linux-kernel - Re: [PATCH] PCI: qcom: configure the parf halt window size to 1GB

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]

Date:   Fri, 23 Jun 2023 12:17:22 +0200
From:   Konrad Dybcio <konrad.dybcio@...aro.org>
To:     Devi Priya <quic_devipriy@...cinc.com>, agross@...nel.org,
        andersson@...nel.org, lpieralisi@...nel.org, kw@...ux.com,
        robh@...nel.org, bhelgaas@...gle.com, mani@...nel.org,
        linux-arm-msm@...r.kernel.org, linux-pci@...r.kernel.org,
        linux-kernel@...r.kernel.org
Cc:     quic_srichara@...cinc.com, quic_sjaganat@...cinc.com,
        quic_kathirav@...cinc.com, quic_anusha@...cinc.com,
        quic_ipkumar@...cinc.com
Subject: Re: [PATCH] PCI: qcom: configure the parf halt window size to 1GB

On 23.06.2023 06:57, Devi Priya wrote:
> Configure the ADDR_BIT_INDEX of PARF_AXI_MSTR_WR_ADDR_HALT_V2 register with
> 0x1E to increase the halt window size to 1GB so that, when new inbound
> posted write transactions whose address crosses 1G address range, the
> controller would halt all the incoming writes until all the previous AXI
> responses are received.
> 
> Signed-off-by: Devi Priya <quic_devipriy@...cinc.com>
> ---
Has this been tested on anything except IPQ9574? Does it concern other
SoCs?


>  This patch depends on the below series which adds support for PCIe 
>  controllers in IPQ9574
>  https://lore.kernel.org/linux-arm-msm/20230519090219.15925-1-quic_devipriy@quicinc.com/
> 
>  drivers/pci/controller/dwc/pcie-qcom.c | 9 +++++++++
>  1 file changed, 9 insertions(+)
> 
> diff --git a/drivers/pci/controller/dwc/pcie-qcom.c b/drivers/pci/controller/dwc/pcie-qcom.c
> index c7579dfa5b1c..26c40e006120 100644
> --- a/drivers/pci/controller/dwc/pcie-qcom.c
> +++ b/drivers/pci/controller/dwc/pcie-qcom.c
> @@ -116,6 +116,8 @@
>  
>  /* PARF_AXI_MSTR_WR_ADDR_HALT register fields */
>  #define EN					BIT(31)
> +#define ADDR_BIT_INDEX				(BIT(0) | BIT(1) | BIT(2) | \
> +						BIT(3) | BIT(4) | BIT(5))
You surely should have the names of these bitfields, mind defining them?

>  
>  /* PARF_LTSSM register fields */
>  #define LTSSM_EN				BIT(8)
> @@ -154,6 +156,8 @@
>  
>  #define QCOM_PCIE_CRC8_POLYNOMIAL		(BIT(2) | BIT(1) | BIT(0))
>  
> +#define PARF_AXI_MSTR_WR_ADDR_HALT_WINDOW_SIZE	0x1e
> +
>  #define QCOM_PCIE_1_0_0_MAX_CLOCKS		4
>  struct qcom_pcie_resources_1_0_0 {
>  	struct clk_bulk_data clks[QCOM_PCIE_1_0_0_MAX_CLOCKS];
> @@ -1126,6 +1130,11 @@ static int qcom_pcie_post_init(struct qcom_pcie *pcie)
>  
>  	writel(0, pcie->parf + PARF_Q2A_FLUSH);
>  
> +	val = readl(pcie->parf + PARF_AXI_MSTR_WR_ADDR_HALT_V2);
> +	val &= ~ADDR_BIT_INDEX;
> +	writel(val | PARF_AXI_MSTR_WR_ADDR_HALT_WINDOW_SIZE, pcie->parf +
> +			PARF_AXI_MSTR_WR_ADDR_HALT_V2);
val |= ..
writel(val, pcie..)

would be more readable

Konrad
> +
>  	dw_pcie_dbi_ro_wr_en(pci);
>  	writel(PCIE_CAP_SLOT_VAL, pci->dbi_base + offset + PCI_EXP_SLTCAP);
>