Re: Kernel halt when connecting re0 to the lan

On Monday 10 November 2003 17:10, Nils Segerdahl wrote:
> Problem:
> when connecting my laptop (Compaq evo N1020v) to the network, the kernel
> halts right after execution of re_diag() in the re-device driver.
> The loopback test fails.
> The interface works ok if I remove the test from the driver, or if I use
> another operating system. It used to work perfect with 4.8-STABLE.
>
> Any suggestions?

hmm I have the very same laptop model and it runs current from last week 
(06.11.2003 i believe) without any trouble, so probably there is something 
different in our configuration... do you use re as module or is it compiled 
in ? (I have it compiled in)
probably full dmesg and kernel config would be helpful so we can see what are 
the differences and trace the cause down better.
here are my dmesg and kernel config:
http://www.bsd.ee/~hadara/dump/dmesg_04.11.2003.txt
http://www.bsd.ee/~hadara/dump/IKALDUS_NODEBUG

You probably should cc Bill Paul too, he was asking for testers when he 
separated support for 8139C+ from rl driver but since it so rare, noone 
answered and bugs might have very well gone in unnoticed.

>
> From dmesg:
>
> re0: <RealTek 8139C+ 10/100BaseTX> port 0x9000-0x90ff mem
> 0xf0019800-0xf00198ff
> irq 11 at device 11.0 on pci0
> re0: Ethernet address: 00:08:02:d6:bf:cd
> miibus0: <MII bus> on re0
> rlphy0: <RealTek internal media interface> on miibus0
> rlphy0:  10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, auto
>

I don't see differences in this part of dmesg's