[ltp] Thinkpad T40 BIOS upgrade, how to update the ECP? (PXE boot)

[Oisin Feeley]

On 3/17/07, Henrique de Moraes Holschuh <hmh@hmh.eng.br> wrote:
> On Fri, 16 Mar 2007, Oisin Feeley wrote:
> > 6. Insert the ECP CD created and tested above.  The machine went
> > through POST, but pressing F12 to get a boot device list results in a
> > jump to a PXE bootloader screen.  I can then press Ctrl-S to access
> > the Setup menu of the Intel Boot Agent 4.1.07  which presents me with
> > some choices, namely:
> > Network Boot Protocol: PXE/RPL (PXE)
> > Boot Order: Use Bios Setup Boot Order (no choices)
> > Show Setup Prompt: Enabled/Disabled (Enabled)
> > Setup Wait Time: 2s (3s)
> > Legacy OS Wakeup Support: Enabled/Disabled (Enabled)
>
> Try entering setup, and fixing the boot order temporarily there to boot from
> CD.  While at it, make triple-sure everything in setup is configured as
> you'd like it to be.

OK, some progress.  In the interim before I checked Daniel and
Henrique's responses I managed to update the BIOS and ECP.  I omitted
some important information in Step 6 above:
If I did not choose Ctrl-S and just let the machine proceed then it
would show a small window which had two choices: Legacy Diskette or
Setup.  If I chose "Setup" then I could change the boot order as
suggested above by Henrique.  I did this and chose to allow network
boot and CD (ATAPI).  [I had also removed the harddrive for testing at
this stage (it was OK, as I thought it would be because I was running
smartd and logging the reports via email on a different machine and
they all seemed fine).]
Rebooting with the ECP update disk created above I was able to update the ECP.

I then left the machine running over night with memtest86+ iterating
and found that it had done 34 passes with no errors this morning.  It
rebooted and all seemed fine.  .....

Then it froze again.  So my problem was not fixed by reseating the
keyboard, checking the CMOS battery and updating the BIOS.

I have now observed that the freeze may be correlated with the
position that the screen is angled at, or the screen being moved. So I
now suspect that the ribbon cable is something that I should try to
replace.

Thank you to Daniel and Henrique for their responses.

Summary:  it does not seem to matter whether one uses
diskette/non-diskette BIOS/ECP update, it is possible to flash the
BIOS before the ECP(hardware controller) in accordance with what Intel
suggest and in disagreement with the current (Mar 2007) advice on
ThinkWiki.  Screen corruption problems may be due to the
connector/cable rather than anything to do with the ATI Radeon
Mobility M7 graphics chip being loose (as suggested in the T40 section
of known problems at ThinkWiki).

I shall update on progress.  Thanks again for the responses.

Oisin Feeley