GCC 3.3.1 vastly different failure #'s (ch 5 lfs 5.0, 2nd pass)

zoltan fmzoltan at mindspring.com
Sat Mar 13 12:57:08 PST 2004

Ken Moffat wrote:
> On Fri, 12 Mar 2004, zoltan wrote:
>>Howdy folks,
>>So are these what would be considered "vastly" different results for the
>>  second pass of GCC 3.3.1 in LFS-5.0 in chapter 5?  Seems to me they
>>are...Am I the only one who's seen this, or is there perhaps something more?
>>BTW, my host system is an LFS 4.1 system, which runs great.  I've
>>followed the LFS-5.0 instructions pretty much to the letter, except I've
>>  replaced --prefix=/tools with --prefix=/my-own-tools globally (right
>>from the get-go, including SPEC file changes, etc. (consistently), so it
>>shouldn't be an issue...
>>Any ideas?
>>                 === g++ Summary ===
>># of expected passes            7016
>># of unexpected failures        1166
>>                 === libstdc++-v3 Summary ===
>># of expected passes            258
>># of unexpected failures        195
> Something like this used to be seen in maybe the back end of last year.
> I don't remember the details, but I think I was one of the people bitten
> by it.  If I'm right, it's either down to using a patch which enforces
> /tools (we were moving from /stage1 to /tools), or you missed changing a
> prefix somewhere.  If I'm wrong, try searching the list archives for gcc
> unexpected failures.
> Ken

Thank you...that was it.  The spec file patch has hard-coded references 
to /tools, which wasn't appropriate for me.  All is well now...


More information about the lfs-support mailing list