**** BEGIN LOGGING AT Sun Mar 07 02:59:58 2010 Mar 07 03:41:58 nico * r20024 /trunk/package/siit/Makefile: [package] siit: remove obsolete dep (builds fine on 2.6.31 since r18191) Mar 07 03:44:50 nico * r20025 /trunk/package/siit/Makefile: [package] siit: use $(KERNEL_MAKEOPTS) Mar 07 03:55:36 nico * r20026 /trunk/package/siit/Makefile: [package] siit: oops, 'make M=... modules' is 2.6 only, use SUBDIRS=... instead Mar 07 05:40:57 juhosg * r20027 /trunk/target/linux/ar71xx/ (6 files in 2 dirs): ar71xx: rename mach-rb-4xx.c to mach-rb4xx.c, and change Kconfig symbol Mar 07 05:41:00 juhosg * r20028 /trunk/target/linux/ar71xx/ (2 files in 2 dirs): ar71xx: make the rb4xx nand driver depends on AR71XX_MACH_RB4XX Mar 07 05:41:03 juhosg * r20029 /trunk/target/linux/ar71xx/ (9 files in 4 dirs): ar71xx: add preliminary support for the RouterBOARD 750 Mar 07 07:22:00 build #22 of ramips is complete: Failure [failed compile_6] Build details are at http://tksite.gotdns.org:8010/builders/ramips/builds/22 Mar 07 07:48:53 <_trine> cshore, ping Mar 07 07:49:25 ping Trine....not here long, going to bed in about 10-20 minutes Mar 07 07:49:32 <_trine> ok Mar 07 07:50:12 <_trine> I just wanted to mention I'm having problem with the pivit root script on my wndr3700 Mar 07 07:51:33 <_trine> it does not run correctly and after the router boots to pivot root the green power leds flashes indefinitely although the router does continue to work in a fashion Mar 07 07:52:07 so it never fully boots? Mar 07 07:52:23 <_trine> we can dicus this another time if you are about to go to bed it's not important Mar 07 07:52:32 <_trine> it does boot Mar 07 07:52:58 <_trine> but it does not work as it should I will need to do more tests when you are around Mar 07 07:53:43 <_trine> the way I do it at the moment works perfectly with the other script so we need to find out what's happening with your script Mar 07 07:54:33 <_trine> one thought that crassed my mind was does your script require a clear flash drive from the beginning Mar 07 07:54:38 <_trine> crossed* Mar 07 07:54:54 <_trine> I would not have thought it did Mar 07 07:55:08 it should not have /bin /usr/sbin /lib etc Mar 07 07:55:32 <_trine> don't worry about it now if your off to bed there is no rush Mar 07 07:55:54 <_trine> this can wait till you have had your sleep Mar 07 07:57:16 <_trine> the problem is probably likely to be that i am using the same flash drive with both scripts Mar 07 07:57:25 <_trine> we'll see later Mar 07 11:34:13 hi, is there somone with MPC8377 experience ? Mar 07 12:53:03 Hi, i need some help.I'm a brazillian student,and i'm couln't find which 'rate control algorithm" openwrt uses. Mar 07 12:53:15 I've searched everywhere,I looked at /dl directory in the "wireless_tools.29" but i could't find it. Mar 07 12:53:18 Tiago: which wifi hw is it ? Mar 07 12:54:16 sorry,hw? Mar 07 12:54:45 kamikaze 8.09 Mar 07 12:55:09 what wifi chip Mar 07 12:55:15 broadcom, atheros ..... Mar 07 12:56:07 broadcom Mar 07 12:56:26 2.4 or 2.6 kenrel ? Mar 07 12:56:41 2.4 uses the broadcom algorithm Mar 07 12:56:47 2.6 uses the b43 algo Mar 07 12:56:56 not sure what b43 uses though :) Mar 07 12:57:22 2.4 Mar 07 12:57:29 thanks a lot.... Mar 07 12:57:49 Tiago: it is a binary only driver from broadcom Mar 07 12:58:49 ok,is there any possible to implement a open source algorithm to