Hi, I am an Outreachy intern working under Rik van Riel on memory allocation latency tracing using tracepoints. The goal of my project is to add tracepoints to code in vmscan.c and compaction.c to gain insight into what happens there and examine latencies using a postprocessing script.
The one here: https://github.com/Jananiravichandran/Analyzing-tracepoints/blob/master/shrink_slab_latencies.py is a very basic script that shows how long direct reclaim and shrinkers take. I intend to keep updating the script as more tracepoints are added in the direct reclaim and compaction code and eventually submit the script itself once I'm done. Suggestions on this are most welcome! As of now, there are no mechanisms to find out how long slowpath and memory compaction take to execute. This patchset adds new tracepoints and also modifies a couple of existing ones to address this and collect some zone information that may be useful. Janani Ravichandran (2): mm: page_alloc.c: Add tracepoints for slowpath mm: compaction.c: Add/Modify direct compaction tracepoints include/trace/events/compaction.h | 38 ++++++++++++++++++++++++++++++++----- include/trace/events/kmem.h | 40 +++++++++++++++++++++++++++++++++++++++ mm/compaction.c | 6 ++++-- mm/page_alloc.c | 5 +++++ 4 files changed, 82 insertions(+), 7 deletions(-) -- 2.7.0
