Dave Hansen [EMAIL PROTECTED] wrote:
|
| The following patch completes the removal of proc_mnt. It fetches
| the mnt on which to do dentry invalidations from the pid_namespace
| in which the task appears.
|
| For now, there is only one pid namespace in mainline so this is
| straightforward. In
Here are a few pidspace patches I have been working on
and would like to seek comments on before sending them
to LKML.
Some of these were included in 2.6.20-rcX-mm1 but were
dropped due to some collisions. Andrew wanted us to
retest and send them out.
Some were already reviewed earlier once
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Subject: [PATCH 2/2] Explicitly set pgid and sid of init
Explicitly set pgid and sid of init process to 1.
Signed-off-by: Sukadev Bhattiprolu [EMAIL PROTECTED]
Cc: Cedric Le Goater [EMAIL PROTECTED]
Cc: Dave Hansen [EMAIL PROTECTED]
Cc: Serge Hallyn
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Subject: [PATCH] attach_pid() with struct pid parameter
attach_pid() currently takes a pid_t and then uses find_pid() to find
the corresponding struct pid. Sometimes we already have the struct pid.
We can then skip find_pid() if attach_pid() were
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Subject: [PATCH] Use struct pid parameter in copy_process()
Modify copy_process() to take a struct pid * parameter instead of a pid_t.
This simplifies the code a bit and also avoids having to call find_pid()
to convert the pid_t to a struct pid.
Signed
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Subject: [PATCH] Use task_pgrp() and task_session() in setpgid/setsid
Use the container-friendly interfaces task_pgrp() and task_session()
in getpgid() and setpgid().
Signed-off-by: Sukadev Bhattiprolu [EMAIL PROTECTED]
Cc: Cedric Le Goater [EMAIL
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Subject: [PATCH] pidspace rocket driver
The process_session() and process_group() values are not really
used by the driver.
Signed-off-by: Sukadev Bhattiprolu [EMAIL PROTECTED]
Cc: Cedric Le Goater [EMAIL PROTECTED]
Cc: Dave Hansen [EMAIL PROTECTED
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Subject: [PATCH] Use task_pgrp() and task_session in binfmt
Use container friendly interfaces task_pgrp() and task_session() in
binfmt files
Signed-off-by: Sukadev Bhattiprolu [EMAIL PROTECTED]
Cc: Cedric Le Goater [EMAIL PROTECTED]
Cc: Dave Hansen
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Subject: [PATCH 2/4] Replace pid_t in autofs with struct pid reference.
Make autofs container-friendly by caching struct pid reference rather
than pid_t and using pid_nr() to retreive a task's pid_t.
ChangeLog:
- Fix Eric Biederman's comments
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Subject: [PATCH 3/4] Replace pid_t in autofs4 with struct pid reference.
Make autofs4 container-friendly by caching struct pid reference rather
than pid_t and using pid_nr() to retreive a task's pid_t.
ChangeLog:
- Fix Eric Biederman's
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Subject: [PATCH 4/4] Use task_pgrp() in autofs/autofs4
Replace process_group(tsk) with pid_nr(task_pgrp(tsk)) in autofs and autofs4.
Signed-off-by: Sukadev Bhattiprolu [EMAIL PROTECTED]
Cc: Cedric Le Goater [EMAIL PROTECTED]
Cc: Dave Hansen [EMAIL
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Subject: [PATCH] pidspace coda fs
Use container-safe interfaces for pid and pgid in coda.
Signed-off-by: Sukadev Bhattiprolu [EMAIL PROTECTED]
Cc: Cedric Le Goater [EMAIL PROTECTED]
Cc: Dave Hansen [EMAIL PROTECTED]
Cc: Serge Hallyn [EMAIL PROTECTED
Eric W. Biederman [EMAIL PROTECTED] wrote:
| [EMAIL PROTECTED] writes:
|
| From: Sukadev Bhattiprolu [EMAIL PROTECTED]
| Subject: [PATCH] Use struct pid parameter in copy_process()
|
| Modify copy_process() to take a struct pid * parameter instead of a pid_t.
| This simplifies the code a bit
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Subject: [PATCH 3/5] Use struct pid parameter in copy_process()
Modify copy_process() to take a struct pid * parameter instead of a pid_t.
