Bugs item #2596114, was opened at 2009-02-13 15:32
Message generated for change (Comment added) made by stmane
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=482468&aid=2596114&group_id=56967

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: SQL/Core
Group: SQL CVS Head
>Status: Closed
Resolution: Fixed
Priority: 5
Private: No
Submitted By: Jens Oehlschlägel (jens_oehl)
Assigned to: Stefan Manegold (stmane)
Summary: POWER(x,2) by factor 100 slower than x*x;

Initial Comment:
POWER does not recognize tuning opportunities with certain constant integer 
exponents.

Oberved under Win32 Feb2009Alpha

Replication and demonstration of speed potential attached.

----------------------------------------------------------------------

>Comment By: Stefan Manegold (stmane)
Date: 2009-04-02 23:06

Message:
added test in
sql/src/test/BugTracker-2009/Tests/POWER_vs_prod.SF-2596114.stable.*


----------------------------------------------------------------------

Comment By: Stefan Manegold (stmane)
Date: 2009-02-14 10:46

Message:
re-opened to remind me that I want to add a test after all ...


----------------------------------------------------------------------

Comment By: Martin Kersten (mlkersten)
Date: 2009-02-14 06:22

Message:
The naming convention for the multiplex variants [bat]mmath
where not obeyed. This made the remap optimizer not aware
of the existence of fast math function implementations.

no test needed.

----------------------------------------------------------------------

Comment By: Martin Kersten (mlkersten)
Date: 2009-02-13 17:14

Message:
Indeed, power(x,2) is  not replaced by x *x, which is btw often down within
the function itself
and you only safe the functionc all overhead.
The real issue is that power(:bat[:oid,:flt],:flt):bat[:oid,:dbl]
is missing in batmath.mx
Fix pending.

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=482468&aid=2596114&group_id=56967

------------------------------------------------------------------------------
_______________________________________________
Monetdb-bugs mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/monetdb-bugs

Reply via email to