Re: [PATCH v4 2/4] fusion: remove iopriority handling

[Hannes Reinecke]

On 10/13/2016 09:53 PM, Adam Manzanares wrote:
> The request priority is now by default coming from the ioc. It was not
> clear what this code was trying to do based upon the iopriority class or
> data. The driver should check that a device supports priorities and use
> them according to the specificiations of ioprio.
> 
> Signed-off-by: Adam Manzanares 
> ---
>  drivers/message/fusion/mptscsih.c | 5 -
>  1 file changed, 5 deletions(-)
> 
> diff --git a/drivers/message/fusion/mptscsih.c 
> b/drivers/message/fusion/mptscsih.c
> index 6c9fc11..4740bb6 100644
> --- a/drivers/message/fusion/mptscsih.c
> +++ b/drivers/message/fusion/mptscsih.c
> @@ -1369,11 +1369,6 @@ mptscsih_qcmd(struct scsi_cmnd *SCpnt)
>   if ((vdevice->vtarget->tflags & MPT_TARGET_FLAGS_Q_YES)
>   && (SCpnt->device->tagged_supported)) {
>   scsictl = scsidir | MPI_SCSIIO_CONTROL_SIMPLEQ;
> - if (SCpnt->request && SCpnt->request->ioprio) {
> - if (((SCpnt->request->ioprio & 0x7) == 1) ||
> - !(SCpnt->request->ioprio & 0x7))
> - scsictl |= MPI_SCSIIO_CONTROL_HEADOFQ;
> - }
>   } else
>   scsictl = scsidir | MPI_SCSIIO_CONTROL_UNTAGGED;
>  
> 
Reviewed-by: Hannes Reinecke 

Cheers,

Hannes
-- 
Dr. Hannes ReineckeTeamlead Storage & Networking
h...@suse.de   +49 911 74053 688
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: F. Imendörffer, J. Smithard, J. Guild, D. Upmanyu, G. Norton
HRB 21284 (AG Nürnberg)
--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH v4 1/4] block: Add iocontext priority to request

[Hannes Reinecke]

On 10/13/2016 09:53 PM, Adam Manzanares wrote:
> Patch adds an association between iocontext ioprio and the ioprio of a
> request. This value is set in blk_rq_set_prio which takes the request and
> the ioc as arguments. If the ioc is valid in blk_rq_set_prio then the
> iopriority of the request is set as the iopriority of the ioc. In
> init_request_from_bio a check is made to see if the ioprio of the bio is
> valid and if so then the request prio comes from the bio.
> 
> Signed-off-by: Adam Manzananares 
> ---
>  block/blk-core.c   |  4 +++-
>  include/linux/blkdev.h | 14 ++
>  2 files changed, 17 insertions(+), 1 deletion(-)
> 
> diff --git a/block/blk-core.c b/block/blk-core.c
> index 14d7c07..361b1b9 100644
> --- a/block/blk-core.c
> +++ b/block/blk-core.c
> @@ -1153,6 +1153,7 @@ static struct request *__get_request(struct 
> request_list *rl, int op,
>  
>   blk_rq_init(q, rq);
>   blk_rq_set_rl(rq, rl);
> + blk_rq_set_prio(rq, ioc);
>   req_set_op_attrs(rq, op, op_flags | REQ_ALLOCED);
>  
>   /* init elvpriv */
> @@ -1656,7 +1657,8 @@ void init_request_from_bio(struct request *req, struct 
> bio *bio)
>  
>   req->errors = 0;
>   req->__sector = bio->bi_iter.bi_sector;
> - req->ioprio = bio_prio(bio);
> + if (ioprio_valid(bio_prio(bio)))
> + req->ioprio = bio_prio(bio);
>   blk_rq_bio_prep(req->q, req, bio);
>  }
>  
> diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h
> index c47c358..9a0ceaa 100644
> --- a/include/linux/blkdev.h
> +++ b/include/linux/blkdev.h
> @@ -934,6 +934,20 @@ static inline unsigned int blk_rq_count_bios(struct 
> request *rq)
>  }
>  
>  /*
> + * blk_rq_set_prio - associate a request with prio from ioc
> + * @rq: request of interest
> + * @ioc: target iocontext
> + *
> + * Assocate request prio with ioc prio so request based drivers
> + * can leverage priority information.
> + */
> +static inline void blk_rq_set_prio(struct request *rq, struct io_context 
> *ioc)
> +{
> + if (ioc)
> + rq->ioprio = ioc->ioprio;
> +}
> +
> +/*
>   * Request issue related functions.
>   */
>  extern struct request *blk_peek_request(struct request_queue *q);
> 
Don't you need to check for 'ioprio_valid()' here, too?

Cheers,

Hannes
-- 
Dr. Hannes ReineckeTeamlead Storage & Networking
h...@suse.de   +49 911 74053 688
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: F. Imendörffer, J. Smithard, J. Guild, D. Upmanyu, G. Norton
HRB 21284 (AG Nürnberg)
--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH v5 2/4] fusion: remove iopriority handling

[Christoph Hellwig]

> Style wise you can further remove the extra parens around
>   SCpnt->device->tagged_supported
> As well as the now redundant braces.

I did send a patch looking just like that earlier :)
--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH v2 1/2] scsi: Handle Unit Attention when issuing SCSI command

[James Bottomley]

On Thu, 2016-10-13 at 15:47 -0300, Gabriel Krisman Bertazi wrote:
> Hi James,
> 
> Thanks for your review.  Please see the v2 below.
> 
> > OK, so really this isn't what you want, because blk_execute_req may
> > have used several of your retries, so you now get a maximum 
> > possible set of retries at UNIT_ATTENTION_RETRIES*retries.  You 
> > need to start from the returned req->retries, which probably means 
> > this loop needs to be inside __scsi_execute.
> 
> Hmm, I was aware of that, but I saw there were other places that may
> have run retries^2 times, like scsi_test_unit_ready and 
> scsi_mode_sense, if I read the code correctly.

There's a lot of this type of retry multiplication, but we really do
need to reduce it, not increase it.

>   But, I see your point and I fixed it on v2.  I also updated the
> second patch to rework these cases.
> 
> Another thing that got me confused is where the blk layer updates
> req->retries.
> 
> What do you think about the v2 below?
> 
> Thanks,
> 
> -- >8 --
> 
> Usually, re-sending the SCSI command is enough to recover from a Unit
> Attention (UA).  This adds a generic retry code to the SCSI command 
> path in case of an UA, before giving up and returning the error 
> condition to the caller.
> 
> I added the UA verification into scsi_execute instead of
> scsi_execute_req because there are at least a few callers that invoke
> scsi_execute directly and would benefit from the internal UA retry.
> Also, I didn't use scsi_normalize_sense to not duplicate 
> functionality with scsi_execute_req_flags.  Instead, scsi_execute 
> uses a small helper function that verifies only the UA condition 
> directly from the raw sense buffer.  If this design is not OK, I can 
> refactor to use scsi_normalize_sense.
> 
> This prevents us from duplicating the retry code in at least a few
> places.  In particular, it fixes an issue found in some IBM 
> enclosures, in which the device may return an Unit Attention during 
> probe, breaking the bind with the ses module:
> 
> scsi 1:0:7:0: Failed to get diagnostic page 0x802
> scsi 1:0:7:0: Failed to bind enclosure -19
> 
> Link: https://patchwork.kernel.org/patch/9336763/
> Suggested-by: Brian King 
> Suggested-by: James Bottomley 
> Signed-off-by: Gabriel Krisman Bertazi 
> ---
>  drivers/scsi/scsi_lib.c| 27 ---
>  include/scsi/scsi_common.h |  9 +
>  2 files changed, 33 insertions(+), 3 deletions(-)
> 
> diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c
> index c71344aebdbb..9c6623abf120 100644
> --- a/drivers/scsi/scsi_lib.c
> +++ b/drivers/scsi/scsi_lib.c
> @@ -187,15 +187,24 @@ int scsi_execute(struct scsi_device *sdev,
> const unsigned char *cmd,
>   struct request *req;
>   int write = (data_direction == DMA_TO_DEVICE);
>   int ret = DRIVER_ERROR << 24;
> + bool priv_sense = false;
>  
> + if (!sense) {
> + sense = kzalloc(SCSI_SENSE_BUFFERSIZE, GFP_NOIO);
> + if (!sense)
> + return ret;
> + priv_sense = true;
> + }

I think here we might be able to stack allocate this and avoid a
potential error return under memory pressure.  The annoying thing is
that blk_execute_rq does exactly this, so now we have an additional
useless allocation bloating the stack, but I think it's worth it to get
rid of priv_sense and all the error legs.

> + retry:
>   req = blk_get_request(sdev->request_queue, write,
> __GFP_RECLAIM);
>   if (IS_ERR(req))
> - return ret;
> + goto free_sense;
>   blk_rq_set_block_pc(req);
>  
>   if (bufflen &&  blk_rq_map_kern(sdev->request_queue,
> req,
>   buffer, bufflen,
> __GFP_RECLAIM))
> - goto out;
> + goto put_req;
>  
>   req->cmd_len = COMMAND_SIZE(cmd[0]);
>   memcpy(req->cmd, cmd, req->cmd_len);
> @@ -210,6 +219,13 @@ int scsi_execute(struct scsi_device *sdev, const
> unsigned char *cmd,
>*/
>   blk_execute_rq(req->q, NULL, req, 1);
>  
> + if (scsi_sense_unit_attention(sense) && req->retries > 0) {
> + memset(sense, 0, SCSI_SENSE_BUFFERSIZE);
> + retries = req->retries - 1;
> + blk_put_request(req);
> + goto retry;
> + }

OK, so this is more theory, but I think you can actually reuse the same
request to go around this loop without doing a get/put.  I've cc'd Jens
to confirm, since no other driver I can find does this, but if it's
legal, it saves freeing and reallocating the request.  You can then
replace the goto with a do { } while (...) which makes the loop obvious
to the next person looking at this.

>   /*
>* Some devices (USB mass-storage in particular) may
> transfer
>* garbage data together with a residue indicating that the
> data
> @@ -222,9 +238,14 @@ int scsi_execute(struct scsi_device *sdev, const

[PATCH RFC 2/2] blk-mq: Avoid memory reclaim when remapping queues

[Gabriel Krisman Bertazi]

While stressing memory and IO at the same time we changed SMT settings,
we were able to consistently trigger deadlocks in the mm system, which
froze the entire machine.

I think that under memory stress conditions, the large allocations
performed by blk_mq_init_rq_map may trigger a reclaim, which stalls
waiting on the block layer remmaping completion, thus deadlocking the
system.  The trace below was collected after the machine stalled,
waiting for the hotplug event completion.

The simplest fix for this is to make allocations in this path
non-reclaimable, with GFP_NOWAIT.  With this patch, We couldn't hit the
issue anymore.

This should apply on top of Jen's for-next branch cleanly.

 Call Trace:
[c00f0160aaf0] [c00f0160ab50] 0xc00f0160ab50 (unreliable)
[c00f0160acc0] [c0016624] __switch_to+0x2e4/0x430
[c00f0160ad20] [c0b1a880] __schedule+0x310/0x9b0
[c00f0160ae00] [c0b1af68] schedule+0x48/0xc0
[c00f0160ae30] [c0b1b4b0] schedule_preempt_disabled+0x20/0x30
[c00f0160ae50] [c0b1d4fc] __mutex_lock_slowpath+0xec/0x1f0
[c00f0160aed0] [c0b1d678] mutex_lock+0x78/0xa0
[c00f0160af00] [d00019413cac] xfs_reclaim_inodes_ag+0x33c/0x380 [xfs]
[c00f0160b0b0] [d00019415164] xfs_reclaim_inodes_nr+0x54/0x70 [xfs]
[c00f0160b0f0] [d000194297f8] xfs_fs_free_cached_objects+0x38/0x60 [xfs]
[c00f0160b120] [c03172c8] super_cache_scan+0x1f8/0x210
[c00f0160b190] [c026301c] shrink_slab.part.13+0x21c/0x4c0
[c00f0160b2d0] [c0268088] shrink_zone+0x2d8/0x3c0
[c00f0160b380] [c026834c] do_try_to_free_pages+0x1dc/0x520
[c00f0160b450] [c026876c] try_to_free_pages+0xdc/0x250
[c00f0160b4e0] [c0251978] __alloc_pages_nodemask+0x868/0x10d0
[c00f0160b6f0] [c0567030] blk_mq_init_rq_map+0x160/0x380
[c00f0160b7a0] [c056758c] blk_mq_map_swqueue+0x33c/0x360
[c00f0160b820] [c0567904] blk_mq_queue_reinit+0x64/0xb0
[c00f0160b850] [c056a16c] blk_mq_queue_reinit_notify+0x19c/0x250
[c00f0160b8a0] [c00f5d38] notifier_call_chain+0x98/0x100
[c00f0160b8f0] [c00c5fb0] __cpu_notify+0x70/0xe0
[c00f0160b930] [c00c63c4] notify_prepare+0x44/0xb0
[c00f0160b9b0] [c00c52f4] cpuhp_invoke_callback+0x84/0x250
[c00f0160ba10] [c00c570c] cpuhp_up_callbacks+0x5c/0x120
[c00f0160ba60] [c00c7cb8] _cpu_up+0xf8/0x1d0
[c00f0160bac0] [c00c7eb0] do_cpu_up+0x120/0x150
[c00f0160bb40] [c06fe024] cpu_subsys_online+0x64/0xe0
[c00f0160bb90] [c06f5124] device_online+0xb4/0x120
[c00f0160bbd0] [c06f5244] online_store+0xb4/0xc0
[c00f0160bc20] [c06f0a68] dev_attr_store+0x68/0xa0
[c00f0160bc60] [c03ccc30] sysfs_kf_write+0x80/0xb0
[c00f0160bca0] [c03cbabc] kernfs_fop_write+0x17c/0x250
[c00f0160bcf0] [c030fe6c] __vfs_write+0x6c/0x1e0
[c00f0160bd90] [c0311490] vfs_write+0xd0/0x270
[c00f0160bde0] [c03131fc] SyS_write+0x6c/0x110
[c00f0160be30] [c0009204] system_call+0x38/0xec

Signed-off-by: Gabriel Krisman Bertazi 
Cc: Brian King 
Cc: linux-bl...@vger.kernel.org
Cc: linux-scsi@vger.kernel.org
---
 block/blk-mq.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/block/blk-mq.c b/block/blk-mq.c
index f19e3b8c9e15..dc19befa48b4 100644
--- a/block/blk-mq.c
+++ b/block/blk-mq.c
@@ -1487,7 +1487,7 @@ static struct blk_mq_tags *blk_mq_init_rq_map(struct 
blk_mq_tag_set *set,
INIT_LIST_HEAD(>page_list);
 
tags->rqs = kzalloc_node(set->queue_depth * sizeof(struct request *),
-GFP_KERNEL | __GFP_NOWARN | __GFP_NORETRY,
+GFP_NOWAIT | __GFP_NOWARN | __GFP_NORETRY,
 set->numa_node);
if (!tags->rqs) {
blk_mq_free_tags(tags);
@@ -1513,7 +1513,7 @@ static struct blk_mq_tags *blk_mq_init_rq_map(struct 
blk_mq_tag_set *set,
 
do {
page = alloc_pages_node(set->numa_node,
-   GFP_KERNEL | __GFP_NOWARN | __GFP_NORETRY | 
__GFP_ZERO,
+   GFP_NOWAIT | __GFP_NOWARN | __GFP_NORETRY | 
__GFP_ZERO,
this_order);
if (page)
break;
@@ -1534,7 +1534,7 @@ static struct blk_mq_tags *blk_mq_init_rq_map(struct 
blk_mq_tag_set *set,
 * Allow kmemleak to scan these pages as they contain pointers
 * to additional allocations like via ops->init_request().
 */
-   kmemleak_alloc(p, order_to_size(this_order), 1, GFP_KERNEL);
+   kmemleak_alloc(p, order_to_size(this_order), 1, GFP_NOWAIT);
entries_per_page = order_to_size(this_order) / 

Re: [PATCH v4 2/4] fusion: remove iopriority handling

[Adam Manzanares]

The 10/13/2016 15:05, Sathya Prakash Veerichetty wrote:
> By removing the code below, we put all the commands for all the types of
> devices (SAS/SATA) as simple-Q (requeue as the device require) and I am
> not sure whether it is the intention of this change.
> 

This is the intention of the change. I don't think the iopriority of the
request is being used correctly. What does it mean to use 0x7 as an 
indicator that a command should be put at the head of the queue? This 
would be clearer if it was using some of the macros from ioprio. If 
0x7 means something special I think this should be some #define in the 
includes of the fusion driver with some documentation.

> -Original Message-
> From: Adam Manzanares [mailto:adam.manzana...@hgst.com]
> Sent: Thursday, October 13, 2016 1:54 PM
> To: ax...@kernel.dk; t...@kernel.org; dan.j.willi...@intel.com;
> h...@suse.de; martin.peter...@oracle.com; mchri...@redhat.com;
> toshi.k...@hpe.com; ming@canonical.com; sathya.prak...@broadcom.com;
> chaitra.basa...@broadcom.com; suganath-prabu.subram...@broadcom.com
> Cc: linux-bl...@vger.kernel.org; linux-...@vger.kernel.org;
> linux-ker...@vger.kernel.org; mpt-fusionlinux@broadcom.com;
> linux-scsi@vger.kernel.org; Adam Manzanares; Adam Manzanares
> Subject: [PATCH v4 2/4] fusion: remove iopriority handling
> 
> The request priority is now by default coming from the ioc. It was not
> clear what this code was trying to do based upon the iopriority class or
> data. The driver should check that a device supports priorities and use
> them according to the specificiations of ioprio.
> 
> Signed-off-by: Adam Manzanares 
> ---
>  drivers/message/fusion/mptscsih.c | 5 -
>  1 file changed, 5 deletions(-)
> 
> diff --git a/drivers/message/fusion/mptscsih.c
> b/drivers/message/fusion/mptscsih.c
> index 6c9fc11..4740bb6 100644
> --- a/drivers/message/fusion/mptscsih.c
> +++ b/drivers/message/fusion/mptscsih.c
> @@ -1369,11 +1369,6 @@ mptscsih_qcmd(struct scsi_cmnd *SCpnt)
>   if ((vdevice->vtarget->tflags & MPT_TARGET_FLAGS_Q_YES)
>   && (SCpnt->device->tagged_supported)) {
>   scsictl = scsidir | MPI_SCSIIO_CONTROL_SIMPLEQ;
> - if (SCpnt->request && SCpnt->request->ioprio) {
> - if (((SCpnt->request->ioprio & 0x7) == 1) ||
> - !(SCpnt->request->ioprio & 0x7))
> - scsictl |= MPI_SCSIIO_CONTROL_HEADOFQ;
> - }
>   } else
>   scsictl = scsidir | MPI_SCSIIO_CONTROL_UNTAGGED;
> 
> --
> 2.1.4
> 
> Western Digital Corporation (and its subsidiaries) E-mail Confidentiality
> Notice & Disclaimer:
> 
> This e-mail and any files transmitted with it may contain confidential or
> legally privileged information of WDC and/or its affiliates, and are
> intended solely for the use of the individual or entity to which they are
> addressed. If you are not the intended recipient, any disclosure, copying,
> distribution or any action taken or omitted to be taken in reliance on it,
> is prohibited. If you have received this e-mail in error, please notify
> the sender immediately and delete the e-mail in its entirety from your
> system.
> --
> To unsubscribe from this list: send the line "unsubscribe linux-ide" in
> the body of a message to majord...@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

Take care,
Adam
--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH v2 1/6] ibmvscsis: Rearrange functions for future patches

[Michael Cyr]

This patch reorders functions in a manner necessary for a follow-on
patch.  It also makes some minor styling changes (mostly removing extra
spaces) and fixes some typos.

There are no code changes in this patch, with one exception: due to the
reordering of the functions, I needed to explicitly declare a function
at the top of the file.  However, this will be removed in the next patch,
since the code requiring the predeclaration will be removed.

Signed-off-by: Michael Cyr 
Signed-off-by: Bryant G. Ly 
---
 drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.c | 796 ---
 1 file changed, 399 insertions(+), 397 deletions(-)

diff --git a/drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.c 
b/drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.c
index 642b739..01a430c 100644
--- a/drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.c
+++ b/drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.c
@@ -22,7 +22,7 @@
  *
  /
 
-#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
+#define pr_fmt(fmt)KBUILD_MODNAME ": " fmt
 
 #include 
 #include 
@@ -61,6 +61,8 @@ static long ibmvscsis_parse_command(struct scsi_info *vscsi,
 
 static void ibmvscsis_adapter_idle(struct scsi_info *vscsi);
 
+static void ibmvscsis_reset_queue(struct scsi_info *vscsi, uint new_state);
+
 static void ibmvscsis_determine_resid(struct se_cmd *se_cmd,
  struct srp_rsp *rsp)
 {
@@ -81,7 +83,7 @@ static void ibmvscsis_determine_resid(struct se_cmd *se_cmd,
}
} else if (se_cmd->se_cmd_flags & SCF_OVERFLOW_BIT) {
if (se_cmd->data_direction == DMA_TO_DEVICE) {
-   /*  residual data from an overflow write */
+   /* residual data from an overflow write */
rsp->flags = SRP_RSP_FLAG_DOOVER;
rsp->data_out_res_cnt = cpu_to_be32(residual_count);
} else if (se_cmd->data_direction == DMA_FROM_DEVICE) {
@@ -101,7 +103,7 @@ static void ibmvscsis_determine_resid(struct se_cmd *se_cmd,
  * and the function returns TRUE.
  *
  * EXECUTION ENVIRONMENT:
- *  Interrupt or Process environment
+ * Interrupt or Process environment
  */
 static bool connection_broken(struct scsi_info *vscsi)
 {
@@ -324,7 +326,7 @@ static struct viosrp_crq *ibmvscsis_cmd_q_dequeue(uint mask,
 }
 
 /**
- * ibmvscsis_send_init_message() -  send initialize message to the client
+ * ibmvscsis_send_init_message() - send initialize message to the client
  * @vscsi: Pointer to our adapter structure
  * @format:Which Init Message format to send
  *
@@ -382,13 +384,13 @@ static long ibmvscsis_check_init_msg(struct scsi_info 
*vscsi, uint *format)
  vscsi->cmd_q.base_addr);
if (crq) {
*format = (uint)(crq->format);
-   rc =  ERROR;
+   rc = ERROR;
crq->valid = INVALIDATE_CMD_RESP_EL;
dma_rmb();
}
} else {
*format = (uint)(crq->format);
-   rc =  ERROR;
+   rc = ERROR;
crq->valid = INVALIDATE_CMD_RESP_EL;
dma_rmb();
}
@@ -397,166 +399,6 @@ static long ibmvscsis_check_init_msg(struct scsi_info 
*vscsi, uint *format)
 }
 
 /**
- * ibmvscsis_establish_new_q() - Establish new CRQ queue
- * @vscsi: Pointer to our adapter structure
- * @new_state: New state being established after resetting the queue
- *
- * Must be called with interrupt lock held.
- */
-static long ibmvscsis_establish_new_q(struct scsi_info *vscsi,  uint new_state)
-{
-   long rc = ADAPT_SUCCESS;
-   uint format;
-
-   vscsi->flags &= PRESERVE_FLAG_FIELDS;
-   vscsi->rsp_q_timer.timer_pops = 0;
-   vscsi->debit = 0;
-   vscsi->credit = 0;
-
-   rc = vio_enable_interrupts(vscsi->dma_dev);
-   if (rc) {
-   pr_warn("reset_queue: failed to enable interrupts, rc %ld\n",
-   rc);
-   return rc;
-   }
-
-   rc = ibmvscsis_check_init_msg(vscsi, );
-   if (rc) {
-   dev_err(>dev, "reset_queue: check_init_msg failed, rc 
%ld\n",
-   rc);
-   return rc;
-   }
-
-   if (format == UNUSED_FORMAT && new_state == WAIT_CONNECTION) {
-   rc = ibmvscsis_send_init_message(vscsi, INIT_MSG);
-   switch (rc) {
-   case H_SUCCESS:
-   case H_DROPPED:
-   case H_CLOSED:
-   rc = ADAPT_SUCCESS;
-   break;
-
-   case H_PARAMETER:
-   case H_HARDWARE:
-   break;
-
-   default:
-   vscsi->state = UNDEFINED;
-   rc = H_HARDWARE;
-   break;
-   }

Re: [PATCH v5 4/4] ata: ATA Command Priority Disabled By Default

[Adam Manzanares]

Hello Tejun,

The 10/13/2016 19:22, Tejun Heo wrote:
> Hello, Adam.
> 
> Sorry about late reply.  Was on vacation.

NP I was on vacation at the end of the week last week.

> 
> On Thu, Oct 13, 2016 at 04:00:31PM -0700, Adam Manzanares wrote:
> > Add a sysfs entry to turn on priority information being passed
> > to a ATA device. By default this feature is turned off.
> > 
> > This patch depends on ata: Enabling ATA Command Priorities
> 
> Looks generally good but can we please use a device attribute name
> which is more specific - ie. enable_ncq_prio?

Will do, I'll also double check the naming scheme of functions and variables 
also. The functions that check if the device has the ncq prio capability might
be too similar to the function that checks if the device attribute is 
enabled.

> 
> Thanks.
> 
> -- 
> tejun

Take care,
Adam
--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH v5 2/4] fusion: remove iopriority handling

[Shaun Tancheff]

On Thu, Oct 13, 2016 at 6:00 PM, Adam Manzanares
 wrote:
> The request priority is now by default coming from the ioc. It was not
> clear what this code was trying to do based upon the iopriority class or
> data. The driver should check that a device supports priorities and use
> them according to the specificiations of ioprio.
>
> Signed-off-by: Adam Manzanares 
> ---
>  drivers/message/fusion/mptscsih.c | 5 -
>  1 file changed, 5 deletions(-)
>
> diff --git a/drivers/message/fusion/mptscsih.c 
> b/drivers/message/fusion/mptscsih.c
> index 6c9fc11..4740bb6 100644
> --- a/drivers/message/fusion/mptscsih.c
> +++ b/drivers/message/fusion/mptscsih.c
> @@ -1369,11 +1369,6 @@ mptscsih_qcmd(struct scsi_cmnd *SCpnt)
> if ((vdevice->vtarget->tflags & MPT_TARGET_FLAGS_Q_YES)
> && (SCpnt->device->tagged_supported)) {
> scsictl = scsidir | MPI_SCSIIO_CONTROL_SIMPLEQ;
> -   if (SCpnt->request && SCpnt->request->ioprio) {
> -   if (((SCpnt->request->ioprio & 0x7) == 1) ||
> -   !(SCpnt->request->ioprio & 0x7))
> -   scsictl |= MPI_SCSIIO_CONTROL_HEADOFQ;
> -   }
> } else
> scsictl = scsidir | MPI_SCSIIO_CONTROL_UNTAGGED;

Style wise you can further remove the extra parens around
  SCpnt->device->tagged_supported
As well as the now redundant braces.

Regards,
Shaun

> --
> 2.1.4
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-block" in
> the body of a message to majord...@vger.kernel.org
> More majordomo info at  
> https://urldefense.proofpoint.com/v2/url?u=http-3A__vger.kernel.org_majordomo-2Dinfo.html=DQIBAg=IGDlg0lD0b-nebmJJ0Kp8A=Wg5NqlNlVTT7Ugl8V50qIHLe856QW0qfG3WVYGOrWzA=ZE7JzxXeXPEWqk9WYm42hZHj8gESRg1QoS5XklfbprM=C0iMyTgYbYl06F1SQ2DqfdESKBtl3Whp5rSnHSBXOc4=
--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH v5 3/4] ata: Enabling ATA Command Priorities

[Tejun Heo]

Hello,

On Thu, Oct 13, 2016 at 04:00:30PM -0700, Adam Manzanares wrote:
> This patch checks to see if an ATA device supports NCQ command priorities.
> If so and the user has specified an iocontext that indicates
> IO_PRIO_CLASS_RT then we build a tf with a high priority command.
> 
> This is done to improve the tail latency of commands that are high
> priority by passing priority to the device.
> 
> Signed-off-by: Adam Manzanares 

Looks good to me.  Once the block changes are applied, I'll pull it
into libata tree and apply the ata part on top.

Thanks.

-- 
tejun
--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH v5 3/4] ata: Enabling ATA Command Priorities

[Adam Manzanares]

This patch checks to see if an ATA device supports NCQ command priorities.
If so and the user has specified an iocontext that indicates
IO_PRIO_CLASS_RT then we build a tf with a high priority command.

This is done to improve the tail latency of commands that are high
priority by passing priority to the device.

Signed-off-by: Adam Manzanares 
---
 drivers/ata/libata-core.c | 35 ++-
 drivers/ata/libata-scsi.c |  6 +-
 drivers/ata/libata.h  |  2 +-
 include/linux/ata.h   |  6 ++
 include/linux/libata.h| 18 ++
 5 files changed, 64 insertions(+), 3 deletions(-)

diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index 223a770..181b530 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -739,6 +739,7 @@ u64 ata_tf_read_block(const struct ata_taskfile *tf, struct 
ata_device *dev)
  * @n_block: Number of blocks
  * @tf_flags: RW/FUA etc...
  * @tag: tag
+ * @class: IO priority class
  *
  * LOCKING:
  * None.
@@ -753,7 +754,7 @@ u64 ata_tf_read_block(const struct ata_taskfile *tf, struct 
ata_device *dev)
  */
 int ata_build_rw_tf(struct ata_taskfile *tf, struct ata_device *dev,
u64 block, u32 n_block, unsigned int tf_flags,
-   unsigned int tag)
+   unsigned int tag, int class)
 {
tf->flags |= ATA_TFLAG_ISADDR | ATA_TFLAG_DEVICE;
tf->flags |= tf_flags;
@@ -785,6 +786,12 @@ int ata_build_rw_tf(struct ata_taskfile *tf, struct 
ata_device *dev,
tf->device = ATA_LBA;
if (tf->flags & ATA_TFLAG_FUA)
tf->device |= 1 << 7;
+
+   if (ata_ncq_prio_enabled(dev)) {
+   if (class == IOPRIO_CLASS_RT)
+   tf->hob_nsect |= ATA_PRIO_HIGH <<
+ATA_SHIFT_PRIO;
+   }
} else if (dev->flags & ATA_DFLAG_LBA) {
tf->flags |= ATA_TFLAG_LBA;
 
@@ -2156,6 +2163,30 @@ static void ata_dev_config_ncq_non_data(struct 
ata_device *dev)
}
 }
 
+static void ata_dev_config_ncq_prio(struct ata_device *dev)
+{
+   struct ata_port *ap = dev->link->ap;
+   unsigned int err_mask;
+
+   err_mask = ata_read_log_page(dev,
+ATA_LOG_SATA_ID_DEV_DATA,
+ATA_LOG_SATA_SETTINGS,
+ap->sector_buf,
+1);
+   if (err_mask) {
+   ata_dev_dbg(dev,
+   "failed to get Identify Device data, Emask 0x%x\n",
+   err_mask);
+   return;
+   }
+
+   if (ap->sector_buf[ATA_LOG_NCQ_PRIO_OFFSET] & BIT(3))
+   dev->flags |= ATA_DFLAG_NCQ_PRIO;
+   else
+   ata_dev_dbg(dev, "SATA page does not support priority\n");
+
+}
+
 static int ata_dev_config_ncq(struct ata_device *dev,
   char *desc, size_t desc_sz)
 {
@@ -2205,6 +2236,8 @@ static int ata_dev_config_ncq(struct ata_device *dev,
ata_dev_config_ncq_send_recv(dev);
if (ata_id_has_ncq_non_data(dev->id))
ata_dev_config_ncq_non_data(dev);
+   if (ata_id_has_ncq_prio(dev->id))
+   ata_dev_config_ncq_prio(dev);
}
 
return 0;
diff --git a/drivers/ata/libata-scsi.c b/drivers/ata/libata-scsi.c
index e207b33..18629e8 100644
--- a/drivers/ata/libata-scsi.c
+++ b/drivers/ata/libata-scsi.c
@@ -50,6 +50,7 @@
 #include 
 #include 
 #include 
+#include 
 
 #include "libata.h"
 #include "libata-transport.h"
@@ -1757,6 +1758,8 @@ static unsigned int ata_scsi_rw_xlat(struct 
ata_queued_cmd *qc)
 {
struct scsi_cmnd *scmd = qc->scsicmd;
const u8 *cdb = scmd->cmnd;
+   struct request *rq = scmd->request;
+   int class = IOPRIO_PRIO_CLASS(req_get_ioprio(rq));
unsigned int tf_flags = 0;
u64 block;
u32 n_block;
@@ -1823,7 +1826,8 @@ static unsigned int ata_scsi_rw_xlat(struct 
ata_queued_cmd *qc)
qc->nbytes = n_block * scmd->device->sector_size;
 
rc = ata_build_rw_tf(>tf, qc->dev, block, n_block, tf_flags,
-qc->tag);
+qc->tag, class);
+
if (likely(rc == 0))
return 0;
 
diff --git a/drivers/ata/libata.h b/drivers/ata/libata.h
index 3b301a4..8f3a559 100644
--- a/drivers/ata/libata.h
+++ b/drivers/ata/libata.h
@@ -66,7 +66,7 @@ extern u64 ata_tf_to_lba48(const struct ata_taskfile *tf);
 extern struct ata_queued_cmd *ata_qc_new_init(struct ata_device *dev, int tag);
 extern int ata_build_rw_tf(struct ata_taskfile *tf, struct ata_device *dev,
   u64 block, u32 n_block, unsigned int tf_flags,
-  unsigned int tag);
+  

[PATCH v5 4/4] ata: ATA Command Priority Disabled By Default

[Adam Manzanares]

Add a sysfs entry to turn on priority information being passed
to a ATA device. By default this feature is turned off.

This patch depends on ata: Enabling ATA Command Priorities

Signed-off-by: Adam Manzanares 
---
 drivers/ata/libahci.c |  1 +
 drivers/ata/libata-core.c |  2 +-
 drivers/ata/libata-scsi.c | 68 +++
 include/linux/libata.h|  8 ++
 4 files changed, 78 insertions(+), 1 deletion(-)

diff --git a/drivers/ata/libahci.c b/drivers/ata/libahci.c
index dcf2c72..383adf7 100644
--- a/drivers/ata/libahci.c
+++ b/drivers/ata/libahci.c
@@ -140,6 +140,7 @@ EXPORT_SYMBOL_GPL(ahci_shost_attrs);
 struct device_attribute *ahci_sdev_attrs[] = {
_attr_sw_activity,
_attr_unload_heads,
+   _attr_enable_prio,
NULL
 };
 EXPORT_SYMBOL_GPL(ahci_sdev_attrs);
diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index 181b530..d0cf987 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -787,7 +787,7 @@ int ata_build_rw_tf(struct ata_taskfile *tf, struct 
ata_device *dev,
if (tf->flags & ATA_TFLAG_FUA)
tf->device |= 1 << 7;
 
-   if (ata_ncq_prio_enabled(dev)) {
+   if (ata_ncq_prio_enabled(dev) && ata_prio_enabled(dev)) {
if (class == IOPRIO_CLASS_RT)
tf->hob_nsect |= ATA_PRIO_HIGH <<
 ATA_SHIFT_PRIO;
diff --git a/drivers/ata/libata-scsi.c b/drivers/ata/libata-scsi.c
index 18629e8..10ba118 100644
--- a/drivers/ata/libata-scsi.c
+++ b/drivers/ata/libata-scsi.c
@@ -271,6 +271,73 @@ DEVICE_ATTR(unload_heads, S_IRUGO | S_IWUSR,
ata_scsi_park_show, ata_scsi_park_store);
 EXPORT_SYMBOL_GPL(dev_attr_unload_heads);
 
+static ssize_t ata_enable_prio_show(struct device *device,
+   struct device_attribute *attr, char *buf)
+{
+   struct scsi_device *sdev = to_scsi_device(device);
+   struct ata_port *ap;
+   struct ata_device *dev;
+   int rc = 0;
+   int enable_prio;
+
+   ap = ata_shost_to_port(sdev->host);
+
+   spin_lock_irq(ap->lock);
+   dev = ata_scsi_find_dev(ap, sdev);
+   if (!dev) {
+   rc = -ENODEV;
+   goto unlock;
+   }
+
+   enable_prio = ata_prio_enabled(dev);
+
+unlock:
+   spin_unlock_irq(ap->lock);
+
+   return rc ? rc : snprintf(buf, 20, "%u\n", enable_prio);
+}
+
+static ssize_t ata_enable_prio_store(struct device *device,
+struct device_attribute *attr,
+const char *buf, size_t len)
+{
+   struct scsi_device *sdev = to_scsi_device(device);
+   struct ata_port *ap;
+   struct ata_device *dev;
+   long int input;
+   unsigned long flags;
+   int rc;
+
+   rc = kstrtol(buf, 10, );
+   if (rc)
+   return rc;
+   if ((input < 0) || (input > 1))
+   return -EINVAL;
+
+   ap = ata_shost_to_port(sdev->host);
+
+   spin_lock_irqsave(ap->lock, flags);
+   dev = ata_scsi_find_dev(ap, sdev);
+   if (unlikely(!dev)) {
+   rc = -ENODEV;
+   goto unlock;
+   }
+
+   if (input)
+   dev->flags |= ATA_DFLAG_ENABLE_PRIO;
+   else
+   dev->flags &= ~ATA_DFLAG_ENABLE_PRIO;
+
+unlock:
+   spin_unlock_irqrestore(ap->lock, flags);
+
+   return rc ? rc : len;
+}
+
+DEVICE_ATTR(enable_prio, S_IRUGO | S_IWUSR,
+   ata_enable_prio_show, ata_enable_prio_store);
+EXPORT_SYMBOL_GPL(dev_attr_enable_prio);
+
 void ata_scsi_set_sense(struct ata_device *dev, struct scsi_cmnd *cmd,
u8 sk, u8 asc, u8 ascq)
 {
@@ -402,6 +469,7 @@ EXPORT_SYMBOL_GPL(dev_attr_sw_activity);
 
 struct device_attribute *ata_common_sdev_attrs[] = {
_attr_unload_heads,
+   _attr_enable_prio,
NULL
 };
 EXPORT_SYMBOL_GPL(ata_common_sdev_attrs);
diff --git a/include/linux/libata.h b/include/linux/libata.h
index 244f261..c8acb16 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -166,6 +166,7 @@ enum {
ATA_DFLAG_UNLOCK_HPA= (1 << 18), /* unlock HPA */
ATA_DFLAG_NCQ_SEND_RECV = (1 << 19), /* device supports NCQ SEND and 
RECV */
ATA_DFLAG_NCQ_PRIO  = (1 << 20), /* device supports NCQ priority */
+   ATA_DFLAG_ENABLE_PRIO   = (1 << 21), /* User enable device priority */
ATA_DFLAG_INIT_MASK = (1 << 24) - 1,
 
ATA_DFLAG_DETACH= (1 << 24),
@@ -544,6 +545,7 @@ typedef void (*ata_postreset_fn_t)(struct ata_link *link, 
unsigned int *classes)
 
 extern struct device_attribute dev_attr_link_power_management_policy;
 extern struct device_attribute dev_attr_unload_heads;
+extern struct device_attribute dev_attr_enable_prio;
 extern struct device_attribute dev_attr_em_message_type;
 extern struct device_attribute 

[PATCH v5 2/4] fusion: remove iopriority handling

[Adam Manzanares]

The request priority is now by default coming from the ioc. It was not
clear what this code was trying to do based upon the iopriority class or
data. The driver should check that a device supports priorities and use
them according to the specificiations of ioprio.

