linux-wireless.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: ian <ischram@telenet.be>
To: "John W. Linville" <linville@tuxdriver.com>
Cc: Johannes Berg <johannes@sipsolutions.net>,
	wireless <linux-wireless@vger.kernel.org>
Subject: Re: letting drivers choose their preferred rate scale
Date: Mon, 03 Sep 2007 02:21:18 +0200	[thread overview]
Message-ID: <46DB537E.8030801@telenet.be> (raw)
In-Reply-To: <20070831180933.GB3352@tuxdriver.com>



John W. Linville wrote:
> On Mon, Aug 27, 2007 at 12:49:09PM +0200, Johannes Berg wrote:
>> On Sat, 2007-08-25 at 14:15 +0200, ian wrote:
>>
>>> I reread the discussion carefully and can't find any specific
>>> comments on the patch which for clarity I will reproduce below.
>> Then I guess I remember wrong and the comment was raised in another
>> thread. The problem we have is that once any rate control algorithm is
>> loaded, it will be default for any future drivers, which is often
>> inappropriate. Hence, it would be great if
>> hw->preferred_rate_control==NULL would indicate "use default" where
>> "default" would be set to "simple" to start with but also changeable.
>> (although the change functionality could just be not usable from
>> userspace for now)
> 
> I ACK this approach.  Anyone have a patch hacked-up to post?
> 

I read up on the code and kernel mechanisms involved and this is
what i came up with

Right now there is no way of knowing what the "default" rc should be
other than that it has a module alias rc80211_default, checking
for this is probably not done, and would be pretty unpractical.
And it wouldn't be settable from userspace in the future either.

We could however replace "simple" with a user settable variable.
(global variable in ieee80211.c like the rate_ctrl_algs and mutex,
with some accessors for a module parameter/debugfs/..?)

we could additionally remove the module_alias from rc80211_simple

I don't know too much of the rest of mac80211 code and conventions,
so feel free to be brutal if this patch makes no sense.
---
Make rc80211_simple the default rate scaling algorithm

Signed-of-by: Ian Schram <ischram@telenet.be>
--- a/net/mac80211/ieee80211_rate.c	2007-09-03 01:26:18.000000000 +0200
+++ b/net/mac80211/ieee80211_rate.c	2007-09-03 01:46:54.000000000 +0200
@@ -62,7 +62,7 @@ ieee80211_try_rate_control_ops_get(const

 	mutex_lock(&rate_ctrl_mutex);
 	list_for_each_entry(alg, &rate_ctrl_algs, list) {
-		if (!name || !strcmp(alg->ops->name, name))
+		if (!strcmp(alg->ops->name, name))
 			if (try_module_get(alg->ops->module)) {
 				ops = alg->ops;
 				break;
@@ -78,11 +78,12 @@ static struct rate_control_ops *
 ieee80211_rate_control_ops_get(const char *name)
 {
 	struct rate_control_ops *ops;
+	const char *try_name = name ? name : "simple";

-	ops = ieee80211_try_rate_control_ops_get(name);
+	ops = ieee80211_try_rate_control_ops_get(try_name);
 	if (!ops) {
-		request_module("rc80211_%s", name ? name : "default");
-		ops = ieee80211_try_rate_control_ops_get(name);
+		request_module("rc80211_%s", try_name);
+		ops = ieee80211_try_rate_control_ops_get(try_name);
 	}
 	return ops;
 }



  reply	other threads:[~2007-09-03  0:22 UTC|newest]

Thread overview: 13+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2007-08-25  2:37 letting drivers choose their preferred rate scale ian
2007-08-25  8:02 ` Johannes Berg
2007-08-25 12:15   ` ian
2007-08-25 14:19     ` Jochen Voss
2007-08-27 10:49     ` Johannes Berg
2007-08-31 18:09       ` John W. Linville
2007-09-03  0:21         ` ian [this message]
2007-09-03  8:20           ` Johannes Berg
2007-09-03  8:38           ` Johannes Berg
2007-09-03 14:06             ` ian
2007-09-13 18:39               ` John W. Linville
2007-09-13 21:42                 ` Ian Schram
2007-09-14 12:08                   ` Johannes Berg

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=46DB537E.8030801@telenet.be \
    --to=ischram@telenet.be \
    --cc=johannes@sipsolutions.net \
    --cc=linux-wireless@vger.kernel.org \
    --cc=linville@tuxdriver.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).