IRC Logs for #crux Friday, 2015-01-30

*** hnode has joined #crux00:01
hnodehi cruxers :) I solve the problem with X/i915 driver :)00:02
korinice!00:02
hnodeis dificult with crux not to learn something every day !00:03
greduan_Indeed00:03
greduan_So what editor do people use?00:09
greduan_I use Vim and Emacs (depending on the task) but right now I've neither so I'm taking it as a chance and I'm looking at some of the alternatives.00:09
hnodeI use vim, but I'm far from taking advantage of00:10
greduan_You mean you are not very advanced?00:11
hnodeNeovim is the new thing in vi ?00:11
hnodeyes, just a normal key stroks,00:11
hnodebut at the moment is the only I have, and now I know how to copy and past with tmux,00:12
hnodeso I`m very happy :)00:12
greduan_I see00:12
greduan_Good :)00:12
greduan_NeoVim is the new thing, yes, but it's not yet stab le so I haven't looked into it00:12
greduan_stable*00:14
hnode:) sorry my english,00:14
hnodeI hope at least is funny :P00:14
greduan_It's fine00:17
greduan_:)00:17
greduan_Where are you from?00:17
hnodePortugal00:18
hnodeYou ?00:18
greduan_I am from Mexico00:18
hnodeNice to meet you !00:19
hnodeIm new user, Im tryng to learn more about system development,00:20
hnodebut just confuring crux is being lot of fun and learning,00:21
greduan_Nice to meet you :)00:21
greduan_I am also a new CRUX user, although I am pretty familiar with Linux.00:21
hnodeand English :D00:21
greduan_Indeed :)00:22
greduan_Almost my whole family speaks English, so I learnt since I was a little kid00:22
hnodeI use linux at least since 2004, but was a apt-get user00:22
greduan_I see00:23
greduan_I use Linux full-time since... Two or three years ago. Before that I always used Windows and then started using a Mac (for half a year), after which I started getting into open source and all that00:24
greduan_So here I am using Linux00:24
hnodefulltime linux since 200600:24
greduan_Cool!00:24
hnodeBut Im watching mircokernels00:24
hnodeI think is a "cleaner" way to do,00:25
greduan_Indeed00:25
hnodeand also to learn,00:25
hnode"learn" depends,00:26
greduan_How do you mean?00:26
hnodeI was thinking about minix, is "easy to learn" in versions contained in the book,00:27
greduan_ah OK00:28
hnodecannot say the same for example about L4, I think for a starter may be difficult00:28
greduan_I see00:29
hnodegreduan_: will this stuff that Im learning here in the next 10 years or so ?00:30
tired890I'm curious how old are you guys?00:34
koriI'm 1600:35
greduan_hnode how do you mean?00:35
hnodemy uptime is 3000:35
greduan_tired890: 17, turning 18 this year00:35
koriturning 17 this year00:36
greduan_kids if you're reading this, don't give your age so readily to somebody on the internet. lol00:36
tired890wow much younger than I expected00:36
hnodekids say the wrong age00:36
greduan_Too high or just miss the target?00:37
greduan_oh I get it00:37
greduan_bbl00:38
*** kori_ has joined #crux00:54
*** kori_ has quit IRC00:54
*** kori has quit IRC00:56
*** kori has joined #crux00:57
*** kori_ has joined #crux01:02
*** kori_ has quit IRC01:02
*** kori_ has joined #crux01:02
*** kori has quit IRC01:02
*** tilman has quit IRC01:03
greduan_back01:08
hnode :)01:08
hnodeI need to configure console printing, it messes with tmux01:09
hnodegreduan: do you have any blog ?01:14
*** dougl has quit IRC01:21
greduanhnode greduan.com01:24
greduanmy blog is in blog.greduan.com01:24
*** cezar has quit IRC01:52
*** nilp has joined #crux01:52
hnodeI get an error building firefox,02:06
hnodeUnifiedProtocols502:06
hnodeSome dependency not meet ?02:06
greduanwow02:32
greduanOnly 400MB of my RAM being used.02:32
greduanI guess YouTube uses up the other 600MB. lol02:32
greduanI'm glad I can't currently watch YouTube with this laptop, so I could find this out02:32
greduanOr I wonder if that's just the buffer/cache that Linux uses02:34
greduanTIL by tapping the Alt key Firefox will toggle the menu bar.03:02
greduanTime to sleep, see you guys later03:32
