From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from a.ns.miles-group.at ([95.130.255.143] helo=radon.swed.at) by merlin.infradead.org with esmtps (Exim 4.76 #1 (Red Hat Linux)) id 1TfXXz-0005gH-8i for linux-mtd@lists.infradead.org; Mon, 03 Dec 2012 15:03:04 +0000 Date: Mon, 3 Dec 2012 16:02:58 +0100 From: Richard Weinberger To: dedekind1@gmail.com Subject: Re: UBI wl_tree_add problems after PEB scrubbed Message-ID: <20121203160258.4a719a67@spider.haslach.nod.at> In-Reply-To: <1354545960.11583.17.camel@sauron.fi.intel.com> References: <50B8CB36.3010604@itwatchdogs.com> <1354545960.11583.17.camel@sauron.fi.intel.com> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Cc: linux-mtd@lists.infradead.org, Zach Sadecki List-Id: Linux MTD discussion mailing list List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Am Mon, 03 Dec 2012 16:46:00 +0200 schrieb Artem Bityutskiy : > /me looks at Richard... /me looks at the code. ;) > OK, I confirm that this is fastmap to blame. Namely, this commit: CONFIG_MTD_UBI_FASTMAP is y and ubi.fm_autoconvert is 1, correct? What tests are you using? ubi-tests? Does the issue still happen if you disable ubi_self_check_all_ff() in __wl_get_peb() And yeah, that's a very stupid bug, I'm very sorry. I'm running ubi-tests now for a while, self_check_ec() triggered once, but I've never seen "UBI assert failed in wl_tree_add at 204" so far. Thanks, //richard