[PATCH net 1/1] amd-xgbe: fix active cable determination

[PATCH net 1/1] amd-xgbe: fix active cable determination

[Thomas Kupper]

When determine the type of SFP, active cables were not handled.

Add the check for active cables as an extension to the passive cable
check.

Signed-off-by: Thomas Kupper <thomas.kupper@gmail.com>
---
  drivers/net/ethernet/amd/xgbe/xgbe-phy-v2.c | 5 +++--
  1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/drivers/net/ethernet/amd/xgbe/xgbe-phy-v2.c 
b/drivers/net/ethernet/amd/xgbe/xgbe-phy-v2.c
index 4064c3e3dd49..1ba550d5c52d 100644
--- a/drivers/net/ethernet/amd/xgbe/xgbe-phy-v2.c
+++ b/drivers/net/ethernet/amd/xgbe/xgbe-phy-v2.c
@@ -1158,8 +1158,9 @@ static void xgbe_phy_sfp_parse_eeprom(struct 
xgbe_prv_data *pdata)
      }

      /* Determine the type of SFP */
-    if (phy_data->sfp_cable == XGBE_SFP_CABLE_PASSIVE &&
-        xgbe_phy_sfp_bit_rate(sfp_eeprom, XGBE_SFP_SPEED_10000))
+    if ((phy_data->sfp_cable == XGBE_SFP_CABLE_PASSIVE ||
+         phy_data->sfp_cable == XGBE_SFP_CABLE_ACTIVE) &&
+         xgbe_phy_sfp_bit_rate(sfp_eeprom, XGBE_SFP_SPEED_10000))
          phy_data->sfp_base = XGBE_SFP_BASE_10000_CR;
      else if (sfp_base[XGBE_SFP_BASE_10GBE_CC] & XGBE_SFP_BASE_10GBE_CC_SR)
          phy_data->sfp_base = XGBE_SFP_BASE_10000_SR;

--
2.34.1

Re: [PATCH net 1/1] amd-xgbe: fix active cable determination

[Jakub Kicinski]

On Thu, 10 Nov 2022 22:03:32 +0100 Thomas Kupper wrote:
> When determine the type of SFP, active cables were not handled.
> 
> Add the check for active cables as an extension to the passive cable
> check.

Is this patch on top of net or net-next or... ? Reportedly it does not
apply to net. Could you rebase, add a Fixes tag and repost CCing Tom
and Raju?

Re: [PATCH net 1/1] amd-xgbe: fix active cable determination

[Thomas Kupper]

Am 10.11.22 um 23:35 schrieb Jakub Kicinski:
> On Thu, 10 Nov 2022 23:20:02 +0100 Thomas Kupper wrote:
>> Am 10.11.22 um 22:57 schrieb Jakub Kicinski:
>>> On Thu, 10 Nov 2022 22:03:32 +0100 Thomas Kupper wrote:
>>>> When determine the type of SFP, active cables were not handled.
>>>>
>>>> Add the check for active cables as an extension to the passive cable
>>>> check.
>>> Is this patch on top of net or net-next or... ? Reportedly it does not
>>> apply to net. Could you rebase, add a Fixes tag and repost CCing Tom
>>> and Raju?
>> I apologise, after reading through all the guidelines I forgot that it
>> was on top of the latest linux-kernel instead of net.
>>
>> Regarding the 'Fixes' tag: active cables don't works for at least since
>> kernel v5.15, to what commit would you suggest do I refer to?
> Which exact sub-version of 5.15 ? Looking at the history of the file
> commit 09c5f6bf11ac988743 seems like a candidate but you'd need to
> double check based on what you know, or just revert and see if that
> fixes your problem (to confirm that's the culprit).

Checking with git blame shows that in commit abf0a1c2b26ad from 
2016-11-10 the whole if, else if ... clause plus a lot more was 
introduced. And since then the handling of the active cables was 
missing. The check (for the passive cable) got moved up in the commit 
you mentioned.

I would then use 'Fixes: abf0a1c2b26ad ...', right? And sent pretty much 
the same mail as the first time, with Tom and Raju CCed? And Patchwork 
will realise that?


Thanks for your help and patience
Thomas

Re: [PATCH net 1/1] amd-xgbe: fix active cable determination

[Jakub Kicinski]

On Fri, 11 Nov 2022 00:00:38 +0100 Thomas Kupper wrote:
> >> I apologise, after reading through all the guidelines I forgot that it
> >> was on top of the latest linux-kernel instead of net.
> >>
> >> Regarding the 'Fixes' tag: active cables don't works for at least since
> >> kernel v5.15, to what commit would you suggest do I refer to?  
> > Which exact sub-version of 5.15 ? Looking at the history of the file
> > commit 09c5f6bf11ac988743 seems like a candidate but you'd need to
> > double check based on what you know, or just revert and see if that
> > fixes your problem (to confirm that's the culprit).  
> 
> Checking with git blame shows that in commit abf0a1c2b26ad from 
> 2016-11-10 the whole if, else if ... clause plus a lot more was 
> introduced. And since then the handling of the active cables was 
> missing. The check (for the passive cable) got moved up in the commit 
> you mentioned.
> 
> I would then use 'Fixes: abf0a1c2b26ad ...', right?

Yup, sounds like it! Make sure you use the exact format from here
https://www.kernel.org/doc/html/v4.12/process/submitting-patches.html#describe-your-changes
don't wrap the line, and don't separate the tags with empty lines.
The Fixes tags are used by automated backport machinery so we need
exact format to be followed.

> And sent pretty much the same mail as the first time, with Tom and
> Raju CCed? And Patchwork will realise that?

You can throw in v2 into the subject tag to avoid any confusion:
[PATCH new v2] and that's it, yes :)