**** BEGIN LOGGING AT Fri Jul 31 02:59:57 2009 Jul 31 03:25:55 k. we got ezstream running Jul 31 04:07:39 Hi. Jul 31 04:07:59 Who tested d-link dir-330 router? Jul 31 05:18:45 good morning Jul 31 05:18:57 good morning Jul 31 05:19:00 how goes this morning Jul 31 07:01:27 juhosg * r17057 /trunk/target/linux/ar71xx/files/arch/mips/ar71xx/mach-wrt160nl.c: [ar71xx] add ethernet suppor for the WRT160NL (thanks to Christian Daniel) Jul 31 07:02:13 jushog, I haven't given that one a serious test yet, but, so far, seems to work, with one gotcha Jul 31 07:02:22 mac addresses aren't right Jul 31 07:02:34 and i'm completely re-writing the trx partition handler Jul 31 07:06:13 groz: i assume that the mac address is stored in the nvram partition, so we can get it from there probably Jul 31 07:06:20 yes Jul 31 07:06:25 but, i found a few other issues Jul 31 07:06:26 example Jul 31 07:06:36 the stuff for building trx from uimage Jul 31 07:06:46 it's not using a multi-part trx Jul 31 07:06:54 so, hard to split off the partitions properly Jul 31 07:07:03 i've re-worked it to use multi-part trx images Jul 31 07:07:21 and now working in the startup code that correctly makes the mtd partitins match the trx Jul 31 07:07:29 groz: you rock Jul 31 07:07:38 then the squash split stuff should suddenly 'just work' Jul 31 07:12:02 but for some reason, it's telling me I dont have a squashfs in my partition now Jul 31 07:12:05 and the boot is failing Jul 31 07:12:18 i'm triple checked all the numbers, i am mapping them correctly now Jul 31 07:12:24 according to what is in the bin file Jul 31 07:18:59 ju, got a dumb question, should I be able to mount the root.squashfs on my dev machine, or is endian a problem to mount it on x86 box ? Jul 31 07:19:10 i'm getting errors trying to mount it and look inside it Jul 31 07:19:38 getting zlib inflate errors Jul 31 07:19:45 groz: we use lzma in squashfs. Jul 31 07:20:06 hmm, ok Jul 31 07:20:11 (and not the unofficial official lzma squash patch either) Jul 31 07:21:36 you can use the unsquashfs-lzma binary that is generated by the buildsystem to peek insinde. Jul 31 07:22:02 well i'm trying to figure out why it keeps failing to mount Jul 31 07:22:16 so, i was just gonna try mount same fs here Jul 31 07:23:05 mtd: partition "rootfs" set to be root filesystem Jul 31 07:23:05 split_squashfs: no squashfs found in "spi0.0" Jul 31 07:23:13 keep getting that message Jul 31 07:23:37 and shortly later, kernel panic, no root Jul 31 07:23:56 No filesystem could mount root, tried: squashfs Jul 31 07:23:56 Kernel panic - not syncing: VFS: Unable to mount root fs on unknown-block(31,2) Jul 31 07:24:24 that's the correct partition Jul 31 07:24:35 and it's showing correct size in the mtd prints Jul 31 09:52:35 iptables v1.4.4: Couldn't load match `recent':File not found Jul 31 09:53:25 i've insmod'd xt_recent.ko Jul 31 10:47:01 florian * r17058 /trunk/target/linux/brcm63xx/image/Makefile: [brcm63xx] generate images for RTA1025W_16-based routers (#5594) Jul 31 10:49:20 <[florian]> russell_: don't you also need its userspace counterpart ? Jul 31 10:49:47 florian * r17059 /trunk/package/fuse/Makefile: [package] disable fuse kernel compilation module, all versions that we have have it mainline (#5584) Jul 31 11:09:29 florian * r17060 /trunk/ (4 files in 4 dirs): [kernel] update to 2.6.27.29 Jul 31 11:10:21 florian * r17061 /trunk/target/linux/ (4 files in 2 dirs): [kernel] refresh 2.6.27 patches Jul 31 11:13:59 florian * r17062 /trunk/ (17 files in 17 dirs): [kernel] update to 2.6.30.4 Jul 31 12:15:52 kaloz * r17063 /trunk/ (include/package-defaults.mk package/opkg/Makefile): use board names instead of arch in packages Jul 31 12:47:11 [florian]: hi florian, haven't been keeping up. news on bcm6345? Jul 31 12:49:54 danage: https://dev.openwrt.org/changeset/17054 Jul 31 13:28:26 hello guys, one