[Mesa-dev] set.c vs. hash_table.c

[Brian Paul]

On 12/09/2012 07:41 PM, Eric Anholt wrote:
> Brian Paul<brianp at vmware.com>  writes:
>
>> Hi Jordan,
>>
>> set.c and hash_table.c look a LOT alike.  Could the functions in set.c
>> be implemented in terms of the hash_table.c functions?  It seems
>> little silly to have so much duplicated code.
>
> The idea of the set code is to save most of the memory in the case that
> all you're doing is storing a set of pointers.  We've got some debug
> code that also wants to use this in ir_validate.
>
> Whether the set code is important enough to have just for these two
> case, I'm not sure.  I wouldn't have imagined that people wrote code
> using many of these object types until I encountered minecraft, though
> :)

I have no issue with the set api/interface.  I was thinking the set 
functions could be implemented in set.h with some inline functions 
that just wrap the hash functions.

-Brian