Commits · 02742710f36763ed6210f03012efc8262c177115 · Lorenzo Albano / LLVM bpEVL

Aug 09, 2005

SCEVAddExpr::get() of an empty list is invalid. · 02742710
Chris Lattner authored Aug 09, 2005
```
llvm-svn: 22724
```
02742710
This is now implemented · 23e3fb9e
Chris Lattner authored Aug 09, 2005
```
llvm-svn: 22723
```
23e3fb9e

Implement: LoopStrengthReduce/share_ivs.ll · a091ff17

Chris Lattner authored Aug 09, 2005

Two changes:
  * Only insert one PHI node for each stride.  Other values are live in
    values.  This cannot introduce higher register pressure than the
    previous approach, and can take advantage of reg+reg addressing modes.
  * Factor common base values out of uses before moving values from the
    base to the immediate fields.  This improves codegen by starting the
    stride-specific PHI node out at a common place for each IV use.

As an example, we used to generate this for a loop in swim:

.LBB_main_no_exit_2E_6_2E_i_no_exit_2E_7_2E_i_2:        ; no_exit.7.i
        lfd f0, 0(r8)
        stfd f0, 0(r3)
        lfd f0, 0(r6)
        stfd f0, 0(r7)
        lfd f0, 0(r2)
        stfd f0, 0(r5)
        addi r9, r9, 1
        addi r2, r2, 8
        addi r5, r5, 8
        addi r6, r6, 8
        addi r7, r7, 8
        addi r8, r8, 8
        addi r3, r3, 8
        cmpw cr0, r9, r4
        bgt .LBB_main_no_exit_2E_6_2E_i_no_exit_2E_7_2E_i_1

now we emit:

.LBB_main_no_exit_2E_6_2E_i_no_exit_2E_7_2E_i_2:        ; no_exit.7.i
        lfdx f0, r8, r2
        stfdx f0, r9, r2
        lfdx f0, r5, r2
        stfdx f0, r7, r2
        lfdx f0, r3, r2
        stfdx f0, r6, r2
        addi r10, r10, 1
        addi r2, r2, 8
        cmpw cr0, r10, r4
        bgt .LBB_main_no_exit_2E_6_2E_i_no_exit_2E_7_2E_i_1

As another more dramatic example, we used to emit this:

.LBB_main_L_90_no_exit_2E_0_2E_i16_no_exit_2E_1_2E_i19_2:       ; no_exit.1.i19
        lfd f0, 8(r21)
        lfd f4, 8(r3)
        lfd f5, 8(r27)
        lfd f6, 8(r22)
        lfd f7, 8(r5)
        lfd f8, 8(r6)
        lfd f9, 8(r30)
        lfd f10, 8(r11)
        lfd f11, 8(r12)
        fsub f10, f10, f11
        fadd f5, f4, f5
        fmul f5, f5, f1
        fadd f6, f6, f7
        fadd f6, f6, f8
        fadd f6, f6, f9
        fmadd f0, f5, f6, f0
        fnmsub f0, f10, f2, f0
        stfd f0, 8(r4)
        lfd f0, 8(r25)
        lfd f5, 8(r26)
        lfd f6, 8(r23)
        lfd f9, 8(r28)
        lfd f10, 8(r10)
        lfd f12, 8(r9)
        lfd f13, 8(r29)
        fsub f11, f13, f11
        fadd f4, f4, f5
        fmul f4, f4, f1
        fadd f5, f6, f9
        fadd f5, f5, f10
        fadd f5, f5, f12
        fnmsub f0, f4, f5, f0
        fnmsub f0, f11, f3, f0
        stfd f0, 8(r24)
        lfd f0, 8(r8)
        fsub f4, f7, f8
        fsub f5, f12, f10
        fnmsub f0, f5, f2, f0
        fnmsub f0, f4, f3, f0
        stfd f0, 8(r2)
        addi r20, r20, 1
        addi r2, r2, 8
        addi r8, r8, 8
        addi r10, r10, 8
        addi r12, r12, 8
        addi r6, r6, 8
        addi r29, r29, 8
        addi r28, r28, 8
        addi r26, r26, 8
        addi r25, r25, 8
        addi r24, r24, 8
        addi r5, r5, 8
        addi r23, r23, 8
        addi r22, r22, 8
        addi r3, r3, 8
        addi r9, r9, 8
        addi r11, r11, 8
        addi r30, r30, 8
        addi r27, r27, 8
        addi r21, r21, 8
        addi r4, r4, 8
        cmpw cr0, r20, r7
        bgt .LBB_main_L_90_no_exit_2E_0_2E_i16_no_exit_2E_1_2E_i19_1

