netdev - Re: [PATCH net v2] r8169: fix rtl8125b PAUSE frames blasting when suspended

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 for Android: free password hash cracker in your pocket

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

Message-ID: <4aeebf95-cf12-4462-80c7-dd1dafddb611@intel.com>
Date: Wed, 29 Nov 2023 15:04:56 -0800
From: Jacob Keller <jacob.e.keller@...el.com>
To: ChunHao Lin <hau@...ltek.com>, <hkallweit1@...il.com>
CC: <nic_swsd@...ltek.com>, <davem@...emloft.net>, <edumazet@...gle.com>,
	<kuba@...nel.org>, <pabeni@...hat.com>, <netdev@...r.kernel.org>,
	<linux-kernel@...r.kernel.org>, <grundler@...omium.org>,
	<stable@...r.kernel.org>
Subject: Re: [PATCH net v2] r8169: fix rtl8125b PAUSE frames blasting when
 suspended



On 11/29/2023 7:53 AM, ChunHao Lin wrote:
> When FIFO reaches near full state, device will issue pause frame.
> If pause slot is enabled(set to 1), in this time, device will issue
> pause frame only once. But if pause slot is disabled(set to 0), device
> will keep sending pause frames until FIFO reaches near empty state.
> 
> When pause slot is disabled, if there is no one to handle receive
> packets, device FIFO will reach near full state and keep sending
> pause frames. That will impact entire local area network.
> 
> This issue can be reproduced in Chromebox (not Chromebook) in
> developer mode running a test image (and v5.10 kernel):
> 1) ping -f $CHROMEBOX (from workstation on same local network)
> 2) run "powerd_dbus_suspend" from command line on the $CHROMEBOX
> 3) ping $ROUTER (wait until ping fails from workstation)
> 
> Takes about ~20-30 seconds after step 2 for the local network to
> stop working.
> 
> Fix this issue by enabling pause slot to only send pause frame once
> when FIFO reaches near full state.
> 

Makes sense. Avoiding the spam is good. The naming is a bit confusing
but I guess that comes from realtek datasheet?

> Fixes: f1bce4ad2f1c ("r8169: add support for RTL8125")
> Reported-by: Grant Grundler <grundler@...omium.org>
> Tested-by: Grant Grundler <grundler@...omium.org>
> Cc: stable@...r.kernel.org
> Signed-off-by: ChunHao Lin <hau@...ltek.com>
> ---
> v2:
> - update comment and title.
> ---
>  drivers/net/ethernet/realtek/r8169_main.c | 7 ++++++-
>  1 file changed, 6 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/net/ethernet/realtek/r8169_main.c b/drivers/net/ethernet/realtek/r8169_main.c
> index 62cabeeb842a..bb787a52bc75 100644
> --- a/drivers/net/ethernet/realtek/r8169_main.c
> +++ b/drivers/net/ethernet/realtek/r8169_main.c
> @@ -196,6 +196,7 @@ enum rtl_registers {
>  					/* No threshold before first PCI xfer */
>  #define	RX_FIFO_THRESH			(7 << RXCFG_FIFO_SHIFT)
>  #define	RX_EARLY_OFF			(1 << 11)
> +#define	RX_PAUSE_SLOT_ON		(1 << 11)	/* 8125b and later */

This confuses me though: RX_EARLY_OFF is (1 << 11) as well.. Is that
from a different set of devices? We're writing to the same register
RxConfig here I think in both cases?

Can you clarify if these are supposed to be the same bit?

>  #define	RXCFG_DMA_SHIFT			8
>  					/* Unlimited maximum PCI burst. */
>  #define	RX_DMA_BURST			(7 << RXCFG_DMA_SHIFT)
> @@ -2306,9 +2307,13 @@ static void rtl_init_rxcfg(struct rtl8169_private *tp)
>  	case RTL_GIGA_MAC_VER_40 ... RTL_GIGA_MAC_VER_53:
>  		RTL_W32(tp, RxConfig, RX128_INT_EN | RX_MULTI_EN | RX_DMA_BURST | RX_EARLY_OFF);
>  		break;
> -	case RTL_GIGA_MAC_VER_61 ... RTL_GIGA_MAC_VER_63:
> +	case RTL_GIGA_MAC_VER_61:
>  		RTL_W32(tp, RxConfig, RX_FETCH_DFLT_8125 | RX_DMA_BURST);
>  		break;

I assume there isn't a VER_62 between these?

> +	case RTL_GIGA_MAC_VER_63:
> +		RTL_W32(tp, RxConfig, RX_FETCH_DFLT_8125 | RX_DMA_BURST |
> +			RX_PAUSE_SLOT_ON);

We add RX_PAUSE_SLOT_ON now for RTL_GIGA_MAC_VER_63 in addition. Makes
sense.
> +		break;
>  	default:
>  		RTL_W32(tp, RxConfig, RX128_INT_EN | RX_DMA_BURST);
>  		break;