- developers - lists.mariadb.org

Re: [Maria-developers] 581857bd999: MDEV-22313: Neither SHOW CREATE USER nor SHOW GRANTS prints a user's default role
by Sergei Golubchik 23 Oct '20

23 Oct '20

Hi, Anel! Looks good, thanks! A couple of small comments below. On Oct 23, Anel Husakovic wrote: > revision-id: 581857bd999 (mariadb-10.1.43-314-g581857bd999) let's fix it in 10.2 and up, we're doing the very last 10.1 release, so better be extra-super-conservative about it. > parent(s): 43ec9370b32 > author: Anel Husakovic <anel(a)mariadb.org> > committer: Anel Husakovic <anel(a)mariadb.org> > timestamp: 2020-10-23 00:20:53 +0200 > message: > > MDEV-22313: Neither SHOW CREATE USER nor SHOW GRANTS prints a user's default role > > diff --git a/mysql-test/t/grant5.test b/mysql-test/t/grant5.test > index 74a69952124..bf97a930097 100644 > --- a/mysql-test/t/grant5.test > +++ b/mysql-test/t/grant5.test > @@ -52,6 +52,23 @@ disconnect u1; > drop user u1@localhost; > drop database mysqltest1; > > +# > +# MDEV-22313: Neither SHOW CREATE USER nor SHOW GRANTS prints a user's default role > +# > +CREATE ROLE test_role; > +CREATE USER test_user; > +GRANT test_role TO test_user; > +SET DEFAULT ROLE test_role FOR test_user; > +SHOW GRANTS FOR test_user; > +SET DEFAULT ROLE NONE for test_user; > +SHOW GRANTS FOR test_user; > +SET DEFAULT ROLE test_role; oops, sorry, my mistake. Please, add `SET ROLE test_role` before SHOW GRANTS. The goal is to see that SET DEFAULT ROLE is shown *after* all grants to root@localhost, but before any grants to test_role. > +SHOW GRANTS; > +SET DEFAULT ROLE NONE; > +SHOW GRANTS; > +DROP USER test_user; > +DROP ROLE test_role; > + > # > # End of 10.1 tests > # > diff --git a/sql/sql_acl.cc b/sql/sql_acl.cc > index cf0b1d87bd7..6c852976061 100644 > --- a/sql/sql_acl.cc > +++ b/sql/sql_acl.cc > @@ -8001,15 +8007,44 @@ static ROLE_GRANT_PAIR *find_role_grant_pair(const LEX_STRING *u, > my_hash_search(&acl_roles_mappings, (uchar*)pair_key.ptr(), key_length); > } > > +static bool show_default_role(THD *thd, const char *hostname, > + ACL_USER *acl_entry, char *buff, size_t buffsize) > +{ > + Protocol *protocol= thd->protocol; > + LEX_STRING def_rolename= acl_entry->default_rolename; > + > + if (def_rolename.length) > + { > + String def_str(buff, buffsize, system_charset_info); > + def_str.length(0); > + def_str.append(STRING_WITH_LEN("SET DEFAULT ROLE ")); > + def_str.append(&def_rolename); > + def_str.append(" FOR '"); > + def_str.append(&acl_entry->user); > + DBUG_ASSERT(!(acl_entry->flags & IS_ROLE)); > + def_str.append(STRING_WITH_LEN("'@'")); > + def_str.append(acl_entry->host.hostname, acl_entry->hostname_length, > + system_charset_info); > + def_str.append('\''); > + protocol->prepare_for_resend(); > + protocol->store(def_str.ptr(),def_str.length(),def_str.charset()); > + if (protocol->write()) > + { > + return TRUE; > + } > + } > + return FALSE; looks good. The hostname argument seems to be unused now, doesn't it? > +} > + > -static bool show_role_grants(THD *thd, const char *username, > - const char *hostname, ACL_USER_BASE *acl_entry, > +static bool show_role_grants(THD *thd, const char *hostname, > + ACL_USER_BASE *acl_entry, > char *buff, size_t buffsize) > { > uint counter; > Protocol *protocol= thd->protocol; > LEX_STRING host= {const_cast<char*>(hostname), strlen(hostname)}; > > - String grant(buff,sizeof(buff),system_charset_info); > + String grant(buff, buffsize, system_charset_info); ouch, indeed. thanks! > for (counter= 0; counter < acl_entry->role_grants.elements; counter++) > { > grant.length(0); Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

1 0

Re: [Maria-developers] 01612f594b3: MDEV-23769: MTR can abort before it prints the test result summary
by Sergei Golubchik 22 Oct '20

22 Oct '20

Hi, Anel! On Oct 22, Anel Husakovic wrote: > Hi Sergei, > thanks for the review. > > Having warning/error is a necessary condition. > Sufficient condition is to call at least 2 tests. > Incorrect behaviour can be seen if: > - Both tests are correct but there is error generated (incorrect option > passed): > So this `./mysql-test/mtr --mysqld=--lock-wait-timeout=-xx 1st` will > generate and write statistic at the end, > but this `./mysql-test/mtr --mysqld=--lock-wait-timeout=-xx 1st 1st > --force` will not generate the statistic. > Note: warning is generated (--mysqld=--lock-wait-timeout=-1) statistic > will be generated Okay, thanks. This would be useful to show in the commit comment, that one can use ./mtr --mysqld=--lock-wait-timeout=-xx 1st 1st --force --parallel 1 to repeat the error. But then I don't understand why this happens. Why one failure is not enough? Why --parallel 2 destroys the effect? Apparently, it's check_warnings_post_shutdown() that causes that, not exit(1). One can change to exit(0) or remove exit(1) completely - the bug will still happen. Why check_warnings_post_shutdown() writing to the master server causes this bug? It looks rather fishy that in your fix it's the worker process that prints the statistics, not the master. Btw, note that check_warnings_post_shutdown() doesn't have any return statement, it very implicitly return "number of bytes sent" - it would be easier to read if it'd return something explicitly. Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

