[PATCH 0/4] Add support for Marvell CN10K CPT block

[PATCH 0/4] Add support for Marvell CN10K CPT block

[Srujana Challa]

The current CPT driver supports OcteonTX2 silicon variants.
The same OcteonTX2 Resource Virtualization Unit(RVU) is
carried forward to the next-gen silicon ie OcteonTX3(CN10K),
with some changes and feature enhancements.

This patch series adds support for CN10K silicon.

Srujana Challa (4):
  crypto: octeontx2: Add mailbox support for CN10K
  crypto: octeontx2: add support to map LMTST region for CN10K
  crypto: octeontx2: add support for CPT operations on CN10K
  crypto: octeontx2: enable and handle ME interrupts

 drivers/crypto/marvell/octeontx2/Makefile     |  13 +-
 drivers/crypto/marvell/octeontx2/cn10k_cpt.c  |  93 ++++++++++
 drivers/crypto/marvell/octeontx2/cn10k_cpt.h  |  36 ++++
 .../marvell/octeontx2/otx2_cpt_common.h       |  23 +++
 .../marvell/octeontx2/otx2_cpt_hw_types.h     |  16 +-
 drivers/crypto/marvell/octeontx2/otx2_cptlf.c |   9 +-
 drivers/crypto/marvell/octeontx2/otx2_cptlf.h |  10 ++
 drivers/crypto/marvell/octeontx2/otx2_cptpf.h |   1 +
 .../marvell/octeontx2/otx2_cptpf_main.c       | 160 +++++++++++++-----
 .../marvell/octeontx2/otx2_cptpf_ucode.c      |  32 +++-
 .../marvell/octeontx2/otx2_cptpf_ucode.h      |   8 +-
 drivers/crypto/marvell/octeontx2/otx2_cptvf.h |   3 +
 .../marvell/octeontx2/otx2_cptvf_main.c       |  49 ++++--
 .../marvell/octeontx2/otx2_cptvf_mbox.c       |  43 +++++
 .../marvell/octeontx2/otx2_cptvf_reqmgr.c     |  17 +-
 15 files changed, 438 insertions(+), 75 deletions(-)
 create mode 100644 drivers/crypto/marvell/octeontx2/cn10k_cpt.c
 create mode 100644 drivers/crypto/marvell/octeontx2/cn10k_cpt.h

-- 
2.29.0

[PATCH 1/4] crypto: octeontx2: Add mailbox support for CN10K

[Srujana Challa]

Mailbox region configuration has some changes on CN10K platform
from OcteonTX2(CN9XX) platform.

On CN10K platform:
The DRAM region allocated to PF is enumerated as PF BAR4 memory.
PF BAR4 contains AF-PF mbox region followed by its VFs mbox region.
AF-PF mbox region base address is configured at RVU_AF_PFX_BAR4_ADDR
PF-VF mailbox base address is configured at
RVU_PF(x)_VF_MBOX_ADDR = RVU_AF_PF()_BAR4_ADDR+64KB. PF access its
mbox region via BAR4, whereas VF accesses PF-VF DRAM mailboxes via
BAR2 indirect access.

On CN9XX platform:
Mailbox region in DRAM is divided into two parts AF-PF mbox region and
PF-VF mbox region i.e all PFs mbox region is contiguous similarly all
VFs.
The base address of the AF-PF mbox region is configured at
RVU_AF_PF_BAR4_ADDR.
AF-PF1 mbox address can be calculated as RVU_AF_PF_BAR4_ADDR * mbox
size.

This patch changes mbox initialization to support both CN9XX and CN10K
platform.
This patch also removes platform specific name from the PF/VF driver name
to make it appropriate for all supported platforms.

Signed-off-by: Srujana Challa <schalla@marvell.com>
---
 drivers/crypto/marvell/octeontx2/Makefile     | 12 ++---
 .../marvell/octeontx2/otx2_cpt_common.h       | 20 +++++++++
 .../marvell/octeontx2/otx2_cpt_hw_types.h     |  3 ++
 drivers/crypto/marvell/octeontx2/otx2_cptpf.h |  1 +
 .../marvell/octeontx2/otx2_cptpf_main.c       | 35 +++++++++------
 drivers/crypto/marvell/octeontx2/otx2_cptvf.h |  3 ++
 .../marvell/octeontx2/otx2_cptvf_main.c       | 45 +++++++++++++------
 .../marvell/octeontx2/otx2_cptvf_mbox.c       | 43 ++++++++++++++++++
 8 files changed, 129 insertions(+), 33 deletions(-)

diff --git a/drivers/crypto/marvell/octeontx2/Makefile b/drivers/crypto/marvell/octeontx2/Makefile
index b9c6201019e0..10e1fe056a9e 100644
--- a/drivers/crypto/marvell/octeontx2/Makefile
+++ b/drivers/crypto/marvell/octeontx2/Makefile
@@ -1,10 +1,10 @@
 # SPDX-License-Identifier: GPL-2.0-only
-obj-$(CONFIG_CRYPTO_DEV_OCTEONTX2_CPT) += octeontx2-cpt.o octeontx2-cptvf.o
+obj-$(CONFIG_CRYPTO_DEV_OCTEONTX2_CPT) += rvu_cptpf.o rvu_cptvf.o
 
-octeontx2-cpt-objs := otx2_cptpf_main.o otx2_cptpf_mbox.o \
-		      otx2_cpt_mbox_common.o otx2_cptpf_ucode.o otx2_cptlf.o
-octeontx2-cptvf-objs := otx2_cptvf_main.o otx2_cptvf_mbox.o otx2_cptlf.o \
-			otx2_cpt_mbox_common.o otx2_cptvf_reqmgr.o \
-			otx2_cptvf_algs.o
+rvu_cptpf-objs := otx2_cptpf_main.o otx2_cptpf_mbox.o \
+		  otx2_cpt_mbox_common.o otx2_cptpf_ucode.o otx2_cptlf.o
+rvu_cptvf-objs := otx2_cptvf_main.o otx2_cptvf_mbox.o otx2_cptlf.o \
+		  otx2_cpt_mbox_common.o otx2_cptvf_reqmgr.o \
+		  otx2_cptvf_algs.o
 
 ccflags-y += -I$(srctree)/drivers/net/ethernet/marvell/octeontx2/af
diff --git a/drivers/crypto/marvell/octeontx2/otx2_cpt_common.h b/drivers/crypto/marvell/octeontx2/otx2_cpt_common.h
index ecedd91a8d85..414427dcfa61 100644
--- a/drivers/crypto/marvell/octeontx2/otx2_cpt_common.h
+++ b/drivers/crypto/marvell/octeontx2/otx2_cpt_common.h
@@ -25,6 +25,9 @@
 #define OTX2_CPT_NAME_LENGTH 64
 #define OTX2_CPT_DMA_MINALIGN 128
 
+/* HW capability flags */
+#define CN10K_MBOX  0
+
 #define BAD_OTX2_CPT_ENG_TYPE OTX2_CPT_MAX_ENG_TYPES
 
 enum otx2_cpt_eng_type {
@@ -116,6 +119,23 @@ static inline u64 otx2_cpt_read64(void __iomem *reg_base, u64 blk, u64 slot,
 			     OTX2_CPT_RVU_FUNC_ADDR_S(blk, slot, offs));
 }
 
+static inline bool is_dev_otx2(struct pci_dev *pdev)
+{
+	if (pdev->device == OTX2_CPT_PCI_PF_DEVICE_ID ||
+	    pdev->device == OTX2_CPT_PCI_VF_DEVICE_ID)
+		return true;
+
+	return false;
+}
+
+static inline void otx2_cpt_set_hw_caps(struct pci_dev *pdev,
+					unsigned long *cap_flag)
+{
+	if (!is_dev_otx2(pdev))
+		__set_bit(CN10K_MBOX, cap_flag);
+}
+
+
 int otx2_cpt_send_ready_msg(struct otx2_mbox *mbox, struct pci_dev *pdev);
 int otx2_cpt_send_mbox_msg(struct otx2_mbox *mbox, struct pci_dev *pdev);
 
diff --git a/drivers/crypto/marvell/octeontx2/otx2_cpt_hw_types.h b/drivers/crypto/marvell/octeontx2/otx2_cpt_hw_types.h
index ecafc42f37a2..391a457f7116 100644
--- a/drivers/crypto/marvell/octeontx2/otx2_cpt_hw_types.h
+++ b/drivers/crypto/marvell/octeontx2/otx2_cpt_hw_types.h
@@ -10,6 +10,8 @@
 /* Device IDs */
 #define OTX2_CPT_PCI_PF_DEVICE_ID 0xA0FD
 #define OTX2_CPT_PCI_VF_DEVICE_ID 0xA0FE
+#define CN10K_CPT_PCI_PF_DEVICE_ID 0xA0F2
+#define CN10K_CPT_PCI_VF_DEVICE_ID 0xA0F3
 
 /* Mailbox interrupts offset */
 #define OTX2_CPT_PF_MBOX_INT	6
@@ -25,6 +27,7 @@
  */
 #define OTX2_CPT_VF_MSIX_VECTORS 1
 #define OTX2_CPT_VF_INTR_MBOX_MASK BIT(0)
+#define CN10K_CPT_VF_MBOX_REGION  (0xC0000)
 
 /* CPT LF MSIX vectors */
 #define OTX2_CPT_LF_MSIX_VECTORS 2
diff --git a/drivers/crypto/marvell/octeontx2/otx2_cptpf.h b/drivers/crypto/marvell/octeontx2/otx2_cptpf.h
index e19af1356f12..5ebba86c65d9 100644
--- a/drivers/crypto/marvell/octeontx2/otx2_cptpf.h
+++ b/drivers/crypto/marvell/octeontx2/otx2_cptpf.h
@@ -47,6 +47,7 @@ struct otx2_cptpf_dev {
 	struct workqueue_struct	*flr_wq;
 	struct cptpf_flr_work   *flr_work;
 
+	unsigned long cap_flag;
 	u8 pf_id;               /* RVU PF number */
 	u8 max_vfs;		/* Maximum number of VFs supported by CPT */
 	u8 enabled_vfs;		/* Number of enabled VFs */
diff --git a/drivers/crypto/marvell/octeontx2/otx2_cptpf_main.c b/drivers/crypto/marvell/octeontx2/otx2_cptpf_main.c
index 58f47e3ab62e..d341aecd3dd2 100644
--- a/drivers/crypto/marvell/octeontx2/otx2_cptpf_main.c
+++ b/drivers/crypto/marvell/octeontx2/otx2_cptpf_main.c
@@ -8,8 +8,8 @@
 #include "otx2_cptpf.h"
 #include "rvu_reg.h"
 
-#define OTX2_CPT_DRV_NAME    "octeontx2-cpt"
-#define OTX2_CPT_DRV_STRING  "Marvell OcteonTX2 CPT Physical Function Driver"
+#define OTX2_CPT_DRV_NAME    "rvu_cptpf"
+#define OTX2_CPT_DRV_STRING  "Marvell RVU CPT Physical Function Driver"
 
 static void cptpf_enable_vfpf_mbox_intr(struct otx2_cptpf_dev *cptpf,
 					int num_vfs)
@@ -284,7 +284,11 @@ static int cptpf_vfpf_mbox_init(struct otx2_cptpf_dev *cptpf, int num_vfs)
 		return -ENOMEM;
 
 	/* Map VF-PF mailbox memory */
-	vfpf_mbox_base = readq(cptpf->reg_base + RVU_PF_VF_BAR4_ADDR);
+	if (test_bit(CN10K_MBOX, &cptpf->cap_flag))
+		vfpf_mbox_base = readq(cptpf->reg_base + RVU_PF_VF_MBOX_ADDR);
+	else
+		vfpf_mbox_base = readq(cptpf->reg_base + RVU_PF_VF_BAR4_ADDR);
+
 	if (!vfpf_mbox_base) {
 		dev_err(dev, "VF-PF mailbox address not configured\n");
 		err = -ENOMEM;
@@ -365,6 +369,8 @@ static int cptpf_register_afpf_mbox_intr(struct otx2_cptpf_dev *cptpf)
 
 static int cptpf_afpf_mbox_init(struct otx2_cptpf_dev *cptpf)
 {
+	struct pci_dev *pdev = cptpf->pdev;
+	resource_size_t offset;
 	int err;
 
 	cptpf->afpf_mbox_wq = alloc_workqueue("cpt_afpf_mailbox",
@@ -373,8 +379,17 @@ static int cptpf_afpf_mbox_init(struct otx2_cptpf_dev *cptpf)
 	if (!cptpf->afpf_mbox_wq)
 		return -ENOMEM;
 
+	offset = pci_resource_start(pdev, PCI_MBOX_BAR_NUM);
+	/* Map AF-PF mailbox memory */
+	cptpf->afpf_mbox_base = devm_ioremap_wc(&pdev->dev, offset, MBOX_SIZE);
+	if (!cptpf->afpf_mbox_base) {
+		dev_err(&pdev->dev, "Unable to map BAR4\n");
+		err = -ENOMEM;
+		goto error;
+	}
+
 	err = otx2_mbox_init(&cptpf->afpf_mbox, cptpf->afpf_mbox_base,
-			     cptpf->pdev, cptpf->reg_base, MBOX_DIR_PFAF, 1);
+			     pdev, cptpf->reg_base, MBOX_DIR_PFAF, 1);
 	if (err)
 		goto error;
 
@@ -607,7 +622,6 @@ static int otx2_cptpf_probe(struct pci_dev *pdev,
 			    const struct pci_device_id *ent)
 {
 	struct device *dev = &pdev->dev;
-	resource_size_t offset, size;
 	struct otx2_cptpf_dev *cptpf;
 	int err;
 
@@ -644,15 +658,6 @@ static int otx2_cptpf_probe(struct pci_dev *pdev,
 	if (err)
 		goto clear_drvdata;
 
-	offset = pci_resource_start(pdev, PCI_MBOX_BAR_NUM);
-	size = pci_resource_len(pdev, PCI_MBOX_BAR_NUM);
-	/* Map AF-PF mailbox memory */
-	cptpf->afpf_mbox_base = devm_ioremap_wc(dev, offset, size);
-	if (!cptpf->afpf_mbox_base) {
-		dev_err(&pdev->dev, "Unable to map BAR4\n");
-		err = -ENODEV;
-		goto clear_drvdata;
-	}
 	err = pci_alloc_irq_vectors(pdev, RVU_PF_INT_VEC_CNT,
 				    RVU_PF_INT_VEC_CNT, PCI_IRQ_MSIX);
 	if (err < 0) {
@@ -660,6 +665,7 @@ static int otx2_cptpf_probe(struct pci_dev *pdev,
 			RVU_PF_INT_VEC_CNT);
 		goto clear_drvdata;
 	}
+	otx2_cpt_set_hw_caps(pdev, &cptpf->cap_flag);
 	/* Initialize AF-PF mailbox */
 	err = cptpf_afpf_mbox_init(cptpf);
 	if (err)
@@ -719,6 +725,7 @@ static void otx2_cptpf_remove(struct pci_dev *pdev)
 /* Supported devices */
 static const struct pci_device_id otx2_cpt_id_table[] = {
 	{ PCI_DEVICE(PCI_VENDOR_ID_CAVIUM, OTX2_CPT_PCI_PF_DEVICE_ID) },
+	{ PCI_DEVICE(PCI_VENDOR_ID_CAVIUM, CN10K_CPT_PCI_PF_DEVICE_ID) },
 	{ 0, }  /* end of table */
 };
 
diff --git a/drivers/crypto/marvell/octeontx2/otx2_cptvf.h b/drivers/crypto/marvell/octeontx2/otx2_cptvf.h
index 4f0a169fddbd..4207e2236903 100644
--- a/drivers/crypto/marvell/octeontx2/otx2_cptvf.h
+++ b/drivers/crypto/marvell/octeontx2/otx2_cptvf.h
@@ -19,11 +19,14 @@ struct otx2_cptvf_dev {
 	struct otx2_mbox	pfvf_mbox;
 	struct work_struct	pfvf_mbox_work;
 	struct workqueue_struct *pfvf_mbox_wq;
+	void *bbuf_base;
+	unsigned long cap_flag;
 };
 
 irqreturn_t otx2_cptvf_pfvf_mbox_intr(int irq, void *arg);
 void otx2_cptvf_pfvf_mbox_handler(struct work_struct *work);
 int otx2_cptvf_send_eng_grp_num_msg(struct otx2_cptvf_dev *cptvf, int eng_type);
 int otx2_cptvf_send_kvf_limits_msg(struct otx2_cptvf_dev *cptvf);
+int otx2_cpt_mbox_bbuf_init(struct otx2_cptvf_dev *cptvf, struct pci_dev *pdev);
 
 #endif /* __OTX2_CPTVF_H */
diff --git a/drivers/crypto/marvell/octeontx2/otx2_cptvf_main.c b/drivers/crypto/marvell/octeontx2/otx2_cptvf_main.c
index 47f378731024..5178e0688d75 100644
--- a/drivers/crypto/marvell/octeontx2/otx2_cptvf_main.c
+++ b/drivers/crypto/marvell/octeontx2/otx2_cptvf_main.c
@@ -7,7 +7,7 @@
 #include "otx2_cptvf_algs.h"
 #include <rvu_reg.h>
 
-#define OTX2_CPTVF_DRV_NAME "octeontx2-cptvf"
+#define OTX2_CPTVF_DRV_NAME "rvu_cptvf"
 
 static void cptvf_enable_pfvf_mbox_intrs(struct otx2_cptvf_dev *cptvf)
 {
@@ -70,6 +70,8 @@ static int cptvf_register_interrupts(struct otx2_cptvf_dev *cptvf)
 
 static int cptvf_pfvf_mbox_init(struct otx2_cptvf_dev *cptvf)
 {
+	struct pci_dev *pdev = cptvf->pdev;
+	resource_size_t offset, size;
 	int ret;
 
 	cptvf->pfvf_mbox_wq = alloc_workqueue("cpt_pfvf_mailbox",
@@ -78,14 +80,39 @@ static int cptvf_pfvf_mbox_init(struct otx2_cptvf_dev *cptvf)
 	if (!cptvf->pfvf_mbox_wq)
 		return -ENOMEM;
 
