From mboxrd@z Thu Jan 1 00:00:00 1970 Subject: Re: [PATCH v2 2/8] genhomedircon: move fallback user to genhomedircon_user_entry_t To: Jason Zaman References: <1460131535-15688-1-git-send-email-jason@perfinion.com> <1461391499-20593-1-git-send-email-jason@perfinion.com> <1461391499-20593-3-git-send-email-jason@perfinion.com> <20160429192314.GA14636@meriadoc.perfinion.com> Cc: selinux@tycho.nsa.gov From: Stephen Smalley Message-ID: <72c34032-8d57-dd8f-a422-bb0e3abffaa3@tycho.nsa.gov> Date: Fri, 29 Apr 2016 16:29:18 -0400 MIME-Version: 1.0 In-Reply-To: <20160429192314.GA14636@meriadoc.perfinion.com> Content-Type: text/plain; charset=utf-8 List-Id: "Security-Enhanced Linux \(SELinux\) mailing list" List-Post: List-Help: On 04/29/2016 03:23 PM, Jason Zaman wrote: > On Fri, Apr 29, 2016 at 12:54:44PM -0400, Stephen Smalley wrote: >> On 04/23/2016 02:04 AM, Jason Zaman wrote: >>> The fallback user is used in all the write functions, making it use a >>> struct allows us to have everything consistent between normal and >>> fallback users. >>> >>> Signed-off-by: Jason Zaman >> >> When you split a patch into a series, the goal is to ensure that the >> code remains in a working state after each patch in the series. >> Otherwise, git bisect will often find broken states in the future. >> So, we have two options: >> 1. You can refactor this patch set so that it does compile after each >> patch (currently breaks on this one at least). >> 2. I can squash them all into one logical change. > > The best is squashing 2-4 into one then. The others are pretty self > contained. 2-4 was a really big change so I split it up even tho it > broke the build but if you'd prefer squashing them thats not a problem. I have to squash 2-5 to keep it buildable at each step.