| 00:06.42 | *** join/#nslu2-linux aadamson (n=ne1h@c-76-97-53-219.hsd1.ga.comcast.net) |
| 00:36.29 | CIA-8 | 03bzhou * r7842 10optware/trunk/make/gpsd.mk: gpsd: relax on stripping python *.so |
| 01:18.14 | *** join/#nslu2-linux aadamson (n=adamson_@c-76-97-53-219.hsd1.ga.comcast.net) |
| 01:28.12 | *** join/#nslu2-linux aadamson (n=ne1h@c-76-97-53-219.hsd1.ga.comcast.net) |
| 01:39.45 | *** join/#nslu2-linux aadamson (n=ne1h@c-76-97-53-219.hsd1.ga.comcast.net) |
| 01:48.28 | CIA-8 | 03bzhou * r7843 10optware/trunk/make/asterisk16.mk: asterisk16: adjusted LDFLAGS for some platforms |
| 01:52.11 | CIA-8 | 03bzhou * r7844 10optware/trunk/ (3 files in 2 dirs): asterisk16: promoted on most platforms |
| 02:03.51 | *** join/#nslu2-linux Sinclair73de_ (n=Sinclair@p57B48C27.dip0.t-ipconnect.de) |
| 02:16.50 | *** join/#nslu2-linux ka6sox-work (n=tking@nslu2-linux/ka6sox) |
| 02:28.22 | *** join/#nslu2-linux mithro (n=tim@ppp246-117.static.internode.on.net) |
| 02:41.03 | aadamson | anyone figured out what's wrong with the Master Makefile and building the slugosbe-image? |
| 02:41.17 | aadamson | for the last 3 or 4 days, I can't build it due to an error right at the start of the make |
| 02:41.40 | aadamson | looks like a file is messed up somewhere and a path is null when it shouldn't be during the make |
| 02:43.10 | aadamson | here is what I get for an error on each build attempt - http://www.pastebin.ca/914149 |
| 02:43.54 | *** join/#nslu2-linux tking_ (n=tking@netblock-66-159-209-57.dslextreme.com) |
| 02:44.01 | aadamson | I added a print statement ot that last py file and it processed a number of directories/files until it hits one that has no value - assume it's null |
| 02:44.23 | aadamson | and then it spits out the errors on the pastebin link above. |
| 02:44.40 | aadamson | It appears that I'm not the only one effected, the main distro person also has the same problem |
| 02:44.51 | aadamson | this is on slugos btw |
| 02:52.57 | mwester | No, nobody has figured that out and fixed it yet. It's an upstream problem, and in the core of the code - there are very few (like one) people who understand that part of it. |
| 02:53.52 | mwester | Somebody upstream changed something, and either they'll fix it, or it will get fixed when somebody building SlugOS finds the issue and pushes a fix upstream (or adds a workaround at the SlugOS level). |
| 02:53.58 | aadamson | mwester, oh, bummer, that's a major issue... well, not really, but a PITA anyway... I poked away at it, and if I knew how it figured out what it built maybe I could figure out how to fix it... what it built as in what order |
| 02:54.09 | mwester | So for now, we either live with it, or someone troubleshoots it themselves. |
| 02:54.32 | aadamson | sorta makes building a new image a bit of a problem to live with :) |
| 02:54.35 | mwester | aadamson: It's not building anything at the point of the failure. |
| 02:55.08 | mwester | So there's nothing wrong with the SlugOS build files themselves -- the problem happens as the build engine (bitbake) initializes itself. |
| 02:55.40 | aadamson | well, I should say that it's going thru some order of things as I mentioned, I put a print in the python code that errors and it goes thru about 50 or more items verifiting there existance and then hits a null or empty name.... |
| 02:56.31 | mwester | It's likely a missing value in a config file, one that's not used or parsed by other than SlugOS build environments, but that may be a red herring as it could also be sensitive to the OS environment, not to mention that the core OE developers are all off partying at some event this weekend, so perhaps nobody is really paying much attention... |
