Re: [PATCH v4] mmc: Make ID freq configurable

From: Chris Ball
Date: Thu Sep 02 2010 - 21:48:32 EST

Hi Hein, thanks for resending this,

On Fri, Sep 03, 2010 at 05:34:53AM +0800, Hein_Tibosch wrote:
> In the latest releases of the mmc driver, the freq during initialization
> is set to a fixed 400 Khz. This was reportedly too fast for several
> users. As there doesn't seem to be an ideal frequency which-works-for-all,
> Pierre suggested to let the driver try several frequencies.
>
> This patch implements that idea. It will try mmc-initialization using
> several frequencies from an array 400, 300, 200 and 100.
> I submitted it earlier but it's now adapted to and tested with kernel
> 2.6.36-rc3.
>
> In case SDIO is broken, it'll still try to detect SDMEM, also at different
> freqs.
>
> Signed-off-by: Hein Tibosch <hein_tibosch@xxxxxxxx>

Reviewed-and-Tested-by: Chris Ball <cjb@xxxxxxxxxx>

I tested by modifying mmc_send_app_op_cond() to return timeouts for
various frequencies, and verified that the code correctly progresses
through the freqs[] table until reaching host->f_min, and can bring
up the card successfully on any of those retries.

In an earlier mail, you said:

> Here it tries 3 frequencies, which takes 290 ms:
> In a meanwhile it is doing other things, so not much time is wasted.

I decided to check this. If my emulated failing of bringing up a card
at a given frequency is correct, 14ms per frequency attempt is spent in
mdelay() rather than msleep(). I don't think this is a problem, though,
because you only take that hit if you have a card inserted that we're
unable to bring up at the default frequency -- the only alternative
would be to give up on the card.

If a host has no card inserted then we still try to detect a card on
all frequencies, but all of the paths in that case are msleep() rather
than mdelay(), so that's okay.

Two minor checkpatch fixes below:

> ---
> diff -Nurp a/drivers/mmc/core/core.c b/drivers/mmc/core/core.c
> --- a/drivers/mmc/core/core.c 2010-08-29 23:36:04.000000000 +0800
> +++ b/drivers/mmc/core/core.c 2010-09-03 04:28:52.000000000 +0800
> @@ -907,12 +907,7 @@ static void mmc_power_up(struct mmc_host
> */
> mmc_delay(10);
>
> - if (host->f_min > 400000) {
> - pr_warning("%s: Minimum clock frequency too high for "
> - "identification mode\n", mmc_hostname(host));
> - host->ios.clock = host->f_min;
> - } else
> - host->ios.clock = 400000;
> + host->ios.clock = host->f_init;
>
> host->ios.power_mode = MMC_POWER_ON;
> mmc_set_ios(host);
> @@ -1404,6 +1399,8 @@ void mmc_rescan(struct work_struct *work
> u32 ocr;
> int err;
> unsigned long flags;
> + int i;
> + unsigned freqs[] = { 400000, 300000, 200000, 100000 };
>
> spin_lock_irqsave(&host->lock, flags);
>
> @@ -1443,55 +1440,64 @@ void mmc_rescan(struct work_struct *work
> if (host->ops->get_cd && host->ops->get_cd(host) == 0)
> goto out;
>
> - mmc_claim_host(host);
> + for (i = 0; i < ARRAY_SIZE(freqs); i++) {
> + mmc_claim_host(host);
>
> - mmc_power_up(host);
> - sdio_reset(host);
> - mmc_go_idle(host);
> + if (freqs[i] >= host->f_min)
> + host->f_init = freqs[i];
> + else if (i && freqs[i-1] <= host->f_min)
> + goto out;
> + else
> + host->f_init = host->f_min;
>
> - mmc_send_if_cond(host, host->ocr_avail);
> + printk ("mmc_rescan: trying %u Hz\n", host->f_init);

Need a loglevel here, and an mmc_hostname. Something like:

pr_info("%s: %s: trying to init card at %u Hz\n",
mmc_hostname(host), __func__, host->f_init);

> + mmc_power_up(host);
> + sdio_reset(host);
> + mmc_go_idle(host);
>
> - /*
> - * First we search for SDIO...
> - */
> - err = mmc_send_io_op_cond(host, 0, &ocr);
> - if (!err) {
> - if (mmc_attach_sdio(host, ocr)) {
> - mmc_claim_host(host);
> - /* try SDMEM (but not MMC) even if SDIO is broken */
> - if (mmc_send_app_op_cond(host, 0, &ocr))
> - goto out_fail;
> + mmc_send_if_cond(host, host->ocr_avail);
> +
> + /*
> + * First we search for SDIO...
> + */
> + err = mmc_send_io_op_cond(host, 0, &ocr);
> + if (!err) {
> + if (mmc_attach_sdio(host, ocr)) {
> + mmc_claim_host(host);
> + /* try SDMEM (but not MMC) even if SDIO is broken */

This breaks 80-chars, so:

/*
* Try SDMEM (but not MMC) even if SDIO
* is broken.
*/

> + if (mmc_send_app_op_cond(host, 0, &ocr))
> + goto out_fail;
> +
> + if (mmc_attach_sd(host, ocr))
> + mmc_power_off(host);
> + }
> + goto out;
> + }
>
> + /*
> + * ...then normal SD...
> + */
> + err = mmc_send_app_op_cond(host, 0, &ocr);
> + if (!err) {
> if (mmc_attach_sd(host, ocr))
> mmc_power_off(host);
> + goto out;
> }
> - goto out;
> - }
>
> - /*
> - * ...then normal SD...
> - */
> - err = mmc_send_app_op_cond(host, 0, &ocr);
> - if (!err) {
> - if (mmc_attach_sd(host, ocr))
> - mmc_power_off(host);
> - goto out;
> - }
> -
> - /*
> - * ...and finally MMC.
> - */
> - err = mmc_send_op_cond(host, 0, &ocr);
> - if (!err) {
> - if (mmc_attach_mmc(host, ocr))
> - mmc_power_off(host);
> - goto out;
> - }
> + /*
> + * ...and finally MMC.
> + */
> + err = mmc_send_op_cond(host, 0, &ocr);
> + if (!err) {
> + if (mmc_attach_mmc(host, ocr))
> + mmc_power_off(host);
> + goto out;
> + }
>
> out_fail:
> - mmc_release_host(host);
> - mmc_power_off(host);
> -
> + mmc_release_host(host);
> + mmc_power_off(host);
> + }
> out:
> if (host->caps & MMC_CAP_NEEDS_POLL)
> mmc_schedule_delayed_work(&host->detect, HZ);
> diff -Nurp a/include/linux/mmc/host.h b/include/linux/mmc/host.h
> --- a/include/linux/mmc/host.h 2010-08-29 23:36:04.000000000 +0800
> +++ b/include/linux/mmc/host.h 2010-09-03 00:52:48.000000000 +0800
> @@ -123,6 +123,7 @@ struct mmc_host {
> const struct mmc_host_ops *ops;
> unsigned int f_min;
> unsigned int f_max;
> + unsigned int f_init;
> u32 ocr_avail;
> struct notifier_block pm_notify;
>
> ---

--
Chris Ball <cjb@xxxxxxxxxx> <http://printf.net/>
One Laptop Per Child
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/