linux-scsi.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: John Garry <john.garry@huawei.com>
To: Hannes Reinecke <hare@suse.de>,
	JBottomley@odin.com, martin.petersen@oracle.com
Cc: linuxarm@huawei.com, zhangfei.gao@linaro.org,
	xuwei5@hisilicon.com, john.garry2@mail.dcu.ie,
	linux-scsi@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH 2/6] hisi_sas: add hisi_sas_slot_abort()
Date: Tue, 16 Feb 2016 15:41:54 +0000	[thread overview]
Message-ID: <56C34342.9050605@huawei.com> (raw)
In-Reply-To: <56C33EC2.3020005@suse.de>

On 16/02/2016 15:22, Hannes Reinecke wrote:
> On 02/16/2016 01:22 PM, John Garry wrote:
>> Add a function to abort a slot (task) in the device
>> (if it is in the task set) and then cleanup and
>> complete the task.
>> The function is called from work queue context as
>> it cannot be called from the context where it is
>> triggered (interrupt).
>>
>> Signed-off-by: John Garry <john.garry@huawei.com>
>> ---
>>   drivers/scsi/hisi_sas/hisi_sas.h      |  1 +
>>   drivers/scsi/hisi_sas/hisi_sas_main.c | 43 +++++++++++++++++++++++++++++++++++
>>   2 files changed, 44 insertions(+)
>>
>> diff --git a/drivers/scsi/hisi_sas/hisi_sas.h b/drivers/scsi/hisi_sas/hisi_sas.h
>> index 02da7e4..a05ce71 100644
>> --- a/drivers/scsi/hisi_sas/hisi_sas.h
>> +++ b/drivers/scsi/hisi_sas/hisi_sas.h
>> @@ -120,6 +120,7 @@ struct hisi_sas_slot {
>>   	dma_addr_t command_table_dma;
>>   	struct hisi_sas_sge_page *sge_page;
>>   	dma_addr_t sge_page_dma;
>> +	struct work_struct abort_slot;
>>   };
>>
>>   struct hisi_sas_tmf_task {
>> diff --git a/drivers/scsi/hisi_sas/hisi_sas_main.c b/drivers/scsi/hisi_sas/hisi_sas_main.c
>> index c600f5e..65509eb 100644
>> --- a/drivers/scsi/hisi_sas/hisi_sas_main.c
>> +++ b/drivers/scsi/hisi_sas/hisi_sas_main.c
>> @@ -15,6 +15,9 @@
>>   #define DEV_IS_GONE(dev) \
>>   	((!dev) || (dev->dev_type == SAS_PHY_UNUSED))
>>
>> +static int hisi_sas_debug_issue_ssp_tmf(struct domain_device *device,
>> +				u8 *lun, struct hisi_sas_tmf_task *tmf);
>> +
>>   static struct hisi_hba *dev_to_hisi_hba(struct domain_device *device)
>>   {
>>   	return device->port->ha->lldd_ha;
>> @@ -113,6 +116,45 @@ static int hisi_sas_task_prep_ata(struct hisi_hba *hisi_hba,
>>   	return hisi_hba->hw->prep_stp(hisi_hba, slot);
>>   }
>>
>> +/*
>> + * This function will issue an abort TMF if a task is still in
>> + * the target. Then it will do the task complete cleanup and
>> + * callbacks.
>> + */
>> +static void hisi_sas_slot_abort(struct work_struct *work)
>> +{
>> +	struct hisi_sas_slot *abort_slot =
>> +		container_of(work, struct hisi_sas_slot, abort_slot);
>> +	struct sas_task *task = abort_slot->task;
>> +	struct hisi_hba *hisi_hba = dev_to_hisi_hba(task->dev);
>> +	struct scsi_cmnd *cmnd = task->uldd_task;
>> +	struct hisi_sas_tmf_task tmf_task;
>> +	struct domain_device *device = task->dev;
>> +	struct hisi_sas_device *sas_dev = device->lldd_dev;
>> +	struct scsi_lun lun;
>> +	int tag = abort_slot->idx, rc;
>> +
>> +	int_to_scsilun(cmnd->device->lun, &lun);
>> +	tmf_task.tmf = TMF_QUERY_TASK;
>> +	tmf_task.tag_of_task_to_be_managed = cpu_to_le16(tag);
>> +
>> +	rc = hisi_sas_debug_issue_ssp_tmf(task->dev, lun.scsi_lun, &tmf_task);
>> +
>> +	/* TMF_RESP_FUNC_SUCC means task is present in the task set */
>> +	if (rc != TMF_RESP_FUNC_SUCC)
>> +		goto out;
>> +	tmf_task.tmf = TMF_ABORT_TASK;
>> +	tmf_task.tag_of_task_to_be_managed = cpu_to_le16(tag);
>> +
>> +	rc = hisi_sas_debug_issue_ssp_tmf(task->dev, lun.scsi_lun, &tmf_task);
>> +out:
>> +	hisi_sas_slot_task_free(hisi_hba, task, abort_slot);
>> +	if (task->task_done)
>> +		task->task_done(task);
>> +	if (sas_dev && sas_dev->running_req)
>> +		sas_dev->running_req--;
>> +}
>> +
>>   static int hisi_sas_task_prep(struct sas_task *task, struct hisi_hba *hisi_hba,
>>   			      int is_tmf, struct hisi_sas_tmf_task *tmf,
>>   			      int *pass)
> Do you really need to query the task first?
> As per SAM a successful return from an ABORT TASK TMF has this meaning:
>
> A response of FUNCTION COMPLETE shall indicate that the task was
> aborted or was not in the task set.
>
> Ie it doesn't matter if the task was present or not.
>
> Cheers,
>
> Hannes
>

