netdev - Re: [net-next,PATCH] net: dsa: microchip: Do not count RX/TX Bytes and discards on KSZ87xx

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: <c97cf296-f947-40d5-a4bc-6c30c2ba753b@mailbox.org>
Date: Sun, 2 Nov 2025 00:15:35 +0100
From: Marek Vasut <marek.vasut@...lbox.org>
To: vtpieter@...il.com, tristram.ha@...rochip.com
Cc: Arun.Ramadoss@...rochip.com, UNGLinuxDriver@...rochip.com,
 Woojung.Huh@...rochip.com, andrew@...n.ch, edumazet@...gle.com,
 kuba@...nel.org, netdev@...r.kernel.org, olteanv@...il.com, pabeni@...hat.com
Subject: Re: [net-next,PATCH] net: dsa: microchip: Do not count RX/TX Bytes
 and discards on KSZ87xx

On 8/13/25 9:03 AM, vtpieter@...il.com wrote:
>>>> Actually that many rx_discards may be a problem I need to find out.
>>>>
>>>> I think you are confused about how those MIB counters are read from
>>>> KSZ8795.  They are not using the code in ksz9477.c but in ksz8.c.
>>>
>>> See [1] TABLE 4-26: PORT MIB COUNTER INDIRECT MEMORY OFFSETS (CONTINUED)
>>> , page 108 at the very end . Notice the table ends at 0x1F
>>> TxMultipleCollision .
>>>
>>> See [2] TABLE 5-6: MIB COUNTERS (CONTINUED) , page 219 at the end of the
>>> table . Notice the table contains four extra entries , 0x80 RxByteCnt ,
>>> 0x81 TxByteCnt , 0x82 RxDropPackets , 0x83 TXDropPackets .
>>>
>>> These entries are present on KSZ9477 and missing on KSZ8795 .
>>>
>>> This is what this patch attempts to address.
>>
>> As I said KSZ8795 MIB counters are not using the code in ksz9477.c and
>> their last counter locations are not the same as KSZ9477.  KSZ9477 uses
>> ksz9477_r_mib_pkt while KSZ8795 uses ksz8795_r_mib_pkt.  The other
>> KSZ8895 and KSZ8863 switches uses ksz8863_r_mib_pkt.
>>
>> The 0x80 and such registers are not used in KSZ8795.  Its registers start
>> at 0x100, 0x101, ...
>>
>> They are in table 4-28.
>   
> Just wanted to chip in that for me, with a KSZ8794, the iproute2
> statistics work as expected as well after commit 0d3edc90c4a0 ("net:
> dsa: microchip: fix KSZ87xx family structure wrt the datasheet"). I'm
> on a 6.12 kernel and I see the following:
I'm sorry for my late reply, and yes, I got the versions mixed up.

The problem is still present in 6.6.115 , it is fixed in 6.12.y :

I can confirm 0d3edc90c4a0 ("net: dsa: microchip: fix KSZ87xx family 
structure wrt the datasheet") fixes the problem, thank you for that!

It would be good to backport this to 6.6.y LTS I think ?

The issue looks like this:

v6.11# cat /proc/net/dev
Inter-|   Receive                                                |  Transmit
  face |bytes    packets errs drop fifo frame compressed multicast|bytes 
    packets errs drop fifo colls carrier compressed
   eth0: 14818693   10334    0    0    0     0          0         0 
335961    2859    0    0    0     0       0          0
   lan1: 18446744073709510320   10324 14786075 14786075    0     0 
    0     10323 18446744073709540220    2849 342812    0 342812     0 
    0          0

v6.12.56# cat /proc/net/dev
Inter-|   Receive                                                |  Transmit
  face |bytes    packets errs drop fifo frame compressed multicast|bytes 
    packets errs drop fifo colls carrier compressed
   eth0: 14707237   10296    0    0    0     0          0         0 
340493    2889    0    0    0     0       0          0
   lan1: 14632865   10281    0    0    0     0          0         0 
335126    2875    0    0    0     0       0          0

Thank you for your help !