➜

This issue tracker has been migrated to GitHub, and is currently read-only.
For more information, see the GitHub FAQs in the Python's Developer Guide.

classification
Title: opcode prediction distorts dynamic execution profile stats
Type: Stage:
Components: Interpreter Core Versions:
process
Status: closed Resolution: fixed
Dependencies: Superseder:
Assigned To: rhettinger Nosy List: aimacintyre, rhettinger
Priority: normal Keywords:

Created on 2004-01-25 06:34 by aimacintyre, last changed 2022-04-11 14:56 by admin. This issue is now closed.

Messages (3)
msg19816 - (view) Author: Andrew I MacIntyre (aimacintyre) * (Python triager) Date: 2004-01-25 06:34 
By defining DYNAMIC_EXECUTION_PROFILE (and DXPAIRS if
you want opcode pair counts), opcode frequencies can be
evaluated for various program runs.

However successful opcode predictions bypass the
counter updates, which happen at the top of the loop.

So the result can be significantly under-represented
opcodes in the total statistics.
msg19817 - (view) Author: Andrew I MacIntyre (aimacintyre) * (Python triager) Date: 2004-01-25 06:47 
Logged In: YES 
user_id=250749

See patch #884022 for a possible fix/work around.
msg19818 - (view) Author: Raymond Hettinger (rhettinger) * (Python committer) Date: 2004-02-08 20:06 
Logged In: YES 
user_id=80475

Applied patch 884022.
History
Date User Action Args
2022-04-11 14:56:02adminsetgithub: 39864
2004-01-25 06:34:00aimacintyrecreate