1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025
|
//! Test cases for fixpoint iteration cycle resolution.
//!
//! These test cases use a generic query setup that allows constructing arbitrary dependency
//! graphs, and attempts to achieve good coverage of various cases.
mod common;
use common::{ExecuteValidateLoggerDatabase, LogDatabase};
use expect_test::expect;
use salsa::{CycleRecoveryAction, Database as Db, DatabaseImpl as DbImpl, Durability, Setter};
#[cfg(not(miri))]
use test_log::test;
#[derive(Clone, Copy, Debug, PartialEq, Eq, salsa::Update)]
enum Value {
N(u8),
OutOfBounds,
TooManyIterations,
}
impl Value {
fn to_value(self) -> Option<u8> {
if let Self::N(val) = self {
Some(val)
} else {
None
}
}
}
/// A vector of inputs a query can evaluate to get an iterator of values to operate on.
///
/// This allows creating arbitrary query graphs between the four queries below (`min_iterate`,
/// `max_iterate`, `min_panic`, `max_panic`) for testing cycle behaviors.
#[salsa::input]
struct Inputs {
#[returns(ref)]
inputs: Vec<Input>,
}
impl Inputs {
fn values(self, db: &dyn Db) -> impl Iterator<Item = Value> + use<'_> {
self.inputs(db).iter().map(|input| input.eval(db))
}
}
/// A single input, evaluating to a single [`Value`].
#[derive(Clone)]
enum Input {
/// a simple value
Value(Value),
/// a simple value, reported as an untracked read
UntrackedRead(Value),
/// minimum of the given inputs, with fixpoint iteration on cycles
MinIterate(Inputs),
/// maximum of the given inputs, with fixpoint iteration on cycles
MaxIterate(Inputs),
/// minimum of the given inputs, panicking on cycles
MinPanic(Inputs),
/// maximum of the given inputs, panicking on cycles
MaxPanic(Inputs),
/// value of the given input, plus one; propagates error values
Successor(Box<Input>),
/// successor, converts error values to zero
SuccessorOrZero(Box<Input>),
}
impl Input {
fn eval(&self, db: &dyn Db) -> Value {
match *self {
Self::Value(value) => value,
Self::UntrackedRead(value) => {
db.report_untracked_read();
value
}
Self::MinIterate(inputs) => min_iterate(db, inputs),
Self::MaxIterate(inputs) => max_iterate(db, inputs),
Self::MinPanic(inputs) => min_panic(db, inputs),
Self::MaxPanic(inputs) => max_panic(db, inputs),
Self::Successor(ref input) => match input.eval(db) {
Value::N(num) => Value::N(num + 1),
other => other,
},
Self::SuccessorOrZero(ref input) => match input.eval(db) {
Value::N(num) => Value::N(num + 1),
_ => Value::N(0),
},
}
}
fn assert(&self, db: &dyn Db, expected: Value) {
assert_eq!(self.eval(db), expected)
}
fn assert_value(&self, db: &dyn Db, expected: u8) {
self.assert(db, Value::N(expected))
}
fn assert_bounds(&self, db: &dyn Db) {
self.assert(db, Value::OutOfBounds)
}
fn assert_count(&self, db: &dyn Db) {
self.assert(db, Value::TooManyIterations)
}
}
const MIN_VALUE: u8 = 10;
const MAX_VALUE: u8 = 245;
const MAX_ITERATIONS: u32 = 3;
/// Recover from a cycle by falling back to `Value::OutOfBounds` if the value is out of bounds,
/// `Value::TooManyIterations` if we've iterated more than `MAX_ITERATIONS` times, or else
/// iterating again.
fn cycle_recover(
_db: &dyn Db,
value: &Value,
count: u32,
_inputs: Inputs,
) -> CycleRecoveryAction<Value> {
if value
.to_value()
.is_some_and(|val| val <= MIN_VALUE || val >= MAX_VALUE)
{
CycleRecoveryAction::Fallback(Value::OutOfBounds)
} else if count > MAX_ITERATIONS {
CycleRecoveryAction::Fallback(Value::TooManyIterations)
} else {
CycleRecoveryAction::Iterate
}
}
/// Fold an iterator of `Value` into a `Value`, given some binary operator to apply to two `u8`.
/// `Value::TooManyIterations` and `Value::OutOfBounds` will always propagate, with
/// `Value::TooManyIterations` taking precedence.
fn fold_values<F>(values: impl IntoIterator<Item = Value>, op: F) -> Value
where
F: Fn(u8, u8) -> u8,
{
values
.into_iter()
.fold(None, |accum, elem| {
let Some(accum) = accum else {
return Some(elem);
};
match (accum, elem) {
(Value::TooManyIterations, _) | (_, Value::TooManyIterations) => {
Some(Value::TooManyIterations)
}
(Value::OutOfBounds, _) | (_, Value::OutOfBounds) => Some(Value::OutOfBounds),
(Value::N(val1), Value::N(val2)) => Some(Value::N(op(val1, val2))),
}
})
.expect("inputs should not be empty")
}
/// Query minimum value of inputs, with cycle recovery.
