R devel - Oct 2008 - Configure error when compiling on "sh" architecture

Amusingly enough, the SuperH (or sh for short) architecture isn't one of
the many I have handy, but someone seems to be porting Fedora to that
arch. They filed a bug against R, which isn't really an R bug, but is
more an inherited bug from autoconf. Specifically, the R configure
script invokes gcc(gfortran) which then passes "-little" to gas when
compiling for backward compatibility. This fails on sh, and they
proposed that it be added to the list of "flags to ignore" in
configure.

See:
https://bugzilla.redhat.com/show_bug.cgi?id=464055
https://bugzilla.redhat.com/show_bug.cgi?id=456886

I've attached the same patch that is in 464055 to this email. Hopefully,
at some point before the sun explodes, that fix will end up in autoconf,
and upstreams will pick it up when they run the fixed autoconf on their
tree.

Thanks,

~tom
-------------- next part --------------
A non-text attachment was scrubbed...
Name: R-2.7.2-filter_asoption.patch
Type: text/x-patch
Size: 505 bytes
Desc: not available
URL:
<https://stat.ethz.ch/pipermail/r-devel/attachments/20081016/3a7457f5/attachment.bin>

Tom "spot" Callaway wrote:
> Amusingly enough, the SuperH (or sh for short) architecture isn't one
> of the many I have handy, but someone seems to be porting Fedora to
> that arch. They filed a bug against R, which isn't really an R bug,
> but is more an inherited bug from autoconf. Specifically, the R
> configure script invokes gcc(gfortran) which then passes
"-little" to
> gas when compiling for backward compatibility. This fails on sh, and
> they proposed that it be added to the list of "flags to ignore"
in
> configure.
> 
> See: https://bugzilla.redhat.com/show_bug.cgi?id=464055 
> https://bugzilla.redhat.com/show_bug.cgi?id=456886
> 
> I've attached the same patch that is in 464055 to this email.
> Hopefully, at some point before the sun explodes, that fix will end
> up in autoconf, and upstreams will pick it up when they run the fixed
> autoconf on their tree.
Right. Until then, I think it needs to be fixed downstreams.

We can't fix configure upstreams because it gets rebuilt by autoconf all 
the time during our build procedures.

I suppose that we _could_ override /usr/share/auto 
conf/autoconf/fortran.m4 with a file of our own, but it
sounds like Bad Luck...

Anyways, as you probably know, we're deep in code freeze for 2.8.0 on 
Monday.

-p

> 
> Thanks,
> 
> ~tom
> 
> 
> ------------------------------------------------------------------------
> 
> 
> ______________________________________________ R-devel at r-project.org
> mailing list https://stat.ethz.ch/mailman/listinfo/r-devel

-- 
    O__  ---- Peter Dalgaard             ?ster Farimagsgade 5, Entr.B
   c/ /'_ --- Dept. of Biostatistics     PO Box 2099, 1014 Cph. K
  (*) \(*) -- University of Copenhagen   Denmark      Ph:  (+45) 35327918
~~~~~~~~~~ - (p.dalgaard at biostat.ku.dk)              FAX: (+45) 35327907