crypto: tegra - Add Tegra Security Engine driver

Add support for Tegra Security Engine which can accelerate various
crypto algorithms. The Engine has two separate instances within for
AES and HASH algorithms respectively.

The driver registers two crypto engines - one for AES and another for
HASH algorithms and these operate independently and both uses the host1x
bus. Additionally, it provides  hardware-assisted key protection for up
to 15 symmetric keys which it can use for the cipher operations.

Signed-off-by: Akhil R <akhilrajeev@nvidia.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

MAINTAINERS

@@ -21702,6 +21702,11 @@ M:	Prashant Gaikwad <pgaikwad@nvidia.com>
 S:	Supported
 F:	drivers/clk/tegra/
 
+TEGRA CRYPTO DRIVERS
+M:	Akhil R <akhilrajeev@nvidia.com>
+S:	Supported
+F:	drivers/crypto/tegra/*
+
 TEGRA DMA DRIVERS
 M:	Laxman Dewangan <ldewangan@nvidia.com>
 M:	Jon Hunter <jonathanh@nvidia.com>

drivers/crypto/Kconfig

@@ -660,6 +660,14 @@ config CRYPTO_DEV_ROCKCHIP_DEBUG
 	  This will create /sys/kernel/debug/rk3288_crypto/stats for displaying
 	  the number of requests per algorithm and other internal stats.
 
+config CRYPTO_DEV_TEGRA
+	tristate "Enable Tegra Security Engine"
+	depends on TEGRA_HOST1X
+	select CRYPTO_ENGINE
+
+	help
+	  Select this to enable Tegra Security Engine which accelerates various
+	  AES encryption/decryption and HASH algorithms.
 
 config CRYPTO_DEV_ZYNQMP_AES
 	tristate "Support for Xilinx ZynqMP AES hw accelerator"

drivers/crypto/Makefile

@@ -41,6 +41,7 @@ obj-$(CONFIG_CRYPTO_DEV_SAHARA) += sahara.o
 obj-$(CONFIG_CRYPTO_DEV_SL3516) += gemini/
 obj-y += stm32/
 obj-$(CONFIG_CRYPTO_DEV_TALITOS) += talitos.o
+obj-$(CONFIG_CRYPTO_DEV_TEGRA) += tegra/
 obj-$(CONFIG_CRYPTO_DEV_VIRTIO) += virtio/
 #obj-$(CONFIG_CRYPTO_DEV_VMX) += vmx/
 obj-$(CONFIG_CRYPTO_DEV_BCM_SPU) += bcm/

drivers/crypto/tegra/Makefile

@@ -0,0 +1,9 @@
+// SPDX-License-Identifier: GPL-2.0-only
+// SPDX-FileCopyrightText: Copyright (c) 2023 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+
+tegra-se-objs := tegra-se-key.o tegra-se-main.o
+
+tegra-se-y += tegra-se-aes.o
+tegra-se-y += tegra-se-hash.o
+
+obj-$(CONFIG_CRYPTO_DEV_TEGRA) += tegra-se.o

