fs/sync.c : Add CAP_SYS_ADMIN checking before sync
PATCH 1/1 fs/sync.c : Add CAP_SYS_ADMIN checking before sync
The process, supposed in one container, can't flush the metadata and data of the all host's partitions without CAP_SYS_ADMIN capability, like sys_mount is doing. The checking will prevent some vicious programs impacting IO sequnces of those partitions, particularly, the ones which can't be accessed in the container.
---fs/sync.c | 3 +++
1 file changed, 3 insertions(+)
fs/sync.c +0 -0
@@ -103,6 +103,9 @@ SYSCALL_DEFINE0(sync)
{
int nowait = 0, wait = 1;
+ if (!capable(CAP_SYS_ADMIN))
+ return -EPERM;
+wakeup_flusher_threads(0, WB_REASON_SYNC);iterate_supers(sync_inodes_one_sb, NULL);iterate_supers(sync_fs_one_sb, &nowait);
--
1.9.1
Quoted 16 lines · Wuqixuan
The process, supposed in one container, can't flush the metadata and data of the all host's partitions without
CAP_SYS_ADMINcapability, likesys_mountis doing. The checking will prevent some vicious programs impacting IO sequnces of those partitions, particularly, the ones which can't be accessed in the container.Signed-off-by Last Wu <[email protected]>---
fs/sync.c| 3 +++1 file changed, 3 insertions(+)fs/sync.c +0
-0
fs/sync.c SYSCALL_DEFINE0(sync) · line 107
Quoted 5 lines
fs/sync.c +2
-0@@ -103,6 +103,9 @@ SYSCALL_DEFINE0(sync) {int nowait =0 , wait =1 ; +if (!capable (CAP_SYS_ADMIN )) +return -EPERM ;
So after this patch I can't call sync as a regular user? (even without containers).
But nothing in sync(2) says about special permissions for this.
Quoted 20 lines · Azat Khuzhin
Quoted 19 lines · Wuqixuan
The process, supposed in one container, can't flush the metadata and data of the all host's partitions without
CAP_SYS_ADMINcapability, likesys_mountis doing. The checking will prevent some vicious programs impacting IO sequnces of those partitions, particularly, the ones which can't be accessed in the container.Signed-off-by Last Wu <[email protected]>---
fs/sync.c| 3 +++1 file changed, 3 insertions(+)fs/sync.c +0
-0@@ -103,6 +103,9 @@ SYSCALL_DEFINE0(sync) {int nowait =0 , wait =1 ;
Quoted 6 lines
fs/sync.c SYSCALL_DEFINE0(sync) · line 107
Quoted 2 lines
fs/sync.c +2
-0+if (!capable (CAP_SYS_ADMIN )) +return -EPERM ;So after this patch I can't call sync as a regular user? (even without containers).
But nothing in sync(2) says about special permissions for this.
Yeah, this solution will break userspace.
A much more generic solution would be to wait for cgroup aware writeback[1]. As temporary hack you can check whether the calling process is in the initial pid namesapce to detect
a container calling sync(2)...
[1] https://lwn.net/Articles/628631/
Signature · 2 lines
Thanks, //richard
Thanks for your both replies. We will waiting for the patchset, and modify as per Richard's suggestion for temporary hack.
Quoted 34 lines · Richard Weinberger
Quoted 20 lines · Azat Khuzhin
Quoted 19 lines · Wuqixuan
The process, supposed in one container, can't flush the metadata and data of the all host's partitions without
CAP_SYS_ADMINcapability, likesys_mountis doing. The checking will prevent some vicious programs impacting IO sequnces of those partitions, particularly, the ones which can't be accessed in the container.Signed-off-by Last Wu <[email protected]>---
fs/sync.c| 3 +++1 file changed, 3 insertions(+)fs/sync.c +0
-0@@ -103,6 +103,9 @@ SYSCALL_DEFINE0(sync) {int nowait =0 , wait =1 ;Quoted 5 lines
fs/sync.c SYSCALL_DEFINE0(sync) · line 107
Quoted 2 lines
fs/sync.c +2
-0+if (!capable (CAP_SYS_ADMIN )) +return -EPERM ;So after this patch I can't call sync as a regular user? (even without containers).
But nothing in sync(2) says about special permissions for this.Yeah, this solution will break userspace.
A much more generic solution would be to wait for cgroup aware writeback[1]. As temporary hack you can check whether the calling process is in the initial pid namesapce to detect
a container calling sync(2)...