Merge (x & y) == 0 pattern to emit a single test instruction.

Merge (x & y) == 0 pattern to emit a single test instruction.

R=srdjan@google.com
BUG=

[srdjan, 2013-04-15 20:27:27]

LGTM (cc-d Regis FYI)

https://codereview.chromium.org/14251023/diff/1/runtime/vm/intermediate_langu...
File runtime/vm/intermediate_language.cc (right):

https://codereview.chromium.org/14251023/diff/1/runtime/vm/intermediate_langu...
runtime/vm/intermediate_language.cc:641: // Remove the current comparison's
input uses.
Would be probable more readable if CompareInstr* would be called instr instead
of comp?

https://codereview.chromium.org/14251023/diff/1/runtime/vm/intermediate_langu...
runtime/vm/intermediate_language.cc:1238: static bool
RecognizeTestPattern(Value* left, Value* right) {
Add comment what pattern is being recognized.

https://codereview.chromium.org/14251023/diff/1/runtime/vm/intermediate_langu...
File runtime/vm/intermediate_language.h (right):

https://codereview.chromium.org/14251023/diff/1/runtime/vm/intermediate_langu...
runtime/vm/intermediate_language.h:2593: class TestSmiInstr : public
ComparisonInstr {
Maybe add comment what this instruction is for.

https://codereview.chromium.org/14251023/diff/1/runtime/vm/intermediate_langu...
runtime/vm/intermediate_language.h:2637: (kind == Token::kNE));
Back to one line?

[regis, 2013-04-15 20:43:38]

LGTM

[Vyacheslav Egorov (Google), 2013-04-16 11:51:16]

https://codereview.chromium.org/14251023/diff/1/runtime/vm/intermediate_langu...
File runtime/vm/intermediate_language.cc (right):

https://codereview.chromium.org/14251023/diff/1/runtime/vm/intermediate_langu...
runtime/vm/intermediate_language.cc:641: // Remove the current comparison's
input uses.
On 2013/04/15 20:27:27, srdjan wrote:
> Would be probable more readable if CompareInstr* would be called instr instead
> of comp?

I think instr is somewhat confusing as well. I renamed it to new_comparison.

https://codereview.chromium.org/14251023/diff/1/runtime/vm/intermediate_langu...
runtime/vm/intermediate_language.cc:1238: static bool
RecognizeTestPattern(Value* left, Value* right) {
On 2013/04/15 20:27:27, srdjan wrote:
> Add comment what pattern is being recognized.

Done.

https://codereview.chromium.org/14251023/diff/1/runtime/vm/intermediate_langu...
File runtime/vm/intermediate_language.h (right):

https://codereview.chromium.org/14251023/diff/1/runtime/vm/intermediate_langu...
runtime/vm/intermediate_language.h:2593: class TestSmiInstr : public
ComparisonInstr {
On 2013/04/15 20:27:27, srdjan wrote:
> Maybe add comment what this instruction is for.

Done.

https://codereview.chromium.org/14251023/diff/1/runtime/vm/intermediate_langu...
runtime/vm/intermediate_language.h:2637: (kind == Token::kNE));
On 2013/04/15 20:27:27, srdjan wrote:
> Back to one line?

Done.

[srdjan, 2013-05-28 08:22:46]

What happened to this CL?

[Vyacheslav Egorov (Google), 2013-05-28 08:24:32]

Oh. I forgot to commit it. I think there was yet another strange regression on
Richards. 

I should rebase & remeasure.