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: <20200430204017.GA62947@bjorn-Precision-5520>
Date:   Thu, 30 Apr 2020 15:40:17 -0500
From:   Bjorn Helgaas <helgaas@...nel.org>
To:     Jim Quinlan <james.quinlan@...adcom.com>
Cc:     Nicolas Saenz Julienne <nsaenzjulienne@...e.de>,
        Lorenzo Pieralisi <lorenzo.pieralisi@....com>,
        Rob Herring <robh@...nel.org>,
        Bjorn Helgaas <bhelgaas@...gle.com>,
        Florian Fainelli <f.fainelli@...il.com>,
        "maintainer:BROADCOM BCM7XXX ARM ARCHITECTURE" 
        <bcm-kernel-feedback-list@...adcom.com>,
        "moderated list:BROADCOM BCM2711/BCM2835 ARM ARCHITECTURE" 
        <linux-rpi-kernel@...ts.infradead.org>,
        "moderated list:BROADCOM BCM2711/BCM2835 ARM ARCHITECTURE" 
        <linux-arm-kernel@...ts.infradead.org>,
        "open list:PCI NATIVE HOST BRIDGE AND ENDPOINT DRIVERS" 
        <linux-pci@...r.kernel.org>,
        open list <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH 5/5] PCI: brcmstb: disable L0s component of ASPM by
 default

On Thu, Apr 30, 2020 at 02:55:22PM -0400, Jim Quinlan wrote:
> From: Jim Quinlan <jquinlan@...adcom.com>
> 
> Some informal internal experiments has shown that the BrcmSTB ASPM L0s
> savings may introduce an undesirable noise signal on some customers'
> boards.  In addition, L0s was found lacking in realized power savings,
> especially relative to the L1 ASPM component.  This is BrcmSTB's
> experience and may not hold for others.  At any rate, we disable L0s
> savings by default unless the DT node has the 'brcm,aspm-en-l0s'
> property.

I assume this works by writing the PCIe Link Capabilities register,
which is read-only via the config space path used by the generic ASPM
code, so that code thinks the device doesn't support L0s at all.

Documentation/devicetree/bindings/pci/rockchip-pcie-host.txt includes
an "aspm-no-l0s" property.  It'd be nice if this could use the same
property.

> Signed-off-by: Jim Quinlan <jquinlan@...adcom.com>
> ---
>  drivers/pci/controller/pcie-brcmstb.c | 14 +++++++++++++-
>  1 file changed, 13 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/pci/controller/pcie-brcmstb.c b/drivers/pci/controller/pcie-brcmstb.c
> index 2bc913c0262c..bc1d514b19e4 100644
> --- a/drivers/pci/controller/pcie-brcmstb.c
> +++ b/drivers/pci/controller/pcie-brcmstb.c
> @@ -44,6 +44,9 @@
>  #define PCIE_RC_CFG_PRIV1_ID_VAL3			0x043c
>  #define  PCIE_RC_CFG_PRIV1_ID_VAL3_CLASS_CODE_MASK	0xffffff
>  
> +#define PCIE_RC_CFG_PRIV1_LINK_CAPABILITY			0x04dc
> +#define  PCIE_RC_CFG_PRIV1_LINK_CAPABILITY_ASPM_SUPPORT_MASK	0xc00
> +
>  #define PCIE_RC_DL_MDIO_ADDR				0x1100
>  #define PCIE_RC_DL_MDIO_WR_DATA				0x1104
>  #define PCIE_RC_DL_MDIO_RD_DATA				0x1108
> @@ -696,7 +699,7 @@ static int brcm_pcie_setup(struct brcm_pcie *pcie)
>  	int num_out_wins = 0;
>  	u16 nlw, cls, lnksta;
>  	int i, ret;
> -	u32 tmp;
> +	u32 tmp, aspm_support;
>  
>  	/* Reset the bridge */
>  	brcm_pcie_bridge_sw_init_set(pcie, 1);
> @@ -806,6 +809,15 @@ static int brcm_pcie_setup(struct brcm_pcie *pcie)
>  		num_out_wins++;
>  	}
>  
> +	/* Only support ASPM L1 unless L0s is explicitly desired */
> +	aspm_support = PCIE_LINK_STATE_L1;
> +	if (of_property_read_bool(pcie->np, "brcm,aspm-en-l0s"))
> +		aspm_support |= PCIE_LINK_STATE_L0S;
> +	tmp = readl(base + PCIE_RC_CFG_PRIV1_LINK_CAPABILITY);
> +	u32p_replace_bits(&tmp, aspm_support,
> +		PCIE_RC_CFG_PRIV1_LINK_CAPABILITY_ASPM_SUPPORT_MASK);
> +	writel(tmp, base + PCIE_RC_CFG_PRIV1_LINK_CAPABILITY);
> +
>  	/*
>  	 * For config space accesses on the RC, show the right class for
>  	 * a PCIe-PCIe bridge (the default setting is to be EP mode).
> -- 
> 2.17.1
> 

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