#[salsa::tracked(cycle_fn=cycle_recover, cycle_initial=min_initial)]
fn min_iterate<'db>(db: &'db dyn Db, inputs: Inputs) -> Value {
fold_values(inputs.values(db), u8::min)
}
fn min_initial(_db: &dyn Db, _inputs: Inputs) -> Value {
Value::N(255)
}
/// Query maximum value of inputs, with cycle recovery.
#[salsa::tracked(cycle_fn=cycle_recover, cycle_initial=max_initial)]
fn max_iterate<'db>(db: &'db dyn Db, inputs: Inputs) -> Value {
fold_values(inputs.values(db), u8::max)
}
fn max_initial(_db: &dyn Db, _inputs: Inputs) -> Value {
Value::N(0)
}
/// Query minimum value of inputs, without cycle recovery.
#[salsa::tracked]
fn min_panic<'db>(db: &'db dyn Db, inputs: Inputs) -> Value {
fold_values(inputs.values(db), u8::min)
}
/// Query maximum value of inputs, without cycle recovery.
#[salsa::tracked]
fn max_panic<'db>(db: &'db dyn Db, inputs: Inputs) -> Value {
fold_values(inputs.values(db), u8::max)
}
fn untracked(num: u8) -> Input {
Input::UntrackedRead(Value::N(num))
}
fn value(num: u8) -> Input {
Input::Value(Value::N(num))
}
// Diagram nomenclature for nodes: Each node is represented as a:xx(ii), where `a` is a sequential
// identifier from `a`, `b`, `c`..., xx is one of the four query kinds:
// - `Ni` for `min_iterate`
// - `Xi` for `max_iterate`
// - `Np` for `min_panic`
// - `Xp` for `max_panic`
//\
// and `ii` is the inputs for that query, represented as a comma-separated list, with each
// component representing an input:
// - `a`, `b`, `c`... where the input is another node,
// - `uXX` for `UntrackedRead(XX)`
// - `vXX` for `Value(XX)`
// - `sY` for `Successor(Y)`
// - `zY` for `SuccessorOrZero(Y)`
//
// We always enter from the top left node in the diagram.
/// a:Np(a) -+
/// ^ |
/// +--------+
///
/// Simple self-cycle, no iteration, should panic.
#[test]
#[should_panic(expected = "dependency graph cycle")]
fn self_panic() {
let mut db = DbImpl::new();
let a_in = Inputs::new(&db, vec![]);
let a = Input::MinPanic(a_in);
a_in.set_inputs(&mut db).to(vec![a.clone()]);
a.eval(&db);
}
/// a:Np(u10, a) -+
/// ^ |
/// +-------------+
///
/// Simple self-cycle with untracked read, no iteration, should panic.
#[test]
#[should_panic(expected = "dependency graph cycle")]
fn self_untracked_panic() {
let mut db = DbImpl::new();
let a_in = Inputs::new(&db, vec![]);
let a = Input::MinPanic(a_in);
a_in.set_inputs(&mut db).to(vec![untracked(10), a.clone()]);
a.eval(&db);
}
/// a:Ni(a) -+
/// ^ |
/// +--------+
///
/// Simple self-cycle, iteration converges on initial value.
#[test]
fn self_converge_initial_value() {
let mut db = DbImpl::new();
let a_in = Inputs::new(&db, vec![]);
let a = Input::MinIterate(a_in);
a_in.set_inputs(&mut db).to(vec![a.clone()]);
a.assert_value(&db, 255);
}
/// a:Ni(b) --> b:Np(a)
/// ^ |
/// +-----------------+
///
/// Two-query cycle, one with iteration and one without.
/// If we enter from the one with iteration, we converge on its initial value.
#[test]
fn two_mixed_converge_initial_value() {
let mut db = DbImpl::new();
let a_in = Inputs::new(&db, vec![]);
let b_in = Inputs::new(&db, vec![]);
let a = Input::MinIterate(a_in);
let b = Input::MinPanic(b_in);
a_in.set_inputs(&mut db).to(vec![b]);
b_in.set_inputs(&mut db).to(vec![a.clone()]);
a.assert_value(&db, 255);
}
/// a:Np(b) --> b:Ni(a)
/// ^ |
/// +-----------------+
///
/// Two-query cycle, one with iteration and one without.