Signed-off-by: Adam Manzanares 
---
 drivers/message/fusion/mptscsih.c | 5 -
 1 file changed, 5 deletions(-)

diff --git a/drivers/message/fusion/mptscsih.c 
b/drivers/message/fusion/mptscsih.c
index 6c9fc11..4740bb6 100644
--- a/drivers/message/fusion/mptscsih.c
+++ b/drivers/message/fusion/mptscsih.c
@@ -1369,11 +1369,6 @@ mptscsih_qcmd(struct scsi_cmnd *SCpnt)
if ((vdevice->vtarget->tflags & MPT_TARGET_FLAGS_Q_YES)
&& (SCpnt->device->tagged_supported)) {
scsictl = scsidir | MPI_SCSIIO_CONTROL_SIMPLEQ;
-   if (SCpnt->request && SCpnt->request->ioprio) {
-   if (((SCpnt->request->ioprio & 0x7) == 1) ||
-   !(SCpnt->request->ioprio & 0x7))
-   scsictl |= MPI_SCSIIO_CONTROL_HEADOFQ;
-   }
} else
scsictl = scsidir | MPI_SCSIIO_CONTROL_UNTAGGED;
 
-- 
2.1.4

--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH v5 1/4] block: Add iocontext priority to request

[Adam Manzanares]

Patch adds an association between iocontext ioprio and the ioprio of a
request. This is done to enable request based drivers the ability to
act on priority information stored in the request. An example being
ATA devices that support command priorities. If the ATA driver discovers
that the device supports command priorities and the request has valid
priority information indicating the request is high priority, then a high
priority command can be sent to the device. This should improve tail
latencies for high priority IO on any device that queues requests
internally and can make use of the priority information stored in the
request.

The ioprio of the request is set in blk_rq_set_prio which takes the
request and the ioc as arguments. If the ioc is valid in blk_rq_set_prio
then the iopriority of the request is set as the iopriority of the ioc.
In init_request_from_bio a check is made to see if the ioprio of the bio
is valid and if so then the request prio comes from the bio.

Signed-off-by: Adam Manzananares 
---
 block/blk-core.c   |  4 +++-
 include/linux/blkdev.h | 14 ++
 2 files changed, 17 insertions(+), 1 deletion(-)

diff --git a/block/blk-core.c b/block/blk-core.c
index 14d7c07..361b1b9 100644
--- a/block/blk-core.c
+++ b/block/blk-core.c
@@ -1153,6 +1153,7 @@ static struct request *__get_request(struct request_list 
*rl, int op,
 
blk_rq_init(q, rq);
blk_rq_set_rl(rq, rl);
+   blk_rq_set_prio(rq, ioc);
req_set_op_attrs(rq, op, op_flags | REQ_ALLOCED);
 
/* init elvpriv */
@@ -1656,7 +1657,8 @@ void init_request_from_bio(struct request *req, struct 
bio *bio)
 
req->errors = 0;
req->__sector = bio->bi_iter.bi_sector;
-   req->ioprio = bio_prio(bio);
+   if (ioprio_valid(bio_prio(bio)))
+   req->ioprio = bio_prio(bio);
blk_rq_bio_prep(req->q, req, bio);
 }
 
diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h
index c47c358..9a0ceaa 100644
--- a/include/linux/blkdev.h
+++ b/include/linux/blkdev.h
@@ -934,6 +934,20 @@ static inline unsigned int blk_rq_count_bios(struct 
request *rq)
 }
 
 /*
+ * blk_rq_set_prio - associate a request with prio from ioc
+ * @rq: request of interest
+ * @ioc: target iocontext
+ *
+ * Assocate request prio with ioc prio so request based drivers
+ * can leverage priority information.
+ */
+static inline void blk_rq_set_prio(struct request *rq, struct io_context *ioc)
+{
+   if (ioc)
+   rq->ioprio = ioc->ioprio;
+}
+
+/*
  * Request issue related functions.
  */
 extern struct request *blk_peek_request(struct request_queue *q);
-- 
2.1.4

--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH v5 0/4] Enabling ATA Command Priorities

[Adam Manzanares]

This patch builds ATA commands with high priority if the iocontext of a process
is set to real time. The goal of the patch is to improve tail latencies of 
workloads that use higher queue depths. This requires setting the iocontext 
ioprio on the request when it is initialized.

This patch has been tested with an Ultrastar HE8 HDD and cuts the 
the p99.99 tail latency of foreground IO from 2s down to 72ms when
using the deadline scheduler. This patch works independently of the
scheduler so it can be used with all of the currently available 
request based schedulers. 

Foreground IO, for the previously described results, is an async fio job 
submitting 4K read requests at a QD of 1 to the HDD. The foreground IO is set 
with the iopriority class of real time. The background workload is another fio
job submitting read requests at a QD of 32 to the same HDD with default 
iopriority.

This feature is enabled for ATA devices by setting the ata enable_prio device 
attribute to 1. An ATA device is also checked to see if the device supports per
command priority.

v5:
 - Updated block patch commit message to explain the why 

v4:
 - Added blk_rq_set_prio function to associate request prio with ioc prio
 - In init_request_from_bio use bio_prio if it is valid
 - Removed queue flag used to enable prio
 - Added ata enable_prio dev attribute to sysfs to enable prioritized commands

v3:
 - Removed null dereference issue in blk-core
 - Renamed queue sysfs entries for clarity
 - Added documentation for sysfs queue entry

v2:
 - Add queue flag to set iopriority going to the request
 - If queue flag set, send iopriority class to ata_build_rw_tf
 - Remove redundant code in ata_ncq_prio_enabled function.


Adam Manzanares (4):
  block: Add iocontext priority to request
  fusion: remove iopriority handling
  ata: Enabling ATA Command Priorities
  ata: ATA Command Priority Disabled By Default

 block/blk-core.c  |  4 ++-
 drivers/ata/libahci.c |  1 +
 drivers/ata/libata-core.c | 35 +-
 drivers/ata/libata-scsi.c | 74 ++-
 drivers/ata/libata.h  |  2 +-
 drivers/message/fusion/mptscsih.c |  5 ---
 include/linux/ata.h   |  6 
 include/linux/blkdev.h| 14 
 include/linux/libata.h| 26 ++
 9 files changed, 158 insertions(+), 9 deletions(-)

-- 
2.1.4

--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH v3 09/17] lpfc: Code cleanup for lpfc_topology parameter

[James Smart]

Code cleanup for lpfc_topology parameter

Signed-off-by: Dick Kennedy 
Signed-off-by: James Smart 
Reviewed-by: Johannes Thumshirn 
---
 drivers/scsi/lpfc/lpfc_attr.c | 7 +++
 1 file changed, 3 insertions(+), 4 deletions(-)

diff --git a/drivers/scsi/lpfc/lpfc_attr.c b/drivers/scsi/lpfc/lpfc_attr.c
index ae76aaa..81b1faf 100644
--- a/drivers/scsi/lpfc/lpfc_attr.c
+++ b/drivers/scsi/lpfc/lpfc_attr.c
@@ -3194,6 +3194,8 @@ LPFC_VPORT_ATTR_R(scan_down, 1, 0, 1,
 # Set loop mode if you want to run as an NL_Port. Value range is [0,0x6].
 # Default value is 0.
 */
+LPFC_ATTR(topology, 0, 0, 6,
+   "Select Fibre Channel topology");
 
 /**
  * lpfc_topology_set - Set the adapters topology field
@@ -3271,11 +3273,8 @@ lpfc_topology_store(struct device *dev, struct 
device_attribute *attr,
phba->brd_no, val);
return -EINVAL;
 }
-static int lpfc_topology = 0;
-module_param(lpfc_topology, int, S_IRUGO);
-MODULE_PARM_DESC(lpfc_topology, "Select Fibre Channel topology");
+
 lpfc_param_show(topology)
-lpfc_param_init(topology, 0, 0, 6)
 static DEVICE_ATTR(lpfc_topology, S_IRUGO | S_IWUSR,
lpfc_topology_show, lpfc_topology_store);
 
-- 
2.5.0


--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH v3 05/17] lpfc: Make lpfc_prot_xxx params per hba parameters

[James Smart]

Make lpfc_prot_mask and lpfc_prot_guard per hba parameters

Signed-off-by: Dick Kennedy 
Signed-off-by: James Smart 
Reviewed-by: Johannes Thumshirn 
---
 drivers/scsi/lpfc/lpfc.h  |  2 ++
 drivers/scsi/lpfc/lpfc_attr.c | 23 ++-
 drivers/scsi/lpfc/lpfc_crtn.h |  2 --
 drivers/scsi/lpfc/lpfc_init.c | 28 +++-
 4 files changed, 31 insertions(+), 24 deletions(-)

diff --git a/drivers/scsi/lpfc/lpfc.h b/drivers/scsi/lpfc/lpfc.h
index b484859..debba5e 100644
--- a/drivers/scsi/lpfc/lpfc.h
+++ b/drivers/scsi/lpfc/lpfc.h
@@ -746,6 +746,8 @@ struct lpfc_hba {
uint32_t cfg_oas_priority;
uint32_t cfg_XLanePriority;
uint32_t cfg_enable_bg;
+   uint32_t cfg_prot_mask;
+   uint32_t cfg_prot_guard;
uint32_t cfg_hostmem_hgp;
uint32_t cfg_log_verbose;
uint32_t cfg_aer_support;
diff --git a/drivers/scsi/lpfc/lpfc_attr.c b/drivers/scsi/lpfc/lpfc_attr.c
index f101990..be81e61 100644
--- a/drivers/scsi/lpfc/lpfc_attr.c
+++ b/drivers/scsi/lpfc/lpfc_attr.c
@@ -4691,12 +4691,15 @@ unsigned int lpfc_fcp_look_ahead = LPFC_LOOK_AHEAD_OFF;
 #  HBA supports DIX Type 1: Host to HBA  Type 1 protection
 #
 */
-unsigned int lpfc_prot_mask = SHOST_DIF_TYPE1_PROTECTION |
- SHOST_DIX_TYPE0_PROTECTION |
- SHOST_DIX_TYPE1_PROTECTION;
-
-module_param(lpfc_prot_mask, uint, S_IRUGO);
-MODULE_PARM_DESC(lpfc_prot_mask, "host protection mask");
+LPFC_ATTR(prot_mask,
+   (SHOST_DIF_TYPE1_PROTECTION |
+   SHOST_DIX_TYPE0_PROTECTION |
+   SHOST_DIX_TYPE1_PROTECTION),
+   0,
+   (SHOST_DIF_TYPE1_PROTECTION |
+   SHOST_DIX_TYPE0_PROTECTION |
+   SHOST_DIX_TYPE1_PROTECTION),
+   "T10-DIF host protection capabilities mask");
 
 /*
 # lpfc_prot_guard: i
@@ -4706,9 +4709,9 @@ MODULE_PARM_DESC(lpfc_prot_mask, "host protection mask");
 #  - Default will result in registering capabilities for all guard types
 #
 */
-unsigned char lpfc_prot_guard = SHOST_DIX_GUARD_IP;
-module_param(lpfc_prot_guard, byte, S_IRUGO);
-MODULE_PARM_DESC(lpfc_prot_guard, "host protection guard type");
+LPFC_ATTR(prot_guard,
+   SHOST_DIX_GUARD_IP, SHOST_DIX_GUARD_CRC, SHOST_DIX_GUARD_IP,
+   "T10-DIF host protection guard type");
 
 /*
  * Delay initial NPort discovery when Clean Address bit is cleared in
@@ -5828,6 +5831,8 @@ lpfc_get_cfgparam(struct lpfc_hba *phba)
phba->cfg_oas_flags = 0;
phba->cfg_oas_priority = 0;
lpfc_enable_bg_init(phba, lpfc_enable_bg);
+   lpfc_prot_mask_init(phba, lpfc_prot_mask);
+   lpfc_prot_guard_init(phba, lpfc_prot_guard);
if (phba->sli_rev == LPFC_SLI_REV4)
phba->cfg_poll = 0;
else
diff --git a/drivers/scsi/lpfc/lpfc_crtn.h b/drivers/scsi/lpfc/lpfc_crtn.h
index bd7576d..16195b7 100644
--- a/drivers/scsi/lpfc/lpfc_crtn.h
+++ b/drivers/scsi/lpfc/lpfc_crtn.h
@@ -397,8 +397,6 @@ extern spinlock_t _dump_buf_lock;
 extern int _dump_buf_done;
 extern spinlock_t pgcnt_lock;
 extern unsigned int pgcnt;
-extern unsigned int lpfc_prot_mask;
-extern unsigned char lpfc_prot_guard;
 extern unsigned int lpfc_fcp_look_ahead;
 
 /* Interface exported by fabric iocb scheduler */
diff --git a/drivers/scsi/lpfc/lpfc_init.c b/drivers/scsi/lpfc/lpfc_init.c
index af64d70..117c69a 100644
--- a/drivers/scsi/lpfc/lpfc_init.c
+++ b/drivers/scsi/lpfc/lpfc_init.c
@@ -6279,34 +6279,36 @@ lpfc_setup_bg(struct lpfc_hba *phba, struct Scsi_Host 
*shost)
uint32_t old_guard;
 
int pagecnt = 10;
-   if (lpfc_prot_mask && lpfc_prot_guard) {
+   if (phba->cfg_prot_mask && phba->cfg_prot_guard) {
lpfc_printf_log(phba, KERN_INFO, LOG_INIT,
"1478 Registering BlockGuard with the "
"SCSI layer\n");
 
-   old_mask = lpfc_prot_mask;
-   old_guard = lpfc_prot_guard;
+   old_mask = phba->cfg_prot_mask;
+   old_guard = phba->cfg_prot_guard;
 
/* Only allow supported values */
-   lpfc_prot_mask &= (SHOST_DIF_TYPE1_PROTECTION |
+   phba->cfg_prot_mask &= (SHOST_DIF_TYPE1_PROTECTION |
SHOST_DIX_TYPE0_PROTECTION |
SHOST_DIX_TYPE1_PROTECTION);
-   lpfc_prot_guard &= (SHOST_DIX_GUARD_IP | SHOST_DIX_GUARD_CRC);
+   phba->cfg_prot_guard &= (SHOST_DIX_GUARD_IP |
+SHOST_DIX_GUARD_CRC);
 
/* DIF Type 1 protection for profiles AST1/C1 is end to end */
-   if (lpfc_prot_mask == SHOST_DIX_TYPE1_PROTECTION)
-   lpfc_prot_mask |= SHOST_DIF_TYPE1_PROTECTION;
+   if (phba->cfg_prot_mask == SHOST_DIX_TYPE1_PROTECTION)
+   phba->cfg_prot_mask |= SHOST_DIF_TYPE1_PROTECTION;
 
-   if 

[PATCH v3 02/17] lpfc: Correct embedded io wq element size

[James Smart]

Correct embedded io wq element size. Embedded element sizes are
128 byte elements

Signed-off-by: Dick Kennedy 
Signed-off-by: James Smart 
Reviewed-by: Johannes Thumshirn 
---
v2 mods:
 change style of selecting wqesize using local variable

 drivers/scsi/lpfc/lpfc_init.c | 14 +-
 1 file changed, 5 insertions(+), 9 deletions(-)

diff --git a/drivers/scsi/lpfc/lpfc_init.c b/drivers/scsi/lpfc/lpfc_init.c
index 734a042..af64d70 100644
--- a/drivers/scsi/lpfc/lpfc_init.c
+++ b/drivers/scsi/lpfc/lpfc_init.c
@@ -7256,6 +7256,7 @@ int
 lpfc_sli4_queue_create(struct lpfc_hba *phba)
 {
struct lpfc_queue *qdesc;
+   uint32_t wqesize;
int idx;
 
/*
@@ -7340,15 +7341,10 @@ lpfc_sli4_queue_create(struct lpfc_hba *phba)
phba->sli4_hba.fcp_cq[idx] = qdesc;
 
/* Create Fast Path FCP WQs */
-   if (phba->fcp_embed_io) {
-   qdesc = lpfc_sli4_queue_alloc(phba,
- LPFC_WQE128_SIZE,
- LPFC_WQE128_DEF_COUNT);
-   } else {
-   qdesc = lpfc_sli4_queue_alloc(phba,
- phba->sli4_hba.wq_esize,
- phba->sli4_hba.wq_ecount);
-   }
+   wqesize = (phba->fcp_embed_io) ?
+   LPFC_WQE128_SIZE : phba->sli4_hba.wq_esize;
+   qdesc = lpfc_sli4_queue_alloc(phba, wqesize,
+   phba->sli4_hba.wq_ecount);
if (!qdesc) {
lpfc_printf_log(phba, KERN_ERR, LOG_INIT,
"0503 Failed allocate fast-path FCP "
-- 
2.5.0


--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH v3 00/17] lpfc: Update driver to version 11.2.0.2

[James Smart]

This patch set updates the lpfc driver to revision 11.2.0.2

The patches were cut against scsi-misc

James Smart (17):
  Fix few small typos in lpfc_scsi.c
   originally posted by Milan Gandhi
  lpfc: Correct embedded io wq element size
  lpfc: Fix sg_reset on SCSI device causing kernel crash
  lpfc: Set driver environment data on adapter
  lpfc: Make lpfc_prot_mask and lpfc_prot_guard per hba parameters
  lpfc: Code clean up for lpfc_iocb_cnt parameter
  lpfc: Code cleanup for lpfc_enable_rrq parameter
  lpfc: Code cleanup for lpfc_aer_support parameter
  lpfc: Code cleanup for lpfc_topology parameter
  lpfc: Code cleanup for lpfc_max_scsicmpl_time parameter
  lpfc: Code cleanup for lpfc_sriov_nr_virtfn parameter
  lpfc: Revise strings with full lpfc parameter name
  lpfc: Fix lost target in pt-to-pt connect
  lpfc: Correct panics with eh_timeout and eh_deadline
  lpfc: Synchronize link speed with boot driver
  lpfc: Fix fw download on SLI-4 FC adapters
  lpfc: lpfc version changed to 11.2.0.2

 drivers/scsi/lpfc/lpfc.h |   6 ++
 drivers/scsi/lpfc/lpfc_attr.c| 160 +++
 drivers/scsi/lpfc/lpfc_bsg.c |  45 +++
 drivers/scsi/lpfc/lpfc_bsg.h |  10 +++
 drivers/scsi/lpfc/lpfc_crtn.h|   2 -
 drivers/scsi/lpfc/lpfc_els.c |   2 +-
 drivers/scsi/lpfc/lpfc_hw4.h |  18 -
 drivers/scsi/lpfc/lpfc_init.c| 116 
 drivers/scsi/lpfc/lpfc_scsi.c|  49 +++-
 drivers/scsi/lpfc/lpfc_sli.c |  41 +-
 drivers/scsi/lpfc/lpfc_version.h |   2 +-
 11 files changed, 276 insertions(+), 175 deletions(-)

-- 
2.5.0


--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH v3 07/17] lpfc: Code cleanup for lpfc_enable_rrq parameter

[James Smart]

Code cleanup for lpfc_enable_rrq parameter

Signed-off-by: Dick Kennedy 
Signed-off-by: James Smart 
Reviewed-by: Johannes Thumshirn 
---
 drivers/scsi/lpfc/lpfc_attr.c | 8 ++--
 1 file changed, 2 insertions(+), 6 deletions(-)

diff --git a/drivers/scsi/lpfc/lpfc_attr.c b/drivers/scsi/lpfc/lpfc_attr.c
index e5092dc..22a66c7 100644
--- a/drivers/scsi/lpfc/lpfc_attr.c
+++ b/drivers/scsi/lpfc/lpfc_attr.c
@@ -2759,18 +2759,14 @@ LPFC_ATTR_R(enable_npiv, 1, 0, 1,
 LPFC_ATTR_R(fcf_failover_policy, 1, 1, 2,
"FCF Fast failover=1 Priority failover=2");
 
-int lpfc_enable_rrq = 2;
-module_param(lpfc_enable_rrq, int, S_IRUGO);
-MODULE_PARM_DESC(lpfc_enable_rrq, "Enable RRQ functionality");
-lpfc_param_show(enable_rrq);
 /*
 # lpfc_enable_rrq: Track XRI/OXID reuse after IO failures
 #  0x0 = disabled, XRI/OXID use not tracked.
 #  0x1 = XRI/OXID reuse is timed with ratov, RRQ sent.
 #  0x2 = XRI/OXID reuse is timed with ratov, No RRQ sent.
 */
-lpfc_param_init(enable_rrq, 2, 0, 2);
-static DEVICE_ATTR(lpfc_enable_rrq, S_IRUGO, lpfc_enable_rrq_show, NULL);
+LPFC_ATTR_R(enable_rrq, 2, 0, 2,
+   "Enable RRQ functionality");
 
 /*
 # lpfc_suppress_link_up:  Bring link up at initialization
-- 
2.5.0


--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH v3 08/17] lpfc: Code cleanup for lpfc_aer_support parameter

[James Smart]

Code cleanup for lpfc_aer_support parameter

Signed-off-by: Dick Kennedy 
Signed-off-by: James Smart 
Reviewed-by: Johannes Thumshirn 
---
 drivers/scsi/lpfc/lpfc_attr.c | 43 +++
 1 file changed, 3 insertions(+), 40 deletions(-)

diff --git a/drivers/scsi/lpfc/lpfc_attr.c b/drivers/scsi/lpfc/lpfc_attr.c
index 22a66c7..ae76aaa 100644
--- a/drivers/scsi/lpfc/lpfc_attr.c
+++ b/drivers/scsi/lpfc/lpfc_attr.c
@@ -3779,6 +3779,9 @@ static DEVICE_ATTR(lpfc_link_speed, S_IRUGO | S_IWUSR,
 #   1  = aer supported and enabled (default)
 # Value range is [0,1]. Default value is 1.
 */
+LPFC_ATTR(aer_support, 1, 0, 1,
+   "Enable PCIe device AER support");
+lpfc_param_show(aer_support)
 
 /**
  * lpfc_aer_support_store - Set the adapter for aer support
@@ -3861,46 +3864,6 @@ lpfc_aer_support_store(struct device *dev, struct 
device_attribute *attr,
return rc;
 }
 
-static int lpfc_aer_support = 1;
-module_param(lpfc_aer_support, int, S_IRUGO);
-MODULE_PARM_DESC(lpfc_aer_support, "Enable PCIe device AER support");
-lpfc_param_show(aer_support)
-
-/**
- * lpfc_aer_support_init - Set the initial adapters aer support flag
- * @phba: lpfc_hba pointer.
- * @val: enable aer or disable aer flag.
- *
- * Description:
- * If val is in a valid range [0,1], then set the adapter's initial
- * cfg_aer_support field. It will be up to the driver's probe_one
- * routine to determine whether the device's AER support can be set
- * or not.
- *
- * Notes:
- * If the value is not in range log a kernel error message, and
- * choose the default value of setting AER support and return.
- *
- * Returns:
- * zero if val saved.
- * -EINVAL val out of range
- **/
-static int
-lpfc_aer_support_init(struct lpfc_hba *phba, int val)
-{
-   if (val == 0 || val == 1) {
-   phba->cfg_aer_support = val;
-   return 0;
-   }
-   lpfc_printf_log(phba, KERN_ERR, LOG_INIT,
-   "2712 lpfc_aer_support attribute value %d out "
-   "of range, allowed values are 0|1, setting it "
-   "to default value of 1\n", val);
-   /* By default, try to enable AER on a device */
-   phba->cfg_aer_support = 1;
-   return -EINVAL;
-}
-
 static DEVICE_ATTR(lpfc_aer_support, S_IRUGO | S_IWUSR,
   lpfc_aer_support_show, lpfc_aer_support_store);
 
-- 
2.5.0


--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH v3 03/17] lpfc: Fix sg_reset on SCSI device causing kernel crash

[James Smart]

Fix sg_reset on SCSI device causing kernel crash

Driver could reference stale node pointers in task mgmt call.
Changed to use resetting cmd and look up node pointer in task mgmt
function.

Signed-off-by: Dick Kennedy 
Signed-off-by: James Smart 
Reviewed-by: Johannes Thumshirn 
---
v2 mods:
 Revised lpfc_send_taskmgmt() to check rdata

 drivers/scsi/lpfc/lpfc_scsi.c | 22 +-
 1 file changed, 13 insertions(+), 9 deletions(-)

diff --git a/drivers/scsi/lpfc/lpfc_scsi.c b/drivers/scsi/lpfc/lpfc_scsi.c
index ca56f28..4c53149e 100644
--- a/drivers/scsi/lpfc/lpfc_scsi.c
+++ b/drivers/scsi/lpfc/lpfc_scsi.c
@@ -4948,26 +4948,30 @@ lpfc_check_fcp_rsp(struct lpfc_vport *vport, struct 
lpfc_scsi_buf *lpfc_cmd)
  *   0x2002 - Success.
  **/
 static int
-lpfc_send_taskmgmt(struct lpfc_vport *vport, struct lpfc_rport_data *rdata,
-   unsigned  tgt_id, uint64_t lun_id,
-   uint8_t task_mgmt_cmd)
+lpfc_send_taskmgmt(struct lpfc_vport *vport, struct scsi_cmnd *cmnd,
+  unsigned int tgt_id, uint64_t lun_id,
+  uint8_t task_mgmt_cmd)
 {
struct lpfc_hba   *phba = vport->phba;
struct lpfc_scsi_buf *lpfc_cmd;
struct lpfc_iocbq *iocbq;
struct lpfc_iocbq *iocbqrsp;
-   struct lpfc_nodelist *pnode = rdata->pnode;
+   struct lpfc_rport_data *rdata;
+   struct lpfc_nodelist *pnode;
int ret;
int status;
 
-   if (!pnode || !NLP_CHK_NODE_ACT(pnode))
+   rdata = lpfc_rport_data_from_scsi_device(cmnd->device);
+   if (!rdata || !rdata->pnode || !NLP_CHK_NODE_ACT(rdata->pnode))
return FAILED;
+   pnode = rdata->pnode;
 
-   lpfc_cmd = lpfc_get_scsi_buf(phba, rdata->pnode);
+   lpfc_cmd = lpfc_get_scsi_buf(phba, pnode);
if (lpfc_cmd == NULL)
return FAILED;
lpfc_cmd->timeout = phba->cfg_task_mgmt_tmo;
lpfc_cmd->rdata = rdata;
+   lpfc_cmd->pCmd = cmnd;
 
status = lpfc_scsi_prep_task_mgmt_cmd(vport, lpfc_cmd, lun_id,
   task_mgmt_cmd);
@@ -5174,7 +5178,7 @@ lpfc_device_reset_handler(struct scsi_cmnd *cmnd)
fc_host_post_vendor_event(shost, fc_get_event_number(),
sizeof(scsi_event), (char *)_event, LPFC_NL_VENDOR_ID);
 
-   status = lpfc_send_taskmgmt(vport, rdata, tgt_id, lun_id,
+   status = lpfc_send_taskmgmt(vport, cmnd, tgt_id, lun_id,
FCP_LUN_RESET);
 
lpfc_printf_vlog(vport, KERN_ERR, LOG_FCP,
@@ -5252,7 +5256,7 @@ lpfc_target_reset_handler(struct scsi_cmnd *cmnd)
fc_host_post_vendor_event(shost, fc_get_event_number(),
sizeof(scsi_event), (char *)_event, LPFC_NL_VENDOR_ID);
 
-   status = lpfc_send_taskmgmt(vport, rdata, tgt_id, lun_id,
+   status = lpfc_send_taskmgmt(vport, cmnd, tgt_id, lun_id,
FCP_TARGET_RESET);
 
lpfc_printf_vlog(vport, KERN_ERR, LOG_FCP,
@@ -5331,7 +5335,7 @@ lpfc_bus_reset_handler(struct scsi_cmnd *cmnd)
if (!match)
continue;
 
-   status = lpfc_send_taskmgmt(vport, ndlp->rport->dd_data,
+   status = lpfc_send_taskmgmt(vport, cmnd,
i, 0, FCP_TARGET_RESET);
 
if (status != SUCCESS) {
-- 
2.5.0


--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH v3 01/17] Fix few small typos in lpfc_scsi.c

[James Smart]

From: "Milan P. Gandhi" 

This patch does a cleanup and fixes few small typos in lpfc_scsi.c

Signed-off-by: Milan P. Gandhi 
Signed-off-by: James Smart 
Reviewed-by: Johannes Thumshirn 
---
 
 drivers/scsi/lpfc/lpfc_scsi.c | 21 -
 1 file changed, 12 insertions(+), 9 deletions(-)

diff --git a/drivers/scsi/lpfc/lpfc_scsi.c b/drivers/scsi/lpfc/lpfc_scsi.c
index d197aa1..ca56f28 100644
--- a/drivers/scsi/lpfc/lpfc_scsi.c
+++ b/drivers/scsi/lpfc/lpfc_scsi.c
@@ -607,7 +607,7 @@ lpfc_sli4_fcp_xri_aborted(struct lpfc_hba *phba,
 }
 
 /**
- * lpfc_sli4_post_scsi_sgl_list - Psot blocks of scsi buffer sgls from a list
+ * lpfc_sli4_post_scsi_sgl_list - Post blocks of scsi buffer sgls from a list
  * @phba: pointer to lpfc hba data structure.
  * @post_sblist: pointer to the scsi buffer list.
  *
@@ -736,7 +736,7 @@ lpfc_sli4_post_scsi_sgl_list(struct lpfc_hba *phba,
 }
 
 /**
- * lpfc_sli4_repost_scsi_sgl_list - Repsot all the allocated scsi buffer sgls
+ * lpfc_sli4_repost_scsi_sgl_list - Repost all the allocated scsi buffer sgls
  * @phba: pointer to lpfc hba data structure.
  *
  * This routine walks the list of scsi buffers that have been allocated and
@@ -857,7 +857,7 @@ lpfc_new_scsi_buf_s4(struct lpfc_vport *vport, int 
num_to_alloc)
psb->data, psb->dma_handle);
kfree(psb);
lpfc_printf_log(phba, KERN_ERR, LOG_FCP,
-   "3368 Failed to allocated IOTAG for"
+   "3368 Failed to allocate IOTAG for"
" XRI:0x%x\n", lxri);
lpfc_sli4_free_xri(phba, lxri);
break;
@@ -1136,7 +1136,7 @@ lpfc_release_scsi_buf(struct lpfc_hba *phba, struct 
lpfc_scsi_buf *psb)
  *
  * This routine does the pci dma mapping for scatter-gather list of scsi cmnd
  * field of @lpfc_cmd for device with SLI-3 interface spec. This routine scans
- * through sg elements and format the bdea. This routine also initializes all
+ * through sg elements and format the bde. This routine also initializes all
  * IOCB fields which are dependent on scsi command request buffer.
  *
  * Return codes:
@@ -1269,13 +1269,16 @@ lpfc_scsi_prep_dma_buf_s3(struct lpfc_hba *phba, struct 
lpfc_scsi_buf *lpfc_cmd)
 
 #ifdef CONFIG_SCSI_LPFC_DEBUG_FS
 
-/* Return if if error injection is detected by Initiator */
+/* Return BG_ERR_INIT if error injection is detected by Initiator */
 #define BG_ERR_INIT0x1
-/* Return if if error injection is detected by Target */
+/* Return BG_ERR_TGT if error injection is detected by Target */
 #define BG_ERR_TGT 0x2
-/* Return if if swapping CSUM<-->CRC is required for error injection */
+/* Return BG_ERR_SWAP if swapping CSUM<-->CRC is required for error injection 
*/
 #define BG_ERR_SWAP0x10
-/* Return if disabling Guard/Ref/App checking is required for error injection 
*/
+/**
+ * Return BG_ERR_CHECK if disabling Guard/Ref/App checking is required for
+ * error injection
+ **/
 #define BG_ERR_CHECK   0x20
 
 /**
@@ -4822,7 +4825,7 @@ wait_for_cmpl:
ret = FAILED;
lpfc_printf_vlog(vport, KERN_ERR, LOG_FCP,
 "0748 abort handler timed out waiting "
-"for abortng I/O (xri:x%x) to complete: "
+"for aborting I/O (xri:x%x) to complete: "
 "ret %#x, ID %d, LUN %llu\n",
 iocb->sli4_xritag, ret,
 cmnd->device->id, cmnd->device->lun);

--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH v3 06/17] lpfc: Code clean up for lpfc_iocb_cnt parameter

[James Smart]

Code clean up for lpfc_iocb_cnt parameter

Signed-off-by: Dick Kennedy 
Signed-off-by: James Smart 
Reviewed-by: Johannes Thumshirn 
---
 drivers/scsi/lpfc/lpfc_attr.c | 8 +---
 1 file changed, 1 insertion(+), 7 deletions(-)

diff --git a/drivers/scsi/lpfc/lpfc_attr.c b/drivers/scsi/lpfc/lpfc_attr.c
index be81e61..e5092dc 100644
--- a/drivers/scsi/lpfc/lpfc_attr.c
+++ b/drivers/scsi/lpfc/lpfc_attr.c
@@ -2827,14 +2827,8 @@ lpfc_txcmplq_hw_show(struct device *dev, struct 
device_attribute *attr,
 static DEVICE_ATTR(txcmplq_hw, S_IRUGO,
 lpfc_txcmplq_hw_show, NULL);
 
-int lpfc_iocb_cnt = 2;
-module_param(lpfc_iocb_cnt, int, S_IRUGO);
-MODULE_PARM_DESC(lpfc_iocb_cnt,
+LPFC_ATTR_R(iocb_cnt, 2, 1, 5,
"Number of IOCBs alloc for ELS, CT, and ABTS: 1k to 5k IOCBs");
-lpfc_param_show(iocb_cnt);
-lpfc_param_init(iocb_cnt, 2, 1, 5);
-static DEVICE_ATTR(lpfc_iocb_cnt, S_IRUGO,
-lpfc_iocb_cnt_show, NULL);
 
 /*
 # lpfc_nodev_tmo: If set, it will hold all I/O errors on devices that disappear
-- 
2.5.0


--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH v3 04/17] lpfc: Set driver environment data on adapter

[James Smart]

Set driver environment data on adapter

Signed-off-by: Dick Kennedy 
Signed-off-by: James Smart 
Reviewed-by: Johannes Thumshirn 
---
 drivers/scsi/lpfc/lpfc_hw4.h | 12 
 drivers/scsi/lpfc/lpfc_sli.c | 29 +
 2 files changed, 41 insertions(+)

diff --git a/drivers/scsi/lpfc/lpfc_hw4.h b/drivers/scsi/lpfc/lpfc_hw4.h
index ee80227..0b2c337 100644
--- a/drivers/scsi/lpfc/lpfc_hw4.h
+++ b/drivers/scsi/lpfc/lpfc_hw4.h
@@ -921,6 +921,7 @@ struct mbox_header {
 #define LPFC_MBOX_OPCODE_GET_PORT_NAME 0x4D
 #define LPFC_MBOX_OPCODE_MQ_CREATE_EXT 0x5A
 #define LPFC_MBOX_OPCODE_GET_VPD_DATA  0x5B
+#define LPFC_MBOX_OPCODE_SET_HOST_DATA 0x5D
 #define LPFC_MBOX_OPCODE_SEND_ACTIVATION   0x73
 #define LPFC_MBOX_OPCODE_RESET_LICENSES0x74
 #define LPFC_MBOX_OPCODE_GET_RSRC_EXTENT_INFO  0x9A
@@ -2919,6 +2920,16 @@ struct lpfc_mbx_set_feature {
 };
 
 
+#define LPFC_SET_HOST_OS_DRIVER_VERSION0x2
+struct lpfc_mbx_set_host_data {
+#define LPFC_HOST_OS_DRIVER_VERSION_SIZE   48
+   struct mbox_header header;
+   uint32_t param_id;
+   uint32_t param_len;
+   uint8_t  data[LPFC_HOST_OS_DRIVER_VERSION_SIZE];
+};
+
+
 struct lpfc_mbx_get_sli4_parameters {
struct mbox_header header;
struct lpfc_sli4_parameters sli4_parameters;
@@ -3313,6 +3324,7 @@ struct lpfc_mqe {
struct lpfc_mbx_get_port_name get_port_name;
struct lpfc_mbx_set_feature  set_feature;
struct lpfc_mbx_memory_dump_type3 mem_dump_type3;
+   struct lpfc_mbx_set_host_data set_host_data;
struct lpfc_mbx_nop nop;
} un;
 };
diff --git a/drivers/scsi/lpfc/lpfc_sli.c b/drivers/scsi/lpfc/lpfc_sli.c
index c532605..cb59f4e 100644
--- a/drivers/scsi/lpfc/lpfc_sli.c
+++ b/drivers/scsi/lpfc/lpfc_sli.c
@@ -47,6 +47,7 @@
 #include "lpfc_compat.h"
 #include "lpfc_debugfs.h"
 #include "lpfc_vport.h"
+#include "lpfc_version.h"
 
 /* There are only four IOCB completion types. */
 typedef enum _lpfc_iocb_type {
@@ -6289,6 +6290,25 @@ lpfc_sli4_repost_els_sgl_list(struct lpfc_hba *phba)
return 0;
 }
 
+void
+lpfc_set_host_data(struct lpfc_hba *phba, LPFC_MBOXQ_t *mbox)
+{
+   uint32_t len;
+
+   len = sizeof(struct lpfc_mbx_set_host_data) -
+   sizeof(struct lpfc_sli4_cfg_mhdr);
+   lpfc_sli4_config(phba, mbox, LPFC_MBOX_SUBSYSTEM_COMMON,
+LPFC_MBOX_OPCODE_SET_HOST_DATA, len,
+LPFC_SLI4_MBX_EMBED);
+
+   mbox->u.mqe.un.set_host_data.param_id = LPFC_SET_HOST_OS_DRIVER_VERSION;
+   mbox->u.mqe.un.set_host_data.param_len = 8;
+   snprintf(mbox->u.mqe.un.set_host_data.data,
+LPFC_HOST_OS_DRIVER_VERSION_SIZE,
+"Linux %s v"LPFC_DRIVER_VERSION,
+(phba->hba_flag & HBA_FCOE_MODE) ? "FCoE" : "FC");
+}
+
 /**
  * lpfc_sli4_hba_setup - SLI4 device intialization PCI function
  * @phba: Pointer to HBA context object.
@@ -6540,6 +6560,15 @@ lpfc_sli4_hba_setup(struct lpfc_hba *phba)
goto out_free_mbox;
}
 
+   lpfc_set_host_data(phba, mboxq);
+
+   rc = lpfc_sli_issue_mbox(phba, mboxq, MBX_POLL);
+   if (rc) {
+   lpfc_printf_log(phba, KERN_WARNING, LOG_MBOX | LOG_SLI,
+   "2134 Failed to set host os driver version %x",
+   rc);
+   }
+
/* Read the port's service parameters. */
rc = lpfc_read_sparam(phba, mboxq, vport->vpi);
if (rc) {
-- 
2.5.0


--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH v3 11/17] lpfc: Code cleanup for lpfc_sriov_nr_virtfn parameter

[James Smart]

Code cleanup for lpfc_sriov_nr_virtfn parameter

Signed-off-by: Dick Kennedy 
Signed-off-by: James Smart 
Reviewed-by: Johannes Thumshirn 
---
 drivers/scsi/lpfc/lpfc_attr.c | 35 +++
 1 file changed, 3 insertions(+), 32 deletions(-)

diff --git a/drivers/scsi/lpfc/lpfc_attr.c b/drivers/scsi/lpfc/lpfc_attr.c
index 3051e66..1b6090c 100644
--- a/drivers/scsi/lpfc/lpfc_attr.c
+++ b/drivers/scsi/lpfc/lpfc_attr.c
@@ -4007,39 +4007,10 @@ lpfc_sriov_nr_virtfn_store(struct device *dev, struct 
device_attribute *attr,
return rc;
 }
 
-static int lpfc_sriov_nr_virtfn = LPFC_DEF_VFN_PER_PFN;
-module_param(lpfc_sriov_nr_virtfn, int, S_IRUGO|S_IWUSR);
-MODULE_PARM_DESC(lpfc_sriov_nr_virtfn, "Enable PCIe device SR-IOV virtual fn");
-lpfc_param_show(sriov_nr_virtfn)
+LPFC_ATTR(sriov_nr_virtfn, LPFC_DEF_VFN_PER_PFN, 0, LPFC_MAX_VFN_PER_PFN,
+   "Enable PCIe device SR-IOV virtual fn");
 
-/**
- * lpfc_sriov_nr_virtfn_init - Set the initial sr-iov virtual function enable
- * @phba: lpfc_hba pointer.
- * @val: link speed value.
- *
- * Description:
- * If val is in a valid range [0,255], then set the adapter's initial
- * cfg_sriov_nr_virtfn field. If it's greater than the maximum, the maximum
- * number shall be used instead. It will be up to the driver's probe_one
- * routine to determine whether the device's SR-IOV is supported or not.
- *
- * Returns:
- * zero if val saved.
- * -EINVAL val out of range
- **/
-static int
-lpfc_sriov_nr_virtfn_init(struct lpfc_hba *phba, int val)
-{
-   if (val >= 0 && val <= LPFC_MAX_VFN_PER_PFN) {
-   phba->cfg_sriov_nr_virtfn = val;
-   return 0;
-   }
-
-   lpfc_printf_log(phba, KERN_ERR, LOG_INIT,
-   "3017 Enabling %d virtual functions is not "
-   "allowed.\n", val);
-   return -EINVAL;
-}
+lpfc_param_show(sriov_nr_virtfn)
 static DEVICE_ATTR(lpfc_sriov_nr_virtfn, S_IRUGO | S_IWUSR,
   lpfc_sriov_nr_virtfn_show, lpfc_sriov_nr_virtfn_store);
 
-- 
2.5.0


--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH v3 15/17] lpfc: Synchronize link speed with boot driver

[James Smart]

Synchronize link speed with boot driver

Link speed settings set by the boot driver are reported by the hw.
Driver will attempt to read them, and if set, will respect their
values.
The driver can override the settings with its own if instructed by
user space (via bsg), with the new values being picked up by the
boot driver.

