From: Stefan Beller Date: Fri, 13 Apr 2018 17:39:03 GMT Subject: Re: Optimizing writes to unchanged files during merges? Message-ID: In-Reply-To: Hi Linus, On Fri, Apr 13, 2018 at 10:14 AM, Linus Torvalds wrote: > > Comments? Because considering the problems this code has had, maybe > "stupid" really is the right approach... Would s/read/xread/ make sense in working_tree_matches ? If read returns unexpectedly due to EINTR or EAGAIN, this is no correctness issue, but you'd have to recompile the kernel. Stefan