linux-kernel - Re: [PATCH v2 7/7] mtd: spi-nor: Rework the disabling of block write protection

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <836fcecd-766c-c7e3-74aa-06a148b146f8@microchip.com>
Date:   Sun, 25 Aug 2019 12:57:35 +0000
From:   <Tudor.Ambarus@...rochip.com>
To:     <boris.brezillon@...labora.com>
CC:     <marek.vasut@...il.com>, <vigneshr@...com>,
        <miquel.raynal@...tlin.com>, <richard@....at>,
        <linux-mtd@...ts.infradead.org>, <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH v2 7/7] mtd: spi-nor: Rework the disabling of block write
 protection

On 08/25/2019 03:24 PM, Boris Brezillon wrote:
> On Sat, 24 Aug 2019 12:00:48 +0000
> <Tudor.Ambarus@...rochip.com> wrote:
> 
>> From: Tudor Ambarus <tudor.ambarus@...rochip.com>
>>
>> Get rid of MFR handling and implement specific manufacturer
>> default_init() fixup hooks.
>>
>> Signed-off-by: Tudor Ambarus <tudor.ambarus@...rochip.com>
>> ---
>>  drivers/mtd/spi-nor/spi-nor.c | 30 ++++++++++++++++++++----------
>>  1 file changed, 20 insertions(+), 10 deletions(-)
>>
>> diff --git a/drivers/mtd/spi-nor/spi-nor.c b/drivers/mtd/spi-nor/spi-nor.c
>> index fc9e14777212..f4e9fcca619f 100644
>> --- a/drivers/mtd/spi-nor/spi-nor.c
>> +++ b/drivers/mtd/spi-nor/spi-nor.c
>> @@ -4146,6 +4146,16 @@ static int spi_nor_parse_sfdp(struct spi_nor *nor,
>>  	return err;
>>  }
>>  
>> +static void atmel_set_default_init(struct spi_nor *nor)
>> +{
>> +	nor->params.disable_block_protection = spi_nor_clear_sr_bp;
>> +}
>> +
>> +static void intel_set_default_init(struct spi_nor *nor)
>> +{
>> +	nor->params.disable_block_protection = spi_nor_clear_sr_bp;
> 
> That's weird: you can unlock blocks but locking is not
> explicitly flagged as supported (SNOR_F_HAS_LOCK is not set). Is that
> expected?

Yes. Manufacturers have different methods for locking/unlocking blocks of
memory. Right now we support just the stm/sr locking operations. sst26vf064b for
example, uses dedicated registers for reading/writing which blocks are
protected, and not the Status Register.

The reason for having disable_block_protection(), is that some spi-nor flashes
are write protected by default after a power-on reset cycle, in order to avoid
inadvertent writes during power-up. Backward compatibility imposes to disable
the write block protection at power-up by default, so that you can erase/write
the memory without having to send an unlock-all command. Which is bad in my
opinion and that's why I proposed https://patchwork.ozlabs.org/patch/1133278/.

Even if sst26vf064b does not yet have the lock ops implemented (SNOR_F_HAS_LOCK
is not set), I would like to be able to interact with it, so to disable the
block protection at power-up. This flash, and others, support a Global Unlock
Command which unlocks the entire memory array in a single cycle. We can't
determine who supports this command purely by manufacturer type, and it's not
discoverable through SFDP, so we'll have to add a nor->info flag for it:
UNLOCK_GLOBAL_BLOCK (see https://patchwork.ozlabs.org/patch/1152606/).

In conclusion, even if SNOR_F_HAS_LOCK is not set (the locking ops are not
implemented), we can still have disable_block_protection() mechanisms to unlock
the entire flash at power-up.

> 
>> +}
>> +
>>  static void macronix_set_default_init(struct spi_nor *nor)
>>  {
>>  	nor->params.quad_enable = macronix_quad_enable;
>> @@ -4173,6 +4183,14 @@ static void spi_nor_manufacturer_init_params(struct spi_nor *nor)
>>  {
>>  	/* Init flash parameters based on MFR */
>>  	switch (JEDEC_MFR(nor->info)) {
>> +	case SNOR_MFR_ATMEL:
>> +		atmel_set_default_init(nor);
>> +		break;
>> +
>> +	case SNOR_MFR_INTEL:
>> +		intel_set_default_init(nor);
>> +		break;
>> +
>>  	case SNOR_MFR_MACRONIX:
>>  		macronix_set_default_init(nor);
>>  		break;
>> @@ -4760,18 +4778,10 @@ int spi_nor_scan(struct spi_nor *nor, const char *name,
>>  	if (info->flags & SPI_S3AN)
>>  		nor->flags |=  SNOR_F_READY_XSR_RDY;
>>  
>> -	if (info->flags & SPI_NOR_HAS_LOCK)
>> +	if (info->flags & SPI_NOR_HAS_LOCK) {
> 
> If this flag implies SR_BP-based locking we should really rename it into
> SPI_NOR_HAS_SR_BP_LOCK to avoid any confusion.

Not only SR-based locking, should be a general flag that indicates that locking
ops are supported whichever they are. I would keep the SPI_NOR_HAS_LOCK and let
the manufacturer set its locking ops using the ->default_init() hook.

> 
>>  		nor->flags |= SNOR_F_HAS_LOCK;
>> -
>> -	/*
>> -	 * Atmel, SST, Intel/Numonyx, and others serial NOR tend to power up
>> -	 * with the software protection bits set.
>> -	 */
>> -	if (JEDEC_MFR(nor->info) == SNOR_MFR_ATMEL ||
>> -	    JEDEC_MFR(nor->info) == SNOR_MFR_INTEL ||
>> -	    JEDEC_MFR(nor->info) == SNOR_MFR_SST ||
>> -	    nor->info->flags & SPI_NOR_HAS_LOCK)
>>  		nor->params.disable_block_protection = spi_nor_clear_sr_bp;
>> +	}
>>  
>>  	/* Init flash parameters based on flash_info struct and SFDP */
>>  	spi_nor_init_params(nor);
> 
>