openwrt? Mar 07 12:59:28 there are such things Mar 07 12:59:33 but not for that hardware Mar 07 15:06:24 nico * r20030 /trunk/toolchain/uClibc/ (4 files in 2 dirs): [toolchain] uClibc: revert to old linuxthreads implementation on x86 (like all others arches) Mar 07 15:10:59 nico * r20031 /trunk/toolchain/uClibc/ (4 files in 4 dirs): [toolchain] uClibc: enhance debug support (closes: #6118) Mar 07 16:29:30 nbd * r20032 /trunk/package/mac80211/ (18 files in 2 dirs): mac80211: update to wireless-testing 2010-03-03 Mar 07 16:29:34 nbd * r20033 /trunk/package/mac80211/patches/560-minstrel_ht.patch: mac80211: implement .rate_update in minstrel_ht Mar 07 16:29:38 nbd * r20034 /trunk/package/mac80211/patches/560-minstrel_ht.patch: mac80211: increase minstrel_ht precision by properly using a-mpdu frame stats Mar 07 16:29:43 nbd * r20035 /trunk/package/mac80211/patches/560-minstrel_ht.patch: mac80211: simplify minstrel_ht code, rewrite packet sampling algorithm for improved performance and a-mpdu length Mar 07 16:36:53 florian * r20036 /trunk/package/acx-mac80211/patches/005-do_not_override_pci.patch: [ar7] add one more acx-mac80211 patch to finally allow its building on ar7 Mar 07 16:37:03 florian * r20037 /trunk/target/linux/ar7/ (2 files in 2 dirs): [ar7] correct the MDIO base address and reset sequence for Titan (#6573) Mar 07 16:50:10 got a missing symbol on ar71xx/ubnt-rspro: MikroTik RouterBOARD 4xx series support (AR71XX_MACH_RB_4XX) [N/y] (NEW) Y Mar 07 16:50:41 try make taget/linux/clean world Mar 07 17:00:45 nbd * r20038 /trunk/package/mac80211/patches/560-minstrel_ht.patch: mac80211: improve minstrel_ht performance by reducing the impact of slow sampling Mar 07 17:59:44 nbd: ping Mar 07 18:01:54 rtz2: pong Mar 07 18:02:51 nbd: I have a (not yet working) patch for the wrt160nl boot problem and I will get a unit in a few days to test it on Mar 07 18:03:16 nbd: somebody told me, he gave you a dump of the new u-boot Mar 07 18:03:20 yes Mar 07 18:03:22 nbd: could you give it to me? Mar 07 18:03:24 sure Mar 07 18:03:28 can you show me your patch? Mar 07 18:03:43 one moment Mar 07 18:03:48 did you change the image to exclude the jffs2 eof marker from the trx area? Mar 07 18:05:43 I excluded the whole rootfs from it, otherwise it would need different handling for jffs2 and squashfs images Mar 07 18:05:47 nbd: http://openwrt.pastebin.com/0UkwyajM Mar 07 18:06:12 nbd: it also includes a new image generation patch for the rdc, which is already working Mar 07 18:06:35 nbd: but there are some problems with the wrt160nl, not sure what they are yet Mar 07 18:06:40 rtz2: one minor thing, can you give your "prepare_image" script a less generic name? Mar 07 18:07:01 xMff: padd_image? Mar 07 18:07:18 if it is arch specific then maybe ar71xx_pad_image Mar 07 18:07:25 xMff: it isn't Mar 07 18:07:28 ok Mar 07 18:07:57 is that script even necessary? Mar 07 18:08:00 and it also adds the jffs2 start markers to the rootfs, so it's not only pading Mar 07 18:08:02 i mean the rootfs image is already padded Mar 07 18:08:05 and it needs to be aligned anyway Mar 07 18:09:58 nbd: why does it have to be aligned? Mar 07 18:10:12 nbd: for squashfs, it shouldn't matter Mar 07 18:10:39 the rootfs start is looked up by the same code for squashfs and jffs2 Mar 07 18:10:41 and we support both Mar 07 18:10:50 so squashfs needs be aligned the