This simplifies the code a bit and also avoids having to call find_pid()
to convert the pid_t to a struct pid
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Subject: [PATCH 2/5] Explicitly set pgid and sid of init process
Explicitly set pgid and sid of init process to 1.
Signed-off-by: Sukadev Bhattiprolu [EMAIL PROTECTED]
Cc: Cedric Le Goater [EMAIL PROTECTED]
Cc: Dave Hansen [EMAIL PROTECTED]
Cc: Serge
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Subject: [PATCH] pidspace rocket driver
The process_session() and process_group() values are not really
used by the driver.
Signed-off-by: Sukadev Bhattiprolu [EMAIL PROTECTED]
Cc: Cedric Le Goater [EMAIL PROTECTED]
Cc: Dave Hansen [EMAIL PROTECTED
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Subject: [PATCH 2/3] Replace pid_t in autofs with struct pid reference.
Make autofs container-friendly by caching struct pid reference rather
than pid_t and using pid_nr() to retreive a task's pid_t.
ChangeLog:
- Fix Eric Biederman's comments
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Subject: [PATCH 3/3] Replace pid_t in autofs4 with struct pid reference.
Make autofs4 container-friendly by caching struct pid reference rather
than pid_t and using pid_nr() to retreive a task's pid_t.
ChangeLog:
- Fix Eric Biederman's
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Subject [RFC][PATCH 2/6] Reimplement pid_nr() using struct pid_nr
Replace the pid_nr() stub with a version that uses struct pid_nr
and returns the pid_t based on the pid namespace of the caller.
Signed-off-by: Sukadev Bhattiprolu [EMAIL PROTECTED]
Cc
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Subject: [RFC][PATCH] Define/use task_pid_ns() wrapper
Use a wrapper to access/set pid_namespace of a task.
Signed-off-by: Sukadev Bhattiprolu [EMAIL PROTECTED]
Cc: Cedric Le Goater [EMAIL PROTECTED]
Cc: Dave Hansen [EMAIL PROTECTED]
Cc: Serge Hallyn
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Subject [RFC][PATCH 0/5] Remove nsproxy-pid_ns
We need to decouple pid namespace from nsproxy to allow getting and
releasing pid namespace independently from other namespaces. This is
required since a process's reference to its pid namespace must exist
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Subject: [RFC][PATCH 3/5] Use pid namespace from struct pid_nrs list
Stop using task-nsproxy-pid_ns. Use pid_namespace from pid-pid_nrs
list instead.
To simplify error handling, this patch moves processing of CLONE_NEWPID
flag, currently
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Subject: [RFC][PATCH]
Remove copy_pid_ns() as its no longer needed.
Signed-off-by: Sukadev Bhattiprolu [EMAIL PROTECTED]
Cc: Cedric Le Goater [EMAIL PROTECTED]
Cc: Dave Hansen [EMAIL PROTECTED]
Cc: Serge Hallyn [EMAIL PROTECTED]
Cc: [EMAIL PROTECTED
This is actually PATCH 5/5.
[EMAIL PROTECTED] [EMAIL PROTECTED] wrote:
|
| From: Sukadev Bhattiprolu [EMAIL PROTECTED]
| Subject: [RFC][PATCH 5/5] Remove copy_pid_ns()
|
| Remove copy_pid_ns() as its no longer needed.
|
| Signed-off-by: Sukadev Bhattiprolu [EMAIL PROTECTED]
| Cc: Cedric Le
struct pid_nrs per
| process, one for each namespace.
|
| Signed-off-by: Cedric Le Goater [EMAIL PROTECTED]
| Signed-off-by: Sukadev Bhattiprolu [EMAIL PROTECTED]
|
| ---
| include/linux/sched.h |1 +
| kernel/fork.c | 22 ++
| kernel/nsproxy.c |5
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Subject: [PATCH 4/5] Remove the likely(pid) check in copy_process
Now that we pass in a struct pid parameter to copy_process()
and even the swapper (pid_t == 0) has a valid struct pid,
we no longer need this check.
Changelog:
Per Eric
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Subject: [PATCH] Kill unused sesssion and group values in rocket driver
The process_session() and process_group() values are not really
used by the driver.
Signed-off-by: Sukadev Bhattiprolu [EMAIL PROTECTED]
Cc: Cedric Le Goater [EMAIL PROTECTED]
Cc
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Subject: [PATCH 1/2] Fix some coding-style errors in autofs
Fix coding style errors (extra spaces, long lines) in autofs
and autofs4 files being modified for container/pidspace issues.
Signed-off-by: Sukadev Bhattiprolu [EMAIL PROTECTED]
Cc: Cedric Le
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Subject: [PATCH 2/2] Replace pid_t in autofs with struct pid reference.
Make autofs container-friendly by caching struct pid reference rather
than pid_t and using pid_nr() to retreive a task's pid_t.
ChangeLog:
- Fix Eric Biederman's comments
of this.
---
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Subject: [PATCH 1/2] Fix some coding-style errors in autofs
Fix coding style errors (extra spaces, long lines) in autofs
and autofs4 files being modified for container/pidspace issues.
Signed-off-by: Sukadev Bhattiprolu [EMAIL PROTECTED]
Cc
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Subject: [RFC][PATCH 02/14] Move alloc_pid call to copy_process
Move alloc_pid() into copy_process(). This will help in
code to support cloning of pid namespace.
Signed-off-by: Sukadev Bhattiprolu [EMAIL PROTECTED]
---
kernel/fork.c | 31
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Subject: [RFC][PATCH 03/14] use pid_nr in procfs
With containers, a process can have different pid_t values in different
pid namespaces. To ensure we get the correct pid_t value in any context,
we should use pid_nr() function rather than directly
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Subject: [RFC][PATCH 05/14] Initialize struct pid_nr for swapper.
Statically initialize a struct pid_nr for the swapper process.
Signed-off-by: Sukadev Bhattiprolu [EMAIL PROTECTED]
---
include/linux/init_task.h | 10 ++
kernel/pid.c
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Subject: [RFC][PATCH 04/14] Add struct pid_nr and pid_nrs list
A struct pid_nr associates a pid_t value with a pid namespace. We attach
a list of struct pid_nr entries to a struct pid (pid-pid_nrs), allowing
the struct pid to take different pid_t
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Subject: [RFC][PATCH 07/14] Remove unused pid_chain from struct pid
Now that we use struct pid_nr entries in the pid_hash table, we no
longer need pid_chain field in struct pid.
Signed-off-by: Sukadev Bhattiprolu [EMAIL PROTECTED]
---
include/linux
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Subject: [RFC][PATCH 10/14] Use pid ns from pid_nrs list
We need to decouple pid namespace from nsproxy to allow getting and
releasing pid namespace independently from other namespaces. This is
required since a process's reference to its pid namespace
This was discussed on containers and we thought it would be useful
to reserve this flag.
---
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Subject: [RFC][PATCH 13/14] Define CLONE_NEWPID flag
Define CLONE_NEWPID flag that will be used to clone pid namespaces.
Signed-off-by: Sukadev Bhattiprolu
PROTECTED]
Signed-off-by: Sukadev Bhattiprolu [EMAIL PROTECTED]
---
include/linux/pid.h |9 --
kernel/pid.c| 73 ++--
2 files changed, 50 insertions(+), 32 deletions(-)
Index: 2.6.21-rc3-mm2/kernel/pid.c
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Subject: [RFC][PATCH 12/14] Remove copy_pid_ns function
Remove the copy_pid_ns() function as we have decoupled pid namespace
from nsproxy and also because we currently disallow unsharing of pid
namespace.
Signed-off-by: Sukadev Bhattiprolu [EMAIL
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Subject: [RFC][PATCH 09/14] Save leaders struct pid before detach_pid()
Save the struct pid of the thread group leader before detaching
our pid. See comments in the code below for more details.
Signed-off-by: Sukadev Bhattiprolu [EMAIL PROTECTED
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Subject: [RFC][PATCH 14/14] Enable cloning pid namespace
When clone() is invoked with CLONE_NEWPID, create a new pid namespace
and then create a new struct pid for the new process. Allocate pid_ts
for the new process in the new pid namespace and all
This was discussed on containers and we thought it would be useful
to reserve this flag.
---
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Subject: [PATCH] Define CLONE_NEWPID flag
Define CLONE_NEWPID flag that will be used to clone pid namespaces.
Signed-off-by: Sukadev Bhattiprolu [EMAIL
Kirill,
We are working on getting a few UTS namespace tests included in Linux
Testsuite Project (LTP) and were wondering if you (or anyone else) had
any test cases for IPC namespace.