/// If we enter from the one with no iteration, we panic.
#[test]
#[should_panic(expected = "dependency graph cycle")]
fn two_mixed_panic() {
let mut db = DbImpl::new();
let a_in = Inputs::new(&db, vec![]);
let b_in = Inputs::new(&db, vec![]);
let a = Input::MinPanic(b_in);
let b = Input::MinIterate(a_in);
a_in.set_inputs(&mut db).to(vec![b]);
b_in.set_inputs(&mut db).to(vec![a.clone()]);
a.eval(&db);
}
/// a:Ni(b) --> b:Xi(a)
/// ^ |
/// +-----------------+
///
/// Two-query cycle, both with iteration.
/// We converge on the initial value of whichever we first enter from.
#[test]
fn two_iterate_converge_initial_value() {
let mut db = DbImpl::new();
let a_in = Inputs::new(&db, vec![]);
let b_in = Inputs::new(&db, vec![]);
let a = Input::MinIterate(a_in);
let b = Input::MaxIterate(b_in);
a_in.set_inputs(&mut db).to(vec![b.clone()]);
b_in.set_inputs(&mut db).to(vec![a.clone()]);
a.assert_value(&db, 255);
b.assert_value(&db, 255);
}
/// a:Xi(b) --> b:Ni(a)
/// ^ |
/// +-----------------+
///
/// Two-query cycle, both with iteration.
/// We converge on the initial value of whichever we enter from.
/// (Same setup as above test, different query order.)
#[test]
fn two_iterate_converge_initial_value_2() {
let mut db = DbImpl::new();
let a_in = Inputs::new(&db, vec![]);
let b_in = Inputs::new(&db, vec![]);
let a = Input::MaxIterate(a_in);
let b = Input::MinIterate(b_in);
a_in.set_inputs(&mut db).to(vec![b.clone()]);
b_in.set_inputs(&mut db).to(vec![a.clone()]);
a.assert_value(&db, 0);
b.assert_value(&db, 0);
}
/// a:Np(b) --> b:Ni(c) --> c:Xp(b)
/// ^ |
/// +-----------------+
///
/// Two-query cycle, enter indirectly at node with iteration, converge on its initial value.
#[test]
fn two_indirect_iterate_converge_initial_value() {
let mut db = DbImpl::new();
let a_in = Inputs::new(&db, vec![]);
let b_in = Inputs::new(&db, vec![]);
let c_in = Inputs::new(&db, vec![]);
let a = Input::MinPanic(a_in);
let b = Input::MinIterate(b_in);
let c = Input::MaxPanic(c_in);
a_in.set_inputs(&mut db).to(vec![b.clone()]);
b_in.set_inputs(&mut db).to(vec![c]);
c_in.set_inputs(&mut db).to(vec![b]);
a.assert_value(&db, 255);
}
/// a:Xp(b) --> b:Np(c) --> c:Xi(b)
/// ^ |
/// +-----------------+
///
/// Two-query cycle, enter indirectly at node without iteration, panic.
#[test]
#[should_panic(expected = "dependency graph cycle")]
fn two_indirect_panic() {
let mut db = DbImpl::new();
let a_in = Inputs::new(&db, vec![]);
let b_in = Inputs::new(&db, vec![]);
let c_in = Inputs::new(&db, vec![]);
let a = Input::MinPanic(a_in);
let b = Input::MinPanic(b_in);
let c = Input::MaxIterate(c_in);
a_in.set_inputs(&mut db).to(vec![b.clone()]);
b_in.set_inputs(&mut db).to(vec![c]);
c_in.set_inputs(&mut db).to(vec![b]);
a.eval(&db);
}
/// a:Np(b) -> b:Ni(v200,c) -> c:Xp(b)
/// ^ |
/// +---------------------+
///
/// Two-query cycle, converges to non-initial value.
#[test]
fn two_converge() {
let mut db = DbImpl::new();
let a_in = Inputs::new(&db, vec![]);
let b_in = Inputs::new(&db, vec![]);
let c_in = Inputs::new(&db, vec![]);
let a = Input::MinPanic(a_in);
let b = Input::MinIterate(b_in);
let c = Input::MaxPanic(c_in);
a_in.set_inputs(&mut db).to(vec![b.clone()]);
b_in.set_inputs(&mut db).to(vec![value(200), c]);
c_in.set_inputs(&mut db).to(vec![b]);
a.assert_value(&db, 200);
}
/// a:Xp(b) -> b:Xi(v20,c) -> c:Xp(sb)
/// ^ |
/// +---------------------+
///
/// Two-query cycle, falls back due to >3 iterations.
