[01:23] <BenC> mjg59: thanks, however the x86_apic.diff was empty
[01:27] <mjg59> Hm.
[01:28] <mjg59> Ok, resent
[03:14] <BenC> mjg59: do I need to backout your old patch before applying these?
[03:15] <mjg59> BenC: Yup
[03:15] <BenC> ok, thanks
[03:15] <mjg59> Oh, hang on
[03:15] <mjg59> No
[03:15] <BenC> ok
[03:15] <mjg59> Just apply them as is
[03:15] <mjg59> (I think)
[03:15] <mjg59> The guy with the softmac prism54 vanished before he could tell me what the firmware needed to be called
[03:16] <BenC> I'm talking about the ioapic patches
[03:16] <mjg59> Yeah
[03:16] <mjg59> I was just mentioning my caveat in the other mail :)
[03:16] <BenC> so the disable_timer_pin_1 thing for ATI remains?
[03:16] <BenC> ah, ok
[03:16] <BenC> I can check the driver file to find that out
[03:18] <mjg59> BenC: disable_timer_pin_1 should still be there (it can be set on boot), but it shouldn't automatically be set anywhere now
[03:18] <mjg59> Instead timer_over_8254 gets set to 0
[03:18] <mjg59> It's actually the inverse of my original patch
[03:18] <BenC> ok, so I need to backout the patch that does that for ATI
[03:18] <mjg59> The one I provided ought to do that
[03:19] <mjg59> -                                               disable_timer_pin_1 = 1;
[03:19] <BenC> yeah, it does
[03:19] <BenC> I wasn't looking at the x86_64 one, my mistake :)
[03:19] <mjg59> Heh
[03:28] <Keybuk> I HATE NETWORK MANAGER
[03:28] <Keybuk> that is all
[03:28] <Keybuk> :p
[03:29] <mjg59> I'd been getting that impression...
[03:29] <mjg59> What's up with it now?
[03:29] <Keybuk> won't play WPA
[03:29] <Keybuk> something about NM, WPA and madwifi causes unhapiness
[03:29] <Keybuk> I think it's not getting the message that it WORKED
[03:29] <Keybuk> so it keeps blacklisting and retrying
[03:30] <Keybuk> while the AP log says "Associated"
[03:30] <mjg59> Hm.
[03:30] <mjg59> I'm sure I've seen other people bitching about that.
[03:30] <mjg59> rml sent a patch to the list last week, IIRC
[03:31] <Keybuk> yeah lots of bitching, no working
[03:31] <Keybuk> I checked the list and we have all the patches
[03:31] <Keybuk> without the patch rml sent, even non-WPA doesn't work
[03:31] <Keybuk> (because wpa_supplicant fouls up the association, I assume)
[03:31] <mjg59> Right
[03:31] <Keybuk> at least normal and WEP work right now
[03:31] <mjg59> And you can connect just using wpa_suppository?
[03:31] <Keybuk> so I've declared success and uploaded it
[03:31] <Keybuk> yup
[03:32] <mjg59> Fun
[03:32] <mjg59> Probably some exciting timing issue
[03:32] <Keybuk> stuck wpa-driver madwifi, wpa-ssid linksys, wpa-pks blahblah in /etc/network/interfaces and that just works
[03:32] <mjg59> Never mind. You've got more broken patches to look forward to.
[03:32] <Keybuk> so don't think it's a suppository issue
[03:32] <Keybuk> indeed
[03:32] <mjg59> Ah. For an equivalent test, shouldn't you be using the wext driver?
[03:33] <mjg59> Or does nm hardcode it to madwifi?
[03:33] <Keybuk> wext doesn't work at all
[03:33] <Keybuk> one of rml's patches it to use madwifi
[03:33] <Keybuk> without that one, it fails to associate to any AP
[03:33] <mjg59> Right
[03:33] <mjg59> Well, my only madwifi card is in London right now
[03:34] <Keybuk> tomorrow I'll get around to installing Umbungo on my Powerbook and test it on that
[03:34] <Keybuk> eek out a whole new world of corner case pain
[03:34] <mjg59> How recent is the powerbook? Airport or express?
[03:34] <Keybuk> neither
[03:34] <mjg59> Ah
[03:34] <Keybuk> pcmcia wireless card
[03:34] <mjg59> Ancient :)
[03:35] <Keybuk> Cheap :)
[03:35] <mjg59> Hmm. I worry about the way that the madwifi tree in the prism54_softmac driver source appears to have been mutilated to look like ieee80211_softmac
[03:35] <mjg59> In a "Does this thing actually work with the in-kernel stack?" sort of way
[03:37] <Keybuk> heh
[03:37] <mjg59> Well, it /loads/ with it
[03:37] <mjg59> But I don't have the hardware to test
[03:37] <mjg59> Fuckit. We'll see who complains.
