[9fans] requesting more information about the Gsoc project

[9fans] requesting more information about the Gsoc project

[Rajika Kumarasiri]

[-- Attachment #1: Type: text/plain, Size: 928 bytes --]

hi,
I am computer science and Engineering student who is intrested in excuting the project  "General purpose boot loader based on Plan 9 kernel and /dev/reboot "under your organization. I am very much intersting in operating systems and low level programming.So I am very much intresting doing this project.
 
I downlaod the source of Plan9 source and played  with it. I really need to know few things to write my project proposal. 
First thing why Plan9 need a new boot loader insted of the existing one.
Second what extra benifits will be achived using  the new boot loader.
I am very much greatful if anybody can update me on these information.
My project informations and personal information can be found here.
http://www.cse.mrt.ac.lk/~danrkuma/
Thanks in advace
Regards,
rajika
 

 
---------------------------------
Bored stiff? Loosen up...
Download and play hundreds of games for free on Yahoo! Games.

[-- Attachment #2: Type: text/html, Size: 1091 bytes --]

Re: [9fans] requesting more information about the Gsoc project

[erik quanstrom]

i don't know of any particular complaints with the functionality of 9load.

as i understand, there's one basic complaint against 9load --- it has
a seperate set of drivers.  russ outlines his ideas on 9load here
http://9fans.net/archive/2005/12/81.  (9load now is able to pxe
load the plan9.ini file.) 

i'm not most people, but i like having plan9.ini.  it elimates guesswork
about which kernel was started.  i wouldn't like a 9load that made 
plan9.ini optional and guessed which kernel to run and which fs to use.

also, there's a very good reason for having a seperate kernel and drivers
for 9load.  pxe booting all but one's core machines is definately the way
to fly, but pxe images are limited in size.  i'm not sure what the maximum
size is (the minimum guarenteed size is 64k), but i have not been able to
directly load modern plan 9 kernels on our hardware.

instead of making 9load as close to the plan 9 cpu kernel as possible,
i would use bios instead of drivers.  the bios already knows how to talk to
pxe-capable nics, and storage with boot rom.

this would require writing a real mode assembler.  (i would resist mightily
porting gas.  you'd spend more time porting libbinutils than writing real code.
;-))

- erik

On Wed Mar 21 05:17:30 EDT 2007, rajikacc@yahoo.com wrote:
> hi,
> I am computer science and Engineering student who is intrested in excuting the project  "General purpose boot loader based on Plan 9 kernel and /dev/reboot "under your organization. I am very much intersting in operating systems and low level programming.So I am very much intresting doing this project.
>  
> I downlaod the source of Plan9 source and played  with it. I really need to know few things to write my project proposal. 
> First thing why Plan9 need a new boot loader insted of the existing one.
> Second what extra benifits will be achived using  the new boot loader.
> I am very much greatful if anybody can update me on these information.
> My project informations and personal information can be found here.
> http://www.cse.mrt.ac.lk/~danrkuma/
> Thanks in advace
> Regards,
> rajika

Re: [9fans] requesting more information about the Gsoc project

[erik quanstrom]

i don't know of any particular complaints with the functionality of 9load.

as i understand, there's one basic complaint against 9load --- it has
a seperate set of drivers.  russ outlines his ideas on 9load here
http://9fans.net/archive/2005/12/81.  (9load now is able to pxe
load the plan9.ini file.) 

i'm not most people, but i like having plan9.ini.  it elimates guesswork
about which kernel was started.  i wouldn't like a 9load that made 
plan9.ini optional and guessed which kernel to run and which fs to use.

also, there's a very good reason for having a seperate kernel and drivers
for 9load.  pxe booting all but one's core machines is definately the way
to fly, but pxe images are limited in size.  i'm not sure what the maximum
size is (the minimum guarenteed size is 64k), but i have not been able to
directly load modern plan 9 kernels on our hardware.

instead of making 9load as close to the plan 9 cpu kernel as possible,
why not use bios instead of drivers.  the bios already knows how to talk to
pxe-capable nics, and storage with boot rom.

this would require writing a real mode assembler.  (i would resist mightily
porting gas.  you'd spend more time porting libbinutils than writing useful
code. ;-))

- erik

On Wed Mar 21 05:17:30 EDT 2007, rajikacc@yahoo.com wrote:
> hi,
> I am computer science and Engineering student who is intrested in excuting the project  "General purpose boot loader based on Plan 9 kernel and /dev/reboot "under your organization. I am very much intersting in operating systems and low level programming.So I am very much intresting doing this project.
>  
> I downlaod the source of Plan9 source and played  with it. I really need to know few things to write my project proposal. 
> First thing why Plan9 need a new boot loader insted of the existing one.
> Second what extra benifits will be achived using  the new boot loader.
> I am very much greatful if anybody can update me on these information.
> My project informations and personal information can be found here.
> http://www.cse.mrt.ac.lk/~danrkuma/
> Thanks in advace
> Regards,
> rajika

[9fans] Re: requesting more information about the Gsoc project

[Darren Bane]

erik quanstrom <quanstro@coraid.com> wrote:

!snip!

> this would require writing a real mode assembler.  (i would resist mightily
> porting gas.  you'd spend more time porting libbinutils than writing useful
> code. ;-))

Russ Cox seems to have at least the start of such a beast on sources, in
/contrib/rsc/86a
-- 
Darren Bane

Re: [9fans] requesting more information about the Gsoc project

[Christopher Nielsen]

On 3/21/07, erik quanstrom <quanstro@coraid.com> wrote:
> i don't know of any particular complaints with the functionality of 9load.

i don't think the complaints have anything to do with functionality. i
think it's more maintainability. you're maintaining drivers in two
separate places, which is tedious and error prone.

-- 
Christopher Nielsen
"They who can give up essential liberty for temporary
safety, deserve neither liberty nor safety." --Benjamin Franklin