[PATCH 1/4] scsi: ufs: core: Fix mcq tag calcualtion

[PATCH 1/4] scsi: ufs: core: Fix mcq tag calcualtion

[Po-Wen Kao]

Transfer command descriptor is allocated in ufshcd_memory_alloc()
and referenced by transfer request descriptor with stride size
sizeof_utp_transfer_cmd_desc()
instead of
sizeof(struct utp_transfer_cmd_desc).

Consequently, computing tag by address offset should also refer to the
same stride.

As suggested, sizeof_utp_transfer_cmd_desc() is further renamed to
ufshcd_get_ucd_size().

Signed-off-by: Po-Wen Kao <powen.kao@mediatek.com>
---
 drivers/ufs/core/ufs-mcq.c | 2 +-
 drivers/ufs/core/ufshcd.c  | 8 ++++----
 include/ufs/ufshcd.h       | 2 +-
 3 files changed, 6 insertions(+), 6 deletions(-)

diff --git a/drivers/ufs/core/ufs-mcq.c b/drivers/ufs/core/ufs-mcq.c
index 31df052fbc41..a39746b2a8be 100644
--- a/drivers/ufs/core/ufs-mcq.c
+++ b/drivers/ufs/core/ufs-mcq.c
@@ -265,7 +265,7 @@ static int ufshcd_mcq_get_tag(struct ufs_hba *hba,
 	addr = (le64_to_cpu(cqe->command_desc_base_addr) & CQE_UCD_BA) -
 		hba->ucdl_dma_addr;
 
-	return div_u64(addr, sizeof(struct utp_transfer_cmd_desc));
+	return div_u64(addr, ufshcd_get_ucd_size(hba));
 }
 
 static void ufshcd_mcq_process_cqe(struct ufs_hba *hba,
diff --git a/drivers/ufs/core/ufshcd.c b/drivers/ufs/core/ufshcd.c
index 3b3cf78d3b10..81c9f07ebfc8 100644
--- a/drivers/ufs/core/ufshcd.c
+++ b/drivers/ufs/core/ufshcd.c
@@ -2823,10 +2823,10 @@ static void ufshcd_map_queues(struct Scsi_Host *shost)
 static void ufshcd_init_lrb(struct ufs_hba *hba, struct ufshcd_lrb *lrb, int i)
 {
 	struct utp_transfer_cmd_desc *cmd_descp = (void *)hba->ucdl_base_addr +
-		i * sizeof_utp_transfer_cmd_desc(hba);
+		i * ufshcd_get_ucd_size(hba);
 	struct utp_transfer_req_desc *utrdlp = hba->utrdl_base_addr;
 	dma_addr_t cmd_desc_element_addr = hba->ucdl_dma_addr +
-		i * sizeof_utp_transfer_cmd_desc(hba);
+		i * ufshcd_get_ucd_size(hba);
 	u16 response_offset = offsetof(struct utp_transfer_cmd_desc,
 				       response_upiu);
 	u16 prdt_offset = offsetof(struct utp_transfer_cmd_desc, prd_table);
@@ -3735,7 +3735,7 @@ static int ufshcd_memory_alloc(struct ufs_hba *hba)
 	size_t utmrdl_size, utrdl_size, ucdl_size;
 
 	/* Allocate memory for UTP command descriptors */
-	ucdl_size = sizeof_utp_transfer_cmd_desc(hba) * hba->nutrs;
+	ucdl_size = ufshcd_get_ucd_size(hba) * hba->nutrs;
 	hba->ucdl_base_addr = dmam_alloc_coherent(hba->dev,
 						  ucdl_size,
 						  &hba->ucdl_dma_addr,
@@ -3835,7 +3835,7 @@ static void ufshcd_host_memory_configure(struct ufs_hba *hba)
 	prdt_offset =
 		offsetof(struct utp_transfer_cmd_desc, prd_table);
 
