/srv/irclogs.ubuntu.com/2011/02/08/#ubuntu-mythtv.txt

DHRI've built a new backend box and imported the from the old one, as per http://www.mythtv.org/wiki/Backend_migration.  It says to restore before running any myth programs.  Well, something in http://www.mythbuntu.org/existing-ubuntu blew away the database contents.  How should I proceed?  I tried stoping the backend and restoring again but that doesn't work.00:03
Zinn[www.mythtv.org] Backend migration - MythTV00:03
DHRa related question: what causes the required schema fix (I'm restoring from a .20 box)?00:05
Newbuntu81Does anyone have any tips on errors preventing me from compiling the analog 2250 driver?  See details at http://ubuntuforums.org/showthread.php?p=10437860#post1043786000:37
Newbuntu81Specifically, it looks like 2 things are causing me issues.00:50
Newbuntu811. /home/michael/saa7164-v4l/v4l/au0828-video.c:185: error: implicit declaration of function 'usb_buffer_free'00:51
Newbuntu812. /home/michael/saa7164-v4l/v4l/au0828-video.c:255: error: implicit declaration of function 'usb_buffer_alloc'00:51
Newbuntu81('usb_buffer_free' AND 'usb_buffer_alloc')00:51
fluvvellnot sure if any kiwis are online, but can someone confirm whether ppa:s are needed for vdpau in NZ?01:06
Newbuntu81What Mythbuntu kernel is everyone running? I have 2.6.35 and was just told i should perhaps downgrade. Just curious what the rest of you are running?01:08
DHRNewBuntu81: that message suggests that a header is missing (which should produce an earlier error message).  Not that I know the code, only the C language.01:08
Newbuntu81Ah thanks DHR. Good info.  I saw something about adding the lines back in but I have no clue what files to edit.01:09
PatrickdkNewbuntu81, have fun compiling that on .35 :)01:09
Patrickdkall kinds of new security stuff went into the .33 kernel01:09
Patrickdkso stuff made for <.33 kernel or maybe .34 kernel can be a bitch to make work01:09
ZinnPatrickdk: Please watch your language.01:09
Patrickdkoh heh01:10
Newbuntu81Hmmm, so everyone here is on .35?01:11
Newbuntu81I'm trying to fix my issues in compiling...the errors are due to: ('usb_buffer_free' AND 'usb_buffer_alloc')01:11
* Patrickdk wonders what he said to have Zinn comment01:11
Newbuntu81In another room they suggested i downgrade the kernel.01:11
PatrickdkI'm not01:11
fluvvellPatrickdk, perhaps the b**ch word ?01:12
DHRzinn is a bot.  It appears to notice clasic swear words.01:12
ZinnHi DHR, something I can help you with today?  I am a bot, use !help to see what I can do.01:12
Patrickdkthat was after zinn commented01:12
ZinnHi Patrickdk, something I can help you with today?  I am a bot, use !help to see what I can do.01:12
Patrickdkhmm01:13
DHRzounds01:13
DHR'sblood01:13
Newbuntu81LOL01:13
fluvvellha, it seems z*nn has big ears01:13
DHRmerde01:13
Newbuntu81So anyone know how to fix 1) 'usb_buffer_free' AND 2) 'usb_buffer_alloc')?01:13
Newbuntu81I saw information saying I needed to add 2 lines to various files. I just don't know which files to edit.01:14
Newbuntu81Here's my errors /log. http://ubuntuforums.org/showthread.php?p=10437860#post1043786001:14
Zinn[ubuntuforums.org] Hauppauge HVR-2250 analog support issues with MythTV - Page 8 - Ubuntu Forums01:14
DHRI'd say that was a symptom.  Your headers (from the kernel, I guess) don't match what the code expected.01:14
Newbuntu81Yep, Zinn, that's my post.01:14
ZinnHi Newbuntu81, something I can help you with today?  I am a bot, use !help to see what I can do.01:14
Patrickdkzinn, I want to do something unsavory to you later.01:15
ZinnHi Patrickdk, something I can help you with today?  I am a bot, use !help to see what I can do.01:15
DHRI assume (hey, I don't know anything) that the drive is "out of tree" so it has trouble keeping up with kernel advances.01:15
Patrickdkzinn is kinky01:15
ZinnHi Patrickdk, something I can help you with today?  I am a bot, use !help to see what I can do.01:15
