summaryrefslogtreecommitdiffstats
path: root/drivers
diff options
context:
space:
mode:
authorLinus Torvalds <torvalds@linux-foundation.org>2012-07-24 13:14:03 -0700
committerLinus Torvalds <torvalds@linux-foundation.org>2012-07-24 13:14:03 -0700
commit62c4d9afa4bcf5315e2745a17a0228bf65b9ba40 (patch)
treea7b9d97283441ea5f0c738fa388e120c4c1491b6 /drivers
parent5fecc9d8f59e765c2a48379dd7c6f5cf88c7d75a (diff)
parentc3d93f880197953f86ab90d9da4744e926b38e33 (diff)
downloadop-kernel-dev-62c4d9afa4bcf5315e2745a17a0228bf65b9ba40.zip
op-kernel-dev-62c4d9afa4bcf5315e2745a17a0228bf65b9ba40.tar.gz
Merge tag 'stable/for-linus-3.6-rc0-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/konrad/xen
Pull Xen update from Konrad Rzeszutek Wilk: "Features: * Performance improvement to lower the amount of traps the hypervisor has to do 32-bit guests. Mainly for setting PTE entries and updating TLS descriptors. * MCE polling driver to collect hypervisor MCE buffer and present them to /dev/mcelog. * Physical CPU online/offline support. When an privileged guest is booted it is present with virtual CPUs, which might have an 1:1 to physical CPUs but usually don't. This provides mechanism to offline/online physical CPUs. Bug-fixes for: * Coverity found fixes in the console and ACPI processor driver. * PVonHVM kexec fixes along with some cleanups. * Pages that fall within E820 gaps and non-RAM regions (and had been released to hypervisor) would be populated back, but potentially in non-RAM regions." * tag 'stable/for-linus-3.6-rc0-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/konrad/xen: xen: populate correct number of pages when across mem boundary (v2) xen PVonHVM: move shared_info to MMIO before kexec xen: simplify init_hvm_pv_info xen: remove cast from HYPERVISOR_shared_info assignment xen: enable platform-pci only in a Xen guest xen/pv-on-hvm kexec: shutdown watches from old kernel xen/x86: avoid updating TLS descriptors if they haven't changed xen/x86: add desc_equal() to compare GDT descriptors xen/mm: zero PTEs for non-present MFNs in the initial page table xen/mm: do direct hypercall in xen_set_pte() if batching is unavailable xen/hvc: Fix up checks when the info is allocated. xen/acpi: Fix potential memory leak. xen/mce: add .poll method for mcelog device driver xen/mce: schedule a workqueue to avoid sleep in atomic context xen/pcpu: Xen physical cpus online/offline sys interface xen/mce: Register native mce handler as vMCE bounce back point x86, MCE, AMD: Adjust initcall sequence for xen xen/mce: Add mcelog support for Xen platform
Diffstat (limited to 'drivers')
-rw-r--r--drivers/tty/hvc/hvc_xen.c15
-rw-r--r--drivers/xen/Kconfig8
-rw-r--r--drivers/xen/Makefile2
-rw-r--r--drivers/xen/mcelog.c414
-rw-r--r--drivers/xen/pcpu.c371
-rw-r--r--drivers/xen/platform-pci.c18
-rw-r--r--drivers/xen/xen-acpi-processor.c9
-rw-r--r--drivers/xen/xenbus/xenbus_xs.c20
8 files changed, 845 insertions, 12 deletions
diff --git a/drivers/tty/hvc/hvc_xen.c b/drivers/tty/hvc/hvc_xen.c
index 944eaeb..1e456dc 100644
--- a/drivers/tty/hvc/hvc_xen.c
+++ b/drivers/tty/hvc/hvc_xen.c
@@ -209,11 +209,10 @@ static int xen_hvm_console_init(void)
info = kzalloc(sizeof(struct xencons_info), GFP_KERNEL | __GFP_ZERO);
if (!info)
return -ENOMEM;
- }
-
- /* already configured */
- if (info->intf != NULL)
+ } else if (info->intf != NULL) {
+ /* already configured */
return 0;
+ }
/*
* If the toolstack (or the hypervisor) hasn't set these values, the
* default value is 0. Even though mfn = 0 and evtchn = 0 are
@@ -259,12 +258,10 @@ static int xen_pv_console_init(void)
info = kzalloc(sizeof(struct xencons_info), GFP_KERNEL | __GFP_ZERO);
if (!info)
return -ENOMEM;
- }
-
- /* already configured */
- if (info->intf != NULL)
+ } else if (info->intf != NULL) {
+ /* already configured */
return 0;
-
+ }
info->evtchn = xen_start_info->console.domU.evtchn;
info->intf = mfn_to_virt(xen_start_info->console.domU.mfn);
info->vtermno = HVC_COOKIE;
diff --git a/drivers/xen/Kconfig b/drivers/xen/Kconfig
index 8d2501e..d4dffcd 100644
--- a/drivers/xen/Kconfig
+++ b/drivers/xen/Kconfig
@@ -196,4 +196,12 @@ config XEN_ACPI_PROCESSOR
called xen_acpi_processor If you do not know what to choose, select
M here. If the CPUFREQ drivers are built in, select Y here.
