On Saturday 23 April 2011, John Stultz wrote:
> @@ -482,7 +485,26 @@ int mmc_sd_setup_card(struct mmc_host *host, struct mmc_card *card,
> /*
> * Fetch switch information from card.
> */
> +#ifdef CONFIG_MMC_PARANOID_SD_INIT
> + for (retries = 1; retries <= 3; retries++) {
> + err = mmc_read_switch(card);
> + if (!err) {
> + if (retries > 1) {
> + printk(KERN_WARNING
> + "%s: recovered\n",
> + mmc_hostname(host));
> + }
> + break;
> + } else {
> + printk(KERN_WARNING
> + "%s: read switch failed (attempt %d)\n",
> + mmc_hostname(host), retries);
> + }
> + }
> +#else
> err = mmc_read_switch(card);
> +#endif
> +
I see no reason to have this as a compile time option. Having retries in here
might be useful, but a better place would be inside of mmc_read_switch()
or mmc_sd_switch() so it is only done if the error comes from the controller,
not for cases where we know it will fail (e.g. !(card->csd.cmdclass & CCC_SWITCH)).
I would also recommend being a bit less noisy, e.g. print a warning only
after either giving up or succeeding a retry.
Arnd
--
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
