On 22/12/15 08:07, Shannon Zhao wrote:

From: Shannon Zhao <redacted>

This patchset adds guest PMU support for KVM on ARM64. It takes
trap-and-emulate approach. When guest wants to monitor one event, it
will be trapped by KVM and KVM will call perf_event API to create a perf
event and call relevant perf_event APIs to get the count value of event.

Use perf to test this patchset in guest. When using "perf list", it
shows the list of the hardware events and hardware cache events perf
supports. Then use "perf stat -e EVENT" to monitor some event. For
example, use "perf stat -e cycles" to count cpu cycles and
"perf stat -e cache-misses" to count cache misses.

I finally feel like we're pretty close to something we could merge. My
current concerns are:

- the 32bit bugs, and the fact that it has obviously not been tested
with a 32bit guest. It would be good to give it a go (you'll probably
have to lie about the nature of the PMU in your QEMU code by pretending
it is a Cortex-A15 PMU, for example).

- The UNDEF path and the patch I've posted today. I've quickly tested
it, and it seems to do the right thing, but this warrant a thorough review.

- I'd like Peter Maydell (cc'd) to give his Ack to the last patch, as we
need the userspace ABI to be agreed upon.

Thanks,

	M.
-- 
Jazz is not dead. It just smells funny...