Bug#658037: D-I tries to install Debian on my cellphone instead of loading the firmware
On Mon, Jan 30, 2012 at 10:55:25PM +0100, Michelle Konzack wrote:
> Package: debian-installer
> Severity: important
> Tags: squeeze
> Hello Maintainers,
> I have 146 IBM eServer xSeries 335 and they are working fine with Lenny
> and Squeeze, but now I have gotten some new ones and failed
So the working ones and the new ones are the same model (being x335
means nothing by itself given IBM loves reusing model numbers on different
> 1) D-I does not ask me for the firmware of the integrated Broadcom NICs
> -> tg3_tso.bin
> 2) does not ask me for the firmware of the integrated SCSI-Controller
> -> LSI 53c1020
It shouldn't ask since the firmware is built in so that you can boot
from it. Not like the network card.
> $ lsmod |grep mpt
> mptspi 9597 0
> mptscsih 14370 1 mptspi
> mptbase 43816 2 mptspi,mptscsih
> scsi_transport_spi 14858 0 mptspi
> scsi_mod 102061 7 mptspi,mptscsih,scsi_transport_spi,usb_storage,sg,sr_mod,libata
> $ dmesg |grep mpt
> [ 94.656803] Fusion MPT base driver 3.04.12
> [ 94.698935] Fusion MPT SPI host driver 3.04.12
> [ 94.698998] mptspi 0000:01:01.0: PCI INT A -> GSI 22 (level, low) -> IRQ 22
> [ 94.699095] mptbase: ioc0: initiating bringup
> [ 134.688010] mptbase: ioc0: ERROR - Doorbell ACK timeout (count 4999), IntStatus=3D80000000!
> [ 135.796011] mptbase: ioc0: ERROR - Diagnostic reset FAILED! (102h)
> [ 135.796016] mptbase: ioc0: WARNING - NOT READY WARNING!
> [ 135.796020] mptbase: ioc0: ERROR - didn't initialize properly! (-1)
> [ 135.796079] mptspi: probe of 0000:01:01.0: failed with error -1
> 2) If I mount the USB-Stick (cellphone or SD card) manualy, D-I detect
> it as an install medium and does not bother installing Debian on it.
> 3) D-I can not mount a fresh bought SD card, or USB Stick, because it
> want to enforce a superdisk on /dev/sda instead of mounting the
> parition /dev/sda1.
> However, the previous owner of the server had FC 8 installed on it and
> it worked, so why can D-I not load the modules and firmware?
> Normaly this bug should be "grave" or even "critial"...
> Note: Don't believe, I throw away 400 Servers of a value
> of more the 5 mio Euro even if D-I is sick!
> Realy, Debian becomes more and more worse!
Sometimes obscure hardware breaks in new kernels. It happens and until
someone with that obscure hardware tests it, no one knows it is broken.
> It was the best server systems and now it degrade it self.
It looks a lot like a bug that was fixed in 2.6.32-12, but maybe it
is different. I seems it would have to be.