LGTM.

On Thu, Sep 25, 2008 at 6:18 PM,  <[EMAIL PROTECTED]> wrote:
> Reviewers: Kasper Lund,
>
> Message:
> I just set a breakpoint on the jump to next eliminator, and only found
> these two cases ever get it.  If it makes the control flow too confusing
> (the switch one is a big of a monster), I'm up for leaving it as is.
> Leaving the jump to next optimization is simple, but I want to get rid
> of the other stuff.
>
> Description:
> Remove two cases of needlessly jumping to the next instruction.  The
> jump eliminator will remove these cases, but it's easy enough to just do
> statically in the code, and not require the eliminator to do it at
> runtime.
>
> Please review this at http://codereview.chromium.org/4277
>
> Affected files:
>  M src/codegen-ia32.cc
>
>
> Index: src/codegen-ia32.cc
> diff --git a/src/codegen-ia32.cc b/src/codegen-ia32.cc
> index
> 1c6fc5b2c696007bd7426d562871a9cd896d1b28..4f237ea1a51132ead91b30ac4fb6525ff108cf49
> 100644
> --- a/src/codegen-ia32.cc
> +++ b/src/codegen-ia32.cc
> @@ -1127,7 +1127,7 @@ void ToBooleanStub::Generate(MacroAssembler* masm) {
>   __ sahf();
>   __ pop(eax);
>   __ j(zero, &false_result);
> -  __ jmp(&true_result);
> +  // Fall through to |true_result|.
>
>   // Return 1/0 for true/false in eax.
>   __ bind(&true_result);
> @@ -1492,7 +1492,7 @@ void GenericBinaryOpStub::Generate(MacroAssembler*
> masm) {
>       __ bind(&slow);
>       __ mov(eax, Operand(esp, 1 * kPointerSize));
>       __ mov(edx, Operand(esp, 2 * kPointerSize));
> -      __ jmp(&call_runtime);
> +      // Fall through to |call_runtime|.
>       break;
>     }
>     case Token::BIT_OR:
>
>
>

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

Reply via email to