amazing. thank you so much. :-)
> On 14/12/2020 23:56 Jacob Keller <jacob.e.kel...@intel.com> wrote:
>
>
> On 12/14/2020 2:43 PM, Nils Hasler wrote:
> > Thanks Jacob.
> >
> >
> > $ ethtool -i eno1
> > driver: e1000e
> > version: 3.2.6-k
> > firmware-version: 0.4-4
> > expansion-rom-version:
> > bus-info: 0000:00:1f.6
> > supports-statistics: yes
> > supports-test: yes
> > supports-eeprom-access: yes
> > supports-register-dump: yes
> > supports-priv-flags: no
> >
>
> Yep, this is indeed the driver that Richard mentioned earlier. This
> driver has some workarounds for known bugs in the timestamp logic that
> can occur with how the hardware latches the SYSTIME registers.
>
> I checked the Linux git history, and the sanitize function that checks
> this does appear to be in the 4.18 tree, and has been around for some time.
>
> However, I noticed that a newer fix from Miroslav, commit e1f65b0d70e9
> ("e1000e: allow non-monotonic SYSTIM readings") landed in Linux 5.0
>
> Per the description:
>
>
> >
> > It seems with some NICs supported by the e1000e driver a SYSTIM reading
> > may occasionally be few microseconds before the previous reading and if
> > enabled also pass e1000e_sanitize_systim() without reaching the maximum
> > number of rereads, even if the function is modified to check three
> > consecutive readings (i.e. it doesn't look like a double read error).
> > This causes an underflow in the timecounter and the PHC time jumps hours
> > ahead.
> >
> > This was observed on 82574, I217 and I219. The fastest way to reproduce
> > it is to run a program that continuously calls the PTP_SYS_OFFSET ioctl
> > on the PHC.
> >
> > Modify e1000e_phc_gettime() to use timecounter_cyc2time() instead of
> > timecounter_read() in order to allow non-monotonic SYSTIM readings and
> > prevent the PHC from jumping.
> >
>
> Without this fix, the time counter can get screwed up just like you
> observed.
>
> I would suggest seeing if a kernel update to include the newer driver
> would fix this issue for you. You need something at least from 5.0 or
> newer, or a distribution which back ported the relevant fix. I don't
> believe Ubuntu has done that.
>
> Hope this helps!
>
> Thanks,
> Jake
--
Dr. Nils Hasler
+49 174 175 66 42
www.TheCaptury.com
The Captury GmbH, Saarbruecken, Germany
Amtsgericht Saarbrücken, HRB 101327
Executive Director: Nils Hasler
_______________________________________________
Linuxptp-users mailing list
Linuxptp-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linuxptp-users