#[test]
fn two_fallback_count() {
let mut db = DbImpl::new();
let a_in = Inputs::new(&db, vec![]);
let b_in = Inputs::new(&db, vec![]);
let c_in = Inputs::new(&db, vec![]);
let a = Input::MaxPanic(a_in);
let b = Input::MaxIterate(b_in);
let c = Input::MaxPanic(c_in);
a_in.set_inputs(&mut db).to(vec![b.clone()]);
b_in.set_inputs(&mut db).to(vec![value(20), c]);
c_in.set_inputs(&mut db)
.to(vec![Input::Successor(Box::new(b))]);
a.assert_count(&db);
}
/// a:Xp(b) -> b:Xi(v20,c) -> c:Xp(zb)
/// ^ |
/// +---------------------+
///
/// Two-query cycle, falls back but fallback does not converge.
#[test]
#[should_panic(expected = "fallback did not converge")]
fn two_fallback_diverge() {
let mut db = DbImpl::new();
let a_in = Inputs::new(&db, vec![]);
let b_in = Inputs::new(&db, vec![]);
let c_in = Inputs::new(&db, vec![]);
let a = Input::MaxPanic(a_in);
let b = Input::MaxIterate(b_in);
let c = Input::MaxPanic(c_in);
a_in.set_inputs(&mut db).to(vec![b.clone()]);
b_in.set_inputs(&mut db).to(vec![value(20), c.clone()]);
c_in.set_inputs(&mut db)
.to(vec![Input::SuccessorOrZero(Box::new(b))]);
a.assert_count(&db);
}
/// a:Xp(b) -> b:Xi(v244,c) -> c:Xp(sb)
/// ^ |
/// +---------------------+
///
/// Two-query cycle, falls back due to value reaching >MAX_VALUE (we start at 244 and each
/// iteration increments until we reach >245).
#[test]
fn two_fallback_value() {
let mut db = DbImpl::new();
let a_in = Inputs::new(&db, vec![]);
let b_in = Inputs::new(&db, vec![]);
let c_in = Inputs::new(&db, vec![]);
let a = Input::MaxPanic(a_in);
let b = Input::MaxIterate(b_in);
let c = Input::MaxPanic(c_in);
a_in.set_inputs(&mut db).to(vec![b.clone()]);
b_in.set_inputs(&mut db).to(vec![value(244), c]);
c_in.set_inputs(&mut db)
.to(vec![Input::Successor(Box::new(b))]);
a.assert_bounds(&db);
}
/// a:Ni(b) -> b:Np(a, c) -> c:Np(v25, a)
/// ^ | |
/// +----------+------------------------+
///
/// Three-query cycle, (b) and (c) both depend on (a). We converge on 25.
#[test]
fn three_fork_converge() {
let mut db = DbImpl::new();
let a_in = Inputs::new(&db, vec![]);
let b_in = Inputs::new(&db, vec![]);
let c_in = Inputs::new(&db, vec![]);
let a = Input::MinIterate(a_in);
let b = Input::MinPanic(b_in);
let c = Input::MinPanic(c_in);
a_in.set_inputs(&mut db).to(vec![b]);
b_in.set_inputs(&mut db).to(vec![a.clone(), c]);
c_in.set_inputs(&mut db).to(vec![value(25), a.clone()]);
a.assert_value(&db, 25);
}
/// a:Ni(b) -> b:Ni(a, c) -> c:Np(v25, b)
/// ^ | ^ |
/// +----------+ +----------+
///
/// Layered cycles. We converge on 25.
#[test]
fn layered_converge() {
let mut db = DbImpl::new();
let a_in = Inputs::new(&db, vec![]);
let b_in = Inputs::new(&db, vec![]);
let c_in = Inputs::new(&db, vec![]);
let a = Input::MinIterate(a_in);
let b = Input::MinIterate(b_in);
let c = Input::MinPanic(c_in);
a_in.set_inputs(&mut db).to(vec![b.clone()]);
b_in.set_inputs(&mut db).to(vec![a.clone(), c]);
c_in.set_inputs(&mut db).to(vec![value(25), b]);
a.assert_value(&db, 25);
}
/// a:Xi(b) -> b:Xi(a, c) -> c:Xp(v25, sb)
/// ^ | ^ |
/// +----------+ +----------+
///
/// Layered cycles. We hit max iterations and fall back.
#[test]
fn layered_fallback_count() {
let mut db = DbImpl::new();
let a_in = Inputs::new(&db, vec![]);
let b_in = Inputs::new(&db, vec![]);
let c_in = Inputs::new(&db, vec![]);
let a = Input::MaxIterate(a_in);
let b = Input::MaxIterate(b_in);
let c = Input::MaxPanic(c_in);
a_in.set_inputs(&mut db).to(vec![b.clone()]);
b_in.set_inputs(&mut db).to(vec![a.clone(), c]);
c_in.set_inputs(&mut db)
.to(vec![value(25), Input::Successor(Box::new(b))]);
a.assert_count(&db);
}
/// a:Xi(b) -> b:Xi(a, c) -> c:Xp(v243, sb)
/// ^ | ^ |
/// +----------+ +----------+
///
/// Layered cycles. We hit max value and fall back.