-	cmd_desc_size = sizeof_utp_transfer_cmd_desc(hba);
+	cmd_desc_size = ufshcd_get_ucd_size(hba);
 	cmd_desc_dma_addr = hba->ucdl_dma_addr;
 
 	for (i = 0; i < hba->nutrs; i++) {
diff --git a/include/ufs/ufshcd.h b/include/ufs/ufshcd.h
index ed9e3d5addb3..8f79cca449e1 100644
--- a/include/ufs/ufshcd.h
+++ b/include/ufs/ufshcd.h
@@ -1136,7 +1136,7 @@ static inline size_t ufshcd_sg_entry_size(const struct ufs_hba *hba)
 	({ (void)(hba); BUILD_BUG_ON(sg_entry_size != sizeof(struct ufshcd_sg_entry)); })
 #endif
 
-static inline size_t sizeof_utp_transfer_cmd_desc(const struct ufs_hba *hba)
+static inline size_t ufshcd_get_ucd_size(const struct ufs_hba *hba)
 {
 	return sizeof(struct utp_transfer_cmd_desc) + SG_ALL * ufshcd_sg_entry_size(hba);
 }
-- 
2.18.0

[PATCH 2/4] scsi: ufs: core: Fix mcq nr_hw_queues

[Po-Wen Kao]

Need to add one to QMAX to obtain number of hardware queue.

Signed-off-by: Po-Wen Kao <powen.kao@mediatek.com>
---
 drivers/ufs/core/ufs-mcq.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/ufs/core/ufs-mcq.c b/drivers/ufs/core/ufs-mcq.c
index a39746b2a8be..5d5bc0bc6e88 100644
--- a/drivers/ufs/core/ufs-mcq.c
+++ b/drivers/ufs/core/ufs-mcq.c
@@ -150,7 +150,7 @@ static int ufshcd_mcq_config_nr_queues(struct ufs_hba *hba)
 	u32 hba_maxq, rem, tot_queues;
 	struct Scsi_Host *host = hba->host;
 
-	hba_maxq = FIELD_GET(MAX_QUEUE_SUP, hba->mcq_capabilities);
+	hba_maxq = FIELD_GET(MAX_QUEUE_SUP, hba->mcq_capabilities) + 1 ;
 
 	tot_queues = UFS_MCQ_NUM_DEV_CMD_QUEUES + read_queues + poll_queues +
 			rw_queues;
-- 
2.18.0

[PATCH 3/4] scsi: ufs: core: Add hwq print for debug

[Po-Wen Kao]

Introduce hwq printing function for debug purpose.
- ufshcd_mcq_print_hwqs()

Signed-off-by: Po-Wen Kao <powen.kao@mediatek.com>
---
 drivers/ufs/core/ufs-mcq.c     | 32 +++++++++++++++++++++++++++++++-
 drivers/ufs/core/ufshcd-priv.h |  9 +++++++++
 drivers/ufs/core/ufshcd.c      | 18 +++++++++++-------
 include/ufs/ufshci.h           | 12 ++++++++++++
 4 files changed, 63 insertions(+), 8 deletions(-)

diff --git a/drivers/ufs/core/ufs-mcq.c b/drivers/ufs/core/ufs-mcq.c
index 5d5bc0bc6e88..d1ff3ccd2085 100644
--- a/drivers/ufs/core/ufs-mcq.c
+++ b/drivers/ufs/core/ufs-mcq.c
@@ -23,7 +23,6 @@
 
 #define MAX_DEV_CMD_ENTRIES	2
 #define MCQ_CFG_MAC_MASK	GENMASK(16, 8)
-#define MCQ_QCFG_SIZE		0x40
 #define MCQ_ENTRY_SIZE_IN_DWORD	8
 #define CQE_UCD_BA GENMASK_ULL(63, 7)
 
@@ -75,6 +74,13 @@ module_param_cb(poll_queues, &poll_queue_count_ops, &poll_queues, 0644);
 MODULE_PARM_DESC(poll_queues,
 		 "Number of poll queues used for r/w. Default value is 1");
 
+static const int mcq_opr_size[] = {
+	MCQ_SQD_SIZE,
+	MCQ_SQIS_SIZE,
+	MCQ_CQD_SIZE,
+	MCQ_CQIS_SIZE,
+};
+
 /**
  * ufshcd_mcq_config_mac - Set the #Max Activ Cmds.
  * @hba: per adapter instance
@@ -237,6 +243,30 @@ static void __iomem *mcq_opr_base(struct ufs_hba *hba,
 	return opr->base + opr->stride * i;
 }
 
