[plug] Does this appear to be a memory leak?

[James Devenish]

In message <4179DC2F.4090201 at tigris.org>
on Sat, Oct 23, 2004 at 12:21:03PM +0800, Tim White wrote:
> firefox windows open and I was running OO.org and FAH so I closed them

Out of curiosity: what is FAH?

> As all gxmms does is control xmms I think this is unreasonable. Also
> mozilla-Thunderbird (after re-opening) is using a far bit but I believe
> that is mainly in the shared librarys kept in memory.

(By the way: 'library' -> 'libraries', 'company' -> 'companies', etc.)

> Is it possible that there is a memory leek in gxmms? (It was running for
> a long time, > 43hrs of CPU time?)

It sounds like a memory leak. Also, if gxmms has CPU time of 43h on a
machine that has only been up for 4 days, I would rank it as an
"intensely active" programme. I assume, therefore, that it is has been
playing music continuously for four days. Actually, 43h of CPU time
suggests to me that it's been affecting the responsiveness of your
machine even without considering the memory problem. Perhaps you
have some kind of visualisation or other calculation-intensive feature
enabled?

>   PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  SWAP COMMAND
>  1873 tim       15   0  179m 3744  16m S  0.0  1.5  43:58.27 176m gxmms
>  5557 tim       15   0  139m  34m  36m S  0.0 14.0   0:08.00 104m mozilla-thunder

Indeed, assuming you have a 32-bit system, both of those sizes look
remarkably large. The other sizes (below) look reasonable. Although I
don't use either of the above products, I would agree with your initial
assessment that they are using more memory than can be reasonably
expected. Hopefully someone else can comment about Thunderbird.

>  1402 root       5 -10 68684  12m  49m S  4.0  5.2 192:30.04  54m XFree86
>  1809 tim       16   0 40380 9260  19m S  0.0  3.6   1:08.43  30m nautilus
> -- after kill
>  5652 tim       15   0 19036 8436  16m S  0.0  3.3   0:00.49  10m gxmms
>