| === hggdh_ is now known as hggdh | ||
| legend2440 | looking at config-2.6.28-15-generic i see that CONFIG_HZ=250 and that # CONFIG_HZ_1000 is not set. would there be any noticeable benefit to recompiling with CONFIG_HZ_1000 enabled? | 04:33 |
|---|---|---|
| maxb | What's the proper way to request a backport of an upstream kernel cset as SAUCE? Straightforward bug on the linux package? | 07:38 |
| amitk | smb: You set LSM_MMAP_MIN_ADDR to 0 for the x86 archs | 08:56 |
| smb | yes | 08:56 |
| amitk | I don't see that documented in the Kconfig | 08:56 |
| amitk | what does that do? | 08:56 |
| amitk | I need to decide what to set it to for ARM | 08:57 |
| smb | Upstream added a hard low barrier for the mmap addresses. So if you set LSM_MMAP_MIN_ADDR to any value you cannot change the mmap_min_addr in the running system to something lower than that | 08:57 |
| smb | There is a DEFAULT_MMAP_MIN_ADDR which sets the default value, and the LSM version is the lower barrier | 08:58 |
| amitk | aah, I get it now | 08:58 |
| smb | Changing that to 0 might be a temporary solution until we know how to allow certain applications access to lower regions with that in effect | 08:58 |
| amitk | and I presume a sysfs interface allows one to change it to any value between DEFAULT_MMAP_MIN_ADDR and LSM_MMAP_MIN_ADDR | 08:59 |
| smb | correct | 08:59 |
| smb | err | 08:59 |
| smb | no | 08:59 |
| smb | you can change it to anything between LSM_MMAP_MIN_ADDR...max | 08:59 |
| amitk | LSM_MMAP_MIN_ADDR <= DEFAULT_MMAP_MIN_ADDR <= max ? | 09:00 |
| smb | The DEFAULT is just the initial starting point after boot | 09:00 |
| amitk | so LSM_MMAP_MIN_ADDR <= DEFAULT_MMAP_MIN_ADDR <= max ? | 09:01 |
| smb | I cannot remember to have seen code to make sure DEFAULT is bigger or equal than LSM_MIN | 09:01 |
| smb | It probably is assumed you act sensible | 09:01 |
| amitk | ok, in any case LSM_MMAP_MIN_ADDR=0 allows userspace policy overrides w/o any sideeffects? | 09:02 |
| smb | I would say so | 09:02 |
| * amitk goes about changing the ARM values to 0 | 09:03 | |
| smb | The thing is maybe, for security reasons, you want to have it set so normal userspace cannot go below a certain value and then define some security framework exceptions for those special apps that really need it | 09:03 |
| amitk | smb: I was hoping that the 0 recommendation was from our security team | 09:04 |
| amitk | and that they are working on policy to enforce it | 09:04 |
| smb | amitk, Well the 0 was from our security team because the upstream change broke the current interface. I still have to get back to them with that proposeal/question | 09:07 |
| amitk | ogra: regarding bug 423767 | 09:18 |
| ubot3` | Malone bug 423767 in linux-fsl-imx51 "please enable rt3070sta in the imx51 kernel" [Medium,New] https://launchpad.net/bugs/423767 | 09:18 |
| amitk | ogra http://pastebin.ubuntu.com/267080/ | 09:18 |
| amitk | that driver is currently x86-only | 09:18 |
| * ogra points to #mobile | 09:18 | |
| === Ng_ is now known as Ng | ||
| amitk | ogra: re: bug 424400 | 09:41 |
| ubot3` | Malone bug 424400 in linux-fsl-imx51 "DM9601 or Pegasus based USB NICs dont find their MAC address under 2.6.31-100-imx51" [High,New] https://launchpad.net/bugs/424400 | 09:41 |
| amitk | is this a bug only with the imx51 flavour or x86 as well? | 09:42 |
| amitk | nevermind, it seems to work on i386 according to you last line in your report | 09:42 |
| ogra | amitk, its not my bug :) | 09:43 |
| amitk | you reported it | 09:44 |
