On 16 December 2012 06:36, NeilBrown <neilb@xxxxxxx> wrote:
>
> hi,
> the commit below causes my wifi to disappear on suspend and never
> reappear.
> Unfortunately the commentary says what the patch does but not why, so I
> cannot tell if simply reverting it is correct, or if something more subtle
> in needed.
Some more background to the patch; If the card slot is configured to
be a non-removable card and it appears that the slot does not contain
a card at boot; "mmc_rescan" should _not_ check at every
suspend/resume cycle if a card has been inserted. It will mean
unnecessary power up of the card slot and thus wast power for every
suspend/resume cycle. Moreover it should not be needed since the card
is not removable.
>
> My wifi chip is a "libertas" driven "Marvell" chip. It is permanently
> attached to port 1 on an OMAP3 'hsmmc' interface. So MMC_CAP_NONREMOVABLE
> seem appropriate and is set.
>
> It is normally inactive and when the suspend code notices this it doesn't
> bother trying to suspend through the driver but just removes the core:
> /*
> * We simply "remove" the card in this case.
> * It will be redetected on resume. (Calling
> * bus_ops->remove() with a claimed host can
> * deadlock.)
> */
> (in mmc_suspend_host). Only it isn't redetected on resume because
> rescan_entered is set.
Correct, but this is how it is supposed to work.
An sdio functional driver (libertas) that handles non-removable sdio
devices, shall implement suspend|resume callbacks for the function
driver. Otherwise mmc_sdio_suspend function (drivers/mmc/core/sdio.c)
will return "-ENOSYS" and trigger the card and the card device to be
removed. That is not what you expect for non-removable device to
happen.
So, I suggest to you implement the suspend|resume callbacks in your
sdio func driver, that would do the trick I believe.
>
> I guess we could clear rescan_entered at this point. Would that make sense?
> It seems to work.
>
> diff --git a/drivers/mmc/core/core.c b/drivers/mmc/core/core.c
> index 06c42cf..65b8935 100644
> --- a/drivers/mmc/core/core.c
> +++ b/drivers/mmc/core/core.c
> @@ -2433,6 +2433,7 @@ int mmc_suspend_host(struct mmc_host *host)
> mmc_claim_host(host);
> mmc_detach_bus(host);
> mmc_power_off(host);
> + host->rescan_entered = 0;
> mmc_release_host(host);
> host->pm_flags = 0;
> err = 0;
>
>
> Thanks,
> NeilBrown
>
>
>
> From 3339d1e33185798a45dbdb5ea6c0bec1c27ca5fd Mon Sep 17 00:00:00 2001
> From: Johan Rudholm <johan.rudholm@xxxxxxxxxxxxxx>
> Date: Thu, 23 Aug 2012 13:40:55 +0200
> Subject: [PATCH] mmc: core: Do not rescan non-removable devices
>
> If MMC_CAP_NONREMOVABLE is set, only issue a detect job on init.
>
> Signed-off-by: Johan Rudholm <johan.rudholm@xxxxxxxxxxxxxx>
> Acked-by: Ulf Hansson <ulf.hansson@xxxxxxxxxx>
> Acked-by: Linus Walleij <linus.walleij@xxxxxxxxxx>
> Signed-off-by: Chris Ball <cjb@xxxxxxxxxx>
>
> diff --git a/drivers/mmc/core/core.c b/drivers/mmc/core/core.c
> index 835c9f0..af2c4d2 100644
> --- a/drivers/mmc/core/core.c
> +++ b/drivers/mmc/core/core.c
> @@ -2053,6 +2053,11 @@ void mmc_rescan(struct work_struct *work)
> if (host->rescan_disable)
> return;
>
> + /* If there is a non-removable card registered, only scan once */
> + if ((host->caps & MMC_CAP_NONREMOVABLE) && host->rescan_entered)
> + return;
> + host->rescan_entered = 1;
> +
> mmc_bus_get(host);
>
> /*
> diff --git a/include/linux/mmc/host.h b/include/linux/mmc/host.h
> index f578a71..d5d9bd4 100644
> --- a/include/linux/mmc/host.h
> +++ b/include/linux/mmc/host.h
> @@ -300,6 +300,7 @@ struct mmc_host {
> #endif
>
> int rescan_disable; /* disable card detection */
> + int rescan_entered; /* used with nonremovable devices */
>
> struct mmc_card *card; /* device attached to this host */
>
Kind regards
Ulf Hansson
--
To unsubscribe from this list: send the line "unsubscribe linux-mmc" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
