Bug#492251: crashes here also, not only on 2.6.25 but also on 2.6.26

[Santiago Garcia Mantinan]

Package: madwifi-source
Version: 1:0.9.4+r3772.20080716-1
Followup-For: Bug #492251

I've been having crashes also on my laptop with the new version of
madwifi-source, I needed to use new versions because the laptop seems to
have a Atheros AR2425 (AR5007EG) which was not supported by stable madwifi
or ath5k. This is the data of the card:

01:00.0 0200: 168c:001c (rev 01)
        Subsystem: 103c:137b
        Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx-
        Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
        Latency: 0, Cache Line Size: 128 bytes
        Interrupt: pin A routed to IRQ 16
        Region 0: Memory at 91300000 (64-bit, non-prefetchable) [size=64K]
        Capabilities: <access denied>
        Kernel driver in use: ath_pci
        Kernel modules: ath5k, ath_pci

MadWifi: ath_attach: Switching rfkill capability off.
wifi0: Atheros AR2425 chip found (MAC 14.2, PHY SChip 7.0, Radio 10.2)
ath_pci: wifi0: Atheros 5424/2424: mem=0x91300000, irq=16

The driver was hanging the computer from time to time, I even tried to
upgrade other machines which had stable madwifi driven cards to the new
version of the driver with similar results, they also crashed and they had
been stable before. This is not only on the 2.6.25 kernel but also on
current 2.6.26, it is a thing with the driver, nothing to do with the kernel
it is happening to me with current packages and current kernel on all my
atheros hardware.

I'm currently testing the new "released" hal code by the OpenWrt.org, DD-WRT
and MakSat guys and seems much more stable, this is only my third day with
but no problems have appeared so far.

I know it is difficult to choose a version of the soft when it comes to a
release, having the stable 0.9.4 wouldn't give support to the guys with new
hardware like me, but on the other hand, providing a unstable driver is not
good either, at least if there is a stable one for most hardware. We are
really advanced on the release cycle but... I'm wondering if having a stable
package plus a development one wouldn't have been a good idea here.

Regards...

-- System Information:
Debian Release: lenny/sid
  APT prefers testing
  APT policy: (990, 'testing'), (500, 'unstable')
Architecture: amd64 (x86_64)

Kernel: Linux 2.6.26-1-amd64 (SMP w/1 CPU core)
Locale: LANG=gl_ES.UTF-8, LC_CTYPE=gl_ES.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/bash

Versions of packages madwifi-source depends on:
ii  bzip2                         1.0.5-1    high-quality block-sorting file co
ii  debhelper                     7.0.15     helper programs for debian/rules
ii  module-assistant              0.10.11.0  tool to make module package creati

madwifi-source recommends no packages.

Versions of packages madwifi-source suggests:
ii  kernel-pac 11.001-0.1                    A utility for building Linux kerne
ii  madwifi-to 1:0.9.4+r3685.20080531+dfsg-1 tools for the Multiband Atheros Dr

-- no debconf information