Why doesn't JUMP_IF_FALSE do POP_TOP ?

John Machin sjmachin at lexicon.net
Tue Jan 14 13:09:36 CET 2003


Tim Peters <tim.one at comcast.net> wrote in message news:<mailman.1042478427.14500.python-list at python.org>...
.
> 
> As to whether it's "minor" or not, of course it depends on the exact
> programs you're measuring.  There's never been a good reason to believe this
> one isn't minor <wink>.

Have you considered possible flow-on effects? An observation, not a
candidate for a "good reason to believe": POP_TOPs can inhibit
jump-target optimisations e.g. a jump to a POP_TOP, which is followed
by another jump. If the POP_TOP isn't there, you can change the target
of the 1st jump. In some cases you get a bonus; the 2nd jump is then
unreachable and can be blown away.




More information about the Python-list mailing list