From mboxrd@z Thu Jan 1 00:00:00 1970 Date: Tue, 2 Feb 2016 20:26:54 +0100 From: Gilles Chanteperdrix Message-ID: <20160202192654.GA29937@hermes.click-hack.org> References: <1453821607-20836-1-git-send-email-henning.schild@siemens.com> <1453902069-18824-1-git-send-email-henning.schild@siemens.com> <56A9F314.9030808@xenomai.org> <20160128215313.7717dafb@md1em3qc> <56AB9D2B.1030905@xenomai.org> <20160129183948.GE24550@hermes.click-hack.org> <20160202130849.44ee20d5@md1em3qc> <56B0B611.4070505@xenomai.org> <20160202173849.1a5e43fc@md1em3qc> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20160202173849.1a5e43fc@md1em3qc> Subject: Re: [Xenomai] [PATCH v2] ipipe x86 mm: handle huge pages in memory pinning List-Id: Discussions about the Xenomai project List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Henning Schild Cc: Jan Kiszka , Xenomai On Tue, Feb 02, 2016 at 05:38:49PM +0100, Henning Schild wrote: > I do not want to keep working on that patch just to hear that it is not > required in the next round. Gilles suggested it might not be and you > are the one that introduced the pinning. > > Maybe a HAVE_ARCH_HUGE_VMAP depends on !IPIPE would also be fine? I did not expect you to use my remark as an excuse for dropping the ball. What I meant is that maybe, there is a way to skip the call to __ipipe_pin_mapping_globally() entirely in the upper layers, for the "huge pages" case. And if there is, this should be preferred over patching vmalloc_sync_one, a function not meant to be used for huge mappings anyway. -- Gilles. https://click-hack.org