+config XEN_MCE_LOG
+ bool "Xen platform mcelog"
+ depends on XEN_DOM0 && X86_64 && X86_MCE
+ default n
+ help
+ Allow kernel fetching MCE error from Xen platform and
+ converting it into Linux mcelog format for mcelog tools
+
endmenu
diff --git a/drivers/xen/Makefile b/drivers/xen/Makefile
index fc34886..d80bea5 100644
--- a/drivers/xen/Makefile
+++ b/drivers/xen/Makefile
@@ -17,7 +17,9 @@ obj-$(CONFIG_XEN_SYS_HYPERVISOR) += sys-hypervisor.o
obj-$(CONFIG_XEN_PVHVM) += platform-pci.o
obj-$(CONFIG_XEN_TMEM) += tmem.o
obj-$(CONFIG_SWIOTLB_XEN) += swiotlb-xen.o
+obj-$(CONFIG_XEN_DOM0) += pcpu.o
obj-$(CONFIG_XEN_DOM0) += pci.o acpi.o
+obj-$(CONFIG_XEN_MCE_LOG) += mcelog.o
obj-$(CONFIG_XEN_PCIDEV_BACKEND) += xen-pciback/
obj-$(CONFIG_XEN_PRIVCMD) += xen-privcmd.o
obj-$(CONFIG_XEN_ACPI_PROCESSOR) += xen-acpi-processor.o
diff --git a/drivers/xen/mcelog.c b/drivers/xen/mcelog.c
new file mode 100644
index 0000000..8feee08
--- /dev/null
+++ b/drivers/xen/mcelog.c
@@ -0,0 +1,414 @@
+/******************************************************************************
+ * mcelog.c
+ * Driver for receiving and transferring machine check error infomation
+ *
+ * Copyright (c) 2012 Intel Corporation
+ * Author: Liu, Jinsong <jinsong.liu@intel.com>
+ * Author: Jiang, Yunhong <yunhong.jiang@intel.com>
+ * Author: Ke, Liping <liping.ke@intel.com>
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License version 2
+ * as published by the Free Software Foundation; or, when distributed
+ * separately from the Linux kernel or incorporated into other
+ * software packages, subject to the following license:
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this source file (the "Software"), to deal in the Software without
+ * restriction, including without limitation the rights to use, copy, modify,
+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
+ * and to permit persons to whom the Software is furnished to do so, subject to
+ * the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
+ * IN THE SOFTWARE.
