Commit Graph

20 Commits

Author SHA1 Message Date
Alexander Kuzmenkov
59ed586b69 fixup 2020-08-03 20:15:08 +03:00
Alexander Kuzmenkov
5214621830 [draft] only fix int16->int32 insert in PODArray 2020-07-30 18:20:47 +03:00
Alexey Milovidov
0c237de00d Fix build 2020-07-15 00:32:14 +03:00
Alexey Milovidov
befbae585b Revert the only change that should affect performance in gcc 2020-07-12 21:31:17 +03:00
Alexey Milovidov
70aae6eb44 Fix error 2020-07-09 04:21:25 +03:00
Alexey Milovidov
c4e925434a Fix build 2020-07-08 06:29:54 +03:00
Alexey Milovidov
a6dbe421ac Fix build 2020-07-08 05:31:29 +03:00
Alexey Milovidov
491bd44955 Improve performance of clang build 2020-07-08 04:41:09 +03:00
Alexey Milovidov
3471e526ff Improvements in PODArray 2020-07-08 04:26:44 +03:00
Alexey Milovidov
671c544b7d Improvements in PODArray 2020-07-08 04:24:38 +03:00
Alexey Milovidov
c34f3254f2 Improvements in PODArray 2020-07-08 04:21:39 +03:00
alexey-milovidov
c256682068
Update PODArray.h 2020-06-20 10:45:51 +03:00
alexey-milovidov
2e604726e0
Update PODArray.h 2020-06-20 08:39:52 +03:00
maxim-babenko
c3dbf7d6bf
Fix PODArrayBase::push_back_raw_many
1) capacity check must be ready for overflow during memcpy
2) reserve must be called with the new size, not just delta
2020-06-17 19:38:23 +03:00
Alexander Kuzmenkov
83aaea7b0f Merge remote-tracking branch 'origin/master' into HEAD 2020-05-26 18:29:42 +03:00
Vasily Nemkov
28a2cc0a82 Functions {to,from}UnixTimestamp64{Milli,Micro,Nano}
* Family of functions to convert DateTime64 to Int64 and back, usefull for
   imporintg/exporting raw timestamp values of certain precision.
 * Docs
 * Tests
 * Fixed ya.make
2020-05-26 16:42:22 +03:00
Alexander Kuzmenkov
9d25ae22b5 Merge remote-tracking branch 'origin/master' into HEAD 2020-05-26 09:07:01 +03:00
Alexey Milovidov
55a02c391d Style 2020-05-26 01:06:02 +03:00
Alexander Kuzmenkov
50df893dc6 Assert that allocator & container are consistent wrt. inline memory
There are lots of places where HashTable is used with
AllocatorWithStackMemory, but the size of allocator inline memory is set
incorrectly, and it's less than the initial HashTable buffer size.
Because of this, HashTable is always allocated on the heap, and the
inline memory becomes a useless dead weight.

For PODArray, we have previously added a helper template that makes sure
these values are in sync, so there was only one such discrepancy left,
in the unit test.
2020-05-20 22:37:13 +03:00
Ivan Lezhankin
06446b4f08 dbms/ → src/ 2020-04-03 18:14:31 +03:00