From: Andre Hedrick <andre@linux-ide.org>
To: Jens Axboe <axboe@suse.de>
Cc: Marcelo Tosatti <marcelo@conectiva.com.br>,
Karsten Weiss <knweiss@gmx.de>,
lkml <linux-kernel@vger.kernel.org>
Subject: Re: Linux 2.4.19-pre3
Date: Wed, 13 Mar 2002 01:00:25 -0800 (PST) [thread overview]
Message-ID: <Pine.LNX.4.10.10203130056210.18254-100000@master.linux-ide.org> (raw)
In-Reply-To: <20020313080946.GC15877@suse.de>
Jens,
Please try again because that is not the real problem.
All you have shown is that we disagree on the method of page walking
between BLOCK v/s IOCTL. This is very minor and I agreed that it is
reasonable to map the IOCTL buffer in to BH or BIO so this is a net zero
of negative point.
How about attempting to describe the differences between the atomic and
what is violated by who and where. I will help you later if you get
stuck.
Regards,
Andre Hedrick
On Wed, 13 Mar 2002, Jens Axboe wrote:
> On Tue, Mar 12 2002, Marcelo Tosatti wrote:
> > So, Jens, could you please explain the problem in the interrupt handlers
> > in detail ?
>
> Ok... It affects all the pio handlers in ide-taskfile.c,
> multi-write/read as well. The address for pio transfers is calculated
> like so:
>
> va = rq->buffer + (rq->nr_sectors - rq->current_nr_sectors) * SECTOR_SIZE;
>
> which is wrong for two reasons. First of all, rq->buffer cannot be
> indexed for the entire nr_sectors range -- it's per definition only the
> first segment in the request, and can as such only be indexed within the
> first current_nr_sectors number of sectors. The above can be grossly out
> of range... Second, nr_sectors and current_nr_sectors are indexing two
> different things -- the former indexes the entire request (all segments)
> while the latter indexes only the first segments. So
>
> foo = rq->nr_sectors - rq->current_nr_sectors;
>
> makes no sense _at all_ and can only be wrong.
>
> So why does 2.4.19-pre3 work for pio at all? For the same reason that
> Andre never found this problem in 2.5 either: the taskfile interrupt
> handlers are _never_ used in pio mode. In 2.5 it was by accident, and
> when the merge happened they did indeed get used. It ate disks, very
> quickly. Take a look at drivers/ide/ide-disk.c, line 64:
>
> #ifdef CONFIG_IDE_TASKFILE_IO
> # undef __TASKFILE__IO /* define __TASKFILE__IO */
> #else /* CONFIG_IDE_TASKFILE_IO */
> # undef __TASKFILE__IO
> #endif /* CONFIG_IDE_TASKFILE_IO */
>
> It's a mess... This really should have been fixed prior to 2.4
> inclusion. Oh well.
>
> --
> Jens Axboe
>
next prev parent reply other threads:[~2002-03-13 9:01 UTC|newest]
Thread overview: 45+ messages / expand[flat|nested] mbox.gz Atom feed top
2002-03-11 21:08 Linux 2.4.19-pre3 Marcelo Tosatti
2002-03-11 22:40 ` Eyal Lebedinsky
2002-03-11 23:03 ` Alan Cox
2002-03-11 22:55 ` Stephan von Krawczynski
2002-03-12 1:50 ` Mike Fedyk
2002-03-12 11:06 ` Stephan von Krawczynski
2002-03-11 23:04 ` Eyal Lebedinsky
2002-03-11 23:14 ` Keith Owens
2002-03-12 15:49 ` DevilKin
2002-03-21 12:06 ` [patch] Don't offer CONFIG_INDYDOG on non-ip22 machines Adrian Bunk
2002-03-22 21:13 ` Marcelo Tosatti
2002-03-11 23:21 ` Linux 2.4.19-pre3 Eyal Lebedinsky
2002-03-12 1:40 ` Takeo Saito
2002-03-12 0:04 ` J.A. Magallon
2002-03-12 0:37 ` Ken Brownfield
2002-03-19 2:41 ` I/O APIC fixed in 2.4.19-pre3 & 2.5.6 (was Re: Linux 2.4.19-pre3) Ken Brownfield
2002-03-19 16:22 ` Bill Davidsen
2002-03-20 0:34 ` Ken Brownfield
2002-03-20 4:36 ` Bill Davidsen
2002-03-12 10:35 ` Linux 2.4.19-pre3 Gerd Knorr
2002-03-12 17:24 ` Trond Myklebust
2002-03-12 17:47 ` Gerd Knorr
2002-03-12 13:01 ` Karsten Weiss
2002-03-12 13:17 ` Jens Axboe
2002-03-12 13:37 ` Alan Cox
2002-03-12 13:46 ` Jens Axboe
2002-03-12 19:02 ` Marcelo Tosatti
2002-03-13 8:09 ` Jens Axboe
2002-03-13 9:00 ` Andre Hedrick [this message]
2002-03-13 9:27 ` Jens Axboe
2002-03-13 9:33 ` Andre Hedrick
2002-03-13 17:32 ` Andre Hedrick
2002-03-13 17:40 ` Jens Axboe
2002-03-13 18:52 ` Andre Hedrick
2002-03-14 4:41 ` Marcelo Tosatti
2002-03-14 6:28 ` Andre Hedrick
2002-03-14 3:32 ` Michal Jaegermann
2002-03-12 14:47 ` J.A. Magallon
2002-03-12 13:31 ` Jurgen Philippaerts
2002-03-12 15:43 ` Geert Uytterhoeven
2002-03-12 21:59 ` Paul Mackerras
2002-03-12 0:51 Mikael Pettersson
2002-03-12 5:13 ` skidley
2002-03-12 21:00 ` Dave Jones
[not found] <fa.npg7nmv.si46bq@ifi.uio.no>
2002-03-12 7:26 ` junkio
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=Pine.LNX.4.10.10203130056210.18254-100000@master.linux-ide.org \
--to=andre@linux-ide.org \
--cc=axboe@suse.de \
--cc=knweiss@gmx.de \
--cc=linux-kernel@vger.kernel.org \
--cc=marcelo@conectiva.com.br \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).