+	if (test_bit(CN10K_MBOX, &cptvf->cap_flag)) {
+		/* For cn10k platform, VF mailbox region is in its BAR2
+		 * register space
+		 */
+		cptvf->pfvf_mbox_base = cptvf->reg_base +
+					CN10K_CPT_VF_MBOX_REGION;
+	} else {
+		offset = pci_resource_start(pdev, PCI_MBOX_BAR_NUM);
+		size = pci_resource_len(pdev, PCI_MBOX_BAR_NUM);
+		/* Map PF-VF mailbox memory */
+		cptvf->pfvf_mbox_base = devm_ioremap_wc(&pdev->dev, offset,
+							size);
+		if (!cptvf->pfvf_mbox_base) {
+			dev_err(&pdev->dev, "Unable to map BAR4\n");
+			ret = -ENOMEM;
+			goto free_wqe;
+		}
+	}
+
 	ret = otx2_mbox_init(&cptvf->pfvf_mbox, cptvf->pfvf_mbox_base,
-			     cptvf->pdev, cptvf->reg_base, MBOX_DIR_VFPF, 1);
+			     pdev, cptvf->reg_base, MBOX_DIR_VFPF, 1);
 	if (ret)
 		goto free_wqe;
 
+	ret = otx2_cpt_mbox_bbuf_init(cptvf, pdev);
+	if (ret)
+		goto destroy_mbox;
+
 	INIT_WORK(&cptvf->pfvf_mbox_work, otx2_cptvf_pfvf_mbox_handler);
 	return 0;
 
+destroy_mbox:
+	otx2_mbox_destroy(&cptvf->pfvf_mbox);
 free_wqe:
 	destroy_workqueue(cptvf->pfvf_mbox_wq);
 	return ret;
@@ -305,7 +332,6 @@ static int otx2_cptvf_probe(struct pci_dev *pdev,
 			    const struct pci_device_id *ent)
 {
 	struct device *dev = &pdev->dev;
-	resource_size_t offset, size;
 	struct otx2_cptvf_dev *cptvf;
 	int ret;
 
@@ -337,15 +363,7 @@ static int otx2_cptvf_probe(struct pci_dev *pdev,
 
 	cptvf->reg_base = pcim_iomap_table(pdev)[PCI_PF_REG_BAR_NUM];
 
-	offset = pci_resource_start(pdev, PCI_MBOX_BAR_NUM);
-	size = pci_resource_len(pdev, PCI_MBOX_BAR_NUM);
-	/* Map PF-VF mailbox memory */
-	cptvf->pfvf_mbox_base = devm_ioremap_wc(dev, offset, size);
-	if (!cptvf->pfvf_mbox_base) {
-		dev_err(&pdev->dev, "Unable to map BAR4\n");
-		ret = -ENODEV;
-		goto clear_drvdata;
-	}
+	otx2_cpt_set_hw_caps(pdev, &cptvf->cap_flag);
 	/* Initialize PF<=>VF mailbox */
 	ret = cptvf_pfvf_mbox_init(cptvf);
 	if (ret)
@@ -392,6 +410,7 @@ static void otx2_cptvf_remove(struct pci_dev *pdev)
 /* Supported devices */
 static const struct pci_device_id otx2_cptvf_id_table[] = {
 	{PCI_VDEVICE(CAVIUM, OTX2_CPT_PCI_VF_DEVICE_ID), 0},
+	{PCI_VDEVICE(CAVIUM, CN10K_CPT_PCI_VF_DEVICE_ID), 0},
 	{ 0, }  /* end of table */
 };
 
@@ -405,6 +424,6 @@ static struct pci_driver otx2_cptvf_pci_driver = {
 module_pci_driver(otx2_cptvf_pci_driver);
 
 MODULE_AUTHOR("Marvell");
-MODULE_DESCRIPTION("Marvell OcteonTX2 CPT Virtual Function Driver");
+MODULE_DESCRIPTION("Marvell RVU CPT Virtual Function Driver");
 MODULE_LICENSE("GPL v2");
 MODULE_DEVICE_TABLE(pci, otx2_cptvf_id_table);
diff --git a/drivers/crypto/marvell/octeontx2/otx2_cptvf_mbox.c b/drivers/crypto/marvell/octeontx2/otx2_cptvf_mbox.c
index 5d73b711cba6..02cb9e44afd8 100644
--- a/drivers/crypto/marvell/octeontx2/otx2_cptvf_mbox.c
+++ b/drivers/crypto/marvell/octeontx2/otx2_cptvf_mbox.c
@@ -5,6 +5,48 @@
 #include "otx2_cptvf.h"
 #include <rvu_reg.h>
 
+int otx2_cpt_mbox_bbuf_init(struct otx2_cptvf_dev *cptvf, struct pci_dev *pdev)
+{
+	struct otx2_mbox_dev *mdev;
+	struct otx2_mbox *otx2_mbox;
+
+	cptvf->bbuf_base = devm_kmalloc(&pdev->dev, MBOX_SIZE, GFP_KERNEL);
+	if (!cptvf->bbuf_base)
+		return -ENOMEM;
+	/*
+	 * Overwrite mbox mbase to point to bounce buffer, so that PF/VF
+	 * prepare all mbox messages in bounce buffer instead of directly
+	 * in hw mbox memory.
+	 */
+	otx2_mbox = &cptvf->pfvf_mbox;
+	mdev = &otx2_mbox->dev[0];
+	mdev->mbase = cptvf->bbuf_base;
+
+	return 0;
+}
+
+static void otx2_cpt_sync_mbox_bbuf(struct otx2_mbox *mbox, int devid)
+{
+	u16 msgs_offset = ALIGN(sizeof(struct mbox_hdr), MBOX_MSG_ALIGN);
+	void *hw_mbase = mbox->hwbase + (devid * MBOX_SIZE);
+	struct otx2_mbox_dev *mdev = &mbox->dev[devid];
+	struct mbox_hdr *hdr;
+	u64 msg_size;
+
+	if (mdev->mbase == hw_mbase)
+		return;
+
+	hdr = hw_mbase + mbox->rx_start;
+	msg_size = hdr->msg_size;
+
+	if (msg_size > mbox->rx_size - msgs_offset)
+		msg_size = mbox->rx_size - msgs_offset;
+
+	/* Copy mbox messages from mbox memory to bounce buffer */
+	memcpy(mdev->mbase + mbox->rx_start,
+	       hw_mbase + mbox->rx_start, msg_size + msgs_offset);
+}
+
 irqreturn_t otx2_cptvf_pfvf_mbox_intr(int __always_unused irq, void *arg)
 {
 	struct otx2_cptvf_dev *cptvf = arg;
@@ -106,6 +148,7 @@ void otx2_cptvf_pfvf_mbox_handler(struct work_struct *work)
 
 	cptvf = container_of(work, struct otx2_cptvf_dev, pfvf_mbox_work);
 	pfvf_mbox = &cptvf->pfvf_mbox;
+	otx2_cpt_sync_mbox_bbuf(pfvf_mbox, 0);
 	mdev = &pfvf_mbox->dev[0];
 	rsp_hdr = (struct mbox_hdr *)(mdev->mbase + pfvf_mbox->rx_start);
 	if (rsp_hdr->num_msgs == 0)
-- 
2.29.0

[PATCH 2/4] crypto: octeontx2: add support to map LMTST region for CN10K

[Srujana Challa]

On CN10K platform transmit/receive buffer alloc and free from/to
hardware had changed to support burst operation. Whereas pervious
silicon's only support single buffer free at a time.
To Support the same firmware allocates a DRAM region for each PF/VF for
storing LMTLINES. These LMTLINES are used to send CPT commands to HW.
PF/VF LMTST region is accessed via BAR4. PFs LMTST region is followed
by its VFs mbox memory. The size of region varies from 2KB to 256KB
based on number of LMTLINES configured.

This patch adds support for mapping of PF/VF LMTST region.