Signed-off-by: Dick Kennedy 
Signed-off-by: James Smart 
---
v3 mods:
 parens cleanup
 use shost_priv() on new code

 drivers/scsi/lpfc/lpfc.h  |  4 
 drivers/scsi/lpfc/lpfc_attr.c |  7 +-
 drivers/scsi/lpfc/lpfc_bsg.c  | 45 
 drivers/scsi/lpfc/lpfc_bsg.h  | 10 
 drivers/scsi/lpfc/lpfc_hw4.h  |  3 +++
 drivers/scsi/lpfc/lpfc_init.c | 54 +++
 6 files changed, 122 insertions(+), 1 deletion(-)

diff --git a/drivers/scsi/lpfc/lpfc.h b/drivers/scsi/lpfc/lpfc.h
index debba5e..8a20b4e 100644
--- a/drivers/scsi/lpfc/lpfc.h
+++ b/drivers/scsi/lpfc/lpfc.h
@@ -648,6 +648,10 @@ struct lpfc_hba {
 #define HBA_FCP_IOQ_FLUSH  0x8000 /* FCP I/O queues being flushed */
 #define HBA_FW_DUMP_OP 0x1 /* Skips fn reset before FW dump */
 #define HBA_RECOVERABLE_UE 0x2 /* Firmware supports recoverable UE */
+#define HBA_FORCED_LINK_SPEED  0x4 /*
+* Firmware supports Forced Link Speed
+* capability
+*/
uint32_t fcp_ring_in_use; /* When polling test if intr-hndlr active*/
struct lpfc_dmabuf slim2p;
 
diff --git a/drivers/scsi/lpfc/lpfc_attr.c b/drivers/scsi/lpfc/lpfc_attr.c
index 3740e5d..c847755 100644
--- a/drivers/scsi/lpfc/lpfc_attr.c
+++ b/drivers/scsi/lpfc/lpfc_attr.c
@@ -3668,7 +3668,12 @@ lpfc_link_speed_store(struct device *dev, struct 
device_attribute *attr,
int nolip = 0;
const char *val_buf = buf;
int err;
-   uint32_t prev_val;
+   uint32_t prev_val, if_type;
+
+   if_type = bf_get(lpfc_sli_intf_if_type, >sli4_hba.sli_intf);
+   if (if_type == LPFC_SLI_INTF_IF_TYPE_2 &&
+   phba->hba_flag & HBA_FORCED_LINK_SPEED)
+   return -EPERM;
 
if (!strncmp(buf, "nolip ", strlen("nolip "))) {
nolip = 1;
diff --git a/drivers/scsi/lpfc/lpfc_bsg.c b/drivers/scsi/lpfc/lpfc_bsg.c
index 05dcc2a..e6a5254 100644
--- a/drivers/scsi/lpfc/lpfc_bsg.c
+++ b/drivers/scsi/lpfc/lpfc_bsg.c
@@ -5185,6 +5185,48 @@ no_dd_data:
return rc;
 }
 
+static int
+lpfc_forced_link_speed(struct fc_bsg_job *job)
+{
+   struct Scsi_Host *shost = job->shost;
+   struct lpfc_vport *vport = shost_priv(shost);
+   struct lpfc_hba *phba = vport->phba;
+   struct forced_link_speed_support_reply *forced_reply;
+   int rc = 0;
+
+   if (job->request_len <
+   sizeof(struct fc_bsg_request) +
+   sizeof(struct get_forced_link_speed_support)) {
+   lpfc_printf_log(phba, KERN_WARNING, LOG_LIBDFC,
+   "0048 Received FORCED_LINK_SPEED request "
+   "below minimum size\n");
+   rc = -EINVAL;
+   goto job_error;
+   }
+
+   forced_reply = (struct forced_link_speed_support_reply *)
+   job->reply->reply_data.vendor_reply.vendor_rsp;
+
+   if (job->reply_len <
+   sizeof(struct fc_bsg_request) +
+   sizeof(struct forced_link_speed_support_reply)) {
+   lpfc_printf_log(phba, KERN_WARNING, LOG_LIBDFC,
+   "0049 Received FORCED_LINK_SPEED reply below "
+   "minimum size\n");
+   rc = -EINVAL;
+   goto job_error;
+   }
+
+   forced_reply->supported = (phba->hba_flag & HBA_FORCED_LINK_SPEED)
+  ? LPFC_FORCED_LINK_SPEED_SUPPORTED
+  : LPFC_FORCED_LINK_SPEED_NOT_SUPPORTED;
+job_error:
+   job->reply->result = rc;
+   if (rc == 0)
+   job->job_done(job);
+   return rc;
+}
+
 /**
  * lpfc_bsg_hst_vendor - process a vendor-specific fc_bsg_job
  * @job: fc_bsg_job to handle
@@ -5227,6 +5269,9 @@ lpfc_bsg_hst_vendor(struct fc_bsg_job *job)
case LPFC_BSG_VENDOR_MENLO_DATA:
rc = lpfc_menlo_cmd(job);
break;
+   case LPFC_BSG_VENDOR_FORCED_LINK_SPEED:
+   rc = lpfc_forced_link_speed(job);
+   break;
default:
rc = -EINVAL;
job->reply->reply_payload_rcv_len = 0;
diff --git a/drivers/scsi/lpfc/lpfc_bsg.h b/drivers/scsi/lpfc/lpfc_bsg.h
index e557bcd..f2247aa 100644
--- a/drivers/scsi/lpfc/lpfc_bsg.h
+++ b/drivers/scsi/lpfc/lpfc_bsg.h
@@ -35,6 +35,7 @@
 #define LPFC_BSG_VENDOR_MENLO_DATA 9
 #define LPFC_BSG_VENDOR_DIAG_MODE_END  10
 #define LPFC_BSG_VENDOR_LINK_DIAG_TEST 11
+#define 

[PATCH v3 10/17] lpfc: Code cleanup for lpfc_max_scsicmpl_time parameter

[James Smart]

Code cleanup for lpfc_max_scsicmpl_time parameter

Signed-off-by: Dick Kennedy 
Signed-off-by: James Smart 
Reviewed-by: Johannes Thumshirn 
---
 drivers/scsi/lpfc/lpfc_attr.c | 6 ++
 1 file changed, 2 insertions(+), 4 deletions(-)

diff --git a/drivers/scsi/lpfc/lpfc_attr.c b/drivers/scsi/lpfc/lpfc_attr.c
index 81b1faf..3051e66 100644
--- a/drivers/scsi/lpfc/lpfc_attr.c
+++ b/drivers/scsi/lpfc/lpfc_attr.c
@@ -4393,12 +4393,10 @@ LPFC_VPORT_ATTR_RW(first_burst_size, 0, 0, 65536,
 # to limit the I/O completion time to the parameter value.
 # The value is set in milliseconds.
 */
-static int lpfc_max_scsicmpl_time;
-module_param(lpfc_max_scsicmpl_time, int, S_IRUGO);
-MODULE_PARM_DESC(lpfc_max_scsicmpl_time,
+LPFC_VPORT_ATTR(max_scsicmpl_time, 0, 0, 6,
"Use command completion time to control queue depth");
+
 lpfc_vport_param_show(max_scsicmpl_time);
-lpfc_vport_param_init(max_scsicmpl_time, 0, 0, 6);
 static int
 lpfc_max_scsicmpl_time_set(struct lpfc_vport *vport, int val)
 {
-- 
2.5.0


--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH v3 13/17] lpfc: Fix lost target in pt-to-pt connect

[James Smart]

Fix lost target in pt-to-pt connect

Change reject code to something that allows a retry

Signed-off-by: Dick Kennedy 
Signed-off-by: James Smart 
Reviewed-by: Johannes Thumshirn 
---
 drivers/scsi/lpfc/lpfc_els.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/scsi/lpfc/lpfc_els.c b/drivers/scsi/lpfc/lpfc_els.c
index b7d54bf..236e4e5 100644
--- a/drivers/scsi/lpfc/lpfc_els.c
+++ b/drivers/scsi/lpfc/lpfc_els.c
@@ -7610,7 +7610,7 @@ lpfc_els_unsol_buffer(struct lpfc_hba *phba, struct 
lpfc_sli_ring *pring,
/* reject till our FLOGI completes */
if ((vport->port_state < LPFC_FABRIC_CFG_LINK) &&
(cmd != ELS_CMD_FLOGI)) {
-   rjt_err = LSRJT_UNABLE_TPC;
+   rjt_err = LSRJT_LOGICAL_BSY;
rjt_exp = LSEXP_NOTHING_MORE;
goto lsrjt;
}
-- 
2.5.0


--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH v3 16/17] lpfc: Fix fw download on SLI-4 FC adapters

[James Smart]

Fix fw download on SLI-4 FC adapters

Driver performs a quick validation of magic numbers in the fw
download image. Driver needed to be updated for more recent
magic numbers.

Signed-off-by: Dick Kennedy 
Signed-off-by: James Smart 
---
v3 mods:
 code cleanup: parsing image header fields

 drivers/scsi/lpfc/lpfc_hw4.h  |  3 ++-
 drivers/scsi/lpfc/lpfc_init.c | 20 +++-
 2 files changed, 13 insertions(+), 10 deletions(-)

diff --git a/drivers/scsi/lpfc/lpfc_hw4.h b/drivers/scsi/lpfc/lpfc_hw4.h
index bbdcb5a..5646699 100644
--- a/drivers/scsi/lpfc/lpfc_hw4.h
+++ b/drivers/scsi/lpfc/lpfc_hw4.h
@@ -3996,7 +3996,8 @@ union lpfc_wqe128 {
struct gen_req64_wqe gen_req;
 };
 
-#define LPFC_GROUP_OJECT_MAGIC_NUM 0xfeaa0001
+#define LPFC_GROUP_OJECT_MAGIC_G5  0xfeaa0001
+#define LPFC_GROUP_OJECT_MAGIC_G6  0xfeaa0003
 #define LPFC_FILE_TYPE_GROUP   0xf7
 #define LPFC_FILE_ID_GROUP 0xa2
 struct lpfc_grp_hdr {
diff --git a/drivers/scsi/lpfc/lpfc_init.c b/drivers/scsi/lpfc/lpfc_init.c
index 53227e5..7be9b8a 100644
--- a/drivers/scsi/lpfc/lpfc_init.c
+++ b/drivers/scsi/lpfc/lpfc_init.c
@@ -10312,6 +10312,7 @@ lpfc_write_firmware(const struct firmware *fw, void 
*context)
int i, rc = 0;
struct lpfc_dmabuf *dmabuf, *next;
uint32_t offset = 0, temp_offset = 0;
+   uint32_t magic_number, ftype, fid, fsize;
 
/* It can be null in no-wait mode, sanity check */
if (!fw) {
@@ -10320,18 +10321,19 @@ lpfc_write_firmware(const struct firmware *fw, void 
*context)
}
image = (struct lpfc_grp_hdr *)fw->data;
 
+   magic_number = be32_to_cpu(image->magic_number);
+   ftype = bf_get_be32(lpfc_grp_hdr_file_type, image);
+   fid = bf_get_be32(lpfc_grp_hdr_id, image),
+   fsize = be32_to_cpu(image->size);
+
INIT_LIST_HEAD(_buffer_list);
-   if ((be32_to_cpu(image->magic_number) != LPFC_GROUP_OJECT_MAGIC_NUM) ||
-   (bf_get_be32(lpfc_grp_hdr_file_type, image) !=
-LPFC_FILE_TYPE_GROUP) ||
-   (bf_get_be32(lpfc_grp_hdr_id, image) != LPFC_FILE_ID_GROUP) ||
-   (be32_to_cpu(image->size) != fw->size)) {
+   if ((magic_number != LPFC_GROUP_OJECT_MAGIC_G5 &&
+magic_number != LPFC_GROUP_OJECT_MAGIC_G6) ||
+   ftype != LPFC_FILE_TYPE_GROUP || fsize != fw->size) {
lpfc_printf_log(phba, KERN_ERR, LOG_INIT,
"3022 Invalid FW image found. "
-   "Magic:%x Type:%x ID:%x\n",
-   be32_to_cpu(image->magic_number),
-   bf_get_be32(lpfc_grp_hdr_file_type, image),
-   bf_get_be32(lpfc_grp_hdr_id, image));
+   "Magic:%x Type:%x ID:%x Size %d %ld\n",
+   magic_number, ftype, fid, fsize, fw->size);
rc = -EINVAL;
goto release_out;
}
-- 
2.5.0


--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH v3 17/17] lpfc: lpfc version changed to 11.2.0.2

[James Smart]

lpfc version changed to 11.2.0.2

Signed-off-by: Dick Kennedy 
Signed-off-by: James Smart 
Reviewed-by: Johannes Thumshirn 
---
 drivers/scsi/lpfc/lpfc_version.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/scsi/lpfc/lpfc_version.h b/drivers/scsi/lpfc/lpfc_version.h
index c9bf20e..50bfc43 100644
--- a/drivers/scsi/lpfc/lpfc_version.h
+++ b/drivers/scsi/lpfc/lpfc_version.h
@@ -18,7 +18,7 @@
  * included with this package. *
  ***/
 
-#define LPFC_DRIVER_VERSION "11.2.0.0."
+#define LPFC_DRIVER_VERSION "11.2.0.2"
 #define LPFC_DRIVER_NAME   "lpfc"
 
 /* Used for SLI 2/3 */
-- 
2.5.0


--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH v3 14/17] lpfc: Correct panics with eh_timeout and eh_deadline

[James Smart]

Correct panics with eh_timeout and eh_deadline

We were having double completions on our SLI-3 version of adapters.
Solved by clearing our command pointer before calling scsi_done.

The eh paths potentially ran simulatenously and would see the non-null
value and invoke scsi_done again.

Signed-off-by: Dick Kennedy 
Signed-off-by: James Smart 
Reviewed-by: Johannes Thumshirn 
---
 drivers/scsi/lpfc/lpfc_scsi.c |  6 +++---
 drivers/scsi/lpfc/lpfc_sli.c  | 12 
 2 files changed, 11 insertions(+), 7 deletions(-)

diff --git a/drivers/scsi/lpfc/lpfc_scsi.c b/drivers/scsi/lpfc/lpfc_scsi.c
index 4c53149e..1b0ef79 100644
--- a/drivers/scsi/lpfc/lpfc_scsi.c
+++ b/drivers/scsi/lpfc/lpfc_scsi.c
@@ -4142,13 +4142,13 @@ lpfc_scsi_cmd_iocb_cmpl(struct lpfc_hba *phba, struct 
lpfc_iocbq *pIocbIn,
 
lpfc_scsi_unprep_dma_buf(phba, lpfc_cmd);
 
-   /* The sdev is not guaranteed to be valid post scsi_done upcall. */
-   cmd->scsi_done(cmd);
-
spin_lock_irqsave(>hbalock, flags);
lpfc_cmd->pCmd = NULL;
spin_unlock_irqrestore(>hbalock, flags);
 
+   /* The sdev is not guaranteed to be valid post scsi_done upcall. */
+   cmd->scsi_done(cmd);
+
/*
 * If there is a thread waiting for command completion
 * wake up the thread.
diff --git a/drivers/scsi/lpfc/lpfc_sli.c b/drivers/scsi/lpfc/lpfc_sli.c
index cb59f4e..27cbd68 100644
--- a/drivers/scsi/lpfc/lpfc_sli.c
+++ b/drivers/scsi/lpfc/lpfc_sli.c
@@ -2677,15 +2677,16 @@ lpfc_sli_iocbq_lookup(struct lpfc_hba *phba,
 
if (iotag != 0 && iotag <= phba->sli.last_iotag) {
cmd_iocb = phba->sli.iocbq_lookup[iotag];
-   list_del_init(_iocb->list);
if (cmd_iocb->iocb_flag & LPFC_IO_ON_TXCMPLQ) {
+   /* remove from txcmpl queue list */
+   list_del_init(_iocb->list);
cmd_iocb->iocb_flag &= ~LPFC_IO_ON_TXCMPLQ;
+   return cmd_iocb;
}
-   return cmd_iocb;
}
 
lpfc_printf_log(phba, KERN_ERR, LOG_SLI,
-   "0317 iotag x%x is out off "
+   "0317 iotag x%x is out of "
"range: max iotag x%x wd0 x%x\n",
iotag, phba->sli.last_iotag,
*(((uint32_t *) >iocb) + 7));
@@ -2720,8 +2721,9 @@ lpfc_sli_iocbq_lookup_by_tag(struct lpfc_hba *phba,
return cmd_iocb;
}
}
+
lpfc_printf_log(phba, KERN_ERR, LOG_SLI,
-   "0372 iotag x%x is out off range: max iotag (x%x)\n",
+   "0372 iotag x%x is out of range: max iotag (x%x)\n",
iotag, phba->sli.last_iotag);
return NULL;
 }
@@ -11808,6 +11810,8 @@ lpfc_sli4_els_wcqe_to_rspiocbq(struct lpfc_hba *phba,
/* Look up the ELS command IOCB and create pseudo response IOCB */
cmdiocbq = lpfc_sli_iocbq_lookup_by_tag(phba, pring,
bf_get(lpfc_wcqe_c_request_tag, wcqe));
+   /* Put the iocb back on the txcmplq */
+   lpfc_sli_ringtxcmpl_put(phba, pring, cmdiocbq);
spin_unlock_irqrestore(>ring_lock, iflags);
 
if (unlikely(!cmdiocbq)) {
-- 
2.5.0


--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH v3 12/17] lpfc: Revise strings with full lpfc parameter name

[James Smart]

Revise strings with full lpfc parameter name

Signed-off-by: Dick Kennedy 
Signed-off-by: James Smart 
Reviewed-by: Johannes Thumshirn 
---
 drivers/scsi/lpfc/lpfc_attr.c | 23 ---
 1 file changed, 12 insertions(+), 11 deletions(-)

diff --git a/drivers/scsi/lpfc/lpfc_attr.c b/drivers/scsi/lpfc/lpfc_attr.c
index 1b6090c..3740e5d 100644
--- a/drivers/scsi/lpfc/lpfc_attr.c
+++ b/drivers/scsi/lpfc/lpfc_attr.c
@@ -2877,9 +2877,9 @@ lpfc_nodev_tmo_init(struct lpfc_vport *vport, int val)
vport->cfg_nodev_tmo = vport->cfg_devloss_tmo;
if (val != LPFC_DEF_DEVLOSS_TMO)
lpfc_printf_vlog(vport, KERN_ERR, LOG_INIT,
-"0407 Ignoring nodev_tmo module "
-"parameter because devloss_tmo is "
-"set.\n");
+"0407 Ignoring lpfc_nodev_tmo module "
+"parameter because lpfc_devloss_tmo "
+"is set.\n");
return 0;
}
 
@@ -2938,8 +2938,8 @@ lpfc_nodev_tmo_set(struct lpfc_vport *vport, int val)
if (vport->dev_loss_tmo_changed ||
(lpfc_devloss_tmo != LPFC_DEF_DEVLOSS_TMO)) {
lpfc_printf_vlog(vport, KERN_ERR, LOG_INIT,
-"0401 Ignoring change to nodev_tmo "
-"because devloss_tmo is set.\n");
+"0401 Ignoring change to lpfc_nodev_tmo "
+"because lpfc_devloss_tmo is set.\n");
return 0;
}
if (val >= LPFC_MIN_DEVLOSS_TMO && val <= LPFC_MAX_DEVLOSS_TMO) {
@@ -2954,7 +2954,7 @@ lpfc_nodev_tmo_set(struct lpfc_vport *vport, int val)
return 0;
}
lpfc_printf_vlog(vport, KERN_ERR, LOG_INIT,
-"0403 lpfc_nodev_tmo attribute cannot be set to"
+"0403 lpfc_nodev_tmo attribute cannot be set to "
 "%d, allowed range is [%d, %d]\n",
 val, LPFC_MIN_DEVLOSS_TMO, LPFC_MAX_DEVLOSS_TMO);
return -EINVAL;
@@ -3005,8 +3005,8 @@ lpfc_devloss_tmo_set(struct lpfc_vport *vport, int val)
}
 
lpfc_printf_vlog(vport, KERN_ERR, LOG_INIT,
-"0404 lpfc_devloss_tmo attribute cannot be set to"
-" %d, allowed range is [%d, %d]\n",
+"0404 lpfc_devloss_tmo attribute cannot be set to "
+"%d, allowed range is [%d, %d]\n",
 val, LPFC_MIN_DEVLOSS_TMO, LPFC_MAX_DEVLOSS_TMO);
return -EINVAL;
 }
@@ -4174,7 +4174,8 @@ lpfc_fcp_imax_init(struct lpfc_hba *phba, int val)
}
 
lpfc_printf_log(phba, KERN_ERR, LOG_INIT,
-   "3016 fcp_imax: %d out of range, using default\n", val);
+   "3016 lpfc_fcp_imax: %d out of range, using default\n",
+   val);
phba->cfg_fcp_imax = LPFC_DEF_IMAX;
 
return 0;
@@ -4324,8 +4325,8 @@ lpfc_fcp_cpu_map_init(struct lpfc_hba *phba, int val)
}
 
lpfc_printf_log(phba, KERN_ERR, LOG_INIT,
-   "3326 fcp_cpu_map: %d out of range, using default\n",
-   val);
+   "3326 lpfc_fcp_cpu_map: %d out of range, using "
+   "default\n", val);
phba->cfg_fcp_cpu_map = LPFC_DRIVER_CPU_MAP;
 
return 0;
-- 
2.5.0


--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH v4 1/4] block: Add iocontext priority to request

[Adam Manzananares]

The 10/13/2016 14:09, Jens Axboe wrote:
> On 10/13/2016 02:06 PM, Dan Williams wrote:
> >On Thu, Oct 13, 2016 at 12:53 PM, Adam Manzanares
> > wrote:
> >>Patch adds an association between iocontext ioprio and the ioprio of a
> >>request. This value is set in blk_rq_set_prio which takes the request and
> >>the ioc as arguments. If the ioc is valid in blk_rq_set_prio then the
> >>iopriority of the request is set as the iopriority of the ioc. In
> >>init_request_from_bio a check is made to see if the ioprio of the bio is
> >>valid and if so then the request prio comes from the bio.
> >>
> >>Signed-off-by: Adam Manzananares 
> >>---
> >> block/blk-core.c   |  4 +++-
> >> include/linux/blkdev.h | 14 ++
> >> 2 files changed, 17 insertions(+), 1 deletion(-)
> >>
> >>diff --git a/block/blk-core.c b/block/blk-core.c
> >>index 14d7c07..361b1b9 100644
> >>--- a/block/blk-core.c
> >>+++ b/block/blk-core.c
> >>@@ -1153,6 +1153,7 @@ static struct request *__get_request(struct 
> >>request_list *rl, int op,
> >>
> >>blk_rq_init(q, rq);
> >>blk_rq_set_rl(rq, rl);
> >>+   blk_rq_set_prio(rq, ioc);
> >>req_set_op_attrs(rq, op, op_flags | REQ_ALLOCED);
> >>
> >>/* init elvpriv */
> >>@@ -1656,7 +1657,8 @@ void init_request_from_bio(struct request *req, 
> >>struct bio *bio)
> >>
> >>req->errors = 0;
> >>req->__sector = bio->bi_iter.bi_sector;
> >>-   req->ioprio = bio_prio(bio);
> >>+   if (ioprio_valid(bio_prio(bio)))
> >>+   req->ioprio = bio_prio(bio);
> >
> >Should we use ioprio_best() here?  If req->ioprio and bio_prio()
> >disagree one side has explicitly asked for a higher priority.
> 
> It's a good question - but if priority has been set in the bio, it makes
> sense that that would take priority over the general setting for the
> task/io context. So I think the patch is correct as-is.
> 
> Adam, you'll want to rewrite the commit message though. A good commit
> message should explain WHY the change is made, not detail the code
> implementation of it.

Got it I'll send something out soon.

> 
> -- 
> Jens Axboe
> 

Take care,
Adam
--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH v4 1/4] block: Add iocontext priority to request

[Jens Axboe]

On 10/13/2016 02:59 PM, Dan Williams wrote:

On Thu, Oct 13, 2016 at 1:24 PM, Jens Axboe  wrote:

On 10/13/2016 02:19 PM, Dan Williams wrote:


On Thu, Oct 13, 2016 at 1:09 PM, Jens Axboe  wrote:


On 10/13/2016 02:06 PM, Dan Williams wrote:



On Thu, Oct 13, 2016 at 12:53 PM, Adam Manzanares
 wrote:



Patch adds an association between iocontext ioprio and the ioprio of a
request. This value is set in blk_rq_set_prio which takes the request
and
the ioc as arguments. If the ioc is valid in blk_rq_set_prio then the
iopriority of the request is set as the iopriority of the ioc. In
init_request_from_bio a check is made to see if the ioprio of the bio
is
valid and if so then the request prio comes from the bio.

Signed-off-by: Adam Manzananares 
---
 block/blk-core.c   |  4 +++-
 include/linux/blkdev.h | 14 ++
 2 files changed, 17 insertions(+), 1 deletion(-)

diff --git a/block/blk-core.c b/block/blk-core.c
index 14d7c07..361b1b9 100644
--- a/block/blk-core.c
+++ b/block/blk-core.c
@@ -1153,6 +1153,7 @@ static struct request *__get_request(struct
request_list *rl, int op,

blk_rq_init(q, rq);
blk_rq_set_rl(rq, rl);
+   blk_rq_set_prio(rq, ioc);
req_set_op_attrs(rq, op, op_flags | REQ_ALLOCED);

/* init elvpriv */
@@ -1656,7 +1657,8 @@ void init_request_from_bio(struct request *req,
struct bio *bio)

req->errors = 0;
req->__sector = bio->bi_iter.bi_sector;
-   req->ioprio = bio_prio(bio);
+   if (ioprio_valid(bio_prio(bio)))
+   req->ioprio = bio_prio(bio);




Should we use ioprio_best() here?  If req->ioprio and bio_prio()
disagree one side has explicitly asked for a higher priority.




It's a good question - but if priority has been set in the bio, it makes
sense that that would take priority over the general setting for the
task/io context. So I think the patch is correct as-is.



Assuming you always trust the kernel to know the right priority...



If it set it in the bio, it better know what it's doing. Besides,
there's nothing stopping the caller from checking the task priority when
it sets it. If we do ioprio_best(), then we are effectively preventing
anyone from submitting a bio with a lower priority than the task has
generally set.


Ah, that makes sense.  Move the ioprio_best() decision out to whatever
code is setting bio_prio() to allow for cases where the kernel knows
best.


Yes, precisely.

--
Jens Axboe

--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

RE: [PATCH v4 2/4] fusion: remove iopriority handling

[Sathya Prakash Veerichetty]

By removing the code below, we put all the commands for all the types of
devices (SAS/SATA) as simple-Q (requeue as the device require) and I am
not sure whether it is the intention of this change.

-Original Message-
From: Adam Manzanares [mailto:adam.manzana...@hgst.com]
Sent: Thursday, October 13, 2016 1:54 PM
To: ax...@kernel.dk; t...@kernel.org; dan.j.willi...@intel.com;
h...@suse.de; martin.peter...@oracle.com; mchri...@redhat.com;
toshi.k...@hpe.com; ming@canonical.com; sathya.prak...@broadcom.com;
chaitra.basa...@broadcom.com; suganath-prabu.subram...@broadcom.com
Cc: linux-bl...@vger.kernel.org; linux-...@vger.kernel.org;
linux-ker...@vger.kernel.org; mpt-fusionlinux@broadcom.com;
linux-scsi@vger.kernel.org; Adam Manzanares; Adam Manzanares
Subject: [PATCH v4 2/4] fusion: remove iopriority handling

The request priority is now by default coming from the ioc. It was not
clear what this code was trying to do based upon the iopriority class or
data. The driver should check that a device supports priorities and use
them according to the specificiations of ioprio.

Signed-off-by: Adam Manzanares 
---
 drivers/message/fusion/mptscsih.c | 5 -
 1 file changed, 5 deletions(-)

diff --git a/drivers/message/fusion/mptscsih.c
b/drivers/message/fusion/mptscsih.c
index 6c9fc11..4740bb6 100644
--- a/drivers/message/fusion/mptscsih.c
+++ b/drivers/message/fusion/mptscsih.c
@@ -1369,11 +1369,6 @@ mptscsih_qcmd(struct scsi_cmnd *SCpnt)
if ((vdevice->vtarget->tflags & MPT_TARGET_FLAGS_Q_YES)
&& (SCpnt->device->tagged_supported)) {
scsictl = scsidir | MPI_SCSIIO_CONTROL_SIMPLEQ;
-   if (SCpnt->request && SCpnt->request->ioprio) {
-   if (((SCpnt->request->ioprio & 0x7) == 1) ||
-   !(SCpnt->request->ioprio & 0x7))
-   scsictl |= MPI_SCSIIO_CONTROL_HEADOFQ;
-   }
} else
scsictl = scsidir | MPI_SCSIIO_CONTROL_UNTAGGED;

--
2.1.4

Western Digital Corporation (and its subsidiaries) E-mail Confidentiality
Notice & Disclaimer:

This e-mail and any files transmitted with it may contain confidential or
legally privileged information of WDC and/or its affiliates, and are
intended solely for the use of the individual or entity to which they are
addressed. If you are not the intended recipient, any disclosure, copying,
distribution or any action taken or omitted to be taken in reliance on it,
is prohibited. If you have received this e-mail in error, please notify
the sender immediately and delete the e-mail in its entirety from your
system.
--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH v4 1/4] block: Add iocontext priority to request

[Dan Williams]

On Thu, Oct 13, 2016 at 1:24 PM, Jens Axboe  wrote:
> On 10/13/2016 02:19 PM, Dan Williams wrote:
>>
>> On Thu, Oct 13, 2016 at 1:09 PM, Jens Axboe  wrote:
>>>
>>> On 10/13/2016 02:06 PM, Dan Williams wrote:


 On Thu, Oct 13, 2016 at 12:53 PM, Adam Manzanares
  wrote:
>
>
> Patch adds an association between iocontext ioprio and the ioprio of a
> request. This value is set in blk_rq_set_prio which takes the request
> and
> the ioc as arguments. If the ioc is valid in blk_rq_set_prio then the
> iopriority of the request is set as the iopriority of the ioc. In
> init_request_from_bio a check is made to see if the ioprio of the bio
> is
> valid and if so then the request prio comes from the bio.
>
> Signed-off-by: Adam Manzananares 
> ---
>  block/blk-core.c   |  4 +++-
>  include/linux/blkdev.h | 14 ++
>  2 files changed, 17 insertions(+), 1 deletion(-)
>
> diff --git a/block/blk-core.c b/block/blk-core.c
> index 14d7c07..361b1b9 100644
> --- a/block/blk-core.c
> +++ b/block/blk-core.c
> @@ -1153,6 +1153,7 @@ static struct request *__get_request(struct
> request_list *rl, int op,
>
> blk_rq_init(q, rq);
> blk_rq_set_rl(rq, rl);
> +   blk_rq_set_prio(rq, ioc);
> req_set_op_attrs(rq, op, op_flags | REQ_ALLOCED);
>
> /* init elvpriv */
> @@ -1656,7 +1657,8 @@ void init_request_from_bio(struct request *req,
> struct bio *bio)
>
> req->errors = 0;
> req->__sector = bio->bi_iter.bi_sector;
> -   req->ioprio = bio_prio(bio);
> +   if (ioprio_valid(bio_prio(bio)))
> +   req->ioprio = bio_prio(bio);



 Should we use ioprio_best() here?  If req->ioprio and bio_prio()
 disagree one side has explicitly asked for a higher priority.
>>>
>>>
>>>
>>> It's a good question - but if priority has been set in the bio, it makes
>>> sense that that would take priority over the general setting for the
>>> task/io context. So I think the patch is correct as-is.
>>
>>
>> Assuming you always trust the kernel to know the right priority...
>
>
> If it set it in the bio, it better know what it's doing. Besides,
> there's nothing stopping the caller from checking the task priority when
> it sets it. If we do ioprio_best(), then we are effectively preventing
> anyone from submitting a bio with a lower priority than the task has
> generally set.

Ah, that makes sense.  Move the ioprio_best() decision out to whatever
code is setting bio_prio() to allow for cases where the kernel knows
best.
--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH] scsi: hpsa: remove unneeded void pointer cast

[Don Brace]

On 10/13/2016 11:10 AM, Javier Martinez Canillas wrote:

It's not necessary to cast the result of kmalloc, since void pointers are
promoted to any other type. This also fixes following coccinelle warning:

casting value returned by memory allocation function to 
(BIG_IOCTL_Command_struct *) is useless.

Signed-off-by: Javier Martinez Canillas 
---

  drivers/scsi/hpsa.c | 3 +--
  1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/drivers/scsi/hpsa.c b/drivers/scsi/hpsa.c
index d007ec18179a..4e82b692298e 100644
--- a/drivers/scsi/hpsa.c
+++ b/drivers/scsi/hpsa.c
@@ -6657,8 +6657,7 @@ static int hpsa_big_passthru_ioctl(struct ctlr_info *h, 
void __user *argp)
return -EINVAL;
if (!capable(CAP_SYS_RAWIO))
return -EPERM;
-   ioc = (BIG_IOCTL_Command_struct *)
-   kmalloc(sizeof(*ioc), GFP_KERNEL);
+   ioc = kmalloc(sizeof(*ioc), GFP_KERNEL);
if (!ioc) {
status = -ENOMEM;
goto cleanup1;

Acked-by: Don Brace 
--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH v4 1/4] block: Add iocontext priority to request

[Jens Axboe]

On 10/13/2016 02:19 PM, Dan Williams wrote:

On Thu, Oct 13, 2016 at 1:09 PM, Jens Axboe  wrote:

On 10/13/2016 02:06 PM, Dan Williams wrote:


On Thu, Oct 13, 2016 at 12:53 PM, Adam Manzanares
 wrote:


Patch adds an association between iocontext ioprio and the ioprio of a
request. This value is set in blk_rq_set_prio which takes the request and
the ioc as arguments. If the ioc is valid in blk_rq_set_prio then the
iopriority of the request is set as the iopriority of the ioc. In
init_request_from_bio a check is made to see if the ioprio of the bio is
valid and if so then the request prio comes from the bio.

