[lm-sensors] describe the patch flow, review process

[lm-sensors] describe the patch flow, review process

[Jim Cromie]

rather than just ask,  Ill guess (hopefully educated ;-)
pls correct / elaborate as necessary.


1.  Jean Delvare is main reviewer of patches here.
    he'd like help.
    half/baked reviews (like the one I just did ;-) are still better 
than none.
    if you give bad advice, youll get corrected,
    you'll learn more, and it wont hurt.

    the dialog will help you, the reviewee, and others
       And I cant resist a quote (boy, arent I clever ;-)
        from http://plato-dialogues.org/plato.htm

       Plato wrote *dialogues* rather than philosophy treatises, and, 
what's more, dialogues in which he never stages himself as a 
participant, it is because his purpose was not to tell his readers what 
/he himself/ thought, what were the answers /he himself/ had given to 
the most fundamental questions in life about *what it means to be a 
(wo)man*, but to teach them *to think by themselves*

    (see, talk is good !  It just has to go somewhere, not over same old 
'talking-points')

    Jean wont mind the 'extra' work of reviewing reviews
       its an 'investment' in a more vibrant lm-sensors community.
       (he still dreams about just playing freecell or Duke-Nukem  ;-)

2.   When Jean approves he:
    a. adds it to his http://khali.linux-fr.org/devel/i2c/linux-2.6/
            with comments in series file.
    b. sends it to GregKH for further review and occaisional pushback.
        that doesnt happen much, Jean is thorough.
             Greg's happy about that.
             Jean would be too.  See 1.

3.   Greg then
    a. puts them here ??  (I dont see them)
        http://www.kernel.org/pub/linux/kernel/people/gregkh/
    b.  pushes them to AKPM for inclusion in -mm
          does he consolidate them to 1 patch ?
             I think not, that would lose the attributions


and this is where I get a bit fuzzy.

2.6.15-mm1 came out recently  
    http://lkml.org/lkml/2006/1/5/149

but none of the stuff in 2.a  is there   (I looked for mine).

Is that because theyre going to Linus  / main / git  instead ?
and if so (not going to -mm)  why ?
        dont they warrant more testing / exposure there ?

Im not second-guessing here (hope thats obvious, but..)
just trying to understand, and more tothe point,
to know where to look, so I can check myself.  
(optimistically thinking here ;-)


Theres a less elaborate version of this at
    http://lm-sensors.org/wiki/FAQ
Ill update at some point, but wont mind if you do..

tia
jimc

[lm-sensors] describe the patch flow, review process

[Greg KH]

On Fri, Jan 06, 2006 at 11:11:15AM -0700, Jim Cromie wrote:
> 3.   Greg then
>     a. puts them here ??  (I dont see them)
>         http://www.kernel.org/pub/linux/kernel/people/gregkh/

Close:
	kernel.org/pub/linuc/kernel/people/gregkh/gregkh-2.6/patches/
Note that there is a lot of non-sensor stuff in there so you probably
only care about:
	kernel.org/pub/linuc/kernel/people/gregkh/gregkh-2.6/patches/i2c/
but not always.

>     b.  pushes them to AKPM for inclusion in -mm

No, Andrew picks them up from:
	kernel.org/pub/linuc/kernel/people/gregkh/gregkh-2.6/gregkh-02-i2c/
which is the staging area for him.

>           does he consolidate them to 1 patch ?

No.

>              I think not, that would lose the attributions
> 
> 
> and this is where I get a bit fuzzy.

I then push them to Linus when it's ok to do so, by taking the patches,
creating a temporary git tree and asking Linus to pull it.  I also post
the patches that I'm asking Linus to pull to this list.

> 2.6.15-mm1 came out recently  
>     http://lkml.org/lkml/2006/1/5/149
> 
> but none of the stuff in 2.a  is there   (I looked for mine).

It takes a while to bubble up :)

> Is that because theyre going to Linus  / main / git  instead ?
> and if so (not going to -mm)  why ?
>         dont they warrant more testing / exposure there ?

If I've already sent the patches to Linus, and he's applied them, then
they get into -mm with the "latest linus patch".

Hope this helps,

greg k-h