Re: [PATCH v18 2/6] remoteproc: Add TEE support

On 6/17/25 06:34, Bjorn Andersson wrote:

On Mon, Jun 16, 2025 at 09:55:26AM +0200, Arnaud Pouliquen wrote:

...

Add a remoteproc TEE (Trusted Execution Environment) driver that will be probed by the TEE bus. If the associated Trusted application is supported on the secure part, this driver offers a client interface to load firmware by the secure part. This firmware could be authenticated by the secure trusted application.

A specificity of the implementation is that the firmware has to be

[...]

...

• }

+} +EXPORT_SYMBOL_GPL(rproc_tee_release_fw);

+/**

• • rproc_tee_load_fw - Load firmware from TEE application
• • @rproc: Pointer to the struct rproc representing the remote processor
• • @fw: Pointer to the firmware structure containing the firmware data and size
• • This function invokes the TA_RPROC_FW_CMD_LOAD_FW TEE client function to load the firmware.
• • It registers the fw->data as a shared memory region with the TEE, and request the TEE to load
• • the firmware. This function can be called twice during the remote processor boot, considering
• • that the TEE application ignores the command if the firmware is already loaded.

"can be called twice"? How does this relate to the TEE application ignoring the command? How does the client know if the application ignored it?

No need that the client is aware. only needed due to the boot sequence that calls rproc_tee_parse_fw() then the rproc_tee_load_fw() I will update the comment to make it more explicit.

...

• • @rproc: Pointer to the struct rproc representing the remote processor
• • @fw: Pointer to the firmware structure containing the firmware data and size

https://docs.kernel.org/doc-guide/kernel-doc.html#function-documentation says function name should have () suffix. Then arguments (here you have them twice). Then the longer description (for rproc_tee_release_fw() you have this before arguments).

Also, please keep it within 80 characters.

...

• • Return: 0 on success, or an error code on failure
• */

+int rproc_tee_load_fw(struct rproc *rproc, const struct firmware *fw) +{

• struct tee_param param[MAX_TEE_PARAM_ARRAY_MEMBER];
• struct rproc_tee *trproc = rproc_to_trproc(rproc);
• struct tee_ioctl_invoke_arg arg;
• struct tee_shm *fw_shm;
• int ret;
• if (!rproc_tee_ctx.dev)

• return -ENODEV;
  

• if (!trproc)

• return -EINVAL;
  

• fw_shm = tee_shm_register_kernel_buf(rproc_tee_ctx.tee_ctx, (void *)fw->data, fw->size);
• if (IS_ERR(fw_shm))

• return PTR_ERR(fw_shm);
  

• rproc_tee_prepare_args(trproc, TA_RPROC_FW_CMD_LOAD_FW, &arg, param, 1);
• /* Provide the address of the firmware image */
• param[1] = (struct tee_param) {

• .attr = TEE_IOCTL_PARAM_ATTR_TYPE_MEMREF_INPUT,
  

• .u.memref = {
  

• 	.shm = fw_shm,
  

• 	.size = fw->size,
  

• 	.shm_offs = 0,
  

• },
  

• };
• ret = tee_client_invoke_func(rproc_tee_ctx.tee_ctx, &arg, param);
• if (ret < 0 || arg.ret != 0) {

• dev_err(rproc_tee_ctx.dev,
  

• 	"TA_RPROC_FW_CMD_LOAD_FW invoke failed TEE err: %#x, ret:%d\n",
  

• 	arg.ret, ret);
  

• if (!ret)
  

• 	ret = -EIO;
  

If ret == 0 and arg.ret == <some error>, then this function will print an error to the log and return success (0). Same with many of the other functions where you have copy pasted this.

If ret == 0 and arg.ret == <some error>, we return -EIO, or I missed something?

...

