2018-06-20 02:59:09

by Jia-Ju Bai

[permalink] [raw]
Subject: [PATCH] staging: rtlwifi: Fix a possible sleep-in-atomic-context bug in _is_fw_read_cmd_down()

The driver may sleep with holding a spinlock.
The function call path (from bottom to top) in Linux-4.16.7 is:

[FUNC] schedule
drivers/staging/rtlwifi/halmac/rtl_halmac.c, 884:
schedule in _is_fw_read_cmd_down
drivers/staging/rtlwifi/halmac/rtl_halmac.c, 912:
_is_fw_read_cmd_down in rtl_halmac_send_h2c
drivers/staging/rtlwifi/halmac/rtl_halmac.c, 907:
_raw_spin_lock_irqsave in rtl_halmac_send_h2c

To fix this bug, schedule() is replaced with mdelay(1).

This bug is found by my static analysis tool (DSAC-2) and checked by
my code review.

Signed-off-by: Jia-Ju Bai <[email protected]>
---
drivers/staging/rtlwifi/halmac/rtl_halmac.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/staging/rtlwifi/halmac/rtl_halmac.c b/drivers/staging/rtlwifi/halmac/rtl_halmac.c
index 66f0a6dfc52c..e921a62e72a3 100644
--- a/drivers/staging/rtlwifi/halmac/rtl_halmac.c
+++ b/drivers/staging/rtlwifi/halmac/rtl_halmac.c
@@ -881,7 +881,7 @@ static bool _is_fw_read_cmd_down(struct rtl_priv *rtlpriv, u8 msgbox_num)
if (valid == 0)
read_down = true;
else
- schedule();
+ mdelay(1);
} while ((!read_down) && (retry_cnts--));

return read_down;
--
2.17.0