+void ufshcd_mcq_print_hwqs(struct ufs_hba *hba, unsigned long bitmap)
+{
+	int id, i;
+	char prefix[15];
+
+	if (!is_mcq_enabled(hba))
+		return;
+
+	for_each_set_bit(id, &bitmap, hba->nr_hw_queues) {
+		snprintf(prefix, sizeof(prefix), "q%d SQCFG: ", id);
+		ufshcd_hex_dump(prefix,
+			hba->mcq_base + MCQ_QCFG_SIZE * id, MCQ_QCFG_SQ_SIZE);
+
+		snprintf(prefix, sizeof(prefix), "q%d CQCFG: ", id);
+		ufshcd_hex_dump(prefix,
+			hba->mcq_base + MCQ_QCFG_SIZE * id + MCQ_QCFG_SQ_SIZE, MCQ_QCFG_CQ_SIZE);
+
+		for (i = 0; i < OPR_MAX ; i++) {
+			snprintf(prefix, sizeof(prefix), "q%d OPR%d: ", id, i);
+			ufshcd_hex_dump(prefix, mcq_opr_base(hba, i, id), mcq_opr_size[i]);
+		}
+	}
+}
+
 u32 ufshcd_mcq_read_cqis(struct ufs_hba *hba, int i)
 {
 	return readl(mcq_opr_base(hba, OPR_CQIS, i) + REG_CQIS);
diff --git a/drivers/ufs/core/ufshcd-priv.h b/drivers/ufs/core/ufshcd-priv.h
index 529f8507a5e4..13534a9a6d0a 100644
--- a/drivers/ufs/core/ufshcd-priv.h
+++ b/drivers/ufs/core/ufshcd-priv.h
@@ -6,6 +6,14 @@
 #include <linux/pm_runtime.h>
 #include <ufs/ufshcd.h>
 
+#define ufshcd_hex_dump(prefix_str, buf, len) do {                       \
+	size_t __len = (len);                                            \
+	print_hex_dump(KERN_ERR, prefix_str,                             \
+		       __len > 4 ? DUMP_PREFIX_OFFSET : DUMP_PREFIX_NONE,\
+		       16, 4, buf, __len, false);                        \
+} while (0)
+
+
 static inline bool ufshcd_is_user_access_allowed(struct ufs_hba *hba)
 {
 	return !hba->shutting_down;
@@ -65,6 +73,7 @@ void ufshcd_compl_one_cqe(struct ufs_hba *hba, int task_tag,
 			  struct cq_entry *cqe);
 int ufshcd_mcq_init(struct ufs_hba *hba);
 int ufshcd_mcq_decide_queue_depth(struct ufs_hba *hba);
+void ufshcd_mcq_print_hwqs(struct ufs_hba *hba, unsigned long bitmap);
 int ufshcd_mcq_memory_alloc(struct ufs_hba *hba);
 void ufshcd_mcq_make_queues_operational(struct ufs_hba *hba);
 void ufshcd_mcq_config_mac(struct ufs_hba *hba, u32 max_active_cmds);
diff --git a/drivers/ufs/core/ufshcd.c b/drivers/ufs/core/ufshcd.c
index 81c9f07ebfc8..a15a5a78160c 100644
--- a/drivers/ufs/core/ufshcd.c
+++ b/drivers/ufs/core/ufshcd.c
@@ -135,13 +135,6 @@ MODULE_PARM_DESC(use_mcq_mode, "Control MCQ mode for controllers starting from U
 		_ret;                                                   \
 	})
 
-#define ufshcd_hex_dump(prefix_str, buf, len) do {                       \
-	size_t __len = (len);                                            \
-	print_hex_dump(KERN_ERR, prefix_str,                             \
-		       __len > 4 ? DUMP_PREFIX_OFFSET : DUMP_PREFIX_NONE,\
-		       16, 4, buf, __len, false);                        \
-} while (0)
-
 int ufshcd_dump_regs(struct ufs_hba *hba, size_t offset, size_t len,
 		     const char *prefix)
 {
@@ -536,6 +529,8 @@ static
 void ufshcd_print_trs(struct ufs_hba *hba, unsigned long bitmap, bool pr_prdt)
 {
 	const struct ufshcd_lrb *lrbp;
+	struct ufs_hw_queue *hwq;
+	struct scsi_cmnd *cmd;
 	int prdt_length;
 	int tag;
 
