From: Joel A Fernandes <[email protected]>
This series is fixes to get MMC working on AM33XX without HWMOD data.
On removal of HWMOD data, interrupt and register properties need to be provided
for the driver to function correctly.
Also instead of specifying offset in match->data, I've added bindings for the
same and specify it directly from DT as this is a hardware description.
Reposting Matt's remaining 2 patches on the bindings and dts.
Joel A Fernandes (3):
MMC: omap_hsmmc: Add reg_offset to DT bindings
ARM: dts: am33xx: Add interrupts and memory resources for MMC
mmc: omap_hsmmc: Add reg-offset to bindings documentation
Matt Porter (2):
ARM: dts: add AM33XX MMC support
mmc: omap_hsmmc: add generic DMA request support to the DT binding
.../devicetree/bindings/mmc/ti-omap-hsmmc.txt | 29 +++++++++++++-
arch/arm/boot/dts/am335x-bone.dts | 7 ++++
arch/arm/boot/dts/am335x-evm.dts | 7 ++++
arch/arm/boot/dts/am335x-evmsk.dts | 7 ++++
arch/arm/boot/dts/am33xx.dtsi | 40 ++++++++++++++++++++
drivers/mmc/host/omap_hsmmc.c | 7 +++-
6 files changed, 95 insertions(+), 2 deletions(-)
--
1.7.9.5
Hi Joel,
On 06/26/2013 05:28 AM, Joel A Fernandes wrote:
> On Tue, Jun 25, 2013 at 8:23 PM, Joel A Fernandes <[email protected]> wrote:
>> From: Joel A Fernandes <[email protected]>
>>
>> This series is fixes to get MMC working on AM33XX without HWMOD data.
>> On removal of HWMOD data, interrupt and register properties need to be provided
>> for the driver to function correctly.
>
> Please don't pull patches authored by me in this series. I have posted
> them with my wrong email address by mistake. I will correct my email
> during the next repost, after a round of review. Matt's patches in the
> series are still good though.
In that case, please repost these in two series to avoid pulling the DTS
files in the MMC tree.
Thanks,
Benoit
On Thursday 27 June 2013 07:56 PM, Joel A Fernandes wrote:
> Hi Benoit,
>
> On Thu, Jun 27, 2013 at 7:03 AM, Benoit Cousson <[email protected]> wrote:
>> Hi Joel,
>>
>> On 06/26/2013 05:28 AM, Joel A Fernandes wrote:
>>> On Tue, Jun 25, 2013 at 8:23 PM, Joel A Fernandes <[email protected]> wrote:
>>>> From: Joel A Fernandes <[email protected]>
>>>>
>>>> This series is fixes to get MMC working on AM33XX without HWMOD data.
>>>> On removal of HWMOD data, interrupt and register properties need to be provided
>>>> for the driver to function correctly.
>>>
>>> Please don't pull patches authored by me in this series. I have posted
>>> them with my wrong email address by mistake. I will correct my email
>>> during the next repost, after a round of review. Matt's patches in the
>>> series are still good though.
>>
>> In that case, please repost these in two series to avoid pulling the DTS
>> files in the MMC tree.
>
> The series has been replaced with the single patch:
> "[PATCH v2] ARM: dts: add AM33XX MMC support"
> http://marc.info/?l=linux-omap&m=137230403815943&w=2
>
> Some patches were squashed and others dropped in the series resulting
> in the single
> patch above. This patch should be good to apply
Hi Joel,
Before pushing mmc dts support for am335x, Can you please let me
know if below 2 patches [1] are merged/accepted,
IIRC these patches are needed for MMC to work properly.
I suggest to hold the dts patch until slave_sg_limit gets merged
otherwise it will cause regression in am335x mmc.
