[PATCH v2 RESEND 0/2] mtd: Harmonize implementations of OTP write and _get_{fact,user}_prot_info

[PATCH v2 RESEND 0/2] mtd: Harmonize implementations of OTP write and _get_{fact,user}_prot_info

[Christian Riesch]

Hi all,

In the discussion on my patchset for the OTP support for
drivers/mtd/chips/cfi_cmdset_0002.c [1-5], Artem requested two changes in the
current code of the OTP write functions and the _get_{fact,user}_prot_info
code.

These two patches are an attempt to make the requested changes.

The first patch adds a retlen parameter to the _get_fact_prot_info and
_get_user_prot_info functions and thus harmonizes the implementation
with those of the write and read functions.

The second patch fixes a problem that I earlier addressed in [1]. After
the discussion about this patch on the mtd mailing list, I think that the
correct behavior of the write function should be the one specified in [6]:
Try to write as many bytes as possible and return the number of bytes
that were written. If no data could be written due to lack of OTP memory,
return -ENOSPC.

Artem, would you please have a look at these patches? I would like to know
if I understood you correctly, or if I missed something here. Please note
that I cannot test these patches since I do not have the hardware. The
patches are compile tested only. If these patches are ok, I will also
respin the OTP support patches for cfi_cmdset_0002.

Changes for v2:
- Fixed buggy cfi_intelext_get_fact_prot_info

Thank you!

Best regards,
Christian

[1] http://patchwork.ozlabs.org/patch/239897/
[2] http://patchwork.ozlabs.org/patch/240010/
[3] http://patchwork.ozlabs.org/patch/240007/
[4] http://patchwork.ozlabs.org/patch/240008/
[5] http://patchwork.ozlabs.org/patch/240009/
[6] http://pubs.opengroup.org/onlinepubs/9699919799/functions/write.html

Cc: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Cc: Kyungmin Park <kyungmin.park@samsung.com>
Cc: Amul Kumar Saha <amul.saha@samsung.com>
Cc: Brian Norris <computersforpeace@gmail.com>

Christian Riesch (2):
  mtd: Add a retlen parameter to _get_{fact,user}_prot_info
  mtd: Fix the behavior of otp write if there is not enough room for
    data

 drivers/mtd/chips/cfi_cmdset_0001.c |   44 +++++++++++++++++++----------------
 drivers/mtd/devices/mtd_dataflash.c |   20 +++++++---------
 drivers/mtd/mtdchar.c               |   18 ++++++++++----
 drivers/mtd/mtdcore.c               |   12 +++++-----
 drivers/mtd/mtdpart.c               |   14 ++++++-----
 drivers/mtd/onenand/onenand_base.c  |   40 ++++++++++++++++---------------
 include/linux/mtd/mtd.h             |   16 ++++++-------
 7 files changed, 89 insertions(+), 75 deletions(-)

-- 
1.7.9.5

Re: [PATCH v2 RESEND 0/2] mtd: Harmonize implementations of OTP write and _get_{fact, user}_prot_info

[Christian Riesch]

Hi,

--On January 28, 2014 09:29 +0100 Christian Riesch 
<christian.riesch@omicron.at> wrote:

> Hi all,
>
> In the discussion on my patchset for the OTP support for
> drivers/mtd/chips/cfi_cmdset_0002.c [1-5], Artem requested two changes in
> the current code of the OTP write functions and the
> _get_{fact,user}_prot_info code.
>
> These two patches are an attempt to make the requested changes.
>
> The first patch adds a retlen parameter to the _get_fact_prot_info and
> _get_user_prot_info functions and thus harmonizes the implementation
> with those of the write and read functions.
>
> The second patch fixes a problem that I earlier addressed in [1]. After
> the discussion about this patch on the mtd mailing list, I think that the
> correct behavior of the write function should be the one specified in [6]:
> Try to write as many bytes as possible and return the number of bytes
> that were written. If no data could be written due to lack of OTP memory,
> return -ENOSPC.
>
> Artem, would you please have a look at these patches? I would like to know
> if I understood you correctly, or if I missed something here. Please note
> that I cannot test these patches since I do not have the hardware. The
> patches are compile tested only. If these patches are ok, I will also
> respin the OTP support patches for cfi_cmdset_0002.
>

Sorry for bugging you (Artem, mtd Maintainers) again, but I would really 
like to get this patchset ([a], [b]) and the OTP support for 
cfi_cmdset_0002.c into mainline. Please have a look at the patches and just 
tell me if the changes in them are acceptable, or if I should rather post 
the OTP support without these two patches. Once I have your answer, I will 
respin these patches and my OTP patches, test them again, and submit them. 
But without your help I cannot continue.

[a] http://patchwork.ozlabs.org/patch/314624/
[b] http://patchwork.ozlabs.org/patch/314625/

Thank you very much!

Christian

> Changes for v2:
> - Fixed buggy cfi_intelext_get_fact_prot_info
>
> Thank you!
>
> Best regards,
> Christian
>
> [1] http://patchwork.ozlabs.org/patch/239897/
> [2] http://patchwork.ozlabs.org/patch/240010/
> [3] http://patchwork.ozlabs.org/patch/240007/
> [4] http://patchwork.ozlabs.org/patch/240008/
> [5] http://patchwork.ozlabs.org/patch/240009/
> [6] http://pubs.opengroup.org/onlinepubs/9699919799/functions/write.html
>
> Cc: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
> Cc: Kyungmin Park <kyungmin.park@samsung.com>
> Cc: Amul Kumar Saha <amul.saha@samsung.com>
> Cc: Brian Norris <computersforpeace@gmail.com>
>
> Christian Riesch (2):
>   mtd: Add a retlen parameter to _get_{fact,user}_prot_info
>   mtd: Fix the behavior of otp write if there is not enough room for
>     data
>
>  drivers/mtd/chips/cfi_cmdset_0001.c |   44
> +++++++++++++++++++----------------  drivers/mtd/devices/mtd_dataflash.c
> |   20 +++++++---------
>  drivers/mtd/mtdchar.c               |   18 ++++++++++----
>  drivers/mtd/mtdcore.c               |   12 +++++-----
>  drivers/mtd/mtdpart.c               |   14 ++++++-----
>  drivers/mtd/onenand/onenand_base.c  |   40
> ++++++++++++++++---------------  include/linux/mtd/mtd.h             |
> 16 ++++++-------
>  7 files changed, 89 insertions(+), 75 deletions(-)
>
> --
> 1.7.9.5
>
>
> ______________________________________________________
> Linux MTD discussion mailing list
> http://lists.infradead.org/mailman/listinfo/linux-mtd/