+ */
+
+#include <linux/init.h>
+#include <linux/types.h>
+#include <linux/kernel.h>
+#include <linux/slab.h>
+#include <linux/fs.h>
+#include <linux/device.h>
+#include <linux/miscdevice.h>
+#include <linux/uaccess.h>
+#include <linux/capability.h>
+#include <linux/poll.h>
+#include <linux/sched.h>
+
+#include <xen/interface/xen.h>
+#include <xen/events.h>
+#include <xen/interface/vcpu.h>
+#include <xen/xen.h>
+#include <asm/xen/hypercall.h>
+#include <asm/xen/hypervisor.h>
+
+#define XEN_MCELOG "xen_mcelog: "
+
+static struct mc_info g_mi;
+static struct mcinfo_logical_cpu *g_physinfo;
+static uint32_t ncpus;
+
+static DEFINE_MUTEX(mcelog_lock);
+
+static struct xen_mce_log xen_mcelog = {
+ .signature = XEN_MCE_LOG_SIGNATURE,
+ .len = XEN_MCE_LOG_LEN,
+ .recordlen = sizeof(struct xen_mce),
+};
+
+static DEFINE_SPINLOCK(xen_mce_chrdev_state_lock);
+static int xen_mce_chrdev_open_count; /* #times opened */
+static int xen_mce_chrdev_open_exclu; /* already open exclusive? */
+
+static DECLARE_WAIT_QUEUE_HEAD(xen_mce_chrdev_wait);
+
+static int xen_mce_chrdev_open(struct inode *inode, struct file *file)
+{
+ spin_lock(&xen_mce_chrdev_state_lock);
+
+ if (xen_mce_chrdev_open_exclu ||
+ (xen_mce_chrdev_open_count && (file->f_flags & O_EXCL))) {
+ spin_unlock(&xen_mce_chrdev_state_lock);
+
+ return -EBUSY;
+ }
+
+ if (file->f_flags & O_EXCL)
+ xen_mce_chrdev_open_exclu = 1;
+ xen_mce_chrdev_open_count++;
+
+ spin_unlock(&xen_mce_chrdev_state_lock);
+
+ return nonseekable_open(inode, file);
+}
+
+static int xen_mce_chrdev_release(struct inode *inode, struct file *file)
+{
+ spin_lock(&xen_mce_chrdev_state_lock);
+
+ xen_mce_chrdev_open_count--;
+ xen_mce_chrdev_open_exclu = 0;
+
+ spin_unlock(&xen_mce_chrdev_state_lock);
+
+ return 0;
+}
+
+static ssize_t xen_mce_chrdev_read(struct file *filp, char __user *ubuf,
+ size_t usize, loff_t *off)
+{
+ char __user *buf = ubuf;
+ unsigned num;
+ int i, err;
+
+ mutex_lock(&mcelog_lock);
+
+ num = xen_mcelog.next;
+
+ /* Only supports full reads right now */
+ err = -EINVAL;
+ if (*off != 0 || usize < XEN_MCE_LOG_LEN*sizeof(struct xen_mce))
+ goto out;
+
+ err = 0;
+ for (i = 0; i < num; i++) {
+ struct xen_mce *m = &xen_mcelog.entry[i];
+
+ err |= copy_to_user(buf, m, sizeof(*m));
+ buf += sizeof(*m);
+ }
+
+ memset(xen_mcelog.entry, 0, num * sizeof(struct xen_mce));
+ xen_mcelog.next = 0;
+
+ if (err)
+ err = -EFAULT;
+
+out:
+ mutex_unlock(&mcelog_lock);
+
+ return err ? err : buf - ubuf;
+}
+
+static unsigned int xen_mce_chrdev_poll(struct file *file, poll_table *wait)
+{
+ poll_wait(file, &xen_mce_chrdev_wait, wait);
+
+ if (xen_mcelog.next)
+ return POLLIN | POLLRDNORM;
+
+ return 0;
+}
+
+static long xen_mce_chrdev_ioctl(struct file *f, unsigned int cmd,
+ unsigned long arg)
+{
+ int __user *p = (int __user *)arg;
+
+ if (!capable(CAP_SYS_ADMIN))
+ return -EPERM;
+
+ switch (cmd) {
+ case MCE_GET_RECORD_LEN:
+ return put_user(sizeof(struct xen_mce), p);
+ case MCE_GET_LOG_LEN:
+ return put_user(XEN_MCE_LOG_LEN, p);
+ case MCE_GETCLEAR_FLAGS: {
+ unsigned flags;
+
+ do {
+ flags = xen_mcelog.flags;
+ } while (cmpxchg(&xen_mcelog.flags, flags, 0) != flags);
+
+ return put_user(flags, p);
+ }
+ default:
+ return -ENOTTY;
+ }
+}
+
+static const struct file_operations xen_mce_chrdev_ops = {
+ .open = xen_mce_chrdev_open,
+ .release = xen_mce_chrdev_release,
+ .read = xen_mce_chrdev_read,
+ .poll = xen_mce_chrdev_poll,
+ .unlocked_ioctl = xen_mce_chrdev_ioctl,
+ .llseek = no_llseek,
+};
+
+static struct miscdevice xen_mce_chrdev_device = {
+ MISC_MCELOG_MINOR,
+ "mcelog",
+ &xen_mce_chrdev_ops,
+};
+
+/*
+ * Caller should hold the mcelog_lock
+ */
+static void xen_mce_log(struct xen_mce *mce)
+{
+ unsigned entry;
+
+ entry = xen_mcelog.next;
+
+ /*
+ * When the buffer fills up discard new entries.
