- developers - lists.mariadb.org

[Maria-developers] Updated (by Guest): Using the Valgrind API in mysqld (23)
by worklog-noreply＠askmonty.org 22 May '09

22 May '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Using the Valgrind API in mysqld CREATION DATE..: Fri, 22 May 2009, 11:43 SUPERVISOR.....: Monty IMPLEMENTOR....: Knielsen COPIES TO......: CATEGORY.......: Server-RawIdeaBin TASK ID........: 23 (http://askmonty.org/worklog/?tid=23) VERSION........: Benchmarks-3.0 STATUS.........: Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 40 (hours remain) ORIG. ESTIMATE.: 40 PROGRESS NOTES: -=-=(Guest - Fri, 22 May 2009, 12:04)=-=- High-Level Specification modified. --- /tmp/wklog.23.old.18061 2009-05-22 12:04:05.000000000 +0300 +++ /tmp/wklog.23.new.18061 2009-05-22 12:04:05.000000000 +0300 @@ -26,3 +26,5 @@ initialised, it is possible to detect problems earlier, speeding up debugging. Such code can be added in more places over time as development and debugging goes on. + +See also a patch here: http://bugs.mysql.com/bug.php?id=44582 -=-=(Knielsen - Fri, 22 May 2009, 11:52)=-=- High-Level Specification modified. --- /tmp/wklog.23.old.17628 2009-05-22 11:52:33.000000000 +0300 +++ /tmp/wklog.23.new.17628 2009-05-22 11:52:33.000000000 +0300 @@ -1 +1,28 @@ +With custom memory allocators, using the Valgrind APIs we can tell Valgrind when +a memory block is allocated (so that data read from memory is marked as undefined +instead of being defined or not at random depending on prior use); and when a +memory block is freed (so that use after freeing can be reported as an error). +In some cases cheking for leaks may also be appropriate. + +Another possibility is to add an explicit check for whether memory is defined. + +One place this would be useful is when calling libz. Due to the design of that +library, Valgrind produces lots of false alarms about using undefined values +(I think the issue is that it runs a few bytes off of initialized memory to +reduce boundary checks in each loop iteration, then after the loop has checks to +avoid using the undefined part of the result). This means we have lots of libz +Valgrind suppressions and continue to add more as new warnings surface. So we +might easily miss a real problem in this area. This could be improved by adding +explicit checks at the call to libz functions that the passed memory is properly +defined. + +Another use is to improve debugging. It is often the case when debugging a +warning about using un-initialised memory that the detection happens long after +the real problem, the un-initialized value being passed along through the code +for a long time before being detected. This makes debugging the problem slow. + +By adding in strategic places code that asserts that a specific value must be +initialised, it is possible to detect problems earlier, speeding up debugging. +Such code can be added in more places over time as development and debugging +goes on. DESCRIPTION: Valgrind (the memcheck tool) has some very useful APIs that can be used in mysqld when testing with Valgrind to improve testing and/or debugging: file:///usr/share/doc/valgrind/html/mc-manual.html#mc-manual.clientreqs file:///usr/share/doc/valgrind/html/mc-manual.html#mc-manual.mempools This worklog is about adding configure checks and headers to allow to use these in a way that continues to work on machines where the Valgrind headers or functionality is missing. It also includes adding some basic Valgrind enhancements: - Adding Valgrind annotations to custom memory allocators so that Valgrind can detect leaks, use-before-init, and use-after-free problems also for these allocators. - Adding checks for definedness in appropriate places (eg. when calling libz). HIGH-LEVEL SPECIFICATION: With custom memory allocators, using the Valgrind APIs we can tell Valgrind when a memory block is allocated (so that data read from memory is marked as undefined instead of being defined or not at random depending on prior use); and when a memory block is freed (so that use after freeing can be reported as an error). In some cases cheking for leaks may also be appropriate. Another possibility is to add an explicit check for whether memory is defined. One place this would be useful is when calling libz. Due to the design of that library, Valgrind produces lots of false alarms about using undefined values (I think the issue is that it runs a few bytes off of initialized memory to reduce boundary checks in each loop iteration, then after the loop has checks to avoid using the undefined part of the result). This means we have lots of libz Valgrind suppressions and continue to add more as new warnings surface. So we might easily miss a real problem in this area. This could be improved by adding explicit checks at the call to libz functions that the passed memory is properly defined. Another use is to improve debugging. It is often the case when debugging a warning about using un-initialised memory that the detection happens long after the real problem, the un-initialized value being passed along through the code for a long time before being detected. This makes debugging the problem slow. By adding in strategic places code that asserts that a specific value must be initialised, it is possible to detect problems earlier, speeding up debugging. Such code can be added in more places over time as development and debugging goes on. See also a patch here: http://bugs.mysql.com/bug.php?id=44582 ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Updated (by Knielsen): Using the Valgrind API in mysqld (23)
by worklog-noreply＠askmonty.org 22 May '09

