* [PATCH] elf_core_dump: use rcu_read_lock() to access ->real_parent
@ 2009-04-29 9:06 Oleg Nesterov
2009-04-29 9:46 ` David Howells
0 siblings, 1 reply; 4+ messages in thread
From: Oleg Nesterov @ 2009-04-29 9:06 UTC (permalink / raw)
To: kpm; +Cc: David Howells, Roland McGrath, linux-kernel
(on top of ptrace-s-parent-real_parent-in-binfmt_elf_fdpicc.patch)
In theory it is not safe to dereference ->parent/real_parent without
tasklist or rcu lock, we can race with re-parenting.
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
fs/binfmt_elf.c | 8 ++++++--
fs/binfmt_elf_fdpic.c | 8 ++++++--
2 files changed, 12 insertions(+), 4 deletions(-)
--- PTRACE/fs/binfmt_elf.c~RCU_PARENT 2009-04-06 00:03:40.000000000 +0200
+++ PTRACE/fs/binfmt_elf.c 2009-04-29 10:46:54.000000000 +0200
@@ -1340,8 +1340,10 @@ static void fill_prstatus(struct elf_prs
prstatus->pr_info.si_signo = prstatus->pr_cursig = signr;
prstatus->pr_sigpend = p->pending.signal.sig[0];
prstatus->pr_sighold = p->blocked.sig[0];
+ rcu_read_lock();
+ prstatus->pr_ppid = task_pid_vnr(rcu_dereference(p->real_parent));
+ rcu_read_unlock();
prstatus->pr_pid = task_pid_vnr(p);
- prstatus->pr_ppid = task_pid_vnr(p->real_parent);
prstatus->pr_pgrp = task_pgrp_vnr(p);
prstatus->pr_sid = task_session_vnr(p);
if (thread_group_leader(p)) {
@@ -1382,8 +1384,10 @@ static int fill_psinfo(struct elf_prpsin
psinfo->pr_psargs[i] = ' ';
psinfo->pr_psargs[len] = 0;
+ rcu_read_lock();
+ psinfo->pr_ppid = task_pid_vnr(rcu_dereference(p->real_parent));
+ rcu_read_unlock();
psinfo->pr_pid = task_pid_vnr(p);
- psinfo->pr_ppid = task_pid_vnr(p->real_parent);
psinfo->pr_pgrp = task_pgrp_vnr(p);
psinfo->pr_sid = task_session_vnr(p);
--- PTRACE/fs/binfmt_elf_fdpic.c~RCU_PARENT 2009-04-28 21:40:49.000000000 +0200
+++ PTRACE/fs/binfmt_elf_fdpic.c 2009-04-29 10:52:31.000000000 +0200
@@ -1387,8 +1387,10 @@ static void fill_prstatus(struct elf_prs
prstatus->pr_info.si_signo = prstatus->pr_cursig = signr;
prstatus->pr_sigpend = p->pending.signal.sig[0];
prstatus->pr_sighold = p->blocked.sig[0];
+ rcu_read_lock();
+ prstatus->pr_ppid = task_pid_vnr(rcu_dereference(p->real_parent));
+ rcu_read_unlock();
prstatus->pr_pid = task_pid_vnr(p);
- prstatus->pr_ppid = task_pid_vnr(p->real_parent);
prstatus->pr_pgrp = task_pgrp_vnr(p);
prstatus->pr_sid = task_session_vnr(p);
if (thread_group_leader(p)) {
@@ -1432,8 +1434,10 @@ static int fill_psinfo(struct elf_prpsin
psinfo->pr_psargs[i] = ' ';
psinfo->pr_psargs[len] = 0;
+ rcu_read_lock();
+ psinfo->pr_ppid = task_pid_vnr(rcu_dereference(p->real_parent));
+ rcu_read_unlock();
psinfo->pr_pid = task_pid_vnr(p);
- psinfo->pr_ppid = task_pid_vnr(p->real_parent);
psinfo->pr_pgrp = task_pgrp_vnr(p);
psinfo->pr_sid = task_session_vnr(p);
^ permalink raw reply [flat|nested] 4+ messages in thread
* [PATCH] elf_core_dump: use rcu_read_lock() to access ->real_parent
@ 2009-04-29 9:07 Oleg Nesterov
2009-04-29 19:46 ` Roland McGrath
0 siblings, 1 reply; 4+ messages in thread
From: Oleg Nesterov @ 2009-04-29 9:07 UTC (permalink / raw)
To: Andrew Morton; +Cc: David Howells, Roland McGrath, linux-kernel
(on top of ptrace-s-parent-real_parent-in-binfmt_elf_fdpicc.patch)
In theory it is not safe to dereference ->parent/real_parent without
tasklist or rcu lock, we can race with re-parenting.
