**** BEGIN LOGGING AT Tue Dec 28 02:59:56 2010 Dec 28 03:00:19 cshore_: Do you know how will I be able to acces the content of this chat tomorrow from any web browser? Dec 28 03:01:56 chsore_: Now, it just finished the compilation. The packaging fails to locate mod_fax.so file. It looks like the compilation did not compile mod_fax even though it is included. :( Dec 28 03:02:32 mazilo: http://logs.nslu2-linux.org/livelogs/openwrt-devel/ Dec 28 03:02:48 I think mod_fax is marked as broken Dec 28 03:02:57 for OpenWRT Dec 28 03:03:08 I know perl was Dec 28 03:04:16 cshore_: Yes and a lot of other modules are also marked broken. But, I tried to fix them. So, I demark some of them, i.e. mod_fax, mod_perl, mod_spandsp, etc. Dec 28 03:11:06 cshore_: on my OpenWRT Maefile for FS git, I have this entry '$(eval $(call BuildPlugin,fax,Fax,mod_fax,,+libjpeg @(!(powerpc)||BROKEN)))' for mod_fax and it shows on the 'make menuconfi'. and I select it to build as a module, but it doesn't build. If it is marked broken, it won't show on 'make menuconfig', right? Dec 28 03:11:48 cshore_: I am trying to build FS git for ARM platform (a Seagate DockStar). Dec 28 03:16:44 mazilo: if it's marked broken it shouldn't show on menuconfig, unless you have selected show broken packages in menuconfig Dec 28 03:17:33 mazilo: fax is supposedly not broken though, you're right Dec 28 03:18:09 it builds for me though, so maybe it's and ARM thing Dec 28 03:18:42 oh wait, I haven't tried since the libtool changes Dec 28 03:18:45 maybe it's that Dec 28 03:18:58 cshore_: No, the 'show broken platforms / package' is not checked. Dec 28 03:19:19 cshore_: Perhaps, it is the ARM thing. ;) Dec 28 03:20:18 cshore_: On Fs-1.0.6. ti won't compile mod_dingaling even though it is checked. OTOH, the mod_dingaling gets compiled for FS git without a problem. Dec 28 03:20:33 cshore_: These are on ARM. Dec 28 03:20:42 cshore_: And, I have no idea why. Dec 28 03:24:23 OK. I call the night and see you all tomorrow. Thank you, cshore_, for your helps and guidance. Dec 28 04:05:29 cshore * r24841 /packages/mail/nullmailer/Makefile: [packages] mail/nullmailer: Fixed sendmail symlink creation to use shell variable for postinst script not Makefile variable Dec 28 04:07:41 cshore * r24842 /trunk/target/linux/brcm63xx/patches-2.6.35/040-bcm963xx_flashmap.patch: Dec 28 04:07:41 [brcm63xx] patches-2.6.36: Fix flashmap spare length calculation from previous Dec 28 04:07:41 commit (for rootfs and rootfs_data mtd partitions). Apparently the interruption Dec 28 04:07:41 in working on it resulted in an incorrect memory of getting it working. **** BEGIN LOGGING AT Tue Dec 28 04:44:34 2010 Dec 28 08:13:11 cshore * r24843 /trunk/target/linux/brcm63xx/patches-2.6.35/040-bcm963xx_flashmap.patch: Dec 28 08:13:11 [brcm63xx] patches-2.6.35: mtd flashmap: Really fixed sparelen and rootfslen. Dec 28 08:13:11 The real rootfslen for OpenWRT images comes from a bigendian uint32_t in Dec 28 08:13:11 reserved1 of the bcm_tag, which is now used to correctly calculate the rootfslen Dec 28 08:13:11 (and thus the sparelen). Dec 28 09:02:10 {Nico}, are you around? Dec 28 09:02:29 I have a question on libgphoto's package in openwrt Dec 28 10:49:37 <[florian]> cshore_: pong Dec 28 10:49:53 [florian]: forgot Dec 28 10:49:57 <[florian]> cshore_: ok Dec 28 11:40:06 nunojpg * r24844 /packages/utils/watchcat/files/watchcat.sh: [packages] watchcat: added missing file Dec 28 13:15:39 build #49 of ps3 is complete: Failure [failed compile_3] Build details are at http://tksite.gotdns.org:8010/builders/ps3/builds/49 Dec 28 14:16:20 I took a look at the OpenWRT wiki on "creating packages" (http://wiki.openwrt.org/doc/devel/packages) which was last modified on 2010/12/20 