#[test]
fn layered_fallback_value() {
let mut db = DbImpl::new();
let a_in = Inputs::new(&db, vec![]);
let b_in = Inputs::new(&db, vec![]);
let c_in = Inputs::new(&db, vec![]);
let a = Input::MaxIterate(a_in);
let b = Input::MaxIterate(b_in);
let c = Input::MaxPanic(c_in);
a_in.set_inputs(&mut db).to(vec![b.clone()]);
b_in.set_inputs(&mut db).to(vec![a.clone(), c]);
c_in.set_inputs(&mut db)
.to(vec![value(243), Input::Successor(Box::new(b))]);
a.assert_bounds(&db);
}
/// a:Ni(b) -> b:Ni(c) -> c:Np(v25, a, b)
/// ^ ^ |
/// +----------+------------------------+
///
/// Nested cycles. We converge on 25.
#[test]
fn nested_converge() {
let mut db = DbImpl::new();
let a_in = Inputs::new(&db, vec![]);
let b_in = Inputs::new(&db, vec![]);
let c_in = Inputs::new(&db, vec![]);
let a = Input::MinIterate(a_in);
let b = Input::MinIterate(b_in);
let c = Input::MinPanic(c_in);
a_in.set_inputs(&mut db).to(vec![b.clone()]);
b_in.set_inputs(&mut db).to(vec![c]);
c_in.set_inputs(&mut db).to(vec![value(25), a.clone(), b]);
a.assert_value(&db, 25);
}
/// a:Ni(b) -> b:Ni(c) -> c:Np(v25, b, a)
/// ^ ^ |
/// +----------+------------------------+
///
/// Nested cycles, inner first. We converge on 25.
#[test]
fn nested_inner_first_converge() {
let mut db = DbImpl::new();
let a_in = Inputs::new(&db, vec![]);
let b_in = Inputs::new(&db, vec![]);
let c_in = Inputs::new(&db, vec![]);
let a = Input::MinIterate(a_in);
let b = Input::MinIterate(b_in);
let c = Input::MinPanic(c_in);
a_in.set_inputs(&mut db).to(vec![b.clone()]);
b_in.set_inputs(&mut db).to(vec![c]);
c_in.set_inputs(&mut db).to(vec![value(25), b, a.clone()]);
a.assert_value(&db, 25);
}
/// a:Xi(b) -> b:Xi(c) -> c:Xp(v25, a, sb)
/// ^ ^ |
/// +----------+-------------------------+
///
/// Nested cycles. We hit max iterations and fall back.
#[test]
fn nested_fallback_count() {
let mut db = DbImpl::new();
let a_in = Inputs::new(&db, vec![]);
let b_in = Inputs::new(&db, vec![]);
let c_in = Inputs::new(&db, vec![]);
let a = Input::MaxIterate(a_in);
let b = Input::MaxIterate(b_in);
let c = Input::MaxPanic(c_in);
a_in.set_inputs(&mut db).to(vec![b.clone()]);
b_in.set_inputs(&mut db).to(vec![c]);
c_in.set_inputs(&mut db)
.to(vec![value(25), a.clone(), Input::Successor(Box::new(b))]);
a.assert_count(&db);
}
/// a:Xi(b) -> b:Xi(c) -> c:Xp(v25, b, sa)
/// ^ ^ |
/// +----------+-------------------------+
///
/// Nested cycles, inner first. We hit max iterations and fall back.
#[test]
fn nested_inner_first_fallback_count() {
let mut db = DbImpl::new();
let a_in = Inputs::new(&db, vec![]);
let b_in = Inputs::new(&db, vec![]);
let c_in = Inputs::new(&db, vec![]);
let a = Input::MaxIterate(a_in);
let b = Input::MaxIterate(b_in);
let c = Input::MaxPanic(c_in);
a_in.set_inputs(&mut db).to(vec![b.clone()]);
b_in.set_inputs(&mut db).to(vec![c]);
c_in.set_inputs(&mut db)
.to(vec![value(25), b, Input::Successor(Box::new(a.clone()))]);
a.assert_count(&db);
}
/// a:Xi(b) -> b:Xi(c) -> c:Xp(v243, a, sb)
/// ^ ^ |
/// +----------+--------------------------+
///
/// Nested cycles. We hit max value and fall back.
