| 1 | /* |
| 2 | * Copyright (C) 2014, 2015 Apple Inc. All rights reserved. |
| 3 | * |
| 4 | * Redistribution and use in source and binary forms, with or without |
| 5 | * modification, are permitted provided that the following conditions |
| 6 | * are met: |
| 7 | * 1. Redistributions of source code must retain the above copyright |
| 8 | * notice, this list of conditions and the following disclaimer. |
| 9 | * 2. Redistributions in binary form must reproduce the above copyright |
| 10 | * notice, this list of conditions and the following disclaimer in the |
| 11 | * documentation and/or other materials provided with the distribution. |
| 12 | * |
| 13 | * THIS SOFTWARE IS PROVIDED BY APPLE INC. ``AS IS'' AND ANY |
| 14 | * EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE |
| 15 | * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR |
| 16 | * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL APPLE INC. OR |
| 17 | * CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, |
| 18 | * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, |
| 19 | * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR |
| 20 | * PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY |
| 21 | * OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT |
| 22 | * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE |
| 23 | * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. |
| 24 | */ |
| 25 | |
| 26 | #include "config.h" |
| 27 | #include "DFGStructureAbstractValue.h" |
| 28 | |
| 29 | #if ENABLE(DFG_JIT) |
| 30 | |
| 31 | #include "DFGGraph.h" |
| 32 | |
| 33 | namespace JSC { namespace DFG { |
| 34 | |
| 35 | #if !ASSERT_DISABLED |
| 36 | void StructureAbstractValue::assertIsRegistered(Graph& graph) const |
| 37 | { |
| 38 | if (isTop()) |
| 39 | return; |
| 40 | |
| 41 | for (unsigned i = size(); i--;) |
| 42 | graph.assertIsRegistered(at(i).get()); |
| 43 | } |
| 44 | #endif // !ASSERT_DISABLED |
| 45 | |
| 46 | void StructureAbstractValue::clobber() |
| 47 | { |
| 48 | // The premise of this approach to clobbering is that anytime we introduce |
| 49 | // a watchable structure into an abstract value, we watchpoint it. You can assert |
| 50 | // that this holds by calling assertIsWatched(). |
| 51 | |
| 52 | if (isTop()) |
| 53 | return; |
| 54 | |
| 55 | setClobbered(true); |
| 56 | |
| 57 | if (m_set.isThin()) { |
| 58 | if (!m_set.singleEntry()) |
| 59 | return; |
| 60 | if (!m_set.singleEntry()->dfgShouldWatch()) |
| 61 | makeTopWhenThin(); |
| 62 | return; |
| 63 | } |
| 64 | |
| 65 | RegisteredStructureSet::OutOfLineList* list = m_set.list(); |
| 66 | for (unsigned i = list->m_length; i--;) { |
| 67 | if (!list->list()[i]->dfgShouldWatch()) { |
| 68 | makeTop(); |
| 69 | return; |
| 70 | } |
| 71 | } |
| 72 | } |
| 73 | |
| 74 | void StructureAbstractValue::observeTransition(RegisteredStructure from, RegisteredStructure to) |
| 75 | { |
| 76 | ASSERT(!from->dfgShouldWatch()); |
| 77 | |
| 78 | if (isTop()) |
| 79 | return; |
| 80 | |
| 81 | if (!m_set.contains(from)) |
| 82 | return; |
| 83 | |
| 84 | if (!m_set.add(to)) |
| 85 | return; |
| 86 | |
| 87 | if (m_set.size() > polymorphismLimit) |
| 88 | makeTop(); |
| 89 | } |
| 90 | |
| 91 | void StructureAbstractValue::observeTransitions(const TransitionVector& vector) |
| 92 | { |
| 93 | if (isTop()) |
| 94 | return; |
| 95 | |
| 96 | RegisteredStructureSet newStructures; |
| 97 | for (unsigned i = vector.size(); i--;) { |
| 98 | ASSERT(!vector[i].previous->dfgShouldWatch()); |
| 99 | |
| 100 | if (!m_set.contains(vector[i].previous)) |
| 101 | continue; |
| 102 | |
| 103 | newStructures.add(vector[i].next); |
| 104 | } |
| 105 | |
| 106 | if (!m_set.merge(newStructures)) |
| 107 | return; |
| 108 | |
| 109 | if (m_set.size() > polymorphismLimit) |
| 110 | makeTop(); |
| 111 | } |
| 112 | |
| 113 | bool StructureAbstractValue::add(RegisteredStructure structure) |
| 114 | { |
| 115 | if (isTop()) |
| 116 | return false; |
| 117 | |
| 118 | if (!m_set.add(structure)) |
| 119 | return false; |
| 120 | |
| 121 | if (m_set.size() > polymorphismLimit) |
| 122 | makeTop(); |
| 123 | |
| 124 | return true; |
| 125 | } |
| 126 | |
| 127 | bool StructureAbstractValue::merge(const RegisteredStructureSet& other) |
| 128 | { |
| 129 | if (isTop()) |
| 130 | return false; |
| 131 | |
| 132 | return mergeNotTop(other); |
| 133 | } |
| 134 | |
| 135 | bool StructureAbstractValue::mergeSlow(const StructureAbstractValue& other) |
| 136 | { |
| 137 | // It isn't immediately obvious that the code below is doing the right thing, so let's go |
| 138 | // through it. |
| 139 | // |
| 140 | // This not clobbered, other not clobbered: Clearly, we don't want to make anything clobbered |
| 141 | // since we just have two sets and we are merging them. mergeNotTop() can handle this just |
| 142 | // fine. |
| 143 | // |
| 144 | // This clobbered, other clobbered: Clobbered means that we have a set of things, plus we |
| 145 | // temporarily have the set of all things but the latter will go away once we hit the next |
| 146 | // invalidation point. This allows us to merge two clobbered sets the natural way. For now |
| 147 | // the set will still be TOP (and so we keep the clobbered bit set), but we know that after |
| 148 | // invalidation, we will have the union of the this and other. |
| 149 | // |
| 150 | // This clobbered, other not clobbered: It's safe to merge in other for both before and after |
| 151 | // invalidation, so long as we leave the clobbered bit set. Before invalidation this has no |
| 152 | // effect since the set will still appear to have all things in it. The way to think about |
| 153 | // what invalidation would do is imagine if we had a set A that was clobbered and a set B |
| 154 | // that wasn't and we considered the following two cases. Note that we expect A to be the |
| 155 | // same at the end in both cases: |
| 156 | // |
| 157 | // A.merge(B) InvalidationPoint |
| 158 | // InvalidationPoint A.merge(B) |
| 159 | // |
| 160 | // The fact that we expect A to be the same in both cases means that we want to merge other |
| 161 | // into this but keep the clobbered bit. |
| 162 | // |
| 163 | // This not clobbered, other clobbered: This is just the converse of the previous case. We |
| 164 | // want to merge other into this and set the clobbered bit. |
| 165 | |
| 166 | bool changed = false; |
| 167 | |
| 168 | if (!isClobbered() && other.isClobbered()) { |
| 169 | setClobbered(true); |
| 170 | changed = true; |
| 171 | } |
| 172 | |
| 173 | changed |= mergeNotTop(other.m_set); |
| 174 | |
| 175 | return changed; |
| 176 | } |
| 177 | |
