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 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151
#![no_std] //! A simple memory allocator, written for educational purposes. //! //! This module was written primarily for the code to be read. The allocator //! enclosed can be used as a memory allocator in a rust program. //! //! ## Usage //! //! ```rust //! use basic_allocator::UnixAllocator; //! //! #[global_allocator] //! static ALLOCATOR: UnixAllocator = UnixAllocator::new(); //! fn main() { //! println!("It works!") //! } //! ``` //! //! See also //! [`core::alloc::GlobalAlloc`](https://doc.rust-lang.org/nightly/core/alloc/trait.GlobalAlloc.html). //! //! ## Major Components //! //! This module has several parts. //! //! ### [`BlockList`](struct.BlockList.html) //! //! A `BlockList` is a linked list of _freed_ memory not returned to the OS, //! which can be reused by the allocator. //! //! The free block starts with a header, and then has unused memory after that. //! The header is 16 bytes, and consists of a pointer to the next block and the //! size of the block as a whole. //! //! ### [`RawAlloc`](struct.RawAlloc.html) //! //! A `RawAlloc` is a single-threaded, non-thread-safe heap and freed memory //! manager, implementing //! [`core::alloc::GlobalAlloc`](https://doc.rust-lang.org/nightly/core/alloc/trait.GlobalAlloc.html). //! However, because it is not thread-safe, it canot be used as a global //! allocator.BlockList //! //! ### [`UnixAllocator`](struct.UnixAllocator.html) //! //! A `UnixAllocator` wraps `RawAlloc` with a spin lock to make it thread-safe, //! allowing it to be used as the global allocator. It also combines `RawAlloc` //! with a unix-specific `UnixHeapGrower` to use virtual memory pages as its //! underlying basis for making those calls. //! //! ## Lesser Components //! //! `HeapGrower` is a simple trait interface meant to abstract over the calls to //! the OS to expand the heap. use core::alloc::{GlobalAlloc, Layout}; use core::fmt; use core::mem::MaybeUninit; use core::ops::Range; use core::ptr::{null_mut, NonNull}; use core::sync::atomic::{AtomicU8, Ordering}; use spin::{Mutex, MutexGuard}; use static_assertions::const_assert; /// The header for our free blocks. /// /// The header includes a pointer to the next free block, and the size of the /// current block (including the header). /// /// We use C representation and align to 16 bytes for... simplicity. This is /// perhaps a stronger constraint that we need, but it does make things simple /// and straightforward. #[repr(C, align(16))] pub struct FreeHeader { next: Option<FreeBlock>, size: usize, } /// We will align to 16 bytes and our headers will be given that much space /// Similarly, all blocks will be at least 16 bytes large, even if they aren't /// aware of it. /// /// This is likely a stronger constraint than is entirely needed, but it does /// simplify things. const HEADER_SIZE: usize = 16; const_assert!(HEADER_SIZE <= core::mem::size_of::<FreeHeader>()); /// An enum for easy comparison of blocks and their order pub enum Relation { Before, AdjacentBefore, Overlapping, AdjacentAfter, After, } impl FreeHeader { /// Construct a header from a freed memory block at `ptr`, with a link to /// the next in `next`, and the size of the block in `size`. /// /// # Safety /// /// This is unsafe because its manipulating raw, freed memory. /// /// To use this safely, `ptr` must point to memory of size `size` not in use /// by or accessible by any program logic. /// /// Further safety constraints are enforced by the invariants of `FreeBlock` /// and `BlockList`. #[allow(clippy::cast_ptr_alignment)] pub unsafe fn from_raw( ptr: NonNull<u8>, next: Option<FreeBlock>, size: usize, ) -> NonNull<FreeHeader> { let header = FreeHeader { next, size }; let raw_ptr: NonNull<FreeHeader> = ptr.cast(); core::ptr::write(ptr.as_ptr() as *mut FreeHeader, header); raw_ptr } } /// A `FreeBlock` is a wrapper around a pointer to a freed block to be /// maintained in a [`BlockList`](struct.BlockList.html). /// /// Invariants are enforced by / inherited from the NonNull strict. /// /// Note that this is very similar to Box, except that it doesn't assume a heap /// or memory allocator, so it doesn't implement Clone or Drop, and it also has /// a 'next'. pub struct FreeBlock { header: NonNull<FreeHeader>, } impl FreeBlock { /// Construct a `FreeBlock` from raw parts: a freed memory block at `ptr` of /// size `size`. This will also write the header appropriately. /// /// # Safety /// /// This is unsafe because its manipulating raw, freed memory. /// /// To use this safely, `ptr` must point to memory of size `size` not in use /// by or accessible by any program logic. /// /// Further safety constraints are enforced by the invariants of `BlockList`. #[must_use] pub unsafe fn from_raw(ptr: NonNull<u8>, next: Option<FreeBlock>, size: usize) -> FreeBlock { if size < HEADER_SIZE { panic!("Can't recapture a block smaller than HEADER_SIZE"); } let header = FreeHeader::from_raw(ptr, next, size); FreeBlock { header } } /// Get the memory covered by this block as a slice. pub fn as_slice(&self) -> &[u8] { unsafe { let size = self.header_view().size; core::slice::from_raw_parts(self.header.as_ptr() as *const u8, size) } } /// Get the pointer range covered by this block. pub fn as_range(&self) -> Range<*const u8> { unsafe { let size = self.header_view().size; let start = self.header.as_ptr() as *const u8; start..(start.add(size)) } } /// Consume this block and return the range of memory covered, and the next /// block in the list. #[must_use] pub fn decompose(mut self) -> (Range<NonNull<u8>>, Option<FreeBlock>) { let next = self.take_next(); let range = unsafe { let size = self.header_view().size; let start: NonNull<u8> = self.header.cast(); let end: NonNull<u8> = NonNull::new_unchecked(self.header.as_ptr().add(size) as *mut u8); start..end }; (range, next) } /// Compare two blocks to see how they are ordered. fn relation(&self, other: &Self) -> Relation { let self_range = self.as_range(); let other_range = other.as_range(); if self_range.end < other_range.start { Relation::Before } else if self_range.end == other_range.start { Relation::AdjacentBefore } else if self_range.start < other_range.end { Relation::Overlapping } else if self_range.start == other_range.end { Relation::AdjacentAfter } else { Relation::After } } /// Get the next block over from this one. fn next(&self) -> Option<&Self> { (&self.header_view().next).into() } /// Get the next block over from this one. fn next_mut(&mut self) -> Option<&mut Self> { unsafe { (&mut self.header_mut().next).into() } } /// Remove the next, and return it #[must_use] fn take_next(&mut self) -> Option<Self> { unsafe { (&mut self.header_mut().next).take() } } /// Set this block's next to new_next, and return the old one #[must_use] fn replace_next(&mut self, new_next: FreeBlock) -> Option<Self> { unsafe { (&mut self.header_mut().next).replace(new_next) } } /// The size of the block, in bytes. pub fn size(&self) -> usize { self.header_view().size } /// An immutable pointer to the header pub fn header_view(&self) -> &FreeHeader { unsafe { self.header.as_ref() } } /// Get a mutable view of the header. /// /// # Safety /// /// This method is unsafe because it allows modifying the size or pointer of /// a free block in safe code, which could lead to corruption. pub unsafe fn header_mut(&mut self) -> &mut FreeHeader { self.header.as_mut() } /// Remove the block after this one from the linked list, and return /// a pointer to that block and its size. /// /// As is required in a linked list, this will set self.next = next.next. /// /// If there is no next, returns (None, 0). #[must_use] pub fn pop_next(&mut self) -> Option<FreeBlock> { let mut next = match self.take_next() { None => { return None; } Some(n) => n, }; // Update this block to look to next's next, cutting next out of the chain if let Some(next_next) = next.take_next() { assert!