Newbuntu81LOL01:15
Newbuntu81The funny part is...this is how I feel at work.01:15
Patrickdkout of tree?01:15
Newbuntu81You ask and ask and ask and get non-sensical answers. LOL.01:16
Patrickdkdhr, all you have to do is make sure the backend isn't running01:16
Patrickdkrestore the db01:16
Patrickdkthen start the backend01:16
Patrickdkafter that, start a frontend, the frontend should upgrade the db01:16
DHRpatrickdk: thanks.01:16
PatrickdkI can't remember if I upgraded my old .15 or if I started over01:17
Patrickdkbut been using myth since like .1101:17
Newbuntu81Patrickdk: Are these tips for me?01:17
DHR"out of tree" means, in the Linux world, code that isn't in the official Linux tree.  When people change the kernel, they make the required changes to the in-tree drivers but they don't do it for out-of tree drivers.01:17
DHRthe Right Thing is to work really hard at getting your driver accepted in the tree.  The usual barrier is quality of your code.  But many folks are too shy (or something) to submit their drivers.01:18
Newbuntu81Oh nevermind. Patrickdk is helping DHR, and both he and DHR are helping me. Wow this gets confusing.01:19
Patrickdkheh, guess so, I just call out mainline or screwed01:19
PatrickdkI have only submitted one thing to the kernel01:20
Patrickdkand they wanted it badly01:20
Patrickdkall I did was email the kernel list saying I patched it, and alex cox took it and applied it01:20
Patrickdktotal of like 3 days01:20
DHRa while ago?  Does Alan still act as a gatekeeper for anything?01:22
PatrickdkI don't think so :)01:22
Patrickdkthis was back when pci/localbus didn't exist, 16bit isa baby :)01:22
Newbuntu81and now we have PCIe01:23
Newbuntu81Are people still using AGP for video cards, or has everyone switched over to PCI Express now?01:23
* Patrickdk wonders if anyone remembers localbus01:23
Patrickdkor microchannel01:23
Patrickdkhell, I have lots of 8bit isa cards still right next to me01:24
ZinnPatrickdk: Please watch your language.01:24
Patrickdkoh, it's hell01:24
ZinnPatrickdk: Please watch your language.01:24
PatrickdkIhaving a hell of a good time :)01:24
ZinnPatrickdk: Please watch your language.01:24
Newbuntu81Has anyone else ran into these errors?01:41
Newbuntu81My issues seem to be with:01:41
Newbuntu811. /home/michael/saa7164-v4l/v4l/au0828-video.c:185: error: implicit declaration of function 'usb_buffer_free'01:41
Newbuntu812. /home/michael/saa7164-v4l/v4l/au0828-video.c:255: error: implicit declaration of function 'usb_buffer_alloc'01:41
Patrickdkyou really can't use google at all can you01:42
Patrickdkhttps://issues.asterisk.org/print_bug_page.php?bug_id=1738301:42
Zinn[issues.asterisk.org] 0017383: [patch] usb_buffer_free/alloc renamed in 2.6.34; compat funcs dropped in 2.6.35-rc0 - Asterisk.org Issue Tracker01:42
Newbuntu81LOL.  I've been searching for days.  That's why I thought hey--these people probably had to do it.  Maybe they know.01:52
patdk-lapgoogle: usb_buffer_free 2.6.35, click url01:52
Newbuntu81Actually the patch failed.01:55
Newbuntu81But maybe from the results I"ll know which 2 files to gedit01:55
Newbuntu81Is there a preferred way to paste in errors in here?01:55
patdk-lappatch?01:56
patdk-lapyou wheren't suppost to apply a PATCH from the url I posted01:56
patdk-lapyou where suppost to READ, and hack it yourself01:56
patdk-lapor are you saying you don't know c at all?01:56
Newbuntu81I don't know C very well...01:57
Newbuntu81But the page had a patch.01:57
Newbuntu81[wget patch] [License OK (v3.0)]01:57
Newbuntu81wget 'https://issues.asterisk.org/file_download.php?file_id=26178&type=bug' -O - | patch -p001:57
patdk-lapthe patch was for asterisk, not video01:57
* patdk-lap notes telephone system != video01:57
Newbuntu81Sorry, I should have known it wouldn't be THAT easy01:59
patdk-lapbut if you read it, it is simpler02:00
patdk-lapjust a search/replace02:00
