**** BEGIN LOGGING AT Sat Sep 01 02:59:56 2007 Sep 01 03:41:54 nico * r8566 /packages/libs/yapsnmp/Makefile: add dependency on python (closes: #2283) Sep 01 05:20:16 nico * r8567 /trunk/include/package.mk: fix .dep_files exclusion in rdep calls Sep 01 05:25:38 nico * r8568 /trunk/target/linux/rdc-2.6/ (3 files in 2 dirs): small rdc-2.6 fixes (from sn9, thanks!) Sep 01 05:28:47 {Nico}: [florian] isn't happy about that change Sep 01 05:33:28 <{Nico}> sn9: why? Sep 01 05:34:15 <[florian]> sn9: what's the problem with it ? this is an usual programming technique to handle errors in drivers instead of having a complex state machine to ensure you were at step x) and should free resources y,z ... Sep 01 05:34:15 <[florian]> sn9: this really needs no patch Sep 01 05:34:15 but there are no resources referenced in this case Sep 01 05:34:15 <[florian]> sn9: I said resources, this can be anything Sep 01 05:34:15 <[florian]> don't bitch about that Sep 01 05:36:00 <{Nico}> sn9: you mean in rdc321x_wdt_probe? Sep 01 05:36:08 right Sep 01 05:36:32 btw, what did you think of my DEPENDS:=-kernel idea? Sep 01 05:38:14 <{Nico}> i don't see anything wrong with removing a goto at this stage Sep 01 05:38:55 neither do i Sep 01 05:39:15 <{Nico}> but maybe errors should be handled in subsequent functions calls Sep 01 05:39:41 this code would not affect that Sep 01 05:40:07 <{Nico}> right Sep 01 05:40:20 <{Nico}> what's your -kernel thing? Sep 01 05:40:40 the part of the patch you didn't apply Sep 01 05:42:19 <{Nico}> is this working? Sep 01 05:42:55 i have not tried a .21 target to find out Sep 01 05:43:21 works fine for .22 Sep 01 05:43:47 <{Nico}> i'm not sure DEPENDS:=-something is implemented Sep 01 05:44:08 i believe i've seen it in some other pkg Sep 01 05:44:46 there has to be some way to specify a pkg dependency conflict Sep 01 05:47:19 i will checkout a separate tree and try it with both .22 and .21 right now Sep 01 05:48:50 <{Nico}> i only see '+' and '@' handled, not '-' Sep 01 05:49:02 <{Nico}> you should try '!' instead Sep 01 05:49:14 i will try Sep 01 05:49:39 what's a .21 target i could test with? Sep 01 05:49:51 <{Nico}> ixp4xx-2.6 Sep 01 08:03:16 hi sn9 ? Sep 01 08:03:22 hello Sep 01 08:03:37 ? Sep 01 08:03:41 hi wrt Sep 01 08:03:47 LoneStar: what? Sep 01 08:03:48 you are always here :) Sep 01 08:04:10 well, i have some questions about the kernel of kamikaze Sep 01 08:04:21 go ahead Sep 01 08:04:28 now I got loader.elf and loader.bin Sep 01 08:04:34 hi sn9 , you are here Sep 01 08:04:41 welcome to #openwrt-devel! Sep 01 08:05:10 can I burn loader.bi to flash, and simply let bootloader jump to that address? Sep 01 08:05:48 on ar7, part of the bootloader is built into the cpu Sep 01 08:06:33 there should be a bootloader to bridge that part and the kernel Sep 01 08:06:44 such as adam2 or pspboot Sep 01 08:07:21 yes, I have one, psbl, from TI Sep 01 08:07:37 it is on 0xb0000000, and take over the system after power on Sep 01 08:07:45 yes Sep 01 08:08:05 how can it jump to our kernel then? Sep 01 08:08:40 so, tnetv is more like regular mips, because ar7 has adam2 or pspboot at 0x90000000 Sep 01 08:09:05 0xbfc00000 is inside the cpu Sep 01 08:09:12 0xb0000000 are the same as 0x90000000 Sep 01 08:09:24 ah, i did not realize that Sep 01 08:09:29 just an uncached area Sep 01 08:10:26 did the psbl come with instructions? Sep 01 08:10:48 you mean a command line? Sep 01 08:11:11 some documentaion saying where the kernel should go Sep 01 08:14:33 yes, some. but seems complex, I modified to code to let it jump to a fix address after a timeout Sep 01 08:14:57 which address? beginning of the flash? Sep 01 08:15:02 my plan was to put linux kernel right there, so bring it up Sep 01 08:15:24 middle of the flash. the beginning is the psbl itself Sep 01 08:16:02 if psbl is linux-aware, short-circuiting it may not be the best idea Sep 01 08:19:07 i found loader.elf was linked to address 0x94600000, does it need to be put right that place to run? Sep 01 08:25:21 you looked at target/linux/ar7-2.6/image/Makefile, right? Sep 01 08:41:02 yes Sep 01 08:41:40 LOADADDR := 0x94600000 Sep 01 08:41:47 {Nico}: you're right, those dependencies don't work at all. replacement patch: http://homepage.mac.com/danielg4/rdc.patch Sep 01 08:43:10 LoneStar: loader.bin contains its own loader, so you jump to the loader, not the kernel Sep 01 08:43:53 but i think the loader will decompress the real kernel and call it in turn Sep 01 08:44:04 won't it? Sep 01 08:45:21 right Sep 01 12:57:10 nbd * r8569 /trunk/package/ppp/files/ppp.sh: allow the user to disable the defaultroute option for ppp Sep 01 14:26:42 re Sep 01 14:53:32 bye Sep 01 15:20:28 nico * r8570 /trunk/scripts/abs2rel.pl: add an abs2rel.pl helper script (returns a path relative to another) Sep 01 21:01:04 blogic * r8571 /trunk/package/hotplug2/files/hotplug2.rules: reduce watchdog trigger from 10 to 5 seconds Sep 01 23:49:12 n8 **** ENDING LOGGING AT Sun Sep 02 02:59:57 2007