Try disabling iAMT in bios. If that helps it may be possible to workaround if the right places can be found, see the patch in comment 7 on https://bugzilla.kernel.org/show_bug.cgi?id=10990
On 2022/08/22 13:17, Walter Alejandro Iglesias wrote: > Hi, > > This bug: > > https://marc.info/?l=openbsd-bugs&m=156682947025229 > > is still present in the latest snapshot. > > Perhaps that bug report was too noisy. Let's explain the issue again. > The machine is a HP dc7700, its ethernet card is a Intel 82566DM > Gigabit, listed in em(4) as supported. With FreeBSD em driver happens > exactly the same. > > The device is auto detected as 100baseTX, ifconfig lets me change it > manually to 1000baseT, but right after the connection is lost (No carrier). > This time I performed the below test to find out if ifconfig was just > showing it wrong. > > In both tests I used netperf(1) (in ports) from the HP desktop to a > Laptop connected to the same router with the same ethernet cables. > 192.168.1.103 is a Thinkpad T410 running OpenBSD (there OpenBSD properly > recognizes the card as 1000baseT, by the way). > > Running Linux in the HP desktop: > > $ netperf -H 192.168.1.103 > MIGRATED TCP STREAM TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to > 192.168.1.103 () port 0 AF_INET : demo > Recv Send Send > Socket Socket Message Elapsed > Size Size Size Time Throughput > bytes bytes bytes secs. 10^6bits/sec > > 16384 16384 16384 10.02 905.51 > > > Running OpenBSD in the HP desktop: > > $ netperf -H 192.168.1.103 > MIGRATED TCP STREAM TEST from (null) (0.0.0.0) port 0 AF_INET to (null) () > port 0 AF_INET : demo > Recv Send Send > Socket Socket Message Elapsed > Size Size Size Time Throughput > bytes bytes bytes secs. 10^6bits/sec > > 16384 16384 16384 10.03 93.68 > > > Same router, same machines, same cables. If this test is telling the > true the difference in speed is real. > > > Walter > >