Signed-off-by: Adam Manzananares 
---
 block/blk-core.c   |  4 +++-
 include/linux/blkdev.h | 14 ++
 2 files changed, 17 insertions(+), 1 deletion(-)

diff --git a/block/blk-core.c b/block/blk-core.c
index 14d7c07..361b1b9 100644
--- a/block/blk-core.c
+++ b/block/blk-core.c
@@ -1153,6 +1153,7 @@ static struct request *__get_request(struct
request_list *rl, int op,

blk_rq_init(q, rq);
blk_rq_set_rl(rq, rl);
+   blk_rq_set_prio(rq, ioc);
req_set_op_attrs(rq, op, op_flags | REQ_ALLOCED);

/* init elvpriv */
@@ -1656,7 +1657,8 @@ void init_request_from_bio(struct request *req,
struct bio *bio)

req->errors = 0;
req->__sector = bio->bi_iter.bi_sector;
-   req->ioprio = bio_prio(bio);
+   if (ioprio_valid(bio_prio(bio)))
+   req->ioprio = bio_prio(bio);



Should we use ioprio_best() here?  If req->ioprio and bio_prio()
disagree one side has explicitly asked for a higher priority.



It's a good question - but if priority has been set in the bio, it makes
sense that that would take priority over the general setting for the
task/io context. So I think the patch is correct as-is.


Assuming you always trust the kernel to know the right priority...


If it set it in the bio, it better know what it's doing. Besides,
there's nothing stopping the caller from checking the task priority when
it sets it. If we do ioprio_best(), then we are effectively preventing
anyone from submitting a bio with a lower priority than the task has
generally set.

Now, this depends on the priority being set in req->ioprio is
exclusively inherited from ioc->ioprio. That's the case for file system
requests, but if someone allocated a request and set the priority
otherwise, then ioprio_best() would be correct.

--
Jens Axboe

--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH v4 1/4] block: Add iocontext priority to request

[Dan Williams]

On Thu, Oct 13, 2016 at 1:09 PM, Jens Axboe  wrote:
> On 10/13/2016 02:06 PM, Dan Williams wrote:
>>
>> On Thu, Oct 13, 2016 at 12:53 PM, Adam Manzanares
>>  wrote:
>>>
>>> Patch adds an association between iocontext ioprio and the ioprio of a
>>> request. This value is set in blk_rq_set_prio which takes the request and
>>> the ioc as arguments. If the ioc is valid in blk_rq_set_prio then the
>>> iopriority of the request is set as the iopriority of the ioc. In
>>> init_request_from_bio a check is made to see if the ioprio of the bio is
>>> valid and if so then the request prio comes from the bio.
>>>
>>> Signed-off-by: Adam Manzananares 
>>> ---
>>>  block/blk-core.c   |  4 +++-
>>>  include/linux/blkdev.h | 14 ++
>>>  2 files changed, 17 insertions(+), 1 deletion(-)
>>>
>>> diff --git a/block/blk-core.c b/block/blk-core.c
>>> index 14d7c07..361b1b9 100644
>>> --- a/block/blk-core.c
>>> +++ b/block/blk-core.c
>>> @@ -1153,6 +1153,7 @@ static struct request *__get_request(struct
>>> request_list *rl, int op,
>>>
>>> blk_rq_init(q, rq);
>>> blk_rq_set_rl(rq, rl);
>>> +   blk_rq_set_prio(rq, ioc);
>>> req_set_op_attrs(rq, op, op_flags | REQ_ALLOCED);
>>>
>>> /* init elvpriv */
>>> @@ -1656,7 +1657,8 @@ void init_request_from_bio(struct request *req,
>>> struct bio *bio)
>>>
>>> req->errors = 0;
>>> req->__sector = bio->bi_iter.bi_sector;
>>> -   req->ioprio = bio_prio(bio);
>>> +   if (ioprio_valid(bio_prio(bio)))
>>> +   req->ioprio = bio_prio(bio);
>>
>>
>> Should we use ioprio_best() here?  If req->ioprio and bio_prio()
>> disagree one side has explicitly asked for a higher priority.
>
>
> It's a good question - but if priority has been set in the bio, it makes
> sense that that would take priority over the general setting for the
> task/io context. So I think the patch is correct as-is.

Assuming you always trust the kernel to know the right priority...
--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH v4 1/4] block: Add iocontext priority to request

[Dan Williams]

On Thu, Oct 13, 2016 at 12:53 PM, Adam Manzanares
 wrote:
> Patch adds an association between iocontext ioprio and the ioprio of a
> request. This value is set in blk_rq_set_prio which takes the request and
> the ioc as arguments. If the ioc is valid in blk_rq_set_prio then the
> iopriority of the request is set as the iopriority of the ioc. In
> init_request_from_bio a check is made to see if the ioprio of the bio is
> valid and if so then the request prio comes from the bio.
>
> Signed-off-by: Adam Manzananares 
> ---
>  block/blk-core.c   |  4 +++-
>  include/linux/blkdev.h | 14 ++
>  2 files changed, 17 insertions(+), 1 deletion(-)
>
> diff --git a/block/blk-core.c b/block/blk-core.c
> index 14d7c07..361b1b9 100644
> --- a/block/blk-core.c
> +++ b/block/blk-core.c
> @@ -1153,6 +1153,7 @@ static struct request *__get_request(struct 
> request_list *rl, int op,
>
> blk_rq_init(q, rq);
> blk_rq_set_rl(rq, rl);
> +   blk_rq_set_prio(rq, ioc);
> req_set_op_attrs(rq, op, op_flags | REQ_ALLOCED);
>
> /* init elvpriv */
> @@ -1656,7 +1657,8 @@ void init_request_from_bio(struct request *req, struct 
> bio *bio)
>
> req->errors = 0;
> req->__sector = bio->bi_iter.bi_sector;
> -   req->ioprio = bio_prio(bio);
> +   if (ioprio_valid(bio_prio(bio)))
> +   req->ioprio = bio_prio(bio);

Should we use ioprio_best() here?  If req->ioprio and bio_prio()
disagree one side has explicitly asked for a higher priority.
--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH v4 1/4] block: Add iocontext priority to request

[Jens Axboe]

On 10/13/2016 02:06 PM, Dan Williams wrote:

On Thu, Oct 13, 2016 at 12:53 PM, Adam Manzanares
 wrote:

Patch adds an association between iocontext ioprio and the ioprio of a
request. This value is set in blk_rq_set_prio which takes the request and
the ioc as arguments. If the ioc is valid in blk_rq_set_prio then the
iopriority of the request is set as the iopriority of the ioc. In
init_request_from_bio a check is made to see if the ioprio of the bio is
valid and if so then the request prio comes from the bio.

Signed-off-by: Adam Manzananares 
---
 block/blk-core.c   |  4 +++-
 include/linux/blkdev.h | 14 ++
 2 files changed, 17 insertions(+), 1 deletion(-)

diff --git a/block/blk-core.c b/block/blk-core.c
index 14d7c07..361b1b9 100644
--- a/block/blk-core.c
+++ b/block/blk-core.c
@@ -1153,6 +1153,7 @@ static struct request *__get_request(struct request_list 
*rl, int op,

blk_rq_init(q, rq);
blk_rq_set_rl(rq, rl);
+   blk_rq_set_prio(rq, ioc);
req_set_op_attrs(rq, op, op_flags | REQ_ALLOCED);

/* init elvpriv */
@@ -1656,7 +1657,8 @@ void init_request_from_bio(struct request *req, struct 
bio *bio)

req->errors = 0;
req->__sector = bio->bi_iter.bi_sector;
-   req->ioprio = bio_prio(bio);
+   if (ioprio_valid(bio_prio(bio)))
+   req->ioprio = bio_prio(bio);


Should we use ioprio_best() here?  If req->ioprio and bio_prio()
disagree one side has explicitly asked for a higher priority.


It's a good question - but if priority has been set in the bio, it makes
sense that that would take priority over the general setting for the
task/io context. So I think the patch is correct as-is.

Adam, you'll want to rewrite the commit message though. A good commit
message should explain WHY the change is made, not detail the code
implementation of it.

--
Jens Axboe

--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH v4 4/4] ata: ATA Command Priority Disabled By Default

[Adam Manzanares]

Add a sysfs entry to turn on priority information being passed
to a ATA device. By default this feature is turned off.

This patch depends on ata: Enabling ATA Command Priorities

Signed-off-by: Adam Manzanares 
---
 drivers/ata/libahci.c |  1 +
 drivers/ata/libata-core.c |  2 +-
 drivers/ata/libata-scsi.c | 68 +++
 include/linux/libata.h|  8 ++
 4 files changed, 78 insertions(+), 1 deletion(-)

diff --git a/drivers/ata/libahci.c b/drivers/ata/libahci.c
index dcf2c72..383adf7 100644
--- a/drivers/ata/libahci.c
+++ b/drivers/ata/libahci.c
@@ -140,6 +140,7 @@ EXPORT_SYMBOL_GPL(ahci_shost_attrs);
 struct device_attribute *ahci_sdev_attrs[] = {
_attr_sw_activity,
_attr_unload_heads,
+   _attr_enable_prio,
NULL
 };
 EXPORT_SYMBOL_GPL(ahci_sdev_attrs);
diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index 181b530..d0cf987 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -787,7 +787,7 @@ int ata_build_rw_tf(struct ata_taskfile *tf, struct 
ata_device *dev,
if (tf->flags & ATA_TFLAG_FUA)
tf->device |= 1 << 7;
 
-   if (ata_ncq_prio_enabled(dev)) {
+   if (ata_ncq_prio_enabled(dev) && ata_prio_enabled(dev)) {
if (class == IOPRIO_CLASS_RT)
tf->hob_nsect |= ATA_PRIO_HIGH <<
 ATA_SHIFT_PRIO;
diff --git a/drivers/ata/libata-scsi.c b/drivers/ata/libata-scsi.c
index 18629e8..10ba118 100644
--- a/drivers/ata/libata-scsi.c
+++ b/drivers/ata/libata-scsi.c
@@ -271,6 +271,73 @@ DEVICE_ATTR(unload_heads, S_IRUGO | S_IWUSR,
ata_scsi_park_show, ata_scsi_park_store);
 EXPORT_SYMBOL_GPL(dev_attr_unload_heads);
 
+static ssize_t ata_enable_prio_show(struct device *device,
+   struct device_attribute *attr, char *buf)
+{
+   struct scsi_device *sdev = to_scsi_device(device);
+   struct ata_port *ap;
+   struct ata_device *dev;
+   int rc = 0;
+   int enable_prio;
+
+   ap = ata_shost_to_port(sdev->host);
+
+   spin_lock_irq(ap->lock);
+   dev = ata_scsi_find_dev(ap, sdev);
+   if (!dev) {
+   rc = -ENODEV;
+   goto unlock;
+   }
+
+   enable_prio = ata_prio_enabled(dev);
+
+unlock:
+   spin_unlock_irq(ap->lock);
+
+   return rc ? rc : snprintf(buf, 20, "%u\n", enable_prio);
+}
+
+static ssize_t ata_enable_prio_store(struct device *device,
+struct device_attribute *attr,
+const char *buf, size_t len)
+{
+   struct scsi_device *sdev = to_scsi_device(device);
+   struct ata_port *ap;
+   struct ata_device *dev;
+   long int input;
+   unsigned long flags;
+   int rc;
+
+   rc = kstrtol(buf, 10, );
+   if (rc)
+   return rc;
+   if ((input < 0) || (input > 1))
+   return -EINVAL;
+
+   ap = ata_shost_to_port(sdev->host);
+
+   spin_lock_irqsave(ap->lock, flags);
+   dev = ata_scsi_find_dev(ap, sdev);
+   if (unlikely(!dev)) {
+   rc = -ENODEV;
+   goto unlock;
+   }
+
+   if (input)
+   dev->flags |= ATA_DFLAG_ENABLE_PRIO;
+   else
+   dev->flags &= ~ATA_DFLAG_ENABLE_PRIO;
+
+unlock:
+   spin_unlock_irqrestore(ap->lock, flags);
+
+   return rc ? rc : len;
+}
+
+DEVICE_ATTR(enable_prio, S_IRUGO | S_IWUSR,
+   ata_enable_prio_show, ata_enable_prio_store);
+EXPORT_SYMBOL_GPL(dev_attr_enable_prio);
+
 void ata_scsi_set_sense(struct ata_device *dev, struct scsi_cmnd *cmd,
u8 sk, u8 asc, u8 ascq)
 {
@@ -402,6 +469,7 @@ EXPORT_SYMBOL_GPL(dev_attr_sw_activity);
 
 struct device_attribute *ata_common_sdev_attrs[] = {
_attr_unload_heads,
+   _attr_enable_prio,
NULL
 };
 EXPORT_SYMBOL_GPL(ata_common_sdev_attrs);
diff --git a/include/linux/libata.h b/include/linux/libata.h
index 244f261..c8acb16 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -166,6 +166,7 @@ enum {
ATA_DFLAG_UNLOCK_HPA= (1 << 18), /* unlock HPA */
ATA_DFLAG_NCQ_SEND_RECV = (1 << 19), /* device supports NCQ SEND and 
RECV */
ATA_DFLAG_NCQ_PRIO  = (1 << 20), /* device supports NCQ priority */
+   ATA_DFLAG_ENABLE_PRIO   = (1 << 21), /* User enable device priority */
ATA_DFLAG_INIT_MASK = (1 << 24) - 1,
 
ATA_DFLAG_DETACH= (1 << 24),
@@ -544,6 +545,7 @@ typedef void (*ata_postreset_fn_t)(struct ata_link *link, 
unsigned int *classes)
 
 extern struct device_attribute dev_attr_link_power_management_policy;
 extern struct device_attribute dev_attr_unload_heads;
+extern struct device_attribute dev_attr_enable_prio;
 extern struct device_attribute dev_attr_em_message_type;
 extern struct device_attribute 

[PATCH v4 3/4] ata: Enabling ATA Command Priorities

[Adam Manzanares]

This patch checks to see if an ATA device supports NCQ command priorities.
If so and the user has specified an iocontext that indicates IO_PRIO_CLASS_RT
then we build a tf with a high priority command.

Signed-off-by: Adam Manzanares 
---
 drivers/ata/libata-core.c | 35 ++-
 drivers/ata/libata-scsi.c |  6 +-
 drivers/ata/libata.h  |  2 +-
 include/linux/ata.h   |  6 ++
 include/linux/libata.h| 18 ++
 5 files changed, 64 insertions(+), 3 deletions(-)

diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index 223a770..181b530 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -739,6 +739,7 @@ u64 ata_tf_read_block(const struct ata_taskfile *tf, struct 
ata_device *dev)
  * @n_block: Number of blocks
  * @tf_flags: RW/FUA etc...
  * @tag: tag
+ * @class: IO priority class
  *
  * LOCKING:
  * None.
@@ -753,7 +754,7 @@ u64 ata_tf_read_block(const struct ata_taskfile *tf, struct 
ata_device *dev)
  */
 int ata_build_rw_tf(struct ata_taskfile *tf, struct ata_device *dev,
u64 block, u32 n_block, unsigned int tf_flags,
-   unsigned int tag)
+   unsigned int tag, int class)
 {
tf->flags |= ATA_TFLAG_ISADDR | ATA_TFLAG_DEVICE;
tf->flags |= tf_flags;
@@ -785,6 +786,12 @@ int ata_build_rw_tf(struct ata_taskfile *tf, struct 
ata_device *dev,
tf->device = ATA_LBA;
if (tf->flags & ATA_TFLAG_FUA)
tf->device |= 1 << 7;
+
+   if (ata_ncq_prio_enabled(dev)) {
+   if (class == IOPRIO_CLASS_RT)
+   tf->hob_nsect |= ATA_PRIO_HIGH <<
+ATA_SHIFT_PRIO;
+   }
} else if (dev->flags & ATA_DFLAG_LBA) {
tf->flags |= ATA_TFLAG_LBA;
 
@@ -2156,6 +2163,30 @@ static void ata_dev_config_ncq_non_data(struct 
ata_device *dev)
}
 }
 
+static void ata_dev_config_ncq_prio(struct ata_device *dev)
+{
+   struct ata_port *ap = dev->link->ap;
+   unsigned int err_mask;
+
+   err_mask = ata_read_log_page(dev,
+ATA_LOG_SATA_ID_DEV_DATA,
+ATA_LOG_SATA_SETTINGS,
+ap->sector_buf,
+1);
+   if (err_mask) {
+   ata_dev_dbg(dev,
+   "failed to get Identify Device data, Emask 0x%x\n",
+   err_mask);
+   return;
+   }
+
+   if (ap->sector_buf[ATA_LOG_NCQ_PRIO_OFFSET] & BIT(3))
+   dev->flags |= ATA_DFLAG_NCQ_PRIO;
+   else
+   ata_dev_dbg(dev, "SATA page does not support priority\n");
+
+}
+
 static int ata_dev_config_ncq(struct ata_device *dev,
   char *desc, size_t desc_sz)
 {
@@ -2205,6 +2236,8 @@ static int ata_dev_config_ncq(struct ata_device *dev,
ata_dev_config_ncq_send_recv(dev);
if (ata_id_has_ncq_non_data(dev->id))
ata_dev_config_ncq_non_data(dev);
+   if (ata_id_has_ncq_prio(dev->id))
+   ata_dev_config_ncq_prio(dev);
}
 
return 0;
diff --git a/drivers/ata/libata-scsi.c b/drivers/ata/libata-scsi.c
index e207b33..18629e8 100644
--- a/drivers/ata/libata-scsi.c
+++ b/drivers/ata/libata-scsi.c
@@ -50,6 +50,7 @@
 #include 
 #include 
 #include 
+#include 
 
 #include "libata.h"
 #include "libata-transport.h"
@@ -1757,6 +1758,8 @@ static unsigned int ata_scsi_rw_xlat(struct 
ata_queued_cmd *qc)
 {
struct scsi_cmnd *scmd = qc->scsicmd;
const u8 *cdb = scmd->cmnd;
+   struct request *rq = scmd->request;
+   int class = IOPRIO_PRIO_CLASS(req_get_ioprio(rq));
unsigned int tf_flags = 0;
u64 block;
u32 n_block;
@@ -1823,7 +1826,8 @@ static unsigned int ata_scsi_rw_xlat(struct 
ata_queued_cmd *qc)
qc->nbytes = n_block * scmd->device->sector_size;
 
rc = ata_build_rw_tf(>tf, qc->dev, block, n_block, tf_flags,
-qc->tag);
+qc->tag, class);
+
if (likely(rc == 0))
return 0;
 
diff --git a/drivers/ata/libata.h b/drivers/ata/libata.h
index 3b301a4..8f3a559 100644
--- a/drivers/ata/libata.h
+++ b/drivers/ata/libata.h
@@ -66,7 +66,7 @@ extern u64 ata_tf_to_lba48(const struct ata_taskfile *tf);
 extern struct ata_queued_cmd *ata_qc_new_init(struct ata_device *dev, int tag);
 extern int ata_build_rw_tf(struct ata_taskfile *tf, struct ata_device *dev,
   u64 block, u32 n_block, unsigned int tf_flags,
-  unsigned int tag);
+  unsigned int tag, int class);
 extern u64 ata_tf_read_block(const struct ata_taskfile *tf,
 

[PATCH v4 2/4] fusion: remove iopriority handling

[Adam Manzanares]

The request priority is now by default coming from the ioc. It was not
clear what this code was trying to do based upon the iopriority class or
data. The driver should check that a device supports priorities and use
them according to the specificiations of ioprio.

Signed-off-by: Adam Manzanares 
---
 drivers/message/fusion/mptscsih.c | 5 -
 1 file changed, 5 deletions(-)

diff --git a/drivers/message/fusion/mptscsih.c 
b/drivers/message/fusion/mptscsih.c
index 6c9fc11..4740bb6 100644
--- a/drivers/message/fusion/mptscsih.c
+++ b/drivers/message/fusion/mptscsih.c
@@ -1369,11 +1369,6 @@ mptscsih_qcmd(struct scsi_cmnd *SCpnt)
if ((vdevice->vtarget->tflags & MPT_TARGET_FLAGS_Q_YES)
&& (SCpnt->device->tagged_supported)) {
scsictl = scsidir | MPI_SCSIIO_CONTROL_SIMPLEQ;
-   if (SCpnt->request && SCpnt->request->ioprio) {
-   if (((SCpnt->request->ioprio & 0x7) == 1) ||
-   !(SCpnt->request->ioprio & 0x7))
-   scsictl |= MPI_SCSIIO_CONTROL_HEADOFQ;
-   }
} else
scsictl = scsidir | MPI_SCSIIO_CONTROL_UNTAGGED;
 
-- 
2.1.4

--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH v4 1/4] block: Add iocontext priority to request

[Adam Manzanares]

Patch adds an association between iocontext ioprio and the ioprio of a
request. This value is set in blk_rq_set_prio which takes the request and
the ioc as arguments. If the ioc is valid in blk_rq_set_prio then the
iopriority of the request is set as the iopriority of the ioc. In
init_request_from_bio a check is made to see if the ioprio of the bio is
valid and if so then the request prio comes from the bio.

Signed-off-by: Adam Manzananares 
---
 block/blk-core.c   |  4 +++-
 include/linux/blkdev.h | 14 ++
 2 files changed, 17 insertions(+), 1 deletion(-)

diff --git a/block/blk-core.c b/block/blk-core.c
index 14d7c07..361b1b9 100644
--- a/block/blk-core.c
+++ b/block/blk-core.c
@@ -1153,6 +1153,7 @@ static struct request *__get_request(struct request_list 
*rl, int op,
 
blk_rq_init(q, rq);
blk_rq_set_rl(rq, rl);
+   blk_rq_set_prio(rq, ioc);
req_set_op_attrs(rq, op, op_flags | REQ_ALLOCED);
 
/* init elvpriv */
@@ -1656,7 +1657,8 @@ void init_request_from_bio(struct request *req, struct 
bio *bio)
 
req->errors = 0;
req->__sector = bio->bi_iter.bi_sector;
-   req->ioprio = bio_prio(bio);
+   if (ioprio_valid(bio_prio(bio)))
+   req->ioprio = bio_prio(bio);
blk_rq_bio_prep(req->q, req, bio);
 }
 
diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h
index c47c358..9a0ceaa 100644
--- a/include/linux/blkdev.h
+++ b/include/linux/blkdev.h
@@ -934,6 +934,20 @@ static inline unsigned int blk_rq_count_bios(struct 
request *rq)
 }
 
 /*
+ * blk_rq_set_prio - associate a request with prio from ioc
+ * @rq: request of interest
+ * @ioc: target iocontext
+ *
+ * Assocate request prio with ioc prio so request based drivers
+ * can leverage priority information.
+ */
+static inline void blk_rq_set_prio(struct request *rq, struct io_context *ioc)
+{
+   if (ioc)
+   rq->ioprio = ioc->ioprio;
+}
+
+/*
  * Request issue related functions.
  */
 extern struct request *blk_peek_request(struct request_queue *q);
-- 
2.1.4

--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH v4 0/4] Enabling ATA Command Priorities

[Adam Manzanares]

This patch builds ATA commands with high priority if the iocontext
of a process is set to real time. The goal of the patch is to
improve tail latencies of workloads that use higher queue depths.
This required setting the iocontext ioprio on the request when 
the request is initialized.

This patch has been tested with an Ultrastar HE8 HDD and cuts the 
the p99.99 tail latency of foreground IO from 2s down to 72ms when
using the deadline scheduler. This patch works independently of the
scheduler so it can be used with all of the currently available 
request based schedulers. 

Foreground IO, for the previously described results, is an async fio job 
submitting 4K read requests at a QD of 1 to the HDD. The foreground IO is set 
with the iopriority class of real time. The background workload is another fio
job submitting read requests at a QD of 32 to the same HDD with default 
iopriority.

This feature is enabled for ATA devices by setting the ata enable_prio device 
attribute to 1. An ATA device is also checked to see if the device supports per
command priority.

v4:
 - Added blk_rq_set_prio function to associate request prio with ioc prio
 - In init_request_from_bio use bio_prio if it is valid
 - Added ata enable_prio dev attribute to sysfs to enable prioritized commands

v3:
 - Removed null dereference issue in blk-core
 - Renamed queue sysfs entries for clarity
 - Added documentation for sysfs queue entry

v2:
 - Add queue flag to set iopriority going to the request
 - If queue flag set, send iopriority class to ata_build_rw_tf
 - Remove redundant code in ata_ncq_prio_enabled function.


Adam Manzanares (4):
  block: Add iocontext priority to request
  fusion: remove iopriority handling
  ata: Enabling ATA Command Priorities
  ata: ATA Command Priority Disabled By Default

 block/blk-core.c  |  4 ++-
 drivers/ata/libahci.c |  1 +
 drivers/ata/libata-core.c | 35 +-
 drivers/ata/libata-scsi.c | 74 ++-
 drivers/ata/libata.h  |  2 +-
 drivers/message/fusion/mptscsih.c |  5 ---
 include/linux/ata.h   |  6 
 include/linux/blkdev.h| 14 
 include/linux/libata.h| 26 ++
 9 files changed, 158 insertions(+), 9 deletions(-)

-- 
2.1.4

--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH 1/1] [4.9-rc] ipr: Fix async error WARN_ON

[Brian King]

Commit afc3f83cb4a5 ("scsi: ipr: Add asynchronous error notification")
introduced the warn on shown below. To fix this, rather than attempting
to send the KOBJ_CHANGE uevent from interrupt context, which is what is
causing the WARN_ON, just wake the ipr worker thread which will send a
KOBJ_CHANGE uevent.

[  142.278120] WARNING: CPU: 15 PID: 0 at kernel/softirq.c:161 
__local_bh_enable_ip+0x7c/0xd0
[  142.278124] Modules linked in: ip6t_rpfilter ip6t_REJECT nf_reject_ipv6 
ipt_REJECT nf_reject_ipv4 xt_conntrack ebtable_nat ebtable_broute bridge stp 
llc ebtable_filter ebtables ip6table_nat nf_conntrack_ipv6 nf_defrag_ipv6 
nf_nat_ipv6 ip6table_mangle ip6table_security ip6table_raw ip6table_filter 
ip6_tables iptable_nat nf_conntrack_ipv4 nf_defrag_ipv4 nf_nat_ipv4 nf_nat 
nf_conntrack iptable_mangle iptable_security iptable_raw iptable_filter ses 
enclosure scsi_transport_sas sg pseries_rng nfsd auth_rpcgss nfs_acl lockd 
grace sunrpc ip_tables xfs libcrc32c sr_mod sd_mod cdrom ipr libata ibmvscsi 
scsi_transport_srp ibmveth dm_mirror dm_region_hash dm_log dm_mod
[  142.278208] CPU: 15 PID: 0 Comm: swapper/15 Not tainted 4.8.0.ipr+ #21
[  142.278213] task: c0010cf24480 task.stack: c0010cfec000
[  142.278217] NIP: c00c0c7c LR: c0881778 CTR: c03c5bf0
[  142.278221] REGS: c0010cfef080 TRAP: 0700   Not tainted  (4.8.0.ipr+)
[  142.278224] MSR: 80029033   CR: 28008022  XER: 
200f
[  142.278236] CFAR: c00c0c20 SOFTE: 0 
GPR00: c0706c78 c0010cfef300 c0f91d00 c0706c78 
GPR04: 0200 c0f7bc80  024000c0 
GPR08:  0001 c0ee1d00 c0a9bdd0 
GPR12: c03c5bf0 ceb22d00 c00100ca3880 c0020ed38400 
GPR16:   c00100940508  
GPR20:    024000c0 
GPR24: c04588e0 c0010863bd00 c0010863bd00 c13773f8 
GPR28: c0f7bc80   c0f7bcd8 
[  142.278290] NIP [c00c0c7c] __local_bh_enable_ip+0x7c/0xd0
[  142.278296] LR [c0881778] _raw_spin_unlock_bh+0x38/0x60
[  142.278299] Call Trace:
[  142.278303] [c0010cfef300] [c0f7bc80] init_net+0x0/0x1900 
(unreliable)
[  142.278310] [c0010cfef320] [c0706c78] peernet2id+0x58/0x80
[  142.278316] [c0010cfef370] [c075caec] 
netlink_broadcast_filtered+0x30c/0x550
[  142.278323] [c0010cfef430] [c0459078] 
kobject_uevent_env+0x588/0x780
[  142.278331] [c0010cfef510] [d3163a6c] 
ipr_process_error+0x11c/0x240 [ipr]
[  142.278337] [c0010cfef5c0] [d3152298] 
ipr_fail_all_ops+0x108/0x220 [ipr]
[  142.278343] [c0010cfef670] [d31643f8] 
ipr_reset_restore_cfg_space+0xa8/0x240 [ipr]
[  142.278350] [c0010cfef6f0] [d3158a00] 
ipr_reset_ioa_job+0x80/0xe0 [ipr]
[  142.278356] [c0010cfef720] [d3153f78] 
ipr_reset_timer_done+0xa8/0xe0 [ipr]
[  142.278363] [c0010cfef770] [c0149c88] call_timer_fn+0x58/0x1c0
[  142.278368] [c0010cfef800] [c0149f60] expire_timers+0x140/0x200
[  142.278373] [c0010cfef870] [c014a0e8] 
run_timer_softirq+0xc8/0x230
[  142.278379] [c0010cfef900] [c00c0844] __do_softirq+0x164/0x3c0
[  142.278384] [c0010cfef9f0] [c00c0f18] irq_exit+0x1a8/0x1c0
[  142.278389] [c0010cfefa20] [c0020b54] timer_interrupt+0xa4/0xe0
[  142.278394] [c0010cfefa50] [c0002414] 
decrementer_common+0x114/0x180


Signed-off-by: Brian King 
---

 drivers/scsi/ipr.c |3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff -puN drivers/scsi/ipr.c~ipr_async_err_fixup drivers/scsi/ipr.c
--- linux-2.6.git/drivers/scsi/ipr.c~ipr_async_err_fixup2016-10-13 
09:57:47.634920565 -0500
+++ linux-2.6.git-bjking1/drivers/scsi/ipr.c2016-10-13 14:42:25.654418345 
-0500
@@ -2586,7 +2586,6 @@ static void ipr_process_error(struct ipr
struct ipr_hostrcb *hostrcb = ipr_cmd->u.hostrcb;
u32 ioasc = be32_to_cpu(ipr_cmd->s.ioasa.hdr.ioasc);
u32 fd_ioasc;
-   char *envp[] = { "ASYNC_ERR_LOG=1", NULL };
 
if (ioa_cfg->sis64)
fd_ioasc = be32_to_cpu(hostrcb->hcam.u.error64.fd_ioasc);
@@ -2607,8 +2606,8 @@ static void ipr_process_error(struct ipr
}
 
list_add_tail(>queue, _cfg->hostrcb_report_q);
+   schedule_work(_cfg->work_q);
hostrcb = ipr_get_free_hostrcb(ioa_cfg);
-   kobject_uevent_env(_cfg->host->shost_dev.kobj, KOBJ_CHANGE, envp);
 
ipr_send_hcam(ioa_cfg, IPR_HCAM_CDB_OP_CODE_LOG_DATA, hostrcb);
 }
_

--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH v2 2/2] scsi: sr: Drop custom handling of unit attention

[Gabriel Krisman Bertazi]

These custom handling are no longer necessary, since we always retry UA
in scsi_execute now.

Signed-off-by: Gabriel Krisman Bertazi 
---
 drivers/scsi/scsi_lib.c | 21 ++---
 drivers/scsi/sr_ioctl.c |  6 ++
 2 files changed, 8 insertions(+), 19 deletions(-)

diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c
index 9c6623abf120..e9324a224b11 100644
--- a/drivers/scsi/scsi_lib.c
+++ b/drivers/scsi/scsi_lib.c
@@ -2317,7 +2317,7 @@ scsi_mode_sense(struct scsi_device *sdev, int dbd, int 
modepage,
unsigned char cmd[12];
int use_10_for_ms;
int header_length;
-   int result, retry_count = retries;
+   int result;
struct scsi_sense_hdr my_sshdr;
 
memset(data, 0, sizeof(*data));
@@ -2396,11 +2396,6 @@ scsi_mode_sense(struct scsi_device *sdev, int dbd, int 
modepage,
data->block_descriptor_length = buffer[3];
}
data->header_length = header_length;
-   } else if ((status_byte(result) == CHECK_CONDITION) &&
-  scsi_sense_valid(sshdr) &&
-  sshdr->sense_key == UNIT_ATTENTION && retry_count) {
-   retry_count--;
-   goto retry;
}
 
return result;
@@ -2434,15 +2429,11 @@ scsi_test_unit_ready(struct scsi_device *sdev, int 
timeout, int retries,
else
sshdr = sshdr_external;
 
-   /* try to eat the UNIT_ATTENTION if there are enough retries */
-   do {
-   result = scsi_execute_req(sdev, cmd, DMA_NONE, NULL, 0, sshdr,
- timeout, retries, NULL);
-   if (sdev->removable && scsi_sense_valid(sshdr) &&
-   sshdr->sense_key == UNIT_ATTENTION)
-   sdev->changed = 1;
-   } while (scsi_sense_valid(sshdr) &&
-sshdr->sense_key == UNIT_ATTENTION && --retries);
+   result = scsi_execute_req(sdev, cmd, DMA_NONE, NULL, 0, sshdr,
+ timeout, retries, NULL);
+   if (sdev->removable && scsi_sense_valid(sshdr) &&
+   sshdr->sense_key == UNIT_ATTENTION)
+   sdev->changed = 1;
 
if (!sshdr_external)
kfree(sshdr);
diff --git a/drivers/scsi/sr_ioctl.c b/drivers/scsi/sr_ioctl.c
index 03054c0e7689..93b5544a5966 100644
--- a/drivers/scsi/sr_ioctl.c
+++ b/drivers/scsi/sr_ioctl.c
@@ -179,8 +179,8 @@ static int sr_play_trkind(struct cdrom_device_info *cdi,
 }
 
 /* We do our own retries because we want to know what the specific
-   error code is.  Normally the UNIT_ATTENTION code will automatically
-   clear after one error */
+   error code is.
+*/
 
 int sr_do_ioctl(Scsi_CD *cd, struct packet_command *cgc)
 {
@@ -220,8 +220,6 @@ int sr_do_ioctl(Scsi_CD *cd, struct packet_command *cgc)
if (!cgc->quiet)
sr_printk(KERN_INFO, cd,
  "disc change detected.\n");
-   if (retries++ < 10)
-   goto retry;
err = -ENOMEDIUM;
break;
case NOT_READY: /* This happens if there is no disc in drive */
-- 
2.7.4

--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH v2 1/2] scsi: Handle Unit Attention when issuing SCSI command

[Gabriel Krisman Bertazi]

Hi James,

Thanks for your review.  Please see the v2 below.

> OK, so really this isn't what you want, because blk_execute_req may
> have used several of your retries, so you now get a maximum possible
> set of retries at UNIT_ATTENTION_RETRIES*retries.  You need to start
> from the returned req->retries, which probably means this loop needs to
> be inside __scsi_execute.

Hmm, I was aware of that, but I saw there were other places that may
have run retries^2 times, like scsi_test_unit_ready and scsi_mode_sense,
if I read the code correctly.  But, I see your point and I fixed it on
v2.  I also updated the second patch to rework these cases.

Another thing that got me confused is where the blk layer updates
req->retries.

What do you think about the v2 below?

Thanks,

-- >8 --

Usually, re-sending the SCSI command is enough to recover from a Unit
Attention (UA).  This adds a generic retry code to the SCSI command path
in case of an UA, before giving up and returning the error condition to
the caller.

I added the UA verification into scsi_execute instead of
scsi_execute_req because there are at least a few callers that invoke
scsi_execute directly and would benefit from the internal UA retry.
Also, I didn't use scsi_normalize_sense to not duplicate functionality
with scsi_execute_req_flags.  Instead, scsi_execute uses a small helper
function that verifies only the UA condition directly from the raw sense
buffer.  If this design is not OK, I can refactor to use
scsi_normalize_sense.

This prevents us from duplicating the retry code in at least a few
places.  In particular, it fixes an issue found in some IBM enclosures,
in which the device may return an Unit Attention during probe, breaking
the bind with the ses module:

scsi 1:0:7:0: Failed to get diagnostic page 0x802
scsi 1:0:7:0: Failed to bind enclosure -19

Link: https://patchwork.kernel.org/patch/9336763/
Suggested-by: Brian King 
Suggested-by: James Bottomley 
Signed-off-by: Gabriel Krisman Bertazi 
---
 drivers/scsi/scsi_lib.c| 27 ---
 include/scsi/scsi_common.h |  9 +
 2 files changed, 33 insertions(+), 3 deletions(-)

diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c
index c71344aebdbb..9c6623abf120 100644
--- a/drivers/scsi/scsi_lib.c
+++ b/drivers/scsi/scsi_lib.c
@@ -187,15 +187,24 @@ int scsi_execute(struct scsi_device *sdev, const unsigned 
char *cmd,
struct request *req;
int write = (data_direction == DMA_TO_DEVICE);
int ret = DRIVER_ERROR << 24;
+   bool priv_sense = false;
 
+   if (!sense) {
+   sense = kzalloc(SCSI_SENSE_BUFFERSIZE, GFP_NOIO);
+   if (!sense)
+   return ret;
+   priv_sense = true;
+   }
+
+ retry:
req = blk_get_request(sdev->request_queue, write, __GFP_RECLAIM);
if (IS_ERR(req))
-   return ret;
+   goto free_sense;
blk_rq_set_block_pc(req);
 
if (bufflen &&  blk_rq_map_kern(sdev->request_queue, req,
buffer, bufflen, __GFP_RECLAIM))
-   goto out;
+   goto put_req;
 
req->cmd_len = COMMAND_SIZE(cmd[0]);
memcpy(req->cmd, cmd, req->cmd_len);
@@ -210,6 +219,13 @@ int scsi_execute(struct scsi_device *sdev, const unsigned 
char *cmd,
 */
blk_execute_rq(req->q, NULL, req, 1);
 
+   if (scsi_sense_unit_attention(sense) && req->retries > 0) {
+   memset(sense, 0, SCSI_SENSE_BUFFERSIZE);
+   retries = req->retries - 1;
+   blk_put_request(req);
+   goto retry;
+   }
+
/*
 * Some devices (USB mass-storage in particular) may transfer
 * garbage data together with a residue indicating that the data
@@ -222,9 +238,14 @@ int scsi_execute(struct scsi_device *sdev, const unsigned 
char *cmd,
if (resid)
*resid = req->resid_len;
ret = req->errors;
- out:
+
+ put_req:
blk_put_request(req);
 
+ free_sense:
+   if (priv_sense)
+   kfree(sense);
+
return ret;
 }
 EXPORT_SYMBOL(scsi_execute);
diff --git a/include/scsi/scsi_common.h b/include/scsi/scsi_common.h
index 20bf7eaef05a..747b632d5b57 100644
--- a/include/scsi/scsi_common.h
+++ b/include/scsi/scsi_common.h
@@ -58,6 +58,15 @@ static inline bool scsi_sense_valid(const struct 
scsi_sense_hdr *sshdr)
return (sshdr->response_code & 0x70) == 0x70;
 }
 
+static inline bool scsi_sense_unit_attention(const char *sense)
+{
+   int resp = sense[0] & 0x7f;
+
+   return ((resp & 0x70) &&
+   ((resp >= 0x72 && (sense[1] & 0xf) == UNIT_ATTENTION) ||
+(resp < 0x72 && (sense[2] & 0xf) == UNIT_ATTENTION)));
+}
+
 extern bool scsi_normalize_sense(const u8 *sense_buffer, int sb_len,
 struct scsi_sense_hdr *sshdr);
 
-- 

Re: [PATCH] ufs: changing maintainer

[Subhash Jadavani]

On 2016-10-13 10:23, Joao Pinto wrote:

Signed-off-by: Joao Pinto 
---
 MAINTAINERS | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/MAINTAINERS b/MAINTAINERS
index 5ac91d8a..d9855b5 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -12084,7 +12084,8 @@ F:  Documentation/scsi/ufs.txt
 F: drivers/scsi/ufs/

 UNIVERSAL FLASH STORAGE HOST CONTROLLER DRIVER DWC HOOKS
-M: Joao Pinto 
+M: Manjunath M Bettegowda 
+M: Prabu Thangamuthu 
 L: linux-scsi@vger.kernel.org
 S: Supported
 F: drivers/scsi/ufs/*dwc*


Looking at the commit text, it gives the impression that we are changing 
the maintainer for whole UFS driver.
Can you make it specific? May be the commit text header should be "ufs: 
changing maintainer for DWC"?


