Re: [PATCH v2 6/8] staging: fsl-mc: don't use raw device io functions

[Arnd Bergmann]

On Tue, Jul 18, 2017 at 4:26 PM, Laurentiu Tudor
 wrote:
>
> Maybe i made an error in my previous emails, but the hi-lo variant is
> the correct one. The command execution is triggered when the _first_
> 32-bit half of the header (header&0x) is written, so that's why
> it must be written last.

I'm pretty sure I just misremembered it then. Thanks for the clarification.

Arnd

Re: [PATCH v2 6/8] staging: fsl-mc: don't use raw device io functions

[Arnd Bergmann]

On Tue, Jul 18, 2017 at 4:26 PM, Laurentiu Tudor
 wrote:
>
> Maybe i made an error in my previous emails, but the hi-lo variant is
> the correct one. The command execution is triggered when the _first_
> 32-bit half of the header (header&0x) is written, so that's why
> it must be written last.

I'm pretty sure I just misremembered it then. Thanks for the clarification.

Arnd

Re: [PATCH v2 6/8] staging: fsl-mc: don't use raw device io functions

[Laurentiu Tudor]

Hi Arnd,

On 07/18/2017 05:18 PM, Arnd Bergmann wrote:
> On Tue, Jul 18, 2017 at 3:37 PM,   wrote:
>> From: Laurentiu Tudor 
>>
>> As raw device io functions are not portable and don't handle byte-order
>> (triggering suspicion that endianness isn't handled well) switch to
>> using the standard api.
>> Since MC expects LE byte-order and the upper layers already take care
>> of that, we need to trick the device io api by doing a LE -> CPU
>> conversion just before calling it. This way, the CPU -> LE conversion
>> done in the api puts the data back in the right byte-order. Obviously,
>> for reads the extra step is mirrored: there's a CPU -> LE conversion
>> following the API call.
>>
>> Signed-off-by: Laurentiu Tudor 
>> ---
>> Notes:
>>  -v2
>>-new patch replacing 
>> https://emea01.safelinks.protection.outlook.com/?url=https%3A%2F%2Flkml.org%2Flkml%2F2017%2F7%2F17%2F419=01%7C01%7Claurentiu.tudor%40nxp.com%7C77381272b4914c9ac64708d4cde7d94e%7C686ea1d3bc2b4c6fa92cd99c5c301635%7C0=FTVLKox6T4i9OFmb%2B5BkSEDQrDrafXznY6nsJ0dgFSk%3D=0
>>
>>   drivers/staging/fsl-mc/bus/mc-sys.c | 21 +++--
>>   1 file changed, 15 insertions(+), 6 deletions(-)
>>
>> diff --git a/drivers/staging/fsl-mc/bus/mc-sys.c 
>> b/drivers/staging/fsl-mc/bus/mc-sys.c
>> index 195d9f3..8a6dc47 100644
>> --- a/drivers/staging/fsl-mc/bus/mc-sys.c
>> +++ b/drivers/staging/fsl-mc/bus/mc-sys.c
>> @@ -126,12 +126,15 @@ static inline void mc_write_command(struct mc_command 
>> __iomem *portal,
>>
>>  /* copy command parameters into the portal */
>>  for (i = 0; i < MC_CMD_NUM_OF_PARAMS; i++)
>> -   __raw_writeq(cmd->params[i], >params[i]);
>> -   /* ensure command params are committed before submitting it */
>> -   wmb();
>> +   /*
>> +* Data is already in the expected LE byte-order. Do an
>> +* extra LE -> CPU conversion so that the CPU -> LE done in
>> +* the device io write api puts it back in the right order.
>> +*/
>> +   writeq_relaxed(le64_to_cpu(cmd->params[i]), 
>> >params[i]);
>>
>>  /* submit the command by writing the header */
>> -   __raw_writeq(cmd->header, >header);
>> +   writeq(le64_to_cpu(cmd->header), >header);
>>   }
>
> Looks good, but just to be sure this is what you intended:
>
> On 32-bit systems, this will now write val>>32 to cmd->header+4,
> followed by writing val&0x to cmd->header.

