00:33.36 | marex | fray: any news on meta-xilinx-standalone ? |
02:58.17 | *** join/#oe davidlt (~davidlt@78-63-27-146.static.zebra.lt) |
04:07.26 | *** join/#oe sakoman (~steve@rrcs-66-91-142-162.west.biz.rr.com) |
04:55.39 | *** join/#oe vmeson (~rmacleod@198-48-226-187.cpe.pppoe.ca) |
05:13.44 | *** join/#oe armpit (~armpit@2601:202:4180:a5c0:5ce5:5745:9142:936e) |
05:34.38 | *** join/#oe AndersD (~AndersD@h83-209-96-136.cust.a3fiber.se) |
05:36.39 | *** join/#oe AndersD_ (~AndersD@h-94-254-17-226.A137.corp.bahnhof.se) |
06:24.55 | *** join/#oe rob_w (~bob@unaffiliated/rob-w/x-1112029) |
06:29.54 | *** join/#oe kaa (~kaa@95.67.115.55) |
06:32.53 | *** join/#oe frsc (~frsc@i59F4BD95.versanet.de) |
06:36.16 | *** join/#oe frsc (~frsc@i59F4BD95.versanet.de) |
06:42.06 | *** join/#oe cpriouzeau (~cpriouzea@165.225.94.93) |
06:57.43 | *** join/#oe BlastuR (~magnus@unaffiliated/blastur) |
07:04.33 | *** join/#oe leon-anavi (~Leon@78.130.197.211) |
07:14.59 | *** join/#oe ao2 (~ao2@host-79-18-147-136.retail.telecomitalia.it) |
07:36.23 | *** join/#oe florian_kc (~florian_k@Maemo/community/contributor/florian) |
07:54.42 | *** join/#oe ChanServ (ChanServ@services.) |
07:54.42 | *** mode/#oe [+o ChanServ] by services. |
08:02.43 | *** join/#oe sakoman (~steve@rrcs-66-91-142-162.west.biz.rr.com) |
08:22.39 | *** join/#oe frsc (~frsc@i59F4BD95.versanet.de) |
08:46.36 | *** join/#oe cpriouzeau (~cpriouzea@165.225.94.93) |
08:51.07 | *** join/#oe tnovotny (~tnovotny@ip4-83-240-26-162.cust.nbox.cz) |
09:34.57 | *** join/#oe flynn378 (sid63564@gateway/web/irccloud.com/x-ogfyhjtmbwimlsbr) |
09:36.59 | *** join/#oe Tartarus (sid72705@gateway/web/irccloud.com/x-jbognfstpnabifbr) |
09:38.29 | *** join/#oe madisox (sid453692@gateway/web/irccloud.com/x-uclhrvwbjwdttzar) |
09:38.30 | *** join/#oe mithro (sid24875@gateway/web/irccloud.com/x-jdxmzyceidyowomu) |
12:11.48 | *** join/#oe berton (~user@2804:14d:4085:87f7:2812:ea42:2a4c:4bba) |
12:17.35 | *** join/#oe frsc (~frsc@i59F4BD95.versanet.de) |
12:24.44 | *** join/#oe florian (~florian_k@Maemo/community/contributor/florian) |
12:41.19 | *** join/#oe bencoh (~bencoh@unaffiliated/bencoh/x-184637) |
12:45.59 | *** join/#oe simosx (~simosx@ubuntu/member/simosx) |
13:13.16 | *** join/#oe vmeson (~rmacleod@198-48-226-187.cpe.pppoe.ca) |
13:43.46 | *** join/#oe Emantor (~Emantor@magratgarlick.emantor.de) |
13:51.17 | *** join/#oe XorA (~XorA@xora.default.ggregory.uk0.bigv.io) |
13:59.16 | *** join/#oe sakoman (~steve@rrcs-66-91-142-162.west.biz.rr.com) |
15:41.23 | *** join/#oe florian_kc (~florian_k@Maemo/community/contributor/florian) |
16:50.52 | *** join/#oe ao2 (~ao2@host-79-18-147-136.retail.telecomitalia.it) |
16:51.56 | *** join/#oe florian__ (~florian_k@Maemo/community/contributor/florian) |
16:54.09 | *** join/#oe vmeson (~rmacleod@198-48-226-187.cpe.pppoe.ca) |
17:15.47 | *** join/#oe moldorcoder7 (~moldorcod@206.166.251.59) |
17:21.16 | *** join/#oe pascalseeland (~pascalsee@p200300c76f1b1200022522fffeda012a.dip0.t-ipconnect.de) |
18:02.37 | *** join/#oe rob_w (~rob@unaffiliated/rob-w/x-1112029) |
18:51.58 | felix_inst | I am struggling a bit with why a config fragment I made does not seem to get applied. I see the fragment pop up in build/tmp/work/plat-linux/.. along with other applied fragments. For whatever reason, though, the one I added is not reflected in .config - why would that be? |
19:10.38 | zeddii | what's your kernel provider / kernel recipe. step 1) does it inherit kernel-yocto ? or include linux-yocto.inc ? |
19:14.26 | felix_inst | this is with a linux-xlnx kernel - checking on the includes |
19:21.49 | felix_inst | it includes linux-yocto.inc |
19:22.32 | zeddii | so that's good. fragments are never ignored or not applied, but that doesn't mean their settings always make it into the .config |
19:23.02 | zeddii | if there's a missing dependency, or another option in the kernel selects/clears a CONFIG, what's in your fragment won't be in the final .config |
19:23.24 | zeddii | I have audit tools for the latest releases that can help explain that, but not for the releases that Xilinx supports. |
