[9fans] Re: Solaris thread scheaduling

[9fans] Re: Solaris thread scheaduling

[saroj]

In article <399D2F7F.AD34496D@compaq.com>,
  Dave Butenhof <David.Butenhof@compaq.com> wrote:

> Even Mac OS (X) will soon have POSIX kernel threads. Cooperative
threading is dying
> out from the world, finally.


If I remember correctly, the new Plan 9 operating system has
threads that do cooperative threading and you use lightweight
processes (rfork) to get time-slicing like regular processes.

- Saroj Mahapatra


Sent via Deja.com http://www.deja.com/
Before you buy.

Re: [9fans] Re: Solaris thread scheaduling

[kilgore]

> > Even Mac OS (X) will soon have POSIX kernel threads. Cooperative
> threading is dying
> > out from the world, finally.

i can hardly wait.

[9fans] Monitors and VGADB

[Mark C. Otto]

I have a Micron 700FGx 17" monitor, and I was trying to work through the
second part of the vgadb(6) to see if vga(8) will recognize my monitor. 
I found an entry in the database for a Micron 17FGx.  I cannot figure
out how to go from the information in the monitor manual to VGADB entry,
particularly how to get shb, ehb, ht, vrs, vre, vt, hsync, and vsync. 
Since the VGADB specifications are in terms of character clocks, this
seems to be from the generation of character terminals, which I cannot
find a description of.  I have been reading the PC Upgrade Bible to get
an idea of how the video cards and monitors work.  Ferraro, Richard E.,
Programming Guide to the EGA, VGA and Super VGA Cards is out of print.  

Could someone could give me a time period in the 9fans archives or an
another book or article that would explain how to go from a monitor
description like the one below to a VGADB monitor entry.  Russ has
talked about updating the graphics part of Plan9, but until then some
conversion formulas would help, especially when I do not have the system
installed to experiment with it. 

Thanks for any references or information someone can provide.
Mark Otto



Display			300x225mm of viewable area.  
Dot pitch		0.26mm (gives only one and does not specify horizontal or
vertical)
Horizontal frequency	30.0-69.0kHz
Vertical frequency	50.0-160.0Hz
Video band width	86MHz
Maximum resolution	1280x1024@60Hz
Preset modes:

					Frequency		Signal Polarity
Mode		Resolution	Horzontal	Vertical	Horz	Vertical
VGA 480@60Hz	640x480		31.47kHz	59.94Hz		-	-
640x480@75Hz	640x480		37.50kHz	75.00Hz		-	-
800x600@75Hz	800x600		46.88kHz	75.00Hz		+	+
832x624@75Hz	832x624		49.73kHz	74.55Hz		-	-
1024x768@70Hz	1024x768	56.48kHz	70.07Hz		-	-
1024x768@75Hz	1024x768	60.24kHz	74.93Hz		-	-
1024x768@75Hz	1024x768	60.02kHz	75.03Hz		+	+
1280x1024@60Hz	1280x1024	63.98kHz	60.02Hz		+	+

There is no mention of interlacing in the monitor manual.

Re: [9fans] Monitors and VGADB

[William Staniewicz]

Mark,

I just installed Plan9 on my pc yesterday. I have the same monitor as you do
and used the vgadb entry for the Micron 17FGx. This is probably not the correct
thing to keep running it but at least it got something up on the screen for me
to work with. I looked at my user's manual for the Micron 700FGx and noticed a
difference in one of the numbers you posted. I wasn't sure if it is a typo, but
what I have is indicated below. BTW, if you got any replies to your posting I would
be interested in how to modify my vgadb correctly too.

/Bill
wstanREMOVETHIS@planet.nl

On Fri, Aug 18, 2000 at 04:14:15PM +0000, Mark C. Otto wrote:
> I have a Micron 700FGx 17" monitor, and I was trying to work through the
> second part of the vgadb(6) to see if vga(8) will recognize my monitor. 
> I found an entry in the database for a Micron 17FGx.  I cannot figure
> out how to go from the information in the monitor manual to VGADB entry,
> Display			300x225mm of viewable area.  
> Dot pitch		0.26mm (gives only one and does not specify horizontal or
> vertical)
> Horizontal frequency	30.0-69.0kHz
> Vertical frequency	50.0-160.0Hz <--- I have 50.0-120.0
> Video band width	86MHz
> Maximum resolution	1280x1024@60Hz
> Preset modes:

Re: [9fans] Re: Solaris thread scheaduling

[dhog]

> > > Cooperative threads are just coroutines.  They're `cooperative'
> > > because if the thread doesn't cooperate by calling the scheduler,
> > > no other thread ever get scheduled.
> > 
> > It's a bit like, "cooperate -- or else!".
> 
> No, it's more like "those who do not cooperate inherit the cpu".

"The frustrated user will cut off their inheritance (yea, unto the 4th generation)".

Re: [9fans] Re: Solaris thread scheaduling

[Douglas A. Gwyn]

> > Cooperative threads are just coroutines.  They're `cooperative'
> > because if the thread doesn't cooperate by calling the scheduler,
> > no other thread ever get scheduled.
dhog@plan9.bell-labs.com wrote:
> It's a bit like, "cooperate -- or else!".

Well, if the tasks are decently coded, that can be a better
approach than preemptive scheduling where a task is interrupted
with things in an inconsistent state and all registers and status
flags have to be carefully preserved.

The operating systems in the Blit family of terminals used
cooperative scheduling.

Re: [9fans] Re: Solaris thread scheaduling

[Lucio De Re]

On Fri, Aug 18, 2000 at 04:53:56PM -0400, dhog@plan9.bell-labs.com wrote:
> 
> > Cooperative threads are just coroutines.  They're `cooperative'
> > because if the thread doesn't cooperate by calling the scheduler,
> > no other thread ever get scheduled.
> 
> It's a bit like, "cooperate -- or else!".

No, it's more like "those who do not cooperate inherit the cpu".

++L

Re: [9fans] Re: Solaris thread scheaduling

[Boyd Roberts]

From: <dhog@plan9.bell-labs.com>

> It's a bit like, "cooperate -- or else!".

yep dhog, you got it down.  score?

Re: [9fans] Re: Solaris thread scheaduling

[dhog]

> Cooperative threads are just coroutines.  They're `cooperative'
> because if the thread doesn't cooperate by calling the scheduler,
> no other thread ever get scheduled.

It's a bit like, "cooperate -- or else!".

Re: [9fans] Re: Solaris thread scheaduling

[Tom Duff]

On Aug 18, 11:34am, rob pike wrote:
> Subject: Re: [9fans] Re: Solaris thread scheaduling
> What, we should use uncooperative threads?
> Adversarial threads? Anarchic threads?
>
> I guess I don't know the terminology.  If POSIX threads
> are a good thing, perhaps I don't want to know what they're
> better than.

Cooperative threads are just coroutines.  They're `cooperative'
because if the thread doesn't cooperate by calling the scheduler,
no other thread ever get scheduled.

-- 
Tom Duff.  FUD in optima forma.

Re: [9fans] Re: Solaris thread scheaduling

[rob pike]

What, we should use uncooperative threads?
Adversarial threads? Anarchic threads?

I guess I don't know the terminology.  If POSIX threads
are a good thing, perhaps I don't want to know what they're
better than.

-rob