From: Kirill Shcherbatov <kshcherbatov@tarantool.org> To: tarantool-patches@freelists.org, korablev@tarantool.org Cc: Kirill Shcherbatov <kshcherbatov@tarantool.org> Subject: [tarantool-patches] [PATCH v2 0/5] sql: do not use OP_Delete+OP_Insert for UPDATES Date: Sat, 29 Dec 2018 13:48:57 +0300 [thread overview] Message-ID: <cover.1546079994.git.kshcherbatov@tarantool.org> (raw) Introduced a new OP_Update opcode making Tarantool native Update operation. In case of UPDATE or REPLACE we can't use new OP_Update as it has a complex SQL-specific semantics: CREATE TABLE tj (s1 INT PRIMARY KEY, s2 INT); INSERT INTO tj VALUES (1, 3),(2, 4),(3,5); CREATE UNIQUE INDEX i ON tj (s2); SELECT * FROM tj; [1, 3], [2, 4], [3, 5] UPDATE OR REPLACE tj SET s2 = s2 + 1; SELECT * FROM tj; [1, 4], [3, 6] I.e. [1, 3] tuple is updated as [1, 4] and have replaced tuple [2, 4]. This logic is implemented as preventive tuples deletion by all corresponding indexes in SQL. The other significant change is forbidden primary key update. It was possible to deal with it the same way like with or REPLACE specifier but we need an atomic UPDATE step for #3691 ticket to support "or IGNORE/or ABORT/or FAIL" specifiers. Reworked tests to make testing avoiding primary key UPDATE where possible. Fixed bug in VDBE - sometimes temporal tuples in memory allocated with sql_vdbe_mem_alloc_region were stored in Mem variable having invalid flags set. Changes in version 2: - Reworked part of code to close tikets #3035, #3918: reused a new routine mpstream_encode_vdbe_mem to encode tuples on region everywhere on Vdbe execution: got rid of routines sqlite3VdbeMsgpackRecordLen and sqlite3VdbeMsgpackRecordPut that became useless. - Fixed few spell mistakes in comments and commit messages - Minor code changes: better mpstream methods names, changed OP_IdxUpdate to OP_Update name. Branch: http://github.com/tarantool/tarantool/tree/kshch/gh-3850-op-update Issue: https://github.com/tarantool/tarantool/issues/3850 Kirill Shcherbatov (5): sql: clean-up vdbe_emit_constraint_checks sql: fix sql_vdbe_mem_alloc_region result memory sql: fix fkey exception for self-referenced table sql: encode tuples with mpstream on Vdbe run sql: do not use OP_Delete+OP_Insert for UPDATES src/box/sql.c | 27 ++- src/box/sql/fkey.c | 58 ++++--- src/box/sql/insert.c | 28 +-- src/box/sql/sqliteInt.h | 14 ++ src/box/sql/update.c | 88 +++++++--- src/box/sql/vdbe.c | 138 +++++++++++++-- src/box/sql/vdbeInt.h | 34 +++- src/box/sql/vdbeaux.c | 74 +------- src/box/sql/vdbemem.c | 65 +++++++ src/mpstream.c | 30 ++++ src/mpstream.h | 11 ++ test/sql-tap/alter2.test.lua | 16 +- test/sql-tap/analyze9.test.lua | 43 ++++- test/sql-tap/bigrow1.test.lua | 12 +- test/sql-tap/check.test.lua | 14 +- test/sql-tap/fkey2.test.lua | 161 +++++++++--------- test/sql-tap/fkey3.test.lua | 25 ++- test/sql-tap/fkey4.test.lua | 32 ++-- test/sql-tap/gh2140-trans.test.lua | 26 +-- .../gh2250-trigger-chain-limit.test.lua | 8 +- test/sql-tap/gh2259-in-stmt-trans.test.lua | 12 +- test/sql-tap/identifier_case.test.lua | 2 +- test/sql-tap/index6.test.lua | 5 +- test/sql-tap/intpkey.test.lua | 23 ++- test/sql-tap/misc1.test.lua | 20 ++- test/sql-tap/quote.test.lua | 6 +- test/sql-tap/table.test.lua | 2 +- test/sql-tap/tkt-a8a0d2996a.test.lua | 4 +- test/sql-tap/tkt2767.test.lua | 16 +- test/sql-tap/tkt2832.test.lua | 10 +- test/sql-tap/tkt3554.test.lua | 15 +- test/sql-tap/trigger2.test.lua | 28 +-- test/sql-tap/trigger7.test.lua | 2 +- test/sql-tap/triggerB.test.lua | 28 +-- test/sql-tap/triggerC.test.lua | 43 ++--- test/sql-tap/triggerD.test.lua | 4 +- test/sql-tap/update.test.lua | 31 ++-- test/sql-tap/with1.test.lua | 24 +-- test/sql/collation.result | 12 +- test/sql/collation.test.lua | 12 +- test/sql/on-conflict.result | 8 +- test/sql/on-conflict.test.lua | 8 +- test/sql/row-count.result | 10 +- test/sql/row-count.test.lua | 10 +- test/sql/triggers.result | 16 +- test/sql/triggers.test.lua | 16 +- 46 files changed, 781 insertions(+), 490 deletions(-) -- 2.19.2
next reply other threads:[~2018-12-29 10:49 UTC|newest] Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top 2018-12-29 10:48 Kirill Shcherbatov [this message] 2018-12-29 10:48 ` [tarantool-patches] [PATCH v2 1/5] sql: clean-up vdbe_emit_constraint_checks Kirill Shcherbatov 2018-12-29 10:48 ` [tarantool-patches] [PATCH v2 2/5] sql: fix sql_vdbe_mem_alloc_region result memory Kirill Shcherbatov 2018-12-29 10:49 ` [tarantool-patches] [PATCH v2 3/5] sql: fix fkey exception for self-referenced table Kirill Shcherbatov 2018-12-29 13:26 ` [tarantool-patches] " n.pettik 2018-12-29 10:49 ` [tarantool-patches] [PATCH v2 4/5] sql: encode tuples with mpstream on Vdbe run Kirill Shcherbatov 2018-12-29 13:26 ` [tarantool-patches] " n.pettik 2018-12-29 15:28 ` Kirill Shcherbatov 2019-01-09 12:29 ` n.pettik 2018-12-29 10:49 ` [tarantool-patches] [PATCH v2 5/5] sql: do not use OP_Delete+OP_Insert for UPDATES Kirill Shcherbatov 2018-12-29 13:35 ` [tarantool-patches] " n.pettik 2018-12-29 15:31 ` Kirill Shcherbatov 2019-01-10 12:30 ` [tarantool-patches] Re: [PATCH v2 0/5] " Kirill Yukhin
Reply instructions: You may reply publicly to this message via plain-text email using any one of the following methods: * Save the following mbox file, import it into your mail client, and reply-to-all from there: mbox Avoid top-posting and favor interleaved quoting: https://en.wikipedia.org/wiki/Posting_style#Interleaved_style * Reply using the --to, --cc, and --in-reply-to switches of git-send-email(1): git send-email \ --in-reply-to=cover.1546079994.git.kshcherbatov@tarantool.org \ --to=kshcherbatov@tarantool.org \ --cc=korablev@tarantool.org \ --cc=tarantool-patches@freelists.org \ --subject='Re: [tarantool-patches] [PATCH v2 0/5] sql: do not use OP_Delete+OP_Insert for UPDATES' \ /path/to/YOUR_REPLY https://kernel.org/pub/software/scm/git/docs/git-send-email.html * If your mail client supports setting the In-Reply-To header via mailto: links, try the mailto: link
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox