From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Michael Kerrisk (man-pages)" Subject: Re: [PATCH] [RFCv3] add manpages for Memory Protection Keys Date: Fri, 3 Jun 2016 14:25:31 -0500 Message-ID: References: <1464826600-17110-1-git-send-email-dave.hansen@intel.com> <647d23bf-a163-deee-d0ec-f961ecfb0b90@gmail.com> <57519A04.5020700@intel.com> Mime-Version: 1.0 Content-Type: text/plain; charset=windows-1252 Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <57519A04.5020700-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org> Sender: linux-man-owner-u79uwXL29TY76Z2rM5mHXA@public.gmane.org To: Dave Hansen , dave-gkUM19QKKo4@public.gmane.org Cc: mtk.manpages-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org, Dave Hansen , linux-man-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, linux-api-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, x86-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org List-Id: linux-man@vger.kernel.org On 06/03/2016 09:53 AM, Dave Hansen wrote: > On 06/02/2016 05:25 PM, Michael Kerrisk (man-pages) wrote: >> The convention for man-pages is that new sentences always start >> of new source lines. (This makes subsequent patches less "noisy", >> since the common unit of change in a text is a sentence.) >> Could you fix this throughout please? > > Yep, I can do that, and I'll also integrate all of your comments, > although I won't respond to all of them individually, I will integrate them. > >>> +no longer be used in any protection-key-related operations. >>> +.PP >>> +.RB ( pkey_alloc ()) >>> +.I flags >>> +may contain zero or more disable operations: >> >> Why "zero or more" rather than "zero or one"? I mean: >> what sense could it make to OR together PKEY_DISABLE_ACCESS and >> PKEY_DISABLE_WRITE? > > This is one of the attributes of the hardware that I carried up in to > the interfaces. The hardware contains two bits: one to write-disable > and one to access-disable. You're allowed to set both at the same time, > even though the "access" bit overrules the "write" bit when set. > > So, it doesn't make a ton of logical sense with these two flags, but it > might if we ever got an "execute disable" feature or some other feature > that could be combined more arbitrarily. So, I have a suggestion. How about tightening the constraint here, so that only one of these flags is allowed for now. (EINVAL if both are specified.) That constraint could always be relaxed later , if desired, and adding it now may allow some wriggle room later in terms of modifying the API or allowing for different architectural choices. Cheers, Michael -- Michael Kerrisk Linux man-pages maintainer; http://www.kernel.org/doc/man-pages/ Linux/UNIX System Programming Training: http://man7.org/training/ -- To unsubscribe from this list: send the line "unsubscribe linux-man" in the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org More majordomo info at http://vger.kernel.org/majordomo-info.html