bsdacct: delete timer with sync intension
PATCH 1/1 bsdacct: delete timer with sync intension
acct_exit_ns --> acct_file_reopen deletes timer without check timer execution on other CPUs. So acct_timeout() can change an unmapped memory.
--- kernel/acct.c | 17 +++++++++-------- 1 files changed, 9 insertions(+), 8 deletions(-)
kernel/acct.c +9 -8
@@ -353,17 +353,18 @@ restart:void acct_exit_ns (struct pid_namespace *ns) { -struct bsd_acct_struct *acct; +struct bsd_acct_struct *acct = ns->bacct; -spin_lock (&acct_lock); - acct = ns->bacct; -if (acct !=NULL ) { -if (acct->file !=NULL ) -acct_file_reopen (acct,NULL ,NULL ); +if (acct ==NULL ) +return ; -kfree (acct); - } +del_timer_sync (&acct->timer); +spin_lock (&acct_lock); +if (acct->file !=NULL ) +acct_file_reopen (acct,NULL ,NULL );spin_unlock (&acct_lock); + +kfree (acct); }/*
Quoted 4 lines · On Thu, 25 Mar 2010 17:35:10 +0300 Vitaliy Gusev <[email protected]> wrote:
acct_exit_ns-->acct_file_reopendeletes timer without check timer execution on other CPUs. Soacct_timeout()can change an unmapped memory.
That sounds ugly.
Quoted 25 lines
--- kernel/acct.c | 17 +++++++++-------- 1 files changed, 9 insertions(+), 8 deletions(-)kernel/acct.c +3
-7@@ -353,17 +353,18 @@ restart:void acct_exit_ns (struct pid_namespace *ns) { -struct bsd_acct_struct *acct; +struct bsd_acct_struct *acct = ns->bacct; -spin_lock (&acct_lock); - acct = ns->bacct; -if (acct !=NULL ) { -if (acct->file !=NULL ) -acct_file_reopen (acct,NULL ,NULL ); +if (acct ==NULL ) +return ; -kfree (acct);
kernel/acct.c restart: · line 369
Quoted 10 lines
kernel/acct.c +6
-1- } +del_timer_sync (&acct->timer); +spin_lock (&acct_lock); +if (acct->file !=NULL ) +acct_file_reopen (acct,NULL ,NULL );spin_unlock (&acct_lock); + +kfree (acct); }
Is this sufficient? acct_file_reopen() does a del_timer(), so acct_timeout() could be running concurrently with acct_file_reopen(), but acct_file_reopen() is merrily altering data at *acct.
Perhaps acct_file_reopen() should be using del_timer_sync()?
check_free_space() is doing a similar thing:
del_timer(&acct->timer); acct->needcheck = 0;
the currently-running timer handler now goes and sets needcheck again!
Methinks the whole thing needs a bit of a rethink, bearing in mind how del_timer() actually works.