[<prev] [next>] [thread-next>] [day] [month] [year] [list]
Message-Id: <20240424033409.2735257-1-libaokun@huaweicloud.com>
Date: Wed, 24 Apr 2024 11:34:04 +0800
From: libaokun@...weicloud.com
To: netfs@...ts.linux.dev
Cc: dhowells@...hat.com,
jlayton@...nel.org,
zhujia.zj@...edance.com,
jefflexu@...ux.alibaba.com,
linux-cachefs@...hat.com,
linux-erofs@...ts.ozlabs.org,
linux-fsdevel@...r.kernel.org,
linux-kernel@...r.kernel.org,
libaokun@...weicloud.com,
Baokun Li <libaokun1@...wei.com>
Subject: [PATCH 0/5] cachefiles: some bugfixes for clean object/send req/poll
From: Baokun Li <libaokun1@...wei.com>
Hello everyone!
Recently we found some bugs while doing tests on cachefiles ondemand mode,
and this patchset is a fix for some of those issues. The following is a
brief overview of the patches, see the patches for more details.
Patch 1-3: After an object has been cleaned up, make sure it has no
outstanding requests and that the corresponding ondemand_object_worker
has exited, otherwise it may use-after-free.
Patch 4: Cyclic allocation of msg_id to avoid msg_id reuse misleading
the daemon to cause hung.
Patch 5: Hold xas_lock during polling to avoid dereferencing reqs causing
use-after-free.
Comments and questions are, as always, welcome.
Thanks,
Baokun
Baokun Li (3):
cachefiles: stop sending new request when dropping object
cachefiles: flush all requests for the object that is being dropped
cachefiles: cyclic allocation of msg_id to avoid reuse
Hou Tao (1):
cachefiles: flush ondemand_object_worker during clean object
Jingbo Xu (1):
cachefiles: add missing lock protection when polling
fs/cachefiles/daemon.c | 4 +-
fs/cachefiles/internal.h | 3 +
fs/cachefiles/ondemand.c | 120 ++++++++++++++++++++++++++-------------
3 files changed, 86 insertions(+), 41 deletions(-)
--
2.39.2
Powered by blists - more mailing lists