mmc: core: Fix WRITE_ZEROES CQE handling

WRITE_ZEROES requests use TRIM, so mark them as needing to be issued
synchronously even when a CQE is being used.  Without this,
mmc_blk_mq_issue_rq() triggers a WARN_ON_ONCE() and fails the request
since we don't have any handling for issuing this asynchronously.

Fixes: f7b6fc327327 ("mmc: core: Support zeroout using TRIM for eMMC")
Reported-by: Jon Hunter <jonathanh@nvidia.com>
Tested-by: Jon Hunter <jonathanh@nvidia.com>
Signed-off-by: Vincent Whitchurch <vincent.whitchurch@axis.com>
Reviewed-by: Avri Altman <avri.altman@wdc.com>
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/20221020130123.4033218-1-vincent.whitchurch@axis.com
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>

diff --git a/drivers/mmc/core/queue.c b/drivers/mmc/core/queue.c
index 86be55d..b396e39 100644 (file)
--- a/drivers/mmc/core/queue.c
+++ b/drivers/mmc/core/queue.c
@@ -48,6 +48,7 @@ static enum mmc_issue_type mmc_cqe_issue_type(struct mmc_host *host,
        case REQ_OP_DRV_OUT:
        case REQ_OP_DISCARD:
        case REQ_OP_SECURE_ERASE:
+       case REQ_OP_WRITE_ZEROES:
                return MMC_ISSUE_SYNC;
        case REQ_OP_FLUSH:
                return mmc_cqe_can_dcmd(host) ? MMC_ISSUE_DCMD : MMC_ISSUE_SYNC;