same way Mar 07 18:12:52 nbd: well, there are 2 lenght fields in the header one could use to store the start of the rootfs Mar 07 18:13:38 nbd: for jffs2, the kernel image gets padded to an erease block, for squashfs, it doesn't Mar 07 18:14:19 we typically don't make a per-filesystem distinction for the image container stuff Mar 07 18:15:53 nbd: well, the script does it automatically and the mtd partition code wouldn't have to care Mar 07 18:16:11 florian * r20039 /trunk/package/ar7-atm/ (34 files in 4 dirs): [ar7] add support for new ar7-atm version D7.05.00.00 (#6515) Mar 07 18:16:13 nbd: I could change it, but I don't see the problem Mar 07 18:17:02 i'd prefer to not have padding rely on the filename Mar 07 18:17:31 e.g. on some platforms we strip the -64k or -128k from the filename if only one of those is possible Mar 07 18:17:47 oh, never mind Mar 07 18:17:51 the script uses the raw names Mar 07 18:17:57 nbd: right Mar 07 18:18:10 nbd: would it be ok that way? Mar 07 18:19:16 well, i don't see the point in adding the script and bypassing parts of the code that do basically the same as what the new script does Mar 07 18:19:39 your patch should work as-is if you remove the invocation of the script and drop the .orig suffix from the rootfs image Mar 07 18:19:42 imho Mar 07 18:20:05 the original code didn't work for some reason, let me think about it Mar 07 18:23:15 ahh, right Mar 07 18:24:23 nbd: I originally wanted to include the rootfs image in the trx file, but this didn't work with the markers already added Mar 07 18:24:34 but that was before I thought of the jffs2 image Mar 07 18:26:48 nbd: I will have to think about this Mar 07 18:27:02 nbd: could you mail me the bootloader? Mar 07 18:35:43 nbd * r20040 /trunk/package/mac80211/patches/560-minstrel_ht.patch: mac80211: fix a few more minstrel_ht bugs (thx to Christian Lamparter) Mar 07 18:35:54 yeah, just a sec Mar 07 18:36:42 http://nbd.name/uboot.bin.hardhatpat Mar 07 18:37:50 nbd: You don't have permission to access /uboot.bin.hardhatpat on this server. Mar 07 18:38:27 fixed Mar 07 18:38:56 nbd: ok, thanks Mar 07 18:43:52 florian * r20041 /packages/net/samba3/patches/100-avm.patch: [packages] fix samba3 logging in nmbd when CONFIG_SAMBA3_DEBUG is enabled (#6389) Mar 07 18:47:24 florian * r20042 /packages/net/samba3/patches/100-avm.patch: [packages] fix the logging file for nmbd Mar 07 22:23:37 ping nbd Mar 07 22:29:39 I got glibc compiling, and many programs compile, however there are 3 (so far) that I've had to #define around some duplicate symbols - all to do with option parsing! Mar 07 22:29:50 .. lzo, git and.. umm. one other Mar 07 22:30:14 (I remembered to svn add the patches anywy) Mar 07 22:41:47 exit Mar 07 22:42:03 sorry wrong keyboard Mar 07 23:05:35 cshore: yay for keyboard switchers :) Mar 07 23:05:49 Yeah, I have to buy one Mar 07 23:06:03 cshore: I only have one keyboard. 