Re: [PATCH] fs: Reorder inode_owner_or_capable() to avoid needless

2017-06-29 Thread Al Viro
On Tue, Jun 20, 2017 at 02:40:24PM -0700, Kees Cook wrote: > Checking for capabilities should be the last operation when performing > access control tests so that PF_SUPERPRIV is set only when it was required > for success (implying that the capability was needed for the operation). Applied

Re: [PATCH] fs: Reorder inode_owner_or_capable() to avoid needless

2017-06-21 Thread Andy Lutomirski
On Tue, Jun 20, 2017 at 2:40 PM, Kees Cook wrote: > Checking for capabilities should be the last operation when performing > access control tests so that PF_SUPERPRIV is set only when it was required > for success (implying that the capability was needed for the operation). > Reviewed-by: Andy Lu

Re: [PATCH] fs: Reorder inode_owner_or_capable() to avoid needless

2017-06-20 Thread Serge E. Hallyn
Quoting Kees Cook (keesc...@chromium.org): > Checking for capabilities should be the last operation when performing > access control tests so that PF_SUPERPRIV is set only when it was required > for success (implying that the capability was needed for the operation). > > Reported-by: Solar Designe

[PATCH] fs: Reorder inode_owner_or_capable() to avoid needless

2017-06-20 Thread Kees Cook
Checking for capabilities should be the last operation when performing access control tests so that PF_SUPERPRIV is set only when it was required for success (implying that the capability was needed for the operation). Reported-by: Solar Designer Cc: Serge E. Hallyn Cc: Andy Lutomirski Signed-o