Re: memory leak #4 & patch

[Derek Robert Price]

Kenneth Lorber wrote:

> This leak is in the diff code - which I understand from a recent thread
> we don't want to change if we can avoid it.  On the other hand this leak
> may have been caused by a previous change to this function (the memory
> leaked is allocated by cvs_temp_name()).  Apply or not as you think
> best.
>  

That looked good too and I don't think it was avoidable.  I put 
cvs_temp_name() in diff/diff3.c before I had thought too much about 
keeping the two as separate as possible.  I expect that a few changes 
will be unavoidable until the GNU diff project actually exports a difflib.

Looks like GNULIB has a mkstemp() function.  That might be a step in the 
right direction.

Anyhow, I commited your fix for the leak, for the nonce.

Derek

--
               *8^)

Email: derek@ximbiot.com

Get CVS support at <http://ximbiot.com>!
--
File not found. Should I fake it? (Y/N)