Re: coreutils build on OSF/1 4.0d

[Matthew Woehlke]

Jim Meyering wrote:
> Matthew Woehlke <address@hidden> wrote:
> > Bruno Haible wrote:
> > > Building a coreutils snapshot (2007-02-21) from http//"meyering.net/cu/
> > > on OSF/1 4.0d with the native cc shows three problems:
> > > 2) "make check" failure of du/slink.
> > > The 'out' file produced by this test has size 0 for all sylimks. The same
> > > happens for /bin/du, so it's not a problem in the code of 'du'. Probably
> > > this type of file system needs to be excluded, like 'xfs' is excluded.
> > > The type of file system in this case is 'nfsv3'. The NFS server in this
> > > case is dedicated hardware designed for use as a file server. This
> > > workaround workd for me (causes the test to be skipped).
> > ...except I haven't had this failure on OSF 4.0g on nfsv3 file systems.
> > Wouldn't it be better to skip on nfsv3 only on OSF 4.0d or earlier?
>
> Yes, that would be better.  Propose a patch!

Ok. Instead of:

- if test -s tmp; then

...do:

+ if test -s tmp && test `uname -v` -le FAILING-OSF-VERSION; then

...replacing "FAILING-OSF-VERSION" with the most recent version of OSF 
known to break (i.e. 4.0d, unless someone has access to 4.0e). How's 
that? (I should be able to test 4.0f for you later, but it might not be 
today.)

--
Matthew
"Lost a planet, Obi Wan has? How embarrassing..."
  -- Yoda (Star Wars II: Attack of the Clones)