#[test]
fn nested_fallback_value() {
let mut db = DbImpl::new();
let a_in = Inputs::new(&db, vec![]);
let b_in = Inputs::new(&db, vec![]);
let c_in = Inputs::new(&db, vec![]);
let a = Input::MaxIterate(a_in);
let b = Input::MaxIterate(b_in);
let c = Input::MaxPanic(c_in);
a_in.set_inputs(&mut db).to(vec![b.clone()]);
b_in.set_inputs(&mut db).to(vec![c.clone()]);
c_in.set_inputs(&mut db).to(vec![
value(243),
a.clone(),
Input::Successor(Box::new(b.clone())),
]);
a.assert_bounds(&db);
b.assert_bounds(&db);
c.assert_bounds(&db);
}
/// a:Xi(b) -> b:Xi(c) -> c:Xp(v243, b, sa)
/// ^ ^ |
/// +----------+--------------------------+
///
/// Nested cycles, inner first. We hit max value and fall back.
#[test]
fn nested_inner_first_fallback_value() {
let mut db = DbImpl::new();
let a_in = Inputs::new(&db, vec![]);
let b_in = Inputs::new(&db, vec![]);
let c_in = Inputs::new(&db, vec![]);
let a = Input::MaxIterate(a_in);
let b = Input::MaxIterate(b_in);
let c = Input::MaxPanic(c_in);
a_in.set_inputs(&mut db).to(vec![b.clone()]);
b_in.set_inputs(&mut db).to(vec![c]);
c_in.set_inputs(&mut db)
.to(vec![value(243), b, Input::Successor(Box::new(a.clone()))]);
a.assert_bounds(&db);
}
/// a:Ni(b) -> b:Ni(c, a) -> c:Np(v25, a, b)
/// ^ ^ | |
/// +----------+--------|------------------+
/// | |
/// +-------------------+
///
/// Nested cycles, double head. We converge on 25.
#[test]
fn nested_double_converge() {
let mut db = DbImpl::new();
let a_in = Inputs::new(&db, vec![]);
let b_in = Inputs::new(&db, vec![]);
let c_in = Inputs::new(&db, vec![]);
let a = Input::MinIterate(a_in);
let b = Input::MinIterate(b_in);
let c = Input::MinPanic(c_in);
a_in.set_inputs(&mut db).to(vec![b.clone()]);
b_in.set_inputs(&mut db).to(vec![c, a.clone()]);
c_in.set_inputs(&mut db).to(vec![value(25), a.clone(), b]);
a.assert_value(&db, 25);
}
// Multiple-revision cycles
/// a:Ni(b) --> b:Np(a)
/// ^ |
/// +-----------------+
///
/// a:Ni(b) --> b:Np(v30)
///
/// Cycle becomes not-a-cycle in next revision.
#[test]
fn cycle_becomes_non_cycle() {
let mut db = DbImpl::new();
let a_in = Inputs::new(&db, vec![]);
let b_in = Inputs::new(&db, vec![]);
let a = Input::MinIterate(a_in);
let b = Input::MinPanic(b_in);
a_in.set_inputs(&mut db).to(vec![b]);
b_in.set_inputs(&mut db).to(vec![a.clone()]);
a.assert_value(&db, 255);
b_in.set_inputs(&mut db).to(vec![value(30)]);
a.assert_value(&db, 30);
}
/// a:Ni(b) --> b:Np(v30)
///
/// a:Ni(b) --> b:Np(a)
/// ^ |
/// +-----------------+
///
/// Non-cycle becomes a cycle in next revision.
#[test]
fn non_cycle_becomes_cycle() {
let mut db = DbImpl::new();
let a_in = Inputs::new(&db, vec![]);
let b_in = Inputs::new(&db, vec![]);
let a = Input::MinIterate(a_in);
let b = Input::MinPanic(b_in);
a_in.set_inputs(&mut db).to(vec![b]);
b_in.set_inputs(&mut db).to(vec![value(30)]);
a.assert_value(&db, 30);
b_in.set_inputs(&mut db).to(vec![a.clone()]);
a.assert_value(&db, 255);
}
/// a:Xi(b) -> b:Xi(c, a) -> c:Xp(v25, a, sb)
/// ^ ^ | |
/// +----------+--------|-------------------+
/// | |
/// +-------------------+
///
/// Nested cycles, double head. We hit max iterations and fall back, then max value on the next
/// revision, then converge on the next.
