optee: use driver internal tee_context for some rpc

Adds a driver private tee_context by moving the tee_context in struct
optee_notif to struct optee. This tee_context was previously used when
doing internal calls to secure world to deliver notification.

The new driver internal tee_context is now also when allocating driver
private shared memory. This decouples the shared memory object from its
original tee_context. This is needed when the life time of such a memory
allocation outlives the client tee_context.

This patch fixes the problem described below:

The addition of a shutdown hook by commit f25889f93184 ("optee: fix tee out
of memory failure seen during kexec reboot") introduced a kernel shutdown
regression that can be triggered after running the OP-TEE xtest suites.

Once the shutdown hook is called it is not possible to communicate any more
with the supplicant process because the system is not scheduling task any
longer. Thus if the optee driver shutdown path receives a supplicant RPC
request from the OP-TEE we will deadlock the kernel's shutdown.

Fixes: f25889f93184 ("optee: fix tee out of memory failure seen during kexec reboot")
Fixes: 217e0250cccb ("tee: use reference counting for tee_context")
Reported-by: Lars Persson <larper@axis.com>
Cc: stable@vger.kernel.org
Reviewed-by: Sumit Garg <sumit.garg@linaro.org>
Signed-off-by: Jens Wiklander <jens.wiklander@linaro.org>

4 files changed, 64 insertions, 67 deletions

diff --git a/drivers/tee/optee/core.c b/drivers/tee/optee/core.c
index 1ca320885fad..17a6f51d3089 100644
--- a/drivers/tee/optee/core.c
+++ b/drivers/tee/optee/core.c
@@ -158,6 +158,7 @@ void optee_remove_common(struct optee *optee)
 	optee_unregister_devices();
 
 	optee_notif_uninit(optee);
+	teedev_close_context(optee->ctx);
 	/*
 	 * The two devices have to be unregistered before we can free the
 	 * other resources.
diff --git a/drivers/tee/optee/ffa_abi.c b/drivers/tee/optee/ffa_abi.c
index 20a1b1a3d965..545f61af1248 100644
--- a/drivers/tee/optee/ffa_abi.c
+++ b/drivers/tee/optee/ffa_abi.c
@@ -424,6 +424,7 @@ static struct tee_shm_pool_mgr *optee_ffa_shm_pool_alloc_pages(void)
  */
 
 static void handle_ffa_rpc_func_cmd_shm_alloc(struct tee_context *ctx,
+					      struct optee *optee,
 					      struct optee_msg_arg *arg)
 {
 	struct tee_shm *shm;
@@ -439,7 +440,7 @@ static void handle_ffa_rpc_func_cmd_shm_alloc(struct tee_context *ctx,
 		shm = optee_rpc_cmd_alloc_suppl(ctx, arg->params[0].u.value.b);
 		break;
 	case OPTEE_RPC_SHM_TYPE_KERNEL:
-		shm = tee_shm_alloc(ctx, arg->params[0].u.value.b,
+		shm = tee_shm_alloc(optee->ctx, arg->params[0].u.value.b,
 				    TEE_SHM_MAPPED | TEE_SHM_PRIV);
 		break;
 	default:
@@ -493,14 +494,13 @@ err_bad_param:
 }
 
 static void handle_ffa_rpc_func_cmd(struct tee_context *ctx,
+				    struct optee *optee,
 				    struct optee_msg_arg *arg)
 {
-	struct optee *optee = tee_get_drvdata(ctx->teedev);
-
 	arg->ret_origin = TEEC_ORIGIN_COMMS;
 	switch (arg->cmd) {
 	case OPTEE_RPC_CMD_SHM_ALLOC:
-		handle_ffa_rpc_func_cmd_shm_alloc(ctx, arg);
+		handle_ffa_rpc_func_cmd_shm_alloc(ctx, optee, arg);
 		break;
 	case OPTEE_RPC_CMD_SHM_FREE:
 		handle_ffa_rpc_func_cmd_shm_free(ctx, optee, arg);
@@ -510,12 +510,12 @@ static void handle_ffa_rpc_func_cmd(struct tee_context *ctx,
 	}
 }
 
