We saw a hard-to-reproduce runtime regression in SPEC Leslie3 that was due the field Prefer_Fuse not being initialized in an instance of the DO_LOOP_INFO class that was constructed with:
DO_LOOP_INFO::DO_LOOP_INFO(DO_LOOP_INFO *dli, MEM_POOL *pool)
I attached a patch to fix this problem.
Also I added:
Comments to caution developers about issue.
Other missing initializations, along with other cosmetic changes so
that one can visually compare the two constructors with tkdiff.
Note that tkdiff can highlight different strings on lines that contain
common strings, however it sometimes common text lines in two files
to synchronize the line-by-line text comparisons (and thus detect
common strings).
Could someone review this change when they have the chance?
Thanks,
Doug
lnopt_main.patch
Description: lnopt_main.patch
------------------------------------------------------------------------------ What Every C/C++ and Fortran developer Should Know! Read this article and learn how Intel has extended the reach of its next-generation tools to help Windows* and Linux* C/C++ and Fortran developers boost performance applications - including clusters. http://p.sf.net/sfu/intel-dev2devmay
_______________________________________________ Open64-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/open64-devel
