oota-llvm.git
18 years agoNew testcase, check that dead code doesn't pessimize instcombine
Chris Lattner [Wed, 10 May 2006 18:56:04 +0000 (18:56 +0000)]
New testcase, check that dead code doesn't pessimize instcombine

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28214 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoFix the PowerPC JIT-only failure on UnitTests/Vector/sumarray-dbl, which is
Chris Lattner [Wed, 10 May 2006 06:38:32 +0000 (06:38 +0000)]
Fix the PowerPC JIT-only failure on UnitTests/Vector/sumarray-dbl, which is
really a bad codegen bug that LLC happens to get lucky with. I must chat with
Nate for the proper fix.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28213 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoTemplatify RegReductionPriorityQueue
Evan Cheng [Wed, 10 May 2006 06:16:44 +0000 (06:16 +0000)]
Templatify RegReductionPriorityQueue

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28212 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoAdd alloca/malloc ctors that don't take array sizes.
Chris Lattner [Wed, 10 May 2006 04:38:35 +0000 (04:38 +0000)]
Add alloca/malloc ctors that don't take array sizes.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28211 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoAdd an assertion for a common error
Chris Lattner [Wed, 10 May 2006 04:32:43 +0000 (04:32 +0000)]
Add an assertion for a common error

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28210 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoWatch out for the following case:
Evan Cheng [Wed, 10 May 2006 02:47:57 +0000 (02:47 +0000)]
Watch out for the following case:
1. Use expects a chain output.
2. Node is expanded into multiple target ops.
3. One of the inner node produces a chain, the outer most one doesn't.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28209 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoFix a load folding bug. It is exposed by a multi- resulting instructions
Evan Cheng [Wed, 10 May 2006 00:05:46 +0000 (00:05 +0000)]
Fix a load folding bug. It is exposed by a multi- resulting instructions
def : Pat<> pattern.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28208 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoFix PR773
Nate Begeman [Tue, 9 May 2006 18:20:51 +0000 (18:20 +0000)]
Fix PR773

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28207 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoThis method doesn't need to be virtual, thanks to Reid for pointing this out.
Chris Lattner [Tue, 9 May 2006 17:29:17 +0000 (17:29 +0000)]
This method doesn't need to be virtual, thanks to Reid for pointing this out.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28206 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoFix a regression in my patch from last night that broke the llvmgcc4 build on
Chris Lattner [Tue, 9 May 2006 16:41:59 +0000 (16:41 +0000)]
Fix a regression in my patch from last night that broke the llvmgcc4 build on
ppc

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28205 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoIndent .data/.text in the .s file
Chris Lattner [Tue, 9 May 2006 16:15:00 +0000 (16:15 +0000)]
Indent .data/.text in the .s file

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28204 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoAnother instruction scheduling test case
Evan Cheng [Tue, 9 May 2006 07:20:24 +0000 (07:20 +0000)]
Another instruction scheduling test case

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28203 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoAdd pseudo dependency to force a def&use operand to be scheduled last (unless
Evan Cheng [Tue, 9 May 2006 07:13:34 +0000 (07:13 +0000)]
Add pseudo dependency to force a def&use operand to be scheduled last (unless
the distance between the def and another use is much longer). This is under
option control for now "-sched-lower-defnuse".

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28201 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoDebugging info
Evan Cheng [Tue, 9 May 2006 06:55:15 +0000 (06:55 +0000)]
Debugging info

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28200 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoRemove a completed entry.
Evan Cheng [Tue, 9 May 2006 06:54:05 +0000 (06:54 +0000)]
Remove a completed entry.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28199 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoTest case for PR770
Evan Cheng [Tue, 9 May 2006 06:48:12 +0000 (06:48 +0000)]
Test case for PR770

