netdev - Re: [PATCH 1/1] net: mvpp2: fix XDP rx queues registering

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] [thread-next>] [day] [month] [year] [list]

Message-ID: <bdc1f03c-036f-ee29-e2a1-a80f640adcc4@eho.link>
Date:   Mon, 6 Dec 2021 16:37:20 +0100
From:   Emmanuel Deloget <emmanuel.deloget@....link>
To:     Louis Amas <louis.amas@....link>
Cc:     andrii@...nel.org, ast@...nel.org, bpf@...r.kernel.org,
        daniel@...earbox.net, davem@...emloft.net, hawk@...nel.org,
        john.fastabend@...il.com, kafai@...com, kpsingh@...nel.org,
        kuba@...nel.org, linux-kernel@...r.kernel.org,
        linux@...linux.org.uk, mw@...ihalf.com, netdev@...r.kernel.org,
        songliubraving@...com, yhs@...com
Subject: Re: [PATCH 1/1] net: mvpp2: fix XDP rx queues registering

Hello,

On 10/11/2021 15:41, Louis Amas wrote:
> The registration of XDP queue information is incorrect because the
> RX queue id we use is invalid. When port->id == 0 it appears to works
> as expected yet it's no longer the case when port->id != 0.
> 
> When we register the XDP rx queue information (using
> xdp_rxq_info_reg() in function mvpp2_rxq_init()) we tell them to use
> rxq->id as the queue id. This value iscomputed as:
> rxq->id = port->id * max_rxq_count + queue_id
> 
> where max_rxq_count depends on the device version. In the MB case,
> this value is 32, meaning that rx queues on eth2 are numbered from
> 32 to 35 - there are four of them.
> 
> Clearly, this is not the per-port queue id that XDP is expecting:
> it wants a value in the range [0..3]. It shall directly use queue_id
> which is stored in rxq->logic_rxq -- so let's use that value instead.
> 
> This is consistent with the remaining part of the code in
> mvpp2_rxq_init().
> 
> Fixes: b27db2274ba8 ("mvpp2: use page_pool allocator")
> Signed-off-by: Louis Amas <louis.amas@....link>
> Signed-off-by: Emmanuel Deloget <emmanuel.deloget@....link>
> Reviewed-by: Marcin Wojtas <mw@...ihalf.com>
> ---
> This is a repost of [1]. The patch itself is not changed, but the
> commit message has been enhanced using part of the explaination in
> order to make it clearer (hopefully) and to incorporate the
> reviewed-by tag from Marcin.
> 
> v1: original patch
> v2: revamped commit description (no change in the patch itself)
> 
> [1] https://lore.kernel.org/bpf/20211109103101.92382-1-louis.amas@eho.link/
> 
>   drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c | 4 ++--
>   1 file changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c b/drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c
> index 587def69a6f7..f0ea377341c6 100644
> --- a/drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c
> +++ b/drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c
> @@ -2959,11 +2959,11 @@ static int mvpp2_rxq_init(struct mvpp2_port *port,
>   	mvpp2_rxq_status_update(port, rxq->id, 0, rxq->size);
>   
>   	if (priv->percpu_pools) {
> -		err = xdp_rxq_info_reg(&rxq->xdp_rxq_short, port->dev, rxq->id, 0);
> +		err = xdp_rxq_info_reg(&rxq->xdp_rxq_short, port->dev, rxq->logic_rxq, 0);
>   		if (err < 0)
>   			goto err_free_dma;
>   
> -		err = xdp_rxq_info_reg(&rxq->xdp_rxq_long, port->dev, rxq->id, 0);
> +		err = xdp_rxq_info_reg(&rxq->xdp_rxq_long, port->dev, rxq->logic_rxq, 0);
>   		if (err < 0)
>   			goto err_unregister_rxq_short;
>   
> 

Is there any update on this patch ? Without it, XDP only partially work 
on a MACCHIATOBin (read: it works on some ports, not on others, as 
described in our analysis sent together with the original patch).

Best regards,

-- Emmanuel Deloget