* todfix
@ 2016-01-04 18:16 Ethan Grammatikidis
2016-01-04 19:05 ` [9front] todfix cinap_lenrek
0 siblings, 1 reply; 3+ messages in thread
From: Ethan Grammatikidis @ 2016-01-04 18:16 UTC (permalink / raw)
To: 9front
i have a 9front vm guest on a linux host. the host is suspended regularly. whenever it's resumed, the guest prints "todfix" and a number on the screen. if i understand the kernel code correctly, this is because the time jumped too much.
/sys/src/9/port/tod.c:210
it's a nuisance because it appears on the screen rather than through kprint, and it appears to be unnecessary; the guest has never failed to get the correct time after the host is resumed. the host has never been suspended more than about 12 hours, so it may not be much of a test. i've just paused the guest by a qemu feature, so if a longer test is wanted, i can keep it paused for a few days or a week or more. the guest's kernel is a little old (april) but i see port/tod.c:210 hasn't changed.up to the last iso.
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [9front] todfix
2016-01-04 18:16 todfix Ethan Grammatikidis
@ 2016-01-04 19:05 ` cinap_lenrek
2016-01-04 21:13 ` Ethan Grammatikidis
0 siblings, 1 reply; 3+ messages in thread
From: cinap_lenrek @ 2016-01-04 19:05 UTC (permalink / raw)
To: 9front
> i have a 9front vm guest on a linux host. the host is suspended regularly.
> whenever it's resumed, the guest prints "todfix" and a number on the screen.
> if i understand the kernel code correctly, this is because the time jumped too much.
so it appears that the tsc keeps ticking while the vm is suspended,
so fastticks() will jump ahead.
do you propose the removal of that iprint()? that would be fine with me.
--
cinap
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [9front] todfix
2016-01-04 19:05 ` [9front] todfix cinap_lenrek
@ 2016-01-04 21:13 ` Ethan Grammatikidis
0 siblings, 0 replies; 3+ messages in thread
From: Ethan Grammatikidis @ 2016-01-04 21:13 UTC (permalink / raw)
To: 9front
On Mon, Jan 4, 2016, at 07:05 PM, cinap_lenrek@felloff.net wrote:
> > i have a 9front vm guest on a linux host. the host is suspended regularly.
> > whenever it's resumed, the guest prints "todfix" and a number on the screen.
> > if i understand the kernel code correctly, this is because the time jumped too much.
>
> so it appears that the tsc keeps ticking while the vm is suspended,
> so fastticks() will jump ahead.
>
> do you propose the removal of that iprint()? that would be fine with me.
>
> --
> cinap
yes, removing the iprint would suit me
--
Wir mussen wissen, wir werden wissen.
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2016-01-04 21:13 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2016-01-04 18:16 todfix Ethan Grammatikidis
2016-01-04 19:05 ` [9front] todfix cinap_lenrek
2016-01-04 21:13 ` Ethan Grammatikidis
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).