linux-kernel - Re: [PATCH] mtd: spi-nor: Check for spi_nor_hwcaps

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] [day] [month] [year] [list]

Date:   Tue, 19 Sep 2017 19:33:15 +0200
From:   Cyrille Pitchen <cyrille.pitchen@...ev4u.fr>
To:     Richard Weinberger <richard@....at>, linux-mtd@...ts.infradead.org
Cc:     boris.brezillon@...e-electrons.com, linux-kernel@...r.kernel.org,
        marek.vasut@...il.com, computersforpeace@...il.com,
        dwmw2@...radead.org
Subject: Re: [PATCH] mtd: spi-nor: Check for spi_nor_hwcaps_read2cmd() return
 value

Hi Richard,

Le 17/09/2017 à 11:57, Richard Weinberger a écrit :
> The function can return a negativ value in case of errors,
> don't use it blindly as array index.
> 
> Detected by CoverityScan CID#1418067 ("Memory - illegal accesses")
> Fixes: f384b352cbf0 ("mtd: spi-nor: parse Serial Flash Discoverable
> Parameters (SFDP) tables")
> Signed-off-by: Richard Weinberger <richard@....at>
> ---
>  drivers/mtd/spi-nor/spi-nor.c | 3 +++
>  1 file changed, 3 insertions(+)
> 
> diff --git a/drivers/mtd/spi-nor/spi-nor.c b/drivers/mtd/spi-nor/spi-nor.c
> index cf1d4a15e10a..d71765739a93 100644
> --- a/drivers/mtd/spi-nor/spi-nor.c
> +++ b/drivers/mtd/spi-nor/spi-nor.c
> @@ -2145,6 +2145,9 @@ static int spi_nor_parse_bfpt(struct spi_nor *nor,
>  
>  		params->hwcaps.mask |= rd->hwcaps;
>  		cmd = spi_nor_hwcaps_read2cmd(rd->hwcaps);
> +		if (cmd < 0)
> +			return -EINVAL;
> +
>  		read = &params->reads[cmd];
>  		half = bfpt.dwords[rd->settings_dword] >> rd->settings_shift;
>  		spi_nor_set_read_settings_from_bfpt(read, half, rd->proto);
> 

This one is a false positive:

static const struct sfdp_bfpt_read sfdp_bfpt_reads[] = {
	/* Fast Read 1-1-2 */
	{
		SNOR_HWCAPS_READ_1_1_2,
		BFPT_DWORD(1), BIT(16),	/* Supported bit */
		BFPT_DWORD(4), 0,	/* Settings */
		SNOR_PROTO_1_1_2,
	},
	[...]
};

static int spi_nor_parse_bfpt(...)
{
	[...]
	for (i = 0; i < ARRAY_SIZE(sfdp_bfpt_reads); i++) {
		const struct sfdp_bfpt_read *rd = &sfdp_bfpt_reads[i];
		[...]
		cmd = spi_nor_hwcaps_read2cmd(rd->hwcaps);
		read = &params->reads[cmd];
		[...]
	}
	[...]
}

sfdp_bfpt_reads[] is a static const array where each element has a valid
.hwcaps value so spi_nor_hwcaps_read2cmd() can actually never return an
out of range 'cmd' result here.


However in spi_nor_select_read(), spi_nor_hwcaps_read2cmd() is called
too but this time its input argument now comes from the spi_nor_scan()
caller so in this later case the code already checks the output of
spi_nor_hwcaps_read2cmd() as it could be -EINVAL.


Do you want us to take this patch anyway to please the code checker ?
I don't think it is mandatory as a fix for 4.14-rc2.

Best regards,

Cyrille