1 0

Re: [Maria-developers] 99c14476251: MDEV-22313: Neither SHOW CREATE USER nor SHOW GRANTS prints a user's default role
by Sergei Golubchik 22 Oct '20

22 Oct '20

Hi, Anel! On Oct 22, Anel Husakovic wrote: > revision-id: 99c14476251 (mariadb-10.1.43-314-g99c14476251) > parent(s): 43ec9370b32 > author: Anel Husakovic <anel(a)mariadb.org> > committer: Anel Husakovic <anel(a)mariadb.org> > timestamp: 2020-10-22 16:49:28 +0200 > message: > > MDEV-22313: Neither SHOW CREATE USER nor SHOW GRANTS prints a user's default role > > diff --git a/mysql-test/t/grant5.test b/mysql-test/t/grant5.test > index 74a69952124..8756648c3e1 100644 > --- a/mysql-test/t/grant5.test > +++ b/mysql-test/t/grant5.test > @@ -53,5 +53,19 @@ drop user u1@localhost; > drop database mysqltest1; > > # > -# End of 10.1 tests > +# MDEV-22313: Neither SHOW CREATE USER nor SHOW GRANTS prints a user's default role > +# > +CREATE ROLE test_role; > +CREATE USER test_user; > +GRANT test_role TO test_user; > +SET DEFAULT ROLE test_role FOR test_user; > +SHOW GRANTS FOR test_user; > +SET DEFAULT ROLE NONE for test_user; > +SHOW GRANTS FOR test_user; > +SHOW GRANTS; The point is to run `SHOW GRANTS` for a user with a default role. E.g. SET DEFAULT ROLE test_role; SHOW GRANTS; SET DEFAULT ROLE NONE; > +DROP USER test_user; > +DROP ROLE test_role; > + > # > +# End of 10.1 tests > +# > \ No newline at end of file new line at the end of file? > diff --git a/sql/sql_acl.cc b/sql/sql_acl.cc > index cf0b1d87bd7..f42f7bdcfaf 100644 > --- a/sql/sql_acl.cc > +++ b/sql/sql_acl.cc > @@ -8001,6 +8007,36 @@ static ROLE_GRANT_PAIR *find_role_grant_pair(const LEX_STRING *u, > my_hash_search(&acl_roles_mappings, (uchar*)pair_key.ptr(), key_length); > } > > +static bool show_default_role(THD *thd, const char *hostname, > + ACL_USER *acl_entry, char *buff) > +{ > + Protocol *protocol= thd->protocol; > + LEX_STRING host= {const_cast<char*>(hostname), strlen(hostname)}; > + > + LEX_STRING def_rolename= acl_entry->default_rolename; > + if (def_rolename.length) > + { > + String def_str(buff,sizeof(buff),system_charset_info); eh? sizeof(buff) is sizeof(char*) that is usually 8 on 64-bit, 4 on 32-bit architecture. You want the size of the buffer here, not the size of the pointer to the buffer. > + def_str.length(0); > + def_str.append(STRING_WITH_LEN("SET DEFAULT ROLE ")); > + def_str.append(&def_rolename); > + def_str.append(" FOR '"); > + def_str.append(acl_entry->user.str, acl_entry->user.length, > + system_charset_info); again, just as above, you can use def_str.append(&acl_entry->user); > + DBUG_ASSERT(!(acl_entry->flags & IS_ROLE)); > + def_str.append(STRING_WITH_LEN("'@'")); > + def_str.append(&host); why not to use def_str.append(acl_entry->host.hostname, acl_entry->hostname_length, system_charset_info); > + def_str.append('\''); > + protocol->prepare_for_resend(); > + protocol->store(def_str.ptr(),def_str.length(),def_str.charset()); > + if (protocol->write()) > + { > + return TRUE; > + } > + } > + return FALSE; > +} Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

1 0

Re: [Maria-developers] 6729bf5c031: MDEV-22313: Neither SHOW CREATE USER nor SHOW GRANTS prints a user's default role
by Sergei Golubchik 22 Oct '20

22 Oct '20

Hi, Anel! On Oct 22, Anel Husakovic wrote: > Hi Sergei, > thank you for your review. > How is easier for you (or to ask a general question: how does the review > process go?) regarding the commits? > Is it better to add a new commit as a review commit and if the patch gets > accepted to squash after, or to rebase commits and send only a single > commit again for the review? One single commit is better, thanks. At least for small patches. > Addressed in the new patch: > https://github.com/MariaDB/server/commit/99c144762514d4dea3b6ec395de96aaf10… Thanks! I'll reply to the commit directly. Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

1 0

Re: [Maria-developers] 83614913d9a: MDEV-16243 [Warning] "Versioned table ...: partition ... is full, add more HISTORY partitions" floods the error log
by Sergei Golubchik 22 Oct '20

22 Oct '20

Hi, Aleksey! On Oct 22, Aleksey Midenkov wrote: > revision-id: 83614913d9a (mariadb-10.4.11-291-g83614913d9a) > parent(s): efb1023b6f4 > author: Aleksey Midenkov <midenok(a)gmail.com> > committer: Sergei Golubchik <serg(a)mariadb.com> > timestamp: 2020-10-22 16:44:13 +0200 > message: > > MDEV-16243 [Warning] "Versioned table ...: partition ... is full, add more HISTORY partitions" floods the error log > > diff --git a/sql/partition_info.cc b/sql/partition_info.cc > index 8989a918c0c..f287cfddcd2 100644 > --- a/sql/partition_info.cc > +++ b/sql/partition_info.cc > @@ -879,9 +879,12 @@ void partition_info::vers_set_hist_part(THD *thd) > } > return; > warn: > - my_error(WARN_VERS_PART_FULL, MYF(ME_WARNING|ME_ERROR_LOG), > - table->s->db.str, table->s->table_name.str, > - vers_info->hist_part->partition_name); > + if (vers_info->hist_part->warn_vers_part_full) > + { > + my_error(WARN_VERS_PART_FULL, MYF(ME_WARNING|ME_ERROR_LOG), > + table->s->db.str, table->s->table_name.str, > + vers_info->hist_part->partition_name); > + } would it make sense to put warn_vers_part_full into vers_info, instead of storing it per-partition? > } > > > Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

