opal,ompi: move container_of definition to its own header #12724
Conversation
wenduwan
force-pushed
the
define_container_of
branch
from
9918983 to
1d112bd
Compare
|
@jsquyres I touched usnic for this PR so it would be great to have your eyes. |
wenduwan
force-pushed
the
define_container_of
branch
from
1d112bd to
f462fe0
Compare
| BEGIN_C_DECLS | ||
|
|
||
| #ifndef container_of | ||
| # define container_of(ptr, type, member) ((type *) (((char *) (ptr)) - offsetof(type, member))) |
There was a problem hiding this comment.
Should we namespace this to opal_container_of so that there is no conflict if another project has the brilliant idea of providing their own copy?
There was a problem hiding this comment.
Hmmmm container_of is quite common in the wild and we are actually not making a fancier option than the default one, so I'm inclined to keep the name(also we are doing ifndef so it should be safe).
There was a problem hiding this comment.
I missed the #ifndef so I guess it's fine and keeps the impact small.
Centralize container_of macro definition so that we don't have to define it everywhere. Signed-off-by: Wenduo Wang <[email protected]>
wenduwan
force-pushed
the
define_container_of
branch
from
f462fe0 to
8acb59d
Compare
| BEGIN_C_DECLS | ||
|
|
||
| #ifndef container_of | ||
| # define container_of(ptr, type, member) ((type *) (((char *) (ptr)) - offsetof(type, member))) |
There was a problem hiding this comment.
I missed the #ifndef so I guess it's fine and keeps the impact small.
|
I'm running more tests. Will open backports today. |
Centralize container_of macro definition so that we don't have to define it everywhere.