[PATCH 4.19 23/38] mmc: core: Limit retries when analyse of SDIO tuples fails

From: Greg Kroah-Hartman
Date: Mon Feb 08 2021 - 11:16:15 EST


From: Fengnan Chang <fengnanchang@xxxxxxxxx>

commit f92e04f764b86e55e522988e6f4b6082d19a2721 upstream.

When analysing tuples fails we may loop indefinitely to retry. Let's avoid
this by using a 10s timeout and bail if not completed earlier.

Signed-off-by: Fengnan Chang <fengnanchang@xxxxxxxxx>
Cc: stable@xxxxxxxxxxxxxxx
Link: https://lore.kernel.org/r/20210123033230.36442-1-fengnanchang@xxxxxxxxx
Signed-off-by: Ulf Hansson <ulf.hansson@xxxxxxxxxx>
Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>
---
drivers/mmc/core/sdio_cis.c | 6 ++++++
1 file changed, 6 insertions(+)

--- a/drivers/mmc/core/sdio_cis.c
+++ b/drivers/mmc/core/sdio_cis.c
@@ -24,6 +24,8 @@
#include "sdio_cis.h"
#include "sdio_ops.h"

+#define SDIO_READ_CIS_TIMEOUT_MS (10 * 1000) /* 10s */
+
static int cistpl_vers_1(struct mmc_card *card, struct sdio_func *func,
const unsigned char *buf, unsigned size)
{
@@ -270,6 +272,8 @@ static int sdio_read_cis(struct mmc_card

do {
unsigned char tpl_code, tpl_link;
+ unsigned long timeout = jiffies +
+ msecs_to_jiffies(SDIO_READ_CIS_TIMEOUT_MS);

ret = mmc_io_rw_direct(card, 0, 0, ptr++, 0, &tpl_code);
if (ret)
@@ -322,6 +326,8 @@ static int sdio_read_cis(struct mmc_card
prev = &this->next;

if (ret == -ENOENT) {
+ if (time_after(jiffies, timeout))
+ break;
/* warn about unknown tuples */
pr_warn_ratelimited("%s: queuing unknown"
" CIS tuple 0x%02x (%u bytes)\n",