From: "Luis R. Rodriguez" <mcgrof@gmail.com>
To: Randy Dunlap <rdunlap@xenotime.net>
Cc: John Linville <linville@tuxdriver.com>,
linux-wireless@vger.kernel.org, Jiri Slaby <jirislaby@gmail.com>,
Nick Kossifidis <mickflemm@gmail.com>
Subject: Re: [PATCH 7/7] ath5k: Add documentation for struct ath5k_rate
Date: Thu, 1 Nov 2007 23:11:31 -0400 [thread overview]
Message-ID: <20071102031131.GA4070@pogo> (raw)
In-Reply-To: <20071101150651.a61a6f0f.rdunlap@xenotime.net>
On Thu, Nov 01, 2007 at 03:06:51PM -0700, Randy Dunlap wrote:
> On Thu, 1 Nov 2007 17:56:40 -0400 Luis R. Rodriguez wrote:
>
> > ** Resending after checkpatch.pl
> >
> > This adds documentation for struct ath5k_rate. This also removes
> > some unused variables, lp_ack_duration and sp_ack_duration which
> > are simply unnecessary. We obviously have information about the rest
> > of the rate values, we can add more as we go, this just starts this up.
> > I'll next target cleaning up the RATE macros, think that may be there
> > the other G mode issues are in.
>
> Thanks for doing this... but one more change, please:
> (just move lines around, see below)
>
> > Changes to ath5k.h
> > Changes-licensed-under: ISC
> >
> > Signed-off-by: Luis R. Rodriguez <mcgrof@gmail.com>
> > ---
> > drivers/net/wireless/ath5k/ath5k.h | 58 +++++++++++++++++++++++++++++++----
> > 1 files changed, 51 insertions(+), 7 deletions(-)
> >
> > diff --git a/drivers/net/wireless/ath5k/ath5k.h b/drivers/net/wireless/ath5k/ath5k.h
> > index c8ab09a..7147fb4 100644
> > --- a/drivers/net/wireless/ath5k/ath5k.h
> > +++ b/drivers/net/wireless/ath5k/ath5k.h
> > @@ -549,17 +549,61 @@ struct ath5k_athchan_2ghz {
> > * used by the rate control algorytm on MadWiFi.
> > */
> >
> > -#define AR5K_MAX_RATES 32 /*max number of rates on the rate table*/
> > +/* Max number of rates on the rate table and what it seems
> > + * Atheros hardware supports */
> > +#define AR5K_MAX_RATES 32
>
> As described in Documentation/kernel-doc-nano-HOWTO.txt, the
> structure fields/members should immediately follow the first line.
> After all of the fields/members, you can add general info, like
> the first paragraph that is here.
>
> Example from that doc file:
>
>
> /**
> * struct my_struct - short description
> * @a: first member
> * @b: second member
> *
> * Longer description
> */
New patch, completes the docs and addresses Randy's points.
This adds documentation for struct ath5k_rate. This also removes
some unused variables, lp_ack_duration and sp_ack_duration which
are simply unnecessary.
We're now just missing information about rate hw values:
3-5
16
17-23
28-32
If anyone knows what those are, please let us know.
Changes to ath5k.h
Changes-licensed-under: ISC
Signed-off-by: Luis R. Rodriguez <mcgrof@gmail.com>
---
drivers/net/wireless/ath5k/ath5k.h | 58 +++++++++++++++++++++++++++++++----
1 files changed, 51 insertions(+), 7 deletions(-)
diff --git a/drivers/net/wireless/ath5k/ath5k.h b/drivers/net/wireless/ath5k/ath5k.h
index c8ab09a..72c24bc 100644
--- a/drivers/net/wireless/ath5k/ath5k.h
+++ b/drivers/net/wireless/ath5k/ath5k.h
@@ -549,17 +549,60 @@ struct ath5k_athchan_2ghz {
* used by the rate control algorytm on MadWiFi.
*/
-#define AR5K_MAX_RATES 32 /*max number of rates on the rate table*/
+/* Max number of rates on the rate table and what it seems
+ * Atheros hardware supports */
+#define AR5K_MAX_RATES 32
+/**
+ * struct ath5k_rate - rate structure
+ * @valid: is this a valid rate for the current mode
+ * @modulation: respective mac80211 modulation
+ * @rate_kbps: rate in kbit/s
+ * @rate_code: hardware rate value, used in &struct ath5k_desc, on RX on
+ * &struct ath5k_rx_status.rs_rate and on TX on
+ * &struct ath5k_tx_status.ts_rate. Seems the ar5xxx harware supports
+ * up to 32 rates, indexed by 1-32. This means we really only need
+ * 6 bits for the rate_code.
+ * @dot11_rate: respective IEEE-802.11 rate value
+ * @control_rate: index of rate assumed to be used to send control frames.
