automake-patches
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Dependency tracking of pre processed Assembler files


From: Ralf Corsepius
Subject: Re: Dependency tracking of pre processed Assembler files
Date: Thu, 12 Oct 2006 06:49:30 +0200

On Wed, 2006-10-11 at 21:42 +0200, Ralf Wildenhues wrote:
> [ moving from automake to -patches, and reordering; this is
>   http://lists.gnu.org/archive/html/automake/2006-10/msg00031.html ]
> 
> * Ralf Corsepius wrote on Wed, Oct 11, 2006 at 11:09:30AM CEST:
> > > http://sources.redhat.com/cgi-bin/gnatsweb.pl?cmd=view%20audit-trail&database=automake&pr=492
> > Thanks - This seems to be fixed.
> 
> Erm, no, I don't think PR 492 is fixed.
I must be missing something. AFAIS with my real world code, automake-CVS
is correctly generating deps and *.o's w/ subdir-objects enabled.

> > On Wed, 2006-10-11 at 09:55 +0200, Ralf Wildenhues wrote:
> > > * Ralf Corsepius wrote on Wed, Oct 11, 2006 at 08:52:42AM CEST:
> > > > OK, I just was about to try automake-CVS + autoconf-2.60 and now am
> > > > facing an issue with a package using subdir-objects and *_SOURCES
> > > > containing *.S + *.c:
> [...]
> > > > /opt/gnu/share/automake-1.9c/am/depend2.am: am__fastdepCCAS does not
> > > > appear in AM_CONDITIONAL
> 
> > > Forgot to add AM_PROG_AS?
> > 
> > Yes, but ... with a simple testcase, I get this warning:
> [...]
> > With my real world test case, I don't get this warning. Probably
> > because I am AC_SUBST'ing CCAS and CCASCOMPILE.
> 
> Thanks for reporting this.  Actually, two bugs here:
> - there is no test for *.S,
I don't understand.

> - above warning is suboptimal because there is too little help
>   (the other warnings you posted had help, but were only present
>   in the test case).
> 
> Further note a small buglet in m4/as.m4:
> - without no-dependencies, we should not compute depmode.
Could you elaborate what this issue is you are trying to fix?
I must be missing something.

> OK to apply these two patches to fix them?  Or should I rather add
> a test case exposure to every possible warning?
> 
> Ralf, is the warning good enough in your real world test case now?
Thanks, I'll try to check, when I get around to testing it (Probably later 
today).

Ralf






reply via email to

[Prev in Thread] Current Thread [Next in Thread]