Re: [RFC 4/7] arm64: pmu: Add function implementation to update event index in userpage.

From: Raphael Gault
Date: Wed May 29 2019 - 08:43:23 EST


Hi Peter,

On 5/29/19 1:32 PM, Peter Zijlstra wrote:
On Wed, May 29, 2019 at 01:25:46PM +0100, Raphael Gault wrote:
Hi Robin, Hi Peter,

On 5/29/19 11:50 AM, Robin Murphy wrote:
On 29/05/2019 11:46, Raphael Gault wrote:
Hi Peter,

On 5/29/19 10:46 AM, Peter Zijlstra wrote:
On Tue, May 28, 2019 at 04:03:17PM +0100, Raphael Gault wrote:
+static int armv8pmu_access_event_idx(struct perf_event *event)
+{
+ÂÂÂ if (!(event->hw.flags & ARMPMU_EL0_RD_CNTR))
+ÂÂÂÂÂÂÂ return 0;
+
+ÂÂÂ /*
+ÂÂÂÂ * We remap the cycle counter index to 32 to
+ÂÂÂÂ * match the offset applied to the rest of
+ÂÂÂÂ * the counter indeces.
+ÂÂÂÂ */
+ÂÂÂ if (event->hw.idx == ARMV8_IDX_CYCLE_COUNTER)
+ÂÂÂÂÂÂÂ return 32;
+
+ÂÂÂ return event->hw.idx;

Is there a guarantee event->hw.idx is never 0? Or should you, just like
x86, use +1 here?


You are right, I should use +1 here. Thanks for pointing that out.

Isn't that already the case though, since we reserve index 0 for the
cycle counter? I'm looking at ARMV8_IDX_TO_COUNTER() here...


Well the current behaviour is correct and takes care of the zero case with
the ARMV8_IDX_CYCLE_COUNTER check. But using ARMV8_IDX_TO_COUNTER() and add
1 would also work. However this seems indeed redundant with the current
value held in event->hw.idx.

Note that whatever you pick now will become ABI. Also note that the
comment/pseudo-code in perf_event_mmap_page suggests to use idx-1 for
the actual hardware access.


Indeed that's true. As for the pseudo-code in perf_event_mmap_page. It is compatible with what I do here. The two approach are only different in form but it is in both case necessary to subtract 1 on the returned value in order to access the correct hardware counter.

Thank you,

--
Raphael Gault