greduanBTW, prologic, I got your links about Gopher. :)03:33
prologicgreat :)03:33
*** mavrick61 has quit IRC03:50
*** mavrick61 has joined #crux03:51
*** hnode has quit IRC04:06
*** nilp has quit IRC04:23
tired890hey guys04:43
tired890I'm copying about 200gigs of data between sda to sdb (different sata drives), when I do, the system becomes sluggish04:43
tired890the OS is running from another third disk (no read/write operations on OS disk)04:44
tired890CPU utilization is unaffected04:44
tired890so I'm wondering, what is causing the system to become sluggish?04:44
tired890could it be a misconfig on the bus or something04:45
tired890?04:45
vlnxtired890: ionice. Set a lower priority to that process, so the system should stay responsive05:12
tired890vlnx, I'm thinking it might be the transparent huge page (memory) thing05:14
tired890the system waits for the disk to write or something05:14
tired890the system is sluggish even tho CPU utilization is less than 20%05:15
tired890when I open top05:15
tired890I see a process taking 10%05:15
tired890mount.ntfs05:15
tired890I'm copying these files over so I reformat my disk as ext405:15
tired890I hope after I do, when I copy these files back the problem will go away05:16
tired890vlnx, what puzzles me is that transfer is being done between two disks (B and C), system is on another disk (A, which is an SSD!).. the bus capacity is more than enough for the relatively slow mechanical disks.. yet when I transfer between the two (with cpu 90% free) the system becomes unresponsive..05:38
tired8907 mins left and then will format the disk as ext4, sure hope this is limited to ntfs only.. (btw the disk is mounted with ntfs-3g, as the kernel built in ntfs support has poor write support)05:41
*** xvee has joined #crux05:41
xveei updated my system and i think i broke everything. any time i try to compile or open a program i get illegal instruction. cflag issue?05:47
*** tired890 has quit IRC05:58
*** tired890 has joined #crux05:59
*** tilman has joined #crux06:05
xveechanged the clfags and rebuilding the system. let us hope06:12
tired890confirmed guys.. it was due to ntfs06:12
tired890I formatted disk as ext4 and now copying back the data06:13
tired890system is not sluggish06:13
*** pejman has joined #crux06:23
*** pejman has quit IRC06:23
*** pejman has joined #crux06:23
*** pejman_ has quit IRC06:24
diverseteK_: how is your project going?06:24
Romstertired890, maybe the disk scheduler07:14
*** tvaalen has joined #crux07:15
xveesort of a weird question, but wehre is the pkgmk.conf file in the bootcd?07:20
Romsterthere is none other than in pkgutils package07:22
Romsterpkgadd pkgutuils or prt-get update pkgutils and rejmerge07:23
xveelet me give that a shot. for some reason, i cant compile anything because of a error in the pkgmk file i believe07:24
xveechecked teh cflags and all and it seems to be fine07:24
xveejust reinstalled everything with the cd. no idea what happened the first time. thanks Romster07:32
*** vlnx has quit IRC07:48
*** nilp has joined #crux07:49
*** vlnx has joined #crux07:51
*** xvee has quit IRC08:16
*** hhhhhhhh has joined #crux08:24
teK_diverse: nothings happening until June08:41
teK_that's for sure.08:42
*** BitPuffin has joined #crux08:51
*** BitPuffin has quit IRC08:55
*** onodera has joined #crux09:29
*** SiFuh_ has joined #crux09:43
*** SiFuh has quit IRC09:45
*** SiFuh_ has quit IRC09:48
diverseteK_: I mean the python one that you are currently working on09:51
*** BitPuffin has joined #crux09:55
*** jue has quit IRC10:57
*** jue has joined #crux11:05
*** leo-unglaub has joined #crux12:07
*** xeirrr has joined #crux12:12
*** Necrosporus has joined #crux12:15
tired890hi Romster12:16
tired890cat /sys/block/sdb/queue/scheduler12:17
tired890noop deadline [cfq]12:17
tired890thats the current scheduler12:18
tired890don't have exp with this though.. what would you recommend ?12:18
NecrosporusHello. I am not able to see non-ascii letters in gtk apps12:18
