[01:53] In kernel 4.2.1 from kernel.ubuntu.com, 0001-base-packing.patch has "+drivers/scsi/scsi_dh.ko" in generic.inclusion-list. I think it's supposed to be "+drivers/scsi/device_handler/scsi_dh.ko", since that's where that file actaully is [01:53] And in 4.2 it has the subdirectory in the path [01:53] Because when I try and build it, I get errors about that file not being found [01:54] 4.2 works, 4.2.1 doesn't [01:58] So, why was that changed? The file is still where it was before, but the path is now wrong === Guest75773 is now known as adrian === adrian is now known as Guest32345 [07:30] apw: i suppose they did not yet contact you? [07:31] apw: i mean manufcaturer/dev of camera SDK with the short URB broblem [07:41] Frogging101-chan (N,BFTL), we build all of those without extras to avoid that kind of things, you should too ... add do_extras_package=false [07:41] silidan, no they did not [10:17] apw: did they contact you? === psivaa is now known as psivaa-afk [10:32] apw: i found an open source capture application which also implements an open source low level access to the camera via libusb, the essential call to libusb_bluk_transfer is definitly made with the right size (1228800 bytes = 1280x960x1byte), the URB in wireshark still report a length of 61504, with the URB containing: "URB Length = 1228800" and "Data Length = 61440" interesting.. [10:38] apw: could this short URBs data length be a limitation of usbmon itself? === psivaa-afk is now known as psivaa [10:58] apw: according to gdb, the libusb_bulk_transfer call returns with exactly the amount of data it expected... [11:45] silidan1, it is possible that is a capture length, strace has a similar thing, you can often change it === psivaa is now known as psivaa-lunch [12:00] apw: any hint how? there seems to be no manpage for usbmon, at least on my system [12:01] silidan1, sorry no [12:12] apw: can you point me where i should ask about usbmon if not here? [12:19] silidan1, sorry i have no additional info on that tool [12:40] does anyone here know if the usbmon kernel module has a limitation on the size of captured frames? [12:41] does anyone here know if the usbmon kernel module has a limitation on the size of captured frames? [12:55] silidan: look at the drivers/usb/mon/ source [12:57] silidan: default buffer for binary capture looks to be ~300KB, and the max 1.2MB, but it looks like that is in total, not per-packet === psivaa-lunch is now known as psivaa [13:49] apw: a little question, lets assume in my app is using libusb, i know i want to receive 30fps from my camera and my camera can provide them. Can I just issue 30 libusb_bulk_transfers every second to achive that? [13:49] silidan, i have no idea, but in their sceario they are sending continious packets without having to receieve one, which makes a difference [13:50] apw: you refer to their winxp scenario? [13:50] yes [13:51] apw: but arent these receiveing bulk transfer also issued from the host controller via URBs? i mean doesnt just the app send the URB for bulk in, and waits for the kernel to basically return it with data in? [13:52] silidan, nope, it sends some vendor specific control message at the start, and never speaks again [13:53] apw: but usb 2.0 devices cant initiate a bulk transfer on their own.. this is done from host side if im not mistaken... like host asks device for bulk transfer... waits till device sendit, and returns the data to the app.. [13:53] silidan, that is not what the trace says [13:54] apw: but the trace just reports the URBs send to the kernel and their contents, the kernel operates based on that data, ie reads from it or writes to it and then just issues a callback if i got that right from reading so far? [13:55] apw: o wait, you said trace, you meant the strace output? [16:00] hey there, can i have some help? [16:01] i have latest ubuntu version available and i want to recompile the kernel with custom parameters [16:01] Why? [16:01] is there any easy way to achieve that? i failed at "sudo apt-get source linux-image-`uname -r`" step [16:02] mikkqu: https://help.ubuntu.com/community/Kernel/Compile [16:02] actually, there are two points why: 1. for learning purposes and to find out how it works in ubuntu, 2. i just want to experiment with option MODULE_FORCE_UNLOAD to be able to unload broken modules [16:02] and I know that's not safe, i just want to get some experience with that [16:04] jpds: I've seen the page, but there is no (yet) instructions for 15.04 which I have [16:04] mikkqu: I can't imagine that they change that much [16:04] I'm stuck with