kernel/linux.git/fs/pstore/pmsg.c, branch v6.1.168

pstore: Revert pmsg_lock back to a normal mutex

2023-05-11T14:03:27+00:00

[ Upstream commit 5239a89b06d6b199f133bf0ffea421683187f257 ] This reverts commit 76d62f24db07f22ccf9bc18ca793c27d4ebef721. So while priority inversion on the pmsg_lock is an occasional problem that an rt_mutex would help with, in uses where logging is writing to pmsg heavily from multiple threads, the pmsg_lock can be heavily contended. After this change landed, it was reported that cases where the mutex locking overhead was commonly adding on the order of 10s of usecs delay had suddenly jumped to ~msec delay with rtmutex. It seems the slight differences in the locks under this level of contention causes the normal mutexes to utilize the spinning optimizations, while the rtmutexes end up in the sleeping slowpath (which allows additional threads to pile on trying to take the lock). In this case, it devolves to a worse case senerio where the lock acquisition and scheduling overhead dominates, and each thread is waiting on the order of ~ms to do ~us of work. Obviously, having tons of threads all contending on a single lock for logging is non-optimal, so the proper fix is probably reworking pstore pmsg to have per-cpu buffers so we don't have contention. Additionally, Steven Rostedt has provided some furhter optimizations for rtmutexes that improves the rtmutex spinning path, but at least in my testing, I still see the test tripping into the sleeping path on rtmutexes while utilizing the spinning path with mutexes. But in the short term, lets revert the change to the rt_mutex and go back to normal mutexes to avoid a potentially major performance regression. And we can work on optimizations to both rtmutexes and finer-grained locking for pstore pmsg in the future. Cc: Wei Wang Cc: Midas Chien Cc: "Chunhui Li (李春辉)" Cc: Steven Rostedt Cc: Kees Cook Cc: Anton Vorontsov Cc: "Guilherme G. Piccoli" Cc: Tony Luck Cc: kernel-team@android.com Fixes: 76d62f24db07 ("pstore: Switch pmsg_lock to an rt_mutex to avoid priority inversion") Reported-by: "Chunhui Li (李春辉)" Signed-off-by: John Stultz Signed-off-by: Kees Cook Link: https://lore.kernel.org/r/20230308204043.2061631-1-jstultz@google.com Signed-off-by: Sasha Levin

pstore: Switch pmsg_lock to an rt_mutex to avoid priority inversion

2022-12-31T12:33:07+00:00

[ Upstream commit 76d62f24db07f22ccf9bc18ca793c27d4ebef721 ] Wei Wang reported seeing priority inversion caused latencies caused by contention on pmsg_lock, and suggested it be switched to a rt_mutex. I was initially hesitant this would help, as the tasks in that trace all seemed to be SCHED_NORMAL, so the benefit would be limited to only nice boosting. However, another similar issue was raised where the priority inversion was seen did involve a blocked RT task so it is clear this would be helpful in that case. Cc: Wei Wang Cc: Midas Chien Cc: Connor O'Brien Cc: Kees Cook Cc: Anton Vorontsov Cc: Colin Cross Cc: Tony Luck Cc: kernel-team@android.com Fixes: 9d5438f462ab ("pstore: Add pmsg - user-space accessible pstore object") Reported-by: Wei Wang Signed-off-by: John Stultz Signed-off-by: Kees Cook Link: https://lore.kernel.org/r/20221214231834.3711880-1-jstultz@google.com Signed-off-by: Sasha Levin

treewide: Replace GPLv2 boilerplate/reference with SPDX - rule 282

2019-06-05T15:36:37+00:00

Based on 1 normalized pattern(s): this software is licensed under the terms of the gnu general public license version 2 as published by the free software foundation and may be copied distributed and modified under those terms this program is distributed in the hope that it will be useful but without any warranty without even the implied warranty of merchantability or fitness for a particular purpose see the gnu general public license for more details extracted by the scancode license scanner the SPDX license identifier GPL-2.0-only has been chosen to replace the boilerplate/reference in 285 file(s). Signed-off-by: Thomas Gleixner Reviewed-by: Alexios Zavras Reviewed-by: Allison Randal Cc: linux-spdx@vger.kernel.org Link: https://lkml.kernel.org/r/20190529141900.642774971@linutronix.de Signed-off-by: Greg Kroah-Hartman