1 0

Re: [Maria-developers] 2830b9c290e: MDEV-23926: Fix warnings generated during compilation of plugin/auth_pam/mapper/pam_user_map.c on MacOS
by Sergei Golubchik 22 Oct '20

22 Oct '20

Hi, Dmitry! On Oct 22, Dmitry Shulga wrote: > revision-id: 2830b9c290e (mariadb-10.2.31-489-g2830b9c290e) > parent(s): 8984d779103 > author: Dmitry Shulga <dmitry.shulga(a)mariadb.com> > committer: Dmitry Shulga <dmitry.shulga(a)mariadb.com> > timestamp: 2020-10-22 00:42:10 +0700 > message: > > MDEV-23926: Fix warnings generated during compilation of plugin/auth_pam/mapper/pam_user_map.c on MacOS > > Compiler warnings like one listed below are generated during server build on MacOS: > > > --- > plugin/auth_pam/CMakeLists.txt | 26 ++++++++++++++++++++++++++ > plugin/auth_pam/config.h.cmake | 1 + > plugin/auth_pam/mapper/pam_user_map.c | 30 +++++++++++++++++++----------- > 3 files changed, 46 insertions(+), 11 deletions(-) Nice! Could you also - in a second follow-up commit, not to confuse with the pure bug fix - do a small cleanup and move all other checks from plugin/auth_pam/CMakeLists.txt into your new config.h.cmake ? Currently they're added to the command line, like IF(HAVE_PAM_EXT_H) ADD_DEFINITIONS(-DHAVE_PAM_EXT_H) ENDIF() but as you've introduced config.h.cmake, there's no need to put them all on the command line anymore. You'll need to include config_auth_pam.h into more source files. And, I suppose. add config_auth_pam.h to .gitignore too. ok to push these two commits when done. Thanks! Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

1 0

Re: [Maria-developers] 65af1d07771: MDEV-23178: Qualified asterisk not supported in INSERT .. RETURNING
by Sergei Golubchik 22 Oct '20

22 Oct '20

Hi, Rucha! On Oct 22, Rucha Deodhar wrote: > revision-id: 65af1d07771 (mariadb-10.5.4-108-g65af1d07771) > parent(s): bbd70fcc43c > author: Rucha Deodhar <rucha.deodhar(a)mariadb.com> > committer: Rucha Deodhar <rucha.deodhar(a)mariadb.com> > timestamp: 2020-08-06 17:23:49 +0530 > message: > > MDEV-23178: Qualified asterisk not supported in INSERT .. RETURNING > > Anaylsis: When we have INSERT/REPLACE...RETURNING and have qualified asterisk, > table_name is not NULL and context->table_list is either NULL or has > incorrect reference because context->table_list has tables from the FROM clause. > context->table_list has incorrect reference (has table from the FROM clause > instead of table we are inserting into) for INSERT/REPLACE...SELECT...RETURNING > because we have a FROM clause from the SELECT statement. > For INSERT/REPLACE...RETURNING it is NULL because there is no FROM clause. > Fix: If table_name is not null, check if we have INSERT/REPLACE...RETURNING. > If so, the reference should be the table we are inserting into and not > the table in the FROM clause. > > diff --git a/mysql-test/main/insert_returning.result b/mysql-test/main/insert_returning.result > index e664e02bedc..e1bab6f0329 100644 > --- a/mysql-test/main/insert_returning.result > +++ b/mysql-test/main/insert_returning.result > @@ -89,6 +89,9 @@ total val1 id1 && id1 id1 UPPER(val1) f(id1) good tests! > diff --git a/sql/sql_base.cc b/sql/sql_base.cc > index 45ce4be3eb5..4ce7b034029 100644 > --- a/sql/sql_base.cc > +++ b/sql/sql_base.cc > @@ -7959,7 +7959,7 @@ insert_fields(THD *thd, Name_resolution_context *context, const char *db_name, > bool any_privileges, uint *hidden_bit_fields) > { > Field_iterator_table_ref field_iterator; > - bool found; > + bool found, is_insert_or_replace_returning= false; > char name_buff[SAFE_NAME_LEN+1]; > DBUG_ENTER("insert_fields"); > DBUG_PRINT("arena", ("stmt arena: %p",thd->stmt_arena)); > @@ -7978,13 +7978,31 @@ insert_fields(THD *thd, Name_resolution_context *context, const char *db_name, > > found= FALSE; > > + if ((thd->lex->sql_command == SQLCOM_INSERT_SELECT || > + thd->lex->sql_command == SQLCOM_REPLACE_SELECT || > + thd->lex->sql_command == SQLCOM_INSERT || > + thd->lex->sql_command == SQLCOM_REPLACE) && > + thd->lex->has_returning()) > + is_insert_or_replace_returning= true; What about INSERT t1 SELECT t2.* FROM t2 RETURNING t1.*; can your is_insert_or_replace_returning distinguish between the first and the second asterisk? Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

1 0

Re: [Maria-developers] 01612f594b3: MDEV-23769: MTR can abort before it prints the test result summary
by Sergei Golubchik 22 Oct '20

