Commit Graph

18465 Commits

Author SHA1 Message Date
Chris Lattner
23993561e2 Instead of making ZERO_EXTEND_INREG nodes, use the helper method in
SelectionDAG to do the job with AND.  Don't legalize Z_E_I anymore as
it is gone


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21266 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-13 02:38:47 +00:00
Chris Lattner
0f2287baa1 Remove all foldings of ZERO_EXTEND_INREG, moving them to work for AND nodes
instead.  OVerall, this increases the amount of folding we can do.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21265 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-13 02:38:18 +00:00
Chris Lattner
1ccae666f5 Add a new helper method which returns the and that is equivalent to what
ZERO_EXTEND_INREG was.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21264 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-13 02:37:19 +00:00
Chris Lattner
ea57610689 Remove the ZERO_EXTEND_INREG node which is redundant with AND
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21263 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-13 02:36:41 +00:00
Nate Begeman
db81ebab89 Fold shift x, [sz]ext(y) -> shift x, y
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21262 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-12 23:32:28 +00:00
Nate Begeman
b882752bd0 Fold shift by size larger than type size to undef
Make llvm undef values generate ISD::UNDEF nodes


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21261 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-12 23:12:17 +00:00
Nate Begeman
9765c25eb7 Implement setcc op, -1 sequences
Remove dead setcc op, 0 sequences
Coming later: generalization of op, imm


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21260 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-12 21:22:28 +00:00
Chris Lattner
1c51c6ac13 promote extload i1 -> extload i8
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21258 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-12 20:30:10 +00:00
Chris Lattner
3464547968 add an argument to allow avoiding deleting phi nodes.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21255 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-12 18:52:14 +00:00
Chris Lattner
0b2fc9b560 add an argument.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21254 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-12 18:51:53 +00:00
Chris Lattner
151c80be81 Get rid of this for_each loop
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21253 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-12 18:51:33 +00:00
Duraid Madina
e8fd25f5c5 * OK, after changing to use liveIn/liveOut instead of IDEFs,
to avoid redundant mov out3=r44 type instructions, we need to
tell the register allocator the truth about out? registers.

FIXME: unfortunately, since the list of allocatable registers is immutable,
we can't simply 'delete r127' from the allocation order, say, if 'out0' is
used. The only correct thing we can do is have a linear order of regs:

out7, out6 ... out2, out1, out0, r32, r33, r34 ... r126, r127

and slide a 'window' of 96 registers along this line, depending on how many
of the out? regs a function actually uses. The only downside of this is
that the out? registers will be allocated _first_, which makes the
resulting assembly ugly. :( Note this in the README. Hope this gets fixed
soon. :) (note the 3rd person speech there)


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21252 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-12 18:42:59 +00:00
Andrew Lenharth
0b04b5d562 Get rid of idefs for arguments (oops)
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21251 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-12 17:47:57 +00:00
Andrew Lenharth
e1c5a00e8b Get rid of idefs for arguments
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21250 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-12 17:35:16 +00:00
Chris Lattner
9c24ba642e Put out* into the allocation order, allowing the register allocator to
coallesce moves into outgoing args.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21249 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-12 15:12:51 +00:00
Chris Lattner
ea6f770fb0 Make sure to realize that calls use their argument regs
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21248 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-12 15:12:19 +00:00
Duraid Madina
ca494fddcd stop emitting IDEFs for args - change to using liveIn/liveOut
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21247 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-12 14:54:44 +00:00
Nate Begeman
7af0248af4 Initial support for allocation condition registers
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21246 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-12 07:04:16 +00:00
Chris Lattner
af2e3e0859 Fix a crash analyzing MultiSource/Benchmarks/MallocBench/gs
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21245 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-12 03:59:27 +00:00
Chris Lattner
fd1f1ee0ba Remove some redundant checks, add a couple of new ones. This allows us to
compile this:

int foo (unsigned long a, unsigned long long g) {
  return a >= g;
}

To:

foo:
        movl 8(%esp), %eax
        cmpl %eax, 4(%esp)
        setae %al
        cmpl $0, 12(%esp)
        sete %cl
        andb %al, %cl
        movzbl %cl, %eax
        ret

instead of:

foo:
        movl 8(%esp), %eax
        cmpl %eax, 4(%esp)
        setae %al
        movzbw %al, %cx
        movl 12(%esp), %edx
        cmpl $0, %edx
        sete %al
        movzbw %al, %ax
        cmpl $0, %edx
        cmove %cx, %ax
        movzbl %al, %eax
        ret


