| 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
| ||
|
Message-ID: <5704FE2C.5000902@arm.com>
Date: Wed, 6 Apr 2016 13:16:44 +0100
From: Julien Grall <julien.grall@....com>
To: Shannon Zhao <shannon.zhao@...aro.org>,
linux-arm-kernel@...ts.infradead.org, sstabellini@...nel.org
Cc: david.vrabel@...rix.com, devicetree@...r.kernel.org,
linux-efi@...r.kernel.org, linux-kernel@...r.kernel.org,
catalin.marinas@....com, will.deacon@....com,
peter.huangpeng@...wei.com, xen-devel@...ts.xen.org,
zhaoshenglong@...wei.com
Subject: Re: [PATCH v10 06/17] Xen: ARM: Add support for mapping platform
device mmio
Hi Shannon,
Sorry to come late in the review process.
On 01/04/2016 16:49, Shannon Zhao wrote:
> Add a bus_notifier for platform bus device in order to map the device
> mmio regions when DOM0 booting with ACPI.
>
> Signed-off-by: Shannon Zhao <shannon.zhao@...aro.org>
> Acked-by: Stefano Stabellini <stefano.stabellini@...citrix.com>
> ---
> drivers/xen/Makefile | 1 +
> drivers/xen/arm-device.c | 141 +++++++++++++++++++++++++++++++++++++++++++++++
> 2 files changed, 142 insertions(+)
> create mode 100644 drivers/xen/arm-device.c
>
> diff --git a/drivers/xen/Makefile b/drivers/xen/Makefile
> index 9b7a35c..415f286 100644
> --- a/drivers/xen/Makefile
> +++ b/drivers/xen/Makefile
> @@ -9,6 +9,7 @@ CFLAGS_features.o := $(nostackp)
>
> CFLAGS_efi.o += -fshort-wchar
>
> +dom0-$(CONFIG_ARM64) += arm-device.o
> dom0-$(CONFIG_PCI) += pci.o
> dom0-$(CONFIG_USB_SUPPORT) += dbgp.o
> dom0-$(CONFIG_XEN_ACPI) += acpi.o $(xen-pad-y)
> diff --git a/drivers/xen/arm-device.c b/drivers/xen/arm-device.c
> new file mode 100644
> index 0000000..76e26e5
> --- /dev/null
> +++ b/drivers/xen/arm-device.c
> @@ -0,0 +1,141 @@
> +/*
> + * Copyright (c) 2015, Linaro Limited, Shannon Zhao
> + *
> + * This program is free software; you can redistribute it and/or modify it
> + * under the terms and conditions of the GNU General Public License,
> + * version 2, as published by the Free Software Foundation.
> + *
> + * This program is distributed in the hope it will be useful, but WITHOUT
> + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
> + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for
> + * more details.
> + *
> + * You should have received a copy of the GNU General Public License
> + * along with this program. If not, see <http://www.gnu.org/licenses/>.
> + */
> +
> +#include <linux/platform_device.h>
> +#include <linux/acpi.h>
> +#include <xen/xen.h>
> +#include <xen/page.h>
> +#include <xen/interface/memory.h>
> +#include <asm/xen/hypervisor.h>
> +#include <asm/xen/hypercall.h>
> +
> +static int xen_unmap_device_mmio(struct resource *resource, unsigned int count)
NITs:
There is multiple resource so s/resource/resources/.
Also the variable could be const.
