[09:43] <ailo> Len-nb: I'm trying to change IRQ for my pci devices, but nothing happens
[09:43] <ailo> I find that there's a lot to learn about IRQ on my account, as I'm reading more about it
[09:43] <ailo> What are they exactly, and how does the kernel handle them
[09:43] <ailo> So, I'm unable to reproduce your config right now
[09:44] <ailo> I mean, I'm unable to reproduce your problem with xruns
[09:45] <ailo> Len-nb: When you changed the order between your audio device and your eth card, you did that in the bios, right?
[09:45] <ailo> You'll be sleeping now, but just let me know when you read this :)
[10:42] <ailo> Anybody notice how flash just changed?
[10:44] <ailo> Everythings handled by the totem plugin, only not everything works anymore
[14:15] <ailo> Len-nb: Well, on this machine, the only way to achieve the same type of situation with the irq's was having the pci cards placed in a certain order to have my wifi be ahead of my audio card. But, with this machine, that seems to have improved performance for me, cause now my audio device is not sharing irq with anything else
[14:17] <ailo> I will have to test on other machines. Also, do we know for sure that the order of the IRQ's matter?
[14:18] <ailo> I know there are two kinds of interrupts. Short and long ones. The short ones will interrupt everything else and spit out it's full message, while long ones will push out their message in parets
[14:18] <ailo> parts*
[14:19] <ailo> I'm not getting any xruns either way
[14:19] <ailo> I was getting xruns because of pulseaudio though
[14:19] <ailo> So, I'm not using the jack bridge
[14:59] <Len-nb> ailo, irq setup is somewhat hardware dependant. there has been a standard chip that has been used since the original ibm pc.
[15:01] <Len-nb> the pc had 8 irqs and then shortly after they found that was not enough and so used two of the same chip.
[15:04] <Len-nb> then someone made another chip that can handle up to 256 irqs, but it handles things the same way and they only use 24 of them.
[15:06] <Len-nb> In any case, if there are more than one interrupt the highest number interrupt blocks lower numbers untill the cpu resets it.
[15:06] <Len-nb> This is a physical thing.
[15:13] <Len-nb> Huh, he's gone
[16:02] <Len-nb> ailo, did you see anything i said?
[16:24] <ailo> Len-nb: Ok, but did you change your irq's using the bios, or you just switched places with your pci cards?
[16:25] <ailo> For me changing irq's in the bios has had no effect
[16:25] <ailo> I will need to see with other machines, if it's the same thing
[16:26] <Len-nb> ailo, on my MB I had to change the card position PCIe is different again
[16:26] <Len-nb> I think it is all done by central chip
[16:27] <Len-nb> I can "reserve" an irq with bios, but not set it for a card.
[16:28] <ailo> I have an old ASUS board on which I remember being able to change IRQ for a bunch of devices. Will try today to see what that does
[16:28] <Len-nb> Mine is a one bit or something like that.
[16:30] <Len-nb> The other odd thing is that the bios shows irqs at boot, but they are different from what linux says they are.
[16:31] <ailo> I have a dim recollection of Linux preparing its own irq
[16:32] <ailo> In which case, the bios has no effect
[16:33] <ailo> Len-nb: Do you have problems with flash-player?
[16:33] <ailo> I'm a bit annoyed by suddenly having totem handling flash. It doesn't do it very well
[16:45] <Len-nb> My wife seems to have had problems... more so with firefox than chromium
[16:45] <ailo> Firefox and chromium behaves a bit differently, but they are both using a totem plugin instead of flash
[16:46] <ailo> I wasn't able to change that in the browser itself
[16:46] <Len-nb> Hmm.
[16:46] <ailo> It claims to be using flash, but it's not
[16:46] <Len-nb> It is true that when I installed there was no flash DL.
[16:47] <ailo> youtube works fine, but any autoplay flash does not autoplay, and stuff that requires a more recent flash does not work. Just like it is using gnash
[16:47] <Len-nb> The thing to do is find a page that uses a part of flash that totom doesn't handle. Then it should download the real one.
[16:47] <ailo> I have flash installed
[16:47] <ailo> This is not a flash problem
[16:47] <ailo> It's a Ubuntu problem
[16:48] <Len-nb> Surprise!
[16:48] <ailo> This happened after I updated last
[16:48] <ailo> I double checked to see if Wheezy was suffering from the same thing. It's gotta be some kind of config somehwere that is new to Ubuntu
[16:48] <Len-nb> Ubunu
[16:49] <Len-nb> toops ubuntu is trying to be all free lic.
[16:50] <ailo> I don't think so. Debian is, by making it possible to install a kernel without non-free firmware
[16:50] <ailo> Ubuntu is not doing that to my knowledge
[16:51] <ailo> Flash is pain in the but
[16:51] <Len-nb> ubuntu tries to make it's ISOs all free lic., but offer the non-free in repo.
[16:52] <ailo> The kernel itself is non-free, so I still don't think so
[16:52] <Len-nb> Esp. for anything that might be illegal in some country
[16:52] <ailo> Also, I had flash working for months. This happened after the final release
[16:53] <ailo> Anyway, I've been trying to find anyone else complaining about it. That is why I'm asking
[16:53] <Len-nb> While I was watching #ubuntu-release they had a list of thing to release post 12.04
[16:54] <Len-nb> I did notice when we first went live dvd that a lot of flash stuff worked right out of the box, but the "you need to install flash" prompt still came up.
[16:55] <Len-nb> Maybe they have "fixed" that.
[16:55] <Len-nb> anyway, I have to go. by now.
[17:47] <ailo> Does someone have 4+ GB RAM? Would be good to see if the -lowlatency pae is a pae, or if the -generic pae is a pae
[18:45] <micahg> umm, both are PAE kernels?
[19:57] <ailo> micahg: According to Ralph on the mail list he's not able to locate all of his memory
[20:01] <micahg> ailo: what's the subect/
[20:01] <micahg> *subject
[20:09] <ailo> micahg: "no audio output using Precise due to Pulseaudio?"
[20:09] <ailo> He just mentions this in one of his replies
[20:11] <micahg> an upgrade from oneiric won't switch to the PAE kernel, the user has to do that