Date: Wed, 10 Jan 2018 13:32:29 -0700 From: Warner Losh <imp@bsdimp.com> To: Mark Johnston <markj@freebsd.org> Cc: Warner Losh <imp@freebsd.org>, src-committers <src-committers@freebsd.org>, svn-src-all@freebsd.org, svn-src-head@freebsd.org Subject: Re: svn commit: r326964 - head/sys/cam Message-ID: <CANCZdfrNfkqkSTC7gsm-0dMstQYH1vayJEnx8hW297Cn6OiP4g@mail.gmail.com> In-Reply-To: <20180110202442.GC78524@raichu> References: <201712190413.vBJ4DMY9090982@repo.freebsd.org> <20180110202442.GC78524@raichu>
next in thread | previous in thread | raw e-mail | index | archive | help
On Wed, Jan 10, 2018 at 1:24 PM, Mark Johnston <markj@freebsd.org> wrote: > On Tue, Dec 19, 2017 at 04:13:22AM +0000, Warner Losh wrote: > > Author: imp > > Date: Tue Dec 19 04:13:22 2017 > > New Revision: 326964 > > URL: https://svnweb.freebsd.org/changeset/base/326964 > > > > Log: > > When doing a dump, the scheduler is normally not running, so this > > changed worked to capture dumps for me. However, the test for > > SCHEDULER_STOPPED() isn't right. We can also call the dump routine > > from ddb, in which case the scheduler is still running. This leads to > > an assertion panic that we're sleeping when we shouldn't. Instead, use > > the proper test for dumping or not. This brings us in line with other > > places that do special things while we're doing polled I/O like this. > > > > Noticed by: pho@ > > Differential Revision: https://reviews.freebsd.org/D13531 > > > > Modified: > > head/sys/cam/cam_periph.c > > > > Modified: head/sys/cam/cam_periph.c > > ============================================================ > ================== > > --- head/sys/cam/cam_periph.c Tue Dec 19 04:06:07 2017 (r326963) > > +++ head/sys/cam/cam_periph.c Tue Dec 19 04:13:22 2017 (r326964) > > @@ -38,6 +38,7 @@ __FBSDID("$FreeBSD$"); > > #include <sys/malloc.h> > > #include <sys/kernel.h> > > #include <sys/bio.h> > > +#include <sys/conf.h> > > #include <sys/lock.h> > > #include <sys/mutex.h> > > #include <sys/buf.h> > > @@ -1158,7 +1159,7 @@ cam_periph_runccb(union ccb *ccb, > > struct bintime *starttime; > > struct bintime ltime; > > int error; > > - bool sched_stopped; > > + bool must_poll; > > struct mtx *periph_mtx; > > struct cam_periph *periph; > > uint32_t timeout = 1; > > @@ -1182,7 +1183,13 @@ cam_periph_runccb(union ccb *ccb, > > devstat_start_transaction(ds, starttime); > > } > > > > - sched_stopped = SCHEDULER_STOPPED(); > > + /* > > + * We must poll the I/O while we're dumping. The scheduler is > normally > > + * stopped for dumping, except when we call doadump from ddb. > While the > > + * scheduler is running in this case, we still need to poll the > I/O to > > + * avoid sleeping waiting for the ccb to complete. > > + */ > > + must_poll = dumping; > > Hmm, unfortunately this introduces a new problem: after a panic we call > adashutdown() as part of the shutdown_post_sync event. This occurs > before we call doadump(), so dumping == 0 and we don't poll for the > spindown command's completion. But since we've panicked, the scheduler > is stopped, and so we just hang. > So "dumping || SCHEDULER_STOPPED()" then is needed? Warner
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CANCZdfrNfkqkSTC7gsm-0dMstQYH1vayJEnx8hW297Cn6OiP4g>