Re: [PATCH v2 11/16] qla2xxx: Add selective command queuing
On 12/29/15, 1:32 AM, "linux-scsi-ow...@vger.kernel.org on behalf of Bart Van Assche" wrote: >On 12/17/2015 08:57 PM, Himanshu Madhani wrote: >>From: Quinn Tran >>queue work element to specific process lessen cache miss > >How about replacing this patch by something like the (untested) patch >below ? The advantage of the patch below is that it doesn't require to >modify other target driver to keep their current behavior. > >Thanks, > >Bart. QT> Bart, Thanks for reviewing. Good point. Will rework patch to preserve current behavior of other drivers. <>
Re: [PATCH v2 11/16] qla2xxx: Add selective command queuing
> drivers/scsi/qla2xxx/qla_isr.c |2 +- > drivers/scsi/qla2xxx/qla_target.c | 13 - > drivers/scsi/qla2xxx/tcm_qla2xxx.c |4 ++-- > drivers/target/target_core_transport.c |5 - > include/target/target_core_base.h |1 + please split out target core changes from your drivers bits into a separate patch! > cmd->cmd_in_wq = 1; > cmd->cmd_flags |= BIT_0; > + cmd->se_cmd.cpuid = -1; As Bart pointed out please use the proper constants. Also make sure this is initialized in the core target code instad of here. -- 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 11/16] qla2xxx: Add selective command queuing
On 12/17/2015 08:57 PM, Himanshu Madhani wrote:
> From: Quinn Tran
>
> queue work element to specific process lessen cache miss
How about replacing this patch by something like the (untested) patch
below ? The advantage of the patch below is that it doesn't require to
modify other target driver to keep their current behavior.
Thanks,
Bart.
---
drivers/scsi/qla2xxx/qla_target.c | 5 -
drivers/scsi/qla2xxx/tcm_qla2xxx.c | 2 +-
drivers/target/target_core_transport.c | 4 +++-
include/target/target_core_base.h | 2 ++
4 files changed, 10 insertions(+), 3 deletions(-)
diff --git a/drivers/scsi/qla2xxx/qla_target.c
b/drivers/scsi/qla2xxx/qla_target.c
index 83214fc..1bd1511 100644
--- a/drivers/scsi/qla2xxx/qla_target.c
+++ b/drivers/scsi/qla2xxx/qla_target.c
@@ -3982,13 +3982,16 @@ static int qlt_handle_cmd_for_atio(struct scsi_qla_host
*vha,
cmd->cmd_in_wq = 1;
cmd->cmd_flags |= BIT_0;
+ cmd->se_cmd.cpuid = (ha->msix_count &&
+!cmd->atio.u.isp24.fcp_cmnd.rddata) ?
+ ha->tgt.rspq_vector_cpuid : WORK_CPU_UNBOUND;
spin_lock(&vha->cmd_list_lock);
list_add_tail(&cmd->cmd_list, &vha->qla_cmd_list);
spin_unlock(&vha->cmd_list_lock);
INIT_WORK(&cmd->work, qlt_do_work);
- queue_work(qla_tgt_wq, &cmd->work);
+ queue_work_on(cmd->se_cmd.cpuid, qla_tgt_wq, &cmd->work);
return 0;
}
diff --git a/drivers/scsi/qla2xxx/tcm_qla2xxx.c
b/drivers/scsi/qla2xxx/tcm_qla2xxx.c
index a6fd02a..e65ba34 100644
--- a/drivers/scsi/qla2xxx/tcm_qla2xxx.c
+++ b/drivers/scsi/qla2xxx/tcm_qla2xxx.c
@@ -419,7 +419,7 @@ static int tcm_qla2xxx_handle_cmd(scsi_qla_host_t *vha,
struct qla_tgt_cmd *cmd,
struct se_cmd *se_cmd = &cmd->se_cmd;
struct se_session *se_sess;
struct qla_tgt_sess *sess;
- int flags = TARGET_SCF_ACK_KREF;
+ int flags = TARGET_SCF_ACK_KREF | TARGET_SCF_USE_CPUID;
if (bidi)
flags |= TARGET_SCF_BIDI_OP;
diff --git a/drivers/target/target_core_transport.c
b/drivers/target/target_core_transport.c
index a8e21fb..3b1a002 100644
--- a/drivers/target/target_core_transport.c
+++ b/drivers/target/target_core_transport.c
@@ -681,7 +681,7 @@ void target_complete_cmd(struct se_cmd *cmd, u8 scsi_status)
cmd->transport_state |= CMD_T_COMPLETE;
spin_unlock_irqrestore(&cmd->t_state_lock, flags);
