linux-kernel - Re: [PATCH net-next 2/2] r8169: disable RTL8126 ZRX-DC timeout

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: <ecfc71d3-47b6-4f17-b081-69452e7884ac@gmail.com>
Date: Fri, 14 Mar 2025 12:33:27 +0100
From: Heiner Kallweit <hkallweit1@...il.com>
To: ChunHao Lin <hau@...ltek.com>, nic_swsd@...ltek.com,
 andrew+netdev@...n.ch, davem@...emloft.net, edumazet@...gle.com,
 kuba@...nel.org, pabeni@...hat.com
Cc: netdev@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH net-next 2/2] r8169: disable RTL8126 ZRX-DC timeout

On 14.03.2025 08:50, ChunHao Lin wrote:
> Disable it due to it dose not meet ZRX-DC specification. If it is enabled,

dose -> does

> device will exit L1 substate every 100ms. Disable it for saving more power
> in L1 substate.
> 
> Signed-off-by: ChunHao Lin <hau@...ltek.com>
> ---
>  drivers/net/ethernet/realtek/r8169_main.c | 16 ++++++++++++++++
>  1 file changed, 16 insertions(+)
> 
> diff --git a/drivers/net/ethernet/realtek/r8169_main.c b/drivers/net/ethernet/realtek/r8169_main.c
> index 3c663fca07d3..dfc96b09b85e 100644
> --- a/drivers/net/ethernet/realtek/r8169_main.c
> +++ b/drivers/net/ethernet/realtek/r8169_main.c
> @@ -2852,6 +2852,21 @@ static u32 rtl_csi_read(struct rtl8169_private *tp, int addr)
>  		RTL_R32(tp, CSIDR) : ~0;
>  }
>  
> +static void rtl_disable_zrxdc_timeout(struct rtl8169_private *tp)
> +{
> +	struct pci_dev *pdev = tp->pci_dev;
> +	u8 val;
> +
> +	if (pdev->cfg_size > 0x0890 &&
> +	    pci_read_config_byte(pdev, 0x0890, &val) == PCIBIOS_SUCCESSFUL &&
> +	    pci_write_config_byte(pdev, 0x0890, val & ~BIT(0)) == PCIBIOS_SUCCESSFUL)
> +		return;
> +
> +	netdev_notice_once(tp->dev,
> +		"No native access to PCI extended config space, falling back to CSI\n");
> +	rtl_csi_write(tp, 0x0890, rtl_csi_read(tp, 0x0890) & ~BIT(0));
> +}
> +

Does the datasheet have a name for this extended config space register and bit 0?
This would be better than using magic numbers.

I think we can factor out the extended config space access to a helper. The same code
we have in another place already. But this can be done as a follow-up.

>  static void rtl_set_aspm_entry_latency(struct rtl8169_private *tp, u8 val)
>  {
>  	struct pci_dev *pdev = tp->pci_dev;
> @@ -3824,6 +3839,7 @@ static void rtl_hw_start_8125d(struct rtl8169_private *tp)
>  
>  static void rtl_hw_start_8126a(struct rtl8169_private *tp)
>  {
> +	rtl_disable_zrxdc_timeout(tp);
>  	rtl_set_def_aspm_entry_latency(tp);
>  	rtl_hw_start_8125_common(tp);
>  }