Signed-off-by: Srujana Challa <schalla@marvell.com>
---
 drivers/crypto/marvell/octeontx2/Makefile     |  5 +-
 drivers/crypto/marvell/octeontx2/cn10k_cpt.c  | 53 +++++++++++++++++++
 drivers/crypto/marvell/octeontx2/cn10k_cpt.h  | 13 +++++
 .../marvell/octeontx2/otx2_cpt_common.h       |  5 +-
 drivers/crypto/marvell/octeontx2/otx2_cptlf.h |  2 +
 .../marvell/octeontx2/otx2_cptpf_main.c       |  5 ++
 .../marvell/octeontx2/otx2_cptvf_main.c       |  6 +++
 7 files changed, 86 insertions(+), 3 deletions(-)
 create mode 100644 drivers/crypto/marvell/octeontx2/cn10k_cpt.c
 create mode 100644 drivers/crypto/marvell/octeontx2/cn10k_cpt.h

diff --git a/drivers/crypto/marvell/octeontx2/Makefile b/drivers/crypto/marvell/octeontx2/Makefile
index 10e1fe056a9e..c242d22008c3 100644
--- a/drivers/crypto/marvell/octeontx2/Makefile
+++ b/drivers/crypto/marvell/octeontx2/Makefile
@@ -2,9 +2,10 @@
 obj-$(CONFIG_CRYPTO_DEV_OCTEONTX2_CPT) += rvu_cptpf.o rvu_cptvf.o
 
 rvu_cptpf-objs := otx2_cptpf_main.o otx2_cptpf_mbox.o \
-		  otx2_cpt_mbox_common.o otx2_cptpf_ucode.o otx2_cptlf.o
+		  otx2_cpt_mbox_common.o otx2_cptpf_ucode.o otx2_cptlf.o \
+		  cn10k_cpt.o
 rvu_cptvf-objs := otx2_cptvf_main.o otx2_cptvf_mbox.o otx2_cptlf.o \
 		  otx2_cpt_mbox_common.o otx2_cptvf_reqmgr.o \
-		  otx2_cptvf_algs.o
+		  otx2_cptvf_algs.o cn10k_cpt.o
 
 ccflags-y += -I$(srctree)/drivers/net/ethernet/marvell/octeontx2/af
diff --git a/drivers/crypto/marvell/octeontx2/cn10k_cpt.c b/drivers/crypto/marvell/octeontx2/cn10k_cpt.c
new file mode 100644
index 000000000000..57cf156934ab
--- /dev/null
+++ b/drivers/crypto/marvell/octeontx2/cn10k_cpt.c
@@ -0,0 +1,53 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/* Copyright (C) 2021 Marvell. */
+
+#include "otx2_cptpf.h"
+#include "otx2_cptvf.h"
+#include "otx2_cptlf.h"
+#include "cn10k_cpt.h"
+
+int cn10k_cptpf_lmtst_init(struct otx2_cptpf_dev *cptpf)
+{
+	struct pci_dev *pdev = cptpf->pdev;
+	resource_size_t size;
+	u64 lmt_base;
+
+	if (!test_bit(CN10K_LMTST, &cptpf->cap_flag))
+		return 0;
+
+	lmt_base = readq(cptpf->reg_base + RVU_PF_LMTLINE_ADDR);
+	if (!lmt_base) {
+		dev_err(&pdev->dev, "PF LMTLINE address not configured\n");
+		return -ENOMEM;
+	}
+	size = pci_resource_len(pdev, PCI_MBOX_BAR_NUM);
+	size -= ((1 + cptpf->max_vfs) * MBOX_SIZE);
+	cptpf->lfs.lmt_base = devm_ioremap_wc(&pdev->dev, lmt_base, size);
+	if (!cptpf->lfs.lmt_base) {
+		dev_err(&pdev->dev,
+			"Mapping of PF LMTLINE address failed\n");
+		return -ENOMEM;
+	}
+
+	return 0;
+}
+
+int cn10k_cptvf_lmtst_init(struct otx2_cptvf_dev *cptvf)
+{
+	struct pci_dev *pdev = cptvf->pdev;
+	resource_size_t offset, size;
+
+	if (!test_bit(CN10K_LMTST, &cptvf->cap_flag))
+		return 0;
+
+	offset = pci_resource_start(pdev, PCI_MBOX_BAR_NUM);
+	size = pci_resource_len(pdev, PCI_MBOX_BAR_NUM);
+	/* Map VF LMILINE region */
+	cptvf->lfs.lmt_base = devm_ioremap_wc(&pdev->dev, offset, size);
+	if (!cptvf->lfs.lmt_base) {
+		dev_err(&pdev->dev, "Unable to map BAR4\n");
+		return -ENOMEM;
+	}
+
+	return 0;
+}
diff --git a/drivers/crypto/marvell/octeontx2/cn10k_cpt.h b/drivers/crypto/marvell/octeontx2/cn10k_cpt.h
new file mode 100644
index 000000000000..b9a8c463eaf3
--- /dev/null
+++ b/drivers/crypto/marvell/octeontx2/cn10k_cpt.h
@@ -0,0 +1,13 @@
+/* SPDX-License-Identifier: GPL-2.0-only
+ * Copyright (C) 2021 Marvell.
+ */
+#ifndef __CN10K_CPT_H
+#define __CN10K_CPT_H
+
+#include "otx2_cptpf.h"
+#include "otx2_cptvf.h"
+
+int cn10k_cptpf_lmtst_init(struct otx2_cptpf_dev *cptpf);
+int cn10k_cptvf_lmtst_init(struct otx2_cptvf_dev *cptvf);
+
+#endif /* __CN10K_CPTLF_H */
diff --git a/drivers/crypto/marvell/octeontx2/otx2_cpt_common.h b/drivers/crypto/marvell/octeontx2/otx2_cpt_common.h
index 414427dcfa61..c5445b05f53c 100644
--- a/drivers/crypto/marvell/octeontx2/otx2_cpt_common.h
+++ b/drivers/crypto/marvell/octeontx2/otx2_cpt_common.h
@@ -27,6 +27,7 @@
 
 /* HW capability flags */
 #define CN10K_MBOX  0
+#define CN10K_LMTST 1
 
 #define BAD_OTX2_CPT_ENG_TYPE OTX2_CPT_MAX_ENG_TYPES
 
@@ -131,8 +132,10 @@ static inline bool is_dev_otx2(struct pci_dev *pdev)
 static inline void otx2_cpt_set_hw_caps(struct pci_dev *pdev,
 					unsigned long *cap_flag)
 {
-	if (!is_dev_otx2(pdev))
+	if (!is_dev_otx2(pdev)) {
 		__set_bit(CN10K_MBOX, cap_flag);
+		__set_bit(CN10K_LMTST, cap_flag);
+	}
 }
 
 
diff --git a/drivers/crypto/marvell/octeontx2/otx2_cptlf.h b/drivers/crypto/marvell/octeontx2/otx2_cptlf.h
index ab1678fc564d..c87c18e31171 100644
--- a/drivers/crypto/marvell/octeontx2/otx2_cptlf.h
+++ b/drivers/crypto/marvell/octeontx2/otx2_cptlf.h
@@ -87,6 +87,8 @@ struct otx2_cptlf_info {
 struct otx2_cptlfs_info {
 	/* Registers start address of VF/PF LFs are attached to */
 	void __iomem *reg_base;
+#define LMTLINE_SIZE  128
+	void __iomem *lmt_base;
 	struct pci_dev *pdev;   /* Device LFs are attached to */
 	struct otx2_cptlf_info lf[OTX2_CPT_MAX_LFS_NUM];
 	struct otx2_mbox *mbox;
diff --git a/drivers/crypto/marvell/octeontx2/otx2_cptpf_main.c b/drivers/crypto/marvell/octeontx2/otx2_cptpf_main.c
index d341aecd3dd2..4ec3a4613e74 100644
--- a/drivers/crypto/marvell/octeontx2/otx2_cptpf_main.c
+++ b/drivers/crypto/marvell/octeontx2/otx2_cptpf_main.c
@@ -6,6 +6,7 @@
 #include "otx2_cpt_common.h"
 #include "otx2_cptpf_ucode.h"
 #include "otx2_cptpf.h"
+#include "cn10k_cpt.h"
 #include "rvu_reg.h"
 
 #define OTX2_CPT_DRV_NAME    "rvu_cptpf"
@@ -677,6 +678,10 @@ static int otx2_cptpf_probe(struct pci_dev *pdev,
 
 	cptpf->max_vfs = pci_sriov_get_totalvfs(pdev);
 
+	err = cn10k_cptpf_lmtst_init(cptpf);
+	if (err)
+		goto unregister_intr;
+
 	/* Initialize CPT PF device */
 	err = cptpf_device_init(cptpf);
 	if (err)
diff --git a/drivers/crypto/marvell/octeontx2/otx2_cptvf_main.c b/drivers/crypto/marvell/octeontx2/otx2_cptvf_main.c
index 5178e0688d75..3411e664cf50 100644
--- a/drivers/crypto/marvell/octeontx2/otx2_cptvf_main.c
+++ b/drivers/crypto/marvell/octeontx2/otx2_cptvf_main.c
@@ -5,6 +5,7 @@
 #include "otx2_cptvf.h"
 #include "otx2_cptlf.h"
 #include "otx2_cptvf_algs.h"
+#include "cn10k_cpt.h"
 #include <rvu_reg.h>
 
 #define OTX2_CPTVF_DRV_NAME "rvu_cptvf"
@@ -364,6 +365,11 @@ static int otx2_cptvf_probe(struct pci_dev *pdev,
 	cptvf->reg_base = pcim_iomap_table(pdev)[PCI_PF_REG_BAR_NUM];
 
 	otx2_cpt_set_hw_caps(pdev, &cptvf->cap_flag);
+
+	ret = cn10k_cptvf_lmtst_init(cptvf);
+	if (ret)
+		goto clear_drvdata;
+
 	/* Initialize PF<=>VF mailbox */
 	ret = cptvf_pfvf_mbox_init(cptvf);
 	if (ret)
-- 
2.29.0

[PATCH 3/4] crypto: octeontx2: add support for CPT operations on CN10K

[Srujana Challa]

CPT result format had changed for CN10K HW to accommodate more
fields. This patch adds support to use new result format and
new LMTST lines for CPT operations on CN10K platform.

Signed-off-by: Srujana Challa <schalla@marvell.com>
---
 drivers/crypto/marvell/octeontx2/cn10k_cpt.c  | 44 ++++++++++++++++++-
 drivers/crypto/marvell/octeontx2/cn10k_cpt.h  | 23 ++++++++++
 .../marvell/octeontx2/otx2_cpt_hw_types.h     | 13 +++++-
 drivers/crypto/marvell/octeontx2/otx2_cptlf.c |  9 +++-
 drivers/crypto/marvell/octeontx2/otx2_cptlf.h |  8 ++++
 .../marvell/octeontx2/otx2_cptpf_main.c       |  2 +-
 .../marvell/octeontx2/otx2_cptpf_ucode.c      | 32 +++++++++++---
 .../marvell/octeontx2/otx2_cptpf_ucode.h      |  8 ++--
 .../marvell/octeontx2/otx2_cptvf_reqmgr.c     | 17 ++++---
 9 files changed, 134 insertions(+), 22 deletions(-)

diff --git a/drivers/crypto/marvell/octeontx2/cn10k_cpt.c b/drivers/crypto/marvell/octeontx2/cn10k_cpt.c
index 57cf156934ab..1499ef75b5c2 100644
--- a/drivers/crypto/marvell/octeontx2/cn10k_cpt.c
+++ b/drivers/crypto/marvell/octeontx2/cn10k_cpt.c
@@ -1,20 +1,57 @@
 // SPDX-License-Identifier: GPL-2.0-only
 /* Copyright (C) 2021 Marvell. */
 
+#include <linux/soc/marvell/octeontx2/asm.h>
 #include "otx2_cptpf.h"
 #include "otx2_cptvf.h"
 #include "otx2_cptlf.h"
 #include "cn10k_cpt.h"
 
+static struct cpt_hw_ops otx2_hw_ops = {
+	.send_cmd = otx2_cpt_send_cmd,
+	.cpt_get_compcode = otx2_cpt_get_compcode,
+	.cpt_get_uc_compcode = otx2_cpt_get_uc_compcode,
+};
+
+static struct cpt_hw_ops cn10k_hw_ops = {
+	.send_cmd = cn10k_cpt_send_cmd,
+	.cpt_get_compcode = cn10k_cpt_get_compcode,
+	.cpt_get_uc_compcode = cn10k_cpt_get_uc_compcode,
+};
+
+void cn10k_cpt_send_cmd(union otx2_cpt_inst_s *cptinst, u32 insts_num,
+			struct otx2_cptlf_info *lf)
+{
+	void __iomem *lmtline = lf->lmtline;
+	u64 val = (lf->slot & 0x7FF);
+	u64 tar_addr = 0;
+
+	/* tar_addr<6:4> = Size of first LMTST - 1 in units of 128b. */
+	tar_addr |= (__force u64)lf->ioreg |
+		    (((OTX2_CPT_INST_SIZE/16) - 1) & 0x7) << 4;
+	/*
+	 * Make sure memory areas pointed in CPT_INST_S
+	 * are flushed before the instruction is sent to CPT
+	 */
+	dma_wmb();
+
+	/* Copy CPT command to LMTLINE */
+	memcpy_toio(lmtline, cptinst, insts_num * OTX2_CPT_INST_SIZE);
+	cn10k_lmt_flush(val, tar_addr);
+}
+
 int cn10k_cptpf_lmtst_init(struct otx2_cptpf_dev *cptpf)
 {
 	struct pci_dev *pdev = cptpf->pdev;
 	resource_size_t size;
 	u64 lmt_base;
 
-	if (!test_bit(CN10K_LMTST, &cptpf->cap_flag))
+	if (!test_bit(CN10K_LMTST, &cptpf->cap_flag)) {
+		cptpf->lfs.ops = &otx2_hw_ops;
 		return 0;
+	}
 
+	cptpf->lfs.ops = &cn10k_hw_ops;
 	lmt_base = readq(cptpf->reg_base + RVU_PF_LMTLINE_ADDR);
 	if (!lmt_base) {
 		dev_err(&pdev->dev, "PF LMTLINE address not configured\n");
@@ -37,9 +74,12 @@ int cn10k_cptvf_lmtst_init(struct otx2_cptvf_dev *cptvf)
 	struct pci_dev *pdev = cptvf->pdev;
 	resource_size_t offset, size;
 
-	if (!test_bit(CN10K_LMTST, &cptvf->cap_flag))
+	if (!test_bit(CN10K_LMTST, &cptvf->cap_flag)) {
+		cptvf->lfs.ops = &otx2_hw_ops;
 		return 0;
+	}
 
+	cptvf->lfs.ops = &cn10k_hw_ops;
 	offset = pci_resource_start(pdev, PCI_MBOX_BAR_NUM);
 	size = pci_resource_len(pdev, PCI_MBOX_BAR_NUM);
 	/* Map VF LMILINE region */
diff --git a/drivers/crypto/marvell/octeontx2/cn10k_cpt.h b/drivers/crypto/marvell/octeontx2/cn10k_cpt.h
index b9a8c463eaf3..c091392b47e0 100644
--- a/drivers/crypto/marvell/octeontx2/cn10k_cpt.h
+++ b/drivers/crypto/marvell/octeontx2/cn10k_cpt.h
@@ -4,9 +4,32 @@
 #ifndef __CN10K_CPT_H
 #define __CN10K_CPT_H
 
+#include "otx2_cpt_common.h"
 #include "otx2_cptpf.h"
 #include "otx2_cptvf.h"
 
+static inline u8 cn10k_cpt_get_compcode(union otx2_cpt_res_s *result)
+{
+	return ((struct cn10k_cpt_res_s *)result)->compcode;
+}
+
+static inline u8 cn10k_cpt_get_uc_compcode(union otx2_cpt_res_s *result)
+{
+	return ((struct cn10k_cpt_res_s *)result)->uc_compcode;
+}
+
+static inline u8 otx2_cpt_get_compcode(union otx2_cpt_res_s *result)
+{
+	return ((struct cn9k_cpt_res_s *)result)->compcode;
+}
+
+static inline u8 otx2_cpt_get_uc_compcode(union otx2_cpt_res_s *result)
+{
+	return ((struct cn9k_cpt_res_s *)result)->uc_compcode;
+}
+
+void cn10k_cpt_send_cmd(union otx2_cpt_inst_s *cptinst, u32 insts_num,
+			struct otx2_cptlf_info *lf);
 int cn10k_cptpf_lmtst_init(struct otx2_cptpf_dev *cptpf);
 int cn10k_cptvf_lmtst_init(struct otx2_cptvf_dev *cptvf);
 
diff --git a/drivers/crypto/marvell/octeontx2/otx2_cpt_hw_types.h b/drivers/crypto/marvell/octeontx2/otx2_cpt_hw_types.h
index 391a457f7116..6f947978e4e8 100644
--- a/drivers/crypto/marvell/octeontx2/otx2_cpt_hw_types.h
+++ b/drivers/crypto/marvell/octeontx2/otx2_cpt_hw_types.h
@@ -138,7 +138,7 @@ enum otx2_cpt_comp_e {
 	OTX2_CPT_COMP_E_FAULT = 0x02,
 	OTX2_CPT_COMP_E_HWERR = 0x04,
 	OTX2_CPT_COMP_E_INSTERR = 0x05,
-	OTX2_CPT_COMP_E_LAST_ENTRY = 0x06
+	OTX2_CPT_COMP_E_WARN = 0x06
 };
 
 /*
@@ -269,13 +269,22 @@ union otx2_cpt_inst_s {
 union otx2_cpt_res_s {
 	u64 u[2];
 
-	struct {
+	struct cn9k_cpt_res_s {
 		u64 compcode:8;
 		u64 uc_compcode:8;
 		u64 doneint:1;
 		u64 reserved_17_63:47;
 		u64 reserved_64_127;
 	} s;
+
+	struct cn10k_cpt_res_s {
+		u64 compcode:7;
+		u64 doneint:1;
+		u64 uc_compcode:8;
+		u64 rlen:16;
+		u64 spi:32;
+		u64 esn;
+	} cn10k;
 };
 