If there are tests for IPC NS, can you share those tests with us ?
And, do you have any plans to push the tests
Dave, Serge,
Here is my current pid namespace patchset. There are still a couple of
issues I am investigating, but appreciate any feedback.
Suka
---
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Subject: [RFC][PATCH 0/16] Enable clone pid namespace
A pid namespace is a view of a particular set
Subject: Define/use task_active_pid_ns() wrapper
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
A process is visible in one or more pid namespaces, as it is known by
some pid_t in every ancestor pid namespace. Every time the process
forks, the child process also gets a pid_t in every ancestor
Subject: Rename pid_nr function
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Rename pid_nr() function to pid_to_nr() which is more descriptive
and will hopefully cause less confusion with new structure/functions
being added to support multiple pid namespaces.
Signed-off-by: Sukadev Bhattiprolu
Subject: Rename child_reaper function.
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Rename the child_reaper() function to task_child_reaper() to be
inline with other task_* functions and to distinguish the function
from struct pid_namspace.child_reaper.
Signed-off-by: Sukadev Bhattiprolu
Subject: Use pid_to_nr() in process info functions
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Use pid_to_nr() function in getppid(), getpgid() and getsid() functions
so they return the correct pid_t for processes in multiple pid namespaces.
Note: We don't need pid_to_nr() in getpid() because
Subject: Use task_pid() to find leader's pid
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Use task_pid() to get leader's pid since find_pid() cannot be used
after detach_pid(). See comments in the code below for more details.
Signed-off-by: Sukadev Bhattiprolu [EMAIL PROTECTED]
---
fs/exec.c
namespace
is the init_pid_ns.
Changelog:
2.6.21-mm2-pidns2:
- [Sukadev Bhattiprolu] Changed is_container_init() calls in {powerpc,
ppc,avr32}/traps.c for the _exception() call to is_global_init().
This way, we kill only the container if the container's init has
Subject: Move alloc_pid call to copy_process
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Move alloc_pid() into copy_process(). This will keep all pid and pid
namespace code together and simplify error handling when we support
multiple pid namespaces.
Changelog:
- [Eric Biederman] Move
Subject: Define/use pid-upid_list list.
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
With multiple pid namespaces, a process would be known by several pid_t
values, one in each pid namespace. To represent this, we introduce a
'struct upid' which associates a single pid_t value with a single pid
Subject: Use pid ns from pid-upid_list
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
We need to decouple pid namespace from nsproxy to allow getting and releasing
pid namespace independently from other namespaces. This is required since a
process's reference to its pid namespace must exist even
Subject: Define CLONE_NEWPID flag
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
This was discussed on containers and we thought it would be useful
to reserve this flag.
TODO: Merge this with the next patch ?
Define CLONE_NEWPID flag that will be used to clone pid namespaces.
Signed-off
Subject: Enable cloning pid namespace
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
When clone() is invoked with CLONE_NEWPID, create a new pid namespace
and then create a new struct pid for the new process. Allocate pid_t's
for the new process in the new pid namespace and all ancestor pid
Subject: Terminate processes in a ns when reaper is exiting.
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
This should actually be considered a part of the previous patch which
enables cloning of pid namespace. Its been separated out for easier
review.
Terminate all processes in a namespace when
Subject: Remove proc_mnt's use for killing inodes
From: Dave Hansen [EMAIL PROTECTED]
We use proc_mnt as a shortcut to find a superblock on which
to go killing /proc inodes. This will break if we ever have
more than one /proc superblock. So, use the superblock list
to go find each of the
-off-by: Sukadev Bhattiprolu [EMAIL PROTECTED]
---
fs/proc/base.c| 32 -
fs/proc/inode.c | 11 +++-
fs/proc/root.c| 52 --
include/linux/pid_namespace.h |1
include/linux
Subject: Enable signaling child reaper from parent ns.
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
The reaper of a child namespace must receive signals from its parent pid
namespace but not receive any signals from its own namespace.
