| led-bandit | lib6 error on update tonight when i got home cant boot into system now | 00:39 |
|---|---|---|
| led-bandit | and rescue mode will not finish loading | 00:39 |
| led-bandit | libc6-dev-i386_2.17-0ubuntu5_amd64.deb to be exact | 00:47 |
| led-bandit | A copy of the C library was found in an unexpected directory: | 00:57 |
| led-bandit | '/lib/i386-linux-gnu/ld-2.17.so.dpkg-new' | 00:57 |
| led-bandit | It is not safe to upgrade the C library in this situation; | 00:57 |
| led-bandit | please remove that copy of the C library or get it out of | 00:57 |
| led-bandit | '/lib/i386-linux-gnu' and try again. | 00:57 |
| led-bandit | has chroot changed for 13.04? | 01:05 |
| bjsnider | led-bandit, http://packages.ubuntu.com/raring/amd64/libc6/filelist | 01:06 |
| bjsnider | that should have been the i386 list, but whatever, just change the url | 01:07 |
| bjsnider | i386 is not appropriate to use on linux | 01:07 |
| bjsnider | just an opinion | 01:07 |
| led-bandit | i use multi lib to run i386 apps | 01:07 |
| led-bandit | obviously somthing borked the install so that my system is not bootable | 01:08 |
| led-bandit | this update was pushed out 6 hours ago so hope others dont start having same problem | 01:09 |
| bandit-led | A copy of the C library was found in an unexpected directory: | 02:59 |
| bandit-led | '/lib/i386-linux-gnu/ld-2.17.so.dpkg-new' | 02:59 |
| bandit-led | It is not safe to upgrade the C library in this situation; | 02:59 |
| bandit-led | please remove that copy of the C library or get it out of | 02:59 |
| bandit-led | '/lib/i386-linux-gnu' and try again. | 02:59 |
| bandit-led | fix http://forum.ubuntu-it.org/viewtopic.php?f=30&t=541711&sid=588a1d84d0d07e6f5079afd9670b5e54&start=40#p4262699 | 02:59 |
| richerVE | Hi, it isn't today the release of 13.04 RC ?? | 03:04 |
| c_smith | Hey, I'm testing Ubuntu 13.04 and just noticed a BIG issue: my HDMI sound is not appearing in the sound menu with the FGLRX 13.3 beta, neither with the Ubuntu supplied ones or the ones in the repos. | 03:11 |
| c_smith | I asked about this a while ago on Ubuntu Gnome's IRC channel with their beta, and they referred me here. | 03:11 |
| c_smith | can't test the open-source drivers as with them I get a kernel panic shortly after boot | 03:12 |
| c_smith | GPU is too new for them to work at all. | 03:13 |
| c_smith | I've found a solution: Use the cards VGA port with an audio cord for the Monitor, not a perfect solution, but I can get by with it | 03:39 |
| valorie | c_smith: did you file a bug report? | 03:53 |
| valorie | the devels may or may not read the chan logs | 03:53 |
| c_smith | valorie, I'm in the process of doing so now. | 03:53 |
| c_smith | just need to retrieve my Launchpad login from Keepass | 03:54 |
| valorie | cool | 03:54 |
| c_smith | though would Pulse fall under the Linux package? | 03:54 |
| c_smith | in this case, that is. | 03:54 |
| valorie | pulseaudio? you were talking about fglrx earlier | 03:55 |
| c_smith | I was. aside from the sound issue, HDMI works. | 03:56 |
| c_smith | kinda confused as to what I should file it under. | 03:56 |
| c_smith | whether it be FGLRX, Pulse, or Linux | 03:56 |
| valorie | you might ask in #ubuntu-kernel | 03:57 |
| valorie | kernel panic sounds important | 03:57 |
| c_smith | alright | 03:58 |
| valorie | very helpful people there too | 03:58 |
| c_smith | the Kernel Panic is unrelated to the issue I was talking about, but it makes the open source AMD drivers useless on this card (and I can pin it to upstream as it also affects every distribution I've tried, Arch, openSUSE, Ubuntu, you name it) | 03:59 |
| c_smith | heck, I've even tried it on PCLinuxOS. which usually lags behind a bit. | 04:00 |
| c_smith | gonna file the current bug I'm talking about in FGLRX, seems to fit best. | 04:01 |
| valorie | ok | 04:02 |
| valorie | sounds like you have two bugs then | 04:03 |
| Wutzan | Hey, wasn't the RC supposed to be out today? | 04:04 |
| c_smith | valorie, true, but my Radeon card is actually pretty new (the Turks line, Radeon HD 6670) so it could be that they just haven't implemented the needed features in the free driver. | 04:06 |
| c_smith | hmmmmm...... looking at the apt-cache policy, the AMD driver versions in the repo and the version from the site vary quite a bit. | 04:07 |
| c_smith | though 13.1 directly from AMD is unusable in 13.04 | 04:08 |
| c_smith | if this works, it tells me it's only something AMD can fix, as it does appear the devs may have rolled back the version | 04:11 |
| c_smith | *people who maintain the FGLRX packages | 04:11 |
| valorie | those Very Important People | 04:14 |
| ShapeShifter499 | hi again | 04:18 |
| ShapeShifter499 | I was checking out the cool looking updated "shutdown" prompt you'd get after hitting the computer's power button, I wasn't planning on shutting down my system but even though I exited out of the prompt by clicking the "X" in the upper left corner of the prompt my system still shut itself down about a minute later, why? is this a bug? | 04:21 |
| === azend is now known as DGNP | ||
| === DGNP is now known as Azend | ||
| === Azend is now known as azend | ||
| ShapeShifter499 | anyone know? | 04:34 |
| valorie | sounds like you selected it without meaning to | 04:36 |
| Paul_MyBB | I would test, but I don't want to lose my work on this essay :P | 04:36 |
| * valorie uses kubuntu, not Unity | 04:36 | |
| ShapeShifter499 | valorie, I don't see how, the shutdown button area is on the right side, opposite of the exit button | 04:40 |
| valorie | ok, haven't seen it | 04:44 |
