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 -~----------~----~----~----~------~----~------~--~---
