lilo and USB keyboard

[Timothy Madden]

Hello

I have a USB keyboard and when lilo starts at boot time I can not select
any entry in the menu. It looks like keyboard is not working yet, as
keyboard leds do not work either.

Eventually the default timeout expires and lilo will boot the default
(first) entry.

Is there a way to get lilo work with my USB keyboard (Hama AK 120
(R9011287, 105-keys + 3 power/standby buttons) romanian version) ?

I really do not think I am the only one to run into this problem, but a
search for lilo on the newsgroup shows no other messages for this problem.

Thank you,
Timothy Madden

[Henrik Carlqvist]

On Sun, 21 Apr 2013 14:29:16 +0300, Timothy Madden wrote:
> I have a USB keyboard and when lilo starts at boot time I can not select
> any entry in the menu.

Check the cmos setup settings of your motherboard if it is possible to
enable usb keyboard input for your motherboard bios. As this is obviously
not already enabled you might have to temporary connect a ps/2 keyboard
to enter and navigate your cmos setup menus.

regards Henrik
--
The address in the header is only to prevent spam. My real address is:
hc351(at)poolhem.se Examples of addresses which go to spammers:
root@localhost postmaster@localhost

[Michael Black]

On Sun, 21 Apr 2013, Henrik Carlqvist wrote:

> On Sun, 21 Apr 2013 14:29:16 +0300, Timothy Madden wrote:
>> I have a USB keyboard and when lilo starts at boot time I can not select
>> any entry in the menu.
>
> Check the cmos setup settings of your motherboard if it is possible to
> enable usb keyboard input for your motherboard bios. As this is obviously
> not already enabled you might have to temporary connect a ps/2 keyboard
> to enter and navigate your cmos setup menus.
>

I was going to suggest that, but what happens with people who have USB
keyboards to begin with? Does the BIOS see the keyboard before it's set
for a USB keyboard?

Michael

[Aragorn]

On Sunday 21 April 2013 13:29, Timothy Madden conveyed the following to
alt.os.linux.slackware...

Short explanation:

You need to enable "USB legacy support" in the BIOS setup program.

Longer explanation:

When the machine boots, the processor is be in real mode. This is
still the case in LILO as well. USB devices however require an I/O
address outside the real mode address space, so the BIOS has to
switch the processor into systems management mode to emulate a PS/2
keyboard from a USB keyboard.

Once the kernel image is loaded and decompressed, none of that
matters anymore as the kernel runs in 32-bit protected mode or
64-bit long mode and uses its own I/O routines, but the crucial part
here is the real mode phase of the machine's hardware boot sequence.
Real mode is the DOS- and legacy BIOS-compatible 8086 emulation mode
of all x86 processors from the 80286 on and upward, and it is the
mode in which - for backward compatibility reasons - the x86
processor boots up at power-up.

--
= Aragorn =
GNU/Linux user #223157 - http://www.linuxcounter.net

[greymausg]

USB keyboard here, Microsoft, no problems. Most keyboards are USB now.


--
maus
.
.
...

[Timothy Madden]

Thank you all for your answers.

I had 'Enable USB Keyboard' and 'Enable USB mouse' in my Phoenix
AwardBIOS set to 'Disabled', and that fixed my problem and lilo works
with my keyboard now.

However, disabled as the USB keyboard was, BIOS /could/ work with it and
I had not need to plug my old PS/2 keyboard in. So my guess would be
people with no PS/2 keyboard to begin with should still be fine :)

Thank you,
Timothy Madden

[Aragorn]

On Sunday 21 April 2013 19:27, Timothy Madden conveyed the following to
alt.os.linux.slackware...

Yes, the _BIOS setup utility_ can handle that because it'll
automatically switch to systems management mode if it detects a USB
keyboard or mouse, but the USB keyboard support menu item is there to
ensure that systems management mode can continue to do the USB-to-PS/2
translation _after_ you exit the BIOS setup utility.

With the option disabled, once the BIOS has completed its routine, it
defaults to standard real mode again - which is what LILO will be
running in then - and then you lose USB keyboard and mouse support.

[william...@earthlink.net]

On 2013-04-21, Timothy Madden <termin...@gmail.com> wrote:

It's a function of your BIOS. There's probably a BIOS setting that will
enable a USB keyboard.

Bud

[william...@earthlink.net]

Mostly BIOSes come from the factory set for USB keyboards, since that's what
you get instead of a PS/2 keyboard.

Bud

[jo...@wexfordpress.com]

The keyboard that came with my latest low end computer was so crummy
that I dug out an ancient PS2 Hewlett Packard keyboard. Now the lights
work and everything.

Not every change is an improvement.

John Culleton

[Pascal Hambourg]

Hello,

Aragorn a ï¿œcrit :

>
> When the machine boots, the processor is be in real mode. This is
> still the case in LILO as well. USB devices however require an I/O
> address outside the real mode address space, so the BIOS has to
> switch the processor into systems management mode to emulate a PS/2
> keyboard from a USB keyboard.

I doubt the address space has anything to do for that matter.
On my PC's, I/O port addresses used by UHCI/OHCI USB controllers are
accessible in real mode. It's just that legacy software just does not
know how to use a USB keyboard or mouse.

[Aragorn]

On Saturday 27 April 2013 11:47, Pascal Hambourg conveyed the following
to alt.os.linux.slackware...

> Hello,
>
> Aragorn a écrit :

They are not normally accessible from within real mode, but it is
possible that your BIOS boots to _unreal_ mode. This is a feature of
newer BIOSes in which the BIOS initiates a preliminary protected mode
setup - i.e. with the pagetables initialized - and then switches back to
real mode without resetting the processor. As such, the processor, in
real mode, has access to the complete protected mode address space.

