Control: tag -1 - moreinfo On Wed, 2018-07-18 at 22:31 +0200, Thomas Leuxner wrote: > * Ben Hutchings <b...@decadent.org.uk> 2018.07.17 22:52: > > > > Jul 14 12:46:18 edi kernel: [ 29.601249] sit: IPv6, IPv4 and MPLS over > > > IPv4 tunneling driver > > > Jul 14 12:46:24 edi kernel: [ 35.920488] sit: non-ECT from 238.163.0.0 > > > with TOS=0x2 > > > Jul 14 12:46:24 edi kernel: [ 35.956189] sit: non-ECT from 238.163.0.0 > > > with TOS=0x2 > > > Jul 14 12:46:24 edi kernel: [ 35.980514] sit: non-ECT from 238.163.0.0 > > > with TOS=0x2 > > > Jul 14 12:46:25 edi kernel: [ 36.207554] sit: non-ECT from 0.12.0.0 > > > with TOS=0xb > > > Jul 14 12:46:25 edi kernel: [ 36.230568] sit: non-ECT from 0.12.0.0 > > > with TOS=0x6 > > > Jul 14 12:46:25 edi kernel: [ 36.413928] sit: non-ECT from 0.12.0.0 > > > with TOS=0x6 > > > Jul 14 12:46:25 edi kernel: [ 36.437173] sit: non-ECT from 0.12.0.0 > > > with TOS=0x6# > > > > I con't see any changes in the sit driver's receive path since the > > previous upstream version (4.9.88), and we don't have any Debian- > > specific patches for it, so I don't think this is the result of a > > regression. > > > > Is it possible that the network configuration applied after reboot is > > not the same as you had applied previously? > > Hi, > > the network configuration hasn't changed. > > > The sit module has a module parameter (log_ecn_error) that controls > > this log message. Perhaps that was set to 0/N previously? > > I could mitigate the noise by applying: > > echo N > /sys/module/sit/parameters/log_ecn_error > > So you may be right, maybe the default has changed?
No, I don't believe the default has changed. I was suggesting that perhaps that parameter had been changed on this system before the reboot. Ben. -- Ben Hutchings Logic doesn't apply to the real world. - Marvin Minsky
signature.asc
Description: This is a digitally signed message part
