Skip to contentSkip to conversation

staging: dgnc: Fix a sleep-in-atomic-context bug in cls_assert_modem_signals()

PATCH New Jia-Ju Bai 1 participant 1 message Sep 15, 2018 Staging

lore
Share & tools
lore

Jia-Ju Bai

author
#0 lore raw

PATCH 1/1 staging: dgnc: Fix a sleep-in-atomic-context bug in cls_assert_modem_signals()

The driver may sleep with holding a spinlock.
The function call path (from bottom to top) in Linux-4.17 is:

[FUNC] usleep_range
drivers/staging/dgnc/dgnc_cls.c, 391:
usleep_range in cls_assert_modem_signals
drivers/staging/dgnc/dgnc_cls.c, 449:
cls_assert_modem_signals in cls_copy_data_from_queue_to_uart
drivers/staging/dgnc/dgnc_cls.c, 406:
_raw_spin_lock_irqsave in cls_copy_data_from_queue_to_uart

To fix this bug, usleep_range() is replaced with udelay().

This bug is found by my static analysis tool DSAC.

---
 drivers/staging/dgnc/dgnc_cls.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/staging/dgnc/dgnc_cls.c +1 -1
index 7e6cbfe4e4ee..a1e98ee7f9e0 100644
@@ -388,7 +388,7 @@ static void cls_assert_modem_signals(struct channel_t *ch)
 	writeb(out, &ch->ch_cls_uart->mcr);
 
 	/* Give time for the UART to actually drop the signals */
-	usleep_range(10, 20);
+	udelay(20);
 }
 
 static void cls_copy_data_from_queue_to_uart(struct channel_t *ch)
Signature · 1 line
2.17.0