question to be sure, when I use a c function that return char *, I need free memory when I done, right? Jul 31 13:39:31 fofware: it depends on your code Jul 31 13:40:45 fofware: if function that return char* do the malloc, returned pointer need to be free somewhere... Jul 31 13:43:05 thanks, the case is for example char *foo = inet_ntoa(struct in_addr ipval) or other functions like this Jul 31 13:44:26 I imagine It do a malloc to return a char, so after I do free(foo), but some times i get error Jul 31 13:45:17 fofware: are you sure that this function never return NULL pointer ? Jul 31 13:46:03 yea, because I show the values before free Jul 31 13:48:54 ok, I will check if each var is not null before free Jul 31 13:52:17 fofware: inet_ntoa may uses a static buffer. Jul 31 13:53:20 fofware: in this case, you don't need to free it, the better is to directly look at source, you will be sure... Jul 31 13:54:43 so, if I do many printf("%s",inet_ntoa(...)); that don't leave memory Jul 31 13:55:28 thanks Jul 31 13:56:20 "The string is returned in a statically allocated buffer, which subsequent calls will overwrite." Jul 31 13:56:27 from man 3 inet_ntoa Jul 31 13:57:47 thanks xMff for this accurate information Jul 31 13:58:43 thanks and where I can find the soureces or help of each functions in library? Jul 31 13:59:30 fofware: on my ubuntu here it's apt-get install manpages-dev manpages-posix manpages-posix-dev Jul 31 13:59:33 fofware: as said by xMff : "man" command is helpful in your case :) Jul 31 13:59:36 then just man 3 functionname Jul 31 14:00:28 ok, thanks Jul 31 14:01:15 <[florian]> danage: nothing new yet Jul 31 14:01:17 that is very good data for me Jul 31 14:02:02 indeed Jul 31 14:02:14 it even tells what headers to use :) Jul 31 14:02:19 rtfm ftw Jul 31 14:13:17 looking at https://dev.openwrt.org/browser/trunk/target/linux/ar71xx/files/arch/mips/ar71xx/mach-rb-4xx.c#L186 Jul 31 14:13:29 ar71xx_add_device_eth defines two nics Jul 31 14:13:50 but the 433uah has three nic Jul 31 14:14:04 what should i change? Jul 31 14:24:40 florian * r17064 /trunk/toolchain/gdb/Makefile: [toolchain] disable warnings as error in the gdb configuration (#5602) Jul 31 15:20:28 Dogge_: the cpu has only 2 nics Jul 31 15:21:49 Dogge_: simply one of them is connected to a switch chip.. but thay doens't give you a 3rd interface, just a 3rd port Jul 31 15:31:47 ok, battling a bit of an issue here, anybody got any pointers fer me on this Jul 31 15:32:05 got the 160nl oing, all the hardware is working, booting from initramfs Jul 31 15:32:24 i can mount the flash partition and use it with jffs Jul 31 15:32:38 but, if i build a squashfs root into my flash partition Jul 31 15:32:50 it seems to map the loacations correctly in the mtd tables Jul 31 15:32:57 but attampting to mount it fails Jul 31 15:33:07 complaining about no squashfs file system in there Jul 31 15:33:45 trying to figure out a way of determining that the squashfs is indeed good, and correctly built Jul 31 15:39:34 which kernel version? Jul 31 15:40:00 28.10 Jul 31 15:40:00 also, pastebin a full bootlog for me Jul 31 15:40:09 ok, hang on, i'm just rebuilding my netboot kernel Jul 31 15:41:52 http://pastebin.com/m5ffbf554 Jul 31 15:41:55 ok, this one I just added the ethernet driver back Jul 31 15:41:59 been working without it Jul 31 15:42:07 cuz it's triggering a fault during boot Jul 31 15:42:21 so that stack trace you see in this one, has not been in it up to now Jul 31 15:42:35 how can i build single binaries for a router running on mips32 core ? this kamikaze thingie can help with that or it builds only whole images Jul 31 15:42:39 and this is an initramfs build right now Jul 31 15:42:53 i have to rebuild a bin file and flash it again to get the boot log from the failed squashfs Jul 31 15:43:29 groz Jul 31 15:43:39 what ethernet driver did you put back? Jul 31 15:43:53 it's the same one jushog commited yesterday Jul 31 15:44:05 ah ok Jul 31 15:44:10 i got the code from hoscotch, it's actually just board specific startup stuff Jul 31 15:44:17 which configures things for this board Jul 31 15:44:39 btw, whoever did that trx-on-ar71xx braindamage at cisco should be shot on sight Jul 31 15:44:41 it's not a driver, it's startup code that correctly initializes this one Jul 31 15:45:00 groz: show me the log when it fails to mount the squashfs :) Jul 31 15:45:15 ok, give me 5 minutes to rebuild a squashfs binary Jul 31 15:45:17 and flash it Jul 31 15:46:12 if i can get past this, then, i'm on the downhill slope with this board, cuz, everything seems to work now, just need to clean up a couple of things Jul 31 15:47:21 . Jul 31 15:49:40 him_: it's designed for whole images, but you can use it or the sdk for compiling simply apps Jul 31 15:49:46 groz: be back in half an hour Jul 31 15:50:23 ok, im just flashing this thing now Jul 31 15:50:27 Kaloz and i need to configure all the mess even if i just want to compile single apps? Jul 31 15:55:31 good morning juhosg Jul 31 15:56:20 hi groz Jul 31 15:56:54 have you used a squash root on ar71 recently ? Jul 31 15:57:14 yes Jul 31 15:57:19 having trouble mounting the squash partition, and, searching for where it may be broken Jul 31 15:58:58 it is working for me on the tew-632brp with .28 and with .30, i flashed it yesterday Jul 31 15:59:05 http://pastebin.com/m5a7dac2b Jul 31 15:59:13 two things on this one, if you can look for just a sec Jul 31 15:59:32 a) the ethernet stuff does cause a stack dump in boot, i'm gonna look at that, after i get my root working Jul 31 16:00:07 b) it seems to be mapping the squash partition in the right places Jul 31 16:00:13 but, failing to mount it Jul 31 16:01:31 my next avenue to try look at whats going on, gonna flash the partition from an initramfs boot Jul 31 16:01:37 with the binary out of the bin dir Jul 31 16:01:41 hm, pastebin.com does not work for me atm Jul 31 16:01:43 and see if i can manually get it to do anything Jul 31 16:01:53 haha, figgers Jul 31 16:02:48 http://www.pastebin.ca/1513871 Jul 31 16:02:52 does that one work ? Jul 31 16:03:00 lotsa pastebins around Jul 31 16:03:26 yeah, this is ok Jul 31 16:03:59 Atheros(R) L2 Ethernet Driver - version 2.2.3 ? Jul 31 16:04:29 the only changes i have in ethernet, was the startup code from hopscotch Jul 31 16:04:35 which you got yesterday too Jul 31 16:04:44 the rest is just a vanilla checkout from a couple days ago Jul 31 16:05:26 well, maybe a week, cant remember exactly Jul 31 16:05:44 but i did svn up a couple days ago, just before i started all this Jul 31 16:06:24 why, is that an old one ? Jul 31 16:07:31 we did not use this driver at all Jul 31 16:07:39 hmmmmmmm Jul 31 16:08:17 i'm gonna check out a fresh copy Jul 31 16:08:47 cuz, i've not been playing with ethernet drivers, other than the startup code from hopscotch yesterday Jul 31 16:08:47 which you committed later Jul 31 16:10:46 weird, something is wrong in your kernel configuration Jul 31 16:10:58 k, i'm gonna start fresh Jul 31 16:11:00 and see what happens Jul 31 16:11:10 i have fussed a bunch with kernel configs over the last 24 hours Jul 31 16:11:31 mostly changing command lines, and removing yaffs because i thought it was breaking the mount Jul 31 16:12:05 when you apply the patch from hopscotch for the trx parser, it adds another kernel config line too Jul 31 16:14:05 fyi, i did re-work that trx parser a bunch, now it correctly handles multi-part trx Jul 31 16:18:09 ok, the rootfs_split should be working, can you check the flash contents from u-boot? Jul 31 16:18:49 'md.b 0xbf140000 0x100' Jul 31 16:18:59 just getting into uboot Jul 31 16:19:42 bf140000: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................ Jul 31 16:19:42 bf140010: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................ Jul 31 16:19:42 bf140020: 73 71 73 68 00 00 01 e4 00 00 00 00 00 00 00 00 sqsh............ Jul 31 16:19:42 bf140030: 00 00 00 00 00 00 00 00 00 00 00 00 00 03 00 00 ................ Jul 31 16:19:42 bf140040: 00 00 00 10 40 01 00 4a 73 14 77 00 00 00 00 08 ....@..Js.w..... Jul 31 16:19:54 hmmm, is that off by 32 ?? Jul 31 16:19:56 i think it may be Jul 31 16:19:56 it i wrong Jul 31 16:20:25 yes, the sqsh should be at 0xbf140000 Jul 31 16:20:37 ok, now I know what I'm hunting for Jul 31 16:20:45 and i know exactly where to look Jul 31 16:20:50 that's in the creation of the trx Jul 31 16:20:58 ok Jul 31 16:21:11 but to double check Jul 31 16:21:17 i can kludge the partition map code Jul 31 16:21:25 and repoint start of position up by 32 bytes Jul 31 16:21:28 see that it then mounts ro Jul 31 16:21:51 squash doesn't have to be aligned Jul 31 16:23:12 I've compiled trunk for my wrt160n and wifi is not working in dmesg I get wl0: wlc_attach: failed with err 13 Jul 31 16:23:15 wl%d: 4.150.10.5 driver failed with code 13 Jul 31 16:24:38 groz: yes, sqashfs must be working at any offset Jul 31 16:25:00 i'll look carefully in a bit at why it got like thiat Jul 31 16:25:05 bigon: have you installed the brcm-wl-mimo driver? Jul 31 16:25:09 i'm suspecting, the uboot bin flasher isn't stripping the header Jul 31 16:25:16 the way they do on the gl Jul 31 16:25:37 so the extra bits from addheader are in flash too Jul 31 16:25:49 and that's offsetting it all by 312 Jul 31 16:25:51 32 Jul 31 16:26:12 u-boot does not strip anything Jul 31 16:26:39 ok, so when I re-worked a few things , was on the assumption the trx is in flash Jul 31 16:26:46 just like it is on the wrt54gl Jul 31 16:26:51 not the entire bin Jul 31 16:27:11 I can confirm this in a couple more minutes Jul 31 16:29:03 it was late last nite, i was getting punchy Jul 31 16:29:14 never occurred to me to just dump flash from uboot Jul 31 16:29:42 Bartman007: mm work better indeed :/ sorry for the noice Jul 31 16:30:38 just an other quick question may I choose the target profile with Generic, Broadcom WiFi (MIMO) ? Jul 31 16:31:13 bigon: yes, that will include it in your images. Jul 31 16:31:14 haha, ju, that was it Jul 31 16:31:28 Bartman007: great thx very much Jul 31 16:52:21 groz: this kind of 'WARNING' usually happens when you are trying to register two devices with the same name Jul 31 16:52:59 i'm doing a fresh build from trunk now Jul 31 16:53:06 just did a fresh checkout Jul 31 16:53:21 bothers me that it's got a driver version in there that you say was never used Jul 31 16:53:29 i musta messed up kernel config somewhere along the way Jul 31 16:55:49 a lot of the last couple days has been spent learning my way around the kamikaze build environment Jul 31 16:55:59 and, i've touched a lot of things that likely dont need touching Jul 31 16:56:11 better to start fresh, now that I know specifically what spots need touching Jul 31 16:58:24 ok, we'll see it Jul 31 16:59:32 example, once i stumbled on the squash split stuff, start to wonder if parsing trx is even necessary Jul 31 16:59:46 <_trine> juhosg: just a quick question have you seen that the ar71 code from trunk does not have an IP address when it boots Jul 31 17:01:27 * _trine thinks hes invisible Jul 31 17:01:35 <_trine> cloak Jul 31 17:02:41 _trine: nope, i have tried the tew-632brp with .28, and with .30 yesterday, but i did not see such problem Jul 31 17:02:52 <_trine> oh Jul 31 17:03:03 <_trine> I'll give it another try then Jul 31 17:03:29 <_trine> that's strange I tried several times too Jul 31 17:06:30 it was r17037, and it is still running on it Jul 31 17:06:48 <_trine> ok I'm gonna give it a try Jul 31 17:07:40 <_trine> I could not see when I tried the code I compiled an IP address at all and I was sniffing and using a couple of computers Jul 31 17:07:55 <_trine> wierd Jul 31 17:10:04 is the 2.6 version for brcm devices stables? Jul 31 17:10:18 yes Jul 31 17:13:37 mm I should gave it a try maybe Jul 31 17:20:57 bigon: support for 11n gear is lacking Jul 31 17:21:08 (as far as I know, haven't checked in a while) Jul 31 17:28:50 questio about bcm6345! Jul 31 17:29:15 hi AndyI, go ahead Jul 31 17:29:51 speak russian? Jul 31 17:30:23 german english some french :) there is a guy from the ukraine here though, he might speak russian Jul 31 17:32:41 possible make work lan driver for bcm6345? Jul 31 17:33:41 AndyI: [florian] is working on it. he said it shouldn't be difficult. we have problems on bcm6345 though: right now, the system crashes after a minute, watchdog reboots it Jul 31 17:35:12 my system not crashed Jul 31 17:35:26 AndyI: !!! good, which one is it? Jul 31 17:35:39 AndyI: did you deactivate PCI? Jul 31 17:35:59 no Jul 31 17:36:25 AndyI: maybe the reason is i deactivated PCI. i will try. which router do you have? Jul 31 17:37:34 Paradyne 6211-A1-302 Jul 31 17:39:25 possible look curent florian patch? Jul 31 17:39:55 [florian]: did you read? you there? Jul 31 17:40:00 AndyI: i'm not sure he has one yet Jul 31 17:41:24 AndyI: we should wait for his response Jul 31 17:41:43 <[florian]> danage: yes I just read Jul 31 17:41:52 on bcm6345 gpio not work Jul 31 17:41:59 <[florian]> danage: I do not think PCI is related to your crashed Jul 31 17:42:04 <[florian]> AndyI: yes, thit is "normal" it is not correctly implemented yet Jul 31 17:42:58 [florian]: anyhow it's good news it's not crashing for andyl. i might have hardware breakage, possibly. i dropped the router on the floor once, in a disassebled state. i will try with the original firmware this weekend to double check. Jul 31 17:43:06 <[florian]> if swap GPIO_DATA_HI_REG & GPIO_DATA_LO_REG gpio work fine Jul 31 17:43:15 :) Jul 31 17:43:41 <[florian]> AndyI: ok, I will patch this Jul 31 17:43:54 <[florian]> danage: sounds like a good thing Jul 31 17:44:31 [florian]: yes it would suck trying to poke a software bug that is hardware breakage... Jul 31 17:44:53 <[florian]> danage: indeed Jul 31 17:44:59 [florian]: what's the status on ethernet, AndyI asked Jul 31 17:45:38 <[florian]> danage: for ethernet, well I am going to have a closer look this weekend about what needs changing Jul 31 17:46:01 <[florian]> danage: according to mbizon the driver needed is different Jul 31 17:46:20 What problems with watchdog? Jul 31 17:46:33 <[florian]> AndyI: we are not sure this is watchdog Jul 31 17:46:36 AndyI: no problems per se. my router crashes and watchdog reboots it Jul 31 17:46:52 <[florian]> AndyI: could you show me a bootlog of your router? Jul 31 17:47:07 AndyI: which image did you flash? Jul 31 17:47:13 in form Jul 31 17:47:20 in forum Jul 31 17:47:30 <[florian]> AndyI: where in the forum? Jul 31 17:47:41 [florian]: is there gpl sources of the driver that we need? Jul 31 17:47:50 waiting Jul 31 17:48:06 <[florian]> danage: we could manage to get it working with only the headers describing registers layout I guess Jul 31 17:48:12 https://forum.openwrt.org/viewtopic.php?pid=91749#p91749 Jul 31 17:48:23 <[florian]> AndyI: thanks! Jul 31 17:49:06 [florian]: ah he found the missing 0 too Jul 31 17:49:08 <[florian]> after eth driver load lan indicator turn on Jul 31 17:49:31 <[florian]> and nothing ;-) Jul 31 17:49:53 AndyI: which image file did you flash? Jul 31 17:50:26 dont understand (please write simple) Jul 31 17:50:36 <[florian]> AndyI: but ethernet does not work, right? Jul 31 17:50:42 juhosg, you still around ? Jul 31 17:50:52 I just did a full build on a fresh checkout from trunk Jul 31 17:51:01 configured it for wrt160nl Jul 31 17:51:04 built from scratch Jul 31 17:51:13 groz: yes, i'm here Jul 31 17:51:15 now, there is a problem with