3 monitors. middle one switches between the two systems Mar 07 23:06:21 (so each box has 2 monitors effectively) Mar 07 23:07:25 cshore: I have a nice little usb one - is a 2 usb port hub, which it switches between the 2 systems + graphics + sound (not used) Mar 07 23:09:08 frogonwheels: I haven't decided how I'm going to configure the systems yet. I'm thinking of making one computer the main one and use the other one headless, except occasionally Mar 07 23:31:40 cshore: ah. I have work windoze box + home linux box Mar 07 23:32:57 I'm all Linux except a VM for things that require Windows...except I recently got a laptop that I'm going to have a small partition with Windows on (dual boot), so that I have Winblows if I need it Mar 07 23:33:27 on physical hardware I mean Mar 07 23:45:00 lars * r20043 /trunk/target/linux/xburst/files-2.6.32/ (arch/mips/jz4740/clock.c sound/soc/jz4740/jz4740-i2s.c): [xburst] Add support for enabling/disabling the aic controllers clock. Mar 07 23:45:47 lars * r20044 /trunk/target/linux/xburst/files-2.6.32/sound/soc/jz4740/ (Kconfig Makefile n516.c): [xburst] Add sound support for the n516. Mar 07 23:51:40 nbd: ping Mar 08 00:18:08 build #26 of brcm_2_4 is complete: Success [build successful] Build details are at http://tksite.gotdns.org:8010/builders/brcm_2_4/builds/26 Mar 08 00:18:14 build #23 of ifxmips is complete: Failure [failed svn shell_19 compile_15] Build details are at http://tksite.gotdns.org:8010/builders/ifxmips/builds/23 Mar 08 00:19:30 build #22 of pxcab is complete: Success [build successful] Build details are at http://tksite.gotdns.org:8010/builders/pxcab/builds/22 Mar 08 00:24:42 rtz2: pong Mar 08 00:31:25 nbd: I did think about the flash generation problem Mar 08 00:32:22 basically, you can't use a generic squashfs image for all boards if you want to avoid the padding Mar 08 00:32:36 because you don't know, where to place the jffs2 start marks Mar 08 00:33:09 I can't decide, if this is worth the effort, but imho avoiding the wast of 64k flash is Mar 08 00:33:26 and this also applys to a bunch of other ar71xx routers Mar 08 00:34:13 i think we should keep it the way it is Mar 08 00:34:19 aside from moving the rootfs out of the trx of course Mar 08 00:34:28 nbd * r20045 /trunk/package/mac80211/patches/560-minstrel_ht.patch: mac80211: small whitespace fix in minstrel_ht Mar 08 00:34:32 nbd * r20046 /trunk/package/mac80211/patches/120-linux-2.6.30-compat.patch: mac80211: remove an obsolete patch which breaks 2.6.32 compat Mar 08 00:36:07 nbd: sorry, but I have to ask why? Mar 08 00:36:34 it works Mar 08 00:36:41 and there are bigger sources of waste to be eliminated first Mar 08 00:37:47 e.g. hardcoded kernel image sizes, gzipped kernel images instead of lzma Mar 08 00:37:49 stuff like that Mar 08 00:40:26 well, I meanted the targets with the hardcoded kernel images when I said other targets which could use this Mar 08 00:41:38 did mean Mar 08 00:43:51 moving the squashfs image right after the kernel without padding doesn't always save 64k Mar 08 00:44:06 in fact quite often it probably wont Mar 08 00:44:41 i mean in cases where it's dynamic Mar 08 00:46:36 i need some sleep now Mar 08 00:46:50 nbd: ok, I admit the padding for the wrt160nl is done in a pretty stupid way :/ Mar 08 00:46:55 nbd: ok, good night Mar 08 00:47:58 yes, for the wrt160nl we can probably save an eraseblock Mar 08 00:48:27 we just need a better padding utility than dd Mar 08 00:48:54 which can pad to (blocksize - offset) Mar 08 00:50:12 well, that was what I tried to create, but I maybe I lumped too much functionality into it Mar 08 01:55:34 nico * r20047 /trunk/package/acx/Makefile: [package] acx: remove wrong CONFIG_PCI dep (should have been PCI_SUPPORT, but that just fails on some targets) **** ENDING LOGGING AT Mon Mar 08 02:59:57 2010