| ogra | and i cant reproduce it with any NIC i have | 09:44 |
| ogra | i moved it out of another bug that got very confusion because it had ten different things going on | 09:44 |
| ogra | its Tobins bug | 09:44 |
| amitk | ok | 09:44 |
| ogra | amitk, tobin has a million of starnge things going on with his setup ... using a chain of USB KVMs, i still think its power related or his board is screwed | 09:46 |
| amitk | ogra: who is Vladimir Vukicevic? | 09:46 |
| ogra | no idea | 09:46 |
| ogra | where do you see him ? | 09:46 |
| amitk | bug 418238 | 09:46 |
| ubot3` | Malone bug 418238 in linux-fsl-imx51 "Ethernet (FEC) not functional on Babbage with 2.6.31" [Medium,Incomplete] https://launchpad.net/bugs/418238 | 09:46 |
| ogra | oh, wait | 09:46 |
| ogra | yeah, i remember him, he has a B1 | 09:47 |
| amitk | FSL? | 09:47 |
| ogra | he hangs around in #ubuntu-armel | 09:47 |
| ogra | i dont think so | 09:47 |
| ogra | FSL doesnt touch the B1 platform anymore | 09:47 |
| ogra | (he is vlad in -armel) | 09:48 |
| amitk | neither do I (touch the B1, that is). Atleast for now. | 09:48 |
| ogra | right, feel free to say its not supported and mark it invalid ... or dup it against the actual NIC bug | 09:49 |
| ogra | i suspect once the NIC works generally it will work on the B1 too | 09:49 |
| amitk | ogra: I don't see any other nic bug | 09:49 |
| ogra | hmm, didnt i file it | 09:50 |
| amitk | so I changed the description on that to be applicable to B2 | 09:50 |
| amitk | https://bugs.edge.launchpad.net/ubuntu/+source/linux-fsl-imx51 | 09:50 |
| ogra | ok | 09:50 |
| ogra | fine with me | 09:50 |
| Twigathy | rar | 12:10 |
| Twigathy | I just got a kernel oops... is this dead hardware or a bug in the kernel? http://paste.lisp.org/display/86753 (Last line is `[ 954.886783] Fixing recursive fault but reboot is needed!') | 12:10 |
| === zul_ is now known as zul | ||
| amitk | lool: ogra http://people.canonical.com/~amitk/mx51/linux-image-2.6.31-100-imx51_2.6.31-100.8_armel.deb has the rt3070 wireless enabled. (Plus the non working sound driver) | 13:19 |
| === bjf-afk is now known as bjf | ||
| ogra | amitk, installing ... | 13:46 |
| lool | Twigathy: no idea | 13:59 |
| lool | amitk: ok thanks | 13:59 |
| ogra | amitk, you didnt break it hard enough, still boots properly ;) | 14:01 |
| amitk | damn. I need to work harder. | 14:01 |
| ogra | heh | 14:01 |
| ogra | http://paste.ubuntu.com/267256/ | 14:02 |
| amitk | ogra: btw, did you insmod the sound driver last time? | 14:04 |
| ogra | no, it was loaded automatically | 14:04 |
| amitk | ok | 14:04 |
| ogra | i rmmodded all sound stuff and modprobed it though | 14:04 |
| ogra | neither gave any msg in dmesg | 14:05 |
| ogra | amm | 14:07 |
| ogra | hmm even | 14:07 |
| ogra | http://paste.ubuntu.com/267260/ | 14:07 |
| ogra | ogra@babbage2:~$ sudo iwlist scan 2>&1|grep ^wlan | 14:08 |
| ogra | wlan4 Interface doesn't support scanning : Network is down | 14:08 |
| ogra | ogra@babbage2:~$ sudo ifconfig wlan4 up | 14:08 |
| ogra | SIOCSIFFLAGS: Unknown error 132 | 14:08 |
| ogra | amitk, so i'D say that driver does noting helpful for us atm | 14:18 |
| amitk | too bad | 14:19 |
| ogra | amitk, oh, it didnt work because the kernal always loads rt3070sta and rt2800usb at the same time .... | 15:09 |
| ogra | amitk, rmmodding both and then modprobing rt3070sta gets me this intresting output: http://paste.ubuntu.com/267302/ | 15:10 |
| amitk | aaah. yet another rt* driver that tries to load multiple drivers | 15:10 |
| amitk | rt* device | 15:10 |
| ogra | well, the outcome of the dmesg info looks very intresting | 15:10 |
| ogra | some kind of ascii art | 15:11 |
| amitk | heh | 15:11 |
| maxb | hi, is a backport of an upstream changeset that fixes the acerhdf fan control driver to work on more kinds of Aspire One something that could be considered for SAUCE in karmic? | 15:51 |
| amitk | maxb: is it already going into 2.6.32? If so, yes. Pointer to a bug and a patch sent to kernel-team@lists.ubuntu.com are appreciated. | 15:52 |
| maxb | according to kernel patchwork it's currently in acpi-test | 15:52 |
| maxb | so aimed at 2.6.32 | 15:53 |
| maxb | http://patchwork.kernel.org/patch/39701/ <-- this is it. So, I should file a bug against "linux" sourcepackage and then email kernel-team? | 15:54 |
| amitk | maxb: yes, thats about right. | 15:55 |
| maxb | ok, will do | 15:56 |
| rtg | bjf, did 2.6.31-203.9 fix your Dove video issues? | 17:12 |
| bjf | rtg, I didn't test that build, I tested the one I had done with the patch | 17:13 |
| rtg | bjf, hint, hint... | 17:13 |
| bjf | rtg, ack | 17:14 |
| * amitk takes a break. Back in 40 mins for the meeting | 17:20 | |
| rtg | amitk, fix committed for bug #359049 | 18:34 |
| ubot3` | Malone bug 359049 in linux-mvl-dove "imx51 udeb hardcodes linux version in vmlinuz binary name" [High,Fix committed] https://launchpad.net/bugs/359049 | 18:34 |
| amitk | rtg: ack, thanks | 18:35 |
| rtg | amitk, thought I'd save you an hour of groveling around in kernel-wedge | 18:36 |
| bjf | rtg, actually you saved me, I was going to tell amitk I'd take care of it | 18:36 |
| amitk | rtg: you probably did. I have no idea what changing bdep from y to n does | 18:37 |
| rtg | amitk, actually, it was suffix, bdep isn't set | 18:37 |
| amitk | aah, ok | 18:38 |
| amitk | I'll pull in that patch to my rebased tree | 18:38 |
| amitk | rtg: does wireless-crda come from a completely separate source package? | 20:39 |
| rtg | amitk, it comes from a couple of git repositories | 20:40 |
| rtg | amitk, see debian/scripts/update-crda-from-git.sh | 20:41 |
| rtg | git://kernel.ubuntu.com/rtg/wireless-crda.git | 20:42 |
| amitk | thanks | 20:42 |
| amitk | rtg: hmm, it doesn't seem to be compiled for arm... | 20:43 |
| rtg | amitk, uh, no, probably not | 20:43 |
| amitk | I wonder how kernel packages are being installed in the arm live images currently. Because dpkg -i linux-image-2.6.31-100-imx51_2.6.31-100.8_armel.deb fails due to lack of wireless-crda | 20:45 |
| amitk | lool: ogra ^ | 20:45 |
| rtg | amitk, good question | 20:45 |
| amitk | rtg: the control file has Architecture: any | 20:48 |
| rtg | amitk, oh, well then its built for any arch supported by the buildd | 20:48 |
| amitk | rmadison wireless-crda | 20:50 |
| amitk | wireless-crda | 1.7 | jaunty | source, amd64, i386 | 20:50 |
| amitk | wireless-crda | 1.10 | karmic | source, amd64, i386 | 20:50 |
| amitk | rtg: any clue why it isn't available for arm ^ ? | 20:50 |
| rtg | amitk, I just tried pciutils, it only shows up for amd64 and i386 | 20:51 |
| amitk | packages.ubuntu.com shows the same | 20:52 |
| amitk | rtg: it could be because arm packages lives on ports.ubuntu.com | 20:59 |
| apw | TheMuso, seems our hda pcbeep channel is not well. if you do enable it the sound is no longer pure tones but a nasty rasp | 21:17 |
| Q-FUNK | ogasawara: anything new? :) | 21:32 |
| Q-FUNK | I seem to notice initramfs hooks to dump a kernel panic to a file with kexec-tools and makedumpfile but where is the file dumped to? | 23:38 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!