[cmaloney]

These exceptions and their subclasses had no vectorcall, and __init__ parsed keywords through PyArg_ParseTupleAndKeywords with a temporary empty tuple on every call. Error messages are unchanged except when more keywords than parameters are given, which now names the unexpected keyword.

The clinic inputs have no docstring, so docstrings are unchanged and these exceptions don't gain a __text_signature__.

pyperf microbenchmark (PGO+LTO, run serially):

Benchmark	base	vectorcall
AttributeError(s, name=s, obj=s)	127 ns	41.9 ns: 3.02x faster
ImportError(s, name=s, path=s)	129 ns	43.4 ns: 2.96x faster
NameError(s, name=s)	88.3 ns	39.3 ns: 2.24x faster
ModuleNotFoundError(s, name=s)	89.9 ns	40.6 ns: 2.21x faster
AttributeError(s)	39.4 ns	32.1 ns: 1.23x faster
UnboundLocalError(s)	38.6 ns	31.8 ns: 1.22x faster
NameError(s)	38.5 ns	32.2 ns: 1.19x faster
ImportError(s)	40.7 ns	34.4 ns: 1.18x faster
attribute_error_to_none()	176 ns	173 ns: 1.02x faster

Benchmark code:

import pyperf

runner = pyperf.Runner()
setup = """s = 'x'
class O: pass
o = O()
def attribute_error_to_none():
    try:
        o.x
    except AttributeError:
        pass
"""
for stmt in ["AttributeError(s)", "AttributeError(s, name=s, obj=s)",
             "NameError(s)", "NameError(s, name=s)",
             "ImportError(s)", "ImportError(s, name=s, path=s)",
             "ModuleNotFoundError(s, name=s)", "UnboundLocalError(s)",
             "attribute_error_to_none()"]:
    runner.timeit(stmt, stmt, setup)

• Issue: Improve performance of exceptions #161242

[eendebakpt]

There is a behavior change for e = ImportError('a'); e.__init__('b', foo=1). The call to e.__init__ fails both on main and the PR, but with the PR e.args is not modified. I think the change is good and no real issue. Maybe not even add a test for or a note in the news entry.

[cmaloney]

Yea, I was trying for a while to find a good way to "forward" an args tuple like the original did but it doesn't make a big performance different (< 5 ns) and arguably the new behavior is better.

[eendebakpt]

I think you plan to tackle this in a followup PR right?

[cmaloney]

Yes, I'm trying to keep PRs to just one of

1. Add a Argument Clinic new code gen + convert minimal example test types to use it
2. Change existing code gen (which can effect a lot very quickly)
3. Move things to Argument Clinic and/or Vectorcall. For this I plan to make a type-refactor issue then do a large migration once Generate vectorcall code to parse arguments using Argument Clinic #87613 is done replacing the much slower PyArg_ParseTupleAndKeywords broadly + handwritten ones that are flat for performance.

For this particular piece it's a Clinic codegen optimization that we can avoid a new tuple some of the time. In general I think the goto + local variable structure of a lot of the clinic codegen is a lot more complex than needed, have lots of thoughts but separate feature chain.