In general, Apache projects tend to have many contributors and few committers. The usual process is that if you have a patch, you post it as an attachment, mark it as available, and get someone (not necessarily a committer) to review it. Once it's been approved, one of the committers will get to it and check it in, usually within a day or two.
To generate a patch, you say "svn diff [files you want in the patch] > patchfilenanme.patch". Then post just the patch, not the whole modified file. This makes it clearer what's being changed and prevents extraneous and unexpected changes from creeping in.
One other process note. If at all possible, your patch should include new or modified unit tests to demonstrate that it's functioning correctly. That's probably neither feasible nor useful here, so don't worry about it on this issue.
This may sound like a fair bit of trouble for a small and simple change. But that's the way Hadoop, Chukwa, and I believe all apache projects are run. It makes it much, MUCH easier to keep the codebase stable and working correctly.