Re: [PATCH 0/2] make weak attributes in {k,u}probes

[Peter Zijlstra]

On Tue, Sep 20, 2022 at 10:32:46PM +0530, Naveen N. Rao wrote:
> Yipeng Zou wrote:
> > We have some function implementation under some arch does nothing.
> > We can mark it with weak attributes to improve.
> 
> That's not always an improvement. See [1] for an example, among many other
> patches to reduce use of __weak functions in the kernel.

More weak 'fun':

  https://lkml.kernel.org/r/20220419203807.655552918@xxxxxxxxxxxxx