Re: [Linuxptp-users] phc offset explodes after a while

[Vick, M. <mat...@in...>]

On 1/30/14, 12:53 AM, "Koehrer Mathias (ETAS/ESW5)"
<mat...@et...> wrote:

>Hi all,
>
>I am very new in using linuxptp.
>I want to synchronize two PC using a separate point-to-point Ethernet
>connection between both PCs.
>This Ethernet connection is used only for PTP, any other network traffic
>is routed via another NIC.
>
>Kernel 3.2.48-rt69 (RT_PREEMPT patch).
>I use the Intel Gigabit CT Desktop adapter (82574L).
>For this is use the latest e1000e driver (Version 2.5.4) and build it
>outside the kernel to have the PTP support.
>One of the PCs (PCA) runs using 64 bit kernel, the other PC (PCB) runs
>using a 32 bit kernel.
>linux-ptp version 1.3.
>
>On PCA I start the following commands:
># ./ptp4l -i eth1 -p /dev/ptp1 -m -H -P
>and after a couple of seconds:
># ./phc2sys -s /dev/ptp1 -w -m
>
>On PCB I do the very same (unless I have to use /dev/ptp0 here).
>
>This works really fine for a while. However after a while the value of
>phc explodes dramatically.
>The value will be -35192325800 which looks like a kind of value overrun!
>In hex this is 0xFFFFFFF7CE5FB958.
>Please see the logfiles below.
>What could be the issue for that?!?
>
>Thanks for any feedback on this!
>
>Regards
>
>Mathias

Mathias,

Thanks for your report on this! We've had a few reports of this now and
are investigating it internally. We currently believe we may have a
hardware bug on 82574L where some reads of the device clock are corrupted.
We are testing a workaround now and will be submitting the patch upstream
once it has passed our internal validation.

Cheers,
Matthew

Matthew Vick
Linux Development
Networking Division
Intel Corporation