22 May '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Using the Valgrind API in mysqld CREATION DATE..: Fri, 22 May 2009, 11:43 SUPERVISOR.....: Monty IMPLEMENTOR....: Knielsen COPIES TO......: CATEGORY.......: Server-RawIdeaBin TASK ID........: 23 (http://askmonty.org/worklog/?tid=23) VERSION........: Benchmarks-3.0 STATUS.........: Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 40 (hours remain) ORIG. ESTIMATE.: 40 PROGRESS NOTES: -=-=(Knielsen - Fri, 22 May 2009, 11:52)=-=- High-Level Specification modified. --- /tmp/wklog.23.old.17628 2009-05-22 11:52:33.000000000 +0300 +++ /tmp/wklog.23.new.17628 2009-05-22 11:52:33.000000000 +0300 @@ -1 +1,28 @@ +With custom memory allocators, using the Valgrind APIs we can tell Valgrind when +a memory block is allocated (so that data read from memory is marked as undefined +instead of being defined or not at random depending on prior use); and when a +memory block is freed (so that use after freeing can be reported as an error). +In some cases cheking for leaks may also be appropriate. + +Another possibility is to add an explicit check for whether memory is defined. + +One place this would be useful is when calling libz. Due to the design of that +library, Valgrind produces lots of false alarms about using undefined values +(I think the issue is that it runs a few bytes off of initialized memory to +reduce boundary checks in each loop iteration, then after the loop has checks to +avoid using the undefined part of the result). This means we have lots of libz +Valgrind suppressions and continue to add more as new warnings surface. So we +might easily miss a real problem in this area. This could be improved by adding +explicit checks at the call to libz functions that the passed memory is properly +defined. + +Another use is to improve debugging. It is often the case when debugging a +warning about using un-initialised memory that the detection happens long after +the real problem, the un-initialized value being passed along through the code +for a long time before being detected. This makes debugging the problem slow. + +By adding in strategic places code that asserts that a specific value must be +initialised, it is possible to detect problems earlier, speeding up debugging. +Such code can be added in more places over time as development and debugging +goes on. DESCRIPTION: Valgrind (the memcheck tool) has some very useful APIs that can be used in mysqld when testing with Valgrind to improve testing and/or debugging: file:///usr/share/doc/valgrind/html/mc-manual.html#mc-manual.clientreqs file:///usr/share/doc/valgrind/html/mc-manual.html#mc-manual.mempools This worklog is about adding configure checks and headers to allow to use these in a way that continues to work on machines where the Valgrind headers or functionality is missing. It also includes adding some basic Valgrind enhancements: - Adding Valgrind annotations to custom memory allocators so that Valgrind can detect leaks, use-before-init, and use-after-free problems also for these allocators. - Adding checks for definedness in appropriate places (eg. when calling libz). HIGH-LEVEL SPECIFICATION: With custom memory allocators, using the Valgrind APIs we can tell Valgrind when a memory block is allocated (so that data read from memory is marked as undefined instead of being defined or not at random depending on prior use); and when a memory block is freed (so that use after freeing can be reported as an error). In some cases cheking for leaks may also be appropriate. Another possibility is to add an explicit check for whether memory is defined. One place this would be useful is when calling libz. Due to the design of that library, Valgrind produces lots of false alarms about using undefined values (I think the issue is that it runs a few bytes off of initialized memory to reduce boundary checks in each loop iteration, then after the loop has checks to avoid using the undefined part of the result). This means we have lots of libz Valgrind suppressions and continue to add more as new warnings surface. So we might easily miss a real problem in this area. This could be improved by adding explicit checks at the call to libz functions that the passed memory is properly defined. Another use is to improve debugging. It is often the case when debugging a warning about using un-initialised memory that the detection happens long after the real problem, the un-initialized value being passed along through the code for a long time before being detected. This makes debugging the problem slow. By adding in strategic places code that asserts that a specific value must be initialised, it is possible to detect problems earlier, speeding up debugging. Such code can be added in more places over time as development and debugging goes on. ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Updated (by Knielsen): Using the Valgrind API in mysqld (23)
by worklog-noreply＠askmonty.org 22 May '09

