<mato>
kensan: No CPU usage, and nothing is happening...
<kensan>
mato: What was the build command?
<mato>
kensan: I just did "make". Lots of stuff happened and now nothing seems to be happening.
<kensan>
mato: The proof tools on GPL 2016 have some issues and do take a long time. For building/running you do not need to perform the SPARK proofs.
<kensan>
mato: Can you try "make NO_PROOF=true"
<mato>
kensan: Do I want a "make distclean" or can I just restart make with that?
<kensan>
mato: You should not need to clean. The necessary steps should be performed.
<mato>
ok, that worked
<kensan>
mato: To test the unikernel in Bochs use this incantation: make SYSTEM=xml/mirage-solo5.xml COMPONENTS="libmutime libdebuglog libxhcidbg sm time dbgserver" NO_PROOF=true emulate
<mato>
kensan: Ok, plain "make NO_PROOF=true emulate" works. I'll look at Mirage in a bit, multitasking between different things. Will get back to you.
<mato>
kensan: works as in IIUC a Linux guest boots up and I get a root shell.
<kensan>
mato: Cool, thanks for testing!
<kensan>
mato: btw, the log output can be found in emulate/serial.out
<kensan>
mato: To get a cleaned up log of only the Solo5 messages you can use this command:
<mato>
kensan: (I haven't forgotten you, distracted by reviewing the ARM64 ukvm PR)
<hannes>
mato: oh nice, the FreeBSD CI turns out to be useful (or yelling there for #include <linux/kvm.h>) ;)
<kensan>
mato: No worries, just trying to preempt pitfalls ;)
<kensan>
Will probably do a writeup/blog post once all the bits (ocaml-freestanding + Mirage) are merged.
<dstolfa>
hannes: It's funny to see people get yelled at for breaking builds on FreeBSD, I often get yelled at for breaking builds on Linux
<dstolfa>
CI is a nice thing :D
<hannes>
I usually avoid to yell and fix code I use which does not compile on FreeBSD...
<dstolfa>
hannes: Well, the only thing that yells at me is the CI, and I'm not much of a person to yell at others :P
<mato>
*phew* massive review submitted
<mato>
I think it's almost beer o'clock :)
argent_smith has quit [Quit: Leaving.]
copy` has joined #mirage
<dstolfa>
mato: Nice. I'm dealing with code that has a comment related to delay(1) in the kernel, which states that the developer "hopes" that it will be enough time to do the things in the taskqueue
<dstolfa>
It sounds like headache o'clock here
argent_smith has joined #mirage
<mato>
dstolfa: Ah yes, the quintessential distributed/operating systems "fix". Just add sleep(1) and hope. :-)
<dstolfa>
I should be in Cambridge starting 4th of July
<hannes>
mato: and then subsequent tweaks to use nanosleep(0.2) instead (or even less)
<mato>
dstolfa: I'll be in the area in late August, we can meet up for some beers if you'll be around.
<hannes>
dstolfa: 4th july!? neat!
<mato>
dstolfa: Or when I'm in Cambridge (I visit semi-regularly)
<dstolfa>
mato: Sounds great! :)
<dstolfa>
hannes: Yeah, spoke to TES today
<dstolfa>
Okay well, I've just touched that code in a pretty significant way. Time to blow up my machine I guess?
<hannes>
dstolfa: oh, TES? not a phd / more permanent?
<dstolfa>
hannes: PhD right after that :)
<dstolfa>
Doing TES during summer
<hannes>
hmm, /me is struggling with wait4 ... i seem to not get sensible data out of its rusage (for a child process -- unless I call getrusage in the parent beforehand)!?!?
<hannes>
(passing WNOHANG as flag...)
<hannes>
ah, oh... it is even documented it seems... hmmm
<hannes>
[pls correct me if I'm wrong -- I try to get rusage of a child process, and I thought wait4 would be the way to go)]
<dstolfa>
hannes: Seems like it would work to me, but I don't know if it's the standard way to go about these things or not
<dstolfa>
These changes... seem to work?
<dstolfa>
Eh, there's the panic
<dstolfa>
Oh, heh. That seems to have worked
brson has joined #mirage
argent_smith has quit [Quit: Leaving.]
AltGr has left #mirage [#mirage]
<hannes>
(nvm, looks like it's the wrong interface for my problem)
dudelson has quit [Ping timeout: 260 seconds]
dudelson has joined #mirage
argent_smith has joined #mirage
balduin has quit [Ping timeout: 240 seconds]
balduin has joined #mirage
balduin has quit [Ping timeout: 260 seconds]
copy` has quit [Quit: Connection closed for inactivity]