From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mail.bootlin.com ([62.4.15.54]) by bombadil.infradead.org with esmtp (Exim 4.90_1 #2 (Red Hat Linux)) id 1fQuQ6-0001s0-Ot for linux-mtd@lists.infradead.org; Thu, 07 Jun 2018 12:53:56 +0000 Date: Thu, 7 Jun 2018 14:53:26 +0200 From: Miquel Raynal To: Abhishek Sahu Cc: Boris Brezillon , David Woodhouse , Brian Norris , Marek Vasut , Richard Weinberger , Cyrille Pitchen , linux-arm-msm@vger.kernel.org, linux-kernel@vger.kernel.org, linux-mtd@lists.infradead.org, Andy Gross , Archit Taneja Subject: Re: [PATCH v3 13/16] mtd: rawnand: qcom: minor code reorganization for bad block check Message-ID: <20180607145326.339170fd@xps13> In-Reply-To: <53caff8799d30b6a81ac10f63a7c56c4@codeaurora.org> References: <1527250904-21988-1-git-send-email-absahu@codeaurora.org> <1527250904-21988-14-git-send-email-absahu@codeaurora.org> <20180526104629.74315561@xps13> <90ae248edf8a06a1d35e2da458f75af5@codeaurora.org> <20180528090352.254022ed@xps13> <53caff8799d30b6a81ac10f63a7c56c4@codeaurora.org> MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable List-Id: Linux MTD discussion mailing list List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Hi Abhishek, On Mon, 28 May 2018 15:40:52 +0530, Abhishek Sahu wrote: > On 2018-05-28 12:33, Miquel Raynal wrote: > > Hi Abhishek, =20 > > >> >> /* implements ecc->read_page() */ =20 > >> >> static int qcom_nandc_read_page(struct mtd_info *mtd, struct >> na= nd_chip *chip, > >> >> uint8_t *buf, int oob_required, int page) > >> >> @@ -2118,6 +2083,7 @@ static int qcom_nandc_block_bad(struct mtd_in= fo >> *mtd, loff_t ofs) > >> >> struct nand_ecc_ctrl *ecc =3D &chip->ecc; > >> >> int page, ret, bbpos, bad =3D 0; > >> >> u32 flash_status; > >> >> + u8 *bbm_bytes_buf =3D chip->data_buf; =20 > >> >> >> page =3D (int)(ofs >> chip->page_shift) & chip->pagemask; > >> >> >> @@ -2128,11 +2094,31 @@ static int qcom_nandc_block_bad(struct >= > mtd_info *mtd, loff_t ofs) =20 > >> >> * that contains the BBM > >> >> */ > >> >> host->use_ecc =3D false; > >> >> + bbpos =3D mtd->writesize - host->cw_size * (ecc->steps - 1); > >> > > Are we sure there is no layout with only 1 step? > >> >> All the layouts are such that, the BBM will come in =20 > >> first byte of spare area. =20 > >> >> For 4 bit ECC, the cw_size is 528 so for 2K page > >> >> 2048 - 528 * 3 =3D 464 =20 > > > My question was more about small page NANDs. But I suppose it works = =20 > > too if ecc->steps =3D=3D 1. > > =20 > Correct Miquel. >=20 > >> >> So for last CW, the 464 is BBM (i.e 2048th byte) in =20 > >> full page. =20 > >> >> > >> >> clear_bam_transaction(nandc); =20 > >> >> - ret =3D copy_last_cw(host, page); > >> >> - if (ret) > >> >> + clear_read_regs(nandc); > >> >> + > >> >> + set_address(host, host->cw_size * (ecc->steps - 1), page); > >> >> + update_rw_regs(host, 1, true); > >> >> + > >> >> + /* > >> >> + * The last codeword data will be copied from NAND device to NAND > >> >> + * controller internal HW buffer. Copy only required BBM size byt= es > >> >> + * from this HW buffer to bbm_bytes_buf which is present at > >> >> + * bbpos offset. > >> >> + */ > >> >> + nandc_set_read_loc(nandc, 0, bbpos, host->bbm_size, 1); > >> >> + config_nand_single_cw_page_read(nandc); > >> >> + read_data_dma(nandc, FLASH_BUF_ACC + bbpos, bbm_bytes_buf, > >> >> + host->bbm_size, 0); > >> >> + > >> >> + ret =3D submit_descs(nandc); > >> >> + free_descs(nandc); > >> >> + if (ret) { > >> >> + dev_err(nandc->dev, "failed to copy bad block bytes\n"); > >> >> goto err; > >> >> + } =20 > >> >> >> flash_status =3D le32_to_cpu(nandc->reg_read_buf[0]); > >> >> >> @@ -2141,12 +2127,10 @@ static int qcom_nandc_block_bad(struct >= > mtd_info *mtd, loff_t ofs) =20 > >> >> goto err; > >> >> } =20 > >> >> >> - bbpos =3D mtd->writesize - host->cw_size * (ecc->steps - 1); = =20 > >> >> - > >> >> - bad =3D nandc->data_buffer[bbpos] !=3D 0xff; > >> >> + bad =3D bbm_bytes_buf[0] !=3D 0xff; > >> > > This is suspect as it still points to the beginning of the data bu= ffer. =20 > >> > Can you please check you did not meant bbm_bytes_buf[bbpos]? > >> > =20 > >> The main thing here is > >> nandc_set_read_loc(nandc, 0, bbpos, host->bbm_size, 1); =20 > >> >> After reading one complete CW from NAND, the data will be still = =20 > >> in NAND HW buffer. =20 > >> >> The above register tells that we need to read data from =20 > >> bbpos of size host->bbm_size (which is 1 byte for 8 bus witdh > >> and 2 byte for 16 bus width) in bbm_bytes_buf. > > > I see: idx 0 in bbm_bytes_buf is the data at offset bbpos. Then =20 > > it's ok. =20 > > >> >> So bbm_bytes_buf[0] will contain the BBM first byte. =20 > >> and bbm_bytes_buf[1] will contain the BBM second byte. =20 > >> >> Regards, =20 > >> Abhishek =20 > >> >> >> >> if (chip->options & NAND_BUSWIDTH_16) =20 > >> >> - bad =3D bad || (nandc->data_buffer[bbpos + 1] !=3D 0xff); > >> >> + bad =3D bad || (bbm_bytes_buf[1] !=3D 0xff); =20 > > > Sorry, my mistake, I did not see the above line. > > > However, technically, the BBM could be located in the first, second o= r =20 > > last page of the block. You should check the three of them are 0xFF > > before declaring the block is not bad. =20 > > > The more I look at the function, the more I wonder if you actually ne= ed =20 > > it. Why does the generic nand_block_bad() implementation in the core > > do not fit? =20 >=20 > The BBM bytes can be accessed in raw mode only for QCOM NAND > Contoller. We started with following patch for initial patches >=20 > http://patchwork.ozlabs.org/patch/508565/ >=20 > I am also not very much sure, how can we go ahead now. > Ideally we need to use generic function only which > requires raw_read. >=20 I see, thanks for pointing this thread. Well for now then let's keep our driver-specific implementation. I will just ask you to do a consistent check as requested above (you can copy code from the core) and add a comment above this function explaining why it is needed (what you just told me). Thanks, Miqu=C3=A8l