Re: [PATCH iwl-next v2 3/3] igc: add support for forcing link speed without autonegotiation

["Abdul Rahim, Faizal" <faizal.abdul.rahim@linux.intel.com>]

On 19/4/2026 12:48 am, Simon Horman wrote:
> From: 'Simon Horman' <horms@kernel.org>
> 
> This is an AI-generated review of your patch. The human sending this
> email has considered the AI review valid, or at least plausible.
> Full review at: https://sashiko.dev
> ---
> igc: add support for forcing link speed without autonegotiation
> 
> This commit adds support to allow users to force 10/100 Mb/s link speed
> and duplex via ethtool when autonegotiation is disabled.
> 
>> diff --git a/drivers/net/ethernet/intel/igc/igc_ethtool.c b/drivers/net/ethernet/intel/igc/igc_ethtool.c
>> index cfcbf2fdad6ea..5bd37d1be1688 100644
>> --- a/drivers/net/ethernet/intel/igc/igc_ethtool.c
>> +++ b/drivers/net/ethernet/intel/igc/igc_ethtool.c
> 
> [ ... ]
> 
>> @@ -2000,6 +2013,41 @@ static int igc_ethtool_get_link_ksettings(struct net_device *netdev,
>>   	return 0;
>>   }
>>   
>> +/**
>> + * igc_handle_autoneg_disabled - Configure forced speed/duplex settings
>> + * @adapter: private driver structure
>> + * @speed: requested speed (must be SPEED_10 or SPEED_100)
>> + * @duplex: requested duplex
>> + *
>> + * Records forced speed/duplex when autoneg is disabled.
>> + * Caller must validate speed before calling this function.
>> + */
>> +static void igc_handle_autoneg_disabled(struct igc_adapter *adapter, u32 speed,
>> +					u8 duplex)
>> +{
>> +	struct igc_mac_info *mac = &adapter->hw.mac;
>> +
>> +	switch (speed) {
>> +	case SPEED_10:
>> +		mac->forced_speed_duplex = (duplex == DUPLEX_FULL) ?
>> +			IGC_FORCED_10F : IGC_FORCED_10H;
>> +		break;
>> +	case SPEED_100:
>> +		mac->forced_speed_duplex = (duplex == DUPLEX_FULL) ?
>> +			IGC_FORCED_100F : IGC_FORCED_100H;
>> +		break;
>> +	default:
>> +		WARN_ONCE(1, "Unsupported speed %u\n", speed);
>> +		return;
>> +	}
>> +
>> +	mac->autoneg_enabled = false;
>> +
>> +	/* Half-duplex cannot support flow control per IEEE 802.3 */
>> +	if (duplex == DUPLEX_HALF)
>> +		adapter->hw.fc.requested_mode = igc_fc_none;
>> +}
> 
> If userspace issues an ethtool command without specifying duplex, the duplex
> value can be DUPLEX_UNKNOWN. In this case, forced_speed_duplex correctly
> defaults to half-duplex since duplex == DUPLEX_FULL evaluates to false.
> 
> However, since DUPLEX_UNKNOWN is not equal to DUPLEX_HALF, the check above
> fails, and the requested flow control mode remains unchanged. Could this leave
> flow control enabled on a forced half-duplex link, violating IEEE 802.3?
> Should this check be based on the resolved mac->forced_speed_duplex state
> instead?
>

You're right, thanks for pointing that out.

That said, it feels simpler to address it with [1]:
if (duplex != DUPLEX_FULL)
     adapter->hw.fc.requested_mode = igc_fc_none;

Rather than [2]:
  if (mac->forced_speed_duplex == IGC_FORCED_10H ||
         mac->forced_speed_duplex == IGC_FORCED_100H)
         adapter->hw.fc.requested_mode = igc_fc_none;

Are you okay with [1] ?

> [ ... ]
> 
>> @@ -2071,14 +2120,20 @@ igc_ethtool_set_link_ksettings(struct net_device *netdev,
>>   		}
>>   	}
>>   
>> +	if (cmd->base.autoneg == AUTONEG_DISABLE &&
>> +	    cmd->base.speed != SPEED_10 && cmd->base.speed != SPEED_100) {
>> +		netdev_info(dev, "Unsupported speed for forced link\n");
>> +		return -EINVAL;
>> +	}
>> +
>>   	while (test_and_set_bit(__IGC_RESETTING, &adapter->state))
>>   		usleep_range(1000, 2000);
> 
> This isn't a bug introduced by this patch, but does this ad-hoc
> synchronization using bit flags circumvent proper locking guarantees? The
> networking subsystem guidelines forbid using bit flags as ad-hoc locks.
> 
> Using a bit flag in a polling loop bypasses lockdep analysis, fairness, and
> strict memory ordering. Could this be replaced with a real synchronization
> mechanism like a mutex?

It looks like a worthwhile cleanup. However, it likely doesn’t belong in 
this series, since the synchronization pattern predates these patches and 
is used throughout the igc driver (set_ringparam, set_pauseparam, 
set_channels, etc.). We could address it in different patch series and 
align the other code paths at the same time ?