drivers/crypto/tegra/tegra-se-key.c

@@ -0,0 +1,156 @@
+// SPDX-License-Identifier: GPL-2.0-only
+// SPDX-FileCopyrightText: Copyright (c) 2023 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+/*
+ * Crypto driver file to manage keys of NVIDIA Security Engine.
+ */
+
+#include <linux/bitops.h>
+#include <linux/module.h>
+#include <crypto/aes.h>
+
+#include "tegra-se.h"
+
+#define SE_KEY_FULL_MASK		GENMASK(SE_MAX_KEYSLOT, 0)
+
+/* Reserve keyslot 0, 14, 15 */
+#define SE_KEY_RSVD_MASK		(BIT(0) | BIT(14) | BIT(15))
+#define SE_KEY_VALID_MASK		(SE_KEY_FULL_MASK & ~SE_KEY_RSVD_MASK)
+
+/* Mutex lock to guard keyslots */
+static DEFINE_MUTEX(kslt_lock);
+
+/* Keyslot bitmask (0 = available, 1 = in use/not available) */
+static u16 tegra_se_keyslots = SE_KEY_RSVD_MASK;
+
+static u16 tegra_keyslot_alloc(void)
+{
+	u16 keyid;
+
+	mutex_lock(&kslt_lock);
+	/* Check if all key slots are full */
+	if (tegra_se_keyslots == GENMASK(SE_MAX_KEYSLOT, 0)) {
+		mutex_unlock(&kslt_lock);
+		return 0;
+	}
+
+	keyid = ffz(tegra_se_keyslots);
+	tegra_se_keyslots |= BIT(keyid);
+
+	mutex_unlock(&kslt_lock);
+
+	return keyid;
+}
+
+static void tegra_keyslot_free(u16 slot)
+{
+	mutex_lock(&kslt_lock);
+	tegra_se_keyslots &= ~(BIT(slot));
+	mutex_unlock(&kslt_lock);
+}
+
+static unsigned int tegra_key_prep_ins_cmd(struct tegra_se *se, u32 *cpuvaddr,
+					   const u32 *key, u32 keylen, u16 slot, u32 alg)
+{
+	int i = 0, j;
+
+	cpuvaddr[i++] = host1x_opcode_setpayload(1);
+	cpuvaddr[i++] = se_host1x_opcode_incr_w(se->hw->regs->op);
+	cpuvaddr[i++] = SE_AES_OP_WRSTALL | SE_AES_OP_DUMMY;
+
+	cpuvaddr[i++] = host1x_opcode_setpayload(1);
+	cpuvaddr[i++] = se_host1x_opcode_incr_w(se->hw->regs->manifest);
+	cpuvaddr[i++] = se->manifest(se->owner, alg, keylen);
+	cpuvaddr[i++] = host1x_opcode_setpayload(1);
+	cpuvaddr[i++] = se_host1x_opcode_incr_w(se->hw->regs->key_dst);
+
+	cpuvaddr[i++] = SE_AES_KEY_DST_INDEX(slot);
+
+	for (j = 0; j < keylen / 4; j++) {
+		/* Set key address */
+		cpuvaddr[i++] = host1x_opcode_setpayload(1);
+		cpuvaddr[i++] = se_host1x_opcode_incr_w(se->hw->regs->key_addr);
+		cpuvaddr[i++] = j;
+
+		/* Set key data */
+		cpuvaddr[i++] = host1x_opcode_setpayload(1);
+		cpuvaddr[i++] = se_host1x_opcode_incr_w(se->hw->regs->key_data);
+		cpuvaddr[i++] = key[j];
+	}
+
+	cpuvaddr[i++] = host1x_opcode_setpayload(1);
+	cpuvaddr[i++] = se_host1x_opcode_incr_w(se->hw->regs->config);
+	cpuvaddr[i++] = SE_CFG_INS;
+
+	cpuvaddr[i++] = host1x_opcode_setpayload(1);
+	cpuvaddr[i++] = se_host1x_opcode_incr_w(se->hw->regs->op);
+	cpuvaddr[i++] = SE_AES_OP_WRSTALL | SE_AES_OP_START |
+			SE_AES_OP_LASTBUF;
+
+	cpuvaddr[i++] = se_host1x_opcode_nonincr(host1x_uclass_incr_syncpt_r(), 1);
+	cpuvaddr[i++] = host1x_uclass_incr_syncpt_cond_f(1) |
+			host1x_uclass_incr_syncpt_indx_f(se->syncpt_id);
+
+	dev_dbg(se->dev, "key-slot %u key-manifest %#x\n",
+		slot, se->manifest(se->owner, alg, keylen));
+
+	return i;
+}
+
+static bool tegra_key_in_kslt(u32 keyid)
+{
+	bool ret;
+
+	if (keyid > SE_MAX_KEYSLOT)
+		return false;
+
+	mutex_lock(&kslt_lock);
+	ret = ((BIT(keyid) & SE_KEY_VALID_MASK) &&
+		(BIT(keyid) & tegra_se_keyslots));
+	mutex_unlock(&kslt_lock);
+
+	return ret;
+}
+
+static int tegra_key_insert(struct tegra_se *se, const u8 *key,
+			    u32 keylen, u16 slot, u32 alg)
+{
+	const u32 *keyval = (u32 *)key;
+	u32 *addr = se->cmdbuf->addr, size;
+
+	size = tegra_key_prep_ins_cmd(se, addr, keyval, keylen, slot, alg);
+
+	return tegra_se_host1x_submit(se, size);
+}
+
+void tegra_key_invalidate(struct tegra_se *se, u32 keyid, u32 alg)
+{
+	u8 zkey[AES_MAX_KEY_SIZE] = {0};
+
+	if (!keyid)
+		return;
+
+	/* Overwrite the key with 0s */
+	tegra_key_insert(se, zkey, AES_MAX_KEY_SIZE, keyid, alg);
+
+	tegra_keyslot_free(keyid);
+}
+
+int tegra_key_submit(struct tegra_se *se, const u8 *key, u32 keylen, u32 alg, u32 *keyid)
+{
+	int ret;
+
+	/* Use the existing slot if it is already allocated */
+	if (!tegra_key_in_kslt(*keyid)) {
+		*keyid = tegra_keyslot_alloc();
+		if (!(*keyid)) {
+			dev_err(se->dev, "failed to allocate key slot\n");
+			return -ENOMEM;
+		}
+	}
+
+	ret = tegra_key_insert(se, key, keylen, *keyid, alg);
+	if (ret)
+		return ret;
+
+	return 0;
+}

