Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.os.linux.hardware > #963 > unrolled thread

HOW2 test a frame-buffer?

Back to article view | Back to comp.os.linux.hardware

Contents

  HOW2 test a frame-buffer? no.top.post@gmail.com - 2012-01-04 14:10 +0000
    Re: HOW2 test a frame-buffer? no.top.post@gmail.com - 2012-01-06 08:15 +0000
      Re: HOW2 test a frame-buffer? pacman@kosh.dhis.org (Alan Curry) - 2012-01-06 08:57 +0000
        Re (2): HOW2 test a frame-buffer? no.top.post@gmail.com - 2012-01-06 16:26 +0000
      Re: HOW2 test a frame-buffer? "David W. Hodgins" <dwhodgins@nomail.afraid.org> - 2012-01-06 23:18 -0500
        Re (2): HOW2 test a frame-buffer? no.top.post@gmail.com - 2012-01-08 05:35 +0000
        Re (3): HOW2 test a frame-buffer? no.top.post@gmail.com - 2012-01-08 15:42 +0000

#963 — HOW2 test a frame-buffer?

An application that I need to have on a HP-mini netbook,
needs a frame buffer.

When I did: cd /dev; ./MAKEDEV  -n -v  fb
it showed :
udev active, devices will be created in /dev/.static/dev/
create fb0      c 29 0 root:video 0660
create fb1      c 29 1 root:video 0660
create fb2      c 29 2 root:video 0660
create fb3      c 29 3 root:video 0660
create fb4      c 29 4 root:video 0660
create fb5      c 29 5 root:video 0660
create fb6      c 29 6 root:video 0660
create fb7      c 29 7 root:video 0660

Because the app. still failed, I did:
ln -s /dev/.static/dev/fb0   /dev/fb0
which made a difference: 
 the VT-display, just blanks.

But I was able to 'key' to other VTs and
Ctrl/C, Ctrl/D ..to free up the VT, which I
had to re-login, after each test cycle.

My slak13 PC shows some info for `fbset`;
but the netbook has got DebianEtch without any
utilities to test the frame-buffer, that I can find.

How would I test the fb* facility independantly
of my application?

== TIA.

[toc] | [next] | [standalone]

#968

In article <je1ml9$3gs$1@dont-email.me>, no.top.post@gmail.com wrote: 

OK, here's the elementary info, which I wanted:--
because the framebuffer is a 'live' file/device you can use file access
to read/write it [probably] and see what it shows.

-> cat <a file>  > /dev/fb0
will write to the display, I tested in VT mode.
Perhaps <video buffer overflow: error> could be an error,
but for my test a 300KB file just <wrote pixels to the top 20% of
the screen.   My more advance question is below.

> An application that I need to have on a HP-mini netbook,
> needs a frame buffer.
> 
> When I did: cd /dev; ./MAKEDEV  -n -v  fb
> it showed :
> udev active, devices will be created in /dev/.static/dev/
> create fb0      c 29 0 root:video 0660
> create fb1      c 29 1 root:video 0660
> create fb2      c 29 2 root:video 0660
> create fb3      c 29 3 root:video 0660
> create fb4      c 29 4 root:video 0660
> create fb5      c 29 5 root:video 0660
> create fb6      c 29 6 root:video 0660
> create fb7      c 29 7 root:video 0660
> 
> Because the app. still failed, I did:
> ln -s /dev/.static/dev/fb0   /dev/fb0
> which made a difference: 
>  the VT-display, just blanks.
> 
> But I was able to 'key' to other VTs and
> Ctrl/C, Ctrl/D ..to free up the VT, which I
> had to re-login, after each test cycle.
> 
> My slak13 PC shows some info for `fbset`;
> but the netbook has got DebianEtch without any
> utilities to test the frame-buffer, that I can find.
> 
> How would I test the fb* facility independantly
> of my application?
> 
> == TIA.

On my PC:slak13 which has the normal /dev/fb0
the above described test is good.