22 Oct '20

Hi, Anel! On Oct 22, Anel Husakovic wrote: > revision-id: 01612f594b3 (mariadb-10.1.43-300-g01612f594b3) > parent(s): 82301aea4f2 > author: Anel Husakovic <anel(a)mariadb.org> > committer: Anel Husakovic <anel(a)mariadb.org> > timestamp: 2020-10-03 02:34:03 +0200 > message: > > MDEV-23769: MTR can abort before it prints the test result summary > > - Patch is solving generating report on warning > - Patch is doing cosmetic fix of `current_test` log file which holds the old log value of test `CURRENT TEST:..` in `mark_log()` in case of `unknown option` and as such > the logic which is using it's content doesn't output valid log content and doesn't generate valid `$test->{'comment'}` message. more is not always better. please remove everything in the commit comment below this line - 400 lines that don't explain anything beyond what your *one line* of the comment already did. And now you have to press PageDown many times to get to my main comment about the fix :) > Here are the test cases observed. Note that only in 6. scenario there is `print $child "HELLO\n";` problem in printing (to delete?). > > 1. Valid test case, but warning in server log > ``` > $ ./mtr MDEV-23769 --force > Logging: ./mtr MDEV-23769 --force > vardir: /home/anel/mariadb/10.1-mtr-in-source/mysql-test/var > Checking leftover processes... > Removing old var directory... > Creating var directory '/home/anel/mariadb/10.1-mtr-in-source/mysql-test/var'... > Checking supported features... > MariaDB Version 10.1.47-MariaDB-debug > - SSL connections supported > - binaries are debug compiled > Sphinx 'indexer' binary not found, sphinx suite will be skipped > Collecting tests... > Installing system database... > > ============================================================================== > > TEST RESULT TIME (ms) or COMMENT > -------------------------------------------------------------------------- > > worker[1] Using MTR_BUILD_THREAD 300, with reserved ports 16000..16019 > show databases; > Database > information_schema > mtr > mysql > performance_schema > test > main.MDEV-23769 [ fail ] Found warnings/errors in server log file! > Test ended at 2020-10-03 02:13:00 > line > 2020-10-03 2:13:00 140309996242496 [Warning] option 'lock_wait_timeout': unsigned value 18446744073709551615 adjusted to 31536000 > ^ Found warnings in /home/anel/mariadb/10.1-mtr-in-source/mysql-test/var/log/mysqld.1.err > ok > > - saving '/home/anel/mariadb/10.1-mtr-in-source/mysql-test/var/log/main.MDEV-23769/' to '/home/anel/mariadb/10.1-mtr-in-source/mysql-test/var/log/main.MDEV-23769/' > -------------------------------------------------------------------------- > The servers were restarted 0 times > Spent 0.000 of 4 seconds executing testcases > mysql-test-run: WARNING: Got errors/warnings while running tests, please examine '/home/anel/mariadb/10.1-mtr-in-source/mysql-test/var/log/warnings' for details. > > Completed: Failed 1/1 tests, 0.00% were successful. > > Failing test(s): main.MDEV-23769 > > The log files in var/log may give you some hint of what went wrong. > > If you want to report this error, please read first the documentation > at http://dev.mysql.com/doc/mysql/en/mysql-test-suite.html > > mysql-test-run: *** ERROR: there were failing test cases > ``` > > 2. Valid test cases (same), but warning in server log (Note reported single test case) > ``` > $ ./mtr MDEV-23769 MDEV-23769 --force > Logging: ./mtr MDEV-23769 MDEV-23769 --force > vardir: /home/anel/mariadb/10.1-mtr-in-source/mysql-test/var > Checking leftover processes... > Removing old var directory... > Creating var directory '/home/anel/mariadb/10.1-mtr-in-source/mysql-test/var'... > Checking supported features... > MariaDB Version 10.1.47-MariaDB-debug > - SSL connections supported > - binaries are debug compiled > Sphinx 'indexer' binary not found, sphinx suite will be skipped > Collecting tests... > Installing system database... > > ============================================================================== > > TEST RESULT TIME (ms) or COMMENT > -------------------------------------------------------------------------- > > worker[1] Using MTR_BUILD_THREAD 300, with reserved ports 16000..16019 > show databases; > Database > information_schema > mtr > mysql > performance_schema > test > main.MDEV-23769 [ fail ] Found warnings/errors in server log file! > Test ended at 2020-10-03 02:15:05 > line > 2020-10-03 2:15:04 139868136609344 [Warning] option 'lock_wait_timeout': unsigned value 18446744073709551615 adjusted to 31536000 > ^ Found warnings in /home/anel/mariadb/10.1-mtr-in-source/mysql-test/var/log/mysqld.1.err > ok > > - saving '/home/anel/mariadb/10.1-mtr-in-source/mysql-test/var/log/main.MDEV-23769/' to '/home/anel/mariadb/10.1-mtr-in-source/mysql-test/var/log/main.MDEV-23769/' > show databases; > Database > information_schema > mtr > mysql > performance_schema > test > main.MDEV-23769 [ fail ] Found warnings/errors in server log file! > Test ended at 2020-10-03 02:15:05 > line > 2020-10-03 2:15:05 140513782085184 [Warning] option 'lock_wait_timeout': unsigned value 18446744073709551615 adjusted to 31536000 > ^ Found warnings in /home/anel/mariadb/10.1-mtr-in-source/mysql-test/var/log/mysqld.1.err > ok > > - saving '/home/anel/mariadb/10.1-mtr-in-source/mysql-test/var/log/main.MDEV-23769/' to '/home/anel/mariadb/10.1-mtr-in-source/mysql-test/var/log/main.MDEV-23769/' > -------------------------------------------------------------------------- > The servers were restarted 1 times > Spent 0.000 of 5 seconds executing testcases > mysql-test-run: WARNING: Got errors/warnings while running tests, please examine '/home/anel/mariadb/10.1-mtr-in-source/mysql-test/var/log/warnings' for details. > > Completed: Failed 2/2 tests, 0.00% were successful. > > Failing test(s): main.MDEV-23769 > > The log files in var/log may give you some hint of what went wrong. > > If you want to report this error, please read first the documentation > at http://dev.mysql.com/doc/mysql/en/mysql-test-suite.html > > mysql-test-run: *** ERROR: there were failing test cases > ``` > > 3. Valid test cases (different test cases), but warning in server log (Note one passed one failed) > > ``` > $ ./mtr MDEV-23769 1st --force > Logging: ./mtr MDEV-23769 1st --force > vardir: /home/anel/mariadb/10.1-mtr-in-source/mysql-test/var > Checking leftover processes... > Removing old var directory... > Creating var directory '/home/anel/mariadb/10.1-mtr-in-source/mysql-test/var'... > Checking supported features... > MariaDB Version 10.1.47-MariaDB-debug > - SSL connections supported > - binaries are debug compiled > Sphinx 'indexer' binary not found, sphinx suite will be skipped > Collecting tests... > Installing system database... > > ============================================================================== > > TEST RESULT TIME (ms) or COMMENT > -------------------------------------------------------------------------- > > worker[1] Using MTR_BUILD_THREAD 300, with reserved ports 16000..16019 > show databases; > Database > information_schema > mtr > mysql > performance_schema > test > main.MDEV-23769 [ fail ] Found warnings/errors in server log file! > Test ended at 2020-10-03 02:17:09 > line > 2020-10-03 2:17:08 140658805515840 [Warning] option 'lock_wait_timeout': unsigned value 18446744073709551615 adjusted to 31536000 > ^ Found warnings in /home/anel/mariadb/10.1-mtr-in-source/mysql-test/var/log/mysqld.1.err > ok > > - saving '/home/anel/mariadb/10.1-mtr-in-source/mysql-test/var/log/main.MDEV-23769/' to '/home/anel/mariadb/10.1-mtr-in-source/mysql-test/var/log/main.MDEV-23769/' > main.1st [ pass ] 1 > -------------------------------------------------------------------------- > The servers were restarted 1 times > Spent 0.001 of 4 seconds executing testcases > mysql-test-run: WARNING: Got errors/warnings while running tests, please examine '/home/anel/mariadb/10.1-mtr-in-source/mysql-test/var/log/warnings' for details. > > Completed: Failed 1/2 tests, 50.00% were successful. > > Failing test(s): main.MDEV-23769 > > The log files in var/log may give you some hint of what went wrong. > > If you want to report this error, please read first the documentation > at http://dev.mysql.com/doc/mysql/en/mysql-test-suite.html > > mysql-test-run: *** ERROR: there were failing test cases > > ``` > > 4. Now for example 3. create artificial failure with warnings in server log (note, both test failed) > > ``` > > $ ./mtr MDEV-23769 1st --force > Logging: ./mtr MDEV-23769 1st --force > vardir: /home/anel/mariadb/10.1-mtr-in-source/mysql-test/var > Checking leftover processes... > Removing old var directory... > Creating var directory '/home/anel/mariadb/10.1-mtr-in-source/mysql-test/var'... > Checking supported features... > MariaDB Version 10.1.47-MariaDB-debug > - SSL connections supported > - binaries are debug compiled > Sphinx 'indexer' binary not found, sphinx suite will be skipped > Collecting tests... > Installing system database... > > ============================================================================== > > TEST RESULT TIME (ms) or COMMENT > -------------------------------------------------------------------------- > > worker[1] Using MTR_BUILD_THREAD 300, with reserved ports 16000..16019 > show databases; > Database > information_schema > mtr > mysql > performance_schema > test > main.MDEV-23769 [ fail ] Found warnings/errors in server log file! > Test ended at 2020-10-03 02:17:52 > line > 2020-10-03 2:17:51 140260493149760 [Warning] option 'lock_wait_timeout': unsigned value 18446744073709551615 adjusted to 31536000 > ^ Found warnings in /home/anel/mariadb/10.1-mtr-in-source/mysql-test/var/log/mysqld.1.err > ok > > - saving '/home/anel/mariadb/10.1-mtr-in-source/mysql-test/var/log/main.MDEV-23769/' to '/home/anel/mariadb/10.1-mtr-in-source/mysql-test/var/log/main.MDEV-23769/' > main.1st [ fail ] > Test ended at 2020-10-03 02:17:52 > > CURRENT_TEST: main.1st > mysqltest: At line 5: End of line junk detected: "show tables in mysql > " > > The result from queries just before the failure was: > show databases; > Database > information_schema > mtr > mysql > performance_schema > test > > - saving '/home/anel/mariadb/10.1-mtr-in-source/mysql-test/var/log/main.1st/' to '/home/anel/mariadb/10.1-mtr-in-source/mysql-test/var/log/main.1st/' > -------------------------------------------------------------------------- > The servers were restarted 1 times > Spent 