+ * This can be used to set override the value on the rate duration
+ * registers. This is only useful if we can override in the harware at
+ * what rate we want to send control frames at. Note that IEEE-802.11
+ * Ch. 9.6 (after IEEE 802.11g changes) defines the rate at which we
+ * should send ACK/CTS, if we change this value we can be breaking
+ * the spec.
+ *
+ * This structure is used to get the RX rate or set the TX rate on the
+ * hardware descriptors. It is also used for internal modulation control
+ * and settings.
+ *
+ * On RX after the &struct ath5k_desc is parsed by the appropriate
+ * ah_proc_rx_desc() the respective hardware rate value is set in
+ * &struct ath5k_rx_status.rs_rate. On TX the desired rate is set in
+ * &struct ath5k_tx_status.ts_rate which is later used to setup the
+ * &struct ath5k_desc correctly. This is the hardware rate map we are
+ * aware of:
+ *
+ * rate_code 1 2 3 4 5 6 7 8
+ * rate_kbps 3000 1000 ? ? ? 2000 500 48000
+ *
+ * rate_code 9 10 11 12 13 14 15 16
+ * rate_kbps 24000 12000 6000 54000 36000 18000 9000 ?
+ *
+ * rate_code 17 18 19 20 21 22 23 24
+ * rate_kbps ? ? ? ? ? ? ? 11000
+ *
+ * rate_code 25 26 27 28 29 30 31 32
+ * rate_kbps 5500 2000 1000 ? ? ? ? ?
+ *
+ */
struct ath5k_rate {
- u8 valid; /* Valid for rate control */
+ u8 valid;
u32 modulation;
- u16 rate_kbps; /* Rate in kbps used in computetxtime */
- u8 rate_code; /* Rate mapping for h/w descriptors */
+ u16 rate_kbps;
+ u8 rate_code;
u8 dot11_rate;
- u8 control_rate; /* Rate for management frames -not used */
- u16 lp_ack_duration;/* long preamble ACK duration -not used */
- u16 sp_ack_duration;/* short preamble ACK duration -not used */
+ u8 control_rate;
};
/* XXX: GRR all this stuff to get leds blinking ??? (check out setcurmode) */
next prev parent reply other threads:[~2007-11-02 3:11 UTC|newest]
Thread overview: 23+ messages / expand[flat|nested] mbox.gz Atom feed top
2007-11-01 4:35 [PATCH 2/7] ath5k: Remove opaque pointers from ath5k_hw_attach() Luis R. Rodriguez
2007-11-01 4:37 ` [PATCH 3/7] ath5k: Clean up ath5k rate duration settings Luis R. Rodriguez
2007-11-01 4:38 ` [PATCH 4/7] ath5k: Fix clock on OFDM timing computation Luis R. Rodriguez
2007-11-01 4:38 ` [PATCH 5/7] ath5k: Move OFDM timings into a helper routine Luis R. Rodriguez
2007-11-01 4:39 ` [PATCH 6/7] ath5k: Clear up settings of AR5K_RSSI_THR register settings Luis R. Rodriguez
2007-11-01 4:40 ` [PATCH 7/7] ath5k: Add documentation for struct ath5k_rate Luis R. Rodriguez
2007-11-01 13:16 ` Nick Kossifidis
2007-11-01 16:47 ` Luis R. Rodriguez
2007-11-01 21:56 ` Luis R. Rodriguez
2007-11-01 22:06 ` Randy Dunlap
2007-11-02 3:11 ` Luis R. Rodriguez [this message]
2007-11-01 13:27 ` [PATCH 6/7] ath5k: Clear up settings of AR5K_RSSI_THR register settings Nick Kossifidis
2007-11-01 21:55 ` Luis R. Rodriguez
2007-11-01 21:53 ` [PATCH 5/7] ath5k: Move OFDM timings into a helper routine Luis R. Rodriguez
2007-11-01 21:57 ` Jiri Slaby
2007-11-01 22:01 ` Luis R. Rodriguez
2007-11-01 22:02 ` Jiri Slaby
2007-11-01 22:21 ` Jiri Slaby
2007-11-01 22:27 ` Nick Kossifidis
2007-11-01 21:50 ` [PATCH 3/7] ath5k: Clean up ath5k rate duration settings Luis R. Rodriguez
2007-11-01 13:06 ` [PATCH 2/7] ath5k: Remove opaque pointers from ath5k_hw_attach() Nick Kossifidis
2007-11-01 17:27 ` Christoph Hellwig
2007-11-01 18:44 ` Luis R. Rodriguez
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=20071102031131.GA4070@pogo \
--to=mcgrof@gmail.com \
--cc=jirislaby@gmail.com \
--cc=linux-wireless@vger.kernel.org \
--cc=linville@tuxdriver.com \
--cc=mickflemm@gmail.com \
--cc=rdunlap@xenotime.net \
/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).