This is the mail archive of the
binutils@sourceware.org
mailing list for the binutils project.
Re: Relative DT_RPATH/DT_RUNPATH vs. sysroot capable ld.
- From: Alan Modra <amodra at gmail dot com>
- To: Hans-Peter Nilsson <hp at bitrange dot com>, Nick Clifton <nickc at redhat dot com>
- Cc: binutils at sourceware dot org
- Date: Fri, 25 Sep 2015 07:51:14 +0930
- Subject: Re: Relative DT_RPATH/DT_RUNPATH vs. sysroot capable ld.
- Authentication-results: sourceware.org; auth=none
- References: <20150924133754 dot GA12440 at bubble dot grove dot modra dot org> <alpine dot BSF dot 2 dot 02 dot 1509241358230 dot 23964 at arjuna dot pair dot com>
On Thu, Sep 24, 2015 at 02:04:13PM -0400, Hans-Peter Nilsson wrote:
> On Thu, 24 Sep 2015, Alan Modra wrote:
> > This fixes two problems. First, the --sysroot option wasn't available
> > with a ld configured without --with-sysroot, a historical accident.
> > This led to people configuring binutils with --with-sysroot=/ in order
> > to enable sysroot support, which exposes a case where ld wrongly
> > prepends the sysroot to a relative path.
> >
> > PR ld/18992
> > * ldmain.c (main): Always enable --sysroot.
> > * emultempl/elf32.em (gld${EMULATION_NAME}_after_open): Don't
> > prepend sysroot to relative rpath/runpath.
>
> I though this would break some case in
> testsuite/ld-scripts/sysroot-prefix.exp but apparently not.
> (I re-checked with --with-sysroot=/tmp/somewherenotfound
> toolchains.)
>
> Still, it could do with some adjustment: parts of the table can
> be folded now that there's no "non-sysroot" (it's now the same
> as sysroot == "/" IIUC).
Yes, I noticed that too. I was hoping someone else would take care of
adjusting the sysroot tests. I think you know who I'm looking at. :)
Preapproved. Actually, you probably ought to be reviewing my
testsuite changes rather than the other way around. If you'd like to
be listed as a testsuite maintainer, you have my vote.
--
Alan Modra
Australia Development Lab, IBM