Further copy_from_user() discussion.

[Vadim Lobanov <vlobanov@speakeasy.net>]

Hi,

Interested by the recent discussions concerning the copy_from_user()
function, I browsed the 2.6.11.7 kernel source, and came up with a few
questions.

1. Is there any particular reason why __copy_from_user_ll() is currently
EXPORT_SYMBOL()ed for i386? At least none of the in-tree modules
currently seem to use it, and __copy_from_user() seems like what most
would want anyway. If __copy_from_user_ll() is unexported, it looks like
we can eliminate the BUG_ON() statement within it.

2. Would it be possible to eliminate the might_sleep() call in
copy_from_user()? It seems that, very soon after, the __copy_from_user()
macro does another might_sleep(), with very few instructions in between.
But there might be some trick here that I'm missing.

Please enlighten. :-)

-Vadim Lobanov