Re: Questions about patches for hurdselect.c

[Svante Signell]

On Fri, 2012-12-07 at 17:35 +0100, Samuel Thibault wrote:
> Svante Signell, le Fri 07 Dec 2012 17:13:00 +0100, a écrit :
> > In the works of rewriting hurdselect.c I have some question on how to
> > write patches. The changed code contains three parts:
> > 1) Introducing three cases: DELAY, POLL, and SELECT
> 
> Why three cases? Isn't there a lot of common code between poll and
> select?

Unfortunately the differences are too large to have them in the same
code path, without loosing the overview. Additionally, the pure delay
code is much faster than today's as reported in
http://lists.gnu.org/archive/html/bug-hurd/2012-11/msg00000.html

> > Doing the first part makes a diff with the original code not so useful,
> 
> But it does not change the semantic, right?
> 
> Your goal should be to make patches both readable and not introduce a
> regression. Splitting patches into 3 to make it more readable while
> keeping no regression when applying only one or two of them is
> completely fine.

OK, will do it in three steps then. The first diff will look strange,
but the two succeeding ones will not, as they will be patches of the
preceding ones.

> > Q3) Part 3) is not yet finished, Should I wait until it is complete?
> 
> No. Submit your work early, that'll save you work if you have to go back
> to change things.

My changes so far to case POLL will be in patch 3. There is still the
shutdown problem in tcp sockets to solve, to return the errors needed
for poll to handle. For pipes I have already proposed a patch, is it OK?