[PATCH v4 1/3] dmaengine: add dma_get_slave_sg_limits()
[PATCH v4 2/3] dma: edma: add device_slave_sg_limits() support
[1]
https://groups.google.com/forum/#!topic/linux.kernel/khTTcPtQpag
>
> Thanks,
>
> -Joel
>
Hi Balaji,
> -----Original Message-----
> From: Krishnamoorthy, Balaji T
> Sent: Thursday, June 27, 2013 11:27 AM
> To: Cousson, Benoit; Fernandes, Joel
> Cc: [email protected]; Tony Lindgren; Nori, Sekhar; Matt Porter; Grant Likely;
> Rob Herring; Vinod Koul; Mark Brown; Russell King; Rob Landley; Andrew
> Morton; Jason Kridner; Koen Kooi; Chris Ball; Devicetree Discuss; Linux OMAP
> List; Linux ARM Kernel List; Linux DaVinci Kernel List; Linux Kernel Mailing List;
> Linux MMC List; Arnd Bergmann
> Subject: Re: [PATCH 0/5] AM33xx: MMC resources from DT without HWMOD
> data
>
> On Thursday 27 June 2013 07:56 PM, Joel A Fernandes wrote:
> > Hi Benoit,
> >
> > On Thu, Jun 27, 2013 at 7:03 AM, Benoit Cousson <[email protected]> wrote:
> >> Hi Joel,
> >>
> >> On 06/26/2013 05:28 AM, Joel A Fernandes wrote:
> >>> On Tue, Jun 25, 2013 at 8:23 PM, Joel A Fernandes <[email protected]>
> wrote:
> >>>> From: Joel A Fernandes <[email protected]>
> >>>>
> >>>> This series is fixes to get MMC working on AM33XX without HWMOD
> data.
> >>>> On removal of HWMOD data, interrupt and register properties need to
> >>>> be provided for the driver to function correctly.
> >>>
> >>> Please don't pull patches authored by me in this series. I have
> >>> posted them with my wrong email address by mistake. I will correct
> >>> my email during the next repost, after a round of review. Matt's
> >>> patches in the series are still good though.
> >>
> >> In that case, please repost these in two series to avoid pulling the
> >> DTS files in the MMC tree.
> >
> > The series has been replaced with the single patch:
> > "[PATCH v2] ARM: dts: add AM33XX MMC support"
> > http://marc.info/?l=linux-omap&m=137230403815943&w=2
> >
> > Some patches were squashed and others dropped in the series resulting
> > in the single patch above. This patch should be good to apply
> Hi Joel,
>
> Before pushing mmc dts support for am335x, Can you please let me
> know if below 2 patches [1] are merged/accepted,
> IIRC these patches are needed for MMC to work properly.
>
> I suggest to hold the dts patch until slave_sg_limit gets merged
> otherwise it will cause regression in am335x mmc.
>
> [PATCH v4 1/3] dmaengine: add dma_get_slave_sg_limits()
> [PATCH v4 2/3] dma: edma: add device_slave_sg_limits() support
Actually this is an independent regression and is currently worked on by Lars-Peter Clausen [1].
Currently not having the MMC DTS merged is also a regression, so I suggest lets take it one
step at a time and get the DTS merged in first. Also there will be overhead of reposting this
DTS patch later again which will not change. What do you say?
Thanks,
-Joel
[1] https://patchwork.kernel.org/patch/2228041/
Adding Benoit's other email as current one is marked Away. Thanks.
> -----Original Message-----
> From: Krishnamoorthy, Balaji T
> Sent: Thursday, June 27, 2013 11:27 AM
> To: Cousson, Benoit; Fernandes, Joel
> Cc: [email protected]; Tony Lindgren; Nori, Sekhar; Matt Porter; Grant Likely;
> Rob Herring; Vinod Koul; Mark Brown; Russell King; Rob Landley; Andrew
> Morton; Jason Kridner; Koen Kooi; Chris Ball; Devicetree Discuss; Linux OMAP
> List; Linux ARM Kernel List; Linux DaVinci Kernel List; Linux Kernel Mailing List;
> Linux MMC List; Arnd Bergmann
> Subject: Re: [PATCH 0/5] AM33xx: MMC resources from DT without HWMOD
> data
>
> On Thursday 27 June 2013 07:56 PM, Joel A Fernandes wrote:
> > Hi Benoit,
> >
> > On Thu, Jun 27, 2013 at 7:03 AM, Benoit Cousson <[email protected]> wrote:
> >> Hi Joel,
> >>
> >> On 06/26/2013 05:28 AM, Joel A Fernandes wrote:
> >>> On Tue, Jun 25, 2013 at 8:23 PM, Joel A Fernandes <[email protected]>
> wrote:
> >>>> From: Joel A Fernandes <[email protected]>
> >>>>
> >>>> This series is fixes to get MMC working on AM33XX without HWMOD
> data.
> >>>> On removal of HWMOD data, interrupt and register properties need to
> >>>> be provided for the driver to function correctly.
> >>>
> >>> Please don't pull patches authored by me in this series. I have
> >>> posted them with my wrong email address by mistake. I will correct
> >>> my email during the next repost, after a round of review. Matt's
> >>> patches in the series are still good though.
> >>
> >> In that case, please repost these in two series to avoid pulling the
> >> DTS files in the MMC tree.
