Removal of removing fixincludes.

Greg Schafer gschafer at zip.com.au
Thu May 15 17:48:21 PDT 2003


On Thu, May 15, 2003 at 05:33:56PM -0600, Gerard Beekmans wrote:
> On May 15, 2003 04:39 pm, Greg Schafer wrote:
> 
> Ignore my previous reply, I think I'm misunderstanding the problem, I'm 
> getting rather confused by all the emails on this thread.
> 
> Here is how I understand the issue:
> 
> gcc-pass1 must be done with fixed includes in order to properly compile Glibc.

No, gcc-pass1 (gcc-3.3) must be done with fixed includes to allow it to even
build (when coming from a glibc-2.2.5 host)

In other words, we should not patch it to disable fixincludes.

> gcc-pass2 must be compiled with the fixedincludes stuff enabled during the 
> build, but the fixed headers should not be installed since everything after 
> gcc-pass2 is done using the toolchain in /stage1 so a 'broken' host glibc 
> will not play a role anymore.

No, we can safely disable fixincludes for gcc-pass2 and we should in fact do
so. (but we're already doing this anyway).

In other words, here we do patch it.

> I've seen a suggestion to install gcc-pass2 without the fixinclude path and 
> after 'make install' remove the installed fixed headers.
> 
> The other suggestion is patching Makefile after 'make' but before 'make 
> install' to suppress the copying of the fixed headers.

Yeah, those are the suggestiion I'm dubious about. Time will tell I s'pose
:-)

Greg
-- 
Unsubscribe: send email to listar at linuxfromscratch.org
and put 'unsubscribe lfs-dev' in the subject header of the message



More information about the lfs-dev mailing list