Re: [PATCH] net: stmmac: fix ethtool per-queue statistics

[Petr Tesařík]

On Fri, 5 Jan 2024 20:54:01 +0100
Andrew Lunn <andrew@xxxxxxx> wrote:

> On Fri, Jan 05, 2024 at 07:10:24PM +0100, Petr Tesarik wrote:
> > Fix per-queue statistics for devices with more than one queue.
> > 
> > The output data pointer is currently reset in each loop iteration,
> > effectively summing all queue statistics in the first four u64 values.
> > 
> > The summary values are not even labeled correctly. For example, if eth0 has
> > 2 queues, ethtool -S eth0 shows:
> > 
> >      q0_tx_pkt_n: 374 (actually tx_pkt_n over all queues)
> >      q0_tx_irq_n: 23  (actually tx_normal_irq_n over all queues)
> >      q1_tx_pkt_n: 462 (actually rx_pkt_n over all queues)
> >      q1_tx_irq_n: 446 (actually rx_normal_irq_n over all queues)
> >      q0_rx_pkt_n: 0
> >      q0_rx_irq_n: 0
> >      q1_rx_pkt_n: 0
> >      q1_rx_irq_n: 0
> > 
> > While touching this code, change the pointer type to u64 and get rid of the
> > weird pointer arithmetic.
> > 
> > Signed-off-by: Petr Tesarik <petr@xxxxxxxxxxx>
> > Fixes: 133466c3bbe1 ("net: stmmac: use per-queue 64 bit statistics where necessary")  
> 
> Hi Petr
> 
> There are a few process things you are missing. Please take a look at
> 
> https://www.kernel.org/doc/html/latest/process/maintainer-netdev.html
> 
> You need to indicate which tree this is for.

Ah. Thank you for the pointer. You see, this is my first netdev patch...

> Additionally, your Signed-off-by comes last.

Oh, I'm sorry for that.

> Patches for stable should ideally be minimal. And obviously correct. See:
> 
> https://www.kernel.org/doc/html/latest/process/stable-kernel-rules.html
> 
> So the bits about changing the pointer type and removing the weird
> arithmetic might be better suited for net-next, not net.

Good. Given that I have meanwhile found out that I will have to change
this whole function to fix the lockup on 32-bit systems, let me just
discard this part, add "net" to the subject and cc stable.

Resending...

Petr T