Re: 2.4.21-pre2 stalls out when running unixbench

[Andrew Morton]

"Stephen C. Tweedie" wrote:
> 
> Hi,
> 
> On Sat, 2003-01-04 at 11:11, Andrew Morton wrote:
> 
> > This is because of differences in how sync() is handled between 2.4.20's
> > ext3 and 2.4.21-pre2's.
> >
> > 2.4.21-pre2:
> >
> >   sync() will start the commit, and will wait on it.  So you know that
> >   when it returns, everything which was dirty is now tight on disk.
> >
> > So yes, running a looping sync while someone else is writing stuff
> > will take much longer in 2.4.21-pre2, because that kernel actually
> > waits on the writeout.
> 
> Actually, I'm wondering if we should back that particular bit out.  For
> a user with a hundred mounted filesystems, syncing each one in order,
> sequentially, is going to suck (and we don't currently have a simple way
> in 2.4 to detect which syncs are on separate spindles and so can be
> parallelised.)
> 

Well personally I prefer slow-and-safe.  But we could make 2.4
do what 2.5 is doing - one pass through the superblocks to start
the syncs and a second pass to wait on them all.

This is fragile stuff though....