OSDN Git Service
Brian Gaeke [Wed, 29 Sep 2004 16:45:47 +0000 (16:45 +0000)]
Recognize FpMOVD as a move.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16586
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Wed, 29 Sep 2004 05:43:32 +0000 (05:43 +0000)]
Do not insert trivially dead select instructions, which allows us to
potentially fold more in one pass.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16583
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Wed, 29 Sep 2004 05:07:12 +0000 (05:07 +0000)]
Fold binary expressions and casts into PHI nodes that have all constant inputs.
This takes something like this:
%A = phi int [ 3, %cond_false.0 ], [ 2, %endif.0.i ], [ 2, %endif.1.i ]
%B = div int %tmp.243, 4
and turns it into:
%A = phi int [ 3/4, %cond_false.0 ], [ 2/4, %endif.0.i ], [ 2/4, %endif.1.i ]
which is later simplified (in this case) into %A = 0.
This triggers thousands of times in spec, for example, 269 times in 176.gcc.
This is tested by InstCombine/add.ll:test23 and set.ll:test18.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16582
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Wed, 29 Sep 2004 05:06:12 +0000 (05:06 +0000)]
New testcases
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16581
91177308-0d34-0410-b5e6-
96231b3b80d8
Nate Begeman [Wed, 29 Sep 2004 05:00:31 +0000 (05:00 +0000)]
Generate better code by being far less clever when it comes to the select instruction. Don't create overlapping register lifetimes
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16580
91177308-0d34-0410-b5e6-
96231b3b80d8
Brian Gaeke [Wed, 29 Sep 2004 03:48:55 +0000 (03:48 +0000)]
add results
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16579
91177308-0d34-0410-b5e6-
96231b3b80d8
Nate Begeman [Wed, 29 Sep 2004 03:45:33 +0000 (03:45 +0000)]
improve Type::BoolTy codegen by eliminating unnecessary clears and sign extends
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16578
91177308-0d34-0410-b5e6-
96231b3b80d8
Brian Gaeke [Wed, 29 Sep 2004 03:34:41 +0000 (03:34 +0000)]
Simplify copyConstantToRegister() for longs, using a pair of recursive calls.
Copy constant-pool entries' addresses into registers before loading out of them,
to avoid errors from the assembler.
Handle loading call args past the 6th one off the stack.
Add IMPLICIT_DEF pseudo-instrs for double and long arguments passed in register
pairs.
Use FpMOVD to copy doubles around instead of the horrible store-load thing we
were doing before.
Handle 'ret double' and 'ret long'.
Fix a bug in handling 'and/or/xor long'.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16577
91177308-0d34-0410-b5e6-
96231b3b80d8
Brian Gaeke [Wed, 29 Sep 2004 03:28:15 +0000 (03:28 +0000)]
Fix bug recognizing moves: isMoveInstr should only treat ORs with %g0 as
moves, not all ORs.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16576
91177308-0d34-0410-b5e6-
96231b3b80d8
Brian Gaeke [Wed, 29 Sep 2004 03:27:30 +0000 (03:27 +0000)]
Use FpMOVD pseudo-instruction to move doubles around.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16575
91177308-0d34-0410-b5e6-
96231b3b80d8
Brian Gaeke [Wed, 29 Sep 2004 03:27:29 +0000 (03:27 +0000)]
Add new FpMOVD pseudo-instruction, used to move doubles around.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16574
91177308-0d34-0410-b5e6-
96231b3b80d8
Brian Gaeke [Wed, 29 Sep 2004 03:26:27 +0000 (03:26 +0000)]
Fix double and long alignment.
Call the FPMover pass after register allocation.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16573
91177308-0d34-0410-b5e6-
96231b3b80d8
Brian Gaeke [Wed, 29 Sep 2004 03:25:40 +0000 (03:25 +0000)]
Put quotes around argument to .section directive.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16572
91177308-0d34-0410-b5e6-
96231b3b80d8
Brian Gaeke [Wed, 29 Sep 2004 03:25:39 +0000 (03:25 +0000)]
Add createSparcV8FPMoverPass().
