From: Aaron Gaudio <agaudio@eng.mc.xerox.com>
To: Daniel Jacobowitz <drow@false.org>
Cc: gdb <gdb@sources.redhat.com>
Subject: Re: Configuring linker flags while cross-compiling gdb
Date: Mon, 13 Jun 2005 19:07:00 -0000 [thread overview]
Message-ID: <1118689574.27729.82.camel@rockhopper> (raw)
In-Reply-To: <20050613153007.GA7054@nevyn.them.org>
[-- Attachment #1: Type: text/plain, Size: 2446 bytes --]
On Mon, 2005-06-13 at 11:30 -0400, Daniel Jacobowitz wrote:
> On Mon, Jun 13, 2005 at 11:26:06AM -0400, Aaron Gaudio wrote:
> > Continuing with my adventure from last week, I am now trying to cross-
> > comile gdb to run on i386-pc-solaris2.10 and build on sparc-sun-
> > solaris2.10.
> >
> > I used the following configure line:
> >
> > $ bash_cv_have_mbstate_t=yes ./configure --build sparc-sun-solaris2.8 --
> > host i386-pc-solaris2.10
> >
> > This seems to get the correct compilers to use. However, when trying to
> > link, it is missing the correct rpath-link (where I have the x86 sysroot
> > to link system libraries), and it is also not linking in a required
> > library, libsocket (I get undefined symbols that manually adding -
> > lsocket to the link command fixes).
>
> These both sound like problems with your cross compiler, not problems
> with GDB.
In the case of libsocket, it is not a cross-compiler problem; you always
have to add -lsocket to the link libs when using things like socket(),
or linking libraries that do (such as X11 libs). The problem is that
when building a native gdb on solaris (with --target=i386-pc-
solaris2.10), configure detects that -lsocket is required and adds it to
the appopriate Makefile macros, but does not when building for x86
Solaris....
>
> No, that's for target libraries. Gdb is a host program. The fact taht
> build != host is irrelevant here; your "target" is i386-pc-solaris2.10,
> and so is your "host". LDFLAGS is the correct place to add flags for
> the host.
>
I've tried this, but it didn't seem to change anything. Part of the
confusion is that the sub-level configures don't get called until you
run a top-level make, and not all environment variables set (such as
LDFLAGS) at configure time get cascaded to the sub-level configure, and
furthermore I'm not sure the sub-level Makefile is handling merging
LDFLAGS from the environment and from the makefile (since I just want to
tell it to add "-lsocket", not manually figure out every single library
that needs to be linked).
Any further help is much appreciated.
--
Aaron Gaudio agaudio @ eng.mc.xerox.com 585-422-6876
While life is yours, live joyously;
None can escape Death's searching eye:
When once this frame of ours they burn,
How shall it e'er again return?
[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 197 bytes --]
next prev parent reply other threads:[~2005-06-13 19:07 UTC|newest]
Thread overview: 6+ messages / expand[flat|nested] mbox.gz Atom feed top
2005-06-13 15:27 Aaron Gaudio
2005-06-13 15:30 ` Daniel Jacobowitz
2005-06-13 19:07 ` Aaron Gaudio [this message]
2005-06-13 19:40 ` Aaron Gaudio
2005-06-13 19:42 ` Daniel Jacobowitz
2005-06-14 7:20 UDOGAN
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1118689574.27729.82.camel@rockhopper \
--to=agaudio@eng.mc.xerox.com \
--cc=drow@false.org \
--cc=gdb@sources.redhat.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox