pmc_reprogram_counter() always sets a sample period based on the value of pmc->counter. However, hsw_hw_config() rejects sample periods less than 2^31 - 1. So for example, if a KVM guest does
struct perf_event_attr attr; memset(&attr, 0, sizeof(attr)); attr.type = PERF_TYPE_RAW; attr.size = sizeof(attr); attr.config = 0x2005101c4; // conditional branches retired IN_TXCP attr.sample_period = 0; int fd = syscall(__NR_perf_event_open, &attr, 0, -1, -1, 0); ioctl(fd, PERF_EVENT_IOC_DISABLE, 0); ioctl(fd, PERF_EVENT_IOC_ENABLE, 0); the guest kernel counts some conditional branch events, then updates the virtual PMU register with a nonzero count. The host reaches pmc_reprogram_counter() with nonzero pmc->counter, triggers EOPNOTSUPP in hsw_hw_config(), prints "kvm_pmu: event creation failed" in pmc_reprogram_counter(), and silently (from the guest's point of view) stops counting events. We fix event counting by forcing attr.sample_period to always be zero for in_tx_cp counters. Sampling doesn't work, but it already didn't work and can't be fixed without major changes to the approach in hsw_hw_config(). Signed-off-by: Robert O'Callahan <rob...@ocallahan.org> --- arch/x86/kvm/pmu.c | 13 ++++++++++--- 1 file changed, 10 insertions(+), 3 deletions(-) diff --git a/arch/x86/kvm/pmu.c b/arch/x86/kvm/pmu.c index 06ce377..026db42 100644 --- a/arch/x86/kvm/pmu.c +++ b/arch/x86/kvm/pmu.c @@ -113,12 +113,19 @@ static void pmc_reprogram_counter(struct kvm_pmc *pmc, u32 type, .config = config, }; + attr.sample_period = (-pmc->counter) & pmc_bitmask(pmc); + if (in_tx) attr.config |= HSW_IN_TX; - if (in_tx_cp) + if (in_tx_cp) { + /* + * HSW_IN_TX_CHECKPOINTED is not supported with nonzero + * period. Just clear the sample period so at least + * allocating the counter doesn't fail. + */ + attr.sample_period = 0; attr.config |= HSW_IN_TX_CHECKPOINTED; - - attr.sample_period = (-pmc->counter) & pmc_bitmask(pmc); + } event = perf_event_create_kernel_counter(&attr, -1, current, intr ? kvm_perf_overflow_intr : -- 2.9.3