IRC Logs for #crux Thursday, 2018-08-23

*** john_cephalopoda has quit IRC02:17
*** blueness_ has joined #crux02:24
*** admincomplex has quit IRC02:25
*** blueness has quit IRC02:25
*** tilman_ has joined #crux03:16
*** abenz has joined #crux03:31
ryuoIs Intel *trying* to commit corporate suicide?03:44
ryuohttps://www.theregister.co.uk/2018/08/21/intel_cpu_patch_licence/03:44
*** groovy2shoes has quit IRC04:23
*** groovy2shoes has joined #crux04:28
jaegervery neat: https://steamcommunity.com/games/221410/announcements/detail/169605585573935056104:34
*** abenz has quit IRC05:57
*** abenz has joined #crux06:11
*** slek has joined #crux06:24
ryuoI'd like to see them try to enforce a no-benchmarking clause.06:34
ryuothat would effectively make something like pcmark illegal if it were enforceable.06:35
*** abenz has quit IRC07:19
*** abenz has joined #crux07:19
*** JanC has quit IRC07:29
*** slek has quit IRC07:48
cruxbot[core.git/3.4]: dhcpcd: update to 7.0.808:36
joacimjaeger: odd to see games like quake and ultimate doom on that list. since those would work a lot better with quakespasm and gzdoom on newer hardware08:39
joacimdidn't expect to see nier automata on that list either. enjoyed that game a lot08:39
joacimi think many of the ports to OS X and Linux are already based on wine anyways08:40
Romsterintel is so butt hurt on losing performance on patching these bugs, and they took shortcuts to be all look at intel we got high performance compared to AMD. Now I can laugh at intel in there faces for being on AMD all these years.09:30
frinnstI have one esxi host patched. didnt see any license stuff09:42
frinnstseems you have to download the license to be 'bound' by it?09:52
frinnstthey should be laughed out of every court they try to enforce that in09:53
*** surrounder has joined #crux10:30
*** abenz has quit IRC11:02
*** john_cephalopoda has joined #crux11:29
*** onodera has joined #crux11:50
*** abenz has joined #crux11:50
*** onodera has quit IRC11:54
*** onodera has joined #crux11:54
*** abenz has quit IRC13:11
*** abenz has joined #crux13:28
*** abenz has quit IRC13:39
*** abenz has joined #crux13:40
*** JanC has joined #crux14:12
*** Anselmo has quit IRC14:57
abenzjaeger: ping15:12
jaegerabenz: I'm sorta around, working16:03
*** slek has joined #crux16:30
*** kori has quit IRC16:31
*** slek has quit IRC16:52
abenzjaeger: let me know when you can spare a few minutes16:55
abenzdo you remember once you helped me fix a build issue on crux16:55
abenzI have a similar prob again.. been on it for the past few hours and cant figure out why it fails16:56
abenzbuilds fine on debian/ubuntu and the usuals..16:56
abenzbut since crux is my main I'd rather figure it out16:56
abenzhttps://pastebin.com/1xfwscwq16:57
abenzyou can see failure in line 32, I think its similar to the prob before, ie echo or something and the flags dont get passed correctly16:57
jaegerMaybe it's a dash vs. bash thing again16:58
jaegerWhat is it you're building? openwrt or similar?17:00
abenzyes openwrt17:01
abenzI can write commands to duplicate? if you can help I'd appreciate it17:01
jaegerI can try... but not sure. That output doesn't really say what's wrong17:03
jaegerIt's getting a wrong -jX arg but why?17:03
abenzjaeger: yes correct. I believe the problem is in constructing the cross compile line17:04
abenzremember you fixed a similar problem before (it was caused by shell built-in echo, it produced a different output as opposed to /bin/echo)17:04
jaegerThat's why I wonder if it's dash vs. bash17:05
abenzgit clone https://git.openwrt.org/openwrt/openwrt.git17:06
abenzcd openwrt17:06
abenz ./scripts/feeds update -a17:06
abenz ./scripts/feeds install -a17:06
abenzmake defconfig17:08
abenzpls put the contents of this file into .config17:10
abenzhttps://pastebin.com/raw/z5ysDh6v17:10
abenzthe .config should be in root (ie inside openwrt dir you just cloned)17:10
abenzafter wards:17:10
abenzmake defconfig17:10
abenzmake -j<your cpu threads+2>17:11
john_cephalopodaThat's a new formula :þ17:17
john_cephalopodaI've seen threads-1, threads+1, threads, but I don't think I've seen threads+2 before :þ17:17
abenzjohn_cephalopoda: this makes my cpu 100% consistently..17:19
abenzI have an 8 core amd btw17:20
john_cephalopodassh times out when I compile with 8 cores on my raspi :þ17:22
john_cephalopodaCould also be OOM, not quite sure. cmake.17:23
abenzcompiling, on a raspi?17:23
jaegerI would expect that since it only has 4 cores and not much RAM17:23
pedjasomething weird is going on with that openwrt git17:24
abenzwhats goign on?17:25
pedjawhen I switch to openwrt-18.06 branch, and set it to track origin/openwrt-18.06, there are a lots of merge conflicts17:26
abenzits supposed to track it automatically?17:26
abenzwhat commands do you issue?17:27
jaegerjohn_cephalopoda: why would you use -j8 (or the cmake equivalent) on an rpi?17:32
john_cephalopodajaeger: Because it has 8 cores.17:32
john_cephalopodaWait, or has it?17:33
jaegerwhich rpi model has 8 cores?17:33
jaegerI'm not aware of any17:33
john_cephalopodaHuh, only 4.17:33
john_cephalopodaI thought it had 8. Rpi B3+17:33
jaegerah. Nope, just 417:33
abenzjohn_cephalopoda: you are compiling openwrt on it?17:35
jaegerabenz: I think it just has a bug in the build system17:35
abenzjaeger: same error on your system?17:35
jaegeryes17:36
john_cephalopodaabenz: I am using CRUX-ARM on it.17:36
pedjaabenz, hm. I re-cloned the repo, and with 'git co -b foo --track origin/foo' no merge conflicts17:36
abenzpedja: why do you need to do that? the local branches track the origin by default no?17:37
abenzjaeger: there are some logs in the build dir17:37
pedjatry it :)17:37
abenzeg build_dir/target-mips_24kc_musl/wolfssl-3.14.417:38
abenzpedja: I have a copy that I cloned ages ago17:38
abenzwhen I do git pull it syncs with origin17:38
abenzisn't that tracking?17:38
jaegerabenz: it seems to finish if I export CONFIG_SHELL=/bin/bash before starting the process17:44
pedjaabenz, afaik, that works only for master branch. could be wrong, though17:45
abenzexport CONFIG_SHELL=/bin/bash17:50
abenzafter running make it still fails for me17:51
abenzanything else you did jaeger ?17:51
pedjaso openwrt Makefile expects that sh is a symlink from bash? I love when they fail to mention that :)17:51
pedjaabenz, are you building master or release?17:52
pedja18.0617:52
abenz18.06 suceeds, master fails17:52
pedjaare you dbugging the issue, or building an image for the device?17:53
abenzit seems a dev found the prob, broken local macros17:53
abenzpedja: both17:54
pedjaah :)17:54
abenzbut a core dev took over so no worries17:54
abenzjaeger: many thanks for your time17:54
abenzyou too pedja and john_cephalopoda for compiling on your rpi :D17:54
abenzthe way I see it, crux forces clean code! so its a win win for everyone17:55
jaegerabenz: maybe need a make clean or something, but I didn't change anything in the code or repo17:59
jaegerjust exported CONFIG_SHELL18:00
abenzjaeger: absolutey right!18:00
abenzI just did a clean and it finished successfully18:00
jaegerok, cool18:00
abenzjaeger: if you were curious:18:04
abenz<jow> this macro does:  if $enable_jobserver = yes; then let enable_jobserver=$CPUCOUNT+1; fi18:04
abenz<jow> however with a posix shell:  $ sh -c 'foo=yes; let foo=$foo+1; echo $foo'18:04
abenz<jow> sh: 1: let: not found18:04
*** leah2 has quit IRC18:09
jaegerAh, there you go18:22
jaegergood to know18:22
*** leah2 has joined #crux18:23
*** onodera has quit IRC20:28
jaegeradding a 1080Ti to my GPU inventory today, woo20:37
*** BurnZeZ has joined #crux20:49
*** ryuo has quit IRC21:01
joacimfound one on sale?21:24
joacimthey're still a bit expensive here21:25
jaegerJust like my RX 580, bought it when it came down to $20 over MSRP21:30
jaegerSpecifically the asus strix one. There were cheaper options but I prefer the asus strix or msi gaming x coolers21:31
*** ryuo has joined #crux21:31
joacimgot my r9 fury at around 350 USD I think22:50
joacimwould like a similarly "nice" card for around the same price22:50
ryuojoacim: does the price make you furious?22:50
ryuo:)22:50
joacimnah already used to my ati rage cards :322:51
ryuolol22:51
jaegerhehe22:51
ryuoslowly making progress on my restoration of this e5530...22:51
ryuojaeger: we found somethng that seems to have worked better for a channel that we couldn't use +r in.22:54
ryuomarking it +s.22:55
ryuoit took awhile but the amount of spam decreased considerably.22:55
ryuoI presume it's because they can't find our channel as easily now.22:55
*** Tazy has joined #crux23:08
jaegermaybe so23:08
joacim+rs for a bit?23:30
joacimthen try -r+s next year?23:31
ryuo+s makes the channel unlisted in /list23:32
ryuothat's the only drawback.23:32
ryuoit depends how much crux relies on /list for discovery by users.23:33
ryuothough honestly, i don't believe I discovered the IRC channel through /list.23:33
ryuoI believe I used distrowatch back in the day and used the website.23:33
joacimgot here through crux.nu i think23:41

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!