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21244 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-12 02:54:39 +00:00
Chris Lattner
5b95ed652f Emit comparisons against the sign bit better. Codegen this:
bool %test1(long %X) {
        %A = setlt long %X, 0
        ret bool %A
}

like this:

test1:
        cmpl $0, 8(%esp)
        setl %al
        movzbl %al, %eax
        ret

instead of:

test1:
        movl 8(%esp), %ecx
        cmpl $0, %ecx
        setl %al
        movzbw %al, %ax
        cmpl $0, 4(%esp)
        setb %dl
        movzbw %dl, %dx
        cmpl $0, %ecx
        cmove %dx, %ax
        movzbl %al, %eax
        ret


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21243 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-12 02:19:10 +00:00
Chris Lattner
08b698e38d Emit long comparison against -1 better. Instead of this (x86):
test2:
        movl 8(%esp), %eax
        notl %eax
        movl 4(%esp), %ecx
        notl %ecx
        orl %eax, %ecx
        cmpl $0, %ecx
        sete %al
        movzbl %al, %eax
        ret

or this (PPC):

_test2:
        nor r2, r4, r4
        nor r3, r3, r3
        or r2, r2, r3
        cntlzw r2, r2
        srwi r3, r2, 5
        blr

Emit this:

test2:
        movl 8(%esp), %eax
        andl 4(%esp), %eax
        cmpl $-1, %eax
        sete %al
        movzbl %al, %eax
        ret

or this:

_test2:
.LBB_test2_0:   ;
        and r2, r4, r3
        cmpwi cr0, r2, -1
        li r3, 1
        li r2, 0
        beq .LBB_test2_2        ;
.LBB_test2_1:   ;
        or r3, r2, r2
.LBB_test2_2:   ;
        blr

it seems like the PPC isel could do better for R32 == -1 case.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21242 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-12 01:46:05 +00:00
Chris Lattner
3b2c1d95a2 canonicalize x <u 1 -> x == 0. On this testcase:
unsigned long long g;
unsigned long foo (unsigned long a) {
  return (a >= g) ? 1 : 0;
}

It changes the ppc code from:

_foo:
.LBB_foo_0:     ; entry
        mflr r11
        stw r11, 8(r1)
        bl "L00000$pb"
"L00000$pb":
        mflr r2
        addis r2, r2, ha16(L_g$non_lazy_ptr-"L00000$pb")
        lwz r2, lo16(L_g$non_lazy_ptr-"L00000$pb")(r2)
        lwz r4, 0(r2)
        lwz r2, 4(r2)
        cmplw cr0, r3, r2
        li r2, 1
        li r3, 0
        bge .LBB_foo_2  ; entry
.LBB_foo_1:     ; entry
        or r2, r3, r3
.LBB_foo_2:     ; entry
        cmplwi cr0, r4, 1
        li r3, 1
        li r5, 0
        blt .LBB_foo_4  ; entry
.LBB_foo_3:     ; entry
        or r3, r5, r5
.LBB_foo_4:     ; entry
        cmpwi cr0, r4, 0
        beq .LBB_foo_6  ; entry
.LBB_foo_5:     ; entry
        or r2, r3, r3
.LBB_foo_6:     ; entry
        rlwinm r3, r2, 0, 31, 31
        lwz r11, 8(r1)
        mtlr r11
        blr


to:

