This is the mail archive of the gcc@gcc.gnu.org mailing list for the GCC project.

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Oopack.C: mainline vs tree-ssa

Hi everyone!

this is on my P4-2400, -O2
(Max=500000 Matrix=5000 Complex=200000 Iterator=500000):

mainline
--------
                        Seconds       Mflops
Test       Iterations     C    OOP     C    OOP  Ratio
----       ----------  -----------  -----------  -----
Max            500000    1.5   1.5  331.1 337.8    1.0
Matrix           5000    1.4   1.5  862.1 856.2    1.0
Complex        200000    1.7  29.6  946.7  54.0   17.5  <<<
Iterator       500000    1.1   1.2  925.9 854.7    1.1

tree-ssa
--------
                        Seconds       Mflops
Test       Iterations     C    OOP     C    OOP  Ratio
----       ----------  -----------  -----------  -----
Max            500000    1.5   1.5  340.1 342.5    1.0
Matrix           5000    1.5   2.2  856.2 578.7    1.5
Complex        200000    1.7   1.4  941.2 1185.2   0.8  <<<
Iterator       500000    1.1   1.1  900.9 877.2    1.0


So, great news for the the Complex test!!!

Looks like there is a minor regression for Matrix, tough...

I'm attaching the whole testcase to this message, in
case someone is interested.

Paolo.

Attachment: oopack_v1p8.C.gz
Description: GNU Zip compressed data

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]