we now emit:

.LBB_main_L_90_no_exit_2E_0_2E_i16_no_exit_2E_1_2E_i19_2:       ; no_exit.1.i19
        lfdx f0, r21, r20
        lfdx f4, r3, r20
        lfdx f5, r27, r20
        lfdx f6, r22, r20
        lfdx f7, r5, r20
        lfdx f8, r6, r20
        lfdx f9, r30, r20
        lfdx f10, r11, r20
        lfdx f11, r12, r20
        fsub f10, f10, f11
        fadd f5, f4, f5
        fmul f5, f5, f1
        fadd f6, f6, f7
        fadd f6, f6, f8
        fadd f6, f6, f9
        fmadd f0, f5, f6, f0
        fnmsub f0, f10, f2, f0
        stfdx f0, r4, r20
        lfdx f0, r25, r20
        lfdx f5, r26, r20
        lfdx f6, r23, r20
        lfdx f9, r28, r20
        lfdx f10, r10, r20
        lfdx f12, r9, r20
        lfdx f13, r29, r20
        fsub f11, f13, f11
        fadd f4, f4, f5
        fmul f4, f4, f1
        fadd f5, f6, f9
        fadd f5, f5, f10
        fadd f5, f5, f12
        fnmsub f0, f4, f5, f0
        fnmsub f0, f11, f3, f0
        stfdx f0, r24, r20
        lfdx f0, r8, r20
        fsub f4, f7, f8
        fsub f5, f12, f10
        fnmsub f0, f5, f2, f0
        fnmsub f0, f4, f3, f0
        stfdx f0, r2, r20
        addi r19, r19, 1
        addi r20, r20, 8
        cmpw cr0, r19, r7
        bgt .LBB_main_L_90_no_exit_2E_0_2E_i16_no_exit_2E_1_2E_i19_1

llvm-svn: 22722

a091ff17

Suck the base value out of the UsersToProcess vector into the BasedUser · 37c24cc9
Chris Lattner authored Aug 08, 2005
```
class to simplify the code.  Fuse two loops.

llvm-svn: 22721
```
37c24cc9

Split MoveLoopVariantsToImediateField out from MoveImmediateValues. The · 37ed895b

Chris Lattner authored Aug 08, 2005

first is a correctness thing, and the later is an optzn thing.  This also
is needed to support a future change.

llvm-svn: 22720

37ed895b

Factor out some common code, and be smarter about when to emit load hi/lo · c92787e1
Nate Begeman authored Aug 08, 2005
```
code sequences.

llvm-svn: 22719
```
c92787e1
A testcase I don't want to break in the future · 319292a6
Chris Lattner authored Aug 08, 2005
```
llvm-svn: 22718
```
319292a6