This is a very early draft :-) and following tests seem
Pavel Emelianov [EMAIL PROTECTED] wrote:
| Index: lx26-21-mm2/kernel/pid.c
| ===
| --- lx26-21-mm2.orig/kernel/pid.c 2007-05-22 16:59:34.0 -0700
| +++ lx26-21-mm2/kernel/pid.c2007-05-22 16:59:46.0 -0700
|
Pavel Emelianov [EMAIL PROTECTED] wrote:
| Index: lx26-21-mm2/kernel/pid.c
| ===
| --- lx26-21-mm2.orig/kernel/pid.c 2007-05-22 16:59:46.0 -0700
| +++ lx26-21-mm2/kernel/pid.c2007-05-22 17:06:48.0 -0700
|
Pavel Emelianov [EMAIL PROTECTED] wrote:
| +/*
| + * Return the pid_t by which the process @pid is known in the pid
| + * namespace @ns.
| + *
| + * Return 0 if:
| + * - @pid is NULL (eg: procfs calls this for task_pgrp(init_task)
| + * which is NULL).
| + *
| + *
Pavel Emelianov [EMAIL PROTECTED] wrote:
| [EMAIL PROTECTED] wrote:
| Pavel Emelianov [EMAIL PROTECTED] wrote:
| | Index: lx26-21-mm2/kernel/pid.c
| | ===
| | --- lx26-21-mm2.orig/kernel/pid.c 2007-05-22
Dave Hansen [EMAIL PROTECTED] wrote:
| On Thu, 2007-05-24 at 13:24 +0400, Pavel Emelianov wrote:
| | +int is_global_init(struct task_struct *tsk)
| | +{
| | + return (task_active_pid_ns(tsk) == init_pid_ns tsk-pid == 1);
| |
| | This can OOPS if you pass arbitrary task to this
Dave Hansen [EMAIL PROTECTED] wrote:
| On Fri, 2007-05-25 at 13:44 -0700, [EMAIL PROTECTED] wrote:
| Dave Hansen [EMAIL PROTECTED] wrote:
| | On Thu, 2007-05-24 at 13:24 +0400, Pavel Emelianov wrote:
| | | +int is_global_init(struct task_struct *tsk)
| | | +{
| | | + return
Dave Hansen [EMAIL PROTECTED] wrote:
| On Thu, 2007-05-31 at 15:48 +0400, Pavel Emelianov wrote:
| +void proc_flush_task(struct task_struct *task)
| +{
| + int i;
| + struct pid *pid;
| + struct upid* upid;
| +
| + pid = task_pid(task);
| + if (!pid)
| +
Like we discussed earlier and Pavel/others had pointed out,
proc_flush_task() in its current place in release_task() is
useless with the new pid namespace code, because task_pid()
for the task is already NULL before the call to proc_flush_task().
So as a simple change I tried to move
Date: Sat, 16 Jun 2007 15:29:14 -0700
From: [EMAIL PROTECTED]
Subject: [PATCH 00/17] Pid-NS(V3) Enable multiple pid namespaces
--
See Changelog below for changes since Pid-NS(V2) (2.6.21-mm2-pidns3).
---
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
A pid namespace is a view of a particular
Date: Sat, 16 Jun 2007 15:29:15 -0700
From: [EMAIL PROTECTED]
Cc: Sukadev Bhattiprolu [EMAIL PROTECTED]
Subject: [PATCH 01/17] Pid-NS(V3) Define and use task_active_pid_ns() wrapper
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
A process is visible in one or more pid namespaces, as it is known
Date: Sat, 16 Jun 2007 15:29:16 -0700
From: [EMAIL PROTECTED]
Cc: Sukadev Bhattiprolu [EMAIL PROTECTED]
Subject: [PATCH 02/17] Pid-NS(V3) Rename pid_nr function
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Rename pid_nr() function to pid_to_nr() which is more descriptive
and will hopefully cause
Subject: [PATCH 05/17] Pid-NS(V3) Use task_pid() to find leaders pid
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Use task_pid() to get leader's pid since find_pid() cannot be used
after detach_pid(). See comments in the code below for more details.
Signed-off-by: Sukadev Bhattiprolu [EMAIL
-pidns1:
- Use 'init_pid_ns.child_reaper' to determine if a given task is the
global init (/sbin/init) process. This would improve performance
and remove dependence on the task_pid().
2.6.21-mm2-pidns2:
- [Sukadev Bhattiprolu] Changed is_container_init() calls
Subject: [PATCH 07/17] Pid-NS(V3) Move alloc_pid call to copy_process
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Move alloc_pid() into copy_process(). This will keep all pid and pid
namespace code together and simplify error handling when we support
multiple pid namespaces.