+ * Assume that the earlier errors are the more
+ * interesting ones:
+ */
+ if (entry >= XEN_MCE_LOG_LEN) {
+ set_bit(XEN_MCE_OVERFLOW,
+ (unsigned long *)&xen_mcelog.flags);
+ return;
+ }
+
+ memcpy(xen_mcelog.entry + entry, mce, sizeof(struct xen_mce));
+
+ xen_mcelog.next++;
+}
+
+static int convert_log(struct mc_info *mi)
+{
+ struct mcinfo_common *mic;
+ struct mcinfo_global *mc_global;
+ struct mcinfo_bank *mc_bank;
+ struct xen_mce m;
+ uint32_t i;
+
+ mic = NULL;
+ x86_mcinfo_lookup(&mic, mi, MC_TYPE_GLOBAL);
+ if (unlikely(!mic)) {
+ pr_warning(XEN_MCELOG "Failed to find global error info\n");
+ return -ENODEV;
+ }
+
+ memset(&m, 0, sizeof(struct xen_mce));
+
+ mc_global = (struct mcinfo_global *)mic;
+ m.mcgstatus = mc_global->mc_gstatus;
+ m.apicid = mc_global->mc_apicid;
+
+ for (i = 0; i < ncpus; i++)
+ if (g_physinfo[i].mc_apicid == m.apicid)
+ break;
+ if (unlikely(i == ncpus)) {
+ pr_warning(XEN_MCELOG "Failed to match cpu with apicid %d\n",
+ m.apicid);
+ return -ENODEV;
+ }
+
+ m.socketid = g_physinfo[i].mc_chipid;
+ m.cpu = m.extcpu = g_physinfo[i].mc_cpunr;
+ m.cpuvendor = (__u8)g_physinfo[i].mc_vendor;
+ m.mcgcap = g_physinfo[i].mc_msrvalues[__MC_MSR_MCGCAP].value;
+
+ mic = NULL;
+ x86_mcinfo_lookup(&mic, mi, MC_TYPE_BANK);
+ if (unlikely(!mic)) {
+ pr_warning(XEN_MCELOG "Fail to find bank error info\n");
+ return -ENODEV;
+ }
+
+ do {
+ if ((!mic) || (mic->size == 0) ||
+ (mic->type != MC_TYPE_GLOBAL &&
+ mic->type != MC_TYPE_BANK &&
+ mic->type != MC_TYPE_EXTENDED &&
+ mic->type != MC_TYPE_RECOVERY))
+ break;
+
+ if (mic->type == MC_TYPE_BANK) {
+ mc_bank = (struct mcinfo_bank *)mic;
+ m.misc = mc_bank->mc_misc;
+ m.status = mc_bank->mc_status;
+ m.addr = mc_bank->mc_addr;
+ m.tsc = mc_bank->mc_tsc;
+ m.bank = mc_bank->mc_bank;
+ m.finished = 1;
+ /*log this record*/
+ xen_mce_log(&m);
+ }
+ mic = x86_mcinfo_next(mic);
+ } while (1);
+
+ return 0;
+}
+
+static int mc_queue_handle(uint32_t flags)
+{
+ struct xen_mc mc_op;
+ int ret = 0;
+
+ mc_op.cmd = XEN_MC_fetch;
+ mc_op.interface_version = XEN_MCA_INTERFACE_VERSION;
+ set_xen_guest_handle(mc_op.u.mc_fetch.data, &g_mi);
+ do {
+ mc_op.u.mc_fetch.flags = flags;
+ ret = HYPERVISOR_mca(&mc_op);
+ if (ret) {
+ pr_err(XEN_MCELOG "Failed to fetch %s error log\n",
+ (flags == XEN_MC_URGENT) ?
