Re: [PATCH v2 1/2] mmc: core: Add sdio_trigger_replug() API

[Ulf Hansson]

On Thu, 17 Oct 2019 at 02:22, Doug Anderson <dianders@xxxxxxxxxxxx> wrote:
>
> Hi,
>
> On Thu, Oct 10, 2019 at 7:11 AM Ulf Hansson <ulf.hansson@xxxxxxxxxx> wrote:
> >
> > On Mon, 22 Jul 2019 at 21:41, Douglas Anderson <dianders@xxxxxxxxxxxx> wrote:
> > >
> > > When using Marvell WiFi SDIO cards, it is not uncommon for Linux WiFi
> > > driver to fully lose the communication channel to the firmware running
> > > on the card.  Presumably the firmware on the card has a bug or two in
> > > it and occasionally crashes.
> > >
> > > The Marvell WiFi driver attempts to recover from this problem.
> > > Specifically the driver has the function mwifiex_sdio_card_reset()
> > > which is called when communcation problems are found.  That function
> > > attempts to reset the state of things by utilizing the mmc_hw_reset()
> > > function.
> > >
> > > The current solution is a bit complex because the Marvell WiFi driver
> > > needs to manually deinit and reinit the WiFi driver around the reset
> > > call.  This means it's going through a bunch of code paths that aren't
> > > normally tested.  However, complexity isn't our only problem.  The
> > > other (bigger) problem is that Marvell WiFi cards are often combo
> > > WiFi/Bluetooth cards and Bluetooth runs on a second SDIO func.  While
> > > the WiFi driver knows that it should re-init its own state around the
> > > mmc_hw_reset() call there is no good way to inform the Bluetooth
> > > driver.  That means that in Linux today when you reset the Marvell
> > > WiFi driver you lose all Bluetooth communication.  Doh!
> >
> > Thanks for a nice description to the problem!
> >
> > In principle it makes mmc_hw_reset() quite questionable to use for
> > SDIO func drivers, at all. However, let's consider that for later.
>
> Yeah, unless you somehow knew that your card would only have one function.
>
>
> > > One way to fix the above problems is to leverage a more standard way
> > > to reset the Marvell WiFi card where we go through the same code paths
> > > as card unplug and the card plug.  In this patch we introduce a new
> > > API call for doing just that: sdio_trigger_replug().  This API call
> > > will trigger an unplug of the SDIO card followed by a plug of the
> > > card.  As part of this the card will be nicely reset.
> >
> > I have been thinking back and forth on this, exploring various
> > options, perhaps adding some callbacks that the core could invoke to
> > inform the SDIO func drivers of what is going on.
> >
> > Although, in the end this boils done to complexity and I think your
> > approach is simply the most superior in regards to this. However, I
> > think there is a few things that we can do to even further simply your
> > approach, let me comment on the code below.
>
> Right.  Unplugging / re-plugging is sorta gross / inelegant, but it is
> definitely simpler and nice that it doesn't add so many new code
> paths.  For cases where you're just trying to re-init things with a
> hammer it works pretty well.
>
>
> > > Signed-off-by: Douglas Anderson <dianders@xxxxxxxxxxxx>
> > > Reviewed-by: Matthias Kaehlcke <mka@xxxxxxxxxxxx>
> > > ---
> > >
> > > Changes in v2:
> > > - s/routnine/routine (Brian Norris, Matthias Kaehlcke).
> > > - s/contining/containing (Matthias Kaehlcke).
> > > - Add Matthias Reviewed-by tag.
> > >
> > >  drivers/mmc/core/core.c       | 28 ++++++++++++++++++++++++++--
> > >  drivers/mmc/core/sdio_io.c    | 20 ++++++++++++++++++++
> > >  include/linux/mmc/host.h      | 15 ++++++++++++++-
> > >  include/linux/mmc/sdio_func.h |  2 ++
> > >  4 files changed, 62 insertions(+), 3 deletions(-)
> > >
> > > diff --git a/drivers/mmc/core/core.c b/drivers/mmc/core/core.c
> > > index 221127324709..5da365b1fdb4 100644
> > > --- a/drivers/mmc/core/core.c
> > > +++ b/drivers/mmc/core/core.c
> > > @@ -2161,6 +2161,12 @@ int mmc_sw_reset(struct mmc_host *host)
> > >  }
> > >  EXPORT_SYMBOL(mmc_sw_reset);
> > >
> > > +void mmc_trigger_replug(struct mmc_host *host)
> > > +{
> > > +       host->trigger_replug_state = MMC_REPLUG_STATE_UNPLUG;
> > > +       _mmc_detect_change(host, 0, false);
> > > +}
> > > +
> > >  static int mmc_rescan_try_freq(struct mmc_host *host, unsigned freq)
> > >  {
> > >         host->f_init = freq;
> > > @@ -2214,6 +2220,11 @@ int _mmc_detect_card_removed(struct mmc_host *host)
> > >         if (!host->card || mmc_card_removed(host->card))
> > >                 return 1;
> > >
> > > +       if (host->trigger_replug_state == MMC_REPLUG_STATE_UNPLUG) {
> > > +               mmc_card_set_removed(host->card);
> > > +               return 1;
> >
> > Do you really need to set state of the card to "removed"?
> >
> > If I understand correctly, what you need is to allow mmc_rescan() to
> > run a second time, in particular for non removable cards.
> >
> > In that path, mmc_rescan should find the card being non-functional,
> > thus it should remove it and then try to re-initialize it again. Etc.
> >
> > Do you want me to send a patch to show you what I mean!?
>
> If you don't mind, that would probably be easiest.  I've totally
> swapped out all of the implementation details of this from my brain
> now, but if I saw a patch from you it would be easy for me to analyze
> it and test it.

Alright, I think I owe you that because of my slow review pase. :-)

Patches are coming soon!

Kind regards
Uffe