(self.replace_next(next_next).is_none()); } Some(next) } /// Insert a new element, after this one, maintaining linked list invariants. /// /// # Safety /// /// `ptr` must be a pointer to valid, freed memory of size `size`. pub unsafe fn insert(&mut self, block: FreeBlock) { let mut inserting = block; let next_next = self.header_mut().next.take(); inserting.header_mut().next = next_next; self.header_mut().next = Some(inserting); } /// Insert a new element, after this one, maintaining linked list invariants /// and merging with either this item and/or the next, depending on /// adjacency. /// /// # Safety /// /// `ptr` must be a pointer to valid, freed memory of size `size`. To /// maintain `BlockList` invariants, `ptr` must also be greater then /// self.header, and less than self.next (or self.next must be null). unsafe fn insert_merge(&mut self, block: FreeBlock) -> usize { let this_end = self.as_range().end; let other_start = block.as_range().start; let (merges, try_next) = if this_end == other_start { self.header_mut().size += block.size(); (1, self) } else { self.insert(block); (0, self.next_mut().unwrap()) }; merges + if try_next.try_merge_next() { 1 } else { 0 } } /// Split off part of this FreeBlock, and return a pointer to the split off /// data. /// /// The returned pointer is to a region of size 'size' that is no longer /// considered free. /// /// Panics if 'size' is greater than this block's size - HEADER_SIZE, as /// there is no way to split off a chunk that large while leaving behind a /// FreeBlock with an intact header. pub fn split(&mut self, size: usize) -> Range<NonNull<u8>> { if size + HEADER_SIZE > self.header_view().size { panic!( "Can't split a block of size {} off of a block of size {} - need {} for header", size, self.size(), HEADER_SIZE, ) } unsafe { let self_size = self.size(); let header = self.header_mut(); header.size -= size; let start = NonNull::new_unchecked((header as *mut FreeHeader as *mut u8).add(header.size)); let end = NonNull::new_unchecked((header as *mut FreeHeader as *mut u8).add(self_size)); // log::trace!( // "Splitting {} bytes off from {:?}:{} to get {:?}", // size, // (header as *mut FreeHeader as *mut u8), // self_size, // ptr, // ); start..end } } /// Attempt to merge this block with the next. /// /// If the next block exaists, is adjacent, and exists directly after this /// block, the two will merge and this will return True; otherwise, this will /// return False. pub fn try_merge_next(&mut self) -> bool { let (next_start, next_size) = match self.next() { None => return false, Some(block) => (block.as_range().start, block.size()), }; if self.as_range().end != next_start { return false; }; unsafe { let header = self.header_mut(); header.size += next_size; let mut next = header.next.take().unwrap(); header.next = next.header_mut().next.take(); } true } } /// A `BlockList` is a linked list of "free" blocks in memory. /// /// Each block should be considered "owned" by the BlockList when inserted, and /// do not hold any sort of payload. They may be split or merged internally. /// /// In this module, thse memory blocks represent freed memory that has not been /// returned to the OS, and provide a "pool" of available memory for reuse by /// the allocator. /// /// It maintains a few internal invariants: /// /// - Each block should link to the next, with the last one linking to null. /// - Each block should have a pointer < next. /// - No two blocks should be precisely adjacent (those should be automatically /// merged on insertion). pub struct BlockList { first: Option<FreeBlock>, } pub struct BlockIter<'list> { next: Option<&'list FreeBlock>, } impl<'list> Iterator for BlockIter<'list> { type Item = &'list FreeBlock; fn next(&mut self) -> Option<Self::Item> { let next = self.next.take()?; self.next = next.next(); Some(next) } } // A BlockList is sendable - as long as the whole "chain" is maintained across // threads, its fine. // // With some tweaking and atomic pointer swapping, we could make a thread-safe // version of BlockList, but that has not been done here; hence, it implements // Send but not Sync. unsafe impl Send for FreeBlock {} impl Default for BlockList { fn default() -> Self { BlockList { first: None } } } impl<'list> IntoIterator for &'list BlockList { type Item = &'list FreeBlock; type IntoIter = BlockIter<'list>; fn into_iter(self) -> Self::IntoIter { BlockIter { next: self.first.as_ref(), } } } // impl<'list> IntoIterator for &'list mut BlockList { // type Item = &'list mut FreeBlock; // type IntoIter = BlockIterMut<'list>; // fn into_iter(self) -> Self::IntoIter { // BlockIterMut { // next: self.first.as_mut(), // } // } // } impl fmt::Display for BlockList { fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { write!(f, "BlockList(")?; let mut start = true; for block in self { if !start { write!(f, ", ")?; } else { start = false; } write!(f, "FreeBlock({:?}, {})", block.header, block.size())?; } write!(f, ")") } } /// Validity contains a representation of all invalid states found in a /// BlockList. #[derive(Default, Debug)] pub struct Validity { /// Number of blocks overlapping other blocks. /// /// This likely indicates corruption. /// /// If there are also out of order blocks, this might undercount. pub overlaps: usize, /// Number of blocks that are directly adjacent to each other, and not /// merged. This shouldn't happen, but isn't totally corrupt. pub adjacents: usize, /// Number of blocks that do not have an address less than their next. /// /// This shouldn't occur. pub out_of_orders: usize, } impl Validity { /// Returns a boolean - a simple check if all cases are 0 pub fn is_valid(&self) -> bool { self.overlaps == 0 && self.adjacents == 0 && self.out_of_orders == 0 } } impl From<Validity> for bool { fn from(v: Validity) -> bool { v.is_valid() } } #[derive(Default, Debug)] pub struct Stats { pub length: usize, pub size: usize, } /// State after a single "apply". pub enum ApplyState<C, R> { // Keep going, and pass C into the next 'apply' Continue(C), // Finish iterating, and return Some(R) Finished(R), // Finish iterating, and return None Fail, } impl BlockList { pub fn iter(&self) -> BlockIter { BlockIter { next: self.first.as_ref(), } } /// Iterate through the blocklist, and apply a function at each step. This /// allows mutating the list as it is traversed, and replaces IterMut, which /// cannot be used due to the links between blocks. /// /// Note that any changes to any block's "next" will be followed at the next /// iteration. pub fn apply<C, R, F: FnMut(&mut FreeBlock, C) -> ApplyState<C, R>>( &mut self, start: C, mut pred: F, ) -> Option<R> { let mut next = self.first.as_mut(); let mut state = start; while let Some(block) = next.take() { state = match pred(block, state) { ApplyState::Continue(c) => c, ApplyState::Finished(r) => return Some(r), ApplyState::Fail => return None, }; next = block.next_mut() } None } /// Check current size of the list, and whether its valid. pub fn stats(&self) -> (Validity, Stats) { let mut validity: Validity = Default::default(); let mut stats: Stats = Default::default(); let mut previous: Option<&FreeBlock> = None; for next in self.iter() { match previous.map(|p| p.relation(&next)) { Some(Relation::Before) => { // This is valid, do nothing. } Some(Relation::AdjacentBefore) => { // Right order, but these should be merged. validity.adjacents += 1; } Some(Relation::Overlapping) => { // This is really bad. validity.overlaps += 1; } Some(Relation::AdjacentAfter) => { // Wrong order, and these should be merged. validity.out_of_orders += 1; validity.adjacents += 1; } Some(Relation::After) => { // Wrong order. validity.out_of_orders += 1; } None => { // This is the first in the list. Valid, do nothing. } } stats.length += 1; stats.size += next.size(); previous = Some(next); } (validity, stats) } /// Find and remove a chunk of size 'size' from the linked list pub fn pop_size(&mut self, size: usize) -> Option<Range<NonNull<u8>>> { // debug!