[net,0/2] net: dsa: Avoid cross-chip syncing of VLAN filtering

Message

Tobias Waldekranz Jan. 24, 2022, 9:09 p.m. UTC

This bug has been latent in the source for quite some time, I suspect
due to the homogeneity of both typical configurations and hardware.

On singlechip systems, this would never be triggered. The only reason
I saw it on my multichip system was because not all chips had the same
number of ports, which means that the misdemeanor alien call turned
into a felony array-out-of-bounds access.

Tobias Waldekranz (2):
  net: dsa: Move VLAN filtering syncing out of dsa_switch_bridge_leave
  net: dsa: Avoid cross-chip syncing of VLAN filtering

 net/dsa/switch.c | 40 +++++++++++++++++++++++++++-------------
 1 file changed, 27 insertions(+), 13 deletions(-)

Comments

Jakub Kicinski Jan. 25, 2022, 6:01 p.m. UTC | #1

On Mon, 24 Jan 2022 22:09:42 +0100 Tobias Waldekranz wrote:
> This bug has been latent in the source for quite some time, I suspect
> due to the homogeneity of both typical configurations and hardware.
> 
> On singlechip systems, this would never be triggered. The only reason
> I saw it on my multichip system was because not all chips had the same
> number of ports, which means that the misdemeanor alien call turned
> into a felony array-out-of-bounds access.

Applied, thanks, 934d0f039959 ("Merge branch
'dsa-avoid-cross-chip-vlan-sync'") in net-next.

Tobias Waldekranz Jan. 25, 2022, 7:05 p.m. UTC | #2

On Tue, Jan 25, 2022 at 10:01, Jakub Kicinski <[e-mail]> wrote:
> On Mon, 24 Jan 2022 22:09:42 +0100 Tobias Waldekranz wrote:
>> This bug has been latent in the source for quite some time, I suspect
>> due to the homogeneity of both typical configurations and hardware.
>> 
>> On singlechip systems, this would never be triggered. The only reason
>> I saw it on my multichip system was because not all chips had the same
>> number of ports, which means that the misdemeanor alien call turned
>> into a felony array-out-of-bounds access.
>
> Applied, thanks, 934d0f039959 ("Merge branch
> 'dsa-avoid-cross-chip-vlan-sync'") in net-next.

Thank you! Is there a particular reason that this was applied to
net-next? I guess my question is really: will it still be considered for
upcoming stable kernel releases?

Jakub Kicinski Jan. 25, 2022, 8:41 p.m. UTC | #3

On Tue, 25 Jan 2022 20:05:45 +0100 Tobias Waldekranz wrote:
> On Tue, Jan 25, 2022 at 10:01, Jakub Kicinski <[e-mail]> wrote:
> > On Mon, 24 Jan 2022 22:09:42 +0100 Tobias Waldekranz wrote:  
> >> This bug has been latent in the source for quite some time, I suspect
> >> due to the homogeneity of both typical configurations and hardware.
> >> 
> >> On singlechip systems, this would never be triggered. The only reason
> >> I saw it on my multichip system was because not all chips had the same
> >> number of ports, which means that the misdemeanor alien call turned
> >> into a felony array-out-of-bounds access.  
> >
> > Applied, thanks, 934d0f039959 ("Merge branch
> > 'dsa-avoid-cross-chip-vlan-sync'") in net-next.  
> 
> Is there a particular reason that this was applied to net-next?

Not sure, there were issues with kernel.org infra during the night,
could be unintentional.

> I guess my question is really: will it still be considered for
> upcoming stable kernel releases?

Only after the next merge window, but yes.

Tobias Waldekranz Jan. 25, 2022, 10:20 p.m. UTC | #4

On Tue, Jan 25, 2022 at 12:41, Jakub Kicinski <[e-mail]> wrote:
> On Tue, 25 Jan 2022 20:05:45 +0100 Tobias Waldekranz wrote:
>> On Tue, Jan 25, 2022 at 10:01, Jakub Kicinski <[e-mail]> wrote:
>> > On Mon, 24 Jan 2022 22:09:42 +0100 Tobias Waldekranz wrote:  
>> >> This bug has been latent in the source for quite some time, I suspect
>> >> due to the homogeneity of both typical configurations and hardware.
>> >> 
>> >> On singlechip systems, this would never be triggered. The only reason
>> >> I saw it on my multichip system was because not all chips had the same
>> >> number of ports, which means that the misdemeanor alien call turned
>> >> into a felony array-out-of-bounds access.  
>> >
>> > Applied, thanks, 934d0f039959 ("Merge branch
>> > 'dsa-avoid-cross-chip-vlan-sync'") in net-next.  
>> 
>> Is there a particular reason that this was applied to net-next?
>
> Not sure, there were issues with kernel.org infra during the night,
> could be unintentional.

Ahh ok, hope it gets sorted quickly!

>> I guess my question is really: will it still be considered for
>> upcoming stable kernel releases?
>
> Only after the next merge window, but yes.

I had a feeling that would be the case. Aright, not optimal, but not a
big deal either. Thanks for taking the time to respond.