glibc-2.8-20080929 build fails in Chapter 5

Jeremy Huntwork jhuntwork at
Wed Oct 15 14:45:18 PDT 2008

Greg Schafer wrote:
> Definitely triggered by the addition of `--disabled-shared' to GCC Pass 1.
> If you're going to use that switch then must also use `--disable-libssp'.


But, wait, why only now? I wasn't seeing this issue with glibc 2.7 and 
gcc-4.2.3. What is currently in the jh branch works with those versions, 
uses --disable-shared and does not use --disable-libssp.

Looks like I have a fair amount of research to do.

> Sorry guys. I've abandoned the old build method you're currently using.
> It's getting harder to support and the Next Gen method is soooo much
> nicer IMNSHO. LFS should really look at getting on board.

Yeah, we probably will. Or, at least, I'll probably look at using it in 
my branch if and when I update it, and maybe then it will trickle back 
to trunk. Unless, of course, someone decides to head down that path 
earlier. It does seem to be the right combination of cross methods and 
native building.

I hit upon something you wrote in your reference build the other day 
which I hadn't noticed before:

I must say, it definitely seems the right approach, and I think it's the 
sort of thing LFS could use more of.


More information about the lfs-dev mailing list