| Coreyon | Is the RC out yet? | 05:11 |
| valorie | !release | 05:15 |
| ubottu | Ubuntu releases a new version every 6 months. Each version is supported for 18 months to 5 years. More info at https://wiki.ubuntu.com/Releases & http://wiki.ubuntu.com/TimeBasedReleases | 05:15 |
| valorie | huh | 05:15 |
| valorie | not here: http://cdimage.ubuntu.com/releases/raring/ | 05:18 |
| lordievader | Good morning. | 05:50 |
| pabs3 | where can I find the alternate installer for Ubuntu raring? | 05:57 |
| bazhang | !alternate | pabs3 | 06:18 |
| ubottu | pabs3: The alternate CD has been discontinued for the main Ubuntu distro, please use and report any bugs in the !LiveCD | 06:18 |
| pabs3 | hmm, does the main installer support preseeding? | 06:19 |
| jackw | when 13.04 is going to be released? | 06:37 |
| jackw | cant wait | 06:37 |
| bazhang | !schedule | jackw | 06:38 |
| ubottu | jackw: Raring Ringtail (13.04) release milestones can be found at https://wiki.ubuntu.com/RaringRingtail/ReleaseSchedule | 06:38 |
| lordievader | jackw: The 25th. | 06:39 |
| nonix4 | Is there more formal/correct way to make "mdadm -S" possible than an infinite loop of "while true; do sudo vgchange -a n /dev/ubuntu-vg; done" while doing that mdadm -S? As in how to disable auto-reactivation of LVM VGs? | 07:53 |
| === Amaranthus is now known as Amaranth | ||
| MonkeyDust | nonix4 i'm not familiar with LVM, but it says here, that they are never automatically activated -- there's also the channel #lvm2 | 11:46 |
| nonix4 | MonkeyDust: iirc the thing that activates such things automatically was related to gvfs-gdu-something... | 11:48 |
| MonkeyDust | nonix4 75+ people in #lvm, better ask there | 11:49 |
| MonkeyDust | nonix4 75+ people in #lvm2, better ask there <-- correction | 11:49 |
| davidbaumann | Hell together. I submitted my first bug yesterday, but I guess I made some mistakes ;) | 11:49 |
| nonix4 | think some gnome channel might be more appropriate place... | 11:49 |
| davidbaumann | 1170463, it's about using ALFA AWUS036NHR on Kubuntu 13.04 x64 | 11:50 |
| yeahuyen | having trouble using ubuntu one to backup my .home directory, getting this error: :Giving up on request after 5 attempts, last status 400 Bad Request | 11:50 |
| yeahuyen | /home directory i mean | 11:51 |
| kvarley | Skype fails to load due to a segmentation fault, any ideas how to fix this? | 11:59 |
| caf4926 | did anyone else notice a problem with Chrome installing in the latest +1 | 12:55 |
| caf4926 | specifically libudev0 (>=147) | 12:58 |
| leolove | Hi | 13:00 |
| leolove | I just updated my 13.04 to new kernel 3.8.0-19 and all of sudden sound stopped working | 13:00 |
| leolove | /proc/asound/cards is empty, card detected in lspci, /blacklist-oss.conf contains blacklist SoundCard | 13:01 |
| leolove | any idea? | 13:03 |
| philinux | leolove: boot the previous kernel and see if all ok | 13:09 |
| leolove | Yes tested, it works | 13:10 |
| leolove | previous version was 3.8.0-18 | 13:10 |
| philinux | leolove: boot into 19 and in terminal ubuntu-bug linux and send off a bug report | 13:11 |
| leolove | philinux done | 13:14 |
| leolove | seems alsa modules not loaded | 13:14 |
| philinux | leolove: hang on | 13:15 |
| philinux | I have sound but /proc/asounds/cards is cat: /proc/asounds/cards: No such file or directory | 13:16 |
| philinux | not sure how relevant that is | 13:17 |
| leolove | wait, let me share the link | 13:18 |
| philinux | leolove: i've just done a sound test in sound settigns and all fine here | 13:18 |
| philinux | leolove: maybe hardware specific | 13:18 |
| philinux | leolove: mine is " Audio device: Intel Corporation 82801I (ICH9 Family) HD Audio Controller (rev 03) | 13:19 |
| leolove | Audio device: Advanced Micro Devices [AMD] FCH Azalia Controller | 13:19 |
| philinux | leolove: I'll check my other pc later | 13:20 |
| philinux | whats the bug report number | 13:20 |
| philinux | I'll be back later | 13:21 |
| leolove | 1170697 | 13:23 |
| leolove | philinux are you there? | 13:34 |
| leolove | I just sudo apt-get install --reinstall linux-image-$(uname -r) and it worked now | 13:34 |
| wilee-nilee | habanany, Have you checked the usb to see if it is full, a persistent will fill up and is really difficult to clean without a hack. | 13:37 |
| habanany | i guess im gonna have to stop the process | 13:38 |
| BluesKaj | Hi all | 13:39 |
| leolove | habanany you are facing issue to erase usb? | 13:40 |
| wilee-nilee | habanany, Updating a persistent is problematic all of the updates are in the casper-rw, kernel upgrade may brick it, so that you have to remove the casper and just have the iso. | 13:40 |
| wilee-nilee | leolove, No a stuck update. | 13:40 |
| habanany | leolove i ll hist going to reboot | 13:41 |
| habanany | just ** | 13:41 |
| leolove | seems ubuntu / AMD had problems with switchable graphics | 13:44 |
| BluesKaj | most switchable graphics are a problem on Linux, period | 13:46 |
| habanany | leolove it seems like terminal was sleeping | 13:46 |
| habanany | leolove i had this apport stop/waiting | 13:53 |
| leolove | BluesKaj I was fighting with tearing switchable graphics and followed every guide. Nothing works but a small trick | 13:53 |
| leolove | habanany okay | 13:54 |
| leolove | so it resolves now? | 13:54 |
| BluesKaj | leolove, then save that trick or an expalanation in text form, so that you can share wih others who may have the same issue | 13:55 |
| BluesKaj | leolove, I know the that the nvidia dual graphics switch seems to work for some using bumblebee | 13:56 |