0.000 of 4 seconds executing testcases > mysql-test-run: WARNING: Got errors/warnings while running tests, please examine '/home/anel/mariadb/10.1-mtr-in-source/mysql-test/var/log/warnings' for details. > > Completed: Failed 2/2 tests, 0.00% were successful. > > Failing test(s): main.MDEV-23769 main.1st > > The log files in var/log may give you some hint of what went wrong. > > If you want to report this error, please read first the documentation > at http://dev.mysql.com/doc/mysql/en/mysql-test-suite.html > > mysql-test-run: *** ERROR: there were failing test cases > > ``` > > 5. Generate unknown option, for the same example as 3. (note: both failed) > ``` > > $ ./mtr MDEV-23769 1st --mysqld=--xx --force > Logging: ./mtr MDEV-23769 1st --mysqld=--xx --force > vardir: /home/anel/mariadb/10.1-mtr-in-source/mysql-test/var > Checking leftover processes... > Removing old var directory... > Creating var directory '/home/anel/mariadb/10.1-mtr-in-source/mysql-test/var'... > Checking supported features... > MariaDB Version 10.1.47-MariaDB-debug > - SSL connections supported > - binaries are debug compiled > Sphinx 'indexer' binary not found, sphinx suite will be skipped > Collecting tests... > Installing system database... > > ============================================================================== > > TEST RESULT TIME (ms) or COMMENT > -------------------------------------------------------------------------- > > worker[1] Using MTR_BUILD_THREAD 300, with reserved ports 16000..16019 > worker[1] mysql-test-run: WARNING: Process [mysqld.1 - pid: 52531, winpid: 52531, exit: 512] died after mysql-test-run waited 0.3 seconds for /home/anel/mariadb/10.1-mtr-in-source/mysql-test/var/run/mysqld.1.pid to be created. > main.MDEV-23769 [ fail ] > Test ended at 2020-10-03 02:19:01 > > CURRENT_TEST: main.MDEV-23769 > > Failed to start mysqld.1 > mysqltest failed but provided no output > > - saving '/home/anel/mariadb/10.1-mtr-in-source/mysql-test/var/log/main.MDEV-23769/' to '/home/anel/mariadb/10.1-mtr-in-source/mysql-test/var/log/main.MDEV-23769/' > ***Warnings generated in error logs during shutdown after running tests: main.MDEV-23769 > > 2020-10-03 2:19:01 140238937720384 [Warning] option 'lock_wait_timeout': unsigned value 18446744073709551615 adjusted to 31536000 > 2020-10-03 2:19:01 140238937720384 [ERROR] /home/anel/mariadb/10.1-mtr-in-source/sql/mysqld: unknown option '--xx' > 2020-10-03 2:19:01 140238937720384 [ERROR] Aborting > > worker[1] mysql-test-run: WARNING: Process [mysqld.1 - pid: 52537, winpid: 52537, exit: 512] died after mysql-test-run waited 0.3 seconds for /home/anel/mariadb/10.1-mtr-in-source/mysql-test/var/run/mysqld.1.pid to be created. > main.1st [ fail ] > Test ended at 2020-10-03 02:19:02 > > CURRENT_TEST: main.1st > > Failed to start mysqld.1 > mysqltest failed but provided no output > > - saving '/home/anel/mariadb/10.1-mtr-in-source/mysql-test/var/log/main.1st/' to '/home/anel/mariadb/10.1-mtr-in-source/mysql-test/var/log/main.1st/' > ***Warnings generated in error logs during shutdown after running tests: main.1st > > 2020-10-03 2:19:02 140048201700928 [ERROR] /home/anel/mariadb/10.1-mtr-in-source/sql/mysqld: unknown option '--xx' > 2020-10-03 2:19:02 140048201700928 [ERROR] Aborting > > Can't use an undefined value as a symbol reference at ./mtr line 657. > -------------------------------------------------------------------------- > The servers were restarted 1 times > Spent 0.000 of 5 seconds executing testcases > > Completed: Failed 2/2 tests, 0.00% were successful. > > Failing test(s): main.MDEV-23769 main.1st > > The log files in var/log may give you some hint of what went wrong. > > If you want to report this error, please read first the documentation > at http://dev.mysql.com/doc/mysql/en/mysql-test-suite.html > > Errors/warnings were found in logfiles during server shutdown after running the > following sequence(s) of tests: > main.MDEV-23769 > main.1st > mysql-test-run: *** ERROR: there were failing test cases > > ``` > > 6. Again the same option, but with 2 same tests -> not perl error: `Can't use an undefined value as a symbol reference at ./mtr line 657. > ` ` print $child "HELLO\n";` (solution delete the line). > > ``` > $ ./mtr MDEV-23769 MDEV-23769 --mysqld=--xx --force > Logging: ./mtr MDEV-23769 MDEV-23769 --mysqld=--xx --force > vardir: /home/anel/mariadb/10.1-mtr-in-source/mysql-test/var > Checking leftover processes... > Removing old var directory... > Creating var directory '/home/anel/mariadb/10.1-mtr-in-source/mysql-test/var'... > Checking supported features... > MariaDB Version 10.1.47-MariaDB-debug > - SSL connections supported > - binaries are debug compiled > Sphinx 'indexer' binary not found, sphinx suite will be skipped > Collecting tests... > Installing system database... > > ============================================================================== > > TEST RESULT TIME (ms) or COMMENT > -------------------------------------------------------------------------- > > worker[1] Using MTR_BUILD_THREAD 300, with reserved ports 16000..16019 > worker[1] mysql-test-run: WARNING: Process [mysqld.1 - pid: 52595, winpid: 52595, exit: 512] died after mysql-test-run waited 0.3 