Date: Fri, 21 Jun 2002 02:00:22 +0200 From: Matthias Andree <matthias.andree@gmx.de> To: freebsd-stable@freebsd.org Subject: Re: ATA tags bug fix committed to -releng4 Message-ID: <m3znxpv6uh.fsf@merlin.emma.line.org> In-Reply-To: <030c01c21831$1ac8b320$0900a8c0@max> ("John Nielsen"'s message of "Thu, 20 Jun 2002 02:04:29 -0600") References: <20020620085158.K54942-100000@alexander.diva.nl> <030c01c21831$1ac8b320$0900a8c0@max>
next in thread | previous in thread | raw e-mail | index | archive | help
"John Nielsen" <stable@jnielsen.net> writes:
> I agree that merging minor bugfixes into a release branch is not good
> practice. But if the ata fixes turn out to be not-so-minor, I wouldn't m=
ind
> seeing them go in in this case.
IMHO, it qualifies as critical patch.
Scenario: assume you're running 4.5-RELEASE-p<MUMBLE>, the "4.5 critical
fixes" branch. Assume your /boot/loader.conf.local lists
hw.ata.tags=3D"1".
If you did now update to the "4.6 critical fixes" branch without this
patch, your ATA stuff would break and let your Queued ("tags") enabled
drives stall, and finally fall back to PIO after some retries, while 4.5
got it right.
The trivial patch (it swaps two lines) reportedly fixes the regression
that came with the MFC, and IMNSHO, PIO (without tags, tags require DMA)
versus UDMA *does* matter.
BTW, Thanks to S=F8ren for fixing this.
--=20
Matthias Andree
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-stable" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?m3znxpv6uh.fsf>