This is the mail archive of the binutils@sources.redhat.com mailing list for the binutils project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: Your patch breaks i386-pe gas


> The last time I went through this, everybody agreed that according to
> the documentation, such as it is, these symbol types should be marked
> as section -2, but the actual tools all marked them as section -1.
> 
> So until proven otherwise, I continue to believe that the old gas
> behaviour matches that of historical COFF assemblers.

You know your COFF, so if you say it is so I'll buy it.

But let's please get to it. What should happen now? Is a #define needed so a 
target can select between "historical" and "documented" behavior? And what if 
any changes should be made to the testsuite?

Mark


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]