invalidate: busy buffer

invalidate: busy buffer

[James Stevenson]

Hi

does anyone know what these mean ?

under

2.4.19-pre8

invalidate: busy buffer

in the dmesg output
got a bunch of these about 15 all together all of a sudden


thanks
    James

--------------------------
Mobile: +44 07779080838
http://www.stev.org
 11:00pm  up 6 days, 10:21,  7 users,  load average: 0.04, 0.15, 0.12

Re: invalidate: busy buffer

[Andrew Rodland]

On Mon, 17 Jun 2002 23:14:48 +0100
"James Stevenson" <mistral@stev.org> wrote:

Something tried to wipe out all of the caches for some device, but
something else was using it at the time. For example, if you try to run
parted on something mounted (and bypass/confuse its mountedness check)
you'll see this. Were you doing anything like that?

> Hi
> 
> does anyone know what these mean ?
> 
> under
> 
> 2.4.19-pre8
> 
> invalidate: busy buffer
> 
> in the dmesg output
> got a bunch of these about 15 all together all of a sudden
> 
> 
> thanks
>     James
> 
> --------------------------
> Mobile: +44 07779080838
> http://www.stev.org
>  11:00pm  up 6 days, 10:21,  7 users,  load average: 0.04, 0.15, 0.12
> 
> 
> 
> -
> To unsubscribe from this list: send the line "unsubscribe
> linux-kernel" in the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/
> 

Re: invalidate: busy buffer

[Andre Tomt]

On Tue, 2002-06-18 at 00:21, Andrew Rodland wrote:
> On Mon, 17 Jun 2002 23:14:48 +0100
> "James Stevenson" <mistral@stev.org> wrote:
> 
> Something tried to wipe out all of the caches for some device, but
> something else was using it at the time. For example, if you try to run
> parted on something mounted (and bypass/confuse its mountedness check)
> you'll see this. Were you doing anything like that?

or run for example hdparm -tT device to do a (imho pretty useless)
"benchmark" of io throughput.

-- 
André Tomt
andre@tomt.net

Re: invalidate: busy buffer

[Andrew Rodland]

On 18 Jun 2002 00:46:14 +0200
Andre Tomt <andre@tomt.net> wrote:

> On Tue, 2002-06-18 at 00:21, Andrew Rodland wrote:
> > On Mon, 17 Jun 2002 23:14:48 +0100
> > "James Stevenson" <mistral@stev.org> wrote:
> > 
> > Something tried to wipe out all of the caches for some device, but
> > something else was using it at the time. For example, if you try to
> > run parted on something mounted (and bypass/confuse its mountedness
> > check) you'll see this. Were you doing anything like that?
> 
> or run for example hdparm -tT device to do a (imho pretty useless)
> "benchmark" of io throughput.
> 
> -- 
> André Tomt
> andre@tomt.net
> 

the first half of hdparm -Tt is useful as a benchmark of how quick other
parts of the kernel are, though... changing the optimization level, or
the difference between vanilla and rmap, non-preempt and preempt, etc,
affects this a good bit. :)

Re: invalidate: busy buffer

[Andrew Rodland]

On Mon, 17 Jun 2002 23:27:35 +0100
"James Stevenson" <mistral@stev.org> wrote:

> 
> > Something tried to wipe out all of the caches for some device, but
> > something else was using it at the time. For example, if you try to
> > run parted on something mounted (and bypass/confuse its mountedness
> > check) you'll see this. Were you doing anything like that?
> 
> well i was running a badblocks over a 40GB ide disk i tend to check
> them every few months.
> apart from that everything else was running as non-root.
> but yes some things would have been using the disk there was a copy
> going on at the same time
> around 400MB
> 

badblocks is another one of those apps that will try to flush out as
much cache as it can; it's not much use trying the sectors repeatedly if
it's just going to get cached versions :)
> 
> > > Hi
> > >
> > > does anyone know what these mean ?
> > >
> > > under
> > >
> > > 2.4.19-pre8
> > >
> > > invalidate: busy buffer
> > >
> > > in the dmesg output
> > > got a bunch of these about 15 all together all of a sudden
> 
> 
>