* [PATCH 1/3] hisi_sas: fix v1 itct masks
2015-12-21 11:35 [PATCH 0/3] hisi_sas v1 hw itct fixes John Garry
@ 2015-12-21 11:35 ` John Garry
2016-01-08 4:31 ` Seymour, Shane M
2016-01-08 4:37 ` Matthew R. Ochs
2015-12-21 11:35 ` [PATCH 2/3] hisi_sas: fix typo in setup_itct_v1_hw() John Garry
` (2 subsequent siblings)
3 siblings, 2 replies; 11+ messages in thread
From: John Garry @ 2015-12-21 11:35 UTC (permalink / raw)
To: JBottomley, martin.petersen
Cc: linuxarm, zhangfei.gao, xuwei5, john.garry2, linux-scsi,
linux-kernel, John Garry
The mask fields are for quad-words, so add ULL suffix.
Also unreferenced ITCT_HDR_BREAK_REPLY and
ITCT_HDR_MAX_BURST are removed.
Fixes: 50af155b6c ("hisi_sas: Add v1 hardware reg")
Signed-off-by: John Garry <john.garry@huawei.com>
---
drivers/scsi/hisi_sas/hisi_sas_v1_hw.c | 29 ++++++++++++-----------------
1 file changed, 12 insertions(+), 17 deletions(-)
diff --git a/drivers/scsi/hisi_sas/hisi_sas_v1_hw.c b/drivers/scsi/hisi_sas/hisi_sas_v1_hw.c
index d543811..0af682c 100644
--- a/drivers/scsi/hisi_sas/hisi_sas_v1_hw.c
+++ b/drivers/scsi/hisi_sas/hisi_sas_v1_hw.c
@@ -247,41 +247,36 @@
/* ITCT header */
/* qw0 */
#define ITCT_HDR_DEV_TYPE_OFF 0
-#define ITCT_HDR_DEV_TYPE_MSK (0x3 << ITCT_HDR_DEV_TYPE_OFF)
+#define ITCT_HDR_DEV_TYPE_MSK (0x3ULL << ITCT_HDR_DEV_TYPE_OFF)
#define ITCT_HDR_VALID_OFF 2
-#define ITCT_HDR_VALID_MSK (0x1 << ITCT_HDR_VALID_OFF)
-#define ITCT_HDR_BREAK_REPLY_ENA_OFF 3
-#define ITCT_HDR_BREAK_REPLY_ENA_MSK (0x1 << ITCT_HDR_BREAK_REPLY_ENA_OFF)
+#define ITCT_HDR_VALID_MSK (0x1ULL << ITCT_HDR_VALID_OFF)
#define ITCT_HDR_AWT_CONTROL_OFF 4
-#define ITCT_HDR_AWT_CONTROL_MSK (0x1 << ITCT_HDR_AWT_CONTROL_OFF)
+#define ITCT_HDR_AWT_CONTROL_MSK (0x1ULL << ITCT_HDR_AWT_CONTROL_OFF)
#define ITCT_HDR_MAX_CONN_RATE_OFF 5
-#define ITCT_HDR_MAX_CONN_RATE_MSK (0xf << ITCT_HDR_MAX_CONN_RATE_OFF)
+#define ITCT_HDR_MAX_CONN_RATE_MSK (0xfULL << ITCT_HDR_MAX_CONN_RATE_OFF)
#define ITCT_HDR_VALID_LINK_NUM_OFF 9
-#define ITCT_HDR_VALID_LINK_NUM_MSK (0xf << ITCT_HDR_VALID_LINK_NUM_OFF)
