Commit Graph

9099 Commits

Author SHA1 Message Date
Chris Lattner
c71d6949b9 Use the TargetInstrInfo::commuteInstruction method to commute instructions
instead of doing it manually.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19685 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-19 07:08:42 +00:00
Chris Lattner
167b10cba4 Finegrainify namespacification
Add default impl of commuteInstruction
Add notes about ugly V9 code.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19684 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-19 06:53:34 +00:00
Chris Lattner
8571637ee9 Codegen long >> 2 to this:
foo:
        mov %EAX, DWORD PTR [%ESP + 4]
        mov %EDX, DWORD PTR [%ESP + 8]
        shrd %EAX, %EDX, 2
        sar %EDX, 2
        ret

instead of this:

test1:
        mov %ECX, DWORD PTR [%ESP + 4]
        shr %ECX, 2
        mov %EDX, DWORD PTR [%ESP + 8]
        mov %EAX, %EDX
        shl %EAX, 30
        or %EAX, %ECX
        sar %EDX, 2
        ret

and long << 2 to this:

foo:
        mov %EAX, DWORD PTR [%ESP + 4]
        mov %ECX, DWORD PTR [%ESP + 8]
***     mov %EDX, %EAX
        shrd %EDX, %ECX, 30
        shl %EAX, 2
        ret

instead of this:

foo:
        mov %EAX, DWORD PTR [%ESP + 4]
        mov %ECX, %EAX
        shr %ECX, 30
        mov %EDX, DWORD PTR [%ESP + 8]
        shl %EDX, 2
        or %EDX, %ECX
        shl %EAX, 2
        ret

The extra copy (marked ***) can be eliminated when I teach the code generator
that shrd32rri8 is really commutative.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19681 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-19 06:18:43 +00:00
Chris Lattner
e34b396ab7 Implement a way of expanding shifts. This applies to targets that offer
select operations or to shifts that are by a constant.  This automatically
implements (with no special code) all of the special cases for shift by 32,
shift by < 32 and shift by > 32.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19679 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-19 04:19:40 +00:00
Chris Lattner
009b55b6f6 X86 shifts mask the amount.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19678 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-19 03:36:30 +00:00
Chris Lattner
d6e496732b Add a hook to find out how the target handles shift amounts that are out of
range.  Either they are undefined (the default), they mask the shift amount
to the size of the register (X86, Alpha, etc), or they extend the shift (PPC).

This defaults to undefined, which is conservatively correct.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19677 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-19 03:36:14 +00:00
Chris Lattner
d5d5682512 Zero is cheaper than sign extend.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19675 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-18 21:57:59 +00:00
Chris Lattner
477c931852 Code to handle FP_EXTEND is dead now. X86 doesn't support any data types to
FP_EXTEND from!


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19674 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-18 20:05:56 +00:00
Chris Lattner
3dd68151dc Remove more dead code.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19673 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-18 19:50:08 +00:00
Chris Lattner
a7a4c0f7c4 The selection dag code handles the promotions from F32 to F64 for us, so we
don't need to even think about F32 in the X86 code anymore.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19672 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-18 19:46:54 +00:00
Chris Lattner
47e9223e06 Fix some fixmes (promoting bools for select and brcond), fix promotion
of zero and sign extends.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19671 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-18 19:27:06 +00:00
Chris Lattner
4a9b4f1943 Keep track of the retval type as well.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19670 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-18 19:26:36 +00:00
Chris Lattner
ef5cd1d3cf Teach legalize to promote copy(from|to)reg, instead of making the isel pass
do it.  This results in better code on X86 for floats (because if strict
precision is not required, we can elide some more expensive double -> float
conversions like the old isel did), and allows other targets to emit
CopyFromRegs that are not legal for arguments.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19668 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-18 17:54:55 +00:00
Chris Lattner
b422aeac9e Fix 124.m88ksim.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19667 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-18 17:35:28 +00:00
Chris Lattner
811482a82b Do not emit loads multiple times, potentially in the wrong places.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19661 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-18 04:18:32 +00:00
Tanya Lattner
0111474bde Minor changes.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19660 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-18 04:15:41 +00:00
Chris Lattner
b71f8fc99d Eliminate bad assertions.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19659 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-18 04:00:54 +00:00
Chris Lattner
4a108668df * Eliminate the TokenSet and just use the ExprMap for both tokens and values.
* Insert some really pedantic assertions that will notice when we emit the
  same loads more than one time, exposing bugs.  This turns a miscompilation in
  bzip2 into a compile-fail.  yaay.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19658 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-18 03:51:59 +00:00
Chris Lattner
82fbfb6984 Teach legalize to promote SetCC results.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19657 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-18 02:59:52 +00:00
Chris Lattner
f30b73b1c6 Allow setcc operations to have nonbool types.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19656 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-18 02:52:03 +00:00
Chris Lattner
883c86f510 Rely on the code in MatchAddress to do this work. Otherwise we fail to
match (X+Y)+(Z << 1), because we match the X+Y first, consuming the index
register, then there is no place to put the Z.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19652 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-18 02:25:52 +00:00
Chris Lattner
6e4e7652d8 Fix the completely broken FP constant folds for setcc's.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19651 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-18 02:11:55 +00:00
Chris Lattner
98a8ba0393 Fix a problem where probing for addressing modes caused expressions to be
emitted too early.  In particular, this fixes
Regression/CodeGen/X86/regpressure.ll:regpressure3.

