17:07 <mwalle> cool jon is alive ;)

17:10 <sb0> yes, and that patch seems to work :)

17:10 <mwalle> ysionneau: well because i haven't pushed the mmu upstream yet, and won't do that until at least the MMU interface register set is stable (eg. what about ASID etc), i'd suggest that you try to get the gdb patch upstream

17:12 <mwalle> or maybe there is some kind of runtime switch to enable that feature

17:16 <mwalle> other gdb backends (like the one on milkymist) needs to be changed too

17:17 <mwalle> https://sourceware.org/gdb/onlinedocs/gdb/Target-Descriptions.html#Target-Descriptions

17:19 <mwalle> ysionneau: so the actual fix should be to add support for that target desciptors to qemu

17:21 <mwalle> ysionneau: you could try if your qemu patch works with the gdb patch if you use the "set tdesc filname <xml desc file>"

17:57 <sb0> larsc, is the linux-milkymist git head supposed to compile?

18:00 <sb0> e.g. the lm32 code still tries to use num_physpages even though it has been replaced

18:10 <larsc> hm

18:11 <larsc> it should, but maybe I have uncommited local changes

18:34 <sb0> reverted to commit 6a2f2e4198eaff63ee75f6085ce9f966c47b4441

18:35 <sb0> it compiles with gcc 4.8.2, and in qemu, freezes at "calibrating delay loop". not sure if gcc can be blamed for that.

18:37 <larsc> that usually means that there are no timer interrupts

18:37 <larsc> I'll take a look when I'm home

18:37 <larsc> in 1h our so

20:52 <GitHub71> [linux-milkymist] larsclausen pushed 1 new commit to master: http://git.io/IthwCw

20:52 <GitHub71> linux-milkymist/master 1f446f9 Lars-Peter Clausen: Fix merge fallout...