summaryrefslogtreecommitdiffstats
path: root/vespalib
Commit message (Collapse)AuthorAgeFilesLines
* Add NewestReplica equality tests and gmock matcher for distinct elements in ↵Tor Brede Vekterli2023-04-201-0/+35
| | | | | | | | | | | a range Add a new `matchers` directory in vespalib which can be used as a repository for generic, reusable GMock matchers. Move distributor tests from using an explicit gtest runner to using `GTest::gmock_main` which serves the same purpose. Need to depend on a gmock target (not just gtest) to be able to link with stuff required for matchers.
* Merge pull request #26721 from vespa-engine/toregge/remove-unused-variableHenning Baldersheim2023-04-121-1/+0
|\ | | | | Remove unused variable.
| * Remove unused variable.Tor Egge2023-04-121-1/+0
| |
* | Don't convert a call to a jump when this breaks reporting ofTor Egge2023-04-122-0/+9
|/ | | | stack frames.
* Remove array size from BufferFreeList.Tor Egge2023-04-1111-21/+16
|
* Adjust BufferState member function comments.Tor Egge2023-04-051-2/+2
|
* Change buffer state accounting from elements to entries.Tor Egge2023-04-0442-575/+548
|
* Rename member functions in BufferTypeBase to track renameTor Egge2023-04-043-5/+5
| | | | of member variables.
* Rename member variables in BufferTypeBase:Tor Egge2023-04-0415-112/+112
| | | | | | minArrays => min_entries maxArrays => max_entries numArraysForNewBuffer => num_entries_for_new_buffer
* Pass number of needed entries to DataStoreBase member functions thatTor Egge2023-04-046-44/+48
| | | | might switch primary buffer for a buffer type.
* Switch from elements to entries for raw alloator alloc parameter.Tor Egge2023-04-0412-32/+32
|
* Adjust ArrayStoreTypeMapper class comment.Tor Egge2023-04-031-1/+1
|
* Use ElemT as template parameter for ArrayStoreTest.Tor Egge2023-04-031-21/+21
|
* Use ElemT as template parameter for ArrayStoreSimpleTypeMapperTor Egge2023-04-032-5/+5
|
* Use ElemT as template parameter for BufferType, SmallArrayBufferType,Tor Egge2023-04-038-109/+112
| | | | LargeArrayBufferType and ArrayStore.
* Workarounds for libc++ 15 are still needed for libc++ 16.Tor Egge2023-04-032-2/+2
|
* Fix typo in optimized.hIkko Eltociear Ashimine2023-03-291-4/+4
| | | postion -> position
* Adjust explicit instantiations of BufferType and Array.Tor Egge2023-03-244-9/+11
|
* Merge pull request #26538 from ↵Tor Brede Vekterli2023-03-2312-95/+243
|\ | | | | | | | | vespa-engine/vekterli/add-capability-checks-to-state-api-handlers Add capability checking to state API handlers
| * Add capability checking to state API handlersTor Brede Vekterli2023-03-2212-95/+243
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This covers both the entry points from the `storagenode` and `searchnode` HTTP servers, though the former is mostly in the name of legacy support. Ideally, capability checking would exist as a property of the HTTP server (Portal) bindings, but the abstractions for the JSON request handling are sufficiently leaky that it ended up making more sense to push things further down the hierarchy. It's always a good thing to move away from using strings with implicit semantics as return types anyway. The `searchnode` state API handler mapping supports fine grained capabilities. The legacy `storagenode` state API forwarding does not; it uses a sledgehammer that expects the union of all possible API capability requirements.
* | Relax check for max prefix.Tor Egge2023-03-231-1/+1
|/
* Add utility wrapper around RE2 possible regex prefix match rangeTor Brede Vekterli2023-03-213-3/+57
| | | | | | For a strictly start-anchored regex, this provides a lower/upper bound pair that constrains the possible prefix range that may contain a string matching the regex.
* Revert "Add utility wrapper around RE2 possible regex prefix match range"Harald Musum2023-03-203-57/+3
|
* Merge pull request #26494 from ↵Tor Brede Vekterli2023-03-203-3/+57
|\ | | | | | | | | vespa-engine/vekterli/add-possible-match-prefix-regex-range-util Add utility wrapper around RE2 possible regex prefix match range
| * Test match range failure case with empty ranges returnedTor Brede Vekterli2023-03-201-0/+5
| |
| * Add utility wrapper around RE2 possible regex prefix match rangeTor Brede Vekterli2023-03-203-3/+52
| | | | | | | | | | | | For a strictly start-anchored regex, this provides a lower/upper bound pair that constrains the possible prefix range that may contain a string matching the regex.
* | Support gcc 11 std::string::size() not being constexpr.Henning Baldersheim2023-03-201-1/+1
| |
* | libstdc++ 11 lacks stream operators for duration. Use same workaroundTor Egge2023-03-192-2/+4
|/ | | | as for old libc++ versions.
* GC convenience constructor to be explicit.Henning Baldersheim2023-03-163-6/+1
|
* - GC unused _activeBuffers member.Henning Baldersheim2023-03-154-9/+8
| | | | - Just keep a very small ChangeVector, it might never be used.
* Merge pull request #26451 from ↵Henning Baldersheim2023-03-151-1/+1
|\ | | | | | | | | vespa-engine/toregge/factor-out-implementation-dependent-sizeof-std-string Factor out implementation dependent sizeof(std::string).
| * Factor out implementation dependent sizeof(std::string).Tor Egge2023-03-151-1/+1
| |
* | Merge pull request #26448 from vespa-engine/havardpe/verify-object-on-attachHåvard Pettersen2023-03-151-1/+4
|\ \ | |/ |/| verify object when attaching it to a handle
| * verify object when attaching it to a handleHåvard Pettersen2023-03-151-1/+4
| |
* | Merge pull request #26447 from ↵Henning Baldersheim2023-03-154-8/+11
|\ \ | | | | | | | | | | | | vespa-engine/toregge/expose-static-size-of-entry-ref-hold-list-deque-in-datastore-base Expose static size of entry ref hold list dequeue in DataStoreBase.
| * | Expose static size of entry ref hold list dequeue in DataStoreBase.Tor Egge2023-03-154-8/+11
| |/
* / Fix format strings.Tor Egge2023-03-151-2/+3
|/
* Assert expected guard value earlier in enable_ref_counted::internal_subrefTor Egge2023-03-141-1/+1
|
* Merge pull request #26441 from vespa-engine/havardpe/avoid-atomic-thread-fenceTor Egge2023-03-141-4/+6
|\ | | | | make TSAN happy (it does not support atomic thread fences)
| * make TSAN happy (it does not support atomic thread fences)Håvard Pettersen2023-03-141-4/+6
| |
* | Prefer std::vectorHenning Baldersheim2023-03-147-64/+2
| |
* | Merge pull request #26440 from ↵Henning Baldersheim2023-03-147-0/+144
|\ \ | | | | | | | | | | | | vespa-engine/havardpe/run-multiple-coroutines-concurrently-from-coroutine enable running sub-co-routines concurrently
| * | enable running sub-co-routines concurrentlyHåvard Pettersen2023-03-147-0/+144
| | |
* | | Merge pull request #26436 from ↵Henning Baldersheim2023-03-142-11/+0
|\ \ \ | | | | | | | | | | | | | | | | vespa-engine/balder/gc-dangerous-docstore-options-with-no-right-to-live summary.cache.initialentries and summary.log.chunk.skipcrconread does…
| * | | summary.cache.initialentries and summary.log.chunk.skipcrconread does not ↵Henning Baldersheim2023-03-142-11/+0
| | |/ | |/| | | | | | | have any purpose.
* / | Add noexceptHenning Baldersheim2023-03-143-56/+55
|/ /
* | Track static cache usageHenning Baldersheim2023-03-142-2/+7
| |
* | Wire in and test static memory usage for caches.Henning Baldersheim2023-03-143-8/+18
| |
* | Use std::vector instead of vespalib::ArrayHenning Baldersheim2023-03-141-6/+6
| |
* | GC unused includesHenning Baldersheim2023-03-141-1/+0
| |