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28198 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoPR 770 - permit coallescing of registers in subset register classes.
Evan Cheng [Tue, 9 May 2006 06:37:48 +0000 (06:37 +0000)]
PR 770 - permit coallescing of registers in subset register classes.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28197 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoAdded sub- register classes information.
Evan Cheng [Tue, 9 May 2006 06:35:30 +0000 (06:35 +0000)]
Added sub- register classes information.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28196 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoAdd sub-register class information.
Evan Cheng [Tue, 9 May 2006 06:34:26 +0000 (06:34 +0000)]
Add sub-register class information.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28195 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoImplement MASM sections correctly, without a "has masm sections flag" and a bunch...
Chris Lattner [Tue, 9 May 2006 05:33:48 +0000 (05:33 +0000)]
Implement MASM sections correctly, without a "has masm sections flag" and a bunch of special case code.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28194 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoImplement MASM sections correctly, without a "has masm sections flag" and a
Chris Lattner [Tue, 9 May 2006 05:33:28 +0000 (05:33 +0000)]
Implement MASM sections correctly, without a "has masm sections flag" and a
bunch of special case code.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28193 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoOh yeah, there are two of these now, unify both.
Chris Lattner [Tue, 9 May 2006 05:24:50 +0000 (05:24 +0000)]
Oh yeah, there are two of these now, unify both.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28192 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoSetting SwitchToSectionDirective properly in the MASM backend permits a bunch
Chris Lattner [Tue, 9 May 2006 05:23:12 +0000 (05:23 +0000)]
Setting SwitchToSectionDirective properly in the MASM backend permits a bunch
of code to be unified.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28191 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoMASM doesn't have one of these.
Chris Lattner [Tue, 9 May 2006 05:21:47 +0000 (05:21 +0000)]
MASM doesn't have one of these.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28190 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoDon't prefix section directives with a tab. Doing so causes blank lines to
Chris Lattner [Tue, 9 May 2006 05:19:59 +0000 (05:19 +0000)]
Don't prefix section directives with a tab.  Doing so causes blank lines to
be emitted to the .s file.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28189 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoMake the masm codepath work like the normal code path.
Chris Lattner [Tue, 9 May 2006 05:15:58 +0000 (05:15 +0000)]
Make the masm codepath work like the normal code path.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28188 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoPreserve prior behavior
Chris Lattner [Tue, 9 May 2006 05:15:24 +0000 (05:15 +0000)]
Preserve prior behavior

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28187 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoThe MASM asmprinter has been fixed, these hacks are no longer needed.
Chris Lattner [Tue, 9 May 2006 05:13:34 +0000 (05:13 +0000)]
The MASM asmprinter has been fixed, these hacks are no longer needed.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28186 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoFix the MASM asmprinter's lies. It does not want to emit code to .text/.data
Chris Lattner [Tue, 9 May 2006 05:12:53 +0000 (05:12 +0000)]
Fix the MASM asmprinter's lies.  It does not want to emit code to .text/.data
it wants it emitted to _text/_data.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28185 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoSplit SwitchSection into SwitchTo{Text|Data}Section methods.
Chris Lattner [Tue, 9 May 2006 04:59:56 +0000 (04:59 +0000)]
Split SwitchSection into SwitchTo{Text|Data}Section methods.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28184 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoSplit SwitchSection into the SwitchTo{Text|Data}Section functions, to better
Chris Lattner [Tue, 9 May 2006 04:59:30 +0000 (04:59 +0000)]
Split SwitchSection into the SwitchTo{Text|Data}Section functions, to better
support assemblers that distinguish the two.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28183 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoSome notes and thoughts to myself
Chris Lattner [Tue, 9 May 2006 04:58:46 +0000 (04:58 +0000)]
Some notes and thoughts to myself

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28182 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoPatch to make some xforms preserve each other. Patch contributed by
Chris Lattner [Tue, 9 May 2006 04:13:41 +0000 (04:13 +0000)]
Patch to make some xforms preserve each other.  Patch contributed by
Domagoj Babic!

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28181 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoUse the -dumpversion option to llvm-gcc which gives us just the version #
Reid Spencer [Tue, 9 May 2006 00:31:01 +0000 (00:31 +0000)]
Use the -dumpversion option to llvm-gcc which gives us just the version #
and is unlikely to change in future releases. This also simplifies the
parsing of the full and major llvm-gcc version numbers in the script.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28180 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoMove some methods out of line so that MutexGuard.h isn't needed in a public header.
Chris Lattner [Mon, 8 May 2006 22:00:52 +0000 (22:00 +0000)]
Move some methods out of line so that MutexGuard.h isn't needed in a public header.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28179 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoMove methods out of line so that MutexGuard.h isn't required in the header.
Chris Lattner [Mon, 8 May 2006 22:00:26 +0000 (22:00 +0000)]
Move methods out of line so that MutexGuard.h isn't required in the header.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28178 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoAnother bad case I noticed
Chris Lattner [Mon, 8 May 2006 21:39:45 +0000 (21:39 +0000)]
Another bad case I noticed

