tech-toolchain archive
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]
Re: [PATCH alt] Install gcc's unwind.h instead of libexecinfo's
On Thu, Jan 16, 2020 at 06:35:17PM +0100, Michał Górny wrote:
> On Thu, 2020-01-16 at 13:59 +0100, Joerg Sonnenberger wrote:
> > On Thu, Jan 16, 2020 at 08:39:04AM +0100, Michał Górny wrote:
> > > The prototypes in libexecinfo's unwind.h do not match those commonly
> > > used (e.g. by gcc, clang, GNU libunwind, LLVM libunwind...), causing
> > > C++ programs to fail to build on type mismatches (e.g. compiler-rt,
> > > libc++abi). Rather than providing our own header, reuse the one
> > > included in gcc.
> >
> > Just replace /usr/include/unwind.h for ${HAVE_LIBGCC_EH} == "yes". It
> > describes a system interface, whether it is part of libgcc_s or libc.
> >
>
> I'm sorry but I don't understand what you mean. Isn't that exactly what
> I did? I've replaced /usr/include/unwind.h coming from libexecinto with
> the one coming from gcc, for HAVE_LIBGCC_EH=yes. For =no, the existing
> behavior of copying it from libunwind remains.
No, it still moves the header to /usr/include/gcc-*.
> I needed to additionally include it in /usr/include/gcc-8/unwind.h to
> address gcc being unable to find it while building distribution.
Is GCC including it with ""? That is special casing the directory
search. An explicit -I=/usr/include is ugly, but would help with that.
Joerg
Home |
Main Index |
Thread Index |
Old Index