• }
• tee_shm_free(fw_shm);
• return ret;

+} +EXPORT_SYMBOL_GPL(rproc_tee_load_fw);

+static int rproc_tee_get_loaded_rsc_table(struct rproc *rproc, phys_addr_t *rsc_pa,

• 			  size_t *table_sz)
  

+{

• struct tee_param param[MAX_TEE_PARAM_ARRAY_MEMBER];
• struct rproc_tee *trproc = rproc_to_trproc(rproc);
• struct tee_ioctl_invoke_arg arg;
• int ret;
• if (!rproc_tee_ctx.dev)

• return -ENODEV;
  

• if (!trproc)

• return -EINVAL;
  

• rproc_tee_prepare_args(trproc, TA_RPROC_FW_CMD_GET_RSC_TABLE, &arg, param, 2);
• param[1].attr = TEE_IOCTL_PARAM_ATTR_TYPE_VALUE_OUTPUT;
• param[2].attr = TEE_IOCTL_PARAM_ATTR_TYPE_VALUE_OUTPUT;
• ret = tee_client_invoke_func(rproc_tee_ctx.tee_ctx, &arg, param);
• if (ret < 0 || arg.ret != 0) {

• dev_err(rproc_tee_ctx.dev,
  

• 	"TA_RPROC_FW_CMD_GET_RSC_TABLE invoke failed TEE err: %#x, ret:%d\n",
  

• 	arg.ret, ret);
  

• return -EIO;
  

• }
• *table_sz = param[2].u.value.a;

What happened to .b?

Not really used for now, but I will add extra code to support addresses and size in 64-bits testing the sizeof phys_addr_t and size_t.

...

• if (*table_sz)

• *rsc_pa = param[1].u.value.a;
  

Ditto

...

• else

• *rsc_pa  = 0;
  

• return 0;

+}

+/**

• • rproc_tee_parse_fw - Get the resource table from TEE application
• • @rproc: Pointer to the struct rproc representing the remote processor
• • @fw: Pointer to the firmware structure containing the firmware data and size
• • This function retrieves the loaded resource table and creates a cached_table copy. Since the
• • firmware image is signed and potentially encrypted, the firmware must be loaded first to
• • access the loaded resource table.
• • @rproc: Pointer to the struct rproc representing the remote processor
• • @fw: Pointer to the firmware structure containing the firmware data and size

Duplicated arguments list...

...

• • Return: 0 on success, or an error code on failure
• */

+int rproc_tee_parse_fw(struct rproc *rproc, const struct firmware *fw) +{

• phys_addr_t rsc_table;
• void __iomem *rsc_va;
• size_t table_sz;
• int ret;
• if (!rproc)

• return -EINVAL;
  

• /* We need first to Load the firmware, to be able to get the resource table. */
• ret = rproc_tee_load_fw(rproc, fw);
• if (ret)

• return ret;
  

• ret = rproc_tee_get_loaded_rsc_table(rproc, &rsc_table, &table_sz);
• if (ret)

• goto release_fw;
  

• /*

• * We assume here that the memory mapping is the same between the TEE and Linux kernel
  

• * contexts. Else a new TEE remoteproc service could be needed to get a copy of the
  

• * resource table
  

• */
  

• rsc_va = ioremap_wc(rsc_table, table_sz);

You're using tee_shm for transferring the image, you don't allow Linux access to any part of the firmware (which is the reason why you need to load the segments in rproc_parse_fw())...

Right, Linux does not understand the format and could not retrieve the resource table in case of encryption.

So just out of curiosity, why is the resource table passed back to Linux using a ioremap of some random/undefined chunk of memory?

It is not a random chunck of memory but the address of the table installed in the remote processor memory. It is the equivalent of get_loaded_rsc_table ops.

...

• if (!rsc_va) {

• dev_err(rproc_tee_ctx.dev, "Unable to map memory region: %pa+%zx\n",
  

• 	&rsc_table, table_sz);
  

• ret = -ENOMEM;
  

• goto release_fw;
  

• }
• /*

• * Create a copy of the resource table to have the same behavior as the ELF loader.
  

• * This cached table will be used by the remoteproc core after the remoteproc stops
  

• * to free resources and for crash recovery to reapply the settings.
  

• * The cached table will be freed by the remoteproc core.
  

