mknod bug (in uClibc)

Rob Landley rob at landley.net
Fri Nov 14 22:13:00 UTC 2008


On Friday 14 November 2008 14:35:49 Khem Raj wrote:
> I posted a patch back in march which got lost in the pile.
>
> http://www.uclibc.org/lists/uclibc/2008-March/019155.html

I think that's overcomplicated.  There's no need for long long here (which 
remains expensive on 32 bit systems) merely for error checking.

The cast will truncate to 32 bits, and if they feed in more than 32 bits 
because the loony gnu people decided that attaching 4 billion devices to a 
single system couldn't possibly be enough, and we truncate that silently 
rather than erroring out, I'm not particularly worried about it.

Keep in mind the current one's been silently truncating to 8 bits plus 8 bits 
all along, and it took this long to bother doing anything about it.  A more 
explicit error message when you ask Linux to do something Linux can't do 
doesn't seem worth the extra size...

Rob



More information about the uClibc mailing list