Hello all,
how do I work with a mb that contains no keyboard controller, but has only USB
for keyboard and mouse?
While booting the kernel I get:
pc_keyb: controller jammed (0xFF)
(a lot of these :-)
and afterwards I cannot use the USB keyboard.
Everything works with a mb that contains a keyboard-controller, but where I use a
USB keyboard.
--
Regards,
Stephan
On Thu, 2003-01-09 at 10:42, Stephan von Krawczynski wrote:
> Hello all,
>
> how do I work with a mb that contains no keyboard controller, but has only USB
> for keyboard and mouse?
> While booting the kernel I get:
>
> pc_keyb: controller jammed (0xFF)
Does your BIOS do keyboard emulation ?
On 09 Jan 2003 17:42:01 +0000
Alan Cox <[email protected]> wrote:
> On Thu, 2003-01-09 at 10:42, Stephan von Krawczynski wrote:
> > Hello all,
> >
> > how do I work with a mb that contains no keyboard controller, but has only
> > USB for keyboard and mouse?
> > While booting the kernel I get:
> >
> > pc_keyb: controller jammed (0xFF)
>
> Does your BIOS do keyboard emulation ?
It is Compaq EVO D510. It has merely nothing of interest in the BIOS (no
keyboard emu). As far as I remember it contains an I845 chipset.
--
Regards,
Stephan
On Thu, 2003-01-09 at 17:39, Stephan von Krawczynski wrote:
> > > pc_keyb: controller jammed (0xFF)
> >
> > Does your BIOS do keyboard emulation ?
>
> It is Compaq EVO D510. It has merely nothing of interest in the BIOS (no
> keyboard emu). As far as I remember it contains an I845 chipset.
Can you use the USB keyboard to configure the BIOS during boot. If so
then it almost certainly has USB bios emulation. Another trivial test
that would be useful is to stick a freedos boot floppy in the box and
see if freedos works
Hi,
I have a question with my linux box. I am using supermicro 370dl3, dual
1GHz. 2 HD one on scsi aic-7892, the other one on Promise 100 TX2.
I am running SuSE 8.0, current kernel version is 2.4.18, coming with
installation. Now I am try to upgrade to 2.4.19. However, I use the 2.4.19
from ftp://ftp.suse.com/pub/people/mantel/linux-2.4.19-41.tar.bz2. It
doest boot up right. when it detect hd on pdc20268 then it give me "lost
interrupt" message. This message can go away if I use command line
"noapic" at boot prompt.
So, I switch to 2.4.19 from ftp.kernel.edu, then this kernel it boot fine
without any argument at boot prompt.
Can anybody guide me to make 2.4.19 from suse ftp site work without put
"noapic" at boot prompt.
One more thing this machine can not shutdown, even it is atx case. I tried
to put acpi on but it give me some table name error.
Thank you.
+AU
On 9 Jan 2003, Alan Cox wrote:
| On Thu, 2003-01-09 at 17:39, Stephan von Krawczynski wrote:
| > > > pc_keyb: controller jammed (0xFF)
| > >
| > > Does your BIOS do keyboard emulation ?
| >
| > It is Compaq EVO D510. It has merely nothing of interest in the BIOS (no
| > keyboard emu). As far as I remember it contains an I845 chipset.
|
| Can you use the USB keyboard to configure the BIOS during boot. If so
| then it almost certainly has USB bios emulation. Another trivial test
| that would be useful is to stick a freedos boot floppy in the box and
| see if freedos works
| -
PS/2 keyboard emulation might not show up in the BIOS Setup menu.
Or do you know that the BIOS doesn't contain PS/2 keyboard emulation?
Have you installed Linux on it? If so, how did you do that?
Once past this hurdle, there are patches that can help.
--
~Randy
>> > > pc_keyb: controller jammed (0xFF)
>> >
>> > Does your BIOS do keyboard emulation ?
>>
>> It is Compaq EVO D510. It has merely nothing of interest in the BIOS (no
>> keyboard emu). As far as I remember it contains an I845 chipset.
>
> Can you use the USB keyboard to configure the BIOS during boot. If so
> then it almost certainly has USB bios emulation. Another trivial test
> that would be useful is to stick a freedos boot floppy in the box and
> see if freedos works
I fail to see the point, Alan. Stephan's BIOS does exactly the
right thing: it emulates BIOS INTs which allow to read buffered
keystrokes, but it does not do SMM tricks to emulate port 0x60.
This is great, now pc_keyb.d must do detection right. It must
not loop endlessly if 0xff is returned from inb(). It's a bug.