--
The Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum,
a Linux Foundation Collaborative Project
--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH v2 2/6] ibmvscsis: Synchronize cmds at tpg_enable_store time

[Michael Cyr]

This patch changes the way the IBM vSCSI server driver manages its
Command/Response Queue (CRQ).  We used to register the CRQ with phyp at
probe time.  Now we wait until tpg_enable_store.  Similarly, when
tpg_enable_store is called to "disable" (i.e. the stored value is 0),
we unregister the queue with phyp.

One consquence to this is that we have no need for the PART_UP_WAIT_ENAB
state, since we can't get an Init Message from the client in our CRQ if
we're waiting to be enabled, since we haven't registered the queue yet.

Signed-off-by: Michael Cyr 
Signed-off-by: Bryant G. Ly 
---
 drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.c | 224 ++-
 drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.h |   2 -
 2 files changed, 38 insertions(+), 188 deletions(-)

diff --git a/drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.c 
b/drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.c
index 01a430c..2ce1d73 100644
--- a/drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.c
+++ b/drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.c
@@ -61,8 +61,6 @@ static long ibmvscsis_parse_command(struct scsi_info *vscsi,
 
 static void ibmvscsis_adapter_idle(struct scsi_info *vscsi);
 
-static void ibmvscsis_reset_queue(struct scsi_info *vscsi, uint new_state);
-
 static void ibmvscsis_determine_resid(struct se_cmd *se_cmd,
  struct srp_rsp *rsp)
 {
@@ -417,7 +415,6 @@ static void ibmvscsis_disconnect(struct work_struct *work)
   proc_work);
u16 new_state;
bool wait_idle = false;
-   long rc = ADAPT_SUCCESS;
 
spin_lock_bh(>intr_lock);
new_state = vscsi->new_state;
@@ -470,30 +467,12 @@ static void ibmvscsis_disconnect(struct work_struct *work)
vscsi->state = new_state;
break;
 
-   /*
-* If this is a transition into an error state.
-* a client is attempting to establish a connection
-* and has violated the RPA protocol.
-* There can be nothing pending on the adapter although
-* there can be requests in the command queue.
-*/
case WAIT_ENABLED:
-   case PART_UP_WAIT_ENAB:
switch (new_state) {
+   /* should never happen */
case ERR_DISCONNECT:
-   vscsi->flags |= RESPONSE_Q_DOWN;
-   vscsi->state = new_state;
-   vscsi->flags &= ~(SCHEDULE_DISCONNECT |
- DISCONNECT_SCHEDULED);
-   ibmvscsis_free_command_q(vscsi);
-   break;
case ERR_DISCONNECT_RECONNECT:
-   ibmvscsis_reset_queue(vscsi, WAIT_ENABLED);
-   break;
-
-   /* should never happen */
case WAIT_IDLE:
-   rc = ERROR;
dev_err(>dev, "disconnect: invalid state %d for 
WAIT_IDLE\n",
vscsi->state);
break;
@@ -630,7 +609,6 @@ static void ibmvscsis_post_disconnect(struct scsi_info 
*vscsi, uint new_state,
break;
 
case WAIT_ENABLED:
-   case PART_UP_WAIT_ENAB:
case WAIT_IDLE:
case WAIT_CONNECTION:
case CONNECTED:
@@ -675,7 +653,6 @@ static long ibmvscsis_handle_init_compl_msg(struct 
scsi_info *vscsi)
case SRP_PROCESSING:
case CONNECTED:
case WAIT_ENABLED:
-   case PART_UP_WAIT_ENAB:
default:
rc = ERROR;
dev_err(>dev, "init_msg: invalid state %d to get init 
compl msg\n",
@@ -698,10 +675,6 @@ static long ibmvscsis_handle_init_msg(struct scsi_info 
*vscsi)
long rc = ADAPT_SUCCESS;
 
switch (vscsi->state) {
-   case WAIT_ENABLED:
-   vscsi->state = PART_UP_WAIT_ENAB;
-   break;
-
case WAIT_CONNECTION:
rc = ibmvscsis_send_init_message(vscsi, INIT_COMPLETE_MSG);
switch (rc) {
@@ -737,7 +710,7 @@ static long ibmvscsis_handle_init_msg(struct scsi_info 
*vscsi)
case UNCONFIGURING:
break;
 
-   case PART_UP_WAIT_ENAB:
+   case WAIT_ENABLED:
case CONNECTED:
case SRP_PROCESSING:
case WAIT_IDLE:
@@ -800,11 +773,10 @@ static long ibmvscsis_init_msg(struct scsi_info *vscsi, 
struct viosrp_crq *crq)
 /**
  * ibmvscsis_establish_new_q() - Establish new CRQ queue
  * @vscsi: Pointer to our adapter structure
- * @new_state: New state being established after resetting the queue
  *
  * Must be called with interrupt lock held.
  */
-static long ibmvscsis_establish_new_q(struct scsi_info *vscsi, uint new_state)
+static long ibmvscsis_establish_new_q(struct scsi_info *vscsi)
 {
long rc = ADAPT_SUCCESS;
uint format;
@@ -816,19 +788,19 @@ static long ibmvscsis_establish_new_q(struct scsi_info 

[PATCH] ufs: changing maintainer

[Joao Pinto]

Signed-off-by: Joao Pinto 
---
 MAINTAINERS | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/MAINTAINERS b/MAINTAINERS
index 5ac91d8a..d9855b5 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -12084,7 +12084,8 @@ F:  Documentation/scsi/ufs.txt
 F: drivers/scsi/ufs/
 
 UNIVERSAL FLASH STORAGE HOST CONTROLLER DRIVER DWC HOOKS
-M: Joao Pinto 
+M: Manjunath M Bettegowda 
+M: Prabu Thangamuthu 
 L: linux-scsi@vger.kernel.org
 S: Supported
 F: drivers/scsi/ufs/*dwc*
-- 
1.8.1.5

--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH v2 0/6] Fixes for ibmvscsis driver

[Michael Cyr]

Various fixes and cleanups for the ibmvscsis driver.
The first is a sort of prequel to the second, which is the primary
change (and the biggest).  The rest are fairly small fixes.

Michael Cyr (6):
  ibmvscsis: Rearrange functions for future patches
  ibmvscsis: Synchronize cmds at tpg_enable_store time
  ibmvscsis: Synchronize cmds at remove time
  ibmvscsis: Clean up properly if target_submit_cmd/tmr fails
  ibmvscsis: Return correct partition name/# to client
  ibmvscsis: Issues from Dan Carpenter/Smatch

 drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.c | 1096 +-
 drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.h |5 +-
 2 files changed, 494 insertions(+), 607 deletions(-)

-- 
2.5.0

--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH] MAINTAINERS: remove defunct iss storage mailing list

[Javier Martinez Canillas]

It appears that the mailing list email address doesn't exist anymore:

: host smtp.hp.com[15.73.96.116] said: 550 5.1.1
: Recipient address rejected: User unknown in
virtual alias table (in reply to RCPT TO command)

Signed-off-by: Javier Martinez Canillas 

---

 MAINTAINERS | 2 --
 1 file changed, 2 deletions(-)

diff --git a/MAINTAINERS b/MAINTAINERS
index 19d8daab5b0d..56ff868dd7d2 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -5600,7 +5600,6 @@ F:drivers/watchdog/hpwdt.c
 
 HEWLETT-PACKARD SMART ARRAY RAID DRIVER (hpsa)
 M: Don Brace 
-L: iss_storage...@hp.com
 L: esc.storage...@microsemi.com
 L: linux-scsi@vger.kernel.org
 S: Supported
@@ -5611,7 +5610,6 @@ F:include/uapi/linux/cciss*.h
 
 HEWLETT-PACKARD SMART CISS RAID DRIVER (cciss)
 M: Don Brace 
-L: iss_storage...@hp.com
 L: esc.storage...@microsemi.com
 L: linux-scsi@vger.kernel.org
 S: Supported
-- 
2.7.4

--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH v3 02/16] scsi: don't use fc_bsg_job::request and fc_bsg_job::reply directly

[Johannes Thumshirn]

On Thu, Oct 13, 2016 at 05:55:11PM +0200, Steffen Maier wrote:
> Hm, still behaves for me like I reported for v2:
> http://marc.info/?l=linux-scsi=147637177902937=2
> 
Well given what you've wrote for v2 it's kinda expected.

Byte,
Johannes

-- 
Johannes Thumshirn  Storage
jthumsh...@suse.de+49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: Felix Imendörffer, Jane Smithard, Graham Norton
HRB 21284 (AG Nürnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH v2 02/16] scsi: don't use fc_bsg_job::request and fc_bsg_job::reply directly

[Johannes Thumshirn]

On Thu, Oct 13, 2016 at 05:15:25PM +0200, Steffen Maier wrote:
> I'm puzzled.
> 
> $ git bisect start fc_bsg master
> Bisecting: 8 revisions left to test after this (roughly 3 steps)
> [005d51510eee6102636d5dbb06310531c5d46151] scsi: fc: implement kref backed
> reference counting
> $ git bisect bad
> Bisecting: 3 revisions left to test after this (roughly 2 steps)
> [bef6da201de1bb81bb4d9511f9a155862efc251f] scsi: Unify interfaces of
> fc_bsg_jobdone and bsg_job_done
> $ git bisect bad
> Bisecting: 1 revision left to test after this (roughly 1 step)
> [3087864ce3d7282f59021245d8a5f83ef1caef18] scsi: don't use
> fc_bsg_job::request and fc_bsg_job::reply directly
> $ git bisect bad
> Bisecting: 0 revisions left to test after this (roughly 0 steps)
> [81aea44720d22d2e0c4a2613ae8b1c256ef6b0cb] scsi: Get rid of struct
> fc_bsg_buffer
> > $ git bisect good
> > 3087864ce3d7282f59021245d8a5f83ef1caef18 is the first bad commit
> > commit 3087864ce3d7282f59021245d8a5f83ef1caef18
> > Author: Johannes Thumshirn 
> > Date:   Wed Oct 12 15:06:28 2016 +0200
> > 
> > scsi: don't use fc_bsg_job::request and fc_bsg_job::reply directly
> > 
> > Don't use fc_bsg_job::request and fc_bsg_job::reply directly, but use
> > helper variables bsg_request and bsg_reply. This will be helpfull  when
> > transitioning to bsg-lib.
> > 
> > Signed-off-by: Johannes Thumshirn 
> > 
> > :04 04 140c4b6829d5cfaec4079716e0795f63f8bc3bd2 
> > 0d9fe225615679550be91fbd9f84c09ab1e280fc M  drivers
> 
> From there (on the reverse bisect path) I get the following Oops,
> except for the full patch set having another stack trace as in my previous
> mail (dying in zfcp code).
> 

[...]

> 
> > @@ -3937,6 +3944,7 @@ fc_bsg_request_handler(struct request_queue *q, 
> > struct Scsi_Host *shost,
> > struct request *req;
> > struct fc_bsg_job *job;
> > enum fc_dispatch_result ret;
> > +   struct fc_bsg_reply *bsg_reply;
> > 
> > if (!get_device(dev))
> > return;
> > @@ -3973,8 +3981,9 @@ fc_bsg_request_handler(struct request_queue *q, 
> > struct Scsi_Host *shost,
> > /* check if we have the msgcode value at least */
> > if (job->request_len < sizeof(uint32_t)) {
> > BUG_ON(job->reply_len < sizeof(uint32_t));
> > -   job->reply->reply_payload_rcv_len = 0;
> > -   job->reply->result = -ENOMSG;
> > +   bsg_reply = job->reply;
> > +   bsg_reply->reply_payload_rcv_len = 0;
> > +   bsg_reply->result = -ENOMSG;
> > job->reply_len = sizeof(uint32_t);
> > fc_bsg_jobdone(job);
> > spin_lock_irq(q->queue_lock);
> > 

Ahm and what exactly can break here? It's just assigning variables. Now
I'm puzzled too.

I'll have to look into it tomorrow.

Byte,
Johannes

-- 
Johannes Thumshirn  Storage
jthumsh...@suse.de+49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: Felix Imendörffer, Jane Smithard, Graham Norton
HRB 21284 (AG Nürnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH] cciss: remove unneeded void pointer cast

[Javier Martinez Canillas]

It's not necessary to cast the result of kmalloc, since void pointers are
promoted to any other type. This also fixes following coccinelle warning:

casting value returned by memory allocation function to (struct 
cciss_scsi_adapter_data_t *) is useless.

Signed-off-by: Javier Martinez Canillas 
---

 drivers/block/cciss_scsi.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/drivers/block/cciss_scsi.c b/drivers/block/cciss_scsi.c
index 1537302e56e3..06692f164f40 100644
--- a/drivers/block/cciss_scsi.c
+++ b/drivers/block/cciss_scsi.c
@@ -703,8 +703,7 @@ cciss_scsi_setup(ctlr_info_t *h)
struct cciss_scsi_adapter_data_t * shba;
 
ccissscsi[h->ctlr].ndevices = 0;
-   shba = (struct cciss_scsi_adapter_data_t *)
-   kmalloc(sizeof(*shba), GFP_KERNEL); 
+   shba = kmalloc(sizeof(*shba), GFP_KERNEL);
if (shba == NULL)
return;
shba->scsi_host = NULL;
-- 
2.7.4

--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH] scsi: hpsa: remove unneeded void pointer cast

[Javier Martinez Canillas]

It's not necessary to cast the result of kmalloc, since void pointers are
promoted to any other type. This also fixes following coccinelle warning:

casting value returned by memory allocation function to 
(BIG_IOCTL_Command_struct *) is useless.

Signed-off-by: Javier Martinez Canillas 
---

 drivers/scsi/hpsa.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/drivers/scsi/hpsa.c b/drivers/scsi/hpsa.c
index d007ec18179a..4e82b692298e 100644
--- a/drivers/scsi/hpsa.c
+++ b/drivers/scsi/hpsa.c
@@ -6657,8 +6657,7 @@ static int hpsa_big_passthru_ioctl(struct ctlr_info *h, 
void __user *argp)
return -EINVAL;
if (!capable(CAP_SYS_RAWIO))
return -EPERM;
-   ioc = (BIG_IOCTL_Command_struct *)
-   kmalloc(sizeof(*ioc), GFP_KERNEL);
+   ioc = kmalloc(sizeof(*ioc), GFP_KERNEL);
if (!ioc) {
status = -ENOMEM;
goto cleanup1;
-- 
2.7.4

--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH v2 5/6] ibmvscsis: Return correct partition name/# to client

[Michael Cyr]

Signed-off-by: Michael Cyr 
Signed-off-by: Bryant G. Ly 
---
 drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.c | 5 -
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.c 
b/drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.c
index cd9f5c7..fe220a1 100644
--- a/drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.c
+++ b/drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.c
@@ -3297,6 +3297,9 @@ static int ibmvscsis_probe(struct vio_dev *vdev,
strncat(vscsi->eye, vdev->name, MAX_EYE);
 
vscsi->dds.unit_id = vdev->unit_address;
+   strncpy(vscsi->dds.partition_name, partition_name,
+   sizeof(vscsi->dds.partition_name));
+   vscsi->dds.partition_num = partition_number;
 
spin_lock_bh(_dev_lock);
list_add_tail(>list, _dev_list);
@@ -3495,7 +3498,7 @@ static int ibmvscsis_get_system_info(void)
 
num = of_get_property(rootdn, "ibm,partition-no", NULL);
if (num)
-   partition_number = *num;
+   partition_number = of_read_number(num, 1);
 
of_node_put(rootdn);
 
-- 
2.5.0

--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH v2 3/6] ibmvscsis: Synchronize cmds at remove time

[Michael Cyr]

This patch adds code to disconnect from the client, which will make sure
any outstanding commands have been completed, before continuing on with
the remove operation.

Signed-off-by: Michael Cyr 
Signed-off-by: Bryant G. Ly 
---
 drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.c | 39 
 drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.h |  3 +++
 2 files changed, 37 insertions(+), 5 deletions(-)

diff --git a/drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.c 
b/drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.c
index 2ce1d73..41af435 100644
--- a/drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.c
+++ b/drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.c
@@ -469,6 +469,18 @@ static void ibmvscsis_disconnect(struct work_struct *work)
 
case WAIT_ENABLED:
switch (new_state) {
+   case UNCONFIGURING:
+   vscsi->state = new_state;
+   vscsi->flags |= RESPONSE_Q_DOWN;
+   vscsi->flags &= ~(SCHEDULE_DISCONNECT |
+ DISCONNECT_SCHEDULED);
+   dma_rmb();
+   if (vscsi->flags & CFG_SLEEPING) {
+   vscsi->flags &= ~CFG_SLEEPING;
+   complete(>unconfig);
+   }
+   break;
+
/* should never happen */
case ERR_DISCONNECT:
case ERR_DISCONNECT_RECONNECT:
@@ -481,6 +493,13 @@ static void ibmvscsis_disconnect(struct work_struct *work)
 
case WAIT_IDLE:
switch (new_state) {
+   case UNCONFIGURING:
+   vscsi->flags |= RESPONSE_Q_DOWN;
+   vscsi->state = new_state;
+   vscsi->flags &= ~(SCHEDULE_DISCONNECT |
+ DISCONNECT_SCHEDULED);
+   ibmvscsis_free_command_q(vscsi);
+   break;
case ERR_DISCONNECT:
case ERR_DISCONNECT_RECONNECT:
vscsi->state = new_state;
@@ -1186,6 +1205,15 @@ static void ibmvscsis_adapter_idle(struct scsi_info 
*vscsi)
free_qs = true;
 
switch (vscsi->state) {
+   case UNCONFIGURING:
+   ibmvscsis_free_command_q(vscsi);
+   dma_rmb();
+   isync();
+   if (vscsi->flags & CFG_SLEEPING) {
+   vscsi->flags &= ~CFG_SLEEPING;
+   complete(>unconfig);
+   }
+   break;
case ERR_DISCONNECT_RECONNECT:
ibmvscsis_reset_queue(vscsi);
pr_debug("adapter_idle, disc_rec: flags 0x%x\n", vscsi->flags);
@@ -3338,6 +3366,7 @@ static int ibmvscsis_probe(struct vio_dev *vdev,
 (unsigned long)vscsi);
 
init_completion(>wait_idle);
+   init_completion(>unconfig);
 
snprintf(wq_name, 24, "ibmvscsis%s", dev_name(>dev));
vscsi->work_q = create_workqueue(wq_name);
@@ -3393,10 +3422,11 @@ static int ibmvscsis_remove(struct vio_dev *vdev)
 
pr_debug("remove (%s)\n", dev_name(>dma_dev->dev));
 
-   /*
-* TBD: Need to handle if there are commands on the waiting_rsp q
-*  Actually, can there still be cmds outstanding to tcm?
-*/
+   spin_lock_bh(>intr_lock);
+   ibmvscsis_post_disconnect(vscsi, UNCONFIGURING, 0);
+   vscsi->flags |= CFG_SLEEPING;
+   spin_unlock_bh(>intr_lock);
+   wait_for_completion(>unconfig);
 
vio_disable_interrupts(vdev);
free_irq(vdev->irq, vscsi);
@@ -3405,7 +3435,6 @@ static int ibmvscsis_remove(struct vio_dev *vdev)
 DMA_BIDIRECTIONAL);
kfree(vscsi->map_buf);
tasklet_kill(>work_task);
-   ibmvscsis_unregister_command_q(vscsi);
ibmvscsis_destroy_command_q(vscsi);
ibmvscsis_freetimer(vscsi);
ibmvscsis_free_cmds(vscsi);
diff --git a/drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.h 
b/drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.h
index 17e0ef4..98b0ca7 100644
--- a/drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.h
+++ b/drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.h
@@ -257,6 +257,8 @@ struct scsi_info {
 #define SCHEDULE_DISCONNECT   0x00400
/* disconnect handler is scheduled */
 #define DISCONNECT_SCHEDULED  0x00800
+   /* remove function is sleeping */
+#define CFG_SLEEPING  0x01000
u32 flags;
/* adapter lock */
spinlock_t intr_lock;
@@ -285,6 +287,7 @@ struct scsi_info {
 
struct workqueue_struct *work_q;
struct completion wait_idle;
+   struct completion unconfig;
struct device dev;
struct vio_dev *dma_dev;
struct srp_target target;
-- 
2.5.0

--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  

[PATCH v2 6/6] ibmvscsis: Issues from Dan Carpenter/Smatch

[Michael Cyr]

Signed-off-by: Michael Cyr 
Signed-off-by: Bryant G. Ly 
---
 drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.c | 13 +++--
 1 file changed, 3 insertions(+), 10 deletions(-)

diff --git a/drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.c 
b/drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.c
index fe220a1..c9fa356 100644
--- a/drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.c
+++ b/drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.c
@@ -1896,14 +1896,7 @@ static long ibmvscsis_mad(struct scsi_info *vscsi, 
struct viosrp_crq *crq)
 
pr_debug("mad: type %d\n", be32_to_cpu(mad->type));
 
-   if (be16_to_cpu(mad->length) < 0) {
-   dev_err(>dev, "mad: length is < 0\n");
-   ibmvscsis_post_disconnect(vscsi,
- ERR_DISCONNECT_RECONNECT, 0);
-   rc = SRP_VIOLATION;
-   } else {
-   rc = ibmvscsis_process_mad(vscsi, iue);
-   }
+   rc = ibmvscsis_process_mad(vscsi, iue);
 
pr_debug("mad: status %hd, rc %ld\n", be16_to_cpu(mad->status),
 rc);
@@ -2523,7 +2516,6 @@ static void ibmvscsis_parse_cmd(struct scsi_info *vscsi,
dev_err(>dev, "0x%llx: parsing SRP descriptor table 
failed.\n",
srp->tag);
goto fail;
-   return;
}
 
cmd->rsp.sol_not = srp->sol_not;
@@ -3282,7 +3274,8 @@ static int ibmvscsis_probe(struct vio_dev *vdev,
INIT_LIST_HEAD(>waiting_rsp);
INIT_LIST_HEAD(>active_q);
 
-   snprintf(vscsi->tport.tport_name, 256, "%s", dev_name(>dev));
+   snprintf(vscsi->tport.tport_name, IBMVSCSIS_NAMELEN, "%s",
+dev_name(>dev));
 
pr_debug("probe tport_name: %s\n", vscsi->tport.tport_name);
 
-- 
2.5.0

--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH v2 4/6] ibmvscsis: Clean up properly if target_submit_cmd/tmr fails

[Michael Cyr]

Signed-off-by: Michael Cyr 
Signed-off-by: Bryant G. Ly 
---
 drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.c | 7 +++
 1 file changed, 7 insertions(+)

diff --git a/drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.c 
b/drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.c
index 41af435..cd9f5c7 100644
--- a/drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.c
+++ b/drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.c
@@ -2560,6 +2560,10 @@ static void ibmvscsis_parse_cmd(struct scsi_info *vscsi,
   data_len, attr, dir, 0);
if (rc) {
dev_err(>dev, "target_submit_cmd failed, rc %d\n", rc);
+   spin_lock_bh(>intr_lock);
+   list_del(>list);
+   ibmvscsis_free_cmd_resources(vscsi, cmd);
+   spin_unlock_bh(>intr_lock);
goto fail;
}
return;
@@ -2639,6 +2643,9 @@ static void ibmvscsis_parse_task(struct scsi_info *vscsi,
if (rc) {
dev_err(>dev, "target_submit_tmr failed, rc 
%d\n",
rc);
+   spin_lock_bh(>intr_lock);
+   list_del(>list);
+   spin_unlock_bh(>intr_lock);
cmd->se_cmd.se_tmr_req->response =
TMR_FUNCTION_REJECTED;
}
-- 
2.5.0

--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH v3 02/16] scsi: don't use fc_bsg_job::request and fc_bsg_job::reply directly

[Steffen Maier]

Hm, still behaves for me like I reported for v2:
http://marc.info/?l=linux-scsi=147637177902937=2

On 10/13/2016 05:00 PM, Johannes Thumshirn wrote:

Don't use fc_bsg_job::request and fc_bsg_job::reply directly, but use
helper variables bsg_request and bsg_reply. This will be helpfull  when
transitioning to bsg-lib.

Signed-off-by: Johannes Thumshirn 
Reviewed-by: Hannes Reinecke 
---
 drivers/s390/scsi/zfcp_fc.c  |   9 +-
 drivers/scsi/bfa/bfad_bsg.c  |  40 +++---
 drivers/scsi/ibmvscsi/ibmvfc.c   |  22 ++--
 drivers/scsi/libfc/fc_lport.c|  23 ++--
 drivers/scsi/lpfc/lpfc_bsg.c | 194 +---
 drivers/scsi/qla2xxx/qla_bsg.c   | 264 ++-
 drivers/scsi/qla2xxx/qla_iocb.c  |   5 +-
 drivers/scsi/qla2xxx/qla_isr.c   |  46 ---
 drivers/scsi/qla2xxx/qla_mr.c|  10 +-
 drivers/scsi/scsi_transport_fc.c |  37 +++---
 10 files changed, 387 insertions(+), 263 deletions(-)



diff --git a/drivers/scsi/scsi_transport_fc.c b/drivers/scsi/scsi_transport_fc.c
index 8ff2067..eafc7555 100644
--- a/drivers/scsi/scsi_transport_fc.c
+++ b/drivers/scsi/scsi_transport_fc.c



@@ -3973,8 +3981,9 @@ enum fc_dispatch_result {
/* check if we have the msgcode value at least */
if (job->request_len < sizeof(uint32_t)) {
BUG_ON(job->reply_len < sizeof(uint32_t));
-   job->reply->reply_payload_rcv_len = 0;
-   job->reply->result = -ENOMSG;
+   bsg_reply = job->reply;
+   bsg_reply->reply_payload_rcv_len = 0;
+   bsg_reply->result = -ENOMSG;
job->reply_len = sizeof(uint32_t);
fc_bsg_jobdone(job);
spin_lock_irq(q->queue_lock);



--
Mit freundlichen Grüßen / Kind regards
Steffen Maier

Linux on z Systems Development

IBM Deutschland Research & Development GmbH
Vorsitzende des Aufsichtsrats: Martina Koederitz
Geschaeftsfuehrung: Dirk Wittkopp
Sitz der Gesellschaft: Boeblingen
Registergericht: Amtsgericht Stuttgart, HRB 243294

--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH v2 02/16] scsi: don't use fc_bsg_job::request and fc_bsg_job::reply directly

[Steffen Maier]

I'm puzzled.

$ git bisect start fc_bsg master
Bisecting: 8 revisions left to test after this (roughly 3 steps)
[005d51510eee6102636d5dbb06310531c5d46151] scsi: fc: implement kref 
backed reference counting

$ git bisect bad
Bisecting: 3 revisions left to test after this (roughly 2 steps)
[bef6da201de1bb81bb4d9511f9a155862efc251f] scsi: Unify interfaces of 
fc_bsg_jobdone and bsg_job_done

$ git bisect bad
Bisecting: 1 revision left to test after this (roughly 1 step)
[3087864ce3d7282f59021245d8a5f83ef1caef18] scsi: don't use 
fc_bsg_job::request and fc_bsg_job::reply directly

$ git bisect bad
Bisecting: 0 revisions left to test after this (roughly 0 steps)
[81aea44720d22d2e0c4a2613ae8b1c256ef6b0cb] scsi: Get rid of struct 
fc_bsg_buffer

$ git bisect good
3087864ce3d7282f59021245d8a5f83ef1caef18 is the first bad commit
commit 3087864ce3d7282f59021245d8a5f83ef1caef18
Author: Johannes Thumshirn 
Date:   Wed Oct 12 15:06:28 2016 +0200

scsi: don't use fc_bsg_job::request and fc_bsg_job::reply directly

Don't use fc_bsg_job::request and fc_bsg_job::reply directly, but use
helper variables bsg_request and bsg_reply. This will be helpfull  when
transitioning to bsg-lib.

Signed-off-by: Johannes Thumshirn 

:04 04 140c4b6829d5cfaec4079716e0795f63f8bc3bd2 
0d9fe225615679550be91fbd9f84c09ab1e280fc M  drivers


From there (on the reverse bisect path) I get the following Oops,
except for the full patch set having another stack trace as in my 
previous mail (dying in zfcp code).



[   46.942452] Oops: 0004 ilc:2 [#1] [   46.942460] PREEMPT  SMP [   46.942465]
[   46.942470] Modules linked in: nf_log_ipv6 xt_pkttype nf_log_ipv4 
nf_log_common xt_LOG xt_limit ip6t_REJECT nf_reject_ipv6 xt_tcpudp 
nf_conntrack_ipv6 nf_defrag_ipv6 ip6table_raw ipt_REJECT nf_reject_ipv4 
iptable_raw xt_CT iptable_filter ip6table_mangle nf_conntrack_netbios_ns 
nf_conntrack_broadcast nf_conntrack_ipv4 nf_defrag_ipv4 ip_tables xt_conntrack 
nf_conntrack ip6table_filter ip6_tables x_tables qeth_l2 ghash_s390 prng ecb 
aes_s390 des_s390 des_generic sha512_s390 sha256_s390 sha1_s390 sha_common 
dm_mod qeth ccwgroup zfcp qdio autofs4
[   46.942547] CPU: 1 PID: 1714 Comm: zfcp_ping Not tainted 4.8.0fcbsg+ #9
[   46.942550] Hardware name: IBM  2964 N96  702
  (z/VM)
[   46.942556] task: 5c988008 task.stack: 5d2ec000
[   46.942560] Krnl PSW : 0704e0018000 007c91ec[   46.942574]  
(fc_bsg_request_handler+0x404/0x4b0)
[   46.942579]
[   46.942583]R:0 T:1 IO:1 EX:1 Key:0 M:1 W:0 P:0 AS:3 CC:2 PM:000:
[   46.942598]  RI:0 EA:3
[   46.942601]
[   46.942601] Krnl GPRS:  ffcb  
8001
[   46.942603]007c8fe8 64398c68 69f967e8 
6a3d8008
[   46.942605]6a5e02c8 698b5490  

[   46.942607]6a9ef5f8 00a36840 007c8fe8 
5d2efa00
[   46.942619] Krnl Code: 007c91de: e55dc08c0003clfhsi  
140(%r12),3[   46.942622]
[   46.942622]007c91e4: a7240004brc 2,7c91ec
 #007c91e8: a7f40001   brc 
15,7c91ea[   46.942629]
[   46.942629]   >007c91ec: 5010b000st  
%r1,0(%r11)
  007c91f0: e54cb004   mvhi
4(%r11),0[   46.942635]
[   46.942635]007c91f6: e54cc08c0004mvhi
140(%r12),4
  007c91fc: b904002c   lgr 
%r2,%r12[   46.942643]
[   46.942643]007c9200: c0e5e2c0brasl   
%r14,7c5780
[   46.942646]
[   46.942647] Call Trace:
[   46.942650] ([<007c8fe8>] fc_bsg_request_handler+0x200/0x4b0)
[   46.942656] ([<006b8e0a>] __blk_run_queue+0x52/0x68)
[   46.942661] ([<006c549a>] blk_execute_rq_nowait+0xf2/0x110)
[   46.942664] ([<006c557a>] blk_execute_rq+0xa2/0x110)
[   46.942668] ([<006de0ee>] bsg_ioctl+0x1f6/0x268)
[   46.942675] ([<0036ca20>] do_vfs_ioctl+0x680/0x6d8)
[   46.942677] ([<0036caf4>] SyS_ioctl+0x7c/0xb0)
[   46.942685] ([<009a541e>] system_call+0xd6/0x270)
[   46.942687] INFO: lockdep is turned off.
[   46.942688] Last Breaking-Event-Address:
[   46.942692]  [<007c91e4>] fc_bsg_request_handler+0x3fc/0x4b0
[   46.942696]  [   46.942698] Kernel panic - not syncing: Fatal exception: 
panic_on_oops




On 10/12/2016 03:06 PM, Johannes Thumshirn wrote:

Don't use fc_bsg_job::request and fc_bsg_job::reply directly, but use
helper variables bsg_request and bsg_reply. This will be helpfull  when
transitioning to bsg-lib.

Signed-off-by: Johannes Thumshirn 
---
 drivers/s390/scsi/zfcp_fc.c  |   9 +-
 drivers/scsi/bfa/bfad_bsg.c  |  40 +++---
 

[PATCH v3 01/16] scsi: Get rid of struct fc_bsg_buffer

[Johannes Thumshirn]

struct fc_bsg_buffer is just a clone of struct bsg_buffer from bsg-lib,
so use this one instead.

Signed-off-by: Johannes Thumshirn 
Reviewed-by: Hannes Reinecke 
---
 drivers/scsi/lpfc/lpfc_bsg.c |  3 ++-
 drivers/scsi/scsi_transport_fc.c |  2 +-
 include/scsi/scsi_transport_fc.h | 12 +++-
 3 files changed, 6 insertions(+), 11 deletions(-)

diff --git a/drivers/scsi/lpfc/lpfc_bsg.c b/drivers/scsi/lpfc/lpfc_bsg.c
index 05dcc2a..1002627 100644
--- a/drivers/scsi/lpfc/lpfc_bsg.c
+++ b/drivers/scsi/lpfc/lpfc_bsg.c
@@ -24,6 +24,7 @@
 #include 
 #include 
 #include 
+#include 
 
 #include 
 #include 
@@ -211,7 +212,7 @@ struct lpfc_dmabufext {
 
 static unsigned int
 lpfc_bsg_copy_data(struct lpfc_dmabuf *dma_buffers,
-  struct fc_bsg_buffer *bsg_buffers,
+  struct bsg_buffer *bsg_buffers,
   unsigned int bytes_to_transfer, int to_buffers)
 {
 
diff --git a/drivers/scsi/scsi_transport_fc.c b/drivers/scsi/scsi_transport_fc.c
index 0f3a386..8ff2067 100644
--- a/drivers/scsi/scsi_transport_fc.c
+++ b/drivers/scsi/scsi_transport_fc.c
@@ -3671,7 +3671,7 @@ static void fc_bsg_softirq_done(struct request *rq)
 }
 
 static int
-fc_bsg_map_buffer(struct fc_bsg_buffer *buf, struct request *req)
+fc_bsg_map_buffer(struct bsg_buffer *buf, struct request *req)
 {
size_t sz = (sizeof(struct scatterlist) * req->nr_phys_segments);
 
diff --git a/include/scsi/scsi_transport_fc.h b/include/scsi/scsi_transport_fc.h
index bf66ea6..921b097 100644
--- a/include/scsi/scsi_transport_fc.h
+++ b/include/scsi/scsi_transport_fc.h
@@ -28,6 +28,7 @@
 #define SCSI_TRANSPORT_FC_H
 
 #include 
+#include 
 #include 
 #include 
 #include 
@@ -624,13 +625,6 @@ struct fc_host_attrs {
 #define fc_host_dev_loss_tmo(x) \
(((struct fc_host_attrs *)(x)->shost_data)->dev_loss_tmo)
 
-
-struct fc_bsg_buffer {
-   unsigned int payload_len;
-   int sg_cnt;
-   struct scatterlist *sg_list;
-};
-
 /* Values for fc_bsg_job->state_flags (bitflags) */
 #define FC_RQST_STATE_INPROGRESS   0
 #define FC_RQST_STATE_DONE 1
@@ -659,8 +653,8 @@ struct fc_bsg_job {
 */
 
/* DMA payloads for the request/response */
-   struct fc_bsg_buffer request_payload;
-   struct fc_bsg_buffer reply_payload;
+   struct bsg_buffer request_payload;
+   struct bsg_buffer reply_payload;
 
void *dd_data;  /* Used for driver-specific storage */
 };
-- 
1.8.5.6

--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH v3 04/16] scsi: Unify interfaces of fc_bsg_jobdone and bsg_job_done

[Johannes Thumshirn]

Unify the interfaces of fc_bsg_jobdone and bsg_job_done. This will reduce the
diff when moving from 'struct fc_bsg_job' to a plain 'struct bsg_job' later
on.

