[PATCH v4 0/9] /proc/kcore improvements

From: Omar Sandoval
Date: Wed Jul 25 2018 - 19:59:31 EST


From: Omar Sandoval <osandov@xxxxxx>

Hi,

This series makes a few improvements to /proc/kcore. It fixes a couple
of small issues in v3 but is otherwise the same. Patches 1, 2, and 3 are
prep patches. Patch 4 is a fix/cleanup. Patch 5 is another prep patch.
Patches 6 and 7 are optimizations to ->read(). Patch 8 makes it possible
to enable CRASH_CORE on any architecture, which is needed for patch 9.
Patch 9 adds vmcoreinfo to /proc/kcore.

Based on v4.18-rc6 + James' patch in the mm tree, and tested with crash
and readelf.

Thanks!

Changes from v3:

- Fixes a mixed up up_write() instead of up_read() in patch 5 reported
by Tetsuo Handa
- Added patch 8 to fix a build failure reported by Stephen Rothwell

Changes from v2:

- Add __init to kclist_add() as per Andrew
- Got rid of conversion of kcore_need_update from int to atomic_t and
just used xchg() instead of atomic_cmpxchg() (split out into a new
patch instead of combining it with the rwlock -> rwsem conversion)
- Add comment about the increase in file size to patch 8

Changes from v1:

- Rebased onto v4.18-rc4 + James' patch
(https://patchwork.kernel.org/patch/10519739/) in the mm tree
- Fix spurious sparse warning (see the report and response in
https://patchwork.kernel.org/patch/10512431/)

Omar Sandoval (9):
proc/kcore: don't grab lock for kclist_add()
proc/kcore: don't grab lock for memory hotplug notifier
proc/kcore: replace kclist_lock rwlock with rwsem
proc/kcore: fix memory hotplug vs multiple opens race
proc/kcore: hold lock during read
proc/kcore: clean up ELF header generation
proc/kcore: optimize multiple page reads
crash_core: use VMCOREINFO_SYMBOL_ARRAY() for swapper_pg_dir
proc/kcore: add vmcoreinfo note to /proc/kcore

fs/proc/Kconfig | 1 +
fs/proc/kcore.c | 534 +++++++++++++++++--------------------
include/linux/crash_core.h | 2 +
include/linux/kcore.h | 2 +-
kernel/crash_core.c | 6 +-
5 files changed, 252 insertions(+), 293 deletions(-)

--
2.18.0