**** BEGIN LOGGING AT Sat Jun 06 02:59:59 2009 Jun 06 03:04:47 russell_: okay, with the last 0x6000 bytes ommitted, it should work Jun 06 03:04:58 0x20000 - 0x6000 - 0x2000 = 0x18000 Jun 06 03:06:04 * russell_ just sitting here thinking about why there are two sections and how CFE deals with that Jun 06 03:06:35 it seems like CFE *may* only be looking at 0x1e000 Jun 06 03:06:46 yep Jun 06 03:06:55 printenv show stuff that doesn't appear there, but it might be cooked in Jun 06 03:06:58 the other thing seems to ba kind of compat-stub Jun 06 03:08:05 i was paranoid in the early days with these things and dd'd a bunch of images off these things when they were fresh out of the box, so the bit at 0x18000 *was* there originally. Jun 06 03:08:44 http://openwrt.pastebin.com/m15ac5f6 Jun 06 03:11:41 i don't have a /dev/mtdblock/ directory, btw Jun 06 03:12:27 it first checks for /dev/mtdblock/N, then for /dev/mtdblockN Jun 06 03:12:34 one is 2.4, the other .26 Jun 06 03:13:01 ah, yeah, i was just going to look at the file ... since clearly the patch didn't show that part. Jun 06 03:14:00 i'm wondering what the utility of the nvram utility would be on a wgt634u Jun 06 03:14:07 * xMff too Jun 06 03:14:22 but it should make no damage at least Jun 06 03:14:35 yeah, reading would be interesting Jun 06 03:14:39 so I'm interested in fixing this, even if it has no use Jun 06 03:14:54 i'm wondering if, in that case, it should read and report what's in the 0x1e000 section Jun 06 03:15:17 then it would neet a complete separate code patch for that Jun 06 03:15:20 *need Jun 06 03:15:21 argh Jun 06 03:15:23 yeah Jun 06 03:15:31 hardly urgent, though Jun 06 03:16:12 the non-corrupting is important though, since it'll bork people's devices Jun 06 03:16:53 right now the init script will skip it's stuff if it detects the wgt Jun 06 03:16:59 right Jun 06 03:20:27 jow * r16350 /branches/8.09/package/nvram-brcm47xx/files/nvram.init: merge r16349 to 8.09 Jun 06 03:29:06 russell_: ping me if the patch works for you, I'll go to bed now Jun 06 03:29:22 okay Jun 06 03:29:52 and THANKYOU for your work on it Jun 06 03:30:08 np Jun 06 04:43:56 juhosg * r16351 /trunk/target/linux/ar71xx/ (6 files in 3 dirs): [ar71xx] make the AP83 and VSC7385 SPI drivers available on 2.6.2[89] Jun 06 06:32:42 juhosg * r16352 /trunk/target/linux/ar71xx/ (6 files in 3 dirs): [ar71xx] sync kernel config, and refresh patches Jun 06 07:59:39 * russell_ seeing some usb errors on wgt634u Jun 06 07:59:49 usb 4-2: device descriptor read/64, error -62 Jun 06 08:00:08 hub 4-0:1.0: unable to enumerate USB device on port 2 Jun 06 08:18:58 russell_: that started when brcm47xx was moved to 2.6.28 Jun 06 08:29:20 q Jun 06 08:29:24 oops Jun 06 08:41:48 juhosg * r16353 /trunk/target/linux/ar71xx/ (10 files in 7 dirs): [ar71xx] add parallel flash driver Jun 06 08:54:32 juhosg * r16354 /trunk/target/linux/ar71xx/ (3 files in 3 dirs): [ar71xx] fix erase status check on some parallel flash chips Jun 06 08:55:08 juhosg * r16355 /trunk/target/linux/ar71xx/ (config-2.6.28 config-2.6.29 config-2.6.30): [ar71xx] enable JEDEC probe Jun 06 09:57:39 juhosg * r16356 /trunk/target/linux/ar71xx/files/arch/mips/ar71xx/mach-ap83.c: [ar71xx] change flash partition map of the AP83 board Jun 06 11:20:53 seems there's a qos dependancy missing on trunk: iptables v1.4.3.2: Couldn't load match `layer7':File not found Jun 06 11:20:57 http://pastebin.com/m3c0584d6 Jun 06 13:02:27 juhosg * r16357 /trunk/target/linux/ar71xx/image/Makefile: [ar71xx] build images for the AP83 board Jun 06 13:48:40 juhosg * r16358 /trunk/target/linux/ar71xx/files/arch/mips/ar71xx/irq.c: [ar71xx] handle PCI_CORE interrupt as well Jun 06 14:53:41 russell_: did you had a chance to test the patch? Jun 06 15:01:42 Hello xMff, do you think a lua patch to add pack/unpack functions like perl can be good idea? Jun 06 15:02:14 sure, if someone provides a patch :) Jun 06 15:02:21 Ok Jun 06 15:02:26 I have one Jun 06 15:02:53 I'm testing it and look like working, I will paste for you Jun 06 15:07:41 xMff, there you go http://openwrt.pastebin.com/d3763b66 Jun 06 15:09:15 I guess now I can done the challenge proccess of chillispot Jun 06 15:11:32 k, will take a look later Jun 06 15:11:45 thanks Jun 06 15:56:05 juhosg * r16359 /trunk/target/linux/ar71xx/files/ (5 files in 3 dirs): [ar71xx] update