Commit message (Collapse) | Author | Age | Files | Lines | |
---|---|---|---|---|---|
* | Merge pull request #31336 from ↵ | Tor Egge | 2024-05-29 | 5 | -66/+114 |
|\ | | | | | | | | | vespa-engine/toregge/factor-out-first-phase-rescorer-from-hit-collector Factor out FirstPhaseRescorer from HitCollector. | ||||
| * | Pass docid to rescorers. | Tor Egge | 2024-05-29 | 2 | -4/+5 |
| | | |||||
| * | Factor out FirstPhaseRescorer from HitCollector. | Tor Egge | 2024-05-29 | 5 | -66/+113 |
| | | |||||
* | | - Add is_search_multi_threaded to proton::matching::BlueprintBuilder for ↵ | Henning Baldersheim | 2024-05-29 | 1 | -1/+1 |
| | | | | | | | | | | | | consistency and readability. - Fix CreateBlueprintVisitorHelper to actually detect singlethreaded usage. | ||||
* | | - Separate priority Q functionality into a separate base class ↵ | Henning Baldersheim | 2024-05-28 | 10 | -49/+83 |
|/ | | | | | | | WeakAndPriorityQueue. - Keep thread safety in the SharedWeakAndpriorityQueue. - Only use thread safe variant if more than 1 thread in the search bundle. | ||||
* | Use shared heap | Henning Baldersheim | 2024-05-28 | 2 | -9/+16 |
| | |||||
* | Merge pull request #31282 from ↵ | Henning Baldersheim | 2024-05-28 | 20 | -192/+233 |
|\ | | | | | | | | | vespa-engine/balder/wire-parallel-heap-to-simple-weakand - Wire in a common shared heap for WeakAnd. | ||||
| * | Prepare heap correctly | Henning Baldersheim | 2024-05-24 | 4 | -10/+10 |
| | | |||||
| * | - Wire in a common shared heap for WeakAnd. | Henning Baldersheim | 2024-05-23 | 18 | -185/+226 |
| | | |||||
* | | Add firstPhaseRank feature (step 1, not yet integrated with match loop). | Tor Egge | 2024-05-24 | 11 | -0/+379 |
|/ | |||||
* | Pass all params in constructor and make members const. | Henning Baldersheim | 2024-05-22 | 3 | -20/+17 |
| | |||||
* | Split ParallellWandMatchParams in 2. | Henning Baldersheim | 2024-05-22 | 4 | -26/+36 |
| | |||||
* | It is already a unique_ptr | Henning Baldersheim | 2024-05-21 | 1 | -10/+7 |
| | |||||
* | Use std::forward recommended by clion. | Henning Baldersheim | 2024-05-21 | 2 | -14/+16 |
| | |||||
* | libc++ 18 still needs workarounds. | Tor Egge | 2024-05-21 | 1 | -1/+1 |
| | |||||
* | Use __attribute__((noinline)) instead of glibc macro __attribute_noinline__ | Tor Egge | 2024-05-20 | 2 | -4/+4 |
| | |||||
* | Remove empty comment. | Tor Egge | 2024-05-16 | 1 | -1/+0 |
| | |||||
* | Add second phase feature. | Tor Egge | 2024-05-16 | 6 | -0/+138 |
| | |||||
* | Avoid unresolved symbols when compiling without optimization. | Tor Egge | 2024-05-16 | 1 | -3/+3 |
| | |||||
* | Merge pull request #31210 from vespa-engine/balder/faster-int8float | Henning Baldersheim | 2024-05-16 | 7 | -23/+29 |
|\ | | | | | Speed up dotproduct for int8. | ||||
| * | Now there is preliminary native support for Int8Float too. | Henning Baldersheim | 2024-05-15 | 1 | -5/+8 |
| | | |||||
| * | Use a valuespace representable by both 8,16,23 and 64 bit floats to easier ↵ | Henning Baldersheim | 2024-05-15 | 1 | -3/+3 |
| | | | | | | | | verify correctness. | ||||
| * | Speed up dotproduct for int8. | Henning Baldersheim | 2024-05-15 | 6 | -15/+18 |
| | | |||||
* | | Merge pull request #31209 from ↵ | Håvard Pettersen | 2024-05-15 | 1 | -1/+1 |
|\ \ | |/ |/| | | | | | vespa-engine/geirst/attribute-weighted-set-blueprint-flow-stats Update flow stats to reflect that reverse hash lookup is always used … | ||||
| * | Update flow stats to reflect that reverse hash lookup is always used when ↵ | Geir Storli | 2024-05-15 | 1 | -1/+1 |
| | | | | | | | | non-strict. | ||||
* | | Use explicit static_cast | Henning Baldersheim | 2024-05-15 | 1 | -3/+3 |
| | | |||||
* | | Speed up bfloat16 to float conversion | Henning Baldersheim | 2024-05-14 | 1 | -4/+17 |
|/ | |||||
* | Merge pull request #31200 from vespa-engine/balder/add-distance-benchmark | Henning Baldersheim | 2024-05-14 | 16 | -38/+167 |
|\ | | | | | Add benchmark for distance functions | ||||
| * | Avoid possible compiler optimisation. | Henning Baldersheim | 2024-05-14 | 1 | -18/+32 |
| | | |||||
| * | Use use method without limit which already use hw optimized instructions. | Henning Baldersheim | 2024-05-14 | 1 | -8/+1 |
| | | |||||
| * | Add benchmark for distance functions | Henning Baldersheim | 2024-05-14 | 16 | -30/+152 |
| | | |||||
* | | Merge pull request #31201 from ↵ | Geir Storli | 2024-05-14 | 6 | -107/+14 |
|\ \ | | | | | | | | | | | | | vespa-engine/havardpe/remove-actual-cost-estimates remove code to estimate actual cost | ||||
| * | | remove code to estimate actual cost | Håvard Pettersen | 2024-05-14 | 6 | -107/+14 |
| | | | |||||
* | | | Merge pull request #31199 from ↵ | Geir Storli | 2024-05-14 | 1 | -2/+2 |
|\ \ \ | |_|/ |/| | | | | | | | | vespa-engine/toregge/adjust-invalid-feature-name-messages-when-setting-up-first-phase-and-second-phrase-rank-feature Adjust invalid feature name message when setting up first phase and | ||||
| * | | Adjust invalid feature name message when setting up first phase and | Tor Egge | 2024-05-14 | 1 | -2/+2 |
| | | | | | | | | | | | | second phase rank feature. | ||||
* | | | Drop short circuit optimisation as it prevents vectorisation. | Henning Baldersheim | 2024-05-13 | 2 | -3/+2 |
|/ / | |||||
* | | Merge pull request #31184 from vespa-engine/balder/unify-code | Henning Baldersheim | 2024-05-13 | 2 | -29/+19 |
|\ \ | | | | | | | Unify code | ||||
| * | | Unify code | Henning Baldersheim | 2024-05-13 | 2 | -29/+19 |
| | | | |||||
* | | | Use an assertless typify in innerloop | Henning Baldersheim | 2024-05-13 | 1 | -1/+1 |
| |/ |/| | |||||
* | | Avoid assert in inner loop. | Henning Baldersheim | 2024-05-13 | 1 | -1/+0 |
|/ | |||||
* | Merge pull request #31169 from ↵ | Geir Storli | 2024-05-10 | 2 | -9/+14 |
|\ | | | | | | | | | vespa-engine/toregge/use-btree-iterator-move-constructor Use btree iterator move constructor. | ||||
| * | Use btree iterator move constructor. | Tor Egge | 2024-05-10 | 2 | -9/+14 |
| | | |||||
* | | use least common term to drive strict phrase search | Håvard Pettersen | 2024-05-10 | 1 | -6/+4 |
|/ | | | | | | | | | | stop using beginId since it only makes sense for the first (or full) docid space remove misleading assert. We could still end up with an infinite loop without triggering the assert if the driving term matched the next document. Also, this assert is not used with other iterators that require strict children. | ||||
* | run non-strict source blender | Håvard Pettersen | 2024-05-07 | 1 | -30/+44 |
| | |||||
* | Remove obsolete workaround for indexing exact-match fields. | Tor Egge | 2024-05-06 | 2 | -3/+1 |
| | |||||
* | Merge pull request #31128 from ↵v8.340.17 | Geir Storli | 2024-05-06 | 3 | -24/+77 |
|\ | | | | | | | | | vespa-engine/havardpe/source-selector-iterator-benchmark baseline source blender integration in benchmark | ||||
| * | baseline source blender integration in benchmark | Håvard Pettersen | 2024-05-06 | 3 | -24/+77 |
| | | |||||
* | | Adjust down the estimate when its unknown and avoid WhiteListBlueprint being ↵ | Geir Storli | 2024-05-06 | 3 | -8/+15 |
|/ | | | | | | | strict. In cases where only attributes without fast-search are searched with e.g. an OR, this change avoids sorting the WhiteListBlueprint first under the root AND, which is very costly to evaluate. | ||||
* | test and adjust some stuff | Håvard Pettersen | 2024-05-03 | 6 | -7/+52 |
| | |||||
* | when is actual non-strict array lookup faster | Håvard Pettersen | 2024-05-02 | 1 | -11/+27 |
| | | | | ... than forced strict btree posting list |