Re: [commit] fix for "info threads" printing multiple headers

[Michael Snyder <msnyder@vmware.com>]

Tom Tromey wrote:
>>>>>> "Pedro" == Pedro Alves <pedro@codesourcery.com> writes:
> 
> Pedro> Thanks!  A few comments below.  If you don't want to
> Pedro> fix these, I'll try to find a bit later on myself.
> 
> Pedro> - get_number_or_range mantains an internal state machine
> Pedro> using static variables.  I think that as long as you
> Pedro> always pass in the same list string, and the list spec
> Pedro> string is correctly formed, you're not hitting stale
> Pedro> state inside get_number_or_range.  It'd be nicer
> Pedro> if get_number_or_range (or a variant which get_number_or_range
> Pedro> would then be implemented on top of) took an additional
> Pedro> struct pointer that pointed to a struct that held
> Pedro> all the currently static state.
> 
> I almost did this when moving stuff to cli-utils, but decided against it
> on the basis of least change.
> 
> If Michael doesn't want to implement this, I'd be happy to.
> Just let me know.

Seems like it would only be useful if the function had to be thread safe.