+#define ITCT_HDR_VALID_LINK_NUM_MSK (0xfULL << ITCT_HDR_VALID_LINK_NUM_OFF)
#define ITCT_HDR_PORT_ID_OFF 13
-#define ITCT_HDR_PORT_ID_MSK (0x7 << ITCT_HDR_PORT_ID_OFF)
+#define ITCT_HDR_PORT_ID_MSK (0x7ULL << ITCT_HDR_PORT_ID_OFF)
#define ITCT_HDR_SMP_TIMEOUT_OFF 16
-#define ITCT_HDR_SMP_TIMEOUT_MSK (0xffff << ITCT_HDR_SMP_TIMEOUT_OFF)
-#define ITCT_HDR_MAX_BURST_BYTES_OFF 16
-#define ITCT_HDR_MAX_BURST_BYTES_MSK (0xffffffff << \
- ITCT_MAX_BURST_BYTES_OFF)
+#define ITCT_HDR_SMP_TIMEOUT_MSK (0xffffULL << ITCT_HDR_SMP_TIMEOUT_OFF)
/* qw1 */
#define ITCT_HDR_MAX_SAS_ADDR_OFF 0
#define ITCT_HDR_MAX_SAS_ADDR_MSK (0xffffffffffffffff << \
ITCT_HDR_MAX_SAS_ADDR_OFF)
/* qw2 */
#define ITCT_HDR_IT_NEXUS_LOSS_TL_OFF 0
-#define ITCT_HDR_IT_NEXUS_LOSS_TL_MSK (0xffff << \
+#define ITCT_HDR_IT_NEXUS_LOSS_TL_MSK (0xffffULL << \
ITCT_HDR_IT_NEXUS_LOSS_TL_OFF)
#define ITCT_HDR_BUS_INACTIVE_TL_OFF 16
-#define ITCT_HDR_BUS_INACTIVE_TL_MSK (0xffff << \
+#define ITCT_HDR_BUS_INACTIVE_TL_MSK (0xffffULL << \
ITCT_HDR_BUS_INACTIVE_TL_OFF)
#define ITCT_HDR_MAX_CONN_TL_OFF 32
-#define ITCT_HDR_MAX_CONN_TL_MSK (0xffff << \
+#define ITCT_HDR_MAX_CONN_TL_MSK (0xffffULL << \
ITCT_HDR_MAX_CONN_TL_OFF)
#define ITCT_HDR_REJ_OPEN_TL_OFF 48
-#define ITCT_HDR_REJ_OPEN_TL_MSK (0xffff << \
- ITCT_REJ_OPEN_TL_OFF)
+#define ITCT_HDR_REJ_OPEN_TL_MSK (0xffffULL << \
+ ITCT_HDR_REJ_OPEN_TL_OFF)
/* Err record header */
#define ERR_HDR_DMA_TX_ERR_TYPE_OFF 0
--
1.9.1
^ permalink raw reply [flat|nested] 11+ messages in thread* RE: [PATCH 1/3] hisi_sas: fix v1 itct masks
2015-12-21 11:35 ` [PATCH 1/3] hisi_sas: fix v1 itct masks John Garry
@ 2016-01-08 4:31 ` Seymour, Shane M
2016-01-08 4:37 ` Matthew R. Ochs
1 sibling, 0 replies; 11+ messages in thread
From: Seymour, Shane M @ 2016-01-08 4:31 UTC (permalink / raw)
To: John Garry, JBottomley, martin.petersen
Cc: linuxarm, zhangfei.gao, xuwei5, john.garry2, linux-scsi, linux-kernel
> The mask fields are for quad-words, so add ULL suffix.
> Also unreferenced ITCT_HDR_BREAK_REPLY and ITCT_HDR_MAX_BURST are removed.
