Syscall and syscall versioning documentation for review

Hello,

I have updated the markup a bit, so this is ready for review by our
Docbook experts. I will update the section on the naming convention when
we reach consensus.

Again, please review:

http://martani.no-ip.org/~pavel/chap-processes.html

Pavel

Index: chap-processes.xml
===================================================================
RCS file: /cvsroot/htdocs/Documentation/internals/en/chap-processes.xml,v
retrieving revision 1.4
diff -u -r1.4 chap-processes.xml
--- chap-processes.xml	3 Mar 2006 12:01:21 -0000	1.4
+++ chap-processes.xml	1 Sep 2006 09:49:12 -0000
 <at>  <at>  -565,12 +565,7  <at>  <at> 
       <title>Traps</title>
       <para>XXX write me</para>
     </sect2>
-
-    <sect2 id="libc_syscall">
-      <title>System call implementation in libc</title>
-      <para>XXX write me</para>
-    </sect2>
-

"Device not configured" FAQ: some incompleteness

From NetBSD FAQ (web site)

"Frequently Asked Hardware Questions

** What does device not configured mean? (top)

  * If this message appears during the autoconfiguration output of system
    boot, it means
    ...

  * If this message appears when you try to access a device node in /dev (e.g.
    a SCSI disk), this means that the driver can't find the specific device
    unit you tried to access, e.g.
    ...
"

That's all about "Device not configured".
But there is another case:

% pfctl -f /etc/pf.conf
pfctl: /dev/pf: Device not configured
%

'modload pf.o' help in this case...

--

-- 
Best regards, Aleksey Cheusov.

Re: "Device not configured" FAQ: some incompleteness

On Sat, 2 Sep 2006, Aleksey Cheusov wrote:
> But there is another case:
>
> % pfctl -f /etc/pf.conf
> pfctl: /dev/pf: Device not configured
> %
>
> 'modload pf.o' help in this case...

I've added something to 
http://www.netbsd.org/Documentation/kernel/index.html#device-not-configured, 
it should be life within an hour.

If you have any suggestions for better wording let me know!

  - Hubert

Re: "Device not configured" FAQ: some incompleteness

On Mon, Sep 04, 2006 at 12:47:25AM +0200, Hubert Feyrer wrote:
> 
> On Sat, 2 Sep 2006, Aleksey Cheusov wrote:
> >But there is another case:
> >
> >% pfctl -f /etc/pf.conf
> >pfctl: /dev/pf: Device not configured
> >%
> >
> >'modload pf.o' help in this case...
> 
> I've added something to 
> http://www.netbsd.org/Documentation/kernel/index.html#device-not-configured, it 
> should be life within an hour.
> 
> If you have any suggestions for better wording let me know!

This is in fact the same case as with the above, just with a pseudo-device
instead of a physical device.

BTW your statement that "Using ktrace(1)  can help you find what's going
on inside a command, and to determine you what's being accessed that may
cause the error message." is misleading. There is no need to use ktrace in
the case mentioned. pfctl tells you what device node did it attempt to
use. (/dev/pf)

Pavel

Re: "Device not configured" FAQ: some incompleteness

On Mon, 4 Sep 2006, Pavel Cahyna wrote:
> This is in fact the same case as with the above, just with a pseudo-device
> instead of a physical device.
>
> BTW your statement that "Using ktrace(1)  can help you find what's going
> on inside a command, and to determine you what's being accessed that may
> cause the error message." is misleading. There is no need to use ktrace in
> the case mentioned. pfctl tells you what device node did it attempt to
> use. (/dev/pf)

Want to send me a patch to update the text?
I'm not 100% sure what you're suggesting to change...

  - Hubert

Re: "Device not configured" FAQ: some incompleteness

On Mon, Sep 04, 2006 at 10:36:27PM +0200, Hubert Feyrer wrote:
> On Mon, 4 Sep 2006, Pavel Cahyna wrote:
> >This is in fact the same case as with the above, just with a pseudo-device
> >instead of a physical device.
> >
> >BTW your statement that "Using ktrace(1)  can help you find what's going
> >on inside a command, and to determine you what's being accessed that may
> >cause the error message." is misleading. There is no need to use ktrace in
> >the case mentioned. pfctl tells you what device node did it attempt to
> >use. (/dev/pf)
> 
> Want to send me a patch to update the text?
> I'm not 100% sure what you're suggesting to change...

