netdev
[Top] [All Lists]

Re: [Prism54] CVS -> bk tree update

To: "Luis R. Rodriguez" <mcgrof@xxxxxxxxxxxxxxxxxxxx>
Subject: Re: [Prism54] CVS -> bk tree update
From: Jeff Garzik <jgarzik@xxxxxxxxx>
Date: Tue, 15 Jun 2004 23:10:28 -0400
Cc: "Luis R. Rodriguez" <mcgrof@xxxxxxxxxxxxxxxxxxxx>, Netdev <netdev@xxxxxxxxxxx>, prism54-devel@xxxxxxxxxxx
In-reply-to: <20040616022008.GM6253@ruslug.rutgers.edu>
References: <20040614191651.GC6253@ruslug.rutgers.edu> <40CE2754.1020109@pobox.com> <20040616022008.GM6253@ruslug.rutgers.edu>
Sender: netdev-bounce@xxxxxxxxxxx
User-agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.6) Gecko/20040510
Luis R. Rodriguez wrote:
Thanks Jeff. I did, and a diff of Andrew's -mm tree Vs our cvs tree
2.6.7-prism54 gives a 1898 line diff. This is excluding spaces,
newlines, and tabbing and all of Andew's .orig, .rej's.


This time it's going to be a *real* pain to provide a changelog and split the
diffs.. so can I just send you that 1898 line diff and then another for
space changes? Everything has already been reviewed on the lists here...

I promise this will be our last big patch too. I'll ask our team
to send patches from now on to netdev after each ChangeLog entry, for
more public review and to not loose sync of our trees. Let us know if
you think we should do otherwise.


Sorry, no. You knew that split-up patches would be needed, once the initial driver merge is complete (which was complete before the most recent series of 17 patches).

This is one of the big downsides to developing in CVS, and it bites people again and again.

Linux kernel development relies on split-up patches for review, testing, and narrowing down which patch in a series introduced a bug. There are real, engineering-related reasons we do things this way.

        Jeff



<Prev in Thread] Current Thread [Next in Thread>