drivers/crypto/tegra/tegra-se-main.c

@@ -0,0 +1,439 @@
+// SPDX-License-Identifier: GPL-2.0-only
+// SPDX-FileCopyrightText: Copyright (c) 2023 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+/*
+ * Crypto driver for NVIDIA Security Engine in Tegra Chips
+ */
+
+#include <linux/clk.h>
+#include <linux/dma-mapping.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/mod_devicetable.h>
+
+#include <crypto/engine.h>
+
+#include "tegra-se.h"
+
+static struct host1x_bo *tegra_se_cmdbuf_get(struct host1x_bo *host_bo)
+{
+	struct tegra_se_cmdbuf *cmdbuf = container_of(host_bo, struct tegra_se_cmdbuf, bo);
+
+	kref_get(&cmdbuf->ref);
+
+	return host_bo;
+}
+
+static void tegra_se_cmdbuf_release(struct kref *ref)
+{
+	struct tegra_se_cmdbuf *cmdbuf = container_of(ref, struct tegra_se_cmdbuf, ref);
+
+	dma_free_attrs(cmdbuf->dev, cmdbuf->size, cmdbuf->addr,
+		       cmdbuf->iova, 0);
+
+	kfree(cmdbuf);
+}
+
+static void tegra_se_cmdbuf_put(struct host1x_bo *host_bo)
+{
+	struct tegra_se_cmdbuf *cmdbuf = container_of(host_bo, struct tegra_se_cmdbuf, bo);
+
+	kref_put(&cmdbuf->ref, tegra_se_cmdbuf_release);
+}
+
+static struct host1x_bo_mapping *
+tegra_se_cmdbuf_pin(struct device *dev, struct host1x_bo *bo, enum dma_data_direction direction)
+{
+	struct tegra_se_cmdbuf *cmdbuf = container_of(bo, struct tegra_se_cmdbuf, bo);
+	struct host1x_bo_mapping *map;
+	int err;
+
+	map = kzalloc(sizeof(*map), GFP_KERNEL);
+	if (!map)
+		return ERR_PTR(-ENOMEM);
+
+	kref_init(&map->ref);
+	map->bo = host1x_bo_get(bo);
+	map->direction = direction;
+	map->dev = dev;
+
+	map->sgt = kzalloc(sizeof(*map->sgt), GFP_KERNEL);
+	if (!map->sgt) {
+		err = -ENOMEM;
+		goto free;
+	}
+
+	err = dma_get_sgtable(dev, map->sgt, cmdbuf->addr,
+			      cmdbuf->iova, cmdbuf->words * 4);
+	if (err)
+		goto free_sgt;
+
+	err = dma_map_sgtable(dev, map->sgt, direction, 0);
+	if (err)
+		goto free_sgt;
+
+	map->phys = sg_dma_address(map->sgt->sgl);
+	map->size = cmdbuf->words * 4;
+	map->chunks = err;
+
+	return map;
+
+free_sgt:
+	sg_free_table(map->sgt);
+	kfree(map->sgt);
+free:
+	kfree(map);
+	return ERR_PTR(err);
+}
+
+static void tegra_se_cmdbuf_unpin(struct host1x_bo_mapping *map)
+{
+	if (!map)
+		return;
+
+	dma_unmap_sgtable(map->dev, map->sgt, map->direction, 0);
+	sg_free_table(map->sgt);
+	kfree(map->sgt);
+	host1x_bo_put(map->bo);
+
+	kfree(map);
+}
+
+static void *tegra_se_cmdbuf_mmap(struct host1x_bo *host_bo)
+{
+	struct tegra_se_cmdbuf *cmdbuf = container_of(host_bo, struct tegra_se_cmdbuf, bo);
+
+	return cmdbuf->addr;
+}
+
+static void tegra_se_cmdbuf_munmap(struct host1x_bo *host_bo, void *addr)
