Re: [PATCH net 2/5] e1000e: do not cache the hwtstamp config before it is applied

[Jacob Keller <jacob.e.keller@intel.com>]

On 11/2/2026 5:04 PM, Linkui Xiao wrote:
> Hi Jake,
> 
> Thank you for applying the patch and for the review.
> 
> On 2026/11/3 06:15, Jacob Keller wrote:
>> On 10/28/2026 2:04 AM, Linkui Xiao wrote:
>>> From: Linkui Xiao <xiaolinkui@kylinos.cn>
>>>
>>> e1000e_config_hwtstamp() stores the requested configuration in
>>> adapter->hwtstamp_config right after the rx filter has been
>>> translated, but the two register writes that follow are verified by
>>> reading the registers back and each of them can fail with -EAGAIN.
>>>
>>> When one of them does, the driver is left holding a configuration the
>>> hardware never accepted:
>>>
>>>   - e1000e_hwtstamp_get() hands it back to user space, so an
>>>     application that reads the configuration back is told that
>>>     hardware time stamping is in effect while it is not;
>>>
>>>   - e1000e_systim_reset() re-applies it after every reset, including
>>>     the ones the watchdog triggers, and keeps failing;
>>>
>>>   - e1000_watchdog() starts polling TSYNCRXCTL_VALID for the cached rx
>>>     filter, and e1000_xmit_frame() still parks skbs that ask for a Tx
>>>     time stamp because it only looks at FLAG_HAS_HW_TIMESTAMP, so they
>>>     are eventually dropped by e1000e_tx_hwtstamp_work() with a
>>>     "clearing Tx timestamp hang" warning.
>>>
>>> Cache the configuration only once both registers have been verified,
>>> so that adapter->hwtstamp_config always describes what the hardware
>>> is programmed with.
>>>
>>
>> Any idea why this got marked as patch 2 out of 5? This doesn't appear on
>> either netdev or iwl with any other patches.
> Regarding your question about the "2/5" marking: it was an oversight on
> my end. I generated this fix from a local branch that contained other
> commits, and `git format-patch` automatically added the series
> numbering. The patch is indeed a standalone fix and was not intended to
> be part of a series posted elsewhere.
> 
> I'll make sure to use a proper invocation for single patches going forward.
> 
> Thanks again for your time and for taking this fix into the Intel Wired
> LAN tree.
> 
> Best regards,
> Linkui

Thanks for clarifying! I mostly wanted to make sure it wasn't some issue
with mails not hitting the list (spam or other issues), and potential
fixes being missed.

The numbering will get fixed and merged with other patches once our
review process completes and Tony finds space for submission to netdev.
That may take a bit.

Thanks,
Jake