19:23.41 | felix_inst | yes, I am having trouble seeing layer priorities because I need to work through petalinux |
19:24.26 | felix_inst | for the option I want, I also tried enabling all required settings in the same fragment to make sure |
19:25.10 | felix_inst | is the fact that I see the cfg file in the build tree enough of an indication that I didn't mess up the recipe? |
19:25.28 | zeddii | you would get a fetcher error if the .cfg wasn't picked up. |
19:25.36 | felix_inst | because that part seemed simple - and I can at least start looking elsewhere if I can cross that one off the list |
19:26.10 | zeddii | try doing a menuconfig on the kernel via bitbake, and search up the option you are trying to set. |
19:26.18 | zeddii | the dependencies and selcted implications will be listed there. |
19:26.41 | felix_inst | that's what I did - should be fine |
19:27.05 | felix_inst | I do get an info saying that ../linux-xlnx/oe-local-files/devtool-fragment.cfg would be removed |
19:27.11 | felix_inst | any idea what that is all about? |
19:28.12 | felix_inst | and that file doesn't exist - sorry if this is a xilinx specific question |
19:28.46 | felix_inst | oh nvm, that's from config not build |
19:29.55 | felix_inst | build seems to have no related messages other that it is compiling from an external source tree |
19:32.12 | felix_inst | and in that tree's config I have the prerequisites set, but not the option I am trying to enable |
19:33.15 | zeddii | hmm. is it using externalsrc ? |
19:33.35 | zeddii | if so, the config steps were inhibited for that in the past. |
19:33.39 | felix_inst | how do I check on that? sorry, utterly new to all of this |
19:36.08 | zeddii | what's the exact message about the external src tree you are seeing ? |
19:37.19 | felix_inst | NOTE: linux-xlnx: compiling from external source tree /home/user/components/yocto/workspace/sources/linux-xlnx |
19:38.22 | felix_inst | i see a "inherit externalsrc" in a kernel .bbappend - so that means it *is* using externalsrc I assume |
19:39.27 | zeddii | yup. |
19:39.51 | zeddii | let me see if I can raise someone else to comment ... fray you around ? |
19:40.21 | felix_inst | awesome, i appreciate the help! |
19:40.35 | zeddii | what's the xilinx release ? if I map it to the base yocto release, we can peek to see if that was before or after the configuration step was inhibited. |
19:40.59 | felix_inst | is there some section in the documentation i should read? i went through config fragements getting me to this point ;) |
19:41.17 | felix_inst | xilinx 2020.1 |
19:42.25 | felix_inst | oh i see - shouldn't I be able to get my option turned on with a defconfig I make? |
19:42.52 | felix_inst | i was just trying to do it in a custom layer to leave the base config untouched if at all possible |
19:43.20 | zeddii | possibly, but it goes through the same process as a fragment, and defconfig will overlay any base stuff completely. |
19:43.55 | felix_inst | hmm, I do have fragments in other layers as well, though - wouldn't those have the same issue? |
19:44.16 | zeddii | possibly. |
19:44.33 | felix_inst | let me check if all of those option apply properly |
19:44.56 | zeddii | i don't know how the petalinux build is structured, if you can find the work-shared/$MACHINE/kernel-source directory, there are logs, etc, to check. |
19:48.16 | felix_inst | i see debug messages searching for the .cfg file - but nothing that looks like a warning or error |
19:49.56 | zeddii | which file exactly did you check ? |
19:50.17 | felix_inst | i grepped all of them in the oe-logs folder |
19:51.34 | zeddii | not sure if anything specific would be in there, this is internal to config merge process, it isn't dumped to what bitbake/oe would pick up. |
19:51.51 | zeddii | there's a ".kernel-meta" subdirectory under the kernel-source directory that would have more info. |
19:52.08 | felix_inst | i am also looking at merge_config_build.log now - but don't see my setting in there |
19:52.30 | felix_inst | yeah that's the file you are talking about I think |
