Skip to contentSkip to conversation

poll annotations - second round

GIT PULL Al Viro 2 participants 2 messages Feb 10, 2018 – Feb 11, 2018

Subsystems (2)
lore
Share & tools
lore

Al Viro

author maintainer
#0 lore raw

Preparations to solving the problems you've mentioned in the original poll
series. After this series, run

for V in IN OUT PRI ERR RDNORM RDBAND WRNORM WRBAND HUP RDHUP NVAL MSG; do

    L=`git grep -l -w POLL$V | grep -v '^t' | grep -v /um/ | grep -v '^sa' | grep -v '/poll.h$'|grep -v '^D'`
    for f in $L; do sed -i "-es/^\([^\"]*\)\(\<POLL$V\>\)/\\1E\\2/" $f; done

done

(ideally - just before -rc1) for bulk search-and-replace. Then cherry-pick
git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs.git#work.poll on top of that and we should be done with that stuff - {de,}mangle_poll() unified, no more separate kernel-side POLL... definitions.

Basically, that's what you suggested wrt KPOLL..., except that we can use EPOLL... instead - they already are arch-independent (and equal to what is currently kernel-side POLL...). After the preparations (in this series) switch to returning EPOLL... from ->poll() instances is completely mechanical and kernel-side POLL... can go away. The last step (killing kernel-side POLL... and unifying {de,}mangle_poll() has to be done after the search-and-replace job, since we need userland-side POLL... for unified {de,}mangle_poll(), thus the cherry-pick at the last step.

After that we get
* POLL{IN,OUT,...} *not* in __poll_t, so any stray instances of ->poll()
still using those will be caught by sparse.
* eventpoll.c and select.c warning-free wrt __poll_t
* no more kernel-side definitions of POLL... - userland ones are
visible through the entire kernel (and used pretty much only for mangle/demangle)
* same behavior as after the first series (i.e. sparc et.al. epoll(2)
working correctly).

The following changes since commit b2fe5fa68642860e7de76167c3111623aa0d5de1:

Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next (2018-01-31 14:31:10 -0800)

are available in the git repository at:

git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs.git work.poll2

for you to fetch changes up to d85e2aa2e34dac793e70b900d865f48c69ecbc27:

annotate ep_scan_ready_list() (2018-02-01 16:30:06 -0500)

----------------------------------------------------------------
Al Viro (7):
smc: missing poll annotations
xen: fix poll misannotation
use linux/poll.h instead of asm/poll.h
add EPOLLNVAL, annotate EPOLL... and event_poll->event preparation to switching ->poll() to returning EPOLL... ep_send_events_proc(): return result via esed->res annotate ep_scan_ready_list()

 drivers/xen/pvcalls-front.h    |  2 +-
 fs/coda/psdev.c                |  2 +-
 fs/debugfs/file.c              |  2 +-
 fs/eventpoll.c                 | 41 +++++++++++++++++++++++------------------
 fs/fcntl.c                     |  2 +-
 include/linux/poll.h           |  3 ++-
 include/uapi/linux/eventpoll.h | 33 +++++++++++++++++----------------
 net/smc/af_smc.c               |  2 +-
 8 files changed, 47 insertions(+), 40 deletions(-)
#1 lore raw
Quoted 2 lines · Al Viro

(ideally - just before -rc1)

Ok, merged and scripted and final patch applied. It's going through the allmodconfig test after that final patch, but I'll make -rc1 once that's done, and hopefully we are done with the whole EPOLL/POLL confusion.

.. and yes, I'm sure we'll be picking up straggling new uses of the POLL* macros for months or years.

Linus