Signed-off-by: Johannes Thumshirn 
Reviewed-by: Hannes Reinecke 
---
 drivers/s390/scsi/zfcp_fc.c  |  2 +-
 drivers/scsi/bfa/bfad_bsg.c  |  6 ++--
 drivers/scsi/ibmvscsi/ibmvfc.c   |  3 +-
 drivers/scsi/libfc/fc_lport.c|  6 ++--
 drivers/scsi/lpfc/lpfc_bsg.c | 68 +++-
 drivers/scsi/qla2xxx/qla_bsg.c   | 66 +-
 drivers/scsi/scsi_transport_fc.c | 22 +++--
 include/scsi/scsi_transport_fc.h |  3 +-
 8 files changed, 116 insertions(+), 60 deletions(-)

diff --git a/drivers/s390/scsi/zfcp_fc.c b/drivers/s390/scsi/zfcp_fc.c
index 40d8f06..87f6330 100644
--- a/drivers/s390/scsi/zfcp_fc.c
+++ b/drivers/s390/scsi/zfcp_fc.c
@@ -892,7 +892,7 @@ static void zfcp_fc_ct_els_job_handler(void *data)
jr->reply_payload_rcv_len = job->reply_payload.payload_len;
jr->reply_data.ctels_reply.status = FC_CTELS_STATUS_OK;
jr->result = zfcp_ct_els->status ? -EIO : 0;
-   fc_bsg_jobdone(job);
+   fc_bsg_jobdone(job, jr->result, jr->reply_payload_rcv_len);
 }
 
 static struct zfcp_fc_wka_port *zfcp_fc_job_wka_port(struct fc_bsg_job *job)
diff --git a/drivers/scsi/bfa/bfad_bsg.c b/drivers/scsi/bfa/bfad_bsg.c
index 25889b9..e49a6c8 100644
--- a/drivers/scsi/bfa/bfad_bsg.c
+++ b/drivers/scsi/bfa/bfad_bsg.c
@@ -3180,7 +3180,8 @@
bsg_reply->reply_payload_rcv_len = job->reply_payload.payload_len;
bsg_reply->result = rc;
 
-   fc_bsg_jobdone(job);
+   fc_bsg_jobdone(job, bsg_reply->result,
+  bsg_reply->reply_payload_rcv_len);
return rc;
 error:
/* free the command buffer */
@@ -3556,7 +3557,8 @@ struct bfad_buf_info *
bsg_reply->result = rc;
 
if (rc == BFA_STATUS_OK)
-   fc_bsg_jobdone(job);
+   fc_bsg_jobdone(job, bsg_reply->result,
+  bsg_reply->reply_payload_rcv_len);
 
return rc;
 }
diff --git a/drivers/scsi/ibmvscsi/ibmvfc.c b/drivers/scsi/ibmvscsi/ibmvfc.c
index f14e9c1..6070361 100644
--- a/drivers/scsi/ibmvscsi/ibmvfc.c
+++ b/drivers/scsi/ibmvscsi/ibmvfc.c
@@ -1945,7 +1945,8 @@ static int ibmvfc_bsg_request(struct fc_bsg_job *job)
ibmvfc_free_event(evt);
spin_unlock_irqrestore(vhost->host->host_lock, flags);
bsg_reply->result = rc;
-   fc_bsg_jobdone(job);
+   fc_bsg_jobdone(job, bsg_reply->result,
+  bsg_reply->reply_payload_rcv_len);
rc = 0;
 out:
dma_unmap_sg(vhost->dev, job->request_payload.sg_list,
diff --git a/drivers/scsi/libfc/fc_lport.c b/drivers/scsi/libfc/fc_lport.c
index 33e2881..000188f 100644
--- a/drivers/scsi/libfc/fc_lport.c
+++ b/drivers/scsi/libfc/fc_lport.c
@@ -1917,7 +1917,8 @@ static void fc_lport_bsg_resp(struct fc_seq *sp, struct 
fc_frame *fp,
-ECONNABORTED : -ETIMEDOUT;
job->reply_len = sizeof(uint32_t);
job->state_flags |= FC_RQST_STATE_DONE;
-   fc_bsg_jobdone(job);
+   fc_bsg_jobdone(job, bsg_reply->result,
+  bsg_reply->reply_payload_rcv_len);
kfree(info);
return;
}
@@ -1952,7 +1953,8 @@ static void fc_lport_bsg_resp(struct fc_seq *sp, struct 
fc_frame *fp,
job->reply_payload.payload_len;
bsg_reply->result = 0;
job->state_flags |= FC_RQST_STATE_DONE;
-   fc_bsg_jobdone(job);
+   fc_bsg_jobdone(job, bsg_reply->result,
+  bsg_reply->reply_payload_rcv_len);
kfree(info);
}
fc_frame_free(fp);
diff --git a/drivers/scsi/lpfc/lpfc_bsg.c b/drivers/scsi/lpfc/lpfc_bsg.c
index 1db9cca..447a7af 100644
--- a/drivers/scsi/lpfc/lpfc_bsg.c
+++ b/drivers/scsi/lpfc/lpfc_bsg.c
@@ -371,7 +371,8 @@ struct lpfc_dmabufext {
 
if (job) {
bsg_reply->result = rc;
-   fc_bsg_jobdone(job);
+   fc_bsg_jobdone(job, bsg_reply->result,
+  bsg_reply->reply_payload_rcv_len);
}
return;
 }
@@ -644,7 +645,8 @@ struct lpfc_dmabufext {
 
if (job) {
bsg_reply->result = rc;
-   fc_bsg_jobdone(job);
+   fc_bsg_jobdone(job, bsg_reply->result,
+  bsg_reply->reply_payload_rcv_len);
}
return;
 }
@@ -1136,7 +1138,8 @@ struct lpfc_dmabufext {
job->dd_data = NULL;
/* complete the job back to userspace */
spin_unlock_irqrestore(>ct_ev_lock, flags);
-   fc_bsg_jobdone(job);
+   fc_bsg_jobdone(job, bsg_reply->result,
+  

[PATCH v3 02/16] scsi: don't use fc_bsg_job::request and fc_bsg_job::reply directly

[Johannes Thumshirn]

Don't use fc_bsg_job::request and fc_bsg_job::reply directly, but use
helper variables bsg_request and bsg_reply. This will be helpfull  when
transitioning to bsg-lib.

Signed-off-by: Johannes Thumshirn 
Reviewed-by: Hannes Reinecke 
---
 drivers/s390/scsi/zfcp_fc.c  |   9 +-
 drivers/scsi/bfa/bfad_bsg.c  |  40 +++---
 drivers/scsi/ibmvscsi/ibmvfc.c   |  22 ++--
 drivers/scsi/libfc/fc_lport.c|  23 ++--
 drivers/scsi/lpfc/lpfc_bsg.c | 194 +---
 drivers/scsi/qla2xxx/qla_bsg.c   | 264 ++-
 drivers/scsi/qla2xxx/qla_iocb.c  |   5 +-
 drivers/scsi/qla2xxx/qla_isr.c   |  46 ---
 drivers/scsi/qla2xxx/qla_mr.c|  10 +-
 drivers/scsi/scsi_transport_fc.c |  37 +++---
 10 files changed, 387 insertions(+), 263 deletions(-)

diff --git a/drivers/s390/scsi/zfcp_fc.c b/drivers/s390/scsi/zfcp_fc.c
index 237688a..4c4023f 100644
--- a/drivers/s390/scsi/zfcp_fc.c
+++ b/drivers/s390/scsi/zfcp_fc.c
@@ -900,8 +900,9 @@ static struct zfcp_fc_wka_port *zfcp_fc_job_wka_port(struct 
fc_bsg_job *job)
u32 preamble_word1;
u8 gs_type;
struct zfcp_adapter *adapter;
+   struct fc_bsg_request *bsg_request = job->request;
 
-   preamble_word1 = job->request->rqst_data.r_ct.preamble_word1;
+   preamble_word1 = bsg_request->rqst_data.r_ct.preamble_word1;
gs_type = (preamble_word1 & 0xff00) >> 24;
 
adapter = (struct zfcp_adapter *) job->shost->hostdata[0];
@@ -938,6 +939,7 @@ static int zfcp_fc_exec_els_job(struct fc_bsg_job *job,
 {
struct zfcp_fsf_ct_els *els = job->dd_data;
struct fc_rport *rport = job->rport;
+   struct fc_bsg_request *bsg_request = job->request;
struct zfcp_port *port;
u32 d_id;
 
@@ -949,7 +951,7 @@ static int zfcp_fc_exec_els_job(struct fc_bsg_job *job,
d_id = port->d_id;
put_device(>dev);
} else
-   d_id = ntoh24(job->request->rqst_data.h_els.port_id);
+   d_id = ntoh24(bsg_request->rqst_data.h_els.port_id);
 
els->handler = zfcp_fc_ct_els_job_handler;
return zfcp_fsf_send_els(adapter, d_id, els, job->req->timeout / HZ);
@@ -983,6 +985,7 @@ int zfcp_fc_exec_bsg_job(struct fc_bsg_job *job)
struct Scsi_Host *shost;
struct zfcp_adapter *adapter;
struct zfcp_fsf_ct_els *ct_els = job->dd_data;
+   struct fc_bsg_request *bsg_request = job->request;
 
shost = job->rport ? rport_to_shost(job->rport) : job->shost;
adapter = (struct zfcp_adapter *)shost->hostdata[0];
@@ -994,7 +997,7 @@ int zfcp_fc_exec_bsg_job(struct fc_bsg_job *job)
ct_els->resp = job->reply_payload.sg_list;
ct_els->handler_data = job;
 
-   switch (job->request->msgcode) {
+   switch (bsg_request->msgcode) {
case FC_BSG_RPT_ELS:
case FC_BSG_HST_ELS_NOLOGIN:
return zfcp_fc_exec_els_job(job, adapter);
diff --git a/drivers/scsi/bfa/bfad_bsg.c b/drivers/scsi/bfa/bfad_bsg.c
index d1ad020..48366d8 100644
--- a/drivers/scsi/bfa/bfad_bsg.c
+++ b/drivers/scsi/bfa/bfad_bsg.c
@@ -3132,7 +3132,9 @@
 static int
 bfad_im_bsg_vendor_request(struct fc_bsg_job *job)
 {
-   uint32_t vendor_cmd = job->request->rqst_data.h_vendor.vendor_cmd[0];
+   struct fc_bsg_request *bsg_request = job->request;
+   struct fc_bsg_reply *bsg_reply = job->reply;
+   uint32_t vendor_cmd = bsg_request->rqst_data.h_vendor.vendor_cmd[0];
struct bfad_im_port_s *im_port =
(struct bfad_im_port_s *) job->shost->hostdata[0];
struct bfad_s *bfad = im_port->bfad;
@@ -3175,8 +3177,8 @@
 
/* Fill the BSG job reply data */
job->reply_len = job->reply_payload.payload_len;
-   job->reply->reply_payload_rcv_len = job->reply_payload.payload_len;
-   job->reply->result = rc;
+   bsg_reply->reply_payload_rcv_len = job->reply_payload.payload_len;
+   bsg_reply->result = rc;
 
job->job_done(job);
return rc;
@@ -3184,9 +3186,9 @@
/* free the command buffer */
kfree(payload_kbuf);
 out:
-   job->reply->result = rc;
+   bsg_reply->result = rc;
job->reply_len = sizeof(uint32_t);
-   job->reply->reply_payload_rcv_len = 0;
+   bsg_reply->reply_payload_rcv_len = 0;
return rc;
 }
 
@@ -3362,18 +3364,20 @@ struct bfad_buf_info *
struct bfad_fcxp*drv_fcxp;
struct bfa_fcs_lport_s *fcs_port;
struct bfa_fcs_rport_s *fcs_rport;
-   uint32_t command_type = job->request->msgcode;
+   struct fc_bsg_request *bsg_request = bsg_request;
+   struct fc_bsg_reply *bsg_reply = job->reply;
+   uint32_t command_type = bsg_request->msgcode;
unsigned long flags;
struct bfad_buf_info *rsp_buf_info;
void *req_kbuf = NULL, *rsp_kbuf = NULL;
int rc = -EINVAL;
 
job->reply_len  = sizeof(uint32_t); /* Atleast uint32_t reply_len 

[PATCH v3 03/16] scsi: fc: Export fc_bsg_jobdone and use it in FC drivers

[Johannes Thumshirn]

Export fc_bsg_jobdone so drivers can use it directly instead of doing
the round-trip via struct fc_bsg_job::job_done() and use it in the LLDDs.

As we've converted all LLDDs over to use fc_bsg_jobdone() directly,
we can remove the function pointer from struct fc_bsg_job as well.

Signed-off-by: Johannes Thumshirn 
Reviewed-by: Hannes Reinecke 
---
 drivers/s390/scsi/zfcp_fc.c  |  2 +-
 drivers/scsi/bfa/bfad_bsg.c  |  4 ++--
 drivers/scsi/ibmvscsi/ibmvfc.c   |  2 +-
 drivers/scsi/libfc/fc_lport.c|  4 ++--
 drivers/scsi/lpfc/lpfc_bsg.c | 38 +-
 drivers/scsi/qla2xxx/qla_bsg.c   | 44 
 drivers/scsi/scsi_transport_fc.c |  5 ++---
 include/scsi/scsi_transport_fc.h |  2 +-
 8 files changed, 50 insertions(+), 51 deletions(-)

diff --git a/drivers/s390/scsi/zfcp_fc.c b/drivers/s390/scsi/zfcp_fc.c
index 4c4023f..40d8f06 100644
--- a/drivers/s390/scsi/zfcp_fc.c
+++ b/drivers/s390/scsi/zfcp_fc.c
@@ -892,7 +892,7 @@ static void zfcp_fc_ct_els_job_handler(void *data)
jr->reply_payload_rcv_len = job->reply_payload.payload_len;
jr->reply_data.ctels_reply.status = FC_CTELS_STATUS_OK;
jr->result = zfcp_ct_els->status ? -EIO : 0;
-   job->job_done(job);
+   fc_bsg_jobdone(job);
 }
 
 static struct zfcp_fc_wka_port *zfcp_fc_job_wka_port(struct fc_bsg_job *job)
diff --git a/drivers/scsi/bfa/bfad_bsg.c b/drivers/scsi/bfa/bfad_bsg.c
index 48366d8..25889b9 100644
--- a/drivers/scsi/bfa/bfad_bsg.c
+++ b/drivers/scsi/bfa/bfad_bsg.c
@@ -3180,7 +3180,7 @@
bsg_reply->reply_payload_rcv_len = job->reply_payload.payload_len;
bsg_reply->result = rc;
 
-   job->job_done(job);
+   fc_bsg_jobdone(job);
return rc;
 error:
/* free the command buffer */
@@ -3556,7 +3556,7 @@ struct bfad_buf_info *
bsg_reply->result = rc;
 
if (rc == BFA_STATUS_OK)
-   job->job_done(job);
+   fc_bsg_jobdone(job);
 
return rc;
 }
diff --git a/drivers/scsi/ibmvscsi/ibmvfc.c b/drivers/scsi/ibmvscsi/ibmvfc.c
index 7c17a7e..f14e9c1 100644
--- a/drivers/scsi/ibmvscsi/ibmvfc.c
+++ b/drivers/scsi/ibmvscsi/ibmvfc.c
@@ -1945,7 +1945,7 @@ static int ibmvfc_bsg_request(struct fc_bsg_job *job)
ibmvfc_free_event(evt);
spin_unlock_irqrestore(vhost->host->host_lock, flags);
bsg_reply->result = rc;
-   job->job_done(job);
+   fc_bsg_jobdone(job);
rc = 0;
 out:
dma_unmap_sg(vhost->dev, job->request_payload.sg_list,
diff --git a/drivers/scsi/libfc/fc_lport.c b/drivers/scsi/libfc/fc_lport.c
index 23bb114..33e2881 100644
--- a/drivers/scsi/libfc/fc_lport.c
+++ b/drivers/scsi/libfc/fc_lport.c
@@ -1917,7 +1917,7 @@ static void fc_lport_bsg_resp(struct fc_seq *sp, struct 
fc_frame *fp,
-ECONNABORTED : -ETIMEDOUT;
job->reply_len = sizeof(uint32_t);
job->state_flags |= FC_RQST_STATE_DONE;
-   job->job_done(job);
+   fc_bsg_jobdone(job);
kfree(info);
return;
}
@@ -1952,7 +1952,7 @@ static void fc_lport_bsg_resp(struct fc_seq *sp, struct 
fc_frame *fp,
job->reply_payload.payload_len;
bsg_reply->result = 0;
job->state_flags |= FC_RQST_STATE_DONE;
-   job->job_done(job);
+   fc_bsg_jobdone(job);
kfree(info);
}
fc_frame_free(fp);
diff --git a/drivers/scsi/lpfc/lpfc_bsg.c b/drivers/scsi/lpfc/lpfc_bsg.c
index 27b5930..1db9cca 100644
--- a/drivers/scsi/lpfc/lpfc_bsg.c
+++ b/drivers/scsi/lpfc/lpfc_bsg.c
@@ -371,7 +371,7 @@ struct lpfc_dmabufext {
 
if (job) {
bsg_reply->result = rc;
-   job->job_done(job);
+   fc_bsg_jobdone(job);
}
return;
 }
@@ -644,7 +644,7 @@ struct lpfc_dmabufext {
 
if (job) {
bsg_reply->result = rc;
-   job->job_done(job);
+   fc_bsg_jobdone(job);
}
return;
 }
@@ -1136,7 +1136,7 @@ struct lpfc_dmabufext {
job->dd_data = NULL;
/* complete the job back to userspace */
spin_unlock_irqrestore(>ct_ev_lock, flags);
-   job->job_done(job);
+   fc_bsg_jobdone(job);
spin_lock_irqsave(>ct_ev_lock, flags);
}
}
@@ -1361,7 +1361,7 @@ struct lpfc_dmabufext {
spin_unlock_irqrestore(>ct_ev_lock, flags);
job->dd_data = NULL;
bsg_reply->result = 0;
-   job->job_done(job);
+   fc_bsg_jobdone(job);
return 0;
 
 job_error:
@@ -1458,7 +1458,7 @@ struct lpfc_dmabufext {
 
if (job) {
bsg_reply->result = rc;
-   job->job_done(job);
+   fc_bsg_jobdone(job);
}
return;
 }
@@ -1886,7 +1886,7 @@ 

[PATCH v3 00/16] Convert FibreChannel bsg code to use bsg-lib

[Johannes Thumshirn]

This series converts the current bsg usage in the FibreChannel drivers over
to use bsg-lib. SAS will follow once FC is in a good enough shape.

I did take some inspiration from a similar patchset from Mike Christie
dating back to 2011 but it's not a 1:1 copy. Patch 15/16 is heavily based
on his series and attribution is given to him in the commit message.

It is currently regression tested on FCoE using the 'fcns' and
'fcrls' utilities.  I'm still trying to figure out how to test the other
LLDDs. So any pointer from the respective maintainers are appreciated
although the LLDD changes are purely mechanical. All they do is change from
'struct fc_bsg_job' to 'struct bsg_job' and corresponding changes in order
to get the series bisectable.

The idea for this change arose when discussing racy sysfs handling the FC
bsg code with Christoph and is a next step in moving all bsg clients to
bsg-lib to eventually clean up the in kernel bsg API.

Changes to v1:
* Reduce the number of individual patches (44 -> 16)
* Fix s390 build failure (forgotten to kill fc_bsg_job from zfcp_ext.h)
* Fix build failure on configs without CONFIG_BLK_DEV_BSGLIB
* Make bsg_job_get() call kref_get_unless_zero() and use it in 
scsi_transport_fc.c

Changes to v2:
* Add Hannes' Reviewd-by tags
* Address Hannes' comment about kref handling in 8/16
* Move CONFIG_BLK_DEV_BSGLIB dependency to patch 12/16 as per 0day results

Johannes Thumshirn (16):
  scsi: Get rid of struct fc_bsg_buffer
  scsi: don't use fc_bsg_job::request and fc_bsg_job::reply directly
  scsi: fc: Export fc_bsg_jobdone and use it in FC drivers
  scsi: Unify interfaces of fc_bsg_jobdone and bsg_job_done
  scsi: fc: provide fc_bsg_to_shost() helper
  scsi: fc: provide fc_bsg_to_rport() helper
  scsi: libfc: don't set FC_RQST_STATE_DONE before calling
fc_bsg_jobdone()
  scsi: fc: implement kref backed reference counting
  block: add reference counting for struct bsg_job
  scsi: change FC drivers to use 'struct bsg_job'
  scsi: fc: Use bsg_destroy_job
  scsi: fc: use bsg_softirq_done
  scsi: fc: use bsg_job_done
  block: add bsg_job_put() and bsg_job_get()
  scsi: fc: move FC transport's bsg code to bsg-lib
  block: unexport bsg_softirq_done() again

 block/bsg-lib.c  |  23 +-
 drivers/s390/scsi/zfcp_ext.h |   4 +-
 drivers/s390/scsi/zfcp_fc.c  |  33 +--
 drivers/scsi/Kconfig |   1 +
 drivers/scsi/bfa/bfad_bsg.c  |  62 +++---
 drivers/scsi/bfa/bfad_im.h   |   4 +-
 drivers/scsi/ibmvscsi/ibmvfc.c   |  40 ++--
 drivers/scsi/libfc/fc_lport.c|  47 ++--
 drivers/scsi/lpfc/lpfc_bsg.c | 375 +++-
 drivers/scsi/lpfc/lpfc_crtn.h|   4 +-
 drivers/scsi/qla2xxx/qla_bsg.c   | 449 ++-
 drivers/scsi/qla2xxx/qla_def.h   |   2 +-
 drivers/scsi/qla2xxx/qla_gbl.h   |   4 +-
 drivers/scsi/qla2xxx/qla_iocb.c  |  13 +-
 drivers/scsi/qla2xxx/qla_isr.c   |  52 +++--
 drivers/scsi/qla2xxx/qla_mr.c|  15 +-
 drivers/scsi/scsi_transport_fc.c | 409 ++-
 include/linux/bsg-lib.h  |   4 +
 include/scsi/libfc.h |   2 +-
 include/scsi/scsi_transport_fc.h |  62 ++
 20 files changed, 748 insertions(+), 857 deletions(-)

-- 
1.8.5.6

--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH v3 10/16] scsi: change FC drivers to use 'struct bsg_job'

[Johannes Thumshirn]

Change FC drivers to use 'struct bsg_job' from bsg-lib.h instead of 'struct
fc_bsg_job' from scsi_transport_fc.h and remove 'struct fc_bsg_job'.

Signed-off-by: Johannes Thumshirn 
Reviewed-by: Hannes Reinecke 
---
 drivers/s390/scsi/zfcp_ext.h |  4 +--
 drivers/s390/scsi/zfcp_fc.c  | 15 
 drivers/scsi/bfa/bfad_bsg.c  | 10 +++---
 drivers/scsi/bfa/bfad_im.h   |  4 +--
 drivers/scsi/ibmvscsi/ibmvfc.c   |  9 ++---
 drivers/scsi/libfc/fc_lport.c| 10 +++---
 drivers/scsi/lpfc/lpfc_bsg.c | 74 
 drivers/scsi/lpfc/lpfc_crtn.h|  4 +--
 drivers/scsi/qla2xxx/qla_bsg.c   | 61 +
 drivers/scsi/qla2xxx/qla_def.h   |  2 +-
 drivers/scsi/qla2xxx/qla_gbl.h   |  4 +--
 drivers/scsi/qla2xxx/qla_iocb.c  |  8 ++---
 drivers/scsi/qla2xxx/qla_isr.c   |  6 ++--
 drivers/scsi/qla2xxx/qla_mr.c|  5 +--
 drivers/scsi/scsi_transport_fc.c | 20 +--
 include/scsi/libfc.h |  2 +-
 include/scsi/scsi_transport_fc.h | 63 ++
 17 files changed, 138 insertions(+), 163 deletions(-)

diff --git a/drivers/s390/scsi/zfcp_ext.h b/drivers/s390/scsi/zfcp_ext.h
index c8fed9f..968a0ab 100644
--- a/drivers/s390/scsi/zfcp_ext.h
+++ b/drivers/s390/scsi/zfcp_ext.h
@@ -84,8 +84,8 @@ extern void zfcp_fc_enqueue_event(struct zfcp_adapter *,
 extern void zfcp_fc_wka_ports_force_offline(struct zfcp_fc_wka_ports *);
 extern int zfcp_fc_gs_setup(struct zfcp_adapter *);
 extern void zfcp_fc_gs_destroy(struct zfcp_adapter *);
-extern int zfcp_fc_exec_bsg_job(struct fc_bsg_job *);
-extern int zfcp_fc_timeout_bsg_job(struct fc_bsg_job *);
+extern int zfcp_fc_exec_bsg_job(struct bsg_job *);
+extern int zfcp_fc_timeout_bsg_job(struct bsg_job *);
 extern void zfcp_fc_sym_name_update(struct work_struct *);
 extern unsigned int zfcp_fc_port_scan_backoff(void);
 extern void zfcp_fc_conditional_port_scan(struct zfcp_adapter *);
diff --git a/drivers/s390/scsi/zfcp_fc.c b/drivers/s390/scsi/zfcp_fc.c
index 1977a66..b1b4129 100644
--- a/drivers/s390/scsi/zfcp_fc.c
+++ b/drivers/s390/scsi/zfcp_fc.c
@@ -13,6 +13,7 @@
 #include 
 #include 
 #include 
+#include 
 #include 
 #include 
 #include "zfcp_ext.h"
@@ -885,7 +886,7 @@ void zfcp_fc_sym_name_update(struct work_struct *work)
 
 static void zfcp_fc_ct_els_job_handler(void *data)
 {
-   struct fc_bsg_job *job = data;
+   struct bsg_job *job = data;
struct zfcp_fsf_ct_els *zfcp_ct_els = job->dd_data;
struct fc_bsg_reply *jr = job->reply;
 
@@ -895,7 +896,7 @@ static void zfcp_fc_ct_els_job_handler(void *data)
fc_bsg_jobdone(job, jr->result, jr->reply_payload_rcv_len);
 }
 
-static struct zfcp_fc_wka_port *zfcp_fc_job_wka_port(struct fc_bsg_job *job)
+static struct zfcp_fc_wka_port *zfcp_fc_job_wka_port(struct bsg_job *job)
 {
u32 preamble_word1;
u8 gs_type;
@@ -925,7 +926,7 @@ static struct zfcp_fc_wka_port *zfcp_fc_job_wka_port(struct 
fc_bsg_job *job)
 
 static void zfcp_fc_ct_job_handler(void *data)
 {
-   struct fc_bsg_job *job = data;
+   struct bsg_job *job = data;
struct zfcp_fc_wka_port *wka_port;
 
wka_port = zfcp_fc_job_wka_port(job);
@@ -934,7 +935,7 @@ static void zfcp_fc_ct_job_handler(void *data)
zfcp_fc_ct_els_job_handler(data);
 }
 
-static int zfcp_fc_exec_els_job(struct fc_bsg_job *job,
+static int zfcp_fc_exec_els_job(struct bsg_job *job,
struct zfcp_adapter *adapter)
 {
struct zfcp_fsf_ct_els *els = job->dd_data;
@@ -957,7 +958,7 @@ static int zfcp_fc_exec_els_job(struct fc_bsg_job *job,
return zfcp_fsf_send_els(adapter, d_id, els, job->req->timeout / HZ);
 }
 
-static int zfcp_fc_exec_ct_job(struct fc_bsg_job *job,
+static int zfcp_fc_exec_ct_job(struct bsg_job *job,
   struct zfcp_adapter *adapter)
 {
int ret;
@@ -980,7 +981,7 @@ static int zfcp_fc_exec_ct_job(struct fc_bsg_job *job,
return ret;
 }
 
-int zfcp_fc_exec_bsg_job(struct fc_bsg_job *job)
+int zfcp_fc_exec_bsg_job(struct bsg_job *job)
 {
struct Scsi_Host *shost;
struct zfcp_adapter *adapter;
@@ -1010,7 +1011,7 @@ int zfcp_fc_exec_bsg_job(struct fc_bsg_job *job)
}
 }
 
-int zfcp_fc_timeout_bsg_job(struct fc_bsg_job *job)
+int zfcp_fc_timeout_bsg_job(struct bsg_job *job)
 {
/* hardware tracks timeout, reset bsg timeout to not interfere */
return -EAGAIN;
diff --git a/drivers/scsi/bfa/bfad_bsg.c b/drivers/scsi/bfa/bfad_bsg.c
index d3094270..cdc25e6 100644
--- a/drivers/scsi/bfa/bfad_bsg.c
+++ b/drivers/scsi/bfa/bfad_bsg.c
@@ -3130,7 +3130,7 @@
 }
 
 static int
-bfad_im_bsg_vendor_request(struct fc_bsg_job *job)
+bfad_im_bsg_vendor_request(struct bsg_job *job)
 {
struct fc_bsg_request *bsg_request = job->request;
struct fc_bsg_reply *bsg_reply = job->reply;
@@ -3314,7 +3314,7 @@ struct bfad_buf_info *
 }
 
 int

[PATCH v3 13/16] scsi: fc: use bsg_job_done

[Johannes Thumshirn]

fc_bsg_jobdone() and bsg_job_done() are 1:1 copies now so use the bsg-lib one
instead of the FC private implementation.

Signed-off-by: Johannes Thumshirn 
Reviewed-by: Hannes Reinecke 
---
 drivers/s390/scsi/zfcp_fc.c  |  2 +-
 drivers/scsi/bfa/bfad_bsg.c  |  4 ++--
 drivers/scsi/ibmvscsi/ibmvfc.c   |  2 +-
 drivers/scsi/libfc/fc_lport.c|  4 ++--
 drivers/scsi/lpfc/lpfc_bsg.c | 38 +-
 drivers/scsi/qla2xxx/qla_bsg.c   | 44 
 drivers/scsi/scsi_transport_fc.c | 41 +++--
 include/scsi/scsi_transport_fc.h |  2 --
 8 files changed, 50 insertions(+), 87 deletions(-)

diff --git a/drivers/s390/scsi/zfcp_fc.c b/drivers/s390/scsi/zfcp_fc.c
index b1b4129..a0f9c82 100644
--- a/drivers/s390/scsi/zfcp_fc.c
+++ b/drivers/s390/scsi/zfcp_fc.c
@@ -893,7 +893,7 @@ static void zfcp_fc_ct_els_job_handler(void *data)
jr->reply_payload_rcv_len = job->reply_payload.payload_len;
jr->reply_data.ctels_reply.status = FC_CTELS_STATUS_OK;
jr->result = zfcp_ct_els->status ? -EIO : 0;
-   fc_bsg_jobdone(job, jr->result, jr->reply_payload_rcv_len);
+   bsg_job_done(job, jr->result, jr->reply_payload_rcv_len);
 }
 
 static struct zfcp_fc_wka_port *zfcp_fc_job_wka_port(struct bsg_job *job)
diff --git a/drivers/scsi/bfa/bfad_bsg.c b/drivers/scsi/bfa/bfad_bsg.c
index cdc25e6..a9a0016 100644
--- a/drivers/scsi/bfa/bfad_bsg.c
+++ b/drivers/scsi/bfa/bfad_bsg.c
@@ -3179,7 +3179,7 @@
bsg_reply->reply_payload_rcv_len = job->reply_payload.payload_len;
bsg_reply->result = rc;
 
-   fc_bsg_jobdone(job, bsg_reply->result,
+   bsg_job_done(job, bsg_reply->result,
   bsg_reply->reply_payload_rcv_len);
return rc;
 error:
@@ -3555,7 +3555,7 @@ struct bfad_buf_info *
bsg_reply->result = rc;
 
if (rc == BFA_STATUS_OK)
-   fc_bsg_jobdone(job, bsg_reply->result,
+   bsg_job_done(job, bsg_reply->result,
   bsg_reply->reply_payload_rcv_len);
 
return rc;
diff --git a/drivers/scsi/ibmvscsi/ibmvfc.c b/drivers/scsi/ibmvscsi/ibmvfc.c
index f59b0a1..78b72c2 100644
--- a/drivers/scsi/ibmvscsi/ibmvfc.c
+++ b/drivers/scsi/ibmvscsi/ibmvfc.c
@@ -1946,7 +1946,7 @@ static int ibmvfc_bsg_request(struct bsg_job *job)
ibmvfc_free_event(evt);
spin_unlock_irqrestore(vhost->host->host_lock, flags);
bsg_reply->result = rc;
-   fc_bsg_jobdone(job, bsg_reply->result,
+   bsg_job_done(job, bsg_reply->result,
   bsg_reply->reply_payload_rcv_len);
rc = 0;
 out:
diff --git a/drivers/scsi/libfc/fc_lport.c b/drivers/scsi/libfc/fc_lport.c
index 9845667..518e3b0 100644
--- a/drivers/scsi/libfc/fc_lport.c
+++ b/drivers/scsi/libfc/fc_lport.c
@@ -1916,7 +1916,7 @@ static void fc_lport_bsg_resp(struct fc_seq *sp, struct 
fc_frame *fp,
bsg_reply->result = (PTR_ERR(fp) == -FC_EX_CLOSED) ?
-ECONNABORTED : -ETIMEDOUT;
job->reply_len = sizeof(uint32_t);
-   fc_bsg_jobdone(job, bsg_reply->result,
+   bsg_job_done(job, bsg_reply->result,
   bsg_reply->reply_payload_rcv_len);
kfree(info);
return;
@@ -1951,7 +1951,7 @@ static void fc_lport_bsg_resp(struct fc_seq *sp, struct 
fc_frame *fp,
bsg_reply->reply_payload_rcv_len =
job->reply_payload.payload_len;
bsg_reply->result = 0;
-   fc_bsg_jobdone(job, bsg_reply->result,
+   bsg_job_done(job, bsg_reply->result,
   bsg_reply->reply_payload_rcv_len);
kfree(info);
}
diff --git a/drivers/scsi/lpfc/lpfc_bsg.c b/drivers/scsi/lpfc/lpfc_bsg.c
index ca21f25..a862437 100644
--- a/drivers/scsi/lpfc/lpfc_bsg.c
+++ b/drivers/scsi/lpfc/lpfc_bsg.c
@@ -371,7 +371,7 @@ struct lpfc_dmabufext {
 
if (job) {
bsg_reply->result = rc;
-   fc_bsg_jobdone(job, bsg_reply->result,
+   bsg_job_done(job, bsg_reply->result,
   bsg_reply->reply_payload_rcv_len);
}
return;
@@ -645,7 +645,7 @@ struct lpfc_dmabufext {
 
if (job) {
bsg_reply->result = rc;
-   fc_bsg_jobdone(job, bsg_reply->result,
+   bsg_job_done(job, bsg_reply->result,
   bsg_reply->reply_payload_rcv_len);
}
return;
@@ -1138,7 +1138,7 @@ struct lpfc_dmabufext {
job->dd_data = NULL;
/* complete the job back to userspace */
spin_unlock_irqrestore(>ct_ev_lock, flags);
-   fc_bsg_jobdone(job, bsg_reply->result,
+   bsg_job_done(job, bsg_reply->result,

[PATCH v3 08/16] scsi: fc: implement kref backed reference counting

[Johannes Thumshirn]

Implement kref backed reference counting instead of rolling our own. This
elimnates the need of the following fields in 'struct fc_bsg_job':
* ref_cnt
* state_flags
* job_lock
bringing us close to unification of 'struct fc_bsg_job' and 'struct bsg_job'.

Signed-off-by: Johannes Thumshirn 
---
 drivers/scsi/scsi_transport_fc.c | 40 +++-
 include/scsi/scsi_transport_fc.h |  4 +---
 2 files changed, 12 insertions(+), 32 deletions(-)

diff --git a/drivers/scsi/scsi_transport_fc.c b/drivers/scsi/scsi_transport_fc.c
index 96b3a2e..610a985 100644
--- a/drivers/scsi/scsi_transport_fc.c
+++ b/drivers/scsi/scsi_transport_fc.c
@@ -3560,16 +3560,12 @@ struct fc_vport *
  * @job:   fc_bsg_job that is to be torn down
  */
 static void
-fc_destroy_bsgjob(struct fc_bsg_job *job)
+fc_destroy_bsgjob(struct kref *kref)
 {
-   unsigned long flags;
+   struct fc_bsg_job *job = container_of(kref, struct fc_bsg_job, kref);
+   struct request *rq = job->req;
 
-   spin_lock_irqsave(>job_lock, flags);
-   if (job->ref_cnt) {
-   spin_unlock_irqrestore(>job_lock, flags);
-   return;
-   }
-   spin_unlock_irqrestore(>job_lock, flags);
+   blk_end_request_all(rq, rq->errors);
 
put_device(job->dev);   /* release reference for the request */
 
@@ -3620,15 +3616,8 @@ void fc_bsg_jobdone(struct fc_bsg_job *job, int result,
 static void fc_bsg_softirq_done(struct request *rq)
 {
struct fc_bsg_job *job = rq->special;
-   unsigned long flags;
 
-   spin_lock_irqsave(>job_lock, flags);
-   job->state_flags |= FC_RQST_STATE_DONE;
-   job->ref_cnt--;
-   spin_unlock_irqrestore(>job_lock, flags);
-
-   blk_end_request_all(rq, rq->errors);
-   fc_destroy_bsgjob(job);
+   kref_put(>kref, fc_destroy_bsgjob);
 }
 
 /**
@@ -3642,24 +3631,18 @@ static void fc_bsg_softirq_done(struct request *rq)
struct Scsi_Host *shost = fc_bsg_to_shost(job);
struct fc_rport *rport = fc_bsg_to_rport(job);
struct fc_internal *i = to_fc_internal(shost->transportt);
-   unsigned long flags;
-   int err = 0, done = 0;
+   int err = 0, inflight = 0;
 
if (rport && rport->port_state == FC_PORTSTATE_BLOCKED)
return BLK_EH_RESET_TIMER;
 
-   spin_lock_irqsave(>job_lock, flags);
-   if (job->state_flags & FC_RQST_STATE_DONE)
-   done = 1;
-   else
-   job->ref_cnt++;
-   spin_unlock_irqrestore(>job_lock, flags);
+   inflight = kref_get_unless_zero(>kref);
 
-   if (!done && i->f->bsg_timeout) {
+   if (inflight && i->f->bsg_timeout) {
/* call LLDD to abort the i/o as it has timed out */
err = i->f->bsg_timeout(job);
if (err == -EAGAIN) {
-   job->ref_cnt--;
+   kref_put(>kref, fc_destroy_bsgjob);
return BLK_EH_RESET_TIMER;
} else if (err)
printk(KERN_ERR "ERROR: FC BSG request timeout - LLD "
@@ -3667,7 +3650,7 @@ static void fc_bsg_softirq_done(struct request *rq)
}
 
/* the blk_end_sync_io() doesn't check the error */
-   if (done)
+   if (!inflight)
return BLK_EH_NOT_HANDLED;
else
return BLK_EH_HANDLED;
@@ -3730,7 +3713,6 @@ static void fc_bsg_softirq_done(struct request *rq)
job->req = req;
if (i->f->dd_bsg_size)
job->dd_data = (void *)[1];
-   spin_lock_init(>job_lock);
bsg_request = (struct fc_bsg_request *)req->cmd;
job->request_len = req->cmd_len;
bsg_reply = req->sense;
@@ -3752,7 +3734,7 @@ static void fc_bsg_softirq_done(struct request *rq)
job->dev = >shost_gendev;
get_device(job->dev);   /* take a reference for the request */
 
-   job->ref_cnt = 1;
+   kref_init(>kref);
 
return 0;
 
diff --git a/include/scsi/scsi_transport_fc.h b/include/scsi/scsi_transport_fc.h
index 9f53fe3..8ae5680 100644
--- a/include/scsi/scsi_transport_fc.h
+++ b/include/scsi/scsi_transport_fc.h
@@ -634,9 +634,7 @@ struct fc_bsg_job {
struct fc_rport *rport;
struct device *dev;
struct request *req;
-   spinlock_t job_lock;
-   unsigned int state_flags;
-   unsigned int ref_cnt;
+   struct kref kref;
 
struct fc_bsg_request *request;
struct fc_bsg_reply *reply;
-- 
1.8.5.6

--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH v3 12/16] scsi: fc: use bsg_softirq_done

[Johannes Thumshirn]

bsg_softirq_done() and fc_bsg_softirq_done() are copies of each other, so
ditch the fc specific one.

