hostap_get_wireless_stats() is the iw_handler_if::get_wireless_stats()
callback of this driver. This callback was not allowed to sleep until
commit a160ee69c6a46 ("wext: let get_wireless_stats() sleep") in v2.6.32.
Remove the therefore pointless in_atomic() check.
Signed-off-by: Sebastian Andrzej Siewior <bige...@linutronix.de>
Cc: Jouni Malinen <j...@w1.fi>
Cc: Kalle Valo <kv...@codeaurora.org>
Cc: "David S. Miller" <da...@davemloft.net>
Cc: Jakub Kicinski <k...@kernel.org>
Cc: linux-wirel...@vger.kernel.org
Cc: net...@vger.kernel.org
---
drivers/net/wireless/intersil/hostap/hostap_ioctl.c | 13 +------------
1 file changed, 1 insertion(+), 12 deletions(-)
diff --git a/drivers/net/wireless/intersil/hostap/hostap_ioctl.c
b/drivers/net/wireless/intersil/hostap/hostap_ioctl.c
index 514c7b01dbf6f..49766b285230c 100644
--- a/drivers/net/wireless/intersil/hostap/hostap_ioctl.c
+++ b/drivers/net/wireless/intersil/hostap/hostap_ioctl.c
@@ -44,19 +44,8 @@ static struct iw_statistics
*hostap_get_wireless_stats(struct net_device *dev)
if (local->iw_mode != IW_MODE_MASTER &&
local->iw_mode != IW_MODE_REPEAT) {
- int update = 1;
-#ifdef in_atomic
- /* RID reading might sleep and it must not be called in
- * interrupt context or while atomic. However, this
- * function seems to be called while atomic (at least in Linux
- * 2.5.59). Update signal quality values only if in suitable
- * context. Otherwise, previous values read from tick timer
- * will be used. */
- if (in_atomic())
- update = 0;
-#endif /* in_atomic */
- if (update && prism2_update_comms_qual(dev) == 0)
+ if (prism2_update_comms_qual(dev) == 0)
wstats->qual.updated = IW_QUAL_ALL_UPDATED |
IW_QUAL_DBM;
--
2.28.0