[03:38] <mjg59> I'm getting really irritated with people bitching that us shipping a proper bcm43xx driver has broken their ndiswrapper setup
[03:38] <mjg59> BenC: Did that set of bcm patches I sent you apply ok?
[03:38] <mjg59> They ought to get things a lot closer to working out of the box, rather than requiring precise timing...
[03:39] <mjg59> The driver exports a number of sysfs interfaces in
[03:39] <mjg59> /sys/class/islsm/$(DEVICE)/
[03:39] <mjg59> HNGH.
[03:39] <mjg59> I'm fairly sure that's not the correct approach.
[03:47] <Keybuk> evil
[03:47] <Keybuk> nothing under /sys/class/net ?
[03:47] <mjg59> Oh, probably
[03:47] <Keybuk> if islsm are the "upper" devices, that's not so bad
[03:47] <Keybuk> a whole subsystem just for one driver
[03:47] <mjg59> Yeah
[03:47] <Keybuk> wooo
[03:48] <mjg59> It's just so you can tweak some values per-interface
[03:48] <mjg59> Which I thought you could do by sticking stuff under the per-device sysfs stuff /anyway/
[03:48] <mjg59> Since ipw* seem to manage it
[03:52] <mjg59> Ok, looks like it has been ported to softmac
[03:52] <mjg59> Hurrah
[03:20] <BenC> yay, my unread bug email list is below 100!
[05:06] <zul> heylo
[11:22] <hub> hi guys, apparently the PPC kernel does not provide the old iBook clamshell or PowerBook G3 modem
[11:22] <hub> while there is a kernel module in the tree
[11:22] <hub> (used to be in older version)
[11:22] <hub> has it just been forgotten?
[11:25] <infinity> please file a bug at launchpad.net/malone (package: linux-source-2.6.15) and mention the module that's gone missing.
[11:25] <infinity> BenC will no doubt attack it with vigor when he's around.
[11:26] <hub> okay. I just wanted to make sure that it was not on purpose
[11:27] <hub> a bug will be filed, off course
[11:27] <infinity> We don't tend to remove many drivers on purpose. :)
[11:27] <infinity> But when we switched to 2.6.15 and upended the packaging completely, some stuff got dropped on the floor, lost, forgotten, etc.
[11:27] <hub> I know
[11:27] <hub> but sometime it is
[11:27] <hub> I understand
[11:28] <hub> no problem
[11:28] <infinity> Also, xandros.com.  Nice.
[11:28] <infinity> Don't let anyone from the dccalliance know you're running the evil ubuntu. :)
[11:30] <hub> infinity: they know that. I wear ubuntu t-shirt, and I run ubuntu/gnome on my laptop
[11:30] <hub> that is the one I'm typing from atm
[11:30] <infinity> :)
[11:30] <infinity> Even I don't wear Ubuntu t-shirts, and I get 'em for free...
[11:31] <infinity> Though maybe that's why I don't wear them.
[11:31] <infinity> I don't wear any of my free Microsoft or IBM t-shirts either.  Could be a pattern.
[11:31] <hub> infinity: I have a t-shirt of my former company atm. it is sort of a competitor
[11:32] <hub> I don't wear MSFT t-shirt because I don't want any
[11:33] <mjg59> infinity: I'll wear the brown one when I get low on them, but that's because it's actually a good t-shirt
[11:34] <infinity> It is rather nice fabric, yes.  Sadly, I popped mine in the dryer without turning it inside out first, and the logo got a bit melted.
[11:34] <infinity> Whoops.
[11:34] <mjg59> Nf. Must tidy my room.
[11:35] <mjg59> Maybe I'll also try to fix vga16fb on the craptop
[11:36] <hub> mjg59: uh oj vga16fb. was a topic around at the office
[11:36] <hub> infinity: I only have on whit t-shirt that I got at ubz
[11:37] <hub> filed bug 36499
[11:37] <hub> for the pmac modem
[11:38] <mjg59> hub: Yeah, I brought it up on dcc-devel
[11:38] <mjg59> (due to ACPI not working too well with non-vga16fb)
[11:38] <hub> yeah, pat told me that
[11:38] <hub> I don't read the dcc mailing list
[11:52] <hub> c'ya