| 178 | bool StructureAbstractValue::mergeNotTop(const RegisteredStructureSet& other) |
| 179 | { |
| 180 | if (!m_set.merge(other)) |
| 181 | return false; |
| 182 | |
| 183 | if (m_set.size() > polymorphismLimit) |
| 184 | makeTop(); |
| 185 | |
| 186 | return true; |
| 187 | } |
| 188 | |
| 189 | void StructureAbstractValue::filter(const RegisteredStructureSet& other) |
| 190 | { |
| 191 | if (isTop()) { |
| 192 | m_set = other; |
| 193 | return; |
| 194 | } |
| 195 | |
| 196 | if (isClobbered()) { |
| 197 | // We have two choices here: |
| 198 | // |
| 199 | // Do nothing: It's legal to keep our set intact, which would essentially mean that for |
| 200 | // now, our set would behave like TOP but after the next invalidation point it wold be |
| 201 | // a finite set again. This may be a good choice if 'other' is much bigger than our |
| 202 | // m_set. |
| 203 | // |
| 204 | // Replace m_set with other and clear the clobber bit: This is also legal, and means that |
| 205 | // we're no longer clobbered. This is usually better because it immediately gives us a |
| 206 | // smaller set. |
| 207 | // |
| 208 | // This scenario should come up rarely. We usually don't do anything to an abstract value |
| 209 | // after it is clobbered. But we apply some heuristics. |
| 210 | |
| 211 | if (other.size() > m_set.size() + clobberedSupremacyThreshold) |
| 212 | return; // Keep the clobbered set. |
| 213 | |
| 214 | m_set = other; |
| 215 | setClobbered(false); |
| 216 | return; |
| 217 | } |
| 218 | |
| 219 | m_set.filter(other); |
| 220 | } |
| 221 | |
| 222 | void StructureAbstractValue::filter(const StructureAbstractValue& other) |
| 223 | { |
| 224 | if (other.isTop()) |
| 225 | return; |
| 226 | |
| 227 | if (other.isClobbered()) { |
| 228 | if (isTop()) |
| 229 | return; |
| 230 | |
| 231 | if (!isClobbered()) { |
| 232 | // See justification in filter(const RegisteredStructureSet&), above. An unclobbered set is |
| 233 | // almost always better. |
| 234 | if (m_set.size() > other.m_set.size() + clobberedSupremacyThreshold) |
| 235 | *this = other; // Keep the clobbered set. |
| 236 | return; |
| 237 | } |
| 238 | |
| 239 | m_set.filter(other.m_set); |
| 240 | return; |
| 241 | } |
| 242 | |
| 243 | filter(other.m_set); |
| 244 | } |
| 245 | |
| 246 | void StructureAbstractValue::filterSlow(SpeculatedType type) |
| 247 | { |
| 248 | if (!(type & SpecCell)) { |
| 249 | clear(); |
| 250 | return; |
| 251 | } |
| 252 | |
| 253 | ASSERT(!isTop()); |
| 254 | |
| 255 | m_set.genericFilter( |
| 256 | [&] (RegisteredStructure structure) { |
| 257 | return !!(speculationFromStructure(structure.get()) & type); |
| 258 | }); |
| 259 | } |
| 260 | |
| 261 | void StructureAbstractValue::filterClassInfoSlow(const ClassInfo* classInfo) |
| 262 | { |
| 263 | ASSERT(!isTop()); |
| 264 | m_set.genericFilter( |
| 265 | [&] (RegisteredStructure structure) { |
| 266 | return structure->classInfo()->isSubClassOf(classInfo); |
| 267 | }); |
| 268 | } |
| 269 | |
| 270 | bool StructureAbstractValue::contains(RegisteredStructure structure) const |
| 271 | { |
| 272 | if (isInfinite()) |
| 273 | return true; |
| 274 | |
| 275 | return m_set.contains(structure); |
| 276 | } |
| 277 | |
| 278 | bool StructureAbstractValue::contains(Structure* structure) const |
| 279 | { |
| 280 | if (isInfinite()) |
| 281 | return true; |
| 282 | |
