From: John Garry <john.garry@huawei.com>
To: Damien Le Moal <damien.lemoal@opensource.wdc.com>,
<jejb@linux.ibm.com>, <martin.petersen@oracle.com>,
<hare@suse.de>, <bvanassche@acm.org>, <hch@lst.de>,
<ming.lei@redhat.com>, <niklas.cassel@wdc.com>
Cc: <axboe@kernel.dk>, <jinpu.wang@cloud.ionos.com>,
<linux-block@vger.kernel.org>, <linux-kernel@vger.kernel.org>,
<linux-ide@vger.kernel.org>, <linux-scsi@vger.kernel.org>,
<linuxarm@huawei.com>
Subject: Re: [PATCH RFC v3 1/7] ata: libata-scsi: Add ata_scsi_queue_internal()
Date: Thu, 27 Oct 2022 11:45:08 +0100 [thread overview]
Message-ID: <8280ec1c-c274-ba8d-b533-3316865fe5c1@huawei.com> (raw)
In-Reply-To: <8ed42281-4400-a0ed-92f1-c57b9de726a4@opensource.wdc.com>
On 27/10/2022 02:42, Damien Le Moal wrote:
> On 10/25/22 19:32, John Garry wrote:
>> Add a function to handle queued ATA internal SCSI cmnds - does much the
>> same as ata_exec_internal_sg() does (which will be fixed up later to
>> actually queue internal cmnds through this function).
>>
>> Signed-off-by: John Garry <john.garry@huawei.com>
>> ---
>> drivers/ata/libata-sata.c | 3 +++
>> drivers/ata/libata-scsi.c | 43 +++++++++++++++++++++++++++++++++++++++
>> drivers/ata/libata.h | 3 ++-
>> include/linux/libata.h | 6 ++++++
>> 4 files changed, 54 insertions(+), 1 deletion(-)
>>
>> diff --git a/drivers/ata/libata-sata.c b/drivers/ata/libata-sata.c
>> index b6806d41a8c5..e8b828c56542 100644
>> --- a/drivers/ata/libata-sata.c
>> +++ b/drivers/ata/libata-sata.c
>> @@ -1258,6 +1258,9 @@ int ata_sas_queuecmd(struct scsi_cmnd *cmd, struct ata_port *ap)
>> {
>> int rc = 0;
>>
>> + if (blk_mq_is_reserved_rq(scsi_cmd_to_rq(cmd)))
>> + return ata_scsi_queue_internal(cmd, ap->link.device);
>> +
>> if (likely(ata_dev_enabled(ap->link.device)))
>> rc = __ata_scsi_queuecmd(cmd, ap->link.device);
>> else {
>> diff --git a/drivers/ata/libata-scsi.c b/drivers/ata/libata-scsi.c
>> index 476e0ef4bd29..30d7c90b0c35 100644
>> --- a/drivers/ata/libata-scsi.c
>> +++ b/drivers/ata/libata-scsi.c
>> @@ -3965,6 +3965,49 @@ static inline ata_xlat_func_t ata_get_xlat_func(struct ata_device *dev, u8 cmd)
>> return NULL;
>> }
>>
>> +unsigned int ata_scsi_queue_internal(struct scsi_cmnd *scmd,
>> + struct ata_device *dev)
>> +{
>> + struct ata_link *link = dev->link;
>> + struct ata_port *ap = link->ap;
>> + struct ata_queued_cmd *qc;
>> +
>> + /* no internal command while frozen */
>> + if (ap->pflags & ATA_PFLAG_FROZEN)
>> + goto did_err;
>> +
>> + /* initialize internal qc */
>> + qc = __ata_qc_from_tag(ap, ATA_TAG_INTERNAL);
>> + link->preempted_tag = link->active_tag;
>> + link->preempted_sactive = link->sactive;
>> + ap->preempted_qc_active = ap->qc_active;
>> + ap->preempted_nr_active_links = ap->nr_active_links;
>> + link->active_tag = ATA_TAG_POISON;
>> + link->sactive = 0;
>> + ap->qc_active = 0;
>> + ap->nr_active_links = 0;
>> +
>> + if (qc->dma_dir != DMA_NONE) {
>> + int n_elem;
>> +
>> + n_elem = 1;
>> + qc->n_elem = n_elem;
>> + qc->sg = scsi_sglist(scmd);
>> + qc->nbytes = qc->sg->length;
>> + ata_sg_init(qc, qc->sg, n_elem);
>> + }
>> +
>> + scmd->submitter = SUBMITTED_BY_BLOCK_LAYER;
>> +
>> + ata_qc_issue(qc);
>
> Arg, no ! This potentially mixes NCQ and non-NCQ commands, which is
> forbidden by ATA spec. You need to use something like:
>
> if (ap->ops->qc_defer) {
> if ((rc = ap->ops->qc_defer(qc)))
> goto defer;
> }
>
> ata_qc_issue(qc);
>
> which is done in __ata_scsi_queuecmd() -> ata_scsi_translate()
>
> Unless you guarantee that ata_scsi_queue_internal() is always called
> from libata EH context ?
This will be called synchronously called from ata_exec_internal_sg(), so
the same rules on NCQ vs non-NCQ would apply here. As I see,
ata_exec_internal_sg() assumes non-NCQ mode and is not multi-thread safe.