11:37 and did not see a section about config like the one implemented by packages/net/samba3 with a Config.in file. Is this already considered obsolete? Dec 28 14:18:08 I am thinking to use this as a 'git pull' option for my FS git package to let OpenWRT dynamically update the source using 'git pull'. Dec 28 15:11:32 mazilo: more likely it just isn't documented. Dec 28 15:16:22 KanjiMonster: Thanks. Dec 28 15:17:29 KanjiMonster: I reckon it is time to look at other Makefile and Config.in files to learn how to implement this option. Dec 28 15:23:55 build #43 of ep93xx is complete: Failure [failed shell_9] Build details are at http://tksite.gotdns.org:8010/builders/ep93xx/builds/43 Dec 28 15:40:05 I see there are DEVEL and BROKEN variables to represent for a development and broken package. Is there any representation for 'obsolete' and/or 'depricated' variables? Dec 28 15:59:49 build #51 of pxcab is complete: Failure [failed compile_3] Build details are at http://tksite.gotdns.org:8010/builders/pxcab/builds/51 Dec 28 16:02:28 Is anyone here who moderates OpenWRT forum? If so, please kindly clean it up from spams (https://forum.openwrt.org/viewtopic.php?id=27885) as well as those spammers. Thanks. Dec 28 16:06:30 build #43 of ifxmips is complete: Failure [failed shell_9] Build details are at http://tksite.gotdns.org:8010/builders/ifxmips/builds/43 **** ENDING LOGGING AT Tue Dec 28 17:22:34 2010 **** BEGIN LOGGING AT Tue Dec 28 17:22:49 2010 Dec 28 21:07:23 where do I find the code which initializes the modem-part (ADSL) on a certain family? Dec 28 21:08:02 by checking the svn? Dec 28 21:08:33 is target/linux//patches/ the right place? Dec 28 21:08:42 which folder. Dec 28 21:09:06 yes Dec 28 21:09:18 any special file? Dec 28 21:09:36 the names are quite general. Dec 28 21:10:18 (or maybe I'm dumb) Dec 28 21:13:34 sasha__: I think you'll find the adsl stuff in package/ar7-atm or package/ifxmips-dsl-api (depending on which target you are looking for) Dec 28 21:14:03 hey KanjiMonster, any news on the newstyle brcm stuff? Dec 28 21:15:27 xMff: I wasn't at home during the holidays, so nothing new yet. Started reimplemening the stuff (partially going along bernhard's original patches, partially using the original broadcom code) to better understand how it works and what it really needed Dec 28 21:16:08 ah I see Dec 28 21:16:15 KanjiMonster: thanks, I'll check it out. Dec 28 21:17:07 I'm about to poke brcm-wl again on 5354 chips Dec 28 21:20:32 xMff: 5354 is ethernet, no? Dec 28 21:22:09 soc Dec 28 21:24:39 the 5354 has an integrated switch, right? Dec 28 21:25:00 build #52 of gemini is complete: Success [build successful] Build details are at http://tksite.gotdns.org:8010/builders/gemini/builds/52 Dec 28 21:25:44 yes Dec 28 21:30:10 there is no integrated wifi, so i still fail to see the connection between the soc and brcm-wl Dec 28 21:30:31 the connection is that brcm-wl does not work there Dec 28 21:30:37 while it does on older socs Dec 28 21:31:11 really? not because the phy is different? Dec 28 21:31:17 and yes, there is integrated wifi Dec 28 21:32:10 soc with integrated wifi? is that even possible? Dec 28 21:32:42 The BCM5354 integrates a high-performance MIPS32 processor, IEEE 802.11 b/g MAC/PHY, 2.4-GHz direct conversion radio, USB 2.0 host controller, SDRAM controller, and a configurable five-port Fast Ethernet (FE) switch. Dec 28 21:33:26 wow. what devices is that in? Dec 28 21:33:47 buffalo whr-g125 for example Dec 28 21:40:24 afaik the "new" broadcom equivalent would be the BCM5356U Dec 28 21:56:45 xMff: nbd said brcm-wl does not work on any of the lp-phy devices. is that one? Dec 28 21:57:02 afaik yes Dec 28 22:07:15 k Dec 28 22:07:34 tab fail Dec 28 22:31:14 hm, apparently