Re: [PATCH V2 00/11] mmc: Add support for drive strength for eMMCs

[Adrian Hunter <adrian.hunter@intel.com>]

On 11/05/15 13:23, Ulf Hansson wrote:
> On 11 May 2015 at 11:29, Adrian Hunter <adrian.hunter@intel.com> wrote:
>> On 06/02/15 14:12, Adrian Hunter wrote:
>>> Hi
>>>
>>> Here is V2 of some patches to enable a host controller to select
>>> driver strength for eMMCs using HS200 or HS400. These are based
>>> on top of the re-tuning series.
>>>
>>> There can be some confusion over the term "driver strength".
>>> SD calls it "drive strength" for the card but "driver type"
>>> for the host. Whereas JEDEC calls it both "driver strength"
>>> and "driver type". The values are the same for both SD
>>> and eMMC:
>>>
>>>   Value   Driver Type   Relative strength
>>>     0       B             x1  default and mandatory support
>>>     1       A             x1.5
>>>     2       C             x0.75
>>>     3       D             x0.5
>>>
>>> Except eMMC also defines value 4:
>>>
>>>   Value   Driver Type   Relative strength
>>>     4                     x1.2
>>>
>>>
>>> Changes in V2:
>>>
>>>     mmc: core: Add function to read driver-strength device property
>>>         Dropped because there are still questions over how to use
>>>         device properties.
>>>
>>>     mmc: sdhci-pci: Add support for drive strength selection for SPT
>>>         Amended to reflect the change above.
>>>
>>>
>>> Adrian Hunter (11):
>>>       mmc: core: Reset driver type to default
>>>       mmc: core: Allow card drive strength to be different to host
>>>       mmc: core: Simplify card drive strength mask
>>>       mmc: core: Add 'card' to drive strength selection callback
>>>       mmc: core: Factor out common code in drive strength selection
>>>       mmc: core: Record card drive strength
>>>       mmc: mmc: Read card's valid driver strength mask
>>>       mmc: mmc: Add driver strength selection
>>>       mmc: sdhci: Add a callback to select drive strength
>>>       mmc: sdhci-pci: Add support for drive strength selection for SPT
>>>       mmc: sdhci-pci: Enable HS400 for some Intel host controllers
>>>
>>>  drivers/mmc/core/core.c            | 39 ++++++++++++++++++
>>>  drivers/mmc/core/core.h            |  2 +
>>>  drivers/mmc/core/mmc.c             | 46 +++++++++++++++++----
>>>  drivers/mmc/core/sd.c              | 69 ++++++++-----------------------
>>>  drivers/mmc/core/sdio.c            | 77 +++++++++++-----------------------
>>>  drivers/mmc/host/sdhci-pci-data.c  |  3 ++
>>>  drivers/mmc/host/sdhci-pci.c       | 84 ++++++++++++++++++++++++++++++++++++++
>>>  drivers/mmc/host/sdhci-pci.h       |  4 ++
>>>  drivers/mmc/host/sdhci.c           | 13 ++++++
>>>  drivers/mmc/host/sdhci.h           |  4 ++
>>>  include/linux/mmc/card.h           |  2 +
>>>  include/linux/mmc/host.h           |  4 +-
>>>  include/linux/mmc/mmc.h            |  4 ++
>>>  include/linux/mmc/sdhci-pci-data.h |  2 +
>>>  14 files changed, 240 insertions(+), 113 deletions(-)
>>
>> Hi Ulf
>>
>> These patches still apply. Would have time to look at them?
> 
> Thanks for the reminder!
> 
> I am walking through my backlog, but I don't want to give you any
> promises about when, since I never seems to be able to keep them. :-)

Hi

Here's another reminder :-)

I would really like these patches and there is really not much to review.

Patch 1 is quite trivial - only 1 line changed.

Patch 2 is a change the select_drive_strength() host op to allow card driver
strength to be different from host driver strength. It has no effect on
current drivers.

Patch 3 is just tidying up.

Patch 4 is another change to the select_drive_strength() host op (to add
'card' as a parameter) and also has no effect on current drivers.

Patch 5 is another tidy up.

Patch 6 is a very small patch to add drive_strength to struct mmc_card.

Patch 7 is a very small change to read eMMC supported driver strengths from
EXT-CSD.

Patch 8 adds driver strength selection for eMMC

Patch 9, 10 implement the select_drive_strength() for Intel SPT with the
sdhci-pci driver.

Patch 11 finally just enables HS400 for Intel

This shouldn't really take that long to review ;-)

Regards
Adrian