Re: [PATCH v4 1/3] perf: Use monotonic clock as a source for timestamps

On 06/11/14 18:51, Pawel Moll wrote:

Until now, perf framework never defined the meaning of the timestamps
captured as PERF_SAMPLE_TIME sample type. The values were obtaining
from local (sched) clock, which is unavailable in userspace. This made
it impossible to correlate perf data with any other events. Other
tracing solutions have the source configurable (ftrace) or just share
a common time domain between kernel and userspace (LTTng).

Follow the trend by using monotonic clock, which is readily available
as POSIX CLOCK_MONOTONIC.

Also add a sysctl "perf_sample_time_clk_id" attribute which can be used
by the user to obtain the clk_id to be used with POSIX clock API (eg.
clock_gettime()) to obtain a time value comparable with perf samples.

Old behaviour can be restored by using "perf_use_local_clock" kernel
parameter.

Don't forget this breaks the relationship to TSC. So you will need something
like below. Also patch 3 needs to be done first and extended to cover TSC so
there is no gap when we cannot get TSC.

diff --git a/arch/x86/kernel/cpu/perf_event.c b/arch/x86/kernel/cpu/perf_event.c
index 143e5f5..b6e833d 100644
--- a/arch/x86/kernel/cpu/perf_event.c
+++ b/arch/x86/kernel/cpu/perf_event.c

@@ -1922,21 +1922,6 @@ void arch_perf_update_userpage(struct

perf_event_mmap_page *userpg, u64 now)
        userpg->cap_user_time_zero = 0;
        userpg->cap_user_rdpmc = x86_pmu.attr_rdpmc;
        userpg->pmc_width = x86_pmu.cntval_bits;
-
-       if (!sched_clock_stable())
-               return;
-
-       data = cyc2ns_read_begin();
-
-       userpg->cap_user_time = 1;
-       userpg->time_mult = data->cyc2ns_mul;
-       userpg->time_shift = data->cyc2ns_shift;
-       userpg->time_offset = data->cyc2ns_offset - now;
-
-       userpg->cap_user_time_zero = 1;
-       userpg->time_zero = data->cyc2ns_offset;
-
-       cyc2ns_read_end(data);
 }

 /*