Repository navigation
Adds stats for the tier 2 optimizer #109329
Description
Currently we have no stats for anything regarding the tier 2 optimizer.
Without them we are making too many guesses about what we should be doing.
The performance numbers tell us that things aren't working as well as they should, although not too badly either.
However, performance numbers tell us nothing about why that is, or what is happening.
For example, #109038 should have increased the important ratio of (number of uops executed)/(traces started) but we have no idea if it actually did.
We need the following stats soon:
- Total micro-ops executed
- Total number of traces started
- Total number of traces created
- Optimization attempts
The following stats would also be nice, but are less urgent:
- Per uop execution counts, like we have for tier 1 instructions.
- Exit reason counts: polymorphism vs branch mis-prediction
- A histogram of uops executed per trace
Linked PRs
Metadata
Metadata
Assignees
Labels
No labels
Activity
brandtbucher commentedon Sep 14, 2023
We've been collecting and dumping stats for all of the counters in your first list for three months now... but they seem to be ignored in the
summarize_stats.pyscript:@mdboom, maybe you know why these aren't being included in the markdown summaries? From a quick skim, it looks like it might be because we ignore any counters that don't start with "Calls to", "Frame", "GC", or "Object". Maybe we should rework this to not ignore new counters?
brandtbucher commentedon Sep 14, 2023
I'd also like to see the reasons why projecting stopped:
nbody's traces should be 68-286 uops, but we cap them at 64, which means no loops are closed)@gvanrossum, any other stats we might like to see?
gvanrossum commentedon Sep 14, 2023
Those lists look pretty exhaustive. I’d also count anything that I deemed worthy of a DPRINTF call.
mdboom commentedon Sep 25, 2023
Seems plausible. Let's coordinate on how to get this done -- I'm happy to take this on as I have a bit more time these days.
pythongh-109329: Support for basic pystats for Tier 2
pythongh-109329: Support for basic pystats for Tier 2
mdboom commentedon Sep 26, 2023
I have a PR for the basics up, and then I will tackle some of the other suggestions in smaller chunks.
Some questions:
Currently the Tier 2 interpreter affects the results for the Tier 1 interpreter by virtue of calling
STAT_INC. Is that just by accident of how we got here? I would assume we want to have completely separate opcode execution counts for the Tier 1 and Tier 2 interpreters, but thought I should confirm."A histogram of uops executed per trace" -- I assume this is a histogram of the number of uops executed per trace, not a count-by-type-of-uop-per-trace?
gvanrossum commentedon Sep 26, 2023
I think it would be more useful to have separate counters per tier. But it would be somewhat complex to do that -- the header files that define them would have to check for
TIER_ONEandTIER_TWO(only one of these should be defined).Sounds like that question is for @brandtbucher.
brandtbucher commentedon Sep 27, 2023
Yeah, we want to see the distribution in number of uops executed per entry into tier two.
That's helpful because we can execute many uops in traces that are statically short (if they close a hot loop) and few uops in traces that are statically long (if we deopt quickly). We want to optimize for "uops executed in tier two before deopting", not necessarily "trace projection length".
To ease implementation, it's fine to bucket these. The distribution will probably be heavily skewed towards small numbers (since we tend to enter and exit "bad" traces more often), so maybe powers of ten or two would work best?
GH-109329: Add tier 2 stats (GH-109913)
pythongh-109329: Add stat for "trace too short"
gh-109329: Add stat for "trace too short" (GH-110402)
gh-109329: Count tier2 opcode misses (#110561)
pythongh-109329: Count tier2 opcode misses (python#110561)
markshannon commentedon Jan 11, 2024
This is all working nicely now.
pythongh-109329: Count tier2 opcode misses (python#110561)
pythonGH-109329: Add tier 2 stats (pythonGH-109913)
pythongh-109329: Add stat for "trace too short" (pythonGH-110402)
pythongh-109329: Count tier2 opcode misses (python#110561)