DHRI don't remember when PCI came in.  Wikipedia says "did not gain significant market penetration until late 1994 in second-generation Pentium PCs" but I think that my P60 systems had PCI.02:19
patdk-lapyep02:19
DHRI recycled them last year so I cannot check.02:19
patdk-lap486's is when pci and localbus came in02:19
patdk-lapbut localbus was faster than pci, and cheaper02:19
patdk-lapbut you could only have like 3 localbus slots per computer, where pci wasn't so limited02:20
patdk-laplocalbus was also a very very long slot (if I remember right, it was a normal 16bit isa, plus the localbus on the end of it)02:20
DHRThe new Myth box I'm setting up has 5 PCI.  For analogue tuner cards.02:20
patdk-lap5 tuners?02:20
patdk-lapI have 3 in my system, hardly any of them are ever busy02:21
DHRFour at the moment.  Need a better splitter.02:21
patdk-lapnothing on tv worth watching02:21
patdk-lapmainly just use mythvideo02:21
DHReverything worth watching can happen at the same time.  And it is recording for more than one person.02:22
patdk-lapmine does 3 people02:22
patdk-lapI opted for 4 pciex16 slots though02:23
patdk-lapneeded the slots for video cards02:23
DHRMy old myth box had 5 pci plus an AGP.  That makes it obsolescent.  I took over one PCI for a SATA controller.02:23
patdk-lapmy new one has 8 sata, only 6 usable02:23
patdk-laphave sata mulpliers in 3 of them02:24
DHRThe tuners I've collected are all PCI.02:24
patdk-lapso 18 usable sata ports :)02:24
patdk-lapexpandable to 26 (or 30 with usb boot)02:24
DHRI'm only using 5 of the 6 in the box I've just built.02:24
DHRmultipliers don't work with all controllers, right?  So many under-advertised options in SATA.  Like: only some controllers support hot swap.02:25
patdk-lapyep02:25
patdk-lapwhat southbridge you have?02:25
DHRP43 chipset.  http://www.gigabyte.com/products/product-page.aspx?pid=3637#ov02:26
patdk-lapthat is a northbridge02:27
patdk-lapnot south02:27
DHRICH1002:27
patdk-lapthat is good then02:27
patdk-lapyou have that jmb368 chip also02:28
patdk-lapit's useless02:28
DHRinexpensive board with 5 PCI.  Trust Gigabyte more than ECS02:28
DHRgood to know.  I've not connected any IDE.02:29
patdk-lapoh they are only using it for ide02:29
patdk-lapprobably ok02:29
patdk-lapmine does ide and two sata02:29
patdk-lapbut only two sata in raid mode (and when not in raid mode, only one sata)02:29
patdk-lapand it's slow02:29
patdk-lapso I just don't use them at all02:29
patdk-lapmine is a beast of a system, atleast for myth02:30
DHRWhy would you use RAID mode?  My understanding is that software RAID is better.02:30
patdk-lapit is02:30
patdk-lapI wouldn't use it, so I don't use it02:30
patdk-lapand the port mulpliers saved me02:31
patdk-lapthey max out at around 240MB/s per sata port02:31
patdk-lapso only really fullspeed of 2-3 drives per port, instead of 502:32
patdk-lapbut acceptable02:32
DHRMy impression (may not be right) is that the southbridges have unadvertised bottlenecks.02:32
patdk-lapthe ich10 fixed many of them02:33
patdk-lapI know ich7/8 and I think 9, where all connected a single pci bus02:33
patdk-lapso all the sata where limited to 100MB/s02:33
patdk-lapbeen awhile since I looked at the ich10 datasheet02:33
DHRwhy do you have such an overspeced machine for Myth.  Does it do something else?02:34
patdk-lapit's a amd 1090T with 16gigs ram02:34
patdk-lapwith 4 gt240 video cards02:35
* mrand remembers localbus and microchannel.02:35
mrand"It's like a multi-lane highway"02:35
patdk-lapit's my primary workstation02:35
DHRI remember s100.  I've still got an Altair.02:35
patdk-lapI use the hdmi ports on each video card to feed a tv02:35
mrandfour tv's?02:36
patdk-lapyep02:36
mrandin one room?02:36
patdk-lapand 4 monitors (for my workstation)02:36
patdk-lap4 rooms02:36
mrandinteresting setup.02:37
patdk-lappower savings are much better having a beast, then several frontends02:37
DHRI like a dedicated myth box.  That way it can sit in the corner, undisturbed, like an appliance.  My workstation gets rebooted several times a month.02:37
