3.16.75-rc1 review patch. If anyone has any objections, please let me know.
------------------
From: Joakim Zhang qiangqing.zhang@nxp.com
commit 247e5356a709eb49a0d95ff2a7f07dac05c8252c upstream.
Current we can meet timeout issue when setting a small bitrate like 10000 as follows on i.MX6UL EVK board (ipg clock = 66MHZ, per clock = 30MHZ):
| root@imx6ul7d:~# ip link set can0 up type can bitrate 10000
A link change request failed with some changes committed already. Interface can0 may have been left with an inconsistent configuration, please check.
| RTNETLINK answers: Connection timed out
It is caused by calling of flexcan_chip_unfreeze() timeout.
Originally the code is using usleep_range(10, 20) for unfreeze operation, but the patch (8badd65 can: flexcan: avoid calling usleep_range from interrupt context) changed it into udelay(10) which is only a half delay of before, there're also some other delay changes.
After double to FLEXCAN_TIMEOUT_US to 100 can fix the issue.
Meanwhile, Rasmus Villemoes reported that even with a timeout of 100, flexcan_probe() fails on the MPC8309, which requires a value of at least 140 to work reliably. 250 works for everyone.
Signed-off-by: Joakim Zhang qiangqing.zhang@nxp.com Reviewed-by: Dong Aisheng aisheng.dong@nxp.com Signed-off-by: Marc Kleine-Budde mkl@pengutronix.de [bwh: Backported to 3.16: adjust context] Signed-off-by: Ben Hutchings ben@decadent.org.uk --- drivers/net/can/flexcan.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
--- a/drivers/net/can/flexcan.c +++ b/drivers/net/can/flexcan.c @@ -157,7 +157,7 @@
#define FLEXCAN_MB_CODE_MASK (0xf0ffffff)
-#define FLEXCAN_TIMEOUT_US (50) +#define FLEXCAN_TIMEOUT_US (250)
/* * FLEXCAN hardware feature flags