[PATCH] Block device bringup using Xenstore

[PATCH] Block device bringup using Xenstore

[Rusty Russell]

[-- Attachment #1: Type: text/plain, Size: 455 bytes --]

This megapatch updates the xenstore daemon, updates the xenbus code, and
hacks the block devices to use the store rather than control messages
for initialization.

We just got this to bring a block device up; tomorrow we'll be cleaning
it up, getting destroy, suspend and resume working.

You'll need Anthony's experimental store-enabled vm-tools: he's sending
email about it now.

Rusty.
-- 
A bad analogy is like a leaky screwdriver -- Richard Braakman

[-- Attachment #2: diff.bz2 --]
[-- Type: application/x-bzip, Size: 19781 bytes --]

[-- Attachment #3: Type: text/plain, Size: 138 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

Re: [PATCH] Block device bringup using Xenstore

[Mike Wray]

Rusty Russell wrote:
> This megapatch updates the xenstore daemon, updates the xenbus code, and
> hacks the block devices to use the store rather than control messages
> for initialization.
> 
> We just got this to bring a block device up; tomorrow we'll be cleaning
> it up, getting destroy, suspend and resume working.
> 
> You'll need Anthony's experimental store-enabled vm-tools: he's sending
> email about it now.
> 
> Rusty.


Block device and network interface setup has been working in the control
tools CVS for several days, and is fully integrated with the xend tools.
It would be more helpful if you would integrate with that code instead
of going off on your own.

Mike

Re: [PATCH] Block device bringup using Xenstore

[Rusty Russell]

On Thu, 2005-07-07 at 11:45 +0100, Mike Wray wrote:
> Rusty Russell wrote:
> > This megapatch updates the xenstore daemon, updates the xenbus code, and
> > hacks the block devices to use the store rather than control messages
> > for initialization.

> Block device and network interface setup has been working in the control
> tools CVS for several days, and is fully integrated with the xend tools.
> It would be more helpful if you would integrate with that code instead
> of going off on your own.

No.  I completely disagree with what you've done.  I disagree with how
you use the store, both here and in most other places.  Your changes
made the xenbus code more complicated to suit a usage model which is a
retrograde step.

Rather than criticise in the abstract, I believe that the way to solve
these kind of disputes is to present a concrete implementation of one's
idea, so people can compare.  I don't expect you to implement my vision
of how things should be, but it's completely incompatible with your
work, and I've finally had the opportunity to implement it.

I will continue to develop this as my conscience dictates.  You should
do the same.  As always, I aim to create code which is convincing in
itself.

Rusty.
-- 
A bad analogy is like a leaky screwdriver -- Richard Braakman

Re: Re: [PATCH] Block device bringup using Xenstore

[Mike Wray]

Rusty Russell wrote:
> On Thu, 2005-07-07 at 11:45 +0100, Mike Wray wrote:
> 
>>Rusty Russell wrote:
>>
>>>This megapatch updates the xenstore daemon, updates the xenbus code, and
>>>hacks the block devices to use the store rather than control messages
>>>for initialization.
> 
> 
>>Block device and network interface setup has been working in the control
>>tools CVS for several days, and is fully integrated with the xend tools.
>>It would be more helpful if you would integrate with that code instead
>>of going off on your own.
> 
> 
> No.  I completely disagree with what you've done.  I disagree with how
> you use the store, both here and in most other places.  Your changes
> made the xenbus code more complicated to suit a usage model which is a
> retrograde step.

What can I say? I was under the impression we were all working together on
modifying xend and the drivers to use xenstore. Going off and doing your
own thing hardly helps. If you had issues with the current code perhaps
it might have been better to have raised them? Up to now you have not voiced
any concerns. It's disappointing you seem to be choosing confrontation rather
than collaboration.

> Rather than criticise in the abstract, I believe that the way to solve
> these kind of disputes is to present a concrete implementation of one's
> idea, so people can compare.  I don't expect you to implement my vision
> of how things should be, but it's completely incompatible with your
> work, and I've finally had the opportunity to implement it.
> 
> I will continue to develop this as my conscience dictates.  You should
> do the same.  As always, I aim to create code which is convincing in
> itself.
> 
> Rusty.

Regards,

Mike

Re: [PATCH] Block device bringup using Xenstore

[Christian Limpach]

On 7/7/05, Rusty Russell <rusty@rustcorp.com.au> wrote:
> This megapatch updates the xenstore daemon, updates the xenbus code, and
> hacks the block devices to use the store rather than control messages
> for initialization.

I've applied the xenstore daemon and xenbus driver changes. 
Unfortunately, the blkback/xenbus.c file is missing from the block
device changes.

     christian