[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Bug#361199: sparc64 tftpboot kernel driver failure during Gigaswift ethernet (aka cassini) detection (#2)



Package: kernel
Severity: critical

It is not possible to bring Gigaswift (aka Cassini) ethernet cards online and therefore continue installation on SunFire V440 (UltraSparc IIIi) when I use boot.img placed in the following directory of debian archives:

/debian/dists/sid/main/installer-sparc/current/images/sparc64/netboot/ 2.6/

The installation fails during network detection process with the following dump:

Apr 4 21:40:00 main-menu[2937]: DEBUG: configure ethernet-card- detection, status: 0 Apr 4 21:40:00 main-menu[2937]: DEBUG: virtual package ethernet-card- detection
Apr  4 21:40:00 kernel: Unable to handle kernel NULL pointer dereference
Apr  4 21:40:00 kernel: tsk->{mm,active_mm}->context = 00000000000001b6
Apr  4 21:40:00 kernel: tsk->{mm,active_mm}->pgd = fffff8133f65e000
Apr  4 21:40:00 kernel:               \|/ ____ \|/
Apr  4 21:40:00 kernel:               "@'/ .. \`@"
Apr  4 21:40:00 kernel:               /_| \__/ |_\
Apr  4 21:40:00 kernel:                  \__U_/
Apr  4 21:40:00 kernel: netcfg(7981): Oops [#1]
Apr 4 21:40:00 kernel: TSTATE: 0000009980009600 TPC: 0000000000412eb4 TNPC: 0000000000412eb8 Y: 00000000 Not tainted
Apr  4 21:40:00 kernel: TPC: <request_irq+0x14/0x2b8>
Apr 4 21:40:00 kernel: g0: fffff8133ecfc000 g1: fffff800006dd000 g2: 00000000000066d8 g3: fffff8133e2c13e0 Apr 4 21:40:00 kernel: g4: fffff8133d30b9c0 g5: 0000000000000000 g6: fffff8133e808000 g7: 00000000000066b8

The cards are recognized by the driver. Driver has explicitly mentioned that MAC address are not set in some kind of firmware memory (?)

Mentioned boot.img uses kernel 2.6.15-1, this bug was probably solved in 2.6.16, so this is something to be checked and boot.img file is to be updated.

UPDATE: 2.6.16 based installer ( from http://people.debian.org/ ~stappers/d-i/images/daily/sparc64/netboot/2.6/ ) hangs during detection as well.

I'm ready to test this problem again upon your request.

Karel Panek






Reply to: