prism54: revert "prism54: set carrier flags correctly"

prism54: revert "prism54: set carrier flags correctly"

[Patrick McHardy]

[-- Attachment #1: Type: text/plain, Size: 0 bytes --]



[-- Attachment #2: x --]
[-- Type: text/plain, Size: 2186 bytes --]

commit 47230a6beff4656a55bf23d92daee27f96c5b5a1
Author: Patrick McHardy <kaber@trash.net>
Date:   Mon May 12 18:25:51 2008 +0200

    prism54: revert "prism54: set carrier flags correctly"

    Commit 7b463ced6 (prism54: set carrier flags correctly) causes a
    regression: in AP-mode, the device comes up without a carrier,
    which (in my case) causes IPv6 to not add a link-local address,
    which in turn causes radvd (configured to announce routes on that
    device) to refuse to start, breaking IPv6 for the entire network.
    Additionally it most likely will be unable to pass packets because
    the queue in never started.

    A patch that seems to fix the issue is available, however the
    wireless developers have not merged it upstream for three weeks
    now and not responded to further email, so this patch reverts
    the faulty change.

    Signed-off-by: Patrick McHardy <kaber@trash.net>

diff --git a/drivers/net/wireless/prism54/isl_ioctl.c b/drivers/net/wireless/prism54/isl_ioctl.c
index 5b375b2..5d7951a 100644
--- a/drivers/net/wireless/prism54/isl_ioctl.c
+++ b/drivers/net/wireless/prism54/isl_ioctl.c
@@ -2081,7 +2081,6 @@ link_changed(struct net_device *ndev, u32 bitrate)
 	islpci_private *priv = netdev_priv(ndev);
 
 	if (bitrate) {
-		netif_carrier_on(ndev);
 		if (priv->iw_mode == IW_MODE_INFRA) {
 			union iwreq_data uwrq;
 			prism54_get_wap(ndev, NULL, (struct sockaddr *) &uwrq,
@@ -2090,10 +2089,8 @@ link_changed(struct net_device *ndev, u32 bitrate)
 		} else
 			send_simple_event(netdev_priv(ndev),
 					  "Link established");
-	} else {
-		netif_carrier_off(ndev);
+	} else
 		send_simple_event(netdev_priv(ndev), "Link lost");
-	}
 }
 
 /* Beacon/ProbeResp payload header */
diff --git a/drivers/net/wireless/prism54/islpci_dev.c b/drivers/net/wireless/prism54/islpci_dev.c
index 04c2638..eb7c1c6 100644
--- a/drivers/net/wireless/prism54/islpci_dev.c
+++ b/drivers/net/wireless/prism54/islpci_dev.c
@@ -387,9 +387,7 @@ islpci_open(struct net_device *ndev)
 	}
 
 	netif_start_queue(ndev);
-
-	/* Turn off carrier unless we know we have associated */
-	netif_carrier_off(ndev);
+/*      netif_mark_up( ndev ); */
 
 	return 0;
 }

Re: prism54: revert "prism54: set carrier flags correctly"

[Luis R. Rodriguez]

Patick, thanks for staying on top of this, however NACK. The fix I
sumbitted and you tested should be merged. I nagged about it too, John
just wanted a resubmit with a log message.

John please accept the patch I posted, sorry I'd resend and add a nice
commit log message but I've been away from interweb access for a few
days now and expect it to be this way for the next two days.

You can add something like "this fixes a regression introduced by $SHA1_SUM"

Thanks,

  Luis



On 5/12/08, Patrick McHardy <kaber@trash.net> wrote:
>

-- 
Sent from Gmail for mobile | mobile.google.com

Re: prism54: revert "prism54: set carrier flags correctly"

[Luis R. Rodriguez]

Oh and if you're wondering how im sending this, its through GPRS on my
trusty bb and through my trusty gmail :) so I do have at least "some"
level of interweb access..

Now if only I could hack on this too...



On 5/12/08, Luis R. Rodriguez <mcgrof@winlab.rutgers.edu> wrote:
> Patick, thanks for staying on top of this, however NACK. The fix I
> sumbitted and you tested should be merged. I nagged about it too, John
> just wanted a resubmit with a log message.
>
> John please accept the patch I posted, sorry I'd resend and add a nice
> commit log message but I've been away from interweb access for a few
> days now and expect it to be this way for the next two days.
>
> You can add something like "this fixes a regression introduced by $SHA1_SUM"
>
> Thanks,
>
>   Luis
>
>
>
> On 5/12/08, Patrick McHardy <kaber@trash.net> wrote:
> >
>
> --
> Sent from Gmail for mobile | mobile.google.com
>

-- 
Sent from Gmail for mobile | mobile.google.com

Re: prism54: revert "prism54: set carrier flags correctly"

[Patrick McHardy]

Luis R. Rodriguez wrote:
> Patick, thanks for staying on top of this, however NACK. The fix I
> sumbitted and you tested should be merged. I nagged about it too, John
> just wanted a resubmit with a log message.


I'm happy either way, I just wanted to resurrect the interest in
this bug :)