Changelog
Subject: [PATCH 10/17] Pid-NS(V3) Define CLONE_NEWPID flag
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
This was discussed on containers and we thought it would be useful
to reserve this flag.
TODO: Merge this with the next patch ?
Define CLONE_NEWPID flag that will be used to clone pid
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Subject: [PATCH 14/17] Pid-NS(V3) [PATCH] Allow task_active_pid_ns() to be NULL
In earlier patches we get the pid_t or active pid namespace of a task/process
from its 'task_pid()'. But when a process is exiting, the task_pid() can be
NULL and we
(), the
signal originated from an ancestor namespace and so the container-
init honors the signal.
Signed-off-by: Sukadev Bhattiprolu [EMAIL PROTECTED]
---
include/linux/pid.h|1
include/linux/sched.h | 11 +
include/linux/signal.h |1
kernel/pid.c | 22
/proc pid entries. Those are a different beast.
Changelog:
2.6.22-rc4-mm2-pidns1:
- [Sukadev Bhattiprolu] Initialize proc_fs_type.fs_supers.
Signed-off-by: Dave Hansen [EMAIL PROTECTED]
Signed-off-by: Sukadev Bhattiprolu [EMAIL PROTECTED]
---
fs/proc/generic.c | 28
Pavel Emelianov [EMAIL PROTECTED] wrote:
| [EMAIL PROTECTED] wrote:
| Subject: [PATCH 08/17] Pid-NS(V3) Define/use pid-upid_list list.
|
| From: Sukadev Bhattiprolu [EMAIL PROTECTED]
|
|
| With multiple pid namespaces, a process would be known by several pid_t
| values, one in each pid
Pavel Emelianov [EMAIL PROTECTED] wrote:
| [EMAIL PROTECTED] wrote:
| Subject: [PATCH 12/17] Pid-NS(V3) Terminate processes in a ns when reaper
is exiting.
|
| From: Sukadev Bhattiprolu [EMAIL PROTECTED]
|
| This should actually be considered a part of the previous patch which
| enables
Pavel Emelianov [EMAIL PROTECTED] wrote:
| This patch opens the multilevel model patches.
|
| The multilevel model idea is basically the same as for the flat one,
| but in this case task may have many virtual pids - one id for each
| sub-namespace this task is visible in. The struct pid carries
Pavel Emelianov [EMAIL PROTECTED] wrote:
| [EMAIL PROTECTED] wrote:
| Pavel Emelianov [EMAIL PROTECTED] wrote:
| | [EMAIL PROTECTED] wrote:
| | Subject: [PATCH 08/17] Pid-NS(V3) Define/use pid-upid_list list.
| |
| | From: Sukadev Bhattiprolu [EMAIL PROTECTED
Pavel Emelianov [EMAIL PROTECTED] wrote:
| [EMAIL PROTECTED] wrote:
| Pavel Emelianov [EMAIL PROTECTED] wrote:
| | This patch opens the flat model patches.
| |
| | The flat model idea is that struct pid has two numbers. The first one
| | (pid-nr) is a global one and is unique in the system.
namespace when mounting/remounting /proc.
|Put this reference when unmounting.
|
| Signed-off-by: Dave Hansen [EMAIL PROTECTED]
| Signed-off-by: Sukadev Bhattiprolu [EMAIL PROTECTED]
| ---
|
| fs/proc/base.c| 30 ++-
| fs/proc/inode.c | 12
Pavel Emelianov [EMAIL PROTECTED] wrote:
| [EMAIL PROTECTED] wrote:
| Pavel Emelianov [EMAIL PROTECTED] wrote:
| | The most important one is moving exit_task_namespaces behind exit_notify
| | in do_exit() to make it possible to see the task's pid namespace to
| | properly notify the parent.
|
Subject: [PATCH 1/6] Define and use task_active_pid_ns() wrapper
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
With multiple pid namespaces, a process is known by some pid_t in
every ancestor pid namespace. Every time the process forks, the
child process also gets a pid_t in every ancestor pid
Subject: [PATCH 2/6] Rename pid_nr function
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Rename pid_nr() function to pid_to_nr() which is more descriptive
and will hopefully cause less confusion with new structure/functions
being added to support multiple pid namespaces.
Signed-off-by: Sukadev
Subject: [PATCH 3/6] Rename child_reaper function.