> >
> > The series has been replaced with the single patch:
> > "[PATCH v2] ARM: dts: add AM33XX MMC support"
> > http://marc.info/?l=linux-omap&m=137230403815943&w=2
> >
> > Some patches were squashed and others dropped in the series resulting
> > in the single patch above. This patch should be good to apply
> Hi Joel,
>
> Before pushing mmc dts support for am335x, Can you please let me
> know if below 2 patches [1] are merged/accepted,
> IIRC these patches are needed for MMC to work properly.
>
> I suggest to hold the dts patch until slave_sg_limit gets merged
> otherwise it will cause regression in am335x mmc.
>
> [PATCH v4 1/3] dmaengine: add dma_get_slave_sg_limits()
> [PATCH v4 2/3] dma: edma: add device_slave_sg_limits() support
>
> [1]
> https://groups.google.com/forum/#!topic/linux.kernel/khTTcPtQpag
>
> >
> > Thanks,
> >
> > -Joel
> >
On Thursday 27 June 2013 11:38 PM, Fernandes, Joel wrote:
> Hi Balaji,
>
<snip>
>>> Some patches were squashed and others dropped in the series resulting
>>> in the single patch above. This patch should be good to apply
>> Hi Joel,
>>
>> Before pushing mmc dts support for am335x, Can you please let me
>> know if below 2 patches [1] are merged/accepted,
>> IIRC these patches are needed for MMC to work properly.
>>
>> I suggest to hold the dts patch until slave_sg_limit gets merged
>> otherwise it will cause regression in am335x mmc.
>>
>> [PATCH v4 1/3] dmaengine: add dma_get_slave_sg_limits()
>> [PATCH v4 2/3] dma: edma: add device_slave_sg_limits() support
>
> Actually this is an independent regression and is currently worked on by Lars-Peter Clausen [1].
>
> Currently not having the MMC DTS merged is also a regression,
No, Right now mmc support is not added to am335x, so it is not really a regression
> so I suggest lets take it one
> step at a time and get the DTS merged in first. Also there will be overhead of reposting this
> DTS patch later again which will not change. What do you say?
>
If it gets enabled via dts, without slave sg limit support,
file read/write will fail, possibly corrupting file system.
> Thanks,
>
> -Joel
>
> [1] https://patchwork.kernel.org/patch/2228041/
>
> -----Original Message-----
> From: Krishnamoorthy, Balaji T
> Sent: Friday, June 28, 2013 10:49 AM
> To: Fernandes, Joel
> Cc: [email protected]; [email protected]; Tony Lindgren; Nori, Sekhar;
> Matt Porter; Grant Likely; Rob Herring; Vinod Koul; Mark Brown; Russell King;
> Rob Landley; Andrew Morton; Jason Kridner; Koen Kooi; Chris Ball; Devicetree
> Discuss; Linux OMAP List; Linux ARM Kernel List; Linux DaVinci Kernel List;
> Linux Kernel Mailing List; Linux MMC List; Arnd Bergmann
> Subject: Re: [PATCH 0/5] AM33xx: MMC resources from DT without HWMOD
> data
>
> On Thursday 27 June 2013 11:38 PM, Fernandes, Joel wrote:
> > Hi Balaji,
> >
>
> <snip>
>
> >>> Some patches were squashed and others dropped in the series
> >>> resulting in the single patch above. This patch should be good to
> >>> apply
> >> Hi Joel,
> >>
> >> Before pushing mmc dts support for am335x, Can you please let me know
> >> if below 2 patches [1] are merged/accepted, IIRC these patches are
> >> needed for MMC to work properly.
> >>
> >> I suggest to hold the dts patch until slave_sg_limit gets merged
> >> otherwise it will cause regression in am335x mmc.
> >>
> >> [PATCH v4 1/3] dmaengine: add dma_get_slave_sg_limits() [PATCH v4
> >> 2/3] dma: edma: add device_slave_sg_limits() support
> >
> > Actually this is an independent regression and is currently worked on by Lars-
> Peter Clausen [1].
> >
> > Currently not having the MMC DTS merged is also a regression,
>
> No, Right now mmc support is not added to am335x, so it is not really a
> regression.
>
> > so I suggest lets take it one
> > step at a time and get the DTS merged in first. Also there will be
> > overhead of reposting this DTS patch later again which will not change. What
> do you say?
> >
>
> If it gets enabled via dts, without slave sg limit support, file read/write will fail,
> possibly corrupting file system.
I agree with this. Thank you.
Tony, Benoit, I guess let's just not apply the DTS yet till the slave-sg limits support
is merged as well. Sorry for the noise. Thanks.
-Joel