seconds for /home/anel/mariadb/10.1-mtr-in-source/mysql-test/var/run/mysqld.1.pid to be created. > main.MDEV-23769 [ fail ] > Test ended at 2020-10-03 02:20:12 > > CURRENT_TEST: main.MDEV-23769 > > Failed to start mysqld.1 > mysqltest failed but provided no output > > - saving '/home/anel/mariadb/10.1-mtr-in-source/mysql-test/var/log/main.MDEV-23769/' to '/home/anel/mariadb/10.1-mtr-in-source/mysql-test/var/log/main.MDEV-23769/' > ***Warnings generated in error logs during shutdown after running tests: main.MDEV-23769 > > 2020-10-03 2:20:12 139815724283456 [Warning] option 'lock_wait_timeout': unsigned value 18446744073709551615 adjusted to 31536000 > 2020-10-03 2:20:12 139815724283456 [ERROR] /home/anel/mariadb/10.1-mtr-in-source/sql/mysqld: unknown option '--xx' > 2020-10-03 2:20:12 139815724283456 [ERROR] Aborting > > worker[1] mysql-test-run: WARNING: Process [mysqld.1 - pid: 52601, winpid: 52601, exit: 512] died after mysql-test-run waited 0.3 seconds for /home/anel/mariadb/10.1-mtr-in-source/mysql-test/var/run/mysqld.1.pid to be created. > main.MDEV-23769 [ fail ] > Test ended at 2020-10-03 02:20:12 > > CURRENT_TEST: main.MDEV-23769 > > Failed to start mysqld.1 > mysqltest failed but provided no output > > - saving '/home/anel/mariadb/10.1-mtr-in-source/mysql-test/var/log/main.MDEV-23769/' to '/home/anel/mariadb/10.1-mtr-in-source/mysql-test/var/log/main.MDEV-23769/' > ***Warnings generated in error logs during shutdown after running tests: main.MDEV-23769 > > 2020-10-03 2:20:12 140650797745728 [Warning] option 'lock_wait_timeout': unsigned value 18446744073709551615 adjusted to 31536000 > 2020-10-03 2:20:12 140650797745728 [ERROR] /home/anel/mariadb/10.1-mtr-in-source/sql/mysqld: unknown option '--xx' > 2020-10-03 2:20:12 140650797745728 [ERROR] Aborting > > Can't use an undefined value as a symbol reference at ./mtr line 657. > -------------------------------------------------------------------------- > The servers were restarted 1 times > Spent 0.000 of 5 seconds executing testcases > > Completed: Failed 2/2 tests, 0.00% were successful. > > Failing test(s): main.MDEV-23769 > > The log files in var/log may give you some hint of what went wrong. > > If you want to report this error, please read first the documentation > at http://dev.mysql.com/doc/mysql/en/mysql-test-suite.html > > Errors/warnings were found in logfiles during server shutdown after running the > following sequence(s) of tests: > main.MDEV-23769 > main.MDEV-23769 > mysql-test-run: *** ERROR: there were failing test cases > > ``` > > diff --git a/mysql-test/disabled.def b/mysql-test/disabled.def > index ee30c4f4d3c..fdb62a7e5f5 100644 > --- a/mysql-test/disabled.def > +++ b/mysql-test/disabled.def > @@ -20,3 +20,4 @@ ssl_crl_clrpath : broken upstream > file_contents : MDEV-6526 these files are not installed anymore > max_statement_time : cannot possibly work, depends on timing > partition_open_files_limit : open_files_limit check broken by MDEV-18360 > +MDEV-23769 : Intentionally disabled, for validating purposes. I don't know if it makes any sense. A dead test that nobody will ever run. May be it's better not to add it at all? In the past I've added tests like that too, the "mtr" and "mtr2" suites. I regret it now :) Tests only help if they're run. > diff --git a/mysql-test/mysql-test-run.pl b/mysql-test/mysql-test-run.pl > index 52920530132..8d761fd70cf 100755 > --- a/mysql-test/mysql-test-run.pl > +++ b/mysql-test/mysql-test-run.pl > @@ -1005,6 +1005,13 @@ sub run_worker ($) { > stop_servers(reverse all_servers()); > if(check_warnings_post_shutdown($server)) { > # Warnings appeared in log file(s) during final server shutdown. > + # Make sure first to run statistics > + my $tests= collect_test_cases($opt_reorder, $opt_suites, \@opt_cases, \@opt_skip_test_list); > + my ($prefix, $fail, $completed, $extra_warnings)= > + run_test_server($server, $tests, $opt_parallel); > + mtr_print_line(); > + print_total_times($opt_parallel) if $opt_report_times; > + mtr_report_stats($prefix, $fail, $completed, $extra_warnings); I don't understand the fix. You're saying that if a warning happens on shutdown then mtr won't print the result summary. This doesn't seem to be the case, I've tried few times to repeat it - modified mtr to always pretend there's a warning on shutdown, and modified mysqld to always issue a warning on shutdown - and in all cases I saw the result summary all right. > exit(1); > } > } > @@ -3971,6 +3978,7 @@ sub run_testcase ($$) { > if (start_servers($tinfo)) > { > report_failure_and_restart($tinfo); > + unlink $path_current_testlog; > return 1; > } > } > diff --git a/mysql-test/t/MDEV-23769.opt b/mysql-test/t/MDEV-23769.opt > new file mode 100644 > index 00000000000..fd8dbcab53f > --- /dev/null > +++ b/mysql-test/t/MDEV-23769.opt > @@ -0,0 +1 @@ > +--lock-wait-timeout=-1 > diff --git a/mysql-test/t/MDEV-23769.test b/mysql-test/t/MDEV-23769.test > new file mode 100644 > index 00000000000..9938c7cf644 > --- /dev/null > +++ b/mysql-test/t/MDEV-23769.test > @@ -0,0 +1,4 @@ > +--echo # > +--echo # MDEV-23769: MTR can abort before it prints the test result summary > +--echo # > +show databases; > Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

