Problems with shutdown

[Douglas Alves]

PC:      Desktop w/ Celeron 2.5Gh, 128k cache, 256Mb RAM DDR333
Printer: Epson Stylus C60 on usb

Since Ubuntu Warty installation, my PC would never power-off after 'Shut
down' in Computer-Log Out. The cabinet switch with which I power-on my
PC doesn't cut power. In Windows (xp) it initiates Shutdown and Power
off (I believe this is a feature of modern PC power sources), so to
power-off the PC from Warty I always had to select 'Restart the
computer', enter Windows and then press either the keyboard Power key or
the PC's cabinet power switch.

My printer has always worked fine with the Gimp, gEditor, and Evolution;
but not with OpenOffice. When it stopped working completely I had to
stop and think, What did I change for this to happen? So I discovered
the following:

I had noticed that at boot up (ever since being installed), Warty never
complained on shpchp or pciehp like it did on an IBM ThinkPad T21 on
which I had installed Warty. My desktop PC has always loaded apm, but
the suggestions given on the list had never solved my no-power-off
problem. This week somebody pointed out the (recent) solution on the
Unofficial Ubuntu Guide of adding 'acpi=off apm=off' to the Ubuntu
kernel line in Grub's menu.lst. I made this change and the PC powers-off
but now the modprobe: FATAL: Errors come up at boot time and the usb
printer is useless.
   With the printer on at startup, the booting seems to pause for about
25 seconds after the modprobe FATAL Errors (before showing 'Configuring
network interfaces'); with the printer off it takes about 4 sec, and in
both cases the printer isn't available. I removed what I had added to
the Grub line, and the PC and printer worked fine.
   I added only the 'acpi=off' to Grub: the FATAL Errors show up and
printer is useless, but the PC shuts down fine (it turns off).
   I removed 'acpi=off' and put only 'apm=off': no F.E., the printer
came to life after Ubuntu loaded ACPI modules. I was able to use the
printer in this session, but it didn't power-off after using 'Shut
down'.
   In all this I haven't followed the advice of changing the
file /etc/modules; though it does list apm.

I've made tests repeating the above conditions and the behavior remains
consistent.

Can the developers see any connection between the known bug of 'shpchp
and pciehp', the apm paramenters added above, and the mal-functioning of
a usb printer?  There must be a relation. As I say, this behavior is
consistent and repeatable.

Joaquin:
   Si necesitas traduccion o deseas comparar notas, puedes escribirme.

Cheers,
-Douglas Alves