This ensures it's updated at all time. It means targets which perform
dynamic stack alignment would know whether it is required and whether frame
pointer register cannot be made available register allocation.
This is a fix for rdar://7625239. Sorry, I can't create a reasonably sized test
case.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96069 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-13 01:56:41 +00:00
Jakob Stoklund Olesen
f0907fe590 Enable the inlinehint attribute in the Inliner.
Functions explicitly marked inline will get an inlining threshold slightly
more aggressive than the default for -O3. This means than -O3 builds are
mostly unaffected while -Os builds will be a bit bigger and faster.

The difference depends entirely on how many 'inline's are sprinkled on the
source.

In the CINT2006 suite, only these tests are significantly affected under -Os:

               Size   Time
471.omnetpp   +1.63% -1.85%
473.astar     +4.01% -6.02%
483.xalancbmk +4.60%  0.00%

Note that 483.xalancbmk runs too quickly to give useful timing results.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96066 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-13 01:51:53 +00:00
Sean Callanan
95a5a7d570 Fixed encodings for invlpg, invept, and invvpid.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96065 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-13 01:48:34 +00:00
Daniel Dunbar
e9a60eb498 MC/AsmParser: Attempt to constant fold expressions up-front. This ensures we avoid fixups for obvious cases like '-(16)'.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96064 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-13 01:28:07 +00:00
Johnny Chen
08b85f371e Added a bunch of saturating add/subtract instructions for disassembly only.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96063 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-13 01:21:01 +00:00
Chris Lattner
f068304b1f rip out the 'heinous' x86 MCCodeEmitter implementation.
We still have the templated X86 JIT emitter, *and* the
almost-copy in X86InstrInfo for getting instruction sizes.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96059 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-13 00:49:29 +00:00
Chris Lattner
a599de2410 remove special cases for vmlaunch, vmresume, vmxoff, and swapgs
fix swapgs to be spelled right.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96058 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-13 00:41:14 +00:00
Bob Wilson
bf9b221c00 Besides removing phi cycles that reduce to a single value, also remove dead
phi cycles.  Adjust a few tests to keep dead instructions from being optimized
away.  This (together with my previous change for phi cycles) fixes Apple
radar 7627077.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96057 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-13 00:31:44 +00:00