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]
Other format: [Raw text]

Re: Problems with gcc + exceptions + -pedantic + libiberty.h



> Why is it that __linux__ etc. means we disregard HAVE_DECL_BASENAME ?

The original problem is subdirectories that don't even test for the
decl.  In those cases, HAVE_DECL_BASENAME is not set at all.  Hence,
given an OS, we must try to guess what the right thing to do is.  I
didn't add that code myself; I would have defaulted to *no* decl, and
force projects to adopt some sort of detection.

This code also breaks when included by a C++ source :-(

However, now that you've pointed it out, the existing logic is flawed
in that a defined HAVE_DECL_* doesn't override the os-based guess.
Patches welcome ;)


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