G'day all,
Been playing with multiple vm's on linux and hacking around my suspend/resume
dropped interrupt issues.
I just noticed the /dev/rtc interface appears to be very single use oriented.
If I run multiple vm's
simultaneously on the same machine, the 1st one picks up the interrupt driven timer and the rest
must revert to userspace based timers.
open("/dev/rtc", O_RDONLY|O_LARGEFILE) = -1 EBUSY (Device or resource busy)
Not having investigated this much further as I'm testing on a fairly low powered device, but I
wonder how this impacts server applications where we desire to run several qemu instances concurrently.
(I have already played with dropping the timer interrupt rate back from 1024 to 256 and it really
does not appear to impact my linux guests, but it makes my windows guests audio unusable. Still
experimenting with that)
Just fishing for comments really.. wondering about some form of shared interrupt timing device to
allow multiple instances to run with full timer resolution or if it really even matters.
Brad
--
"Human beings, who are almost unique in having the ability
to learn from the experience of others, are also remarkable
for their apparent disinclination to do so." -- Douglas Adams
_______________________________________________
Qemu-devel mailing list
Qemu-devel@nongnu.org
http://lists.nongnu.org/mailman/listinfo/qemu-devel