--- index.xml.~1.11.~	Mon Sep  4 22:46:30 2006
+++ index.xml	Mon Sep  4 23:29:37 2006
 <at>  <at>  -662,7 +662,7  <at>  <at> 
             node in <code>/dev</code> (e.g. a SCSI disk), this means
             that the driver can't find the specific device unit you
             tried to access, e.g. accessing a SCSI disk that isn't
-            there.
+            there, or the driver is not compiled in the kernel.
           </para>

           <para>
 <at>  <at>  -676,18 +676,33  <at>  <at> 
             <code>/var/run/dmesg.boot</code> (a saved copy of the
             boot time autoconfiguration output).
           </para>
-        </listitem>

Re: "Device not configured" FAQ: some incompleteness

> On Mon, Sep 04, 2006 at 10:36:27PM +0200, Hubert Feyrer wrote:
 >> On Mon, 4 Sep 2006, Pavel Cahyna wrote:
 >> >This is in fact the same case as with the above, just with a pseudo-device
 >> >instead of a physical device.
 >> >
 >> >BTW your statement that "Using ktrace(1)  can help you find what's going
 >> >on inside a command, and to determine you what's being accessed that may
 >> >cause the error message." is misleading. There is no need to use ktrace in
 >> >the case mentioned. pfctl tells you what device node did it attempt to
 >> >use. (/dev/pf)
 >> 
 >> Want to send me a patch to update the text?
 >> I'm not 100% sure what you're suggesting to change...

> --- index.xml.~1.11.~	Mon Sep  4 22:46:30 2006
> +++ index.xml	Mon Sep  4 23:29:37 2006

Now it looks much better and clear. Thank you both!

--

-- 
Best regards, Aleksey Cheusov.

NetBSD Guide section 12.8: some incompletness

12.8. Using audio CDs with NetBSD
subsection 2. To read ("rip") audio tracks)

Sorry for being intrusive, but this section is somewhat incomplete too.

Things described is there just don't work (at least with
cdparanoia) because the ordinary users have no write permissions to
/dev/cd0X devices required to grab CDs.
The default permissions are root:operator - 640.

I see the following solutions for this:
- chmod +s /usr/pkg/bin/cdparanoia
  that allows wheel members to grab CDs.
  For obvious reasons this is probably the worst method.
- chmod g+w
  Bad too. CD grabhbing users should not be able to reade sdX/wdX.
- chmod o+rw
  No comments
- Configuring /etc/ttyaction in order users to take ownership
  of cdXY device files.
  I didn't try this because I don't know how to configure this in case of
  XDM login.

- groupadd cdrom
  useradd -G cdrom <username>
  chgrp cdrom /dev/cdXY
  chmod g+w /dev/cdXY

  In my veiw this way is the best one, but there are a few questions.
  Can permissions and ownership of /dev files be overriden

Re: NetBSD Guide section 12.8: some incompletness

On Thu, Sep 07, 2006 at 11:43:01PM +0300, Aleksey Cheusov wrote:
> 12.8. Using audio CDs with NetBSD
> subsection 2. To read ("rip") audio tracks)
> 
> Sorry for being intrusive, but this section is somewhat incomplete too.
> 
> Things described is there just don't work (at least with
> cdparanoia) because the ordinary users have no write permissions to
> /dev/cd0X devices required to grab CDs.
> The default permissions are root:operator - 640.
> 
> I see the following solutions for this:
> - chmod +s /usr/pkg/bin/cdparanoia
>   that allows wheel members to grab CDs.
>   For obvious reasons this is probably the worst method.
> - chmod g+w
>   Bad too. CD grabhbing users should not be able to reade sdX/wdX.

how are permissions of cdX related to sdX/wdX ?

Pavel

Re: NetBSD Guide section 12.8: some incompletness

> On Thu, Sep 07, 2006 at 11:43:01PM +0300, Aleksey Cheusov wrote:
 >> 12.8. Using audio CDs with NetBSD
 >> subsection 2. To read ("rip") audio tracks)
 >> 
 >> Sorry for being intrusive, but this section is somewhat incomplete too.
 >> 
 >> Things described is there just don't work (at least with
 >> cdparanoia) because the ordinary users have no write permissions to
 >> /dev/cd0X devices required to grab CDs.
 >> The default permissions are root:operator - 640.
 >> 
 >> I see the following solutions for this:
 >> - chmod +s /usr/pkg/bin/cdparanoia
 >>   that allows wheel members to grab CDs.
I missed 
chmod o-rx /usr/pkg/bin/cdparanoia
here

 >>   For obvious reasons this is probably the worst method.
 >> - chmod g+w
 >>   Bad too. CD grabhbing users should not be able to reade sdX/wdX.

> how are permissions of cdX related to sdX/wdX ?
and
useradd -G operator <username>
here.

--

-- 
Best regards, Aleksey Cheusov.