_foo:
.LBB_foo_0:     ; entry
        mflr r11
        stw r11, 8(r1)
        bl "L00000$pb"
"L00000$pb":
        mflr r2
        addis r2, r2, ha16(L_g$non_lazy_ptr-"L00000$pb")
        lwz r2, lo16(L_g$non_lazy_ptr-"L00000$pb")(r2)
        lwz r4, 0(r2)
        lwz r2, 4(r2)
        cmplw cr0, r3, r2
        li r2, 1
        li r3, 0
        bge .LBB_foo_2  ; entry
.LBB_foo_1:     ; entry
        or r2, r3, r3
.LBB_foo_2:     ; entry
        cntlzw r3, r4
        srwi r3, r3, 5
        cmpwi cr0, r4, 0
        beq .LBB_foo_4  ; entry
.LBB_foo_3:     ; entry
        or r2, r3, r3
.LBB_foo_4:     ; entry
        rlwinm r3, r2, 0, 31, 31
        lwz r11, 8(r1)
        mtlr r11
        blr


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21241 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-12 00:28:49 +00:00
Nate Begeman
9f833d3085 Implement bitfield clears
Implement divide by negative power of two


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21240 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-12 00:10:02 +00:00
Nate Begeman
ef9531efed Update PPC readme. Remove things that are done or aren't ppc specific
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21232 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-11 20:48:57 +00:00
Chris Lattner
bd0781ed12 Teach the dag mechanism that this:
long long test2(unsigned A, unsigned B) {
        return ((unsigned long long)A << 32) + B;
}

is equivalent to this:

long long test1(unsigned A, unsigned B) {
        return ((unsigned long long)A << 32) | B;
}

Now they are both codegen'd to this on ppc:

_test2:
        blr

or this on x86:

test2:
        movl 4(%esp), %edx
        movl 8(%esp), %eax
        ret


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21231 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-11 20:29:59 +00:00
Chris Lattner
ee27f57a6a Fix expansion of shifts by exactly NVT bits on arch's (like X86) that have
masking shifts.

This fixes the miscompilation of this:

long long test1(unsigned A, unsigned B) {
        return ((unsigned long long)A << 32) | B;
}

into this:

test1:
        movl 4(%esp), %edx
        movl %edx, %eax
        orl 8(%esp), %eax
        ret

allowing us to generate this instead:

test1:
        movl 4(%esp), %edx
        movl 8(%esp), %eax
        ret


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21230 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-11 20:08:52 +00:00
Chris Lattner
30e82431b3 IA64 supports this operation.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21228 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-11 18:55:36 +00:00
Chris Lattner
5eef9f3bc9 ORo sets CR0
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21227 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-11 15:03:48 +00:00
Chris Lattner
6b4ea2cfa2 Revert the previous patch, which I didn't mean to check in.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21226 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-11 15:03:41 +00:00
Chris Lattner
26d4fdb968 Fix a minor bug (ORo didn't mark that it set CR0).
Refactor how . instructions are handled.  In particular, instead of passing
the RC flag all the way up the inheritance hierarchy, just make a new tblgen
class 'DOT' which can be added to an instruction definition.

For example, instead of this:

-def AND  : XForm_6<31,  28, 0, 0, 0, (ops GPRC:$rA, GPRC:$rS, GPRC:$rB),
-let Defs = [CR0] in
-def ANDo : XForm_6<31,  28, 1, 0, 0, (ops GPRC:$rA, GPRC:$rS, GPRC:$rB),
-                   "and. $rA, $rS, $rB">;

We now have this:

+def AND  : XForm_6<31,  28, 0, 0, (ops GPRC:$rA, GPRC:$rS, GPRC:$rB),
                    "and $rA, $rS, $rB">;


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21225 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-11 15:01:39 +00:00
Duraid Madina
21478e55db hmm, should probably change addImm() to take 64-bit arguments one day anyway.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21224 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-11 07:16:39 +00:00
Duraid Madina
2d4c883118 rename addU64Imm() to addImm64()
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21223 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-11 07:14:41 +00:00
Nate Begeman
c7bd4827cd Add recording variants of ISD::AND and ISD::OR. This kills almost 1000
(1.5%) instructions in 186.crafty