Thanks,
John
>
>> +
>> + return 0;
>> +did_err:
>> + scmd->result = (DID_ERROR << 16);
>> + scsi_done(scmd);
>> + return 0;
>> +}
>> +
>> int __ata_scsi_queuecmd(struct scsi_cmnd *scmd, struct ata_device *dev)
>> {
>> u8 scsi_op = scmd->cmnd[0];
>> diff --git a/drivers/ata/libata.h b/drivers/ata/libata.h
>> index 0c2df1e60768..15cd1cd618b8 100644
>> --- a/drivers/ata/libata.h
>> +++ b/drivers/ata/libata.h
>> @@ -82,7 +82,6 @@ extern int ata_port_probe(struct ata_port *ap);
>> extern void __ata_port_probe(struct ata_port *ap);
>> extern unsigned int ata_read_log_page(struct ata_device *dev, u8 log,
>> u8 page, void *buf, unsigned int sectors);
>> -
>> #define to_ata_port(d) container_of(d, struct ata_port, tdev)
>>
>> /* libata-acpi.c */
>> @@ -130,6 +129,8 @@ extern int ata_scsi_user_scan(struct Scsi_Host *shost, unsigned int channel,
>> void ata_scsi_sdev_config(struct scsi_device *sdev);
>> int ata_scsi_dev_config(struct scsi_device *sdev, struct ata_device *dev);
>> int __ata_scsi_queuecmd(struct scsi_cmnd *scmd, struct ata_device *dev);
>> +unsigned int ata_scsi_queue_internal(struct scsi_cmnd *scmd,
>> + struct ata_device *dev);
>>
>> /* libata-eh.c */
>> extern unsigned int ata_internal_cmd_timeout(struct ata_device *dev, u8 cmd);
>> diff --git a/include/linux/libata.h b/include/linux/libata.h
>> index 827d5838cd23..8938b584520f 100644
>> --- a/include/linux/libata.h
>> +++ b/include/linux/libata.h
>> @@ -764,7 +764,9 @@ struct ata_link {
>>
>> struct device tdev;
>> unsigned int active_tag; /* active tag on this link */
>> + unsigned int preempted_tag;
>> u32 sactive; /* active NCQ commands */
>> + u32 preempted_sactive;
>>
>> unsigned int flags; /* ATA_LFLAG_xxx */
>>
>> @@ -857,6 +859,10 @@ struct ata_port {
>> #ifdef CONFIG_ATA_ACPI
>> struct ata_acpi_gtm __acpi_init_gtm; /* use ata_acpi_init_gtm() */
>> #endif
>> +
>> + u64 preempted_qc_active;
>> + int preempted_nr_active_links;
>> +
>> /* owned by EH */
>> u8 sector_buf[ATA_SECT_SIZE] ____cacheline_aligned;
>> };
>
next prev parent reply other threads:[~2022-10-27 10:45 UTC|newest]
Thread overview: 30+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-10-25 10:32 [PATCH RFC v3 0/7] blk-mq/libata/scsi: SCSI driver tagging improvements Part II John Garry
2022-10-25 10:32 ` [PATCH RFC v3 1/7] ata: libata-scsi: Add ata_scsi_queue_internal() John Garry
2022-10-27 1:42 ` Damien Le Moal
2022-10-27 10:45 ` John Garry [this message]
2022-10-27 22:24 ` Damien Le Moal
2022-10-25 10:32 ` [PATCH RFC v3 2/7] ata: libata-scsi: Add ata_internal_queuecommand() John Garry
2022-10-27 1:45 ` Damien Le Moal
2022-10-27 9:56 ` John Garry
2022-10-27 13:02 ` Hannes Reinecke
2022-10-27 17:23 ` John Garry
2022-10-27 22:35 ` Damien Le Moal
2022-10-28 8:14 ` John Garry
2022-10-28 8:26 ` Damien Le Moal
2022-10-27 22:25 ` Damien Le Moal
2022-10-28 8:01 ` John Garry
2022-10-28 8:07 ` Damien Le Moal
2022-10-28 8:33 ` John Garry
2022-10-31 5:59 ` Damien Le Moal
2022-11-02 9:52 ` John Garry
2022-11-02 10:07 ` Damien Le Moal
2022-11-02 11:12 ` Hannes Reinecke
2022-11-02 11:25 ` Damien Le Moal
2022-11-07 10:12 ` Hannes Reinecke
2022-11-07 13:29 ` Damien Le Moal
2022-11-07 14:34 ` Hannes Reinecke
2022-10-25 10:32 ` [PATCH RFC v3 3/7] ata: libata: Make space for ATA queue command in scmd payload John Garry
2022-10-25 10:32 ` [PATCH RFC v3 4/7] ata: libata: Add ata_internal_timeout() John Garry
2022-10-25 10:32 ` [PATCH RFC v3 5/7] ata: libata: Queue ATA internal commands as requests John Garry
2022-10-25 10:32 ` [PATCH RFC v3 6/7] scsi: mvsas: Remove internal tag handling John Garry
2022-10-25 10:32 ` [PATCH RFC v3 7/7] scsi: hisi_sas: Remove internal tag handling for reserved commands 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=8280ec1c-c274-ba8d-b533-3316865fe5c1@huawei.com \
--to=john.garry@huawei.com \
--cc=axboe@kernel.dk \
--cc=bvanassche@acm.org \
--cc=damien.lemoal@opensource.wdc.com \
--cc=hare@suse.de \
--cc=hch@lst.de \
--cc=jejb@linux.ibm.com \
--cc=jinpu.wang@cloud.ionos.com \
--cc=linux-block@vger.kernel.org \
--cc=linux-ide@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-scsi@vger.kernel.org \
--cc=linuxarm@huawei.com \
--cc=martin.petersen@oracle.com \
--cc=ming.lei@redhat.com \
--cc=niklas.cassel@wdc.com \
/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