On Sun, Mar 11, 2018 at 01:15:59PM -0700, Linus Torvalds wrote:
> On Sun, Mar 11, 2018 at 3:55 AM, Dominik Brodowski
> <li...@dominikbrodowski.net> wrote:
> > Here is a first set of patches which reduce the number of syscall
> > invocations
> > from within the kernel.
> This all looks ok to me. I think there may be some room for cleanup
> and bikeshedding later (the "ksys_mmap_pgoff()" name made me go "that
> sounds more like a real helper than a kernel syscall", for example),
> but on the whole I think it's best to start with this kind of fairly
> direct translation.
> In fact, because it all looks *so* mechanical, I'd even be willing to
> just pull this into 4.16, if that makes the real work that this
> depends on easier?
> Anyway, ack from me, qith the only question being how you want to
> handle the patches. Direct pull request to me, go through the -mm
> tree, or just keep them in your own git tree and then build upon it
> and send it all later?
Thanks for your feedback! As I think this deserves at least another round of
review and some testing by 0day, I've set up a git tree at
This will allow me to re-base and tweak the patches for a while, and add
some more similar patches on top. For example, besides the first set of
35+1 patches in the aforementioned branch "syscalls-next" there are a few
other patches -- including one you might find familiar -- in "syscalls-WIP"
which are not fully ready yet.
The patches in that repository also include the following changes since
- rebase to v4.15-rc5
- sys_ioperm already got its SYSCALL_DEFINE3
- add ACK from Richard Weinberger
- modify CC -> Cc (suggested by Ingo Molnar)
- update comment in include/linux/syscalls.h (suggested by Ingo Molnar and
- separate declarations from definitions with newlines in
include/linux/syscalls.h; add comment on ksys_close() (suggested by
- expand commit messages (suggested by Christoph Hellwig)
- add patch 36: fs: add ksys_open() wrapper; remove in-kernel calls to