This also improves the 2nd basic block in 164.gzip:flush_block, which went from

.LBBflush_block_1:      # loopentry.1.i
        movzx %EAX, WORD PTR [dyn_ltree + 20]
        movzx %ECX, WORD PTR [dyn_ltree + 16]
        mov DWORD PTR [%ESP + 32], %ECX
        movzx %ECX, WORD PTR [dyn_ltree + 12]
        movzx %EDX, WORD PTR [dyn_ltree + 8]
        movzx %EBX, WORD PTR [dyn_ltree + 4]
        mov DWORD PTR [%ESP + 36], %EBX
        movzx %EBX, WORD PTR [dyn_ltree]
        add DWORD PTR [%ESP + 36], %EBX
        add %EDX, DWORD PTR [%ESP + 36]
        add %ECX, %EDX
        add DWORD PTR [%ESP + 32], %ECX
        add %EAX, DWORD PTR [%ESP + 32]
        movzx %ECX, WORD PTR [dyn_ltree + 24]
        add %EAX, %ECX
        mov %ECX, 0
        mov %EDX, %ECX

to

.LBBflush_block_1:      # loopentry.1.i
        movzx %EAX, WORD PTR [dyn_ltree]
        movzx %ECX, WORD PTR [dyn_ltree + 4]
        add %ECX, %EAX
        movzx %EAX, WORD PTR [dyn_ltree + 8]
        add %EAX, %ECX
        movzx %ECX, WORD PTR [dyn_ltree + 12]
        add %ECX, %EAX
        movzx %EAX, WORD PTR [dyn_ltree + 16]
        add %EAX, %ECX
        movzx %ECX, WORD PTR [dyn_ltree + 20]
        add %ECX, %EAX
        movzx %EAX, WORD PTR [dyn_ltree + 24]
        add %ECX, %EAX
        mov %EAX, 0
        mov %EDX, %EAX

... which results in less spilling in the function.

This change alone speeds up 164.gzip from 37.23s to 36.24s on apoc.  The
default isel takes 37.31s.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19650 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-18 01:06:26 +00:00
Chris Lattner
e9fe2bcac2 Fix indentation.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19649 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-17 23:25:45 +00:00
Chris Lattner
869e0434ae Don't bother using max here.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19647 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-17 23:02:13 +00:00
Chris Lattner
90d1be7eef Do not give token factor nodes outrageous weights
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19645 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-17 22:56:09 +00:00
Chris Lattner
d3948116b8 Non-volatile loads can be freely reordered against each other. This fixes
X86/reg-pressure.ll again, and allows us to do nice things in other cases.
For example, we now codegen this sort of thing:

int %loadload(int *%X, int* %Y) {
  %Z = load int* %Y
  %Y = load int* %X      ;; load between %Z and store
  %Q = add int %Z, 1
  store int %Q, int* %Y
  ret int %Y
}

Into this:

loadload:
        mov %EAX, DWORD PTR [%ESP + 4]
        mov %EAX, DWORD PTR [%EAX]
        mov %ECX, DWORD PTR [%ESP + 8]
        inc DWORD PTR [%ECX]
        ret

where we weren't able to form the 'inc [mem]' before.  This also lets the
instruction selector emit loads in any order it wants to, which can be good
for register pressure as well.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19644 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-17 22:19:26 +00:00
Chris Lattner
5c65981a57 Two changes:
1. Fold  [mem] += (1|-1) into inc [mem]/dec [mem] to save some icache space.
 2. Do not let token factor nodes prevent forming '[mem] op= val' folds.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19643 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-17 22:10:42 +00:00
Chris Lattner
a651cf67b7 Don't call SelectionDAG.getRoot() directly, go through a forwarding method.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19642 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-17 19:43:36 +00:00
Chris Lattner
e10269ba30 Refactor load/op/store folding into it's own method, no functionality changes.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19641 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-17 19:25:26 +00:00
Chris Lattner
0afa8e348e Implement a target independent optimization to codegen arguments only into
the basic block that uses them if possible.  This is a big win on X86, as it
lets us fold the argument loads into instructions and reduce register pressure
(by not loading all of the arguments in the entry block).

