Commit graph

16 commits

Author SHA1 Message Date
Marko Mäkelä
be85d3e61b Merge 10.2 into 10.3 2019-05-14 17:18:46 +03:00
Vicențiu Ciorbaru
cb248f8806 Merge branch '5.5' into 10.1 2019-05-11 22:19:05 +03:00
Vicențiu Ciorbaru
5543b75550 Update FSF Address
* Update wrong zip-code
2019-05-11 21:29:06 +03:00
Marko Mäkelä
34841d2305 Merge bb-10.2-ext into 10.3 2017-12-12 09:57:17 +02:00
Sergey Vojtovich
1029b22feb MDEV-13728 - Import MySQL 5.7 atomic operations for MSVC and Solaris
gcc_sync.h, solaris.h, generic-msvc.h copied verbatim.
2017-11-27 18:35:53 +04:00
Sergey Vojtovich
c10e523d78 MDEV-11212 - Clean-up MariaDB atomic operations
No more black magic for gcc atomic bultins. Fixes some warnings.
2016-12-13 16:38:20 +04:00
Sergey Vojtovich
81f280789b MDEV-11212 - Clean-up MariaDB atomic operations
Moved gcc specific code to gcc_builtins.h.
Moved intptr into the black magic code block.
Moved definition of atomic operations for "long" out of black magic code block.
2016-12-13 16:38:20 +04:00
Sergey Vojtovich
8ff3b892ae MDEV-11212 - Clean-up MariaDB atomic operations
Removed MY_ATOMIC_MODE_DUMMY and WITH_ATOMIC_OPS: MariaDB is not functional
without atomic operations anymore.
2016-12-13 16:38:20 +04:00
Sergei Golubchik
66fd45afce MDEV-7398 mysqld segfaults on FreeBSD 10.1 i386 when built with clang 3.4
in cmake tests let's treat clang like gcc (same options,
same builtins) in many cases.

* don't check the compiler when
  * testing for -fvisibility=hidden support
  * testing for HAVE_ABI_CXA_DEMANGLE
  * testing for HAVE_GCC_ATOMIC_BUILTINS
  * when removing options with string(replace)
  * when running ${CC} --version (ignore the error instead)
* run ABI checks for clang
* use "canonical" gcc flags for clang
* fix groonga too

Also:

* add cmake detection for gcc __atomic_* builtins. they might be
  supported (__ATOMIC_SEQ_CST is defined), but not for all operand
  sizes. In particular, 64-bit atomic load is problematic on i386
* cache check results for Windows
* remove the test for HAVE_CXXABI_H (HAVE_ABI_CXA_DEMANGLE is
  suffifient)
2015-06-16 23:58:21 +02:00
Sergey Vojtovich
9836fd5d13 MDEV-5766 - my_atomic_load does memory writes
my_atomic_load() is implemented as __sync_fetch_and_or(var, 0) which
writes or-ed value back to var. Memory writes as such have worse
performance and scalability than reads.

gcc 4.7 and up offers better facility for atomic loads/stores. Use it
whenever it is available.
2014-03-07 11:43:06 +04:00
Kent Boortz
02e07e3b51 Updated/added copyright headers 2011-06-30 17:46:53 +02:00
Davi Arnaut
93e38e8a3e Bug#22320: my_atomic-t unit test fails
Bug#52261: 64 bit atomic operations do not work on Solaris i386
           gcc in debug compilation

One of the various problems was that the source operand to
CMPXCHG8b was marked as a input/output operand, causing GCC
to use the EBX register as the destination register for the
CMPXCHG8b instruction. This could lead to crashes as the EBX
register is also implicitly used by the instruction, causing
the value to be potentially garbaged and a protection fault
once the value is used to access a position in memory.

Another problem was the lack of proper clobbers for the atomic
operations and, also, a discrepancy between the implementations
for the Compare and Set operation. The specific problems are
described and fixed by Kristian Nielsen patches:

Patch: 1

Fix bugs in my_atomic_cas*(val,cmp,new) that *cmp is accessed
after CAS succeds.

In the gcc builtin implementation, problem was that *cmp was
read again after atomic CAS to check if old *val == *cmp;
this fails if CAS is successful and another thread modifies
*cmp in-between.

In the x86-gcc implementation, problem was that *cmp was set
also in the case of successful CAS; this means there is a
window where it can clobber a value written by another thread
after successful CAS.

Patch 2:

Add a GCC asm "memory" clobber to primitives that imply a
memory barrier.

This signifies to GCC that any potentially aliased memory
must be flushed before the operation, and re-read after the
operation, so that read or modification in other threads of
such memory values will work as intended.

In effect, it makes these primitives work as memory barriers
for the compiler as well as the CPU. This is better and more
correct than adding "volatile" to variables.

include/atomic/gcc_builtins.h:
  Do not read from *cmp after the operation as it might be
  already gone if the operation was successful.
include/atomic/nolock.h:
  Prefer system provided atomics over the broken x86 asm.
include/atomic/x86-gcc.h:
  Do not mark source operands as input/output operands.
  Add proper memory clobbers.
include/my_atomic.h:
  Add notes about my_atomic_add and my_atomic_cas behaviors.
unittest/mysys/my_atomic-t.c:
  Remove work around, if it fails, there is either a problem
  with the atomic operations code or the specific compiler
  version should be black-listed.
2010-07-23 09:37:10 -03:00
Davi Arnaut
cd41cd953d Bug#22320: my_atomic-t unit test fails
The atomic operations implementation on 5.1 has a few problems,
which might cause tests to abort randomly. Since no code in 5.1
uses atomic operations, simply remove the code.
2010-07-05 09:00:39 -03:00
Marc Alff
382ae22290 WL#2595 kernel-independent atomic operations
Backport from 6.0.14 to 5.6.0

Original code from Sergei Golubchik
2009-11-17 17:11:32 -07:00
Mats Kindahl
d47710c8dc WL#5016: Fix header file include guards
Adding header include file guards to files that are missing such.
2009-09-23 23:32:31 +02:00
unknown
26af3c7e35 Bug#33728 Atomic builtins
Use compiler provided atomic builtins as a 'backend' for
MySQL's atomic primitives. The builtins are available on
a handful of platforms and compilers.


configure.in:
  Check if the compiler provides atomic builtins and
  that __sync_lock_test_and_set stores the correct value.
include/atomic/nolock.h:
  Use the atomic builtins if available.
include/atomic/gcc_builtins.h:
  Implement the atomic ADD, SWAP, CAS, STORE (or operation
  optimized away) and LOAD primitives using atomic builtins
  provided by the compiler.
2008-01-11 20:34:36 -02:00