diff options
author | Steven Rostedt (VMware) <rostedt@goodmis.org> | 2020-04-22 12:25:41 -0400 |
---|---|---|
committer | Steven Rostedt (VMware) <rostedt@goodmis.org> | 2020-06-29 11:42:47 -0400 |
commit | 5da7cd11d0811c35a6988d416053b5421bc61521 (patch) | |
tree | 3d00d9c09c9eb6b8c7bcc17d5ee6174b16d3b72d /drivers/target | |
parent | 0b4f8ddc0cc235f30431ed5c7d533b24e995d267 (diff) | |
download | linux-stable-5da7cd11d0811c35a6988d416053b5421bc61521.tar.gz linux-stable-5da7cd11d0811c35a6988d416053b5421bc61521.tar.bz2 linux-stable-5da7cd11d0811c35a6988d416053b5421bc61521.zip |
x86/ftrace: Only have the builtin ftrace_regs_caller call direct hooks
If a direct hook is attached to a function that ftrace also has a function
attached to it, then it is required that the ftrace_ops_list_func() is used
to iterate over the registered ftrace callbacks. This will also include the
direct ftrace_ops helper, that tells ftrace_regs_caller where to return to
(the direct callback and not the function that called it).
As this direct helper is only to handle the case of ftrace callbacks
attached to the same function as the direct callback, the ftrace callback
allocated trampolines (used to only call them), should never be used to
return back to a direct callback.
Only copy the portion of the ftrace_regs_caller that will return back to
what called it, and not the portion that returns back to the direct caller.
The direct ftrace_ops must then pick the ftrace_regs_caller builtin function
as its own trampoline to ensure that it will never have one allocated for
it (which would not include the handling of direct callbacks).
Link: http://lkml.kernel.org/r/20200422162750.495903799@goodmis.org
Cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
Diffstat (limited to 'drivers/target')
0 files changed, 0 insertions, 0 deletions