Chris Lattner
bfff18a869
Fix two bugs: one where a condition was mistakenly swapped, and another
...
where we folded (X & 254) -> X < 1 instead of X < 2. These problems were
latent problems exposed by the latest patch.
llvm-svn: 16528
2004-09-27 19:29:18 +00:00
Misha Brukman
d409d38151
SparcV8 int regs are not only 32-bits in width, but they are 32-bit aligned!
...
llvm-svn: 16526
2004-09-27 18:22:18 +00:00
Chris Lattner
d7dc1ecd42
The system ranlib on darwin occasionally adds two extra newlines to the
...
end of files, breaking the CFE build. As a gross hack around this,
ignore any trailing garbage on bytecode files. Thanks to Brian for digging
in and identifying the problem.
llvm-svn: 16525
2004-09-27 16:59:06 +00:00
Chris Lattner
1023b8726e
Fold: (setcc (shr X, ShAmt), CI), where 'cc' is eq or ne. This xform
...
triggers often, for example:
6x in povray, 1x in gzip, 279x in gcc, 1x in crafty, 8x in eon, 11x in perlbmk,
362x in gap, 4x in vortex, 14 in m88ksim, 211x in 126.gcc, 1x in compress,
11x in ijpeg, and 4x in 147.vortex.
llvm-svn: 16521
2004-09-27 16:18:50 +00:00
Nate Begeman
8656a156cf
Correct some BuildMI arguments for the upcoming simple scheduler
...
llvm-svn: 16519
2004-09-27 05:08:17 +00:00
Misha Brukman
a68d76ebfa
Fix the copy-pasto that Brian noticed: V8 int regs are 32-bits wide, not 64.
...
llvm-svn: 16518
2004-09-26 21:07:43 +00:00
Reid Spencer
f71143c3e5
Updated to reflect changes in the interface of TimeValue::now().
...
llvm-svn: 16515
2004-09-25 08:32:37 +00:00
Reid Spencer
1cc1994edf
Added stub implementations of TimeValue concept for remaining platforms.
...
llvm-svn: 16513
2004-09-25 05:03:54 +00:00
Reid Spencer
8b2f9a298d
Wrap to 80 cols.
...
llvm-svn: 16512
2004-09-25 05:03:22 +00:00
Reid Spencer
ca141a5b28
Initial implementation of the TimeValue abstraction.
...
llvm-svn: 16511
2004-09-24 23:25:19 +00:00
Chris Lattner
7e794273f5
Implement shift-and combinations, implementing InstCombine/and.ll:test19-21
...
These combinations trigger 4 times in povray, 7x in gcc, 4x in gap, and 2x in bzip2.
llvm-svn: 16508
2004-09-24 15:21:34 +00:00
Chris Lattner
e1b4d2a470
Move LHSI->hasOneUse() into the arms of the conditional, reindenting code.
...
No functionality changes here.
llvm-svn: 16505
2004-09-23 21:52:49 +00:00
Chris Lattner
8fc5af4da9
Implement Transforms/InstCombine/and.ll:test18, a case that occurs 20 times
...
in perlbmk
llvm-svn: 16504
2004-09-23 21:46:38 +00:00
Chris Lattner
bdcf41a8a2
Implement select.ll:test16: fold load (select C, X, null) -> load X
...
llvm-svn: 16499
2004-09-23 15:46:00 +00:00
Reid Spencer
9904928ea3
Patch for MINGW. Patch provided by Henrik Bach.
...
llvm-svn: 16495
2004-09-23 14:47:10 +00:00
Nate Begeman
49cf74b26c
Fix the last of the major PPC GEP folding deficiencies. This will allow
...
the ISel to use indexed and non-zero immediate offsets for GEPs that have
more than one use. This is common for instruction sequences such as a load
followed by a modify and store to the same address.
llvm-svn: 16493
2004-09-23 05:31:33 +00:00
Misha Brukman
379a05b52a
Use the V8/V9 shared register file description
...
llvm-svn: 16485
2004-09-22 21:48:50 +00:00
Misha Brukman
3c08658d56
Combine the F2 and F3 instruction classes into one file for simplicity
...
llvm-svn: 16484
2004-09-22 21:38:42 +00:00
Misha Brukman
efce1ef9c3
Fix file header path
...
llvm-svn: 16483
2004-09-22 21:29:12 +00:00
Misha Brukman
009aaf9968
Prettify formatting of the file, adjust paths to making V8 a subdir of Sparc
...
llvm-svn: 16482
2004-09-22 20:09:29 +00:00
Misha Brukman
d55f854a5d
V8 is now a subdirectory of Sparc; adjust paths accordingly
...
llvm-svn: 16481
2004-09-22 20:08:52 +00:00
Nate Begeman
033b816171
add optimized code sequences for setcc x, 0
...
