The patch below does not apply to the 5.15-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to stable@vger.kernel.org.
Possible dependencies:
406e14808ee6 ("mmc: block: Remove error check of hw_reset on reset") b71597edfaad ("mmc: core: improve API to make clear mmc_hw_reset is for cards")
thanks,
greg k-h
------------------ original commit in Linus's tree ------------------
From 406e14808ee695cbae1eafa5fd3ac563c29470ab Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Christian=20L=C3=B6hle?= CLoehle@hyperstone.com Date: Thu, 13 Oct 2022 11:16:37 +0000 Subject: [PATCH] mmc: block: Remove error check of hw_reset on reset
Before switching back to the right partition in mmc_blk_reset there used to be a check if hw_reset was even supported. This return value was removed, so there is no reason to check. Furthermore ensure part_curr is not falsely set to a valid value on reset or partition switch error.
As part of this change the code paths of mmc_blk_reset calls were checked to ensure no commands are issued after a failed mmc_blk_reset directly without going through the block layer.
Fixes: fefdd3c91e0a ("mmc: core: Drop superfluous validations in mmc_hw|sw_reset()") Cc: stable@vger.kernel.org Signed-off-by: Christian Loehle cloehle@hyperstone.com Reviewed-by: Adrian Hunter adrian.hunter@intel.com Link: https://lore.kernel.org/r/e91be6199d04414a91e20611c81bfe1d@hyperstone.com Signed-off-by: Ulf Hansson ulf.hansson@linaro.org
diff --git a/drivers/mmc/core/block.c b/drivers/mmc/core/block.c index 54cd009aee50..db6d8a099910 100644 --- a/drivers/mmc/core/block.c +++ b/drivers/mmc/core/block.c @@ -134,6 +134,7 @@ struct mmc_blk_data { * track of the current selected device partition. */ unsigned int part_curr; +#define MMC_BLK_PART_INVALID UINT_MAX /* Unknown partition active */ int area_type;
/* debugfs files (only in main mmc_blk_data) */ @@ -987,33 +988,39 @@ static unsigned int mmc_blk_data_timeout_ms(struct mmc_host *host, return ms; }
+/* + * Attempts to reset the card and get back to the requested partition. + * Therefore any error here must result in cancelling the block layer + * request, it must not be reattempted without going through the mmc_blk + * partition sanity checks. + */ static int mmc_blk_reset(struct mmc_blk_data *md, struct mmc_host *host, int type) { int err; + struct mmc_blk_data *main_md = dev_get_drvdata(&host->card->dev);
if (md->reset_done & type) return -EEXIST;
md->reset_done |= type; err = mmc_hw_reset(host->card); + /* + * A successful reset will leave the card in the main partition, but + * upon failure it might not be, so set it to MMC_BLK_PART_INVALID + * in that case. + */ + main_md->part_curr = err ? MMC_BLK_PART_INVALID : main_md->part_type; + if (err) + return err; /* Ensure we switch back to the correct partition */ - if (err) { - struct mmc_blk_data *main_md = - dev_get_drvdata(&host->card->dev); - int part_err; - - main_md->part_curr = main_md->part_type; - part_err = mmc_blk_part_switch(host->card, md->part_type); - if (part_err) { - /* - * We have failed to get back into the correct - * partition, so we need to abort the whole request. - */ - return -ENODEV; - } - } - return err; + if (mmc_blk_part_switch(host->card, md->part_type)) + /* + * We have failed to get back into the correct + * partition, so we need to abort the whole request. + */ + return -ENODEV; + return 0; }
static inline void mmc_blk_reset_success(struct mmc_blk_data *md, int type) @@ -1871,8 +1878,9 @@ static void mmc_blk_mq_rw_recovery(struct mmc_queue *mq, struct request *req) return;
/* Reset before last retry */ - if (mqrq->retries + 1 == MMC_MAX_RETRIES) - mmc_blk_reset(md, card->host, type); + if (mqrq->retries + 1 == MMC_MAX_RETRIES && + mmc_blk_reset(md, card->host, type)) + return;
/* Command errors fail fast, so use all MMC_MAX_RETRIES */ if (brq->sbc.error || brq->cmd.error)