Uploaded image for project: 'MariaDB Server'
  1. MariaDB Server
  2. MDEV-34996

Buildbot MSAN options in buildbot rather than server

Details

    Description

      There are a number of buildbot configuration items related to MSAN that are in buildbot rather than the server.

      These include:

      • ensuring the compiler is msan compatible and erroring
      • adding -fsanitzer=memory as a linker flag
      • using -fsanitize-memory-track-origins -U_FORTIFY_SOURCE
      • using -stdlib=libc++ as a C++ CFLAG
      • -fdebug-macro as a compile flag under debug - (common but not strictly msan)
      • no SECURITY_HARDENED with MSAN otherwise it bootstrap fails 

        ==24876==WARNING: MemorySanitizer: use-of-uninitialized-value
        		 
            #0 0x5619b9c4cab1 in strmake /source/strings/strmake.c:36:9
        		 
            #1 0x5619b9ac735a in fn_format /source/mysys/mf_format.c:97:9
        		 
            #2 0x5619b81efd5b in Create_tmp_table::start(THD*, TMP_TABLE_PARAM*, st_mysql_const_lex_string const*) /source/sql/sql_select.cc:18903:3
        		 
            #3 0x5619b81f92d6 in create_tmp_table_for_schema(THD*, TMP_TABLE_PARAM*, st_schema_table const&, long long, st_mysql_const_lex_string const&, bool, bool) /source/sql/sql_select.cc:19816:22
        		 
            #4 0x5619b82802c6 in create_schema_table(THD*, TABLE_LIST*) /source/sql/sql_show.cc:8386:9
        		 
            #5 0x5619b8282e82 in mysql_schema_table(THD*, LEX*, TABLE_LIST*) /source/sql/sql_show.cc:8594:16
        		 
            #6 0x5619b7f3e0c2 in open_and_process_table(THD*, TABLE_LIST*, unsigned int*, unsigned int, Prelocking_strategy*, bool, Open_table_context*) /source/sql/sql_base.cc:3697:15
        		 
            #7 0x5619b7f3e0c2 in open_tables(THD*, DDL_options_st const&, TABLE_LIST**, unsigned int*, unsigned int, Prelocking_strategy*) /source/sql/sql_base.cc:4303:14
        		 
            #8 0x5619b7f46d29 in open_and_lock_tables(THD*, DDL_options_st const&, TABLE_LIST*, bool, unsigned int, Prelocking_strategy*) /source/sql/sql_base.cc:5250:7
        		 
            #9 0x5619b8006c34 in open_and_lock_tables(THD*, TABLE_LIST*, bool, unsigned int) /source/sql/sql_base.h:509:10
        		 
            #10 0x5619b80b6ec8 in mysql_execute_command(THD*) /source/sql/sql_parse.cc:5171:13
        		 
            #11 0x5619b80a3dbe in mysql_parse(THD*, char*, unsigned int, Parser_state*, bool, bool) /source/sql/sql_parse.cc:8236:18
        		 
            #12 0x5619b80a28b5 in bootstrap(st_mysql_file*) /source/sql/sql_parse.cc:1087:5
        		 
            #13 0x5619b7de3158 in mysqld_main(int, char**) /source/sql/mysqld.cc:5675:26
        		 
            #14 0x7f4e75646249  (/lib/x86_64-linux-gnu/libc.so.6+0x27249) (BuildId: 30f067a5eccdb3bdce0b9ec7638ebd86992b9c8b)
        		 
            #15 0x7f4e75646304 in __libc_start_main (/lib/x86_64-linux-gnu/libc.so.6+0x27304) (BuildId: 30f067a5eccdb3bdce0b9ec7638ebd86992b9c8b)
        		 
            #16 0x5619b7d3fdd0 in _start (/build/sql/mariadbd+0x717dd0) (BuildId: 7e347ffdc10dc4f03bb7510c02a7b9e86856d32f)

      Attachments

        Issue Links

          blocks

          New Feature - A new feature of the product, which has yet to be developed. MDBF-793 Retire MSAN clang-15 builder, upgrade to clang-19

          • Major - Major loss of function.
          • In Progress
          causes

          Bug - A problem which impairs or prevents the functions of the product. MDEV-36156 building MSAN -stdlib=libc++ test requires -fsanitize=memory

          • Major - Major loss of function.
          • Closed
          relates to

          Bug - A problem which impairs or prevents the functions of the product. MDEV-20377 Make WITH_MSAN more usable

          • Major - Major loss of function.
          • Closed

          Activity

            marko Marko Mäkelä added a comment -

            Thank you, this looks OK, except for one thing.

            I think that changing the logic around WITH_DBUG_TRACE (which affects CMAKE_BUILD_TYPE=Debug) needs to be addressed in a separate ticket that should be linked with MDEV-29613. It has nothing directly to do with MemorySanitizer, but with the performance overhead of some logging that is not strictly necessary while running the regression test suite.

            marko Marko Mäkelä added a comment - Thank you, this looks OK, except for one thing. I think that changing the logic around WITH_DBUG_TRACE (which affects CMAKE_BUILD_TYPE=Debug ) needs to be addressed in a separate ticket that should be linked with MDEV-29613 . It has nothing directly to do with MemorySanitizer, but with the performance overhead of some logging that is not strictly necessary while running the regression test suite.

            People

              danblack Daniel Black
              danblack Daniel Black
              Votes:
              0 Vote for this issue
              Watchers:
              3 Start watching this issue

              Dates

                Created:
                Updated:
                Resolved:

                Git Integration

                  Error rendering 'com.xiplink.jira.git.jira_git_plugin:git-issue-webpanel'. Please contact your Jira administrators.