22 May '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Using the Valgrind API in mysqld CREATION DATE..: Fri, 22 May 2009, 11:43 SUPERVISOR.....: Monty IMPLEMENTOR....: Knielsen COPIES TO......: CATEGORY.......: Server-RawIdeaBin TASK ID........: 23 (http://askmonty.org/worklog/?tid=23) VERSION........: Benchmarks-3.0 STATUS.........: Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 40 (hours remain) ORIG. ESTIMATE.: 40 PROGRESS NOTES: -=-=(Knielsen - Fri, 22 May 2009, 11:52)=-=- High-Level Specification modified. --- /tmp/wklog.23.old.17628 2009-05-22 11:52:33.000000000 +0300 +++ /tmp/wklog.23.new.17628 2009-05-22 11:52:33.000000000 +0300 @@ -1 +1,28 @@ +With custom memory allocators, using the Valgrind APIs we can tell Valgrind when +a memory block is allocated (so that data read from memory is marked as undefined +instead of being defined or not at random depending on prior use); and when a +memory block is freed (so that use after freeing can be reported as an error). +In some cases cheking for leaks may also be appropriate. + +Another possibility is to add an explicit check for whether memory is defined. + +One place this would be useful is when calling libz. Due to the design of that +library, Valgrind produces lots of false alarms about using undefined values +(I think the issue is that it runs a few bytes off of initialized memory to +reduce boundary checks in each loop iteration, then after the loop has checks to +avoid using the undefined part of the result). This means we have lots of libz +Valgrind suppressions and continue to add more as new warnings surface. So we +might easily miss a real problem in this area. This could be improved by adding +explicit checks at the call to libz functions that the passed memory is properly +defined. + +Another use is to improve debugging. It is often the case when debugging a +warning about using un-initialised memory that the detection happens long after +the real problem, the un-initialized value being passed along through the code +for a long time before being detected. This makes debugging the problem slow. + +By adding in strategic places code that asserts that a specific value must be +initialised, it is possible to detect problems earlier, speeding up debugging. +Such code can be added in more places over time as development and debugging +goes on. DESCRIPTION: Valgrind (the memcheck tool) has some very useful APIs that can be used in mysqld when testing with Valgrind to improve testing and/or debugging: file:///usr/share/doc/valgrind/html/mc-manual.html#mc-manual.clientreqs file:///usr/share/doc/valgrind/html/mc-manual.html#mc-manual.mempools This worklog is about adding configure checks and headers to allow to use these in a way that continues to work on machines where the Valgrind headers or functionality is missing. It also includes adding some basic Valgrind enhancements: - Adding Valgrind annotations to custom memory allocators so that Valgrind can detect leaks, use-before-init, and use-after-free problems also for these allocators. - Adding checks for definedness in appropriate places (eg. when calling libz). HIGH-LEVEL SPECIFICATION: With custom memory allocators, using the Valgrind APIs we can tell Valgrind when a memory block is allocated (so that data read from memory is marked as undefined instead of being defined or not at random depending on prior use); and when a memory block is freed (so that use after freeing can be reported as an error). In some cases cheking for leaks may also be appropriate. Another possibility is to add an explicit check for whether memory is defined. One place this would be useful is when calling libz. Due to the design of that library, Valgrind produces lots of false alarms about using undefined values (I think the issue is that it runs a few bytes off of initialized memory to reduce boundary checks in each loop iteration, then after the loop has checks to avoid using the undefined part of the result). This means we have lots of libz Valgrind suppressions and continue to add more as new warnings surface. So we might easily miss a real problem in this area. This could be improved by adding explicit checks at the call to libz functions that the passed memory is properly defined. Another use is to improve debugging. It is often the case when debugging a warning about using un-initialised memory that the detection happens long after the real problem, the un-initialized value being passed along through the code for a long time before being detected. This makes debugging the problem slow. By adding in strategic places code that asserts that a specific value must be initialised, it is possible to detect problems earlier, speeding up debugging. Such code can be added in more places over time as development and debugging goes on. ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] New (by Knielsen): Using the Valgrind API in mysqld (23)
by worklog-noreply＠askmonty.org 22 May '09