#[test]
fn nested_double_multiple_revisions() {
let mut db = DbImpl::new();
let a_in = Inputs::new(&db, vec![]);
let b_in = Inputs::new(&db, vec![]);
let c_in = Inputs::new(&db, vec![]);
let a = Input::MaxIterate(a_in);
let b = Input::MaxIterate(b_in);
let c = Input::MaxPanic(c_in);
a_in.set_inputs(&mut db).to(vec![b.clone()]);
b_in.set_inputs(&mut db).to(vec![c, a.clone()]);
c_in.set_inputs(&mut db).to(vec![
value(25),
a.clone(),
Input::Successor(Box::new(b.clone())),
]);
a.assert_count(&db);
// next revision, we hit max value instead
c_in.set_inputs(&mut db).to(vec![
value(243),
a.clone(),
Input::Successor(Box::new(b.clone())),
]);
a.assert_bounds(&db);
// and next revision, we converge
c_in.set_inputs(&mut db)
.to(vec![value(240), a.clone(), b.clone()]);
a.assert_value(&db, 240);
// one more revision, without relevant changes
a_in.set_inputs(&mut db).to(vec![b]);
a.assert_value(&db, 240);
}
/// a:Ni(b) -> b:Ni(c) -> c:Ni(a)
/// ^ |
/// +---------------------------+
///
/// In a cycle with some LOW durability and some HIGH durability inputs, changing a LOW durability
/// input still re-executes the full cycle in the next revision.
#[test]
fn cycle_durability() {
let mut db = DbImpl::new();
let a_in = Inputs::new(&db, vec![]);
let b_in = Inputs::new(&db, vec![]);
let c_in = Inputs::new(&db, vec![]);
let a = Input::MinIterate(a_in);
let b = Input::MinIterate(b_in);
let c = Input::MinIterate(c_in);
a_in.set_inputs(&mut db)
.with_durability(Durability::LOW)
.to(vec![b.clone()]);
b_in.set_inputs(&mut db)
.with_durability(Durability::HIGH)
.to(vec![c]);
c_in.set_inputs(&mut db)
.with_durability(Durability::HIGH)
.to(vec![a.clone()]);
a.assert_value(&db, 255);
// next revision, we converge instead
a_in.set_inputs(&mut db)
.with_durability(Durability::LOW)
.to(vec![value(45), b]);
a.assert_value(&db, 45);
}
/// a:Np(v59, b) -> b:Ni(v60, c) -> c:Np(b)
/// ^ |
/// +---------------------+
///
/// If nothing in a cycle changed in the new revision, no part of the cycle should re-execute.
#[test]
fn cycle_unchanged() {
let mut db = ExecuteValidateLoggerDatabase::default();
let a_in = Inputs::new(&db, vec![]);
let b_in = Inputs::new(&db, vec![]);
let c_in = Inputs::new(&db, vec![]);
let a = Input::MinPanic(a_in);
let b = Input::MinIterate(b_in);
let c = Input::MinPanic(c_in);
a_in.set_inputs(&mut db).to(vec![value(59), b.clone()]);
b_in.set_inputs(&mut db).to(vec![value(60), c]);
c_in.set_inputs(&mut db).to(vec![b.clone()]);
a.assert_value(&db, 59);
b.assert_value(&db, 60);
db.assert_logs_len(5);
// next revision, we change only A, which is not part of the cycle and the cycle does not
// depend on.
a_in.set_inputs(&mut db).to(vec![value(45), b.clone()]);
b.assert_value(&db, 60);
db.assert_logs(expect![[r#"
[
"salsa_event(DidValidateMemoizedValue { database_key: min_iterate(Id(1)) })",
]"#]]);
a.assert_value(&db, 45);
}
/// a:Np(v59, b) -> b:Ni(v60, c) -> c:Np(d) -> d:Ni(v61, b, e) -> e:Np(d)
/// ^ | ^ |
/// +--------------------------+ +--------------+
///
/// If nothing in a nested cycle changed in the new revision, no part of the cycle should
/// re-execute.
#[test]
fn cycle_unchanged_nested() {
let mut db = ExecuteValidateLoggerDatabase::default();
let a_in = Inputs::new(&db, vec![]);
let b_in = Inputs::new(&db, vec![]);
let c_in = Inputs::new(&db, vec![]);
let d_in = Inputs::new(&db, vec![]);
let e_in = Inputs::new(&db, vec![]);
let a = Input::MinPanic(a_in);
let b = Input::MinIterate(b_in);
let c = Input::MinPanic(c_in);
let d = Input::MinIterate(d_in);
let e = Input::MinPanic(e_in);
a_in.set_inputs(&mut db).to(vec![value(59), b.clone()]);
b_in.set_inputs(&mut db).to(vec![value(60), c.clone()]);
c_in.set_inputs(&mut db).to(vec![d.clone()]);
d_in.set_inputs(&mut db)
.to(vec![value(61), b.clone(), e.clone()]);
e_in.set_inputs(&mut db).to(vec![d.clone()]);
a.assert_value(&db, 59);
b.assert_value(&db, 60);
db.assert_logs_len(13);
// next revision, we change only A, which is not part of the cycle and the cycle does not
// depend on.