patdk-lapI only reboot about once a month, for updates02:37
patdk-lapI wanted the power to attempt to decode 4 hd streams at once02:38
patdk-lapbut I'm happy if it can do 202:38
DHRmy old backend hasn't had updates in several years.  Still on Fedora Core 5.  For one thing, TV-Out support for that card disappeared for several years.02:38
patdk-lapheh, I hated the tv out on the 35002:38
patdk-lapused it for about a week02:39
patdk-lapand dropped it02:39
mrandpatdk-lap: I'm not sure your are saving much power compared to a diskless Zotac, but you've got better control/options/power.02:39
mrand(horsepower, I mean)02:39
patdk-lapmrand, not sure, I would have to check02:39
patdk-lapbut it's not using much more power than my old system02:40
patdk-lapa p4 D02:40
DHRP4 D slurps power02:40
mrandHeh.  I have my daughter on a P4.  It's loud and hot.02:40
patdk-lapI have stacks and stacks of dells here that I fix02:41
patdk-lapI keep replacing the bad caps on broken motherboards02:41
patdk-lapand have a perfectly good computer02:41
DHRduring the P4 years I only bought AMD.02:41
patdk-lapwife wants me to start tossing them out now02:41
DHRReplacing bad caps if pretty labour intensive, isn't it?  you need to replace dozens per board, don't you?02:42
patdk-laphave 5 p4 xeon dual systems :)02:42
patdk-lapdhr, most of them was around 1602:42
patdk-laptook me about an hour per board02:42
mrandyou a EE?  Saw you talking about reading the ich datasheet.02:43
patdk-lapI started with electricity when I was a year old02:44
patdk-laplearned it, did a lot with it02:44
patdk-lapbut around 14 or so, I learned, well, computers can manage electricity so much better :)02:44
patdk-lapbut I still have racks and racks of amd and national semiconductors, and some ti books02:45
patdk-lapand I normally always read the datasheets for chips, expectially for server boards I'm looking at, to see how the slots are connected (shared bus or not) and what bandwidth is available form them to cpu/memory02:45
mrandYeah.  Intel is locking stuff down pretty tight now-a-days.  I have to request access just to ask them a question!02:46
mrandstuff=documentaiton02:46
patdk-lapya, used to be I could easily get a datasheet and even power budget worksheet from their website02:46
patdk-lapbut this last time, it took me forever to find the datasheet, and no powerbudgets02:47
DHRsometimes its what's not in the datasheet.  I couldn't figure out why my Acer Revo 3610, which came with 4G of RAM, couldn't see it all.  Turns out that intel didn't bother to drive all the address pins but see if you can find that in the datasheet.02:47
patdk-lapit's in there02:48
patdk-laphow much could it see?02:48
patdk-lapand where you in 32 or 64bit mode?02:48
DHR4G of address space, some for the PCI02:48
DHR64-bit.  But PAE means that it shouldn't matter.02:48
patdk-lappae?02:48
patdk-lappae doesn't exist in 6402:48
patdk-lapand not all cpu's support pae02:49
patdk-lapmy 64bit cpu only supported 3gig ram, and 32bit mode didn't support pae02:49
patdk-lapram limited was motherboard issue02:49
patdk-lapbut yes, all cpu's have only limited amount of address pins02:49
DHRall Intel CPUs since Pentium Pro support PAE.  PAE only needed for 32, but that's why I said it didn't matter whether I used 64 or 32.02:49
patdk-lapmy cpu clearly said no pae support02:50
DHRWhich CPU?02:50
Newbuntu81Why is software raid better than hardware? I thought hardware anything was better. (i.e. hardware encoding)02:52
DHRreal hardware RAID can have advantages.  But that is expensive.  The cheap stuff has no advantages, at least in the Linux world.02:53
patdk-lapmaybe motherboard issue again02:53
patdk-laphttp://ark.intel.com/Product.aspx?id=2751902:53
Zinn[ark.intel.com] Intel® Pentium® D Processor 940 (4M Cache, 3.20 GHz, 800 MHz FSB)with SPEC Code(s)SL8WQ, SL94Q, SL95W02:53
patdk-lapbut booting any pae enabled kernel, would crash on boot02:53
DHRIt will surely have PAE on-chip.02:53
patdk-lap64bit and 32bit where ok02:53
DHRthat's odd.02:54
patdk-lapya02:54
* patdk-lap is scared of the memory layout of current systems02:55
DHRchipsets and BIOSes and solder masks can all impose stupid limitations.02:55
patdk-lapexpecially 64bit02:55
mranduh, how would solder masks impose a stupid limitation?02:55
DHRI meant traces: not enough lines.02:56
patdk-lapyou have ems block, or umb area02:56
patdk-lapyou have reserved spots for addin cards02:56
patdk-lapthen random pci/agp area02:56
patdk-lapthen random pcie areas in 64bit land02:56
DHRmost PCI reservations seem to be 32-bits so insignificant, if annoying.02:57
patdk-lapall this latency memory readdressing :(02:57
DHRevery montherboard is NUMA02:58
DHRyou didn't even mention the new fangle PCI virtualization hardware, surely adding another level of latency02:59
patdk-lapheh?02:59
patdk-lapI'm already missing one or two thing that screw up the memory map, that I can't remember03:00
DHRsince it is on-chip, it may not matter (gate delays are so short these days)03:00
patdk-lapI used to write protected mode startup code back on the 386/486, would hate to do it these days with how screwed up the memory ranges are03:00
DHRIO MMU AMD-Vi, Intel VT-d03:03
DHRhttp://en.wikipedia.org/wiki/IOMMU03:03
Zinn[en.wikipedia.org] IOMMU - Wikipedia, the free encyclopedia03:03
DHRThe thing that gets me is SMM.  Completely behind the back of the OS.  Makes for uncontrollable jitter, for example.03:04
DHRCan also create holes in OSes: apparently some BIOS SMM code makes assumptions about the MTRRs that users can falsify.03:04
* DHR ducks out for a bit03:06
qwebirc52383Evening all03:52
qwebirc52383Can anyone point me to the MD5 Checksum for mythbuntu-10.10-desktop-amd6403:53
boneliferhttp://mythbuntu.org/downloads03:54
Zinn[mythbuntu.org] Download | Mythbuntu03:54
boneliferTrouble Shooting link03:54
qwebirc52383awesome ty03:54
boneliferMD5SUMS03:54
qwebirc52383didnt see it in the dropdown, thanks03:54
boneliferAfter a fresh install of 10.10 I experience a standard black login screen. I login and can't "startx", I do a "sudo shutdown ..." to restart and I'm in the frontend. It does that everytime I update/install anything.03:56
boneliferthe splashscreen also doesn't have the usual mythbuntu, but install says "Ubuntu 10.10" above the dots.03:56
boneliferwould I have better luck just upgrading?03:57
boneliferthe MD5 sum is right03:57
boneliferfor the iso03:58
superm1the text splash screen isn't overridden from ubuntu one03:59
superm1due to technical reasons03:59
superm1but the graphical one is03:59
boneliferby graphical I assume you mean the purple background.04:00
boneliferwhich is what I was talking about, not really important. the unstable reboot after upgrades is though04:00
bonelifers/upgrades/install04:01
superm1i'm meaning the one that shows a mythbuntu picture04:01
superm1rather than a line of text with ubunt04:02
boneliferyes, that's the one that didn't show, instead it shows Ubuntu 10.1004:02
superm1so during install did you pick the nvidia driver?04:02
boneliferyes04:02
superm1perhaps it's not working properly with your card04:02
superm1have you checked the X log after a boot?04:03
boneliferno, that was the only graphical problem I have, so I didn't bother. Again, the having to reboot after rebooting after an upgrade/install of new program is what's bothering me04:04
boneliferwas just wondering if anybody's seen that problem. Right now I'm justing going to go with a 10.04 install and revisit this later if no one can point me in the right direction.04:05
superm1i dont think i've ever seen that problem myself04:06
bonelifernot really that important if it isn't a off the top of somebody's head sort of thing.04:06
boneliferprobably a nvidia driver thing. as startx tried to start when I issued the command, but then just stopped at something I can't remember and I ctrl-c'd out of it and restarted.04:07
superm1you shouldn't generally be using startx to login to the GUI04:07
superm1you should be using gdm04:07
superm1if you are having to use startx for some reason, you should figure out what's wrong with gdm (and fix that)04:08
boneliferneither worked though04:08
superm1so if gdm doesn't work immediately after boot, check /var/log/Xorg.0.log04:09
superm1and /var/log/syslog04:09
superm1one of them should say why04:09
superm1assuming you didn't disable the upstart job or something04:09
boneliferjust a standard fresh install04:09
superm1right, but that doesn't preclude potential problems with compatibility with the nvidia driver04:10
superm1it can certainly cause problems like that if it's not liking your hardware04:10
superm1again, it would be documented in the logs what's happening though04:10
boneliferASUS GT-220.  I'll keep that a text file for when I try again mid next week.04:11
superm1GT220 should work fine out of the box04:11
superm1(with that nvidia driver that is)04:11
boneliferthat's why I took the leap to get the audio via hdmi for when the system is finished. doing the alsa upgrade via the upgrade script available, left me seeing the ASUS audio, but not able to see or use the builtin. Which the 10.10 install fixed. I'll give it a few days and a fresh set of eyes/mind and your log suggestion to see what's up.04:13
boneliferthe only other strange thing that happened to me was that when scanning for channels it recieved a "100%" signal, and I ended up with channels that were just static, definately don't want to go through and manually disable 60+ channels04:18
boneliferthanks for the suggestions04:19
boneliferthe whole scan thing is weird since 10.04 works right.04:21
baggar11does mythtv still do straight passthrough for all audio types, without upconverting 2 channel to 5.1? When I uncheck the upconvert option, I lose 2 channel audio. The 5.1 stuff still plays fine.15:59
rhpot1991baggar11: it should, you want your speakers set to 5.1, no upconverting and then enable the digital passthrough options16:01
rhpot1991you could also try toggling the stereo pcm only checkbox16:01
mrand!stab linux audio16:22
* Zinn stabs linux audio with a rusty spork.16:22
baggar11rhpot1991: hmm, don't recall seeing the stereo pcm only checkbox, I'll check again, thanks.17:57
rhpot1991baggar11: its in the advenaced section18:01
tgm4883!logs18:38
ZinnMythTV logs are stored in /var/log/mythtv/   You can use mythbuntu-log-grabber from the Applications menu to automatically post the most relevant logs to our pastebin.18:38
Newbuntu81I've made some progress since yesterday, I believe resolving the usb_buffer_alloc and usb_buffer_free errors.  Has anyone with an HVR 2250 run into issues with dvb_net.c when compiling?  All helpful eyes are appreciated.  Details at http://ubuntuforums.org/showthread.php?p=10440736#post10440736, post #74 (pg 8).  Thanks!22:43
mrandNewbuntu81: I'd suggest contacting the last few pepole that checked that file into whatever repository you'll pulling from and ask them.22:57
Newbuntu81does anyone know how to pull a few specific files from a .tar.bz2 file and copy them into a specific folder?23:42
Newbuntu81it would be a sudo cp .....23:43
tgm4883untar, then copy?23:43
tgm4883or man tar23:43
Newbuntu81in the folder named "/usr/src/linux-source-2.6.35", is the zipped file named "linux-source-2.6.35.tar.bz2".  I need to pull out all *.h and *.c files from this directory inside the .tar.gz2 folder named "/linux-source-2.6.35/drivers/media/dvb/dvb-core/"23:45
Newbuntu81so you can't do it all in one step? :-)23:46
mrandNewbuntu81:  I don't remember that you can use a wildcard when un-tar'ing.  You may have to extract everything then grab the files23:47
Newbuntu81i'm going to ask in myth too. thanks mrand23:51
Newbuntu81man tar i guess is the command23:54

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!