+{
+}
+
+static const struct host1x_bo_ops tegra_se_cmdbuf_ops = {
+	.get = tegra_se_cmdbuf_get,
+	.put = tegra_se_cmdbuf_put,
+	.pin = tegra_se_cmdbuf_pin,
+	.unpin = tegra_se_cmdbuf_unpin,
+	.mmap = tegra_se_cmdbuf_mmap,
+	.munmap = tegra_se_cmdbuf_munmap,
+};
+
+static struct tegra_se_cmdbuf *tegra_se_host1x_bo_alloc(struct tegra_se *se, ssize_t size)
+{
+	struct tegra_se_cmdbuf *cmdbuf;
+	struct device *dev = se->dev->parent;
+
+	cmdbuf = kzalloc(sizeof(*cmdbuf), GFP_KERNEL);
+	if (!cmdbuf)
+		return NULL;
+
+	cmdbuf->addr = dma_alloc_attrs(dev, size, &cmdbuf->iova,
+				       GFP_KERNEL, 0);
+	if (!cmdbuf->addr)
+		return NULL;
+
+	cmdbuf->size = size;
+	cmdbuf->dev  = dev;
+
+	host1x_bo_init(&cmdbuf->bo, &tegra_se_cmdbuf_ops);
+	kref_init(&cmdbuf->ref);
+
+	return cmdbuf;
+}
+
+int tegra_se_host1x_submit(struct tegra_se *se, u32 size)
+{
+	struct host1x_job *job;
+	int ret;
+
+	job = host1x_job_alloc(se->channel, 1, 0, true);
+	if (!job) {
+		dev_err(se->dev, "failed to allocate host1x job\n");
+		return -ENOMEM;
+	}
+
+	job->syncpt = host1x_syncpt_get(se->syncpt);
+	job->syncpt_incrs = 1;
+	job->client = &se->client;
+	job->class = se->client.class;
+	job->serialize = true;
+	job->engine_fallback_streamid = se->stream_id;
+	job->engine_streamid_offset = SE_STREAM_ID;
+
+	se->cmdbuf->words = size;
+
+	host1x_job_add_gather(job, &se->cmdbuf->bo, size, 0);
+
+	ret = host1x_job_pin(job, se->dev);
+	if (ret) {
+		dev_err(se->dev, "failed to pin host1x job\n");
+		goto job_put;
+	}
+
+	ret = host1x_job_submit(job);
+	if (ret) {
+		dev_err(se->dev, "failed to submit host1x job\n");
+		goto job_unpin;
+	}
+
+	ret = host1x_syncpt_wait(job->syncpt, job->syncpt_end,
+				 MAX_SCHEDULE_TIMEOUT, NULL);
+	if (ret) {
+		dev_err(se->dev, "host1x job timed out\n");
+		return ret;
+	}
+
+	host1x_job_put(job);
+	return 0;
+
+job_unpin:
+	host1x_job_unpin(job);
+job_put:
+	host1x_job_put(job);
+
+	return ret;
+}
+
+static int tegra_se_client_init(struct host1x_client *client)
+{
+	struct tegra_se *se = container_of(client, struct tegra_se, client);
+	int ret;
+
+	se->channel = host1x_channel_request(&se->client);
+	if (!se->channel) {
+		dev_err(se->dev, "host1x channel map failed\n");
+		return -ENODEV;
+	}
+
+	se->syncpt = host1x_syncpt_request(&se->client, 0);
+	if (!se->syncpt) {
+		dev_err(se->dev, "host1x syncpt allocation failed\n");
+		ret = -EINVAL;
+		goto channel_put;
+	}
+
+	se->syncpt_id =  host1x_syncpt_id(se->syncpt);
+
+	se->cmdbuf = tegra_se_host1x_bo_alloc(se, SZ_4K);
+	if (!se->cmdbuf) {
+		ret = -ENOMEM;
+		goto syncpt_put;
+	}
+