| 02:56.53 | mwester | Do you have that list of things it's testing for? |
| 02:57.00 | aadamson | but alas, I'm certainly no expert on it... figured it would get fixed by now, as it's been out for the last 3 or 4 days. Yes, I heard that Fosdem was going on |
| 02:57.08 | aadamson | I can re-create that easy enough |
| 02:57.11 | aadamson | hang on |
| 02:59.29 | aadamson | http://www.pastebin.ca/914701 - if you scroll to the top, you'll see it going thru valid items until it fails with the last *path is* statement |
| 03:00.02 | aadamson | so you can see that it actually is going thru some of the build, but mostly just verifing existance, etc |
| 03:00.38 | aadamson | you see that line 142 where path is NONE - that's the killer |
| 03:00.56 | mwester | Yeah, and strange. |
| 03:00.58 | aadamson | but I don't know what order or what it's doing to get there or I'd go poke at the distribution. |
| 03:01.26 | mwester | All the others are dev tools from the sanity checker. Then it goes and checks for something completely else. |
| 03:01.34 | aadamson | all the python code is doing that's being called is verifing paths |
| 03:01.39 | mwester | But that's the sanity checker, which I think can be disabled. |
| 03:04.18 | aadamson | I also found that you *dont* want to build from scratch anymore... meaing, create a directory, wget the Makefile and then do a make update - whew, the openwrt guys have done something in their build, it stops you and asks you for all kinds of information that I have no idea what to enter or means |
| 03:04.31 | aadamson | good thing I backed everything up before I went down that path |
| 03:07.01 | mwester | I remove the references to openwrt in the Makefile. |
| 03:07.18 | aadamson | yea, prolly what I should do... |
| 03:14.40 | aadamson | hmm, I commented out the Inherit += sanity from the conf/sanity.conf file and it got past the error, but I didn't let it build |
| 03:18.55 | *** join/#nslu2-linux ka6sox-laptop (n=tom@nslu2-linux/ka6sox) |
| 03:23.51 | mwester | Add the following to your slugos/conf/local.conf file: |
| 03:24.28 | mwester | SANITY_ABIFILE := "${TMPDIR}/abi_version" |
| 03:24.41 | mwester | SANITY_ABI := "" |
| 03:24.53 | mwester | See if that gets it working again. |
| 03:25.18 | aadamson | ok, but before I go there, I think I might be onto something... I found where it failes in the sanity check... let me debug for a few mins, it's in a check of requreid utilitis |
| 03:25.28 | aadamson | utilities... sheesh |
| 03:26.01 | mwester | It's in the code that checks if the version of the tmp directory structure has changed. |
| 03:26.35 | mwester | It fetches the value of the varaible "SANITY_ABIFILE", which doesn't exist in the environment. |
| 03:26.49 | mwester | (it's defined in the source file, not in the environment) |
| 03:27.13 | mwester | The value is, of course, returned as a null value --- which is then passed to os.path.exists to test for existence. |
| 03:27.20 | aadamson | ok, well, you obviously know this way better than I... I'll go try your lines above...be right back |
| 03:28.58 | aadamson | those lines did indeed get it working again, is that a work around or the real fix that was needed? |
| 03:30.42 | mwester | That's a workaround, IMO the code is broken. |
| 03:31.41 | aadamson | well, as always - huge thanks! get's be going again anway. I suppose that it might break again once the real fix is made, but until then, I'm working ... did that happen in a make update somewhere along the line? I guess so as I've updated no tools |
