From: Chunkeey@web.de
To: Joerg Albert <jal2@gmx.de>, "John W. Linville" <linville@tuxdriver.com>
Cc: Johannes Berg <johannes@sipsolutions.net>,
"linux-wireless@vger.kernel.org" <linux-wireless@vger.kernel.org>
Subject: Re: [PATCH] ar9170: added phy register initialisation from eeprom values
Date: Mon, 31 Aug 2009 11:53:37 +0200 [thread overview]
Message-ID: <1789900589@web.de> (raw)
> "Joerg Albert" <jal2@gmx.de> wrote:
>
> This patch adds the initialisation of some PHY registers
> from the modal_header[] values in the EEPROM (see otus/hal/hpmain.c, line 333 ff.)
>
> Signed-off-by: Joerg Albert <jal2@gmx.de>
meh, no hardware... and only a shitty kiosk with web-interface here.
testing has to wait until the weekend.
> diff --git a/drivers/net/wireless/ath/ar9170/phy.c b/drivers/net/wireless/ath/ar9170/phy.c
> index cb8b5cd..47a5e5c 100644
> --- a/drivers/net/wireless/ath/ar9170/phy.c
> +++ b/drivers/net/wireless/ath/ar9170/phy.c
> @@ -396,6 +396,131 @@ static struct ar9170_phy_init ar5416_phy_init[] = {
> { 0x1c9384, 0xf3307ff0, 0xf3307ff0, 0xf3307ff0, 0xf3307ff0, }
> };
>
> +/* look up a certain register in ar5416_phy_init[] and return the init. value
> + for the band and bandwidth given. Return 0 if register address not found. */
See Documentation/CodingStyle - Chapter 8
The preferred style for long (multi-line) comments is:
/*
* look up a certain register in ar5416_phy_init[] and return the init. value
* for the band and bandwidth given. Return 0 if register address not found.
*/
> +u32 ar9170_get_default_phy_reg_val(int reg, bool is_2ghz, bool is_40mhz)
please consider this instead:
static u32 ar9170_get_default_phy_reg_val(u32 reg, [...])
(static function, and reg is a u32)
> +{
> + struct ar9170_phy_init *p;
> + struct ar9170_phy_init *endp =
> + ar5416_phy_init+ARRAY_SIZE(ar5416_phy_init);
(see Codingstyle: 3.1: Spaces
=> Use one space around (on each side of) most binary [...] operators)
> +
> + for (p = ar5416_phy_init; p < endp; p++)
a extra { } wouldn't hurt. The following statements also has multiple lines.
> + if (p->reg == reg) {
> + if (is_2ghz)
> + return is_40mhz ? p->_2ghz_40 : p->_2ghz_20;
> + else
> + return is_40mhz ? p->_5ghz_40 : p->_5ghz_20;
> + }
> + return 0;
> +}
hmm, it's a bit odd to use pointers over a fixed array,
what about (unfortunately space and line damaged... and untested):
{
unsigned int i;
for (i = 0; i < ARRAY_SIZE(ar5416_phy_init); i++) {
if (ar5416_phy_init[i].reg != reg)
continue;
if (is_2ghz) {
if (is_40mhz)
return ar5416_phy_init[i]._2ghz_40;
else
return ar5416_phy_init[i]._2ghz_20;
} else {
if (is_40mhz)
return ar5416_phy_init[i]._5ghz_40;
else
return ar5416_phy_init[i]._5ghz_20;
}
}
return 0;
}
this follows the looks of the rest of the code. (e.g ar9170_init_phy)
(of course, either version should be fine.
so stay with yours if you have doubts.)
> +/* initialize some phy regs from eeprom values in modal_header[]
> + acc. to band and bandwith */
(multi-line comment, but I guess you know what do here now...)
> +static int ar9170_init_phy_from_eeprom(struct ar9170 *ar,
> + bool is_2ghz, bool is_40mhz)
> +{
> + const u8 xpd2pd[16] = {
> + 0x2, 0x2, 0x2, 0x1, 0x2, 0x2, 0x6, 0x2,
> + 0x2, 0x3, 0x7, 0x2, 0xB, 0x2, 0x2, 0x2
> + };
static const u8 xpd2pd ?
> + u32 defval, newval; /* two separate for debugging the changes */
> + /* pointer to the modal_header acc. to band */
> + struct ar9170_eeprom_modal *m = ar->eeprom.modal_header +
> + (is_2ghz ? 1 : 0);
what about:
struct ar9170_eeprom_modal *m = &ar->eeprom.modal_header[is_2ghz]; ?
> + ar9170_regwrite_begin(ar);
> +
> + /* ant common control (index 0) */
> + defval = ar9170_get_default_phy_reg_val(0x1c5964, is_2ghz, is_40mhz);
> + newval = le32_to_cpu(m->antCtrlCommon);
> + ar9170_regwrite(0x1c5964, newval);
> +
> + /* ant control chain 0 (index 1) */
> + defval = ar9170_get_default_phy_reg_val(0x1c5960, is_2ghz, is_40mhz);
> + newval = le32_to_cpu(m->antCtrlChain[0]);
> + ar9170_regwrite(0x1c5960, newval);
> +
> + /* ant control chain 2 (index 2) */
> + defval = ar9170_get_default_phy_reg_val(0x1c7960, is_2ghz, is_40mhz);
> + newval = le32_to_cpu(m->antCtrlChain[1]);
> + ar9170_regwrite(0x1c7960, newval);
> +
> + /* SwSettle (index 3) */
> + if (!is_40mhz) {
> + defval = ar9170_get_default_phy_reg_val(0x1c5844,
> + is_2ghz, is_40mhz);
> + newval = (defval & ~0x3f80) | ((m->switchSettling & 0x7f)<<7);
(well )<<7 is a bit tight, but it looks like you ran out of space in this line?)
> + ar9170_regwrite(0x1c5844, newval);
> + }
> +
> + /* adcDesired, pdaDesired (index 4) */
> + defval = ar9170_get_default_phy_reg_val(0x1c5850, is_2ghz, is_40mhz);
> + newval = (defval & ~0xffff) | ((u8)m->pgaDesiredSize << 8) |
> + ((u8)m->adcDesiredSize);
> + ar9170_regwrite(0x1c5850, newval);
> +
> + /* TxEndToXpaOff, TxFrameToXpaOn (index 5) */
> + defval = ar9170_get_default_phy_reg_val(0x1c5834, is_2ghz, is_40mhz);
> + newval = (m->txEndToXpaOff << 24) | (m->txEndToXpaOff << 16) |
> + (m->txFrameToXpaOn << 8) | m->txFrameToXpaOn;
> + ar9170_regwrite(0x1c5834, newval);
> +
> + /* TxEndToRxOn (index 6) */
> + defval = ar9170_get_default_phy_reg_val(0x1c5828, is_2ghz, is_40mhz);
> + newval = (defval & ~0xff0000) | (m->txEndToRxOn << 16);
> + ar9170_regwrite(0x1c5828, newval);
> +
> + /* thresh62 (index 7) */
> + defval = ar9170_get_default_phy_reg_val(0x1c8864, is_2ghz, is_40mhz);
> + newval = (defval & ~0x7f000) | (m->thresh62 << 12);
> + ar9170_regwrite(0x1c8864, newval);
> +
> + /* tx/rx attenuation chain 0 (index 8) */
> + defval = ar9170_get_default_phy_reg_val(0x1c5848, is_2ghz, is_40mhz);
> + newval = (defval & ~0x3f000) | ((m->txRxAttenCh[0] & 0x3f) << 12);
> + ar9170_regwrite(0x1c5848, newval);
> +
> + /* tx/rx attenuation chain 2 (index 9) */
> + defval = ar9170_get_default_phy_reg_val(0x1c7848, is_2ghz, is_40mhz);
> + newval = (defval & ~0x3f000) | ((m->txRxAttenCh[1] & 0x3f) << 12);
> + ar9170_regwrite(0x1c7848, newval);
> +
> + /* tx/rx margin chain 0 (index 10) */
> + defval = ar9170_get_default_phy_reg_val(0x1c620c, is_2ghz, is_40mhz);
> + newval = (defval & ~0xfc0000) | ((m->rxTxMarginCh[0] & 0x3f) << 18);
> + /* bsw margin chain 0 for 5GHz only */
> + if (!is_2ghz)
> + newval = (newval & ~0x3c00) | ((m->bswMargin[0] & 0xf) << 10);
> + ar9170_regwrite(0x1c620c, newval);
> +
> + /* tx/rx margin chain 2 (index 11) */
> + defval = ar9170_get_default_phy_reg_val(0x1c820c, is_2ghz, is_40mhz);
> + newval = (defval & ~0xfc0000) | ((m->rxTxMarginCh[1] & 0x3f) << 18);
> + ar9170_regwrite(0x1c820c, newval);
> +
> + /* iqCall, iqCallq chain 0 (index 12) */
> + defval = ar9170_get_default_phy_reg_val(0x1c5920, is_2ghz, is_40mhz);
> + newval = (defval & ~0x7ff) | (((u8)m->iqCalICh[0] & 0x3f) << 5) |
> + ((u8)m->iqCalQCh[0] & 0x1f);
> + ar9170_regwrite(0x1c5920, newval);
> +
> + /* iqCall, iqCallq chain 2 (index 13) */
> + defval = ar9170_get_default_phy_reg_val(0x1c7920, is_2ghz, is_40mhz);
> + newval = (defval & ~0x7ff) | (((u8)m->iqCalICh[1] & 0x3f) << 5) |
> + ((u8)m->iqCalQCh[1] & 0x1f);
> + ar9170_regwrite(0x1c7920, newval);
> +
> + /* xpd gain mask (index 14) */
> + defval = ar9170_get_default_phy_reg_val(0x1c6258, is_2ghz, is_40mhz);
> + newval = (defval & ~0xf0000) | (xpd2pd[m->xpdGain & 0xf] << 16);
> + ar9170_regwrite(0x1c6258, newval);
(cannot test this now. it looks good, though.)
It's amazing how much **** you can _cut_ from the vendor driver.
BTW: does this patch help the 1-stage fw stability, or is it still broken?
Regards,
Chr
________________________________________________________________
Neu: WEB.DE Doppel-FLAT mit Internet-Flatrate + Telefon-Flatrate
für nur 19,99 Euro/mtl.!* http://produkte.web.de/go/02/
next reply other threads:[~2009-08-31 9:53 UTC|newest]
Thread overview: 7+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-08-31 9:53 Chunkeey [this message]
2009-08-31 19:34 ` [PATCH] ar9170: added phy register initialisation from eeprom values Joerg Albert
2009-08-31 21:37 ` Christian Lamparter
2009-09-02 23:02 ` [PATCH v2] " Joerg Albert
2009-09-03 18:56 ` Christian Lamparter
[not found] ` <4A9EF532.8070500@gmx.de>
2009-09-03 16:16 ` [PATCH] " Christian Lamparter
-- strict thread matches above, loose matches on Subject: below --
2009-08-30 23:15 Joerg Albert
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=1789900589@web.de \
--to=chunkeey@web.de \
--cc=jal2@gmx.de \
--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).