On Wireshark with 6GB RAM, I've observed `definitionReachesRank` to be
the slowest predicate in the IR. It seems that the implementation was
slow because the optimizer failed to eliminate the common
`reachesRank - 1` subexpression. This led to context being pushed into
the `not`, which got implemented as `MATERIALIZE`. That wouldn't
normally be a disaster, but this is one of the largest predicates in the
IR SSA construction, and iteration 2 was very slow.
Before:
(1505s) Starting to evaluate predicate SSAConstruction::DefUse::definitionReachesRank#ffff#cur_delta/4[1]@93f592 (iteration 1)
(1535s) Tuple counts for SSAConstruction::DefUse::definitionReachesRank#ffff#cur_delta:
130670697 ~0% {4} r1 = SCAN project#SSAConstruction::DefUse::hasDefinitionAtRank#fffff AS I OUTPUT I.<0>, I.<1>, I.<2>, (I.<2> + 1)
130670697 ~6% {5} r2 = JOIN r1 WITH SSAConstruction::DefUse::exitRank#fff AS R ON FIRST 2 OUTPUT r1.<0>, r1.<1>, r1.<2>, r1.<3>, R.<2>
130670697 ~6% {5} r3 = SELECT r2 ON r2.<3> <= r2.<4>
130670697 ~0% {4} r4 = SCAN r3 OUTPUT r3.<0>, r3.<1>, r3.<2>, r3.<3>
return r4
(1535s) - SSAConstruction::DefUse::definitionReachesRank#ffff_delta has 130670697 rows (order for disjuncts: delta=<standard>).
(1535s) Starting to evaluate predicate SSAConstruction::DefUse::definitionReachesRank#ffff#cur_delta/4[2]@866c14 (iteration 2)
(1626s) Tuple counts for SSAConstruction::DefUse::definitionReachesRank#ffff#cur_delta:
261341394 ~107% {4} r1 = JOIN SSAConstruction::DefUse::definitionReachesRank#ffff#prev_delta AS L WITH SSAConstruction::DefUse::definitionReachesRank#ffff#join_rhs AS R ON FIRST 3 OUTPUT R.<0>, R.<1>, R.<2>, (1 + L.<3>)
261341394 ~107% {4} r2 = r1 AND NOT SSAConstruction::DefUse::definitionReachesRank#ffff#prev AS R(r1.<0>, r1.<1>, r1.<2>, r1.<3>)
130670697 ~0% {5} r3 = SCAN r2 OUTPUT r2.<0>, r2.<1>, (r2.<3> - 1), r2.<2>, r2.<3>
106034590 ~1% {4} r4 = JOIN r3 WITH project#SSAConstruction::DefUse::hasDefinitionAtRank#fffff AS R ON FIRST 3 OUTPUT r3.<0>, r3.<1>, r3.<3>, r3.<4>
106034590 {4} r5 = MATERIALIZE r4 AS antijoin_rhs
24636107 ~3% {4} r6 = r2 AND NOT r5(r2.<0>, r2.<1>, r2.<2>, r2.<3>)
24636107 ~0% {5} r7 = JOIN r6 WITH SSAConstruction::DefUse::exitRank#fff AS R ON FIRST 2 OUTPUT r6.<0>, r6.<1>, r6.<2>, r6.<3>, R.<2>
2749441 ~0% {5} r8 = SELECT r7 ON r7.<3> <= r7.<4>
2749441 ~4% {4} r9 = SCAN r8 OUTPUT r8.<0>, r8.<1>, r8.<2>, r8.<3>
return r9
(1626s) - SSAConstruction::DefUse::definitionReachesRank#ffff_delta has 2749441 rows (order for disjuncts: delta=<standard>).
After:
(12s) Tuple counts for SSAConstruction::DefUse::definitionReachesRank#ffff#cur_delta:
130670697 ~0% {4} r1 = SCAN project#SSAConstruction::DefUse::hasDefinitionAtRank#fffff AS I OUTPUT I.<0>, I.<1>, I.<2>, (I.<2> + 1)
return r1
(12s) - SSAConstruction::DefUse::definitionReachesRank#ffff_delta has 130670697 rows (order for disjuncts: delta=<standard>).
(12s) Starting to evaluate predicate SSAConstruction::DefUse::definitionReachesRank#ffff#cur_delta/4[2]@fff64c (iteration 2)
(34s) Tuple counts for SSAConstruction::DefUse::definitionReachesRank#ffff#cur_delta:
108784031 ~0% {4} r1 = SSAConstruction::DefUse::definitionReachesRank#ffff#prev_delta AS L AND NOT SSAConstruction::DefUse::exitRank#fff AS R(L.<0>, L.<1>, L.<3>)
2749441 ~5% {4} r2 = r1 AND NOT project#SSAConstruction::DefUse::hasDefinitionAtRank#fffff AS R(r1.<0>, r1.<1>, r1.<3>)
2749441 ~4% {4} r3 = SCAN r2 OUTPUT r2.<0>, r2.<1>, r2.<2>, (r2.<3> + 1)
2749441 ~4% {4} r4 = r3 AND NOT SSAConstruction::DefUse::definitionReachesRank#ffff#prev AS R(r3.<0>, r3.<1>, r3.<2>, r3.<3>)
return r4
(34s) - SSAConstruction::DefUse::definitionReachesRank#ffff_delta has 2749441 rows (order for disjuncts: delta=<standard>).