the driver never leaves wl_start_txqwork() Dec 28 22:36:35 build #51 of ppc44x is complete: Exception [exception failed slave lost shell_13 compile_12] Build details are at http://tksite.gotdns.org:8010/builders/ppc44x/builds/51 Dec 28 22:36:37 build #48 of sibyte is complete: Exception [exception failed slave lost shell_13 compile_12] Build details are at http://tksite.gotdns.org:8010/builders/sibyte/builds/48 Dec 28 22:36:42 build #50 of ramips is complete: Exception [exception failed slave lost shell_13 compile_12] Build details are at http://tksite.gotdns.org:8010/builders/ramips/builds/50 Dec 28 22:38:35 xMff: you have access to the source? or is that in the wrapper? Dec 28 22:38:51 its in the wrapper Dec 28 22:39:35 i should take a look as well, then Dec 28 22:42:31 http://paste.openwrt.org/d5c276ff1 Dec 28 22:43:03 then you can enable debug messages on-target with wl -i wl0 msglevel foo Dec 28 22:47:43 <[florian]> sn9: thanks for the rdc stuff! Dec 28 22:49:01 last i checked the rdc target, eth was nonfunctional. i should look into that again, but ar7 and brcm47xx have a higher priority for me atm Dec 28 22:52:31 <[florian]> sn9: someone apparently reported the same issue without providing more informations yet Dec 28 22:52:36 <[florian]> sn9: so I will look into it Dec 28 22:52:45 which issue? Dec 28 22:53:26 <[florian]> sn9: that ethernet is apparently non functionnal Dec 28 23:01:29 xMff: i likely need working ar7 before i can work on brcm47xx again Dec 28 23:02:36 [florian]: last i left ar7, it was obvious that all the fixed phy stuff needed to be replaced with some kind of driver for cpmac's built-in phy Dec 28 23:03:23 <[florian]> sn9: yes, except that I do not know any hardware needing to support the internal phy apart the titan variant Dec 28 23:03:36 <[florian]> sn9: we'd better need a switch driver for gpio-connected switches Dec 28 23:03:46 the fritzboxen need it Dec 28 23:04:11 <[florian]> talking about that, someone reported to me that your vlynq changes did break on fritzbox! Dec 28 23:04:16 <[florian]> that was from Svenska Dec 28 23:04:23 <[florian]> he is not online unfortunately Dec 28 23:04:53 i am aware Dec 28 23:05:21 i left that in an intermediate state with uncommitted fixes Dec 28 23:06:44 <[florian]> ok, once you have some, let me know Dec 28 23:09:56 i do, but i need eth to work on all devices Dec 28 23:10:13 it's slightly intertwined Dec 28 23:12:22 someone had a "fritzbox wlan sl" that used the built-in phy on a 7300 Dec 28 23:12:41 i have a speedport w501v that uses it on a 7200 Dec 28 23:14:00 <[florian]> well you could probably create a simple phy driver for the internal phy if the generic phy driver + irq line does not do the job Dec 28 23:19:45 well, i have no immediate access to a device that needs that, but the issue was that under the current regime, fixed phy is needed _either_ in the case of internal phy _or_ in the case of adm6996, and not otherwise Dec 28 23:20:25 <[florian]> right Dec 29 01:02:43 anyone having issues with the password functions? I'm not sure if it's the extroot or uClibc but passwd and sudo hang. I'm investigating Dec 29 01:06:18 [florian]: any idea why a switch on 6358 would say 'failed to enable switch' ? Dec 29 01:06:50 [florian]: stock uses SPI from the bootlog Dec 29 01:08:02 passwd issue is an extroot problem, probably due to ext4 kernel module Dec 29 01:08:20 I'm going to rebuild with ext3 Dec 29 02:12:33 ext3 and ext2 modules work. I've opened a ticket and assigned to Kaloz as I don't know if/when I will have a chance to work on it further, I don't know kernel filesystem internals, and Kaloz wants ext4 only. Dec 29 02:23:27 #8541, more trac spam **** ENDING LOGGING AT Wed Dec 29 02:59:57 2010