NecrosporusI guess I might be lacking needed fonts or gtk settings12:18
teK_diverse: other problems and missing feedback from the module's author12:28
teK_so I'm pretty much stuck atm..12:29
*** Necrosporus has quit IRC12:31
*** BitPuffin has quit IRC12:32
*** Necrosporus has joined #crux12:34
*** BitPuffin has joined #crux12:47
*** alancio_ has quit IRC12:52
diverseteK_: ah, I know that feeling...13:08
teK_it sucks because it works for the author13:13
teK_fun factÖ13:13
teK_fun fact:13:13
teK_test.py segfaults13:13
teK_but I also noticed that running %load test.py in ipython and exiting succeeds. BUT13:13
teK_pasting test.py, running it and afterwards deleting x makes ipython core dump, too:13:13
teK_In [2]: del x13:14
teK_In [3]: [EOF]13:14
teK_zsh: segmentation fault  ipython --no-banner --no-confirm-exit.13:14
teK_on del x, some shared lib-based destructor is called and fails13:14
teK_so I might have hit a bug in the upstream library, that the cython based module relies upon13:14
teK_good thing: the C++ version was written by a colleague of the python version's author :O13:15
*** xeirrr has left #crux ("Leaving")13:16
*** Necrosporus has left #crux ("Leaving")13:17
diverseteK_: does that mean you will take the chance to rewrite it in C++?13:18
teK_no. no. no. NOOO GOD NO13:19
frinnstlol13:19
frinnstchicken13:19
teK_I hope they will talk to each other because they share an office or at least a floor13:19
teK_frinnst: obviously you are taking your life / spare time as a role model for mine; so I suppose you have the time to do that for me :))13:20
jaegertired890: HDD or SSD?13:20
tired890HDDs13:20
tired890copying from HDD1 to HDD2, system on SSD13:20
jaegercfq is fine for 99% of things, then13:21
jaegerset the SSD's scheduler to noop13:21
tired890thanks13:21
jaegerI use http://jaeger.morpheus.net/linux/files/99-block-scheduler.rules in /etc/udev/rules.d13:21
diverseteK_: so you need to convince the colleague to convince the python author to do what you want?13:21
teK_diverse:  I hope they discover a bug in the C++-module and fix it right away. Problem will be that it works for them but fails for me on two different distros in -HEAD and the last stable release13:22
tired890jaeger, changed ssd sched. to noop. Thanks for tip13:23
jaegernp13:24
diverseteK_: also by "shared" did you mean a destructor for something reference counted?13:28
teK_it's in the c++'s .so-library, I meant by shared13:29
diverseoh, that "shared"13:29
teK_python has no desctructors ;))13:29
teK_well it has.. but you know what I mean -_-13:29
*** Feksclaus has joined #crux13:32
diverseteK_: I never bothered with python much, but I figure that to mean you can't define destructors for a specific action when something drops out of scope.13:33
diversethe GC just does regular dropping13:35
*** greduan has joined #crux13:35
teK_the c++ constructor fails, that's all13:35
diverseright, now that makes sense to me13:35
diverseI hope that problem gets solved soon13:36
diversewhat a pain13:37
greduanGood morning. :)13:38
tired890morning greduan13:38
tired890oh my !!13:39
tired890compile -j150 on tmpfs13:39
teK_diverse: I hope so too :)13:39
tired890^ WIN13:39
teK_tired890: run make -j for maximum fun WIN :P13:39
tired890lo13:39
tired890lol*13:39
tired890teK_, I couldn't believe it13:40
tired890I mean oh man13:40
tired890hold on let me find a vid I'm sure someone did this13:41
*** nilp has quit IRC13:41
diversehow are you able to run that many jobs, just because of tmpfs?13:41
frinnstgood idea with a udev rule, jaeger13:42
frinnstim gonna steal that :)13:42
tired890diverse, in ram the speed is so fast that using a higher number of jobs is better13:42
tired890you will need more ram tho13:42
tired890the compile is crazy fast I mean even compared to an SSD13:43
frinnsti *really* doubt things will go any faster with a silly amount of jobs13:43
tired890I wonder how this would perform on a DDR4 system13:43