- queue_work(target_completion_wq, &cmd->work);
+ queue_work_on(cmd->cpuid, target_completion_wq, &cmd->work);
}
EXPORT_SYMBOL(target_complete_cmd);
@@ -1390,6 +1390,8 @@ int target_submit_cmd_map_sgls(struct se_cmd *se_cmd,
struct se_session *se_sess
*/
transport_init_se_cmd(se_cmd, se_tpg->se_tpg_tfo, se_sess,
data_length, data_dir, task_attr, sense);
+ if (!(flags & TARGET_SCF_USE_CPUID))
+ se_cmd->cpuid = WORK_CPU_UNBOUND;
if (flags & TARGET_SCF_UNKNOWN_SIZE)
se_cmd->unknown_data_length = 1;
/*
diff --git a/include/target/target_core_base.h
b/include/target/target_core_base.h
index e9e507a..38a1871 100644
--- a/include/target/target_core_base.h
+++ b/include/target/target_core_base.h
@@ -186,6 +186,7 @@ enum target_sc_flags_table {
TARGET_SCF_BIDI_OP = 0x01,
TARGET_SCF_ACK_KREF = 0x02,
TARGET_SCF_UNKNOWN_SIZE = 0x04,
+ TARGET_SCF_USE_CPUID= 0x08,
};
/* fabric independent task management function values */
@@ -519,6 +520,7 @@ struct se_cmd {
unsigned intt_prot_nents;
sense_reason_t pi_err;
sector_tbad_sector;
+ int cpuid;
};
struct se_ua {
--
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
Re: [PATCH v2 11/16] qla2xxx: Add selective command queuing
On 12/17/2015 08:57 PM, Himanshu Madhani wrote: diff --git a/drivers/scsi/qla2xxx/qla_target.c b/drivers/scsi/qla2xxx/qla_target.c index 9a4aed0..d3cd271 100644 --- a/drivers/scsi/qla2xxx/qla_target.c +++ b/drivers/scsi/qla2xxx/qla_target.c @@ -3982,13 +3982,24 @@ static int qlt_handle_cmd_for_atio(struct scsi_qla_host *vha, cmd->cmd_in_wq = 1; cmd->cmd_flags |= BIT_0; + cmd->se_cmd.cpuid = -1; A second comment about this patch: this initialization code must be moved to the target core (transport_init_se_cmd()). Otherwise the cpuid member variable won't be initialized when target_complete_cmd() is called for another target driver than tcm_qla2xxx. Bart. -- 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 11/16] qla2xxx: Add selective command queuing
On 12/17/2015 08:57 PM, Himanshu Madhani wrote:
From: Quinn Tran
queue work element to specific process lessen cache miss
Signed-off-by: Quinn Tran
Signed-off-by: Himanshu Madhani
---
drivers/scsi/qla2xxx/qla_isr.c |2 +-
drivers/scsi/qla2xxx/qla_target.c | 13 -
drivers/scsi/qla2xxx/tcm_qla2xxx.c |4 ++--
drivers/target/target_core_transport.c |5 -
include/target/target_core_base.h |1 +
5 files changed, 20 insertions(+), 5 deletions(-)
diff --git a/drivers/scsi/qla2xxx/qla_isr.c b/drivers/scsi/qla2xxx/qla_isr.c
index cf0fe8e..3e89122 100644
--- a/drivers/scsi/qla2xxx/qla_isr.c
+++ b/drivers/scsi/qla2xxx/qla_isr.c
@@ -3302,7 +3302,7 @@ static void qla_irq_affinity_notify(struct
irq_affinity_notify *notify,
}
}
-void qla_irq_affinity_release(struct kref *ref)
+static void qla_irq_affinity_release(struct kref *ref)
{
struct irq_affinity_notify *notify =
container_of(ref, struct irq_affinity_notify, kref);
diff --git a/drivers/scsi/qla2xxx/qla_target.c
b/drivers/scsi/qla2xxx/qla_target.c
index 9a4aed0..d3cd271 100644
--- a/drivers/scsi/qla2xxx/qla_target.c
+++ b/drivers/scsi/qla2xxx/qla_target.c
@@ -3982,13 +3982,24 @@ static int qlt_handle_cmd_for_atio(struct scsi_qla_host
*vha,
cmd->cmd_in_wq = 1;
cmd->cmd_flags |= BIT_0;
+ cmd->se_cmd.cpuid = -1;
Please use WORK_CPU_UNBOUND instead of -1. That will allow to convert
the if (cpuid == -1) queue_work() else queue_work_on() constructs into a
single queue_work_on() call.
Bart.
--
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
