Zsh Mailing List Archive
Messages sorted by:
Re: PATCH: ldd and LD_DEBUG completions
- X-seq: zsh-workers 23624
- From: Oliver Kiddle <okiddle@xxxxxxxxxxx>
- To: Zsh workers <zsh-workers@xxxxxxxxxx>
- Subject: Re: PATCH: ldd and LD_DEBUG completions
- Date: Fri, 29 Jun 2007 17:08:42 +0100
- Domainkey-signature: a=rsa-sha1; q=dns; c=nofws; s=s1024; d=yahoo.co.uk; h=Received:X-YMail-OSG:Received:In-reply-to:From:References:To:Subject:Date:Message-ID; b=VngO6rUl/zM2W3Vk4aWjc5V/J30gw4YUgk0i434VZiXyqDENXRVMniLyS6b6WhlZMpPelKrypPzkp9tNRsW+F4KWQ+fB4fVHDqlaL2sLIlrFEX9OdrOlxQE00bh0srtF/3PUU3iqxaKM76xjNQvFiiQ3FIR1z/jsKV3rvNnmvU8= ;
- In-reply-to: <070629073552.ZM32277@xxxxxxxxxxxxxxxxxxxxxx>
- Mailing-list: contact zsh-workers-help@xxxxxxxxxx; run by ezmlm
- References: <15204.1183047166@thecus> <200706281621.l5SGLHKR015636@xxxxxxxxxxxxxx> <16925.1183113761@thecus> <20070629121135.GA25663@xxxxxxxxxxx> <070629073552.ZM32277@xxxxxxxxxxxxxxxxxxxxxx>
> On Jun 29, 8:11am, Clint Adams wrote:
> } Get that feature into GNU binutils and we have a deal.
> The zsh build system will still have to know how to deal with the lack
> of that option ... and apparently also know how to deal with having a
> compiler that appears to implement it but ignores it.
Not for a Debian package specific change it wouldn't.
-z ignore appears to be a linker option as opposed to being a compiler
option. I'm not sure how to detect the Sun linker given that I think
you can have gcc with the Sun linker. One option would be:
compile a dummy .so
compile a dummy binary linked against the .so explicitly with -z ignore
rm the .so
run the binary and see if you get an ld error.
It's actually now possible to get Sun studio for Linux. I just tried
compiling zsh with it. The only problem is that the compiler doesn't
handle the -rdynamic option. It is still using the GNU linker and
passing -export-dynamic instead of -rdynamic seems to do the trick.
That's line 2328 of configure.ac. What's the best way to handle that
from configure. We could check $ac_compiler_gnu but there's also the
Intel compiler to consider. Should we add a check for the Sun compiler
Messages sorted by: