Re: [PATCH edk2-platforms] Silicon/Socionext/SynQuacer: add configurable eMMC support

[Ard Biesheuvel <ard.biesheuvel@linaro.org>]

On 30 January 2018 at 11:47, Leif Lindholm <leif.lindholm@linaro.org> wrote:
> On Tue, Jan 30, 2018 at 11:14:31AM +0000, Ard Biesheuvel wrote:
>> On 30 January 2018 at 11:00, Leif Lindholm <leif.lindholm@linaro.org> wrote:
>> > On Tue, Jan 30, 2018 at 10:32:40AM +0000, Ard Biesheuvel wrote:
>> >> Implement support for the SynQuacer eMMC controller. This involves an
>> >> implementation of the SD/MMC override protocol to handle a couple of
>> >> quirks that would otherwise prevent this IP from being driven by the
>> >> generic SDHCI driver.
>> >>
>> >> Also, add a HII page to the PlatformDxe driver that allows eMMC support
>> >> to be enabled, and wire it up for both DeveloperBox and EVB.
>> >>
>> >> Contributed-under: TianoCore Contribution Agreement 1.1
>> >> Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
>> >> ---
>> >> Now that the core support for the SD/MMC override protocol is finally
>> >> merged, resubmit this again. I dropped Leif's R-b given that I have
>> >> now added DeveloperBox, as well as a HII option to enable eMMC.
>> >
>> > Couple of minor comments/suggestions below and a question.
>> >
>> >>  Platform/Socionext/DeveloperBox/DeveloperBox.dsc                                    |   8 +
>> >>  Platform/Socionext/DeveloperBox/DeveloperBox.fdf                                    |   7 +
>> >>  Platform/Socionext/SynQuacerEvalBoard/SynQuacerEvalBoard.dsc                        |   8 +
>> >>  Platform/Socionext/SynQuacerEvalBoard/SynQuacerEvalBoard.fdf                        |   7 +
>> >>  Silicon/Socionext/SynQuacer/DeviceTree/SynQuacer.dtsi                               |   1 -
>> >>  Silicon/Socionext/SynQuacer/DeviceTree/SynQuacerEvalBoard.dts                       |   4 -
>> >>  Silicon/Socionext/SynQuacer/Drivers/PlatformDxe/Emmc.c                              | 203 ++++++++++++++++++++
>> >>  Silicon/Socionext/SynQuacer/Drivers/PlatformDxe/PlatformDxe.c                       |   5 +
>> >>  Silicon/Socionext/SynQuacer/Drivers/PlatformDxe/PlatformDxe.h                       |   9 +
>> >>  Silicon/Socionext/SynQuacer/Drivers/PlatformDxe/PlatformDxe.inf                     |   4 +
>> >>  Silicon/Socionext/SynQuacer/Drivers/PlatformDxe/PlatformDxeHii.uni                  |   6 +
>> >>  Silicon/Socionext/SynQuacer/Drivers/PlatformDxe/PlatformDxeHii.vfr                  |   8 +
>> >>  Silicon/Socionext/SynQuacer/Include/Platform/VarStore.h                             |   6 +-
>> >>  Silicon/Socionext/SynQuacer/Library/SynQuacerDtbLoaderLib/SynQuacerDtbLoaderLib.c   |  23 ++-
>> >>  Silicon/Socionext/SynQuacer/Library/SynQuacerDtbLoaderLib/SynQuacerDtbLoaderLib.inf |   1 +
>> >>  15 files changed, 287 insertions(+), 13 deletions(-)
>> >>
>> >
>> >> +/**
>> >> +
>> >> +  Override function for SDHCI capability bits
>> >> +
>> >> +  @param[in]      PassThru              A pointer to the
>> >> +                                        EFI_SD_MMC_PASS_THRU_PROTOCOL instance.
>> >> +  @param[in]      ControllerHandle      The EFI_HANDLE of the controller.
>> >> +  @param[in]      Slot                  The 0 based slot index.
>> >> +  @param[in,out]  SdMmcHcSlotCapability The SDHCI capability structure.
>> >> +
>> >> +  @retval EFI_SUCCESS           The override function completed successfully.
>> >> +  @retval EFI_NOT_FOUND         The specified controller or slot does not exist.
>> >> +  @retval EFI_INVALID_PARAMETER SdMmcHcSlotCapability is NULL
>> >> +
>> >> +**/
>> >> +STATIC
>> >> +EFI_STATUS
>> >> +EFIAPI
>> >> +SynQuacerSdMmcCapability (
>> >> +  IN      EFI_HANDLE                      ControllerHandle,
>> >> +  IN      UINT8                           Slot,
>> >> +  IN  OUT VOID                            *SdMmcHcSlotCapability
>> >> +  )
>> >> +{
>> >> +  UINT64 Capability;
>> >> +
>> >> +  if (ControllerHandle != mSdMmcControllerHandle || Slot != 0) {
>> >
>> > This test pattern repeats below, does it suggest a macro?
>> >
>>
>> I don't see how that would clear things up tbh, and the pattern occurs
>> only twice
>>
>> #define IS_OUR_QUIRKY_SDMMC_CONTROLLER(Handle, Slot) \
>>   ((Handle) == mSdMmcControllerHandle && (Slot) == 0)
>>
>> if (!IS_OUR_QUIRKY_SDMMC_CONTROLLER(ControllerHandle, Slot) {
>>   return EFI_SUCCESS;
>> }
>>
>> I can change it if you want, or add a comment if the condition is not
>> self-explanatory enough.
>
> It's just an awful lot of logical operations on a single line.
> 'ControllerHandle != mSdMmcControllerHandle' is reasonably easy to
> figure out, but '|| Slot != 0' looks a bit random there.
>
> A comment would be sufficient.
>
> Another option would be to reduce the number of logical operations by
> two by doing
>   if (ControllerHandle == mSdMmcControllerHandle && Slot == 0) {
> and doing the body inside the if-statement.
>
> That's a bit uglier in the next function, but I would expect it
> follows the paradigm of "handle most likely case first"?
>

Actually, come to think of it, the slot number is completely
redundant. The non-discoverable SDHCI controller we expose only
implements a single slot, so something is terribly wrong if
ControllerHandle == mSdMmcControllerHandle && Slot != 0.

So I will reduce the sequence to

if (ControllerHandle != mSdMmcControllerHandle) {
  return EFI_SUCCESS;
}

ASSERT (Slot == 0);

instead. Ok?

>> >> diff --git a/Silicon/Socionext/SynQuacer/Drivers/PlatformDxe/PlatformDxeHii.uni b/Silicon/Socionext/SynQuacer/Drivers/PlatformDxe/PlatformDxeHii.uni
>> >> index b274d12ed2c6..2eca8bbba8c3 100644
>> >> --- a/Silicon/Socionext/SynQuacer/Drivers/PlatformDxe/PlatformDxeHii.uni
>> >> +++ b/Silicon/Socionext/SynQuacer/Drivers/PlatformDxe/PlatformDxeHii.uni
>> >> @@ -27,3 +27,9 @@
>> >>
>> >>  #string STR_PCIE_MAX_SPEED_UNLIMITED   #language en-US "Unlimited"
>> >>  #string STR_PCIE_MAX_SPEED_GEN1        #language en-US "Gen1 (2.5 GT/s)"
>> >> +
>> >> +#string STR_EMMC_ENABLE_PROMPT         #language en-US "Enable on-board eMMC"
>> >> +#string STR_EMMC_ENABLE_HELP           #language en-US "Enable the on-board eMMC for booting and for use by the OS."
>> >> +
>> >> +#string STR_EMMC_DISABLED              #language en-US "Disabled"
>> >> +#string STR_EMMC_ENABLED               #language en-US "Enabled"
>> >
>> > Perhaps a random question, but ...
>> > Why am I seeing this in cleartext in the patch? Is it really a unicode file?
>>
>> Given that we support UTF-8 in .uni files these days, and the fact
>> that all characters used are in the 7-bit ASCII range, it doesn't
>> really make a difference, I guess.
>
> Fair enough.
>
> /
>     Leif