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16571
91177308-0d34-0410-b5e6-
96231b3b80d8
Brian Gaeke [Wed, 29 Sep 2004 03:24:34 +0000 (03:24 +0000)]
Pass which converts FpMOVD (double move pseudoinstructions) to pairs
of FMOVS instrs.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16570
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Wed, 29 Sep 2004 03:18:20 +0000 (03:18 +0000)]
More testcases for recent instcombine patches
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16569
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Wed, 29 Sep 2004 03:16:24 +0000 (03:16 +0000)]
Hrm, really, all tests passed without this, but it is scary to think how...
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16568
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Wed, 29 Sep 2004 03:09:18 +0000 (03:09 +0000)]
Remove debugging printout
Instcombine (setcc (truncate X), C1).
This occurs THOUSANDS of times in many benchmarks. Particularlly common
seem to be things like (seteq (cast bool X to int), int 0)
This turns it into (seteq bool %X, false), which then becomes (not %X).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16567
91177308-0d34-0410-b5e6-
96231b3b80d8
Nate Begeman [Wed, 29 Sep 2004 02:35:05 +0000 (02:35 +0000)]
To go along with sabre's improved InstCombining, improve recognition of
integers that we can use as immediate values in instructions.
Example from yacr2:
- lis r10, -1
- ori r10, r10, 65535
- add r28, r28, r10
+ addi r28, r28, -1
addi r7, r7, 1
addi r9, r9, 1
b .LBB_main_9 ; loopentry.1.i214
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16566
91177308-0d34-0410-b5e6-
96231b3b80d8
Reid Spencer [Wed, 29 Sep 2004 00:01:17 +0000 (00:01 +0000)]
Improve validity checking of windows path names and fix file creation
problems.
Patch contributed by Jeff Cohen. Thanks Jeff!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16565
91177308-0d34-0410-b5e6-
96231b3b80d8
Reid Spencer [Tue, 28 Sep 2004 23:58:03 +0000 (23:58 +0000)]
Use llvm::sys::Path to destroy the file.
Patch contributed by Jeff Cohen. Thanks Jeff!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16564
91177308-0d34-0410-b5e6-
96231b3b80d8
Reid Spencer [Tue, 28 Sep 2004 23:56:20 +0000 (23:56 +0000)]
Win32 implementation of TimeValue::now().
Patch contributed by Jeff Cohen. Thanks Jeff!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16563
91177308-0d34-0410-b5e6-
96231b3b80d8
Reid Spencer [Tue, 28 Sep 2004 23:54:12 +0000 (23:54 +0000)]
FreeBSD implementation of the TimeValue::now() method.
Patch kindly contributed by Jeff Cohen. Thanks Jeff!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16562
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Tue, 28 Sep 2004 22:33:08 +0000 (22:33 +0000)]
Fold (X setcc C1) | (X setcc C2)
This implements or.ll:test1[89]
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16561
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Tue, 28 Sep 2004 22:31:49 +0000 (22:31 +0000)]
New testcase
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16560
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Tue, 28 Sep 2004 21:48:02 +0000 (21:48 +0000)]
Fold (and (setcc X, C1), (setcc X, C2))
This is important for several reasons:
1. Benchmarks have lots of code that looks like this (perlbmk in particular):
%tmp.2.i = setne int %tmp.0.i, 128 ; <bool> [#uses=1]
%tmp.6343 = seteq int %tmp.0.i, 1 ; <bool> [#uses=1]
%tmp.63 = and bool %tmp.2.i, %tmp.6343 ; <bool> [#uses=1]
we now fold away the setne, a clear improvement.
2. In the more important cases, such as (X >= 10) & (X < 20), we now produce
smaller code: (X-10) < 10.
3. Perhaps the nicest effect of this patch is that it really helps out the
code generators. In particular, for a 'range test' like the above,
instead of generating this on X86 (the difference on PPC is even more
pronounced):
cmp %EAX, 50
setge %CL
cmp %EAX, 100
setl %AL
and %CL, %AL
cmp %CL, 0
we now generate this:
add %EAX, -50
cmp %EAX, 50
Furthermore, this causes setcc's to be folded into branches more often.