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28177 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoadd a note
Chris Lattner [Mon, 8 May 2006 21:24:21 +0000 (21:24 +0000)]
add a note

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28176 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoMake the case I just checked in stronger. Now we compile this:
Chris Lattner [Mon, 8 May 2006 21:18:59 +0000 (21:18 +0000)]
Make the case I just checked in stronger.  Now we compile this:

short test2(short X, short x) {
  int Y = (short)(X+x);
  return Y >> 1;
}

to:

_test2:
        add r2, r3, r4
        extsh r2, r2
        srawi r3, r2, 1
        blr

instead of:

_test2:
        add r2, r3, r4
        extsh r2, r2
        srwi r2, r2, 1
        extsh r3, r2
        blr

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28175 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoImplement and_sext.ll:test3, generating:
Chris Lattner [Mon, 8 May 2006 20:59:41 +0000 (20:59 +0000)]
Implement and_sext.ll:test3, generating:

_test4:
        srawi r3, r3, 16
        blr

instead of:

_test4:
        srwi r2, r3, 16
        extsh r3, r2
        blr

for:

short test4(unsigned X) {
  return (X >> 16);
}

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28174 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agonew testcase
Chris Lattner [Mon, 8 May 2006 20:58:58 +0000 (20:58 +0000)]
new testcase

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28173 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoYet more readme updating
Nate Begeman [Mon, 8 May 2006 20:54:02 +0000 (20:54 +0000)]
Yet more readme updating

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28172 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoCompile this:
Chris Lattner [Mon, 8 May 2006 20:51:54 +0000 (20:51 +0000)]
Compile this:

short test4(unsigned X) {
  return (X >> 16);
}

to:

_test4:
        movl 4(%esp), %eax
        sarl $16, %eax
        ret

instead of:

_test4:
        movl $-65536, %eax
        andl 4(%esp), %eax
        sarl $16, %eax
        ret

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28171 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoNew note about something bad happening in target independent optimizers
Nate Begeman [Mon, 8 May 2006 20:08:28 +0000 (20:08 +0000)]
New note about something bad happening in target independent optimizers

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28170 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoProving once again that I am not as smart as the compiler
Nate Begeman [Mon, 8 May 2006 19:09:24 +0000 (19:09 +0000)]
Proving once again that I am not as smart as the compiler

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28169 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoFold more shifts into inserts, and update the README
Nate Begeman [Mon, 8 May 2006 17:38:32 +0000 (17:38 +0000)]
Fold more shifts into inserts, and update the README

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28168 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoFold shifts with undef operands.
Chris Lattner [Mon, 8 May 2006 17:29:49 +0000 (17:29 +0000)]
Fold shifts with undef operands.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28167 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoWhen tracking demanded bits, if any bits from the sext of an SRA are demanded,
Chris Lattner [Mon, 8 May 2006 17:22:53 +0000 (17:22 +0000)]
When tracking demanded bits, if any bits from the sext of an SRA are demanded,
then so is the input sign bit.  This fixes mediabench/g721 on X86.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28166 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoMake emission of jump tables a bit less conservative; they are now required
Nate Begeman [Mon, 8 May 2006 16:51:36 +0000 (16:51 +0000)]
Make emission of jump tables a bit less conservative; they are now required
to be only 31.25% dense, rather than 75% dense.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28165 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoFixing truncate. Previously we were emitting truncate from r16 to r8 as
Evan Cheng [Mon, 8 May 2006 08:01:26 +0000 (08:01 +0000)]
Fixing truncate. Previously we were emitting truncate from r16 to r8 as
movw. That is we promote the destination operand to r16. So
        %CH = TRUNC_R16_R8 %BP
is emitted as
        movw %bp, %cx.

This is incorrect. If %cl is live, it would be clobbered.
Ideally we want to do the opposite, that is emitted it as
        movb ??, %ch
But this is not possible since %bp does not have a r8 sub-register.

We are now defining a new register class R16_ which is a subclass of R16
containing only those 16-bit registers that have r8 sub-registers (i.e.
AX - DX). We isel the truncate to two instructions, a MOV16to16_ to copy the
value to the R16_ class, followed by a TRUNC_R16_R8.

