• Previous message: R. Timothy Edwards: "magic patch"
• In reply to: R. Timothy Edwards: "magic patch"
• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

right, i figured the default magic memory allocator does the correct
thing, but i wasn't sure.


On Tue, 14 May 2002, R. Timothy Edwards wrote:

> Dear Conrad,
>    The CDPROCESSED patch looks okay.  However, note that magic does a
> "one-buffered malloc" in which a call to freeMagic (or FREE) frees the
> *previous* alloc buffer, and then sets the buffer to the current free
> request.  This allows one item to be addressed after freeing.  It's
> admittedly a bit quirky, but it makes freeing of linked lists very
> simple.  Personally, I'd like to get rid of it everywhere, but going
> through the code and changing it everywhere will be a royal pain.
> Anyway, it means that your patch to the FREE call effectively does
> nothing.
> 					Regards,
> 					Tim
>

• Next message: Yonggang Cui: "layers in SCMOS"
• Previous message: R. Timothy Edwards: "magic patch"
• In reply to: R. Timothy Edwards: "magic patch"
• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]