Re: [PATCH] Documentation: coresight: Update the generic device names

From: Suzuki K Poulose
Date: Fri Jun 07 2019 - 04:53:08 EST

Hi Leo,

A Coresight PMU works the same way as any other PMU, i.e the name of the PMU is
listed along with configuration options within forward slashes '/'. Since a
Coresight system will typically have more than one sink, the name of the sink to
-work with needs to be specified as an event option. Names for sink to choose
-from are listed in sysFS under ($SYSFS)/bus/coresight/devices:
+work with needs to be specified as an event option.
+On newer kernels the available sinks are listed in sysFS under:
- root@linaro-nano:~# ls /sys/bus/coresight/devices/
- 20010000.etf 20040000.funnel 20100000.stm 22040000.etm
- 22140000.etm 230c0000.funnel 23240000.etm 20030000.tpiu
- 20070000.etr 20120000.replicator 220c0000.funnel
- 23040000.etm 23140000.etm 23340000.etm
+ root@localhost:/sys/bus/event_source/devices/cs_etm/sinks# ls
+ tmc_etf0 tmc_etr0 tpiu0
- root@linaro-nano:~# perf record -e cs_etm/@20070000.etr/u --per-thread program
+On older kernels, this may need to be found from the list of coresight devices,
+available under ($SYSFS)/bus/coresight/devices/:
+ root@localhost:/sys/bus/coresight/devices# ls
+ etm0 etm1 etm2 etm3 etm4 etm5 funnel0 funnel1 funnel2 replicator0 stm0 tmc_etf0 tmc_etr0 tpiu0
+ root@linaro-nano:~# perf record -e cs_etm/@tmc_etr0/u --per-thread program

On the "older" kernels you are referring to one would find the original naming
convention. Everything else looks good to me.

True, but do we care what we see there ? All we care about is the location,
where to find them. I could fix it, if you think thats needed.

IIUC, either the old kernel or newer kernel, both we can find the event
from ($SYSFS)/bus/event_source/devices/cs_etm/sinks/; the only
difference between them is the naming convention.

The cs_etm/sinks was only added with the CPU-wide trace support. So, if someone
refers to this document alone and then tries to do something on on older kernel,
which is quite possible for a production device running a stable kernel, {s,}he
might be surprised.

So the doc can use the same location to find event for both new and
old kernel, and explain the naming convention difference?

My question is really, does the naming convention matter ? What you see
under the directory is the name. But yes, I am open to add a section to
explain the fact that we changed the naming scheme, if everyone agrees
to it.