vfs.git pile 13 - killing __copy_in_user()

There used to be 6 places in the entire tree calling __copy_in_user(),
all of them bogus. Four got killed off in work.drm branch, this takes care of the remaining ones and kills the definition of that sucker.

The following changes since commit aeba03903063e9fea885d32318e1546d9b7e83a8:

mga: switch compat ioctls to drm_ioctl_kernel() (2017-07-04 13:16:28 -0400)

are available in the git repository at:

git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs.git work.__copy_in_user

for you to fetch changes up to 119d0312c766773ca3238b9d926077664eed22be:

kill __copy_in_user() (2017-07-04 13:31:27 -0400)

----------------------------------------------------------------
Al Viro (3):
sanitize do_i2c_smbus_ioctl()
Merge branch 'work.drm' into work.__copy_in_user kill __copy_in_user()

 fs/compat_ioctl.c       | 19 +++++++++----------
 include/linux/uaccess.h |  6 ------
 2 files changed, 9 insertions(+), 16 deletions(-)

This branch is garbage.

You seem to have merged in some *old* drm branch into it, so it has all those drm commits that I merged earlier, except they are not the *same* commits.

So I'm not pulling this. Ugly duplicate commits and just unnecessary conflicts showing up again.

Linus

Huh?

<starts cursing>

42bf8be378270f85ca926452ec473e65ea26d94b is where vfs.git #work.__copy_in_user is 119d0312c766773ca3238b9d926077664eed22be is what mentioned in the pull request and that's where the local branch is.

OK, force-pushed, should propagate in a few. I wonder how I'd missed the warning spewed by git request-pull - it did produce one...