[PATCH v2 00/10] firmware: arm_ffa: Refactoring and initial/minor v1.1 update

[PATCH v2 00/10] firmware: arm_ffa: Refactoring and initial/minor v1.1 update

[Sudeep Holla]

Hi All,

This series is just some refactoring in preparation to add FF-A v1.1
support. It doesn't have any memory layout or notification changes
supported in v1.1 yet.

Regards,
Sudeep

v1[1]->v2:
	- Merged dropping of ffa_ops in optee_ffa structure and using
	  ffa_dev->ops into single patch
	- Added separate patch(didn't fit any patch strictly to fit in)
	  to rename ffa_dev_ops as ffa_ops as suggested by Sumit
	- Fixed some minor comments, handling size > structure size in
	  partition_info_get and added extra parameter to ffa_features
	  to get both possible output/interface properties.

[1] https://lore.kernel.org/all/20220830100700.344594-1-sudeep.holla@arm.com/

Sudeep Holla (10):
  firmware: arm_ffa: Add pointer to the ffa_dev_ops in struct ffa_dev
  tee: optee: Drop ffa_ops in optee_ffa structure using ffa_dev->ops directly
  firmware: arm_ffa: Remove ffa_dev_ops_get()
  firmware: arm_ffa: Add support for querying FF-A features
  firmware: arm_ffa: Use FFA_FEATURES to detect if native versions are supported
  firmware: arm_ffa: Make memory apis ffa_device independent
  firmware: arm_ffa: Rename ffa_dev_ops as ffa_ops
  firmware: arm_ffa: Add v1.1 get_partition_info support
  firmware: arm_ffa: Set up 32bit execution mode flag using partiion property
  firmware: arm_ffa: Split up ffa_ops into info, message and memory operations

 drivers/firmware/arm_ffa/bus.c    |   4 +-
 drivers/firmware/arm_ffa/driver.c | 130 +++++++++++++++++++++++-------
 drivers/tee/optee/ffa_abi.c       |  46 +++++------
 drivers/tee/optee/optee_private.h |   1 -
 include/linux/arm_ffa.h           |  36 ++++++---
 5 files changed, 150 insertions(+), 67 deletions(-)

-- 
2.37.3


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH v2 01/10] firmware: arm_ffa: Add pointer to the ffa_dev_ops in struct ffa_dev

[Sudeep Holla]

Currently ffa_dev_ops_get() is the way to fetch the ffa_dev_ops pointer.
It checks if the ffa_dev structure pointer is valid before returning the
ffa_dev_ops pointer.

Instead, the pointer can be made part of the ffa_dev structure and since
the core driver is incharge of creating ffa_device for each identified
partition, there is no need to check for the validity explicitly if the
pointer is embedded in the structure.

Add the pointer to the ffa_dev_ops in the ffa_dev structure itself and
initialise the same as part of creation of the device.

Reviewed-by: Jens Wiklander <jens.wiklander@linaro.org>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
---
 drivers/firmware/arm_ffa/bus.c    | 4 +++-
 drivers/firmware/arm_ffa/driver.c | 2 +-
 include/linux/arm_ffa.h           | 7 +++++--
 3 files changed, 9 insertions(+), 4 deletions(-)

diff --git a/drivers/firmware/arm_ffa/bus.c b/drivers/firmware/arm_ffa/bus.c
index 641a91819088..69328041fbc3 100644
--- a/drivers/firmware/arm_ffa/bus.c
+++ b/drivers/firmware/arm_ffa/bus.c
@@ -167,7 +167,8 @@ bool ffa_device_is_valid(struct ffa_device *ffa_dev)
 	return valid;
 }
 
-struct ffa_device *ffa_device_register(const uuid_t *uuid, int vm_id)
+struct ffa_device *ffa_device_register(const uuid_t *uuid, int vm_id,
+				       const struct ffa_dev_ops *ops)
 {
 	int ret;
 	struct device *dev;
@@ -183,6 +184,7 @@ struct ffa_device *ffa_device_register(const uuid_t *uuid, int vm_id)
 	dev_set_name(&ffa_dev->dev, "arm-ffa-%04x", vm_id);
 
 	ffa_dev->vm_id = vm_id;
+	ffa_dev->ops = ops;
 	uuid_copy(&ffa_dev->uuid, uuid);
 
 	ret = device_register(&ffa_dev->dev);
diff --git a/drivers/firmware/arm_ffa/driver.c b/drivers/firmware/arm_ffa/driver.c
index ec731e9e942b..213665e5ad0e 100644
--- a/drivers/firmware/arm_ffa/driver.c
+++ b/drivers/firmware/arm_ffa/driver.c
@@ -688,7 +688,7 @@ static void ffa_setup_partitions(void)
 		 * as part of the discovery API, we need to pass the
 		 * discovered UUID here instead.
 		 */
-		ffa_dev = ffa_device_register(&uuid_null, tpbuf->id);
+		ffa_dev = ffa_device_register(&uuid_null, tpbuf->id, &ffa_ops);
 		if (!ffa_dev) {
 			pr_err("%s: failed to register partition ID 0x%x\n",
 			       __func__, tpbuf->id);
diff --git a/include/linux/arm_ffa.h b/include/linux/arm_ffa.h
index e5c76c1ef9ed..91b47e42b73d 100644
--- a/include/linux/arm_ffa.h
+++ b/include/linux/arm_ffa.h
@@ -17,6 +17,7 @@ struct ffa_device {
 	bool mode_32bit;
 	uuid_t uuid;
 	struct device dev;
+	const struct ffa_dev_ops *ops;
 };
 
 #define to_ffa_dev(d) container_of(d, struct ffa_device, dev)
@@ -47,7 +48,8 @@ static inline void *ffa_dev_get_drvdata(struct ffa_device *fdev)
 }
 
 #if IS_REACHABLE(CONFIG_ARM_FFA_TRANSPORT)
-struct ffa_device *ffa_device_register(const uuid_t *uuid, int vm_id);
+struct ffa_device *ffa_device_register(const uuid_t *uuid, int vm_id,
+				       const struct ffa_dev_ops *ops);
 void ffa_device_unregister(struct ffa_device *ffa_dev);
 int ffa_driver_register(struct ffa_driver *driver, struct module *owner,
 			const char *mod_name);
@@ -57,7 +59,8 @@ const struct ffa_dev_ops *ffa_dev_ops_get(struct ffa_device *dev);
 
 #else
 static inline
-struct ffa_device *ffa_device_register(const uuid_t *uuid, int vm_id)
+struct ffa_device *ffa_device_register(const uuid_t *uuid, int vm_id,
+				       const struct ffa_dev_ops *ops)
 {
 	return NULL;
 }
-- 
2.37.3


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH v2 02/10] tee: optee: Drop ffa_ops in optee_ffa structure using ffa_dev->ops directly

[Sudeep Holla]

Now that the ffa_device structure holds the pointer to ffa_dev_ops,
there is no need to obtain the same through ffa_dev_ops_get().

Just use the ffa_dev->ops directly. Since the ffa_device itself carries
ffa_dev_ops now, there is no need to keep a copy in optee_ffa structure.

Drop ffa_ops in the optee_ffa structure as it is not needed anymore.

Reviewed-by: Jens Wiklander <jens.wiklander@linaro.org>
Reviewed-by: Sumit Garg <sumit.garg@linaro.org>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
---
 drivers/tee/optee/ffa_abi.c       | 15 +++++----------
 drivers/tee/optee/optee_private.h |  1 -
 2 files changed, 5 insertions(+), 11 deletions(-)

diff --git a/drivers/tee/optee/ffa_abi.c b/drivers/tee/optee/ffa_abi.c
index 7ab31740cff8..3d4079575ccd 100644
--- a/drivers/tee/optee/ffa_abi.c
+++ b/drivers/tee/optee/ffa_abi.c
@@ -271,8 +271,8 @@ static int optee_ffa_shm_register(struct tee_context *ctx, struct tee_shm *shm,
 				  unsigned long start)
 {
 	struct optee *optee = tee_get_drvdata(ctx->teedev);
-	const struct ffa_dev_ops *ffa_ops = optee->ffa.ffa_ops;
 	struct ffa_device *ffa_dev = optee->ffa.ffa_dev;
+	const struct ffa_dev_ops *ffa_ops = ffa_dev->ops;
 	struct ffa_mem_region_attributes mem_attr = {
 		.receiver = ffa_dev->vm_id,
 		.attrs = FFA_MEM_RW,
@@ -314,8 +314,8 @@ static int optee_ffa_shm_unregister(struct tee_context *ctx,
 				    struct tee_shm *shm)
 {
 	struct optee *optee = tee_get_drvdata(ctx->teedev);
-	const struct ffa_dev_ops *ffa_ops = optee->ffa.ffa_ops;
 	struct ffa_device *ffa_dev = optee->ffa.ffa_dev;
+	const struct ffa_dev_ops *ffa_ops = ffa_dev->ops;
 	u64 global_handle = shm->sec_world_id;
 	struct ffa_send_direct_data data = {
 		.data0 = OPTEE_FFA_UNREGISTER_SHM,
@@ -342,7 +342,7 @@ static int optee_ffa_shm_unregister_supp(struct tee_context *ctx,
 					 struct tee_shm *shm)
 {
 	struct optee *optee = tee_get_drvdata(ctx->teedev);
-	const struct ffa_dev_ops *ffa_ops = optee->ffa.ffa_ops;
+	const struct ffa_dev_ops *ffa_ops = optee->ffa.ffa_dev->ops;
 	u64 global_handle = shm->sec_world_id;
 	int rc;
 
@@ -529,8 +529,8 @@ static int optee_ffa_yielding_call(struct tee_context *ctx,
 				   struct optee_msg_arg *rpc_arg)
 {
 	struct optee *optee = tee_get_drvdata(ctx->teedev);
-	const struct ffa_dev_ops *ffa_ops = optee->ffa.ffa_ops;
 	struct ffa_device *ffa_dev = optee->ffa.ffa_dev;
+	const struct ffa_dev_ops *ffa_ops = ffa_dev->ops;
 	struct optee_call_waiter w;
 	u32 cmd = data->data0;
 	u32 w4 = data->data1;
@@ -793,11 +793,7 @@ static int optee_ffa_probe(struct ffa_device *ffa_dev)
 	u32 sec_caps;
 	int rc;
 
-	ffa_ops = ffa_dev_ops_get(ffa_dev);
-	if (!ffa_ops) {
-		pr_warn("failed \"method\" init: ffa\n");
-		return -ENOENT;
-	}
+	ffa_ops = ffa_dev->ops;
 
 	if (!optee_ffa_api_is_compatbile(ffa_dev, ffa_ops))
 		return -EINVAL;
@@ -821,7 +817,6 @@ static int optee_ffa_probe(struct ffa_device *ffa_dev)
 
 	optee->ops = &optee_ffa_ops;
 	optee->ffa.ffa_dev = ffa_dev;
-	optee->ffa.ffa_ops = ffa_ops;
 	optee->rpc_param_count = rpc_param_count;
 
 	teedev = tee_device_alloc(&optee_ffa_clnt_desc, NULL, optee->pool,
diff --git a/drivers/tee/optee/optee_private.h b/drivers/tee/optee/optee_private.h
index a33d98d17cfd..04ae58892608 100644
--- a/drivers/tee/optee/optee_private.h
+++ b/drivers/tee/optee/optee_private.h
@@ -111,7 +111,6 @@ struct optee_smc {
  */
 struct optee_ffa {
 	struct ffa_device *ffa_dev;
-	const struct ffa_dev_ops *ffa_ops;
 	/* Serializes access to @global_ids */
 	struct mutex mutex;
 	struct rhashtable global_ids;
-- 
2.37.3


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH v2 03/10] firmware: arm_ffa: Remove ffa_dev_ops_get()

[Sudeep Holla]

The only user of this exported ffa_dev_ops_get() was OPTEE driver which
now uses ffa_dev->ops directly, there are no other users for this.

Also, since any ffa driver can use ffa_dev->ops directly, there will be
no need for ffa_dev_ops_get(), so just remove ffa_dev_ops_get().

Reviewed-by: Jens Wiklander <jens.wiklander@linaro.org>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
---
 drivers/firmware/arm_ffa/driver.c | 9 ---------
 include/linux/arm_ffa.h           | 6 ------
 2 files changed, 15 deletions(-)

diff --git a/drivers/firmware/arm_ffa/driver.c b/drivers/firmware/arm_ffa/driver.c
index 213665e5ad0e..04e7cbb1b9aa 100644
--- a/drivers/firmware/arm_ffa/driver.c
+++ b/drivers/firmware/arm_ffa/driver.c
@@ -644,15 +644,6 @@ static const struct ffa_dev_ops ffa_ops = {
 	.memory_lend = ffa_memory_lend,
 };
 
-const struct ffa_dev_ops *ffa_dev_ops_get(struct ffa_device *dev)
-{
-	if (ffa_device_is_valid(dev))
-		return &ffa_ops;
-
-	return NULL;
-}
-EXPORT_SYMBOL_GPL(ffa_dev_ops_get);
-
 void ffa_device_match_uuid(struct ffa_device *ffa_dev, const uuid_t *uuid)
 {
 	int count, idx;
diff --git a/include/linux/arm_ffa.h b/include/linux/arm_ffa.h
index 91b47e42b73d..556f50f27fb1 100644
--- a/include/linux/arm_ffa.h
+++ b/include/linux/arm_ffa.h
@@ -55,7 +55,6 @@ int ffa_driver_register(struct ffa_driver *driver, struct module *owner,
 			const char *mod_name);
 void ffa_driver_unregister(struct ffa_driver *driver);
 bool ffa_device_is_valid(struct ffa_device *ffa_dev);
-const struct ffa_dev_ops *ffa_dev_ops_get(struct ffa_device *dev);
 
 #else
 static inline
@@ -79,11 +78,6 @@ static inline void ffa_driver_unregister(struct ffa_driver *driver) {}
 static inline
 bool ffa_device_is_valid(struct ffa_device *ffa_dev) { return false; }
 
-static inline
-const struct ffa_dev_ops *ffa_dev_ops_get(struct ffa_device *dev)
-{
-	return NULL;
-}
 #endif /* CONFIG_ARM_FFA_TRANSPORT */
 
 #define ffa_register(driver) \
-- 
2.37.3


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH v2 04/10] firmware: arm_ffa: Add support for querying FF-A features

[Sudeep Holla]

Add support for FFA_FEATURES to discover properties supported at the
FF-A interface. This interface can be used to query:
 - If an FF-A interface is implemented by the component at the higher EL,
 - If an implemented FF-A interface also implements any optional features
   described in its interface definition, and
 - Any implementation details exported by an implemented FF-A interface
   as described in its interface definition.

Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
---
 drivers/firmware/arm_ffa/driver.c | 26 ++++++++++++++++++++++++++
 1 file changed, 26 insertions(+)

diff --git a/drivers/firmware/arm_ffa/driver.c b/drivers/firmware/arm_ffa/driver.c
index 04e7cbb1b9aa..81b8d578b6ea 100644
--- a/drivers/firmware/arm_ffa/driver.c
+++ b/drivers/firmware/arm_ffa/driver.c
@@ -571,6 +571,32 @@ static int ffa_memory_reclaim(u64 g_handle, u32 flags)
 	return 0;
 }
 
+static int ffa_features(u32 func_feat_id, u32 input_props,
+			u32 *if_props_1, u32 *if_props_2)
+{
+	ffa_value_t id;
+
+	if (!ARM_SMCCC_IS_FAST_CALL(func_feat_id) && input_props) {
+		pr_err("%s: Invalid Parameters: %x, %x", __func__,
+		       func_feat_id, input_props);
+		return ffa_to_linux_errno(FFA_RET_INVALID_PARAMETERS);
+	}
+
+	invoke_ffa_fn((ffa_value_t){
+		.a0 = FFA_FEATURES, .a1 = func_feat_id, .a2 = input_props,
+		}, &id);
+
+	if (id.a0 == FFA_ERROR)
+		return ffa_to_linux_errno((int)id.a2);
+
+	if (if_props_1)
+		*if_props_1 = id.a2;
+	if (if_props_2)
+		*if_props_2 = id.a3;
+
+	return 0;
+}
+
 static u32 ffa_api_version_get(void)
 {
 	return drv_info->version;
-- 
2.37.3


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Re: [PATCH v2 04/10] firmware: arm_ffa: Add support for querying FF-A features

[Jens Wiklander]

On Fri, Sep 2, 2022 at 2:40 PM Sudeep Holla <sudeep.holla@arm.com> wrote:
>
> Add support for FFA_FEATURES to discover properties supported at the
> FF-A interface. This interface can be used to query:
>  - If an FF-A interface is implemented by the component at the higher EL,
>  - If an implemented FF-A interface also implements any optional features
>    described in its interface definition, and
>  - Any implementation details exported by an implemented FF-A interface
>    as described in its interface definition.
>
> Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
> ---
>  drivers/firmware/arm_ffa/driver.c | 26 ++++++++++++++++++++++++++
>  1 file changed, 26 insertions(+)

Reviewed-by: Jens Wiklander <jens.wiklander@linaro.org>

>
> diff --git a/drivers/firmware/arm_ffa/driver.c b/drivers/firmware/arm_ffa/driver.c
> index 04e7cbb1b9aa..81b8d578b6ea 100644
> --- a/drivers/firmware/arm_ffa/driver.c
> +++ b/drivers/firmware/arm_ffa/driver.c
> @@ -571,6 +571,32 @@ static int ffa_memory_reclaim(u64 g_handle, u32 flags)
>         return 0;
>  }
>
> +static int ffa_features(u32 func_feat_id, u32 input_props,
> +                       u32 *if_props_1, u32 *if_props_2)
> +{
> +       ffa_value_t id;
> +
> +       if (!ARM_SMCCC_IS_FAST_CALL(func_feat_id) && input_props) {
> +               pr_err("%s: Invalid Parameters: %x, %x", __func__,
> +                      func_feat_id, input_props);
> +               return ffa_to_linux_errno(FFA_RET_INVALID_PARAMETERS);
> +       }
> +
> +       invoke_ffa_fn((ffa_value_t){
> +               .a0 = FFA_FEATURES, .a1 = func_feat_id, .a2 = input_props,
> +               }, &id);
> +
> +       if (id.a0 == FFA_ERROR)
> +               return ffa_to_linux_errno((int)id.a2);
> +
> +       if (if_props_1)
> +               *if_props_1 = id.a2;
> +       if (if_props_2)
> +               *if_props_2 = id.a3;
> +
> +       return 0;
> +}
> +
>  static u32 ffa_api_version_get(void)
>  {
>         return drv_info->version;
> --
> 2.37.3
>

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH v2 05/10] firmware: arm_ffa: Use FFA_FEATURES to detect if native versions are supported

[Sudeep Holla]

Currently, the ffa_dev->mode_32bit is use to detect if the native 64-bit
or 32-bit versions of FF-A ABI needs to be used. However for the FF-A
memory ABIs, it is not dependent on the ffa_device(i.e. the partition)
itself, but the partition manager(SPM).

So, the FFA_FEATURES can be use to detect if the native 64bit ABIs are
supported or not and appropriate calls can be made based on that.

Use FFA_FEATURES to detect if native versions of MEM_LEND or MEM_SHARE
are implemented and make of the same to use native memory ABIs later on.

Reviewed-by: Jens Wiklander <jens.wiklander@linaro.org>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
---
 drivers/firmware/arm_ffa/driver.c | 22 ++++++++++++++++------
 1 file changed, 16 insertions(+), 6 deletions(-)

diff --git a/drivers/firmware/arm_ffa/driver.c b/drivers/firmware/arm_ffa/driver.c
index 81b8d578b6ea..37a8ee304508 100644
--- a/drivers/firmware/arm_ffa/driver.c
+++ b/drivers/firmware/arm_ffa/driver.c
@@ -163,6 +163,7 @@ struct ffa_drv_info {
 	struct mutex tx_lock; /* lock to protect Tx buffer */
 	void *rx_buffer;
 	void *tx_buffer;
+	bool mem_ops_native;
 };
 
 static struct ffa_drv_info *drv_info;
@@ -597,6 +598,13 @@ static int ffa_features(u32 func_feat_id, u32 input_props,
 	return 0;
 }
 
+static void ffa_set_up_mem_ops_native_flag(void)
+{
+	if (!ffa_features(FFA_FN_NATIVE(MEM_LEND), 0, NULL, NULL) ||
+	    !ffa_features(FFA_FN_NATIVE(MEM_SHARE), 0, NULL, NULL))
+		drv_info->mem_ops_native = true;
+}
+
 static u32 ffa_api_version_get(void)
 {
 	return drv_info->version;
@@ -638,10 +646,10 @@ static int ffa_sync_send_receive(struct ffa_device *dev,
 static int
 ffa_memory_share(struct ffa_device *dev, struct ffa_mem_ops_args *args)
 {
-	if (dev->mode_32bit)
-		return ffa_memory_ops(FFA_MEM_SHARE, args);
+	if (drv_info->mem_ops_native)
+		return ffa_memory_ops(FFA_FN_NATIVE(MEM_SHARE), args);
 
-	return ffa_memory_ops(FFA_FN_NATIVE(MEM_SHARE), args);
+	return ffa_memory_ops(FFA_MEM_SHARE, args);
 }
 
 static int
@@ -654,10 +662,10 @@ ffa_memory_lend(struct ffa_device *dev, struct ffa_mem_ops_args *args)
 	 * however on systems without a hypervisor the responsibility
 	 * falls to the calling kernel driver to prevent access.
 	 */
-	if (dev->mode_32bit)
-		return ffa_memory_ops(FFA_MEM_LEND, args);
+	if (drv_info->mem_ops_native)
+		return ffa_memory_ops(FFA_FN_NATIVE(MEM_LEND), args);
 
-	return ffa_memory_ops(FFA_FN_NATIVE(MEM_LEND), args);
+	return ffa_memory_ops(FFA_MEM_LEND, args);
 }
 
 static const struct ffa_dev_ops ffa_ops = {
@@ -768,6 +776,8 @@ static int __init ffa_init(void)
 
 	ffa_setup_partitions();
 
+	ffa_set_up_mem_ops_native_flag();
+
 	return 0;
 free_pages:
 	if (drv_info->tx_buffer)
-- 
2.37.3


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH v2 06/10] firmware: arm_ffa: Make memory apis ffa_device independent

[Sudeep Holla]

There is a requirement to make memory APIs independent of the ffa_device.
One of the use-case is to have a common memory driver that manages the
memory for all the ffa_devices. That common memory driver won't be a
ffa_driver or won't have any ffa_device associated with it. So having
these memory APIs accessible without a ffa_device is needed and should
be possible as most of these are handled by the partition manager(SPM
or hypervisor).

Drop the ffa_device argument to the memory APIs and make them ffa_device
independent.

Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
---
 drivers/firmware/arm_ffa/driver.c | 6 ++----
 drivers/tee/optee/ffa_abi.c       | 2 +-
 include/linux/arm_ffa.h           | 6 ++----
 3 files changed, 5 insertions(+), 9 deletions(-)

diff --git a/drivers/firmware/arm_ffa/driver.c b/drivers/firmware/arm_ffa/driver.c
index 37a8ee304508..e4fd35773071 100644
--- a/drivers/firmware/arm_ffa/driver.c
+++ b/drivers/firmware/arm_ffa/driver.c
@@ -643,8 +643,7 @@ static int ffa_sync_send_receive(struct ffa_device *dev,
 				       dev->mode_32bit, data);
 }
 
-static int
-ffa_memory_share(struct ffa_device *dev, struct ffa_mem_ops_args *args)
+static int ffa_memory_share(struct ffa_mem_ops_args *args)
 {
 	if (drv_info->mem_ops_native)
 		return ffa_memory_ops(FFA_FN_NATIVE(MEM_SHARE), args);
@@ -652,8 +651,7 @@ ffa_memory_share(struct ffa_device *dev, struct ffa_mem_ops_args *args)
 	return ffa_memory_ops(FFA_MEM_SHARE, args);
 }
 
-static int
-ffa_memory_lend(struct ffa_device *dev, struct ffa_mem_ops_args *args)
+static int ffa_memory_lend(struct ffa_mem_ops_args *args)
 {
 	/* Note that upon a successful MEM_LEND request the caller
 	 * must ensure that the memory region specified is not accessed
diff --git a/drivers/tee/optee/ffa_abi.c b/drivers/tee/optee/ffa_abi.c
index 3d4079575ccd..7257b42d0545 100644
--- a/drivers/tee/optee/ffa_abi.c
+++ b/drivers/tee/optee/ffa_abi.c
@@ -294,7 +294,7 @@ static int optee_ffa_shm_register(struct tee_context *ctx, struct tee_shm *shm,
 	if (rc)
 		return rc;
 	args.sg = sgt.sgl;
-	rc = ffa_ops->memory_share(ffa_dev, &args);
+	rc = ffa_ops->memory_share(&args);
 	sg_free_table(&sgt);
 	if (rc)
 		return rc;
diff --git a/include/linux/arm_ffa.h b/include/linux/arm_ffa.h
index 556f50f27fb1..eafab07c9f58 100644
--- a/include/linux/arm_ffa.h
+++ b/include/linux/arm_ffa.h
@@ -262,10 +262,8 @@ struct ffa_dev_ops {
 	int (*sync_send_receive)(struct ffa_device *dev,
 				 struct ffa_send_direct_data *data);
 	int (*memory_reclaim)(u64 g_handle, u32 flags);
-	int (*memory_share)(struct ffa_device *dev,
-			    struct ffa_mem_ops_args *args);
-	int (*memory_lend)(struct ffa_device *dev,
-			   struct ffa_mem_ops_args *args);
+	int (*memory_share)(struct ffa_mem_ops_args *args);
+	int (*memory_lend)(struct ffa_mem_ops_args *args);
 };
 
 #endif /* _LINUX_ARM_FFA_H */
-- 
2.37.3


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Re: [PATCH v2 06/10] firmware: arm_ffa: Make memory apis ffa_device independent

[Jens Wiklander]

On Fri, Sep 2, 2022 at 2:40 PM Sudeep Holla <sudeep.holla@arm.com> wrote:
>
> There is a requirement to make memory APIs independent of the ffa_device.
> One of the use-case is to have a common memory driver that manages the
> memory for all the ffa_devices. That common memory driver won't be a
> ffa_driver or won't have any ffa_device associated with it. So having
> these memory APIs accessible without a ffa_device is needed and should
> be possible as most of these are handled by the partition manager(SPM
> or hypervisor).
>
> Drop the ffa_device argument to the memory APIs and make them ffa_device
> independent.
>
> Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
> ---
>  drivers/firmware/arm_ffa/driver.c | 6 ++----
>  drivers/tee/optee/ffa_abi.c       | 2 +-
>  include/linux/arm_ffa.h           | 6 ++----
>  3 files changed, 5 insertions(+), 9 deletions(-)

Acked-by: Jens Wiklander <jens.wiklander@linaro.org>

>
> diff --git a/drivers/firmware/arm_ffa/driver.c b/drivers/firmware/arm_ffa/driver.c
> index 37a8ee304508..e4fd35773071 100644
> --- a/drivers/firmware/arm_ffa/driver.c
> +++ b/drivers/firmware/arm_ffa/driver.c
> @@ -643,8 +643,7 @@ static int ffa_sync_send_receive(struct ffa_device *dev,
>                                        dev->mode_32bit, data);
>  }
>
> -static int
> -ffa_memory_share(struct ffa_device *dev, struct ffa_mem_ops_args *args)
> +static int ffa_memory_share(struct ffa_mem_ops_args *args)
>  {
>         if (drv_info->mem_ops_native)
>                 return ffa_memory_ops(FFA_FN_NATIVE(MEM_SHARE), args);
> @@ -652,8 +651,7 @@ ffa_memory_share(struct ffa_device *dev, struct ffa_mem_ops_args *args)
>         return ffa_memory_ops(FFA_MEM_SHARE, args);
>  }
>
> -static int
> -ffa_memory_lend(struct ffa_device *dev, struct ffa_mem_ops_args *args)
> +static int ffa_memory_lend(struct ffa_mem_ops_args *args)
>  {
>         /* Note that upon a successful MEM_LEND request the caller
>          * must ensure that the memory region specified is not accessed
> diff --git a/drivers/tee/optee/ffa_abi.c b/drivers/tee/optee/ffa_abi.c
> index 3d4079575ccd..7257b42d0545 100644
> --- a/drivers/tee/optee/ffa_abi.c
> +++ b/drivers/tee/optee/ffa_abi.c
> @@ -294,7 +294,7 @@ static int optee_ffa_shm_register(struct tee_context *ctx, struct tee_shm *shm,
>         if (rc)
>                 return rc;
>         args.sg = sgt.sgl;
> -       rc = ffa_ops->memory_share(ffa_dev, &args);
> +       rc = ffa_ops->memory_share(&args);
>         sg_free_table(&sgt);
>         if (rc)
>                 return rc;
> diff --git a/include/linux/arm_ffa.h b/include/linux/arm_ffa.h
> index 556f50f27fb1..eafab07c9f58 100644
> --- a/include/linux/arm_ffa.h
> +++ b/include/linux/arm_ffa.h
> @@ -262,10 +262,8 @@ struct ffa_dev_ops {
>         int (*sync_send_receive)(struct ffa_device *dev,
>                                  struct ffa_send_direct_data *data);
>         int (*memory_reclaim)(u64 g_handle, u32 flags);
> -       int (*memory_share)(struct ffa_device *dev,
> -                           struct ffa_mem_ops_args *args);
> -       int (*memory_lend)(struct ffa_device *dev,
> -                          struct ffa_mem_ops_args *args);
> +       int (*memory_share)(struct ffa_mem_ops_args *args);
> +       int (*memory_lend)(struct ffa_mem_ops_args *args);
>  };
>
>  #endif /* _LINUX_ARM_FFA_H */
> --
> 2.37.3
>

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH v2 07/10] firmware: arm_ffa: Rename ffa_dev_ops as ffa_ops

[Sudeep Holla]

Except the message APIs, all other APIs are ffa_device independent and can
be used without any associated ffa_device from a non ffa_driver.

In order to reflect the same, just rename ffa_dev_ops as ffa_ops to
avoid any confusion or to keep it simple.

Suggested-by: Sumit Garg <sumit.garg@linaro.org>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
---
 drivers/firmware/arm_ffa/bus.c    |  2 +-
 drivers/firmware/arm_ffa/driver.c |  2 +-
 drivers/tee/optee/ffa_abi.c       | 14 +++++++-------
 include/linux/arm_ffa.h           |  8 ++++----
 4 files changed, 13 insertions(+), 13 deletions(-)

diff --git a/drivers/firmware/arm_ffa/bus.c b/drivers/firmware/arm_ffa/bus.c
index 69328041fbc3..99d439480612 100644
--- a/drivers/firmware/arm_ffa/bus.c
+++ b/drivers/firmware/arm_ffa/bus.c
@@ -168,7 +168,7 @@ bool ffa_device_is_valid(struct ffa_device *ffa_dev)
 }
 
 struct ffa_device *ffa_device_register(const uuid_t *uuid, int vm_id,
-				       const struct ffa_dev_ops *ops)
+				       const struct ffa_ops *ops)
 {
 	int ret;
 	struct device *dev;
diff --git a/drivers/firmware/arm_ffa/driver.c b/drivers/firmware/arm_ffa/driver.c
index e4fd35773071..2532e0f16cc9 100644
--- a/drivers/firmware/arm_ffa/driver.c
+++ b/drivers/firmware/arm_ffa/driver.c
@@ -666,7 +666,7 @@ static int ffa_memory_lend(struct ffa_mem_ops_args *args)
 	return ffa_memory_ops(FFA_MEM_LEND, args);
 }
 
-static const struct ffa_dev_ops ffa_ops = {
+static const struct ffa_ops ffa_ops = {
 	.api_version_get = ffa_api_version_get,
 	.partition_info_get = ffa_partition_info_get,
 	.mode_32bit_set = ffa_mode_32bit_set,
diff --git a/drivers/tee/optee/ffa_abi.c b/drivers/tee/optee/ffa_abi.c
index 7257b42d0545..2ce5b87dfb27 100644
--- a/drivers/tee/optee/ffa_abi.c
+++ b/drivers/tee/optee/ffa_abi.c
@@ -272,7 +272,7 @@ static int optee_ffa_shm_register(struct tee_context *ctx, struct tee_shm *shm,
 {
 	struct optee *optee = tee_get_drvdata(ctx->teedev);
 	struct ffa_device *ffa_dev = optee->ffa.ffa_dev;
-	const struct ffa_dev_ops *ffa_ops = ffa_dev->ops;
+	const struct ffa_ops *ffa_ops = ffa_dev->ops;
 	struct ffa_mem_region_attributes mem_attr = {
 		.receiver = ffa_dev->vm_id,
 		.attrs = FFA_MEM_RW,
@@ -315,7 +315,7 @@ static int optee_ffa_shm_unregister(struct tee_context *ctx,
 {
 	struct optee *optee = tee_get_drvdata(ctx->teedev);
 	struct ffa_device *ffa_dev = optee->ffa.ffa_dev;
-	const struct ffa_dev_ops *ffa_ops = ffa_dev->ops;
+	const struct ffa_ops *ffa_ops = ffa_dev->ops;
 	u64 global_handle = shm->sec_world_id;
 	struct ffa_send_direct_data data = {
 		.data0 = OPTEE_FFA_UNREGISTER_SHM,
@@ -342,7 +342,7 @@ static int optee_ffa_shm_unregister_supp(struct tee_context *ctx,
 					 struct tee_shm *shm)
 {
 	struct optee *optee = tee_get_drvdata(ctx->teedev);
-	const struct ffa_dev_ops *ffa_ops = optee->ffa.ffa_dev->ops;
+	const struct ffa_ops *ffa_ops = optee->ffa.ffa_dev->ops;
 	u64 global_handle = shm->sec_world_id;
 	int rc;
 
@@ -530,7 +530,7 @@ static int optee_ffa_yielding_call(struct tee_context *ctx,
 {
 	struct optee *optee = tee_get_drvdata(ctx->teedev);
 	struct ffa_device *ffa_dev = optee->ffa.ffa_dev;
-	const struct ffa_dev_ops *ffa_ops = ffa_dev->ops;
+	const struct ffa_ops *ffa_ops = ffa_dev->ops;
 	struct optee_call_waiter w;
 	u32 cmd = data->data0;
 	u32 w4 = data->data1;
@@ -652,7 +652,7 @@ static int optee_ffa_do_call_with_arg(struct tee_context *ctx,
  */
 
 static bool optee_ffa_api_is_compatbile(struct ffa_device *ffa_dev,
-					const struct ffa_dev_ops *ops)
+					const struct ffa_ops *ops)
 {
 	struct ffa_send_direct_data data = { OPTEE_FFA_GET_API_VERSION };
 	int rc;
@@ -687,7 +687,7 @@ static bool optee_ffa_api_is_compatbile(struct ffa_device *ffa_dev,
 }
 
 static bool optee_ffa_exchange_caps(struct ffa_device *ffa_dev,
-				    const struct ffa_dev_ops *ops,
+				    const struct ffa_ops *ops,
 				    u32 *sec_caps,
 				    unsigned int *rpc_param_count)
 {
@@ -783,7 +783,7 @@ static void optee_ffa_remove(struct ffa_device *ffa_dev)
 
 static int optee_ffa_probe(struct ffa_device *ffa_dev)
 {
-	const struct ffa_dev_ops *ffa_ops;
+	const struct ffa_ops *ffa_ops;
 	unsigned int rpc_param_count;
 	struct tee_shm_pool *pool;
 	struct tee_device *teedev;
diff --git a/include/linux/arm_ffa.h b/include/linux/arm_ffa.h
index eafab07c9f58..4c4b06783035 100644
--- a/include/linux/arm_ffa.h
+++ b/include/linux/arm_ffa.h
@@ -17,7 +17,7 @@ struct ffa_device {
 	bool mode_32bit;
 	uuid_t uuid;
 	struct device dev;
-	const struct ffa_dev_ops *ops;
+	const struct ffa_ops *ops;
 };
 
 #define to_ffa_dev(d) container_of(d, struct ffa_device, dev)
@@ -49,7 +49,7 @@ static inline void *ffa_dev_get_drvdata(struct ffa_device *fdev)
 
 #if IS_REACHABLE(CONFIG_ARM_FFA_TRANSPORT)
 struct ffa_device *ffa_device_register(const uuid_t *uuid, int vm_id,
-				       const struct ffa_dev_ops *ops);
+				       const struct ffa_ops *ops);
 void ffa_device_unregister(struct ffa_device *ffa_dev);
 int ffa_driver_register(struct ffa_driver *driver, struct module *owner,
 			const char *mod_name);
@@ -59,7 +59,7 @@ bool ffa_device_is_valid(struct ffa_device *ffa_dev);
 #else
 static inline
 struct ffa_device *ffa_device_register(const uuid_t *uuid, int vm_id,
-				       const struct ffa_dev_ops *ops)
+				       const struct ffa_ops *ops)
 {
 	return NULL;
 }
@@ -254,7 +254,7 @@ struct ffa_mem_ops_args {
 	struct ffa_mem_region_attributes *attrs;
 };
 
-struct ffa_dev_ops {
+struct ffa_ops {
 	u32 (*api_version_get)(void);
 	int (*partition_info_get)(const char *uuid_str,
 				  struct ffa_partition_info *buffer);
-- 
2.37.3


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Re: [PATCH v2 07/10] firmware: arm_ffa: Rename ffa_dev_ops as ffa_ops

[Sumit Garg]

On Fri, 2 Sept 2022 at 18:10, Sudeep Holla <sudeep.holla@arm.com> wrote:
>
> Except the message APIs, all other APIs are ffa_device independent and can
> be used without any associated ffa_device from a non ffa_driver.
>
> In order to reflect the same, just rename ffa_dev_ops as ffa_ops to
> avoid any confusion or to keep it simple.
>
> Suggested-by: Sumit Garg <sumit.garg@linaro.org>
> Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
> ---
>  drivers/firmware/arm_ffa/bus.c    |  2 +-
>  drivers/firmware/arm_ffa/driver.c |  2 +-
>  drivers/tee/optee/ffa_abi.c       | 14 +++++++-------
>  include/linux/arm_ffa.h           |  8 ++++----
>  4 files changed, 13 insertions(+), 13 deletions(-)
>

Reviewed-by: Sumit Garg <sumit.garg@linaro.org>

-Sumit

> diff --git a/drivers/firmware/arm_ffa/bus.c b/drivers/firmware/arm_ffa/bus.c
> index 69328041fbc3..99d439480612 100644
> --- a/drivers/firmware/arm_ffa/bus.c
> +++ b/drivers/firmware/arm_ffa/bus.c
> @@ -168,7 +168,7 @@ bool ffa_device_is_valid(struct ffa_device *ffa_dev)
>  }
>
>  struct ffa_device *ffa_device_register(const uuid_t *uuid, int vm_id,
> -                                      const struct ffa_dev_ops *ops)
> +                                      const struct ffa_ops *ops)
>  {
>         int ret;
>         struct device *dev;
> diff --git a/drivers/firmware/arm_ffa/driver.c b/drivers/firmware/arm_ffa/driver.c
> index e4fd35773071..2532e0f16cc9 100644
> --- a/drivers/firmware/arm_ffa/driver.c
> +++ b/drivers/firmware/arm_ffa/driver.c
> @@ -666,7 +666,7 @@ static int ffa_memory_lend(struct ffa_mem_ops_args *args)
>         return ffa_memory_ops(FFA_MEM_LEND, args);
>  }
>
> -static const struct ffa_dev_ops ffa_ops = {
> +static const struct ffa_ops ffa_ops = {
>         .api_version_get = ffa_api_version_get,
>         .partition_info_get = ffa_partition_info_get,
>         .mode_32bit_set = ffa_mode_32bit_set,
> diff --git a/drivers/tee/optee/ffa_abi.c b/drivers/tee/optee/ffa_abi.c
> index 7257b42d0545..2ce5b87dfb27 100644
> --- a/drivers/tee/optee/ffa_abi.c
> +++ b/drivers/tee/optee/ffa_abi.c
> @@ -272,7 +272,7 @@ static int optee_ffa_shm_register(struct tee_context *ctx, struct tee_shm *shm,
>  {
>         struct optee *optee = tee_get_drvdata(ctx->teedev);
>         struct ffa_device *ffa_dev = optee->ffa.ffa_dev;
> -       const struct ffa_dev_ops *ffa_ops = ffa_dev->ops;
> +       const struct ffa_ops *ffa_ops = ffa_dev->ops;
>         struct ffa_mem_region_attributes mem_attr = {
>                 .receiver = ffa_dev->vm_id,
>                 .attrs = FFA_MEM_RW,
> @@ -315,7 +315,7 @@ static int optee_ffa_shm_unregister(struct tee_context *ctx,
>  {
>         struct optee *optee = tee_get_drvdata(ctx->teedev);
>         struct ffa_device *ffa_dev = optee->ffa.ffa_dev;
> -       const struct ffa_dev_ops *ffa_ops = ffa_dev->ops;
> +       const struct ffa_ops *ffa_ops = ffa_dev->ops;
>         u64 global_handle = shm->sec_world_id;
>         struct ffa_send_direct_data data = {
>                 .data0 = OPTEE_FFA_UNREGISTER_SHM,
> @@ -342,7 +342,7 @@ static int optee_ffa_shm_unregister_supp(struct tee_context *ctx,
>                                          struct tee_shm *shm)
>  {
>         struct optee *optee = tee_get_drvdata(ctx->teedev);
> -       const struct ffa_dev_ops *ffa_ops = optee->ffa.ffa_dev->ops;
> +       const struct ffa_ops *ffa_ops = optee->ffa.ffa_dev->ops;
>         u64 global_handle = shm->sec_world_id;
>         int rc;
>
> @@ -530,7 +530,7 @@ static int optee_ffa_yielding_call(struct tee_context *ctx,
>  {
>         struct optee *optee = tee_get_drvdata(ctx->teedev);
>         struct ffa_device *ffa_dev = optee->ffa.ffa_dev;
> -       const struct ffa_dev_ops *ffa_ops = ffa_dev->ops;
> +       const struct ffa_ops *ffa_ops = ffa_dev->ops;
>         struct optee_call_waiter w;
>         u32 cmd = data->data0;
>         u32 w4 = data->data1;
> @@ -652,7 +652,7 @@ static int optee_ffa_do_call_with_arg(struct tee_context *ctx,
>   */
>
>  static bool optee_ffa_api_is_compatbile(struct ffa_device *ffa_dev,
> -                                       const struct ffa_dev_ops *ops)
> +                                       const struct ffa_ops *ops)
>  {
>         struct ffa_send_direct_data data = { OPTEE_FFA_GET_API_VERSION };
>         int rc;
> @@ -687,7 +687,7 @@ static bool optee_ffa_api_is_compatbile(struct ffa_device *ffa_dev,
>  }
>
>  static bool optee_ffa_exchange_caps(struct ffa_device *ffa_dev,
> -                                   const struct ffa_dev_ops *ops,
> +                                   const struct ffa_ops *ops,
>                                     u32 *sec_caps,
>                                     unsigned int *rpc_param_count)
>  {
> @@ -783,7 +783,7 @@ static void optee_ffa_remove(struct ffa_device *ffa_dev)
>
>  static int optee_ffa_probe(struct ffa_device *ffa_dev)
>  {
> -       const struct ffa_dev_ops *ffa_ops;
> +       const struct ffa_ops *ffa_ops;
>         unsigned int rpc_param_count;
>         struct tee_shm_pool *pool;
>         struct tee_device *teedev;
> diff --git a/include/linux/arm_ffa.h b/include/linux/arm_ffa.h
> index eafab07c9f58..4c4b06783035 100644
> --- a/include/linux/arm_ffa.h
> +++ b/include/linux/arm_ffa.h
> @@ -17,7 +17,7 @@ struct ffa_device {
>         bool mode_32bit;
>         uuid_t uuid;
>         struct device dev;
> -       const struct ffa_dev_ops *ops;
> +       const struct ffa_ops *ops;
>  };
>
>  #define to_ffa_dev(d) container_of(d, struct ffa_device, dev)
> @@ -49,7 +49,7 @@ static inline void *ffa_dev_get_drvdata(struct ffa_device *fdev)
>
>  #if IS_REACHABLE(CONFIG_ARM_FFA_TRANSPORT)
>  struct ffa_device *ffa_device_register(const uuid_t *uuid, int vm_id,
> -                                      const struct ffa_dev_ops *ops);
> +                                      const struct ffa_ops *ops);
>  void ffa_device_unregister(struct ffa_device *ffa_dev);
>  int ffa_driver_register(struct ffa_driver *driver, struct module *owner,
>                         const char *mod_name);
> @@ -59,7 +59,7 @@ bool ffa_device_is_valid(struct ffa_device *ffa_dev);
>  #else
>  static inline
>  struct ffa_device *ffa_device_register(const uuid_t *uuid, int vm_id,
> -                                      const struct ffa_dev_ops *ops)
> +                                      const struct ffa_ops *ops)
>  {
>         return NULL;
>  }
> @@ -254,7 +254,7 @@ struct ffa_mem_ops_args {
>         struct ffa_mem_region_attributes *attrs;
>  };
>
> -struct ffa_dev_ops {
> +struct ffa_ops {
>         u32 (*api_version_get)(void);
>         int (*partition_info_get)(const char *uuid_str,
>                                   struct ffa_partition_info *buffer);
> --
> 2.37.3
>

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Re: [PATCH v2 07/10] firmware: arm_ffa: Rename ffa_dev_ops as ffa_ops

[Jens Wiklander]

On Fri, Sep 2, 2022 at 2:40 PM Sudeep Holla <sudeep.holla@arm.com> wrote:
>
> Except the message APIs, all other APIs are ffa_device independent and can
> be used without any associated ffa_device from a non ffa_driver.
>
> In order to reflect the same, just rename ffa_dev_ops as ffa_ops to
> avoid any confusion or to keep it simple.
>
> Suggested-by: Sumit Garg <sumit.garg@linaro.org>
> Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
> ---
>  drivers/firmware/arm_ffa/bus.c    |  2 +-
>  drivers/firmware/arm_ffa/driver.c |  2 +-
>  drivers/tee/optee/ffa_abi.c       | 14 +++++++-------
>  include/linux/arm_ffa.h           |  8 ++++----
>  4 files changed, 13 insertions(+), 13 deletions(-)

Reviewed-by: Jens Wiklander <jens.wiklander@linaro.org>

>
> diff --git a/drivers/firmware/arm_ffa/bus.c b/drivers/firmware/arm_ffa/bus.c
> index 69328041fbc3..99d439480612 100644
> --- a/drivers/firmware/arm_ffa/bus.c
> +++ b/drivers/firmware/arm_ffa/bus.c
> @@ -168,7 +168,7 @@ bool ffa_device_is_valid(struct ffa_device *ffa_dev)
>  }
>
>  struct ffa_device *ffa_device_register(const uuid_t *uuid, int vm_id,
> -                                      const struct ffa_dev_ops *ops)
> +                                      const struct ffa_ops *ops)
>  {
>         int ret;
>         struct device *dev;
> diff --git a/drivers/firmware/arm_ffa/driver.c b/drivers/firmware/arm_ffa/driver.c
> index e4fd35773071..2532e0f16cc9 100644
> --- a/drivers/firmware/arm_ffa/driver.c
> +++ b/drivers/firmware/arm_ffa/driver.c
> @@ -666,7 +666,7 @@ static int ffa_memory_lend(struct ffa_mem_ops_args *args)
>         return ffa_memory_ops(FFA_MEM_LEND, args);
>  }
>
> -static const struct ffa_dev_ops ffa_ops = {
> +static const struct ffa_ops ffa_ops = {
>         .api_version_get = ffa_api_version_get,
>         .partition_info_get = ffa_partition_info_get,
>         .mode_32bit_set = ffa_mode_32bit_set,
> diff --git a/drivers/tee/optee/ffa_abi.c b/drivers/tee/optee/ffa_abi.c
> index 7257b42d0545..2ce5b87dfb27 100644
> --- a/drivers/tee/optee/ffa_abi.c
> +++ b/drivers/tee/optee/ffa_abi.c
> @@ -272,7 +272,7 @@ static int optee_ffa_shm_register(struct tee_context *ctx, struct tee_shm *shm,
>  {
>         struct optee *optee = tee_get_drvdata(ctx->teedev);
>         struct ffa_device *ffa_dev = optee->ffa.ffa_dev;
> -       const struct ffa_dev_ops *ffa_ops = ffa_dev->ops;
> +       const struct ffa_ops *ffa_ops = ffa_dev->ops;
>         struct ffa_mem_region_attributes mem_attr = {
>                 .receiver = ffa_dev->vm_id,
>                 .attrs = FFA_MEM_RW,
> @@ -315,7 +315,7 @@ static int optee_ffa_shm_unregister(struct tee_context *ctx,
>  {
>         struct optee *optee = tee_get_drvdata(ctx->teedev);
>         struct ffa_device *ffa_dev = optee->ffa.ffa_dev;
> -       const struct ffa_dev_ops *ffa_ops = ffa_dev->ops;
> +       const struct ffa_ops *ffa_ops = ffa_dev->ops;
>         u64 global_handle = shm->sec_world_id;
>         struct ffa_send_direct_data data = {
>                 .data0 = OPTEE_FFA_UNREGISTER_SHM,
> @@ -342,7 +342,7 @@ static int optee_ffa_shm_unregister_supp(struct tee_context *ctx,
>                                          struct tee_shm *shm)
>  {
>         struct optee *optee = tee_get_drvdata(ctx->teedev);
> -       const struct ffa_dev_ops *ffa_ops = optee->ffa.ffa_dev->ops;
> +       const struct ffa_ops *ffa_ops = optee->ffa.ffa_dev->ops;
>         u64 global_handle = shm->sec_world_id;
>         int rc;
>
> @@ -530,7 +530,7 @@ static int optee_ffa_yielding_call(struct tee_context *ctx,
>  {
>         struct optee *optee = tee_get_drvdata(ctx->teedev);
>         struct ffa_device *ffa_dev = optee->ffa.ffa_dev;
> -       const struct ffa_dev_ops *ffa_ops = ffa_dev->ops;
> +       const struct ffa_ops *ffa_ops = ffa_dev->ops;
>         struct optee_call_waiter w;
>         u32 cmd = data->data0;
>         u32 w4 = data->data1;
> @@ -652,7 +652,7 @@ static int optee_ffa_do_call_with_arg(struct tee_context *ctx,
>   */
>
>  static bool optee_ffa_api_is_compatbile(struct ffa_device *ffa_dev,
> -                                       const struct ffa_dev_ops *ops)
> +                                       const struct ffa_ops *ops)
>  {
>         struct ffa_send_direct_data data = { OPTEE_FFA_GET_API_VERSION };
>         int rc;
> @@ -687,7 +687,7 @@ static bool optee_ffa_api_is_compatbile(struct ffa_device *ffa_dev,
>  }
>
>  static bool optee_ffa_exchange_caps(struct ffa_device *ffa_dev,
> -                                   const struct ffa_dev_ops *ops,
> +                                   const struct ffa_ops *ops,
>                                     u32 *sec_caps,
>                                     unsigned int *rpc_param_count)
>  {
> @@ -783,7 +783,7 @@ static void optee_ffa_remove(struct ffa_device *ffa_dev)
>
>  static int optee_ffa_probe(struct ffa_device *ffa_dev)
>  {
> -       const struct ffa_dev_ops *ffa_ops;
> +       const struct ffa_ops *ffa_ops;
>         unsigned int rpc_param_count;
>         struct tee_shm_pool *pool;
>         struct tee_device *teedev;
> diff --git a/include/linux/arm_ffa.h b/include/linux/arm_ffa.h
> index eafab07c9f58..4c4b06783035 100644
> --- a/include/linux/arm_ffa.h
> +++ b/include/linux/arm_ffa.h
> @@ -17,7 +17,7 @@ struct ffa_device {
>         bool mode_32bit;
>         uuid_t uuid;
>         struct device dev;
> -       const struct ffa_dev_ops *ops;
> +       const struct ffa_ops *ops;
>  };
>
>  #define to_ffa_dev(d) container_of(d, struct ffa_device, dev)
> @@ -49,7 +49,7 @@ static inline void *ffa_dev_get_drvdata(struct ffa_device *fdev)
>
>  #if IS_REACHABLE(CONFIG_ARM_FFA_TRANSPORT)
>  struct ffa_device *ffa_device_register(const uuid_t *uuid, int vm_id,
> -                                      const struct ffa_dev_ops *ops);
> +                                      const struct ffa_ops *ops);
>  void ffa_device_unregister(struct ffa_device *ffa_dev);
>  int ffa_driver_register(struct ffa_driver *driver, struct module *owner,
>                         const char *mod_name);
> @@ -59,7 +59,7 @@ bool ffa_device_is_valid(struct ffa_device *ffa_dev);
>  #else
>  static inline
>  struct ffa_device *ffa_device_register(const uuid_t *uuid, int vm_id,
> -                                      const struct ffa_dev_ops *ops)
> +                                      const struct ffa_ops *ops)
>  {
>         return NULL;
>  }
> @@ -254,7 +254,7 @@ struct ffa_mem_ops_args {
>         struct ffa_mem_region_attributes *attrs;
>  };
>
> -struct ffa_dev_ops {
> +struct ffa_ops {
>         u32 (*api_version_get)(void);
>         int (*partition_info_get)(const char *uuid_str,
>                                   struct ffa_partition_info *buffer);
> --
> 2.37.3
>

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH v2 08/10] firmware: arm_ffa: Add v1.1 get_partition_info support

[Sudeep Holla]

FF-A v1.1 adds support to discovery the UUIDs of the partitions that was
missing in v1.0 and which the driver workarounds by using UUIDs supplied
by the ffa_drivers.

Add the v1.1 get_partition_info support and disable the workaround if
the detected FF-A version is greater than v1.0.

Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
---
 drivers/firmware/arm_ffa/driver.c | 42 +++++++++++++++++++++++++------
 include/linux/arm_ffa.h           |  1 +
 2 files changed, 35 insertions(+), 8 deletions(-)

diff --git a/drivers/firmware/arm_ffa/driver.c b/drivers/firmware/arm_ffa/driver.c
index 2532e0f16cc9..dd6ab2f81580 100644
--- a/drivers/firmware/arm_ffa/driver.c
+++ b/drivers/firmware/arm_ffa/driver.c
@@ -264,18 +264,24 @@ static int ffa_rxtx_unmap(u16 vm_id)
 	return 0;
 }
 
+#define PARTITION_INFO_GET_RETURN_COUNT_ONLY	BIT(0)
+
 /* buffer must be sizeof(struct ffa_partition_info) * num_partitions */
 static int
 __ffa_partition_info_get(u32 uuid0, u32 uuid1, u32 uuid2, u32 uuid3,
 			 struct ffa_partition_info *buffer, int num_partitions)
 {
-	int count;
+	int idx, count, flags = 0, size;
 	ffa_value_t partition_info;
 
+	if (!buffer || !num_partitions) /* Just get the count for now */
+		flags = PARTITION_INFO_GET_RETURN_COUNT_ONLY;
+
 	mutex_lock(&drv_info->rx_lock);
 	invoke_ffa_fn((ffa_value_t){
 		      .a0 = FFA_PARTITION_INFO_GET,
 		      .a1 = uuid0, .a2 = uuid1, .a3 = uuid2, .a4 = uuid3,
+		      .a5 = flags,
 		      }, &partition_info);
 
 	if (partition_info.a0 == FFA_ERROR) {
@@ -285,8 +291,18 @@ __ffa_partition_info_get(u32 uuid0, u32 uuid1, u32 uuid2, u32 uuid3,
 
 	count = partition_info.a2;
 
+	if (drv_info->version > FFA_VERSION_1_0) {
+		size = partition_info.a3;
+		if (size > sizeof(*buffer))
+			size = sizeof(*buffer);
+	} else {
+		size = 8; /* FFA_VERSION_1_0 lacks size in the response */
+	}
+
 	if (buffer && count <= num_partitions)
-		memcpy(buffer, drv_info->rx_buffer, sizeof(*buffer) * count);
+		for (idx = 0; idx < count; idx++)
+			memcpy(buffer + idx, drv_info->rx_buffer + idx * size,
+			       size);
 
 	ffa_rx_release();
 
@@ -681,6 +697,14 @@ void ffa_device_match_uuid(struct ffa_device *ffa_dev, const uuid_t *uuid)
 	int count, idx;
 	struct ffa_partition_info *pbuf, *tpbuf;
 
+	/*
+	 * FF-A v1.1 provides UUID for each partition as part of the discovery
+	 * API, the discovered UUID must be populated in the device's UUID and
+	 * there is no need to copy the same from the driver table.
+	 */
+	if (drv_info->version > FFA_VERSION_1_0)
+		return;
+
 	count = ffa_partition_probe(uuid, &pbuf);
 	if (count <= 0)
 		return;
@@ -694,6 +718,7 @@ void ffa_device_match_uuid(struct ffa_device *ffa_dev, const uuid_t *uuid)
 static void ffa_setup_partitions(void)
 {
 	int count, idx;
+	uuid_t uuid;
 	struct ffa_device *ffa_dev;
 	struct ffa_partition_info *pbuf, *tpbuf;
 
@@ -704,14 +729,15 @@ static void ffa_setup_partitions(void)
 	}
 
 	for (idx = 0, tpbuf = pbuf; idx < count; idx++, tpbuf++) {
-		/* Note that the &uuid_null parameter will require
+		import_uuid(&uuid, (u8 *)tpbuf->uuid);
+
+		/* Note that if the UUID will be uuid_null, that will require
 		 * ffa_device_match() to find the UUID of this partition id
-		 * with help of ffa_device_match_uuid(). Once the FF-A spec
-		 * is updated to provide correct UUID here for each partition
-		 * as part of the discovery API, we need to pass the
-		 * discovered UUID here instead.
+		 * with help of ffa_device_match_uuid(). FF-A v1.1 and above
+		 * provides UUID here for each partition as part of the
+		 * discovery API and the same is passed.
 		 */
-		ffa_dev = ffa_device_register(&uuid_null, tpbuf->id, &ffa_ops);
+		ffa_dev = ffa_device_register(&uuid, tpbuf->id, &ffa_ops);
 		if (!ffa_dev) {
 			pr_err("%s: failed to register partition ID 0x%x\n",
 			       __func__, tpbuf->id);
diff --git a/include/linux/arm_ffa.h b/include/linux/arm_ffa.h
index 4c4b06783035..09567ffd1f49 100644
--- a/include/linux/arm_ffa.h
+++ b/include/linux/arm_ffa.h
@@ -107,6 +107,7 @@ struct ffa_partition_info {
 /* partition can send and receive indirect messages. */
 #define FFA_PARTITION_INDIRECT_MSG	BIT(2)
 	u32 properties;
+	u32 uuid[4];
 };
 
 /* For use with FFA_MSG_SEND_DIRECT_{REQ,RESP} which pass data via registers */
-- 
2.37.3


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Re: [PATCH v2 08/10] firmware: arm_ffa: Add v1.1 get_partition_info support

[Jens Wiklander]

On Fri, Sep 2, 2022 at 2:40 PM Sudeep Holla <sudeep.holla@arm.com> wrote:
>
> FF-A v1.1 adds support to discovery the UUIDs of the partitions that was
> missing in v1.0 and which the driver workarounds by using UUIDs supplied
> by the ffa_drivers.
>
> Add the v1.1 get_partition_info support and disable the workaround if
> the detected FF-A version is greater than v1.0.
>
> Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
> ---
>  drivers/firmware/arm_ffa/driver.c | 42 +++++++++++++++++++++++++------
>  include/linux/arm_ffa.h           |  1 +
>  2 files changed, 35 insertions(+), 8 deletions(-)
>
> diff --git a/drivers/firmware/arm_ffa/driver.c b/drivers/firmware/arm_ffa/driver.c
> index 2532e0f16cc9..dd6ab2f81580 100644
> --- a/drivers/firmware/arm_ffa/driver.c
> +++ b/drivers/firmware/arm_ffa/driver.c
> @@ -264,18 +264,24 @@ static int ffa_rxtx_unmap(u16 vm_id)
>         return 0;
>  }
>
> +#define PARTITION_INFO_GET_RETURN_COUNT_ONLY   BIT(0)
> +
>  /* buffer must be sizeof(struct ffa_partition_info) * num_partitions */
>  static int
>  __ffa_partition_info_get(u32 uuid0, u32 uuid1, u32 uuid2, u32 uuid3,
>                          struct ffa_partition_info *buffer, int num_partitions)
>  {
> -       int count;
> +       int idx, count, flags = 0, size;
>         ffa_value_t partition_info;
>
> +       if (!buffer || !num_partitions) /* Just get the count for now */
> +               flags = PARTITION_INFO_GET_RETURN_COUNT_ONLY;
> +
>         mutex_lock(&drv_info->rx_lock);
>         invoke_ffa_fn((ffa_value_t){
>                       .a0 = FFA_PARTITION_INFO_GET,
>                       .a1 = uuid0, .a2 = uuid1, .a3 = uuid2, .a4 = uuid3,
> +                     .a5 = flags,
>                       }, &partition_info);
>
>         if (partition_info.a0 == FFA_ERROR) {
> @@ -285,8 +291,18 @@ __ffa_partition_info_get(u32 uuid0, u32 uuid1, u32 uuid2, u32 uuid3,
>
>         count = partition_info.a2;
>
> +       if (drv_info->version > FFA_VERSION_1_0) {
> +               size = partition_info.a3;
> +               if (size > sizeof(*buffer))
> +                       size = sizeof(*buffer);

Below when calculating the address in the source buffer with
"drv_info->rx_buffer + idx * size" you should use the size from
partition_info.a3 even if it's larger than sizeof(*buffer). The amount
of bytes to copy looks correct though. Does that make sense?

Cheers,
Jens

> +       } else {
> +               size = 8; /* FFA_VERSION_1_0 lacks size in the response */
> +       }
> +
>         if (buffer && count <= num_partitions)
> -               memcpy(buffer, drv_info->rx_buffer, sizeof(*buffer) * count);
> +               for (idx = 0; idx < count; idx++)
> +                       memcpy(buffer + idx, drv_info->rx_buffer + idx * size,
> +                              size);
>
>         ffa_rx_release();
>
> @@ -681,6 +697,14 @@ void ffa_device_match_uuid(struct ffa_device *ffa_dev, const uuid_t *uuid)
>         int count, idx;
>         struct ffa_partition_info *pbuf, *tpbuf;
>
> +       /*
> +        * FF-A v1.1 provides UUID for each partition as part of the discovery
> +        * API, the discovered UUID must be populated in the device's UUID and
> +        * there is no need to copy the same from the driver table.
> +        */
> +       if (drv_info->version > FFA_VERSION_1_0)
> +               return;
> +
>         count = ffa_partition_probe(uuid, &pbuf);
>         if (count <= 0)
>                 return;
> @@ -694,6 +718,7 @@ void ffa_device_match_uuid(struct ffa_device *ffa_dev, const uuid_t *uuid)
>  static void ffa_setup_partitions(void)
>  {
>         int count, idx;
> +       uuid_t uuid;
>         struct ffa_device *ffa_dev;
>         struct ffa_partition_info *pbuf, *tpbuf;
>
> @@ -704,14 +729,15 @@ static void ffa_setup_partitions(void)
>         }
>
>         for (idx = 0, tpbuf = pbuf; idx < count; idx++, tpbuf++) {
> -               /* Note that the &uuid_null parameter will require
> +               import_uuid(&uuid, (u8 *)tpbuf->uuid);
> +
> +               /* Note that if the UUID will be uuid_null, that will require
>                  * ffa_device_match() to find the UUID of this partition id
> -                * with help of ffa_device_match_uuid(). Once the FF-A spec
> -                * is updated to provide correct UUID here for each partition
> -                * as part of the discovery API, we need to pass the
> -                * discovered UUID here instead.
> +                * with help of ffa_device_match_uuid(). FF-A v1.1 and above
> +                * provides UUID here for each partition as part of the
> +                * discovery API and the same is passed.
>                  */
> -               ffa_dev = ffa_device_register(&uuid_null, tpbuf->id, &ffa_ops);
> +               ffa_dev = ffa_device_register(&uuid, tpbuf->id, &ffa_ops);
>                 if (!ffa_dev) {
>                         pr_err("%s: failed to register partition ID 0x%x\n",
>                                __func__, tpbuf->id);
> diff --git a/include/linux/arm_ffa.h b/include/linux/arm_ffa.h
> index 4c4b06783035..09567ffd1f49 100644
> --- a/include/linux/arm_ffa.h
> +++ b/include/linux/arm_ffa.h
> @@ -107,6 +107,7 @@ struct ffa_partition_info {
>  /* partition can send and receive indirect messages. */
>  #define FFA_PARTITION_INDIRECT_MSG     BIT(2)
>         u32 properties;
> +       u32 uuid[4];
>  };
>
>  /* For use with FFA_MSG_SEND_DIRECT_{REQ,RESP} which pass data via registers */
> --
> 2.37.3
>

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Re: [PATCH v2 08/10] firmware: arm_ffa: Add v1.1 get_partition_info support

[Sudeep Holla]

On Wed, Sep 07, 2022 at 10:38:05AM +0200, Jens Wiklander wrote:
> On Fri, Sep 2, 2022 at 2:40 PM Sudeep Holla <sudeep.holla@arm.com> wrote:
> >
> > FF-A v1.1 adds support to discovery the UUIDs of the partitions that was
> > missing in v1.0 and which the driver workarounds by using UUIDs supplied
> > by the ffa_drivers.
> >
> > Add the v1.1 get_partition_info support and disable the workaround if
> > the detected FF-A version is greater than v1.0.
> >
> > Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
> > ---
> >  drivers/firmware/arm_ffa/driver.c | 42 +++++++++++++++++++++++++------
> >  include/linux/arm_ffa.h           |  1 +
> >  2 files changed, 35 insertions(+), 8 deletions(-)
> >
> > diff --git a/drivers/firmware/arm_ffa/driver.c b/drivers/firmware/arm_ffa/driver.c
> > index 2532e0f16cc9..dd6ab2f81580 100644
> > --- a/drivers/firmware/arm_ffa/driver.c
> > +++ b/drivers/firmware/arm_ffa/driver.c
> > @@ -264,18 +264,24 @@ static int ffa_rxtx_unmap(u16 vm_id)
> >         return 0;
> >  }
> >
> > +#define PARTITION_INFO_GET_RETURN_COUNT_ONLY   BIT(0)
> > +
> >  /* buffer must be sizeof(struct ffa_partition_info) * num_partitions */
> >  static int
> >  __ffa_partition_info_get(u32 uuid0, u32 uuid1, u32 uuid2, u32 uuid3,
> >                          struct ffa_partition_info *buffer, int num_partitions)
> >  {
> > -       int count;
> > +       int idx, count, flags = 0, size;
> >         ffa_value_t partition_info;
> >
> > +       if (!buffer || !num_partitions) /* Just get the count for now */
> > +               flags = PARTITION_INFO_GET_RETURN_COUNT_ONLY;
> > +
> >         mutex_lock(&drv_info->rx_lock);
> >         invoke_ffa_fn((ffa_value_t){
> >                       .a0 = FFA_PARTITION_INFO_GET,
> >                       .a1 = uuid0, .a2 = uuid1, .a3 = uuid2, .a4 = uuid3,
> > +                     .a5 = flags,
> >                       }, &partition_info);
> >
> >         if (partition_info.a0 == FFA_ERROR) {
> > @@ -285,8 +291,18 @@ __ffa_partition_info_get(u32 uuid0, u32 uuid1, u32 uuid2, u32 uuid3,
> >
> >         count = partition_info.a2;
> >
> > +       if (drv_info->version > FFA_VERSION_1_0) {
> > +               size = partition_info.a3;
> > +               if (size > sizeof(*buffer))
> > +                       size = sizeof(*buffer);
> 
> Below when calculating the address in the source buffer with
> "drv_info->rx_buffer + idx * size" you should use the size from
> partition_info.a3 even if it's larger than sizeof(*buffer). The amount
> of bytes to copy looks correct though. Does that make sense?
>

Good spot, I missed to see that. It makes complete sense. I will fix it.

-- 
Regards,
Sudeep

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH v2 09/10] firmware: arm_ffa: Set up 32bit execution mode flag using partiion property

[Sudeep Holla]

FF-A v1.1 adds a flag in the partition properties to indicate if the
partition runs in the AArch32 or AArch64 execution state. Use the same
to set-up the 32-bit execution flag mode in the ffa_dev automatically
and ignore any requests to do the same from the ffa_driver.

Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
---
 drivers/firmware/arm_ffa/driver.c | 13 ++++++++++++-
 include/linux/arm_ffa.h           |  2 ++
 2 files changed, 14 insertions(+), 1 deletion(-)

diff --git a/drivers/firmware/arm_ffa/driver.c b/drivers/firmware/arm_ffa/driver.c
index dd6ab2f81580..dbbe15592173 100644
--- a/drivers/firmware/arm_ffa/driver.c
+++ b/drivers/firmware/arm_ffa/driver.c
@@ -647,11 +647,19 @@ static int ffa_partition_info_get(const char *uuid_str,
 	return 0;
 }
 
-static void ffa_mode_32bit_set(struct ffa_device *dev)
+static void _ffa_mode_32bit_set(struct ffa_device *dev)
 {
 	dev->mode_32bit = true;
 }
 
+static void ffa_mode_32bit_set(struct ffa_device *dev)
+{
+	if (drv_info->version > FFA_VERSION_1_0)
+		return;
+
+	_ffa_mode_32bit_set(dev);
+}
+
 static int ffa_sync_send_receive(struct ffa_device *dev,
 				 struct ffa_send_direct_data *data)
 {
@@ -743,6 +751,9 @@ static void ffa_setup_partitions(void)
 			       __func__, tpbuf->id);
 			continue;
 		}
+
+		if (tpbuf->properties & FFA_PARTITION_AARCH64_EXEC)
+			_ffa_mode_32bit_set(ffa_dev);
 	}
 	kfree(pbuf);
 }
diff --git a/include/linux/arm_ffa.h b/include/linux/arm_ffa.h
index 09567ffd1f49..5964b6104996 100644
--- a/include/linux/arm_ffa.h
+++ b/include/linux/arm_ffa.h
@@ -106,6 +106,8 @@ struct ffa_partition_info {
 #define FFA_PARTITION_DIRECT_SEND	BIT(1)
 /* partition can send and receive indirect messages. */
 #define FFA_PARTITION_INDIRECT_MSG	BIT(2)
+/* partition runs in the AArch64 execution state. */
+#define FFA_PARTITION_AARCH64_EXEC	BIT(8)
 	u32 properties;
 	u32 uuid[4];
 };
-- 
2.37.3


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Re: [PATCH v2 09/10] firmware: arm_ffa: Set up 32bit execution mode flag using partiion property

[Jens Wiklander]

On Fri, Sep 2, 2022 at 2:40 PM Sudeep Holla <sudeep.holla@arm.com> wrote:
>
> FF-A v1.1 adds a flag in the partition properties to indicate if the
> partition runs in the AArch32 or AArch64 execution state. Use the same
> to set-up the 32-bit execution flag mode in the ffa_dev automatically
> and ignore any requests to do the same from the ffa_driver.
>
> Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
> ---
>  drivers/firmware/arm_ffa/driver.c | 13 ++++++++++++-
>  include/linux/arm_ffa.h           |  2 ++
>  2 files changed, 14 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/firmware/arm_ffa/driver.c b/drivers/firmware/arm_ffa/driver.c
> index dd6ab2f81580..dbbe15592173 100644
> --- a/drivers/firmware/arm_ffa/driver.c
> +++ b/drivers/firmware/arm_ffa/driver.c
> @@ -647,11 +647,19 @@ static int ffa_partition_info_get(const char *uuid_str,
>         return 0;
>  }
>
> -static void ffa_mode_32bit_set(struct ffa_device *dev)
> +static void _ffa_mode_32bit_set(struct ffa_device *dev)
>  {
>         dev->mode_32bit = true;
>  }
>
> +static void ffa_mode_32bit_set(struct ffa_device *dev)
> +{
> +       if (drv_info->version > FFA_VERSION_1_0)
> +               return;
> +
> +       _ffa_mode_32bit_set(dev);
> +}
> +
>  static int ffa_sync_send_receive(struct ffa_device *dev,
>                                  struct ffa_send_direct_data *data)
>  {
> @@ -743,6 +751,9 @@ static void ffa_setup_partitions(void)
>                                __func__, tpbuf->id);
>                         continue;
>                 }
> +
> +               if (tpbuf->properties & FFA_PARTITION_AARCH64_EXEC)

Shouldn't this be !(tpbuf->properties & FFA_PARTITION_AARCH64_EXEC)?

Before this, we used 64-bit calling convention by default for FF-A
v1.0. I suppose we would like to keep that behaviour.

Cheers,
Jens

> +                       _ffa_mode_32bit_set(ffa_dev);
>         }
>         kfree(pbuf);
>  }
> diff --git a/include/linux/arm_ffa.h b/include/linux/arm_ffa.h
> index 09567ffd1f49..5964b6104996 100644
> --- a/include/linux/arm_ffa.h
> +++ b/include/linux/arm_ffa.h
> @@ -106,6 +106,8 @@ struct ffa_partition_info {
>  #define FFA_PARTITION_DIRECT_SEND      BIT(1)
>  /* partition can send and receive indirect messages. */
>  #define FFA_PARTITION_INDIRECT_MSG     BIT(2)
> +/* partition runs in the AArch64 execution state. */
> +#define FFA_PARTITION_AARCH64_EXEC     BIT(8)
>         u32 properties;
>         u32 uuid[4];
>  };
> --
> 2.37.3
>

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Re: [PATCH v2 09/10] firmware: arm_ffa: Set up 32bit execution mode flag using partiion property

[Sudeep Holla]

On Wed, Sep 07, 2022 at 11:01:27AM +0200, Jens Wiklander wrote:
> On Fri, Sep 2, 2022 at 2:40 PM Sudeep Holla <sudeep.holla@arm.com> wrote:
> >
> > FF-A v1.1 adds a flag in the partition properties to indicate if the
> > partition runs in the AArch32 or AArch64 execution state. Use the same
> > to set-up the 32-bit execution flag mode in the ffa_dev automatically
> > and ignore any requests to do the same from the ffa_driver.
> >
> > Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
> > ---
> >  drivers/firmware/arm_ffa/driver.c | 13 ++++++++++++-
> >  include/linux/arm_ffa.h           |  2 ++
> >  2 files changed, 14 insertions(+), 1 deletion(-)
> >
> > diff --git a/drivers/firmware/arm_ffa/driver.c b/drivers/firmware/arm_ffa/driver.c
> > index dd6ab2f81580..dbbe15592173 100644
> > --- a/drivers/firmware/arm_ffa/driver.c
> > +++ b/drivers/firmware/arm_ffa/driver.c
> > @@ -647,11 +647,19 @@ static int ffa_partition_info_get(const char *uuid_str,
> >         return 0;
> >  }
> >
> > -static void ffa_mode_32bit_set(struct ffa_device *dev)
> > +static void _ffa_mode_32bit_set(struct ffa_device *dev)
> >  {
> >         dev->mode_32bit = true;
> >  }
> >
> > +static void ffa_mode_32bit_set(struct ffa_device *dev)
> > +{
> > +       if (drv_info->version > FFA_VERSION_1_0)
> > +               return;
> > +
> > +       _ffa_mode_32bit_set(dev);
> > +}
> > +
> >  static int ffa_sync_send_receive(struct ffa_device *dev,
> >                                  struct ffa_send_direct_data *data)
> >  {
> > @@ -743,6 +751,9 @@ static void ffa_setup_partitions(void)
> >                                __func__, tpbuf->id);
> >                         continue;
> >                 }
> > +
> > +               if (tpbuf->properties & FFA_PARTITION_AARCH64_EXEC)
> 
> Shouldn't this be !(tpbuf->properties & FFA_PARTITION_AARCH64_EXEC)?
> 
> Before this, we used 64-bit calling convention by default for FF-A
> v1.0. I suppose we would like to keep that behaviour.
> 

Correct, it is stupid mistake on my part. I did notice this but somehow
messed up not to add the change I made to test with OPTEE. Thanks for
noticing this.

-- 
Regards,
Sudeep

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH v2 10/10] firmware: arm_ffa: Split up ffa_ops into info, message and memory operations

[Sudeep Holla]

In preparation to make memory operations accessible for a non
ffa_driver/device, it is better to split the ffa_ops into different
categories of operations: info, message and memory. The info and memory
are ffa_device independent and can be used without any associated
ffa_device from a non ffa_driver.

However, we don't export these info and memory APIs yet without the user.
The first users of these APIs can export them.

Reviewed-by: Jens Wiklander <jens.wiklander@linaro.org>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
---
 drivers/firmware/arm_ffa/driver.c | 16 +++++++++++++--
 drivers/tee/optee/ffa_abi.c       | 33 +++++++++++++++++--------------
 include/linux/arm_ffa.h           | 14 ++++++++++++-
 3 files changed, 45 insertions(+), 18 deletions(-)

diff --git a/drivers/firmware/arm_ffa/driver.c b/drivers/firmware/arm_ffa/driver.c
index dbbe15592173..639ae3911387 100644
--- a/drivers/firmware/arm_ffa/driver.c
+++ b/drivers/firmware/arm_ffa/driver.c
@@ -690,16 +690,28 @@ static int ffa_memory_lend(struct ffa_mem_ops_args *args)
 	return ffa_memory_ops(FFA_MEM_LEND, args);
 }
 
-static const struct ffa_ops ffa_ops = {
+static const struct ffa_info_ops ffa_drv_info_ops = {
 	.api_version_get = ffa_api_version_get,
 	.partition_info_get = ffa_partition_info_get,
+};
+
+static const struct ffa_msg_ops ffa_drv_msg_ops = {
 	.mode_32bit_set = ffa_mode_32bit_set,
 	.sync_send_receive = ffa_sync_send_receive,
+};
+
+static const struct ffa_mem_ops ffa_drv_mem_ops = {
 	.memory_reclaim = ffa_memory_reclaim,
 	.memory_share = ffa_memory_share,
 	.memory_lend = ffa_memory_lend,
 };
 
+static const struct ffa_ops ffa_drv_ops = {
+	.info_ops = &ffa_drv_info_ops,
+	.msg_ops = &ffa_drv_msg_ops,
+	.mem_ops = &ffa_drv_mem_ops,
+};
+
 void ffa_device_match_uuid(struct ffa_device *ffa_dev, const uuid_t *uuid)
 {
 	int count, idx;
@@ -745,7 +757,7 @@ static void ffa_setup_partitions(void)
 		 * provides UUID here for each partition as part of the
 		 * discovery API and the same is passed.
 		 */
-		ffa_dev = ffa_device_register(&uuid, tpbuf->id, &ffa_ops);
+		ffa_dev = ffa_device_register(&uuid, tpbuf->id, &ffa_drv_ops);
 		if (!ffa_dev) {
 			pr_err("%s: failed to register partition ID 0x%x\n",
 			       __func__, tpbuf->id);
diff --git a/drivers/tee/optee/ffa_abi.c b/drivers/tee/optee/ffa_abi.c
index 2ce5b87dfb27..0828240f27e6 100644
--- a/drivers/tee/optee/ffa_abi.c
+++ b/drivers/tee/optee/ffa_abi.c
@@ -272,7 +272,7 @@ static int optee_ffa_shm_register(struct tee_context *ctx, struct tee_shm *shm,
 {
 	struct optee *optee = tee_get_drvdata(ctx->teedev);
 	struct ffa_device *ffa_dev = optee->ffa.ffa_dev;
-	const struct ffa_ops *ffa_ops = ffa_dev->ops;
+	const struct ffa_mem_ops *mem_ops = ffa_dev->ops->mem_ops;
 	struct ffa_mem_region_attributes mem_attr = {
 		.receiver = ffa_dev->vm_id,
 		.attrs = FFA_MEM_RW,
@@ -294,14 +294,14 @@ static int optee_ffa_shm_register(struct tee_context *ctx, struct tee_shm *shm,
 	if (rc)
 		return rc;
 	args.sg = sgt.sgl;
-	rc = ffa_ops->memory_share(&args);
+	rc = mem_ops->memory_share(&args);
 	sg_free_table(&sgt);
 	if (rc)
 		return rc;
 
 	rc = optee_shm_add_ffa_handle(optee, shm, args.g_handle);
 	if (rc) {
-		ffa_ops->memory_reclaim(args.g_handle, 0);
+		mem_ops->memory_reclaim(args.g_handle, 0);
 		return rc;
 	}
 
@@ -315,7 +315,8 @@ static int optee_ffa_shm_unregister(struct tee_context *ctx,
 {
 	struct optee *optee = tee_get_drvdata(ctx->teedev);
 	struct ffa_device *ffa_dev = optee->ffa.ffa_dev;
-	const struct ffa_ops *ffa_ops = ffa_dev->ops;
+	const struct ffa_msg_ops *msg_ops = ffa_dev->ops->msg_ops;
+	const struct ffa_mem_ops *mem_ops = ffa_dev->ops->mem_ops;
 	u64 global_handle = shm->sec_world_id;
 	struct ffa_send_direct_data data = {
 		.data0 = OPTEE_FFA_UNREGISTER_SHM,
@@ -327,11 +328,11 @@ static int optee_ffa_shm_unregister(struct tee_context *ctx,
 	optee_shm_rem_ffa_handle(optee, global_handle);
 	shm->sec_world_id = 0;
 
-	rc = ffa_ops->sync_send_receive(ffa_dev, &data);
+	rc = msg_ops->sync_send_receive(ffa_dev, &data);
 	if (rc)
 		pr_err("Unregister SHM id 0x%llx rc %d\n", global_handle, rc);
 
-	rc = ffa_ops->memory_reclaim(global_handle, 0);
+	rc = mem_ops->memory_reclaim(global_handle, 0);
 	if (rc)
 		pr_err("mem_reclaim: 0x%llx %d", global_handle, rc);
 
@@ -342,7 +343,7 @@ static int optee_ffa_shm_unregister_supp(struct tee_context *ctx,
 					 struct tee_shm *shm)
 {
 	struct optee *optee = tee_get_drvdata(ctx->teedev);
-	const struct ffa_ops *ffa_ops = optee->ffa.ffa_dev->ops;
+	const struct ffa_mem_ops *mem_ops;
 	u64 global_handle = shm->sec_world_id;
 	int rc;
 
@@ -353,7 +354,8 @@ static int optee_ffa_shm_unregister_supp(struct tee_context *ctx,
 	 */
 
 	optee_shm_rem_ffa_handle(optee, global_handle);
-	rc = ffa_ops->memory_reclaim(global_handle, 0);
+	mem_ops = optee->ffa.ffa_dev->ops->mem_ops;
+	rc = mem_ops->memory_reclaim(global_handle, 0);
 	if (rc)
 		pr_err("mem_reclaim: 0x%llx %d", global_handle, rc);
 
@@ -530,7 +532,7 @@ static int optee_ffa_yielding_call(struct tee_context *ctx,
 {
 	struct optee *optee = tee_get_drvdata(ctx->teedev);
 	struct ffa_device *ffa_dev = optee->ffa.ffa_dev;
-	const struct ffa_ops *ffa_ops = ffa_dev->ops;
+	const struct ffa_msg_ops *msg_ops = ffa_dev->ops->msg_ops;
 	struct optee_call_waiter w;
 	u32 cmd = data->data0;
 	u32 w4 = data->data1;
@@ -541,7 +543,7 @@ static int optee_ffa_yielding_call(struct tee_context *ctx,
 	/* Initialize waiter */
 	optee_cq_wait_init(&optee->call_queue, &w);
 	while (true) {
-		rc = ffa_ops->sync_send_receive(ffa_dev, data);
+		rc = msg_ops->sync_send_receive(ffa_dev, data);
 		if (rc)
 			goto done;
 
@@ -576,7 +578,7 @@ static int optee_ffa_yielding_call(struct tee_context *ctx,
 		 * OP-TEE has returned with a RPC request.
 		 *
 		 * Note that data->data4 (passed in register w7) is already
-		 * filled in by ffa_ops->sync_send_receive() returning
+		 * filled in by ffa_mem_ops->sync_send_receive() returning
 		 * above.
 		 */
 		cond_resched();
@@ -654,12 +656,13 @@ static int optee_ffa_do_call_with_arg(struct tee_context *ctx,
 static bool optee_ffa_api_is_compatbile(struct ffa_device *ffa_dev,
 					const struct ffa_ops *ops)
 {
+	const struct ffa_msg_ops *msg_ops = ops->msg_ops;
 	struct ffa_send_direct_data data = { OPTEE_FFA_GET_API_VERSION };
 	int rc;
 
-	ops->mode_32bit_set(ffa_dev);
+	msg_ops->mode_32bit_set(ffa_dev);
 
-	rc = ops->sync_send_receive(ffa_dev, &data);
+	rc = msg_ops->sync_send_receive(ffa_dev, &data);
 	if (rc) {
 		pr_err("Unexpected error %d\n", rc);
 		return false;
@@ -672,7 +675,7 @@ static bool optee_ffa_api_is_compatbile(struct ffa_device *ffa_dev,
 	}
 
 	data = (struct ffa_send_direct_data){ OPTEE_FFA_GET_OS_VERSION };
-	rc = ops->sync_send_receive(ffa_dev, &data);
+	rc = msg_ops->sync_send_receive(ffa_dev, &data);
 	if (rc) {
 		pr_err("Unexpected error %d\n", rc);
 		return false;
@@ -694,7 +697,7 @@ static bool optee_ffa_exchange_caps(struct ffa_device *ffa_dev,
 	struct ffa_send_direct_data data = { OPTEE_FFA_EXCHANGE_CAPABILITIES };
 	int rc;
 
-	rc = ops->sync_send_receive(ffa_dev, &data);
+	rc = ops->msg_ops->sync_send_receive(ffa_dev, &data);
 	if (rc) {
 		pr_err("Unexpected error %d", rc);
 		return false;
diff --git a/include/linux/arm_ffa.h b/include/linux/arm_ffa.h
index 5964b6104996..5f02d2e6b9d9 100644
--- a/include/linux/arm_ffa.h
+++ b/include/linux/arm_ffa.h
@@ -257,16 +257,28 @@ struct ffa_mem_ops_args {
 	struct ffa_mem_region_attributes *attrs;
 };
 
-struct ffa_ops {
+struct ffa_info_ops {
 	u32 (*api_version_get)(void);
 	int (*partition_info_get)(const char *uuid_str,
 				  struct ffa_partition_info *buffer);
+};
+
+struct ffa_msg_ops {
 	void (*mode_32bit_set)(struct ffa_device *dev);
 	int (*sync_send_receive)(struct ffa_device *dev,
 				 struct ffa_send_direct_data *data);
+};
+
+struct ffa_mem_ops {
 	int (*memory_reclaim)(u64 g_handle, u32 flags);
 	int (*memory_share)(struct ffa_mem_ops_args *args);
 	int (*memory_lend)(struct ffa_mem_ops_args *args);
 };
 
+struct ffa_ops {
+	const struct ffa_info_ops *info_ops;
+	const struct ffa_msg_ops *msg_ops;
+	const struct ffa_mem_ops *mem_ops;
+};
+
 #endif /* _LINUX_ARM_FFA_H */
-- 
2.37.3


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Re: [PATCH v2 00/10] firmware: arm_ffa: Refactoring and initial/minor v1.1 update

[Sudeep Holla]

Hi Jens,

On Fri, Sep 02, 2022 at 01:40:22PM +0100, Sudeep Holla wrote:
> Hi All,
> 
> This series is just some refactoring in preparation to add FF-A v1.1
> support. It doesn't have any memory layout or notification changes
> supported in v1.1 yet.
> 

Thanks for the review of v1. There are some trivial changes that touches
optee files in v2 that needs official ack/review from you to take it
via my tree for v6.1. Please provide the same if you are OK especially
for patch 06,07/10 and even others if possible.

-- 
Regards,
Sudeep

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel