[PATCH/RFC] tmio_mmc: keep card-detect interrupts enabled

[PATCH/RFC] tmio_mmc: keep card-detect interrupts enabled

[Guennadi Liakhovetski]

On SuperH platforms the SDHI controller does not produce any command IRQs 
after a completed IO. This leads to card-detect interrupts staying 
disabled. Do not disable card-detect interrupts on DATA IRQs.

Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
---

Marked as RFC because I'm not really sure this is a correct approach to 
fix this problem, and whether this will have negative effect on other 
tmio_mmc MFD users.

diff --git a/drivers/mmc/host/tmio_mmc.h b/drivers/mmc/host/tmio_mmc.h
index c676767..0b31d44 100644
--- a/drivers/mmc/host/tmio_mmc.h
+++ b/drivers/mmc/host/tmio_mmc.h
@@ -55,10 +55,8 @@
 /* Define some IRQ masks */
 /* This is the mask used at reset by the chip */
 #define TMIO_MASK_ALL           0x837f031d
-#define TMIO_MASK_READOP  (TMIO_STAT_RXRDY | TMIO_STAT_DATAEND | \
-		TMIO_STAT_CARD_REMOVE | TMIO_STAT_CARD_INSERT)
-#define TMIO_MASK_WRITEOP (TMIO_STAT_TXRQ | TMIO_STAT_DATAEND | \
-		TMIO_STAT_CARD_REMOVE | TMIO_STAT_CARD_INSERT)
+#define TMIO_MASK_READOP  (TMIO_STAT_RXRDY | TMIO_STAT_DATAEND)
+#define TMIO_MASK_WRITEOP (TMIO_STAT_TXRQ | TMIO_STAT_DATAEND)
 #define TMIO_MASK_CMD     (TMIO_STAT_CMDRESPEND | TMIO_STAT_CMDTIMEOUT | \
 		TMIO_STAT_CARD_REMOVE | TMIO_STAT_CARD_INSERT)
 #define TMIO_MASK_IRQ     (TMIO_MASK_READOP | TMIO_MASK_WRITEOP | TMIO_MASK_CMD)

Re: [PATCH/RFC] tmio_mmc: keep card-detect interrupts enabled

[Guennadi Liakhovetski]

Hi Andrew

I originally submitted this [1] patch on the 6th of November, then 
accidentally the MMC list got dropped from CC and the discussion moved to 
another irrelevant list, Ian's last reply is at [2]. Unfortunately, Ian 
hasn't replied to my reminders since then, and I don't see this patch 
anywhere upstream. Is there any way we could get it into 2.6.33?

Thanks
Guennadi

[1] http://thread.gmane.org/gmane.linux.kernel.mmc/704
[2] http://thread.gmane.org/gmane.linux.drivers.video-input-infrastructure/11981/focus=11995

On Wed, 16 Dec 2009, Guennadi Liakhovetski wrote:

> On Sat, 28 Nov 2009, Guennadi Liakhovetski wrote:
> 
> > On Tue, 10 Nov 2009, Ian Molton wrote:
> > 
> > > Well if they are only masked they shouldnt stop being asserted. But we
> > > should unmask them again.
> > > 
> > > Im not really sure we should mask them anyway, with the card possibly
> > > being gone... Will need to look into it further.
> > 
> > Hi Ian
> > 
> > What's the status of this patch? Are you going to queue it for 2.6.33?
> 
> Ian, ping? I still don't see this patch in any (Linus' / next) upstream 
> tree and the merge window will be closing soon. Any update?
> 
> Thanks
> Guennadi
> 
> > 
> > Thanks
> > Guennadi
> > 
> > > 
> > > 2009/11/9 Guennadi Liakhovetski <g.liakhovetski@gmx.de>:
> > > > (re-adding accidentally dropped ML)
> > > >
> > > > On Mon, 9 Nov 2009, Ian Molton wrote:
> > > >
> > > >> Well, I presume we want to know when the card gets removed :)
> > > >
> > > > Sure, that's why we shouldn't mask those interrupts:-) If they do get
> > > > masked and missed, I do not know, if the interrupt remains pending in this
> > > > case, because they never get detected then:)
> > > >
> > > >>
> > > >> 2009/11/9 Guennadi Liakhovetski <g.liakhovetski@gmx.de>:
> > > >> > Hi Ian
> > > >> >
> > > >> > Why did you drop all CCs?
> > > >> >
> > > >> > On Mon, 9 Nov 2009, Ian Molton wrote:
> > > >> >
> > > >> >> I havent looked at the consequences for the driver if a insert IRQ
> > > >> >> occurs during IO, however it seems logical that we should not
> > > >> >> permanently mask the IRQ.
> > > >> >>
> > > >> >> I presume that the IRQ remains pending?
> > > >> >
> > > >> > Don't know, never checked. Is this important to know?
> > > >> >
> > > >> > Thanks
> > > >> > Guennadi
> > > >> >
> > > >> >>
> > > >> >> 2009/11/6 Guennadi Liakhovetski <g.liakhovetski@gmx.de>:
> > > >> >> > On SuperH platforms the SDHI controller does not produce any command IRQs
> > > >> >> > after a completed IO. This leads to card-detect interrupts staying
> > > >> >> > disabled. Do not disable card-detect interrupts on DATA IRQs.
> > > >> >> >
> > > >> >> > Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
> > > >> >> > ---
> > > >> >> >
> > > >> >> > Marked as RFC because I'm not really sure this is a correct approach to
> > > >> >> > fix this problem, and whether this will have negative effect on other
> > > >> >> > tmio_mmc MFD users.
> > > >> >> >
> > > >> >> > diff --git a/drivers/mmc/host/tmio_mmc.h b/drivers/mmc/host/tmio_mmc.h
> > > >> >> > index c676767..0b31d44 100644
> > > >> >> > --- a/drivers/mmc/host/tmio_mmc.h
> > > >> >> > +++ b/drivers/mmc/host/tmio_mmc.h
> > > >> >> > @@ -55,10 +55,8 @@
> > > >> >> > ═/* Define some IRQ masks */
> > > >> >> > ═/* This is the mask used at reset by the chip */
> > > >> >> > ═#define TMIO_MASK_ALL ═ ═ ═ ═ ═ 0x837f031d
> > > >> >> > -#define TMIO_MASK_READOP ═(TMIO_STAT_RXRDY | TMIO_STAT_DATAEND | \
> > > >> >> > - ═ ═ ═ ═ ═ ═ ═ TMIO_STAT_CARD_REMOVE | TMIO_STAT_CARD_INSERT)
> > > >> >> > -#define TMIO_MASK_WRITEOP (TMIO_STAT_TXRQ | TMIO_STAT_DATAEND | \
> > > >> >> > - ═ ═ ═ ═ ═ ═ ═ TMIO_STAT_CARD_REMOVE | TMIO_STAT_CARD_INSERT)
> > > >> >> > +#define TMIO_MASK_READOP ═(TMIO_STAT_RXRDY | TMIO_STAT_DATAEND)
> > > >> >> > +#define TMIO_MASK_WRITEOP (TMIO_STAT_TXRQ | TMIO_STAT_DATAEND)
> > > >> >> > ═#define TMIO_MASK_CMD ═ ═ (TMIO_STAT_CMDRESPEND | TMIO_STAT_CMDTIMEOUT | \
> > > >> >> > ═ ═ ═ ═ ═ ═ ═ ═TMIO_STAT_CARD_REMOVE | TMIO_STAT_CARD_INSERT)
> > > >> >> > ═#define TMIO_MASK_IRQ ═ ═ (TMIO_MASK_READOP | TMIO_MASK_WRITEOP | TMIO_MASK_CMD)
> > > >
> > > > Thanks
> > > > Guennadi
> > > > ---
> > > > Guennadi Liakhovetski, Ph.D.
> > > > Freelance Open-Source Software Developer
> > > > http://www.open-technology.de/
> > > >
> > > 
> > > 
> > > 
> > > -- 
> > > Ian Molton
> > > Linux, Automotive, and other hacking:
> > > http://www.mnementh.co.uk/
> > > 
> > 
> > ---
> > Guennadi Liakhovetski
> > 
> 
> ---
> Guennadi Liakhovetski, Ph.D.
> Freelance Open-Source Software Developer
> http://www.open-technology.de/
> 

---
Guennadi Liakhovetski

Re: [PATCH/RFC] tmio_mmc: keep card-detect interrupts enabled

[Andrew Morton]

On Thu, 17 Dec 2009 21:25:52 +0100 (CET)
Guennadi Liakhovetski <g.liakhovetski@gmx.de> wrote:

> I originally submitted this [1] patch on the 6th of November, then 
> accidentally the MMC list got dropped from CC and the discussion moved to 
> another irrelevant list, Ian's last reply is at [2]. Unfortunately, Ian 
> hasn't replied to my reminders since then, and I don't see this patch 
> anywhere upstream. Is there any way we could get it into 2.6.33?

Please redo and retest it against curent mainline and do a full resend.

Re: [PATCH/RFC] tmio_mmc: keep card-detect interrupts enabled

[Guennadi Liakhovetski]

Hi Andrew

On Thu, 17 Dec 2009, Andrew Morton wrote:

> On Thu, 17 Dec 2009 21:25:52 +0100 (CET)
> Guennadi Liakhovetski <g.liakhovetski@gmx.de> wrote:
> 
> > I originally submitted this [1] patch on the 6th of November, then 
> > accidentally the MMC list got dropped from CC and the discussion moved to 
> > another irrelevant list, Ian's last reply is at [2]. Unfortunately, Ian 
> > hasn't replied to my reminders since then, and I don't see this patch 
> > anywhere upstream. Is there any way we could get it into 2.6.33?
> 
> Please redo and retest it against curent mainline and do a full resend.

As I tried to re-test, I found out, that some other patches from Ian are 
currently missing in the mainline, so, even though this my specific patch 
doesn't depend on those changes, unfortunately, I currently cannot test 
tmio_mmc on sh. Will resubmit as soon as those patches hit mainline.

Thanks
Guennadi
---
Guennadi Liakhovetski, Ph.D.
Freelance Open-Source Software Developer
http://www.open-technology.de/

[PATCH] tmio_mmc: keep card-detect interrupts enabled

[Guennadi Liakhovetski]

On SuperH platforms the SDHI controller does not produce any command IRQs after
a completed IO. This leads to card-detect interrupts staying disabled. Do not
disable card-detect interrupts on DATA IRQs.

Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
---
On Thu, 17 Dec 2009, Andrew Morton wrote:

> On Thu, 17 Dec 2009 21:25:52 +0100 (CET)
> Guennadi Liakhovetski <g.liakhovetski@gmx.de> wrote:
> 
> > I originally submitted this [1] patch on the 6th of November, then 
> > accidentally the MMC list got dropped from CC and the discussion moved to 
> > another irrelevant list, Ian's last reply is at [2]. Unfortunately, Ian 
> > hasn't replied to my reminders since then, and I don't see this patch 
> > anywhere upstream. Is there any way we could get it into 2.6.33?
> 
> Please redo and retest it against curent mainline and do a full resend.

Re-testing this patch has again become possible now, after Magnus has 
posted his INTC and the "tmio_mmc: use 100ms mmc_detect_change() delay" 
patches. Successfully tested on Migo-R (sh7722) and MS7724SE (sh7724). 
Andrew, please apply.

 drivers/mmc/host/tmio_mmc.h |    6 ++----
 1 files changed, 2 insertions(+), 4 deletions(-)

diff --git a/drivers/mmc/host/tmio_mmc.h b/drivers/mmc/host/tmio_mmc.h
index 692dc23..dafecfb 100644
--- a/drivers/mmc/host/tmio_mmc.h
+++ b/drivers/mmc/host/tmio_mmc.h
@@ -55,10 +55,8 @@
 /* Define some IRQ masks */
 /* This is the mask used at reset by the chip */
 #define TMIO_MASK_ALL           0x837f031d
-#define TMIO_MASK_READOP  (TMIO_STAT_RXRDY | TMIO_STAT_DATAEND | \
-		TMIO_STAT_CARD_REMOVE | TMIO_STAT_CARD_INSERT)
-#define TMIO_MASK_WRITEOP (TMIO_STAT_TXRQ | TMIO_STAT_DATAEND | \
-		TMIO_STAT_CARD_REMOVE | TMIO_STAT_CARD_INSERT)
+#define TMIO_MASK_READOP  (TMIO_STAT_RXRDY | TMIO_STAT_DATAEND)
+#define TMIO_MASK_WRITEOP (TMIO_STAT_TXRQ | TMIO_STAT_DATAEND)
 #define TMIO_MASK_CMD     (TMIO_STAT_CMDRESPEND | TMIO_STAT_CMDTIMEOUT | \
 		TMIO_STAT_CARD_REMOVE | TMIO_STAT_CARD_INSERT)
 #define TMIO_MASK_IRQ     (TMIO_MASK_READOP | TMIO_MASK_WRITEOP | TMIO_MASK_CMD)
-- 
1.6.2.4

Re: [PATCH] tmio_mmc: keep card-detect interrupts enabled

[Andrew Morton]

On Tue, 16 Feb 2010 10:34:55 +0100 (CET)
Guennadi Liakhovetski <g.liakhovetski@gmx.de> wrote:

> On SuperH platforms the SDHI controller does not produce any command IRQs after
> a completed IO. This leads to card-detect interrupts staying disabled. Do not
> disable card-detect interrupts on DATA IRQs.
> 

Why is the current code disabling card-detect interrupts on DATA IRQs? 
Was that just a bug?