+	ret = se->hw->init_alg(se);
+	if (ret) {
+		dev_err(se->dev, "failed to register algorithms\n");
+		goto cmdbuf_put;
+	}
+
+	return 0;
+
+cmdbuf_put:
+	tegra_se_cmdbuf_put(&se->cmdbuf->bo);
+syncpt_put:
+	host1x_syncpt_put(se->syncpt);
+channel_put:
+	host1x_channel_put(se->channel);
+
+	return ret;
+}
+
+static int tegra_se_client_deinit(struct host1x_client *client)
+{
+	struct tegra_se *se = container_of(client, struct tegra_se, client);
+
+	se->hw->deinit_alg(se);
+	tegra_se_cmdbuf_put(&se->cmdbuf->bo);
+	host1x_syncpt_put(se->syncpt);
+	host1x_channel_put(se->channel);
+
+	return 0;
+}
+
+static const struct host1x_client_ops tegra_se_client_ops = {
+	.init = tegra_se_client_init,
+	.exit = tegra_se_client_deinit,
+};
+
+static int tegra_se_host1x_register(struct tegra_se *se)
+{
+	INIT_LIST_HEAD(&se->client.list);
+	se->client.dev = se->dev;
+	se->client.ops = &tegra_se_client_ops;
+	se->client.class = se->hw->host1x_class;
+	se->client.num_syncpts = 1;
+
+	host1x_client_register(&se->client);
+
+	return 0;
+}
+
+static int tegra_se_probe(struct platform_device *pdev)
+{
+	struct device *dev = &pdev->dev;
+	struct tegra_se *se;
+	int ret;
+
+	se = devm_kzalloc(dev, sizeof(*se), GFP_KERNEL);
+	if (!se)
+		return -ENOMEM;
+
+	se->dev = dev;
+	se->owner = TEGRA_GPSE_ID;
+	se->hw = device_get_match_data(&pdev->dev);
+
+	se->base = devm_platform_ioremap_resource(pdev, 0);
+	if (IS_ERR(se->base))
+		return PTR_ERR(se->base);
+
+	dma_set_mask_and_coherent(dev, DMA_BIT_MASK(39));
+	platform_set_drvdata(pdev, se);
+
+	se->clk = devm_clk_get_enabled(se->dev, NULL);
+	if (IS_ERR(se->clk))
+		return dev_err_probe(dev, PTR_ERR(se->clk),
+				"failed to enable clocks\n");
+
+	if (!tegra_dev_iommu_get_stream_id(dev, &se->stream_id))
+		return dev_err_probe(dev, -ENODEV,
+				"failed to get IOMMU stream ID\n");
+
+	writel(se->stream_id, se->base + SE_STREAM_ID);
+
+	se->engine = crypto_engine_alloc_init(dev, 0);
+	if (!se->engine)
+		return dev_err_probe(dev, -ENOMEM, "failed to init crypto engine\n");
+
+	ret = crypto_engine_start(se->engine);
+	if (ret) {
+		crypto_engine_exit(se->engine);
+		return dev_err_probe(dev, ret, "failed to start crypto engine\n");
+	}
+
+	ret = tegra_se_host1x_register(se);
+	if (ret) {
+		crypto_engine_stop(se->engine);
+		crypto_engine_exit(se->engine);
+		return dev_err_probe(dev, ret, "failed to init host1x params\n");
+	}
+
+	return 0;
+}
+
+static int tegra_se_remove(struct platform_device *pdev)
+{
+	struct tegra_se *se = platform_get_drvdata(pdev);
+
+	crypto_engine_stop(se->engine);
+	crypto_engine_exit(se->engine);
+	iommu_fwspec_free(se->dev);