| BluesKaj | aka optimus | 13:57 |
| leolove | BluesKaj, when we run amdconfig --initial to generate xorg.conf somehow the tearfree desktop option dissappear. | 13:58 |
| leolove | Just delete the xorg.conf and restart the pc, it will use fglrx but without your discrete graphics. You can then enable tearfree | 13:59 |
| uictamale | Hey all, I just installed 13.04 and my background is solid white and doesn't change when I pick a new background. Any ideas? | 14:15 |
| slinnky | w00 laaaag | 14:19 |
| genii-around | All the available nvidia drivers now properly install/deinstall and work for me again on x86_64 with the -19-generic kernel and headers, if anyone's interested. | 15:06 |
| genii-around | After re-enabling xorg-edgers the 313 from there also works fine. | 15:10 |
| BluesKaj | ok genii-around , thanks for the heads up | 15:14 |
| philinux | leolove: audio sorted? | 15:14 |
| BluesKaj | genii-around, think I'll pass on the xorg-edgers 313 ppa . Past experience with that driver wasn't good. | 15:31 |
| genii-around | BluesKaj: I'm currently using it, but of course Your Mileage May Vary! | 15:37 |
| Dreaman | http://picbg.net/img.php?file=8dad8206e6727931.png my ubuntu 13.04 rc bulgarian and iron maiden :) | 15:37 |
| BluesKaj | genii-around, yup :) | 15:37 |
| Xajinal | Does anyone know if Raring Ringtail has speed improvements? | 15:40 |
| wilee-nilee | Xajinal, What is your definition of speed improvements? | 15:42 |
| Xajinal | Ignoring the laggy UI (I use a tiling WM), I mean improvements to the compilation of core packages | 15:43 |
| Xajinal | The set of CFlags used in previous versions hasn't been very optimized | 15:43 |
| Xajinal | I get noticeable speed improvements compiling from source as opposed to the default Ubuntu packages | 15:44 |
| BluesKaj | genii-around, the 310.44 driver seems to be working fine . It didn't in the past . | 15:54 |
| Xajinal | Hmm... seems no-one knows what CFLAGs are | 15:55 |
| genii-around | BluesKaj: I tested the -experimental-304, -experimental-310, and the 313-updates which all installed/ran/deinstalled fine and survived reboot, then got brave and re-enabled the xorg-edgers and installed that, which seems to be working fine so far | 15:57 |
| BluesKaj | genii-around, did you get a list of messages during the reboot after installing the non-default drivers and then a stall , which resulted in a complete shutdown shortly after ? | 16:00 |
| BluesKaj | but the startup was normal afterwards | 16:02 |
| alankila | Xajinal: if you claim improvements, produce measurements for performance. Subjective assessments are useless because human observers are generally extremely biased | 16:10 |
| alankila | in general it is not unthinkable that some compilation flags help performance but they may also harm performance, it's not always a win... unfortunately. | 16:11 |
| alankila | and since we went 64 bits the bar was raised very high with respect to CPU capabilities. SSE2 can be assumed to be present etc. | 16:12 |
| Xajinal | alankila: I have results from tests naturally, but previously when I've filed bug reports to launchpad they been ignored | 16:13 |
| jtaylor | Xajinal: example for an CFLAGS? | 16:13 |
| Xajinal | jtaylor: CFLAGS are variables which specify additional parameters to the compiler | 16:14 |
| alankila | as a rule I'm personally interested in improvements larger than 10 %. That is something that begins to be noticeable to disinterested observers | 16:14 |
| jtaylor | I know what they are | 16:14 |
| Xajinal | jtaylor: Why were you asking then? I'm confused | 16:14 |
| jtaylor | 17:43 <Xajinal> The set of CFlags used in previous versions hasn't been very optimized | 16:15 |
| Xajinal | I don't have the source to hand, but when I looked in previous versions (particularly 12.04 when I became interested) they seemed crude | 16:15 |
| alankila | he just wants to see one good cflags line that you claim to improve performance of some package | 16:16 |
| Xajinal | Although I understand a lot of Ubuntu software really just comes from Debian, which is cautious in its use of optimization | 16:16 |
| Xajinal | I don't have the files available, on my gentoo box at the moment | 16:17 |
| Xajinal | I'll file another launchpad ticket I guess, though responses to those seem pretty low | 16:17 |
| alankila | yes, debian has certain diseases which boil down to being extremely infatuated with obsolete hardware | 16:17 |
| jtaylor | Can you point me to an existing launchpad ticket? | 16:18 |
| genii-around | BluesKaj: Nope, went smooth as silk | 16:18 |
| Xajinal | Search for "CFLAGS" on the tracker | 16:18 |
| jtaylor | Xajinal: more specific please | 16:19 |
| BluesKaj | ok genii-around thanks , it must be a reboot glitch of some kind. I'm going to test it again , brb. | 16:19 |
| jtaylor | you must also take into account that ubuntu is a binary distribution, so you can't use stuff like -march=haswell | 16:20 |
| alankila | I can't find anything relevant with respect to cflags query on the launchpad bugs. 275 bugs, 12300 pages, nothing seemingly relevant for 'cflags optimization'. Giving up. | 16:21 |
| alankila | I've seen that some bunch of benchmarks show some 10-15 % improvement when built with -O3 than -O2 | 16:24 |
| jtaylor | the set of application that profits from O3 is quite limited | 16:25 |
| jtaylor | the most major improvement is the activation of the autovectorizer | 16:25 |
| jtaylor | and that only helps if you do some kind of number crunching | 16:25 |
| alankila | http://compcert.inria.fr/compcert-C.html#perfs is a fairly typical result | 16:25 |