Note that the row counts are exactly the same before and after.
The `cpp/missing-case-in-switch` performed badly on some snapshots, to
the extent where it was as slow as the most expensive IR stages
(example: ChakraCore). This commit makes it faster, removing a
`pragma[noopt]` along the way.
The intermediate tuple counts on a customer codebase drop from 84M to
3M, while the content hash of `getAMissingCase` is the same.
Before:
(124s) Tuple counts for Stmt::EnumSwitch::getAMissingCase#ff#antijoin_rhs:
20867789 ~0% {3} r1 = JOIN Stmt::SwitchStmt::getASwitchCase_dispred#ff AS L WITH Stmt::EnumSwitch::getAMissingCase#ff#shared AS R ON FIRST 1 OUTPUT L.<1>, R.<0>, R.<1>
20122830 ~0% {3} r2 = JOIN r1 WITH Stmt::SwitchCase::getExpr_dispred#ff AS R ON FIRST 1 OUTPUT R.<1>, r1.<1>, r1.<2>
20122830 ~0% {3} r3 = JOIN r2 WITH Expr::Expr::getValue_dispred#ff AS R ON FIRST 1 OUTPUT r2.<2>, r2.<1>, R.<1>
83961918 ~0% {4} r4 = JOIN r3 WITH Enum::EnumConstant::getInitializer_dispred#ff AS R ON FIRST 1 OUTPUT R.<1>, r3.<1>, r3.<0>, r3.<2>
83961918 ~0% {4} r5 = JOIN r4 WITH initialisers AS R ON FIRST 1 OUTPUT R.<2>, r4.<3>, r4.<1>, r4.<2>
234348 ~185% {2} r6 = JOIN r5 WITH Expr::Expr::getValue_dispred#ff AS R ON FIRST 2 OUTPUT r5.<2>, r5.<3>
return r6
...
(124s) Tuple counts for Stmt::EnumSwitch::getAMissingCase#ff:
663127 ~4% {2} r1 = Stmt::EnumSwitch::getAMissingCase#ff#shared AS L AND NOT Stmt::EnumSwitch::getAMissingCase#ff#antijoin_rhs AS R(L.<0>, L.<1>)
return r1
(124s) Registering Stmt::EnumSwitch::getAMissingCase#ff + [] with content 2060ff326cvhihcsvoph6k9divuv4
(124s) >>> Wrote relation Stmt::EnumSwitch::getAMissingCase#ff with 663127 rows and 2 columns.
After:
(5s) Tuple counts for Stmt::EnumSwitch::getAMissingCase_dispred#ff#antijoin_rhs:
746029 ~0% {2} r1 = JOIN Stmt::EnumSwitch::getAMissingCase_dispred#ff#shared AS L WITH Enum::Enum::getAnEnumConstant_dispred#ff AS R ON FIRST 1 OUTPUT R.<1>, L.<1>
3116197 ~2% {3} r2 = JOIN r1 WITH Enum::EnumConstant::getInitializer_dispred#ff AS R ON FIRST 1 OUTPUT R.<1>, r1.<1>, r1.<0>
3116197 ~0% {3} r3 = JOIN r2 WITH initialisers AS R ON FIRST 1 OUTPUT R.<2>, r2.<1>, r2.<2>
3116197 ~311% {3} r4 = JOIN r3 WITH Expr::Expr::getValue_dispred#ff AS R ON FIRST 1 OUTPUT r3.<1>, R.<1>, r3.<2>
234348 ~185% {2} r5 = JOIN r4 WITH Stmt::EnumSwitch::matchesValue#ff AS R ON FIRST 2 OUTPUT r4.<0>, r4.<2>
return r5
(5s) Registering Stmt::EnumSwitch::getAMissingCase_dispred#ff#antijoin_rhs + [] with content 173483d71508vl534mvlr1g0ehi12
(5s) >>> Wrote relation Stmt::EnumSwitch::getAMissingCase_dispred#ff#antijoin_rhs with 82902 rows and 2 columns.
(5s) Starting to evaluate predicate Stmt::EnumSwitch::getAMissingCase_dispred#ff/2@ae4c0b
(5s) Tuple counts for Stmt::EnumSwitch::getAMissingCase_dispred#ff:
746029 ~2% {2} r1 = JOIN Stmt::EnumSwitch::getAMissingCase_dispred#ff#shared AS L WITH Enum::Enum::getAnEnumConstant_dispred#ff AS R ON FIRST 1 OUTPUT L.<1>, R.<1>
663127 ~4% {2} r2 = r1 AND NOT Stmt::EnumSwitch::getAMissingCase_dispred#ff#antijoin_rhs AS R(r1.<0>, r1.<1>)
return r2
(5s) Registering Stmt::EnumSwitch::getAMissingCase_dispred#ff + [] with content 2060ff326cvhihcsvoph6k9divuv4
(5s) >>> Wrote relation Stmt::EnumSwitch::getAMissingCase_dispred#ff with 663127 rows and 2 columns.
