Add test_b(Operand, immediate) to ia32 assembler, and use it where possible. ...

Add test_b(Operand, immediate) to ia32 assembler, and use it where possible.  Improve comparison to a constant one-character string.  Use CmpInstanceType in more places on ia32.  Add IsObjectJSObjectType and IsInstanceJSObjectType to ia32 macro assembler, using a single branch for a range test.

Committed: http://code.google.com/p/v8/source/detail?r=4795

[Lasse Reichstein, 2010-06-04 07:46:23]

LGTM

http://codereview.chromium.org/2586001/diff/1/4
File src/ia32/builtins-ia32.cc (right):

http://codereview.chromium.org/2586001/diff/1/4#newcode616
src/ia32/builtins-ia32.cc:616: __ CmpObjectType(ebx, FIRST_JS_OBJECT_TYPE, ecx);
This seems to be comparing the instance type with a range, and not using the
value after that.
We should be able to reduce that to a single unsigned comparison and jump. Could
we add a macro that did range-comparison, e.g.,
 CmpObjectTypeRange(ebx, FIRST_JS_OBJECT_TYPE, LAST_JS_OBJECT_TYPE, ecx)
where ecx is used as scratch (no guarantees on the value afterwards).

http://codereview.chromium.org/2586001/diff/1/5
File src/ia32/codegen-ia32.cc (right):

http://codereview.chromium.org/2586001/diff/1/5#newcode2735
src/ia32/codegen-ia32.cc:2735: // If the length is 0 then the subtraction gave
-1, setting
Comment is out of sync (there is no longer a subtraction).

http://codereview.chromium.org/2586001/diff/1/5#newcode6311
src/ia32/codegen-ia32.cc:6311: __ CmpInstanceType(map.reg(),
LAST_JS_OBJECT_TYPE);
Again a range comparison.

http://codereview.chromium.org/2586001/diff/1/5#newcode8325
src/ia32/codegen-ia32.cc:8325: __ CmpInstanceType(map.reg(),
LAST_JS_OBJECT_TYPE);
And again a range test. I think we use it often enough that it would be worth it
do make a special macro for it.

http://codereview.chromium.org/2586001/diff/1/5#newcode12849
src/ia32/codegen-ia32.cc:12849: ASSERT(kStringEncodingMask == kAsciiStringTag);
Also assert that they are zero.
Perhaps use STATIC_ASSERT, the variables should be compile time constants.

[William Hesse, 2010-06-04 10:42:08]

Comments addressed.

http://codereview.chromium.org/2586001/diff/1/4
File src/ia32/builtins-ia32.cc (right):

http://codereview.chromium.org/2586001/diff/1/4#newcode616
src/ia32/builtins-ia32.cc:616: __ CmpObjectType(ebx, FIRST_JS_OBJECT_TYPE, ecx);
On 2010/06/04 07:46:23, Lasse Reichstein wrote:
> This seems to be comparing the instance type with a range, and not using the
> value after that.
> We should be able to reduce that to a single unsigned comparison and jump.
Could
> we add a macro that did range-comparison, e.g.,
>  CmpObjectTypeRange(ebx, FIRST_JS_OBJECT_TYPE, LAST_JS_OBJECT_TYPE, ecx)
> where ecx is used as scratch (no guarantees on the value afterwards).
> 

Done.

http://codereview.chromium.org/2586001/diff/1/5
File src/ia32/codegen-ia32.cc (right):

http://codereview.chromium.org/2586001/diff/1/5#newcode2735
src/ia32/codegen-ia32.cc:2735: // If the length is 0 then the subtraction gave
-1, setting
On 2010/06/04 07:46:23, Lasse Reichstein wrote:
> Comment is out of sync (there is no longer a subtraction).

Done.