Wolfgang Grandegger schrieb:
> Hi Markus,
Hi Wolfgang.
Are we talking of the same? Think not.
Should I have changed the subject?
> Markus Plessing wrote:
>> Hi Wolfgang,
>>
>> Wolfgang Grandegger schrieb:
>>> Hi Mathias
>>>> +static void esd_usb2_tx_done_msg(struct esd_usb2_net *net, +
>>>> struct esd_usb2_msg *msg) +{ + struct net_device_stats *stats =
>>>> &net->netdev->stats; + + if
>>>> (!netif_device_present(net->netdev)) + return; + + if
>>>> (msg->msg.txdone.status == 0) { + stats->tx_packets++; +
>>>> stats->tx_bytes += msg->msg.txdone.hnd & 0xf; + } +} [...] + +
>>>> netdev->trans_start = jiffies; + + can_get_echo_skb(netdev,
>>>> context->echo_index);
>>> This functions is called when the USB transfer has completed,
>>> right? can_get_echo_skb() should be called when the TX done
>>> notification arrives (in esd_usb2_tx_done_msg). This is *wrong*
>>> in the ems_usb driver as well and should be fixed. Wolfgang.
>> The ems_usb driver increments the tx_packets and tx_bytes before
>> the call to can_get_echo_skb. Is there anything more to do? I don't
>> see a need of a fix in ems_usb or am I wrong?
>
> can_get_echo_skb() should be called in esd_usb2_tx_done_msg().
[...] out of context above [...]
>>> This is *wrong* in the ems_usb driver as well and should be fixed.
[...] out of context [...]
You mentioned that the issue rised in *esd usb2 driver* is also a issue
for the *ems_usb driver*. IMO this is not the case.
Can you please clarify this for *ems_usb*? Thanks :-)
> Wolfgang.
Have a nice day.
Markus
_______________________________________________
Socketcan-core mailing list
[email protected]
https://lists.berlios.de/mailman/listinfo/socketcan-core