>
> Fixes: 50af155b6c ("hisi_sas: Add v1 hardware reg")
>
> Signed-off-by: John Garry <john.garry@huawei.com>
Reviewed-by: Shane Seymour <shane.seymour@hpe.com>
^ permalink raw reply [flat|nested] 11+ messages in thread* Re: [PATCH 1/3] hisi_sas: fix v1 itct masks
2015-12-21 11:35 ` [PATCH 1/3] hisi_sas: fix v1 itct masks John Garry
2016-01-08 4:31 ` Seymour, Shane M
@ 2016-01-08 4:37 ` Matthew R. Ochs
1 sibling, 0 replies; 11+ messages in thread
From: Matthew R. Ochs @ 2016-01-08 4:37 UTC (permalink / raw)
To: John Garry
Cc: James Bottomley, Martin K. Petersen, linuxarm, zhangfei.gao,
xuwei5, john.garry2, linux-scsi, linux-kernel
Reviewed-by: Matthew R. Ochs <mrochs@linux.vnet.ibm.com>
^ permalink raw reply [flat|nested] 11+ messages in thread
* [PATCH 2/3] hisi_sas: fix typo in setup_itct_v1_hw()
2015-12-21 11:35 [PATCH 0/3] hisi_sas v1 hw itct fixes John Garry
2015-12-21 11:35 ` [PATCH 1/3] hisi_sas: fix v1 itct masks John Garry
@ 2015-12-21 11:35 ` John Garry
2016-01-08 4:30 ` Seymour, Shane M
2016-01-08 4:40 ` Matthew R. Ochs
2015-12-21 11:35 ` [PATCH 3/3] hisi_sas: use u64 for qw0 in free_device_v1_hw() John Garry
2016-01-08 17:13 ` [PATCH 0/3] hisi_sas v1 hw itct fixes Martin K. Petersen
3 siblings, 2 replies; 11+ messages in thread
From: John Garry @ 2015-12-21 11:35 UTC (permalink / raw)
To: JBottomley, martin.petersen
Cc: linuxarm, zhangfei.gao, xuwei5, john.garry2, linux-scsi,
linux-kernel, John Garry
We were doing a arithmetic comparison instead
of logical shift by accident.
Mis-programming the itct did not seem to make
a difference to operation.
Fixes: abda97c2fe874 ("hisi_sas: Add dev_found")
Signed-off-by: John Garry <john.garry@huawei.com>
---
drivers/scsi/hisi_sas/hisi_sas_v1_hw.c | 8 ++++----
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/drivers/scsi/hisi_sas/hisi_sas_v1_hw.c b/drivers/scsi/hisi_sas/hisi_sas_v1_hw.c
index 0af682c..38ff575 100644
--- a/drivers/scsi/hisi_sas/hisi_sas_v1_hw.c
+++ b/drivers/scsi/hisi_sas/hisi_sas_v1_hw.c
@@ -528,10 +528,10 @@ static void setup_itct_v1_hw(struct hisi_hba *hisi_hba,
itct->sas_addr = __swab64(itct->sas_addr);
/* qw2 */
- itct->qw2 = cpu_to_le64((500 < ITCT_HDR_IT_NEXUS_LOSS_TL_OFF) |
- (0xff00 < ITCT_HDR_BUS_INACTIVE_TL_OFF) |
- (0xff00 < ITCT_HDR_MAX_CONN_TL_OFF) |
- (0xff00 < ITCT_HDR_REJ_OPEN_TL_OFF));
+ itct->qw2 = cpu_to_le64((500ULL << ITCT_HDR_IT_NEXUS_LOSS_TL_OFF) |
+ (0xff00ULL << ITCT_HDR_BUS_INACTIVE_TL_OFF) |
+ (0xff00ULL << ITCT_HDR_MAX_CONN_TL_OFF) |
+ (0xff00ULL << ITCT_HDR_REJ_OPEN_TL_OFF));
}
static void free_device_v1_hw(struct hisi_hba *hisi_hba,
--
1.9.1
^ permalink raw reply [flat|nested] 11+ messages in thread* RE: [PATCH 2/3] hisi_sas: fix typo in setup_itct_v1_hw()
2015-12-21 11:35 ` [PATCH 2/3] hisi_sas: fix typo in setup_itct_v1_hw() John Garry
@ 2016-01-08 4:30 ` Seymour, Shane M
2016-01-08 4:40 ` Matthew R. Ochs
1 sibling, 0 replies; 11+ messages in thread
From: Seymour, Shane M @ 2016-01-08 4:30 UTC (permalink / raw)
To: John Garry, JBottomley, martin.petersen
Cc: linuxarm, zhangfei.gao, xuwei5, john.garry2, linux-scsi, linux-kernel
> We were doing a arithmetic comparison instead of logical shift by accident.
