[ruby-core:72333] [Ruby trunk - Feature #11405] [Assigned] [PATCH] hash.c: minor speedups to int/fixnum keys

Joseph Jones liked your message with Boxer. On December 10, 2015 at 03:34:29 MST, [e-mail] wrote:Issue #11405 has been updated by Yusuke Endoh.Status changed from Closed to AssignedAssignee set to Eric WongThis caused a lot of trivial hash conflicts of Fixnums that is >= 16384.~~~~$ ./miniruby -ve 'p 16384.hash; p 16385.hash'ruby 2.3.0dev (2015-12-10 master 52945) [x86_64-linux]11048010433492078001104801043349207800~~~~Other pairs:~~~~p [16386, 16387].map {|n| n.hash }.uniq #=> [2133837449075777600]p [16388, 16389].map {|n| n.hash }.uniq #=> [3903799135928350277]p [16390, 16391].map {|n| n.hash }.uniq #=> [-3836391686716480155]p [16392, 16393].map {|n| n.hash }.uniq #=> [1714559302010572050]p [16394, 16395].map {|n| n.hash }.uniq #=> [2147130354083423794]p [16396, 16397].map {|n| n.hash }.uniq #=> [-679539024000319657]p [16398, 16399].map {|n| n.hash }.uniq #=> [4056286416392832887]p [16400, 16401].map {|n| n.hash }.uniq #=> [2733766810351706956]p [16402, 16403].map {|n| n.hash }.uniq #=> [-1228876631044862612]p [16404, 16405].map {|n| n.hash }.uniq #=> [1418226818996216529]~~~~Unlike the name suggests, rb_objid_hash is also used to calculate a hash value of Fixnum.IMO, we must fix this issue before 2.3.0 release.-- Yusuke Endoh ----------------------------------------Feature #11405: [PATCH] hash.c: minor speedups to int/fixnum keyshttps://bugs.ruby-lang.org/issues/11405#change-55440* Author: Eric Wong* Status: Assigned* Priority: Normal* Assignee: Eric Wong----------------------------------------Noticed with [ruby-core:70159] [Bug #11396]The low bits of Ruby object IDs are rarely populated in the currentimplementation, so ensure the get used.Early versions of this patch redundantly shifted static symbols in`any_hash`, causing regressions with static symbols in `hash_aref_sym`* hash.c (any_hash): skip rb_objid_hash for static syms (rb_num_hash_start): extract from rb_ident_hash (rb_objid_hash): call rb_num_hash_start (rb_ident_hash): ditto~~~target 0: a (ruby 2.3.0dev (2015-07-30 trunk 51437) [x86_64-linux]target 1: b (ruby 2.3.0dev (2015-07-30 patch 51437) [x86_64-linux]benchmark results from Xeon E3-1230 v3 @ 3.30GHz (turbo disabled):minimum results in each 10 measurements.Execution time (sec)name a bhash_aref_dsym 0.316 0.300hash_aref_dsym_long 5.106 5.063hash_aref_fix 0.304 0.297hash_aref_flo 0.061 0.060hash_aref_miss 0.433 0.430hash_aref_str 0.408 0.396hash_aref_sym 0.312 0.306hash_aref_sym_long 0.482 0.469hash_flatten 0.385 0.273hash_ident_flo 0.036 0.037hash_ident_num 0.277 0.276hash_ident_obj 0.291 0.284hash_ident_str 0.289 0.286hash_ident_sym 0.285 0.281hash_keys 0.269 0.271hash_shift 0.020 0.016hash_values 0.264 0.264loop_whileloop2 0.101 0.099vm2_bighash* 3.066 2.972Speedup ratio: compare with the result of `a' (greater is better)name bhash_aref_dsym 1.052hash_aref_dsym_long 1.008hash_aref_fix 1.024hash_aref_flo 1.015hash_aref_miss 1.007hash_aref_str 1.031hash_aref_sym 1.018hash_aref_sym_long 1.027hash_flatten 1.410hash_ident_flo 0.994hash_ident_num 1.001hash_ident_obj 1.022hash_ident_str 1.012hash_ident_sym 1.016hash_keys 0.992hash_shift 1.237hash_values 1.001loop_whileloop2 1.013vm2_bighash* 1.032~~~I will commit in a few weeks if everyone is OK with this.---Files--------------------------------0001-hash.c-improve-integer-fixnum-hashing.patch (5.17 KB)-- https://bugs.ruby-lang.org/