+ "urgnet" : "nonurgent");
+ break;
+ }
+
+ if (mc_op.u.mc_fetch.flags & XEN_MC_NODATA ||
+ mc_op.u.mc_fetch.flags & XEN_MC_FETCHFAILED)
+ break;
+ else {
+ ret = convert_log(&g_mi);
+ if (ret)
+ pr_warning(XEN_MCELOG
+ "Failed to convert this error log, "
+ "continue acking it anyway\n");
+
+ mc_op.u.mc_fetch.flags = flags | XEN_MC_ACK;
+ ret = HYPERVISOR_mca(&mc_op);
+ if (ret) {
+ pr_err(XEN_MCELOG
+ "Failed to ack previous error log\n");
+ break;
+ }
+ }
+ } while (1);
+
+ return ret;
+}
+
+/* virq handler for machine check error info*/
+static void xen_mce_work_fn(struct work_struct *work)
+{
+ int err;
+
+ mutex_lock(&mcelog_lock);
+
+ /* urgent mc_info */
+ err = mc_queue_handle(XEN_MC_URGENT);
+ if (err)
+ pr_err(XEN_MCELOG
+ "Failed to handle urgent mc_info queue, "
+ "continue handling nonurgent mc_info queue anyway.\n");
+
+ /* nonurgent mc_info */
+ err = mc_queue_handle(XEN_MC_NONURGENT);
+ if (err)
+ pr_err(XEN_MCELOG
+ "Failed to handle nonurgent mc_info queue.\n");
+
+ /* wake processes polling /dev/mcelog */
+ wake_up_interruptible(&xen_mce_chrdev_wait);
+
+ mutex_unlock(&mcelog_lock);
+}
+static DECLARE_WORK(xen_mce_work, xen_mce_work_fn);
+
+static irqreturn_t xen_mce_interrupt(int irq, void *dev_id)
+{
+ schedule_work(&xen_mce_work);
+ return IRQ_HANDLED;
+}
+
+static int bind_virq_for_mce(void)
+{
+ int ret;
+ struct xen_mc mc_op;
+
+ memset(&mc_op, 0, sizeof(struct xen_mc));
+
+ /* Fetch physical CPU Numbers */
+ mc_op.cmd = XEN_MC_physcpuinfo;
+ mc_op.interface_version = XEN_MCA_INTERFACE_VERSION;
+ set_xen_guest_handle(mc_op.u.mc_physcpuinfo.info, g_physinfo);
+ ret = HYPERVISOR_mca(&mc_op);
+ if (ret) {
+ pr_err(XEN_MCELOG "Failed to get CPU numbers\n");
+ return ret;
+ }
+
+ /* Fetch each CPU Physical Info for later reference*/
+ ncpus = mc_op.u.mc_physcpuinfo.ncpus;
+ g_physinfo = kcalloc(ncpus, sizeof(struct mcinfo_logical_cpu),
+ GFP_KERNEL);
+ if (!g_physinfo)
+ return -ENOMEM;
+ set_xen_guest_handle(mc_op.u.mc_physcpuinfo.info, g_physinfo);
+ ret = HYPERVISOR_mca(&mc_op);
+ if (ret) {
+ pr_err(XEN_MCELOG "Failed to get CPU info\n");
+ kfree(g_physinfo);
+ return ret;
+ }
+
+ ret = bind_virq_to_irqhandler(VIRQ_MCA, 0,
+ xen_mce_interrupt, 0, "mce", NULL);
+ if (ret < 0) {
+ pr_err(XEN_MCELOG "Failed to bind virq\n");
+ kfree(g_physinfo);
+ return ret;
+ }
+
+ return 0;
+}
+
+static int __init xen_late_init_mcelog(void)
+{
+ /* Only DOM0 is responsible for MCE logging */
+ if (xen_initial_domain()) {
+ /* register character device /dev/mcelog for xen mcelog */
+ if (misc_register(&xen_mce_chrdev_device))
+ return -ENODEV;
+ return bind_virq_for_mce();
+ }
+
+ return -ENODEV;
+}
+device_initcall(xen_late_init_mcelog);
diff --git a/drivers/xen/pcpu.c b/drivers/xen/pcpu.c
new file mode 100644
index 0000000..067fcfa
--- /dev/null
+++ b/drivers/xen/pcpu.c
@@ -0,0 +1,371 @@
