Merge lp://qastaging/~stewart/percona-server/5.6-doublewrite into lp://qastaging/percona-server/5.6
Proposed by
Stewart Smith
Status: | Work in progress |
---|---|
Proposed branch: | lp://qastaging/~stewart/percona-server/5.6-doublewrite |
Merge into: | lp://qastaging/percona-server/5.6 |
Diff against target: |
1047 lines (+400/-55) 22 files modified
Percona-Server/mysql-test/suite/innodb/r/percona_innodb_doublewrite_file.result (+13/-0) Percona-Server/mysql-test/suite/innodb/t/percona_innodb_doublewrite_file-master.opt (+1/-0) Percona-Server/mysql-test/suite/innodb/t/percona_innodb_doublewrite_file.test (+12/-0) Percona-Server/mysql-test/suite/sys_vars/r/innodb_doublewrite_file_basic.result (+3/-0) Percona-Server/mysql-test/suite/sys_vars/t/innodb_doublewrite_file_basic-master.opt (+1/-0) Percona-Server/mysql-test/suite/sys_vars/t/innodb_doublewrite_file_basic.test (+2/-0) Percona-Server/storage/innobase/buf/buf0buf.cc (+2/-1) Percona-Server/storage/innobase/buf/buf0dblwr.cc (+68/-26) Percona-Server/storage/innobase/buf/buf0rea.cc (+3/-1) Percona-Server/storage/innobase/dict/dict0load.cc (+3/-2) Percona-Server/storage/innobase/fil/fil0fil.cc (+15/-14) Percona-Server/storage/innobase/fsp/fsp0fsp.cc (+3/-3) Percona-Server/storage/innobase/handler/ha_innodb.cc (+9/-0) Percona-Server/storage/innobase/include/mtr0log.ic (+3/-2) Percona-Server/storage/innobase/include/srv0srv.h (+2/-0) Percona-Server/storage/innobase/include/srv0start.h (+3/-0) Percona-Server/storage/innobase/include/trx0sys.h (+24/-0) Percona-Server/storage/innobase/include/trx0sys.ic (+35/-0) Percona-Server/storage/innobase/row/row0mysql.cc (+1/-1) Percona-Server/storage/innobase/srv/srv0srv.cc (+2/-0) Percona-Server/storage/innobase/srv/srv0start.cc (+163/-0) Percona-Server/storage/innobase/trx/trx0sys.cc (+32/-5) |
To merge this branch: | bzr merge lp://qastaging/~stewart/percona-server/5.6-doublewrite |
Related bugs: |
Reviewer | Review Type | Date Requested | Status |
---|---|---|---|
Laurynas Biveinis (community) | Needs Fixing | ||
Review via email:
|
Description of the change
Port innodb_
To post a comment you must log in.
Unmerged revisions
- 310. By Stewart Smith
-
Port innodb_
doublewrite_ file to Percona Server 5.6 This was a little bit involved as the doublewrite buffer code has been
refactored in MySQL 5.6 out into buf0dblwr.cc.One design difference in the 5.6 patch versus 5.5 and before is that where
InnoDB code has needed to be duplicated, I have instead refactored that code
to accept tablespace id as a parameter rather than copy&pasting it. With
this method, we should get merge conflicts if this code is changed upstream,
making it easier to ensure our end result code is correct. - 309. By Stewart Smith
-
merge 5.6.7
I have logged bugs for the review points that apply to 5.1 and 5.5 too.
I think that for 5.6 everything except for bug 1078181 must be fixed at the MP.
- The main issue with this feature in general is bug 1078181. doublewrite_ space() should be used everywhere it create_ low header should follow InnoDB coding style. SPACE_FIRST_ ID/SRV_ EXTRA_SYS_ SPACE_FIRST_ ID in several user_tablespace _id looks like UNIV_UNLIKELY (bug 1078206). dummy_create( ) is missing comments and misformatted in doublewrite_ space.
- The testcase must do some actual work with data pages. For
example, the workload must be larger than the log capacity. (bug
1078187).
- The sys_vars testcase should be a bit more involved. (bug 1078188).
- trx_sys_
applies. (bug 1078191).
- buf_dblwr_get space_id arg should have comment.
- buf_dblwr_
- s/SRV_LOG_
places (bug 1078201).
- Chunk at 410--411 looks wrong: fil_is_
a correct check already.
- Bogus UNIV_LIKELY/
- trx_sys_
trx0sys.h.
- For initalizations like at 255--261 and 296--302 personally I prefer
the ternary operator.
- trx_sys_sys_space has wrong header comment. "Check if a space is a
system tablespace", likewise for trx_sys_
- Style violation at line 766.
- Too long lines at 769+, 780, 822, 825+, 857, 874
- Spurious diff at lines 428, 436, 495, 507, 576.