Re: [PATCH v5 bpf-next 2/4] ftrace: Allow IPMODIFY and DIRECT ops on the same function

["Naveen N. Rao" <naveen.n.rao@linux.ibm.com>]

Hi Song,

Song Liu wrote:
> 
> 
>> On Sep 9, 2022, at 4:58 AM, Naveen N. Rao <naveen.n.rao@linux.ibm.com> wrote:
>> 
>> Song Liu wrote:
> 
> [...]
> 
>>> +
>>> /**
>>>  * register_ftrace_function - register a function for profiling
>>>  * @ops:	ops structure that holds the function for profiling.
>>> @@ -8018,14 +8206,15 @@ int register_ftrace_function(struct ftrace_ops *ops)
>>> {
>>> 	int ret;
>>> -	ftrace_ops_init(ops);
>>> -
>>> -	mutex_lock(&ftrace_lock);
>>> -
>>> -	ret = ftrace_startup(ops, 0);
>>> +	lock_direct_mutex();
>> 
>> Trying to enable ftrace direct on powerpc, this is resulting in a hung task when testing samples/ftrace/ftrace-direct-modify.c
>> 
>> Essentially, the sample calls modify_ftrace_direct(), which grabs direct_mutex before calling ftrace_modify_direct_caller()->register_ftrace_function().
>> 
> 
> Thanks for the report. Would the following change fix the issue?
> 
> Song
> 
> diff --git i/kernel/trace/ftrace.c w/kernel/trace/ftrace.c
> index bc921a3f7ea8..2f1e6cfa834e 100644
> --- i/kernel/trace/ftrace.c
> +++ w/kernel/trace/ftrace.c
> @@ -5496,7 +5496,7 @@ int __weak ftrace_modify_direct_caller(struct ftrace_func_entry *entry,
>         if (ret)
>                 goto out_lock;
> 
> -       ret = register_ftrace_function(&stub_ops);
> +       ret = register_ftrace_function_nolock(&stub_ops);
>         if (ret) {
>                 ftrace_set_filter_ip(&stub_ops, ip, 1, 0);
>                 goto out_lock;
> 

That fixes it for me.
Reported-and-Tested-by: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com>


Thanks!
- Naveen