* Re: 2.6.25 kernel & compat-wireless-2009-09-14
2009-09-17 0:56 2.6.25 kernel & compat-wireless-2009-09-14 Nikolai ZHUBR
@ 2009-09-16 23:54 ` Luis R. Rodriguez
2009-09-17 3:01 ` Re[2]: " Nikolai ZHUBR
0 siblings, 1 reply; 6+ messages in thread
From: Luis R. Rodriguez @ 2009-09-16 23:54 UTC (permalink / raw)
To: Nikolai ZHUBR; +Cc: linux-wireless
On Wed, Sep 16, 2009 at 5:56 PM, Nikolai ZHUBR <zhubr@mail.ru> wrote:
> Hello all,
>
> After some more digging I think I've found an incompatibility.
>
> device_add in 2.6.25.20 wants some bus_id, but bus_id seems to not
> be assigned anymore, so device_add fails. Therefore, wiphy_register
> fails, and then clearly ieee80211_register_hw fails too.
>
> Any ideas how to properly fix this?
FIgure out what the bus_id is used for first.
Luis
^ permalink raw reply [flat|nested] 6+ messages in thread
* 2.6.25 kernel & compat-wireless-2009-09-14
@ 2009-09-17 0:56 Nikolai ZHUBR
2009-09-16 23:54 ` Luis R. Rodriguez
0 siblings, 1 reply; 6+ messages in thread
From: Nikolai ZHUBR @ 2009-09-17 0:56 UTC (permalink / raw)
To: linux-wireless
Hello all,
After some more digging I think I've found an incompatibility.
device_add in 2.6.25.20 wants some bus_id, but bus_id seems to not
be assigned anymore, so device_add fails. Therefore, wiphy_register
fails, and then clearly ieee80211_register_hw fails too.
Any ideas how to properly fix this?
Thank you,
Nikolai
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: Re[2]: 2.6.25 kernel & compat-wireless-2009-09-14
2009-09-17 3:01 ` Re[2]: " Nikolai ZHUBR
@ 2009-09-17 2:04 ` Johannes Berg
2009-09-17 3:38 ` Re[4]: " Nikolai ZHUBR
0 siblings, 1 reply; 6+ messages in thread
From: Johannes Berg @ 2009-09-17 2:04 UTC (permalink / raw)
To: Nikolai ZHUBR; +Cc: Luis R. Rodriguez, linux-wireless
[-- Attachment #1: Type: text/plain, Size: 980 bytes --]
On Thu, 2009-09-17 at 06:01 +0300, Nikolai ZHUBR wrote:
> Thursday, September 17, 2009, 2:54:28 AM, Luis R. Rodriguez wrote:
> >> device_add in 2.6.25.20 wants some bus_id, but bus_id seems to not
> >> be assigned anymore, so device_add fails. Therefore, wiphy_register
> >> fails, and then clearly ieee80211_register_hw fails too.
> >>
> >> Any ideas how to properly fix this?
>
> > FIgure out what the bus_id is used for first.
>
> Hmm, don't know, but the following helps and wlan0 appears:
>
> --- everything.orig/net/wireless/core.c 2009-09-16 23:45:40.000000000 +0400
> +++ everything/net/wireless/core.c 2009-09-16 23:48:22.000000000 +0400
> @@ -350,6 +350,7 @@
>
> /* give it a proper name */
> dev_set_name(&rdev->wiphy.dev, PHY_NAME "%d", rdev->wiphy_idx);
> + snprintf(rdev->wiphy.dev.bus_id, BUS_ID_SIZE, PHY_NAME "%d", rdev->wiphy_idx);
Isn't that exactly what dev_set_name() is/was supposed to do?
johannes
[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 801 bytes --]
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re[2]: 2.6.25 kernel & compat-wireless-2009-09-14
2009-09-16 23:54 ` Luis R. Rodriguez
@ 2009-09-17 3:01 ` Nikolai ZHUBR
2009-09-17 2:04 ` Johannes Berg
0 siblings, 1 reply; 6+ messages in thread
From: Nikolai ZHUBR @ 2009-09-17 3:01 UTC (permalink / raw)
To: Luis R. Rodriguez; +Cc: linux-wireless
Thursday, September 17, 2009, 2:54:28 AM, Luis R. Rodriguez wrote:
>> device_add in 2.6.25.20 wants some bus_id, but bus_id seems to not
>> be assigned anymore, so device_add fails. Therefore, wiphy_register
>> fails, and then clearly ieee80211_register_hw fails too.
>>
>> Any ideas how to properly fix this?
> FIgure out what the bus_id is used for first.
Hmm, don't know, but the following helps and wlan0 appears:
--- everything.orig/net/wireless/core.c 2009-09-16 23:45:40.000000000 +0400
+++ everything/net/wireless/core.c 2009-09-16 23:48:22.000000000 +0400
@@ -350,6 +350,7 @@
/* give it a proper name */
dev_set_name(&rdev->wiphy.dev, PHY_NAME "%d", rdev->wiphy_idx);
+ snprintf(rdev->wiphy.dev.bus_id, BUS_ID_SIZE, PHY_NAME "%d", rdev->wiphy_idx);
mutex_init(&rdev->mtx);
mutex_init(&rdev->devlist_mtx);
> Luis
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re[4]: 2.6.25 kernel & compat-wireless-2009-09-14
2009-09-17 2:04 ` Johannes Berg
@ 2009-09-17 3:38 ` Nikolai ZHUBR
0 siblings, 0 replies; 6+ messages in thread
From: Nikolai ZHUBR @ 2009-09-17 3:38 UTC (permalink / raw)
To: Johannes Berg; +Cc: Luis R. Rodriguez, linux-wireless
Thursday, September 17, 2009, 5:04:48 AM, Johannes Berg wrote:
>>
>> --- everything.orig/net/wireless/core.c 2009-09-16 23:45:40.000000000 +0400
>> +++ everything/net/wireless/core.c 2009-09-16 23:48:22.000000000 +0400
>> @@ -350,6 +350,7 @@
>>
>> /* give it a proper name */
>> dev_set_name(&rdev->wiphy.dev, PHY_NAME "%d", rdev->wiphy_idx);
>> + snprintf(rdev->wiphy.dev.bus_id, BUS_ID_SIZE, PHY_NAME "%d", rdev->wiphy_idx);
> Isn't that exactly what dev_set_name() is/was supposed to do?
Well, probably yes, but still it doesn't set bus_id, I've checked.
Probably it needs some correction.
> johannes
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re[2]: 2.6.25 kernel & compat-wireless-2009-09-14
2009-09-17 12:57 ` Luis R. Rodriguez
@ 2009-09-17 16:29 ` Nikolai ZHUBR
0 siblings, 0 replies; 6+ messages in thread
From: Nikolai ZHUBR @ 2009-09-17 16:29 UTC (permalink / raw)
To: Luis R. Rodriguez; +Cc: linux-wireless, Senthil Balasubramanian, Gaurav Jauhar
Thursday, September 17, 2009, 3:57:10 PM, Luis R. Rodriguez wrote:
> On Thu, Sep 17, 2009 at 6:22 AM, Nikolai ZHUBR <zhubr@mail.ru> wrote:
>> Hello people,
>>
>> So finally I got it mostly working and usable as AP:
>>
>> Hardware:
>> * ASUS WL-500gP (mips) with stock wireless card removed (broadcom).
>> * New shining Ralink RT2600 Mini-PCI card
>> Software:
>> * kernel 2.6.25.20 (openwrt 8.09.1)
>> * compat-wireless-2009-09-14 (without rfkill_backport and with ugly bus_id fix)
>> libnl-1.1
>> hostapd 0.6.9
>>
>> The led is now even blinking nicely.
>> Strange though that selecting static IP in the client didn't work,
>> but using DHCP works fine (it's ok for me now).
> That's great! Care to send some patches for this?
Well, it appears the only thing that really needs fixing is dev_set_name()
compatability function, but fixing it properly without making things worse
is a bit beyond my capability at this point. I think someone more familiar
with the code should have a look (a hint is below).
Regarding rfkill_backport, the issue is probably openwrt-specific, becase
there seems to be no way to load modules with circular dependencies without
modprobe (as far as I understand it). Commenting out CONFIG_RFKILL_BACKPORT
in config.mk "fixes" this and I think this issue can not be easily solved
inside compat-wireless package.
Nikolai
--- everything.orig/net/wireless/core.c 2009-09-16 23:45:40.000000000 +0400
+++ everything/net/wireless/core.c 2009-09-16 23:48:22.000000000 +0400
@@ -350,6 +350,7 @@
/* give it a proper name */
dev_set_name(&rdev->wiphy.dev, PHY_NAME "%d", rdev->wiphy_idx);
+ snprintf(rdev->wiphy.dev.bus_id, BUS_ID_SIZE, PHY_NAME "%d", rdev->wiphy_idx);
mutex_init(&rdev->mtx);
mutex_init(&rdev->devlist_mtx);
> Luis
^ permalink raw reply [flat|nested] 6+ messages in thread
end of thread, other threads:[~2009-09-17 15:22 UTC | newest]
Thread overview: 6+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2009-09-17 0:56 2.6.25 kernel & compat-wireless-2009-09-14 Nikolai ZHUBR
2009-09-16 23:54 ` Luis R. Rodriguez
2009-09-17 3:01 ` Re[2]: " Nikolai ZHUBR
2009-09-17 2:04 ` Johannes Berg
2009-09-17 3:38 ` Re[4]: " Nikolai ZHUBR
-- strict thread matches above, loose matches on Subject: below --
2009-09-17 13:22 Nikolai ZHUBR
2009-09-17 12:57 ` Luis R. Rodriguez
2009-09-17 16:29 ` Re[2]: " Nikolai ZHUBR
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.