diversefrinnst: that's what I'm thinking13:43
frinnstand disks are not really the bottleneck13:43
tired890what is?13:44
frinnstthe cpu :)13:44
tired890disks require lots of reads and writes when compiling big programs13:44
tired890ops13:44
tired890I mean compiling requires lots of reads and writes13:44
tired890and disks can be bottlenecks13:44
frinnstdepends, if you dont use -pipe, then yeah13:44
diverseit's still all handled by the processor in question...13:44
*** teK__ has joined #crux13:45
*** teK__ has quit IRC13:45
greduanIs the Flash version the latest available version?13:50
frinnstit was a day or so ago13:52
frinnstso unless something was released very recently, yes13:52
greduanHmm...13:53
greduanFor some reason whenever I open a YouTube video, all it does is try to give me the highest quality, and then it tries again when there's no higher quality and it just crashes.13:55
greduanRather, tells me a problem has occurred13:55
frinnsthtml5 ftw13:56
frinnstseems to work pretty well with the firefox beta, too13:56
greduanStill haven't setup HTML513:56
greduanI did download GST13:56
greduanBut haven't enabled it in Firefox13:56
frinnsttoggle media.mediasource.enabled to true in about:conf13:57
frinnstabout:config13:57
frinnstseems to work in firefox 35 too13:57
greduanYou know what it is13:57
greduanit was built disabling gstreamer13:58
greduan--disable-gstreamer13:58
greduangotta compile13:58
greduanI also gotta compile the kernel to the latest version13:58
greduanAny tips for compiling 3.18.5?14:01
frinnstyeah but you dont need gstreamer support for youtube nor for flash14:02
frinnstfirefox bundles webm14:02
greduanfrinnst: That I did not know14:02
greduanNow I do14:02
greduanBut then if YouTube changed their default to HTML5 and stuff, why is it not using it with me?14:03
greduanAnyway, kernel first14:03
frinnstfirefox in windows also uses flash as default14:06
greduanI wonder if it's doing it on a platform basis...14:06
tired890I'm compiling the kernel in tmpfs with different -j to see if it makes a difference ;)14:07
greduanmake oldconfig has a pretty bad interface.14:08
greduanIf only it had colors to let me know where the old option ended and where the new one starts14:08
*** asie has joined #crux14:09
tired890done with -j8, I used time to measure how long it took:14:09
tired890make -j8 all  1097.05s user 60.83s system 729% cpu 2:38.82 total14:09
greduanThe kernel?14:10
tired890yes14:10
tired8903.18.414:10
greduandamn14:10
greduanthat is damn fast14:10
tired890greduan, yes its done on a tmpfs !14:10
tired890now I will do it again with -j100 ..14:10
greduanBy default CRUX doesn't use initrd or initramfs by default, right?14:13
tired890-j100 fail14:14
tired890its slower, you were right frinnst14:15
tired890will try again with -j2014:15
frinnstgreduan: the boot-cd uses an initrd14:15
tired890make -j100 all  1142.91s user 61.80s system 723% cpu 2:46.54 total14:15
greduanfrinnst: OK, thanks. :)14:16
*** jdolan has quit IRC14:17
frinnstand users are expected to build their own kernel so its not correct to say that crux uses initrd by default or not14:17
greduanRight14:18
greduanI meant if the install instructions, have the initrd or initramfs stuff done14:19
tired890make -j20 all  1115.76s user 59.23s system 727% cpu 2:41.56 total14:20
tired890need more cores14:21
*** jdolan has joined #crux14:40
*** BitPuffin has quit IRC14:51
thetornainbowi still need to bump my 3.18.2 kernel - haven't had any time to myself for the past two weeks it feels like14:51
greduanDo I need to compile ext2/ext3 support? I remember seeing somewhere that ext4 can be used for ext2,3 and 4?14:53
thetornainbowthere's no good reason to compile it unless you need it :)14:54
greduanI can replace /boot/vmlinuz without worrying about the current kernel right?15:04
greduanBy which I mean, it won't affect my current boot, will it?15:04
jaegerfrinnst: feel free :)15:07