 /*
diff --git a/drivers/crypto/marvell/octeontx2/otx2_cptlf.c b/drivers/crypto/marvell/octeontx2/otx2_cptlf.c
index 34aba1532761..c8350fcd60fa 100644
--- a/drivers/crypto/marvell/octeontx2/otx2_cptlf.c
+++ b/drivers/crypto/marvell/octeontx2/otx2_cptlf.c
@@ -379,9 +379,14 @@ int otx2_cptlf_init(struct otx2_cptlfs_info *lfs, u8 eng_grp_mask, int pri,
 	for (slot = 0; slot < lfs->lfs_num; slot++) {
 		lfs->lf[slot].lfs = lfs;
 		lfs->lf[slot].slot = slot;
-		lfs->lf[slot].lmtline = lfs->reg_base +
-			OTX2_CPT_RVU_FUNC_ADDR_S(BLKADDR_LMT, slot,
+		if (lfs->lmt_base)
+			lfs->lf[slot].lmtline = lfs->lmt_base +
+						(slot * LMTLINE_SIZE);
+		else
+			lfs->lf[slot].lmtline = lfs->reg_base +
+				OTX2_CPT_RVU_FUNC_ADDR_S(BLKADDR_LMT, slot,
 						 OTX2_CPT_LMT_LF_LMTLINEX(0));
+
 		lfs->lf[slot].ioreg = lfs->reg_base +
 			OTX2_CPT_RVU_FUNC_ADDR_S(BLKADDR_CPT0, slot,
 						 OTX2_CPT_LF_NQX(0));
diff --git a/drivers/crypto/marvell/octeontx2/otx2_cptlf.h b/drivers/crypto/marvell/octeontx2/otx2_cptlf.h
index c87c18e31171..b691b6c1d5c4 100644
--- a/drivers/crypto/marvell/octeontx2/otx2_cptlf.h
+++ b/drivers/crypto/marvell/octeontx2/otx2_cptlf.h
@@ -84,6 +84,13 @@ struct otx2_cptlf_info {
 	struct otx2_cptlf_wqe *wqe;       /* Tasklet work info */
 };
 
+struct cpt_hw_ops {
+	void (*send_cmd)(union otx2_cpt_inst_s *cptinst, u32 insts_num,
+			 struct otx2_cptlf_info *lf);
+	u8 (*cpt_get_compcode)(union otx2_cpt_res_s *result);
+	u8 (*cpt_get_uc_compcode)(union otx2_cpt_res_s *result);
+};
+
 struct otx2_cptlfs_info {
 	/* Registers start address of VF/PF LFs are attached to */
 	void __iomem *reg_base;
@@ -92,6 +99,7 @@ struct otx2_cptlfs_info {
 	struct pci_dev *pdev;   /* Device LFs are attached to */
 	struct otx2_cptlf_info lf[OTX2_CPT_MAX_LFS_NUM];
 	struct otx2_mbox *mbox;
+	struct cpt_hw_ops *ops;
 	u8 are_lfs_attached;	/* Whether CPT LFs are attached */
 	u8 lfs_num;		/* Number of CPT LFs */
 	u8 kcrypto_eng_grp_num;	/* Kernel crypto engine group number */
diff --git a/drivers/crypto/marvell/octeontx2/otx2_cptpf_main.c b/drivers/crypto/marvell/octeontx2/otx2_cptpf_main.c
index 4ec3a4613e74..1fb04f9bb7ac 100644
--- a/drivers/crypto/marvell/octeontx2/otx2_cptpf_main.c
+++ b/drivers/crypto/marvell/octeontx2/otx2_cptpf_main.c
@@ -586,7 +586,7 @@ static int cptpf_sriov_enable(struct pci_dev *pdev, int num_vfs)
 	if (ret)
 		goto disable_intr;
 
-	ret = otx2_cpt_create_eng_grps(cptpf->pdev, &cptpf->eng_grps);
+	ret = otx2_cpt_create_eng_grps(cptpf, &cptpf->eng_grps);
 	if (ret)
 		goto disable_intr;
 
diff --git a/drivers/crypto/marvell/octeontx2/otx2_cptpf_ucode.c b/drivers/crypto/marvell/octeontx2/otx2_cptpf_ucode.c
index a531f4c8b441..dff34b3ec09e 100644
--- a/drivers/crypto/marvell/octeontx2/otx2_cptpf_ucode.c
+++ b/drivers/crypto/marvell/octeontx2/otx2_cptpf_ucode.c
@@ -16,6 +16,8 @@
 #define LOADFVC_MAJOR_OP 0x01
 #define LOADFVC_MINOR_OP 0x08
 
+#define CTX_FLUSH_TIMER_CNT 0xFFFFFF
+
 struct fw_info_t {
 	struct list_head ucodes;
 };
@@ -666,7 +668,8 @@ static int reserve_engines(struct device *dev,
 static void ucode_unload(struct device *dev, struct otx2_cpt_ucode *ucode)
 {
 	if (ucode->va) {
-		dma_free_coherent(dev, ucode->size, ucode->va, ucode->dma);
+		dma_free_coherent(dev, OTX2_CPT_UCODE_SZ, ucode->va,
+				  ucode->dma);
 		ucode->va = NULL;
 		ucode->dma = 0;
 		ucode->size = 0;
@@ -685,7 +688,7 @@ static int copy_ucode_to_dma_mem(struct device *dev,
 	u32 i;
 
 	/*  Allocate DMAable space */
-	ucode->va = dma_alloc_coherent(dev, ucode->size, &ucode->dma,
+	ucode->va = dma_alloc_coherent(dev, OTX2_CPT_UCODE_SZ, &ucode->dma,
 				       GFP_KERNEL);
 	if (!ucode->va)
 		return -ENOMEM;
@@ -1100,11 +1103,12 @@ int otx2_cpt_get_eng_grp(struct otx2_cpt_eng_grps *eng_grps, int eng_type)
 	return eng_grp_num;
 }
 
-int otx2_cpt_create_eng_grps(struct pci_dev *pdev,
+int otx2_cpt_create_eng_grps(struct otx2_cptpf_dev *cptpf,
 			     struct otx2_cpt_eng_grps *eng_grps)
 {
 	struct otx2_cpt_uc_info_t *uc_info[OTX2_CPT_MAX_ETYPES_PER_GRP] = {  };
 	struct otx2_cpt_engines engs[OTX2_CPT_MAX_ETYPES_PER_GRP] = { {0} };
+	struct pci_dev *pdev = cptpf->pdev;
 	struct fw_info_t fw_info;
 	int ret;
 
@@ -1180,6 +1184,23 @@ int otx2_cpt_create_eng_grps(struct pci_dev *pdev,
 	eng_grps->is_grps_created = true;
 
 	cpt_ucode_release_fw(&fw_info);
+
+	if (is_dev_otx2(pdev))
+		return 0;
+	/*
+	 * Configure engine group mask to allow context prefetching
+	 * for the groups.
+	 */
+	otx2_cpt_write_af_reg(&cptpf->afpf_mbox, pdev, CPT_AF_CTL,
+			      OTX2_CPT_ALL_ENG_GRPS_MASK << 3 | BIT_ULL(16),
+			      BLKADDR_CPT0);
+	/*
+	 * Set interval to periodically flush dirty data for the next
+	 * CTX cache entry. Set the interval count to maximum supported
+	 * value.
+	 */
+	otx2_cpt_write_af_reg(&cptpf->afpf_mbox, pdev, CPT_AF_CTX_FLUSH_TIMER,
+			      CTX_FLUSH_TIMER_CNT, BLKADDR_CPT0);
 	return 0;
 
 delete_eng_grp:
@@ -1460,9 +1481,10 @@ int otx2_cpt_discover_eng_capabilities(struct otx2_cptpf_dev *cptpf)
 		iq_cmd.cptr.s.grp = otx2_cpt_get_eng_grp(&cptpf->eng_grps,
 							 etype);
 		otx2_cpt_fill_inst(&inst, &iq_cmd, rptr_baddr);
-		otx2_cpt_send_cmd(&inst, 1, &cptpf->lfs.lf[0]);
+		lfs->ops->send_cmd(&inst, 1, &cptpf->lfs.lf[0]);
 
-		while (result->s.compcode == OTX2_CPT_COMPLETION_CODE_INIT)
+		while (lfs->ops->cpt_get_compcode(result) ==
+						OTX2_CPT_COMPLETION_CODE_INIT)
 			cpu_relax();
 
 		cptpf->eng_caps[etype].u = be64_to_cpup(rptr);
diff --git a/drivers/crypto/marvell/octeontx2/otx2_cptpf_ucode.h b/drivers/crypto/marvell/octeontx2/otx2_cptpf_ucode.h
index 6b0d432de0af..fe019ab730b2 100644
--- a/drivers/crypto/marvell/octeontx2/otx2_cptpf_ucode.h
+++ b/drivers/crypto/marvell/octeontx2/otx2_cptpf_ucode.h
@@ -23,11 +23,13 @@
 /* Microcode version string length */
 #define OTX2_CPT_UCODE_VER_STR_SZ   44
 
-/* Maximum number of supported engines/cores on OcteonTX2 platform */
-#define OTX2_CPT_MAX_ENGINES        128
+/* Maximum number of supported engines/cores on OcteonTX2/CN10K platform */
+#define OTX2_CPT_MAX_ENGINES        144
 
 #define OTX2_CPT_ENGS_BITMASK_LEN   BITS_TO_LONGS(OTX2_CPT_MAX_ENGINES)
 