the flash Jul 31 17:51:26 Atheros AR71xx SPI Controller driver version 0.2.3 Jul 31 17:51:26 m25p80 spi0.0: m25p80 (1024 Kbytes) Jul 31 17:51:30 that's what i get now Jul 31 17:51:36 <[florian]> yes driver load fine and not work Jul 31 17:51:40 Atheros AR71xx SPI Controller driver version 0.2.2 Jul 31 17:51:40 m25p80 spi0.0: en25p64 (8192 Kbytes) Jul 31 17:51:45 that's what i get with my older one Jul 31 17:51:58 so, that driver update to .3 from .2 changed something Jul 31 17:52:06 just now gonna start looking for what Jul 31 17:52:08 weird Jul 31 17:52:11 but, you may be interested in that Jul 31 17:52:39 i'll track it down Jul 31 17:52:55 hehe, but it's always something isn't it :) Jul 31 17:52:56 AndyI: there are many files in /bin/ folder - which one? openwrt-96345GW2-squashfs-bc300-cfe.bin? Jul 31 17:54:24 i add description for my router and build image Jul 31 17:55:35 $(call Image/Build/CFE,$(1),R,6345,bc221,,bc221,-y 5) Jul 31 17:56:01 [florian]: do you think that could lead to different results than flashing the bc300-image? Jul 31 17:56:43 <[florian]> danage: yes it can Jul 31 17:57:25 groz: this change should not break anything: https://dev.openwrt.org/changeset/16767/trunk/target/linux/ar71xx/files/drivers/spi/ar71xx_spi.c Jul 31 17:57:52 <[florian]> why lan driver dont work? Jul 31 17:57:52 groz: can you try the driver without the change? Jul 31 17:58:07 <[florian]> AndyI: not the same driver is requied Jul 31 17:58:59 <[florian]> different hardware? Jul 31 17:59:34 <[florian]> AndyI: for ethernet, yes Jul 31 18:01:18 <[florian]> There is a possibility it to write? It will write? Jul 31 18:01:42 <[florian]> AndyI: I need to find some headers for it Jul 31 18:02:36 bbl Jul 31 18:05:16 <[florian]> What in them should be? Jul 31 18:05:48 <[florian]> AndyI: register description of the Ethernet MAC Jul 31 18:06:15 <[florian]> and PCI code Jul 31 18:20:07 <[florian]> http://download.francofon.fr/firmwarefon/fonera/sources/extracted/target/linux/brcm63xx-2.6/patches/001-bcm963xx.patch Jul 31 18:21:51 <[florian]> You looked it? Jul 31 18:23:28 <[florian]> AndyI: no, thanks for the link Jul 31 18:24:25 that looks like it is just some code from an old fon tarball Jul 31 18:24:45 <[florian]> that guy behing francofon.fr is just a jerk actually Jul 31 18:24:46 so it's a much older version of our code Jul 31 18:25:13 <[florian]> we never had all of the code he added Jul 31 18:26:19 <[florian]> Look my patches? Jul 31 18:27:10 <[florian]> AndyI: which patches? Jul 31 18:27:17 <[florian]> AndyI: on the forum? they are all in svn Jul 31 18:28:01 <[florian]> no new Jul 31 18:28:33 <[florian]> How to transfer? Jul 31 18:28:35 <[florian]> AndyI: that helps, thanks ! Jul 31 18:31:24 <[florian]> ? Jul 31 18:33:15 <[florian]> AndyI: it provides me with interesting information Jul 31 18:33:51 <[florian]> looking my new patches? Jul 31 18:35:46 <[florian]> AndyI: I am looking at this one: http://download.francofon.fr/firmwarefon/fonera/sources/extracted/target/linux/brcm63xx-2.6/patches/001-bcm963xx.patch Jul 31 18:37:22 <[florian]> How to transfer? my patches? Jul 31 18:38:00 <[florian]> AndyI: which patches and where, send them by mail Jul 31 18:38:18 AndyI: give patches :) Jul 31 18:38:36 <[florian]> address please? Jul 31 18:38:44 <[florian]> florian at openwrt dot org Jul 31 18:38:52 <[florian]> ok Jul 31 18:39:18 <[florian]> thanks ! Jul 31 18:40:14 <[florian]> sended! please look it! Jul 31 18:40:42 <[florian]> ok! Jul 31 18:40:42 <[florian]> about gpio? Jul 31 18:41:21 <[florian]> please write simple! (sorry my bad English)! Jul 31 18:41:25 <[florian]> AndyI: the offset for the GPIO registers is wrong for bcm6345 Jul 31 18:41:38 <[florian]> AndyI: that is why you have to nvert GpioHi/Lo Jul 31 18:41:40 <[florian]> yes Jul 31 18:42:15 <[florian]> after this change all led possible controlled! Jul 31 18:42:23 <[florian]> AndyI: ok Jul 31 18:43:06 [florian]: i'll be back tomorrow night, do a testflash of the original hardware and if the router is still fine, i'll setup remote console access for you, ok? Jul 31 18:43:22 <[florian]> danage: perfect Jul 31 18:43:37 ^i meant firmware up there, not hardware Jul 31 18:43:52 and: back home Jul 31 18:44:43 <[florian]> please look "setup pin multiplexing depending on board enabled device" in board_bcm963xx.c Jul 31 18:45:00 <[florian]> AndyI: ok Jul 31 18:46:46 <[florian]> possible save GPIO_MODE_REG before mask and dont write 0 for other hardware Jul 31 18:47:01 <[florian]> AndyI: of course Jul 31 18:47:52 <[florian]> your fix it? Jul 31 18:49:43 <[florian]> my uname -a CYGWIN_NT-5.2 server00 1.5.25(0.156/4/2) 2008-06-12 19:34 i686 Cygwin ;-) Jul 31 18:50:03 <[florian]> it`s normal? Jul 31 18:53:08 <[florian]> bcm6345 dont have spi Jul 31 18:55:32 <[florian]> AndyI: I did not fix this yes Jul 31 18:55:41 <[florian]> AndyI: I know bcm6345 does not have SPI, why do you ask? Jul 31 18:56:11 <[florian]> possible dont register SPI Jul 31 18:56:22 <[florian]> AndyI: what does it change? Jul 31 18:56:42 <[florian]> AndyI: the platform_device is filled with wrong resource informations, so the driver is not registered Jul 31 18:56:51 <[florian]> AndyI: I still did not receive your email Jul 31 18:58:08 <[florian]> write message for my pasha_inter at online dot magadan dot su Jul 31 18:59:25 <[florian]> AndyI: done Jul 31 19:02:57 <[florian]> AndyI: got them, it was greylisted Jul 31 19:03:54 <[florian]> ok Jul 31 19:04:56 <[florian]> About SPI it is simple for appearance! Jul 31 19:14:27 <[florian]> http://www.winline.com.cn/en/files/3.pdf Jul 31 19:16:39 <[florian]> AndyI: interesting :) Jul 31 19:17:40 <[florian]> http://halloserv.de/~ago/bcm63xxdrivers.tbz Jul 31 19:18:15 <[florian]> "interesting" about? Jul 31 19:29:15 <[florian]> AndyI: it is the same SPI core, and probably the same Ethernet core as well Jul 31 19:30:37 <[florian]> dont understand ;-( Jul 31 19:30:56 <[florian]> AndyI: there are the exact same SPI and Ethernet in BCM6345 Jul 31 19:32:01 <[florian]> where? Jul 31 19:32:13 <[florian]> AndyI: in the pdf you sent me Jul 31 19:32:34 <[florian]> understand! Jul 31 19:33:12 <[florian]> look my patches? Jul 31 19:33:50 <[florian]> AndyI: yes, they break 6338/48/58 due to change in EI_SENSE_SHFT etc ... Jul 31 19:34:44 <[florian]> AndyI: these are just comments, it does not hurt ;) Jul 31 19:35:54 <[florian]> it`s add for information for future use Jul 31 19:36:08 <[florian]> AndyI: ok, but I will do it differently Jul 31 19:37:41 <[florian]> When to wait for changes in SVN? Jul 31 19:38:30 <[florian]> AndyI: 10 minutes Jul 31 19:38:53 <[florian]> i write some patches for add CYGWIN support Jul 31 19:39:08 <[florian]> AndyI: ok, send them to me as well :p Jul 31 19:39:18 florian, I already have a great patch for cygwin support Jul 31 19:39:30 its really simple to Jul 31 19:39:35 apt-get install ubuntu Jul 31 19:39:37 <[florian]> groz: oh, well, you got commit access, go ahead ;) Jul 31 19:39:39 :) Jul 31 19:39:39 <[florian]> groz: lol Jul 31 19:40:11 * groz had a long day Jul 31 19:40:20 ;-) Jul 31 19:45:19 all patch fix simple error Jul 31 19:57:56 <[florian]> AndyI: you sent it? Jul 31 19:58:35 <[florian]> AndyI: ok, got the cygwin patches, got to test them first :p Jul 31 19:59:44 <[florian]> AndyI: did quilt compile for you ? Jul 31 20:04:53 <[florian]> explain Jul 31 20:05:18 <[florian]> AndyI: quilt is a tool that openwrt builds, did everything compile fine for you? Jul 31 20:06:29 <[florian]> quilt compile with patch Jul 31 20:06:44 <[florian]> AndyI: where is the patch for quilt? Jul 31 20:07:26 <[florian]> please wait! Jul 31 20:08:14 <[florian]> AndyI: ok :p Jul 31 20:10:23 <[florian]> http://pastebin.com/d6586deee