Aug 08, 2005
- Allow tools with "consume after" options (like lli) to take more positional · d09a9a78
  Chris Lattner authored Aug 08, 2005
```
opts than they take directly.  Thanks to John C for pointing this problem
out to me!

llvm-svn: 22717
```
  d09a9a78
- Remove getImmediateForOpcode, which is now dead. · 64068eb7
  Chris Lattner authored Aug 08, 2005
```
Patch by Jim Laskey.

llvm-svn: 22716
```
  64068eb7
- Add new immediate handling support for mul/div. · 25388199
  Chris Lattner authored Aug 08, 2005
```
Patch by Jim Laskey!

llvm-svn: 22715
```
  25388199
- Add support for OR/XOR/SUB immediates that are handled with the new immediate · 8e9dc319
  Chris Lattner authored Aug 08, 2005
```
way.  This allows ORI/ORIS pairs, for example.

llvm-svn: 22714
```
  8e9dc319
- Modify the ISD::AND opcode case to use new immediate constant predicates. · fd0fe76b
  Chris Lattner authored Aug 08, 2005
```
Includes wider support for rotate and mask cases.

Patch by Jim Laskey.

I've requested that Jim add new regression tests the newly handled cases.

llvm-svn: 22712
```
  fd0fe76b
- Modify the ISD::ADD opcode case to use new immediate constant predicates. · 81e0e3e9
  Chris Lattner authored Aug 08, 2005
```
Includes support for 32-bit constants using addi/addis.

Patch by Jim Laskey.

llvm-svn: 22711
```
  81e0e3e9
- Modify existing support functions to use new immediate constant predicates. · 4c54dae2
  Chris Lattner authored Aug 08, 2005
```
Patch by Jim Laskey

llvm-svn: 22710
```
  4c54dae2
- Add support predicates for future immediate constant changes. · 3cc070cc
  Chris Lattner authored Aug 08, 2005
```
Patch by Jim Laskey

llvm-svn: 22709
```
  3cc070cc
- Move IsRunOfOnes to a more logical place and rename to a proper predicate form · f2267ed5
  Chris Lattner authored Aug 08, 2005
```
(lowercase isXXX).

Patch by Jim Laskey.

llvm-svn: 22708
```
  f2267ed5
- Fix JIT encoding of ppc mfocrf instruction; the operands were reversed · 9a838678
  Nate Begeman authored Aug 08, 2005
```
llvm-svn: 22707
```
  9a838678
- Use the new 'moveBefore' method to simplify some code. Really, which is · 9f269e40
  Chris Lattner authored Aug 08, 2005
```
easier to understand?  :)

llvm-svn: 22706
```
  9f269e40
- Reject command lines that have too many positional arguments passed (e.g., · d380d841
  Chris Lattner authored Aug 08, 2005
```
'opt x y').  This fixes PR493.

Patch contributed by Owen Anderson!

llvm-svn: 22705
```
  d380d841
- Not all constants are legal immediates in load/store instructions. · 14203e85
  Chris Lattner authored Aug 08, 2005
```
llvm-svn: 22704
```
  14203e85
- new testcase, not implemented yet · ac86efb6
  Chris Lattner authored Aug 08, 2005
```
llvm-svn: 22703
```
  ac86efb6
- Implement LoopStrengthReduce/share_code_in_preheader.ll by having one · c70bbc0c
  Chris Lattner authored Aug 08, 2005
```
rewriter for all code inserted into the preheader, which is never flushed.

llvm-svn: 22702
```
  c70bbc0c
- It is better to not depend on CSE to share multiplies due to IV insertion. · 03b6275f
  Chris Lattner authored Aug 08, 2005
```
This testcase checks that only one mul is present in the output code, as it
should be.

llvm-svn: 22701
```
  03b6275f
- These are both implemented by a recent LSR patch · 7eecc6cb
  Chris Lattner authored Aug 08, 2005
```
llvm-svn: 22700
```
  7eecc6cb
- Implement a simple optimization for the termination condition of the loop. · 9bfa6f87
  Chris Lattner authored Aug 08, 2005
```
The termination condition actually wants to use the post-incremented value
of the loop, not a new indvar with an unusual base.

On PPC, for example, this allows us to compile
LoopStrengthReduce/exit_compare_live_range.ll to:

_foo:
        li r2, 0
.LBB_foo_1:     ; no_exit
        li r5, 0
        stw r5, 0(r3)
        addi r2, r2, 1
        cmpw cr0, r2, r4
        bne .LBB_foo_1  ; no_exit
        blr

instead of:

_foo:
        li r2, 1                ;; IV starts at 1, not 0
.LBB_foo_1:     ; no_exit
        li r5, 0
        stw r5, 0(r3)
        addi r5, r2, 1
        cmpw cr0, r2, r4
        or r2, r5, r5           ;; Reg-reg copy, extra live range
        bne .LBB_foo_1  ; no_exit
        blr

This implements LoopStrengthReduce/exit_compare_live_range.ll

llvm-svn: 22699
```
  9bfa6f87
- add new helper function · 24a0a43c
  Chris Lattner authored Aug 08, 2005
```
llvm-svn: 22698
```
  24a0a43c
- add a new helper method · 9dae86a5
  Chris Lattner authored Aug 08, 2005
```
llvm-svn: 22697
```
  9dae86a5
- Handle 64-bit constant exprs on 64-bit targets. · 88e2d2ee
  Chris Lattner authored Aug 08, 2005
```
llvm-svn: 22696
```
  88e2d2ee
Aug 07, 2005
- All stats are "Number of ..." · 579b20b7
  Chris Lattner authored Aug 07, 2005
```
llvm-svn: 22694
```
  579b20b7
- Add some simple folds that occur in bitfield cases. Fix a minor bug in · 2c14cf7b
  Chris Lattner authored Aug 07, 2005
```
isHighOnes, where it would consider 0 to have high ones.

llvm-svn: 22693
```
  2c14cf7b
- Fix typoCVS: ---------------------------------------------------------------------- · 134ebd08
  Chris Lattner authored Aug 07, 2005
```
llvm-svn: 22692
```
  134ebd08
- add a small simplification that can be exposed after promotion/expansion · 0c26a0b9
  Chris Lattner authored Aug 07, 2005
```
llvm-svn: 22691
```
  0c26a0b9
- * Use the new PHINode::hasConstantValue method to simplify some code · f4dd8c44
  Chris Lattner authored Aug 07, 2005
```
* Teach this code to move allocas out of the loop when tail call eliminating
  a call marked 'tail'.  This implements TailCallElim/move_alloca_for_tail_call.ll
* Do not perform this transformation if a call is marked 'tail' and if there
  are allocas that we cannot move out of the loop in #2.  Doing so would increase
  the stack usage of the function.  This implements fixes
  PR615 and TailCallElim/dont-tce-tail-marked-call.ll.

llvm-svn: 22690
```
  f4dd8c44
- New testcases for PR615 · d670e5c8
  Chris Lattner authored Aug 07, 2005
```
llvm-svn: 22689
```
  d670e5c8
Aug 06, 2005