llvm-svn: 16478
2004-09-22 04:40:25 +00:00
Chris Lattner
b121ae1cec
Do not fold (X + C1 != C2) if there are other users of the add. Doing
...
this transformation used to take a loop like this:
int Array[1000];
void test(int X) {
int i;
for (i = 0; i < 1000; ++i)
Array[i] += X;
}
Compiled to LLVM is:
no_exit: ; preds = %entry, %no_exit
%indvar = phi uint [ 0, %entry ], [ %indvar.next, %no_exit ] ; <uint> [#uses=2]
%tmp.4 = getelementptr [1000 x int]* %Array, int 0, uint %indvar ; <int*> [#uses=2]
%tmp.7 = load int* %tmp.4 ; <int> [#uses=1]
%tmp.9 = add int %tmp.7, %X ; <int> [#uses=1]
store int %tmp.9, int* %tmp.4
*** %indvar.next = add uint %indvar, 1 ; <uint> [#uses=2]
*** %exitcond = seteq uint %indvar.next, 1000 ; <bool> [#uses=1]
br bool %exitcond, label %return, label %no_exit
and turn it into a loop like this:
no_exit: ; preds = %entry, %no_exit
%indvar = phi uint [ 0, %entry ], [ %indvar.next, %no_exit ] ; <uint> [#uses=3]
%tmp.4 = getelementptr [1000 x int]* %Array, int 0, uint %indvar ; <int*> [#uses=2]
%tmp.7 = load int* %tmp.4 ; <int> [#uses=1]
%tmp.9 = add int %tmp.7, %X ; <int> [#uses=1]
store int %tmp.9, int* %tmp.4
*** %indvar.next = add uint %indvar, 1 ; <uint> [#uses=1]
*** %exitcond = seteq uint %indvar, 999 ; <bool> [#uses=1]
br bool %exitcond, label %return, label %no_exit
Note that indvar.next and indvar can no longer be coallesced. In machine
code terms, this patch changes this code:
.LBBtest_1: # no_exit
mov %EDX, OFFSET Array
mov %ESI, %EAX
add %ESI, DWORD PTR [%EDX + 4*%ECX]
mov %EDX, OFFSET Array
mov DWORD PTR [%EDX + 4*%ECX], %ESI
mov %EDX, %ECX
inc %EDX
cmp %ECX, 999
mov %ECX, %EDX
jne .LBBtest_1 # no_exit
into this:
.LBBtest_1: # no_exit
mov %EDX, OFFSET Array
mov %ESI, %EAX
add %ESI, DWORD PTR [%EDX + 4*%ECX]
mov %EDX, OFFSET Array
mov DWORD PTR [%EDX + 4*%ECX], %ESI
inc %ECX
cmp %ECX, 1000
jne .LBBtest_1 # no_exit
We need better instruction selection to get this:
.LBBtest_1: # no_exit
add DWORD PTR [Array + 4*%ECX], EAX
inc %ECX
cmp %ECX, 1000
jne .LBBtest_1 # no_exit
... but at least there is less register juggling
llvm-svn: 16473
2004-09-21 21:35:23 +00:00
Alkis Evlogimenos
89dd63733a
The real x87 floating point registers should not be allocatable. They
...
are only used by the stackifier when transforming FPn register
allocations to the real stack file x87 registers.
llvm-svn: 16472
2004-09-21 21:22:11 +00:00
Misha Brukman
6b17bf7193
s/ISel/PPC64ISel/ to have unique class names for debugging via gdb because the
...
C++ front-end in gcc does not mangle classes in anonymous namespaces correctly.
llvm-svn: 16471
2004-09-21 18:22:33 +00:00
Misha Brukman
87201ce8f9
s/ISel/PPC32ISel/ to have unique class names for debugging via gdb because the
...
C++ front-end in gcc does not mangle classes in anonymous namespaces correctly.
llvm-svn: 16470
2004-09-21 18:22:19 +00:00
Misha Brukman
43bd39e04e
s/ISel/X86ISel/ to have unique class names for debugging via gdb because the C++
...
front-end in gcc does not mangle classes in anonymous namespaces correctly.
llvm-svn: 16469
2004-09-21 18:21:21 +00:00
Chris Lattner
0f28cce60c
Make sure to set the operand list
...
llvm-svn: 16466
2004-09-21 17:30:54 +00:00
Chris Lattner
42618551d5
Fix potential miscompilations: InstCombine/2004-09-20-BadLoadCombine*.llx
...
llvm-svn: 16447
2004-09-20 10:15:10 +00:00
Alkis Evlogimenos
d59cebf87a
Fix loop condition so that we don't decrement off the beginning of the
...
list.
llvm-svn: 16440
2004-09-20 06:42:58 +00:00
Chris Lattner
4f2cf030e8
'Pass' should now not be derived from by clients. Instead, they should derive
...
from ModulePass. Instead of implementing Pass::run, then should implement
ModulePass::runOnModule.
llvm-svn: 16436
2004-09-20 04:48:05 +00:00
Chris Lattner
79e523de04
'Pass' should now not be derived from by clients. Instead, they should derive
...
from ModulePass. Instead of implementing Pass::run, then should implement
ModulePass::runOnModule.
