LGTM.

We need to devise a test for GC-during-regexp-matching.

Does it lint?

Does it run on Mac?


http://codereview.chromium.org/13783/diff/1/4
File src/assembler-ia32.h (right):

http://codereview.chromium.org/13783/diff/1/4#newcode187
Line 187: inline explicit Immediate(Label* value);
I think this is rather confusing.  It's for emitting an offset to the
start of the code object, but it looks like it's for emitting the
address of the label itself.  Since constructors can't be renamed I'd
prefer a static method that returns the result of a private constructor.
  That way you can call it CodeObjectRelativeOffset(Label*).

http://codereview.chromium.org/13783/diff/1/4#newcode753
Line 753: inline void emit_label(Label* label);
The name of this is very generic.  It should have a longer name that
reflects what it does, eg. emit_code_relative_offset.

http://codereview.chromium.org/13783/diff/1/5
File src/assembler.h (right):

http://codereview.chromium.org/13783/diff/1/5#newcode178
Line 178: INTERNAL_REFERENCE,  // An address inside the same function.
I liked it better before.

http://codereview.chromium.org/13783/diff/1/7
File src/execution.h (right):

http://codereview.chromium.org/13783/diff/1/7#newcode124
Line 124: // stack overflow, the handle the interruption.
this sense make no

http://codereview.chromium.org/13783/diff/1/9
File src/regexp-macro-assembler-ia32.cc (right):

http://codereview.chromium.org/13783/diff/1/9#newcode198
Line 198: switch (mode_) {
I liked it better before with if instead of switch.  Since there are
only two possibilities you might as well tell the compiler and the
reader that.  One of them can optimize better and the other can
understand better.

http://codereview.chromium.org/13783/diff/1/9#newcode215
Line 215: switch (mode_) {
Ditto and more places below.

http://codereview.chromium.org/13783

--~--~---------~--~----~------------~-------~--~----~
v8-dev mailing list
[email protected]
http://groups.google.com/group/v8-dev
-~----------~----~----~----~------~----~------~--~---

Reply via email to