Re: [Kgdb-bugreport] kdb: core for kgdb back end (1 of 2)

[Dan C. <dan...@or...>]

On Mon, Mar 02, 2015 at 10:02:27AM +0000, Daniel Thompson wrote:
> On Thu, 2015-02-26 at 14:16 +0300, Dan Carpenter wrote:
> >    353                                          break;
> >    354                                  kdb_printf("%s ", p_tmp);
> > 
> > Presumably we print the same thing over and over "count" times.
> 
> kallsyms_symbol_next() internally maintains a static iterator  (the
> second argument is used to reset-or-not the iterator) so it will not
> yield the same result when called.
> 
> 
> > 
> >    355                                  *(p_tmp + len) = '\0';
> >                                                   ^^^
> > "len" was calculated back on line 337.  We don't need to update it?  I'm
> > not sure I understand this line.
> 
> Whilst kallsyms_symbol_next() maintains a static iterator to record the
> position of the search within the symbol table it does *not* record the
> search string that it is looking for. Line 355 shortens the search
> string (which currently contains a search *result* rather than the
> search target) back to its original value ready to do the next search.
> 

Ah.  You're, of course, correct.  I missed the static.  Thanks for
taking the time on this.

regards,
dan carpenter