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21222 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-11 06:34:10 +00:00
Duraid Madina
5ef2ec9929 assorted fixes:
* clean up immediates (we use 14, 22 and 64 bit immediates now. sane.)
  * fold r0/f0/f1 registers into comparisons against 0/0.0/1.0
  * fix nasty thinko - didn't use two-address form of conditional add
    for extending bools to integers, so occasionally there would be
    garbage in the result. it's amazing how often zeros are just
    sitting around in registers ;) - this should fix a bunch of tests.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21221 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-11 05:55:56 +00:00
Reid Spencer
c30088f961 Ensure that the arguments passed to sys::Program::ExecuteAndWait include
the program name as the first argument. Thanks go to Markus Oberhumer for
noticing this problem.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21220 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-11 05:48:04 +00:00
Jeff Cohen
01c5513683 Eliminate tabs
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21216 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-11 03:44:22 +00:00
Jeff Cohen
3bdfdfd10f Eliminate major source of VC++ "possible loss of data" warnings.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21215 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-11 03:38:28 +00:00
Nate Begeman
c7c16575fe Fix libcall code to not pass a NULL Chain to LowerCallTo
Fix libcall code to not crash or assert looking for an ADJCALLSTACKUP node
  when it is known that there is no ADJCALLSTACKDOWN to match.
Expand i64 multiply when ISD::MULHU is legal for the target.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21214 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-11 03:01:51 +00:00
Chris Lattner
2bb6f41282 Don't bother sign/zext_inreg'ing the result of an and operation if we know
the result does change as a result of the extend.

This improves codegen for Alpha on this testcase:

int %a(ushort* %i) {
        %tmp.1 = load ushort* %i
        %tmp.2 = cast ushort %tmp.1 to int
        %tmp.4 = and int %tmp.2, 1
        ret int %tmp.4
}

Generating:

a:
        ldgp $29, 0($27)
        ldwu $0,0($16)
        and $0,1,$0
        ret $31,($26),1

instead of:

a:
        ldgp $29, 0($27)
        ldwu $0,0($16)
        and $0,1,$0
        addl $0,0,$0
        ret $31,($26),1

btw, alpha really should switch to livein/outs for args :)


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21213 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-10 23:37:16 +00:00
Chris Lattner
01ff7216dd Teach legalize to deal with targets that don't support some SEXTLOAD/ZEXTLOADs
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21212 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-10 22:54:25 +00:00
Chris Lattner
bf9add4745 The first argument to ExecuteAndWait should be the program name, but pointed
out by Markus F.X.J. Oberhumer.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21211 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-10 20:59:38 +00:00
Chris Lattner
1a130b9a03 fix this testcase so the regex doesn't match the function name
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21210 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-10 20:45:35 +00:00
Chris Lattner
6841dec8c1 don't zextload fp values!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21209 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-10 17:40:35 +00:00
Duraid Madina
9696a90ee6 * store immediate values as int64_t, not int. come on, we should be happy
when there are immediates, let's not worry about the memory overhead of
this :)

* add addU64Imm(uint64_t val) to machineinstrbuilder

(seriously: this seems required to support 64-bit immediates cleanly. if it
_really_ gets on your nerves, feel free to pull it out ;) )

coming up next week: "all your floating point constants are belong to us"


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21208 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-10 09:18:55 +00:00
Nate Begeman
709c806a1e Fix another fixme: factor out the constant fp generation code.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21207 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-10 06:06:10 +00:00
Nate Begeman
c5b1cd22b5 Fix 64 bit argument loading that straddles the args in regs / args on stack
boundary.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21206 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-10 05:53:14 +00:00
Chris Lattner
232ee95a09 Until we have a dag combiner, promote using zextload's instead of extloads.
This gives the optimizer a bit of information about the top-part of the
value.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21205 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-10 04:33:47 +00:00
Chris Lattner
45278e37eb Fold zext_inreg(zextload), likewise for sext's
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21204 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-10 04:33:08 +00:00
Chris Lattner
edeecfcbd5 add a simple xform
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21203 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-10 04:04:49 +00:00