From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from lithops.sigma-star.at ([195.201.40.130]) by bombadil.infradead.org with esmtps (Exim 4.90_1 #2 (Red Hat Linux)) id 1g7GTv-00031d-Du for linux-mtd@lists.infradead.org; Tue, 02 Oct 2018 08:56:54 +0000 From: Richard Weinberger To: =?utf-8?B?0JzRj9GB0L7QtdC00L7QsiDQkC7Qri4=?= , linux-mtd@lists.infradead.org Cc: "David Oberhollenzer (david.oberhollenzer@sigma-star.at)" Subject: Re: How to organize UBI volumes and power crash Date: Tue, 02 Oct 2018 10:56:23 +0200 Message-ID: <10059393.qm9dNCvvdv@blindfold> In-Reply-To: <5BB331EC.60204@ircoc.vrn.ru> References: <5BB32167.1040902@ircoc.vrn.ru> <5BB331EC.60204@ircoc.vrn.ru> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="UTF-8" List-Id: Linux MTD discussion mailing list List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Am Dienstag, 2. Oktober 2018, 10:53:00 CEST schrieb =D0=9C=D1=8F=D1=81=D0= =BE=D0=B5=D0=B4=D0=BE=D0=B2 =D0=90.=D0=AE.: > Thank ypu, Richard. >=20 > > Both UBI and UBIFS must not corrupt upon a powercut, if you face such = an > issue, please report. This could be a bug or a hardware problem. >=20 > But http://www.linux-mtd.infradead.org/doc/ubifs.html#L_unstable_bits=20 > says about such possible problems (unstable bits). This problem is highly theoretical and so far I have never seen it. So far anyone that claimed to suffer from that was wrong and it turned out = that it was some other issue. It is time to remove that paragraph from the website since it confuses user over and over. David, can you please file and merge a patch? =20 > > The general recommendation is having the MTD partition and therefore= =20 > the UBI device as large as possible such that the wear leveling large. > 02.10.2018 11:41, Richard Weinberger =D0=BD=D0=B0=D0=BF=D0=B8=D1=81=D0=B0= =D0=BB: >=20 > Do you mean one MTD partition for first UBI volume, and second MTD=20 > partition for another one? No, just one UBI device on one MTD partition with multiple UBI volumes. Thanks, //richard