public inbox for gcc-bugs@sourceware.org help / color / mirror / Atom feed
From: "rguenth at gcc dot gnu.org" <gcc-bugzilla@gcc.gnu.org> To: gcc-bugs@gcc.gnu.org Subject: [Bug tree-optimization/102943] [12 Regression] Jump threader compile-time hog with 521.wrf_r Date: Thu, 10 Mar 2022 11:37:25 +0000 [thread overview] Message-ID: <bug-102943-4-0eHMbwJYQM@http.gcc.gnu.org/bugzilla/> (raw) In-Reply-To: <bug-102943-4@http.gcc.gnu.org/bugzilla/> https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102943 Richard Biener <rguenth at gcc dot gnu.org> changed: What |Removed |Added ---------------------------------------------------------------------------- Last reconfirmed|2022-01-18 00:00:00 |2022-3-10 --- Comment #35 from Richard Biener <rguenth at gcc dot gnu.org> --- So I've re-measured -Ofast -march=znver2 -flto on todays trunk with release checking (built with GCC 7, not bootstrapped) and the largest LTRANS (ltrans22 at the moment) unit still has tree VRP : 15.52 ( 20%) 0.03 ( 5%) 15.57 ( 20%) 28M ( 4%) backwards jump threading : 16.17 ( 21%) 0.00 ( 0%) 16.15 ( 21%) 1475k ( 0%) TOTAL : 77.29 0.59 77.92 744M and the 2nd largest (ltrans86 at the moment) alias stmt walking : 7.70 ( 16%) 0.03 ( 8%) 7.70 ( 16%) 703k ( 0%) tree VRP : 8.25 ( 18%) 0.01 ( 3%) 8.27 ( 17%) 14M ( 3%) backwards jump threading : 8.79 ( 19%) 0.00 ( 0%) 8.82 ( 19%) 1645k ( 0%) TOTAL : 46.97 0.38 47.38 438M so it's still by far jump-threading/VRP dominating compile-times (I wonder if we should separate "old" and "new" [E]VRP timevars). Given that VRP shows up as well it's more likely the underlying ranger infrastructure? perf thrown on ltrans22 shows Samples: 302K of event 'cycles', Event count (approx.): 331301505627 Overhead Samples Command Shared Object Symbol 10.34% 31299 lto1-ltrans lto1 [.] bitmap_get_aligned_chunk 7.44% 22540 lto1-ltrans lto1 [.] bitmap_bit_p 3.17% 9593 lto1-ltrans lto1 [.] get_immediate_dominator 2.87% 8668 lto1-ltrans lto1 [.] determine_value_range 2.36% 7143 lto1-ltrans lto1 [.] ranger_cache::propagate_cache 2.32% 7031 lto1-ltrans lto1 [.] bitmap_set_bit 2.20% 6664 lto1-ltrans lto1 [.] operand_compare::operand_equal_p 1.88% 5692 lto1-ltrans lto1 [.] bitmap_set_aligned_chunk 1.79% 5390 lto1-ltrans lto1 [.] number_of_iterations_exit_assumptions 1.66% 5048 lto1-ltrans lto1 [.] get_continuation_for_phi callgraph info in perf is a mixed bag, but maybe it helps to pinpoint things: - 10.20% 10.18% 30364 lto1-ltrans lto1 [.] bitmap_get_aligned_chunk # - 10.18% 0xffffffffffffffff # + 9.16% ranger_cache::propagate_cache # + 1.01% ranger_cache::fill_block_cache - 7.84% 7.83% 23509 lto1-ltrans lto1 [.] bitmap_bit_p # - 6.20% 0xffffffffffffffff # + 1.85% fold_using_range::range_of_range_op # + 1.64% ranger_cache::range_on_edge # + 1.29% gimple_ranger::range_of_expr and the most prominent get_immediate_dominator calls are from back_propagate_equivalences which does FOR_EACH_IMM_USE_FAST (use_p, iter, lhs) ... /* Profiling has shown the domination tests here can be fairly expensive. We get significant improvements by building the set of blocks that dominate BB. We can then just test for set membership below. We also initialize the set lazily since often the only uses are going to be in the same block as DEST. */ if (!domby) { domby = BITMAP_ALLOC (NULL); basic_block bb = get_immediate_dominator (CDI_DOMINATORS, dest); while (bb) { bitmap_set_bit (domby, bb->index); bb = get_immediate_dominator (CDI_DOMINATORS, bb); } } /* This tests if USE_STMT does not dominate DEST. */ if (!bitmap_bit_p (domby, gimple_bb (use_stmt)->index)) continue; I think that "optimization" is flawed - a dominance check is cheap if the DFS numbers are up-to-date: bool dominated_by_p (enum cdi_direction dir, const_basic_block bb1, const_basic_block bb2) { unsigned int dir_index = dom_convert_dir_to_idx (dir); struct et_node *n1 = bb1->dom[dir_index], *n2 = bb2->dom[dir_index]; gcc_checking_assert (dom_computed[dir_index]); if (dom_computed[dir_index] == DOM_OK) return (n1->dfs_num_in >= n2->dfs_num_in && n1->dfs_num_out <= n2->dfs_num_out); return et_below (n1, n2); } it's just the fallback that is not. Also recoding _all_ dominators of 'dest' is expensive for a large CFG but you'll only ever need dominators up to the definition of 'lhs' which we know will dominate all use_stmt so if that does _not_ dominate e->dest no use will (but I think that's always the case in the current code). Note the caller iterates over simple equivalences on an edge so this bitmap is populated multiple times (but if we cache it we cannot prune from the top). For FP we have usually multiple equivalences so caching pays off more than pruning for WRF. Note this is only a minor part of the slowness, I'm testing a patch for this part. Note for WRF always going the "slow" dominated_by_p way is as fast as caching.