Also, fix some undefined behavior, expecting | on booleans to evaluate
left-to-right.
llvm-svn: 16435
2004-09-20 04:47:19 +00:00
Chris Lattner
135419193e
Finegrainify namespacification
...
'Pass' should now not be derived from by clients. Instead, they should derive
from ModulePass. Instead of implementing Pass::run, then should implement
ModulePass::runOnModule.
llvm-svn: 16434
2004-09-20 04:44:31 +00:00
Chris Lattner
cd671065be
Prototype more accurately
...
llvm-svn: 16433
2004-09-20 04:43:57 +00:00
Chris Lattner
3e86084641
Prototype these functions more accurately
...
llvm-svn: 16432
2004-09-20 04:43:15 +00:00
Reid Spencer
45101043ef
Put in a #error in the event that we don't have an mmap that can map a file
...
into memor. This is just a reminder that the ReadFileIntoAddressSpace
function needs to be properly converted to lib/System and implemented via
read/write if there's no mmap of file support.
llvm-svn: 16428
2004-09-20 04:13:43 +00:00
Chris Lattner
e6f13093e6
Make isSafeToLoadUnconditionally a bit smarter, implementing PR362 and
...
Regression/Transforms/InstCombine/CPP_min_max.llx
llvm-svn: 16409
2004-09-19 19:18:10 +00:00
Chris Lattner
d6d99dfa3f
Fix a nasty iterator invalidation problem I introduced yesterday. This
...
unfortunately is the cause of a bunch of failures from tonight, and the
reason the tester is running so slow :(
llvm-svn: 16407
2004-09-19 19:01:06 +00:00
Chris Lattner
855a4ff4dd
Remove a whole bunch of horrible hacky code that was used to promote allocas
...
whose addresses where used by trivial phi nodes and select instructions. This
is now performed by the instcombine pass, which is more powerful, is much
simpler, and is faster. This allows the deletion of a bunch of code, two
FIXME's and two gotos.
llvm-svn: 16406
2004-09-19 18:51:51 +00:00
Chris Lattner
f62ea8ef4b
Make instruction combining a bit more aggressive in the face of volatile
...
loads, and implement two new transforms: InstCombine/load.ll:test[56].
llvm-svn: 16404
2004-09-19 18:43:46 +00:00
Reid Spencer
1bdd0f0a08
Minor correction to Signals implementation.
...
Patch submitted by Jeff Cohen. Thanks Jeff!
llvm-svn: 16401
2004-09-19 05:37:39 +00:00
Chris Lattner
9864df96ba
Add comment
...
llvm-svn: 16400
2004-09-19 01:05:16 +00:00
Chris Lattner
6455c51ab6
Fix the inliner to always delete any edges from the external call node to
...
a function being deleted. Due to optimizations done while inlining, there
can be edges from the external call node to a function node that were not
apparent any longer.
This fixes the compiler crash while compiling 175.vpr
llvm-svn: 16399
2004-09-18 21:37:03 +00:00
Chris Lattner
824a21868b
Add CallGraphNode::removeAnyCallEdgeTo method
...
llvm-svn: 16398
2004-09-18 21:34:34 +00:00
Reid Spencer
91e6f5e975
Use the /dev/zero device as the device on which the pages are mapped.
...
Patch contributed by Henrik Bach. Thanks Henrik!
llvm-svn: 16397
2004-09-18 19:34:09 +00:00
Reid Spencer
0e86336edc
Porting of Unix implementation to Win32.
...
Patch contributed by Jeff Cohen. Thanks Jeff!
llvm-svn: 16396
2004-09-18 19:29:16 +00:00
Reid Spencer
36e3cbfd3b
Get rid of file descriptor leak in create_file.
...
llvm-svn: 16395
2004-09-18 19:25:11 +00:00
Chris Lattner
37b6c4f2d2
Convert this pass to be a CallGraphSCCPass instead of a Pass, which eliminates
...
the worklist and makes it more efficient. This does not change functionality
at all.
llvm-svn: 16390
2004-09-18 00:34:13 +00:00
Chris Lattner
475dc2c93d
Make sure to remove the Select instruction as well
...
llvm-svn: 16389
2004-09-18 00:32:40 +00:00
Chris Lattner
85d5ccc006
When changing a function, make sure to update the CallGraphNode for the
...
function, not just the CallGraph.
llvm-svn: 16388
2004-09-18 00:27:20 +00:00