These combinations trigger dozens of times in the spec benchmarks, particularly
in 176.gcc, 186.crafty, 253.perlbmk, 254.gap, & 099.go.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16559
91177308-0d34-0410-b5e6-
96231b3b80d8
Misha Brukman [Tue, 28 Sep 2004 21:46:18 +0000 (21:46 +0000)]
* Add `deplibs' keyword for specifying a list of dependent libraries
* Convert tabs to spaces
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16558
91177308-0d34-0410-b5e6-
96231b3b80d8
Misha Brukman [Tue, 28 Sep 2004 21:45:54 +0000 (21:45 +0000)]
Add `deplibs' keyword for specifying a list of dependent libraries
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16557
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Tue, 28 Sep 2004 21:39:35 +0000 (21:39 +0000)]
New testcases
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16556
91177308-0d34-0410-b5e6-
96231b3b80d8
Nate Begeman [Tue, 28 Sep 2004 21:29:00 +0000 (21:29 +0000)]
Add support for the isLoad and isStore flags, needed by the instruction scheduler
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16555
91177308-0d34-0410-b5e6-
96231b3b80d8
Nate Begeman [Tue, 28 Sep 2004 21:01:45 +0000 (21:01 +0000)]
Add support for the isLoad and isStore flags, needed by the instruction scheduler
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16554
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Tue, 28 Sep 2004 18:38:01 +0000 (18:38 +0000)]
Turn the hasDelaySlot flag into the M_DELAY_SLOT_FLAG
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16553
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Tue, 28 Sep 2004 18:37:00 +0000 (18:37 +0000)]
Be consistent with our naming
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16552
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Tue, 28 Sep 2004 18:34:14 +0000 (18:34 +0000)]
Capture delay slot info
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16551
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Tue, 28 Sep 2004 18:33:50 +0000 (18:33 +0000)]
Capture information about whether the target instructions have delay slots
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16550
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Tue, 28 Sep 2004 18:22:15 +0000 (18:22 +0000)]
Implement X / C1 / C2 folding
Implement (setcc (shl X, C1), C2) folding.
The second one occurs several dozen times in spec. The first was added
just in case. :)
These are tested by shift.ll:test2[12], and div.ll:test5
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16549
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Tue, 28 Sep 2004 18:21:01 +0000 (18:21 +0000)]
New testcase
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16548
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Tue, 28 Sep 2004 18:18:35 +0000 (18:18 +0000)]
New testcases
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16547
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Tue, 28 Sep 2004 17:54:07 +0000 (17:54 +0000)]
shl is always zero extending, so always use a zero extending shift right.
This latent bug was exposed by recent changes, and is tested as:
llvm/test/Regression/Transforms/InstCombine/2004-09-28-BadShiftAndSetCC.llx
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16546
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Tue, 28 Sep 2004 17:50:05 +0000 (17:50 +0000)]
New testcase for a latent bug in instcombine that I have been chasing for
1.5 days :(
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16545
91177308-0d34-0410-b5e6-
96231b3b80d8
Misha Brukman [Tue, 28 Sep 2004 16:58:12 +0000 (16:58 +0000)]
* InstructionReader.cpp was absorbed into Reader.cpp
* Wrap at 80 cols
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16543
91177308-0d34-0410-b5e6-
96231b3b80d8
Misha Brukman [Tue, 28 Sep 2004 16:57:46 +0000 (16:57 +0000)]
Fix doxygen comment
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16542
91177308-0d34-0410-b5e6-
96231b3b80d8
Brian Gaeke [Tue, 28 Sep 2004 16:04:00 +0000 (16:04 +0000)]
Touch output files before reading or writing them, so that they are
always guaranteed to exist. This fixes PR444. Thanks to Alkis
for reporting the bug and testing the patch.
AddRecord used to return a big list, but that return value was never
used. So now it doesn't return anything.
Create the WebDir if it does not exist.
