There was an error in the B5PHY init specifications.
Signed-off-by: Larry Finger <[email protected]>
---
John,
This patch is meant for wireless-2.6, 'upstream', and 'upstream-fixes'.
Larry
Index: wireless-2.6/drivers/net/wireless/bcm43xx/bcm43xx_phy.c
===================================================================
--- wireless-2.6.orig/drivers/net/wireless/bcm43xx/bcm43xx_phy.c
+++ wireless-2.6/drivers/net/wireless/bcm43xx/bcm43xx_phy.c
@@ -806,7 +806,7 @@ static void bcm43xx_phy_initb5(struct bc
/* Force to channel 7, even if not supported. */
bcm43xx_radio_selectchannel(bcm, 7, 0);
- if (radio->version != 0x2050) {
+ if (radio->version == 0x2050) {
bcm43xx_radio_write16(bcm, 0x0075, 0x0080);
bcm43xx_radio_write16(bcm, 0x0079, 0x0081);
}
Joe,
Joseph Jezak wrote:
> David Woodhouse wrote:
>> On Fri, 2007-03-02 at 11:30 -0600, Larry Finger wrote:
>>> There was an error in the B5PHY init specifications.
>> This patch doesn't fix the machine check in bcm43xx_phy_initb5 which
>> Pavel Roskin and I reported a couple of weeks ago. To get rid of that
>> crash, I still have to revert an earlier 'spec update' patch
>> (740ac4fb08866d702be90f167665d03759bd27d0).
>>
>
> Yeah, the issue is that the Rev1 cards have to be init'd with g mode
> off. I'm not sure anyone has picked up this fix yet. The relevant
> spec fix is here: http://bcm-v4.sipsolutions.net/802.11/PHY/calinit
I think the bug may be even deeper. In bcm43xx_phy_connect, the G Mode Enable bit of TM State Low is
only written for phy->rev >= 5. I have not found any other place where that bit is set. From what I
read in the specs, that bit should exist and be set for all G PHY's. Is that correct? I think I'm
beginning to understand the confusion between phy->connected in the code and gmode in the specs.
Larry
On Mon, Mar 12, 2007 at 04:16:34PM +0100, David Woodhouse wrote:
> On Mon, 2007-03-12 at 10:53 -0400, John W. Linville wrote:
> >
> > FWIW, by inspection it looks like the mac80211-based driver is
> > (trying?) to implement this change.
> >
> > David, have you tried the mac80211 version? Does it still have the
> > same crash?
>
> Should the one in 2.6.20-1.2982.fc7 be OK? I can try that relatively
> easily; anything else might need to wait till I get home.
Yes, I think that will do. It certainly looks to be up-to-date.
Thanks,
John
--
John W. Linville
[email protected]
John W. Linville wrote:
> On Mon, Mar 12, 2007 at 12:42:18PM -0500, Larry Finger wrote:
>> David Woodhouse wrote:
>>> On Mon, 2007-03-12 at 10:53 -0400, John W. Linville wrote:
>>>> FWIW, by inspection it looks like the mac80211-based driver is
>>>> (trying?) to implement this change.
>>>>
>>>> David, have you tried the mac80211 version? Does it still have the
>>>> same crash?
>>> Should the one in 2.6.20-1.2982.fc7 be OK? I can try that relatively
>>> easily; anything else might need to wait till I get home.
>>>
>> I don't think it will improve anything. I don't currently have a copy of wireless-dev, but the code
>> in the mb tree is identical with that of wireless-2.6. I'm currently trying to figure out the spec
>> at http://bcm-v4.sipsolutions.net/802.11/PHY/calinit and what we need to change.
>
> Huh? Are you looking in drivers/net/wireless/mac80211/bcm43xx?
> It certainly seems to differ quite a bit from what is in
> drivers/net/wireless/bcm43xx (i.e. no the softmac-based version).
I was, but the conclusion is the same. The generate_xxatt_list is not relevant and only the last if
clause is important. The second argument to the second call to wireless_core_reset is just a
different way to say "non-zero".
I've been looking at the code for a while now, and I still don't see where the code doesn't match
the spec. Either it is a subtle point, or I'm not even close to sharp today.
Larry
On Fri, 2007-03-02 at 11:30 -0600, Larry Finger wrote:
> There was an error in the B5PHY init specifications.