+#define OTX2_CPT_UCODE_SZ           (64 * 1024)
+
 /* Microcode types */
 enum otx2_cpt_ucode_type {
 	OTX2_CPT_AE_UC_TYPE = 1,  /* AE-MAIN */
@@ -153,7 +155,7 @@ int otx2_cpt_init_eng_grps(struct pci_dev *pdev,
 			   struct otx2_cpt_eng_grps *eng_grps);
 void otx2_cpt_cleanup_eng_grps(struct pci_dev *pdev,
 			       struct otx2_cpt_eng_grps *eng_grps);
-int otx2_cpt_create_eng_grps(struct pci_dev *pdev,
+int otx2_cpt_create_eng_grps(struct otx2_cptpf_dev *cptpf,
 			     struct otx2_cpt_eng_grps *eng_grps);
 int otx2_cpt_disable_all_cores(struct otx2_cptpf_dev *cptpf);
 int otx2_cpt_get_eng_grp(struct otx2_cpt_eng_grps *eng_grps, int eng_type);
diff --git a/drivers/crypto/marvell/octeontx2/otx2_cptvf_reqmgr.c b/drivers/crypto/marvell/octeontx2/otx2_cptvf_reqmgr.c
index d5c1c1b7c7e4..811ded72ce5f 100644
--- a/drivers/crypto/marvell/octeontx2/otx2_cptvf_reqmgr.c
+++ b/drivers/crypto/marvell/octeontx2/otx2_cptvf_reqmgr.c
@@ -320,7 +320,7 @@ static int process_request(struct pci_dev *pdev, struct otx2_cpt_req_info *req,
 			     cpt_req->dlen, false);
 
 	/* Send CPT command */
-	otx2_cpt_send_cmd(&cptinst, 1, lf);
+	lf->lfs->ops->send_cmd(&cptinst, 1, lf);
 
 	/*
 	 * We allocate and prepare pending queue entry in critical section
@@ -349,13 +349,14 @@ int otx2_cpt_do_request(struct pci_dev *pdev, struct otx2_cpt_req_info *req,
 			       &lfs->lf[cpu_num]);
 }
 
-static int cpt_process_ccode(struct pci_dev *pdev,
+static int cpt_process_ccode(struct otx2_cptlfs_info *lfs,
 			     union otx2_cpt_res_s *cpt_status,
 			     struct otx2_cpt_inst_info *info,
 			     u32 *res_code)
 {
-	u8 uc_ccode = cpt_status->s.uc_compcode;
-	u8 ccode = cpt_status->s.compcode;
+	u8 uc_ccode = lfs->ops->cpt_get_uc_compcode(cpt_status);
+	u8 ccode = lfs->ops->cpt_get_compcode(cpt_status);
+	struct pci_dev *pdev = lfs->pdev;
 
 	switch (ccode) {
 	case OTX2_CPT_COMP_E_FAULT:
@@ -389,6 +390,7 @@ static int cpt_process_ccode(struct pci_dev *pdev,
 		return 1;
 
 	case OTX2_CPT_COMP_E_GOOD:
+	case OTX2_CPT_COMP_E_WARN:
 		/*
 		 * Check microcode completion code, it is only valid
 		 * when completion code is CPT_COMP_E::GOOD
@@ -426,7 +428,7 @@ static int cpt_process_ccode(struct pci_dev *pdev,
 	return 0;
 }
 
-static inline void process_pending_queue(struct pci_dev *pdev,
+static inline void process_pending_queue(struct otx2_cptlfs_info *lfs,
 					 struct otx2_cpt_pending_queue *pqueue)
 {
 	struct otx2_cpt_pending_entry *resume_pentry = NULL;
@@ -436,6 +438,7 @@ static inline void process_pending_queue(struct pci_dev *pdev,
 	struct otx2_cpt_inst_info *info = NULL;
 	struct otx2_cpt_req_info *req = NULL;
 	struct crypto_async_request *areq;
+	struct pci_dev *pdev = lfs->pdev;
 	u32 res_code, resume_index;
 
 	while (1) {
@@ -476,7 +479,7 @@ static inline void process_pending_queue(struct pci_dev *pdev,
 			goto process_pentry;
 		}
 
-		if (cpt_process_ccode(pdev, cpt_status, info, &res_code)) {
+		if (cpt_process_ccode(lfs, cpt_status, info, &res_code)) {
 			spin_unlock_bh(&pqueue->lock);
 			return;
 		}
@@ -529,7 +532,7 @@ static inline void process_pending_queue(struct pci_dev *pdev,
 
 void otx2_cpt_post_process(struct otx2_cptlf_wqe *wqe)
 {
-	process_pending_queue(wqe->lfs->pdev,
+	process_pending_queue(wqe->lfs,
 			      &wqe->lfs->lf[wqe->lf_num].pqueue);
 }
 
-- 
2.29.0

[PATCH 4/4] crypto: octeontx2: enable and handle ME interrupts

[Srujana Challa]

Adds master enable (ME) interrupt handler in PF. Upon
receiving ME interrupt for a VF, PF clears it's transaction
pending bit.

Signed-off-by: Srujana Challa <schalla@marvell.com>
---
 .../marvell/octeontx2/otx2_cptpf_main.c       | 118 ++++++++++++++----
 1 file changed, 95 insertions(+), 23 deletions(-)

diff --git a/drivers/crypto/marvell/octeontx2/otx2_cptpf_main.c b/drivers/crypto/marvell/octeontx2/otx2_cptpf_main.c
index 1fb04f9bb7ac..146a55ac4b9b 100644
--- a/drivers/crypto/marvell/octeontx2/otx2_cptpf_main.c
+++ b/drivers/crypto/marvell/octeontx2/otx2_cptpf_main.c
@@ -63,45 +63,66 @@ static void cptpf_disable_vfpf_mbox_intr(struct otx2_cptpf_dev *cptpf,
 	}
 }
 
-static void cptpf_enable_vf_flr_intrs(struct otx2_cptpf_dev *cptpf)
+static void cptpf_enable_vf_flr_me_intrs(struct otx2_cptpf_dev *cptpf,
+					 int num_vfs)
 {
-	/* Clear interrupt if any */
+	/* Clear FLR interrupt if any */
 	otx2_cpt_write64(cptpf->reg_base, BLKADDR_RVUM, 0, RVU_PF_VFFLR_INTX(0),
-			~0x0ULL);
-	otx2_cpt_write64(cptpf->reg_base, BLKADDR_RVUM, 0, RVU_PF_VFFLR_INTX(1),
-			~0x0ULL);
+			 INTR_MASK(num_vfs));
 
 	/* Enable VF FLR interrupts */
 	otx2_cpt_write64(cptpf->reg_base, BLKADDR_RVUM, 0,
-			 RVU_PF_VFFLR_INT_ENA_W1SX(0), ~0x0ULL);
+			 RVU_PF_VFFLR_INT_ENA_W1SX(0), INTR_MASK(num_vfs));
+	/* Clear ME interrupt if any */
+	otx2_cpt_write64(cptpf->reg_base, BLKADDR_RVUM, 0, RVU_PF_VFME_INTX(0),
+			 INTR_MASK(num_vfs));
+	/* Enable VF ME interrupts */
+	otx2_cpt_write64(cptpf->reg_base, BLKADDR_RVUM, 0,
+			 RVU_PF_VFME_INT_ENA_W1SX(0), INTR_MASK(num_vfs));
+
+	if (num_vfs <= 64)
+		return;
+
+	otx2_cpt_write64(cptpf->reg_base, BLKADDR_RVUM, 0, RVU_PF_VFFLR_INTX(1),
+			 INTR_MASK(num_vfs - 64));
 	otx2_cpt_write64(cptpf->reg_base, BLKADDR_RVUM, 0,
-			 RVU_PF_VFFLR_INT_ENA_W1SX(1), ~0x0ULL);
+			 RVU_PF_VFFLR_INT_ENA_W1SX(1), INTR_MASK(num_vfs - 64));
+
+	otx2_cpt_write64(cptpf->reg_base, BLKADDR_RVUM, 0, RVU_PF_VFME_INTX(1),
+			 INTR_MASK(num_vfs - 64));
+	otx2_cpt_write64(cptpf->reg_base, BLKADDR_RVUM, 0,
+			 RVU_PF_VFME_INT_ENA_W1SX(1), INTR_MASK(num_vfs - 64));
 }
 
-static void cptpf_disable_vf_flr_intrs(struct otx2_cptpf_dev *cptpf,
+static void cptpf_disable_vf_flr_me_intrs(struct otx2_cptpf_dev *cptpf,
 				       int num_vfs)
 {
 	int vector;
 
 	/* Disable VF FLR interrupts */
 	otx2_cpt_write64(cptpf->reg_base, BLKADDR_RVUM, 0,
-			 RVU_PF_VFFLR_INT_ENA_W1CX(0), ~0x0ULL);
+			 RVU_PF_VFFLR_INT_ENA_W1CX(0), INTR_MASK(num_vfs));
+	vector = pci_irq_vector(cptpf->pdev, RVU_PF_INT_VEC_VFFLR0);
+	free_irq(vector, cptpf);
+
+	/* Disable VF ME interrupts */
 	otx2_cpt_write64(cptpf->reg_base, BLKADDR_RVUM, 0,
-			 RVU_PF_VFFLR_INT_ENA_W1CX(1), ~0x0ULL);
+			 RVU_PF_VFME_INT_ENA_W1CX(0), INTR_MASK(num_vfs));
+	vector = pci_irq_vector(cptpf->pdev, RVU_PF_INT_VEC_VFME0);
+	free_irq(vector, cptpf);
 