Due to bug 770, the register colaescer is not going to coalesce between R16 and
R16_. That will be fixed later so we can eliminate the MOV16to16_. Right now, it
can only be eliminated if we are lucky that source and destination registers are
the same.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28164 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoMove the definition of value_use_iterator::getOperandNo to User.h where the
Chris Lattner [Mon, 8 May 2006 05:59:36 +0000 (05:59 +0000)]
Move the definition of value_use_iterator::getOperandNo to User.h where the
definition of the User class is available, this fixes the  build with some
compiler versions.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28163 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoUpdate some stuff now that the new rlwimi code has gone in
Nate Begeman [Mon, 8 May 2006 02:52:38 +0000 (02:52 +0000)]
Update some stuff now that the new rlwimi code has gone in

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28162 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoFix PR772
Nate Begeman [Mon, 8 May 2006 01:35:01 +0000 (01:35 +0000)]
Fix PR772

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28161 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoRemove unncessary include
Nate Begeman [Mon, 8 May 2006 01:33:11 +0000 (01:33 +0000)]
Remove unncessary include

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28160 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoThis test passes now, remove xfail marker
Chris Lattner [Sun, 7 May 2006 18:16:31 +0000 (18:16 +0000)]
This test passes now, remove xfail marker

Change test to be a positive test instead of a negative test

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28159 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoTypo's
Evan Cheng [Sun, 7 May 2006 10:10:20 +0000 (10:10 +0000)]
Typo's

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28158 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoUnlike Unix, Windows won't let a file be implicitly replaced via renaming without...
Jeff Cohen [Sun, 7 May 2006 02:51:51 +0000 (02:51 +0000)]
Unlike Unix, Windows won't let a file be implicitly replaced via renaming without explicit permission.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28157 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoNew rlwimi implementation, which is superior to the old one. There are
Nate Begeman [Sun, 7 May 2006 00:23:38 +0000 (00:23 +0000)]
New rlwimi implementation, which is superior to the old one. There are
still a couple missed optimizations, but we now generate all the possible
rlwimis for multiple inserts into the same bitfield.  More regression tests
to come.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28156 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoUse ComputeMaskedBits to determine # sign bits as a fallback. This allows us
Chris Lattner [Sat, 6 May 2006 23:48:13 +0000 (23:48 +0000)]
Use ComputeMaskedBits to determine # sign bits as a fallback.  This allows us
to handle all kinds of stuff, including silly things like:
sextinreg(setcc,i16) -> setcc.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28155 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoAdd some more sign propagation cases
Chris Lattner [Sat, 6 May 2006 23:40:29 +0000 (23:40 +0000)]
Add some more sign propagation cases

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28154 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoApply bug fix supplied by Greg Pettyjohn for a bug he found: '<invalid>' is not...
Jeff Cohen [Sat, 6 May 2006 23:25:53 +0000 (23:25 +0000)]
Apply bug fix supplied by Greg Pettyjohn for a bug he found:  '<invalid>' is not a legal path on Windows.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28153 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoSimplify some code, add a couple minor missed folds
Chris Lattner [Sat, 6 May 2006 23:06:26 +0000 (23:06 +0000)]
Simplify some code, add a couple minor missed folds

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28152 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoconstant fold sign_extend_inreg
Chris Lattner [Sat, 6 May 2006 23:05:41 +0000 (23:05 +0000)]
constant fold sign_extend_inreg

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28151 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoremove cases handled elsewhere
Chris Lattner [Sat, 6 May 2006 22:43:44 +0000 (22:43 +0000)]
remove cases handled elsewhere

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28150 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoAdd some more simple sign bit propagation cases.
Chris Lattner [Sat, 6 May 2006 22:39:59 +0000 (22:39 +0000)]
Add some more simple sign bit propagation cases.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28149 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoFix some loose ends in MASM support.
Jeff Cohen [Sat, 6 May 2006 21:27:14 +0000 (21:27 +0000)]
Fix some loose ends in MASM support.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28148 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agonew testcase we handle right now.
Chris Lattner [Sat, 6 May 2006 18:15:50 +0000 (18:15 +0000)]
new testcase we handle right now.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28147 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoUse the new TargetLowering::ComputeNumSignBits method to eliminate
Chris Lattner [Sat, 6 May 2006 09:30:03 +0000 (09:30 +0000)]
Use the new TargetLowering::ComputeNumSignBits method to eliminate
sign_extend_inreg operations.  Though ComputeNumSignBits is still rudimentary,
this is enough to compile this:

short test(short X, short x) {
  int Y = X+x;
  return (Y >> 1);
}
short test2(short X, short x) {
  int Y = (short)(X+x);
  return Y >> 1;
}

into:

