Hi Peter, On Mon, Sep 29, 2025 at 02:50:45PM -0400, Peter Xu wrote: > On Thu, Sep 18, 2025 at 08:53:42PM +0530, Arun Menon wrote: > > The post_save() function call is responsible for cleaning up resources > > after the device state has been saved. > > Currently it is infallible, and does not return an error. > > > > It is called regardless of whether there is a preceding error from > > save or subsection save. That is, save and post_save() together are > > considered to be an atomic logical operation. > > > > It should not be confused as a counterpart of the post_load() function > > because post_load() does some sanity checks and returns an error if it > > fails. This commit, therefore, renames post_save() to cleanup_save() > > and makes it a void function. > > > > Reviewed-by: Fabiano Rosas <[email protected]> > > Suggested-by: Akihiko Odaki <[email protected]> > > Signed-off-by: Arun Menon <[email protected]> > > I'll need to drop this one patch because it breaks Rust. Please feel free > to send it separately or just leave post_save() as-is for now.
Thank you for the review. I see it now. I used the --enable-rust flag to change my configuration. I have removed the patch and resent the series (v15) > > PS: IMHO post_save() is still a good name to me, pairing well with > pre_save() and all *_load*() functions too. Dropping the retval should > already imply it cannot fail with/without a name change (and also because > modules can do more than "cleanups" in post_save()..). Yes, for now I have kept things as is because we do not essentially need to do something to the existing post_save() call; the dropping of return value can be looked into in a future patch. Thank you. > > -- > Peter Xu > Regards, Arun Menon