• */
  

• rproc->cached_table = kmemdup((__force void *)rsc_va, table_sz, GFP_KERNEL);

Is rsc_va DDR? And if so, wouldn't memremap() be a more accurate choice above. If not, why isn't this kzalloc() + memcpy_fromio()?

Here we explicitly want to make a local copy as done in rproc_elf_load_rsc_table(). using kzalloc() + memcpy_fromio() seems a better implementation I test your proposal, thanks

...

• iounmap(rsc_va);
• if (!rproc->cached_table) {

• ret = -ENOMEM;
  

• goto release_fw;
  

• }
• rproc->table_ptr = rproc->cached_table;
• rproc->table_sz = table_sz;
• return 0;

+release_fw:

• rproc_tee_release_fw(rproc);
• return ret;

+} +EXPORT_SYMBOL_GPL(rproc_tee_parse_fw);

+/**

• • rproc_tee_find_loaded_rsc_table - Find the loaded resource table loaded by the TEE application
• • @rproc: Pointer to the struct rproc representing the remote processor
• • @fw: Pointer to the firmware structure containing the firmware data and size
• • This function retrieves the physical address and size of the resource table loaded by the TEE
• • application.
• • @rproc: Pointer to the struct rproc representing the remote processor
• • @fw: Pointer to the firmware structure containing the firmware data and size

More argument duplication.

...

• • Return: pointer to the resource table if found, or NULL if not found or size is 0
• */

+struct resource_table *rproc_tee_find_loaded_rsc_table(struct rproc *rproc,

• 				       const struct firmware *fw)
  

+{

• phys_addr_t rsc_table;
• size_t table_sz;
• int ret;
• ret = rproc_tee_get_loaded_rsc_table(rproc, &rsc_table, &table_sz);
• if (ret)

• return NULL;
  

• rproc->table_sz = table_sz;
• if (!table_sz)

• return NULL;
  

• /*

• * At this step the memory area that contains the resource table should have been registered
  

• * by the remote proc platform driver and allocated by rproc_alloc_registered_carveouts().
  

• */
  

• return (struct resource_table *)rproc_pa_to_va(rproc, rsc_table, table_sz, NULL);

rproc_pa_to_va() return type is void *, do you really need this explicit typecast?

...

+} +EXPORT_SYMBOL_GPL(rproc_tee_find_loaded_rsc_table);

+/**

• • rproc_tee_start - Request the TEE application to start the remote processor
• • This function invokes the TA_RPROC_FW_CMD_START command to start the remote processor.
• • @rproc: Pointer to the struct rproc representing the remote processor

kernel-doc ordering...

...

• • Return: Returns 0 on success, -EINVAL or -EIO on failure
• */

+int rproc_tee_start(struct rproc *rproc) +{

• struct tee_param param[MAX_TEE_PARAM_ARRAY_MEMBER];
• struct rproc_tee *trproc = rproc_to_trproc(rproc);
• struct tee_ioctl_invoke_arg arg;
• int ret = 0;

First access of ret is an assignment, no need to zero-initialize it here.

...

• if (!trproc)

• return -EINVAL;
  

There's an inconsistency in that rproc_tee_ctx.dev is used without first checking that it's valid in this function...

...

• rproc_tee_prepare_args(trproc, TA_RPROC_FW_CMD_START, &arg, param, 0);
• ret = tee_client_invoke_func(rproc_tee_ctx.tee_ctx, &arg, param);
• if (ret < 0 || arg.ret != 0) {

• dev_err(rproc_tee_ctx.dev,
  

• 	"TA_RPROC_FW_CMD_START invoke failed TEE err: %#x, ret:%d\n", arg.ret, ret);
  

• if (!ret)
  

• 	return  -EIO;
  

Why not assigning ret and falling through, like in most other functions?

...