Right. That's how it should happen.

> You said earlier that the command is triggered when the final
> four bytes are written, but it looks like the order is wrong now.
>
> Should you use io-64-nonatomic-lo-hi.h instead of
> io-64-nonatomic-hi-lo.h then?
>

Maybe i made an error in my previous emails, but the hi-lo variant is
the correct one. The command execution is triggered when the _first_ 
32-bit half of the header (header&0x) is written, so that's why 
it must be written last.

---
Thanks & Best Regards, Laurentiu

Re: [PATCH v2 6/8] staging: fsl-mc: don't use raw device io functions

[Laurentiu Tudor]

Hi Arnd,

On 07/18/2017 05:18 PM, Arnd Bergmann wrote:
> On Tue, Jul 18, 2017 at 3:37 PM,   wrote:
>> From: Laurentiu Tudor 
>>
>> As raw device io functions are not portable and don't handle byte-order
>> (triggering suspicion that endianness isn't handled well) switch to
>> using the standard api.
>> Since MC expects LE byte-order and the upper layers already take care
>> of that, we need to trick the device io api by doing a LE -> CPU
>> conversion just before calling it. This way, the CPU -> LE conversion
>> done in the api puts the data back in the right byte-order. Obviously,
>> for reads the extra step is mirrored: there's a CPU -> LE conversion
>> following the API call.
>>
>> Signed-off-by: Laurentiu Tudor 
>> ---
>> Notes:
>>  -v2
>>-new patch replacing 
>> https://emea01.safelinks.protection.outlook.com/?url=https%3A%2F%2Flkml.org%2Flkml%2F2017%2F7%2F17%2F419=01%7C01%7Claurentiu.tudor%40nxp.com%7C77381272b4914c9ac64708d4cde7d94e%7C686ea1d3bc2b4c6fa92cd99c5c301635%7C0=FTVLKox6T4i9OFmb%2B5BkSEDQrDrafXznY6nsJ0dgFSk%3D=0
>>
>>   drivers/staging/fsl-mc/bus/mc-sys.c | 21 +++--
>>   1 file changed, 15 insertions(+), 6 deletions(-)
>>
>> diff --git a/drivers/staging/fsl-mc/bus/mc-sys.c 
>> b/drivers/staging/fsl-mc/bus/mc-sys.c
>> index 195d9f3..8a6dc47 100644
>> --- a/drivers/staging/fsl-mc/bus/mc-sys.c
>> +++ b/drivers/staging/fsl-mc/bus/mc-sys.c
>> @@ -126,12 +126,15 @@ static inline void mc_write_command(struct mc_command 
>> __iomem *portal,
>>
>>  /* copy command parameters into the portal */
>>  for (i = 0; i < MC_CMD_NUM_OF_PARAMS; i++)
>> -   __raw_writeq(cmd->params[i], >params[i]);
>> -   /* ensure command params are committed before submitting it */
>> -   wmb();
>> +   /*
>> +* Data is already in the expected LE byte-order. Do an
>> +* extra LE -> CPU conversion so that the CPU -> LE done in
>> +* the device io write api puts it back in the right order.
>> +*/
>> +   writeq_relaxed(le64_to_cpu(cmd->params[i]), 
>> >params[i]);
>>
>>  /* submit the command by writing the header */
>> -   __raw_writeq(cmd->header, >header);
>> +   writeq(le64_to_cpu(cmd->header), >header);
>>   }
>
> Looks good, but just to be sure this is what you intended:
>
> On 32-bit systems, this will now write val>>32 to cmd->header+4,
> followed by writing val&0x to cmd->header.

Right. That's how it should happen.

> You said earlier that the command is triggered when the final
> four bytes are written, but it looks like the order is wrong now.
>
> Should you use io-64-nonatomic-lo-hi.h instead of
> io-64-nonatomic-hi-lo.h then?
>

Maybe i made an error in my previous emails, but the hi-lo variant is
the correct one. The command execution is triggered when the _first_ 
32-bit half of the header (header&0x) is written, so that's why 
it must be written last.

---
Thanks & Best Regards, Laurentiu

Re: [PATCH v2 6/8] staging: fsl-mc: don't use raw device io functions

[Arnd Bergmann]

On Tue, Jul 18, 2017 at 3:37 PM,   wrote:
> From: Laurentiu Tudor 
>
> As raw device io functions are not portable and don't handle byte-order
> (triggering suspicion that endianness isn't handled well) switch to
> using the standard api.
> Since MC expects LE byte-order and the upper layers already take care
> of that, we need to trick the device io api by doing a LE -> CPU
> conversion just before calling it. This way, the CPU -> LE conversion
> done in the api puts the data back in the right byte-order. Obviously,
> for reads the extra step is mirrored: there's a CPU -> LE conversion
> following the API call.
>
> Signed-off-by: Laurentiu Tudor 
> ---
> Notes:
> -v2
>   -new patch replacing https://lkml.org/lkml/2017/7/17/419
>
>  drivers/staging/fsl-mc/bus/mc-sys.c | 21 +++--
>  1 file changed, 15 insertions(+), 6 deletions(-)
>
> diff --git a/drivers/staging/fsl-mc/bus/mc-sys.c 
> b/drivers/staging/fsl-mc/bus/mc-sys.c
> index 195d9f3..8a6dc47 100644
> --- a/drivers/staging/fsl-mc/bus/mc-sys.c
> +++ b/drivers/staging/fsl-mc/bus/mc-sys.c
> @@ -126,12 +126,15 @@ static inline void mc_write_command(struct mc_command 
> __iomem *portal,
>
> /* copy command parameters into the portal */
> for (i = 0; i < MC_CMD_NUM_OF_PARAMS; i++)
> -   __raw_writeq(cmd->params[i], >params[i]);
> -   /* ensure command params are committed before submitting it */
> -   wmb();
> +   /*
> +* Data is already in the expected LE byte-order. Do an
> +* extra LE -> CPU conversion so that the CPU -> LE done in
> +* the device io write api puts it back in the right order.
> +*/
> +   writeq_relaxed(le64_to_cpu(cmd->params[i]), 
> >params[i]);
>
> /* submit the command by writing the header */
> -   __raw_writeq(cmd->header, >header);
> +   writeq(le64_to_cpu(cmd->header), >header);
>  }

Looks good, but just to be sure this is what you intended:

On 32-bit systems, this will now write val>>32 to cmd->header+4,
followed by writing val&0x to cmd->header.

You said earlier that the command is triggered when the final
four bytes are written, but it looks like the order is wrong now.

Should you use io-64-nonatomic-lo-hi.h instead of
io-64-nonatomic-hi-lo.h then?

Arnd

Re: [PATCH v2 6/8] staging: fsl-mc: don't use raw device io functions

[Arnd Bergmann]

On Tue, Jul 18, 2017 at 3:37 PM,   wrote:
> From: Laurentiu Tudor 
>
> As raw device io functions are not portable and don't handle byte-order
> (triggering suspicion that endianness isn't handled well) switch to
> using the standard api.
> Since MC expects LE byte-order and the upper layers already take care
> of that, we need to trick the device io api by doing a LE -> CPU
> conversion just before calling it. This way, the CPU -> LE conversion
> done in the api puts the data back in the right byte-order. Obviously,
> for reads the extra step is mirrored: there's a CPU -> LE conversion
> following the API call.
>
> Signed-off-by: Laurentiu Tudor 
> ---
> Notes:
> -v2
>   -new patch replacing https://lkml.org/lkml/2017/7/17/419
>
>  drivers/staging/fsl-mc/bus/mc-sys.c | 21 +++--
>  1 file changed, 15 insertions(+), 6 deletions(-)
>
> diff --git a/drivers/staging/fsl-mc/bus/mc-sys.c 
> b/drivers/staging/fsl-mc/bus/mc-sys.c
> index 195d9f3..8a6dc47 100644
> --- a/drivers/staging/fsl-mc/bus/mc-sys.c
> +++ b/drivers/staging/fsl-mc/bus/mc-sys.c
> @@ -126,12 +126,15 @@ static inline void mc_write_command(struct mc_command 
> __iomem *portal,
>
> /* copy command parameters into the portal */
> for (i = 0; i < MC_CMD_NUM_OF_PARAMS; i++)
> -   __raw_writeq(cmd->params[i], >params[i]);
> -   /* ensure command params are committed before submitting it */
> -   wmb();
> +   /*
> +* Data is already in the expected LE byte-order. Do an
> +* extra LE -> CPU conversion so that the CPU -> LE done in
> +* the device io write api puts it back in the right order.
> +*/
> +   writeq_relaxed(le64_to_cpu(cmd->params[i]), 
> >params[i]);
>
> /* submit the command by writing the header */
> -   __raw_writeq(cmd->header, >header);
> +   writeq(le64_to_cpu(cmd->header), >header);
>  }

Looks good, but just to be sure this is what you intended:

On 32-bit systems, this will now write val>>32 to cmd->header+4,
followed by writing val&0x to cmd->header.

You said earlier that the command is triggered when the final
four bytes are written, but it looks like the order is wrong now.

Should you use io-64-nonatomic-lo-hi.h instead of
io-64-nonatomic-hi-lo.h then?

Arnd

[PATCH v2 6/8] staging: fsl-mc: don't use raw device io functions

[laurentiu.tudor]

From: Laurentiu Tudor 

As raw device io functions are not portable and don't handle byte-order
(triggering suspicion that endianness isn't handled well) switch to
using the standard api.
Since MC expects LE byte-order and the upper layers already take care
of that, we need to trick the device io api by doing a LE -> CPU
conversion just before calling it. This way, the CPU -> LE conversion
done in the api puts the data back in the right byte-order. Obviously,
for reads the extra step is mirrored: there's a CPU -> LE conversion
following the API call.

Signed-off-by: Laurentiu Tudor 
---
Notes:
-v2
  -new patch replacing https://lkml.org/lkml/2017/7/17/419

 drivers/staging/fsl-mc/bus/mc-sys.c | 21 +++--
 1 file changed, 15 insertions(+), 6 deletions(-)

diff --git a/drivers/staging/fsl-mc/bus/mc-sys.c 
b/drivers/staging/fsl-mc/bus/mc-sys.c
index 195d9f3..8a6dc47 100644
--- a/drivers/staging/fsl-mc/bus/mc-sys.c
+++ b/drivers/staging/fsl-mc/bus/mc-sys.c
@@ -126,12 +126,15 @@ static inline void mc_write_command(struct mc_command 
__iomem *portal,
 
/* copy command parameters into the portal */
for (i = 0; i < MC_CMD_NUM_OF_PARAMS; i++)
-   __raw_writeq(cmd->params[i], >params[i]);
-   /* ensure command params are committed before submitting it */
-   wmb();
+   /*
+* Data is already in the expected LE byte-order. Do an
+* extra LE -> CPU conversion so that the CPU -> LE done in
+* the device io write api puts it back in the right order.
+*/
+   writeq_relaxed(le64_to_cpu(cmd->params[i]), >params[i]);
 
/* submit the command by writing the header */
-   __raw_writeq(cmd->header, >header);
+   writeq(le64_to_cpu(cmd->header), >header);
 }
 
 /**
@@ -151,14 +154,20 @@ static inline enum mc_cmd_status mc_read_response(struct 
mc_command __iomem *
enum mc_cmd_status status;
 
/* Copy command response header from MC portal: */
-   resp->header = __raw_readq(>header);
+   resp->header = cpu_to_le64(readq_relaxed(>header));
status = mc_cmd_hdr_read_status(resp);
if (status != MC_CMD_STATUS_OK)
return status;
 
/* Copy command response data from MC portal: */
for (i = 0; i < MC_CMD_NUM_OF_PARAMS; i++)
-   resp->params[i] = __raw_readq(>params[i]);
+   /*
+* Data is expected to be in LE byte-order. Do an
+* extra CPU -> LE to revert the LE -> CPU done in
+* the device io read api.
+*/
+   resp->params[i] =
+   cpu_to_le64(readq_relaxed(>params[i]));
 
return status;
 }
-- 
2.9.4

[PATCH v2 6/8] staging: fsl-mc: don't use raw device io functions

[laurentiu.tudor]

From: Laurentiu Tudor 

As raw device io functions are not portable and don't handle byte-order
(triggering suspicion that endianness isn't handled well) switch to
using the standard api.
Since MC expects LE byte-order and the upper layers already take care
of that, we need to trick the device io api by doing a LE -> CPU
conversion just before calling it. This way, the CPU -> LE conversion
done in the api puts the data back in the right byte-order. Obviously,
for reads the extra step is mirrored: there's a CPU -> LE conversion
following the API call.

Signed-off-by: Laurentiu Tudor 
---
Notes:
-v2
  -new patch replacing https://lkml.org/lkml/2017/7/17/419

 drivers/staging/fsl-mc/bus/mc-sys.c | 21 +++--
 1 file changed, 15 insertions(+), 6 deletions(-)

diff --git a/drivers/staging/fsl-mc/bus/mc-sys.c 
b/drivers/staging/fsl-mc/bus/mc-sys.c
index 195d9f3..8a6dc47 100644
--- a/drivers/staging/fsl-mc/bus/mc-sys.c
+++ b/drivers/staging/fsl-mc/bus/mc-sys.c
@@ -126,12 +126,15 @@ static inline void mc_write_command(struct mc_command 
__iomem *portal,
 
/* copy command parameters into the portal */
for (i = 0; i < MC_CMD_NUM_OF_PARAMS; i++)
-   __raw_writeq(cmd->params[i], >params[i]);
-   /* ensure command params are committed before submitting it */
-   wmb();
+   /*
+* Data is already in the expected LE byte-order. Do an
+* extra LE -> CPU conversion so that the CPU -> LE done in
+* the device io write api puts it back in the right order.
+*/
+   writeq_relaxed(le64_to_cpu(cmd->params[i]), >params[i]);
 
/* submit the command by writing the header */
-   __raw_writeq(cmd->header, >header);
+   writeq(le64_to_cpu(cmd->header), >header);
 }
 
 /**
@@ -151,14 +154,20 @@ static inline enum mc_cmd_status mc_read_response(struct 
mc_command __iomem *
enum mc_cmd_status status;
 
/* Copy command response header from MC portal: */
-   resp->header = __raw_readq(>header);
+   resp->header = cpu_to_le64(readq_relaxed(>header));
status = mc_cmd_hdr_read_status(resp);
if (status != MC_CMD_STATUS_OK)
return status;
 
/* Copy command response data from MC portal: */
for (i = 0; i < MC_CMD_NUM_OF_PARAMS; i++)
-   resp->params[i] = __raw_readq(>params[i]);
+   /*
+* Data is expected to be in LE byte-order. Do an
+* extra CPU -> LE to revert the LE -> CPU done in
+* the device io read api.
+*/
+   resp->params[i] =
+   cpu_to_le64(readq_relaxed(>params[i]));
 
return status;
 }
-- 
2.9.4