This patch doesn't fix the machine check in bcm43xx_phy_initb5 which
Pavel Roskin and I reported a couple of weeks ago. To get rid of that
crash, I still have to revert an earlier 'spec update' patch
(740ac4fb08866d702be90f167665d03759bd27d0).
--
dwmw2
David Woodhouse wrote:
> On Fri, 2007-03-02 at 11:30 -0600, Larry Finger wrote:
>> There was an error in the B5PHY init specifications.
>
> This patch doesn't fix the machine check in bcm43xx_phy_initb5 which
> Pavel Roskin and I reported a couple of weeks ago. To get rid of that
> crash, I still have to revert an earlier 'spec update' patch
> (740ac4fb08866d702be90f167665d03759bd27d0).
>
Yeah, the issue is that the Rev1 cards have to be init'd with g mode
off. I'm not sure anyone has picked up this fix yet. The relevant
spec fix is here: http://bcm-v4.sipsolutions.net/802.11/PHY/calinit
-Joe
On Mon, Mar 12, 2007 at 12:42:18PM -0500, Larry Finger wrote:
> David Woodhouse wrote:
> > On Mon, 2007-03-12 at 10:53 -0400, John W. Linville wrote:
> >> FWIW, by inspection it looks like the mac80211-based driver is
> >> (trying?) to implement this change.
> >>
> >> David, have you tried the mac80211 version? Does it still have the
> >> same crash?
> >
> > Should the one in 2.6.20-1.2982.fc7 be OK? I can try that relatively
> > easily; anything else might need to wait till I get home.
> >
>
> I don't think it will improve anything. I don't currently have a copy of wireless-dev, but the code
> in the mb tree is identical with that of wireless-2.6. I'm currently trying to figure out the spec
> at http://bcm-v4.sipsolutions.net/802.11/PHY/calinit and what we need to change.
Huh? Are you looking in drivers/net/wireless/mac80211/bcm43xx?
It certainly seems to differ quite a bit from what is in
drivers/net/wireless/bcm43xx (i.e. no the softmac-based version).
John
--
John W. Linville
[email protected]
On Mon, Mar 12, 2007 at 09:23:30AM -0400, Joseph Jezak wrote:
> David Woodhouse wrote:
> > On Fri, 2007-03-02 at 11:30 -0600, Larry Finger wrote:
> >> There was an error in the B5PHY init specifications.
> >
> > This patch doesn't fix the machine check in bcm43xx_phy_initb5 which
> > Pavel Roskin and I reported a couple of weeks ago. To get rid of that
> > crash, I still have to revert an earlier 'spec update' patch
> > (740ac4fb08866d702be90f167665d03759bd27d0).
> >
>
> Yeah, the issue is that the Rev1 cards have to be init'd with g mode
> off. I'm not sure anyone has picked up this fix yet. The relevant
> spec fix is here: http://bcm-v4.sipsolutions.net/802.11/PHY/calinit
FWIW, by inspection it looks like the mac80211-based driver is
(trying?) to implement this change.
David, have you tried the mac80211 version? Does it still have the
same crash?
John
--
John W. Linville
[email protected]
On Mon, 2007-03-12 at 10:53 -0400, John W. Linville wrote:
>
> FWIW, by inspection it looks like the mac80211-based driver is
> (trying?) to implement this change.
>
> David, have you tried the mac80211 version? Does it still have the
> same crash?
Should the one in 2.6.20-1.2982.fc7 be OK? I can try that relatively
easily; anything else might need to wait till I get home.
--
dwmw2
David Woodhouse wrote:
> On Mon, 2007-03-12 at 10:53 -0400, John W. Linville wrote:
>> FWIW, by inspection it looks like the mac80211-based driver is
>> (trying?) to implement this change.
>>
>> David, have you tried the mac80211 version? Does it still have the
>> same crash?
>
> Should the one in 2.6.20-1.2982.fc7 be OK? I can try that relatively
> easily; anything else might need to wait till I get home.
>
I don't think it will improve anything. I don't currently have a copy of wireless-dev, but the code
in the mb tree is identical with that of wireless-2.6. I'm currently trying to figure out the spec
at http://bcm-v4.sipsolutions.net/802.11/PHY/calinit and what we need to change.
Larry