Date: Sun, 28 Oct 2001 20:03:21 +1030 From: Greg Lehey <grog@FreeBSD.org> To: Poul-Henning Kamp <phk@critter.freebsd.dk> Cc: mjacob@feral.com, Jonathan Lemon <jlemon@flugsvamp.com>, "Andrey A. Chernov" <ache@nagual.pp.ru>, cvs-committers@FreeBSD.org, cvs-all@FreeBSD.org Subject: Re: Causing known breakage (was: cvs commit: src/sys/kern kern_conf.c subr_disk.c) Message-ID: <20011028200321.B88146@monorchid.lemis.com> In-Reply-To: <31874.1004258544@critter.freebsd.dk>; from phk@critter.freebsd.dk on Sun, Oct 28, 2001 at 09:42:24AM %2B0100 References: <20011028190225.A88146@monorchid.lemis.com> <31874.1004258544@critter.freebsd.dk>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sunday, 28 October 2001 at 9:42:24 +0100, Poul-Henning Kamp wrote: > In message <20011028190225.A88146@monorchid.lemis.com>, Greg Lehey writes: >> On Sunday, 28 October 2001 at 9:28:44 +0100, Poul-Henning Kamp wrote: >>> In message <20011028085000.O8085@wantadilla.lemis.com>, Greg Lehey writes: >>> >>>> I can't be bothered looking for a rule that says you shouldn't commit >>>> anything which will knowingly cause breakage, but if it's not in the >>>> book, it should be. First ensure that cause of the breakage gets >>>> fixed, or fix it yourself. >>> >>> The kernel itself has been whining at making this mistake since >>> 11 sep 2000. >>> >>> That should be more than sufficient warning for anyone. >> >> So why haven't you fixed it? > > Because it is not my task to fix all broken drivers. Correct. So what did you mean with the statement "That should be more than sufficient warning for anyone."? > The message is "WARNING: driver mistake ..." for a very good reason. Correct. > For instance, I can see the bogus code (mentioned earlier) in asr.c > but since I have no hardware I cannot test my hypothesis that the > entire driverinit function can be removed. Fine. >> Seriously, though, you've got to admit that's rather irrelevant to the >> statement. > > Huh ? I made a point. Your answer was irrelevant. The point is, if you introduce this code, *you* produce panics which could be avoided. I've been trying to discuss how to avoid them, and as usual you've been trying to ignore the fact. Let me quote what I said above, because I'm sure it's already out the other ear: I can't be bothered looking for a rule that says you shouldn't commit anything which will knowingly cause breakage, but if it's not in the book, it should be. First ensure that cause of the breakage gets fixed, or fix it yourself. Looking back at that statement, it might be less than 100% clear. With "first", I mean "before you commit your panic code". Obviously it would be better to ensure that the person who did the breakage fixes it. But if he doesn't, that doesn't give you the right to deliberately break things when there's an alternative. Greg -- See complete headers for address and phone numbers To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe cvs-all" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20011028200321.B88146>