On 24.04.2024 11:16, Roger Pau Monné wrote: > On Tue, Apr 23, 2024 at 04:32:14PM +0200, Jan Beulich wrote: >> Drop the inner scope that was left from earlier if/else removal. Take >> the opportunity and make the paging_unlock() invocation common to >> success and error paths, though. > > TBH I'm not sure I prefer the fact to continue function execution > after an error is found, I specially dislike that you have to add a > !rc check to the nestedhvm conditional block, and because anything > that we further add to the function would also need a !rc check. > >> >> Signed-off-by: Jan Beulich <jbeul...@suse.com> > > Acked-by: Roger Pau Monné <roger....@citrix.com>
Thanks. > Albeit I do prefer the extra call to paging_unlock() and early return > from the function in case of error. Which puts me in the middle of your preference and the one George voiced in the context of what is now cc950c49ae6a ("x86/PoD: tie together P2M update and increment of entry count"). Doing the extra adjustment was merely in the hope of meeting his desires ... Jan