-static void optee_handle_ffa_rpc(struct tee_context *ctx, u32 cmd,
-				 struct optee_msg_arg *arg)
+static void optee_handle_ffa_rpc(struct tee_context *ctx, struct optee *optee,
+				 u32 cmd, struct optee_msg_arg *arg)
 {
 	switch (cmd) {
 	case OPTEE_FFA_YIELDING_CALL_RETURN_RPC_CMD:
-		handle_ffa_rpc_func_cmd(ctx, arg);
+		handle_ffa_rpc_func_cmd(ctx, optee, arg);
 		break;
 	case OPTEE_FFA_YIELDING_CALL_RETURN_INTERRUPT:
 		/* Interrupt delivered by now */
@@ -582,7 +582,7 @@ static int optee_ffa_yielding_call(struct tee_context *ctx,
 		 * above.
 		 */
 		cond_resched();
-		optee_handle_ffa_rpc(ctx, data->data1, rpc_arg);
+		optee_handle_ffa_rpc(ctx, optee, data->data1, rpc_arg);
 		cmd = OPTEE_FFA_YIELDING_CALL_RESUME;
 		data->data0 = cmd;
 		data->data1 = 0;
@@ -793,7 +793,9 @@ static int optee_ffa_probe(struct ffa_device *ffa_dev)
 {
 	const struct ffa_dev_ops *ffa_ops;
 	unsigned int rpc_arg_count;
+	struct tee_shm_pool *pool;
 	struct tee_device *teedev;
+	struct tee_context *ctx;
 	struct optee *optee;
 	int rc;
 
@@ -813,12 +815,12 @@ static int optee_ffa_probe(struct ffa_device *ffa_dev)
 	if (!optee)
 		return -ENOMEM;
 
-	optee->pool = optee_ffa_config_dyn_shm();
-	if (IS_ERR(optee->pool)) {
-		rc = PTR_ERR(optee->pool);
-		optee->pool = NULL;
-		goto err;
+	pool = optee_ffa_config_dyn_shm();
+	if (IS_ERR(pool)) {
+		rc = PTR_ERR(pool);
+		goto err_free_optee;
 	}
+	optee->pool = pool;
 
 	optee->ops = &optee_ffa_ops;
 	optee->ffa.ffa_dev = ffa_dev;
@@ -829,7 +831,7 @@ static int optee_ffa_probe(struct ffa_device *ffa_dev)
 				  optee);
 	if (IS_ERR(teedev)) {
 		rc = PTR_ERR(teedev);
-		goto err;
+		goto err_free_pool;
 	}
 	optee->teedev = teedev;
 
@@ -837,50 +839,57 @@ static int optee_ffa_probe(struct ffa_device *ffa_dev)
 				  optee);
 	if (IS_ERR(teedev)) {
 		rc = PTR_ERR(teedev);
-		goto err;
+		goto err_unreg_teedev;
 	}
 	optee->supp_teedev = teedev;
 
 	rc = tee_device_register(optee->teedev);
 	if (rc)
-		goto err;
+		goto err_unreg_supp_teedev;
 
 	rc = tee_device_register(optee->supp_teedev);
 	if (rc)
-		goto err;
+		goto err_unreg_supp_teedev;
 
 	rc = rhashtable_init(&optee->ffa.global_ids, &shm_rhash_params);
 	if (rc)
-		goto err;
+		goto err_unreg_supp_teedev;
 	mutex_init(&optee->ffa.mutex);
 	mutex_init(&optee->call_queue.mutex);
 	INIT_LIST_HEAD(&optee->call_queue.waiters);
 	optee_supp_init(&optee->supp);
 	ffa_dev_set_drvdata(ffa_dev, optee);
+	ctx = teedev_open(optee->teedev);
+	if (IS_ERR(ctx))
+		goto err_rhashtable_free;
+	optee->ctx = ctx;
 	rc = optee_notif_init(optee, OPTEE_DEFAULT_MAX_NOTIF_VALUE);
-	if (rc) {
-		optee_ffa_remove(ffa_dev);
-		return rc;
-	}
+	if (rc)
+		goto err_close_ctx;
 
 	rc = optee_enumerate_devices(PTA_CMD_GET_DEVICES);
-	if (rc) {
-		optee_ffa_remove(ffa_dev);
-		return rc;
-	}
+	if (rc)
+		goto err_unregister_devices;
 
