Re: [reiserfs-dev] Re: [PATCH] sparc32: wrong type of nlink_t

[Oleg Drokin <green@namesys.com>]

Hello!

   I am sorry but yor mailer corrupted the message or something like that
   so I cannot understand what do you mean.

   Ah, I see now. My latest version of a patch does a cast 
   this way (yes, I noticed that problem).
#define REISERFS_LINK_MAX (nlink_t)((((nlink_t) -1) > 0)?~0:((1u<<(sizeof(nlink_t)*8-1))-1))

Bye,
    Oleg
On Thu, Sep 05, 2002 at 11:09:12AM -0500, Dave Kleikamp wrote:
> On Thursday 05 September 2002 04:54, Oleg Drokin wrote:
> 
> > +/* Find maximal number, that nlink_t can hold. GCC is able to
> > calculate this +   value at compile time, so do not worry about extra
> > CPU overhead. */ +#define REISERFS_LINK_MAX ((((nlink_t) -1) >> 0)?~0:((1u<<(sizeof(nlink_t)*8-1))-1))
> 
> Shouldn't this be:
> 
> #define REISERFS_LINK_MAX ((((nlink_t) -1) >> 0)?(nlink_t) ~0:((1u<<(sizeof(nlink_t)*8-1))-1))
> 
> if nlink_t is u16, ~0 would still be 0xffffffff (assuming 32 bits)
> -- 
> David Kleikamp
> IBM Linux Technology Center
>