Messages in this thread | | | Date | Tue, 7 Sep 1999 06:06:37 +0200 (CEST) | From | Mike Galbraith <> | Subject | Re: v2.3.17pre1 - Patches, Complaints, Questions and Jubilations |
| |
On Tue, 7 Sep 1999, David Weinehall wrote:
> I've been looking through the IKD for other interesting things, and I > thought that we would probably also want the CONFIG_MEMLEAK option in the > kernel. What's your opinion on this one? When I get an answer, I'll > extract the applicable parts (either just the deadlock detection code, or > both of them), port them to v2.3, and then post a patch here. If that's ok > with everyone, that is...
IMHO, memleak has too many ifdefs and is needed far too infrequently to be included in the main tree. (howto remove suggestions welcome)
It also has some interesting and undocumented features that could cause recurring oops reports from (network) module users. It's a nice tool, but you have to use it carefully. (teaching the kernel to not unload modules which have active memory allocations would cost even more ugly ifdefs for no real benefit)
-Mike
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |