HEADS-UP: starting to commit linuxolator (SoC 2006) changes...

Alexander Leidinger Alexander at Leidinger.net
Tue Aug 15 16:05:22 UTC 2006


Quoting Suleiman Souhlal <ssouhlal at FreeBSD.org> (Tue, 15 Aug 2006 17:29:33 +0200):

> Alexander Leidinger wrote:
> > Quoting Suleiman Souhlal <ssouhlal at FreeBSD.org> (Tue, 15 Aug 2006 14:53:56 +0200):

> >>Alexander Leidinger wrote:

> >>>So this gives us:
> >>> - coverity reports for the code *before the end of the SoC*
> >>
> >>Why the rush? I'm sure Roman will be around even when the SoC is over.
> > 
> > 
> > Yes, but first he will take a vacation, and then he will not be
> > available as much as he is now. And maybe Coverity will detect a bug
> > which we search currently but can't find (don't know how likely this
> > is, but it is at least possible).
> 
> Then maybe we need to change our Coverity setup so that it's able to 
> look at trees in perforce as well.

Do we have the processing power to do X more branches?

> >>Also, I'm not asking you to wait a month, just a couple of days until 
> >>more people have had a chance to look at the changes.
> > 
> > 
> > I thought about this before deciding to commit it. Most of the code was
> > available in P4. Those people with real interest already had a chance
> > to look at it, or they didn't had time to do so. If they didn't had
> > time to have a look at it, who knows if they have time to do it now?
> 
> Not really. I haven't looked at the changes in p4 not because I'm not 
> interested, but because they were a work in progress. I'd rather look at 
> "commit canditates".

There where reviews of p4 changes. So it's not completely unreviewed.

> > There's always someone who says "oh... if you waited 2 days longer...".
> > 
> > That's not the only reason. Tomorrow my holiday is over and I have to
> > catch up at work. Today I have plenty of time to commit urgent issues
> > (e.g., tinderbox breakage). And at the upcomming WE I may not have
> > enough time to do what I do today.
> 
> If you don't have time to commit something you can always ask someone 
> else to do it for you.

People tend to have enough on their plate already. And there where some
parts in the entire patchset which I didn't committed because of a
reason. So I wanted to handle this myself.

> > Additionally, the work is a little bit stalled currently, we need more
> > testers of the new code. There are a lot of people out there which want
> > to test, but don't know how to patch or fear to do something wrong when
> > they patch. With the approach I've taken, they just need to run a
> > sysctl (compat.linux.osrelease=2.6.16) and they can test. All which
> > don't want to test just don't need to care about it.
> 
> You could have released a sys/ snapshot tarball?

And then people do a cvsup or something else which breaks it and they
request help... I'm around on at least current@ since 3-current and if
people just need to turn a switch to test something, everything is fine.

> >>> - no change in behavior in the default case, since the new calls
> >>>   aren't used by glibs as long as... see following entry
> >>> - easy testing possibility (sysctl compat.linux.osrelease=2.6.16,
> >>>   defaults to 2.4.2)
> >>> - more eyes on the code
> >>
> >>Those are not valid reasons to commit unreviewed and potentially wrong
> >>changes.
> > 
> > 
> > Uhm... how many percent of code committed to current is reviewed before
> > it is commited?
> 
> Not enough.
> 
> > How many percent of the code I committed is not
> > reviewed and how much is this related to the amount of unreviewed code
> > committed in a busy week?
> 
> I don't want to sound harsh or like I'm doubting Roman's ability, but 
> most of the unreviewed code that gets committed comes from people who 
> have more experience than Roman and who usually "own" the part of the 
> kernel they are touching.

This doesn't help when the code in question isn't owned like the
linuxolator. Because most of the new code will not affect the rest of
the system (only linux programs, if at all), I don't think committing
this now was a big mistake.

Bye,
Alexander.

-- 
   And the list goes on, here are some more that I have recently added.
Sometimes I think Calvin and Hobbes have more wisdom than I.  Scarry...
http://www.Leidinger.net  Alexander @ Leidinger.net: PGP ID = B0063FE7
http://www.FreeBSD.org     netchild @ FreeBSD.org  : PGP ID = 72077137


More information about the freebsd-current mailing list