| 283 | return m_set.toStructureSet().contains(structure); |
| 284 | } |
| 285 | |
| 286 | bool StructureAbstractValue::isSubsetOf(const RegisteredStructureSet& other) const |
| 287 | { |
| 288 | if (isInfinite()) |
| 289 | return false; |
| 290 | |
| 291 | return m_set.isSubsetOf(other); |
| 292 | } |
| 293 | |
| 294 | bool StructureAbstractValue::isSubsetOf(const StructureAbstractValue& other) const |
| 295 | { |
| 296 | if (isTop()) |
| 297 | return false; |
| 298 | |
| 299 | if (other.isTop()) |
| 300 | return true; |
| 301 | |
| 302 | if (isClobbered() == other.isClobbered()) |
| 303 | return m_set.isSubsetOf(other.m_set); |
| 304 | |
| 305 | // Here it gets tricky. If in doubt, return false! |
| 306 | |
| 307 | if (isClobbered()) |
| 308 | return false; // A clobbered set is never a subset of an unclobbered set. |
| 309 | |
| 310 | // An unclobbered set is currently a subset of a clobbered set, but it may not be so after |
| 311 | // invalidation. |
| 312 | return m_set.isSubsetOf(other.m_set); |
| 313 | } |
| 314 | |
| 315 | bool StructureAbstractValue::isSupersetOf(const RegisteredStructureSet& other) const |
| 316 | { |
| 317 | if (isInfinite()) |
| 318 | return true; |
| 319 | |
| 320 | return m_set.isSupersetOf(other); |
| 321 | } |
| 322 | |
| 323 | bool StructureAbstractValue::overlaps(const RegisteredStructureSet& other) const |
| 324 | { |
| 325 | if (isInfinite()) |
| 326 | return true; |
| 327 | |
| 328 | return m_set.overlaps(other); |
| 329 | } |
| 330 | |
| 331 | bool StructureAbstractValue::overlaps(const StructureAbstractValue& other) const |
| 332 | { |
| 333 | if (other.isInfinite()) |
| 334 | return true; |
| 335 | |
| 336 | return overlaps(other.m_set); |
| 337 | } |
| 338 | |
| 339 | bool StructureAbstractValue::isSubClassOf(const ClassInfo* classInfo) const |
| 340 | { |
| 341 | if (isInfinite()) |
| 342 | return false; |
| 343 | |
| 344 | // Note taht this function returns true if the structure set is empty. |
| 345 | for (const RegisteredStructure structure : m_set) { |
| 346 | if (!structure->classInfo()->isSubClassOf(classInfo)) |
| 347 | return false; |
| 348 | } |
| 349 | return true; |
| 350 | } |
| 351 | |
| 352 | bool StructureAbstractValue::equalsSlow(const StructureAbstractValue& other) const |
| 353 | { |
| 354 | ASSERT(m_set.m_pointer != other.m_set.m_pointer); |
| 355 | ASSERT(!isTop()); |
| 356 | ASSERT(!other.isTop()); |
| 357 | |
| 358 | return m_set == other.m_set |
| 359 | && isClobbered() == other.isClobbered(); |
| 360 | } |
| 361 | |
| 362 | void StructureAbstractValue::dumpInContext(PrintStream& out, DumpContext* context) const |
| 363 | { |
| 364 | if (isClobbered()) |
| 365 | out.print("Clobbered:" ); |
| 366 | |
| 367 | if (isTop()) |
| 368 | out.print("TOP" ); |
| 369 | else |
| 370 | out.print(inContext(m_set.toStructureSet(), context)); |
| 371 | } |
| 372 | |
| 373 | void StructureAbstractValue::dump(PrintStream& out) const |
| 374 | { |
| 375 | dumpInContext(out, 0); |
| 376 | } |
| 377 | |
| 378 | void StructureAbstractValue::validateReferences(const TrackedReferences& trackedReferences) const |
| 379 | { |
| 380 | if (isTop()) |
| 381 | return; |
| 382 | m_set.validateReferences(trackedReferences); |
| 383 | } |
| 384 | |
| 385 | } } // namespace JSC::DFG |
| 386 | |
| 387 | #endif // ENABLE(DFG_JIT) |
| 388 | |
| 389 | |