On Fri, Apr 26, 2013 at 05:34:45PM +0300, Dan Carpenter wrote: > > I hate these "mostly NUL terminated" strings. MAX_DESC_LEN > is 100 characters so it's very rarely going to be filled all the > way. The special case where ->nd_ps_desc isn't terminated will > never be tested. > Yes, I think the original code picked the size of MAX_DESC_LEN randomly. The longest value I've seen for nd_ps_desc has been around 40. > > It seems likely to me that changing this struct is going to break > things. But maybe we could change the other side and say that the > desc can be 99 characters and a NUL maximum. Probably no one will > notice a change like that. > That seems like a great solution. The one place where nd_ps_desc is used now assumes it to work that way anyhow. -- Bill _______________________________________________ devel mailing list devel@xxxxxxxxxxxxxxxxxxxxxx http://driverdev.linuxdriverproject.org/mailman/listinfo/devel
