Re: 2.6 Cursor problem (was init acting the maggot):-(

2005-07-02 Thread Declan Moriarty
Recently, Somebody Somewhere wrote these words
 On Fri, 1 Jul 2005, Declan Moriarty wrote:
 
Did the cursor work adequately on 2.6 in the past ?
  
  Never. But I never used 2.6 much.
 
Which version of 2.6 are you now using, and which was your
previous working version ?
 
  2.6.12.1.
 
  2.6.0, 2.6.9,  2.6.10 all showed the same thing.
 
 
  So, this isn't a regression within 2.6.

No, it not. It appears to be a regression 2.4 -- 2.6 :-)

 
  Next bit of spare time I'll stick the framebuffer back in. I have
  memories that 1. The radeon card responds strangely to the radeon
  module.  2. I have downgraded the monitor to a cheap/nasty type
  which may not do what the previous one did.
 
  Perhaps try CONFIG_FB_SOFT_CURSOR and CONFIG_FB_MODE_HELPERS with
  CONFIG_FB_RADEON and CONFIG_FB_RADEON_I2C and make sure any other
  framebuffers are not set.  Apart from some junk in the early visual
  output up to maybe 2.6.7, and a missing penguin around 2.6.9 or
  2.6.10, the radeon has worked well here as a console framebuffer
  during 2.6.
 
  In xorg.conf, I use
   Driver  radeon
 (not ati, not fb)
 
  You may find that the bootargs I quoted for the radeon are wrong -
  another box of mine uses video=radeonfb:[EMAIL PROTECTED] - they both seem
  to give me a 128x48 framebuffer console, at least when I don't use a
  graphical login.
 
 

I can fart around with arguments and options for the video. I'm not
scared of that. I did get fed up trying to master this curcor thing, so
I stuck in the ALFS stuff, wiped my 3.3 libs, logs includes, and
executables, and had chapter 6  7 of LFS-6.0 build themselves last
evening. The thing knew best, of course. It stuck in UDEV(!) after me
loudly say6ing That will never work - ever. Of course it decided I
wanted HOTPLUG (Aaaargh!) and I just have hung on the bootscripts for
some reason.

That will give me an LFS-6.0, and then I have the BLFS-6.0 profile there
for another bit of work. At least I will prove if the problem is system
specific.

Mind you, I have to send this  reboot into a 2.6 kernel to build. I got
an error from the toolchain (Glibc IIRC)
FATAL: kernel too old when I was running 2.4.22

-- 

With best Regards,


Declan Moriarty.
-- 
http://linuxfromscratch.org/mailman/listinfo/blfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page


2.6 Cursor problem (was init acting the maggot):-(

2005-06-30 Thread Declan Moriarty
Recently, Somebody Somewhere wrote these words
 On Wed, Jun 29, 2005 at 06:43:49PM +0100, Declan Moriarty wrote:
  
  Should I choose a 2.6 kernel, the loadkeys script doesn't start, and
  I get this curious messing with no cursor.
 

This is infuriating


I rebuilt the 2.4 and 2.6 kernels on LFS-5.0 here. Only a 1% change,
mainly in usb because the modules were barfing over hotplug.

Spamd is now starting under the 2.4 kernel and all seems well there.
Loadkeys is running again under the 2.6 kernel. Neither have framebuffer
enabled for the console. The console cursor problem remains. Further, If
I change from a console into X and back out, the cursor vanishes :-((.

At any stage, running 'setfont same font' restores the cursor action.
No dri enabled either, btw.


I think the init  issue was the poor (idiotic) naming convention of
kernels and maps in the common /boot partition. I now have
vmlinuz-version and System.map-version which was definitely not what I
had before.  


-- 

With best Regards,


Declan Moriarty.
-- 
http://linuxfromscratch.org/mailman/listinfo/blfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page