19:53.21 | zeddii | then, the kernel recipes may be clobbering the SRC_URI or something else. bitbake would have errored if it wasnt found at that level. |
19:53.30 | felix_inst | are bsp.cfg and user.cfg somehow special? |
19:53.32 | zeddii | and the merge never "ignores" a file. |
19:53.34 | zeddii | nope. |
19:53.52 | zeddii | something is injecting them from the petalinux side of things |
19:53.59 | zeddii | but they aren't special outside of that. |
19:54.10 | felix_inst | but are they somehow not subject to the inhibit you mentioned earlier? |
19:54.19 | felix_inst | or declared exempt or something? |
19:54.37 | felix_inst | I am wondering if I can append those .cfg files - that's why I am asking |
19:54.44 | felix_inst | instead of making a new fragment |
19:54.55 | zeddii | if you can find them, then probably. |
19:55.06 | zeddii | and no, either the configuration tasks run, or they don't |
19:55.22 | zeddii | and it does sound like they are running, so they've made sure the task is running with their externalsrc setup. |
19:55.27 | felix_inst | they do in some cases, not in others ;) |
19:57.00 | felix_inst | that file shows a whole bunch of conflicts, though. but those problems are all in the xilinx/avnet setup. that's scary |
19:58.24 | zeddii | conflics/overrides are fun, just the way things work. to both be standalone and/or additive. |
19:59.16 | felix_inst | this one is avnet overriding xlnx - but I am using the exact versions for which the contraption was developed |
19:59.40 | felix_inst | so I am a bit surprised these things didn't get ironed out in development - or am i missing something here? |
20:00.15 | zeddii | I don't understand the setup and petalinux, so I can't say. |
20:00.16 | felix_inst | screw it, I will do the config manually and add a note for now... |
20:00.51 | felix_inst | I am also a bit overwhelmed figuring that out |
20:01.25 | zeddii | if you are in that .kernel-meta directory in the kernel source dir. |
20:01.32 | zeddii | there's also a file 'config.queue' |
20:01.41 | zeddii | see if your fragment is listed in there. |
20:01.54 | felix_inst | i see config.baseline and config.new |
20:02.00 | marex | reads about scc and cringes ... |
20:02.21 | zeddii | has no shits to give |
20:02.27 | zeddii | felix_inst: those are different files. |
20:02.28 | marex | felix_inst: you dont happen to have two fragments named exactly the same , or named differently but in the same kmeta directory , included in the build ? |
20:02.43 | marex | zeddii: I am not gonna complain about meta-xilinx this time :-) |
20:02.45 | marex | I promise |
20:02.55 | felix_inst | marex, they have different names |
20:03.16 | zeddii | config.baseline and config.new are not files from the standard OE/yocto components |
20:03.17 | marex | OK, it was one thing which I ran into and which was caused me a great deal of pain |
20:03.30 | zeddii | they must be petalinux junkers |
20:03.48 | zeddii | what's the path of the directory that you are finding them ? |
20:04.06 | zeddii | doesn't mind some meta-xilinx or petalinux shade. it's all good fun. |
20:05.05 | felix_inst | build/tmp/work-shared/uz3eg-iocc/kernel-source |
20:05.25 | zeddii | and then .kernel-meta ? |
20:05.38 | felix_inst | yes |
20:06.41 | felix_inst | hang on, found a config.queue |
20:06.49 | zeddii | and there's a config.queue file there ? are they listed in that ? I think they may be commiting some other crime. Another thing you can do is a: bitbake -e linux-xlnx and see what the SRC_URI contains. |
20:06.56 | felix_inst | configs//./axisfifo.cfg # non-hardware |
20:07.01 | felix_inst | mine is listed in there |
20:07.19 | zeddii | ok. so it will be applied, but other things can still impact the final.confg |
20:07.40 | zeddii | what comes after it in the file ? those can still turn things off, etc, after that fragment. |
20:08.11 | felix_inst | https://pastebin.com/80pCS4KV |
20:08.30 | felix_inst | oh i see, so those are the configs I gotta work through and make sure my setting stays active? |
20:08.48 | zeddii | the queue is merged in that order, last one through the gate "wins" |