> Mis-programming the itct did not seem to make a difference to operation.
>
> Fixes: abda97c2fe874 ("hisi_sas: Add dev_found")
>
> Signed-off-by: John Garry <john.garry@huawei.com>
Reviewed-by: Shane Seymour <shane.seymour@hpe.com>
^ permalink raw reply [flat|nested] 11+ messages in thread* Re: [PATCH 2/3] hisi_sas: fix typo in setup_itct_v1_hw()
2015-12-21 11:35 ` [PATCH 2/3] hisi_sas: fix typo in setup_itct_v1_hw() John Garry
2016-01-08 4:30 ` Seymour, Shane M
@ 2016-01-08 4:40 ` Matthew R. Ochs
1 sibling, 0 replies; 11+ messages in thread
From: Matthew R. Ochs @ 2016-01-08 4:40 UTC (permalink / raw)
To: John Garry
Cc: James Bottomley, Martin K. Petersen, linuxarm, zhangfei.gao,
xuwei5, john.garry2, linux-scsi, linux-kernel
Reviewed-by: Matthew R. Ochs <mrochs@linux.vnet.ibm.com>
^ permalink raw reply [flat|nested] 11+ messages in thread
* [PATCH 3/3] hisi_sas: use u64 for qw0 in free_device_v1_hw()
2015-12-21 11:35 [PATCH 0/3] hisi_sas v1 hw itct fixes John Garry
2015-12-21 11:35 ` [PATCH 1/3] hisi_sas: fix v1 itct masks John Garry
2015-12-21 11:35 ` [PATCH 2/3] hisi_sas: fix typo in setup_itct_v1_hw() John Garry
@ 2015-12-21 11:35 ` John Garry
2016-01-08 4:29 ` Seymour, Shane M
2016-01-08 4:51 ` Matthew R. Ochs
2016-01-08 17:13 ` [PATCH 0/3] hisi_sas v1 hw itct fixes Martin K. Petersen
3 siblings, 2 replies; 11+ messages in thread
From: John Garry @ 2015-12-21 11:35 UTC (permalink / raw)
To: JBottomley, martin.petersen
Cc: linuxarm, zhangfei.gao, xuwei5, john.garry2, linux-scsi,
linux-kernel, John Garry
By reading in itct.qw0 into a 32b variable the top
32 bits were being lost.
In practice this was OK as they were zeroes.
Fixes: 27a3f229 ("hisi_sas: Add cq interrupt")
Signed-off-by: John Garry <john.garry@huawei.com>
---
drivers/scsi/hisi_sas/hisi_sas_v1_hw.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/drivers/scsi/hisi_sas/hisi_sas_v1_hw.c b/drivers/scsi/hisi_sas/hisi_sas_v1_hw.c
index 38ff575..057fdeb 100644
--- a/drivers/scsi/hisi_sas/hisi_sas_v1_hw.c
+++ b/drivers/scsi/hisi_sas/hisi_sas_v1_hw.c
@@ -539,7 +539,8 @@ static void free_device_v1_hw(struct hisi_hba *hisi_hba,
{
u64 dev_id = sas_dev->device_id;
struct hisi_sas_itct *itct = &hisi_hba->itct[dev_id];
- u32 qw0, reg_val = hisi_sas_read32(hisi_hba, CFG_AGING_TIME);
+ u64 qw0;
+ u32 reg_val = hisi_sas_read32(hisi_hba, CFG_AGING_TIME);
reg_val |= CFG_AGING_TIME_ITCT_REL_MSK;
hisi_sas_write32(hisi_hba, CFG_AGING_TIME, reg_val);
--
1.9.1
^ permalink raw reply [flat|nested] 11+ messages in thread* RE: [PATCH 3/3] hisi_sas: use u64 for qw0 in free_device_v1_hw()
2015-12-21 11:35 ` [PATCH 3/3] hisi_sas: use u64 for qw0 in free_device_v1_hw() John Garry
@ 2016-01-08 4:29 ` Seymour, Shane M
2016-01-08 4:51 ` Matthew R. Ochs
1 sibling, 0 replies; 11+ messages in thread
From: Seymour, Shane M @ 2016-01-08 4:29 UTC (permalink / raw)
To: John Garry, JBottomley, martin.petersen
Cc: linuxarm, zhangfei.gao, xuwei5, john.garry2, linux-scsi, linux-kernel
> By reading in itct.qw0 into a 32b variable the top
> 32 bits were being lost.
