Re: crosstool-0.32 released

["Robert P. J. Day" <rpjday at mindspring dot com>]

On Sun, 1 May 2005, Dan Kegel wrote:

> Robert P. J. Day wrote:
> > but the entries for "sh3" still show failure across the board, despite
> > the fact i've built a chain for that.  thoughts?
>
> It's not that dire.
>
> gcc-4.0.0 fails to build sh toolchains simply
> because I haven't backported patches from glibc cvs
> to make it comptible with gcc-4.0.0.  It's not hard.
>
> gcc-3.4.3 + glibc-2.3.x seem to build
> ok sh3 toolchains:
>
> sh3-gcc-3.4.3-glibc-2.3.2.dat.txt:toolchain=PASS
> sh3-gcc-3.4.3-glibc-2.3.3.dat.txt:toolchain=PASS
> sh3-gcc-3.4.3-glibc-2.3.4.dat.txt:toolchain=PASS
> sh3-gcc-3.4.3-glibc-2.3.5.dat.txt:toolchain=PASS
>
> The build matrix shows kernel failing there,
> but that doesn't mean the toolchain is
> neccessarily bad.
>
> Also, gcc-3.3.5 + glibc-2.3.x seem to build ok
> *sh4* toolchains, but for some time now, they
> fail to build libstdc++.  See e.g.
> http://kegel.com/crosstool/crosstool-0.28-rc37/buildlogs/0.28/sh3-gcc-3.3.4-glibc-2.3.2.log.txt
> http://kegel.com/crosstool/crosstool-0.32/buildlogs/sh3-gcc-3.3.5-glibc-2.3.5.log.txt
> which both fail in the same spot.
> That might be worth looking into, but only if you're
> really determined to use gcc-3.3.x and can't move
> up to gcc-3.4.x.
>
> Which versions of gcc, glibc, and Linux work for you,
> and are you using sanititzed headers?

let me go back and do a fresh build and test with ct-0.32, and i'll
report back.

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