Jul 31 20:11:02 <[florian]> Whence you have learnt it? Jul 31 20:11:54 <[florian]> AndyI: learn what? Jul 31 20:12:29 <[florian]> quilt Jul 31 20:12:59 <[florian]> AndyI: I already tried building under cygwin :p Jul 31 20:13:04 <[florian]> and it failed with quilt Jul 31 20:13:08 <[florian]> and for mtd-utils Jul 31 20:14:24 florian * r17065 /packages/utils/pcsc-lite/ (. Makefile files/ files/pcscd.init): [package] add pcsc-lite (#5603) Jul 31 20:15:27 <[florian]> ok! Jul 31 20:15:45 <[florian]> AndyI: do you have more patches for cygwin? Jul 31 20:15:45 florian * r17066 /trunk/target/linux/brcm63xx/files/ (5 files in 2 dirs): [brcm63xx] more bcm6345 fixes and definitions, thanks to AndyI Jul 31 20:15:58 <[florian]> yes Jul 31 20:16:19 <[florian]> AndyI: ok Jul 31 20:21:19 florian * r17067 /packages/utils/ccid/ (. Makefile): [package] add ccid (#5604) Jul 31 20:22:45 juhosg, you may wanna take a peek here Jul 31 20:22:48 http://openwrt.pastebin.ca/1514135 Jul 31 20:22:55 that's the one we've been waiting for Jul 31 20:23:07 wrt160nl booted from squashfs, all hardware working Jul 31 20:24:06 all i got left to do now, figure out why i'm not getting the mini-fo on the jffs, and that's probably to do with the jffs puking hundreds and hudreds of errors regarding 'dont like these sectors' Jul 31 20:24:46 <[florian]> groz: even 3g? Jul 31 20:25:06 drivers are loaded, device is detected Jul 31 20:25:14 not in this boot, but was in the prior one Jul 31 20:25:19 it's unplugged right now Jul 31 20:25:45 <[florian]> interesting Jul 31 20:26:21 hmm, well, what did I do now Jul 31 20:26:29 haha, plugged it in, and it's not recognized now Jul 31 20:26:49 i'll have to recheck all the modules, cuz, it was working earlier, before I rebuilt from a fresh checkout on trunk Jul 31 20:27:19 <[florian]> that's just awesome Jul 31 20:27:24 <[florian]> how cheap is the wrt160nl? Jul 31 20:27:42 they run about $100 around here, but, I expect them to be cheaper before long Jul 31 20:28:01 linksys will keep a premium on them over the wrt54gl, as long as the gl remains a good seller Jul 31 20:28:02 <[florian]> a new wrt54g in perspective ;) Jul 31 20:28:19 well, i actually had a conversation on monday, with a linksys rep Jul 31 20:28:27 err, actually, cisco rep Jul 31 20:28:30 and what I was told Jul 31 20:28:42 wrt54gl is by far thier best selling single router ever Jul 31 20:28:51 it's not on the end of life roadmap yet Jul 31 20:29:05 but, the wrt160nl is currently on the roadmap as a long life, gpl product Jul 31 20:29:12 seen not as a replacement for the gl Jul 31 20:29:16 but as an upgrade from the gl Jul 31 20:29:17 <[florian]> well now that they understood why its such a bestseller, no need to have it replaced Jul 31 20:29:28 oh, here's the thing Jul 31 20:29:30 but no gigabit switch Jul 31 20:29:45 they brought a load into the us a month ago, 40K pieces Jul 31 20:29:57 now they have a supply problem, and next month delivery gonna be late, on the gl Jul 31 20:30:02 they have a shortage on one chip for it Jul 31 20:30:12 the 40K inventory in usa, is already sold out Jul 31 20:30:35 personally, i see no need for a gigabit switch on this little router Jul 31 20:30:47 simply because, it's not up to processing gigabit anyways Jul 31 20:30:54 and as an edge device on a typical network connection Jul 31 20:31:01 the 100mbit is more than adaquate Jul 31 20:31:12 i'll have it running here with one port plugged into my gigabit stuff Jul 31 20:31:23 for the one in the house Jul 31 20:31:32 the one in the van, heck, dont even need ethernet at all Jul 31 20:31:41 3g modem for internet, and wifi for the netbooks Jul 31 20:31:50 my private wifi hotspot, to go Jul 31 20:32:32 <[florian]> really interesting indeed Jul 31 20:32:52 florian * r17068 /packages/libs/libnfc/ (. Makefile): [package] add libnfc (#5605) Jul 31 20:34:34