------- Comment #30 from zadeck at naturalbridge dot com 2008-01-22 13:58 ------- Subject: Re: [4.3 Regression] gfortran.dg/array_constructor_9.f90
Kenneth Zadeck wrote: > zadeck at naturalbridge dot com wrote: > >> ------- Comment #27 from zadeck at naturalbridge dot com 2008-01-21 15:36 >> ------- >> Subject: Re: [4.3 Regression] gfortran.dg/array_constructor_9.f90 >> >> bonzini at gnu dot org wrote: >> >> >>> ------- Comment #26 from bonzini at gnu dot org 2008-01-21 14:54 ------- >>> Subject: Re: [4.3 Regression] gfortran.dg/array_constructor_9.f90 >>> >>> Agreed. I think we'd better revert the patch now. >>> >>> >>> >>> >>> >> i give. >> >> i will prepare a revert this afternoon. i need to go run some errands now. >> >> kenny >> >> >> >> > This patch reverses most of the rd2 patch that speeds up pr3400 and pr26854. > Unfortunately that patch caused some regressions that are very difficult > to resolved. These were reported in pr34884. > > This patch reverses most of rd2.diff. It leaves in the fixes to the > comments and one fix to a latent bug. The patch has been tested on > ia-64, ppc-32, x86-64 and x86-32. > > 2008-01-22 Kenneth Zadeck <[EMAIL PROTECTED]> > > PR rtl-optimization/26854 > PR rtl-optimization/34400 > PR rtl-optimization/34884 > * ddg.c (create_ddg_dep_from_intra_loop_link): Use > DF_RD->gen. > * df.h (df_changeable_flags.DF_RD_NO_TRIM): Deleted > (df_rd_bb_info.expanded_lr_out): Deleted > * loop_invariant.c (find_defs): Deleted DF_RD_NO_TRIM flag. > * loop_iv.c (iv_analysis_loop_init): Ditto. * df-problems.c > (df_rd_free_bb_info, df_rd_alloc, df_rd_confluence_n, > df_rd_bb_local_compute, df_rd_transfer_function, df_rd_free): > Removed code to allocate, initialize or free expanded_lr_out. > (df_rd_bb_local_compute_process_def): Restructured to make more > understandable. > (df_rd_confluence_n): Removed code to no apply invalidate_by_call > sets if the sets are being trimmed. > > > Ok to commit. > committed as revision 131719 kenny -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=34884
