From: Bartlomiej Zolnierkiewicz <B.Zolnierkiewicz@elka.pw.edu.pl>
To: Mark Haverkamp <markh@osdl.org>
Cc: linux-kernel <linux-kernel@vger.kernel.org>,
Cliff White <cliffw@osdl.org>
Subject: Re: ide taskfile and cdrom hang
Date: Sun, 1 Feb 2004 21:48:20 +0100 [thread overview]
Message-ID: <200402012148.20590.bzolnier@elka.pw.edu.pl> (raw)
In-Reply-To: <1075502193.26342.61.camel@markh1.pdx.osdl.net>
Hi,
On Friday 30 of January 2004 23:36, Mark Haverkamp wrote:
> We have some test machines here at OSDL that have a problem with the ide
> cdrom driver hanging when we cat the /proc/ide/hda/identify file. After
> 30 seconds the console displays: "hda: lost interrupt" which reoccurs
> every 30 seconds forever. We noticed it on 2.6.2-rc2-mm1 but It looks
> like this has been a problem for a while. Our test machines just changed
> their configuration to use make defconfig. I found that if
> CONFIG_IDE_TASKFILE_IO is N then the hang doesn't occur. Is this a
> common problem or are there just certain drives that won't work with
> taskfile i/o enabled? I've included my .config, lspci as attachments.
> The cdrom model is CD-224E.
Does 'hdparm -I /dev/hda" work? If so can you try this patch?
It reverts taskfile code to use old (non-taskfile) order of accessing
port registers for ATAPI identify command. This is a good starting point.
If CONFIG_IDE_TASKFILE_IO is N code just ignores BUSY status of a drive
and reads some random data. I would like to have this problem solved
at least instead of adding the same workaround to CONFIG_IDE_TASKFILE_IO.
Cheers,
--bart
linux-2.6.2-rc2-root/drivers/ide/ide-taskfile.c | 14 ++++++++++++++
1 files changed, 14 insertions(+)
diff -puN drivers/ide/ide-taskfile.c~ide_tf_identify_debug drivers/ide/ide-taskfile.c
--- linux-2.6.2-rc2/drivers/ide/ide-taskfile.c~ide_tf_identify_debug 2004-02-01 21:06:58.494873136 +0100
+++ linux-2.6.2-rc2-root/drivers/ide/ide-taskfile.c 2004-02-01 21:15:06.268720240 +0100
@@ -142,6 +142,20 @@ ide_startstop_t do_rw_taskfile (ide_driv
void debug_taskfile(drive, task);
#endif /* CONFIG_IDE_TASK_IOCTL_DEBUG */
+ if (taskfile->command == WIN_PIDENTIFY) {
+ hwif->OUTB(taskfile->feature, IDE_FEATURE_REG);
+
+ if (IDE_CONTROL_REG)
+ hwif->OUTB(drive->ctl, IDE_CONTROL_REG);
+ SELECT_MASK(drive, 0);
+
+ hwif->OUTB(taskfile->sector_count, IDE_NSECTOR_REG);
+
+ ide_execute_command(drive, taskfile->command, task->handler,
+ WAIT_WORSTCASE, NULL);
+ return ide_started;
+ }
+
/* ALL Command Block Executions SHALL clear nIEN, unless otherwise */
if (IDE_CONTROL_REG) {
/* clear nIEN */
_
next prev parent reply other threads:[~2004-02-01 20:44 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
2004-01-30 22:36 Mark Haverkamp
2004-02-01 20:48 ` Bartlomiej Zolnierkiewicz [this message]
2004-02-02 18:46 ` Mark Haverkamp
2004-02-02 19:45 ` Bartlomiej Zolnierkiewicz
2004-02-02 20:45 ` Mark Haverkamp
2004-02-02 21:35 ` Bartlomiej Zolnierkiewicz
2004-02-02 21:44 ` Mark Haverkamp
2004-02-02 23:37 ` Bartlomiej Zolnierkiewicz
2004-02-02 23:52 ` Mark Haverkamp
2004-02-03 0:13 ` Bartlomiej Zolnierkiewicz
2004-02-03 0:16 ` Mark Haverkamp
2004-02-03 1:09 ` Bartlomiej Zolnierkiewicz
[not found] <1075502193.26342.61.camel@markh1.pdx.osdl.net.suse.lists.linux.kernel>
2004-01-31 0:46 ` Andi Kleen
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=200402012148.20590.bzolnier@elka.pw.edu.pl \
--to=b.zolnierkiewicz@elka.pw.edu.pl \
--cc=cliffw@osdl.org \
--cc=linux-kernel@vger.kernel.org \
--cc=markh@osdl.org \
/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
all inboxes | Powered by JetHome®