[Intel-wired-lan] [RfC] fix auto-negotiation after reconnect

Neftin, Sasha sasha.neftin at intel.com
Thu Jan 17 07:43:50 UTC 2019


On 1/16/2019 19:33, Jan-Marek Glogowski wrote:
> 
> 
> Am 15.01.19 um 16:43 schrieb Neftin, Sasha:
>> You can make changes and resend patches 1 and 3. There is 'u8' type use instead of 'u16' - I not
>> sure in benefit of it. Debug prints from patch 3 are essential I thought.
> 
> Regarding the u8: after I used the the reference of &cmd->base.duplex in the
> e1000e_get_speed_and_duplex_copper call, the compiler complained about duplex being u8 in the UAPI
> struct, but u16 as the duplex reference parameter.
> 
> The same actually happened when I tried the &cmd->base.speed, where the UAPI is u32, but internally
> e1000e uses u16 in a lot of places.
> 
> I decided to keep u16 speed and change the few u8 duplex places. Really don't care, but if you have
> a preference, I'll change it before sending v2, using a temporary duplex variable.
There is no too much benefit from these patches (IMHO). I won't stop 
you. In case you consider resubmit please be consistent and replace the 
type in all relevant places.
> 
>> Regard your problem. Fujitsu can contacts our customer representative.
> 
> Already in touch with Fujitsu.
>
Good.

> Thanks for your help!
> 
> Jan-Marek
> _______________________________________________
> Intel-wired-lan mailing list
> Intel-wired-lan at osuosl.org
> https://lists.osuosl.org/mailman/listinfo/intel-wired-lan
> 
Sasha


More information about the Intel-wired-lan mailing list