@@ -574,7 +569,16 @@ void ufshcd_print_trs(struct ufs_hba *hba, unsigned long bitmap, bool pr_prdt)
 		if (pr_prdt)
 			ufshcd_hex_dump("UPIU PRDT: ", lrbp->ucd_prdt_ptr,
 				ufshcd_sg_entry_size(hba) * prdt_length);
+
+		if (is_mcq_enabled(hba)) {
+			cmd = lrbp->cmd;
+			if (!cmd)
+				return;
+			hwq = ufshcd_mcq_req_to_hwq(hba, scsi_cmd_to_rq(cmd));
+			ufshcd_mcq_print_hwqs(hba, 1 << hwq->id);
+		}
 	}
+
 }
 
 static void ufshcd_print_tmrs(struct ufs_hba *hba, unsigned long bitmap)
diff --git a/include/ufs/ufshci.h b/include/ufs/ufshci.h
index 11424bb03814..027a2e884f89 100644
--- a/include/ufs/ufshci.h
+++ b/include/ufs/ufshci.h
@@ -185,6 +185,18 @@ static inline u32 ufshci_version(u32 major, u32 minor)
 				CRYPTO_ENGINE_FATAL_ERROR |\
 				UIC_LINK_LOST)
 
+/* MCQ size */
+#define MCQ_QCFG_SIZE			0x40
+#define MCQ_QCFG_SQ_SIZE		0x20
+#define MCQ_QCFG_CQ_SIZE		0x20
+
+enum {
+	MCQ_SQD_SIZE		= 0x14,
+	MCQ_SQIS_SIZE		= 0x08,
+	MCQ_CQD_SIZE		= 0x08,
+	MCQ_CQIS_SIZE		= 0x0C,
+};
+
 /* HCS - Host Controller Status 30h */
 #define DEVICE_PRESENT				0x1
 #define UTP_TRANSFER_REQ_LIST_READY		0x2
-- 
2.18.0

[PATCH 4/4] scsi: ufs: core: Remove redundant check

[Po-Wen Kao]

is_mcq_supported() already check on use_mcq_mode.

Signed-off-by: Po-Wen Kao <powen.kao@mediatek.com>
---
 drivers/ufs/core/ufshcd.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/ufs/core/ufshcd.c b/drivers/ufs/core/ufshcd.c
index a15a5a78160c..21e3bf5d8f08 100644
--- a/drivers/ufs/core/ufshcd.c
+++ b/drivers/ufs/core/ufshcd.c
@@ -8548,7 +8548,7 @@ static int ufshcd_device_init(struct ufs_hba *hba, bool init_dev_params)
 			hba->scsi_host_added = true;
 		}
 		/* MCQ may be disabled if ufshcd_alloc_mcq() fails */
-		if (is_mcq_supported(hba) && use_mcq_mode)
+		if (is_mcq_supported(hba))
 			ufshcd_config_mcq(hba);
 	}
 
-- 
2.18.0

Re: [PATCH 1/4] scsi: ufs: core: Fix mcq tag calcualtion

[Bart Van Assche]

On 2/21/23 07:29, Po-Wen Kao wrote:
> Transfer command descriptor is allocated in ufshcd_memory_alloc()
> and referenced by transfer request descriptor with stride size
> sizeof_utp_transfer_cmd_desc()
> instead of
> sizeof(struct utp_transfer_cmd_desc).
> 
> Consequently, computing tag by address offset should also refer to the
> same stride.
> 
> As suggested, sizeof_utp_transfer_cmd_desc() is further renamed to
> ufshcd_get_ucd_size().

This patch includes two changes:
(1) a bug fix.
(2) a name change for a function.

A Linux kernel patch should include only one logical change. Please 
split this patch.

Thanks,

Bart.