Fix a typo in a comment.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16541
91177308-0d34-0410-b5e6-
96231b3b80d8
John Criswell [Tue, 28 Sep 2004 14:52:58 +0000 (14:52 +0000)]
Corrected spelling of Makefile variable, thereby re-enabling profile
builds for projects.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16540
91177308-0d34-0410-b5e6-
96231b3b80d8
Alkis Evlogimenos [Tue, 28 Sep 2004 14:42:44 +0000 (14:42 +0000)]
Add includes and use std:: for standard library calls to make code
compile on windows. This patch was contributed by Paolo Invernizzi.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16539
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Tue, 28 Sep 2004 03:45:22 +0000 (03:45 +0000)]
Bug fixed
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16538
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Tue, 28 Sep 2004 03:36:39 +0000 (03:36 +0000)]
Testcase for LLVM PR445
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16537
91177308-0d34-0410-b5e6-
96231b3b80d8
Alkis Evlogimenos [Tue, 28 Sep 2004 02:53:15 +0000 (02:53 +0000)]
Since we use alloca now make sure we include the proper headers for it.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16536
91177308-0d34-0410-b5e6-
96231b3b80d8
Alkis Evlogimenos [Tue, 28 Sep 2004 02:47:38 +0000 (02:47 +0000)]
Use alloca instead of a C99 style array. This should fix the
compilation problem in windows.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16535
91177308-0d34-0410-b5e6-
96231b3b80d8
Alkis Evlogimenos [Tue, 28 Sep 2004 02:40:37 +0000 (02:40 +0000)]
Pull assignment out of for loop conditional in order for this to
compile under windows. Patch contributed by Paolo Invernizzi!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16534
91177308-0d34-0410-b5e6-
96231b3b80d8
Alkis Evlogimenos [Tue, 28 Sep 2004 02:38:58 +0000 (02:38 +0000)]
Fix includes. Patch contributed by Paolo Invernizzi!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16533
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Tue, 28 Sep 2004 02:29:40 +0000 (02:29 +0000)]
New testcase that crashes the C++ FE, encountered while working on PR445
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16532
91177308-0d34-0410-b5e6-
96231b3b80d8
Alkis Evlogimenos [Tue, 28 Sep 2004 01:59:17 +0000 (01:59 +0000)]
Use class instead of struct for defining classes. This unbreaks the
build on windows. Patch contributed by Paolo Invernizzi!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16531
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Mon, 27 Sep 2004 21:51:25 +0000 (21:51 +0000)]
Patch contributed by Patrick Meredith:
added notes on the fact that the current implementation uses
sbyte* for va_list. Updated all occurances of valist to va_list (it was
inconsistant and find/replace is so easy ;-) ). Added <...> around all
occurances of va_list in the intrinsic functions to match the vaarg and
vanext instructions and to further show that va_list is a variable type.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16530
91177308-0d34-0410-b5e6-
96231b3b80d8
Misha Brukman [Mon, 27 Sep 2004 20:35:59 +0000 (20:35 +0000)]
Fix grammar.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16529
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Mon, 27 Sep 2004 19:29:18 +0000 (19:29 +0000)]
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.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16528
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Mon, 27 Sep 2004 19:25:20 +0000 (19:25 +0000)]
New testcase
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16527
91177308-0d34-0410-b5e6-
96231b3b80d8
Misha Brukman [Mon, 27 Sep 2004 18:22:18 +0000 (18:22 +0000)]
SparcV8 int regs are not only 32-bits in width, but they are 32-bit aligned!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16526
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Mon, 27 Sep 2004 16:59:06 +0000 (16:59 +0000)]
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.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16525
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Mon, 27 Sep 2004 16:41:01 +0000 (16:41 +0000)]
Hrm, if there is an error loading a file, try printing a message so the
user knows that...
