aio: rcu_read_lock protection for new rcu_dereference calls
PATCH 1/1 aio: rcu_read_lock protection for new rcu_dereference calls
Patch "aio: fix rcu sparse warnings introduced by ioctx table lookup patch" (77d30b14d24e557f89c41980011d72428514d729 in linux-next.git) introduced a couple of new rcu_dereference calls which are not protected by rcu_read_lock and result in following warnings during syscall fuzzing(trinity):
[ 471.646379] ===============================
[ 471.649727] [ INFO: suspicious RCU usage. ]
[ 471.653919] 3.11.0-next-20130906+ #496 Not tainted
[ 471.657792] -------------------------------
[ 471.661235] fs/aio.c:503 suspicious rcu_dereference_check() usage!
[ 471.665968]
[ 471.665968] other info that might help us debug this:
[ 471.665968]
[ 471.672141]
[ 471.672141] rcu_scheduler_active = 1, debug_locks = 1
[ 471.677549] 1 lock held by trinity-child0/3774:
[ 471.681675] #0: (&(&mm->ioctx_lock)->rlock){+.+...}, at: [<c119ba1a>] SyS_io_setup+0x63a/0xc70
[ 471.688721]
[ 471.688721] stack backtrace:
[ 471.692488] CPU: 1 PID: 3774 Comm: trinity-child0 Not tainted 3.11.0-next-20130906+ #496
[ 471.698437] Hardware name: Bochs Bochs, BIOS Bochs 01/01/2011
[ 471.703151] 00000000 00000000 c58bbf30 c18a814b de2234c0 c58bbf58 c10a4ec6 c1b0d824
[ 471.709544] c1b0f60e 00000001 00000001 c1af61b0 00000000 cb670ac0 c3aca000 c58bbfac
[ 471.716251] c119bc7c 00000002 00000001 00000000 c119b8dd 00000000 c10cf684 c58bbfb4
[ 471.722902] Call Trace:
[ 471.724859] [<c18a814b>] dump_stack+0x4b/0x66
[ 471.728772] [<c10a4ec6>] lockdep_rcu_suspicious+0xc6/0x100
[ 471.733716] [<c119bc7c>] SyS_io_setup+0x89c/0xc70
[ 471.737806] [<c119b8dd>] ? SyS_io_setup+0x4fd/0xc70
[ 471.741689] [<c10cf684>] ? __audit_syscall_entry+0x94/0xe0
[ 471.746080] [<c18b1fcc>] syscall_call+0x7/0xb
[ 471.749723] [<c1080000>] ? task_fork_fair+0x240/0x260--- fs/aio.c | 6 ++++++ 1 file changed, 6 insertions(+)
fs/aio.c +6 -0
@@ -470,6 +470,7 @@ static int ioctx_add_table(struct kioctx *ctx, struct mm_struct *mm)struct aio_ring *ring;spin_lock (&mm->ioctx_lock); +rcu_read_lock (); table =rcu_dereference (mm->ioctx_table);while (1 ) {
@@ -478,6 +479,7 @@ static int ioctx_add_table(struct kioctx *ctx, struct mm_struct *mm)if (!table->table[i]) { ctx->id = i; table->table[i] = ctx; +rcu_read_unlock ();spin_unlock (&mm->ioctx_lock); ring =kmap_atomic (ctx->ring_pages[0 ]);
@@ -488,6 +490,7 @@ static int ioctx_add_table(struct kioctx *ctx, struct mm_struct *mm) new_nr = (table ? table->nr :1 ) *4 ; +rcu_read_unlock ();spin_unlock (&mm->ioctx_lock); table = kzalloc(sizeof(*table) + sizeof(struct kioctx *) *
@@ -498,6 +501,7 @@ static int ioctx_add_table(struct kioctx *ctx, struct mm_struct *mm) table->nr = new_nr;spin_lock (&mm->ioctx_lock); +rcu_read_lock (); old =rcu_dereference (mm->ioctx_table);if (!old) {
@@ -622,10 +626,12 @@ static void kill_ioctx(struct mm_struct *mm, struct kioctx *ctx)struct kioctx_table *table;spin_lock (&mm->ioctx_lock); +rcu_read_lock (); table =rcu_dereference (mm->ioctx_table);WARN_ON (ctx != table->table[ctx->id]); table->table[ctx->id] =NULL ; +rcu_read_unlock ();spin_unlock (&mm->ioctx_lock);/* percpu_ref_kill() will do the necessary call_rcu() */
Signature · 7 lines
1.8.4 -- To unsubscribe, send a message with 'unsubscribe linux-aio' in the body to [email protected]. For more info on Linux AIO, see: http://www.kvack.org/aio/ Don't email: <a href=mailto:"[email protected]">[email protected]</a>
Quoted 33 lines · Artem Savkov
Patch "aio: fix rcu sparse warnings introduced by ioctx table lookup patch" (
77d30b14d24e557f89c41980011d72428514d729in linux-next.git) introduced a couple of newrcu_dereferencecalls which are not protected byrcu_read_lockand result in following warnings during syscall fuzzing(trinity):[ 471.646379] =============================== [ 471.649727] [ INFO: suspicious RCU usage. ] [ 471.653919] 3.11.0-next-20130906+ #496 Not tainted [ 471.657792] ------------------------------- [ 471.661235] fs/aio.c:503 suspicious rcu_dereference_check() usage! [ 471.665968] [ 471.665968] other info that might help us debug this: [ 471.665968] [ 471.672141] [ 471.672141] rcu_scheduler_active = 1, debug_locks = 1 [ 471.677549] 1 lock held by trinity-child0/3774: [ 471.681675] #0: (&(&mm->ioctx_lock)->rlock){+.+...}, at: [<c119ba1a>] SyS_io_setup+0x63a/0xc70 [ 471.688721] [ 471.688721] stack backtrace: [ 471.692488] CPU: 1 PID: 3774 Comm: trinity-child0 Not tainted 3.11.0-next-20130906+ #496 [ 471.698437] Hardware name: Bochs Bochs, BIOS Bochs 01/01/2011 [ 471.703151] 00000000 00000000 c58bbf30 c18a814b de2234c0 c58bbf58 c10a4ec6 c1b0d824 [ 471.709544] c1b0f60e 00000001 00000001 c1af61b0 00000000 cb670ac0 c3aca000 c58bbfac [ 471.716251] c119bc7c 00000002 00000001 00000000 c119b8dd 00000000 c10cf684 c58bbfb4 [ 471.722902] Call Trace: [ 471.724859] [<c18a814b>] dump_stack+0x4b/0x66 [ 471.728772] [<c10a4ec6>] lockdep_rcu_suspicious+0xc6/0x100 [ 471.733716] [<c119bc7c>] SyS_io_setup+0x89c/0xc70 [ 471.737806] [<c119b8dd>] ? SyS_io_setup+0x4fd/0xc70 [ 471.741689] [<c10cf684>] ? __audit_syscall_entry+0x94/0xe0 [ 471.746080] [<c18b1fcc>] syscall_call+0x7/0xb [ 471.749723] [<c1080000>] ? task_fork_fair+0x240/0x260
Quoted 1 line
Signed-off-by Artem Savkov <[email protected]>
Quoted 53 lines
--- fs/aio.c | 6 ++++++ 1 file changed, 6 insertions(+)fs/aio.c +6
-0@@ -470,6 +470,7 @@ static int ioctx_add_table(struct kioctx *ctx, struct mm_struct *mm)struct aio_ring *ring;spin_lock (&mm->ioctx_lock); +rcu_read_lock (); table =rcu_dereference (mm->ioctx_table);while (1 ) {@@ -478,6 +479,7 @@ static int ioctx_add_table(struct kioctx *ctx, struct mm_struct *mm)if (!table->table[i]) { ctx->id = i; table->table[i] = ctx; +rcu_read_unlock ();spin_unlock (&mm->ioctx_lock); ring =kmap_atomic (ctx->ring_pages[0 ]);@@ -488,6 +490,7 @@ static int ioctx_add_table(struct kioctx *ctx, struct mm_struct *mm) new_nr = (table ? table->nr :1 ) *4 ; +rcu_read_unlock ();spin_unlock (&mm->ioctx_lock); table = kzalloc(sizeof(*table) + sizeof(struct kioctx *) *@@ -498,6 +501,7 @@ static int ioctx_add_table(struct kioctx *ctx, struct mm_struct *mm) table->nr = new_nr;spin_lock (&mm->ioctx_lock); +rcu_read_lock (); old =rcu_dereference (mm->ioctx_table);if (!old) {@@ -622,10 +626,12 @@ static void kill_ioctx(struct mm_struct *mm, struct kioctx *ctx)struct kioctx_table *table;spin_lock (&mm->ioctx_lock); +rcu_read_lock (); table =rcu_dereference (mm->ioctx_table);WARN_ON (ctx != table->table[ctx->id]); table->table[ctx->id] =NULL ; +rcu_read_unlock ();spin_unlock (&mm->ioctx_lock);/* percpu_ref_kill() will do the necessary call_rcu() */
Quoted 4 lines · Artem Savkov
Patch "aio: fix rcu sparse warnings introduced by ioctx table lookup patch" (
77d30b14d24e557f89c41980011d72428514d729in linux-next.git) introduced a couple of newrcu_dereferencecalls which are not protected byrcu_read_lockand result in following warnings during syscall fuzzing(trinity):
Applied. Thanks Artem.
-ben
Signature · 7 lines
"Thought is the essence of where you are now." -- To unsubscribe, send a message with 'unsubscribe linux-aio' in the body to [email protected]. For more info on Linux AIO, see: http://www.kvack.org/aio/ Don't email: <a href=mailto:"[email protected]">[email protected]</a>