From mboxrd@z Thu Jan 1 00:00:00 1970 From: Pavel Machek Subject: Re: [PATCH v8 2/4] fpga manager: add sysfs interface document Date: Sun, 15 Feb 2015 23:40:06 +0100 Message-ID: <20150215224006.GA5626@amd> References: <20150113200032.GA16205@obsidianresearch.com> <20150113222450.GA17475@obsidianresearch.com> <20150115184726.GA23247@obsidianresearch.com> <20150115204502.591bca1d@lxorguk.ukuu.org.uk> <20150121160151.453ba403@lxorguk.ukuu.org.uk> <20150121202700.GB4942@obsidianresearch.com> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Return-path: Content-Disposition: inline In-Reply-To: <20150121202700.GB4942@obsidianresearch.com> Sender: linux-doc-owner@vger.kernel.org To: Jason Gunthorpe Cc: Pantelis Antoniou , One Thousand Gnomes , atull , Greg Kroah-Hartman , hpa@zytor.com, Michal Simek , Michal Simek , Randy Dunlap , Linux Kernel Mailing List , devicetree@vger.kernel.org, robh+dt@kernel.org, Grant Likely , iws@ovro.caltech.edu, linux-doc@vger.kernel.org, Mark Brown , philip@balister.org, rubini@gnudd.com, Steffen Trumtrar , jason@lakedaemon.net, kyle.teske@ni.com, nico@linaro.org, Felipe Balbi , m.chehab@samsung.com, davidb@codeaurora.org, Rob Landley , davem@davemloft.net, cesarb@cesarb.net, sameo@linux.intel.com, Andrew Morton List-Id: devicetree@vger.kernel.org On Wed 2015-01-21 13:27:00, Jason Gunthorpe wrote: > On Wed, Jan 21, 2015 at 06:33:12PM +0200, Pantelis Antoniou wrote: > > Hi Alan, > >=20 > > > On Jan 21, 2015, at 18:01 , One Thousand Gnomes wrote: > > >=20 > > > On Thu, 15 Jan 2015 22:54:46 +0200 > > > Pantelis Antoniou wrote: > > >=20 > > >> Hi Alan, > > >>=20 > > >>> On Jan 15, 2015, at 22:45 , One Thousand Gnomes wrote: > > >>>=20 > > >>> On Thu, 15 Jan 2015 11:47:26 -0700 > > >>> Jason Gunthorpe wrote: > > >>>> It is a novel idea, my concern would be that embedding the FPG= A in the > > >>>> DT makes it permanent unswappable kernel memory. > > >>>> Not having the kernel hold the FPGA is best for many uses. > > >>>=20 > > >>> If you have a filesysytem before the FPGA is set up then it bel= ongs in > > >>> the file system. As you presumably loaded the kernel from somew= here there > > >>> ought to be a file system (even an initrd). > > >>>=20 > > >>=20 > > >> Request firmware does not imply keeping it around. You can alway= s re-request > > >> when reloading (although there=E2=80=99s a nasty big of caching = that needs to be > > >> resolved with the firmware loader). > > >=20 > > > Which comes down to the same thing. Unless you can prove that the= re is a > > > path to recover the firmware file that does not have any dependan= cies > > > upon the firmware executing (and those can be subtle and horrid a= t times) > > > you need to keep it around for suspend/resume at least and potent= ially > > > any unexpected error/reset. > > >=20 > >=20 > > In that case the only safe place to put it is in the kernel image i= tself, which > > is something the firmware loader already supports. >=20 > My point is that the current firmware layer is overly cautious and > FPGAs are very big. My current project on small Xilinx device has a > 10MB programming file. The biggest Xilinx device today has a max > bitfile size around 122MB. >=20 > So keeping that much memory pinned in the kernel when I can prove it > is uncessary for my system (either because there is no suspend/resume > possibility, or because I know the CPU can always access the > filesytem) is very undesirable. Well, your current device aalso has 1GB RAM, no? > Other systems might have to take the ram hit. I'd say the general case is "store bitstream in RAM" we can add optimalizations later. Pavel --=20 (english) http://www.livejournal.com/~pavelmachek (cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses= /blog.html