RFR(L): 8186027: C2: loop strip mining
nils.eliasson at oracle.com
Mon Oct 23 14:16:35 UTC 2017
Sorry for the delay.
First - It's a very impressive work you have done!
Currently your patch doesn't apply cleanly. The fix of JDK-8189067
I have run your code (based on jdk10 before JDK-8189067) through
testing. I encountered a minor build problem on solaris_x64 (patch
below), otherwise it was stable with no encountered test failures.
I have also run performance testing with the conclusion that no
significant regression can be seen. In some benchmarks like
scimark.sparse.large that has a known safepointing issue
(https://bugs.openjdk.java.net/browse/JDK-8177704), very good results
can be seen.
scimark.sparse.large using G1:
-XX:-UseCountedLoopSafepoints (default) ~86 ops/m
-XX:+UseCountedLoopSafepoints ~106 ops/m
-XX:+UseCountedLoopSafepoints -XX:LoopStripMiningIter=1000 ~111 ops/m
The positive results leads us to the conclusion that we would like
UseCountedLoopSafepoints to bedefault true, and LoopStripMiningIter
default to 1000.
- product(bool, UseCountedLoopSafepoints, false,
+ product(bool, UseCountedLoopSafepoints, true,
- product(uintx, LoopStripMiningIter, 0,
+ product(uintx, LoopStripMiningIter, 1000,
solaris_x64 complained about type conversion:
@@ -1729,7 +1729,7 @@
Node* l = cl->outer_loop();
Node* tail = cl->outer_loop_tail();
IfNode* le = cl->outer_loop_end();
- Node* sfpt = cl->outer_safepoint();
+ Node* sfpt = (Node*) cl->outer_safepoint();
@@ -144,7 +144,7 @@
assert(iter_estimate > 0, "broken");
if ((jlong)scaled_iters != scaled_iters_long || iter_estimate <=
// Remove outer loop and safepoint (too few iterations)
- Node* outer_sfpt = inner_cl->outer_safepoint();
+ Node* outer_sfpt = (Node*) inner_cl->outer_safepoint();
In the TraceLoopOpts print out I suggest changing space to underscore to
conform with how the other print outs look:
"PreMainPost Loop: N153/N130 limit_check predicated counted [0,int),+1
(26 iters) has_sfpt strip mined"
- tty->print(" strip mined");
+ tty->print(" strip_mined");
When your patch is updated, I will do some additional functional
testing. Also, a second reviewer is required.
On 2017-10-11 15:53, Roland Westrelin wrote:
>> I have started reviewing and testing I will sponsor your change when the
>> full review is completed.
More information about the hotspot-compiler-dev