Date: Sat, 27 Dec 2008 23:12:29 -0700 (MST) From: "M. Warner Losh" <imp@bsdimp.com> To: alfred@freebsd.org Cc: svn-src-head@freebsd.org, svn-src-all@freebsd.org, scottl@samsco.org, src-committers@freebsd.org, mav@freebsd.org Subject: Re: svn commit: r186182 - head/sys/dev/ata Message-ID: <20081227.231229.139568214.imp@bsdimp.com> In-Reply-To: <20081228053530.GF18389@elvis.mu.org> References: <20081227003826.GV18389@elvis.mu.org> <72A0CF67-F521-4C4B-955D-A80A77CC1165@samsco.org> <20081228053530.GF18389@elvis.mu.org>
next in thread | previous in thread | raw e-mail | index | archive | help
In message: <20081228053530.GF18389@elvis.mu.org>
Alfred Perlstein <alfred@freebsd.org> writes:
: * Scott Long <scottl@samsco.org> [081226 18:34] wrote:
: >
: > On Dec 26, 2008, at 5:38 PM, Alfred Perlstein <alfred@freebsd.org>
: > wrote:
: > >
: > >Easy enough to stash a "once" varible in the generic device struct
: > >and warn when returning from an isr when INVARIANTS or something is
: > >turned on.
: > >
: > >Then you'd only get one warning per device once it happens.
: > >
: > >
: >
: > I prefer to do nothing. It's not unsafe or erroneous to access cfg
: > registers. But if some kind of message does get added, I insist that
: > it drop all pretenses and say, "you're too poor to run freebsd, come
: > back when you can afford better hardware.". :)
:
: Well that could be the result of a driver that has a "bug" that
: makes it access cfg space when it does not need to, as what seems
: to have just been fixed.
I think we need to have better tools to peer into what's going on
inside a driver (or other well defined areas of the kernel). Maybe
dtrace might be the right way to do this?
Warner
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20081227.231229.139568214.imp>