On Mon, Jul 26, 2021 at 10:48:57AM -0700, Darrick J. Wong wrote: > On Mon, Jul 26, 2021 at 04:07:11PM +1000, Dave Chinner wrote: > > From: Dave Chinner <dchinner@xxxxxxxxxx> > > > > We force iclogs in several places - we need them all to have the > > same cache flush semantics, so start by factoring out the iclog > > force into a common helper. > > > > Signed-off-by: Dave Chinner <dchinner@xxxxxxxxxx> > > Looks like a pretty simple hoist. > > I also wonder about the removal of the assertion in xlog_unmount_write, > but I /think/ that's ok because at that point we've turned off > everything else that could write the log and forced it, which means that > the log is clean, and therefore all iclogs ought to be in ACTIVE state? *nod* As it is, this assert goes away real soon in the "Kill XLOG_STATE_IOERROR" patch I wanted to get into this cycle, but progress on that is being held up by working through these g/482 failures. Hence I don't think that it's actually a big issue to lose assert now... Cheers, Dave -- Dave Chinner david@xxxxxxxxxxxxx
