Re: glib 1.2 patch.

> The only for an application to add a node to a hash table
> is g_hash_table_insert (). g_hash_table_insert () sets
> the key and value fields. This just seems like a waste
> of (tiny amounts of) CPU cycles to me. 

Morten uses Purify extensively to check our applications.  

Since the node is never actually released but put on the free list, as
far as purify is concerned the data kept on the hash table is valid,
and makes harder to debug leaks (ie, you might leak but purify wont
catch this since there is still a reference to it).

Best wishes,

[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]