IRC logs for #openrisc Monday, 2015-10-12

--- Log opened Mon Oct 12 00:00:22 2015
blueCmdpoke53281: indeed!06:02
blueCmdpoke53281: have you seen anything like the behaviour in ?06:03
blueCmdpoke53281: weird lock-ups06:03
-!- orsonmmz|away is now known as orsonmmz07:21
blueCmdpoke53281: did you ever get virtio_net to work? the tar ball seems to indicate that you're using ethoc08:44
-!- orsonmmz is now known as orsonmmz|away10:11
-!- orsonmmz|away is now known as orsonmmz10:31
olofk_orsonmmz == Orson, the Icarus VHDL guy?11:59
orsonmmzolofk_: aye, hello ;)11:59
olofk_orsonmmz: Hi :)11:59
olofk_It was an interesting talk. I've been really waiting on a multi-language simulator12:00
orsonmmzolofk_: good job on organization, you guys deserve a really big beer for the event ;)12:00
orsonmmzolofk_: haha, I am afraid there is still a long way to go :P12:01
olofk_Yeah, I guess. I filed a few bugs a while ago (at sourceforge. Where do you want your bug reports, btw?)12:01
olofk_Adding support to Icarus would be pretty sweet, but I understand the problem of reimplementing the VHDL spec :/12:02
orsonmmzolofk_: I guess, sourceforge is the best place12:02
olofk_ghdl otoh is ADA and.... well, let's just say it's ADA12:02
orsonmmzunfortunately it is a bit broken, and it does not allow to change e-mail address for an account12:03
orsonmmzso I am getting notifications to a non-existing mail address :P12:03
orsonmmzso just from time to time, when I get a few spare minutes I peek there and try to fix things12:03
orsonmmzolofk_: true, Ada is an obstacle for us, but we would rather code in C and link it with the existing Ada code12:04
olofk_orsonmmz: No stress. It's bug 941-945 IIRC that I filed12:04
orsonmmzthat part is not very clear for the moment though..12:04
olofk_Yeah, if you can solve it at link time, it might be the least painful way12:05
olofk_That ghdl guy (can't remember his name. Something dutch...) is pretty active on the mailing list for the next VHDL revision, so I think ghdl is pretty up to date when it comes to language support12:06
-!- olofk_ is now known as olofk12:08
orsonmmzolofk: true, from what I can see, ghdl is quite complete12:09
orsonmmzolofk: and he seems to be cooperative, so I believe there is a chance to take advantage of Icarus & GHDL combined12:09
olofkThat's awesome! I really like cooperatin12:10
olofkandrzejr: Did you get a chance to try out barebox/uboot?12:52
* robtaylor waves from geneva airport13:30
* olofk waves back from Göteborg14:33
-!- orsonmmz is now known as orsonmmz|away15:50
poke53281blueCmd: I have implemented virtio_net and it is working.15:55
poke53281I have not activated it yet.15:56
poke53281But I remember, that you had this problem earlier. Let's look for some endian issues in my implementation.16:00
poke53281blueCmd: Sorry have to leave. Will look at your asciicinema later.16:08
poke53281blueCmd: But the virtio network has a lot of features. And I support the simplest ones. Maybe you can tune the device to send and receive only ethernet packets.18:57
poke53281blueCmd: Nope, I have never seen such weird lookups19:02
poke53281What patches do you use?19:02
poke53281with atomics?19:02
andrzejrolofk, not yet, trying some other ideas first19:10
andrzejrolofk, the kernel freeze issue is solved by not enabling i&dcache in kernel23:02
andrzejris there anything that could sneak through wb_transactor? For example, the transactor has not reported the memory sizing error (64MB instead of 128MB), which is not surprising - it only checks the location it tests.23:05
andrzejralso, unlike atlys, nexys4 currently uses a bus arbiter in front of the DDR controller's WB interface. Is there anything that could go wrong at this stage? In the past I've seen dbus bursts starving ibus.23:11
andrzejra newbie question: linux boot stops at "100.00 BogoMIPS (lpj=500000)". How to make it go further?23:13
--- Log closed Tue Oct 13 00:00:24 2015

Generated by 2.15.2 by Marius Gedminas - find it at!