[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <CAL_JsqKUjUN2c8XzCoL7xepb5xZHLQktqTSekYva6bGEZ5Sx2g@mail.gmail.com>
Date: Thu, 20 May 2021 09:00:59 -0500
From: Rob Herring <robh@...nel.org>
To: Wang Xingang <wangxingang5@...wei.com>
Cc: Will Deacon <will@...nel.org>, Joerg Roedel <joro@...tes.org>,
Lorenzo Pieralisi <lorenzo.pieralisi@....com>,
Frank Rowand <frowand.list@...il.com>,
Bjorn Helgaas <helgaas@...nel.org>,
Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
Linux IOMMU <iommu@...ts.linux-foundation.org>,
"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
PCI <linux-pci@...r.kernel.org>,
linux-arm-kernel <linux-arm-kernel@...ts.infradead.org>,
xieyingtai@...wei.com
Subject: Re: [PATCH v3] iommu/of: Fix pci_request_acs() before enumerating PCI devices
On Thu, May 20, 2021 at 2:28 AM Wang Xingang <wangxingang5@...wei.com> wrote:
>
> From: Xingang Wang <wangxingang5@...wei.com>
>
> When booting with devicetree, the pci_request_acs() is called after the
> enumeration and initialization of PCI devices, thus the ACS is not
> enabled. And ACS should be enabled when IOMMU is detected for the
> PCI host bridge, so add check for IOMMU before probe of PCI host and call
> pci_request_acs() to make sure ACS will be enabled when enumerating PCI
> devices.
>
> Fixes: 6bf6c24720d33 ("iommu/of: Request ACS from the PCI core when
> configuring IOMMU linkage")
> Signed-off-by: Xingang Wang <wangxingang5@...wei.com>
> ---
> drivers/iommu/of_iommu.c | 1 -
> drivers/pci/controller/pci-host-common.c | 17 +++++++++++++++++
> 2 files changed, 17 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/iommu/of_iommu.c b/drivers/iommu/of_iommu.c
> index a9d2df001149..54a14da242cc 100644
> --- a/drivers/iommu/of_iommu.c
> +++ b/drivers/iommu/of_iommu.c
> @@ -205,7 +205,6 @@ const struct iommu_ops *of_iommu_configure(struct device *dev,
> .np = master_np,
> };
>
> - pci_request_acs();
> err = pci_for_each_dma_alias(to_pci_dev(dev),
> of_pci_iommu_init, &info);
> } else {
> diff --git a/drivers/pci/controller/pci-host-common.c b/drivers/pci/controller/pci-host-common.c
> index d3924a44db02..5904ad0bd9ae 100644
> --- a/drivers/pci/controller/pci-host-common.c
> +++ b/drivers/pci/controller/pci-host-common.c
This file is generally only for ECAM compliant hosts. Are those the
only hosts we need to support this? From the looks of dts files with
iommu-map, that would be dropping support in lots of cases.
Perhaps in devm_of_pci_bridge_init() or one of the functions it calls
is the better place.
> @@ -49,6 +49,21 @@ static struct pci_config_window *gen_pci_init(struct device *dev,
> return cfg;
> }
>
> +static void pci_host_enable_acs(struct pci_host_bridge *bridge)
> +{
> + struct device_node *np = bridge->dev.parent->of_node;
> + static bool acs_enabled;
> +
> + if (!np || acs_enabled)
> + return;
> +
> + /* Detect IOMMU and make sure ACS will be enabled */
> + if (of_property_read_bool(np, "iommu-map")) {
> + acs_enabled = true;
> + pci_request_acs();
Given this function just sets a variable, I don't think you need the
static acs_enabled here.
> + }
> +}
> +
> int pci_host_common_probe(struct platform_device *pdev)
> {
> struct device *dev = &pdev->dev;
> @@ -81,6 +96,8 @@ int pci_host_common_probe(struct platform_device *pdev)
> bridge->ops = (struct pci_ops *)&ops->pci_ops;
> bridge->msi_domain = true;
>
> + pci_host_enable_acs(bridge);
> +
> return pci_host_probe(bridge);
> }
> EXPORT_SYMBOL_GPL(pci_host_common_probe);
> --
> 2.19.1
>
Powered by blists - more mailing lists