next prev parent reply other threads:[~2022-03-10 11:37 UTC|newest] Thread overview: 57+ messages / expand[flat|nested] mbox.gz Atom feed top 2021-10-26 11:13 [Bug tree-optimization/102943] New: VRP " rguenth at gcc dot gnu.org 2021-10-26 11:15 ` [Bug tree-optimization/102943] [12 Regression] " rguenth at gcc dot gnu.org 2021-10-26 11:25 ` rguenth at gcc dot gnu.org 2021-10-26 11:49 ` rguenth at gcc dot gnu.org 2021-10-26 14:57 ` pinskia at gcc dot gnu.org 2021-10-26 14:58 ` marxin at gcc dot gnu.org 2021-10-26 15:06 ` marxin at gcc dot gnu.org 2021-10-30 6:31 ` aldyh at gcc dot gnu.org 2021-10-31 20:06 ` hubicka at gcc dot gnu.org 2021-11-02 7:25 ` [Bug tree-optimization/102943] [12 Regression] Jump " rguenth at gcc dot gnu.org 2021-11-02 7:29 ` aldyh at gcc dot gnu.org 2021-11-03 10:57 ` aldyh at gcc dot gnu.org 2021-11-03 10:58 ` aldyh at gcc dot gnu.org 2021-11-03 13:17 ` rguenther at suse dot de 2021-11-03 14:33 ` amacleod at redhat dot com 2021-11-03 14:42 ` rguenther at suse dot de 2021-11-04 14:40 ` cvs-commit at gcc dot gnu.org 2021-11-04 14:40 ` cvs-commit at gcc dot gnu.org 2021-11-04 14:40 ` cvs-commit at gcc dot gnu.org 2021-11-04 15:24 ` aldyh at gcc dot gnu.org 2021-11-04 17:00 ` Jan Hubicka 2021-11-04 17:00 ` hubicka at kam dot mff.cuni.cz 2021-11-05 9:08 ` aldyh at gcc dot gnu.org 2021-11-05 11:10 ` marxin at gcc dot gnu.org 2021-11-05 11:13 ` aldyh at gcc dot gnu.org 2021-11-05 11:23 ` marxin at gcc dot gnu.org 2021-11-05 17:16 ` cvs-commit at gcc dot gnu.org 2021-11-07 17:17 ` hubicka at gcc dot gnu.org 2021-11-07 18:16 ` aldyh at gcc dot gnu.org 2021-11-07 18:59 ` Jan Hubicka 2021-11-07 18:59 ` hubicka at kam dot mff.cuni.cz 2021-11-12 22:14 ` hubicka at gcc dot gnu.org 2021-11-14 9:58 ` hubicka at gcc dot gnu.org 2021-11-26 12:38 ` cvs-commit at gcc dot gnu.org 2021-11-30 10:55 ` aldyh at gcc dot gnu.org 2021-12-09 20:17 ` hubicka at gcc dot gnu.org 2022-01-03 8:47 ` rguenth at gcc dot gnu.org 2022-01-03 11:20 ` hubicka at kam dot mff.cuni.cz 2022-01-19 7:06 ` rguenth at gcc dot gnu.org 2022-03-10 11:37 ` rguenth at gcc dot gnu.org [this message] 2022-03-10 12:40 ` cvs-commit at gcc dot gnu.org 2022-03-10 13:22 ` rguenth at gcc dot gnu.org 2022-03-10 13:42 ` cvs-commit at gcc dot gnu.org 2022-03-10 13:45 ` rguenth at gcc dot gnu.org 2022-03-10 13:49 ` rguenth at gcc dot gnu.org 2022-03-10 14:01 ` amacleod at redhat dot com 2022-03-10 14:17 ` amacleod at redhat dot com 2022-03-10 14:23 ` rguenth at gcc dot gnu.org 2022-03-10 14:26 ` rguenth at gcc dot gnu.org 2022-03-10 14:33 ` amacleod at redhat dot com 2022-03-10 14:36 ` amacleod at redhat dot com 2022-03-16 19:48 ` amacleod at redhat dot com 2022-03-17 11:14 ` rguenth at gcc dot gnu.org 2022-03-17 13:05 ` amacleod at redhat dot com 2022-03-17 14:18 ` hubicka at kam dot mff.cuni.cz 2022-03-17 20:44 ` cvs-commit at gcc dot gnu.org 2022-03-23 10:40 ` rguenth at gcc dot gnu.org
Reply instructions: You may reply publicly to this message via plain-text email using any one of the following methods: * Save the following mbox file, import it into your mail client, and reply-to-all from there: mbox Avoid top-posting and favor interleaved quoting: https://en.wikipedia.org/wiki/Posting_style#Interleaved_style * Reply using the --to, --cc, and --in-reply-to switches of git-send-email(1): git send-email \ --in-reply-to=bug-102943-4-0eHMbwJYQM@http.gcc.gnu.org/bugzilla/ \ --to=gcc-bugzilla@gcc.gnu.org \ --cc=gcc-bugs@gcc.gnu.org \ /path/to/YOUR_REPLY https://kernel.org/pub/software/scm/git/docs/git-send-email.html * If your mail client supports setting the In-Reply-To header via mailto: links, try the mailto: linkBe sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox; as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).