This is the mail archive of the libc-alpha@sourceware.org mailing list for the glibc project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: vfprintf typing problem


From: Roland McGrath <roland@hack.frob.com>
Date: Mon,  2 Apr 2012 11:22:59 -0700 (PDT)

>> ERANGE for individual object overflows, EOVERFLOW for
>> overall total length overflow.  I'm just following the
>> existing convention and this is codified in the testsuite.
> 
> I don't see any rationale behind the existing ERANGE behavior.  The cases
> that provoke it are ones that by the rest of the specification would result
> in producing too much output and hence are the EOVERFLOW case.  So
> detecting that early by another means is certainly fine, but the errno
> value ought to be EOVERFLOW uniformly, despite the present behavior.

Ok, I'll adjust all the ERANGE cases to be EOVERFLOW, thanks.


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