[Bug tree-optimization/71347] [7 regression] Performance drop after r235513 on x86-64 in 32-bit mode.
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=71347 amker at gcc dot gnu.org changed: What|Removed |Added Status|UNCONFIRMED |RESOLVED Resolution|--- |FIXED --- Comment #7 from amker at gcc dot gnu.org --- Fixed.
[Bug tree-optimization/71347] [7 regression] Performance drop after r235513 on x86-64 in 32-bit mode.
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=71347 --- Comment #6 from amker at gcc dot gnu.org --- Author: amker Date: Fri Jul 15 08:53:48 2016 New Revision: 238366 URL: https://gcc.gnu.org/viewcvs?rev=238366&root=gcc&view=rev Log: gcc/testsuite PR tree-optimization/71347 * gcc.dg/tree-ssa/pr71347.c: XFAIL on ia64, arm, m68k and sparc. Modified: trunk/gcc/testsuite/ChangeLog trunk/gcc/testsuite/gcc.dg/tree-ssa/pr71347.c
[Bug tree-optimization/71347] [7 regression] Performance drop after r235513 on x86-64 in 32-bit mode.
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=71347 --- Comment #5 from amker at gcc dot gnu.org --- (In reply to Rainer Orth from comment #4) > Created attachment 38744 [details] > pr71347.c.214t.optimized > > The new testcase FAILs on sparc*-sun-solaris2.* (both 32 and 64-bit): > > FAIL: gcc.dg/tree-ssa/pr71347.c scan-tree-dump-not optimized ".* = MEM.*;" > > Dump attached. > > Rainer Thanks for reporting. I am going to disable this on various targets according to https://gcc.gnu.org/ml/gcc-patches/2016-06/msg01377.html
[Bug tree-optimization/71347] [7 regression] Performance drop after r235513 on x86-64 in 32-bit mode.
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=71347 Rainer Orth changed: What|Removed |Added CC||ro at gcc dot gnu.org --- Comment #4 from Rainer Orth --- Created attachment 38744 --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=38744&action=edit pr71347.c.214t.optimized The new testcase FAILs on sparc*-sun-solaris2.* (both 32 and 64-bit): FAIL: gcc.dg/tree-ssa/pr71347.c scan-tree-dump-not optimized ".* = MEM.*;" Dump attached. Rainer
[Bug tree-optimization/71347] [7 regression] Performance drop after r235513 on x86-64 in 32-bit mode.
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=71347 --- Comment #3 from amker at gcc dot gnu.org --- Author: amker Date: Fri Jun 17 09:26:05 2016 New Revision: 237552 URL: https://gcc.gnu.org/viewcvs?rev=237552&root=gcc&view=rev Log: PR tree-optimization/71347 * tree-ssa-loop-ivopts.c (determine_group_iv_cost_address): Compute cost for all uses in group. PR tree-optimization/71347 * gcc.dg/tree-ssa/pr71347.c: New test. Added: trunk/gcc/testsuite/gcc.dg/tree-ssa/pr71347.c Modified: trunk/gcc/ChangeLog trunk/gcc/testsuite/ChangeLog trunk/gcc/tree-ssa-loop-ivopts.c
[Bug tree-optimization/71347] [7 regression] Performance drop after r235513 on x86-64 in 32-bit mode.
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=71347
--- Comment #2 from amker at gcc dot gnu.org ---
Thanks for reporting this.
The dump after IVOPT now is:
:
# prephitmp_21 = PHI
# prephitmp_23 = PHI
# ivtmp.17_16 = PHI
_6 = prephitmp_21 * prephitmp_23;
_4 = (void *) ivtmp.17_16;
MEM[base: _4, offset: 0B] = _6;
ivtmp.17_9 = ivtmp.17_16 + 8;
if (ivtmp.17_9 != _26)
goto ;
else
goto ;
:
_5 = (void *) ivtmp.17_9;
pretmp_20 = MEM[base: _5, offset: 4294967288B];
pretmp_22 = X[1];
goto ;
That patch skips computing cost for sub iv_uses in a group:
Group 0:
Type: ADDRESS
Use 0.0:
At stmt:X[i_18] = _6;
At pos: X[i_18]
IV struct:
Type: double *
Base: (double *) (&X + 16)
Step: 8
Object: (void *) &X
Biv: N
Overflowness wrto loop niter: Overflow
Use 0.1:
At stmt:pretmp_20 = X[_15];
At pos: X[_15]
IV struct:
Type: double *
Base: (double *) (&X + 16)
Step: 8
Object: (void *) &X
Biv: N
Overflowness wrto loop niter: Overflow
Though use 0.0/0.1 have same {base, step}, but there are in different program
point, so if iv_cand is increased before use 0.0, the first use is transformed
into: MEM[var_before, 0], and use 0.1 will be transformed into: MEM[var_after,
-8]. Now the two memory reference have different expressions, though access to
the same object. Afterwards, DOM failed to CSE them.
It is hard to decide which sub iv_use we should compute the cost, and which one
we can skip computing the cost. Maybe I should revert this part of code which
was introduced to save a small amount of compilation time.
[Bug tree-optimization/71347] [7 regression] Performance drop after r235513 on x86-64 in 32-bit mode.
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=71347 Richard Biener changed: What|Removed |Added CC||amker.cheng at gmail dot com Target Milestone|--- |7.0
[Bug tree-optimization/71347] [7 regression] Performance drop after r235513 on x86-64 in 32-bit mode.
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=71347 --- Comment #1 from Yuri Rumyantsev --- Created attachment 38600 --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=38600&action=edit test-case to reproduce Need to be compiled with -O2 -m32 -march=slm -ffast-math options on x64-64.
