general questions about sysroot in building a toolchain

Robert P. J. Day rpjday@mindspring.com
Thu Sep 29 13:44:00 GMT 2005


On Thu, 29 Sep 2005, Maxim Osipov wrote:

> By definition, --with-sysroot=dir option shall produce relocatable
> toolchain with if dir is inside of installation tree (or dir is
> omitted).
>
> In all toolchains I built, that worked fine. If it doesn't for you -
> you've found a bug in toolchain. But try to move toolchain first,
> probably you just miss something ;)

no, i didn't miss something, i was just trying to be overly tricky.
:-P  it seems to work fine if i copy the entire results directory,
which leads to two more questions to close off this issue (which is
admittedly starting to wander a bit far afield).

1) how do the cross-compiling gcc and ld handle being relocated in
terms of finding the new location of sys-root?  if i examine the
cross-compiling gcc, i can see embedded strings that refer to the
*full* pathname of the original location of sys-root. if that's the
first location that's checked and sys-root is no longer there, what
formula do these programs use to find the new location?

(just from the strings, i can see a couple values embedded in that
gcc:

--prefix=(snip)/sh3-unknown-linux-gnu
--with-sysroot=(snip)/sh3-unknown-linux-gnu/sh3-unknown-linux-gnu/sys-root

is that how it determines the relative location of sys-root if the
first attempt fails?)

2) what if i wanted to move the sys-root directory independently of
the toolchain?  is there a run-time option to handle that?

thanks for all the info so far.

rday

------
Want more information?  See the CrossGCC FAQ, http://www.objsw.com/CrossGCC/
Want to unsubscribe? Send a note to crossgcc-unsubscribe@sources.redhat.com



More information about the crossgcc mailing list