_test:
        add r2, r3, r4
        srawi r3, r2, 1
        blr
_test2:
        add r2, r3, r4
        extsh r2, r2
        srawi r3, r2, 1
        blr

instead of:

_test:
        add r2, r3, r4
        srawi r2, r2, 1
        extsh r3, r2
        blr
_test2:
        add r2, r3, r4
        extsh r2, r2
        srawi r2, r2, 1
        extsh r3, r2
        blr

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28146 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoAdd some really really simple code for computing sign-bit propagation.
Chris Lattner [Sat, 6 May 2006 09:27:13 +0000 (09:27 +0000)]
Add some really really simple code for computing sign-bit propagation.

This will certainly be enhanced in the future.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28145 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoAdd some new methods for computing sign bit information.
Chris Lattner [Sat, 6 May 2006 09:26:22 +0000 (09:26 +0000)]
Add some new methods for computing sign bit information.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28144 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoWhen inserting casts, be careful of where we put them. We cannot insert
Chris Lattner [Sat, 6 May 2006 09:10:37 +0000 (09:10 +0000)]
When inserting casts, be careful of where we put them.  We cannot insert
a cast immediately before a PHI node.

This fixes Regression/CodeGen/Generic/2006-05-06-GEP-Cast-Sink-Crash.ll

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28143 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agonew testcase
Chris Lattner [Sat, 6 May 2006 09:09:47 +0000 (09:09 +0000)]
new testcase

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28142 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoMove some code around.
Chris Lattner [Sat, 6 May 2006 09:00:16 +0000 (09:00 +0000)]
Move some code around.

Make the "fold (and (cast A), (cast B)) -> (cast (and A, B))" transformation
only apply when both casts really will cause code to be generated.  If one or
both doesn't, then this xform doesn't remove a cast.

This fixes Transforms/InstCombine/2006-05-06-Infloop.ll

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28141 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agonew testcase from ghostscript that inf looped instcombine
Chris Lattner [Sat, 6 May 2006 08:58:06 +0000 (08:58 +0000)]
new testcase from ghostscript that inf looped instcombine

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28140 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoTeach the X86 backend about non-i32 inline asm register classes.
Chris Lattner [Sat, 6 May 2006 00:29:37 +0000 (00:29 +0000)]
Teach the X86 backend about non-i32 inline asm register classes.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28139 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoFold (trunc (srl x, c)) -> (srl (trunc x), c)
Chris Lattner [Sat, 6 May 2006 00:11:52 +0000 (00:11 +0000)]
Fold (trunc (srl x, c)) -> (srl (trunc x), c)

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28138 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoFold trunc(any_ext). This gives stuff like:
Chris Lattner [Fri, 5 May 2006 22:56:26 +0000 (22:56 +0000)]
Fold trunc(any_ext).  This gives stuff like:

27,28c27
<       movzwl %di, %edi
<       movl %edi, %ebx
---
>       movw %di, %bx

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28137 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoShrink shifts when possible.
Chris Lattner [Fri, 5 May 2006 22:53:17 +0000 (22:53 +0000)]
Shrink shifts when possible.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28136 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoImplement ComputeMaskedBits/SimplifyDemandedBits for ISD::TRUNCATE
Chris Lattner [Fri, 5 May 2006 22:32:12 +0000 (22:32 +0000)]
Implement ComputeMaskedBits/SimplifyDemandedBits for ISD::TRUNCATE

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28135 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoPrint a grouping around inline asm blocks so that we can tell when we are
Chris Lattner [Fri, 5 May 2006 21:50:04 +0000 (21:50 +0000)]
Print a grouping around inline asm blocks so that we can tell when we are
using them.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28134 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoPrint *some* grouping around inline asm blocks so we know where they are.
Chris Lattner [Fri, 5 May 2006 21:48:50 +0000 (21:48 +0000)]
Print *some* grouping around inline asm blocks so we know where they are.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28133 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoIndent multiline asm strings more nicely
Chris Lattner [Fri, 5 May 2006 21:47:05 +0000 (21:47 +0000)]
Indent multiline asm strings more nicely

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28132 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoTeach the code generator to use cvtss2sd as extload f32 -> f64
Chris Lattner [Fri, 5 May 2006 21:35:18 +0000 (21:35 +0000)]
Teach the code generator to use cvtss2sd as extload f32 -> f64

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28131 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoFold (fpext (load x)) -> (extload x)
Chris Lattner [Fri, 5 May 2006 21:34:35 +0000 (21:34 +0000)]
Fold (fpext (load x)) -> (extload x)

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28130 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoMore aggressively sink GEP offsets into loops. For example, before we
Chris Lattner [Fri, 5 May 2006 21:17:49 +0000 (21:17 +0000)]
More aggressively sink GEP offsets into loops.  For example, before we
generated:

        movl 8(%esp), %eax
        movl %eax, %edx
        addl $4316, %edx
        cmpb $1, %cl
        ja LBB1_2       #cond_false
LBB1_1: #cond_true
        movl L_QuantizationTables720$non_lazy_ptr, %ecx
        movl %ecx, (%edx)
        movl L_QNOtoQuantTableShift720$non_lazy_ptr, %edx
        movl %edx, 4460(%eax)
        ret
...

Now we generate:

        movl 8(%esp), %eax
        cmpb $1, %cl
        ja LBB1_2       #cond_false
LBB1_1: #cond_true
        movl L_QuantizationTables720$non_lazy_ptr, %ecx
        movl %ecx, 4316(%eax)
        movl L_QNOtoQuantTableShift720$non_lazy_ptr, %ecx
        movl %ecx, 4460(%eax)
        ret

... which uses one fewer register.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28129 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoFix an infinite loop compiling oggenc last night.
Chris Lattner [Fri, 5 May 2006 20:51:30 +0000 (20:51 +0000)]
Fix an infinite loop compiling oggenc last night.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28128 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoNeed extload patterns after Chris' DAG combiner changes
Evan Cheng [Fri, 5 May 2006 08:23:07 +0000 (08:23 +0000)]
Need extload patterns after Chris' DAG combiner changes

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28127 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoImplement InstCombine/cast.ll:test29
Chris Lattner [Fri, 5 May 2006 06:39:07 +0000 (06:39 +0000)]
Implement InstCombine/cast.ll:test29

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28126 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoNew testcase
Chris Lattner [Fri, 5 May 2006 06:38:40 +0000 (06:38 +0000)]
New testcase

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28125 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoFold some common code.
Chris Lattner [Fri, 5 May 2006 06:32:04 +0000 (06:32 +0000)]
Fold some common code.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28124 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoImplement:
Chris Lattner [Fri, 5 May 2006 06:31:05 +0000 (06:31 +0000)]
Implement:

  // fold (and (sext x), (sext y)) -> (sext (and x, y))
  // fold (or  (sext x), (sext y)) -> (sext (or  x, y))
  // fold (xor (sext x), (sext y)) -> (sext (xor x, y))
  // fold (and (aext x), (aext y)) -> (aext (and x, y))
  // fold (or  (aext x), (aext y)) -> (aext (or  x, y))
  // fold (xor (aext x), (aext y)) -> (aext (xor x, y))

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28123 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoPull and through and/or/xor. This compiles some bitfield code to:
Chris Lattner [Fri, 5 May 2006 06:10:43 +0000 (06:10 +0000)]
Pull and through and/or/xor.  This compiles some bitfield code to:

        mov EAX, DWORD PTR [ESP + 4]
        mov ECX, DWORD PTR [EAX]
        mov EDX, ECX
        add EDX, EDX
        or EDX, ECX
        and EDX, -2147483648
        and ECX, 2147483647
        or EDX, ECX
        mov DWORD PTR [EAX], EDX
        ret

instead of:

        sub ESP, 4
        mov DWORD PTR [ESP], ESI
        mov EAX, DWORD PTR [ESP + 8]
        mov ECX, DWORD PTR [EAX]
        mov EDX, ECX
        add EDX, EDX
        mov ESI, ECX
        and ESI, -2147483648
        and EDX, -2147483648
        or EDX, ESI
        and ECX, 2147483647
        or EDX, ECX
        mov DWORD PTR [EAX], EDX
        mov ESI, DWORD PTR [ESP]
        add ESP, 4
        ret

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28122 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoImplement a variety of simplifications for ANY_EXTEND.
Chris Lattner [Fri, 5 May 2006 05:58:59 +0000 (05:58 +0000)]
Implement a variety of simplifications for ANY_EXTEND.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28121 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoFactor some code, add these transformations:
Chris Lattner [Fri, 5 May 2006 05:51:50 +0000 (05:51 +0000)]
Factor some code, add these transformations:

  // fold (and (trunc x), (trunc y)) -> (trunc (and x, y))
  // fold (or  (trunc x), (trunc y)) -> (trunc (or  x, y))
  // fold (xor (trunc x), (trunc y)) -> (trunc (xor x, y))

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28120 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoBetter implementation of truncate. ISel matches it to a pseudo instruction
Evan Cheng [Fri, 5 May 2006 05:40:20 +0000 (05:40 +0000)]
Better implementation of truncate. ISel matches it to a pseudo instruction
that gets emitted as movl (for r32 to i16, i8) or a movw (for r16 to i8). And
if the destination gets allocated a subregister of the source operand, then
the instruction will not be emitted at all.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28119 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoNew note, Nate, please check to see if I'm full of it :)
Chris Lattner [Fri, 5 May 2006 05:36:15 +0000 (05:36 +0000)]
New note, Nate, please check to see if I'm full of it :)

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28118 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoFix VC++ compilation error.
Jeff Cohen [Fri, 5 May 2006 01:47:05 +0000 (01:47 +0000)]
Fix VC++ compilation error.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28117 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoSomehow, I missed this part of the checkin a couple days ago
Nate Begeman [Fri, 5 May 2006 01:13:11 +0000 (01:13 +0000)]
Somehow, I missed this part of the checkin a couple days ago

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28116 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoSink noop copies into the basic block that uses them. This reduces the number
Chris Lattner [Fri, 5 May 2006 01:04:50 +0000 (01:04 +0000)]
Sink noop copies into the basic block that uses them.  This reduces the number
of cross-block live ranges, and allows the bb-at-a-time selector to always
coallesce these away, at isel time.

This reduces the load on the coallescer and register allocator.  For example
on a codec on X86, we went from:

   1643 asm-printer           - Number of machine instrs printed
    419 liveintervals         - Number of loads/stores folded into instructions
   1144 liveintervals         - Number of identity moves eliminated after coalescing
   1022 liveintervals         - Number of interval joins performed
    282 liveintervals         - Number of intervals after coalescing
   1304 liveintervals         - Number of original intervals
     86 regalloc              - Number of times we had to backtrack
1.90232 regalloc              - Ratio of intervals processed over total intervals
     40 spiller               - Number of values reused
    182 spiller               - Number of loads added
    121 spiller               - Number of stores added
    132 spiller               - Number of register spills
      6 twoaddressinstruction - Number of instructions commuted to coalesce
    360 twoaddressinstruction - Number of two-address instructions

to:

   1636 asm-printer           - Number of machine instrs printed
    403 liveintervals         - Number of loads/stores folded into instructions
   1155 liveintervals         - Number of identity moves eliminated after coalescing
   1033 liveintervals         - Number of interval joins performed
    279 liveintervals         - Number of intervals after coalescing
   1312 liveintervals         - Number of original intervals
     76 regalloc              - Number of times we had to backtrack
1.88998 regalloc              - Ratio of intervals processed over total intervals
      1 spiller               - Number of copies elided
     41 spiller               - Number of values reused
    191 spiller               - Number of loads added
    114 spiller               - Number of stores added
    128 spiller               - Number of register spills
      4 twoaddressinstruction - Number of instructions commuted to coalesce
    356 twoaddressinstruction - Number of two-address instructions

On this testcase, this change provides a modest reduction in spill code,
regalloc iterations, and total instructions emitted.  It increases the number
of register coallesces.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28115 91177308-0d34-0410-b5e6-96231b3b80d8

18 years agoAdd a helper method.
Chris Lattner [Fri, 5 May 2006 00:51:42 +0000 (00:51 +0000)]
Add a helper method.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28114 91177308-0d34-0410-b5e6-96231b3b80d8