From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S933749AbcBZUoI (ORCPT ); Fri, 26 Feb 2016 15:44:08 -0500 Received: from down.free-electrons.com ([37.187.137.238]:35290 "EHLO mail.free-electrons.com" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1754706AbcBZUoG (ORCPT ); Fri, 26 Feb 2016 15:44:06 -0500 Date: Fri, 26 Feb 2016 21:43:52 +0100 From: Boris Brezillon To: Stefan Agner Cc: David Woodhouse , Brian Norris , linux-mtd@lists.infradead.org, Daniel Mack , Haojian Zhuang , Robert Jarzmik , Kukjin Kim , Krzysztof Kozlowski , linux-samsung-soc@vger.kernel.org, linux-arm-kernel@lists.infradead.org, Ralf Baechle , linux-mips@linux-mips.org, Nicolas Ferre , Jean-Christophe Plagniol-Villard , Alexandre Belloni , Wenyou Yang , Josh Wu , Ezequiel Garcia , Maxime Ripard , Chen-Yu Tsai , linux-sunxi@googlegroups.com, Kyungmin Park , Greg Kroah-Hartman , devel@driverdev.osuosl.org, linux-kernel@vger.kernel.org, punnaiah choudary kalluri , Priit Laes , Kamal Dasu , bcm-kernel-feedback-list@broadcom.com, linux-api@vger.kernel.org Subject: Re: [PATCH v3 48/52] mtd: nand: vf610: switch to mtd_ooblayout_ops Message-ID: <20160226214352.5eb24c09@bbrezillon> In-Reply-To: <43666dfa2308e6575d47c76b688174ef@agner.ch> References: <1456448280-27788-1-git-send-email-boris.brezillon@free-electrons.com> <1456448280-27788-49-git-send-email-boris.brezillon@free-electrons.com> <43666dfa2308e6575d47c76b688174ef@agner.ch> X-Mailer: Claws Mail 3.11.1 (GTK+ 2.24.27; x86_64-pc-linux-gnu) MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Hi Stephan, On Fri, 26 Feb 2016 11:43:46 -0800 Stefan Agner wrote: > Hi Boris, > > On 2016-02-25 16:57, Boris Brezillon wrote: > > Implementing the mtd_ooblayout_ops interface is the new way of exposing > > ECC/OOB layout to MTD users. > > > > Signed-off-by: Boris Brezillon > > --- > > drivers/mtd/nand/vf610_nfc.c | 34 ++++------------------------------ > > 1 file changed, 4 insertions(+), 30 deletions(-) > > > > diff --git a/drivers/mtd/nand/vf610_nfc.c b/drivers/mtd/nand/vf610_nfc.c > > index 293feb1..da34de1 100644 > > --- a/drivers/mtd/nand/vf610_nfc.c > > +++ b/drivers/mtd/nand/vf610_nfc.c > > @@ -175,34 +175,6 @@ static inline struct vf610_nfc *mtd_to_nfc(struct > > mtd_info *mtd) > > return container_of(mtd_to_nand(mtd), struct vf610_nfc, chip); > > } > > > > -static struct nand_ecclayout vf610_nfc_ecc45 = { > > - .eccbytes = 45, > > - .eccpos = {19, 20, 21, 22, 23, > > - 24, 25, 26, 27, 28, 29, 30, 31, > > - 32, 33, 34, 35, 36, 37, 38, 39, > > - 40, 41, 42, 43, 44, 45, 46, 47, > > - 48, 49, 50, 51, 52, 53, 54, 55, > > - 56, 57, 58, 59, 60, 61, 62, 63}, > > - .oobfree = { > > - {.offset = 2, > > - .length = 17} } > > -}; > > - > > -static struct nand_ecclayout vf610_nfc_ecc60 = { > > - .eccbytes = 60, > > - .eccpos = { 4, 5, 6, 7, 8, 9, 10, 11, > > - 12, 13, 14, 15, 16, 17, 18, 19, > > - 20, 21, 22, 23, 24, 25, 26, 27, > > - 28, 29, 30, 31, 32, 33, 34, 35, > > - 36, 37, 38, 39, 40, 41, 42, 43, > > - 44, 45, 46, 47, 48, 49, 50, 51, > > - 52, 53, 54, 55, 56, 57, 58, 59, > > - 60, 61, 62, 63 }, > > - .oobfree = { > > - {.offset = 2, > > - .length = 2} } > > -}; > > - > > static inline u32 vf610_nfc_read(struct vf610_nfc *nfc, uint reg) > > { > > return readl(nfc->regs + reg); > > @@ -781,14 +753,16 @@ static int vf610_nfc_probe(struct platform_device *pdev) > > if (mtd->oobsize > 64) > > mtd->oobsize = 64; > > > > + /* > > + * mtd->ecclayout is not specified here because we're using the > > + * default large page ECC layout defined in NAND core. > > + */ > > if (chip->ecc.strength == 32) { > > nfc->ecc_mode = ECC_60_BYTE; > > chip->ecc.bytes = 60; > > > When I add a printk in nand_ooblayout_free_lp, I get this values: > free offset 2 lenth 62 > > This seems wrong to me. It seems that nand_ooblayout_free_lp needs steps > to be set to calculate the free bytes correctly. With chip->ecc.steps = > 1 the calculated values look good to me. Actually, ecc->steps is calculated in nand_scan_tail(), but it happens a bit too late (after mtd_ooblayout_count_freebytes() is called). I reordered things in nand_scan_tail() to address that. > > I quickly tried to run mtd_oobtest.ko test, however that faild: > Kernel BUG at 80399764 [verbose debug info unavailable] > Internal error: Oops - BUG: 0 [#1] ARM > Modules linked in: mtd_oobtest(+) > CPU: 0 PID: 676 Comm: insmod Tainted: G W > 4.5.0-rc1-00097-g4c60bde-dirty #320 > Hardware name: Freescale Vybrid VF5xx/VF6xx (Device Tree) > task: 8db27180 ti: 8ce74000 task.ti: 8ce74000 > PC is at nand_fill_oob+0x48/0xac > LR is at mtd_ooblayout_free+0x5c/0x74 > > mtd_ooblayout_set_databytes seems to return -ERANGE... > > Do you know what is going on here? A bug in the mtd_ooblayout_set/get_bytes() implementation. Just fixed it and pushed the changes on this branch [1]. The bugfix patches are here [2], [3]. Sorry for the inconvenience (I made some changes I thought were trivial and didn't retest the whole thing, which is bad :-/), and thanks for testing. Boris [1]https://github.com/bbrezillon/linux-0day/commits/nand/ecclayout [2]https://github.com/bbrezillon/linux-0day/commit/070393f6269fef4ba1929b5ab74b19466d3b5862 [3]https://github.com/bbrezillon/linux-0day/commit/51fcc44a6c98cd2dbbd22279ca4cd393528ea42a -- Boris Brezillon, Free Electrons Embedded Linux and Kernel engineering http://free-electrons.com