On Thu, Feb 14, 2019 at 3:31 AM Yue Hu <zbest...@gmail.com> wrote: > > From 08ac260f36b4c11eb33fa58466462a5e1027bdb3 Mon Sep 17 00:00:00 2001 > From: Yue Hu <zbest...@gmail.com> > Date: Thu, 14 Feb 2019 19:34:16 +0800 > Subject: [PATCH] pstore/ram: Avoid possible failure of dummy device unregister > > If create dummy platform device failed, dummy variable should not > been set as NULL. Because the following platform_device_unregister() > will judge if the dummy is NULL or not, if dummy is NULL nothing > happen for unregister flow. > > Signed-off-by: Yue Hu <zbest...@gmail.com>
I've gotten this email a few times today. :) I already replied that this isn't what we want to do: we have to mark it NULL since prior to that it was an ERR_PTR, and will break the unregister. -Kees > --- > fs/pstore/ram.c | 1 - > 1 file changed, 1 deletion(-) > > diff --git a/fs/pstore/ram.c b/fs/pstore/ram.c > index 4a06675..61bc573 100644 > --- a/fs/pstore/ram.c > +++ b/fs/pstore/ram.c > @@ -994,7 +994,6 @@ static void __init ramoops_register_dummy(void) > if (IS_ERR(dummy)) { > pr_info("could not create platform device: %ld\n", > PTR_ERR(dummy)); > - dummy = NULL; > ramoops_unregister_dummy(); > } > } > -- > 1.9.1 > -- Kees Cook