-
Notifications
You must be signed in to change notification settings - Fork 4.1k
Bug #80818: default log-tc-size too small on POWER - 3 * ( 64K page size) minimium needed #80
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Conversation
size) minimium needed This is a slightly modified version of the patch contributed by Daniel Black, adding support for AArch64 in addition to POWER when choosing the page size for TC log.
Hi, thank you for your contribution. Please confirm this code is submitted under the terms of the OCA (Oracle's Contribution Agreement) you have previously signed by cutting and pasting the following text as a comment: |
I confirm the code being submitted is offered under the terms of the OCA, and that I am authorized to contribute it. |
Hi, thank you for your contribution. Your code has been assigned to an internal queue. Please follow |
I'm concerned that the defined value for two platforms is odd while the defined value for all the other platforms is a power of two. Is this a valid concern? Thanks |
It looks like @akopytov modified the value in the original patch provided by @grooverdan from 65536 to 65535 but, from what I can tell, there's no public indication of why the modification was made. It would be great to know if this was intentional. Thanks! |
@setharnold @tyhicks 65535 wasn't intentional. I don't remember whether that was a typo, or I just picked an early version of the patch by @grooverdan. Anyway, I noticed and fixed it later in my MySQL branch. |
Definitely typo - https://bugs.mysql.com/file.php?id=23825&bug_id=80818 I actually did 65536 and earlier I had 64*1024 Just yesterday I did ask Ubuntu (https://bugs.launchpad.net/ubuntu/+source/mysql-5.7/+bug/1706291) to carry the corrected patch of @akopytov or the backport in #160 |
* Only report unstable tests and don't fail the build * Show contents of log files after test failures * Add a --flakes option * Start a list of known flaky tests: flaky-tests.txt * Add more flaky tests from mysql#82 * One more flaky test * Fix and check the path to flaky-tests.txt * Ensure --flakes gets a non-empty arg Oddly, `--flakes` is ignored, as is `--flakes=""`, because getopt doesn't call the handler in either of those cases. However: `--flakes ""` will now trigger the `die` added here. The idea here is to catch the error corrected in fe801e631d5a. * Fix path for flaky tests file This fixes the path for the flaky tests, fixes the logic for reading the hash, decreases parallelism further and adds a bunch of tests to the flakes set. * Add flaky tests seen in mysql#80 --------- Co-authored-by: Patrick Reynolds <[email protected]> Signed-off-by: Dirkjan Bussink <[email protected]>
) * Truncate CONSTRAINT name until it fits in length Signed-off-by: Shlomi Noach <[email protected]> * replace first character with '_' Signed-off-by: Shlomi Noach <[email protected]> * adapt tests Signed-off-by: Shlomi Noach <[email protected]> * adapt check_constraints test Signed-off-by: Shlomi Noach <[email protected]> * use erase() and insert() Signed-off-by: Shlomi Noach <[email protected]> * typo Signed-off-by: Shlomi Noach <[email protected]> * stopping condition Signed-off-by: Shlomi Noach <[email protected]> * stopping condition Signed-off-by: Shlomi Noach <[email protected]> * calculate autogenerated FK name in sql_table Signed-off-by: Shlomi Noach <[email protected]> * new_name.length() Signed-off-by: Shlomi Noach <[email protected]> * consistent evaluation of new fk name Signed-off-by: Shlomi Noach <[email protected]> * consolidate abit of the code Signed-off-by: Shlomi Noach <[email protected]> * ensure UTF table name correctness Signed-off-by: Shlomi Noach <[email protected]> * early return condition Signed-off-by: Shlomi Noach <[email protected]> * Update sql/sql_table.cc Co-authored-by: Patrick Reynolds <[email protected]> * Update sql/sql_table.h Co-authored-by: Patrick Reynolds <[email protected]> --------- Signed-off-by: Shlomi Noach <[email protected]> Co-authored-by: Patrick Reynolds <[email protected]>
* Only report unstable tests and don't fail the build * Show contents of log files after test failures * Add a --flakes option * Start a list of known flaky tests: flaky-tests.txt * Add more flaky tests from mysql#82 * One more flaky test * Fix and check the path to flaky-tests.txt * Ensure --flakes gets a non-empty arg Oddly, `--flakes` is ignored, as is `--flakes=""`, because getopt doesn't call the handler in either of those cases. However: `--flakes ""` will now trigger the `die` added here. The idea here is to catch the error corrected in fe801e631d5a. * Fix path for flaky tests file This fixes the path for the flaky tests, fixes the logic for reading the hash, decreases parallelism further and adds a bunch of tests to the flakes set. * Add flaky tests seen in mysql#80 --------- Co-authored-by: Patrick Reynolds <[email protected]>