a_in.set_inputs(&mut db).to(vec![value(45), b.clone()]);
b.assert_value(&db, 60);
db.assert_logs(expect![[r#"
[
"salsa_event(DidValidateMemoizedValue { database_key: min_iterate(Id(1)) })",
]"#]]);
a.assert_value(&db, 45);
}
/// +--------------------------------+
/// | v
/// a:Np(v59, b) -> b:Ni(v60, c) -> c:Np(d, e) -> d:Ni(v61, b, e) -> e:Ni(d)
/// ^ | ^ |
/// +-----------------------------+ +--------------+
///
/// If nothing in a nested cycle changed in the new revision, no part of the cycle should
/// re-execute.
#[test_log::test]
fn cycle_unchanged_nested_intertwined() {
// We run this test twice in order to catch some subtly different cases; see below.
for i in 0..1 {
let mut db = ExecuteValidateLoggerDatabase::default();
let a_in = Inputs::new(&db, vec![]);
let b_in = Inputs::new(&db, vec![]);
let c_in = Inputs::new(&db, vec![]);
let d_in = Inputs::new(&db, vec![]);
let e_in = Inputs::new(&db, vec![]);
let a = Input::MinPanic(a_in);
let b = Input::MinIterate(b_in);
let c = Input::MinPanic(c_in);
let d = Input::MinIterate(d_in);
let e = Input::MinIterate(e_in);
a_in.set_inputs(&mut db).to(vec![value(59), b.clone()]);
b_in.set_inputs(&mut db).to(vec![value(60), c.clone()]);
c_in.set_inputs(&mut db).to(vec![d.clone(), e.clone()]);
d_in.set_inputs(&mut db)
.to(vec![value(61), b.clone(), e.clone()]);
e_in.set_inputs(&mut db).to(vec![d.clone()]);
a.assert_value(&db, 59);
b.assert_value(&db, 60);
// First time we run this test, don't fetch c/d/e here; this means they won't get marked
// `verified_final` in R6 (this revision), which will leave us in the next revision (R7)
// with a chain of could-be-provisional memos from the previous revision which should be
// final but were never confirmed as such; this triggers the case in `deep_verify_memo`
// where we need to double-check `validate_provisional` after traversing dependencies.
//
// Second time we run this test, fetch everything in R6, to check the behavior of
// `maybe_changed_after` with all validated-final memos.
if i == 1 {
c.assert_value(&db, 60);
d.assert_value(&db, 60);
e.assert_value(&db, 60);
}
db.assert_logs_len(15 + i);
// next revision, we change only A, which is not part of the cycle and the cycle does not
// depend on.
a_in.set_inputs(&mut db).to(vec![value(45), b.clone()]);
b.assert_value(&db, 60);
db.assert_logs(expect![[r#"
[
"salsa_event(DidValidateMemoizedValue { database_key: min_iterate(Id(1)) })",
]"#]]);
a.assert_value(&db, 45);
}
}
/// Provisional query results in a cycle should still be cached within a single iteration.
///
/// a:Ni(v59, b) -> b:Np(v60, c, c, c) -> c:Np(a)
/// ^ |
/// +------------------------------------------+
#[test]
fn repeat_provisional_query() {
let mut db = ExecuteValidateLoggerDatabase::default();
let a_in = Inputs::new(&db, vec![]);
let b_in = Inputs::new(&db, vec![]);
let c_in = Inputs::new(&db, vec![]);
let a = Input::MinIterate(a_in);
let b = Input::MinPanic(b_in);
let c = Input::MinPanic(c_in);
a_in.set_inputs(&mut db).to(vec![value(59), b.clone()]);
b_in.set_inputs(&mut db)
.to(vec![value(60), c.clone(), c.clone(), c]);
c_in.set_inputs(&mut db).to(vec![a.clone()]);
a.assert_value(&db, 59);
db.assert_logs(expect![[r#"
[
"salsa_event(WillExecute { database_key: min_iterate(Id(0)) })",
"salsa_event(WillExecute { database_key: min_panic(Id(1)) })",
"salsa_event(WillExecute { database_key: min_panic(Id(2)) })",
"salsa_event(WillIterateCycle { database_key: min_iterate(Id(0)), iteration_count: IterationCount(1), fell_back: false })",
"salsa_event(WillExecute { database_key: min_panic(Id(1)) })",
"salsa_event(WillExecute { database_key: min_panic(Id(2)) })",
]"#]]);
}
|