20:08.49 | felix_inst | and bottom is highest prio? |
20:08.57 | felix_inst | k |
20:09.07 | zeddii | so if you are unlucky something is impacting the option you are trying to change. |
20:09.58 | zeddii | there are some ways to bounce up the priority of yours via the KERNEL_FEATURES variable, but I'd have to know the version to get the incantation right. |
20:10.30 | zeddii | but somewhere in the kernel recipe chain, the SRC_URI being appended to bring those in. |
20:10.33 | felix_inst | oh so that is to bump up prio of part of a layer but not all of the layer? |
20:10.48 | marex | zeddii: is there some more detailed documentation for the inner workings of SCC ? |
20:10.54 | zeddii | these don't really follow that, it's just SRC_URI ordered. |
20:10.59 | marex | it seems it is not just me struggling with the tool |
20:11.30 | zeddii | we had it documented about 8 years ago, but they chucked all those in the bin for simpler docs. |
20:11.39 | zeddii | but scc isn't in play here. |
20:12.03 | zeddii | for the most part anyway. |
20:13.21 | felix_inst | i grepped for CONFIG_XIL_AXIS_FIFO in all .cfg files and no other ones contain that |
20:13.28 | felix_inst | so I don't think anything else overwrites |
20:13.43 | felix_inst | am i not looking for the final config in the right spot? |
20:14.00 | zeddii | that's the kernel build dir, which i'm not sure what they set it to. |
20:14.19 | felix_inst | build/tmp/work/uz3eg_iocc-xilinx-linux/linux-xlnx/5.4+git999-r0/linux-xlnx-5.4+git999/.config:4551:# CONFIG_XIL_AXIS_FIFO is not set |
20:14.35 | felix_inst | build/tmp/work-shared/uz3eg-iocc/kernel-build-artifacts/.config:4550:# CONFIG_XIL_AXIS_FIFO is not set |
20:14.53 | felix_inst | these are the two .config files with the setting I care about - and it's disabled in either one |
20:14.57 | zeddii | yep. that's the one. so something is still missing to turn it on or something else is unselecting it. |
20:15.07 | zeddii | what's the Kconfig block look like for it ? |
20:15.13 | zeddii | i.e. the source code. |
20:16.07 | felix_inst | https://pastebin.com/HwT01sW7 |
20:16.25 | felix_inst | not much there - and OF is enabled in the final config it seems |
20:17.02 | zeddii | yah. I can't imagine it is off. |
20:17.48 | zeddii | and that's in the externasrc tree ? or is it an out of tree module ? |
20:19.13 | felix_inst | the source is in the external tree |
20:19.48 | zeddii | that matters in the sense that even if dependencies are on, the Kconfig may not be reachable. i.e. some other surrounding menu option is disabled. |
20:19.55 | felix_inst | and the config.new in that tree is a synlink to one of the two config files that i just mentioned |
20:20.06 | felix_inst | the first one |
20:20.17 | felix_inst | so that's how they connect I suppose |
20:20.29 | zeddii | in my linux-yocto 5.10, the driver appears to be in staging, same for the petalinux tree ? |
20:20.33 | *** join/#oe sr105 (~sr105@096-058-179-191.res.spectrum.com) |
20:20.39 | felix_inst | yes |
20:20.48 | felix_inst | mine is 5.4 |
20:21.14 | zeddii | staging isn't always available, so it maybe just not reachable. unless you've enabled it, or you know that other staging drivers are on and working. |
20:21.20 | *** join/#oe florian__ (~florian_k@Maemo/community/contributor/florian) |
20:21.37 | felix_inst | others are enabled, yes |
20:21.45 | zeddii | all of those Kconfig's are protected by "if STAGING" |
20:22.00 | felix_inst | staging is enabled as well |
20:23.00 | *** join/#oe Emantor (~Emantor@magratgarlick.emantor.de) |
20:23.22 | *** join/#oe dlan (~dennis@gentoo/developer/dlan) |
20:23.55 | *** join/#oe kmaincent (~kmaincent@shells.bootlin.com) |
20:25.15 | zeddii | can you pastebin the merge_config_build.log file ? |
20:25.42 | felix_inst | https://pastebin.com/yMduiKtC |
20:26.55 | zeddii | and your fragment was called what (again, I can't find it in the history). |
20:27.07 | felix_inst | axisfifo.cfg |
20:27.56 | felix_inst | https://pastebin.com/Aqnx2mte |
20:28.41 | felix_inst | and here is linux-xlnx_%.bbappend |
20:28.46 | felix_inst | https://pastebin.com/YXbs0i4P |
20:29.03 | zeddii | am I overlooking it in that log, I don't see it. but I had to allow ads on pastebin, so it is twitching ads and giving me headache :D |
20:29.18 | felix_inst | no, it is not in that file |
20:29.34 | felix_inst | but it is in config.queue |
20:30.00 | zeddii | yah. so then they are doing something else wierd with the merge_config call. I wonder if they've gone and circumvented the standard one. |
20:30.10 | zeddii | if it ain't in that file, you'll never get it on. |
20:30.40 | felix_inst | ah ok, that's a good hint then |
20:31.00 | felix_inst | i'll see how the things in merge_config end up there |
20:31.02 | zeddii | if you look at the bitbake logs, there should be a log from do_kernel_configme |
20:31.24 | zeddii | if that isn't there, they may have done their own. |
20:31.35 | zeddii | but then again, the log file is there .. |
20:31.36 | zeddii | hmm. |
20:32.59 | felix_inst | i see the buildstate, but no log file |
20:33.07 | felix_inst | *buildstats |
20:34.45 | zeddii | what bitbake logs are in the temp/ ? |
20:34.57 | zeddii | just trying to make sure we are looking at the right thing . |
20:35.04 | felix_inst | would that be in components/yocto/workspace/sources/linux-xlnx/oe-logs? |
20:36.25 | zeddii | not sure. that's a petalinux thing, the yocto one is somewhere like: |
20:36.40 | zeddii | tmp/work/qemux86_64-poky-linux/linux-yocto-stable/5.10.37-r0/temp |
20:36.53 | felix_inst | is there a file that you know must be there? than I can at least confirm we're looking at the same folder |
20:37.13 | zeddii | it's full of log.do* and run.do* |
20:38.03 | felix_inst | ok, found the folder |
20:38.33 | *** join/#oe marex_ (~marex@195.140.253.37) |
20:39.07 | felix_inst | log.do_kernel_link_images / log.do_kernel_metadata / log.do_kernel_version_sanity_check are the kernel ones |
20:39.57 | zeddii | hah. what are they up to .... |
21:38.47 | *** join/#oe tlwoerner (~tlwoerner@unaffiliated/tlwoerner) |
21:53.27 | felix_inst | deleting the linux-xlnx:do_kernel_configme entry in locked-sigs didn't do the trick either |
21:56.21 | felix_inst | i don't think the avnet fragment take effect either |
21:57.02 | felix_inst | https://github.com/Avnet/meta-avnet/blob/master/recipes-kernel/linux/files/uz3eg/bsp.cfg#L5 |
21:57.15 | felix_inst | build/tmp/work/uz3eg_iocc-xilinx-linux/linux-xlnx/5.4+git999-r0/linux-xlnx-5.4+git999$ cat .config | grep -i usb_acm |
21:57.15 | felix_inst | # CONFIG_USB_ACM is not set |
21:57.44 | felix_inst | i'll check if i set this up in a weird way somehow |
22:23.05 | *** join/#oe moldorcoder7 (~moldorcod@206.166.251.59) |
22:49.14 | tlwoerner | just spent an enjoyable hour or so looking over the slides that have been submitted for next week's summit |
22:49.37 | tlwoerner | *cough* get your slides in! *cough* |
22:49.48 | tlwoerner | :-) |
22:58.14 | dl9pf | tlwoerner: when are the trainer login details sent out? |
23:00.15 | tlwoerner | dl9pf: David was sending them out today, but be lives in PST. i'll check with him |
23:00.22 | tlwoerner | halstead: are the trainer accounts setup? |
23:01.48 | halstead | tlwoerner, Waiting on dreyna to finish his setup. |
23:02.24 | tlwoerner | halstead: okay, thanks for the update :-) |
23:02.29 | halstead | dl9pf, tlwoerner If I'm reading the schedule correctly he expects me to have accounts ready for trainers some time tomorrow. |
23:03.19 | tlwoerner | halstead: dl9pf: ah okay, maybe i'm confusing David's "having my stuff ready" vs "having trainer account ready" |
23:03.27 | tlwoerner | thanks :-) |
23:03.33 | dl9pf | okidok |
23:04.03 | *** join/#oe Jin^eLD (~jin@45.62.236.211) |
23:04.15 | Jin^eLD | re |
23:04.21 | Jin^eLD | tlwoerner: thanks for adding the wiki entry |
23:13.47 | tlwoerner | Jin^eLD: no worries, the project is glad for the entry! |
23:14.15 | tlwoerner | RP: your email inspired Jin^eLD to make an entry to https://wiki.yoctoproject.org/wiki/Project_Users#Others |
23:14.35 | Jin^eLD | heh, you make it sound like I was the only one who reacted? :P |
23:16.01 | Jin^eLD | still need to ask the current company that I am working at, so I'll probably have one more soon |