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
fs/binfmt_elf.c | 8 ++++++--
fs/binfmt_elf_fdpic.c | 8 ++++++--
2 files changed, 12 insertions(+), 4 deletions(-)
--- PTRACE/fs/binfmt_elf.c~RCU_PARENT 2009-04-06 00:03:40.000000000 +0200
+++ PTRACE/fs/binfmt_elf.c 2009-04-29 10:46:54.000000000 +0200
@@ -1340,8 +1340,10 @@ static void fill_prstatus(struct elf_prs
prstatus->pr_info.si_signo = prstatus->pr_cursig = signr;
prstatus->pr_sigpend = p->pending.signal.sig[0];
prstatus->pr_sighold = p->blocked.sig[0];
+ rcu_read_lock();
+ prstatus->pr_ppid = task_pid_vnr(rcu_dereference(p->real_parent));
+ rcu_read_unlock();
prstatus->pr_pid = task_pid_vnr(p);
- prstatus->pr_ppid = task_pid_vnr(p->real_parent);
prstatus->pr_pgrp = task_pgrp_vnr(p);
prstatus->pr_sid = task_session_vnr(p);
if (thread_group_leader(p)) {
@@ -1382,8 +1384,10 @@ static int fill_psinfo(struct elf_prpsin
psinfo->pr_psargs[i] = ' ';
psinfo->pr_psargs[len] = 0;
+ rcu_read_lock();
+ psinfo->pr_ppid = task_pid_vnr(rcu_dereference(p->real_parent));
+ rcu_read_unlock();
psinfo->pr_pid = task_pid_vnr(p);
- psinfo->pr_ppid = task_pid_vnr(p->real_parent);
psinfo->pr_pgrp = task_pgrp_vnr(p);
psinfo->pr_sid = task_session_vnr(p);
--- PTRACE/fs/binfmt_elf_fdpic.c~RCU_PARENT 2009-04-28 21:40:49.000000000 +0200
+++ PTRACE/fs/binfmt_elf_fdpic.c 2009-04-29 10:52:31.000000000 +0200
@@ -1387,8 +1387,10 @@ static void fill_prstatus(struct elf_prs
prstatus->pr_info.si_signo = prstatus->pr_cursig = signr;
prstatus->pr_sigpend = p->pending.signal.sig[0];
prstatus->pr_sighold = p->blocked.sig[0];
+ rcu_read_lock();
+ prstatus->pr_ppid = task_pid_vnr(rcu_dereference(p->real_parent));
+ rcu_read_unlock();
prstatus->pr_pid = task_pid_vnr(p);
- prstatus->pr_ppid = task_pid_vnr(p->real_parent);
prstatus->pr_pgrp = task_pgrp_vnr(p);
prstatus->pr_sid = task_session_vnr(p);
if (thread_group_leader(p)) {
@@ -1432,8 +1434,10 @@ static int fill_psinfo(struct elf_prpsin
psinfo->pr_psargs[i] = ' ';
psinfo->pr_psargs[len] = 0;
+ rcu_read_lock();
+ psinfo->pr_ppid = task_pid_vnr(rcu_dereference(p->real_parent));
+ rcu_read_unlock();
psinfo->pr_pid = task_pid_vnr(p);
- psinfo->pr_ppid = task_pid_vnr(p->real_parent);
psinfo->pr_pgrp = task_pgrp_vnr(p);
psinfo->pr_sid = task_session_vnr(p);
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [PATCH] elf_core_dump: use rcu_read_lock() to access ->real_parent
2009-04-29 9:06 [PATCH] elf_core_dump: use rcu_read_lock() to access ->real_parent Oleg Nesterov
@ 2009-04-29 9:46 ` David Howells
0 siblings, 0 replies; 4+ messages in thread
From: David Howells @ 2009-04-29 9:46 UTC (permalink / raw)
To: Oleg Nesterov; +Cc: dhowells, kpm, Roland McGrath, linux-kernel
Oleg Nesterov <oleg@redhat.com> wrote:
> (on top of ptrace-s-parent-real_parent-in-binfmt_elf_fdpicc.patch)
>
> In theory it is not safe to dereference ->parent/real_parent without
> tasklist or rcu lock, we can race with re-parenting.
>
> Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Acked-by: David Howells <dhowells@redhat.com>
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [PATCH] elf_core_dump: use rcu_read_lock() to access ->real_parent
2009-04-29 9:07 Oleg Nesterov
@ 2009-04-29 19:46 ` Roland McGrath
0 siblings, 0 replies; 4+ messages in thread
From: Roland McGrath @ 2009-04-29 19:46 UTC (permalink / raw)
To: Oleg Nesterov; +Cc: Andrew Morton, David Howells, linux-kernel
Acked-by: Roland McGrath <roland@redhat.com>
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2009-04-29 19:47 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2009-04-29 9:06 [PATCH] elf_core_dump: use rcu_read_lock() to access ->real_parent Oleg Nesterov
2009-04-29 9:46 ` David Howells
-- strict thread matches above, loose matches on Subject: below --
2009-04-29 9:07 Oleg Nesterov
2009-04-29 19:46 ` Roland McGrath
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox