Re: [PATCH] locking/lockdep: Fix potential buffer overrun problem in stack_trace[]

[Waiman Long <longman@redhat.com>]

On 12/19/19 9:57 PM, Bart Van Assche wrote:
> On 2019-12-19 10:28, Waiman Long wrote:
>> If the lockdep code is really running out of the stack_trace entries,
>> there is a possiblity that buffer overrun can happen and corrupt the
>              ^^^^^^^^^^
>              possibility?
>> data immediately after stack_trace[].
>>
>> If there is less than LOCK_TRACE_SIZE_IN_LONGS entries left before
>> the call to save_trace(), the max_entries computation will leave it
>> with a very large positive number because of its unsigned nature. The
>> subsequent call to stack_trace_save() will then corrupt the data after
>> stack_trace[]. Fix that by changing max_entries to a signed integer
>> and check for negative value before calling stack_trace_save().
>>
>> Fixes: 12593b7467f9 ("locking/lockdep: Reduce space occupied by stack traces")
>> Signed-off-by: Waiman Long <longman@redhat.com>
>> ---
>>  kernel/locking/lockdep.c | 7 +++----
>>  1 file changed, 3 insertions(+), 4 deletions(-)
>>
>> diff --git a/kernel/locking/lockdep.c b/kernel/locking/lockdep.c
>> index 32282e7112d3..56e260a7582f 100644
>> --- a/kernel/locking/lockdep.c
>> +++ b/kernel/locking/lockdep.c
>> @@ -482,7 +482,7 @@ static struct lock_trace *save_trace(void)
>>  	struct lock_trace *trace, *t2;
>>  	struct hlist_head *hash_head;
>>  	u32 hash;
>> -	unsigned int max_entries;
>> +	int max_entries;
>>  
>>  	BUILD_BUG_ON_NOT_POWER_OF_2(STACK_TRACE_HASH_SIZE);
>>  	BUILD_BUG_ON(LOCK_TRACE_SIZE_IN_LONGS >= MAX_STACK_TRACE_ENTRIES);
>> @@ -490,10 +490,8 @@ static struct lock_trace *save_trace(void)
>>  	trace = (struct lock_trace *)(stack_trace + nr_stack_trace_entries);
>>  	max_entries = MAX_STACK_TRACE_ENTRIES - nr_stack_trace_entries -
>>  		LOCK_TRACE_SIZE_IN_LONGS;
>> -	trace->nr_entries = stack_trace_save(trace->entries, max_entries, 3);
>>  
>> -	if (nr_stack_trace_entries >= MAX_STACK_TRACE_ENTRIES -
>> -	    LOCK_TRACE_SIZE_IN_LONGS - 1) {
>> +	if (max_entries < 0) {
>>  		if (!debug_locks_off_graph_unlock())
>>  			return NULL;
>>  
>> @@ -502,6 +500,7 @@ static struct lock_trace *save_trace(void)
>>  
>>  		return NULL;
>>  	}
>> +	trace->nr_entries = stack_trace_save(trace->entries, max_entries, 3);
>>  
>>  	hash = jhash(trace->entries, trace->nr_entries *
>>  		     sizeof(trace->entries[0]), 0);
> I'm not sure whether it is useful to call stack_trace_save() if
> max_entries == 0. How about changing the "max_entries < 0" test into
> "max_entries <= 0"?

I have actually added some instrumentation code to check the
distribution of stack trace lengths. I did get hits (about 40) on
zero-length stack traces after system bootup. But I am fine changing it
to <= 0.

Cheers,
Longman