linux-vr-cvsupdate

[linux-vr-cvsupdate] Update to module linux

[Bradley D. L. <br...@us...>]

Update of /cvsroot/linux-vr/linux/drivers/video
In directory slayer.i.sourceforge.net:/tmp/cvs-serv27626/drivers/video

Removed Files:
      Tag: linux-sgi
	vga_font.c 
Log Message:
Remove from linux-sgi branch.

[linux-vr-cvsupdate] Update to module linux

[Bradley D. L. <br...@us...>]

Update of /cvsroot/linux-vr/linux
In directory slayer.i.sourceforge.net:/tmp/cvs-serv28162

Modified Files:
	Makefile 
Log Message:
Merge with linux-sgi-2000-07-13.

[linux-vr-cvsupdate] Update to module linux

[Bradley D. L. <br...@us...>]

Update of /cvsroot/linux-vr/linux/arch/mips
In directory slayer.i.sourceforge.net:/tmp/cvs-serv28162/arch/mips

Modified Files:
	Makefile 
Log Message:
Merge with linux-sgi-2000-07-13.

[linux-vr-cvsupdate] Update to module linux

[Bradley D. L. <br...@us...>]

Update of /cvsroot/linux-vr/linux/drivers/video
In directory slayer.i.sourceforge.net:/tmp/cvs-serv28162/drivers/video

Modified Files:
	Makefile 
Log Message:
Merge with linux-sgi-2000-07-13.

[linux-vr-cvsupdate] Update to module linux

[Michael K. <mf...@us...>]

Update of /cvsroot/linux-vr/linux/include/asm-mips
In directory slayer.i.sourceforge.net:/tmp/cvs-serv6875

Modified Files:
	vr41xx-platdep.h 
Log Message:
More Aero 15xx platdep defines from Jorj Bauer.

[linux-vr-cvsupdate] Update to module linux

[Harald K. <ha...@us...>]

Update of /cvsroot/linux-vr/linux/arch/mips/r39xx
In directory slayer.i.sourceforge.net:/tmp/cvs-serv11038/linux/arch/mips/r39xx

Modified Files:
	time.c 
Log Message:
updates

[linux-vr-cvsupdate] Update to module linux

[Michael K. <mf...@us...>]

Update of /cvsroot/linux-vr/linux/arch/mips
In directory slayer.i.sourceforge.net:/tmp/cvs-serv22795/arch/mips

Modified Files:
	config.in 
Log Message:
Added support for suspend mode, in addition to hibernate.  Suspend is not
as deep a power savings mode, and doesn't cut power to most of the device
the way hibernate does, but at least it's usable on non-ROM booted devices.

Added a semaphore for DMA usage.  The sound driver is not yet using it,
though.

[linux-vr-cvsupdate] Update to module linux

[Michael K. <mf...@us...>]

Update of /cvsroot/linux-vr/linux/arch/mips/vr41xx
In directory slayer.i.sourceforge.net:/tmp/cvs-serv22795/arch/mips/vr41xx

Modified Files:
	kbd.c power.c setup.c time.c 
Log Message:
Added support for suspend mode, in addition to hibernate.  Suspend is not
as deep a power savings mode, and doesn't cut power to most of the device
the way hibernate does, but at least it's usable on non-ROM booted devices.

Added a semaphore for DMA usage.  The sound driver is not yet using it,
though.

[linux-vr-cvsupdate] Update to module linux

[Michael K. <mf...@us...>]

Update of /cvsroot/linux-vr/linux/include/asm-mips
In directory slayer.i.sourceforge.net:/tmp/cvs-serv12286/include/asm-mips

Modified Files:
	bootinfo.h 
Log Message:
Fix machine group numpering, so VR41xx doesn't get listed as Origin...
Add IBM Workpad to the device name list
Add spurious interrupt count to /proc/cpuinfo

Apparently, there's a spurious int happening on Workpad, looking into it...

[linux-vr-cvsupdate] Update to module linux

[Michael K. <mf...@us...>]

Update of /cvsroot/linux-vr/linux/arch/mips/vr41xx
In directory slayer.i.sourceforge.net:/tmp/cvs-serv12286/arch/mips/vr41xx

Modified Files:
	prom.c 
Log Message:
Fix machine group numpering, so VR41xx doesn't get listed as Origin...
Add IBM Workpad to the device name list
Add spurious interrupt count to /proc/cpuinfo

Apparently, there's a spurious int happening on Workpad, looking into it...

[linux-vr-cvsupdate] Update to module linux

[Michael K. <mf...@us...>]

Update of /cvsroot/linux-vr/linux/arch/mips/kernel
In directory slayer.i.sourceforge.net:/tmp/cvs-serv12286/arch/mips/kernel

Modified Files:
	proc.c 
Log Message:
Fix machine group numpering, so VR41xx doesn't get listed as Origin...
Add IBM Workpad to the device name list
Add spurious interrupt count to /proc/cpuinfo

Apparently, there's a spurious int happening on Workpad, looking into it...

[linux-vr-cvsupdate] Update to module linux

[Michael K. <mf...@us...>]

Update of /cvsroot/linux-vr/linux/drivers/char
In directory slayer.i.sourceforge.net:/tmp/cvs-serv17756/drivers/char

Modified Files:
	console.c 
Log Message:
Add a kernel command line parameter to override the default default color.
You can now use defcolor=0xf0 to make the kernel bootup messages display
black on white.

Also added 6 pixel fontwidth support to fbcon-cfb16

[linux-vr-cvsupdate] Update to module linux

[Michael K. <mf...@us...>]

Update of /cvsroot/linux-vr/linux/drivers/video
In directory slayer.i.sourceforge.net:/tmp/cvs-serv17756/drivers/video

Modified Files:
	fbcon-cfb16.c fonts.c 
Log Message:
Add a kernel command line parameter to override the default default color.
You can now use defcolor=0xf0 to make the kernel bootup messages display
black on white.

Also added 6 pixel fontwidth support to fbcon-cfb16

[linux-vr-cvsupdate] Update to module linux

[Michael K. <mf...@us...>]

Update of /cvsroot/linux-vr/linux/arch/mips/vr41xx
In directory slayer.i.sourceforge.net:/tmp/cvs-serv18284/arch/mips/vr41xx

Modified Files:
	irq.c kbd.c power.c serial.c time.c 
Log Message:
Mostly minor cleanups and slight enhancements:
Removed the cli/sti from around pm_send in pm_do_suspend, adjusted some of
the pm_request callbacks accordingly.
Added nested ird disable/enable support.
Mostly comment changes in time.c: yes, lost_ticks replacement was correct.

[linux-vr-cvsupdate] Update to module linux

[Bradley D. L. <br...@us...>]

Update of /cvsroot/linux-vr/linux/include/linux
In directory slayer.i.sourceforge.net:/tmp/cvs-serv10284/include/linux

Modified Files:
	fb.h 
Log Message:
RFC: Provide get operations for backlight and contrast settings.
Applications (e.g. a GUI contrast control) may need to determine
the current backlight and contrast settings for proper operation.

[linux-vr-cvsupdate] Update to module linux

[Michael K. <mf...@us...>]

Update of /cvsroot/linux-vr/linux/drivers/video
In directory slayer.i.sourceforge.net:/tmp/cvs-serv6511/drivers/video

Modified Files:
	vrc4171fb.c 
Log Message:
Add some diagnostic code to determine if a VRC4171(A) chip is really present.
Eventually, the driver just won't register the frame buffer if no ichip found.

[linux-vr-cvsupdate] Update to module linux

[Michael K. <mf...@us...>]

Update of /cvsroot/linux-vr/linux/arch/mips/kernel
In directory slayer.i.sourceforge.net:/tmp/cvs-serv8485/arch/mips/kernel

Modified Files:
	r2300_switch.S r4k_switch.S traps.c 
Log Message:
Last few bits of ll/sc emulation.  This is not a complete duplication of
the functionality of those instructions on CPUs that have them, but it
should suffice for any user program that is using them in a sane manner
to synchronize between 2 or more user threads.

This is completely untested at this point.

RE: [linux-vr-cvsupdate] Update to module linux

[Harald K. <Har...@ho...>]

On 19-Jul-00 Michael Klar wrote:
> Update of /cvsroot/linux-vr/linux/arch/mips/kernel
> In directory slayer.i.sourceforge.net:/tmp/cvs-serv8485/arch/mips/kernel
> 
> Modified Files:
>       r2300_switch.S r4k_switch.S traps.c 
> Log Message:
> Last few bits of ll/sc emulation.  This is not a complete duplication of
> the functionality of those instructions on CPUs that have them, but it
> should suffice for any user program that is using them in a sane manner
> to synchronize between 2 or more user threads.
> 
> This is completely untested at this point.

Thanks. I've hacked something similar which I wanted to test little bit. I ran a
few crashmes on one of my DECstations only to find out that there's something
seriously wrong with regard to R3000 processors. My R3000 DECstation freezes
shortly after starting crashme regardless of having the llsc emulation on or off.

The really bad news is that it does so as well with the SGI source tree
(2.4.0-test5-pre2). Looks like time to step back and review the R3000 code.

Just beeing curious, would a VR box survive crashme-2.2?

-- 
Regards,
Harald

RE: [linux-vr-cvsupdate] Update to module linux

[Mike K. <mf...@po...>]

> Just beeing curious, would a VR box survive crashme-2.2?

Where can I find this crashme I keep hearing about?  I've been wanting to
run it against Linux VR for a while now.

Mike Klar

RE: [linux-vr-cvsupdate] Update to module linux

[Harald K. <Har...@ho...>]

On 20-Jul-00 Mike Klar wrote:
>> Just beeing curious, would a VR box survive crashme-2.2?
> 
> Where can I find this crashme I keep hearing about?  I've been wanting to
> run it against Linux VR for a while now.

You can find it in several places. Try asking
http://ftpsearch.lycos.com/?form=medium and search for crashme-2.2

crashme-2.4 which seems to be the latest version doesn't compile with glibc-2.0.7
crashme-2.2 seems to crash fine ;-)

Further tests with 2.2 show:

R4400SC DECstation with 2.4.0-test5-pre2 freezes
R3000A DECstation with 2.4.0-test5-pre2 freezes
R3000A DECstation with linux-vr-2.4.0-test4-pre3 freezes
R3912 Sharp Mobilon with linux-vr-2.4.0-test4-pre3 freezes

There's something seriously going wrong here.

-- 
Regards,
Harald

[linux-vr-cvsupdate] Update to module linux

[Bradley D. L. <br...@us...>]

Update of /cvsroot/linux-vr/linux/drivers/video
In directory slayer.i.sourceforge.net:/tmp/cvs-serv15951/drivers/video

Modified Files:
	vr4181fb.c 
Log Message:
Adjust for new backlight/contrast ioctl defines.

[linux-vr-cvsupdate] Update to module linux

[Bradley D. L. <br...@us...>]

Update of /cvsroot/linux-vr/linux/arch/mips/kernel
In directory slayer.i.sourceforge.net:/tmp/cvs-serv16163/arch/mips/kernel

Modified Files:
	mips_ksyms.c 
Log Message:
Export isa_slot_offset - used by MTD.

[linux-vr-cvsupdate] Update to module linux

[Michael K. <mf...@us...>]

Update of /cvsroot/linux-vr/linux/arch/mips/kernel
In directory slayer.i.sourceforge.net:/tmp/cvs-serv3575

Modified Files:
	traps.c 
Log Message:
OK, ll/sc emulation _almost_ worked, SC opcode was wrong.  Tested now,
appears to work, verified that link gets broken on context switch.

[linux-vr-cvsupdate] Update to module linux

[Michael K. <mf...@us...>]

Update of /cvsroot/linux-vr/linux/arch/mips/r39xx
In directory slayer.i.sourceforge.net:/tmp/cvs-serv27182/arch/mips/r39xx

Modified Files:
	kbd.c kbd_no.c kbd_uart.c kbd_velo.c 
Log Message:
A bunch of console-related fixes and enhancements:
 - Added a blank keymap, made that the default keymap for devices without
   keyboards.  Was going to make it default keymap for all devices, but that
   would make it mandatory to run loadkeys, which is a 124K binary...
 - Fixed the scancode <-> keycode translation stuff to return something
   slightly more reasonable than error, except when user actually tries to
   change the translation (we don't need to translate, the keymap can
   address any scancode we choose to send it).
 - Moved the console pm_register call slightly, so it registers when the
   first VT is used, not only when a second is used...
 - Modified an assumption in vt.c that anything except for PPC or M68K is a
   PC.  loadkeys should work flawlessly now (at least it did in my
   testing).  Previously, it would ignore scancode 0 setting.
 - Fixed a slight problem with the new fontwidth support I added last week
   to cfb16
 - Added 2,4,6,10,12,14,16 pixel fontwidth support to cfb4
 - Cleaned up the VRC4171 chip id probing in the VRC4171 fb device

[linux-vr-cvsupdate] Update to module linux

[Michael K. <mf...@us...>]

Update of /cvsroot/linux-vr/linux/drivers/char
In directory slayer.i.sourceforge.net:/tmp/cvs-serv27182/drivers/char

Modified Files:
	Makefile console.c vt.c 
Added Files:
	blank_keymap.c 
Log Message:
A bunch of console-related fixes and enhancements:
 - Added a blank keymap, made that the default keymap for devices without
   keyboards.  Was going to make it default keymap for all devices, but that
   would make it mandatory to run loadkeys, which is a 124K binary...
 - Fixed the scancode <-> keycode translation stuff to return something
   slightly more reasonable than error, except when user actually tries to
   change the translation (we don't need to translate, the keymap can
   address any scancode we choose to send it).
 - Moved the console pm_register call slightly, so it registers when the
   first VT is used, not only when a second is used...
 - Modified an assumption in vt.c that anything except for PPC or M68K is a
   PC.  loadkeys should work flawlessly now (at least it did in my
   testing).  Previously, it would ignore scancode 0 setting.
 - Fixed a slight problem with the new fontwidth support I added last week
   to cfb16
 - Added 2,4,6,10,12,14,16 pixel fontwidth support to cfb4
 - Cleaned up the VRC4171 chip id probing in the VRC4171 fb device