From mboxrd@z Thu Jan 1 00:00:00 1970 From: zhaoshenglong@huawei.com (Shannon Zhao) Date: Wed, 18 Nov 2015 14:03:54 +0800 Subject: [Xen-devel] [PATCH 06/13] Xen: ARM: Add support for mapping amba device mmio In-Reply-To: <20151117144040.GI3003@char.us.oracle.com> References: <1447754231-7772-1-git-send-email-shannon.zhao@linaro.org> <1447754231-7772-7-git-send-email-shannon.zhao@linaro.org> <20151117144040.GI3003@char.us.oracle.com> Message-ID: <564C14CA.9080601@huawei.com> To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org On 2015/11/17 22:40, Konrad Rzeszutek Wilk wrote: > On Tue, Nov 17, 2015 at 05:57:04PM +0800, shannon.zhao at linaro.org wrote: >> From: Shannon Zhao >> >> Add a bus_notifier for AMBA bus device in order to map the device >> mmio regions when DOM0 booting with ACPI. >> >> Signed-off-by: Shannon Zhao >> --- >> drivers/xen/Makefile | 1 + >> drivers/xen/amba.c | 99 ++++++++++++++++++++++++++++++++++++++++++++++++++++ >> 2 files changed, 100 insertions(+) >> create mode 100644 drivers/xen/amba.c >> >> diff --git a/drivers/xen/Makefile b/drivers/xen/Makefile >> index 2f867e7..139bd0b 100644 >> --- a/drivers/xen/Makefile >> +++ b/drivers/xen/Makefile >> @@ -12,6 +12,7 @@ CFLAGS_features.o := $(nostackp) >> CFLAGS_efi.o += -fshort-wchar >> >> dom0-$(CONFIG_ARM64) += platform.o >> +dom0-$(CONFIG_ARM_AMBA) += amba.o >> dom0-$(CONFIG_PCI) += pci.o >> dom0-$(CONFIG_USB_SUPPORT) += dbgp.o >> dom0-$(CONFIG_XEN_ACPI) += acpi.o $(xen-pad-y) >> diff --git a/drivers/xen/amba.c b/drivers/xen/amba.c >> new file mode 100644 >> index 0000000..e491c8e >> --- /dev/null >> +++ b/drivers/xen/amba.c >> @@ -0,0 +1,99 @@ >> +/* >> + * Copyright (c) 2015, Linaro Limited. >> + * >> + * 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, write to the Free Software Foundation, Inc., 59 Temple >> + * Place - Suite 330, Boston, MA 02111-1307 USA. >> + * >> + * Author: Shannon Zhao >> + */ >> + >> +#include >> +#include >> +#include >> +#include >> +#include >> +#include >> + >> +static int xen_map_amba_device_mmio(struct amba_device *adev) >> +{ >> + int rc = 0; >> + struct resource *r = &adev->res; >> + >> + if (resource_type(r) == IORESOURCE_MEM) > > This is like the 'xen_map_platform_device_mmio' code. Why don't you call > that? Or at least make the 'xen_map_platform_device_mmio' internal parts be > visible to both of these functions. > Ok, will do. >> + { >> + int j; >> + int nr = DIV_ROUND_UP(resource_size(r), PAGE_SIZE); >> + xen_pfn_t *gpfns = kmalloc(sizeof(xen_pfn_t) * nr, GFP_KERNEL); >> + xen_ulong_t *idxs = kmalloc(sizeof(xen_ulong_t) * nr, >> + GFP_KERNEL); >> + int *errs = kmalloc(sizeof(int) * nr, GFP_KERNEL); >> + struct xen_add_to_physmap_range xatp; >> + >> + for (j = 0; j < nr; j++) { >> + gpfns[j] = (r->start >> PAGE_SHIFT) + j; >> + idxs[j] = (r->start >> PAGE_SHIFT) + j; >> + errs[j] = 0; >> + } >> + >> + 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 != 0) >> + return rc; >> + } >> + >> + return rc; >> +} >> + >> +static int xen_amba_notifier(struct notifier_block *nb, >> + unsigned long action, void *data) >> +{ >> + struct amba_device *adev = to_amba_device(data); >> + int r = 0; >> + >> + if (!acpi_disabled && (action == BUS_NOTIFY_ADD_DEVICE)) >> + r = xen_map_amba_device_mmio(adev); >> + > > What about the removal of a bus device? No need to handle that? > I have thought about removal before. I think there is little(or no) chance for AMBA and platform bus devices to be removed. It's not like the PCI devices which will be hot-unplug. Maybe I'm not right. If so, it's fine to add a case to handle the removal. Thanks, -- Shannon