| 03:33.27 | mwester | I'm not sure when the change was made, but it was pulled in with a make update at some point. |
| 03:35.27 | aadamson | ah, I see exactly... in line about 149 there is an attempt to get that ABIFILE variable, but it's defined at the begging of the code and not in the env space... |
| 03:36.33 | CIA-8 | 03bzhou * r7845 10optware/trunk/make/gpsd.mk: gpsd: worked around an -O2 bug of mipsel gcc 3.4.6 by lowering to -O1 on that file alone |
| 03:39.37 | aadamson | mwester: - Thanks much ! if you know the main distro person, he and I were chatting about this issue earlier today as he had the same problem. |
| 03:44.34 | *** join/#nslu2-linux eiffel (n=eiffel@dslb-084-061-134-051.pools.arcor-ip.net) |
| 03:55.13 | *** join/#nslu2-linux Ragnor (n=Ragnor@dslb-084-060-167-093.pools.arcor-ip.net) |
| 04:03.06 | mwester | aadamson: :) Actually fixing that on the build server is probably something I should do, but I think I'll wait until I get some feedback on the bug I just opened on this problem, just in case the workaround has some side-effects or other bad behavior. |
| 04:06.39 | aadamson | ok, fair enough. I'll check back in a few days and see what you might have found... for now, I'm back able to build the entire tree again :) |
| 04:30.01 | *** join/#nslu2-linux dumfrac (n=gordon@71-208-115-201.hlrn.qwest.net) |
| 04:30.05 | *** join/#nslu2-linux ByronT (n=byron-po@nslu2-linux/ByronT) |
| 05:32.43 | *** join/#nslu2-linux DrZimmerman (n=theo@tbaumgartner.broker.freenet6.net) |
| 05:39.45 | *** join/#nslu2-linux yvasilev (n=yvasilev@189.140.145.248) |
| 05:40.52 | *** join/#nslu2-linux ka6sox-laptop (n=tom@nslu2-linux/ka6sox) |
| 05:45.57 | CIA-8 | 03bzhou * r7846 10optware/trunk/ (3 files in 2 dirs): python25: 2.5.1 -> 2.5.2 |
| 05:49.42 | *** join/#nslu2-linux ka6sox-laptop (n=tom@nslu2-linux/ka6sox) |
| 05:58.19 | *** join/#nslu2-linux Sinclair73de_ (n=Sinclair@p57B49A78.dip0.t-ipconnect.de) |
| 06:21.46 | CIA-8 | 03bzhou * r7847 10optware/trunk/platforms/packages-syno-x07.mk: asterisk16: demoted on syno-x07 |
| 06:26.14 | *** join/#nslu2-linux VoodooZ (n=palm3@76.10.166.127) |
| 06:34.49 | *** join/#nslu2-linux ByronT_ (n=byron-po@nslu2-linux/ByronT) |
| 07:00.34 | CIA-8 | 03bzhou * r7848 10optware/trunk/ (3 files in 2 dirs): asterisk14: adjust LDFLAGS for angstrom toolchains |
| 07:07.23 | *** join/#nslu2-linux ch4os_ (n=ch4os@unaffiliated/ch4os/x-059673) |
| 07:08.11 | *** join/#nslu2-linux drath_ (i=vmaster@p5B07C2F6.dip.t-dialin.net) |
| 07:23.55 | *** join/#nslu2-linux erikja (n=erikja@0x57314066.abnxx10.adsl-dhcp.tele.dk) |
| 07:32.11 | *** join/#nslu2-linux jacques (n=jacques@nslu2-linux/jacques) |
| 08:41.26 | *** join/#nslu2-linux DrZimmerman (n=theo@tbaumgartner.broker.freenet6.net) |
| 09:24.23 | CIA-8 | 03dyoung * r7849 10optware/trunk/sources/crosstool/ts72xx-cross334.sh: repair ts72xx toolchain build under newer gccs |
| 13:20.07 | CIA-8 | 03dyoung * r7850 10optware/trunk/make/libcapi20.mk: version 3-cm -> 3.0.5-cm |
| 14:05.30 | *** join/#nslu2-linux GPSFan (n=kenm@12.10.255.246) |
| 14:37.26 | aadamson | mwester: if you have the abliity, go look at the back thread on the #oe channel, they posted a patch for the SANITY_ABIFILE fix |
| 14:37.46 | aadamson | mwester: - http://paste.lisp.org/display/56373 |
| 14:48.42 | *** join/#nslu2-linux kizoku (n=albus@ti0182a340-1885.bb.online.no) |
| 14:53.28 | *** join/#nslu2-linux Reedy_Boy (n=Reedy_bo@wikimedia/Reedy-Boy) |
| 15:02.48 | CIA-8 | 03osas * r7851 10optware/trunk/make/asterisk16.mk: asterisk16: clean up default config |
