I won't apply this version, please send v2 with the changes below.

On Mon, 16 Nov 2026 14:45:43 +0400, Mirian Shilakadze wrote:
> Mirian Shilakadze (6):
>   ve: tell ve_setup_task() which VE the task comes from
>   ve: hide only the tasks the host sends into a VE
>   ve: hide the tasks that set up a starting VE from the rest of it
>   cgroup: give the CT root file writes and hidden roots only to setup
>     tasks
>   fs: give the mount relaxations of a pseudosuper VE only to setup tasks
>   ve: give the remaining pseudosuper relaxations only to setup tasks

The design holds: the setup mark reaches only what vzctl and CRIU run
before an exec, every way into another task's fds goes through
task_still_dumpable(), and no ve0 caller loses a relaxation. What to
change for v2:

1. Patch 2 (and 3), code: a host task that moves from a starting VE
   straight into a running one keeps VD_VE_SETUP_TASK and is visible
   there. Keep it hidden, as it was before the series. Do not key
   this on the mm mark alone: CRIU-restored processes keep
   VD_VE_SETUP_TASK for life.

2. Patch 2, commit message: say why a container task must not get
   VD_VE_ENTER_TASK, and that a task entering its nested VE is no
   longer hidden there.

3. Patch 3, comments: the VD_VE_SETUP_TASK comment in mm_types.h says
   setup tasks are the only ones with the pseudosuper relaxations,
   which patch 5 makes untrue; the ve_current_is_pseudosuper()
   kernel-doc leaves out that it is true for ve0.

4. Patch 4, comment: the comment above ve_pseudosuper_write() still
   says dropping pseudosuper is allowed from any context.

5. Patch 5: do what Sashiko asked (the subject says "only to setup
   tasks" while the mount limit also lets VD_VE_ENTER_TASK through;
   a helper next to ve_current_is_pseudosuper() instead of the
   open-coded host_entered), and update the "In case of
   @is_pseudosuper set" comment in ve_devmnt_process().

6. Patch 6, comment: devcgroup_legacy_check_permission() still says
   everything is allowed in pseudosuper state.

7. Cover letter: "vzctl and CRIU need no change, the vzctl start path
   was already fixed under VSTOR-151174" is two sentences glued with a
   comma, and the series also works with older libvzctl. Say what
   VSTOR-151174 changed or drop the clause. Put blank lines around the
   "- 1/6" list.

Details are in the replies to the patches.

-- 
Konstantin Khorenko,
Virtuozzo Linux Kernel Team

_______________________________________________
Devel mailing list
[email protected]
https://lists.openvz.org/mailman/listinfo/devel

Reply via email to