Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
MIME-Version: 1.0
References: <20210811084924.52293-1-daniel@0x0f.com> <20210816101143.2a64d7b9@xps13>
In-Reply-To: <20210816101143.2a64d7b9@xps13>
From:   Daniel Palmer <daniel@0x0f.com>
Date:   Mon, 23 Aug 2021 23:19:02 +0900
Message-ID: <CAFr9PXnna+b3ChVUftT7YbU1kYR=5JDcik3bMNqzKK-LW=GQzw@mail.gmail.com>
Subject: Re: [PATCH v3] mtd: spinand: add support for Foresee FS35ND0*G parts
To:     Miquel Raynal <miquel.raynal@bootlin.com>
Cc:     linux-mtd@lists.infradead.org, richard@nod.at,
        Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
        linux-arm-kernel <linux-arm-kernel@lists.infradead.org>
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

Hi Miquel,

> > +     /*
> > +      * The datasheet says *successful* with 4 bits flipped.
> > +      * nandbiterrs always complains that the read reported
> > +      * successful but the data is incorrect.
> > +      */
> > +     case FS35ND01G_S1Y2_STATUS_ECC_4_BITFLIPS:
> > +             return 4;
>
> This is a real issue. Can you use the nandflipbits tool from the
> mtd-utils package (you should take a recent version of the package) and
> try to observe what happens when you insert a 4th bitflip in a section?
>
> I generally believe the tool more than the datasheet :)

Maybe I'm using it incorrectly but I can't get a 4 bit flipped
situation to happen.

I erased the paged so it's all 0xFF:

# flash_erase /dev/mtd0 0x8000000 1
Erasing 128 Kibyte @ 8000000 -- 100 % complete
# nanddump --bb=dumpbad -s 0x8000000 -l 1 -c -p /dev/mtd0
ECC failed: 0
ECC corrected: 6234
Number of bad blocks: 0
Number of bbt blocks: 0
Block size 131072, page size 2048, OOB size 64
Dumping data starting at 0x08000000 and ending at 0x08000001...
ECC: 3 corrected bitflip(s) at offset 0x08000000
0x08000000: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff  |................|
0x08000010: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff  |................|
0x08000020: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff  |................|
0x08000030: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff  |................|
0x08000040: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff  |................|
0x08000050: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff  |................|
0x08000060: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff  |................|

Then used nandflipbits to flip a bunch of bits in the first byte and
then a few other bytes:

# nanddump --bb=dumpbad -s 0x8000000 -l 1 -c -p /dev/mtd0
ECC failed: 0
ECC corrected: 6246
Number of bad blocks: 0
Number of bbt blocks: 0
Block size 131072, page size 2048, OOB size 64
Dumping data starting at 0x08000000 and ending at 0x08000001...
ECC: 3 corrected bitflip(s) at offset 0x08000000
0x08000000: f0 ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff  |................|
0x08000010: eb ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff  |................|
0x08000020: ef ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff  |................|
0x08000030: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff  |................|
0x08000040: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff  |................|
0x08000050: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff  |................|
0x08000060: ef ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff  |................|

Anyhow,
I think we should probably return -EBADMSG if the 4 bit flips status
appears as nandbiterrs always complains that the data is wrong.

Cheers,

Daniel