| alankila | Oh bummer | 16:26 |
| alankila | this doesn't have -O2 result | 16:26 |
| alankila | never mind. For some reason I thought it did have | 16:26 |
| genii-around | You could always just use apt-build or such | 16:26 |
| alankila | anyway what you can see from this is that even a stupid compiler like compcert (relatively speaking) gives most of the bang of the bug, and gcc's extensive optimizations in O3 mode tend to make few % of difference. This is, I guess, why most people aren't interested in cflags | 16:27 |
| alankila | bang for the buck too. Need to concentrate on what I'm writing. | 16:27 |
| jtaylor | a few percent in exchange to frequent very hard to find bugs | 16:28 |
| alankila | It is annoyingly hard to find proper benchmarks where someone has extensively studied variety of workloads and program types at large set of compiler flags | 16:30 |
| alankila | all the search results are dominated by various gentoo related hits which isn't helping. No hard data anywhere. | 16:31 |
| jtaylor | its not a very useful thing to do | 16:31 |
| jtaylor | it depends on too many factors | 16:31 |
| alankila | it is, however, in principle a scriptable problem. Take a hardware, run battery of standard tests under various CFLAGS, collect results, publish. Whenever compiler updates, repeat. | 16:32 |
| jtaylor | phoronix does this type of thing | 16:32 |
| alankila | then slice the data in two dimensions, once through cflags and once through compiler versions | 16:32 |
| jtaylor | but it doesn't mean the results you get from that can be used to draw any conclusions | 16:32 |
| alankila | iirc the last time I saw someone do this cflags didn't matter basically one bit but compiler version had huge impact | 16:32 |
| jtaylor | unless the difference is huge you can only say on this exact system that is better | 16:32 |
| alankila | also gcc versions don't always improve performance. sometimes it regresses. | 16:33 |
| alankila | the one parameter that might be very interesting is link time code generation and/or whole program optimization related stuff | 16:34 |
| alankila | that has a real chance of making a big impact | 16:34 |
| jtaylor | In my experience it does not help much | 16:34 |
| alankila | gcc has been very poor with this though, for instance taking huge amounts of memory to do link time code generation for real programs | 16:34 |
| alankila | oh so you have tried it? I've some android programs where I'd like to try it but IDK if the more primitive ARM cpus would benefit more | 16:34 |
| alankila | these intel processors with their fancy speculative executions and all that might really eliminate the impact of subroutine calls | 16:35 |
| jtaylor | it of course depends on the software | 16:35 |
| jtaylor | but usually if there is a real bottleneck due to missing lto it has been fixed in other ways | 16:35 |
| jtaylor | like source file concatination or just putting the small functions into headers | 16:35 |
| alankila | though one really nice thing about link time code generation is that you can write the code in a normal way and let the compiler inline it rather than put half of your implementation into a header | 16:36 |
| alankila | so I imagine ltcg will fix that at least, even if it just gives same performance as you had before | 16:36 |
| jtaylor | yes, its certainly good | 16:36 |
| alankila | it's kind of criminal how poor gcc has been and for how long. optimizing-wise, it has always been beaten by icc that is apparently done in fraction of time, and quality of error messages only got better after clang lit a fire under gcc's tail. Competition is good, even for free software. And MSVC, despite being much loathed, has done LTCG forever. I suspect that compiler beats GCC too in terms of executable code performance. | 16:40 |
| jtaylor | I don'T believe the icc benchmarks | 16:40 |
| jtaylor | haven't seen a single one that convinces me its any good | 16:40 |
| alankila | this is largely uninformed speculation on my part, but I hope that the C++ switch will help over time. | 16:40 |
| jtaylor | gcc has advanced a lot in recent times | 16:41 |
| alankila | oh, I've seen some math-related benchmarks where it was generally ahead though later GCC versions have had a habit of catching up with ICC | 16:41 |
| jtaylor | e.g. the autovectorizer which was always the strongpoint of icc, is really good now | 16:41 |
| jtaylor | the difference is its not enabled by default | 16:41 |
| jtaylor | also icc favors performance over correctness | 16:41 |
| alankila | still not? that seems a shame | 16:41 |
| jtaylor | gcc always favors correctness | 16:42 |
| jtaylor | but allows to disable that | 16:42 |
| jtaylor | thats usually the issue with benchmarks | 16:42 |
| jtaylor | not comparable options | 16:42 |
| alankila | so -ffast-math for win? | 16:42 |
| jtaylor | gcc with options equivalent to icc's default is very likely as fast or faster | 16:42 |
| alankila | I actually use that option sometimes. Anything is better than call to some libm intrinsic. Those library functions are so heavy and slow... :-/ | 16:42 |
| alankila | sorry, libm function | 16:43 |
| jtaylor | glibc has improved in that respect a lot too | 16:43 |
| jtaylor | also many non math functions are much better now | 16:43 |
| jtaylor | e.g. memcpy is vectorized and superfast now | 16:43 |
| jtaylor | independend of compilation | 16:43 |
| jtaylor | its all runtime detected | 16:43 |
| jtaylor | very neat | 16:43 |
