Hi Jesper, Jesper Wallin wrote on Mon, Jul 22, 2019 at 06:09:03PM +0200: > On Mon, Jul 22, 2019 at 03:23:16PM +0200, Ingo Schwarze wrote: >> 3. Jesper, including a patch according to the best of your >> understanding is always welcome. Even if it turns out to be a >> bad patch, because often even a bad patch helps to understand >> what the OP thinks the problem might be: the saying is, bad >> patches trigger good ones. > I'm sorry for late reply, I've been busy/offline the last few days.
No problem at all. > The lessons I take with me from this: > 1. Give myself more time to fully understand both the issue itself > and what the code actually does, before trying to fix it. > 2. Test, test with ktrace, test again and then run more tests. ;-) Sounds reasonable! Testing is important and often neglected. But make sure that doesn't cause bugs to not get reported at all because the process causes too much work or takes too long. :) When running out of time, in particular for bugs that seem important or when it is unclear if and when you might find more time to look at the issue, sending a preliminary report or a preliminary patch with a remark like "i'm trying to work on a patch" or "i only did rudimentary testing so far and ran out of time" might make sense to avoid needless delays. Yours, Ingo
