* [PATCH net] amd-xgbe: read link status twice to avoid inconsistencies
@ 2025-05-14 19:41 Raju Rangoju
2025-05-15 0:07 ` Andrew Lunn
0 siblings, 1 reply; 3+ messages in thread
From: Raju Rangoju @ 2025-05-14 19:41 UTC (permalink / raw)
To: andrew+netdev, davem, edumazet, kuba, pabeni
Cc: netdev, linux-kernel, Shyam-sundar.S-k, Raju Rangoju
The link status is latched low, so read the register twice to get the
current status and avoid link inconsistencies.
As per IEEE 802.3 "Table 22-8 - Status register bit definitions"
1.2 Link Status 1 = link is up 0 = link is down RO/LL
Fixes: 4f3b20bfbb75 ("amd-xgbe: add support for rx-adaptation")
Signed-off-by: Raju Rangoju <Raju.Rangoju@amd.com>
---
drivers/net/ethernet/amd/xgbe/xgbe-phy-v2.c | 4 ++++
1 file changed, 4 insertions(+)
diff --git a/drivers/net/ethernet/amd/xgbe/xgbe-phy-v2.c b/drivers/net/ethernet/amd/xgbe/xgbe-phy-v2.c
index 268399dfcf22..d233e3faa1a9 100644
--- a/drivers/net/ethernet/amd/xgbe/xgbe-phy-v2.c
+++ b/drivers/net/ethernet/amd/xgbe/xgbe-phy-v2.c
@@ -2914,6 +2914,10 @@ static int xgbe_phy_link_status(struct xgbe_prv_data *pdata, int *an_restart)
}
/* check again for the link and adaptation status */
+ /* Link status is latched low, so read once to clear
+ * and then read again to get current state
+ */
+ reg = XMDIO_READ(pdata, MDIO_MMD_PCS, MDIO_STAT1);
reg = XMDIO_READ(pdata, MDIO_MMD_PCS, MDIO_STAT1);
if ((reg & MDIO_STAT1_LSTATUS) && pdata->rx_adapt_done)
return 1;
--
2.34.1
^ permalink raw reply related [flat|nested] 3+ messages in thread
* Re: [PATCH net] amd-xgbe: read link status twice to avoid inconsistencies
2025-05-14 19:41 [PATCH net] amd-xgbe: read link status twice to avoid inconsistencies Raju Rangoju
@ 2025-05-15 0:07 ` Andrew Lunn
2025-06-25 9:42 ` Rangoju, Raju
0 siblings, 1 reply; 3+ messages in thread
From: Andrew Lunn @ 2025-05-15 0:07 UTC (permalink / raw)
To: Raju Rangoju, g
Cc: andrew+netdev, davem, edumazet, kuba, pabeni, netdev,
linux-kernel, Shyam-sundar.S-k
On Thu, May 15, 2025 at 01:11:45AM +0530, Raju Rangoju wrote:
> The link status is latched low, so read the register twice to get the
> current status and avoid link inconsistencies.
>
> As per IEEE 802.3 "Table 22-8 - Status register bit definitions"
> 1.2 Link Status 1 = link is up 0 = link is down RO/LL
>
> Fixes: 4f3b20bfbb75 ("amd-xgbe: add support for rx-adaptation")
> Signed-off-by: Raju Rangoju <Raju.Rangoju@amd.com>
> ---
> drivers/net/ethernet/amd/xgbe/xgbe-phy-v2.c | 4 ++++
> 1 file changed, 4 insertions(+)
>
> diff --git a/drivers/net/ethernet/amd/xgbe/xgbe-phy-v2.c b/drivers/net/ethernet/amd/xgbe/xgbe-phy-v2.c
> index 268399dfcf22..d233e3faa1a9 100644
> --- a/drivers/net/ethernet/amd/xgbe/xgbe-phy-v2.c
> +++ b/drivers/net/ethernet/amd/xgbe/xgbe-phy-v2.c
> @@ -2914,6 +2914,10 @@ static int xgbe_phy_link_status(struct xgbe_prv_data *pdata, int *an_restart)
> }
>
> /* check again for the link and adaptation status */
> + /* Link status is latched low, so read once to clear
> + * and then read again to get current state
> + */
> + reg = XMDIO_READ(pdata, MDIO_MMD_PCS, MDIO_STAT1);
Since this is not a phylib user, i don't care too much, but:
https://elixir.bootlin.com/linux/v6.14.6/source/drivers/net/phy/phy_device.c#L2514
/* The link state is latched low so that momentary link
* drops can be detected. Do not double-read the status
* in polling mode to detect such short link drops except
* the link was already down.
*/
So you don't care about short link drops? You don't want to tell user
space the link went away and came back? It never happened.
Andrew
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [PATCH net] amd-xgbe: read link status twice to avoid inconsistencies
2025-05-15 0:07 ` Andrew Lunn
@ 2025-06-25 9:42 ` Rangoju, Raju
0 siblings, 0 replies; 3+ messages in thread
From: Rangoju, Raju @ 2025-06-25 9:42 UTC (permalink / raw)
To: Andrew Lunn, g
Cc: andrew+netdev, davem, edumazet, kuba, pabeni, netdev,
linux-kernel, Shyam-sundar.S-k
On 5/15/2025 5:37 AM, Andrew Lunn wrote:
> On Thu, May 15, 2025 at 01:11:45AM +0530, Raju Rangoju wrote:
>> The link status is latched low, so read the register twice to get the
>> current status and avoid link inconsistencies.
>>
>> As per IEEE 802.3 "Table 22-8 - Status register bit definitions"
>> 1.2 Link Status 1 = link is up 0 = link is down RO/LL
>>
>> Fixes: 4f3b20bfbb75 ("amd-xgbe: add support for rx-adaptation")
>> Signed-off-by: Raju Rangoju <Raju.Rangoju@amd.com>
>> ---
>> drivers/net/ethernet/amd/xgbe/xgbe-phy-v2.c | 4 ++++
>> 1 file changed, 4 insertions(+)
>>
>> diff --git a/drivers/net/ethernet/amd/xgbe/xgbe-phy-v2.c b/drivers/net/ethernet/amd/xgbe/xgbe-phy-v2.c
>> index 268399dfcf22..d233e3faa1a9 100644
>> --- a/drivers/net/ethernet/amd/xgbe/xgbe-phy-v2.c
>> +++ b/drivers/net/ethernet/amd/xgbe/xgbe-phy-v2.c
>> @@ -2914,6 +2914,10 @@ static int xgbe_phy_link_status(struct xgbe_prv_data *pdata, int *an_restart)
>> }
>>
>> /* check again for the link and adaptation status */
>> + /* Link status is latched low, so read once to clear
>> + * and then read again to get current state
>> + */
>> + reg = XMDIO_READ(pdata, MDIO_MMD_PCS, MDIO_STAT1);
>
>
> Since this is not a phylib user, i don't care too much, but:
>
> https://elixir.bootlin.com/linux/v6.14.6/source/drivers/net/phy/phy_device.c#L2514
>
> /* The link state is latched low so that momentary link
> * drops can be detected. Do not double-read the status
> * in polling mode to detect such short link drops except
> * the link was already down.
> */
>
> So you don't care about short link drops? You don't want to tell user
> space the link went away and came back? It never happened.
Thanks Andrew for point this out. I'll re-spin v2 addressing these.
>
> Andrew
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2025-06-25 9:42 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2025-05-14 19:41 [PATCH net] amd-xgbe: read link status twice to avoid inconsistencies Raju Rangoju
2025-05-15 0:07 ` Andrew Lunn
2025-06-25 9:42 ` Rangoju, Raju
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).