This is the mail archive of the gcc-patches@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]

Re: CSE fold_rtx subreg patch


> Have your tried this on anything other than the trivial example?  Any
> noticable improvement/regression?  What about your original fold_rtx
> patch?

The assembly resulting from compiling gcc for a x86 with the original
fold_rtx patch (which propagates the SUBREGs) installed is only slightly
different from that resulting from using an unpatched version of gcc.
In a couple of places it's marginally better, in a couple of other
places it's marginally worst.

The assembly resulting from compiling gcc with the second patch (which
doesn't propagate the SUBREGs) installed is quite different in terms of
register names.  Once you get pass the register name differences it seems
to be the same story as above.

> Do you have a preference between them?

No.

> Basically I'm being a little cautious here.  I don't know which scheme
> is better.  The issue of not optimizing SUBREGs as well as REGs may be
> a non-issue in reality.  I just don't know.  So, I'd like a little more
> info on what happens with real code :-)

Having played with this a little more I'm incline to believe that it's
a non-issue.  All things considered I would shelve both patches until
someone comes up with some real code which one of the patches helps.

-- John
-------------------------------------------------------------------------
|   Feith Systems  |   Voice: 1-215-646-8000  |  Email: john@feith.com  |
|    John Wehle    |     Fax: 1-215-540-5495  |                         |
-------------------------------------------------------------------------



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