On the HP:Win7:netbook with CF:Debian:Etch
-> cat <file>    >   /dev/.static/dev/fb0
== <-bash: no such device: /dev/.static/dev/>

And yet
-> ls -l /dev/.static/dev/fb*
shows all 8 of them.

-> mknod /dev/fb0  c  29  0
-> ls -l /dev/fb0  == good
-> cat <file>    >   /dev/fb0
== <-bash: no such device: /dev/.static/dev/>

So I'm wondering why, although `ls -l` sees </dev/*/fb0>,
apparently `bash` does not ?

== TIA.

[toc] | [prev] | [next] | [standalone]

#969

In article <je6aj6$gm$1@dont-email.me>,  <no.top.post@gmail.com> wrote:
>
>-> mknod /dev/fb0  c  29  0
>-> ls -l /dev/fb0  == good
>-> cat <file>    >   /dev/fb0
>== <-bash: no such device: /dev/.static/dev/>
>
>So I'm wondering why, although `ls -l` sees </dev/*/fb0>,
>apparently `bash` does not ?
>

Pay close attention to the error message. "No such device" (ENODEV) isn't the
same as "No such file or directory" (ENOENT).

ENODEV usually means the corresponding kernel driver is not present. The
variant "No such device or address" (ENXIO) will sometimes appear when a
driver is loaded but the device itself wasn't detected.

You need to load the fb kernel module for your hardware.

-- 
Alan Curry

[toc] | [prev] | [next] | [standalone]

#970 — Re (2): HOW2 test a frame-buffer?

In article <je6d29$ck$1@speranza.aioe.org>, pacman@kosh.dhis.org (Alan Curry) wrote: 

> In article <je6aj6$gm$1@dont-email.me>,  <no.top.post@gmail.com> wrote:
> >
> >-> mknod /dev/fb0  c  29  0
> >-> ls -l /dev/fb0  == good
> >-> cat <file>    >   /dev/fb0
> >== <-bash: no such device: /dev/.static/dev/>
> >
> >So I'm wondering why, although `ls -l` sees </dev/*/fb0>,
> >apparently `bash` does not ?
> >
> 
> Pay close attention to the error message. "No such device" (ENODEV) isn't the
> same as "No such file or directory" (ENOENT).
> 
> ENODEV usually means the corresponding kernel driver is not present. The
> variant "No such device or address" (ENXIO) will sometimes appear when a
> driver is loaded but the device itself wasn't detected.
> 
I don't think a fb is 'detectable' except by the human eye.
On PC:slak13 `cat <file>  > /dev/fb0` "shows" and fb apparently uses the
same video-memory as the non-fb-display.

> You need to load the fb kernel module for your hardware.
> 
I need to know how to test if the kernel has fb, and
I need to know how to test if the fb-module can be loaded,
if not already in the kernel.
> -- 
> Alan Curry

The PC with Debian [not running now] but tested for fb
shows /boot/System.map* has 16 x "framebuffer"
/boot/config* has 2 x "FRAMEBUFFER".
and the dud-netbook has the same count.
The dud-netbook has `lsmod | grep frame` == nill.

I'm guessing that boot/System.map & config
tell that the kernel HAS fb facilities.

The most confusing for me is the sub-dir:
/dev/.static/dev/  which has the 8 fb*

==TIA.

[toc] | [prev] | [next] | [standalone]

#972

On Fri, 06 Jan 2012 03:15:35 -0500, <no.top.post@gmail.com> wrote:

> On the HP:Win7:netbook with CF:Debian:Etch
> -> cat <file>    >   /dev/.static/dev/fb0
> == <-bash: no such device: /dev/.static/dev/>
>
> And yet
> -> ls -l /dev/.static/dev/fb*
> shows all 8 of them.

Note that the error is not "No such file or directory", which
you would get if the filename was not present.

The underlying device is not there, meaning the frame buffer
module has not been loaded, so the tty is in text only mode.

Try adding vga=795 as a kernel boot parameter.

Regards, Dave Hodgins

-- 
Change nomail.afraid.org to ody.ca to reply by email.
(nomail.afraid.org has been set up specifically for
use in usenet. Feel free to use it yourself.)

[toc] | [prev] | [next] | [standalone]

#973 — Re (2): HOW2 test a frame-buffer?

In article <op.v7o4xkyna3w0dxdave@hodgins.homeip.net>, "David W. Hodgins" <dwhodgins@nomail.afraid.org> wrote: 

> On Fri, 06 Jan 2012 03:15:35 -0500, <no.top.post@gmail.com> wrote:
> 
> > On the HP:Win7:netbook with CF:Debian:Etch
> > -> cat <file>    >   /dev/.static/dev/fb0
> > == <-bash: no such device: /dev/.static/dev/>
> >
> > And yet
> > -> ls -l /dev/.static/dev/fb*
> > shows all 8 of them.
> 
> Note that the error is not "No such file or directory", which
> you would get if the filename was not present.
> 
> The underlying device is not there, meaning the frame buffer
> module has not been loaded, so the tty is in text only mode.
> 
> Try adding vga=795 as a kernel boot parameter.
> 
> Regards, Dave Hodgins
> ================
I'm copying this manually from CF:Debian:Etch to post
via PC:slak13 
-> lsmod | wc -l == 41 
-> grep: "vi", "vga", "fb"  == none
and PC:slak13 has got 43, and none show in `lsmod`, yet
slak runs 'fb'. So I guess the fb-driver is IN the slak-kernel,
not as a module.
I've extended the DebEtch:boot/grub/menu.lst   to:-
kernel /b...  root=/dev/sdb2  vga=795 ro

But I still get ":NO such device" when I `cat FIle`
to *fb0, although *fb0 is seen by `ls -l`
-- 
When I 'live edited` another DebEtch's grub entry with a SEPARATE
line: vga=795
it showed "Unrecognised command"
which is good.
So now I'll put the 'vga=795` before the 'root='

That's BETTER: "You passed an undefined node number..."
"Enter mode number or 'scan': 
<see menu of valid modes: VESA VGA>
80x25...80x60
-> repeated: 'scan' == shows unstable screen & increases menu from 7 to 9 items,
last/8 being 00x02
-> select '8' = blank screen.
---> select '7'/40x25 =BIG text
keeps BIGtext in linux, but still: -> ./fbTst = "No such device"

Reboot:-
-> select <80x60> == DOES INITIALLY show small text while booting.
--> select <80x34> == after <set clock again> restores to bigger text
-> repeated: 'scan' which always shows unstable screen
-> select '8' = blank screen.

Ok, a different DebEtch1 has got /dev/fb*  [but no mc]
and DebEtch2 [which I've been concentrating on
had got the STRANGE /dev/.static/dev/fb*

-> try goog: Debian+grub+framebuffer 
-> goog: grub+framebuffer+"No+such+device"

Thanks !

[toc] | [prev] | [next] | [standalone]

#975 — Re (3): HOW2 test a frame-buffer?

In article <op.v7o4xkyna3w0dxdave@hodgins.homeip.net>, "David W. Hodgins" <dwhodgins@nomail.afraid.org> wrote: 

> On Fri, 06 Jan 2012 03:15:35 -0500, <no.top.post@gmail.com> wrote:
> 
> > On the HP:Win7:netbook with CF:Debian:Etch
> > -> cat <file>    >   /dev/.static/dev/fb0
> > == <-bash: no such device: /dev/.static/dev/>
> >
> > And yet
> > -> ls -l /dev/.static/dev/fb*
> > shows all 8 of them.
> 
> Note that the error is not "No such file or directory", which
> you would get if the filename was not present.
> 
> The underlying device is not there, meaning the frame buffer
> module has not been loaded, so the tty is in text only mode.
> 
> Try adding vga=795 as a kernel boot parameter.
> 
> Regards, Dave Hodgins
> ================
OK, I followed your lead, to some google articles.
vga=865 does it for me.
It's amazing to see this HP-mini break free from the
the Micro$loth monopoly.

Thanks.

[toc] | [prev] | [standalone]

Back to top | Article view | comp.os.linux.hardware

csiph-web