[PATCH v4 0/6] support other fsl SoCs with usbmisc + small fixes

Alexander Shishkin alexander.shishkin at linux.intel.com
Tue Mar 26 23:08:05 EST 2013


Michael Grzeschik <mgr at pengutronix.de> writes:

> Hi Alexander, Fabio, Greg,
>
> On Mon, Mar 25, 2013 at 10:11:15AM -0300, Fabio Estevam wrote:
>> Hi Alexander,
>> 
>> On Thu, Jan 24, 2013 at 7:42 AM, Alexander Shishkin
>> <alexander.shishkin at linux.intel.com> wrote:
>> > Peter Chen <peter.chen at freescale.com> writes:
>> >
>> >> On Tue, Nov 27, 2012 at 05:16:55PM +0100, Michael Grzeschik wrote:
>> >>> Nearly every SoC from Freescale has this non-core usb registers. This series
>> >>> adds support for more users of this driver.
>> >>>
>> >>> This series is based on Peter Chen's work. Its needed to merge his master branch
>> >>> before applying this series:
>> >>>
>> >>> https://github.com/hzpeterchen/linux-usb.git
>> >>
>> >> I have tested it at i.mx6q sabrelite board, it works good.
>> >>
>> >> I have pushed your commit to my git, please cc me
>> >> your coming chipidea patches, thanks.
>> >>
>> >> Alex, please add:
>> >>
>> >> Reviewed-by: Peter Chen <peter.chen at freescale.com>
>> >> Tested-by: Peter Chen <peter.chen at freescale.com>
>> >
>> > Looks good, queueing this one for submission.
>> 
>> Could you please let us know what is preventing this series to be
>> applied?

I think the main problem is that there is a bunch of patchsets with
interdependencies that are quite obscure at least to me. Also, I see a
whole lot of patches that were posted some time ago, commented on, never
reposted.
Right now, I'm looking at one patchset which you have sent last
friday. So far it looks like a good candidate, although I see a few
comments there that suggest that it still needs fixing.

> That's what our chipidea usb disaster currently looks like in detail:
>

this patchset

> http://git.pengutronix.de/git/mgr/linux.git v3.9/topic/usb-phy
> 4414a59 Merge 'v3.9/topic/usb-phy' into usb-3.9
> a806027 usb: otg: mxs-phy: remove clkgate/sftrst clearing
> 026881b usb: otg: mxs-phy: Improve mxs phy initialization
> ed42fa9 usb: mxs-phy: register phy at usb phy library
> d2783aa usb: phy: remove NULL pointer check in usb_add_phy()

looks like Felipe's call, doesn't it? I also can't seem to find these in
my inbox. If you want me to ack them, you should at least send them to
me. :)

Felipe, btw, is also often not copied on chipidea-related phy patches
that I see. And just in case, Cc'ing him on this thread.

> /*
>  * The series usb-ci-udc has already been posted some times
>  * but it currently got no replies by Alexander.
>  */
> http://git.pengutronix.de/git/mgr/linux.git v3.9/topic/usb-ci-udc
> e548c9d Merge 'v3.9/topic/usb-ci-udc' into usb-3.9
> 80538f3 usb: chipidea: udc: add multiple td support to hardware_{en,de}queue
> 02e1c7b usb: chipidea: udc: manage dynamic amount of tds with an linked list
> 0ebb8f4 usb: chipidea: udc: prepare qhead with dma_alloc_coherent
> 39be042 usb: chipidea: udc: add the define TD_PAGE_COUNT and fix all users
> c2f0562 usb: chipidea: udc: use gfp_flags in _hardware_enqueue
> f818222 usb: chipidea: udc: move _ep_queue into an unlocked function
> 8aea3b0 usb: chipidea: udc: configure iso endpoints
> 8e61a12 usb: chipidea: udc: rework ep_enable cap setting
> e5ceed7 usb: chipidea: udc: move ZLT flag change to ep_enable
> 2e33742 usb: chipidea: udc: read status of td only once in hardware_dequeue
> 516803b usb: chipidea: udc: only clear active and halted bits in qhead
> ee95a49 usb: chipidea: udc: use {read,write}l to handle mapped data

Yes, that's the one. You were in discussion with Peter about some of
these patches, so I haven't had a close look before.

> http://git.pengutronix.de/git/mgr/linux.git v3.9/topic/imx-usbmisc
> 1174002 Merge 'v3.9/topic/imx-usbmisc' into usb-3.9
> f406383 usb: chipidea: usbmisc: skip clocks on i.MX6
> d5c8d53 usb: chipidea: usbmisc: use module_platform_driver

These two do not exist in my inbox.

> 105d5d0 usb: chipidea: usbmisc: add support for ahb, ipg and per clock

This one is ok, but it was posted in a patchset, which claims to depend
on Peter's vbus detection patchset, which is still a work in progress.

> http://git.pengutronix.de/git/mgr/linux.git v3.9/topic/imx-usb
> df11f8b Merge 'v3.9/topic/imx-usb' into usb-3.9
> 9a967c4 usb: chipidea: udc: add force-full-speed option

The only version that I have of this patch in my inbox has a problem
wrt DEVLC vs PORTSC in lpm/nolpm devices. I have seen you fix this
problem for the phy mode configuration patch, but not for this one.

> a02af86 USB: chipidea: set STS bit in serial mode

Doesn't exist in my inbox.

> af34326 usb: chipidea: ci13xxx_imx: add 2nd and 3rd clock to support
> imx5x and newer

This patch started a discussion which didn't end in a revised version of
the patch in my inbox.

> 5b7f6d8 usb: chipidea: pci: mark platformdata as static

That obviously should go in. It is not applied because it was posted as
a part of the patchset that required fixing.

>
> /*
>  * All above branches depent on the lower branches of Alexander and Peter
>  * which we currently have rebased on v3.9-rc3 and on top of each other.
>  * It is named alex-and-peter-for-3.9.
>  */
> http://git.pengutronix.de/git/mgr/linux.git alex-and-peter-for-3.9
>
> /*
>  * Peters branch already contain the usbmisc series,
>  * but they never made it to ci-for-greg, although
>  * Alexander already mentioned (24 Jan 2013) to queue them.
>  */
> https://github.com/hzpeterchen/linux-usb.git master^^
> 4d36511 Merge 'v3.9/topic/usb-peterchen'
> 90314f3 usb: chipidea: usbmisc: add post handling and errata fix for mx25
> edde2ad usb: chipidea: usbmisc: add mx53 support
> 6150722 usb: chipidea: usbmisc: prepare driver to handle more than one soc
> 542384e usb: chipidea: usbmisc: fix a potential race condition
> cdffb44 usb: chipidea: usbmisc: unset global varibale usbmisc on driver remove
> b8e47e3 usb: chipidea: usbmisc: rename file, struct and functions to usbmisc_imx
> c5863f9 usb: chipidea: imx: select usb id pin using syscon interface
> 8c8a1f5 usb: chipidea: udc: fix the oops when plugs in usb cable after rmmod gadget
> c703723 usb: chipidea: imx: add getting vbus regulator code
> e897d66 usb: chipidea: delete the delayed work
> 4c0be3e usb: chipidea: add vbus regulator control
> f1e52f8 usb: chipidea: udc: retire the flag CI13_PULLUP_ON_VBUS
> af4e743 usb: chipidea: udc: add pullup/pulldown dp at hw_device_state
> 1be0011 usb: chipidea: add otg id switch and vbus connect/disconnect detect
> 7c3c62d usb: chipidea: add otg file

This patchset is a work in progress, it can't be applied anywhere yet.

> 618bfec Revert "USB: chipidea: add vbus detect for udc"

This one, though, I will pick.

> bfe5248 USB chipidea i.MX: use devm_usb_get_phy_by_phandle to get phy
> 705620f USB mxs-phy: Register phy with framework
> f5e9796 USB chipidea i.MX: introduce dr_mode property
> dc622c4 USB chipidea: introduce dual role mode pdata flags
> b1d4a9e USB: chipidea: add PTW and PTS handling
> 747eee3 USB: chipidea: ci13xxx-imx: create dynamic platformdata

> 2f2bb89 USB: add devicetree helpers for determining dr_mode and
> phy_type

This guy was commented on and didn't show up in my inbox after that.

> 4f30039 USB: move bulk of otg/otg.c to phy/phy.c

this one seems to be in Felipe's testing branch

> 760d9a8 nop-xceiv: make nop to use usb_add_phy_dev

doesn't exist in my inbox and looks like Felipe's call.

>
> /*
>  * This branch is already one month old and stays untouched
>  * /
> https://github.com/hzpeterchen/linux-ci.git ci-for-greg
> e0fe4d5 Merge 'ci-for-greg'
> e160513 usb: chipidea: don't redefine __ffs()
> 93b27aa usb: chipidea: fix precedence bug in ci_requests_show()
> 67d0531 usb: chipidea: move debug files creation/removal to the core
> 5a958dd usb: chipidea: move role to debugfs
> 038429c usb: chipidea: convert debug entries in sysfs to debugfs
> 25dc0fb usb: chipidea: replace interrupt accounting with tracepoints
> dc43e8c usb: chipidea: convert events to tracepoints
> a84a5bf usb: chipidea: trim include list in the core
> f2bf55e usb: chipidea: trim include list in udc code
> adf1889 usb: chipidea: drop redundant includes

Yes, I am still waiting for the patches that can be actually
applied. There is one fix from Felipe and a few that I mentioned above
and probably also one or two from Svetoslav.

So, if you feel like including any of the above patches in this round,
please add them on top of the patchset that you're currently resending
or at least point me at the fixed versions (that could be, for example,
the force fullspeed patch) or the patches that can be taken out of
context of the patchsets that they were sent in.

Regards,
--
Alex


More information about the devicetree-discuss mailing list