On 2025-08-15 09:55:00 [+0200], Paul Menzel wrote: > > Therefore, fetch the timestamp directly from the interrupt handler. > > > > The work queue code stays for the Intel 82576. Tested on Intel i210. > > Excuse my ignorance, I do not understand the first sentence in the last > line. Is it because the driver support different models? Why not change it > for Intel 82576 too?
The 82576 does not have an interrupt event for this as far as I remember. > Do you have a reproducer for the issue, so others can test. The issue is that the workqueue can be delayed and ptp starts complaining. If the timestamp can be retrieved directly, there is no reason for the delay. Sebastian
