linux-kernel - Re: [PATCH v2] net: dsa: realtek-smi: fix indirect reg access for ports>3

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]

Message-ID: <b2ba44bc-57fb-b756-d693-7d896de90f3d@bang-olufsen.dk>
Date:   Fri, 26 Nov 2021 09:34:35 +0000
From:   Alvin Šipraga <ALSI@...g-olufsen.dk>
To:     "luizluca@...il.com" <luizluca@...il.com>
CC:     "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        "stable@...r.kernel.org" <stable@...r.kernel.org>,
        "netdev@...r.kernel.org" <netdev@...r.kernel.org>
Subject: Re: [PATCH v2] net: dsa: realtek-smi: fix indirect reg access for
 ports>3

Hi Luiz,

Thanks for your patch.

You needn't cc the stable list, since rtl8365mb doesn't exist in any 
stable release just yet. If you send a v3, just target net:

[PATCH net v3] net: dsa: ...

Then the fix will land in 5.16. :-)

On 11/26/21 09:12, luizluca@...il.com wrote:
> From: Luiz Angelo Daros de Luca <luizluca@...il.com>
> 
> This switch family can have up to 8 ports {0..7}. However,
> INDIRECT_ACCESS_ADDRESS_PHYNUM_MASK was using 2 bits instead of 3,
> dropping the most significant bit during indirect register reads and
> writes. Reading or writing ports 4, 5, 6, and 7 registers was actually
> manipulating, respectively, ports 0, 1, 2, and 3 registers.

Nice catch. Out of curiosity can you share what switch you are testing? 
So far the driver only advertises support for RTL8365MB-VC. Since that 
switch only uses PHY addresses 0~3, it shouldn't be affected by a 
narrower (2-bit) PHYNUM_MASK, right? Are you able to add words to the 
effect of "... now this fixes the driver to work with RTL83xxxx"?

The change is OK except for some comments below:

> 
> rtl8365mb_phy_{read,write} will now returns -EINVAL if phy is greater
> than 7.

I don't think this is really necessary: a valid (device tree) 
configuration should never specify a PHY with address greater than 7. Or 
am I missing something?

> 
> v2:
> - fix affected ports in commit message

The changelog shouldn't end up in the final commit message - please move 
it out in v3.

> 
> Fixes: 4af2950c50c8 ("net: dsa: realtek-smi: add rtl8365mb subdriver for RTL8365MB-VC")
> Signed-off-by: Luiz Angelo Daros de Luca <luizluca@...il.com>
> ---
>   drivers/net/dsa/rtl8365mb.c | 7 ++++++-
>   1 file changed, 6 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/net/dsa/rtl8365mb.c b/drivers/net/dsa/rtl8365mb.c
> index baaae97283c5..f4414ac74b61 100644
> --- a/drivers/net/dsa/rtl8365mb.c
> +++ b/drivers/net/dsa/rtl8365mb.c
> @@ -107,6 +107,7 @@
>   #define RTL8365MB_LEARN_LIMIT_MAX_8365MB_VC	2112
>   
>   /* Family-specific data and limits */
> +#define RTL8365MB_PHYADDRMAX	7
>   #define RTL8365MB_NUM_PHYREGS	32
>   #define RTL8365MB_PHYREGMAX	(RTL8365MB_NUM_PHYREGS - 1)
>   #define RTL8365MB_MAX_NUM_PORTS	(RTL8365MB_CPU_PORT_NUM_8365MB_VC + 1)
> @@ -176,7 +177,7 @@
>   #define RTL8365MB_INDIRECT_ACCESS_STATUS_REG			0x1F01
>   #define RTL8365MB_INDIRECT_ACCESS_ADDRESS_REG			0x1F02
>   #define   RTL8365MB_INDIRECT_ACCESS_ADDRESS_OCPADR_5_1_MASK	GENMASK(4, 0)
> -#define   RTL8365MB_INDIRECT_ACCESS_ADDRESS_PHYNUM_MASK		GENMASK(6, 5)
> +#define   RTL8365MB_INDIRECT_ACCESS_ADDRESS_PHYNUM_MASK		GENMASK(7, 5)
>   #define   RTL8365MB_INDIRECT_ACCESS_ADDRESS_OCPADR_9_6_MASK	GENMASK(11, 8)
>   #define   RTL8365MB_PHY_BASE					0x2000
>   #define RTL8365MB_INDIRECT_ACCESS_WRITE_DATA_REG		0x1F03
> @@ -679,6 +680,8 @@ static int rtl8365mb_phy_read(struct realtek_smi *smi, int phy, int regnum)
>   	u16 val;
>   	int ret;
>   
> +	if (phy > RTL8365MB_PHYADDRMAX)
> +		return -EINVAL;
>   	if (regnum > RTL8365MB_PHYREGMAX)

If you decide to keep these check, please add a newline after your returns.

>   		return -EINVAL;
>   
> @@ -704,6 +707,8 @@ static int rtl8365mb_phy_write(struct realtek_smi *smi, int phy, int regnum,
>   	u32 ocp_addr;
>   	int ret;
>   
> +	if (phy > RTL8365MB_PHYADDRMAX)
> +		return -EINVAL;
>   	if (regnum > RTL8365MB_PHYREGMAX)
>   		return -EINVAL;
>   
>