 	pr_info("initialized driver\n");
 	return 0;
-err:
-	/*
-	 * tee_device_unregister() is safe to call even if the
-	 * devices hasn't been registered with
-	 * tee_device_register() yet.
-	 */
+
+err_unregister_devices:
+	optee_unregister_devices();
+	optee_notif_uninit(optee);
+err_close_ctx:
+	teedev_close_context(ctx);
+err_rhashtable_free:
+	rhashtable_free_and_destroy(&optee->ffa.global_ids, rh_free_fn, NULL);
+	optee_supp_uninit(&optee->supp);
+	mutex_destroy(&optee->call_queue.mutex);
+err_unreg_supp_teedev:
 	tee_device_unregister(optee->supp_teedev);
+err_unreg_teedev:
 	tee_device_unregister(optee->teedev);
-	if (optee->pool)
-		tee_shm_pool_free(optee->pool);
+err_free_pool:
+	tee_shm_pool_free(pool);
+err_free_optee:
 	kfree(optee);
 	return rc;
 }
diff --git a/drivers/tee/optee/optee_private.h b/drivers/tee/optee/optee_private.h
index 46f74ab07c7e..92bc47bef95f 100644
--- a/drivers/tee/optee/optee_private.h
+++ b/drivers/tee/optee/optee_private.h
@@ -53,7 +53,6 @@ struct optee_call_queue {
 
 struct optee_notif {
 	u_int max_key;
-	struct tee_context *ctx;
 	/* Serializes access to the elements below in this struct */
 	spinlock_t lock;
 	struct list_head db;
@@ -134,9 +133,10 @@ struct optee_ops {
 /**
  * struct optee - main service struct
  * @supp_teedev:	supplicant device
+ * @teedev:		client device
  * @ops:		internal callbacks for different ways to reach secure
  *			world
- * @teedev:		client device
+ * @ctx:		driver internal TEE context
  * @smc:		specific to SMC ABI
  * @ffa:		specific to FF-A ABI
  * @call_queue:		queue of threads waiting to call @invoke_fn
@@ -152,6 +152,7 @@ struct optee {
 	struct tee_device *supp_teedev;
 	struct tee_device *teedev;
 	const struct optee_ops *ops;
+	struct tee_context *ctx;
 	union {
 		struct optee_smc smc;
 		struct optee_ffa ffa;
diff --git a/drivers/tee/optee/smc_abi.c b/drivers/tee/optee/smc_abi.c
index 449d6a72d289..bacd1a1d79ee 100644
--- a/drivers/tee/optee/smc_abi.c
+++ b/drivers/tee/optee/smc_abi.c
@@ -622,6 +622,7 @@ static void handle_rpc_func_cmd_shm_free(struct tee_context *ctx,
 }
 
 static void handle_rpc_func_cmd_shm_alloc(struct tee_context *ctx,
+					  struct optee *optee,
 					  struct optee_msg_arg *arg,
 					  struct optee_call_ctx *call_ctx)
 {
@@ -651,7 +652,8 @@ static void handle_rpc_func_cmd_shm_alloc(struct tee_context *ctx,
 		shm = optee_rpc_cmd_alloc_suppl(ctx, sz);
 		break;
 	case OPTEE_RPC_SHM_TYPE_KERNEL:
-		shm = tee_shm_alloc(ctx, sz, TEE_SHM_MAPPED | TEE_SHM_PRIV);
+		shm = tee_shm_alloc(optee->ctx, sz,
+				    TEE_SHM_MAPPED | TEE_SHM_PRIV);
 		break;
 	default:
 		arg->ret = TEEC_ERROR_BAD_PARAMETERS;
@@ -747,7 +749,7 @@ static void handle_rpc_func_cmd(struct tee_context *ctx, struct optee *optee,
 	switch (arg->cmd) {
 	case OPTEE_RPC_CMD_SHM_ALLOC:
 		free_pages_list(call_ctx);
-		handle_rpc_func_cmd_shm_alloc(ctx, arg, call_ctx);
+		handle_rpc_func_cmd_shm_alloc(ctx, optee, arg, call_ctx);
 		break;
 	case OPTEE_RPC_CMD_SHM_FREE:
 		handle_rpc_func_cmd_shm_free(ctx, arg);
@@ -776,7 +778,7 @@ static void optee_handle_rpc(struct tee_context *ctx,
 
