-debug-only!

llvm-svn: 16868

llvm-svn: 16864

llvm-svn: 16863

llvm-svn: 16858

* Do not lead dangling dead constants prevent optimization
* Iterate global optimization while we're making progress.

These changes allow us to be more aggressive, handling cases like
GlobalOpt/iterate.llx without a problem (turning it into 'ret int 0').

llvm-svn: 16857

we know it is dead.

llvm-svn: 16855

254.gap.

llvm-svn: 16853

optimizations to trigger much more often.  This allows the elimination of
several dozen more global variables in Programs/External.  Note that we only
do this for non-constant globals: constant globals will already be optimized
out if the accesses to them permit it.

This implements Transforms/GlobalOpt/globalsra.llx

llvm-svn: 16842

This comes up when doing adds to bitfield elements.

llvm-svn: 16836

This triggers in cases of bitfield additions, opening opportunities for
future improvements.

llvm-svn: 16834

llvm-svn: 16814

llvm-svn: 16804

llvm-svn: 16803

* Instead of handling dead functions specially, just nuke them.
* Be more aggressive about cleaning up after constification, in
  particular, handle getelementptr instructions and constantexprs.
* Be a little bit more structured about how we process globals.

*** Delete globals that are only stored to, and never read.  These are
    clearly not useful, so they should go.  This implements deadglobal.llx

This last one triggers quite a few times.  In particular, 2208 in the
external tests, 1865 of which are in 252.eon.  This shrinks eon from
1995094 to 1732341 bytes of bytecode.

llvm-svn: 16802

simplifications of the resultant program to avoid making later passes
do it all.

This allows us to constify globals that just have the same constant that
they are initialized stored into them.

Suprisingly this comes up ALL of the freaking time, dozens of times in
SPEC, 30 times in vortex alone.

For example, on 256.bzip2, it allows us to constify these two globals:

%smallMode = internal global ubyte 0             ; <ubyte*> [#uses=8]
%verbosity = internal global int 0               ; <int*> [#uses=49]

Which (with later optimizations) results in the bytecode file shrinking
from 82286 to 69686 bytes!  Lets hear it for IPO :)

For the record, it's nuking lots of "if (verbosity > 2) { do lots of stuff }"
code.

llvm-svn: 16793

llvm-svn: 16769

an instruction if it can be hoisted to a common dominator of the block.
This implements: test/Regression/Transforms/TailDup/MergeTest.ll

llvm-svn: 16758

llvm-svn: 16622

llvm-svn: 16621

llvm-svn: 16620

that always prints when linking programs to libstdc++ :(

llvm-svn: 16603

llvm-svn: 16598

* SubOne/AddOne functions always return ConstantInt, declare them as such
* Pull code for handling setcc X, cst, where cst is at the end of the range,
  or cc is LE or GE up earlier in visitSetCondInst.  This reduces #iterations
  in some cases.
* Fold: (div X, C1) op C2 -> range check, implementing div.ll:test6 - test9.

llvm-svn: 16588

potentially fold more in one pass.

llvm-svn: 16583

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.

llvm-svn: 16582

llvm-svn: 16568

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).

llvm-svn: 16567

This implements or.ll:test1[89]

llvm-svn: 16561

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.

llvm-svn: 16559

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

llvm-svn: 16549

This latent bug was exposed by recent changes, and is tested as:
llvm/test/Regression/Transforms/InstCombine/2004-09-28-BadShiftAndSetCC.llx

llvm-svn: 16546

compile on windows. This patch was contributed by Paolo Invernizzi.

llvm-svn: 16539

compile under windows. Patch contributed by Paolo Invernizzi!

llvm-svn: 16534

where we folded (X & 254) -> X < 1 instead of X < 2.  These problems were
latent problems exposed by the latest patch.

llvm-svn: 16528

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

These combinations trigger 4 times in povray, 7x in gcc, 4x in gap, and 2x in bzip2.

llvm-svn: 16508

No functionality changes here.

llvm-svn: 16505

in perlbmk

llvm-svn: 16504

llvm-svn: 16499

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