10:51 <lekernel> kristianpaul: no, there's no framebuffer yet

10:52 <lekernel> kristianpaul: have you been able to fix your clang build problem?

10:52 <lekernel> I have cmake version 2.8.8

10:52 <lekernel> what's yours?

11:05 <kristianpaul> not fixed afaik...

11:05 <kristianpaul> t2.8.7-0ubuntu4...

11:06 <kristianpaul> from last ubuntu lts

11:18 <lekernel> try autocrap maybe (llvm has both)

11:21 <lekernel> iirc JP has written some autocrap instructions in one of the readme's

11:24 <kristianpaul> k

11:25 <kristianpaul> bbl going to work !

11:39 <stekern> 2.8.7 should work fine (it does here at least)

11:45 <lekernel> stekern: you have built the modified lm32 version? works for you?

11:47 <stekern> yes

14:56 <kristianpaul> stekern: odd for me then..

14:56 <kristianpaul> stekern: lekernel what is your distro?

14:56 <kristianpaul> i'll try on trusty debian :)

15:01 <lekernel> fedora

15:06 <kristianpaul> F17?

15:06 <kristianpaul> had u tried the urjtag recently btw?

15:06 <kristianpaul> i know it got ported but dunno if works well and the usb bugs the other day too

15:48 <lekernel> f16

15:48 <lekernel> haven't reinstalled urjtag for a while...

15:54 <kristianpaul> ;-)

19:25 <kristianpaul> following wolfspra1l qoute about cpu processing mp3 at low mhz

19:25 <kristianpaul> http://opencores.org/articles,1016920279

20:12 <Fallenou> http://pastebin.com/DHU8yf9c < ahahah

20:12 <Fallenou> boom

20:13 <Fallenou> (sorry for the basic instinct of liking when it just blows)

20:37 <kristianpaul> oh wow an activate mmu command !

20:38 <Fallenou> ;)

20:38 <Fallenou> wait a sec

20:39 <kristianpaul> i tought that was decided just in sinthesis

20:41 <Fallenou> MMU logic is included or not at synthesis time indeed

20:41 <Fallenou> the "mmu activate" means going into user mode (which means virtual addresses gets translated into physical ones)

20:41 <Fallenou> and "mmu disactivate" means going into kernel mode

20:41 <Fallenou> which disactivates address translations

20:42 <kristianpaul> what is that usefull for? (activate and desactivate)

20:43 <Fallenou> you know, usually you have 2 (or 4 or other number) of "mode"

20:43 <Fallenou> privileged mode / user mode

20:43 <Fallenou> something like this

20:43 <Fallenou> that we could call "kernel mode"/"user mode"

20:44 <Fallenou> when Linux kernel code is running, physical addresses are used

20:44 <Fallenou> kernel address space is not virtual, it's physical

20:44 <Fallenou> basically when kernel code runs, mmu is disabled

20:44 <Fallenou> and when you switch back to user processus, you activate back the MMU

20:45 <Fallenou> so you therefore need a way to activate/disactivate the mmu :)

20:45 <kristianpaul> so this switch happens all time?

20:45 <Fallenou> yes

20:45 <kristianpaul> ohh

20:45 <Fallenou> at each IRQ, each syscall

20:45 <Fallenou> and upon page fault (which is an exception)

21:00 <Fallenou> http://pastebin.com/gtp1EtMR :)

21:37 <wpwrak> nice ;-)

21:51 <Fallenou> hum hum SoC hangs if a memory store triggers a page fault :)

21:51 <Fallenou> I will need to simulate that one

21:52 <Fallenou> only the "page fault during memory load" is handled correctly atm so

21:52 <Fallenou> would have been too easy if everything was working :p

21:53 <Fallenou> http://pastebin.com/549beWPV ^^"

21:53 <Fallenou> time to sleep

21:53 <Fallenou> gn8 !

23:48 <qwebirc45162> i can tell you guys are technocrats. especially by the quote that was put on the box. so of course you would find it humorous that there's no way to hook up the milkymist to a computer. not unless of course you have the money for the vga converter. 500 dollars for nothing. im probably selling it (and i just got it today)