Re: 2.5.70: CODA breaks boot

[Arnaldo Carvalho de Melo <acme@conectiva.com.br>]

Em Wed, May 28, 2003 at 10:00:06PM +0100, Riley Williams escreveu:
> Hi Andrew.
> 
>  >> ...it oopses in kmem_cache_create, called from release_console_sem and
>  >>  coda_init_inodecache.
> 
>  > You'll be needing this one.
>  > 
>  >  fs/coda/inode.c |    6 +++---
>  >  1 files changed, 3 insertions(+), 3 deletions(-)
>  > 
>  > diff -puN fs/coda/inode.c~coda-typo-fix fs/coda/inode.c
>  > --- 25/fs/coda/inode.c~coda-typo-fix	2003-05-27 22:27:11.000000000 -0700
>  > +++ 25-akpm/fs/coda/inode.c	2003-05-27 22:27:27.000000000 -0700
>  > @@ -69,9 +69,9 @@ static void init_once(void * foo, kmem_c
>  >  int coda_init_inodecache(void)
>  >  {
>  >  	coda_inode_cachep = kmem_cache_create("coda_inode_cache",
>  > -					     sizeof(struct coda_inode_info),
>  > +				sizeof(struct coda_inode_info),
>  > -					     0, SLAB_HWCACHE_ALIGN||SLAB_RECLAIM_ACCOUNT,
                                                                  ^^
                                                             logical OR
>  > +				0, SLAB_HWCACHE_ALIGN|SLAB_RECLAIM_ACCOUNT,
                                                     ^
                                                 bit or
>  > -					     init_once, NULL);
>  > +				init_once, NULL);
>  >  	if (coda_inode_cachep == NULL)
>  >  		return -ENOMEM;
>  >  	return 0;
> 
> That patch has me puzzled. Other than changing the white space, what actual
> change to the code does it make? I can't see any.

gotcha? 8)

- Arnaldo