("pop_size({})", size); let first_size = self.first.as_ref()?.size(); // debug!(" pop_size got first"); if first_size == size { // debug!(" First block at {:?} is big enough", first.header); let (range, next) = self.first.take()?.decompose(); self.first = next; return Some(range); } else if first_size >= size + HEADER_SIZE { let split = self.first.as_mut()?.split(size); // debug!( // " Split off from first block at {:?} to {:?}", // first.header, split, // ); return Some(split); } self.apply((), |previous, ()| { let next_size: usize = match previous.next() { None => return ApplyState::Fail, Some(next) => next.size(), }; // log::trace!(" Checking block at {:?} Size {}", next.header, next.size()); if next_size == size { // This block is just right - let's pop it out of the chain and return it let block = previous.pop_next().unwrap(); let (range, next) = block.decompose(); assert!(next.is_none()); return ApplyState::Finished(range); // log::trace!(" Found correctly sized block at {:?}", ptr); } if next_size < size + HEADER_SIZE { // This block is too small to be split, skip it return ApplyState::Continue(()); } // This block is bigger than we need, split it // log::trace!(" Found big block at {:?}", next.header); let ptr = previous.next_mut().unwrap().split(size); ApplyState::Finished(ptr) }) } /// Add a block to the linked list. Takes ownership of ptr. /// /// # Safety /// /// `ptr` must point to valid, reachable memory of at least `size`, and /// ownership of that memory must be transferred to `BlockList` when this /// method is called. pub unsafe fn add_block(&mut self, ptr: NonNull<u8>, size: usize) { let mut new_block = FreeBlock::from_raw(ptr, None, size); let first: &FreeBlock = match self.first { None => { // There are no blocks in this list, so we make this the head of // the list and return self.first = Some(new_block); return; } Some(ref p) => p, }; // We keep the list in sorted order, by pointer, to enable merging. match new_block.relation(first) { Relation::Before => { // This block is well before the first one in the list, so we // add this to the head of the list new_block.header_mut().next = self.first.take(); self.first = Some(new_block); return; } Relation::AdjacentBefore => { // This block is just before the first block in the list, so we // merge the two into a single block new_block.header_mut().next = self.first.take(); let merged = new_block.try_merge_next(); self.first = Some(new_block); assert!(merged, "They were adjacent, they should merge"); return; } Relation::Overlapping => { // These blocks both claim the same memory panic!("Overlapping memory blocks OH NO"); } Relation::AdjacentAfter => { // This block is just after the first block in the list, so we // merge the two into a single block. This block isn't part of // the list yet, and 'previous' already correctly points to the // next block, so all we need to do is increase the 'previous' // block size. let first = self.first.as_mut().unwrap(); first.header_mut().size += size; // Now that 'previous' has grown, it's possible that 'previous' // is now adjacent to 'next', so we try and merge them. This may // or may not actually happen, and either way, we're left with a // valid list afterwards. first.try_merge_next(); return; } _ => {} } // Loop through the list of blocks, to find where this one should be // inserted. Once its place in the list is found, we merge with the // previous and/or next if we can, and if not, insert it into // the list. self.apply(new_block, |previous, new_block| { // By construction, previous < new_block. Now we check previous.next // to see if previous < new_block < next, in which case we insert // and merge, or if next < new_block, we continue iterating through // the list. let next = match previous.next() { Some(n) => n, None => { // previous < new_block, and nothing previous.insert_merge(new_block); return ApplyState::Finished(()); } }; if next.header.cast() < ptr { // next < pointer, so we continue iterating return ApplyState::Continue(new_block); } // If we are here, it means previous < ptr < next. // Time to insert_merge previous.insert_merge(new_block); ApplyState::Finished(()) }); } pub fn len(&self) -> usize { self.iter().count() } pub fn is_empty(&self) -> bool { self.first.is_none() } } // Round up value to the nearest multiple of increment fn round_up(value: usize, increment: usize) -> usize { if value == 0 { return 0; } increment * ((value - 1) / increment + 1) } pub trait HeapGrower { /// Grow the heap by at least size. Returns a pointer and the size of the /// memory available at that pointer. /// /// # Safety /// /// This is pretty much entirely unsafe. /// /// For this to function properly with the other types in this module: /// /// - The return value may be (null, 0), indicating allocation failure. /// - The return value may be (ptr, new_size), where new_size >= size, and /// where the memory pointed to by ptr must be available and untracked by /// any other rust code, including the allocator itself. unsafe fn grow_heap(&mut self, size: usize) -> (*mut u8, usize); } /// UnixHeapGrower uses virtual memory to grow the heap upon request. #[derive(Default)] pub struct UnixHeapGrower { // Just for tracking, not really needed pages: usize, growths: usize, } impl HeapGrower for UnixHeapGrower { unsafe fn grow_heap(&mut self, size: usize) -> (*mut u8, usize) { if size == 0 { return (null_mut(), 0); } let pagesize = sysconf::page::pagesize(); let to_allocate = round_up(size, pagesize); let ptr = libc::mmap( // Address we want the memory at. We don't care, so null it is. null_mut(), // Amount of memory to allocate to_allocate, // We want read/write access to this memory libc::PROT_WRITE | libc::PROT_READ, // MAP_ANON: We don't want a file descriptor, we're just going to // use the memory. // // MAP_PRIVATE: We're not sharing this with any other process. // // Well, I'm pretty unsure about these choices, but they seem to work... libc::MAP_ANON | libc::MAP_PRIVATE, // The file descriptor we want memory mapped. We don't want a memory // mapped file, so 0 it is. 0, 0, ); if ptr.is_null() { // panic!("No memory allocated!"); return (ptr as *mut u8, 0); } self.pages += to_allocate / pagesize; self.growths += 1; (ptr as *mut u8, to_allocate) } } /// A raw allocator, capable of growing the heap, returning pointers to new /// allocations, and tracking and reusing freed memory. /// /// Note: It never returns memory to the OS; that is not implemented. /// /// This roughly corresponds to the /// [`AllocRef`](https://doc.rust-lang.org/nightly/core/alloc/trait.AllocRef.html) /// trait in Rust nightly, but does not directly implement that trait (although /// it probably could... TODO!) pub struct RawAlloc<G> { pub grower: G, pub blocks: BlockList, } impl<G: HeapGrower + Default> Default for RawAlloc<G> { fn default() -> Self { RawAlloc { grower: G::default(), blocks: BlockList::default(), } } } impl<G: HeapGrower> RawAlloc<G> { /// Create a new `RawAlloc` #[allow(dead_code)] pub fn new(grower: G) -> Self { RawAlloc { grower, blocks: BlockList::default(), } } /// Get statistics on this allocator, and verify validity of the BlockList pub fn stats(&self) -> (Validity, Stats) { self.blocks.stats() } /// Calculate the minimum size of a block to be allocated for the given layout. pub fn block_size(layout: Layout) -> usize { // We align everything to 16 bytes, and all blocks are at least 16 bytes. // Its pretty wasteful, but easy! let aligned_layout = layout .align_to(16) .expect("Whoa, serious memory issues") .pad_to_align(); // log::trace!