Signed-off-by: Johannes Thumshirn 
Reviewed-by: Hannes Reinecke 
---
 block/bsg-lib.c  |  3 ++-
 drivers/scsi/scsi_transport_fc.c | 15 ++-
 include/linux/bsg-lib.h  |  1 +
 3 files changed, 5 insertions(+), 14 deletions(-)

diff --git a/block/bsg-lib.c b/block/bsg-lib.c
index 9f1e8fd..6661f82 100644
--- a/block/bsg-lib.c
+++ b/block/bsg-lib.c
@@ -85,12 +85,13 @@ void bsg_job_done(struct bsg_job *job, int result,
  * bsg_softirq_done - softirq done routine for destroying the bsg requests
  * @rq: BSG request that holds the job to be destroyed
  */
-static void bsg_softirq_done(struct request *rq)
+void bsg_softirq_done(struct request *rq)
 {
struct bsg_job *job = rq->special;
 
kref_put(>kref, bsg_destroy_job);
 }
+EXPORT_SYMBOL_GPL(bsg_softirq_done);
 
 static int bsg_map_buffer(struct bsg_buffer *buf, struct request *req)
 {
diff --git a/drivers/scsi/scsi_transport_fc.c b/drivers/scsi/scsi_transport_fc.c
index d2d0a27..1ae6d86 100644
--- a/drivers/scsi/scsi_transport_fc.c
+++ b/drivers/scsi/scsi_transport_fc.c
@@ -3591,17 +3591,6 @@ void fc_bsg_jobdone(struct bsg_job *job, int result,
 EXPORT_SYMBOL_GPL(fc_bsg_jobdone);
 
 /**
- * fc_bsg_softirq_done - softirq done routine for destroying the bsg requests
- * @rq:BSG request that holds the job to be destroyed
- */
-static void fc_bsg_softirq_done(struct request *rq)
-{
-   struct bsg_job *job = rq->special;
-
-   kref_put(>kref, fc_destroy_bsgjob);
-}
-
-/**
  * fc_bsg_job_timeout - handler for when a bsg request timesout
  * @req:   request that timed out
  */
@@ -4035,7 +4024,7 @@ enum fc_dispatch_result {
 
q->queuedata = shost;
queue_flag_set_unlocked(QUEUE_FLAG_BIDI, q);
-   blk_queue_softirq_done(q, fc_bsg_softirq_done);
+   blk_queue_softirq_done(q, bsg_softirq_done);
blk_queue_rq_timed_out(q, fc_bsg_job_timeout);
blk_queue_rq_timeout(q, FC_DEFAULT_BSG_TIMEOUT);
 
@@ -4081,7 +4070,7 @@ enum fc_dispatch_result {
 
q->queuedata = rport;
queue_flag_set_unlocked(QUEUE_FLAG_BIDI, q);
-   blk_queue_softirq_done(q, fc_bsg_softirq_done);
+   blk_queue_softirq_done(q, bsg_softirq_done);
blk_queue_rq_timed_out(q, fc_bsg_job_timeout);
blk_queue_rq_timeout(q, BLK_DEFAULT_SG_TIMEOUT);
 
diff --git a/include/linux/bsg-lib.h b/include/linux/bsg-lib.h
index 67f7de6..09f3044 100644
--- a/include/linux/bsg-lib.h
+++ b/include/linux/bsg-lib.h
@@ -70,5 +70,6 @@ int bsg_setup_queue(struct device *dev, struct request_queue 
*q, char *name,
bsg_job_fn *job_fn, int dd_job_size);
 void bsg_request_fn(struct request_queue *q);
 void bsg_destroy_job(struct kref *kref);
+void bsg_softirq_done(struct request *rq);
 
 #endif
-- 
1.8.5.6

--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH v3 05/16] scsi: fc: provide fc_bsg_to_shost() helper

[Johannes Thumshirn]

Provide fc_bsg_to_shost() helper that will become handy when we're moving from
struct fc_bsg_job to a plain struct bsg_job. Also use this little helper in
the LLDDs.

Signed-off-by: Johannes Thumshirn 
Reviewed-by: Hannes Reinecke 
---
 drivers/s390/scsi/zfcp_fc.c  |  4 +--
 drivers/scsi/bfa/bfad_bsg.c  |  6 ++---
 drivers/scsi/ibmvscsi/ibmvfc.c   |  4 +--
 drivers/scsi/libfc/fc_lport.c|  2 +-
 drivers/scsi/lpfc/lpfc_bsg.c | 32 
 drivers/scsi/qla2xxx/qla_bsg.c   | 54 
 drivers/scsi/scsi_transport_fc.c |  2 +-
 include/scsi/scsi_transport_fc.h |  5 
 8 files changed, 56 insertions(+), 53 deletions(-)

diff --git a/drivers/s390/scsi/zfcp_fc.c b/drivers/s390/scsi/zfcp_fc.c
index 87f6330..813c286 100644
--- a/drivers/s390/scsi/zfcp_fc.c
+++ b/drivers/s390/scsi/zfcp_fc.c
@@ -905,7 +905,7 @@ static struct zfcp_fc_wka_port *zfcp_fc_job_wka_port(struct 
fc_bsg_job *job)
preamble_word1 = bsg_request->rqst_data.r_ct.preamble_word1;
gs_type = (preamble_word1 & 0xff00) >> 24;
 
-   adapter = (struct zfcp_adapter *) job->shost->hostdata[0];
+   adapter = shost_priv(fc_bsg_to_shost(job));
 
switch (gs_type) {
case FC_FST_ALIAS:
@@ -987,7 +987,7 @@ int zfcp_fc_exec_bsg_job(struct fc_bsg_job *job)
struct zfcp_fsf_ct_els *ct_els = job->dd_data;
struct fc_bsg_request *bsg_request = job->request;
 
-   shost = job->rport ? rport_to_shost(job->rport) : job->shost;
+   shost = job->rport ? rport_to_shost(job->rport) : fc_bsg_to_shost(job);
adapter = (struct zfcp_adapter *)shost->hostdata[0];
 
if (!(atomic_read(>status) & ZFCP_STATUS_COMMON_OPEN))
diff --git a/drivers/scsi/bfa/bfad_bsg.c b/drivers/scsi/bfa/bfad_bsg.c
index e49a6c8..d3094270 100644
--- a/drivers/scsi/bfa/bfad_bsg.c
+++ b/drivers/scsi/bfa/bfad_bsg.c
@@ -3135,8 +3135,7 @@
struct fc_bsg_request *bsg_request = job->request;
struct fc_bsg_reply *bsg_reply = job->reply;
uint32_t vendor_cmd = bsg_request->rqst_data.h_vendor.vendor_cmd[0];
-   struct bfad_im_port_s *im_port =
-   (struct bfad_im_port_s *) job->shost->hostdata[0];
+   struct bfad_im_port_s *im_port = shost_priv(fc_bsg_to_shost(job));
struct bfad_s *bfad = im_port->bfad;
struct request_queue *request_q = job->req->q;
void *payload_kbuf;
@@ -3358,8 +3357,7 @@ struct bfad_buf_info *
 bfad_im_bsg_els_ct_request(struct fc_bsg_job *job)
 {
struct bfa_bsg_data *bsg_data;
-   struct bfad_im_port_s *im_port =
-   (struct bfad_im_port_s *) job->shost->hostdata[0];
+   struct bfad_im_port_s *im_port = shost_priv(fc_bsg_to_shost(job));
struct bfad_s *bfad = im_port->bfad;
bfa_bsg_fcpt_t *bsg_fcpt;
struct bfad_fcxp*drv_fcxp;
diff --git a/drivers/scsi/ibmvscsi/ibmvfc.c b/drivers/scsi/ibmvscsi/ibmvfc.c
index 6070361..02df1f1 100644
--- a/drivers/scsi/ibmvscsi/ibmvfc.c
+++ b/drivers/scsi/ibmvscsi/ibmvfc.c
@@ -1708,7 +1708,7 @@ static void ibmvfc_bsg_timeout_done(struct ibmvfc_event 
*evt)
  **/
 static int ibmvfc_bsg_timeout(struct fc_bsg_job *job)
 {
-   struct ibmvfc_host *vhost = shost_priv(job->shost);
+   struct ibmvfc_host *vhost = shost_priv(fc_bsg_to_shost(job));
unsigned long port_id = (unsigned long)job->dd_data;
struct ibmvfc_event *evt;
struct ibmvfc_tmf *tmf;
@@ -1821,7 +1821,7 @@ static int ibmvfc_bsg_plogi(struct ibmvfc_host *vhost, 
unsigned int port_id)
  **/
 static int ibmvfc_bsg_request(struct fc_bsg_job *job)
 {
-   struct ibmvfc_host *vhost = shost_priv(job->shost);
+   struct ibmvfc_host *vhost = shost_priv(fc_bsg_to_shost(job));
struct fc_rport *rport = job->rport;
struct ibmvfc_passthru_mad *mad;
struct ibmvfc_event *evt;
diff --git a/drivers/scsi/libfc/fc_lport.c b/drivers/scsi/libfc/fc_lport.c
index 000188f..a3fb355 100644
--- a/drivers/scsi/libfc/fc_lport.c
+++ b/drivers/scsi/libfc/fc_lport.c
@@ -2092,7 +2092,7 @@ int fc_lport_bsg_request(struct fc_bsg_job *job)
struct fc_bsg_request *bsg_request = job->request;
struct fc_bsg_reply *bsg_reply = job->reply;
struct request *rsp = job->req->next_rq;
-   struct Scsi_Host *shost = job->shost;
+   struct Scsi_Host *shost = fc_bsg_to_shost(job);
struct fc_lport *lport = shost_priv(shost);
struct fc_rport *rport;
struct fc_rport_priv *rdata;
diff --git a/drivers/scsi/lpfc/lpfc_bsg.c b/drivers/scsi/lpfc/lpfc_bsg.c
index 447a7af..bfcc37d 100644
--- a/drivers/scsi/lpfc/lpfc_bsg.c
+++ b/drivers/scsi/lpfc/lpfc_bsg.c
@@ -384,7 +384,7 @@ struct lpfc_dmabufext {
 static int
 lpfc_bsg_send_mgmt_cmd(struct fc_bsg_job *job)
 {
-   struct lpfc_vport *vport = (struct lpfc_vport *)job->shost->hostdata;
+   struct lpfc_vport *vport = shost_priv(fc_bsg_to_shost(job));
struct lpfc_hba *phba = 

[PATCH v3 06/16] scsi: fc: provide fc_bsg_to_rport() helper

[Johannes Thumshirn]

Provide fc_bsg_to_rport() helper that will become handy when we're moving
from struct fc_bsg_job to a plain struct bsg_job. Also move all LLDDs to use
the new helper.

Signed-off-by: Johannes Thumshirn 
Reviewed-by: Hannes Reinecke 
---
 drivers/s390/scsi/zfcp_fc.c  | 5 +++--
 drivers/scsi/ibmvscsi/ibmvfc.c   | 2 +-
 drivers/scsi/libfc/fc_lport.c| 4 ++--
 drivers/scsi/lpfc/lpfc_bsg.c | 4 ++--
 drivers/scsi/qla2xxx/qla_bsg.c   | 4 ++--
 drivers/scsi/scsi_transport_fc.c | 3 ++-
 include/scsi/scsi_transport_fc.h | 5 +
 7 files changed, 17 insertions(+), 10 deletions(-)

diff --git a/drivers/s390/scsi/zfcp_fc.c b/drivers/s390/scsi/zfcp_fc.c
index 813c286..1977a66 100644
--- a/drivers/s390/scsi/zfcp_fc.c
+++ b/drivers/s390/scsi/zfcp_fc.c
@@ -938,7 +938,7 @@ static int zfcp_fc_exec_els_job(struct fc_bsg_job *job,
struct zfcp_adapter *adapter)
 {
struct zfcp_fsf_ct_els *els = job->dd_data;
-   struct fc_rport *rport = job->rport;
+   struct fc_rport *rport = fc_bsg_to_rport(job);
struct fc_bsg_request *bsg_request = job->request;
struct zfcp_port *port;
u32 d_id;
@@ -986,8 +986,9 @@ int zfcp_fc_exec_bsg_job(struct fc_bsg_job *job)
struct zfcp_adapter *adapter;
struct zfcp_fsf_ct_els *ct_els = job->dd_data;
struct fc_bsg_request *bsg_request = job->request;
+   struct fc_rport *rport = fc_bsg_to_rport(job);
 
-   shost = job->rport ? rport_to_shost(job->rport) : fc_bsg_to_shost(job);
+   shost = rport ? rport_to_shost(rport) : fc_bsg_to_shost(job);
adapter = (struct zfcp_adapter *)shost->hostdata[0];
 
if (!(atomic_read(>status) & ZFCP_STATUS_COMMON_OPEN))
diff --git a/drivers/scsi/ibmvscsi/ibmvfc.c b/drivers/scsi/ibmvscsi/ibmvfc.c
index 02df1f1..4c73fc7 100644
--- a/drivers/scsi/ibmvscsi/ibmvfc.c
+++ b/drivers/scsi/ibmvscsi/ibmvfc.c
@@ -1822,7 +1822,7 @@ static int ibmvfc_bsg_plogi(struct ibmvfc_host *vhost, 
unsigned int port_id)
 static int ibmvfc_bsg_request(struct fc_bsg_job *job)
 {
struct ibmvfc_host *vhost = shost_priv(fc_bsg_to_shost(job));
-   struct fc_rport *rport = job->rport;
+   struct fc_rport *rport = fc_bsg_to_rport(job);
struct ibmvfc_passthru_mad *mad;
struct ibmvfc_event *evt;
union ibmvfc_iu rsp_iu;
diff --git a/drivers/scsi/libfc/fc_lport.c b/drivers/scsi/libfc/fc_lport.c
index a3fb355..9ba340c 100644
--- a/drivers/scsi/libfc/fc_lport.c
+++ b/drivers/scsi/libfc/fc_lport.c
@@ -2107,7 +2107,7 @@ int fc_lport_bsg_request(struct fc_bsg_job *job)
 
switch (bsg_request->msgcode) {
case FC_BSG_RPT_ELS:
-   rport = job->rport;
+   rport = fc_bsg_to_rport(job);
if (!rport)
break;
 
@@ -2117,7 +2117,7 @@ int fc_lport_bsg_request(struct fc_bsg_job *job)
break;
 
case FC_BSG_RPT_CT:
-   rport = job->rport;
+   rport = fc_bsg_to_rport(job);
if (!rport)
break;
 
diff --git a/drivers/scsi/lpfc/lpfc_bsg.c b/drivers/scsi/lpfc/lpfc_bsg.c
index bfcc37d..dae7cc3 100644
--- a/drivers/scsi/lpfc/lpfc_bsg.c
+++ b/drivers/scsi/lpfc/lpfc_bsg.c
@@ -386,7 +386,7 @@ struct lpfc_dmabufext {
 {
struct lpfc_vport *vport = shost_priv(fc_bsg_to_shost(job));
struct lpfc_hba *phba = vport->phba;
-   struct lpfc_rport_data *rdata = job->rport->dd_data;
+   struct lpfc_rport_data *rdata = fc_bsg_to_rport(job)->dd_data;
struct lpfc_nodelist *ndlp = rdata->pnode;
struct fc_bsg_reply *bsg_reply = job->reply;
struct ulp_bde64 *bpl = NULL;
@@ -660,7 +660,7 @@ struct lpfc_dmabufext {
 {
struct lpfc_vport *vport = shost_priv(fc_bsg_to_shost(job));
struct lpfc_hba *phba = vport->phba;
-   struct lpfc_rport_data *rdata = job->rport->dd_data;
+   struct lpfc_rport_data *rdata = fc_bsg_to_rport(job)->dd_data;
struct lpfc_nodelist *ndlp = rdata->pnode;
struct fc_bsg_request *bsg_request = job->request;
struct fc_bsg_reply *bsg_reply = job->reply;
diff --git a/drivers/scsi/qla2xxx/qla_bsg.c b/drivers/scsi/qla2xxx/qla_bsg.c
index 109b852..917eafe 100644
--- a/drivers/scsi/qla2xxx/qla_bsg.c
+++ b/drivers/scsi/qla2xxx/qla_bsg.c
@@ -264,7 +264,7 @@
uint16_t nextlid = 0;
 
if (bsg_request->msgcode == FC_BSG_RPT_ELS) {
-   rport = bsg_job->rport;
+   rport = fc_bsg_to_rport(bsg_job);
fcport = *(fc_port_t **) rport->dd_data;
host = rport_to_shost(rport);
vha = shost_priv(host);
@@ -2485,7 +2485,7 @@
bsg_reply->reply_payload_rcv_len = 0;
 
if (bsg_request->msgcode == FC_BSG_RPT_ELS) {
-   rport = bsg_job->rport;
+   rport = fc_bsg_to_rport(bsg_job);
host = rport_to_shost(rport);
vha = shost_priv(host);
} else {

[PATCH v3 09/16] block: add reference counting for struct bsg_job

[Johannes Thumshirn]

Add reference counting to 'struct bsg_job' so we can implement a reuqest
timeout handler for bsg_jobs, which is needed for Fibre Channel.

Signed-off-by: Johannes Thumshirn 
Reviewed-by: Hannes Reinecke 
---
 block/bsg-lib.c | 7 +--
 include/linux/bsg-lib.h | 2 ++
 2 files changed, 7 insertions(+), 2 deletions(-)

diff --git a/block/bsg-lib.c b/block/bsg-lib.c
index 650f427..632fb40 100644
--- a/block/bsg-lib.c
+++ b/block/bsg-lib.c
@@ -32,8 +32,10 @@
  * bsg_destroy_job - routine to teardown/delete a bsg job
  * @job: bsg_job that is to be torn down
  */
-static void bsg_destroy_job(struct bsg_job *job)
+static void bsg_destroy_job(struct kref *kref)
 {
+   struct bsg_job *job = container_of(kref, struct bsg_job, kref);
+
put_device(job->dev);   /* release reference for the request */
 
kfree(job->request_payload.sg_list);
@@ -84,7 +86,7 @@ static void bsg_softirq_done(struct request *rq)
struct bsg_job *job = rq->special;
 
blk_end_request_all(rq, rq->errors);
-   bsg_destroy_job(job);
+   kref_put(>kref, bsg_destroy_job);
 }
 
 static int bsg_map_buffer(struct bsg_buffer *buf, struct request *req)
@@ -142,6 +144,7 @@ static int bsg_create_job(struct device *dev, struct 
request *req)
job->dev = dev;
/* take a reference for the request */
get_device(job->dev);
+   kref_init(>kref);
return 0;
 
 failjob_rls_rqst_payload:
diff --git a/include/linux/bsg-lib.h b/include/linux/bsg-lib.h
index a226652..58e0717 100644
--- a/include/linux/bsg-lib.h
+++ b/include/linux/bsg-lib.h
@@ -40,6 +40,8 @@ struct bsg_job {
struct device *dev;
struct request *req;
 
+   struct kref kref;
+
/* Transport/driver specific request/reply structs */
void *request;
void *reply;
-- 
1.8.5.6

--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH v3 14/16] block: add bsg_job_put() and bsg_job_get()

[Johannes Thumshirn]

Add bsg_job_put() and bsg_job_get() so don't need to export
bsg_destroy_job() any more.

Signed-off-by: Johannes Thumshirn 
Reviewed-by: Hannes Reinecke 
---
 block/bsg-lib.c  | 17 ++---
 drivers/scsi/scsi_transport_fc.c |  4 ++--
 include/linux/bsg-lib.h  |  3 ++-
 3 files changed, 18 insertions(+), 6 deletions(-)

diff --git a/block/bsg-lib.c b/block/bsg-lib.c
index 6661f82..803ec40 100644
--- a/block/bsg-lib.c
+++ b/block/bsg-lib.c
@@ -32,7 +32,7 @@
  * bsg_destroy_job - routine to teardown/delete a bsg job
  * @job: bsg_job that is to be torn down
  */
-void bsg_destroy_job(struct kref *kref)
+static void bsg_destroy_job(struct kref *kref)
 {
struct bsg_job *job = container_of(kref, struct bsg_job, kref);
struct request *rq = job->req;
@@ -45,7 +45,18 @@ void bsg_destroy_job(struct kref *kref)
kfree(job->reply_payload.sg_list);
kfree(job);
 }
-EXPORT_SYMBOL_GPL(bsg_destroy_job);
+
+void bsg_job_put(struct bsg_job *job)
+{
+   kref_put(>kref, bsg_destroy_job);
+}
+EXPORT_SYMBOL_GPL(bsg_job_put);
+
+int bsg_job_get(struct bsg_job *job)
+{
+   return kref_get_unless_zero(>kref);
+}
+EXPORT_SYMBOL_GPL(bsg_job_get);
 
 /**
  * bsg_job_done - completion routine for bsg requests
@@ -89,7 +100,7 @@ void bsg_softirq_done(struct request *rq)
 {
struct bsg_job *job = rq->special;
 
-   kref_put(>kref, bsg_destroy_job);
+   bsg_job_put(job);
 }
 EXPORT_SYMBOL_GPL(bsg_softirq_done);
 
diff --git a/drivers/scsi/scsi_transport_fc.c b/drivers/scsi/scsi_transport_fc.c
index 720ddc9..7fae045 100644
--- a/drivers/scsi/scsi_transport_fc.c
+++ b/drivers/scsi/scsi_transport_fc.c
@@ -3571,13 +3571,13 @@ struct fc_vport *
if (rport && rport->port_state == FC_PORTSTATE_BLOCKED)
return BLK_EH_RESET_TIMER;
 
-   inflight = kref_get_unless_zero(>kref);
+   inflight = bsg_job_get(job);
 
if (inflight && i->f->bsg_timeout) {
/* call LLDD to abort the i/o as it has timed out */
err = i->f->bsg_timeout(job);
if (err == -EAGAIN) {
-   kref_put(>kref, bsg_destroy_job);
+   bsg_job_put(job);
return BLK_EH_RESET_TIMER;
} else if (err)
printk(KERN_ERR "ERROR: FC BSG request timeout - LLD "
diff --git a/include/linux/bsg-lib.h b/include/linux/bsg-lib.h
index 09f3044..b708db9 100644
--- a/include/linux/bsg-lib.h
+++ b/include/linux/bsg-lib.h
@@ -69,7 +69,8 @@ void bsg_job_done(struct bsg_job *job, int result,
 int bsg_setup_queue(struct device *dev, struct request_queue *q, char *name,
bsg_job_fn *job_fn, int dd_job_size);
 void bsg_request_fn(struct request_queue *q);
-void bsg_destroy_job(struct kref *kref);
 void bsg_softirq_done(struct request *rq);
+void bsg_job_put(struct bsg_job *job);
+int __must_check bsg_job_get(struct bsg_job *job);
 
 #endif
-- 
1.8.5.6

--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH v3 11/16] scsi: fc: Use bsg_destroy_job

[Johannes Thumshirn]

fc_destroy_bsgjob() and bsg_destroy_job() are now 1:1 copies, so use the
later. As bsg_destroy_job() comes from bsg-lib we need to select it in Kconfig
once CONFOG_SCSI_FC_ATTRS is active.

Signed-off-by: Johannes Thumshirn 
Reviewed-by: Hannes Reinecke 
---
 block/bsg-lib.c  |  7 +--
 drivers/scsi/Kconfig |  1 +
 drivers/scsi/scsi_transport_fc.c | 23 ++-
 include/linux/bsg-lib.h  |  1 +
 4 files changed, 9 insertions(+), 23 deletions(-)

diff --git a/block/bsg-lib.c b/block/bsg-lib.c
index 632fb40..9f1e8fd 100644
--- a/block/bsg-lib.c
+++ b/block/bsg-lib.c
@@ -32,9 +32,12 @@
  * bsg_destroy_job - routine to teardown/delete a bsg job
  * @job: bsg_job that is to be torn down
  */
-static void bsg_destroy_job(struct kref *kref)
+void bsg_destroy_job(struct kref *kref)
 {
struct bsg_job *job = container_of(kref, struct bsg_job, kref);
+   struct request *rq = job->req;
+
+   blk_end_request_all(rq, rq->errors);
 
put_device(job->dev);   /* release reference for the request */
 
@@ -42,6 +45,7 @@ static void bsg_destroy_job(struct kref *kref)
kfree(job->reply_payload.sg_list);
kfree(job);
 }
+EXPORT_SYMBOL_GPL(bsg_destroy_job);
 
 /**
  * bsg_job_done - completion routine for bsg requests
@@ -85,7 +89,6 @@ static void bsg_softirq_done(struct request *rq)
 {
struct bsg_job *job = rq->special;
 
-   blk_end_request_all(rq, rq->errors);
kref_put(>kref, bsg_destroy_job);
 }
 
diff --git a/drivers/scsi/Kconfig b/drivers/scsi/Kconfig
index 3e2bdb9..2774756 100644
--- a/drivers/scsi/Kconfig
+++ b/drivers/scsi/Kconfig
@@ -263,6 +263,7 @@ config SCSI_SPI_ATTRS
 config SCSI_FC_ATTRS
tristate "FiberChannel Transport Attributes"
depends on SCSI && NET
+   select BLK_DEV_BSGLIB
select SCSI_NETLINK
help
  If you wish to export transport-specific information about
diff --git a/drivers/scsi/scsi_transport_fc.c b/drivers/scsi/scsi_transport_fc.c
index eadc80f..d2d0a27 100644
--- a/drivers/scsi/scsi_transport_fc.c
+++ b/drivers/scsi/scsi_transport_fc.c
@@ -30,6 +30,7 @@
 #include 
 #include 
 #include 
+#include 
 #include 
 #include 
 #include 
@@ -3554,26 +3555,6 @@ struct fc_vport *
  * BSG support
  */
 
-
-/**
- * fc_destroy_bsgjob - routine to teardown/delete a fc bsg job
- * @job:   fc_bsg_job that is to be torn down
- */
-static void
-fc_destroy_bsgjob(struct kref *kref)
-{
-   struct bsg_job *job = container_of(kref, struct bsg_job, kref);
-   struct request *rq = job->req;
-
-   blk_end_request_all(rq, rq->errors);
-
-   put_device(job->dev);   /* release reference for the request */
-
-   kfree(job->request_payload.sg_list);
-   kfree(job->reply_payload.sg_list);
-   kfree(job);
-}
-
 /**
  * fc_bsg_jobdone - completion routine for bsg requests that the LLD has
  *  completed
@@ -3642,7 +3623,7 @@ static void fc_bsg_softirq_done(struct request *rq)
/* call LLDD to abort the i/o as it has timed out */
err = i->f->bsg_timeout(job);
if (err == -EAGAIN) {
-   kref_put(>kref, fc_destroy_bsgjob);
+   kref_put(>kref, bsg_destroy_job);
return BLK_EH_RESET_TIMER;
} else if (err)
printk(KERN_ERR "ERROR: FC BSG request timeout - LLD "
diff --git a/include/linux/bsg-lib.h b/include/linux/bsg-lib.h
index 58e0717..67f7de6 100644
--- a/include/linux/bsg-lib.h
+++ b/include/linux/bsg-lib.h
@@ -69,5 +69,6 @@ void bsg_job_done(struct bsg_job *job, int result,
 int bsg_setup_queue(struct device *dev, struct request_queue *q, char *name,
bsg_job_fn *job_fn, int dd_job_size);
 void bsg_request_fn(struct request_queue *q);
+void bsg_destroy_job(struct kref *kref);
 
 #endif
-- 
1.8.5.6

--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH v3 16/16] block: unexport bsg_softirq_done() again

[Johannes Thumshirn]

Unexport bsg_softirq_done() again, we don't need it outside of bsg-lib.c
anymore now that scsi_transport_fc is a pure bsg-lib client.

Signed-off-by: Johannes Thumshirn 
Reviewed-by: Hannes Reinecke 
---
 block/bsg-lib.c | 3 +--
 include/linux/bsg-lib.h | 1 -
 2 files changed, 1 insertion(+), 3 deletions(-)

diff --git a/block/bsg-lib.c b/block/bsg-lib.c
index 803ec40..2d1df5c 100644
--- a/block/bsg-lib.c
+++ b/block/bsg-lib.c
@@ -96,13 +96,12 @@ void bsg_job_done(struct bsg_job *job, int result,
  * bsg_softirq_done - softirq done routine for destroying the bsg requests
  * @rq: BSG request that holds the job to be destroyed
  */
-void bsg_softirq_done(struct request *rq)
+static void bsg_softirq_done(struct request *rq)
 {
struct bsg_job *job = rq->special;
 
bsg_job_put(job);
 }
-EXPORT_SYMBOL_GPL(bsg_softirq_done);
 
 static int bsg_map_buffer(struct bsg_buffer *buf, struct request *req)
 {
diff --git a/include/linux/bsg-lib.h b/include/linux/bsg-lib.h
index b708db9..657a718 100644
--- a/include/linux/bsg-lib.h
+++ b/include/linux/bsg-lib.h
@@ -69,7 +69,6 @@ void bsg_job_done(struct bsg_job *job, int result,
 int bsg_setup_queue(struct device *dev, struct request_queue *q, char *name,
bsg_job_fn *job_fn, int dd_job_size);
 void bsg_request_fn(struct request_queue *q);
-void bsg_softirq_done(struct request *rq);
 void bsg_job_put(struct bsg_job *job);
 int __must_check bsg_job_get(struct bsg_job *job);
 
-- 
1.8.5.6

--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH v3 15/16] scsi: fc: move FC transport's bsg code to bsg-lib

[Johannes Thumshirn]

Now that all conversions are done, move the FibreChannel bsg code over to the
bsg library.

This patch is derived from work done by Mike Christie in 2011 [1] but only the
iscsi parts got merged back then.

[1] http://marc.info/?l=linux-scsi=131149780921009=2

Signed-off-by: Johannes Thumshirn 
Reviewed-by: Hannes Reinecke 
---
 drivers/scsi/scsi_transport_fc.c | 287 ++-
 1 file changed, 43 insertions(+), 244 deletions(-)

diff --git a/drivers/scsi/scsi_transport_fc.c b/drivers/scsi/scsi_transport_fc.c
index 7fae045..916d488 100644
--- a/drivers/scsi/scsi_transport_fc.c
+++ b/drivers/scsi/scsi_transport_fc.c
@@ -3591,111 +3591,12 @@ struct fc_vport *
return BLK_EH_HANDLED;
 }
 
-static int
-fc_bsg_map_buffer(struct bsg_buffer *buf, struct request *req)
-{
-   size_t sz = (sizeof(struct scatterlist) * req->nr_phys_segments);
-
-   BUG_ON(!req->nr_phys_segments);
-
-   buf->sg_list = kzalloc(sz, GFP_KERNEL);
-   if (!buf->sg_list)
-   return -ENOMEM;
-   sg_init_table(buf->sg_list, req->nr_phys_segments);
-   buf->sg_cnt = blk_rq_map_sg(req->q, req, buf->sg_list);
-   buf->payload_len = blk_rq_bytes(req);
-   return 0;
-}
-
-
-/**
- * fc_req_to_bsgjob - Allocate/create the fc_bsg_job structure for the
- *   bsg request
- * @shost: SCSI Host corresponding to the bsg object
- * @rport: (optional) FC Remote Port corresponding to the bsg object
- * @req:   BSG request that needs a job structure
- */
-static int
-fc_req_to_bsgjob(struct Scsi_Host *shost, struct fc_rport *rport,
-   struct request *req)
-{
-   struct fc_internal *i = to_fc_internal(shost->transportt);
-   struct request *rsp = req->next_rq;
-   struct bsg_job *job;
-   struct fc_bsg_request *bsg_request;
-   struct fc_bsg_reply *bsg_reply;
-   int ret;
-
-   BUG_ON(req->special);
-
-   job = kzalloc(sizeof(struct bsg_job) + i->f->dd_bsg_size,
-   GFP_KERNEL);
-   if (!job)
-   return -ENOMEM;
-
-   /*
-* Note: this is a bit silly.
-* The request gets formatted as a SGIO v4 ioctl request, which
-* then gets reformatted as a blk request, which then gets
-* reformatted as a fc bsg request. And on completion, we have
-* to wrap return results such that SGIO v4 thinks it was a scsi
-* status.  I hope this was all worth it.
-*/
-
-   req->special = job;
-   job->req = req;
-   if (i->f->dd_bsg_size)
-   job->dd_data = (void *)[1];
-   bsg_request = (struct fc_bsg_request *)req->cmd;
-   job->request_len = req->cmd_len;
-   bsg_reply = req->sense;
-   job->reply_len = SCSI_SENSE_BUFFERSIZE; /* Size of sense buffer
-* allocated */
-   if (req->bio) {
-   ret = fc_bsg_map_buffer(>request_payload, req);
-   if (ret)
-   goto failjob_rls_job;
-   }
-   if (rsp && rsp->bio) {
-   ret = fc_bsg_map_buffer(>reply_payload, rsp);
-   if (ret)
-   goto failjob_rls_rqst_payload;
-   }
-   if (rport)
-   job->dev = >dev;
-   else
-   job->dev = >shost_gendev;
-   get_device(job->dev);   /* take a reference for the request */
-
-   kref_init(>kref);
-
-   return 0;
-
-
-failjob_rls_rqst_payload:
-   kfree(job->request_payload.sg_list);
-failjob_rls_job:
-   kfree(job);
-   return -ENOMEM;
-}
-
-
-enum fc_dispatch_result {
-   FC_DISPATCH_BREAK,  /* on return, q is locked, break from q loop */
-   FC_DISPATCH_LOCKED, /* on return, q is locked, continue on */
-   FC_DISPATCH_UNLOCKED,   /* on return, q is unlocked, continue on */
-};
-
-
 /**
  * fc_bsg_host_dispatch - process fc host bsg requests and dispatch to LLDD
- * @q: fc host request queue
  * @shost: scsi host rport attached to
  * @job:   bsg job to be processed
  */
-static enum fc_dispatch_result
-fc_bsg_host_dispatch(struct request_queue *q, struct Scsi_Host *shost,
-struct bsg_job *job)
+static int fc_bsg_host_dispatch(struct Scsi_Host *shost, struct bsg_job *job)
 {
struct fc_internal *i = to_fc_internal(shost->transportt);
struct fc_bsg_request *bsg_request = job->request;
@@ -3756,7 +3657,7 @@ enum fc_dispatch_result {
 
ret = i->f->bsg_request(job);
if (!ret)
-   return FC_DISPATCH_UNLOCKED;
+   return 0;
 
 fail_host_msg:
/* return the errno failure code as the only status */
@@ -3766,7 +3667,7 @@ enum fc_dispatch_result {
job->reply_len = sizeof(uint32_t);
bsg_job_done(job, bsg_reply->result,
   bsg_reply->reply_payload_rcv_len);
-   return FC_DISPATCH_UNLOCKED;
+   return 0;
 }
 
 
@@ -3790,14 

[PATCH v3 07/16] scsi: libfc: don't set FC_RQST_STATE_DONE before calling fc_bsg_jobdone()

[Johannes Thumshirn]

Don't set FC_RQST_STATE_DONE before calling fc_bsg_jobdone() as
fc_bsg_jobdone() calls blk_complete_requeust() which raises a soft-IRQ that
ends up in fc_bsg_sofirq_done() and fc_bsg_softirq_done() sets the
FC_RQST_STATE_DONE flag.

Signed-off-by: Johannes Thumshirn 
Reviewed-by: Hannes Reinecke 
---
 drivers/scsi/libfc/fc_lport.c | 2 --
 1 file changed, 2 deletions(-)

diff --git a/drivers/scsi/libfc/fc_lport.c b/drivers/scsi/libfc/fc_lport.c
index 9ba340c..34570cf 100644
--- a/drivers/scsi/libfc/fc_lport.c
+++ b/drivers/scsi/libfc/fc_lport.c
@@ -1916,7 +1916,6 @@ static void fc_lport_bsg_resp(struct fc_seq *sp, struct 
fc_frame *fp,
bsg_reply->result = (PTR_ERR(fp) == -FC_EX_CLOSED) ?
-ECONNABORTED : -ETIMEDOUT;
job->reply_len = sizeof(uint32_t);
-   job->state_flags |= FC_RQST_STATE_DONE;
fc_bsg_jobdone(job, bsg_reply->result,
   bsg_reply->reply_payload_rcv_len);
kfree(info);
@@ -1952,7 +1951,6 @@ static void fc_lport_bsg_resp(struct fc_seq *sp, struct 
fc_frame *fp,
bsg_reply->reply_payload_rcv_len =
job->reply_payload.payload_len;
bsg_reply->result = 0;
-   job->state_flags |= FC_RQST_STATE_DONE;
fc_bsg_jobdone(job, bsg_reply->result,
   bsg_reply->reply_payload_rcv_len);
kfree(info);
-- 
1.8.5.6

--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH v2 08/16] scsi: fc: implement kref backed reference counting

[Johannes Thumshirn]

On Thu, Oct 13, 2016 at 01:42:06PM +0200, Hannes Reinecke wrote:
> On 10/12/2016 03:06 PM, Johannes Thumshirn wrote:
> > Implement kref backed reference counting instead of rolling our own. This
> > elimnates the need of the following fields in 'struct fc_bsg_job':
> > * ref_cnt
> > * state_flags
> > * job_lock
> > bringing us close to unification of 'struct fc_bsg_job' and 'struct 
> > bsg_job'.
> > 
> > Signed-off-by: Johannes Thumshirn 
> > ---
> >  drivers/scsi/scsi_transport_fc.c | 38 
> > +-
> >  include/scsi/scsi_transport_fc.h |  4 +---
> >  2 files changed, 10 insertions(+), 32 deletions(-)
> > 
> > diff --git a/drivers/scsi/scsi_transport_fc.c 
> > b/drivers/scsi/scsi_transport_fc.c
> > index 96b3a2e..b0e28af 100644
> > --- a/drivers/scsi/scsi_transport_fc.c
> > +++ b/drivers/scsi/scsi_transport_fc.c
> > @@ -3560,16 +3560,9 @@ fc_vport_sched_delete(struct work_struct *work)
> >   * @job:   fc_bsg_job that is to be torn down
> >   */
> >  static void
> > -fc_destroy_bsgjob(struct fc_bsg_job *job)
> > +fc_destroy_bsgjob(struct kref *kref)
> >  {
> > -   unsigned long flags;
> > -
> > -   spin_lock_irqsave(>job_lock, flags);
> > -   if (job->ref_cnt) {
> > -   spin_unlock_irqrestore(>job_lock, flags);
> > -   return;
> > -   }
> > -   spin_unlock_irqrestore(>job_lock, flags);
> > +   struct fc_bsg_job *job = container_of(kref, struct fc_bsg_job, kref);
> >  
> > put_device(job->dev);   /* release reference for the request */
> >  
> > @@ -3620,15 +3613,9 @@ EXPORT_SYMBOL_GPL(fc_bsg_jobdone);
> >  static void fc_bsg_softirq_done(struct request *rq)
> >  {
> > struct fc_bsg_job *job = rq->special;
> > -   unsigned long flags;
> > -
> > -   spin_lock_irqsave(>job_lock, flags);
> > -   job->state_flags |= FC_RQST_STATE_DONE;
> > -   job->ref_cnt--;
> > -   spin_unlock_irqrestore(>job_lock, flags);
> >  
> > blk_end_request_all(rq, rq->errors);
> > -   fc_destroy_bsgjob(job);
> > +   kref_put(>kref, fc_destroy_bsgjob);
> >  }
> >  
> >  /**
> Hmm. blk_end_request_all() (potentially) triggers a recursion into all
> .end_io callbacks, which might end up doing god-knows-what.
> With some delays in doing so
> During that time we have no idea that bsg_softirq_done() is actually
> running, and we might clash with eg. timeouts or somesuch.
> 
> Maybe it's an idea to move blk_end_request_all into the kref destroy
> callback; that way we're guaranteed to call it only once and would avoid
> this situation.