-- Pete
On Thu, 2003-01-09 at 19:16, Pete Zaitcev wrote:
> I fail to see the point, Alan. Stephan's BIOS does exactly the
> right thing: it emulates BIOS INTs which allow to read buffered
> keystrokes, but it does not do SMM tricks to emulate port 0x60.
> This is great, now pc_keyb.d must do detection right. It must
> not loop endlessly if 0xff is returned from inb(). It's a bug.
Thats what I wanted to verify.
On Thu, Jan 09, 2003 at 11:42:47AM +0100, Stephan von Krawczynski wrote:
> Hello all,
>
> how do I work with a mb that contains no keyboard controller, but has only USB
> for keyboard and mouse?
> While booting the kernel I get:
>
> pc_keyb: controller jammed (0xFF)
>
> (a lot of these :-)
>
> and afterwards I cannot use the USB keyboard.
> Everything works with a mb that contains a keyboard-controller, but where I use a
> USB keyboard.
Get 2.5. ;) It should work without a kbd controller ... you can even
disable it in the kernel config ...
--
Vojtech Pavlik
SuSE Labs
On Thu, 9 Jan 2003 23:24:59 +0100
Vojtech Pavlik <[email protected]> wrote:
> On Thu, Jan 09, 2003 at 11:42:47AM +0100, Stephan von Krawczynski wrote:
> > Hello all,
> >
> > how do I work with a mb that contains no keyboard controller, but has only
> > USB for keyboard and mouse?
> > While booting the kernel I get:
> >
> > pc_keyb: controller jammed (0xFF)
> >
> > (a lot of these :-)
> >
> > and afterwards I cannot use the USB keyboard.
> > Everything works with a mb that contains a keyboard-controller, but where I
> > use a USB keyboard.
>
> Get 2.5. ;) It should work without a kbd controller ... you can even
> disable it in the kernel config ...
Nice idea, but not acceptable as this setup is for production use, you simply
won't do that.
It would be helpful if there was a kernel parameter for disabling the
keyboard(-check) in 2.4. We found out that disabling it as kernel patch is not
the right way, as standard setups with keyboard controller do not work any
longer afterwards. This is a setup where user should be able to choose...
The box contains a BIOS where I can type around with USB-keyboard, btw.
--
Regards,
Stephan
On Thu, Jan 16, 2003 at 12:03:24PM +0100, Stephan von Krawczynski wrote:
> On Thu, 9 Jan 2003 23:24:59 +0100
> Vojtech Pavlik <[email protected]> wrote:
>
> > On Thu, Jan 09, 2003 at 11:42:47AM +0100, Stephan von Krawczynski wrote:
> > > Hello all,
> > >
> > > how do I work with a mb that contains no keyboard controller, but has only
> > > USB for keyboard and mouse?
> > > While booting the kernel I get:
> > >
> > > pc_keyb: controller jammed (0xFF)
> > >
> > > (a lot of these :-)
> > >
> > > and afterwards I cannot use the USB keyboard.
> > > Everything works with a mb that contains a keyboard-controller, but where I
> > > use a USB keyboard.
> >
> > Get 2.5. ;) It should work without a kbd controller ... you can even
> > disable it in the kernel config ...
>
> Nice idea, but not acceptable as this setup is for production use, you simply
> won't do that.
> It would be helpful if there was a kernel parameter for disabling the
> keyboard(-check) in 2.4. We found out that disabling it as kernel patch is not
> the right way, as standard setups with keyboard controller do not work any
> longer afterwards. This is a setup where user should be able to choose...
> The box contains a BIOS where I can type around with USB-keyboard, btw.
Anyway, did you try 2.5? I just would like to know if the keyboard
controller is properly not-detected and the system doesn't crash there?
--
Vojtech Pavlik
SuSE Labs
On Thu, 16 Jan 2003, Stephan von Krawczynski wrote:
| On Thu, 9 Jan 2003 23:24:59 +0100
| Vojtech Pavlik <[email protected]> wrote:
|
| > On Thu, Jan 09, 2003 at 11:42:47AM +0100, Stephan von Krawczynski wrote:
| > > Hello all,
| > >
| > > how do I work with a mb that contains no keyboard controller, but has only
| > > USB for keyboard and mouse?
| > > While booting the kernel I get:
| > >
| > > pc_keyb: controller jammed (0xFF)
| > >
| > > (a lot of these :-)
| > >
| > > and afterwards I cannot use the USB keyboard.
| > > Everything works with a mb that contains a keyboard-controller, but where I
| > > use a USB keyboard.
| >
| > Get 2.5. ;) It should work without a kbd controller ... you can even
| > disable it in the kernel config ...
|
| Nice idea, but not acceptable as this setup is for production use, you simply
| won't do that.
| It would be helpful if there was a kernel parameter for disabling the
| keyboard(-check) in 2.4. We found out that disabling it as kernel patch is not
| the right way, as standard setups with keyboard controller do not work any
| longer afterwards. This is a setup where user should be able to choose...
| The box contains a BIOS where I can type around with USB-keyboard, btw.