22 May '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Using the Valgrind API in mysqld CREATION DATE..: Fri, 22 May 2009, 11:43 SUPERVISOR.....: Monty IMPLEMENTOR....: Knielsen COPIES TO......: CATEGORY.......: Server-RawIdeaBin TASK ID........: 23 (http://askmonty.org/worklog/?tid=23) VERSION........: Benchmarks-3.0 STATUS.........: Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 40 (hours remain) ORIG. ESTIMATE.: 40 PROGRESS NOTES: DESCRIPTION: Valgrind (the memcheck tool) has some very useful APIs that can be used in mysqld when testing with Valgrind to improve testing and/or debugging: file:///usr/share/doc/valgrind/html/mc-manual.html#mc-manual.clientreqs file:///usr/share/doc/valgrind/html/mc-manual.html#mc-manual.mempools This worklog is about adding configure checks and headers to allow to use these in a way that continues to work on machines where the Valgrind headers or functionality is missing. It also includes adding some basic Valgrind enhancements: - Adding Valgrind annotations to custom memory allocators so that Valgrind can detect leaks, use-before-init, and use-after-free problems also for these allocators. - Adding checks for definedness in appropriate places (eg. when calling libz). ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Newly Added (by Knielsen): Using the Valgrind API in mysqld (23)
by worklog-noreply＠askmonty.org 22 May '09

22 May '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Using the Valgrind API in mysqld CREATION DATE..: Fri, 22 May 2009, 11:43 SUPERVISOR.....: Monty IMPLEMENTOR....: Knielsen COPIES TO......: CATEGORY.......: Server-RawIdeaBin TASK ID........: 23 (http://askmonty.org/worklog/?tid=23) VERSION........: Benchmarks-3.0 STATUS.........: Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 40 (hours remain) ORIG. ESTIMATE.: 40 PROGRESS NOTES: DESCRIPTION: Valgrind (the memcheck tool) has some very useful APIs that can be used in mysqld when testing with Valgrind to improve testing and/or debugging: file:///usr/share/doc/valgrind/html/mc-manual.html#mc-manual.clientreqs file:///usr/share/doc/valgrind/html/mc-manual.html#mc-manual.mempools This worklog is about adding configure checks and headers to allow to use these in a way that continues to work on machines where the Valgrind headers or functionality is missing. It also includes adding some basic Valgrind enhancements: - Adding Valgrind annotations to custom memory allocators so that Valgrind can detect leaks, use-before-init, and use-after-free problems also for these allocators. - Adding checks for definedness in appropriate places (eg. when calling libz). ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] New (by Knielsen): Using the Valgrind API in mysqld (23)
by worklog-noreply＠askmonty.org 22 May '09

22 May '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Using the Valgrind API in mysqld CREATION DATE..: Fri, 22 May 2009, 11:43 SUPERVISOR.....: Monty IMPLEMENTOR....: Knielsen COPIES TO......: CATEGORY.......: Server-RawIdeaBin TASK ID........: 23 (http://askmonty.org/worklog/?tid=23) VERSION........: Benchmarks-3.0 STATUS.........: Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 40 (hours remain) ORIG. ESTIMATE.: 40 PROGRESS NOTES: DESCRIPTION: Valgrind (the memcheck tool) has some very useful APIs that can be used in mysqld when testing with Valgrind to improve testing and/or debugging: file:///usr/share/doc/valgrind/html/mc-manual.html#mc-manual.clientreqs file:///usr/share/doc/valgrind/html/mc-manual.html#mc-manual.mempools This worklog is about adding configure checks and headers to allow to use these in a way that continues to work on machines where the Valgrind headers or functionality is missing. It also includes adding some basic Valgrind enhancements: - Adding Valgrind annotations to custom memory allocators so that Valgrind can detect leaks, use-before-init, and use-after-free problems also for these allocators. - Adding checks for definedness in appropriate places (eg. when calling libz). ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Updated (by Knielsen): Implement mysql-test output parser for Buildbot (22)
by worklog-noreply＠askmonty.org 22 May '09

22 May '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Implement mysql-test output parser for Buildbot CREATION DATE..: Thu, 21 May 2009, 22:19 SUPERVISOR.....: Knielsen IMPLEMENTOR....: COPIES TO......: CATEGORY.......: Other TASK ID........: 22 (http://askmonty.org/worklog/?tid=22) VERSION........: Benchmarks-3.0 STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 30 (hours remain) ORIG. ESTIMATE.: 30 PROGRESS NOTES: -=-=(Knielsen - Thu, 21 May 2009, 22:33)=-=- High Level Description modified. --- /tmp/wklog.22.old.27679 2009-05-21 22:33:35.000000000 +0300 +++ /tmp/wklog.22.new.27679 2009-05-21 22:33:35.000000000 +0300 @@ -9,3 +9,6 @@ http://djmitche.github.com/buildbot/docs/0.7.10/#Writing-New-BuildSteps +Later, once we get the infrastructure to write Buildbot results into a MySQL +database, we want to extend this to also insert into the database all test +failures and the mysqltest failure output (for cross-reference search). DESCRIPTION: Like in Pushbuild at MySQL AB, we want to have buildbot parse the output of mysql-test-run for test errors so we can display on the front page the name of any tests that failed. The parser can also count the number of tests completed so far so Buildbot can provide more accurate completion estimates. Buildbot already has support for plugging in such modules. See eg. http://djmitche.github.com/buildbot/docs/0.7.10/#Writing-New-BuildSteps Later, once we get the infrastructure to write Buildbot results into a MySQL database, we want to extend this to also insert into the database all test failures and the mysqltest failure output (for cross-reference search). ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] bzr commit into MariaDB 5.1, with Maria 1.5:maria branch (knielsen:2703)
by knielsen＠knielsen-hq.org 21 May '09

21 May '09

#At lp:maria 2703 knielsen(a)knielsen-hq.org 2009-05-20 Fix accessing ulong enum option as uint, failing on 64-bit big-endian. modified: mysys/my_getopt.c === modified file 'mysys/my_getopt.c' --- a/mysys/my_getopt.c 2009-04-25 10:05:32 +0000 +++ b/mysys/my_getopt.c 2009-05-20 15:34:34 +0000 @@ -647,7 +647,7 @@ static int setval(const struct my_option return EXIT_OUT_OF_MEMORY; break; case GET_ENUM: - if (((*(int*)result_pos)= find_type(argument, opts->typelib, 2) - 1) < 0) + if (((*(ulong *)result_pos)= find_type(argument, opts->typelib, 2) - 1) < 0) return EXIT_ARGUMENT_INVALID; break; case GET_SET: @@ -983,7 +983,7 @@ static void init_one_value(const struct *((int*) variable)= (int) getopt_ll_limit_value((int) value, option, NULL); break; case GET_ENUM: - *((uint*) variable)= (uint) value; + *((ulong*) variable)= (uint) value; break; case GET_UINT: *((uint*) variable)= (uint) getopt_ull_limit_value((uint) value, option, NULL); @@ -1221,7 +1221,7 @@ void my_print_variables(const struct my_ } break; case GET_ENUM: - printf("%s\n", get_type(optp->typelib, *(uint*) value)); + printf("%s\n", get_type(optp->typelib, *(ulong*) value)); break; case GET_STR: case GET_STR_ALLOC: /* fall through */

2 1

[Maria-developers] bzr commit into MariaDB 5.1, with Maria 1.5:maria branch (knielsen:2699)
by knielsen＠knielsen-hq.org 20 May '09