( // "Alignment: {}@{} -> {}@{}", // layout.size(), // layout.align(), // aligned_layout.size(), // aligned_layout.align() // ); aligned_layout.size() } //////////////////////////////////////////////////////////// // Functions for implementing GlobalAlloc /// Allocate space for something fitting in layout /// /// # Safety /// /// This is very unsafe. See GlobalAlloc for details. pub unsafe fn alloc(&mut self, layout: Layout) -> *mut u8 { let needed_size = RawAlloc::<G>::block_size(layout); // log::trace!("Allocating {} bytes", needed_size); if let Some(range) = self.blocks.pop_size(needed_size) { // log::trace!("Popped off a block of size {} at {:?}", needed_size, ptr); return range.start.as_ptr(); } let (ptr, size) = self.grower.grow_heap(needed_size); // log::trace!("Grew to size {}", needed_size); if size == needed_size { // log::trace!(" exactly as needed"); return ptr; } let free_ptr = NonNull::new_unchecked(ptr.add(needed_size)); if size >= needed_size + HEADER_SIZE { // log::trace!("Adding block of size {}", size - needed_size); self.blocks.add_block(free_ptr, size - needed_size); } else if size > needed_size { // Uh-oh. We have a bit of extra free memory, but not enough to add // a header and call it a new free block. This could happen if our // page size was not a multiple of 16. Weird. // // Log it and leak it, I guess... log::warn!("Leaking {} bytes at {:?}", size - needed_size, free_ptr); } ptr } /// Deallocate (or "free") a memory block. /// /// # Safety /// /// This is very unsafe. See GlobalAlloc for details. pub unsafe fn dealloc(&mut self, ptr: *mut u8, layout: Layout) { let size = RawAlloc::<G>::block_size(layout); self.blocks.add_block(NonNull::new_unchecked(ptr), size); } } /// A thread-safe allocator, using a spin lock around a RawAlloc. /// /// Thread-safety is required for an allocator to be used as a global allocator, /// so that was easy to add with a spin lock. pub struct GenericAllocator<G> { // Values: // - 0: Untouched // - 1: Initialization in progress // - 2: Initialized init: AtomicU8, raw: MaybeUninit<Mutex<RawAlloc<G>>>, } impl<G: HeapGrower + Default> Default for GenericAllocator<G> { fn default() -> Self { Self::new() } } impl<G> GenericAllocator<G> { pub const fn new() -> Self { GenericAllocator { init: AtomicU8::new(0), raw: MaybeUninit::uninit(), } } } impl<G: HeapGrower + Default> GenericAllocator<G> { /// Get a reference to the underlying RawAlloc. /// /// # Safety /// /// This is unsafe because it blocks allocation while the mutex guard is in /// place. pub unsafe fn get_raw(&self) -> MutexGuard<RawAlloc<G>> { // First, we check // // The ordering here is SeqCst because that's the safest, if not the // most efficient. This could probably be downgraded, but would require // some analysis and understanding to do so. let mut state = self.init.compare_and_swap(0, 1, Ordering::SeqCst); // log::info!("state: {}", state); if state == 0 { // We haven't initialized, so we do that now. // We cast the raw pointer to be let raw_loc: *const Mutex<RawAlloc<G>> = self.raw.as_ptr(); let raw_mut: *mut Mutex<RawAlloc<G>> = raw_loc as *mut Mutex<RawAlloc<G>>; raw_mut.write(Mutex::new(RawAlloc::default())); let mx: &mut Mutex<RawAlloc<G>> = raw_mut.as_mut().unwrap(); // Let other threads know that the mutex and raw allocator are now initialized, // and they are free to use the mutex to access the raw allocator self.init.store(2, Ordering::SeqCst); return mx.lock(); } while state == 1 { // log::info!("Spinning!"); // Spin while we wait for the state to become 2 core::sync::atomic::spin_loop_hint(); state = self.init.load(Ordering::SeqCst); } let ptr = self.raw.as_ptr().as_ref().unwrap(); ptr.lock() } pub fn stats(&self) -> (Validity, Stats) { unsafe { self.get_raw().stats() } } } #[derive(Default)] pub struct UnixAllocator { alloc: GenericAllocator<UnixHeapGrower>, } impl UnixAllocator { pub const fn new() -> Self { UnixAllocator { alloc: GenericAllocator::new(), } } pub fn stats(&self) -> (Validity, Stats) { self.alloc.stats() } } unsafe impl GlobalAlloc for UnixAllocator { unsafe fn alloc(&self, layout: Layout) -> *mut u8 { self.alloc.get_raw().alloc(layout) } unsafe fn dealloc(&self, ptr: *mut u8, layout: Layout) { self.alloc.get_raw().dealloc(ptr, layout) } } pub struct ToyHeap { pub page_size: usize, pub size: usize, pub heap: [u8; 256 * 1024], } impl Default for ToyHeap { fn default() -> Self { ToyHeap { page_size: 64, size: 0, heap: [0; 256 * 1024], } } } impl HeapGrower for ToyHeap { unsafe fn grow_heap(&mut self, size: usize) -> (*mut u8, usize) { if self.size + size > self.heap.len() { return (null_mut(), 0); } let allocating = round_up(size, self.page_size); let ptr = self.heap.as_mut_ptr().add(self.size); self.size += allocating; (ptr, allocating) } } #[cfg(test)] mod tests { use super::*; use test_env_log::test; #[test] fn test_basic() { let toy_heap = ToyHeap::default(); let mut allocator = RawAlloc::new(toy_heap); const BLOCKS: usize = 3; let layouts: [Layout; BLOCKS] = [ Layout::from_size_align(64, 16).unwrap(), Layout::from_size_align(64, 16).unwrap(), Layout::from_size_align(224, 16).unwrap(), ]; let pointers: [*mut u8; BLOCKS] = unsafe { let mut pointers = [null_mut(); BLOCKS]; for (i, &l) in layouts.iter().enumerate() { pointers[i] = allocator.alloc(l); let (validity, _stats) = allocator.stats(); assert!(validity.is_valid()); } pointers }; for i in 0..BLOCKS - 1 { let l = layouts[i]; let expected = unsafe { pointers[i].add(l.size()) }; let found = pointers[i + 1]; assert_eq!(expected, found); } let toy_heap = &allocator.grower; let page_size = toy_heap.page_size; // Toy heap should be the same size as the blocks requested let total_allocated: usize = layouts.iter().map(|l| l.size()).sum(); let page_space = round_up(total_allocated, page_size); assert_eq!(toy_heap.size, page_space); //////////////////////////////////////////////////////////// // Deallocation // Deallocate the second pointer unsafe { allocator.dealloc(pointers[1], layouts[1]) }; let (validity, _stats) = allocator.stats(); assert!(validity.is_valid()); // Check that the block list is as expected assert!(allocator.blocks.first.is_some()); let first = allocator .blocks .first .as_mut() .expect("This should not be null"); assert_eq!(first.size(), layouts[1].size()); let next_exists = first.next().is_some(); log::info!("dealloc: {}", allocator.blocks); // We should still have the remainder left over from the last page // allocation assert!(next_exists); // The block list now has 1 64-byte block on it log::info!("post-alloc: {}", allocator.blocks); //////////////////////////////////////////////////////////// // Allocation with a block list unsafe { // Allocate 112 bytes, more than fits in the block on the block list let newp = allocator.alloc(Layout::from_size_align(112, 16).unwrap()); let (validity, _stats) = allocator.stats(); assert!(validity.is_valid()); assert_eq!( newp, pointers[2].add(round_up(layouts[2].size(), page_size)) ); log::info!("p112: {}", allocator.blocks); // Allocate 32 bytes, which should fit in the block let p32 = allocator.alloc(Layout::from_size_align(32, 16).unwrap()); let (validity, _stats) = allocator.stats(); assert!(validity.is_valid()); // The algorithm returns the second half of the block assert_eq!(p32, pointers[1].add(32)); // We should now still have 32 bytes in 1 block in the block list (plus page leftovers) // Allocate 8 bytes and another 16 bytes, which should both fit in the block // and completely consume it - because the 8 bytes should expand to 16 log::info!("p32: {}", allocator.blocks); let p8 = allocator.alloc(Layout::from_size_align(16, 4).unwrap()); let (validity, _stats) = allocator.stats(); assert!(validity.is_valid()); log::info!("p8: {}", allocator.blocks); let p16 = allocator.alloc(Layout::from_size_align(8, 1).unwrap()); let (validity, _stats) = allocator.stats(); assert!(validity.is_valid()); // The algorithm returns the second half of the block log::info!("p16: {}", allocator.blocks); assert_eq!(p8, pointers[1].add(16)); assert_eq!(p16, pointers[1]); log::info!("done: {}", allocator.blocks); }; } }