This _could_ explain the panic with zfcp. Fixed that for v3.

-- 
Johannes Thumshirn  Storage
jthumsh...@suse.de+49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: Felix Imendörffer, Jane Smithard, Graham Norton
HRB 21284 (AG Nürnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH 12/24] libfc: Clarify ramp-down messages

[Johannes Thumshirn]

On Thu, Oct 13, 2016 at 03:10:48PM +0200, Hannes Reinecke wrote:
> When the queue depth is reduced we should print out the reason
> for this; it might be due to a queue full condition.
> 
> Signed-off-by: Hannes Reinecke 
> ---

Acked-by: Johannes Thumshirn 
-- 
Johannes Thumshirn  Storage
jthumsh...@suse.de+49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: Felix Imendörffer, Jane Smithard, Graham Norton
HRB 21284 (AG Nürnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH 10/24] libfc: Rework PRLI handling

[Johannes Thumshirn]

On Thu, Oct 13, 2016 at 03:10:46PM +0200, Hannes Reinecke wrote:
> PRLI is only required if the port is acting as an initiator; ports
> which support target functionality only do not need to send PRLI.
> At the same time the PRLI state is only used if the port initiated
> a PRLI transfer; if we received a PRLI request we should _not_
> change the state as this would cause our PRLI response to be dropped.
> And when we receive a PRLI response we need to check if an image
> pair has been established; if not the remote port cannot act as a
> target for us and we need to disable target functionality.
> 
> Signed-off-by: Hannes Reinecke 
> ---

Acked-by: Johannes Thumshirn 
-- 
Johannes Thumshirn  Storage
jthumsh...@suse.de+49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: Felix Imendörffer, Jane Smithard, Graham Norton
HRB 21284 (AG Nürnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: Patch: [Open-FCoE] Revert commit 3e22760d4db6fd89e0be46c3d132390a251da9c6 due to performance issues

[Johannes Thumshirn]

On Thu, Oct 13, 2016 at 08:43:30AM -0400, Laurence Oberman wrote:
> 
> 
> Hello
> 
> This patch reverts commit 3e22760d4db6fd89e0be46c3d132390a251da9c6.
> 
> This revert came about because of efforts by Ewan Milne, Curtis Taylor and I.
> In researching this issue, significant performance issues were seen on large 
> CPU count 
> systems using the software FCOE stack.
> Hannes also weighed in.
> 
> The same was not apparent on much smaller low count CPU systems.
> The behavior introduced by commit 3e22760d4db6fd89e0be46c3d132390a251da9c6 
> lands sup with large
> count CPU systems seeing continual blk_requeue_request() calls due to 
> ML_QUEUE_HOST_BUSY.
> 
> From Ewan:
> 
> fc_exch_alloc() used to try all the available exchange managers in the
> list for an available exchange id, but this was changed in 2010 so that
> if the first matched exchange manager couldn't allocate one, it fails
> and we end up returning host busy.  This was due to commit:
> 
> Setting the ddp_min module parameter to fcoe to 128MB prevents the ->match
> function from permitting the use of the offload exchange manager for the 
> frame,
> and we no longer see the problem with host busy status, since it uses the
> larger non-offloaded pool.
> 
> Reverting commit 3e22760d4db6fd89e0be46c3d132390a251da9c6 was tested to also
> prevent the host busy issue due to failing allocations.
> 
> Suggested-by: Ewan Milne 
> Suggested-by: Curtis Taylor 
> Tested-by: Laurence Oberman 
> Signed-off-by: Laurence Oberman 
> 

Acked-by: Johannes Thumshirn 

-- 
Johannes Thumshirn  Storage
jthumsh...@suse.de+49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: Felix Imendörffer, Jane Smithard, Graham Norton
HRB 21284 (AG Nürnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH 05/24] libfc: sanitize E_D_TOV and R_A_TOV setting

[Johannes Thumshirn]

On Thu, Oct 13, 2016 at 03:10:41PM +0200, Hannes Reinecke wrote:
> When setting the FCP timeout we need to ensure a lower boundary
> for E_D_TOV and R_A_TOV, otherwise we'd be getting spurious I/O
> issues due to the fcp timer firing too early.
> 
> Signed-off-by: Hannes Reinecke 
> ---

Acked-by: Johannes Thumshirn 
-- 
Johannes Thumshirn  Storage
jthumsh...@suse.de+49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: Felix Imendörffer, Jane Smithard, Graham Norton
HRB 21284 (AG Nürnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH 21/24] fcoe: catch invalid values for the 'enabled' attribute

[Johannes Thumshirn]

On Thu, Oct 13, 2016 at 03:10:57PM +0200, Hannes Reinecke wrote:
> The 'enabled' sysfs attribute only accepts the values '0' and '1',
> so we should error out any other values.
> 
> Signed-off-by: Hannes Reinecke 
> ---

Acked-by: Johannes Thumshirn 

-- 
Johannes Thumshirn  Storage
jthumsh...@suse.de+49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: Felix Imendörffer, Jane Smithard, Graham Norton
HRB 21284 (AG Nürnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850
--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH 15/24] libfc: don't fail sequence abort for completed exchanges

[Hannes Reinecke]

If a sequence should be aborted the exchange might already
be completed (eg if the response is still queued in the rx
queue), so this shouldn't considered as an error.

Signed-off-by: Hannes Reinecke 
Acked-by: Johannes Thumshirn 
---
 drivers/scsi/libfc/fc_fcp.c | 40 ++--
 1 file changed, 30 insertions(+), 10 deletions(-)

diff --git a/drivers/scsi/libfc/fc_fcp.c b/drivers/scsi/libfc/fc_fcp.c
index 780d9f0..d43c925 100644
--- a/drivers/scsi/libfc/fc_fcp.c
+++ b/drivers/scsi/libfc/fc_fcp.c
@@ -258,6 +258,17 @@ static void fc_fcp_timer_set(struct fc_fcp_pkt *fsp, 
unsigned long delay)
mod_timer(>timer, jiffies + delay);
 }
 
+static void fc_fcp_abort_done(struct fc_fcp_pkt *fsp)
+{
+   fsp->state |= FC_SRB_ABORTED;
+   fsp->state &= ~FC_SRB_ABORT_PENDING;
+
+   if (fsp->wait_for_comp)
+   complete(>tm_done);
+   else
+   fc_fcp_complete_locked(fsp);
+}
+
 /**
  * fc_fcp_send_abort() - Send an abort for exchanges associated with a
  *  fcp_pkt
@@ -265,6 +276,8 @@ static void fc_fcp_timer_set(struct fc_fcp_pkt *fsp, 
unsigned long delay)
  */
 static int fc_fcp_send_abort(struct fc_fcp_pkt *fsp)
 {
+   int rc;
+
if (!fsp->seq_ptr)
return -EINVAL;
 
@@ -272,7 +285,16 @@ static int fc_fcp_send_abort(struct fc_fcp_pkt *fsp)
put_cpu();
 
fsp->state |= FC_SRB_ABORT_PENDING;
-   return fsp->lp->tt.seq_exch_abort(fsp->seq_ptr, 0);
+   rc = fsp->lp->tt.seq_exch_abort(fsp->seq_ptr, 0);
+   /*
+* ->seq_exch_abort() might return -ENXIO if
+* the sequence is already completed
+*/
+   if (rc == -ENXIO) {
+   fc_fcp_abort_done(fsp);
+   rc = 0;
+   }
+   return rc;
 }
 
 /**
@@ -729,15 +751,8 @@ static void fc_fcp_abts_resp(struct fc_fcp_pkt *fsp, 
struct fc_frame *fp)
ba_done = 0;
}
 
-   if (ba_done) {
-   fsp->state |= FC_SRB_ABORTED;
-   fsp->state &= ~FC_SRB_ABORT_PENDING;
-
-   if (fsp->wait_for_comp)
-   complete(>tm_done);
-   else
-   fc_fcp_complete_locked(fsp);
-   }
+   if (ba_done)
+   fc_fcp_abort_done(fsp);
 }
 
 /**
@@ -1245,6 +1260,11 @@ static int fc_fcp_pkt_abort(struct fc_fcp_pkt *fsp)
return FAILED;
}
 
+   if (fsp->state & FC_SRB_ABORTED) {
+   FC_FCP_DBG(fsp, "target abort cmd  completed\n");
+   return SUCCESS;
+   }
+
init_completion(>tm_done);
fsp->wait_for_comp = 1;
 
-- 
1.8.5.6

--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH 24/24] fcoe: filter out frames from invalid vlans

[Hannes Reinecke]

Any multicase address is set on all interfaces, the base interface
and any VLAN interfaces on top of this. So we might receive frames
which are not destined for us.

Signed-off-by: Hannes Reinecke 
---
 drivers/scsi/fcoe/fcoe_ctlr.c | 12 +++-
 1 file changed, 11 insertions(+), 1 deletion(-)

diff --git a/drivers/scsi/fcoe/fcoe_ctlr.c b/drivers/scsi/fcoe/fcoe_ctlr.c
index d773b46..4aacd60 100644
--- a/drivers/scsi/fcoe/fcoe_ctlr.c
+++ b/drivers/scsi/fcoe/fcoe_ctlr.c
@@ -2754,11 +2754,21 @@ static int fcoe_ctlr_vn_recv(struct fcoe_ctlr *fip, 
struct sk_buff *skb)
struct fc_rport_priv rdata;
struct fcoe_rport frport;
} buf;
-   int rc;
+   int rc, vlan_id = 0;
 
fiph = (struct fip_header *)skb->data;
sub = fiph->fip_subcode;
 
+   if (fip->lp->vlan)
+   vlan_id = skb_vlan_tag_get_id(skb);
+
+   if (vlan_id && vlan_id != fip->lp->vlan) {
+   LIBFCOE_FIP_DBG(fip, "vn_recv drop frame sub %x vlan %d\n",
+   sub, vlan_id);
+   rc = -EAGAIN;
+   goto drop;
+   }
+
rc = fcoe_ctlr_vn_parse(fip, skb, );
if (rc) {
LIBFCOE_FIP_DBG(fip, "vn_recv vn_parse error %d\n", rc);
-- 
1.8.5.6

--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH 23/24] fcoe: correct sending FIP VLAN packets on VLAN 0

[Hannes Reinecke]

The FIP VLAN frame consists of an ethernet header followed
by the FIP VLAN frame, so we need to skip the ethernet header
if we want to check the FIP opcode.

Signed-off-by: Hannes Reinecke 
---
 drivers/scsi/fcoe/fcoe.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/scsi/fcoe/fcoe.c b/drivers/scsi/fcoe/fcoe.c
index 9876fca..cf4adaa 100644
--- a/drivers/scsi/fcoe/fcoe.c
+++ b/drivers/scsi/fcoe/fcoe.c
@@ -590,7 +590,8 @@ static void fcoe_fip_send(struct fcoe_ctlr *fip, struct 
sk_buff *skb)
 * Use default VLAN for FIP VLAN discovery protocol
 */
frame = (struct fip_frame *)skb->data;
-   if (frame->fip.fip_op == ntohs(FIP_OP_VLAN) &&
+   if (ntohs(frame->eth.h_proto) == ETH_P_FIP &&
+   ntohs(frame->fip.fip_op) == FIP_OP_VLAN &&
fcoe->realdev != fcoe->netdev)
skb->dev = fcoe->realdev;
else
-- 
1.8.5.6

--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH 20/24] fcoe: set default TC priority

[Hannes Reinecke]

If DCB is not enabled or compiled in we still should be setting
a sane default priority. So put FCoE frames in priority class
'interactive' and FIP frames in priority class 'besteffort'.

Signed-off-by: Hannes Reinecke 
---
 drivers/scsi/fcoe/fcoe.c | 8 ++--
 1 file changed, 6 insertions(+), 2 deletions(-)

diff --git a/drivers/scsi/fcoe/fcoe.c b/drivers/scsi/fcoe/fcoe.c
index c907661..9876fca 100644
--- a/drivers/scsi/fcoe/fcoe.c
+++ b/drivers/scsi/fcoe/fcoe.c
@@ -2168,6 +2168,8 @@ static bool fcoe_match(struct net_device *netdev)
  */
 static void fcoe_dcb_create(struct fcoe_interface *fcoe)
 {
+   int ctlr_prio = TC_PRIO_BESTEFFORT;
+   int fcoe_prio = TC_PRIO_INTERACTIVE;
 #ifdef CONFIG_DCB
int dcbx;
u8 fup, up;
@@ -2194,10 +2196,12 @@ static void fcoe_dcb_create(struct fcoe_interface *fcoe)
fup = dcb_getapp(netdev, );
}
 
-   fcoe->priority = ffs(up) ? ffs(up) - 1 : 0;
-   ctlr->priority = ffs(fup) ? ffs(fup) - 1 : fcoe->priority;
+   fcoe_prio = ffs(up) ? ffs(up) - 1 : 0;
+   ctlr_prio = ffs(fup) ? ffs(fup) - 1 : fcoe_prio;
}
 #endif
+   fcoe->priority = fcoe_prio;
+   ctlr->priority = ctlr_prio;
 }
 
 enum fcoe_create_link_state {
-- 
1.8.5.6

--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH 19/24] libfc: Check xid when looking up REC exchanges

[Hannes Reinecke]

We currently can only lookup the local xid, so we need
to reject REC with empty rxid.

Signed-off-by: Hannes Reinecke 
---
 drivers/scsi/libfc/fc_exch.c | 16 
 1 file changed, 12 insertions(+), 4 deletions(-)

diff --git a/drivers/scsi/libfc/fc_exch.c b/drivers/scsi/libfc/fc_exch.c
index c62fc27..4c69a99 100644
--- a/drivers/scsi/libfc/fc_exch.c
+++ b/drivers/scsi/libfc/fc_exch.c
@@ -2000,8 +2000,7 @@ static void fc_exch_els_rec(struct fc_frame *rfp)
enum fc_els_rjt_reason reason = ELS_RJT_LOGIC;
enum fc_els_rjt_explan explan;
u32 sid;
-   u16 rxid;
-   u16 oxid;
+   u16 xid, rxid, oxid;
 
lport = fr_dev(rfp);
rp = fc_frame_payload_get(rfp, sizeof(*rp));
@@ -2012,9 +2011,18 @@ static void fc_exch_els_rec(struct fc_frame *rfp)
rxid = ntohs(rp->rec_rx_id);
oxid = ntohs(rp->rec_ox_id);
 
-   ep = fc_exch_lookup(lport,
-   sid == fc_host_port_id(lport->host) ? oxid : rxid);
explan = ELS_EXPL_OXID_RXID;
+   if (sid == fc_host_port_id(lport->host))
+   xid = oxid;
+   else
+   xid = rxid;
+   if (xid == FC_XID_UNKNOWN) {
+   FC_LPORT_DBG(lport,
+"REC request from %x: invalid rxid %x oxid %x\n",
+sid, rxid, oxid);
+   goto reject;
+   }
+   ep = fc_exch_lookup(lport, xid);
if (!ep) {
FC_LPORT_DBG(lport,
 "REC request from %x: rxid %x oxid %x not found\n",
-- 
1.8.5.6

--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH 22/24] fcoe: FIP debugging

[Hannes Reinecke]

Add additional statements for debugging FIP frames.

Signed-off-by: Hannes Reinecke 
---
 drivers/scsi/fcoe/fcoe_ctlr.c | 48 +++
 1 file changed, 44 insertions(+), 4 deletions(-)

diff --git a/drivers/scsi/fcoe/fcoe_ctlr.c b/drivers/scsi/fcoe/fcoe_ctlr.c
index 05573c3..d773b46 100644
--- a/drivers/scsi/fcoe/fcoe_ctlr.c
+++ b/drivers/scsi/fcoe/fcoe_ctlr.c
@@ -801,6 +801,8 @@ int fcoe_ctlr_els_send(struct fcoe_ctlr *fip, struct 
fc_lport *lport,
return -EINPROGRESS;
 drop:
kfree_skb(skb);
+   LIBFCOE_FIP_DBG(fip, "drop els_send op %u d_id %x\n",
+   op, ntoh24(fh->fh_d_id));
return -EINVAL;
 }
 EXPORT_SYMBOL(fcoe_ctlr_els_send);
@@ -2428,6 +2430,8 @@ static void fcoe_ctlr_vn_probe_req(struct fcoe_ctlr *fip,
switch (fip->state) {
case FIP_ST_VNMP_CLAIM:
case FIP_ST_VNMP_UP:
+   LIBFCOE_FIP_DBG(fip, "vn_probe_req: send reply, state %x\n",
+   fip->state);
fcoe_ctlr_vn_send(fip, FIP_SC_VN_PROBE_REP,
  frport->enode_mac, 0);
break;
@@ -2442,15 +2446,21 @@ static void fcoe_ctlr_vn_probe_req(struct fcoe_ctlr 
*fip,
 */
if (fip->lp->wwpn > rdata->ids.port_name &&
!(frport->flags & FIP_FL_REC_OR_P2P)) {
+   LIBFCOE_FIP_DBG(fip, "vn_probe_req: "
+   "port_id collision\n");
fcoe_ctlr_vn_send(fip, FIP_SC_VN_PROBE_REP,
  frport->enode_mac, 0);
break;
}
/* fall through */
case FIP_ST_VNMP_START:
+   LIBFCOE_FIP_DBG(fip, "vn_probe_req: "
+   "restart VN2VN negotiation\n");
fcoe_ctlr_vn_restart(fip);
break;
default:
+   LIBFCOE_FIP_DBG(fip, "vn_probe_req: ignore state %x\n",
+   fip->state);
break;
}
 }
@@ -2472,9 +2482,12 @@ static void fcoe_ctlr_vn_probe_reply(struct fcoe_ctlr 
*fip,
case FIP_ST_VNMP_PROBE1:
case FIP_ST_VNMP_PROBE2:
case FIP_ST_VNMP_CLAIM:
+   LIBFCOE_FIP_DBG(fip, "vn_probe_reply: restart state %x\n",
+   fip->state);
fcoe_ctlr_vn_restart(fip);
break;
case FIP_ST_VNMP_UP:
+   LIBFCOE_FIP_DBG(fip, "vn_probe_reply: send claim notify\n");
fcoe_ctlr_vn_send_claim(fip);
break;
default:
@@ -2517,6 +2530,7 @@ static void fcoe_ctlr_vn_add(struct fcoe_ctlr *fip, 
struct fc_rport_priv *new)
if ((ids->port_name != -1 && ids->port_name != new->ids.port_name) ||
(ids->node_name != -1 && ids->node_name != new->ids.node_name)) {
mutex_unlock(>rp_mutex);
+   LIBFCOE_FIP_DBG(fip, "vn_add rport logoff %6.6x\n", port_id);
lport->tt.rport_logoff(rdata);
mutex_lock(>rp_mutex);
}
@@ -2525,8 +2539,9 @@ static void fcoe_ctlr_vn_add(struct fcoe_ctlr *fip, 
struct fc_rport_priv *new)
mutex_unlock(>rp_mutex);
 
frport = fcoe_ctlr_rport(rdata);
-   LIBFCOE_FIP_DBG(fip, "vn_add rport %6.6x %s\n",
-   port_id, frport->fcoe_len ? "old" : "new");
+   LIBFCOE_FIP_DBG(fip, "vn_add rport %6.6x %s state %d\n",
+   port_id, frport->fcoe_len ? "old" : "new",
+   rdata->rp_state);
*frport = *fcoe_ctlr_rport(new);
frport->time = 0;
 }
@@ -2569,6 +2584,7 @@ static void fcoe_ctlr_vn_claim_notify(struct fcoe_ctlr 
*fip,
struct fcoe_rport *frport = fcoe_ctlr_rport(new);
 
if (frport->flags & FIP_FL_REC_OR_P2P) {
+   LIBFCOE_FIP_DBG(fip, "send probe req for P2P/REC\n");
fcoe_ctlr_vn_send(fip, FIP_SC_VN_PROBE_REQ, fcoe_all_vn2vn, 0);
return;
}
@@ -2576,25 +2592,37 @@ static void fcoe_ctlr_vn_claim_notify(struct fcoe_ctlr 
*fip,
case FIP_ST_VNMP_START:
case FIP_ST_VNMP_PROBE1:
case FIP_ST_VNMP_PROBE2:
-   if (new->ids.port_id == fip->port_id)
+   if (new->ids.port_id == fip->port_id) {
+   LIBFCOE_FIP_DBG(fip, "vn_claim_notify: "
+   "restart, state %d\n",
+   fip->state);
fcoe_ctlr_vn_restart(fip);
+   }
break;
case FIP_ST_VNMP_CLAIM:
case FIP_ST_VNMP_UP:
if (new->ids.port_id == fip->port_id) {
if (new->ids.port_name > fip->lp->wwpn) {
+   LIBFCOE_FIP_DBG(fip, "vn_claim_notify: "
+   "restart, port_id collision\n");

[PATCH 16/24] libfc: Do not drop out-of-order frames

[Hannes Reinecke]

When receiving packets from the network we cannot guarantee any
frame ordering, so we should be receiving all valid frames and
let the upper layers deal with it.

Signed-off-by: Hannes Reinecke 
Acked-by: Johannes Thumshirn 
---
 drivers/scsi/libfc/fc_exch.c | 3 ---
 1 file changed, 3 deletions(-)

diff --git a/drivers/scsi/libfc/fc_exch.c b/drivers/scsi/libfc/fc_exch.c
index 5868824..c62fc27 100644
--- a/drivers/scsi/libfc/fc_exch.c
+++ b/drivers/scsi/libfc/fc_exch.c
@@ -1592,9 +1592,6 @@ static void fc_exch_recv_seq_resp(struct fc_exch_mgr *mp, 
struct fc_frame *fp)
if (fc_sof_is_init(sof)) {
sp->ssb_stat |= SSB_ST_RESP;
sp->id = fh->fh_seq_id;
-   } else if (sp->id != fh->fh_seq_id) {
-   atomic_inc(>stats.seq_not_found);
-   goto rel;
}
 
f_ctl = ntoh24(fh->fh_f_ctl);
-- 
1.8.5.6

--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH 17/24] libfc: reset timeout on queue full

[Hannes Reinecke]

When we're receiving a timeout we should be checking for queue
full status; if there are still some packets pending we should
be resetting the counter to ensure we're not missing out any
packets which are still queued.

Signed-off-by: Hannes Reinecke 
Acked-by: Johannes Thumshirn 
---
 drivers/scsi/libfc/fc_fcp.c | 20 +---
 include/scsi/libfc.h|  3 ++-
 2 files changed, 19 insertions(+), 4 deletions(-)

diff --git a/drivers/scsi/libfc/fc_fcp.c b/drivers/scsi/libfc/fc_fcp.c
index d43c925..fb2ebc7 100644
--- a/drivers/scsi/libfc/fc_fcp.c
+++ b/drivers/scsi/libfc/fc_fcp.c
@@ -254,8 +254,10 @@ static inline void fc_fcp_unlock_pkt(struct fc_fcp_pkt 
*fsp)
  */
 static void fc_fcp_timer_set(struct fc_fcp_pkt *fsp, unsigned long delay)
 {
-   if (!(fsp->state & FC_SRB_COMPL))
+   if (!(fsp->state & FC_SRB_COMPL)) {
mod_timer(>timer, jiffies + delay);
+   fsp->timer_delay = delay;
+   }
 }
 
 static void fc_fcp_abort_done(struct fc_fcp_pkt *fsp)
@@ -932,6 +934,11 @@ static void fc_fcp_resp(struct fc_fcp_pkt *fsp, struct 
fc_frame *fp)
 * Wait a at least one jiffy to see if it is delivered.
 * If this expires without data, we may do SRR.
 */
+   if (fsp->lp->qfull) {
+   FC_FCP_DBG(fsp, "tgt %6.6x queue busy retry\n",
+  fsp->rport->port_id);
+   return;
+   }
FC_FCP_DBG(fsp, "tgt %6.6x xfer len %zx data underrun "
   "len %x, data len %x\n",
   fsp->rport->port_id,
@@ -1434,8 +1441,15 @@ static void fc_fcp_timeout(unsigned long data)
if (fsp->cdb_cmd.fc_tm_flags)
goto unlock;
 
-   FC_FCP_DBG(fsp, "fcp timeout, flags %x state %x\n",
-  rpriv->flags, fsp->state);
+   if (fsp->lp->qfull) {
+   FC_FCP_DBG(fsp, "fcp timeout, resetting timer delay %d\n",
+  fsp->timer_delay);
+   setup_timer(>timer, fc_fcp_timeout, (unsigned long)fsp);
+   fc_fcp_timer_set(fsp, fsp->timer_delay);
+   goto unlock;
+   }
+   FC_FCP_DBG(fsp, "fcp timeout, delay %d flags %x state %x\n",
+  fsp->timer_delay, rpriv->flags, fsp->state);
fsp->state |= FC_SRB_FCP_PROCESSING_TMO;
 
if (rpriv->flags & FC_RP_FLAGS_REC_SUPPORTED)
diff --git a/include/scsi/libfc.h b/include/scsi/libfc.h
index 8cb752f..f5aa54b 100644
--- a/include/scsi/libfc.h
+++ b/include/scsi/libfc.h
@@ -355,7 +355,8 @@ struct fc_fcp_pkt {
 
/* Timeout/error related information */
struct timer_list timer;
-   int   wait_for_comp;
+   int   wait_for_comp;
+   int   timer_delay;
u32   recov_retry;
struct fc_seq *recov_seq;
struct completion tm_done;
-- 
1.8.5.6

--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH 18/24] libfc: wait for E_D_TOV when out-of-order sequence is received

[Hannes Reinecke]

When detecting an out-of-order sequence we should be waiting for
E_D_TOV before trying to abort the sequence.
The response might still be stuck in the queue somewhere.

Signed-off-by: Hannes Reinecke 
---
 drivers/scsi/libfc/fc_fcp.c | 38 --
 1 file changed, 20 insertions(+), 18 deletions(-)

diff --git a/drivers/scsi/libfc/fc_fcp.c b/drivers/scsi/libfc/fc_fcp.c
index fb2ebc7..c033946 100644
--- a/drivers/scsi/libfc/fc_fcp.c
+++ b/drivers/scsi/libfc/fc_fcp.c
@@ -460,6 +460,22 @@ static inline struct fc_frame *fc_fcp_frame_alloc(struct 
fc_lport *lport,
 }
 
 /**
+ * get_fsp_rec_tov() - Helper function to get REC_TOV
+ * @fsp: the FCP packet
+ *
+ * Returns rec tov in jiffies as rpriv->e_d_tov + 1 second
+ */
+static inline unsigned int get_fsp_rec_tov(struct fc_fcp_pkt *fsp)
+{
+   struct fc_rport_libfc_priv *rpriv = fsp->rport->dd_data;
+   unsigned int e_d_tov = FC_DEF_E_D_TOV;
+
+   if (rpriv && rpriv->e_d_tov > e_d_tov)
+   e_d_tov = rpriv->e_d_tov;
+   return msecs_to_jiffies(e_d_tov) + HZ;
+}
+
+/**
  * fc_fcp_recv_data() - Handler for receiving SCSI-FCP data from a target
  * @fsp: The FCP packet the data is on
  * @fp: The data frame
@@ -562,8 +578,10 @@ static void fc_fcp_recv_data(struct fc_fcp_pkt *fsp, 
struct fc_frame *fp)
 * and completes the transfer, call the completion handler.
 */
if (unlikely(fsp->state & FC_SRB_RCV_STATUS) &&
-   fsp->xfer_len == fsp->data_len - fsp->scsi_resid)
+   fsp->xfer_len == fsp->data_len - fsp->scsi_resid) {
+   FC_FCP_DBG( fsp, "complete out-of-order sequence\n" );
fc_fcp_complete_locked(fsp);
+   }
return;
 err:
fc_fcp_recovery(fsp, host_bcode);
@@ -943,7 +961,7 @@ static void fc_fcp_resp(struct fc_fcp_pkt *fsp, struct 
fc_frame *fp)
   "len %x, data len %x\n",
   fsp->rport->port_id,
   fsp->xfer_len, expected_len, fsp->data_len);
-   fc_fcp_timer_set(fsp, 2);
+   fc_fcp_timer_set(fsp, get_fsp_rec_tov(fsp));
return;
}
fsp->status_code = FC_DATA_OVRRUN;
@@ -1152,22 +1170,6 @@ static int fc_fcp_pkt_send(struct fc_lport *lport, 
struct fc_fcp_pkt *fsp)
 }
 
 /**
- * get_fsp_rec_tov() - Helper function to get REC_TOV
- * @fsp: the FCP packet
- *
- * Returns rec tov in jiffies as rpriv->e_d_tov + 1 second
- */
-static inline unsigned int get_fsp_rec_tov(struct fc_fcp_pkt *fsp)
-{
-   struct fc_rport_libfc_priv *rpriv = fsp->rport->dd_data;
-   unsigned int e_d_tov = FC_DEF_E_D_TOV;
-
-   if (rpriv && rpriv->e_d_tov > e_d_tov)
-   e_d_tov = rpriv->e_d_tov;
-   return msecs_to_jiffies(e_d_tov) + HZ;
-}
-
-/**
  * fc_fcp_cmd_send() - Send a FCP command
  * @lport: The local port to send the command on
  * @fsp:   The FCP packet the command is on
-- 
1.8.5.6

--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH 21/24] fcoe: catch invalid values for the 'enabled' attribute

[Hannes Reinecke]

The 'enabled' sysfs attribute only accepts the values '0' and '1',
so we should error out any other values.

Signed-off-by: Hannes Reinecke 
---
 drivers/scsi/fcoe/fcoe_sysfs.c | 12 ++--
 1 file changed, 10 insertions(+), 2 deletions(-)

diff --git a/drivers/scsi/fcoe/fcoe_sysfs.c b/drivers/scsi/fcoe/fcoe_sysfs.c
index 9e6baac..9cf3d56 100644
--- a/drivers/scsi/fcoe/fcoe_sysfs.c
+++ b/drivers/scsi/fcoe/fcoe_sysfs.c
@@ -335,16 +335,24 @@ static ssize_t store_ctlr_enabled(struct device *dev,
  const char *buf, size_t count)
 {
struct fcoe_ctlr_device *ctlr = dev_to_ctlr(dev);
+   bool enabled;
int rc;
 
+   if (*buf == '1')
+   enabled = true;
+   else if (*buf == '0')
+   enabled = false;
+   else
+   return -EINVAL;
+
switch (ctlr->enabled) {
case FCOE_CTLR_ENABLED:
-   if (*buf == '1')
+   if (enabled)
return count;
ctlr->enabled = FCOE_CTLR_DISABLED;
break;
case FCOE_CTLR_DISABLED:
-   if (*buf == '0')
+   if (!enabled)
return count;
ctlr->enabled = FCOE_CTLR_ENABLED;
break;
-- 
1.8.5.6

--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH 14/24] libfc: quarantine timed out xids

[Hannes Reinecke]

When a sequence times out we have no idea what happened to the
frame. And we do not know if we will ever receive the frame.
Hence we cannot re-use the xid as we would risk data corruption
if the xid had been re-used and the timed out frame would be
received after that.
So we need to quarantine the xid until the lport is reset.
Yes, I know this will (eventually) deplete the xid pool.
But for now it's the safest method.

Signed-off-by: Hannes Reinecke 
---
 drivers/scsi/libfc/fc_exch.c | 33 ++---
 drivers/scsi/libfc/fc_fcp.c  | 13 +++--
 include/scsi/libfc.h |  1 +
 3 files changed, 30 insertions(+), 17 deletions(-)

diff --git a/drivers/scsi/libfc/fc_exch.c b/drivers/scsi/libfc/fc_exch.c
index e8784a7..5868824 100644
--- a/drivers/scsi/libfc/fc_exch.c
+++ b/drivers/scsi/libfc/fc_exch.c
@@ -94,6 +94,7 @@ struct fc_exch_pool {
 struct fc_exch_mgr {
struct fc_exch_pool __percpu *pool;
mempool_t   *ep_pool;
+   struct fc_lport *lport;
enum fc_class   class;
struct kref kref;
u16 min_xid;
@@ -408,6 +409,8 @@ static int fc_exch_done_locked(struct fc_exch *ep)
return rc;
 }
 
+static struct fc_exch fc_quarantine_exch;
+
 /**
  * fc_exch_ptr_get() - Return an exchange from an exchange pool
  * @pool:  Exchange Pool to get an exchange from
@@ -452,14 +455,17 @@ static void fc_exch_delete(struct fc_exch *ep)
 
/* update cache of free slot */
index = (ep->xid - ep->em->min_xid) >> fc_cpu_order;
-   if (pool->left == FC_XID_UNKNOWN)
-   pool->left = index;
-   else if (pool->right == FC_XID_UNKNOWN)
-   pool->right = index;
-   else
-   pool->next_index = index;
-
-   fc_exch_ptr_set(pool, index, NULL);
+   if (!(ep->state & FC_EX_QUARANTINE)) {
+   if (pool->left == FC_XID_UNKNOWN)
+   pool->left = index;
+   else if (pool->right == FC_XID_UNKNOWN)
+   pool->right = index;
+   else
+   pool->next_index = index;
+   fc_exch_ptr_set(pool, index, NULL);
+   } else {
+   fc_exch_ptr_set(pool, index, _quarantine_exch);
+   }
list_del(>ex_list);
spin_unlock_bh(>lock);
fc_exch_release(ep);/* drop hold for exch in mp */
@@ -916,14 +922,14 @@ static inline struct fc_exch *fc_exch_alloc(struct 
fc_lport *lport,
  */
 static struct fc_exch *fc_exch_find(struct fc_exch_mgr *mp, u16 xid)
 {
+   struct fc_lport *lport = mp->lport;
struct fc_exch_pool *pool;
struct fc_exch *ep = NULL;
u16 cpu = xid & fc_cpu_mask;
 
if (cpu >= nr_cpu_ids || !cpu_possible(cpu)) {
-   printk_ratelimited(KERN_ERR
-   "libfc: lookup request for XID = %d, "
-   "indicates invalid CPU %d\n", xid, cpu);
+   pr_err("host%u: lport %6.6x: xid %d invalid CPU %d\n:",
+  lport->host->host_no, lport->port_id, xid, cpu);
return NULL;
}
 
@@ -931,6 +937,10 @@ static struct fc_exch *fc_exch_find(struct fc_exch_mgr 
*mp, u16 xid)
pool = per_cpu_ptr(mp->pool, cpu);
spin_lock_bh(>lock);
ep = fc_exch_ptr_get(pool, (xid - mp->min_xid) >> fc_cpu_order);
+   if (ep == _quarantine_exch) {
+   FC_LPORT_DBG(lport, "xid %x quarantined\n", xid);
+   ep = NULL;
+   }
if (ep) {
WARN_ON(ep->xid != xid);
fc_exch_hold(ep);
@@ -2429,6 +2439,7 @@ struct fc_exch_mgr *fc_exch_mgr_alloc(struct fc_lport 
*lport,
return NULL;
 
mp->class = class;
+   mp->lport = lport;
/* adjust em exch xid range for offload */
mp->min_xid = min_xid;
 
diff --git a/drivers/scsi/libfc/fc_fcp.c b/drivers/scsi/libfc/fc_fcp.c
index f7700cc..780d9f0 100644
--- a/drivers/scsi/libfc/fc_fcp.c
+++ b/drivers/scsi/libfc/fc_fcp.c
@@ -1529,13 +1529,14 @@ static void fc_fcp_rec_resp(struct fc_seq *seq, struct 
fc_frame *fp, void *arg)
   fsp->rport->port_id, rjt->er_reason,
   rjt->er_explan);
/*
-* If no data transfer, the command frame got dropped
-* so we just retry.  If data was transferred, we
-* lost the response but the target has no record,
-* so we abort and retry.
+* If response got lost or is stuck in the
+* queue somewhere we have no idea if and when
+* the response will be received. So quarantine
+* the xid and retry the command.
 */
-   if (rjt->er_explan == ELS_EXPL_OXID_RXID &&

Re: [PATCH v2 05/16] scsi: fc: provide fc_bsg_to_shost() helper

[Hannes Reinecke]

On 10/12/2016 03:06 PM, Johannes Thumshirn wrote:
> Provide fc_bsg_to_shost() helper that will become handy when we're moving from
> struct fc_bsg_job to a plain struct bsg_job. Also use this little helper in
> the LLDDs.
> 
> Signed-off-by: Johannes Thumshirn 
> ---
>  drivers/s390/scsi/zfcp_fc.c  |  4 +--
>  drivers/scsi/bfa/bfad_bsg.c  |  6 ++---
>  drivers/scsi/ibmvscsi/ibmvfc.c   |  4 +--
>  drivers/scsi/libfc/fc_lport.c|  2 +-
>  drivers/scsi/lpfc/lpfc_bsg.c | 32 
>  drivers/scsi/qla2xxx/qla_bsg.c   | 54 
> 
>  drivers/scsi/scsi_transport_fc.c |  2 +-
>  include/scsi/scsi_transport_fc.h |  5 
>  8 files changed, 56 insertions(+), 53 deletions(-)
> 
Reviewed-by: Hannes Reinecke 

Cheers,

Hannes
-- 
Dr. Hannes ReineckeTeamlead Storage & Networking
h...@suse.de   +49 911 74053 688
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: F. Imendörffer, J. Smithard, J. Guild, D. Upmanyu, G. Norton
HRB 21284 (AG Nürnberg)
--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html