• }
• return 0;

+} +EXPORT_SYMBOL_GPL(rproc_tee_start);

+/**

• • rproc_tee_stop - Request the TEE application to start the remote processor
• • This function invokes the TA_RPROC_FW_CMD_STOP command to stop the remote processor.
• • @rproc: Pointer to the struct rproc representing the remote processor
• • Return: Returns 0 on success, -EINVAL or -EIO on failure
• */

+int rproc_tee_stop(struct rproc *rproc) +{

• struct tee_param param[MAX_TEE_PARAM_ARRAY_MEMBER];
• struct rproc_tee *trproc = rproc_to_trproc(rproc);
• struct tee_ioctl_invoke_arg arg;
• int ret;
• if (!trproc)

• return -EINVAL;
  

• rproc_tee_prepare_args(trproc, TA_RPROC_FW_CMD_STOP, &arg, param, 0);
• ret = tee_client_invoke_func(rproc_tee_ctx.tee_ctx, &arg, param);
• if (ret < 0 || arg.ret != 0) {

• dev_err(rproc_tee_ctx.dev,
  

• 	"TA_RPROC_FW_CMD_STOP invoke failed TEE err: %#x, ret:%d\n", arg.ret, ret);
  

• if (!ret)
  

• 	ret = -EIO;
  

• }
• return ret;

+} +EXPORT_SYMBOL_GPL(rproc_tee_stop);

+static const struct tee_client_device_id rproc_tee_id_table[] = {

• {UUID_INIT(0x80a4c275, 0x0a47, 0x4905, 0x82, 0x85, 0x14, 0x86, 0xa9, 0x77, 0x1a, 0x08)},
• {}

+};

+/**

• • rproc_tee_register - Register a remote processor controlled by a TEE application.
• • This function registers a remote processor that will be managed by a TEE application,by opening
• • a session with the TEE client.
• • @dev: Pointer to client rproc device
• • @rproc: Pointer to the struct rproc representing the remote processor
• • @rproc_id: ID of the remote processor
• • Return: Returns 0 on success, or an error code on failure
• */

+int rproc_tee_register(struct device *dev, struct rproc *rproc, unsigned int rproc_id) +{

• struct tee_param param[MAX_TEE_PARAM_ARRAY_MEMBER];
• struct tee_ioctl_open_session_arg sess_arg;
• struct tee_client_device *tee_device;
• struct rproc_tee *trproc;
• struct device_link *link;
• int ret;
• spin_lock(&ctx_lock);

Why is this a spin_lock?

To my understanding the purpose of ctx_lock is to ensure mutual exclusion of access to rproc_tee_ctx, but this doesn't look like it's done from a context that isn't able to tolerate a mutex.

In particular during boot, if you have multiple remoteprocs being registred, you're going to waste precious CPU cycles just spinning here.

And if it is a spinlock because you might enter here from some interrupt context, how do you ensure this won't deadlock?

Right I need to use mutex instead and beter protect all the API from concurrent access to better manage bind/unbind.

...

• /*

• * Test if the device has been probed by the TEE bus. In case of failure, we ignore the
  

• * reason. The bus could be not yet probed or the service not available in the secure
  

• * firmware.The assumption in such a case is that the TEE remoteproc is not probed.
  

• */
  

• if (!rproc_tee_ctx.dev) {

• ret = -EPROBE_DEFER;
  

• goto out;
  

• }
• trproc = kzalloc(sizeof(*trproc), GFP_KERNEL);
• if (!trproc) {

• ret = -ENOMEM;
  

• goto out;
  

• }
• tee_device = to_tee_client_device(rproc_tee_ctx.dev);
• memset(&sess_arg, 0, sizeof(sess_arg));
• memcpy(sess_arg.uuid, tee_device->id.uuid.b, TEE_IOCTL_UUID_LEN);
• sess_arg.clnt_login = TEE_IOCTL_LOGIN_REE_KERNEL;
• sess_arg.num_params = 1;
• param[0] = (struct tee_param) {

• .attr = TEE_IOCTL_PARAM_ATTR_TYPE_VALUE_INPUT,
  

• .u.value.a = rproc_id,
  

• };

I'm not familiar with the details of the tee calling convention, but do you really need 4 entries in the param array and if so, would it make sense to zero-initialize the other elements - like you do in all other functions?

In rproc_tee_prepare_args(), we reinitialize the arguments and parameters because they may be reused from a previous message. This allow to set only non zero value.

THere as sess_arg.num_params = 1 only param[0] is valid, other param should be ignored. In this context use memeset seems over protection here.