flash locking code Jun 06 16:01:10 juhosg * r16360 /trunk/target/linux/ar71xx/base-files/ (etc/defconfig/ap83/ etc/defconfig/ap83/network lib/ar71xx.sh): [ar71xx] add default network configuration for the AP83 board Jun 06 16:40:56 jow * r16361 /packages/utils/collectd/Makefile: [packages] collectd: link against libm to prevent unresolved references to __isnan (#5241) Jun 06 17:12:20 jow * r16362 /packages/utils/collectd/Makefile: [packages] collectd: add build- and package-dependency to zlib (#4983) Jun 06 17:17:38 jow * r16363 /packages/utils/collectd/Makefile: [packages] collectd: remove redundant build dependencies Jun 06 17:44:50 juhosg * r16364 /trunk/target/linux/ar71xx/base-files/lib/upgrade/ (. platform.sh): [ar71xx] add experimental sysupgrade support Jun 06 18:10:26 jow * r16365 /packages/libs/axtls/Makefile: [packages] axtls: Capitalize axtls' SSL submenu (#5295) Jun 06 18:24:25 xMff:ping Jun 06 18:26:26 xMff: I have issues to handle charater "\0" in lua can you giveme a tip? Jun 06 18:27:14 what issues? Jun 06 18:29:32 i guess It cut string when i use it like param Jun 06 18:30:08 that's a bit unspecific Jun 06 18:33:51 ok, when i pack something myvar = pack(...) some charaters are "\0", in myvar I check the len of myvar and I have the right len Jun 06 18:34:37 but if I use myvar like param .... somefunction(myvar) Jun 06 18:34:59 in somefunction myvar is truncated Jun 06 18:37:02 so length(myvar) is different within your function? Jun 06 18:38:02 keep in mind that print() etc. will truncate the string on output Jun 06 18:38:12 $ lua -e 'print("test1 \000 test2")' Jun 06 18:38:13 test1 Jun 06 18:38:29 yes, that is the issue Jun 06 18:38:49 so you should bring it in a format the shell or whatever can handle Jun 06 18:38:54 i did check and the length is equal Jun 06 18:38:56 like hex-encoded Jun 06 18:39:26 but like i guess md5 is not working well I trying to use Jun 06 18:39:28 use string.byte() to get the byte values Jun 06 18:39:58 you use something that can't cope with \0 you should avoid that Jun 06 18:40:31 md5sum of shell and when i do io.popen("echo "..myvar.." | md5sum") Jun 06 18:40:38 yeah of course Jun 06 18:40:42 that'll never work Jun 06 18:40:52 in this case cut myvar Jun 06 18:40:55 also it's really unsafe code Jun 06 18:41:20 try using the %q pattern for string.format to pass stuff to the shell Jun 06 18:41:42 $ lua -e 'print(string.format("%q", "test1 \000 test2"))' Jun 06 18:41:42 "test1 \000 test2" Jun 06 18:42:26 yes, but I want check the values tha md5 give me becaus I guess md5 lib is working bad Jun 06 18:42:35 %q will escape special bytes and enclose the date in double quotes Jun 06 18:43:30 even better is this: Jun 06 18:43:30 fd=io.open("/tmp/dump", "w") fd:write(myvar) fd:close() os.execute("md5sum /tmp/dump") Jun 06 18:44:12 but if I use %q md5sum maybe assume all like character Jun 06 18:44:14 Oh! Jun 06 18:44:18 yeha Jun 06 18:44:24 that is good Idea Jun 06 18:44:44 I will do that to check md5sum and md5 lib Jun 06 18:47:25 thanks xMff Jun 06 18:57:31 xMff, other question... md5sum suposse give me the same value in all systems right? Jun 06 18:59:05 because I'm using the same values in router and in some php script to compare and md5 give me different values Jun 06 19:03:31 juhosg * r16366 /trunk/target/linux/ar71xx/files/arch/mips/ar71xx/mach-ap81.c: [ar71xx] setup mac address on the AP81 board Jun 06 19:04:14 forget md5 of php and shell givme the same value Jun 06 20:26:59 xMff: after find the issue, I did find I not need pack function I can use string.char(...) to get good result Jun 06 23:01:33 xMff: ping Jun 06 23:02:20 xMff: xor is inplemented in LUA or I need write one? Jun 06 23:15:51 is there a reason samba2 is still packaged other than history? Jun 07 01:19:06 swalker: size Jun 07 01:21:13 (monitor mode enabled on mon0) Jun 07 01:21:23 whats the best way to grab the interface? Jun 07 01:27:57 thepeople_mobile: I don't think that's an issue anymore considering the combined samba3 package is ~7k smaller than just the samba2-server package Jun 07 01:28:23 lol Jun 07 01:28:27 nice Jun 07 01:30:33 thepeople_mobile: the numbers are different for brcm* rather than x86 but the combined samba3 package is still smaller than the three samba2 packages combined **** ENDING LOGGING AT Sun Jun 07 02:59:57 2009