1 0

Re: [Maria-developers] 8c745467560: MDEV-23836: Assertion `! is_set() || m_can_overwrite_status' in
by Sergei Golubchik 21 Oct '20

21 Oct '20

Hi, Rucha! what happens in this bug? Why does the server need to open_table() after my_ok()? On Oct 21, Rucha Deodhar wrote: > revision-id: 8c745467560 (mariadb-10.5.4-226-g8c745467560) > parent(s): 308f8350c7b > author: Rucha Deodhar <rucha.deodhar(a)mariadb.com> > committer: Rucha Deodhar <rucha.deodhar(a)mariadb.com> > timestamp: 2020-10-20 11:53:22 +0530 > message: > > MDEV-23836: Assertion `! is_set() || m_can_overwrite_status' in > Diagnostics_area::set_error_status (interrupted ALTER TABLE under LOCK) > > Analysis: Assertion failure happens because we don't ignore kill query > even if 'OK' is already sent. > Fix: Check if we already sent 'OK'. If sent, ignore kill query. > > diff --git a/mysql-test/main/alter_table.result b/mysql-test/main/alter_table.result > index 197e52b35ca..204bfa5249c 100644 > --- a/mysql-test/main/alter_table.result > +++ b/mysql-test/main/alter_table.result > @@ -3322,5 +3322,23 @@ Note 1176 Key 'x' doesn't exist in table 't1' > unlock tables; > drop table t1; > # > +# MDEV-23836: Assertion `! is_set() || m_can_overwrite_status' in > +# Diagnostics_area::set_error_status (interrupted ALTER TABLE under LOCK) > +# > +set @max_session_mem_used_save= @@max_session_mem_used; > +CREATE TABLE t1 (a INT); > +SELECT * FROM t1; > +a > +ALTER TABLE x MODIFY xx INT; > +ERROR 42S02: Table 'test.x' doesn't exist > +SET SESSION max_session_mem_used= 8192; > +LOCK TABLE t1 WRITE; > +ALTER TABLE t1 CHANGE COLUMN IF EXISTS b c INT; > +Warnings: > +Note 1054 Unknown column 'b' in 't1' > +set SESSION max_session_mem_used = @max_session_mem_used_save; > +UNLOCK TABLES; > +DROP TABLE t1; > +# > # End of 10.5 tests > # > diff --git a/mysql-test/main/alter_table.test b/mysql-test/main/alter_table.test > index 6fe61a3222a..d09c4378339 100644 > --- a/mysql-test/main/alter_table.test > +++ b/mysql-test/main/alter_table.test > @@ -2529,6 +2529,27 @@ alter table t1 rename key if exists x to xx; > unlock tables; > drop table t1; > > +--echo # > +--echo # MDEV-23836: Assertion `! is_set() || m_can_overwrite_status' in > +--echo # Diagnostics_area::set_error_status (interrupted ALTER TABLE under LOCK) > +--echo # > + > +set @max_session_mem_used_save= @@max_session_mem_used; > + > +CREATE TABLE t1 (a INT); > +SELECT * FROM t1; > + > +--error ER_NO_SUCH_TABLE > +ALTER TABLE x MODIFY xx INT; > + > +SET SESSION max_session_mem_used= 8192; > +LOCK TABLE t1 WRITE; > +ALTER TABLE t1 CHANGE COLUMN IF EXISTS b c INT; > + > +set SESSION max_session_mem_used = @max_session_mem_used_save; > +UNLOCK TABLES; > +DROP TABLE t1; > + > --echo # > --echo # End of 10.5 tests > --echo # > diff --git a/sql/sql_base.cc b/sql/sql_base.cc > index e6d0ab13e75..e8d39917ba6 100644 > --- a/sql/sql_base.cc > +++ b/sql/sql_base.cc > @@ -1682,7 +1682,9 @@ bool open_table(THD *thd, TABLE_LIST *table_list, Open_table_context *ot_ctx) > > if (!(flags & MYSQL_OPEN_IGNORE_KILLED) && thd->killed) > { > - thd->send_kill_message(); > + /* If we already sent 'ok', we can ignore any kill query statements */ > + if (! thd->get_stmt_da()->is_set()) > + thd->send_kill_message(); > DBUG_RETURN(TRUE); > } > Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

1 0

Re: [Maria-developers] 1060cc4235a: MDEV-23500: Assertion `!is_set() || (m_status == DA_OK_BULK && is_bulk_op())'
by Sergei Golubchik 21 Oct '20

21 Oct '20

Hi, Rucha! On Oct 21, Rucha Deodhar wrote: > revision-id: 1060cc4235a (mariadb-10.2.31-438-g1060cc4235a) > parent(s): 80075ba011f > author: Rucha Deodhar <rucha.deodhar(a)mariadb.com> > committer: Rucha Deodhar <rucha.deodhar(a)mariadb.com> > timestamp: 2020-09-22 17:12:57 +0530 > message: > > MDEV-23500: Assertion `!is_set() || (m_status == DA_OK_BULK && is_bulk_op())' > failed in Diagnostics_area::set_ok_status upon multi-table update > > Analysis: When NULL field is copied to NOT-NULL so there is error but > this error is not reported so we get the failure. > Fix: Return the error state. > > diff --git a/mysql-test/r/update.result b/mysql-test/r/update.result > index 9e19abc4e9c..c1d37fc36e7 100644 > --- a/mysql-test/r/update.result > +++ b/mysql-test/r/update.result > @@ -731,4 +731,14 @@ INSERT INTO t2 VALUES > UPDATE t1,t2 SET t1.i1 = -39 WHERE t2.d1 <> t1.i1 AND t2.d1 = t1.d2; > ERROR 22007: Incorrect datetime value: '19' for column `test`.`t1`.`i1` at row 1 > DROP TABLE t1,t2; > +# > +# MDEV-23500: Assertion `!is_set() || (m_status == DA_OK_BULK && > +# is_bulk_op())' failed in Diagnostics_area::set_ok_status upon > +# multi-table update > +# > +CREATE TABLE t (f INT NOT NULL); > +INSERT INTO t VALUES (0),(0); > +UPDATE t AS t1 LEFT JOIN t AS t2 ON t1.f <=> t2.f SET t2.f = NULL; > +ERROR 01000: Data truncated for column 'f' at row 4 with a single-table update I get MariaDB [test]> update t set f=NULL; ERROR 1048 (23000): Column 'f' cannot be null should multi-table update produce the same error? > +DROP TABLE t; Regards, Sergei VP of MariaDB Server Engineering and security(a)mariadb.org

1 0