**** BEGIN LOGGING AT Thu May 24 02:59:57 2007 **** BEGIN LOGGING AT Thu May 24 07:35:47 2007 May 24 08:17:02 kaloz * r7326 /trunk/target/linux/aruba-2.6/patches/000-aruba.patch: cleanup patch May 24 09:10:04 loswillios: any news on the madwifi testing? May 24 09:43:57 kaloz * r7327 /trunk/target/linux/aruba-2.6/patches/ (000-aruba.patch 011-ap6x_serial_fix.patch): fix aruba compile for .21 May 24 10:02:00 nbd: nope. trying to revert to r7314 atm, I accidently updated yesterday May 24 10:03:29 wouldn't it make more sense to wait for the kernel-upgrade and mac80211/dadwifi anyway? May 24 10:04:42 no, dadwifi has a long way to go May 24 10:04:53 it's not even close to being ready May 24 10:05:12 ok May 24 10:44:26 <[mbm]> big long post to the forum about how the device works and various troubleshooting May 24 10:45:51 <[mbm]> btw, saw a ticket earlier that brcm-2.4 was broken; anyone looked into that? May 24 10:46:07 not yet May 24 10:46:17 link? May 24 10:47:39 <[mbm]> https://dev.openwrt.org/ticket/1741 .. think it's just a bad compile May 24 10:48:16 <[mbm]> hmm thought there was another bug as well May 24 10:52:39 <[mbm]> must be thinking of #1710 .. but that's a brcm47xx-2.6 issue May 24 10:52:59 [mbm]: can you test aruba? May 24 10:53:13 <[mbm]> yeah I'll take a look tomorrow when I'm in the office May 24 10:53:28 okie :) May 24 10:53:31 <[mbm]> (had some complaints that it was broken) May 24 10:53:48 sure, as we were upgrading :p May 24 10:53:57 btw, that whole code needs some cleanup May 24 10:54:13 and i'm not bitching about the aruba parts, the idt code itself is ugly May 24 10:54:20 <[mbm]> yeah, not the least bit surprised by the breakage, but it did mean tracking down a stable version since pre1 doesn't have aruba-2.6 May 24 10:54:55 [mbm]: topic :) but if i've fixed .21 right, it should be better May 24 10:55:02 <[mbm]> yep. at some point we should extend the atheros-2.6 and drop the aruba target May 24 10:55:25 atheros has to be merged mainline as well as idt after cleanup May 24 10:55:44 aruba should be a mix of patches to apply on top of that two May 24 10:55:57 just to provide the platform specific stuff and the glue to run on both soc May 24 10:56:19 like we need a routerboard target that uses idt and adm5120 May 24 10:56:57 <[mbm]> well, apart from the ap70 aruba is atheros-2.6; and the idt is similar enough to atheros that the kernel only needs a few tweaks May 24 10:57:24 <[florian]> Kaloz: would you like to add things like (if (machine_is_rb500/100) like routerboard did ? May 24 10:57:32 well, the pci stuf fon idt is totally different May 24 10:58:08 <[mbm]> irq handling has also changed May 24 10:58:09 [florian]: machine_is_foo is ARM thingie :) the right thing is to switch MACH_TYPE like we do on atheros and aruba May 24 10:58:27 <[florian]> Kaloz: and on adm5120 May 24 10:59:09 [florian]: dunno how different adm5120 stuff are.. bootloader differences aren't soc differences :) May 24 10:59:27 <[florian]> Kaloz: that's right, we just set the machine_type to the right board we have detected May 24 10:59:37 [florian]: why? May 24 10:59:45 <[florian]> Kaloz: because it's nicer May 24 10:59:49 [florian]: honestly other then nand, what's different on adm5120? May 24 11:00:02 [florian]: and totally not mergable upstrem May 24 11:00:06 <[florian]> Kaloz: what a question ! the number of leds :) May 24 11:00:09 <[mbm]> for once I agree with florian - setting mach type is nice May 24 11:00:17 [florian]: bootloader stuff are not the same as soc differences May 24 11:00:57 [florian]: i will get popcorn and watch how many people will go to kill you when you introduce machtypes for leds.. :P May 24 11:01:08 <[florian]> Kaloz: ;) May 24 11:01:22 [mbm]: it's nice, for what we are doing on aruba and atheros May 24 11:01:33 [mbm]: but it's not nice for bootloader stuff :P May 24 11:01:47 but whatever :) May 24 11:02:01 <[florian]> Kaloz: for adm5120 it is possible because we gather board specific informations from the bootloader itself May 24 11:03:42 <[florian]> and I find it more having a boot time routine detecting which bootloader you are booting from May 24 11:04:03 <[florian]> Kaloz: and for instance, we need to check the type of bootloader so that the flash size is set up accordingly May 24 11:04:42 ? May 24 11:04:50 aren't you doing CFI? May 24 11:05:10 <[florian]> we do May 24 11:05:28 <[florian]> but I think it did not work very well for edimax boards May 24 11:05:42 edimax board we don't really want to support May 24 11:05:43 :p May 24 11:05:51 iirc all of them have 2mb flash May 24 11:06:25 <[florian]> I think so May 24 11:06:30 <[florian]> anyway, have to go May 24 12:44:45 kaloz * r7328 /trunk/target/linux/ (18 files in 12 dirs): fix Magicbox for .21 and convert to the new structure May 24 14:43:40 nbd * r7329 /trunk/target/linux/rb532-2.6/ (5 files in 4 dirs): fix rb532 compile May 24 14:52:13 nbd * r7330 /trunk/target/linux/rb532-2.6/files/include/asm-mips/rc32434/irq.h: don't redefine MIPS_CPU_IRQ_BASE May 24 15:59:32 nbd * r7331 /trunk/ (10 files in 7 dirs): fix ar7 compile May 24 21:02:52 kaloz * r7332 /trunk/target/linux/sibyte-2.6/ (5 files in 3 dirs): fix .21 for sibyte.. because we can.. May 24 21:57:33 florian * r7333 /trunk/target/linux/au1000-2.6/ (8 files in 2 dirs): Update au1000 for .21, build an SREC initramfs image so that it can be tested directly with yamon May 24 23:30:34 florian * r7334 /trunk/target/linux/ (4 files in 4 dirs): add generic GPIO to MIPS boards, use it for au1000 and for the MTX-1 led driver as well **** ENDING LOGGING AT Fri May 25 02:59:57 2007