[PATCH] linux 2.6.x.x net/sched/sch_api.c -- more comment reviews

[PATCH] linux 2.6.x.x net/sched/sch_api.c -- more comment reviews

[S P]

OK,
I used 'diff -Nru' this time.
Sorry about last time.  Ignorance~~~  :)

Tense in the doc goes back and forth, making it a
little difficult to understand.

I hope I understood the documentation w/ small changes
I'm submitting.

- S.P.


diff -Nru ../linus-2.6/net/sched/sch_api.c sch_api.c
--- ../linus-2.6/net/sched/sch_api.c    2004-09-02
14:47:49.000000000 -0700
+++ sch_api.c   2004-09-13 14:53:30.726090760 -0700
@@ -78,7 +78,7 @@
    checks and part of work, which is common to all
qdiscs
    and to provide rtnetlink notifications.

-   All real intelligent work is done inside qdisc
modules.
+   All real intelligent work is done inside each
qdisc modules.



@@ -87,29 +87,29 @@
    ---dequeue

    dequeue usually returns a skb to send. It is
allowed to return NULL,
-   but it does not mean that queue is empty, it just
means that
-   discipline does not want to send anything this
time.
+   but it does not mean the queue is empty; it means
that
+   the discipline does not want to send anything this
time.
    Queue is really empty if q->q.qlen == 0.
-   For complicated disciplines with multiple queues
q->q is not
-   real packet queue, but however q->q.qlen must be
valid.
+   For complicated disciplines with multiple queues,
q->q is not
+   real packet queue whereas q->q.qlen must be valid.

    ---enqueue

-   enqueue returns 0, if packet was enqueued
successfully.
-   If packet (this one or another one) was dropped,
it returns
-   not zero error code.
+   enqueue returns 0, if packet enqueues
successfully.
+   If packet (this one or another one) is dropped, it
returns
+   non-zero error code.
    NET_XMIT_DROP       - this packet dropped
-     Expected action: do not backoff, but wait until
queue will clear.
-   NET_XMIT_CN         - probably this packet
enqueued, but another one dropped.
-     Expected action: backoff or ignore
-   NET_XMIT_POLICED    - dropped by police.
-     Expected action: backoff or error to real-time
apps.
+     Expected action: do not back-off, but wait until
queue clears.
+   NET_XMIT_CN         - probably this packet is
enqueued, but another packet isdropped.
+     Expected action: back-off or ignore
+   NET_XMIT_POLICED    - packet is dropped by police.
+     Expected action: back-off or error to real-time
apps.

    Auxiliary routines:

    ---requeue

-   requeues once dequeued packet. It is used for
non-standard or
+   requeues once packet is dequeued. It is used for
non-standard or
    just buggy devices, which can defer output even if
dev->tbusy=0.

    ---reset



	
		
__________________________________
Do you Yahoo!?
New and Improved Yahoo! Mail - 100MB free storage!
http://promotions.yahoo.com/new_mail 

Re: [PATCH] linux 2.6.x.x net/sched/sch_api.c -- more comment reviews

[David S. Miller]

On Mon, 13 Sep 2004 14:59:44 -0700 (PDT)
S P <speattle@yahoo.com> wrote:

Your email client linewraps the patches and also transforms
tabs into spaces making your patches un-applyable.  Please fix
for future submissions, thanks.

Now, onto the patch itself.  I think we're adding more tense
errors than we're removing.  For example:

> -   All real intelligent work is done inside qdisc
> modules.
> +   All real intelligent work is done inside each
> qdisc modules.

'each' indicates singularity, yes "modules" is still
plural.  I would change it instead to:

	All the real intelligent work is done inside
	the qdisc modules.

> -   but it does not mean that queue is empty, it just
> means that
> -   discipline does not want to send anything this
> time.
> +   but it does not mean the queue is empty; it means
> that
> +   the discipline does not want to send anything this
> time.

This one looks fine.

> -   For complicated disciplines with multiple queues
> q->q is not
> -   real packet queue, but however q->q.qlen must be
> valid.
> +   For complicated disciplines with multiple queues,
> q->q is not
> +   real packet queue whereas q->q.qlen must be valid.

Looks fine, we're missing an articles here.
So maybe the final version of this verse is:

	For complicated disciplines with multiple queues,
	q->q is not a real packet queue whereas q->q.qlen
	must be valid.

The rest looks fine.

Re: [PATCH] linux 2.6.x.x net/sched/sch_api.c -- more comment reviews

[S P]

 
> Please fix
> for future submissions, thanks.
will use attachments.

> 
> Now, onto the patch itself.  I think we're adding
> more tense
> errors than we're removing.  For example:
> 
> > -   All real intelligent work is done inside qdisc
> > modules.
> > +   All real intelligent work is done inside each
> > qdisc modules.
> 
> 'each' indicates singularity, yes "modules" is still
> plural.  I would change it instead to:
> 
> 	All the real intelligent work is done inside
> 	the qdisc modules.

I think I meant inside each of the module.  Not sure
if your change conveys that.  I did notice it's still
akward, but I didn't want to take away the emphasis of
individuality if the original author meant it.



> Looks fine, we're missing an articles here.
> So maybe the final version of this verse is:
> 
> 	For complicated disciplines with multiple queues,
> 	q->q is not a real packet queue whereas q->q.qlen
> 	must be valid.
> 
> The rest looks fine.
Yup, I thought I lost that 'a' somewhere.  And, the
comma.  (so the code readers don't run out of breathe)
> 
> 
Thanks.
So should I resubmit w/ the changes?
Wasn't sure if you just applied w/ the changes or are
returning for me to apply the fixes.

:)


		
__________________________________
Do you Yahoo!?
Yahoo! Mail - 50x more storage than other providers!
http://promotions.yahoo.com/new_mail