Re: [PATCH 2/2] Perform check in iov_iter_fault_in_readable() by check_readable_bytes()

[Andi Kleen]

Vitaly Mayatskikh <v.mayatskih@xxxxxxxxx> writes:

> This patch changes iov_iter_fault_in_readable() to use check_readable_bytes()
> instead of fault_in_pages_readable(). It allows iov_iter_fault_in_readable()
> callers to know how much data is accessible and proceed with it. It makes
> sys_write() more POSIX-friendly.

Can you describe how it makes it POSIX friendly?

My understanding was that EFAULT behaviour was undefined in POSIX.

The obvious hole in the patch is that all these checks are not race
free -- they don't pin pages -- so if there's a parallel unmap even
with your change they can still fail in the middle.

-Andi 

-- 
ak@xxxxxxxxxxxxxxx -- Speaking for myself only.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/