Hi Pankaj Bansal,
I noticed the issue which you mentioned in "[PATCH 3/4] can: flexcan: modify the interrupt register handling", I also reported and Marc has
fixed it in linux-can-next.
And I noticed the range of a bit time may not correct:
+static const struct can_bittiming_const extended_flexcan_bittiming_const = {
+ .name = DRV_NAME,
+ .tseg1_min = 2,
+ .tseg1_max = *96*,
+ .tseg2_min = 2,
+ .tseg2_max = 32,
+ .sjw_max = 16,
+ .brp_min = 1,
+ .brp_max = 1024,
+ .brp_inc = 1,
+};
+
+static const struct can_bittiming_const flexcan_data_bittiming_const = {
+ .name = DRV_NAME,
+ .tseg1_min = 2,
+ .tseg1_max = *39*,
+ .tseg2_min = 2,
+ .tseg2_max = 8,
+ .sjw_max = 4,
+ .brp_min = 1,
+ .brp_max = 1024,
+ .brp_inc = 1,
+};
I do not know how many bits of PROPSEG and PSEG1 are LX2160A realized.
But in i.MX8QM/QXP:
CBT register: EPROPSEG(6 bit), EPSEG(5 bit) but is can't reach 96
FDCBT register: FPROPSEG(5 bit), EPSEG(3 bit) but is can't reach 39
You can check the code in /drivers/net/can/dev.c for the reason:
Can_cal_bittiming:
| 211 bt->prop_seg = tseg1 / 2;
| 212 bt->phase_seg1 = tseg1 - bt->prop_seg;
| 213 bt->phase_seg2 = tseg2;
Now I am also upstreaming the CAN FD feature for i.MX8, I will send out the V3.
The way is not the same, shall we discuss which is better? This may speed up the process.
Best Regards,
Joakim Zhang
> -----Original Message-----
> From: linux-can-owner@xxxxxxxxxxxxxxx <linux-can-owner@xxxxxxxxxxxxxxx>
> On Behalf Of Pankaj Bansal
> Sent: 2019年5月10日 13:34
> To: Wolfgang Grandegger <wg@xxxxxxxxxxxxxx>; Marc Kleine-Budde
> <mkl@xxxxxxxxxxxxxx>
> Cc: linux-can@xxxxxxxxxxxxxxx; Varun Sethi <V.Sethi@xxxxxxx>
> Subject: RE: [PATCH 0/4] net: can: flexcan: can FD Format (FDF) changes
>
> Hi Marc et. Al,
>
> These changes are pending from a long time and the regular changes to flexcan
> code make it very hard to maintain these changes. I request you to please
> review these changes and merge. I will promply reply to any queries you have.
>
> Regards,
> Pankaj Bansal
>
> > -----Original Message-----
> > From: Pankaj Bansal
> > Sent: Friday, 10 May, 2019 10:58 AM
> > To: Wolfgang Grandegger <wg@xxxxxxxxxxxxxx>; Marc Kleine-Budde
> > <mkl@xxxxxxxxxxxxxx>
> > Cc: linux-can@xxxxxxxxxxxxxxx; Pankaj Bansal <pankaj.bansal@xxxxxxx>;
> > Varun Sethi <V.Sethi@xxxxxxx>
> > Subject: [PATCH 0/4] net: can: flexcan: can FD Format (FDF) changes
> >
> > This is my second attempt to add FD support in flexcan. my earlier
> > patches can be referred here:
> >
> > [1]
> > https://eur01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.
> >
> spinics.net%2Flists%2Flinux-can%2Fmsg00348.html&data=02%7C01%7C
> qia
> >
> ngqing.zhang%40nxp.com%7C041eb587964148921bbc08d6d509245f%7C68
> 6ea1d3bc
> >
> 2b4c6fa92cd99c5c301635%7C0%7C0%7C636930632570755870&sdata=
> 0DY3hiWb
> > VgXMfZwW%2B0gbAfkSXvIFvtOyaka33E1uEbs%3D&reserved=0
> >
> > Some of the patches from above series have already been merged.
> >
> > There has been a significant change in flexcan code since i sent above
> > patches
> > namely:
> >
> > commit cbffaf7aa09e ("can: flexcan: Always use last mailbox for TX")
> >
> > with this patch, the flexcan FD supports breaks (if one were to apply
> > the above
> > patches[1])
> >
> > To fix FD supprt with above patch, i have added changes in this series:
> >
> > 3/4: can: flexcan: modify the interrupt register handling
> >
> > The rest of the patches are from old patch series.
> >
> > Cc: Varun Sethi <V.Sethi@xxxxxxx>
> >
> > Pankaj Bansal (4):
> > net: can: flexcan: use CAN FD frames for Tx/Rx
> > can: flexcan: move can set bititings after module control
> > can: flexcan: modify the interrupt register handling
> > net: can: flexcan: can FD Format (FDF) changes
> >
> > drivers/net/can/flexcan.c | 316 ++++++++++++++++++++++++++-------
> > drivers/net/can/rx-offload.c | 32 +++-
> > include/linux/can/rx-offload.h | 6 +-
> > 3 files changed, 276 insertions(+), 78 deletions(-)
> >
> > --
> > 2.17.1