For this (contrived to show the optimization) testcase:

int %argtest(int %A, int %B) {
        %X = sub int 12345, %A
        br label %L
L:
        %Y = add int %X, %B
        ret int %Y
}

we used to produce:

argtest:
        mov %ECX, DWORD PTR [%ESP + 4]
        mov %EAX, 12345
        sub %EAX, %ECX
        mov %EDX, DWORD PTR [%ESP + 8]
.LBBargtest_1:  # L
        add %EAX, %EDX
        ret


now we produce:

argtest:
        mov %EAX, 12345
        sub %EAX, DWORD PTR [%ESP + 4]
.LBBargtest_1:  # L
        add %EAX, DWORD PTR [%ESP + 8]
        ret

This also fixes the FIXME in the code.

BTW, this occurs in real code.  164.gzip shrinks from 8623 to 8608 lines of
.s file.  The stack frame in huft_build shrinks from 1644->1628 bytes,
inflate_codes shrinks from 116->108 bytes, and inflate_block from 2620->2612,
due to fewer spills.

Take that alkis. :-)


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19639 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-17 17:55:19 +00:00
Chris Lattner
ae3ffba466 Fix a major regression last night that prevented us from producing [mem] op= reg
operations.

The body of the if is less indented but unmodified in this patch.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19638 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-17 17:49:14 +00:00
Chris Lattner
068a81e9fc Refactor code into a new method.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19635 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-17 17:15:02 +00:00
Chris Lattner
2a4e508281 Codegen this:
int %foo(int %X) {
        %T = add int %X, 13
        %S = mul int %T, 3
        ret int %S
}

as this:

        mov %ECX, DWORD PTR [%ESP + 4]
        lea %EAX, DWORD PTR [%ECX + 2*%ECX + 39]
        ret

instead of this:

        mov %ECX, DWORD PTR [%ESP + 4]
        mov %EAX, %ECX
        add %EAX, 13
        imul %EAX, %EAX, 3
        ret


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19633 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-17 06:48:02 +00:00
Tanya Lattner
7a866d9a43 Added tmp instructions to preserve ssa.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19632 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-17 06:47:26 +00:00
Chris Lattner
4ff348befa Fix test/Regression/CodeGen/X86/2005-01-17-CycleInDAG.ll and 132.ijpeg.
Do not fold a load into an operation if it will induce a cycle in the DAG.

Repeat after me: dAg.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19631 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-17 06:26:58 +00:00
Chris Lattner
a3fd1c569c Delete PHI nodes that are not dead but are locked in a cycle of single
useness.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19629 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-17 05:10:15 +00:00
Chris Lattner
b352fa5853 Move code out of indentation one level to make it easier to read.
Disable the xform for < > cases.  It turns out that the following is being
miscompiled:

bool %test(sbyte %S) {
        %T = cast sbyte %S to uint
        %V = setgt uint %T, 255
        ret bool %V
}


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19628 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-17 03:20:02 +00:00
Chris Lattner
cb1aa8d772 Do not fold a load into a comparison that is used by more than one place.
The comparison will probably be folded, so this is not ok to do.
This fixed 197.parser.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19624 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-17 01:34:14 +00:00
Chris Lattner
c98279d371 Do not codegen 'xor bool, true' as 'not reg'. not reg inverts the upper bits
of the bytereg.  This fixes yacr2, 300.twolf and probably others.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19622 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-17 00:23:16 +00:00
Chris Lattner
4df0de975a Set up the shift and setcc types.
If we emit a load because we followed a token chain to get to it, try to
fold it into its single user if possible.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19620 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-17 00:00:33 +00:00
Chris Lattner
714b69d047 Shift and setcc types default to the pointer type.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19619 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-16 23:59:48 +00:00
Chris Lattner
8ac532c55e Implement legalize of call nodes.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19617 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-16 19:46:48 +00:00
Tanya Lattner
7600587e66 Added paramters to a few functions in order to allow me to change the functions to preserve SSA
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19615 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-16 08:51:10 +00:00
Chris Lattner
da2ce11407 * Adjust to changes in TargetLowering interfaces.
* Remove custom promotion for bool and byte select ops.  Legalize now
  promotes them for us.
* Allow folding ConstantPoolIndexes into EXTLOAD's, useful for float immediates.
* Declare which operations are not supported better.
* Add some hacky code for TRUNCSTORE to pretend that we have truncstore
  for i16 types.  This is useful for testing promotion code because I can
  just remove 16-bit registers all together and verify that programs work.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19614 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-16 07:34:08 +00:00
Chris Lattner
55ba8fba75 Revamp supported ops. Instead of just being supported or not, we now keep
track of how to deal with it, and provide the target with a hook that they
can use to legalize arbitrary operations in arbitrary ways.

Implement custom lowering for a couple of ops, implement promotion for select
operations (which x86 needs).


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19613 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-16 07:29:19 +00:00
Chris Lattner
171453a284 add method stub
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19612 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-16 07:28:41 +00:00
Chris Lattner
4b7899343e Don't mash stuff together.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19611 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-16 07:28:31 +00:00
Chris Lattner
cba82f9339 Use enums, move virtual dtor out of line.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19610 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-16 07:28:11 +00:00
Chris Lattner
f8161d83f0 Implement some more missing promotions.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19606 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-16 05:06:12 +00:00