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16524
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Mon, 27 Sep 2004 16:21:26 +0000 (16:21 +0000)]
Testcases for rev 250 of InstructionCombining.cpp
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16522
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Mon, 27 Sep 2004 16:18:50 +0000 (16:18 +0000)]
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.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16521
91177308-0d34-0410-b5e6-
96231b3b80d8
Alkis Evlogimenos [Mon, 27 Sep 2004 07:35:19 +0000 (07:35 +0000)]
Improve warning
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16520
91177308-0d34-0410-b5e6-
96231b3b80d8
Nate Begeman [Mon, 27 Sep 2004 05:08:17 +0000 (05:08 +0000)]
Correct some BuildMI arguments for the upcoming simple scheduler
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16519
91177308-0d34-0410-b5e6-
96231b3b80d8
Misha Brukman [Sun, 26 Sep 2004 21:07:43 +0000 (21:07 +0000)]
Fix the copy-pasto that Brian noticed: V8 int regs are 32-bits wide, not 64.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16518
91177308-0d34-0410-b5e6-
96231b3b80d8
Reid Spencer [Sat, 25 Sep 2004 16:00:07 +0000 (16:00 +0000)]
Add some missing #includes
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16517
91177308-0d34-0410-b5e6-
96231b3b80d8
Reid Spencer [Sat, 25 Sep 2004 15:59:41 +0000 (15:59 +0000)]
Qualify Path with sys:: namespace so this file compiles.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16516
91177308-0d34-0410-b5e6-
96231b3b80d8
Reid Spencer [Sat, 25 Sep 2004 08:32:37 +0000 (08:32 +0000)]
Updated to reflect changes in the interface of TimeValue::now().
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16515
91177308-0d34-0410-b5e6-
96231b3b80d8
Reid Spencer [Sat, 25 Sep 2004 08:29:54 +0000 (08:29 +0000)]
Clean up the interface of TimeValue:
- get rid of unneeded constructors
- get rid of duplicate methods/constructors/operators
- normalize to LLVM coding standards
- wrap to 80 columns.
Many thanks to Alkis Evlogimenos for his suggestions.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16514
91177308-0d34-0410-b5e6-
96231b3b80d8
Reid Spencer [Sat, 25 Sep 2004 05:03:54 +0000 (05:03 +0000)]
Added stub implementations of TimeValue concept for remaining platforms.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16513
91177308-0d34-0410-b5e6-
96231b3b80d8
Reid Spencer [Sat, 25 Sep 2004 05:03:22 +0000 (05:03 +0000)]
Wrap to 80 cols.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16512
91177308-0d34-0410-b5e6-
96231b3b80d8
Reid Spencer [Fri, 24 Sep 2004 23:25:19 +0000 (23:25 +0000)]
Initial implementation of the TimeValue abstraction.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16511
91177308-0d34-0410-b5e6-
96231b3b80d8
John Criswell [Fri, 24 Sep 2004 21:19:06 +0000 (21:19 +0000)]
Updated the last two header files so that they are configured with
AC_CONFIG_HEADERS. This should prevent LLVM from needlessly re-compiling
on a re-configure.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16510
91177308-0d34-0410-b5e6-
96231b3b80d8
John Criswell [Fri, 24 Sep 2004 18:28:00 +0000 (18:28 +0000)]
Modified hash_map and hash_set configuration so that they are not
regenerated on every run of configure.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16509
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Fri, 24 Sep 2004 15:21:34 +0000 (15:21 +0000)]
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.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16508
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Fri, 24 Sep 2004 15:18:43 +0000 (15:18 +0000)]
Add some tests for shr-and folding
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16507
91177308-0d34-0410-b5e6-
96231b3b80d8
John Criswell [Fri, 24 Sep 2004 13:28:51 +0000 (13:28 +0000)]
Configure Datatypes.h.in with AC_CONFIG_HEADERS. This should prevent it
from being re-generated if the new version is identical to the old version.
Hence, it should save us some recompiling after re-configures.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16506
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Thu, 23 Sep 2004 21:52:49 +0000 (21:52 +0000)]
Move LHSI->hasOneUse() into the arms of the conditional, reindenting code.