I asked 1 week ago if the system BIOS does PS/2 keyboard emulation.
I think you are saying here that it does, so I'll try to confirm that:
Does the USB keyboard work when talking to BIOS Setup?
Does the USB keyboard work if NO Linux USB drivers are loaded?
Or are you using the USB keyboard only by having Linux USB drivers
loaded?
I posted a patch to 2.4.20 on 2002-Dec-04 that might work for you.
It's available at
http://www.osdl.org/archive/rddunlap/patches/kbc_option_2420.patch
It might work for you. If you try it out, please let me know how it
does for you.
--
~Randy
On Thu, 16 Jan 2003 08:18:04 -0800 (PST)
"Randy.Dunlap" <[email protected]> wrote:
> I posted a patch to 2.4.20 on 2002-Dec-04 that might work for you.
> It's available at
> http://www.osdl.org/archive/rddunlap/patches/kbc_option_2420.patch
>
> It might work for you. If you try it out, please let me know how it
> does for you.
Hello Randy,
we checked your patch and it works as you expected, only it is not what we are
looking for. We would like to be able to make _one_ kernel, that can be used on
boards with PS/2 keyboard or USB keyboard, but without a PS/2 keyboard check
that takes as long as it does in current version (and gives _one_ warning, but
not tens).
Do you think this is solvable?
--
Regards,
Stephan
> On Thu, 16 Jan 2003 08:18:04 -0800 (PST)
> "Randy.Dunlap" <[email protected]> wrote:
>
>> I posted a patch to 2.4.20 on 2002-Dec-04 that might work for you. It's
>> available at
>> http://www.osdl.org/archive/rddunlap/patches/kbc_option_2420.patch
>>
>> It might work for you. If you try it out, please let me know how it does
>> for you.
>
> Hello Randy,
>
> we checked your patch and it works as you expected, only it is not what we
> are looking for. We would like to be able to make _one_ kernel, that can be
> used on boards with PS/2 keyboard or USB keyboard, but without a PS/2
> keyboard check that takes as long as it does in current version (and gives
> _one_ warning, but not tens).
> Do you think this is solvable?
OK, I see. I'm willing to try a few more things on this if you are
willing to test them...are you?
Is the "controller jammed" message the only one that you are seeing?
~Randy
>> On Thu, 16 Jan 2003 08:18:04 -0800 (PST)
>> "Randy.Dunlap" <[email protected]> wrote:
>>
>>> I posted a patch to 2.4.20 on 2002-Dec-04 that might work for you. It's
>>> available at
>>> http://www.osdl.org/archive/rddunlap/patches/kbc_option_2420.patch
>>>
>>> It might work for you. If you try it out, please let me know how it does
>>> for you.
>>
>> Hello Randy,
>>
>> we checked your patch and it works as you expected, only it is not what we
>> are looking for. We would like to be able to make _one_ kernel, that can
>> be used on boards with PS/2 keyboard or USB keyboard, but without a PS/2
>> keyboard check that takes as long as it does in current version (and gives
>> _one_ warning, but not tens).
>> Do you think this is solvable?
>
> OK, I see. I'm willing to try a few more things on this if you are willing
> to test them...are you?
>
> Is the "controller jammed" message the only one that you are seeing?
On a "jammed" system, please boot Linux with this string added to the
command line: "kbd-reset"
and tell me if you see any other keyboard messages, such as one that
begins with "initialize_kbd:" or these: "Keyboard timed out",
"keyboard: Timeout" or "keyboard: Too many NACKs".
Thanks,
~Randy
On Sat, 25 Jan 2003 21:29:35 -0800 (PST)
Randy Dunlap <[email protected]> wrote:
> OK, I see. I'm willing to try a few more things on this if you are
> willing to test them...are you?
Hello Randy,
thanks for your comments. Of course we will test any given patch. Only I have
to beg for some patience, as the questionable hardware is not in my central
office, so expect a day or two delay for my answers to your questions and
patches.
--
Regards,
Stephan