Remove 'type' argument from access_ok() function

2019-01-04T02:57:57+00:00

Nobody has actually used the type (VERIFY_READ vs VERIFY_WRITE) argument of the user address range verification function since we got rid of the old racy i386-only code to walk page tables by hand. It existed because the original 80386 would not honor the write protect bit when in kernel mode, so you had to do COW by hand before doing any user access. But we haven't supported that in a long time, and these days the 'type' argument is a purely historical artifact. A discussion about extending 'user_access_begin()' to do the range checking resulted this patch, because there is no way we're going to move the old VERIFY_xyz interface to that model. And it's best done at the end of the merge window when I've done most of my merges, so let's just get this done once and for all. This patch was mostly done with a sed-script, with manual fix-ups for the cases that weren't of the trivial 'access_ok(VERIFY_xyz' form. There were a couple of notable cases: - csky still had the old "verify_area()" name as an alias. - the iter_iov code had magical hardcoded knowledge of the actual values of VERIFY_{READ,WRITE} (not that they mattered, since nothing really used it) - microblaze used the type argument for a debug printout but other than those oddities this should be a total no-op patch. I tried to fix up all architectures, did fairly extensive grepping for access_ok() uses, and the changes are trivial, but I may have missed something. Any missed conversion should be trivially fixable, though. Signed-off-by: Linus Torvalds

pstore: Create common record initializer

2017-05-31T17:13:44+00:00

In preparation for setting timestamps in the pstore core, create a common initializer routine, instead of using static initializers. Signed-off-by: Kees Cook

pstore: Remove unused vmalloc.h in pmsg

2017-04-27T21:48:59+00:00

Since the vmalloc code has been removed from write_pmsg() in the commit "5bf6d1b pstore/pmsg: drop bounce buffer", remove the unused header vmalloc.h. Signed-off-by: Geliang Tang Signed-off-by: Kees Cook

pstore: Remove write_buf() callback

2017-03-07T22:01:02+00:00

Now that write() and write_buf() are functionally identical, this removes write_buf(), and renames write_buf_user() to write_user(). Additionally adds sanity-checks for pstore_info's declared functions and flags at registration time. Signed-off-by: Kees Cook

pstore: Replace arguments for write_buf_user() API

2017-03-07T22:01:01+00:00

Removes argument list in favor of pstore record, though the user buffer remains passed separately since it must carry the __user annotation. Signed-off-by: Kees Cook

pstore/pmsg: drop bounce buffer

2016-09-08T22:01:10+00:00

Removing a bounce buffer copy operation in the pmsg driver path is always better. We also gain in overall performance by not requesting a vmalloc on every write as this can cause precious RT tasks, such as user facing media operation, to stall while memory is being reclaimed. Added a write_buf_user to the pstore functions, a backup platform write_buf_user that uses the small buffer that is part of the instance, and implemented a ramoops write_buf_user that only supports PSTORE_TYPE_PMSG. Signed-off-by: Mark Salyzyn Signed-off-by: Kees Cook

pstore: add pstore unregister

2015-10-22T15:59:18+00:00

pstore doesn't support unregistering yet. It was marked as TODO. This patch adds some code to fix it: 1) Add functions to unregister kmsg/console/ftrace/pmsg. 2) Add a function to free compression buffer. 3) Unmap the memory and free it. 4) Add a function to unregister pstore filesystem. Signed-off-by: Geliang Tang Acked-by: Kees Cook [Removed __exit annotation from ramoops_remove(). Reported by Arnd Bergmann] Signed-off-by: Tony Luck