thetornainbowright, the vmlinuz is read by the bootloader, you can replace it with the new one. although i've gotten into the habit of making a vmlinuz-backup and keeping last known good config with a secondary entry in GRUB (or whatever bootloader you like), just in case i bork something15:08
thetornainbowalthough i haven't done that in a while15:08
greduanIndeed15:09
greduanI do plan to keep around the old kernel, just because I can15:09
greduanI have a 500MB /boot for a reason.15:11
greduanWhich is an ext2 so I probably shouldn't have disabled ext2 support in the kernel. lol15:11
greduanIs it?15:11
greduanNow I need to make sure15:11
*** BitPuffin has joined #crux15:11
frinnstits hard to mount and copy your kernel to it if you lack support for it15:11
jaegerthe ext4 driver can handle ext2, ext3, and ext415:11
greduanIt is15:11
jaegerif that option is selected15:11
greduanjaeger: That option is selected, so good.15:12
greduanOh yeah, in fstab, besides my own entries the only required entry is the devpts one, ja?15:13
frinnstand /sys & /proc15:14
jaegerneither /sys nor /proc are needed in fstab15:14
*** onodera has quit IRC15:14
frinnstah, right. modern stuff15:14
greduanOK good lol15:14
greduanBut they are in the kernel right?15:14
jaegeryeah, they're still used, just mounted by /etc/rc15:14
greduanOK15:15
*** SiFuh has joined #crux15:15
greduanAnyone use(d) or tried the TextAdept before?15:15
jaegernot I15:16
greduanTextAdept editor*15:16
*** SiFuh has quit IRC15:20
*** SiFuh has joined #crux15:20
*** SiFuh has quit IRC15:33
*** SiFuh has joined #crux15:34
*** SiFuh has quit IRC15:39
*** SiFuh has joined #crux15:39
*** SiFuh has quit IRC15:44
*** SiFuh has joined #crux15:45
greduanSo everywhere I've seen they tell me to copy the kernel under arch/x86/, can I copy the kernel under arch/x86_64/ ?15:46
greduanPros and cons?15:47
greduanMy processor is a 64bit so I don't need the 32bit compatibility15:47
*** dougl has joined #crux15:47
frinnstyou need to rebuild the toolchain if you intend to exclude 32bit support in the kernel15:48
greduanI see15:49
*** SiFuh has quit IRC15:50
*** SiFuh has joined #crux15:50
greduanbrb, hopefully it boots correctly15:53
*** greduan has quit IRC15:53
*** SiFuh has quit IRC15:54
*** SiFuh has joined #crux15:55
*** greduan has joined #crux15:59
greduanIt did not work15:59
*** leo-unglaub has quit IRC16:00
jaegerdon't remove 32-bit, it isn't worth the effort16:00
jaegerjust don't install anything from compat-32 :)16:00
*** SiFuh has quit IRC16:00
*** SiFuh has joined #crux16:01
greduanjaeger: I didn't16:02
greduanThe boot worked but I guess I didn't compile support for my GPU16:02
greduanalso lilo tells me about /dev/mapper/control failing to open16:02
greduanRecompile, here I o16:04
greduancome*16:04
greduandunno how I missed 75% of the keystrokes. lol16:04
greduanrecompiling16:10
*** SiFuh has quit IRC16:10
*** SiFuh has joined #crux16:11
greduandang no KMS16:38
*** lnds has joined #crux16:46
greduanwhat's the best way to find out which SCSI low-level drivers I actually need?16:53
greduanI only want the kernel to compile with support for the harddrive it's currently using16:53
thetornainbowyou could use the bootstick and lsmod to see what's currently being used, but sometimes it's dangerous cause you won't catch everything17:02
*** quasinoxen has quit IRC17:07
greduanhmm17:09
greduanDunno why, no KMS17:09
joacimis this an actual scsi controller? or just some sata controller? if sata, you probably only need to enable ahci17:17
greduanTBH I've no idea. It works though. lol17:18
greduanI figured out the KMS stuff BTW17:21
tired890greduan, I always copy from x86_64 ..17:21
tired890I thought its the same thing?17:21
tired890I mean you select it from the menuconfig..17:21
greduantired890 that's why I was asking. but apparently there is a reason between those folders17:22
tired890greduan, no they are the same17:25
tired8902f20202dae57bf327325bb254189fd92  x86/boot/bzImage17:26
tired8902f20202dae57bf327325bb254189fd92  x86/boot/bzImage17:26
tired890x86_64 is only a link17:26
greduanoh17:26
greduanwell then17:26
tired890file x86_64/boot/bzImage17:27
tired890x86_64/boot/bzImage: symbolic link to ../../x86/boot/bzImage17:27
greduanI see17:28
jaegeryou need scsi support, scsi disk support, your controller, and your filesystem17:29
jaegereasiest way to find out which controller it is is to use lspci. lspci -k | grep "in use" is my preferred way17:30
cruxbot[opt.git/3.1]: [notify] apache: update to 2.4.1217:40
cruxbot[opt.git/3.1]: dropbear: update to 2015.6717:40
cruxbot[opt.git/3.1]: spectrwm: update to 2.6.217:40
tired890would it be alot of work to make prt-get do certain compiles in a tmpfs? say firefox or so17:46
tilmanalot of work?17:47
tilmani'm imagining it right now17:47
tired890tilman, imagining it?17:48
tilmanthe alot17:49
tilmanworking hard17:49
cruxbot[core.git/3.1]: btrfs-progs: updated to 3.18.217:49
*** BitPuffin has quit IRC17:49
jaegeralot wonder why have to work so hard17:49
tilmani forget the url17:49
jaegerhttp://hyperboleandahalf.blogspot.com/2010/04/alot-is-better-than-you-at-everything.html17:49
tired890I'm using fakeroot17:49
tilman=)17:50
jaegerI keep that one around, hehe17:50
tired890lame17:50
tired890:P17:50
tilmanyou misspelled lmao17:50
frinnsttired890: prt-get.conf - PKGMK_WORK_DIR=17:50
tired890I'm trying to learn to do it the right way..17:50
tired890fredrick, you are the man!17:50
tired890so edit this then edit it back..17:51
*** frinnst has quit IRC17:55
*** ZDM- has joined #crux18:00
*** dougl has quit IRC18:01
*** khanku has quit IRC18:07
*** xetver has joined #crux18:07
*** darfo has quit IRC18:08
*** diverse has quit IRC18:09
*** khanku has joined #crux18:09
tired890to answer my question (on how to switch to normal and tmpfs compiles) is to use "pkgmk -cf tmpfs.conf" , the alternate config file has dirs in tmpfs. If this can be passed to prt-get to fully automate it it would be cool18:10
*** krue has quit IRC18:12
*** frinnst has joined #crux18:15
*** pips- has joined #crux18:20
*** darfo has joined #crux18:20
*** diverse has joined #crux18:21
*** fnords has joined #crux18:22
cruxbot[xorg.git/3.1]: xkeyboard-config: update to 2.1418:39
*** leo-unglaub has joined #crux18:39
cruxbot[contrib.git/3.1]: pmwiki: update to 2.2.7218:40
cruxbot[core.git/3.1]: dhcpcd: update to 6.7.118:43
cruxbot[core.git/3.1]: libdevmapper: update to 1.02.93-218:43
*** leo-unglaub has quit IRC18:46
*** dougl has joined #crux18:56
greduanSo something that the kernel help for each entry tells you (most of the time) is something like "If unsure, say Y" or something like that. Does that mean make it a module or * (built-in)?18:59
greduanSo should it =y or =m or it doesn't matter?19:00
frinnstprobably not19:02
frinnstdepends on what it is19:02
frinnstfilesystem and stuff needed during boot needs to be built-in unless you use a initrd19:03
greduanI see19:03
*** pekka10 has quit IRC19:51
*** dougl has quit IRC19:57
*** hhhhhhhh_ has joined #crux20:09
*** leetspete1 has quit IRC20:38
*** hhhhhhhh_ has quit IRC20:58
*** BitPuffin has joined #crux21:01
*** dougl has joined #crux21:05
*** JSchilli1 has joined #crux21:14
*** leetspete1 has joined #crux21:21
*** jdolan has quit IRC21:25
*** jdolan has joined #crux21:26
greduanCompiling again21:34
greduanTook the time to go over every single option and enable only that which my hardware supports or I find important.21:34
greduanOr that it recommended to have Y if unsure.21:34
greduanHopefully I got the SCSI correctly, and hopefully I got KMS correctly as well21:35
greduanrebooting21:59
greduan /sbin/start_udev: line 20: /proc/sys/kernel/hotpluc: No such file or directory22:05
*** asie has quit IRC22:06
greduanAnd /dev/mapper/control: open failed: No such device22:06
*** lnds has quit IRC22:06
greduanI've CONFIG_SCSI_PROC_FS=y22:13
*** jdolan has quit IRC22:18
greduanCONFIG_DEVTMPFS=y CONFIG_DEVTMPFS_MOUNT=y22:19
greduanwhat am I missing here?22:19
tired890hi greduan22:24
tired890not sure if I can be helpful, but I think you removed some needed stuff. Btw I was once trying to do that (religiously check every thing and remove anything I thought wasn't necessary).. difference in boot time or memory footprint is minimal..22:25
tired890if you have any kernel known to get your device up and running then check loaded modules/drivers to identify the needed driver22:26
tired890I start with the .conf included with stock crux, remove unneeded FS support, remove clearly unneeded drivers (eg I see realtek NIC drivers while I have atheros etc) and keep things I'm not sure about22:28
tired890what would help is faster compiles for all that experimenting.. for that try this (given you have 1.5g free ram):22:31
tired890mount -t tmpfs -o size=2g tmpfs /mnt/tmpfs22:32
tired890cp linux_kernel.tar.xz /mnt/tmpfs22:32
tired890and fire away!22:32
greduanI also start with the stock crux .config, that's what I did with 3.12 but there's no such file for 3.18.5 :/22:35
tired890just copy the .conf file22:36
tired890copy it from 3.12 to 3.1822:36
greduanI did do that, but I went through all the options in order to check all the "(NEW)" stuff22:36
greduanI also had to go through my GPU stuff again because that stuff got reset, some of it, it seems22:37
tired890with stock crux does your scsi thing work ?22:38
greduanIt works right now22:38
greduanExcept it's telling me bout the /proc and /sys stuff22:39
tired890cool22:39
tired890oh22:39
greduanBut I didn't change any of those settings I don't think22:39
greduanI also just onfirmed that I enabled /proc and /sys22:39
greduanAlso, I can boot correctly, which means the necessary stuff to boot from my harddrive is present22:46
*** tired890 has quit IRC22:49
*** tired890 has joined #crux22:51
greduanI think it has to do with one option I disabled (I think) which I m trying to find right now22:51
tired890gl mate :)22:52
greduanlol thx22:52
tired890do make a copy of that .conf once all sorted! I just rebooted into 3.18.5, upgraded from 3.18.4 .. just copied the .conf file and make all...etc done! headache only once, then its smooth sailing :D22:53
teK_diverse: bug found and fixed by the author :))22:53
greduantired890: Indeed22:55
greduanfound it22:56
tired890what is it? (for IRC logger!)22:56
greduanDevice Drivers > Multiple devices driver support (RAID and LVM) > device mapper support22:57
tired890tmpfs compile time!22:57
greduanwill do22:57
greduanafk23:02
*** jdolan has joined #crux23:12
greduan back23:33
greduanLooks like that device mapper stuff didn't fix the dev mapper stuff23:33
greduanit's telling me about incompatible libdevmapper and kernel (or something along those lines I guess)23:33
tired890oh dear23:34
greduanTried recompiling libdevmapper23:36
greduandidn't work23:36
frinnstwhen using a kernel config from an older version, you can run "make oldconfig" - you will get prompted for the new stuff23:47
frinnstits usually much more simple than to run menuconfig and hunt down the new stuff23:49
tired890:)23:49
greduanYeh but the interface was so time consuming. Half of the time I was trying to figure out what was the new text tht appeared.23:50
greduanCompiling with tmpfs this time23:50
tired890:D23:53
tired890keep in mind though greduan, you might need the built sources for module building later23:54
tired890for example I needed to rebuild vboxdrv (virtualbox kernel module) so when I rebooted into 3.18.5 and tried to recompile it failed asking for the new kernel src23:54
tired890so after compile is done, copy the folder linux-3.18.5 to /usr/src23:55
greduanI just need the sources, right?23:56
greduanI was compiling with /usr/src/linux-3.18.5 before copying it to tmpfs23:56
*** nilp has joined #crux23:57
tired890built I believe, IIRC it needs kconfig or something23:58
tired890yup if you have it built there then that should be ok.23:58

Generated by irclog2html.py 2.11.0 by Marius Gedminas - find it at mg.pov.lt!