I think that it should be ok to the abort without first querying. I'll 
just test this to double-check. (This would make the first patch in the 
series superflous)

Cheers,
John



  reply	other threads:[~2016-02-16 15:43 UTC|newest]

Thread overview: 26+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-02-16 12:22 [PATCH 0/6] hisi_sas: add abort and retry feature John Garry
2016-02-16 12:22 ` [PATCH 1/6] hisi_sas: add TMF_RESP_FUNC_SUCC check John Garry
2016-02-16 15:20   ` Hannes Reinecke
2016-02-16 12:22 ` [PATCH 2/6] hisi_sas: add hisi_sas_slot_abort() John Garry
2016-02-16 15:22   ` Hannes Reinecke
2016-02-16 15:41     ` John Garry [this message]
2016-02-18  9:30       ` John Garry
2016-02-16 12:22 ` [PATCH 3/6] hisi_sas: use slot abort in v1 hw John Garry
2016-02-16 15:31   ` Hannes Reinecke
2016-02-16 16:13     ` John Garry
2016-02-18  7:16       ` Hannes Reinecke
2016-02-18  9:52         ` John Garry
2016-02-16 12:22 ` [PATCH 4/6] hisi_sas: use slot abort in v2 hw John Garry
2016-02-16 15:32   ` Hannes Reinecke
2016-02-16 16:58     ` John Garry
2016-02-16 12:22 ` [PATCH 5/6] hisi_sas: add hisi_sas_slave_configure() John Garry
2016-02-16 15:33   ` Hannes Reinecke
2016-02-16 16:56     ` John Garry
2016-02-18  7:40       ` Hannes Reinecke
2016-02-18 10:12         ` John Garry
2016-02-18 10:30           ` Hannes Reinecke
2016-02-18 10:57             ` John Garry
2016-02-19 10:46               ` John Garry
2016-02-19 14:31                 ` Hannes Reinecke
2016-02-22 10:02                   ` John Garry
2016-02-16 12:22 ` [PATCH 6/6] hisi_sas: update driver version to 1.3 John Garry

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=56C34342.9050605@huawei.com \
    --to=john.garry@huawei.com \
    --cc=JBottomley@odin.com \
    --cc=hare@suse.de \
    --cc=john.garry2@mail.dcu.ie \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-scsi@vger.kernel.org \
    --cc=linuxarm@huawei.com \
    --cc=martin.petersen@oracle.com \
    --cc=xuwei5@hisilicon.com \
    --cc=zhangfei.gao@linaro.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).