+	host1x_client_unregister(&se->client);
+
+	return 0;
+}
+
+static const struct tegra_se_regs tegra234_aes1_regs = {
+	.config = SE_AES1_CFG,
+	.op = SE_AES1_OPERATION,
+	.last_blk = SE_AES1_LAST_BLOCK,
+	.linear_ctr = SE_AES1_LINEAR_CTR,
+	.aad_len = SE_AES1_AAD_LEN,
+	.cryp_msg_len = SE_AES1_CRYPTO_MSG_LEN,
+	.manifest = SE_AES1_KEYMANIFEST,
+	.key_addr = SE_AES1_KEY_ADDR,
+	.key_data = SE_AES1_KEY_DATA,
+	.key_dst = SE_AES1_KEY_DST,
+	.result = SE_AES1_CMAC_RESULT,
+};
+
+static const struct tegra_se_regs tegra234_hash_regs = {
+	.config = SE_SHA_CFG,
+	.op = SE_SHA_OPERATION,
+	.manifest = SE_SHA_KEYMANIFEST,
+	.key_addr = SE_SHA_KEY_ADDR,
+	.key_data = SE_SHA_KEY_DATA,
+	.key_dst = SE_SHA_KEY_DST,
+	.result = SE_SHA_HASH_RESULT,
+};
+
+static const struct tegra_se_hw tegra234_aes_hw = {
+	.regs = &tegra234_aes1_regs,
+	.kac_ver = 1,
+	.host1x_class = 0x3b,
+	.init_alg = tegra_init_aes,
+	.deinit_alg = tegra_deinit_aes,
+};
+
+static const struct tegra_se_hw tegra234_hash_hw = {
+	.regs = &tegra234_hash_regs,
+	.kac_ver = 1,
+	.host1x_class = 0x3d,
+	.init_alg = tegra_init_hash,
+	.deinit_alg = tegra_deinit_hash,
+};
+
+static const struct of_device_id tegra_se_of_match[] = {
+	{
+		.compatible = "nvidia,tegra234-se-aes",
+		.data = &tegra234_aes_hw
+	}, {
+		.compatible = "nvidia,tegra234-se-hash",
+		.data = &tegra234_hash_hw,
+	},
+	{ },
+};
+MODULE_DEVICE_TABLE(of, tegra_se_of_match);
+
+static struct platform_driver tegra_se_driver = {
+	.driver = {
+		.name	= "tegra-se",
+		.of_match_table = tegra_se_of_match,
+	},
+	.probe		= tegra_se_probe,
+	.remove		= tegra_se_remove,
+};
+
+static int tegra_se_host1x_probe(struct host1x_device *dev)
+{
+	return host1x_device_init(dev);
+}
+
+static int tegra_se_host1x_remove(struct host1x_device *dev)
+{
+	host1x_device_exit(dev);
+
+	return 0;
+}
+
+static struct host1x_driver tegra_se_host1x_driver = {
+	.driver = {
+		.name = "tegra-se-host1x",
+	},
+	.probe = tegra_se_host1x_probe,
+	.remove = tegra_se_host1x_remove,
+	.subdevs = tegra_se_of_match,
+};
+
+static int __init tegra_se_module_init(void)
+{
+	int ret;
+
+	ret = host1x_driver_register(&tegra_se_host1x_driver);
+	if (ret)
+		return ret;
+
+	return platform_driver_register(&tegra_se_driver);
+}
+
+static void __exit tegra_se_module_exit(void)
+{
+	host1x_driver_unregister(&tegra_se_host1x_driver);
+	platform_driver_unregister(&tegra_se_driver);
+}
+
+module_init(tegra_se_module_init);
+module_exit(tegra_se_module_exit);
+
+MODULE_DESCRIPTION("NVIDIA Tegra Security Engine Driver");
+MODULE_AUTHOR("Akhil R <akhilrajeev@nvidia.com>");
+MODULE_LICENSE("GPL");