This method is used by systems management mode, by the way, because it's
the only way to run firmware code while the operating system is halted
and have that firmware able to access all of the RAM in the machine.

[Pascal Hambourg]

Aragorn a ï¿œcrit :

> On Saturday 27 April 2013 11:47, Pascal Hambourg conveyed the following
> to alt.os.linux.slackware...
>

>> Aragorn a ï¿œcrit :

>>> When the machine boots, the processor is be in real mode. This is
>>> still the case in LILO as well. USB devices however require an I/O
>>> address outside the real mode address space, so the BIOS has to
>>> switch the processor into systems management mode to emulate a PS/2
>>> keyboard from a USB keyboard.
>> I doubt the address space has anything to do for that matter.
>> On my PC's, I/O port addresses used by UHCI/OHCI USB controllers are
>> accessible in real mode. It's just that legacy software just does not
>> know how to use a USB keyboard or mouse.
>
> They are not normally accessible from within real mode

What are you talking about ? UHCI/OHCI USB controllers I/O ports are in
the 64KiB I/O address space (not the 4GiB memory address space), which
are all accessible in real mode through simple IN/OUT instructions.

[Aragorn]

On Sunday 28 April 2013 12:40, Pascal Hambourg conveyed the following to
alt.os.linux.slackware...

> Aragorn a écrit :

>
>> On Saturday 27 April 2013 11:47, Pascal Hambourg conveyed the
>> following to alt.os.linux.slackware...
>>

>>> Aragorn a écrit :

>>>
>>>> When the machine boots, the processor is be in real mode. This
>>>> is still the case in LILO as well. USB devices however require
>>>> an I/O address outside the real mode address space, so the BIOS
>>>> has to switch the processor into systems management mode to
>>>> emulate a PS/2 keyboard from a USB keyboard.
>>>
>>> I doubt the address space has anything to do for that matter.
>>> On my PC's, I/O port addresses used by UHCI/OHCI USB controllers are
>>> accessible in real mode. It's just that legacy software just does
>>> not know how to use a USB keyboard or mouse.
>>
>> They are not normally accessible from within real mode
>
> What are you talking about ? UHCI/OHCI USB controllers I/O ports are
> in the 64KiB I/O address space (not the 4GiB memory address space),
> which are all accessible in real mode through simple IN/OUT
> instructions.

http://en.wikipedia.org/wiki/Unreal_mode

<quote>

Unreal mode is still extensively used by BIOS code. In particular
the System Management Mode in Intel 386SL and later processors
places the processor in unreal mode.

</quote


http://en.wikipedia.org/wiki/System_Management_Mode

<quote>

* Emulate a PS/2 mouse or keyboard by converting the messages from
USB versions of those peripherals to the messages that would have
been generated had PS/2 versions of such hardware been connected.

</quote>

I believe that this last Wikipedia page was recently rewritten, because
I distinctly remember some reference to the user of unreal mode (or SMM)
for PS/2 mouse/keyboard emulation due to the USB I/O range being outside
of real mode's reach.

[Pascal Hambourg]

Aragorn a ï¿œcrit :

> On Sunday 28 April 2013 12:40, Pascal Hambourg conveyed the following to
> alt.os.linux.slackware...
>

>> Aragorn a ï¿œcrit :

>>> They are not normally accessible from within real mode
>> What are you talking about ? UHCI/OHCI USB controllers I/O ports are
>> in the 64KiB I/O address space (not the 4GiB memory address space),
>> which are all accessible in real mode through simple IN/OUT
>> instructions.
>
> http://en.wikipedia.org/wiki/Unreal_mode

> http://en.wikipedia.org/wiki/System_Management_Mode

Thanks but I already read those pages and found no mention of what you
claim.

> I believe that this last Wikipedia page was recently rewritten, because
> I distinctly remember some reference to the user of unreal mode (or SMM)
> for PS/2 mouse/keyboard emulation due to the USB I/O range being outside
> of real mode's reach.

I found no such change in the page history. That cannot be correct,
because the I/O port range is not restricted in real mode.

[Aragorn]

On Monday 29 April 2013 00:31, Pascal Hambourg conveyed the following to
alt.os.linux.slackware...

> Aragorn a écrit :

>
>> On Sunday 28 April 2013 12:40, Pascal Hambourg conveyed the following
>> to alt.os.linux.slackware...
>>

>>> Aragorn a écrit :

>>>
>>>> They are not normally accessible from within real mode
>>>
>>> What are you talking about ? UHCI/OHCI USB controllers I/O ports are
>>> in the 64KiB I/O address space (not the 4GiB memory address space),
>>> which are all accessible in real mode through simple IN/OUT
>>> instructions.
>>
>> http://en.wikipedia.org/wiki/Unreal_mode
>> http://en.wikipedia.org/wiki/System_Management_Mode
>
> Thanks but I already read those pages and found no mention of what you
> claim.
>
>> I believe that this last Wikipedia page was recently rewritten,
>> because I distinctly remember some reference to the user of unreal
>> mode (or SMM) for PS/2 mouse/keyboard emulation due to the USB I/O
>> range being outside of real mode's reach.
>
> I found no such change in the page history. That cannot be correct,
> because the I/O port range is not restricted in real mode.

Fine, then let's leave it at that. It doesn't really matter.

The point is that a legacy BIOS needs system management mode (which
itself implies unreal mode) in order to be able to use USB input devices
as PS/2 input devices.