20 May '09

#At lp:maria 2699 knielsen(a)knielsen-hq.org 2009-05-11 [merge] Merge Toby Thain's Solaris fixes. modified: BUILD/SETUP.sh BUILD/compile-solaris-amd64 BUILD/compile-solaris-amd64-debug* BUILD/compile-solaris-amd64-forte BUILD/compile-solaris-amd64-forte-debug* storage/innobase/include/univ.i storage/pbxt/src/Makefile.am storage/pbxt/src/lock_xt.h storage/pbxt/src/pbms.h === modified file 'BUILD/SETUP.sh' --- a/BUILD/SETUP.sh 2009-05-06 12:03:24 +0000 +++ b/BUILD/SETUP.sh 2009-05-11 15:31:30 +0000 @@ -80,7 +80,13 @@ path=`dirname $0` . "$path/check-cpu" export AM_MAKEFLAGS -AM_MAKEFLAGS="-j 6" +# Default to a parallel build, but only if AM_MAKEFLAGS is not set. +# (So buildbots can easily disable this behaviour if required.) +if test -z "$AM_MAKEFLAGS" +then + AM_MAKEFLAGS="-j 6" +fi + # SSL library to use.--with-ssl will select our bundled yaSSL # implementation of SSL. To use openSSl you will nee too point out === modified file 'BUILD/compile-solaris-amd64' --- a/BUILD/compile-solaris-amd64 2007-04-12 11:20:38 +0000 +++ b/BUILD/compile-solaris-amd64 2009-05-09 04:01:53 +0000 @@ -1,55 +1,32 @@ -#!/usr/bin/bash +#!/bin/sh -function _find_mysql_root () ( - while [ "x$PWD" != "x/" ]; do - # Check if some directories are present - if [ -d BUILD -a -d sql -a -d mysys ]; then - echo "$PWD" - return 0 - fi - cd .. - done - return 1 -) +# Build setup for Solaris 10 + +# Make the Sun Freeware packages use the bundled Perl, instead of their own: +# ln -s `which perl` /usr/local/bin +# Your $PATH needs to include (in this order): +# /usr/local/bin:/usr/sfw/bin:/usr/ccs/bin +# (For Sun Freeware, bundled GNU utilities, Solaris ar, etc.) +# +# Required packages from http://sunfreeware.com/indexintel10.html : +# (The GNU m4 bundled with Solaris is too old.) +# automake-1.10.2-sol10-x86-local.gz +# autoconf-2.63-sol10-x86-local.gz +# m4-1.4.12-sol10-x86-local.gz +# libsigsegv-2.6-sol10-x86-local.gz +# libtool-1.5.24-sol10-x86-local.gz +# ( how to install these packages: + # wget ftp://ftp.sunfreeware.com/pub/freeware/intel/10/automake-1.10.2-sol10-x86-l… + # gunzip automake-1.10.2-sol10-x86-local.gz + # pkgadd -d automake-1.10.2-sol10-x86-local +# ) -make -k clean || true -/bin/rm -f */.deps/*.P config.cache - path=`dirname $0` -. "$path/autorun.sh" - -warning_flags="-Wimplicit -Wreturn-type -Wswitch -Wtrigraphs -Wcomment -W -Wchar-subscripts -Wformat -Wparentheses -Wsign-compare -Wwrite-strings -Wunused" -compiler_flags="-g -O3 -fno-omit-frame-pointer" +. "$path/SETUP.sh" +extra_flags="$amd64_cflags -D__sun -m64 -mtune=athlon64" +extra_configs="$amd64_configs $max_configs --with-libevent" -export CC CXX CFLAGS CXXFLAGS LDFLAGS LIBS -CC="gcc" -CXX="gcc" -CFLAGS="$warning_flags $compiler_flags" -CXXFLAGS="" -LDFLAGS="-O3 -g -static-libgcc" -LIBS=-lmtmalloc -root=$(_find_mysql_root) +LDFLAGS="-lmtmalloc -static-libgcc" +export LDFLAGS -$root/configure \ - --prefix=/usr/local/mysql \ - --localstatedir=/usr/local/mysql/data \ - --libexecdir=/usr/local/mysql/bin \ - --with-extra-charsets=complex \ - --enable-thread-safe-client \ - --enable-local-infile \ - --with-zlib-dir=bundled \ - --with-big-tables \ - --with-readline \ - --with-archive-storage-engine \ - --with-named-curses=-lcurses \ - --with-big-tables \ - --with-innodb \ - --with-berkeley-db \ - --with-example-storage-engine \ - --with-blackhole-storage-engine \ - --with-ndbcluster \ - --with-federated-storage-engine \ - --with-csv-storage-engine \ - --with-ssl \ - --with-embedded-server \ - --disable-shared +. "$path/FINISH.sh" === modified file 'BUILD/compile-solaris-amd64-debug' (properties changed: -x to +x) --- a/BUILD/compile-solaris-amd64-debug 2007-11-10 10:03:07 +0000 +++ b/BUILD/compile-solaris-amd64-debug 2009-05-09 04:01:53 +0000 @@ -1,10 +1,11 @@ -#! /bin/sh +#!