Ksawlii/kernel_samsung_a53x
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

proc/fd: In fdinfo seq_show don't use get_files_struct

Browse source 
[ Upstream commit 775e0656b27210ae668e33af00bece858f44576f ]

When discussing[1] exec and posix file locks it was realized that none
of the callers of get_files_struct fundamentally needed to call
get_files_struct, and that by switching them to helper functions
instead it will both simplify their code and remove unnecessary
increments of files_struct.count.  Those unnecessary increments can
result in exec unnecessarily unsharing files_struct which breaking
posix locks, and it can result in fget_light having to fallback to
fget reducing system performance.

Instead hold task_lock for the duration that task->files needs to be
stable in seq_show.  The task_lock was already taken in
get_files_struct, and so skipping get_files_struct performs less work
overall, and avoids the problems with the files_struct reference
count.

[1] https://lkml.kernel.org/r/20180915160423.GA31461@redhat.com
Suggested-by: Oleg Nesterov <oleg@redhat.com>
Acked-by: Christian Brauner <christian.brauner@ubuntu.com>
v1: https://lkml.kernel.org/r/20200817220425.9389-12-ebiederm@xmission.com
Link: https://lkml.kernel.org/r/20201120231441.29911-17-ebiederm@xmission.com
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
This commit is contained in:
Eric W. Biederman 2020-11-20 17:14:34 -06:00 • committed by Ksawlii
parent d817a14392
commit 111c37fdf7
1 changed files with 5 additions and 4 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

9
fs/proc/fd.c
View file

@ -29,9 +29,8 @@ static int seq_show(struct seq_file *m, void *v)
if (!task) if (!task)
return -ENOENT; return -ENOENT;
files = get_files_struct(task); task_lock(task);
put_task_struct(task); files = task->files;
if (files) { if (files) {
unsigned int fd = proc_fd(m->private); unsigned int fd = proc_fd(m->private);
@ -48,8 +47,9 @@ static int seq_show(struct seq_file *m, void *v)
ret = 0; ret = 0;
} }
spin_unlock(&files->file_lock); spin_unlock(&files->file_lock);
put_files_struct(files);
} }
task_unlock(task);
put_task_struct(task);
if (ret) if (ret)
return ret; return ret;
@ -59,6 +59,7 @@ static int seq_show(struct seq_file *m, void *v)
real_mount(file->f_path.mnt)->mnt_id, real_mount(file->f_path.mnt)->mnt_id,
file_inode(file)->i_ino); file_inode(file)->i_ino);
/* show_fd_locks() never deferences files so a stale value is safe */
show_fd_locks(m, file, files); show_fd_locks(m, file, files);
if (seq_has_overflowed(m)) if (seq_has_overflowed(m))
goto out; goto out;