> +{
> + unsigned int i, j, nr;
> + int rc = 0;
> + struct resource *r;
> + struct xen_remove_from_physmap xrp;
> +
> + for (i = 0; i < count; i++) {
> + r = &resource[i];
> + nr = DIV_ROUND_UP(resource_size(r), XEN_PAGE_SIZE);
> + if ((resource_type(r) != IORESOURCE_MEM) || (nr == 0))
> + continue;
> +
> + for (j = 0; j < nr; j++) {
> + xrp.domid = DOMID_SELF;
> + xrp.gpfn = XEN_PFN_DOWN(r->start) + j;
> + rc = HYPERVISOR_memory_op(XENMEM_remove_from_physmap,
> + &xrp);
> + if (rc)
> + return rc;
> + }
> + }
> +
> + return rc;
> +}
> +
> +static int xen_map_device_mmio(struct resource *resource, unsigned int count)
Ditto
> +{
> + unsigned int i, j, nr;
> + int rc = 0;
> + struct resource *r;
> + xen_pfn_t *gpfns;
> + xen_ulong_t *idxs;
> + int *errs;
> + struct xen_add_to_physmap_range xatp;
> +
> + for (i = 0; i < count; i++) {
> + r = &resource[i];
> + nr = DIV_ROUND_UP(resource_size(r), XEN_PAGE_SIZE);
> + if ((resource_type(r) != IORESOURCE_MEM) || (nr == 0))
> + continue;
> +
> + gpfns = kzalloc(sizeof(xen_pfn_t) * nr, GFP_KERNEL);
> + idxs = kzalloc(sizeof(xen_ulong_t) * nr, GFP_KERNEL);
> + errs = kzalloc(sizeof(int) * nr, GFP_KERNEL);
> + if (!gpfns || !idxs || !errs) {
> + kfree(gpfns);
> + kfree(idxs);
> + kfree(errs);
> + return -ENOMEM;
> + }
> +
> + for (j = 0; j < nr; j++) {
I would add a comment to explain that the regions are always mapped 1:1
to DOM0, and this is fine because the memory map for DOM0 is the same as
the host (except for the RAM).
> + gpfns[j] = XEN_PFN_DOWN(r->start) + j;
> + idxs[j] = XEN_PFN_DOWN(r->start) + j;
> + }
> +
> + xatp.domid = DOMID_SELF;
> + xatp.size = nr;
> + xatp.space = XENMAPSPACE_dev_mmio;
> +
> + set_xen_guest_handle(xatp.gpfns, gpfns);
> + set_xen_guest_handle(xatp.idxs, idxs);
> + set_xen_guest_handle(xatp.errs, errs);
> +
> + rc = HYPERVISOR_memory_op(XENMEM_add_to_physmap_range, &xatp);
> + kfree(gpfns);
> + kfree(idxs);
> + kfree(errs);
> + if (rc)
> + return rc;
Shouldn't we redo the mapping if the hypercall fails?
> + }
> +
> + return rc;
> +}
> +
> +static int xen_platform_notifier(struct notifier_block *nb,
> + unsigned long action, void *data)
> +{
> + struct platform_device *pdev = to_platform_device(data);
> + int r = 0;
> +
> + if (pdev->num_resources == 0 || pdev->resource == NULL)
> + return NOTIFY_OK;
> +
> + switch (action) {
> + case BUS_NOTIFY_ADD_DEVICE:
> + r = xen_map_device_mmio(pdev->resource, pdev->num_resources);
> + break;
> + case BUS_NOTIFY_DEL_DEVICE:
> + r = xen_unmap_device_mmio(pdev->resource, pdev->num_resources);
> + break;
> + default:
> + return NOTIFY_DONE;
> + }
> + if (r)
> + dev_err(&pdev->dev, "Platform: Failed to %s device %s MMIO!\n",
> + action == BUS_NOTIFY_ADD_DEVICE ? "map" :
> + (action == BUS_NOTIFY_DEL_DEVICE ? "unmap" : "?"),
> + pdev->name);
> +
> + return NOTIFY_OK;
> +}
> +
> +static struct notifier_block platform_device_nb = {
> + .notifier_call = xen_platform_notifier,
> +};
> +
> +static int __init register_xen_platform_notifier(void)
> +{
> + if (!xen_initial_domain() || acpi_disabled)
> + return 0;
> +
> + return bus_register_notifier(&platform_bus_type, &platform_device_nb);
> +}
> +
> +arch_initcall(register_xen_platform_notifier);
>
Regards,
--
Julien Grall
Powered by blists - more mailing lists