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Rename the child_reaper() function to task_child_reaper() to be
similar to other task_* functions and to distinguish the function
from 'struct pid_namspace.child_reaper'.
Signed-off-by: Sukadev
Subject: [PATCH 4/6] Use pid_to_nr() in process info functions
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Use pid_to_nr() function in getppid(), getpgid() and getsid() functions
so they return the correct pid_t for processes in multiple pid namespaces.
Note: We don't need pid_to_nr
Subject: [PATCH 5/6] Use task_pid() to find leader's pid
From: Sukadev Bhattiprolu [EMAIL PROTECTED]
Use task_pid() to get leader's pid since find_pid() cannot be used
after detach_pid(). See comments in the code below for more details.
Signed-off-by: Sukadev Bhattiprolu [EMAIL PROTECTED
:
- Use 'init_pid_ns.child_reaper' to determine if a given task is the
global init (/sbin/init) process. This would improve performance
and remove dependence on the task_pid().
2.6.21-mm2-pidns2:
- [Sukadev Bhattiprolu] Changed is_container_init() calls in {powerpc
Pavel Emelianov [EMAIL PROTECTED] wrote:
| This is submition for inclusion of hierarchical, not kconfig
| configurable, zero overheaded ;) pid namespaces.
|
| The overall idea is the following:
|
| The namespace are organized as a tree - once a task is cloned
| with CLONE_NEWPIDS (yes, I've
Pavel Emelianov [EMAIL PROTECTED] wrote:
| This is submition for inclusion of hierarchical, not kconfig
| configurable, zero overheaded ;) pid namespaces.
|
| The overall idea is the following:
|
| The namespace are organized as a tree - once a task is cloned
| with CLONE_NEWPIDS (yes, I've also
Pavel Emelianov [EMAIL PROTECTED] wrote:
| When searching the task by numerical id on may need to find
| it using global pid (as it is done now in kernel) or by its
| virtual id, e.g. when sending a signal to a task from one
| namespace the sender will specify the task's virtual id.
|
|
Pavel Emelianov [EMAIL PROTECTED] wrote:
| When user send signal from (say) init namespace to any task in a sub
| namespace the siginfo struct must not carry the sender's pid value, as
| this value may refer to some task in the destination namespace and thus
| may confuse the application.
Also,
I am not able to find a specific patch that this might be in,
but what happens when the child-reaper of a container exits ?
Do you terminate all processes in the container ? I thought
that was discussed earlier and the consensus was to terminate
all processes in that container and its subordinate
Cedric Le Goater [EMAIL PROTECTED] wrote:
| Pavel Emelianov wrote:
| struct pid_namespace will have the kmem_cache to allocate
| the pids from, the parent, as they are hierarchical, and
| the level of nesting value.
|
| struct pid will have a variable length array of pid_number-s
| one for
Pavel Emelianov [EMAIL PROTECTED] wrote:
| When showing pid to user or getting the pid numerical id for in-kernel
| use the value of this id may differ depending on the namespace.
|
| This set of helpers is used to get the global pid nr, the virtual (i.e.
| seen by task in its namespace) nr and
Pavel Emelianov [EMAIL PROTECTED] wrote:
| [EMAIL PROTECTED] wrote:
| Pavel Emelianov [EMAIL PROTECTED] wrote:
| | This is submition for inclusion of hierarchical, not kconfig
| | configurable, zero overheaded ;) pid namespaces.
| |
| | The overall idea is the following:
| |
| | The
Pavel Emelianov [EMAIL PROTECTED] wrote:
| [EMAIL PROTECTED] wrote:
| Subject: [PATCH 2/6] Rename pid_nr function
|
| From: Sukadev Bhattiprolu [EMAIL PROTECTED]
|
| Rename pid_nr() function to pid_to_nr() which is more descriptive
| and will hopefully cause less confusion with new
Pavel Emelianov [EMAIL PROTECTED] wrote:
| Serge E. Hallyn wrote:
| Quoting Pavel Emelianov ([EMAIL PROTECTED]):
| [EMAIL PROTECTED] wrote:
| Pavel Emelianov [EMAIL PROTECTED] wrote:
| | [EMAIL PROTECTED] wrote:
| | Subject: [PATCH 2/6] Rename pid_nr function
| |
| | From: Sukadev
1 - 100 of 1057 matches
Mail list logo
