[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <fe937905-9641-4c21-8160-4c8943f97c9e@solid-run.com>
Date: Wed, 20 Mar 2024 16:41:09 +0000
From: Josua Mayer <josua@...id-run.com>
To: Jiri Pirko <jiri@...nulli.us>
CC: Andrew Lunn <andrew@...n.ch>, Florian Fainelli <f.fainelli@...il.com>,
Vladimir Oltean <olteanv@...il.com>, "David S. Miller" <davem@...emloft.net>,
Eric Dumazet <edumazet@...gle.com>, Jakub Kicinski <kuba@...nel.org>, Paolo
Abeni <pabeni@...hat.com>, "netdev@...r.kernel.org" <netdev@...r.kernel.org>,
"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] net: dsa: mv88e6xxx: add warning for truncated mdio bus
id
Am 20.03.24 um 17:03 schrieb Jiri Pirko:
> Wed, Mar 20, 2024 at 03:33:24PM CET, josua@...id-run.com wrote:
>> Am 20.03.24 um 15:09 schrieb Jiri Pirko:
>>> Wed, Mar 20, 2024 at 02:48:55PM CET, josua@...id-run.com wrote:
>>>> mv88e6xxx supports multiple mdio buses as children, e.g. to model both
>>>> internal and external phys. If the child buses mdio ids are truncated,
>>>> they might collide which each other leading to an obscure error from
>>>> kobject_add.
>>>>
>>>> The maximum length of bus id is currently defined as 61
>>>> (MII_BUS_ID_SIZE). Truncation can occur on platforms with long node
>>>> names and multiple levels before the parent bus on whiich the dsa switch
>>> s/whiich/which/
>>>
>>>
>>>> sits such as on CN9130 [1].
>>>>
>>>> Test whether the return value of snprintf exceeds the maximum bus id
>>>> length and print a warning.
>>>>
>>>> [1]
>>>> [ 8.324631] mv88e6085 f212a200.mdio-mii:04: switch 0x1760 detected: Marvell 88E6176, revision 1
>>>> [ 8.389516] mv88e6085 f212a200.mdio-mii:04: Truncated bus-id may collide.
>>>> [ 8.592367] mv88e6085 f212a200.mdio-mii:04: Truncated bus-id may collide.
>>>> [ 8.623593] sysfs: cannot create duplicate filename '/devices/platform/cp0/cp0:config-space@...00000/f212a200.mdio/mdio_bus/f212a200.mdio-mii/f212a200.mdio-mii:04/mdio_bus/!cp0!config-space@...00000!mdio@...200!ethernet-switch@...di'
>>>> [ 8.785480] kobject: kobject_add_internal failed for !cp0!config-space@...00000!mdio@...200!ethernet-switch@...di with -EEXIST, don't try to register things with the same name in the same directory.
>>>> [ 8.936514] libphy: mii_bus /cp0/config-space@...00000/mdio@...200/ethernet-switch@...di failed to register
>>>> [ 8.946300] mdio_bus !cp0!config-space@...00000!mdio@...200!ethernet-switch@...di: __mdiobus_register: -22
>>>> [ 8.956003] mv88e6085 f212a200.mdio-mii:04: Cannot register MDIO bus (-22)
>>>> [ 8.965329] mv88e6085: probe of f212a200.mdio-mii:04 failed with error -22
>>>>
>>>> Signed-off-by: Josua Mayer <josua@...id-run.com>
>>> This is not bug fix, assume you target net-next. Please:
>>> 1) Next time, indicate that in the patch subject like this:
>>> [patch net-next] xxx
>>> 2) net-next is currently closed, repost next week.
>> Correct, thanks - will do.
>> Just for future reference for those occasional contributors -
>> is there such a thing as an lkml calendar?
