#milkymist IRC log for Wednesday, 2012-12-12

mwallelekernel: what do you think about putting the lm32 core into an own git tree and include it as a submodule?11:43
lekernelhow many projects use it?11:56
lekernelthis sort of things has administrative overhead... then we could do the same for softusb, navre, etc.11:57
lekernelmain users of LM32 that I know of are milkymist, milkymist-ng and white rabbit (which lives on ohwr.org)11:57
Action: Fallenou likes git submodules12:59
Fallenouit actually reduces administrative overhead12:59
Fallenouyou do modifications only in one lm32 git repo12:59
Fallenouand then each project using lm32 can chose the lm32 SHA-1 it want to use13:00
Fallenouinstead of porting lm32 modifications etc13:00
larscfwiw I agree with mwalle and Fallenou13:11
kristianpaulsubmodules sounds like less forks and branchs, nice16:28
mwallelekernel_: well at least the module could be shared between milkymist and milkymist-ng18:57
mwalleand it is functional by its own, you can now even simulate it using only iverlog18:58
mwalleand it may encourage other users by setting the bar lower, eg. simple forking of the core18:59
mwallelarsc: wpwrak: should you be able to flush the cache in user mode?19:12
larsca application needs to be able to flush the cache19:13
mwalle(directly through writing to the DCC register)19:14
larscon some architectures this is one via a syscall though19:14
larscthe question is could anything bad happen if it is possible to directly flush the caches19:17
GitHub90[migen] sbourdeauducq pushed 1 new commit to master: https://github.com/milkymist/migen/commit/a7227d7d2bbf8d58de1a12180f40ea82cf47390021:09
GitHub90migen/master a7227d7 Sebastien Bourdeauducq: Remove ActorNode21:09
larscbtw. 3.7 worked almost out of the box, only had to apply one minor fixup21:11
GitHub44[migen] sbourdeauducq pushed 1 new commit to master: https://github.com/milkymist/migen/commit/28b4d99d31b4fb55f375d57553fc267af28e1bb721:29
GitHub44migen/master 28b4d99 Sebastien Bourdeauducq: replace some forgotten is_abstract()21:29
mwallelekernel_: btw whats the status on lm32-gcc and g++?21:30
mwallei saw that lattice included the gcc/g++ 4.3.0 in their micobuilder21:31
mwallehttp://www.codethink.co.uk/commandments << thats nice21:31
GitHub97[milkymist-ng] sbourdeauducq pushed 1 new commit to master: http://git.io/YZqeXA21:41
GitHub97milkymist-ng/master 3986790 Sebastien Bourdeauducq: Remove ActorNode21:41
larscseems to be a nice company21:44
lekernel_mwalle: no progress. 4.3 is "old"22:00
mwallelekernel_: "old" means it works?22:01
lekernel_afaik C++ works on 4.x22:01
lekernel_C works on 4.5, C++ has bugs22:02
lekernel_and everything is broken since 4.622:02
mwallehum, ok22:03
lekernel_s/4.x/4.3 and 4.422:03
lekernel_there was this patch http://lists.milkymist.org/pipermail/devel-milkymist.org/2012-August/003144.html22:07
lekernel_but it was never finished unfortunately22:07
GitHub34[milkymist] mwalle created mmu-cleanup (+25 new commits): https://github.com/mwalle/milkymist/compare/83c735b1ac66^...00ae9073e05023:54
GitHub34milkymist/mmu-cleanup 83c735b Michael Walle: lm32: replace clogb2 by builtin $clog2...23:54
GitHub34milkymist/mmu-cleanup 531a968 Michael Walle: lm32: rename mem array in lm32_dp_ram...23:54
GitHub34milkymist/mmu-cleanup 1fad281 Michael Walle: lm32: remove unneeded parameter in lm32_dp_ram...23:54
mwallewarning: internal error: dangerous error << hilarious :)23:58
mwalleis it a warning or an error?23:58
mwallethe severity seems to increase after ever colon :)23:59
--- Thu Dec 13 201200:00

Generated by irclog2html.py 2.9.2 by Marius Gedminas - find it at mg.pov.lt!