| alankila | I kinda think much of C library shouldn't even exist. Only stuff that's required to interface with linux kernel should be there, everything else should be inlinable functions of the compiler | 16:43 |
| jtaylor | no | 16:44 |
| jtaylor | the opposite is the case now | 16:44 |
| jtaylor | gcc inlines are often slower than glibc | 16:44 |
| alankila | well gcc inlines could be improved to take whatever glibc does | 16:44 |
| jtaylor | because glibc can imply much more on the system than gcc | 16:44 |
| jtaylor | not in the same way | 16:44 |
| alankila | but the point is, the call overhead can be pretty huge and there's no way to do partial evaluation there | 16:44 |
| jtaylor | function call cost is not so high on amd64 | 16:44 |
| alankila | so even if you know that some value is not zero, say, the function must test it anyway because it can't be specialized | 16:45 |
| alankila | this is largely the issue with the math functions :-/ | 16:45 |
| alankila | they have to do a lot of this sort of crap before they get to the actual operation | 16:45 |
| jtaylor | true | 16:45 |
| jtaylor | for so smaller math operations its not so nice | 16:45 |
| jtaylor | but more expensive operations like log sin,cos etc the operation cost is pretty high compared to the check | 16:46 |
| alankila | I guess if you need transcendentals you've lost the game no matter what... it tends to go like that | 16:46 |
| jtaylor | though I'm not sure if glibc actually beats IMKL yet | 16:46 |
| jtaylor | problably not | 16:46 |
| alankila | I wonder if we could write x86-64 vm that reads the binary and shared library and uses dynamic inlining and dead code elimination and gets higher performance than executing it on bare metal will | 16:49 |
| alankila | there used to be a VM like this which actually ran some powerpc code faster than bare metal could, which is pretty impressive I guess. In that case it was reportedly the fact that it rewrote the code without branches as much as possible, which were apparently very slow on the hardware | 16:50 |
| alankila | nothing of that sort helps with intel because the processor is so smart, but perhaps we can run dumber processors if we have better optimizing runtimes and get more cores for lower watts | 16:50 |
| alankila | What really kills me about the stuff like SSE/SSE2 SIMD is that so few places can actually use any of that capability. If GCC stilll doesn't generate vectorized code by default, you can bet it sits unused for almost all programs. Only if programmer has taken the time to write some intrinsics or assembly, or build it with nonstandard options will anything use those SIMD instructions. | 16:52 |
| alankila | for instance I am a java programmer and I know for a fact that this dynamic compiler can't use the SSE unit for anything. It's pretty horrible. A much simpler CPU would do just as well if you employ it to run JVM. | 16:53 |
| jtaylor | in standard software opportunity to vectorize does not occur that often | 16:53 |
| alankila | yes, precisely. | 16:53 |
| jtaylor | with AVX2 it might get better | 16:54 |
| jtaylor | it has much better operations for scatter/gather and blending | 16:54 |
| alankila | which is why I think there's a case for Very Simple cpu. The sort of core that has no bells and whistles other than getting stuff our stupid code generators actually does emit to run fast | 16:54 |
| jtaylor | that is a bit what intel tried with itanium | 16:55 |
| alankila | iirc with C you have to use the 'restrict' keyword to allow vectorizing. C is so ill specced that it must assume that pointers can overlap and it only has the rule that pointers of different type of values do not overlap | 16:55 |
| jtaylor | it is a much simpler cpu, one hoped to be able to do the optimizing in the compiler | 16:55 |
| jtaylor | that didn't work out well | 16:55 |
| alankila | much good that does for you if you want to SIMD some dot product loop, everything will probably be float32 type | 16:55 |
| jtaylor | gcc will do runtime aliasing checks to get around that | 16:56 |
| jtaylor | alignment is probably a larger issue | 16:56 |
| alankila | yes, itanium went wrong somewhere but it doesn't eliminate the point of putting a lot of performance per watt out by having simpler ISA which fits actual use cases better... | 16:56 |
| alankila | oh. Man, that's news to me. So GCC will check this sort of thing? That's crazy. | 16:56 |
| jtaylor | it can also backfire :) | 16:57 |
| alankila | yes the alignment is a real problem too. you want movaps, not movups | 16:57 |
| jtaylor | if you compile c++11 array<4, int> it will stupidly add alignment and aliasing checks which is much more expensive than 4 scaler operations :) | 16:57 |
| alankila | but for instance C++ new operator had no way to specify alignment for allocation. All you know is that the address will be divisible by 4. Great. | 16:58 |
| jtaylor | C11 has a aligned_malloc | 16:58 |
| alankila | a good reason to look into that then. | 16:58 |
| jtaylor | we should probably move to a PM, kind of off topic | 16:59 |
| alankila | I might actually go play a game... | 16:59 |
| ChogyDan | meh, unity is not working for me | 17:01 |
| ChogyDan | anyone know of log files I can check, or anything? | 17:01 |
| ChogyDan | or how I can restart unity | 17:02 |
| wololo_226 | hi, I'm trying to install Kubuntu 13.04, but installer stalls at the "Preparation to install" screen | 17:08 |
| alo21 | hi... I have a problem with my wireless card on 13.04 live mode | 17:15 |
| ChogyDan2 | Im still having unity issues. Any help would be appreciated | 17:17 |
