reiser4 went into -mm

reiser4 went into -mm

[Hans Reiser]

Andrew Morton was very helpful in the merge, and gave useful advise 
about our changes to the core kernel.

Now things get political.  If you guys on this list want to see reiser4 
get into mainline, you'll need to watch for the appropriate moments and 
push for it to happen, because reiser4 is fast, and that means 
threatening to many, and the knives are going to come out.

Of course, we still need to get to where we have been in -mm and there 
have been no valid bug reports for a week before I think we can ask to 
go into the official kernel.....

Hans

Re: reiser4 went into -mm

[Sander]

Hans Reiser wrote (ao):
> Andrew Morton was very helpful in the merge, and gave useful advise
> about our changes to the core kernel.

Very nice :-)

Can you give an eta on quota support for Reiser4?

Re: reiser4 went into -mm

[Hans Reiser]

Sander wrote:

>Hans Reiser wrote (ao):
>  
>
>>Andrew Morton was very helpful in the merge, and gave useful advise
>>about our changes to the core kernel.
>>    
>>
>
>Very nice :-)
>
>Can you give an eta on quota support for Reiser4?
>
>
>  
>
Probably not soon, we are busy earning money and such for the next month 
or two.

Hasn

Re: reiser4 went into -mm

[mjt]

On Thu, Aug 19, 2004 at 11:50:35PM -0700, Hans Reiser wrote:

>Of course, we still need to get to where we have been in -mm and there 
>have been no valid bug reports for a week before I think we can ask to 
>go into the official kernel.....

What does this mean for the future of Reiser4 development?

Still going to have auto-snapshots against the -mm kernels where
Reiser4 is merged?

I hope so, but I'd also tend to use the vanilla -mm trees unless/until
there is a bug I stumble upon and only then look to the auto-snapshots
for a solution. Is this a good method?

Seems to me that's the most realistic method now, as the new users will
just stick with -mm :)

-- 
mjt

Re: reiser4 went into -mm

[Hans Reiser]

Markus Törnqvist wrote:

>On Thu, Aug 19, 2004 at 11:50:35PM -0700, Hans Reiser wrote:
>
>  
>
>>Of course, we still need to get to where we have been in -mm and there 
>>have been no valid bug reports for a week before I think we can ask to 
>>go into the official kernel.....
>>    
>>
>
>What does this mean for the future of Reiser4 development?
>
>Still going to have auto-snapshots against the -mm kernels where
>Reiser4 is merged?
>
>I hope so, but I'd also tend to use the vanilla -mm trees unless/until
>there is a bug I stumble upon and only then look to the auto-snapshots
>for a solution. Is this a good method?
>
>Seems to me that's the most realistic method now, as the new users will
>just stick with -mm :)
>
>  
>
Yes, use -mm, we will send our fixes into -mm.

Hans

Re: reiser4 went into -mm

[David Dabbs]

>Andrew Morton was very helpful in the merge, and gave useful advise 
>about our changes to the core kernel.
>
>Now things get political.  If you guys on this list want to see reiser4 
>get into mainline, you'll need to watch for the appropriate moments and 
>push for it to happen, because reiser4 is fast, and that means 
>threatening to many, and the knives are going to come out.
>
>Of course, we still need to get to where we have been in -mm and there 
>have been no valid bug reports for a week before I think we can ask to 
>go into the official kernel.....
>
>Hans

Hans, the notes akpm put into the 2.6.8.1-mm2 ChangeLog include the following:

If you see results that are much different from those at
www.namesys.com/benchmarks.html, let us know.

but the latest benchmarks available there are from March against 2.6.5-rc2.
Does this mean that neither R4 nor other filesystems' performance has changed since then or do you expect potential users to run their own updated benchmarks?

David

Re: reiser4 went into -mm

[Hans Reiser]

David Dabbs wrote:

>>Andrew Morton was very helpful in the merge, and gave useful advise 
>>about our changes to the core kernel.
>>
>>Now things get political.  If you guys on this list want to see reiser4 
>>get into mainline, you'll need to watch for the appropriate moments and 
>>push for it to happen, because reiser4 is fast, and that means 
>>threatening to many, and the knives are going to come out.
>>
>>Of course, we still need to get to where we have been in -mm and there 
>>have been no valid bug reports for a week before I think we can ask to 
>>go into the official kernel.....
>>
>>Hans
>>    
>>
>
>Hans, the notes akpm put into the 2.6.8.1-mm2 ChangeLog include the following:
>
>If you see results that are much different from those at
>www.namesys.com/benchmarks.html, let us know.
>
>but the latest benchmarks available there are from March against 2.6.5-rc2.
>Does this mean that neither R4 nor other filesystems' performance has changed since then or do you expect potential users to run their own updated benchmarks?
>
>David
>
>
>
>  
>
If they have changed, I want to know about it.....;-)

David, we should work on getting your key comparison patch evaluated.  
Zam, please look into it when you can, and send me your todo list 
because I fear it is very full.

Re: reiser4 went into -mm

[David Dabbs]

>
>Hans, the notes akpm put into the 2.6.8.1-mm2 ChangeLog include the following:
>
>If you see results that are much different from those at
>www.namesys.com/benchmarks.html, let us know.
>
>but the latest benchmarks available there are from March against 2.6.5-rc2.
>Does this mean that neither R4 nor other filesystems' performance has changed since then or do you expect potential users to run their own updated benchmarks?
>
>David
>
>
>If they have changed, I want to know about it.....;-)
>
>David, we should work on getting your key comparison patch evaluated.  
>Zam, please look into it when you can, and send me your todo list 
>because I fear it is very full.
>

If I were at home and not at my in-laws' place in western Mass. I would be finalizing this. What has held it up on my end is the intermittent inability to 
finish a mongo run without some sort of problem. This is why I was inquiring about mongo's mounting/unmounting. If it is simply that my hardware is dodgy, then I'll need to find another platform on whichg to test. 

When I return home at the end of the month I'll base the patch on the 2.6.8.1-mmX, test it and forward a copy.

David