mmap(): MAP_ANON vs. MAP_ANONYMOUS

[Richard Hansen]

NetBSD developers,

The Austin Group (POSIX standards body) is considering standardizing
mmap() with anonymous memory:

http://austingroupbugs.net/view.php?id=850

The current draft of the new wording (see comment #2281) uses
MAP_ANONYMOUS as the name of the mmap() flag. NetBSD only defines
MAP_ANON, but otherwise seems conformant to the proposed wording.

Would you be opposed to defining MAP_ANONYMOUS as a synonym for
MAP_ANON? Would you prefer POSIX to specify MAP_ANON instead?

Thanks,
Richard

[Justin Cormack]

Post by Richard Hansen
NetBSD developers,
The Austin Group (POSIX standards body) is considering standardizing
http://austingroupbugs.net/view.php?id=850
The current draft of the new wording (see comment #2281) uses
MAP_ANONYMOUS as the name of the mmap() flag. NetBSD only defines
MAP_ANON, but otherwise seems conformant to the proposed wording.
Would you be opposed to defining MAP_ANONYMOUS as a synonym for
MAP_ANON? Would you prefer POSIX to specify MAP_ANON instead?

I would not be against us adding a synonym; I didn't realise FreeBSD
already had. I admit to forgetting which system uses which and having
to look it up frequently, but slightly prefer MAP_ANON.

Justin

[Matt Thomas]

Post by Justin Cormack

Post by Richard Hansen
NetBSD developers,
The Austin Group (POSIX standards body) is considering standardizing
http://austingroupbugs.net/view.php?id=850
The current draft of the new wording (see comment #2281) uses
MAP_ANONYMOUS as the name of the mmap() flag. NetBSD only defines
MAP_ANON, but otherwise seems conformant to the proposed wording.
Would you be opposed to defining MAP_ANONYMOUS as a synonym for
MAP_ANON? Would you prefer POSIX to specify MAP_ANON instead?

I would not be against us adding a synonym; I didn't realise FreeBSD
already had. I admit to forgetting which system uses which and having
to look it up frequently, but slightly prefer MAP_ANON.

Already done. MAP_ANONYNOUS has the value and MAP_ANON refers to
MAP_ANONYMOUS.