| 15:29.28 | aadamson | mwester, I'd hold off on any changes just yet as of 10:29 EST (US TIME), they are still going back and forth on the ABIFILE fix... |
| 15:48.39 | CIA-8 | 03bzhou * r7852 10optware/trunk/platforms/ (packages-angstrombe.mk packages-angstromle.mk): asterisk14-chan-capi: promoted on platforms with angstrom toolchain |
| 16:40.39 | *** join/#nslu2-linux osas (n=nnnnnnnn@nslu2-linux/osas) |
| 17:01.56 | *** join/#nslu2-linux erikja (n=erikja@0x57314066.abnxx10.adsl-dhcp.tele.dk) |
| 17:15.11 | *** join/#nslu2-linux dumfrac (n=gordon@nslu2-linux/dumfrac) |
| 17:21.02 | *** join/#nslu2-linux mwester-laptop (n=mwester@gw.mwester.net) |
| 17:22.45 | *** join/#nslu2-linux mwester (n=mwester@gw.mwester.net) |
| 18:13.06 | *** join/#nslu2-linux yvasilev (n=yvasilev@189.140.145.248) |
| 18:31.09 | *** join/#nslu2-linux DrZimmerman (n=theo@tbaumgartner.broker.freenet6.net) |
| 18:58.25 | aadamson | mwester: so if I update will I pick up the recent changes for the sanity fixes? |
| 18:59.02 | mwester | Sounds like it - give it a try ;) |
| 18:59.47 | aadamson | and then comment out those lines in the local.conf. but will sanity.conf be downloaded in the mix. Sounds like the work around is they put them in sanity.conf until the more formal fixes are done |
| 18:59.53 | aadamson | that how your read the bouncing ball? |
| 19:00.49 | mwester | Dunno, I haven't updated. But yes, sanity.conf will be downloaded along with everything else. Basically if it's in the openembedded directory, it's part of OE -- over-rides and local configs are in the slugos directory. |
| 19:01.12 | aadamson | k, well, we'll know shortly. |
| 19:01.35 | aadamson | ah, did indeed get a new sanity.conf |
| 19:02.13 | aadamson | and a new sanity.class |
| 19:03.50 | aadamson | nope, didn't fix it |
| 19:03.55 | aadamson | now a different error |
| 19:04.25 | mwester | Well, then it fixed the original problem. ;) |
| 19:06.42 | aadamson | probably, I"m chatting with RP over on oe right now about this |
| 19:07.05 | aadamson | here are the results now - http://www.pastebin.ca/915361 |
| 19:08.06 | mwester | Oh, that. |
| 19:08.22 | mwester | Yeah, that's easily worked around, I think. |
| 19:09.24 | aadamson | yeah, I could just put the variables back in my local.conf file and 'm sure that would fix it :) |
| 19:09.44 | mwester | No, just the one. |
| 19:14.37 | *** join/#nslu2-linux VoodooZ (n=palm3@76.10.166.127) |
| 19:18.44 | aadamson | mwester: again, many thanks! |
| 19:25.28 | *** join/#nslu2-linux pleemans (n=peter@d51A5E76A.access.telenet.be) |
| 19:41.37 | *** join/#nslu2-linux dumfrac (n=gordon@nslu2-linux/dumfrac) |
| 20:02.33 | *** join/#nslu2-linux VoodooZ_ (n=palm3@76.10.166.127) |
| 20:04.32 | *** join/#nslu2-linux mithro (n=tim@ppp246-117.static.internode.on.net) |
| 20:10.22 | *** join/#nslu2-linux ch4os_ (n=ch4os@unaffiliated/ch4os/x-059673) |
| 21:04.24 | *** join/#nslu2-linux yvasilev (n=yvasilev@189.140.145.248) |
| 21:58.25 | *** join/#nslu2-linux Reedy_Boy (n=Reedy_bo@wikimedia/Reedy-Boy) |
| 22:04.37 | CIA-8 | 03dyoung * r7853 10optware/trunk/platforms/toolchain-ts72xx.mk: finish repairing toolchain build |
| 22:05.07 | CIA-8 | 03dyoung * r7854 10optware/trunk/make/microcom.mk: both official mirrors are down, so use ours |
| 23:10.33 | *** join/#nslu2-linux Zword (n=Zword@bos94-3-82-226-234-106.fbx.proxad.net) |
| 23:24.35 | *** join/#nslu2-linux ByronT (n=byron-po@nslu2-linux/ByronT) |
| 23:48.38 | *** join/#nslu2-linux Zword (n=Zword@bos94-3-82-226-234-106.fbx.proxad.net) |