| alo21 | when I go to 'software source', into tab, 'Additional drivers', I select on the b43 driver. During the installation of that driver, my PC crashes | 17:17 |
| alo21 | my first question is: Do I have a network connection to install my b43 driver, or the files are in the kernel yet? | 17:19 |
| ChogyDan2 | alo21: is this on 13.04? my guess would be you need a net connection | 17:20 |
| alo21 | ChogyDan2, yes.. is on 13.04. So you think that my PC will not crash if I connect it to the net? | 17:21 |
| ChogyDan2 | man, this sucks. This is the worst state my Ubuntu install has been in in years :( | 17:23 |
| LukeL | 13.04 is for testing | 17:30 |
| ChogyDan | unity seems to be failing. Anyone know where relevant log files are? or other ways to see what is happening? | 17:44 |
| Volstar | how am I supposed to install Kubuntu 13.04 while installer is broken? how it is possible such bugs slips through testing? | 17:49 |
| wilee-nilee | Volstar, Have you summed the iso or disc? | 17:50 |
| Volstar | wilee-nilee: iso, I tried beta 2 and the latest daily build, so I wasted two dvd | 17:50 |
| wilee-nilee | !md5 | 17:51 |
| ubottu | To verify your Ubuntu ISO image (or other files for which an MD5 checksum is provided), see http://help.ubuntu.com/community/HowToMD5SUM or http://www.linuxquestions.org/linux/answers/LQ_ISO/Checking_the_md5sum_in_Windows | 17:51 |
| Volstar | wilee-nilee: it's not possible to download two images with errors, but I check this in a moment | 17:51 |
| wilee-nilee | Volstar, Make sure they are burned slow as well if the sum matches | 17:51 |
| Volstar | wilee-nilee: I used 8x which was the lowest possible speed | 17:52 |
| Volstar | the beta md5 sum is correct | 17:53 |
| Volstar | however, there's a chance I burned beta twice instead of the latest daily build | 17:54 |
| wilee-nilee | Volstar, So in the past all kubuntu/ubuntu installs have gone fine, and you have looked on the web for any problems with 13.04 and your hardware? | 17:54 |
| Volstar | but, I also booted into desktop and updated ubiquity, so the ubiquity from the latest build should be used, right? | 17:54 |
| wilee-nilee | not sure on that. | 17:55 |
| Volstar | wilee-nille: I found forum threads about ubiquity to be broken in 13.04 beta/beta 2 | 17:55 |
| === yofel_ is now known as yofel | ||
| Volstar | wilee-nilee: but those people were saying it was fixed later | 17:55 |
| wilee-nilee | Volstar, Ah, I have not had to install it again mine is from a earlier install. | 17:56 |
| Volstar | wilee-nilee: maybe I'll do the same, but I will make sure I have burned the latest daily build as well | 17:56 |
| wilee-nilee | Volstar, I'm not a real tester I install than clone it for an easy reload. | 17:57 |
| wilee-nilee | if needed. | 17:57 |
| Volstar | clone? | 17:58 |
| wilee-nilee | Volstar, Yeah with clonezilla. | 17:58 |
| Volstar | wilee-nilee: it seems I indeed burned beta twice, so I'll try the latest build now | 18:02 |
| BluesKaj | I hope the ubiquity freezing and not recognizing locale and amd HW issue is solved before the official release , just tried a daily a few days ago and it still exists for my setup | 18:02 |
| wilee-nilee | BluesKaj, Yeah, that would be nice with the server downloads being bad enough on releases. | 18:03 |
| BluesKaj | seems it's not being taken seriously , at least that my perception, wilee-nilee | 18:07 |
| BluesKaj | it's been known for over 4 months, but it might difficult to recreate , so hard to solve | 18:08 |
| LukeL | everyone speaks english and uses intel | 18:13 |
| genii-around | Well, maybe only 98% | 18:13 |
| sungyo | sshfs#iam@server:/home/foo /home/foo/boo fuse defaults,idmap=user,allow_other,users,gid=1000,uid=1000,IdentityFile=/root/.ssh/id_rsa 0 0 | 18:36 |
| sungyo | Im ubuntu 13.04 and that's my content of 'fstab' file | 18:36 |
| sungyo | and it's not working. | 18:36 |
| sungyo | When i try to mount /home/foo/boo, It saks me the password of server. | 18:37 |
| sungyo | And key file also works well. | 18:38 |
| sungyo | Only, after booting, it's not mounted. | 18:38 |
| sungyo | What can be the problem? | 18:38 |
| sungyo | (And I'll take my sleep. Does anyone answer me, I'll check and try it.) | 18:40 |
| Volstar | I've tried the latest build, but the installer is still broken | 18:42 |
| === jack is now known as Guest30154 | ||
| Paul_MyBB | Hallo | 19:06 |
| Paul_MyBB | Does anyone know where I can find the location of the PNG/SVG used for the Ubuntu launcher icon? | 19:06 |
| bjsnider | icons are usually kept in /usr/share | 19:27 |
| bjsnider | the default location of arch-indep stuff | 19:27 |
| trism | I think it's /usr/share/unity/6/launcher_bfb.png but he already left | 19:27 |
| ChogyDan | qum_: it just seems strange that you are running kernel 3.7 | 19:38 |
| qum_ | Well ChogyDan, i'm kinda new to linux | 19:38 |
| qum_ | I just downloaded ubuntu yesterday, installed today and here i am, trying to get things "right" :D | 19:39 |
| wilee-nilee | qum_, Did you get the drivers from nvidia? | 19:40 |
| guntbert | qum_: for someone new to linux using a beta release is not really recommended | 19:40 |
| ChogyDan | qum_: or did you get them from the edgers ppa? | 19:40 |
| ChogyDan | qum_: you should probably try to just do a general upgrade or something | 19:40 |
| qum_ | wilee-nilee: i think i got them from edgers... | 19:40 |
| qum_ | i got them from the additional drivers page | 19:41 |
| qum_ | after adding their ppa | 19:41 |
| qum_ | i think :D | 19:41 |
| wilee-nilee | qum_, Cool, enjoy. ;) | 19:41 |
| qum_ | and, what exactly do u mean by general upgrade? | 19:41 |