What effect will this change have for current users of this driver?


> ---
> On Thu, 17 Dec 2009, Andrew Morton wrote:
> 
> > On Thu, 17 Dec 2009 21:25:52 +0100 (CET)
> > Guennadi Liakhovetski <g.liakhovetski@gmx.de> wrote:
> > 
> > > I originally submitted this [1] patch on the 6th of November, then 
> > > accidentally the MMC list got dropped from CC and the discussion moved to 
> > > another irrelevant list, Ian's last reply is at [2]. Unfortunately, Ian 
> > > hasn't replied to my reminders since then, and I don't see this patch 
> > > anywhere upstream. Is there any way we could get it into 2.6.33?
> > 
> > Please redo and retest it against curent mainline and do a full resend.
> 
> Re-testing this patch has again become possible now, after Magnus has 
> posted his INTC and the "tmio_mmc: use 100ms mmc_detect_change() delay" 
> patches. Successfully tested on Migo-R (sh7722) and MS7724SE (sh7724). 
> Andrew, please apply.
> 
>  drivers/mmc/host/tmio_mmc.h |    6 ++----
>  1 files changed, 2 insertions(+), 4 deletions(-)
> 
> diff --git a/drivers/mmc/host/tmio_mmc.h b/drivers/mmc/host/tmio_mmc.h
> index 692dc23..dafecfb 100644
> --- a/drivers/mmc/host/tmio_mmc.h
> +++ b/drivers/mmc/host/tmio_mmc.h
> @@ -55,10 +55,8 @@
>  /* Define some IRQ masks */
>  /* This is the mask used at reset by the chip */
>  #define TMIO_MASK_ALL           0x837f031d
> -#define TMIO_MASK_READOP  (TMIO_STAT_RXRDY | TMIO_STAT_DATAEND | \
> -		TMIO_STAT_CARD_REMOVE | TMIO_STAT_CARD_INSERT)
> -#define TMIO_MASK_WRITEOP (TMIO_STAT_TXRQ | TMIO_STAT_DATAEND | \
> -		TMIO_STAT_CARD_REMOVE | TMIO_STAT_CARD_INSERT)
> +#define TMIO_MASK_READOP  (TMIO_STAT_RXRDY | TMIO_STAT_DATAEND)
> +#define TMIO_MASK_WRITEOP (TMIO_STAT_TXRQ | TMIO_STAT_DATAEND)
>  #define TMIO_MASK_CMD     (TMIO_STAT_CMDRESPEND | TMIO_STAT_CMDTIMEOUT | \
>  		TMIO_STAT_CARD_REMOVE | TMIO_STAT_CARD_INSERT)
>  #define TMIO_MASK_IRQ     (TMIO_MASK_READOP | TMIO_MASK_WRITEOP | TMIO_MASK_CMD)
> -- 
> 1.6.2.4

Re: [PATCH] tmio_mmc: keep card-detect interrupts enabled

[Guennadi Liakhovetski]

On Tue, 16 Feb 2010, Andrew Morton wrote:

> On Tue, 16 Feb 2010 10:34:55 +0100 (CET)
> Guennadi Liakhovetski <g.liakhovetski@gmx.de> wrote:
> 
> > On SuperH platforms the SDHI controller does not produce any command IRQs after
> > a completed IO. This leads to card-detect interrupts staying disabled. Do not
> > disable card-detect interrupts on DATA IRQs.
> > 
> 
> Why is the current code disabling card-detect interrupts on DATA IRQs? 
> Was that just a bug?
> 
> What effect will this change have for current users of this driver?

Unfortunately, I cannot answer these questions, best I can do is refer you 
to this reply from Ian:

http://article.gmane.org/gmane.linux.drivers.video-input-infrastructure/11995

Thanks
Guennadi
---
Guennadi Liakhovetski, Ph.D.
Freelance Open-Source Software Developer
http://www.open-technology.de/