-	/* Clear interrupt if any */
-	otx2_cpt_write64(cptpf->reg_base, BLKADDR_RVUM, 0, RVU_PF_VFFLR_INTX(0),
-			 ~0x0ULL);
-	otx2_cpt_write64(cptpf->reg_base, BLKADDR_RVUM, 0, RVU_PF_VFFLR_INTX(1),
-			 ~0x0ULL);
+	if (num_vfs <= 64)
+		return;
 
-	vector = pci_irq_vector(cptpf->pdev, RVU_PF_INT_VEC_VFFLR0);
+	otx2_cpt_write64(cptpf->reg_base, BLKADDR_RVUM, 0,
+			 RVU_PF_VFFLR_INT_ENA_W1CX(1), INTR_MASK(num_vfs - 64));
+	vector = pci_irq_vector(cptpf->pdev, RVU_PF_INT_VEC_VFFLR1);
 	free_irq(vector, cptpf);
 
-	if (num_vfs > 64) {
-		vector = pci_irq_vector(cptpf->pdev, RVU_PF_INT_VEC_VFFLR1);
-		free_irq(vector, cptpf);
-	}
+	otx2_cpt_write64(cptpf->reg_base, BLKADDR_RVUM, 0,
+			 RVU_PF_VFME_INT_ENA_W1CX(1), INTR_MASK(num_vfs - 64));
+	vector = pci_irq_vector(cptpf->pdev, RVU_PF_INT_VEC_VFME1);
+	free_irq(vector, cptpf);
 }
 
 static void cptpf_flr_wq_handler(struct work_struct *work)
@@ -173,11 +194,38 @@ static irqreturn_t cptpf_vf_flr_intr(int __always_unused irq, void *arg)
 	return IRQ_HANDLED;
 }
 
+static irqreturn_t cptpf_vf_me_intr(int __always_unused irq, void *arg)
+{
+	struct otx2_cptpf_dev *cptpf = arg;
+	int reg, vf, num_reg = 1;
+	u64 intr;
+
+	if (cptpf->max_vfs > 64)
+		num_reg = 2;
+
+	for (reg = 0; reg < num_reg; reg++) {
+		intr = otx2_cpt_read64(cptpf->reg_base, BLKADDR_RVUM, 0,
+				       RVU_PF_VFME_INTX(reg));
+		if (!intr)
+			continue;
+		for (vf = 0; vf < 64; vf++) {
+			if (!(intr & BIT_ULL(vf)))
+				continue;
+			otx2_cpt_write64(cptpf->reg_base, BLKADDR_RVUM, 0,
+					 RVU_PF_VFTRPENDX(reg), BIT_ULL(vf));
+			/* Clear interrupt */
+			otx2_cpt_write64(cptpf->reg_base, BLKADDR_RVUM, 0,
+					 RVU_PF_VFME_INTX(reg), BIT_ULL(vf));
+		}
+	}
+	return IRQ_HANDLED;
+}
+
 static void cptpf_unregister_vfpf_intr(struct otx2_cptpf_dev *cptpf,
 				       int num_vfs)
 {
 	cptpf_disable_vfpf_mbox_intr(cptpf, num_vfs);
-	cptpf_disable_vf_flr_intrs(cptpf, num_vfs);
+	cptpf_disable_vf_flr_me_intrs(cptpf, num_vfs);
 }
 
 static int cptpf_register_vfpf_intr(struct otx2_cptpf_dev *cptpf, int num_vfs)
@@ -203,6 +251,15 @@ static int cptpf_register_vfpf_intr(struct otx2_cptpf_dev *cptpf, int num_vfs)
 			"IRQ registration failed for VFFLR0 irq\n");
 		goto free_mbox0_irq;
 	}
+	vector = pci_irq_vector(pdev, RVU_PF_INT_VEC_VFME0);
+	/* Register VF ME interrupt handler */
+	ret = request_irq(vector, cptpf_vf_me_intr, 0, "CPTPF ME0", cptpf);
+	if (ret) {
+		dev_err(dev,
+			"IRQ registration failed for PFVF mbox0 irq\n");
+		goto free_flr0_irq;
+	}
+
 	if (num_vfs > 64) {
 		vector = pci_irq_vector(pdev, RVU_PF_INT_VEC_VFPF_MBOX1);
 		ret = request_irq(vector, otx2_cptpf_vfpf_mbox_intr, 0,
@@ -210,7 +267,7 @@ static int cptpf_register_vfpf_intr(struct otx2_cptpf_dev *cptpf, int num_vfs)
 		if (ret) {
 			dev_err(dev,
 				"IRQ registration failed for PFVF mbox1 irq\n");
-			goto free_flr0_irq;
+			goto free_me0_irq;
 		}
 		vector = pci_irq_vector(pdev, RVU_PF_INT_VEC_VFFLR1);
 		/* Register VF FLR interrupt handler */
@@ -221,15 +278,30 @@ static int cptpf_register_vfpf_intr(struct otx2_cptpf_dev *cptpf, int num_vfs)
 				"IRQ registration failed for VFFLR1 irq\n");
 			goto free_mbox1_irq;
 		}
+		vector = pci_irq_vector(pdev, RVU_PF_INT_VEC_VFME1);
+		/* Register VF FLR interrupt handler */
+		ret = request_irq(vector, cptpf_vf_me_intr, 0, "CPTPF ME1",
+				  cptpf);
+		if (ret) {
+			dev_err(dev,
+				"IRQ registration failed for VFFLR1 irq\n");
+			goto free_flr1_irq;
+		}
 	}
 	cptpf_enable_vfpf_mbox_intr(cptpf, num_vfs);
-	cptpf_enable_vf_flr_intrs(cptpf);
+	cptpf_enable_vf_flr_me_intrs(cptpf, num_vfs);
 
 	return 0;
 
+free_flr1_irq:
+	vector = pci_irq_vector(pdev, RVU_PF_INT_VEC_VFFLR1);
+	free_irq(vector, cptpf);
 free_mbox1_irq:
 	vector = pci_irq_vector(pdev, RVU_PF_INT_VEC_VFPF_MBOX1);
 	free_irq(vector, cptpf);
+free_me0_irq:
+	vector = pci_irq_vector(pdev, RVU_PF_INT_VEC_VFME0);
+	free_irq(vector, cptpf);
 free_flr0_irq:
 	vector = pci_irq_vector(pdev, RVU_PF_INT_VEC_VFFLR0);
 	free_irq(vector, cptpf);
-- 
2.29.0

Re: [PATCH 0/4] Add support for Marvell CN10K CPT block

[Herbert Xu]

On Tue, May 25, 2021 at 04:57:14PM +0530, Srujana Challa wrote:
> The current CPT driver supports OcteonTX2 silicon variants.
> The same OcteonTX2 Resource Virtualization Unit(RVU) is
> carried forward to the next-gen silicon ie OcteonTX3(CN10K),
> with some changes and feature enhancements.
> 
> This patch series adds support for CN10K silicon.
> 
> Srujana Challa (4):
>   crypto: octeontx2: Add mailbox support for CN10K
>   crypto: octeontx2: add support to map LMTST region for CN10K
>   crypto: octeontx2: add support for CPT operations on CN10K
>   crypto: octeontx2: enable and handle ME interrupts
> 
>  drivers/crypto/marvell/octeontx2/Makefile     |  13 +-
>  drivers/crypto/marvell/octeontx2/cn10k_cpt.c  |  93 ++++++++++
>  drivers/crypto/marvell/octeontx2/cn10k_cpt.h  |  36 ++++
>  .../marvell/octeontx2/otx2_cpt_common.h       |  23 +++
>  .../marvell/octeontx2/otx2_cpt_hw_types.h     |  16 +-
>  drivers/crypto/marvell/octeontx2/otx2_cptlf.c |   9 +-
>  drivers/crypto/marvell/octeontx2/otx2_cptlf.h |  10 ++
>  drivers/crypto/marvell/octeontx2/otx2_cptpf.h |   1 +
>  .../marvell/octeontx2/otx2_cptpf_main.c       | 160 +++++++++++++-----
>  .../marvell/octeontx2/otx2_cptpf_ucode.c      |  32 +++-
>  .../marvell/octeontx2/otx2_cptpf_ucode.h      |   8 +-
>  drivers/crypto/marvell/octeontx2/otx2_cptvf.h |   3 +
>  .../marvell/octeontx2/otx2_cptvf_main.c       |  49 ++++--
>  .../marvell/octeontx2/otx2_cptvf_mbox.c       |  43 +++++
>  .../marvell/octeontx2/otx2_cptvf_reqmgr.c     |  17 +-
>  15 files changed, 438 insertions(+), 75 deletions(-)
>  create mode 100644 drivers/crypto/marvell/octeontx2/cn10k_cpt.c
>  create mode 100644 drivers/crypto/marvell/octeontx2/cn10k_cpt.h

All applied.  Thanks.
-- 
Email: Herbert Xu <herbert@gondor.apana.org.au>
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt