Re: [EXTERNAL] Re: [EXTERNAL] Re: [PATCH v4 2/5] soc: ti: pruss: Add pruss_{request,release}_mem_region() API

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 




On 21/03/2023 11:48, Md Danish Anwar wrote:
> Hi Roger,
> 
> On 21/03/23 14:54, Roger Quadros wrote:
>> Hi,
>>
>> On 21/03/2023 07:23, Md Danish Anwar wrote:
>>> Hi Andrew, Roger,
>>>
>>> On 20/03/23 21:48, Andrew Davis wrote:
>>>> On 3/20/23 12:11 AM, Md Danish Anwar wrote:
>>>>> Hi Roger,
>>>>>
>>>>> On 17/03/23 14:26, Roger Quadros wrote:
>>>>>> Hi Andrew & Danish,
>>>>>>
>>>>>>
>>>>>> On 13/03/2023 13:11, MD Danish Anwar wrote:
>>>>>>> From: "Andrew F. Davis" <afd@xxxxxx>
>>>>>>>
>>>>>>> Add two new API - pruss_request_mem_region() & pruss_release_mem_region(),
>>>>>>> to the PRUSS platform driver to allow client drivers to acquire and release
>>>>>>> the common memory resources present within a PRU-ICSS subsystem. This
>>>>>>> allows the client drivers to directly manipulate the respective memories,
>>>>>>> as per their design contract with the associated firmware.
>>>>>>>
>>>>>>> Co-developed-by: Suman Anna <s-anna@xxxxxx>
>>>>>>> Signed-off-by: Suman Anna <s-anna@xxxxxx>
>>>>>>> Signed-off-by: Andrew F. Davis <afd@xxxxxx>
>>>>>>> Co-developed-by: Grzegorz Jaszczyk <grzegorz.jaszczyk@xxxxxxxxxx>
>>>>>>> Signed-off-by: Grzegorz Jaszczyk <grzegorz.jaszczyk@xxxxxxxxxx>
>>>>>>> Signed-off-by: MD Danish Anwar <danishanwar@xxxxxx>
>>>>>>> Reviewed-by: Roger Quadros <rogerq@xxxxxxxxxx>
>>>>>>> ---
>>>>>>>   drivers/soc/ti/pruss.c           | 77 ++++++++++++++++++++++++++++++++
>>>>>>>   include/linux/pruss_driver.h     | 27 +++--------
>>>>>>>   include/linux/remoteproc/pruss.h | 39 ++++++++++++++++
>>>>>>
>>>>>>
>>>>>> We have these 2 header files and I think anything that deals with
>>>>>> 'struct pruss' should go in include/linux/pruss_driver.h
>>>>>>
>>>>>> Anything that deals with pru_rproc (i.e. struct rproc) should go in
>>>>>> include/linux/remoteproc/pruss.h
>>>>>>
>>>>>> Do you agree?
>>>>>>
>>>>>
>>>>> I agree with you Roger but Andrew is the right person to comment here as he is
>>>>> the author of this and several other patches.
>>>>>
>>>>> Hi Andrew, Can you please comment on this?
>>>>>
>>>>
>>>> Original idea was a consumer driver (like "ICSSG Ethernet Driver" in your other
>>>> series) could just
>>>>
>>>> #include <linux/remoteproc/pruss.h>
>>>>
>>>> and get everything they need, and nothing they do not.
>>>>
>>>
>>> If we plan on continuing the original idea, then I think keeping the header
>>> files as it is will be the best. Because if we move anything that deals with
>>> 'struct pruss' to include/linux/pruss_driver.h and anything that deals with
>>> pru_rproc (i.e. struct rproc) to include/linux/remoteproc/pruss.h, then the
>>> consumer drivers will need to do,
>>>
>>> #include <linux/remoteproc/pruss.h>
>>> #include <linux/pruss_driver.h>
>>>
>>> Roger, should I keep the header files arrangement as it is?
>>>
>>
>> OK but can we please rename one of them to something else so they don't
>> sound very similar. Maybe you could use Andrew's suggestion below.
>>
> 
> Yes sure, I'll rename the header files to reduce confusion. The pruss_driver.h
> is located in include/linux, implying it's not internal to PRUSS. So I will
> keep this header file name as it is.
> 
> There are total 3 pruss related header files.
> 
> 1. include/linux/pruss_driver.h (Public header file, not internal to PRUSS,
> will keep it as it is. This exists to allow communication between the pruss
> core and the pru rproc driver which live in different subsystems.)

Andrew asked you to rename this to pruss_internal.h

> 2. include/linux/remoteproc/pruss.h (Public header file, not internal to PRUSS,
> will keep it as it is. Only this header file needs to be included by client
> drivers.)
> 3. drivers/soc/ti/pruss.h (Internal to PRUSS, I will rename this to
> pruss_internal.h, this file has private definitions and APIs to modify PRUSS
> CFG space. This file is private to pruss.c)

No point in changing this file's name as this is not visible elsewhere.

> 
> Please let me know if the above looks OK.
> 
>>>> pruss_driver.h (which could be renamed pruss_internal.h) exists to allow
>>>> comunication between the pruss core and the pru rproc driver which live
>>>> in different subsystems.
>>>>
>>>> Andrew
>>>>
>>>>>>>   3 files changed, 121 insertions(+), 22 deletions(-)
>>>>>>>
>>>>>>> diff --git a/drivers/soc/ti/pruss.c b/drivers/soc/ti/pruss.c
>>>>>>> index a169aa1ed044..c8053c0d735f 100644
>>>>>>> --- a/drivers/soc/ti/pruss.c
>>>>>>> +++ b/drivers/soc/ti/pruss.c
>>>>>>> @@ -88,6 +88,82 @@ void pruss_put(struct pruss *pruss)
>>>>>>>   }
>>>>>>>   EXPORT_SYMBOL_GPL(pruss_put);
>>>>>>>   +/**
>>>>>>> + * pruss_request_mem_region() - request a memory resource
>>>>>>> + * @pruss: the pruss instance
>>>>>>> + * @mem_id: the memory resource id
>>>>>>> + * @region: pointer to memory region structure to be filled in
>>>>>>> + *
>>>>>>> + * This function allows a client driver to request a memory resource,
>>>>>>> + * and if successful, will let the client driver own the particular
>>>>>>> + * memory region until released using the pruss_release_mem_region()
>>>>>>> + * API.
>>>>>>> + *
>>>>>>> + * Return: 0 if requested memory region is available (in such case pointer to
>>>>>>> + * memory region is returned via @region), an error otherwise
>>>>>>> + */
>>>>>>> +int pruss_request_mem_region(struct pruss *pruss, enum pruss_mem mem_id,
>>>>>>> +                 struct pruss_mem_region *region)
>>>>>>> +{
>>>>>>> +    if (!pruss || !region || mem_id >= PRUSS_MEM_MAX)
>>>>>>> +        return -EINVAL;
>>>>>>> +
>>>>>>> +    mutex_lock(&pruss->lock);
>>>>>>> +
>>>>>>> +    if (pruss->mem_in_use[mem_id]) {
>>>>>>> +        mutex_unlock(&pruss->lock);
>>>>>>> +        return -EBUSY;
>>>>>>> +    }
>>>>>>> +
>>>>>>> +    *region = pruss->mem_regions[mem_id];
>>>>>>> +    pruss->mem_in_use[mem_id] = region;
>>>>>>> +
>>>>>>> +    mutex_unlock(&pruss->lock);
>>>>>>> +
>>>>>>> +    return 0;
>>>>>>> +}
>>>>>>> +EXPORT_SYMBOL_GPL(pruss_request_mem_region);
>>>>>>> +
>>>>>>> +/**
>>>>>>> + * pruss_release_mem_region() - release a memory resource
>>>>>>> + * @pruss: the pruss instance
>>>>>>> + * @region: the memory region to release
>>>>>>> + *
>>>>>>> + * This function is the complimentary function to
>>>>>>> + * pruss_request_mem_region(), and allows the client drivers to
>>>>>>> + * release back a memory resource.
>>>>>>> + *
>>>>>>> + * Return: 0 on success, an error code otherwise
>>>>>>> + */
>>>>>>> +int pruss_release_mem_region(struct pruss *pruss,
>>>>>>> +                 struct pruss_mem_region *region)
>>>>>>> +{
>>>>>>> +    int id;
>>>>>>> +
>>>>>>> +    if (!pruss || !region)
>>>>>>> +        return -EINVAL;
>>>>>>> +
>>>>>>> +    mutex_lock(&pruss->lock);
>>>>>>> +
>>>>>>> +    /* find out the memory region being released */
>>>>>>> +    for (id = 0; id < PRUSS_MEM_MAX; id++) {
>>>>>>> +        if (pruss->mem_in_use[id] == region)
>>>>>>> +            break;
>>>>>>> +    }
>>>>>>> +
>>>>>>> +    if (id == PRUSS_MEM_MAX) {
>>>>>>> +        mutex_unlock(&pruss->lock);
>>>>>>> +        return -EINVAL;
>>>>>>> +    }
>>>>>>> +
>>>>>>> +    pruss->mem_in_use[id] = NULL;
>>>>>>> +
>>>>>>> +    mutex_unlock(&pruss->lock);
>>>>>>> +
>>>>>>> +    return 0;
>>>>>>> +}
>>>>>>> +EXPORT_SYMBOL_GPL(pruss_release_mem_region);
>>>>>>> +
>>>>>>>   static void pruss_of_free_clk_provider(void *data)
>>>>>>>   {
>>>>>>>       struct device_node *clk_mux_np = data;
>>>>>>> @@ -290,6 +366,7 @@ static int pruss_probe(struct platform_device *pdev)
>>>>>>>           return -ENOMEM;
>>>>>>>         pruss->dev = dev;
>>>>>>> +    mutex_init(&pruss->lock);
>>>>>>>         child = of_get_child_by_name(np, "memories");
>>>>>>>       if (!child) {
>>>>>>> diff --git a/include/linux/pruss_driver.h b/include/linux/pruss_driver.h
>>>>>>> index 86242fb5a64a..22b4b37d2536 100644
>>>>>>> --- a/include/linux/pruss_driver.h
>>>>>>> +++ b/include/linux/pruss_driver.h
>>>>>>> @@ -9,37 +9,18 @@
>>>>>>>   #ifndef _PRUSS_DRIVER_H_
>>>>>>>   #define _PRUSS_DRIVER_H_
>>>>>>>   +#include <linux/mutex.h>
>>>>>>>   #include <linux/remoteproc/pruss.h>
>>>>>>>   #include <linux/types.h>
>>>>>>>   -/*
>>>>>>> - * enum pruss_mem - PRUSS memory range identifiers
>>>>>>> - */
>>>>>>> -enum pruss_mem {
>>>>>>> -    PRUSS_MEM_DRAM0 = 0,
>>>>>>> -    PRUSS_MEM_DRAM1,
>>>>>>> -    PRUSS_MEM_SHRD_RAM2,
>>>>>>> -    PRUSS_MEM_MAX,
>>>>>>> -};
>>>>>>> -
>>>>>>> -/**
>>>>>>> - * struct pruss_mem_region - PRUSS memory region structure
>>>>>>> - * @va: kernel virtual address of the PRUSS memory region
>>>>>>> - * @pa: physical (bus) address of the PRUSS memory region
>>>>>>> - * @size: size of the PRUSS memory region
>>>>>>> - */
>>>>>>> -struct pruss_mem_region {
>>>>>>> -    void __iomem *va;
>>>>>>> -    phys_addr_t pa;
>>>>>>> -    size_t size;
>>>>>>> -};
>>>>>>> -
>>>>>>>   /**
>>>>>>>    * struct pruss - PRUSS parent structure
>>>>>>>    * @dev: pruss device pointer
>>>>>>>    * @cfg_base: base iomap for CFG region
>>>>>>>    * @cfg_regmap: regmap for config region
>>>>>>>    * @mem_regions: data for each of the PRUSS memory regions
>>>>>>> + * @mem_in_use: to indicate if memory resource is in use
>>>>>>> + * @lock: mutex to serialize access to resources
>>>>>>>    * @core_clk_mux: clk handle for PRUSS CORE_CLK_MUX
>>>>>>>    * @iep_clk_mux: clk handle for PRUSS IEP_CLK_MUX
>>>>>>>    */
>>>>>>> @@ -48,6 +29,8 @@ struct pruss {
>>>>>>>       void __iomem *cfg_base;
>>>>>>>       struct regmap *cfg_regmap;
>>>>>>>       struct pruss_mem_region mem_regions[PRUSS_MEM_MAX];
>>>>>>> +    struct pruss_mem_region *mem_in_use[PRUSS_MEM_MAX];
>>>>>>> +    struct mutex lock; /* PRU resource lock */
>>>>>>>       struct clk *core_clk_mux;
>>>>>>>       struct clk *iep_clk_mux;
>>>>>>>   };
>>>>>>> diff --git a/include/linux/remoteproc/pruss.h
>>>>>>> b/include/linux/remoteproc/pruss.h
>>>>>>> index 93a98cac7829..33f930e0a0ce 100644
>>>>>>> --- a/include/linux/remoteproc/pruss.h
>>>>>>> +++ b/include/linux/remoteproc/pruss.h
>>>>>>> @@ -44,6 +44,28 @@ enum pru_ctable_idx {
>>>>>>>       PRU_C31,
>>>>>>>   };
>>>>>>>   +/*
>>>>>>> + * enum pruss_mem - PRUSS memory range identifiers
>>>>>>> + */
>>>>>>> +enum pruss_mem {
>>>>>>> +    PRUSS_MEM_DRAM0 = 0,
>>>>>>> +    PRUSS_MEM_DRAM1,
>>>>>>> +    PRUSS_MEM_SHRD_RAM2,
>>>>>>> +    PRUSS_MEM_MAX,
>>>>>>> +};
>>>>>>> +
>>>>>>> +/**
>>>>>>> + * struct pruss_mem_region - PRUSS memory region structure
>>>>>>> + * @va: kernel virtual address of the PRUSS memory region
>>>>>>> + * @pa: physical (bus) address of the PRUSS memory region
>>>>>>> + * @size: size of the PRUSS memory region
>>>>>>> + */
>>>>>>> +struct pruss_mem_region {
>>>>>>> +    void __iomem *va;
>>>>>>> +    phys_addr_t pa;
>>>>>>> +    size_t size;
>>>>>>> +};
>>>>>>> +
>>>>>>>   struct device_node;
>>>>>>>   struct rproc;
>>>>>>>   struct pruss;
>>>>>>> @@ -52,6 +74,10 @@ struct pruss;
>>>>>>>     struct pruss *pruss_get(struct rproc *rproc);
>>>>>>>   void pruss_put(struct pruss *pruss);
>>>>>>> +int pruss_request_mem_region(struct pruss *pruss, enum pruss_mem mem_id,
>>>>>>> +                 struct pruss_mem_region *region);
>>>>>>> +int pruss_release_mem_region(struct pruss *pruss,
>>>>>>> +                 struct pruss_mem_region *region);
>>>>>>>     #else
>>>>>>>   @@ -62,6 +88,19 @@ static inline struct pruss *pruss_get(struct rproc
>>>>>>> *rproc)
>>>>>>>     static inline void pruss_put(struct pruss *pruss) { }
>>>>>>>   +static inline int pruss_request_mem_region(struct pruss *pruss,
>>>>>>> +                       enum pruss_mem mem_id,
>>>>>>> +                       struct pruss_mem_region *region)
>>>>>>> +{
>>>>>>> +    return -EOPNOTSUPP;
>>>>>>> +}
>>>>>>> +
>>>>>>> +static inline int pruss_release_mem_region(struct pruss *pruss,
>>>>>>> +                       struct pruss_mem_region *region)
>>>>>>> +{
>>>>>>> +    return -EOPNOTSUPP;
>>>>>>> +}
>>>>>>> +
>>>>>>>   #endif /* CONFIG_TI_PRUSS */
>>>>>>>     #if IS_ENABLED(CONFIG_PRU_REMOTEPROC)
>>>>>>


cheers,
-roger



[Index of Archives]     [Linux Sound]     [ALSA Users]     [ALSA Devel]     [Linux Audio Users]     [Linux Media]     [Kernel]     [Photo Sharing]     [Gimp]     [Yosemite News]     [Linux Media]

  Powered by Linux