downloading the kernel source archive for my currently using kernel [16:04] there is a necessary step for obtaining the source: sudo apt-get source linux-image-`uname -r` [16:04] and it's just don't work for me for some reason [16:05] Picking 'linux-lts-raring' as source package instead of 'linux-image-3.8.0-19-generic' [16:05] E: Unable to find a source package for linux-lts-raring [16:05] maybe I need to have some special repository enabled for that? [16:05] and I have no clue why it suggesting me to pick "linux-lts-raring" which is not found anyway [16:07] mikkqu: are you on ubuntu 12.04? [16:07] no, i'm on the latest available, (not lts) - 15.04 [16:07] according to lsb_release -a [16:07] and running a 3.8 kernel? [16:08] yes, it's just by default. i didn't touch anything [16:08] since i've installed the latest upgrade [16:08] hm 15.04 should use 3.19 [16:09] use apt-get source linux and you should get the latest source for 15.04 [16:09] hmm.. that makes sense actually. i've been tricked that 3.8 is latter than 3.19 [16:09] if you need 3.8 you have to download it manually from launchpad [16:09] but now I see that it's not [16:10] i think i should switch my kernel to 3.19 first, then [16:14] hmm.. i 100% remember i did "sudo apt-get dist-upgrade" (actually, two times) [16:14] why my kernel wasn't switched automatically? [16:15] maybe you removed the linux-generic package so it didn't upgrade [16:16] ops, looks like you're right [16:18] rtg: whats the numbering scheme for the raspi tags in ubuntu-wily git tree [16:18] rtg: 1008 is the starting ABI and will just grow [16:18] ? [16:18] arges, monotonically increasing ? [16:19] rtg: ok, just making sure (since its the first one). thanks [16:20] arges, we used to have some package name conflicts, so haveing an ABI number like this is a bit historical. blame it on ppisati (cause he's not here) [16:21] gotcha [17:11] jpds, https://wiki.ubuntu.com/Kernel/BuildYourOwnKernel is often an easier starting point for folks wanting to build their own kernel === mike_ is now known as Guest52120 === Guest52120 is now known as mikkqu [17:23] jsalisbury, https://bugs.launchpad.net/ubuntu/wily/+source/linux/+bug/1440103/comments/16 [17:23] Ubuntu bug 1440103 in linux (Ubuntu Vivid) "[Hyper-V] Kernel panic not functional on Vivid" [Medium,In progress] [17:25] rtg, awesome. Thanks! [17:29] thanks guys, i've updated to the latest kernel and i've been able to download ubuntu kernel sources with apt-get source [17:30] there is a directory called "linux/", and some gz archive "diff.gz", i guess it's diff between ubuntu and vanilla linux kernel [17:30] i think, it's already applied to the files in "linux/" directory? no need to apply it, so it's just for reference? [17:31] when i'm trying to run "fakeroot debian/rules editconfigs" i still getting a lot of errors [17:32] i'm getting the same errors as in this question, created a month ago. would be great if you could take a look [17:32] http://unix.stackexchange.com/questions/225020/check-config-errors-on-linux-3-19-0-kernel-configured-under-ubuntu-15-04-x86-64 [17:41] ok, it worked in the "old-fashioned" way [17:45] mikkqu, i owuld normally recommend you start from the git repository rather than a source packag [17:45] mikkqu, as that is what we all do when working with the very same source [17:52] apw: hmm, yeah, git might be a more convenient way, but I'll try to stick to packaging method just this one time [17:52] how it'll work well. i should probably modify grub somehow, to be able to choose between kernels in case of fail [17:53] mikkqu, some of the permissions bits get lost in the package on things you use when modifying the package [17:53] mikkqu, so don't be supprised [17:54] apw: i guess, in case of permission problems, the worst case is just the broken build near the end [17:55] but something still could go wrong [17:57] never thought kernel build takes so much time. is it possible to shorten the build time with ccache or something like this? in case if you have to build it a lot? [18:19] mikkqu, to build it "the way the archive does" is tricky to do those tricks [18:19] mikkqu, then again i normally build in a PPA because that uses someone elses compute [18:32] oh crap! build actually failed, after about an hour of work [20:32] What process runs bluetooth or is it in kernel? [20:57] The thing crashes and respawn. [20:58] And says bluetooth process is it talking about bluetoothd? [21:22] Mbp === root is now known as Guest53172 [22:36] Looks like btusb is disabled [22:36] It doesn't load where is the block list?