 	switch (OPTEE_SMC_RETURN_GET_RPC_FUNC(param->a0)) {
 	case OPTEE_SMC_RPC_FUNC_ALLOC:
-		shm = tee_shm_alloc(ctx, param->a1,
+		shm = tee_shm_alloc(optee->ctx, param->a1,
 				    TEE_SHM_MAPPED | TEE_SHM_PRIV);
 		if (!IS_ERR(shm) && !tee_shm_get_pa(shm, 0, &pa)) {
 			reg_pair_from_64(&param->a1, &param->a2, pa);
@@ -954,57 +956,34 @@ static irqreturn_t notif_irq_thread_fn(int irq, void *dev_id)
 {
 	struct optee *optee = dev_id;
 
-	optee_smc_do_bottom_half(optee->notif.ctx);
+	optee_smc_do_bottom_half(optee->ctx);
 
 	return IRQ_HANDLED;
 }
 
 static int optee_smc_notif_init_irq(struct optee *optee, u_int irq)
 {
-	struct tee_context *ctx;
 	int rc;
 
-	ctx = teedev_open(optee->teedev);
-	if (IS_ERR(ctx))
-		return PTR_ERR(ctx);
-
-	optee->notif.ctx = ctx;
 	rc = request_threaded_irq(irq, notif_irq_handler,
 				  notif_irq_thread_fn,
 				  0, "optee_notification", optee);
 	if (rc)
-		goto err_close_ctx;
+		return rc;
 
 	optee->smc.notif_irq = irq;
 
 	return 0;
-
-err_close_ctx:
-	teedev_close_context(optee->notif.ctx);
-	optee->notif.ctx = NULL;
-
-	return rc;
 }
 
 static void optee_smc_notif_uninit_irq(struct optee *optee)
 {
-	if (optee->notif.ctx) {
-		optee_smc_stop_async_notif(optee->notif.ctx);
+	if (optee->smc.sec_caps & OPTEE_SMC_SEC_CAP_ASYNC_NOTIF) {
+		optee_smc_stop_async_notif(optee->ctx);
 		if (optee->smc.notif_irq) {
 			free_irq(optee->smc.notif_irq, optee);
 			irq_dispose_mapping(optee->smc.notif_irq);
 		}
-
-		/*
-		 * The thread normally working with optee->notif.ctx was
-		 * stopped with free_irq() above.
-		 *
-		 * Note we're not using teedev_close_context() or
-		 * tee_client_close_context() since we have already called
-		 * tee_device_put() while initializing to avoid a circular
-		 * reference counting.
-		 */
-		teedev_close_context(optee->notif.ctx);
 	}
 }
 
@@ -1366,6 +1345,7 @@ static int optee_probe(struct platform_device *pdev)
 	struct optee *optee = NULL;
 	void *memremaped_shm = NULL;
 	struct tee_device *teedev;
+	struct tee_context *ctx;
 	u32 max_notif_value;
 	u32 sec_caps;
 	int rc;
@@ -1446,9 +1426,13 @@ static int optee_probe(struct platform_device *pdev)
 	optee->pool = pool;
 
 	platform_set_drvdata(pdev, optee);
+	ctx = teedev_open(optee->teedev);
+	if (IS_ERR(ctx))
+		goto err_supp_uninit;
+	optee->ctx = ctx;
 	rc = optee_notif_init(optee, max_notif_value);
 	if (rc)
-		goto err_supp_uninit;
+		goto err_close_ctx;
 
 	if (sec_caps & OPTEE_SMC_SEC_CAP_ASYNC_NOTIF) {
 		unsigned int irq;
@@ -1496,6 +1480,8 @@ err_disable_shm_cache:
 	optee_unregister_devices();
 err_notif_uninit:
 	optee_notif_uninit(optee);
+err_close_ctx:
+	teedev_close_context(ctx);
 err_supp_uninit:
 	optee_supp_uninit(&optee->supp);
 	mutex_destroy(&optee->call_queue.mutex);