On Wed, Feb 22, 2017 at 7:08 PM, Taeung Song <treeze.taeung@xxxxxxxxx> wrote:
The output of perf-annotate has a problem.
It is so confusing that the output is mixed with
both source code and assembly code.
IMHO, we need readable annotate view based on source code,
not mixed view. (not depending on 'objdump -S')
And to do that, we can collect actual source code per function(sym)
using addr2line() and we can handle 'struct source_code'
that contains each line of code.
In near future, it would be used for new annotate view based on
actual source code per function(sym).
I think this is just a preparation so you'd be better sending it with
your new annotate view patchset later.