>>>> ---
>>>> drivers/net/dsa/mv88e6xxx/chip.c | 6 ++++--
>>>> 1 file changed, 4 insertions(+), 2 deletions(-)
>>>>
>>>> diff --git a/drivers/net/dsa/mv88e6xxx/chip.c b/drivers/net/dsa/mv88e6xxx/chip.c
>>>> index 614cabb5c1b0..1c40f7631ab1 100644
>>>> --- a/drivers/net/dsa/mv88e6xxx/chip.c
>>>> +++ b/drivers/net/dsa/mv88e6xxx/chip.c
>>>> @@ -3731,10 +3731,12 @@ static int mv88e6xxx_mdio_register(struct mv88e6xxx_chip *chip,
>>>>
>>>> if (np) {
>>>> bus->name = np->full_name;
>>>> - snprintf(bus->id, MII_BUS_ID_SIZE, "%pOF", np);
>>>> + if (snprintf(bus->id, MII_BUS_ID_SIZE, "%pOF", np) >= MII_BUS_ID_SIZE)
>>>> + dev_warn(chip->dev, "Truncated bus-id may collide.\n");
>>> How about instead of warn&fail fallback to some different name in this
>>> case?
>> Duplicate could be avoided by truncating from the start,
>> however I don't know if that is a good idea.
>> It affects naming of paths in sysfs, and the root cause is
>> difficult to spot.
>>>> } else {
>>>> bus->name = "mv88e6xxx SMI";
>>>> - snprintf(bus->id, MII_BUS_ID_SIZE, "mv88e6xxx-%d", index++);
>>>> + if (snprintf(bus->id, MII_BUS_ID_SIZE, "mv88e6xxx-%d", index++) >= MII_BUS_ID_SIZE)
>>> How exactly this may happen?
>> It can happen on switch nodes at deep levels in the device-tree,
> Read again, my question is about the else branch.
Oh!
The else case occurs when the switch node does
not have a child node named "mdio":
/* Always register one mdio bus for the internal/default mdio
* bus. This maybe represented in the device tree, but is
* optional.
*/
child = of_get_child_by_name(np, "mdio");
err = mv88e6xxx_mdio_register(chip, child, false);
of_node_put(child);
In this case child is passed to np as NULL.
For example if we have no "mdio" child node, but do have "mdio-external",
the else case creates an mdio bus named "mv88e6xxx-%d".
Then we would end up - in my example - with these two:
cp0!config-space@...00000!mdio@...200!ethernet-switch@...v88e6xxx-0
cp0!config-space@...00000!mdio@...200!ethernet-switch@...dio-external
Truncated to MII_BUS_ID_SIZE:
cp0!config-space@...00000!mdio@...200!ethernet-switch@...v8
cp0!config-space@...00000!mdio@...200!ethernet-switch@...di
They are not duplicates, but too close for comfort.
Different device may exceed maximum size again.
>> while describing both internal and external mdio buses of a switch.
>> E.g. Documentation/devicetree/bindings/net/dsa/marvell,mv88e6xxx.yaml
>>
>> On CN9130 platform device-tree looks like this:
>>
>> / {
>> cp0 {
>> config-space@...00000 {
>> mdio@...200 {
>> ethernet-switch@4 {
>> mdio { ... };
>> mdio-external { ... };
>> };
>> };
>> };
>> };
>> };
>>
>> For mdio-external child all the names alone, without separators,
>> make up 66 characters, exceeding: MII_BUS_ID_SIZE:
>> cp0config-space@...00000mdio@...200ethernet-switch@...io-external
>>
>> With separators ('!') we have:
>> cp0!config-space@...00000!mdio@...200!ethernet-switch@...dio
>> cp0!config-space@...00000!mdio@...200!ethernet-switch@...dio-external
>> Truncated to MII_BUS_ID_SIZE:
>> cp0!config-space@...00000!mdio@...200!ethernet-switch@...di
>> cp0!config-space@...00000!mdio@...200!ethernet-switch@...di
>> They become duplicates.
>>
>>>> + dev_warn(chip->dev, "Truncated bus-id may collide.\n");
>> Another option (imo) is to force the issue and return error code.
>> Then the only way out would be increase of MII_BUS_ID_SIZE.
>>>> }
>>>>
>>>> bus->read = mv88e6xxx_mdio_read;
>>>>
>>>> ---
>>>> base-commit: e8f897f4afef0031fe618a8e94127a0934896aba
>>>> change-id: 20240320-mv88e6xxx-truncate-busid-34a1d2769bbf
>>>>
>>>> Sincerely,
>>>> --
>>>> Josua Mayer <josua@...id-run.com>
>>>>
>>>>
Powered by blists - more mailing lists