/bin/sh + path=`dirname $0` . "$path/SETUP.sh" -amd64_cflags="-m64 -mtune=athlon64" -extra_flags="$amd64_cflags $debug_cflags $max_cflags" -c_warnings="$c_warnings $debug_extra_warnings" -cxx_warnings="$cxx_warnings $debug_extra_warnings" -extra_configs="$amd64_configs $debug_configs $max_configs --enable-thread-safe-client" +extra_flags="$amd64_cflags -D__sun -m64 -mtune=athlon64 $debug_cflags" +extra_configs="$amd64_configs $debug_configs $max_configs --with-libevent" + +LDFLAGS="-lmtmalloc -static-libgcc" +export LDFLAGS . "$path/FINISH.sh" === modified file 'BUILD/compile-solaris-amd64-forte' --- a/BUILD/compile-solaris-amd64-forte 2008-09-30 20:57:48 +0000 +++ b/BUILD/compile-solaris-amd64-forte 2009-05-09 04:01:53 +0000 @@ -1,53 +1,29 @@ -#! /bin/sh +#!/bin/sh + +# See file compile-solaris-amd64 for basic pre-requisites. + +# This build uses the Sun Studio compilers (cc, CC), available from: +# http://developers.sun.com/sunstudio/downloads/index.jsp +# Note that you may want to apply current patches, as the downloaded version +# is typically out of date. Download the PKG version if you intend to patch! + +# After installing, add /opt/SUNWspro/bin to your $PATH + -gmake -k clean || true -/bin/rm -f */.deps/*.P config.cache - path=`dirname $0` -. "$path/autorun.sh" +. "$path/SETUP.sh" -# For "optimal" code for this computer add -fast to EXTRA -# To compile 64 bit, add -m64 to EXTRA_64_BIT +extra_flags="-m64 -mt -D_FORTEC_ -xbuiltin=%all -xlibmil -xlibmopt -fns=no -xprefetch=auto -xprefetch_level=3" +extra_configs="$max_configs --with-libevent" -EXTRA_64_BIT="-m64" -EXTRA="-fast" +warnings="" +c_warnings="" +cxx_warnings="" +base_cxxflags="-noex" -# -# The following should not need to be touched -# - -export CC CXX CFLAGS CXXFLAGS LIBS -STD="-g -mt -D_FORTEC_ $EXTRA $EXTRA_64_BIT" -ASFLAGS="$EXTRA_64_BIT" -CC=cc-5.0 -CFLAGS="-Xa -xstrconst $STD" +CC=cc +CFLAGS="-xstrconst" CXX=CC -CXXFLAGS="-noex $STD" -LIBS=-lmtmalloc -./configure \ - --prefix=/usr/local/mysql \ - --localstatedir=/usr/local/mysql/data \ - --libexecdir=/usr/local/mysql/bin \ - --with-extra-charsets=complex \ - --enable-thread-safe-client \ - --enable-local-infile \ - --with-zlib-dir=bundled \ - --with-big-tables \ - --with-readline \ - --with-archive-storage-engine \ - --with-named-curses=-lcurses \ - --with-big-tables \ - --with-innodb \ - --with-example-storage-engine \ - --with-blackhole-storage-engine \ - --with-federated-storage-engine \ - --with-csv-storage-engine \ - --with-ssl \ - --enable-assembler - -# Not including: -# --with-ndbcluster -# --with-berkeley-db +LDFLAGS="-lmtmalloc" -gmake -j4 -test $? = 0 && make test +. "$path/FINISH.sh" === modified file 'BUILD/compile-solaris-amd64-forte-debug' (properties changed: -x to +x) --- a/BUILD/compile-solaris-amd64-forte-debug 2008-09-30 20:57:48 +0000 +++ b/BUILD/compile-solaris-amd64-forte-debug 2009-05-09 04:01:53 +0000 @@ -1,54 +1,27 @@ -#! /bin/sh +#!/bin/sh -gmake -k clean || true -/bin/rm -f */.deps/*.P config.cache - path=`dirname $0` -. "$path/autorun.sh" +. "$path/SETUP.sh" -# To compile 64 bit, add -m64 to EXTRA_64_BIT -EXTRA_64_BIT="-m64" +# Take only #define options - the others are gcc specific. +# (real fix is for SETUP.sh not to put gcc specific options in $debug_cflags) +DEFS="" +for F in $debug_cflags ; do + expr "$F" : "^-D" && DEFS="$DEFS $F" +done +debug_cflags="-O0 -g $DEFS" + +extra_flags="-m64 -mt -D_FORTEC_ -xlibmopt -fns=no $debug_cflags" +extra_configs="$max_configs --with-libevent $debug_configs" + +warnings="" +c_warnings="" +cxx_warnings="" +base_cxxflags="-noex" -# For "optimal" code for this computer add -fast to EXTRA. Note that -# this causes problem with debugging the program since -fast implies -# -xO5. -EXTRA="" - -# -# The following should not need to be touched -# - -export CC CXX CFLAGS CXXFLAGS -STD="-g -mt -D_FORTEC_ $EXTRA $EXTRA_64_BIT $debug_cflags" -ASFLAGS="$EXTRA_64_BIT" -CC=cc-5.0 -CFLAGS="-Xa -xstrconst $STD" +CC=cc +CFLAGS="-xstrconst" CXX=CC -CXXFLAGS="-noex $STD" -./configure \ - --prefix=/usr/local/mysql \ - --localstatedir=/usr/local/mysql/data \ - --libexecdir=/usr/local/mysql/bin \ - --with-extra-charsets=complex \ - --enable-thread-safe-client \ - --enable-local-infile \ - --with-zlib-dir=bundled \ - --with-big-tables \ - --with-readline \ - --with-archive-storage-engine \ - --with-named-curses=-lcurses \ - --with-big-tables \ - --with-innodb \ - --with-example-storage-engine \ - --with-blackhole-storage-engine \ - --with-federated-storage-engine \ - --with-csv-storage-engine \ - --with-ssl \ - --with-debug \ - --enable-assembler - -# Not including: -# --with-ndbcluster -# --with-berkeley-db +LDFLAGS="-lmtmalloc" -gmake -j4 +. "$path/FINISH.sh" === modified file 'storage/innobase/include/univ.i' --- a/storage/innobase/include/univ.i 2009-05-06 12:03:24 +0000 +++ b/storage/innobase/include/univ.i 2009-05-11 15:31:30 +0000 @@ -55,10 +55,10 @@ of the 32-bit x86 assembler in mutex ope # define UNIV_CAN_USE_X86_ASSEMBLER # endif -/* We only try to do explicit inlining of functions with gcc and -Microsoft Visual C++ */ +/* Enable explicit inlining of functions only for compilers known to +support it. */ -# if !defined(__GNUC__) +# if !defined(__GNUC__) && !defined(__SUNPRO_C) # undef UNIV_MUST_NOT_INLINE /* Remove compiler warning */ # define UNIV_MUST_NOT_INLINE # endif === modified file 'storage/pbxt/src/Makefile.am' --- a/storage/pbxt/src/Makefile.am 2009-04-01 06:40:11 +0000 +++ b/storage/pbxt/src/Makefile.am 2009-05-09 04:01:53 +0000 @@ -46,7 +46,7 @@ libpbxt_la_CFLAGS = $(AM_CFLAGS) -DMYSQ EXTRA_LIBRARIES = libpbxt.a noinst_LIBRARIES = libpbxt.a libpbxt_a_SOURCES = $(libpbxt_la_SOURCES) -libpbxt_a_CXXFLAGS = $(AM_CXXFLAGS) -Wno-long-long +libpbxt_a_CXXFLAGS = $(AM_CXXFLAGS) libpbxt_a_CFLAGS = $(AM_CFLAGS) -std=c99 EXTRA_DIST = CMakeLists.txt === modified file 'storage/pbxt/src/lock_xt.h' --- a/storage/pbxt/src/lock_xt.h 2009-04-02 10:03:14 +0000 +++ b/storage/pbxt/src/lock_xt.h 2009-05-09 04:01:53 +0000 @@ -208,9 +208,9 @@ inline void xt_atomic_dec2(volatile xtWo #elif defined(__GNUC__) __sync_fetch_and_sub(mptr, 1); #elif defined(XT_SPL_SOLARIS_LIB) - val1 = atomic_dec_16_nv(mptr); + atomic_dec_16_nv(mptr); #else - val1 = --(*mptr); + --(*mptr); #endif } === modified file 'storage/pbxt/src/pbms.h' --- a/storage/pbxt/src/pbms.h 2009-03-26 12:18:01 +0000 +++ b/storage/pbxt/src/pbms.h 2009-05-09 04:01:53 +0000 @@ -782,21 +782,28 @@ private: void deleteTempFiles() { - struct dirent entry; + struct dirent *entry; struct dirent *result; DIR *odir; int err; + size_t sz; char temp_file[100]; +#ifdef XT_SOLARIS + sz = sizeof(struct dirent) + pathconf("/tmp/", _PC_NAME_MAX); // Solaris, see readdir(3C) +#else + sz = sizeof(struct dirent); +#endif + entry = (struct dirent*)malloc(sz); if (!(odir = opendir("/tmp/"))) return; - err = readdir_r(odir, &entry, &result); + err = readdir_r(odir, entry, &result); while (!err && result) { const char **prefix = temp_prefix; while (*prefix) { - if (startsWith(entry.d_name, *prefix)) { - int pid = atoi(entry.d_name + strlen(*prefix)); + if (startsWith(entry->d_name, *prefix)) { + int pid = atoi(entry->d_name + strlen(*prefix)); /* If the process does not exist: */ if (kill(pid, 0) == -1 && errno == ESRCH) { @@ -807,9 +814,10 @@ private: prefix++; } - err = readdir_r(odir, &entry, &result); + err = readdir_r(odir, entry, &result); } closedir(odir); + free(entry); } }; #endif // PBMS_API

3 3

[Maria-developers] compile issue and the -lz flag
by Don Kehn 20 May '09

20 May '09

Hi all: when compiling on x64 systems (i.e. CentOS & Fedora core 10) and using the BUILD/compile-pentium64-max, I'm getting the -lz error "/usr/bin/ld: cannot find -lz" and when I simplify the ./comfigure line to a --without-zlib it works. So, after modifing the configure flags that results into: ./configure --prefix=/u02/local/mysql-5.1.34 --enable-assembler --with-extra-charsets=complex --enable-thread-safe-client --with-big-tables --with-readline --with-ssl --with-plugins=max --with-embedded-server --with-mysqld-ldflags=-all-static --with-client-ldflags=-all-static --without-zlib --enable-local-infile It still has problems with the -lz flag, so does anybody know what is bringing the -lz into the configure? Note: I've also tried --with-zlibdir=/usr/lib64, but same error. Thanks tons.

4 11