code.py, specifically, was getting messed up by inconsequential GCC
objdump errors on Clang -g3 generated binaries.
Now stderr from child processes is just redirected to /dev/null when
-v/--verbose is not provided.
If we actually depended on redirecting stderr->stdout these scripts
would have been broken when -v/--verbose was provided anyways. Not
really sure what the original code was trying to do...
The original idea was to allow merging a whole bunch of different csv
results into a single lfs.csv file, but this never really happened. It's
much easier to operate on smaller context-specific csv files, where the
field prefix:
- Doesn't really add much information
- Requires more typing
- Is confusing in how it doesn't match the table field names.
We can always use summary.py -fcode_size=size to add prefixes when
necessary anyways.
We already rely on this symbol in these scripts, so might use it to
display the mathematically correct ratio for new entries.
This has the added benefit of ordering new entries vs extremely big
changes correctly:
$ ./scripts/code.py -u test.after.csv -d test.before.csv
function (1 added, 0 removed) osize nsize dsize
test_a - 49 +49 (+∞%)
test_b 19 719 +700 (+3684.2%)
test_c 91 191 +100 (+109.9%)
TOTAL 110 959 +849 (+771.8%)
This is a bit more complicated, but make testmarks really showed how
confusing this could get.
Now, instead of:
suite passed time
test_alloc 304/304 1.6 (100.0%)
test_badblocks 6880/6880 1323.3 (100.0%)
... snip ...
test_rbyd 385878/385878 592.7 (100.0%)
test_relocations 7899/7899 318.8 (100.0%)
TOTAL 548206/548206 6229.7 (100.0%)
Percents/notes are interspersed next to their relevant fields:
suite passed time
test_alloc 304/304 (100.0%) 1.6
test_badblocks 6880/6880 (100.0%) 1323.3
... snip ...
test_rbyd 385878/385878 (100.0%) 592.7
test_relocations 7899/7899 (100.0%) 318.8
TOTAL 548206/548206 (100.0%) 6229.7
Note has no effect on scripts with only a single field (code.py, etc).
But it does make multi-field diffs a bit more readable:
$ ./scripts/stack.py -u after.stack.csv -d before.stack.csv -p
function frame limit
lfsr_bd_sync 8 (+100.0%) 216 (+100.0%)
lfsr_bd_flush 40 (+25.0%) 208 (+4.0%)
... snip ...
lfsr_file_flush 32 (+0.0%) 2424 (-0.3%)
lfsr_file_flush_ 216 (-3.6%) 2392 (-0.3%)
TOTAL 9008 (+0.4%) 2600 (-0.3%)
Now, fractions are considered equal if they have the same ratio:
- 6/6 == 12/12 => True
- 3/6 == 3/12 => False
- 1/6 == 2/12 => True
It's interesting to note this implementation is actually more
numerically stable than float comparison, though that wasn't really the
goal.
The main reason for this is to allow other fields to take over when
sorting multi-field fractional data: cov (lines + branches), testmarks
(passed + time), etc. Before, sorting would usually stop after
mismatched fraction fields, which wasn't all that useful.
This matches how diff percentages are rendered, and simplifies the
internal table rendering by making Frac less of a special case. It also
allows for other type notes in the future.
One concern is how all the notes are shoved to the side, which may make
it a bit harder to find related percentages. If this becomes annoying we
should probably look into interspersing all notes (including diff
percentages) between the relevant columns.
Before:
function lines branches
lfsr_rbyd_appendattr 230/231 99.6% 172/192 89.6%
lfsr_rbyd_p_recolor 33/34 97.1% 11/12 91.7%
lfs_alloc 40/42 95.2% 21/24 87.5%
lfsr_rbyd_appendcompaction 54/57 94.7% 39/42 92.9%
...
After:
function lines branches
lfsr_rbyd_appendattr 230/231 172/192 (99.6%, 89.6%)
lfsr_rbyd_p_recolor 33/34 11/12 (97.1%, 91.7%)
lfs_alloc 40/42 21/24 (95.2%, 87.5%)
lfsr_rbyd_appendcompaction 54/57 39/42 (94.7%, 92.9%)
...
Previously, with -d/--diff, we would only show non-zero percentages. But
this was ambiguous/confusing when dealing with multiple results
(stack.py, summary.py, etc).
To help with this, I've switched to showing all percentages unless all
percentages are zero (no change). This matches the -d/--diff row-hiding
logic, so by default all rows should show all percentages.
Note -p/--percent did not change, as it already showed all percentages
all of the time.
Note there's a bit of subtlety here, field _types_ are still infered,
but the intention of the fields, i.e. if the field contains data vs
row name/other properties, must be unambiguous in the scripts.
There is still a _tiny_ bit of inference. For most scripts only one
of --by or --fields is strictly needed, since this makes the purpose of
the other fields unambiguous.
The reason for this change is so the scripts are a bit more reliable,
but also because this simplifies the data parsing/inference a bit.
Oh, and this also changes field inference to use the csv.DictReader's
fieldnames field instead of only inspecting the returned dicts. This
should also save a bit of O(n) overhead when parsing CSV files.
The whitespace sensitivity of field args was starting to be a problem,
mostly for advanced plotmpl.py usage (which tbf might be appropriately
described as "super hacky" in how it uses CLI parameters):
./scripts/plotmpl.py \
-Dcase=" \
bench_rbyd_attr_append, \
bench_rbyd_attr_remove, \
bench_rbyd_attr_fetch, \
..."
This may present problems when parsing CSV files with whitespace, in
theory, maybe. But given the scope of these scripts for littlefs...
just don't do that. Thanks.
With the quantity of data being output by bench.py now, filtering ASAP
while parsing CSV files is a valuable optimization. And thanks to how
CSV files are structured, we can even avoid ever loading the full
contents into RAM.
This does end up with use filtering for defines redundantly in a few
places, but this is well worth the saved overhead from early filtering.
Also tried to clean up the plot.py/plotmpl.py's data folding path,
though that may have been wasted effort.
- Renamed struct_.py -> structs.py again.
- Removed lfs.csv, instead prefering script specific csv files.
- Added *-diff make rules for quick comparison against a previous
result, results are now implicitly written on each run.
For example, `make code` creates lfs.code.csv and prints the summary, which
can be followed by `make code-diff` to compare changes against the saved
lfs.code.csv without overwriting.
- Added nargs=? support for -s and -S, now uses a per-result _sort
attribute to decide sort if fields are unspecified.
- Fixed added/removed count in scripts when an entry has no field in
the expected results
- Fixed a python-sort-type issue when by-field is missing in a result
- Changed --(tool)-tool to --(tool)-path in scripts, this seems to be
a more common name for this sort of flag.
- Changed BUILDDIR to not have implicit slash, makes Makefile internals
a bit more readable.
- Fixed some outdated names hidden in less-often used ifdefs.
Based loosely on Linux's perf tool, perfbd.py uses trace output with
backtraces to aggregate and show the block device usage of all functions
in a program, propagating block devices operation cost up the backtrace
for each operation.
This combined with --trace-period and --trace-freq for
sampling/filtering trace events allow the bench-runner to very
efficiently record the general cost of block device operations with very
little overhead.
Adopted this as the default side-effect of make bench, replacing
cycle-based performance measurements which are less important for
littlefs.