Thanks, Arnaud

...

• ret = tee_client_open_session(rproc_tee_ctx.tee_ctx, &sess_arg, param);
• if (ret < 0 || sess_arg.ret != 0) {

• dev_err(dev, "tee_client_open_session failed, err: %#x\n", sess_arg.ret);
  

• ret = -EINVAL;
  

• goto free_tproc;
  

• }
• trproc->rproc_id = rproc_id;
• trproc->session_id = sess_arg.session;
• trproc->rproc = rproc;
• /* Create device link between the rproc device and the TEE device */

This comment would be more useful if it documented why the link is created - the fact that a link is added between the two devices can be read on the next line...

...

• link = device_link_add(dev, rproc_tee_ctx.dev, DL_FLAG_AUTOREMOVE_CONSUMER);
• if (!link) {

• ret = -ENOMEM;
  

• goto close_tee;
  

• }
• list_add_tail(&trproc->node, &rproc_tee_ctx.sessions);
• goto out;

Please don't use goto to skip over the error handling. ret is going to be 0 here, so unlock and return 0 here to make it clear that this is the point of successful return.

...

+close_tee:

• if (tee_client_close_session(rproc_tee_ctx.tee_ctx, trproc->session_id))

• dev_err(rproc_tee_ctx.dev, "tee_client_close_session failed\n");
  

+free_tproc:

• kfree(trproc);

+out:

• spin_unlock(&ctx_lock);
• return ret;

+} +EXPORT_SYMBOL_GPL(rproc_tee_register);

+/**

• • rproc_tee_unregister - Register a remote processor controlled by a TEE application.
• • This function unregisters a remote processor previously registered by the rproc_tee_register()
• • function.
• • @dev: Pointer to client rproc device
• • @rproc: Pointer to the struct rproc representing the remote processor
• • Return: Returns 0 on success, or an error code on failure
• */

+int rproc_tee_unregister(struct device *dev, struct rproc *rproc) +{

• struct rproc_tee *trproc = rproc_to_trproc(rproc);
• int ret;
• if (!trproc)

• return -EINVAL;
  

• spin_lock(&ctx_lock);

This lock is taken after you have traversed the rproc_tee_ctx.session list in a racy fashion.

...

• ret = tee_client_close_session(rproc_tee_ctx.tee_ctx, trproc->session_id);
• if (ret < 0)

• dev_err(rproc_tee_ctx.dev, "tee_client_close_session failed, err: %#x\n", ret);
  

• spin_unlock(&ctx_lock);
• list_del(&trproc->node);

Although, I might misunderstand your locking scheme, because here you're modifying the sessions list immediately after leaving the mutual exclusion region?

...

• kfree(trproc);
• return ret;

+} +EXPORT_SYMBOL_GPL(rproc_tee_unregister);

+static int rproc_tee_ctx_match(struct tee_ioctl_version_data *ver, const void *data) +{

• /* Today we support only the OP-TEE, could be extend to other tees */
• return (ver->impl_id == TEE_IMPL_ID_OPTEE);

+}

+static int rproc_tee_probe(struct device *dev) +{

• struct tee_context *tee_ctx;
• /* Open context with TEE driver */
• tee_ctx = tee_client_open_context(NULL, rproc_tee_ctx_match, NULL, NULL);
• if (IS_ERR(tee_ctx))

• return PTR_ERR(tee_ctx);
  

• spin_lock(&ctx_lock);
• rproc_tee_ctx.dev = dev;
• rproc_tee_ctx.tee_ctx = tee_ctx;
• INIT_LIST_HEAD(&rproc_tee_ctx.sessions);
• spin_unlock(&ctx_lock);
• return 0;

+}

+static int rproc_tee_remove(struct device *dev) +{

• spin_lock(&ctx_lock);
• tee_client_close_context(rproc_tee_ctx.tee_ctx);
• rproc_tee_ctx.dev = NULL;
• rproc_tee_ctx.tee_ctx = NULL;
• spin_unlock(&ctx_lock);
• return 0;