) * Truncate CONSTRAINT name until it fits in length Signed-off-by: Shlomi Noach <[email protected]> * replace first character with '_' Signed-off-by: Shlomi Noach <[email protected]> * adapt tests Signed-off-by: Shlomi Noach <[email protected]> * adapt check_constraints test Signed-off-by: Shlomi Noach <[email protected]> * use erase() and insert() Signed-off-by: Shlomi Noach <[email protected]> * typo Signed-off-by: Shlomi Noach <[email protected]> * stopping condition Signed-off-by: Shlomi Noach <[email protected]> * stopping condition Signed-off-by: Shlomi Noach <[email protected]> * calculate autogenerated FK name in sql_table Signed-off-by: Shlomi Noach <[email protected]> * new_name.length() Signed-off-by: Shlomi Noach <[email protected]> * consistent evaluation of new fk name Signed-off-by: Shlomi Noach <[email protected]> * consolidate abit of the code Signed-off-by: Shlomi Noach <[email protected]> * ensure UTF table name correctness Signed-off-by: Shlomi Noach <[email protected]> * early return condition Signed-off-by: Shlomi Noach <[email protected]> * Update sql/sql_table.cc Co-authored-by: Patrick Reynolds <[email protected]> * Update sql/sql_table.h Co-authored-by: Patrick Reynolds <[email protected]> --------- Signed-off-by: Shlomi Noach <[email protected]> Co-authored-by: Patrick Reynolds <[email protected]>
* Only report unstable tests and don't fail the build * Show contents of log files after test failures * Add a --flakes option * Start a list of known flaky tests: flaky-tests.txt * Add more flaky tests from mysql#82 * One more flaky test * Fix and check the path to flaky-tests.txt * Ensure --flakes gets a non-empty arg Oddly, `--flakes` is ignored, as is `--flakes=""`, because getopt doesn't call the handler in either of those cases. However: `--flakes ""` will now trigger the `die` added here. The idea here is to catch the error corrected in fe801e631d5a. * Fix path for flaky tests file This fixes the path for the flaky tests, fixes the logic for reading the hash, decreases parallelism further and adds a bunch of tests to the flakes set. * Add flaky tests seen in mysql#80 --------- Co-authored-by: Patrick Reynolds <[email protected]> Signed-off-by: Dirkjan Bussink <[email protected]>
) * Truncate CONSTRAINT name until it fits in length Signed-off-by: Shlomi Noach <[email protected]> * replace first character with '_' Signed-off-by: Shlomi Noach <[email protected]> * adapt tests Signed-off-by: Shlomi Noach <[email protected]> * adapt check_constraints test Signed-off-by: Shlomi Noach <[email protected]> * use erase() and insert() Signed-off-by: Shlomi Noach <[email protected]> * typo Signed-off-by: Shlomi Noach <[email protected]> * stopping condition Signed-off-by: Shlomi Noach <[email protected]> * stopping condition Signed-off-by: Shlomi Noach <[email protected]> * calculate autogenerated FK name in sql_table Signed-off-by: Shlomi Noach <[email protected]> * new_name.length() Signed-off-by: Shlomi Noach <[email protected]> * consistent evaluation of new fk name Signed-off-by: Shlomi Noach <[email protected]> * consolidate abit of the code Signed-off-by: Shlomi Noach <[email protected]> * ensure UTF table name correctness Signed-off-by: Shlomi Noach <[email protected]> * early return condition Signed-off-by: Shlomi Noach <[email protected]> * Update sql/sql_table.cc Co-authored-by: Patrick Reynolds <[email protected]> * Update sql/sql_table.h Co-authored-by: Patrick Reynolds <[email protected]> --------- Signed-off-by: Shlomi Noach <[email protected]> Co-authored-by: Patrick Reynolds <[email protected]>
* Only report unstable tests and don't fail the build * Show contents of log files after test failures * Add a --flakes option * Start a list of known flaky tests: flaky-tests.txt * Add more flaky tests from mysql#82 * One more flaky test * Fix and check the path to flaky-tests.txt * Ensure --flakes gets a non-empty arg Oddly, `--flakes` is ignored, as is `--flakes=""`, because getopt doesn't call the handler in either of those cases. However: `--flakes ""` will now trigger the `die` added here. The idea here is to catch the error corrected in fe801e631d5a. * Fix path for flaky tests file This fixes the path for the flaky tests, fixes the logic for reading the hash, decreases parallelism further and adds a bunch of tests to the flakes set. * Add flaky tests seen in mysql#80 --------- Co-authored-by: Patrick Reynolds <[email protected]> Signed-off-by: Dirkjan Bussink <[email protected]>
) * Truncate CONSTRAINT name until it fits in length Signed-off-by: Shlomi Noach <[email protected]> * replace first character with '_' Signed-off-by: Shlomi Noach <[email protected]> * adapt tests Signed-off-by: Shlomi Noach <[email protected]> * adapt check_constraints test Signed-off-by: Shlomi Noach <[email protected]> * use erase() and insert() Signed-off-by: Shlomi Noach <[email protected]> * typo Signed-off-by: Shlomi Noach <[email protected]> * stopping condition Signed-off-by: Shlomi Noach <[email protected]> * stopping condition Signed-off-by: Shlomi Noach <[email protected]> * calculate autogenerated FK name in sql_table Signed-off-by: Shlomi Noach <[email protected]> * new_name.length() Signed-off-by: Shlomi Noach <[email protected]> * consistent evaluation of new fk name Signed-off-by: Shlomi Noach <[email protected]> * consolidate abit of the code Signed-off-by: Shlomi Noach <[email protected]> * ensure UTF table name correctness Signed-off-by: Shlomi Noach <[email protected]> * early return condition Signed-off-by: Shlomi Noach <[email protected]> * Update sql/sql_table.cc Co-authored-by: Patrick Reynolds <[email protected]> * Update sql/sql_table.h Co-authored-by: Patrick Reynolds <[email protected]> --------- Signed-off-by: Shlomi Noach <[email protected]> Co-authored-by: Patrick Reynolds <[email protected]>
This is a slightly modified version of the patch contributed by Daniel
Black, adding support for AArch64 in addition to POWER when choosing the
page size for TC log.