[05:14] <ppisati> moin
[07:55] <ppisati> smb: morning
[07:56] <smb> ppisati, apw, morning
[08:13] <apw> smb, moin
[08:13] <apw> ppisati, morning
[08:13] <ppisati> apw: hi
[08:13]  * ppisati feels hungry again...
[09:36] <ppisati> brb
[09:39] <ara> hello
[09:39] <ara> maybe someone should be looking at this bug? https://bugs.launchpad.net/ubuntu/+source/apport/+bug/1229611
[09:40] <ara> it looks like the apport hook for linux-lts-raring does not exist
[09:48] <apw> ara thanks for the heads up
[09:49] <ara> apw, np
[10:13] <ppisati> infinity: you need to add 'generic-lpae' to the Calxeda Highbank entry in flash-kernel/db/all.db
[10:14] <ppisati> infinity: since we switched from parsing /proc/cpuinfo to device-tree (/proc/device-tree/model)
[10:14] <ppisati> infinity: some entries have a different string id
[10:14] <ppisati> ubuntu@m00:~$ sudo flash-kernel
[10:14] <ppisati> Kernel suffix generic-lpae does not match any of the expected flavors (generic highbank), therefore not writing it to flash.
[10:14] <ppisati> ubuntu@m00:~$ cat /proc/device-tree/model 
[10:14] <ppisati> Calxeda Highbankubuntu@m00:~$
[10:15] <ppisati> infinity: and yes, it's a midway box
[10:15] <ppisati> infinity: problem is that, this way, we can't distinguish between ecx1000 and ecx2000 now
[10:16] <ppisati> infinity: uhm...
[10:32] <ppisati> but do we care in some install an lpae kernel on a non-lpae hw? do we have any seat belt implemented on x86 for this kind of scenario? i think we should follow the same path...
[10:32] <ppisati> infinity: ^
[10:32] <ppisati> and with that said,
[10:32]  * ppisati -> off for a bit
[11:18]  * apw idly wonders how the heck we coped before VMs
[11:21] <ogra_> manpower
[11:22] <ogra_> lots of 
[11:22] <apw> ogra_, heh perhaps so ... perhaps so
[14:51] <jsalisbury> **
[14:51] <jsalisbury> ** Ubuntu Kernel Team Meeting - Today @ 17:00 UTC - #ubuntu-meeting
[14:51] <jsalisbury> **
[14:55] <rtg> apw, pushed unstable, looks like its gonna build. I disabled overlayfs and lttng for FTBS (and dropped VESA modularization until I can fix it)
[14:55] <apw> rtg, ack
[16:07] <infinity> ppisati: No, we don't want that.  We want Calxeda to release their proper firmware that stops lying and pretending that Midway is Highbank.
[16:08] <ppisati> infinity: fine for me
[16:08] <infinity> ppisati: All the testing I've been doing on Midway for them involved a uboot fdt hack to fix the model.
[16:08] <infinity> ppisati: As instructed by Rob.
[16:08] <infinity> ppisati: I'm told their new firmware will happen this week.
[16:10] <robher> infinity, ppisati: why don't you look at capability flags indicating lpae is supported or not.
[16:11] <infinity> robher: Because that would be yet another mangling/rewrite of what flash-kernel does, a bit late in the cycle.
[16:11] <robher> model is a bad thing to look at to begin with. OEMs could put whatever crap they want there.
[16:12] <infinity> robher: But, sure, in the ideal utopia where all the kernels we carry are generic and generic-lpae, and no subarches matter anymore, that would be the sane thing to do, and we can scrap all this mess.
[16:13] <infinity> robher: I was told you guys are releasing your new firmware that stops pretending midway is highbank so, at least for you, this will be solved this week.
[16:14] <robher> infinity: yes that is right.
[16:14] <infinity> robher: For the bigger picture, I agree it's all a horrible mess, due to ARM's history of being horrible and messy, but we're not rewriting and solving all of that for this cycle.
[16:56] <jsalisbury> ##
[16:56] <jsalisbury> ## Kernel team meeting in 5 minutes
[16:56] <jsalisbury> ##
[19:04]  * rtg -> lunch