From: Tejun Heo <htejun@gmail.com>
To: albertl@mail.com
Cc: jgarzik@pobox.com, alan@lxorguk.ukuu.org.uk, axboe@suse.de,
forrest.zhao@intel.com, efalk@google.com,
linux-ide@vger.kernel.org
Subject: Re: [PATCH 08/11] libata-eh-fw: implement new EH scheduling from PIO
Date: Thu, 18 May 2006 20:49:30 +0900 [thread overview]
Message-ID: <446C5F4A.3030307@gmail.com> (raw)
In-Reply-To: <446C4FA8.5020105@tw.ibm.com>
Albert Lee wrote:
> Tejun Heo wrote:
>> PIO executes without holding host_set lock, so it cannot be
>> synchronized using the same mechanism as interrupt driven execution.
>> port_task framework makes sure that EH is not entered until PIO task
>> is flushed, so PIO task can be sure the qc in progress won't go away
>> underneath it. One thing it cannot be sure of is whether the qc has
>> already been scheduled for EH by another exception condition while
>> host_set lock was released.
>>
>> This patch makes ata_poll_qc-complete() handle such conditions
>> properly and make it freeze the port if HSM violation is detected
>> during PIO execution.
>>
>> Signed-off-by: Tejun Heo <htejun@gmail.com>
>>
>> ---
>>
>> drivers/scsi/libata-core.c | 23 +++++++++++++++++++----
>> 1 files changed, 19 insertions(+), 4 deletions(-)
>>
>> 52240d1d208615e461c41a5c297f521a8a892f0d
>> diff --git a/drivers/scsi/libata-core.c b/drivers/scsi/libata-core.c
>> index 436ff5b..86d6a7d 100644
>> --- a/drivers/scsi/libata-core.c
>> +++ b/drivers/scsi/libata-core.c
>> @@ -3429,16 +3429,31 @@ skip_map:
>> * LOCKING:
>> * None. (grabs host lock)
>> */
>> -
>> void ata_poll_qc_complete(struct ata_queued_cmd *qc)
>> {
>> struct ata_port *ap = qc->ap;
>> unsigned long flags;
>>
>> spin_lock_irqsave(&ap->host_set->lock, flags);
>> - ap->flags &= ~ATA_FLAG_NOINTR;
>> - ata_irq_on(ap);
>> - ata_qc_complete(qc);
>> +
>> + if (ap->ops->error_handler) {
>> + /* EH might have kicked in while host_set lock is released */
>> + qc = ata_qc_from_tag(ap, qc->tag);
>> + if (qc) {
>> + if (!(qc->err_mask & AC_ERR_HSM)) {
>> + ap->flags &= ~ATA_FLAG_NOINTR;
>> + ata_irq_on(ap);
>> + ata_qc_complete(qc);
>> + } else
>> + ata_port_freeze(ap);
>
> I don't quite understand here. For the AC_ERR_HSM error, how does the
> EH get triggered after ata_port_freeze()? Do we wait until time out?
>
/**
* ata_port_freeze - abort & freeze port
* @ap: ATA port to freeze
*
* Abort and freeze @ap.
^^^^^^
*
* LOCKING:
* spin_lock_irqsave(host_set lock)
*
* RETURNS:
* Number of aborted commands.
*/
int ata_port_freeze(struct ata_port *ap)
{
int nr_aborted;
WARN_ON(!ap->ops->error_handler);
nr_aborted = ata_port_abort(ap);
^^^^^^^^^^^^^^^^^^^
__ata_port_freeze(ap);
return nr_aborted;
}
--
tejun
next prev parent reply other threads:[~2006-05-18 11:49 UTC|newest]
Thread overview: 20+ messages / expand[flat|nested] mbox.gz Atom feed top
2006-05-11 12:27 [PATCHSET 02/11] new EH framework, take 3 Tejun Heo
2006-05-11 12:27 ` [PATCH 03/11] libata-eh-fw: use special reserved tag and qc for internal commands Tejun Heo
2006-05-11 12:27 ` [PATCH 01/11] libata-eh-fw: add flags and operations for new EH Tejun Heo
2006-05-11 12:27 ` [PATCH 04/11] libata-eh-fw: update ata_qc_from_tag() to enforce normal/EH qc ownership Tejun Heo
2006-05-11 12:27 ` [PATCH 02/11] libata-eh-fw: clear SError in ata_std_postreset() Tejun Heo
2006-05-11 12:27 ` [PATCH 11/11] libata-eh-fw: update SCSI command completion path for new EH Tejun Heo
2006-05-11 12:27 ` [PATCH 09/11] libata-eh-fw: update ata_scsi_error() " Tejun Heo
2006-05-11 12:27 ` [PATCH 05/11] libata-eh-fw: implement new EH scheduling via error completion Tejun Heo
2006-05-11 12:27 ` [PATCH 08/11] libata-eh-fw: implement new EH scheduling from PIO Tejun Heo
2006-05-18 10:42 ` Albert Lee
2006-05-18 11:49 ` Tejun Heo [this message]
2006-05-19 7:31 ` Albert Lee
2006-05-11 12:27 ` [PATCH 10/11] libata-eh-fw: update ata_exec_internal() for new EH Tejun Heo
2006-05-11 12:27 ` [PATCH 07/11] libata-eh-fw: implement freeze/thaw Tejun Heo
2006-05-16 10:15 ` Albert Lee
2006-05-16 10:30 ` Tejun Heo
2006-05-16 10:43 ` Albert Lee
2006-05-16 11:17 ` Albert Lee
2006-05-11 12:27 ` [PATCH 06/11] libata-eh-fw: implement ata_port_schedule_eh() and ata_port_abort() Tejun Heo
2006-05-13 22:01 ` [PATCHSET 02/11] new EH framework, take 3 Jeff Garzik
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=446C5F4A.3030307@gmail.com \
--to=htejun@gmail.com \
--cc=alan@lxorguk.ukuu.org.uk \
--cc=albertl@mail.com \
--cc=axboe@suse.de \
--cc=efalk@google.com \
--cc=forrest.zhao@intel.com \
--cc=jgarzik@pobox.com \
--cc=linux-ide@vger.kernel.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;
as well as URLs for NNTP newsgroup(s).