> In practice this was OK as they were zeroes.
>
> Fixes: 27a3f229 ("hisi_sas: Add cq interrupt")
>
> Signed-off-by: John Garry <john.garry@huawei.com>
Reviewed-by: Shane Seymour <shane.seymour@hpe.com>
^ permalink raw reply [flat|nested] 11+ messages in thread* Re: [PATCH 3/3] hisi_sas: use u64 for qw0 in free_device_v1_hw()
2015-12-21 11:35 ` [PATCH 3/3] hisi_sas: use u64 for qw0 in free_device_v1_hw() John Garry
2016-01-08 4:29 ` Seymour, Shane M
@ 2016-01-08 4:51 ` Matthew R. Ochs
1 sibling, 0 replies; 11+ messages in thread
From: Matthew R. Ochs @ 2016-01-08 4:51 UTC (permalink / raw)
To: John Garry
Cc: James Bottomley, Martin K. Petersen, linuxarm, zhangfei.gao,
xuwei5, john.garry2, linux-scsi, linux-kernel
> On Dec 21, 2015, at 5:35 AM, John Garry <john.garry@huawei.com> wrote:
>
> By reading in itct.qw0 into a 32b variable the top
> 32 bits were being lost.
> In practice this was OK as they were zeroes.
>
> Fixes: 27a3f229 ("hisi_sas: Add cq interrupt")
>
> Signed-off-by: John Garry <john.garry@huawei.com>
> ---
> drivers/scsi/hisi_sas/hisi_sas_v1_hw.c | 3 ++-
> 1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/scsi/hisi_sas/hisi_sas_v1_hw.c b/drivers/scsi/hisi_sas/hisi_sas_v1_hw.c
> index 38ff575..057fdeb 100644
> --- a/drivers/scsi/hisi_sas/hisi_sas_v1_hw.c
> +++ b/drivers/scsi/hisi_sas/hisi_sas_v1_hw.c
> @@ -539,7 +539,8 @@ static void free_device_v1_hw(struct hisi_hba *hisi_hba,
> {
> u64 dev_id = sas_dev->device_id;
> struct hisi_sas_itct *itct = &hisi_hba->itct[dev_id];
> - u32 qw0, reg_val = hisi_sas_read32(hisi_hba, CFG_AGING_TIME);
> + u64 qw0;
This should probably be a __le64, but not a big deal.
Reviewed-by: Matthew R. Ochs <mrochs@linux.vnet.ibm.com>
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [PATCH 0/3] hisi_sas v1 hw itct fixes
2015-12-21 11:35 [PATCH 0/3] hisi_sas v1 hw itct fixes John Garry
` (2 preceding siblings ...)
2015-12-21 11:35 ` [PATCH 3/3] hisi_sas: use u64 for qw0 in free_device_v1_hw() John Garry
@ 2016-01-08 17:13 ` Martin K. Petersen
3 siblings, 0 replies; 11+ messages in thread
From: Martin K. Petersen @ 2016-01-08 17:13 UTC (permalink / raw)
To: John Garry
Cc: JBottomley, martin.petersen, linuxarm, zhangfei.gao, xuwei5,
john.garry2, linux-scsi, linux-kernel, Shane M Seymour,
Matthew R. Ochs
>>>>> "John" == John Garry <john.garry@huawei.com> writes:
John> This patchset introduces some minor fixes for v1 hw itct
John> configuration.
Applied to 4.5/scsi-queue.
--
Martin K. Petersen Oracle Linux Engineering
^ permalink raw reply [flat|nested] 11+ messages in thread