No functionality changes here.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16505
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Thu, 23 Sep 2004 21:46:38 +0000 (21:46 +0000)]
Implement Transforms/InstCombine/and.ll:test18, a case that occurs 20 times
in perlbmk
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16504
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Thu, 23 Sep 2004 21:42:49 +0000 (21:42 +0000)]
New testcase for a pattern that occurs 20 times in perlbmk
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16503
91177308-0d34-0410-b5e6-
96231b3b80d8
Reid Spencer [Thu, 23 Sep 2004 16:03:48 +0000 (16:03 +0000)]
Add a blurb about the use of class vs. struct keyword in class decls.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16500
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Thu, 23 Sep 2004 15:46:00 +0000 (15:46 +0000)]
Implement select.ll:test16: fold load (select C, X, null) -> load X
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16499
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Thu, 23 Sep 2004 15:45:20 +0000 (15:45 +0000)]
load null is undefined behavior, this should fold
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16498
91177308-0d34-0410-b5e6-
96231b3b80d8
Reid Spencer [Thu, 23 Sep 2004 14:49:45 +0000 (14:49 +0000)]
Change Value from a "struct" to a "class" so that VC 7.1 doesn't generate
missing symbols when its referenced as a class.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16496
91177308-0d34-0410-b5e6-
96231b3b80d8
Reid Spencer [Thu, 23 Sep 2004 14:47:10 +0000 (14:47 +0000)]
Patch for MINGW. Patch provided by Henrik Bach.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16495
91177308-0d34-0410-b5e6-
96231b3b80d8
Reid Spencer [Thu, 23 Sep 2004 06:07:24 +0000 (06:07 +0000)]
Missed one $*.a -> $@ conversion for the Release build.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16494
91177308-0d34-0410-b5e6-
96231b3b80d8
Nate Begeman [Thu, 23 Sep 2004 05:31:33 +0000 (05:31 +0000)]
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.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16493
91177308-0d34-0410-b5e6-
96231b3b80d8
Alkis Evlogimenos [Thu, 23 Sep 2004 04:39:30 +0000 (04:39 +0000)]
Update email address
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16492
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Thu, 23 Sep 2004 04:36:01 +0000 (04:36 +0000)]
Grrr... I shouldn't have to do this one :)
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16491
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Thu, 23 Sep 2004 04:30:05 +0000 (04:30 +0000)]
Update credits
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16490
91177308-0d34-0410-b5e6-
96231b3b80d8
Reid Spencer [Thu, 23 Sep 2004 00:58:06 +0000 (00:58 +0000)]
Correct the rules for making shared libraries per libtool 1.5.10
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16489
91177308-0d34-0410-b5e6-
96231b3b80d8
Misha Brukman [Wed, 22 Sep 2004 21:48:50 +0000 (21:48 +0000)]
Use the V8/V9 shared register file description
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16485
91177308-0d34-0410-b5e6-
96231b3b80d8
Misha Brukman [Wed, 22 Sep 2004 21:38:42 +0000 (21:38 +0000)]
Combine the F2 and F3 instruction classes into one file for simplicity
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16484
91177308-0d34-0410-b5e6-
96231b3b80d8
Misha Brukman [Wed, 22 Sep 2004 21:29:12 +0000 (21:29 +0000)]
Fix file header path
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16483
91177308-0d34-0410-b5e6-
96231b3b80d8
Misha Brukman [Wed, 22 Sep 2004 20:09:29 +0000 (20:09 +0000)]
Prettify formatting of the file, adjust paths to making V8 a subdir of Sparc
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16482
91177308-0d34-0410-b5e6-
96231b3b80d8
Misha Brukman [Wed, 22 Sep 2004 20:08:52 +0000 (20:08 +0000)]
V8 is now a subdirectory of Sparc; adjust paths accordingly
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16481
91177308-0d34-0410-b5e6-
96231b3b80d8
Reid Spencer [Wed, 22 Sep 2004 15:32:08 +0000 (15:32 +0000)]
Update to add the HAVE_UINT64_T test and the STACK_DIRECTION indicator.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16480
91177308-0d34-0410-b5e6-
96231b3b80d8
Reid Spencer [Wed, 22 Sep 2004 15:28:32 +0000 (15:28 +0000)]
The alloca function, strangely enough, is found in the malloc.h header file
on MINGW platform. Provide an #elseif case to #include malloc.h for this
platform if malloc.h is found.
Patch provided by Henrik Bach. Thanks Henrik!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16479
91177308-0d34-0410-b5e6-
96231b3b80d8
Nate Begeman [Wed, 22 Sep 2004 04:40:25 +0000 (04:40 +0000)]
add optimized code sequences for setcc x, 0
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16478
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Tue, 21 Sep 2004 21:35:23 +0000 (21:35 +0000)]
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
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@16473
91177308-0d34-0410-b5e6-
96231b3b80d8