| qum_ | wilee-nilee: i'll try mate, i just want to be sure i got the best working drivers for me :D | 19:42 |
| wilee-nilee | qum_, I think you probably fine if it all looks good. | 19:42 |
| wilee-nilee | you're* | 19:42 |
| ChogyDan | qum_: I don't see how you got a 3.7 kernel, unless you are outdated. I figured 3.8 was the kernel for awhile now | 19:42 |
| === Guest30154 is now known as wN | ||
| wilee-nilee | 3.7 is the latest kernel in my raring | 19:43 |
| qum_ | mmm, ChogyDan, maybe from some update or links i pressed? :( | 19:43 |
| bjsnider | he's new to linux and he's using an unstable distro, and xorg-edgers, which is terribly experimental | 19:43 |
| bjsnider | 2 strikes | 19:43 |
| wilee-nilee | as of yeterday anyway | 19:43 |
| qum_ | So guys, can u instruct me what to do pls? | 19:43 |
| ChogyDan | wilee-nilee: is the kernel labeled as 3.8 or something? | 19:44 |
| bjsnider | talk about diving into the deep end of the pool | 19:44 |
| qum_ | bjsnider: talk about helping me out m8) | 19:44 |
| wilee-nilee | ChogyDan, I'm in precise when I looked in raring boot it says 3.7. | 19:45 |
| bjsnider | latest kernel appears to be 3.8.0.19.35 | 19:45 |
| ChogyDan | what bjsnider said for me as well | 19:45 |
| qum_ | okey, so i should google for how to upgrade the kernel to 3.8.0.19.35? | 19:45 |
| ChogyDan | no qum_, you need to learn basic upgrade stuff. That should do it for you. You probably downloaded the wrong disk or something | 19:46 |
| bjsnider | http://packages.ubuntu.com/raring/linux-image | 19:47 |
| qum_ | ChogyDan: can u be so kind and point out in link to what should i download? I don't mind formating and installing ubuntu again, it's better to do it now :D | 19:47 |
| bjsnider | since we're at rc, it might be ok to stay with raring, but it won't be as well documented as quantal | 19:48 |
| bjsnider | but i would ppa-purge xorg-edgers | 19:49 |
| bjsnider | that is for more experienced users who are experimenting with bleeding-edge xorg stuff | 19:49 |
| ChogyDan | qum_: it looks like this page is it: http://releases.ubuntu.com/13.04/ | 19:49 |
| qum_ | ChogyDan: so that i would be safe, i should download it, burn to disc and install from fresh? | 19:50 |
| qum_ | Can't i just upgrade the kernel somehow without format? I mean without leaving chance to some bugs later? | 19:51 |
| ChogyDan | qum_: here is a command: sudo apt-get update && sudo apt-get install linux-generic ubuntu-desktop && sudo apt-get dist-upgrade | 19:52 |
| ChogyDan | qum_: i have to go a bit. My unity is crashing | 19:53 |
| wilee-nilee | thats the command | 19:53 |
| qum_ | ty | 19:53 |
| qum_ | cya soon :D | 19:53 |
| qum_ | ChogyDan: i did it and it didn't upgrade anything... | 19:54 |
| ChogyDan | qum_: I don't know, but you may be better off just reinstalling. the upgrade may take up just as much bandwidth, and take longer to install | 19:56 |
| qum_ | ChogyDan: i got a pretty good bandwith. | 19:56 |
| qum_ | and if i can upgrade from inside without any bugs later i prefer it. | 19:56 |
| ChogyDan | why? | 19:57 |
| qum_ | well, that way i won't format the system, less time :D | 19:58 |
| ChogyDan | qum_: a fresh install takes less time than a large upgrade. And checking out why your system isn't upgrading is that exciting.. | 20:01 |
| qum_ | lol | 20:01 |
| qum_ | ChogyDan: kk, so i downloaded http://releases.ubuntu.com/13.04/ubuntu-13.04-beta2-desktop-i386.iso i'll burn it to a disc and then just install? | 20:02 |
| qum_ | wilee-nilee: is that what he ment? | 20:06 |
| wilee-nilee | qum_, I guess not sure why a new install is needed though honestly, you have all the repos a update and dist-upgrade should suffice. | 20:08 |
| qum_ | wilee-nilee: so what should i do?:( | 20:20 |
| wilee-nilee | qum_, You are running the ubuntu unity desktop correct? | 20:20 |
| wilee-nilee | panel on the left side of the screen? | 20:21 |
| qum_ | if u mean if i got the unity launcher on the left, then yes :D | 20:21 |
| wilee-nilee | qum_, This command should get you the kernel upgrade and anything else, I assume the ppa nvidia drivers will be okay, that is a bit out my area. sudo apt-get update && sudo apt-get install linux-generic ubuntu-desktop && sudo apt-get dist-upgrade | 20:23 |
| wilee-nilee | graphics I'm not real up on I have never needed driver beyond a install. | 20:23 |
| wilee-nilee | qum_, Removing the ppa if things go south and the drivers and using the regular repos drivers I would get acquainted with is all, just to be informed. ;) | 20:26 |
| qum_ | sec mate | 20:27 |
| qum_ | wilee-nilee: when i type sudo apt-get install linux-generic ubuntu-desktop it says ubuntu-desktop is already the newest version. | 20:34 |
| qum_ | linux-generic is already the newest version | 20:34 |
| qum_ | and no action is taken | 20:34 |
| wilee-nilee | qum_, what do you get from uname -r in the terninal? | 20:37 |
| qum_ | 3.7.0-7-generic | 20:38 |
| wilee-nilee | qum_, Did you run this sudo apt-get dist-upgrade | 20:38 |
| qum_ | yes | 20:38 |
| qum_ | and i wrote u what i got | 20:38 |
| qum_ | ubuntu-desktop is already the newest version. | 20:38 |
| qum_ | linux-generic is already the newest version | 20:39 |
| qum_ | 0 upgraded, 0 newly installed, 0 to remove and 0 not upgraded. | 20:39 |
| qum_ | that's what i get when i type it | 20:39 |
| wilee-nilee | qum_, Not sure than myself I ran the whole command and am at 3.8.0-19-generic I would assume it is in the regular repos, honestly I don't really worry about kernels mine have been up to date except for raring today, and now is, so I'm not the best help on this. ;) | 20:41 |
| qum_ | heh | 20:42 |
| wilee-nilee | I would hate to give you wrong info. ;) | 20:43 |
| genii-around | Did you type in sudo apt-get update ...first? | 20:44 |
| qum_ | ye si did | 20:44 |
| genii-around | maybe try linux-image-generic instead of linux-generic | 20:46 |
| qum_ | sudo apt-get install linux-image-generic ? | 20:46 |
| genii-around | Yes | 20:47 |
| qum_ | i get the same | 20:48 |
| genii-around | Perhaps the Israeli local mirror is not very up to date | 20:49 |
| qum_ | perhaps :( | 20:51 |
| wilee-nilee | genii-around, I wonderd that myself. ;) | 20:51 |
| qum_ | well... :D | 20:52 |
| qum_ | wb ChogyDan | 20:52 |
| ChogyDan | meh, nothing is fixed. | 20:53 |
| gh__ | hi, anyone having problems doing "apt-get update" with ubuntu 13.04? I have a hash mismatch error on the universe package list | 20:56 |
| gh__ | the hash mismatch happens with many different mirrors | 20:56 |
| tgm4883 | gh__, did you try blowing away the hash and retrying? | 20:59 |
| IdleOne | working fine here with ca.archive | 21:00 |
| gh__ | tgm4883: removing /var/lib/apt/lists/ ? yes | 21:01 |
| MonkeyDust | gh__ type sudo apt-get install pastebinit;sudo apt-get update | pastebinit and paste the url here | 21:01 |
| tgm4883 | gh__, I'd just remove the one that it's complaining about, but yes | 21:02 |
| ChogyDan | fwiw, I think this command fixed my unity issue: dconf reset -f /org/compiz/ && setsid unity | 21:03 |
| ChogyDan | so it must have been a compiz misconfiguration oO | 21:03 |
| gh__ | MonkeyDust: http://paste.ubuntu.com/5722699/ the error message is in french.. | 21:04 |
| MonkeyDust | gh__ np Ispeak french | 21:06 |
| gh__ | it keeps happening even if I disable -updates and -backports | 21:06 |
| gh__ | I'm in a university, could it be a network problem? | 21:07 |
| tgm4883 | gh__, did you try 'rm /var/lib/apt/lists/partial/fr.archive.ubuntu.com_ubuntu_dists_raring_universe_binary-i386_Packages' | 21:08 |
| gh__ | tgm4883: I just nuked the whole folder and tried again, the same keeps happening | 21:09 |
| tgm4883 | gh__, well thats a subfolder of the one you specified earlier | 21:09 |
| tgm4883 | /var/lib/apt/lists/partial vs /var/lib/apt/lists | 21:10 |
| gh__ | I nuked /var/lib/apt/lists/ :) | 21:10 |
| MonkeyDust | gh__ I have that same file, guess everybody does, not sure what it is | 21:11 |
| MonkeyDust | gh__ guess you can simply ignore that error | 21:12 |
| gh__ | well, the whole universe repository is unavailable | 21:12 |
| gh__ | which mean I cannot install a lot of stuff | 21:12 |
| MonkeyDust | what's an example of a program in the universe repo? | 21:19 |
| tgm4883 | most things are in universe and/or multiverse I think | 21:20 |
| MonkeyDust | gh__ can you install audacious? it's a mediaplayer in the universe | 21:21 |
| gh__ | it's not in universe, since it comes by default with lubuntu | 21:22 |
| gh__ | but for instance I can not install ghc, or tree | 21:22 |
| MonkeyDust | yes, they are in universe too | 21:23 |
| MonkeyDust | audacious is too, type apt-cache policy [package] to find out | 21:23 |
| MonkeyDust | gh__ type sudo apt-get install tree|pastebinit and paste the url here | 21:24 |
| gh__ | it proposes me packages with "tree" in the name. | 21:25 |
| MonkeyDust | gh__ I read write and speak 4 languages, among which french, so you can paste the output | 21:26 |
| gh__ | it happens on 3 machines in the same room | 21:26 |
| MonkeyDust | gh__ are there more ubuntu pc's outside that room? | 21:27 |
| gh__ | yes but they are not using raring | 21:28 |
| gh__ | and they don't have this problem | 21:28 |
| MonkeyDust | hm | 21:28 |
| gh__ | I'm reading this in a forum: "Possibly a proxy? I had to change the proxy that I was going through at the company I work at. I assume that somehow the file I tried to download was corrupt and every time I requested it, the proxy said "I've got that!" and sent me a incomplete/corrupt file. " | 21:29 |
| gh__ | I believe my university does have a proxy | 21:29 |
| gh__ | maybe I just have to wait another day | 21:29 |
| ChogyDan | gh__: change repo maybe? | 21:29 |
| gh__ | ChogyDan: already done, same effect | 21:30 |
| gh__ | dowloading the same .bz2 file by ssh from a remove location (in another country), I have the same md5 hash. so the files are the same. | 21:35 |
| MonkeyDust | gh__ yes, it must be local, as I can install tree from here (to stick to your example) | 21:36 |
| gh__ | ok, it's getting late :) | 21:39 |
| gh__ | thanks a lot for your help | 21:39 |
| gh__ | MonkeyDust | 21:39 |
| richerVE | Greetings, Ubuntu 13.04 RC isn't supposed to be released yesterday? | 22:25 |
| valorie | richerVE: yes | 22:37 |
| valorie | we don't know what held up the publishing of the images | 22:38 |
| valorie | although I assume it was necessary, so all we can do is wait | 22:38 |
| bekks | https://wiki.ubuntu.com/RaringRingtail/ReleaseSchedule | 22:39 |
| bekks | There was no holding up - yesterday was just the final freeze. | 22:39 |
| valorie | bekks: RC was scheduled yesterday | 22:44 |
| === Ian__ is now known as Ian_Corne | ||
| bekks | valorie: And the release date is still 25th :P | 23:40 |
| bekks | So I dont see any hold up there. | 23:40 |
| valorie | right | 23:40 |
| valorie | in https://wiki.ubuntu.com/ReleaseCandidate they sort of give themselves an out, but I know that in the past, there has been an actual, named RC "release" | 23:41 |
| valorie | not this time around, I guess | 23:42 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!
