On Sat, 20 May 2023 23:58:01 +0530
Naveen N Rao [off-list ref] wrote:

quoted

At the begining of ftrace_replace_code(), don't you miss:

	if (unlikely(ftrace_disabled))
		return;  

ftrace_disabled is a static, so it can't be tested here. FWIW, x86 also 
implements ftrace_replace_code(), and doesn't test ftrace_disabled.

Looking at the call path, it looks like it gets tested much before, so 
may not be necessary. It sure would be good to get Steven's view on 
this.

Yes, ftrace_disabled should not be used in arch code. It should be
caught before then.

Thanks,

-- Steve