Mathieu Poirier | fdb433e | 2016-03-03 15:18:22 -0700 | [diff] [blame] | 1 | * ETMv4/PTM - decoder updates to handle advanced configuration. |
| 2 | -> Certain (currently unused by perf / current hardware) configuration settings |
| 3 | can alter the format of the trace output. One example is Return Stack - |
| 4 | settable in the control registers for PTM/ETMv4, and removes some inline |
| 5 | addresses. Decoder must use a follower to correctly trace when this is set. |
| 6 | |
| 7 | * ITM packet processing and decode. |
| 8 | -> ITM is primarily an M class SW trace module. I wouldn't expect to see it on |
| 9 | systems with STM, unless a companion M class was present. |
| 10 | |
| 11 | *Data trace - ETMv4 / ETMv3 |
| 12 | -> Differing solutions to data trace in v4/v3 - v4 is separate trace stream |
| 13 | completely, output at trace ID <instruction_trace_ID>+1. ETMv3 is inline with |
| 14 | the instruction trace. |
| 15 | |
| 16 | Cortex-A cores do not support this architecturally. On R and M profile cores it |
Mike Leach | 76c875a | 2017-06-09 08:11:35 +0100 | [diff] [blame] | 17 | is an option. There are scenarios in future that could see linux on R cores, plus |
Mathieu Poirier | fdb433e | 2016-03-03 15:18:22 -0700 | [diff] [blame] | 18 | on something like Juno it is possible to switch on trace for the SCP |
Mike Leach | 76c875a | 2017-06-09 08:11:35 +0100 | [diff] [blame] | 19 | (M class processor). So at some point data trace may be required. |