[Webkit-unassigned] [Bug 54807] Always explicitly use "signed" keyword to declare signed bitfields

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=54807





--- Comment #18 from Yong Li <yong.li.webkit at gmail.com>  2011-02-23 13:02:31 PST ---
(In reply to comment #17)
> (From update of attachment 83495 [details])
> View in context: https://bugs.webkit.org/attachment.cgi?id=83495&action=review
> 
> Looks good to me. Since I'm not normally working on either side of this code, deferring review to others.
> 
> > Source/JavaScriptCore/assembler/ARMAssembler.h:254
> > +            unsigned m_used : 1;
> 
> Could it be a bool?
> 
> > Source/JavaScriptCore/assembler/ARMv7Assembler.h:532
> > +        unsigned m_used : 1;
> 
> Ditto.
> 
> > Source/JavaScriptCore/assembler/MIPSAssembler.h:207
> > +        unsigned m_used : 1;
> 
> Ditto.
> 
> > Source/JavaScriptCore/assembler/X86Assembler.h:262
> >          bool m_used : 1;
> 
> This one is a bool!
> 
> > Source/JavaScriptCore/bytecode/StructureStubInfo.h:144
> > +        unsigned seen : 1;
> 
> Could this be a bool, too?

I also noticed this. I'm not sure which one we should use. Compilers could probably pack "bool" bitfields differently.

I just did a test with VC. 

struct A {
    unsigned x : 31;
    bool y : 1;
};

sizeof(A) = 8

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.