On Tue, Sep 06, 2022 at 08:43:43AM +0200, Miroslav Lichvar wrote: > On Mon, Sep 05, 2022 at 06:06:23PM +0000, Zdenek Chaloupka via Linuxptp-devel > wrote: > > It seems that the reason for that is that this particular GMs do not reply > > to Delay_Req messages with PTP minor version set to 1. Below find a > > fragment of the Delay_Req message from wireshark dump. As soon as I change > > the PTP minor version of the Delay_Req to 0, the Delay_Resp messages are > > transmitted by GM. > > > > I am not an expert on the 1588-2019 protocol, so not sure what should be > > the expected behavior of the GM. > > > > Any ideas what should be the expected behavior? > > To me that looks like a bug in the GM PTP implementation. > > In 1588-2008 that field is reserved, requiring implementations to set > it to 0 on transmission and ignore on reception. > > In 1588-2019 it is set to 1 on transmission and ignored on reception.
Yes, the 1588 standard is crystal clear on this point. This now the second hardware implementation that is buggy WRT the version field. Going forward, my plan is not to work around such bugs. The release notes for the up coming 4.0 linuxptp version will include a prominent notice about the minor version, and about possible consequences. Users of broken hardware will have the option to stick with version 3.1.1 I can maintain that version with bug fixes indefinitely. Of course, users will have a second option of reverting the one commit for their deployments. Thanks, Richard _______________________________________________ Linuxptp-devel mailing list Linuxptp-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/linuxptp-devel
