MARK CALLAGHAN writes:
> What about ARM which has a much weaker memory model compared to x86?
Yes. There is in fact a bug about it:
https://mariadb.atlassian.net/browse/MDEV-7658
- Kristian.
___
Mailing list: https://launchpad.net/~maria-develo
What about ARM which has a much weaker memory model compared to x86?
On Sat, May 23, 2015 at 12:46 AM, Kristian Nielsen wrote:
> Stewart Smith writes:
>
> >> It is probably still incorrect for some other architectures (eg.
> multi-core
> >> ARM).
> >
> > What about POWER? It's officially suppor
Stewart Smith writes:
>> It is probably still incorrect for some other architectures (eg. multi-core
>> ARM).
>
> What about POWER? It's officially supported by Maria now :)
There is special code for POWER (#ifdef), so it should work...
- Kristian.
__
Kristian Nielsen writes:
> MARK CALLAGHAN writes:
>
>> Did this ever get resolved? I enjoyed reading this, you did a nice job
>> explaining the problem.
>
> It was resolved for x86 in the sense that the code was reverted to the old
> behaviour. The code is semantically incorrect (uses the wrong
MARK CALLAGHAN writes:
> Did this ever get resolved? I enjoyed reading this, you did a nice job
> explaining the problem.
It was resolved for x86 in the sense that the code was reverted to the old
behaviour. The code is semantically incorrect (uses the wrong barriers), but
it happens to generate
Did this ever get resolved? I enjoyed reading this, you did a nice job
explaining the problem.
Linux had some fun in this area recently --
https://groups.google.com/forum/#!topic/mechanical-sympathy/QbmpZxp6C64
The InnoDB thread that checked waiting threads once per second for wakeup
was a hack t
Sergey Vojtovich writes:
> Never seen this before. InnoDB of 5.5 and InnoDB/XtraDB of 10.0 both have:
>
> UNIV_INTERN
> void
> mutex_set_waiters(
> /*==*/
> ib_mutex_t* mutex, /*!< in: mutex */
> ulint n) /*!< in: value to set */
> {
> volat
Kristian,
thanks for your answers! My comments inline.
On Thu, Nov 06, 2014 at 06:48:18PM +0100, Kristian Nielsen wrote:
> Sergey Vojtovich writes:
>
> >> # First spin wait, hoping to get the mutex without yielding.
> >> os_compare_and_swap_ulint(&mutex->waiters, 0, 1);
> > Where did you se
Sergey Vojtovich writes:
>> # First spin wait, hoping to get the mutex without yielding.
>> os_compare_and_swap_ulint(&mutex->waiters, 0, 1);
> Where did you see "os_compare_and_swap_ulint(&mutex->waiters, 0, 1)"?
> FWICS waiters of InnoDB mutex are non-atomic loads/stores.
mutex_enter_func(
Hi Kristian,
On Thu, Nov 06, 2014 at 03:12:55PM +0100, Kristian Nielsen wrote:
> I have been debugging some test failures seen in Buildbot:
>
> https://mariadb.atlassian.net/browse/MDEV-7026
>
> After some debugging, this is what I think is going on. I will refer to
> MariaDB 5.5 sources, th
I have been debugging some test failures seen in Buildbot:
https://mariadb.atlassian.net/browse/MDEV-7026
After some debugging, this is what I think is going on. I will refer to
MariaDB 5.5 sources, though the problem is also in 10.0.
InnoDB/XtraDB mutexes are taken in mutex_enter_func() and
11 matches
Mail list logo