Consolidate the GPOpt stuff to all use the Subtarget, instead of still · 983a415b

Chris Lattner authored Aug 05, 2005

depending on the command line option.  Now the command line option just
sets the subtarget as appropriate.  G5 opts will now default to on on
G5-enabled nightly testers among other machines.

llvm-svn: 22688

983a415b

Aug 05, 2005
- adjust to change in getSubtarget() api · 158acab9
  Chris Lattner authored Aug 05, 2005
```
llvm-svn: 22687
```
  158acab9
- Since getSubtarget() always provides a const Subtarget, dont' require the user · b5887385
  Chris Lattner authored Aug 05, 2005
```
to pass it in.  Also, since it always returns a non-null pointer, make it
return a reference instead for easier use.

llvm-svn: 22686
```
  b5887385
- Enable gp optimizations by default when available, even when a target triple · 431b8d80
  Chris Lattner authored Aug 05, 2005
```
is available, since the target triple doesn't specify whether to use gpopts
or not.

llvm-svn: 22685
```
  431b8d80
- teach TestRunner about prcontext · d01cf9cf
  Chris Lattner authored Aug 05, 2005
```
llvm-svn: 22683
```
  d01cf9cf
- two simple testcases loopreduce should handle but does not yet currently · 8a9f9f47
  Chris Lattner authored Aug 05, 2005
```
llvm-svn: 22682
```
  8a9f9f47