+}

+MODULE_DEVICE_TABLE(tee, rproc_tee_id_table);

+static struct tee_client_driver rproc_tee_fw_driver = {

• .id_table = rproc_tee_id_table,
• .driver = {

• .name		= KBUILD_MODNAME,
  

• .bus		= &tee_bus_type,
  

• .probe		= rproc_tee_probe,
  

• .remove		= rproc_tee_remove,
  

• },

+};

+static int __init rproc_tee_fw_mod_init(void) +{

• return driver_register(&rproc_tee_fw_driver.driver);

+}

+static void __exit rproc_tee_fw_mod_exit(void) +{

• driver_unregister(&rproc_tee_fw_driver.driver);

+}

+module_init(rproc_tee_fw_mod_init); +module_exit(rproc_tee_fw_mod_exit);

+MODULE_DESCRIPTION(" remote processor TEE module");

Why is there a space in the start of the description?

...

+MODULE_LICENSE("GPL"); diff --git a/include/linux/remoteproc_tee.h b/include/linux/remoteproc_tee.h new file mode 100644 index 000000000000..659bd77a4f12 --- /dev/null +++ b/include/linux/remoteproc_tee.h @@ -0,0 +1,90 @@ +/* SPDX-License-Identifier: GPL-2.0-or-later */ +/*

• • Copyright(c) 2024 STMicroelectronics

Year?

...

• */

+#ifndef REMOTEPROC_TEE_H +#define REMOTEPROC_TEE_H

+#include <linux/tee_drv.h> +#include <linux/firmware.h> +#include <linux/remoteproc.h>

+struct rproc; +struct rproc_tee;

rproc_tee is not used in the API, so there shouldn't be a need for forward declaring it.

struct rproc is defined in linux/remoteproc.h, so that should be fine to omit as well.

...

+#if IS_ENABLED(CONFIG_REMOTEPROC_TEE)

+int rproc_tee_register(struct device *dev, struct rproc *rproc, unsigned int rproc_id); +int rproc_tee_unregister(struct device *dev, struct rproc *rproc); +int rproc_tee_parse_fw(struct rproc *rproc, const struct firmware *fw); +int rproc_tee_load_fw(struct rproc *rproc, const struct firmware *fw); +void rproc_tee_release_fw(struct rproc *rproc); +struct resource_table *rproc_tee_find_loaded_rsc_table(struct rproc *rproc,

• 				       const struct firmware *fw);
  

+int rproc_tee_start(struct rproc *rproc); +int rproc_tee_stop(struct rproc *rproc);

+#else

+static inline int rproc_tee_register(struct device *dev, struct rproc *rproc, unsigned int rproc_id) +{

• return -ENODEV;

+}

+static inline int rproc_tee_parse_fw(struct rproc *rproc, const struct firmware *fw) +{

• /* This shouldn't be possible */
• WARN_ON(1);
• return 0;

+}

+static inline int rproc_tee_unregister(struct device *dev, struct rproc *rproc) +{

• /* This shouldn't be possible */
• WARN_ON(1);
• return 0;

+}

+static inline int rproc_tee_load_fw(struct rproc *rproc, const struct firmware *fw)

Double space after the ','

Regards, Bjorn

...

+{

• /* This shouldn't be possible */
• WARN_ON(1);
• return 0;

+}

+static inline int rproc_tee_start(struct rproc *rproc) +{

• /* This shouldn't be possible */
• WARN_ON(1);
• return 0;

+}

+static inline int rproc_tee_stop(struct rproc *rproc) +{

• /* This shouldn't be possible */
• WARN_ON(1);
• return 0;

+}

+static inline void rproc_tee_release_fw(struct rproc *rproc) +{

• /* This shouldn't be possible */
• WARN_ON(1);

+}

+static inline struct resource_table * +rproc_tee_find_loaded_rsc_table(struct rproc *rproc, const struct firmware *fw) +{

• /* This shouldn't be possible */
• WARN_ON(1);
• return NULL;

+} +#endif /* CONFIG_REMOTEPROC_TEE */

+#endif /* REMOTEPROC_TEE_H */

2.25.1