+/******************************************************************************
+ * pcpu.c
+ * Management physical cpu in dom0, get pcpu info and provide sys interface
+ *
+ * Copyright (c) 2012 Intel Corporation
+ * Author: Liu, Jinsong <jinsong.liu@intel.com>
+ * Author: Jiang, Yunhong <yunhong.jiang@intel.com>
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License version 2
+ * as published by the Free Software Foundation; or, when distributed
+ * separately from the Linux kernel or incorporated into other
+ * software packages, subject to the following license:
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this source file (the "Software"), to deal in the Software without
+ * restriction, including without limitation the rights to use, copy, modify,
+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
+ * and to permit persons to whom the Software is furnished to do so, subject to
+ * the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
+ * IN THE SOFTWARE.
+ */
+
+#include <linux/interrupt.h>
+#include <linux/spinlock.h>
+#include <linux/cpu.h>
+#include <linux/stat.h>
+#include <linux/capability.h>
+
+#include <xen/xen.h>
+#include <xen/xenbus.h>
+#include <xen/events.h>
+#include <xen/interface/platform.h>
+#include <asm/xen/hypervisor.h>
+#include <asm/xen/hypercall.h>
+
+#define XEN_PCPU "xen_cpu: "
+
+/*
+ * @cpu_id: Xen physical cpu logic number
+ * @flags: Xen physical cpu status flag
+ * - XEN_PCPU_FLAGS_ONLINE: cpu is online
+ * - XEN_PCPU_FLAGS_INVALID: cpu is not present
+ */
+struct pcpu {
+ struct list_head list;
+ struct device dev;
+ uint32_t cpu_id;
+ uint32_t flags;
+};
+
+static struct bus_type xen_pcpu_subsys = {
+ .name = "xen_cpu",
+ .dev_name = "xen_cpu",
+};
+
+static DEFINE_MUTEX(xen_pcpu_lock);
+
+static LIST_HEAD(xen_pcpus);
+
+static int xen_pcpu_down(uint32_t cpu_id)
+{
+ struct xen_platform_op op = {
+ .cmd = XENPF_cpu_offline,
+ .interface_version = XENPF_INTERFACE_VERSION,
+ .u.cpu_ol.cpuid = cpu_id,
+ };
+
+ return HYPERVISOR_dom0_op(&op);
+}
+
+static int xen_pcpu_up(uint32_t cpu_id)
+{
+ struct xen_platform_op op = {
+ .cmd = XENPF_cpu_online,
+ .interface_version = XENPF_INTERFACE_VERSION,
+ .u.cpu_ol.cpuid = cpu_id,
+ };
+
+ return HYPERVISOR_dom0_op(&op);
+}
+
+static ssize_t show_online(struct device *dev,
+ struct device_attribute *attr,
+ char *buf)
+{
+ struct pcpu *cpu = container_of(dev, struct pcpu, dev);
+
+ return sprintf(buf, "%u\n", !!(cpu->flags & XEN_PCPU_FLAGS_ONLINE));
+}
+
+static ssize_t __ref store_online(struct device *dev,
+ struct device_attribute *attr,
+ const char *buf, size_t count)
+{
+ struct pcpu *pcpu = container_of(dev, struct pcpu, dev);
+ unsigned long long val;
+ ssize_t ret;
+
+ if (!capable(CAP_SYS_ADMIN))
+ return -EPERM;
+
+ if (kstrtoull(buf, 0, &val) < 0)
+ return -EINVAL;
+
+ switch (val) {
+ case 0:
+ ret = xen_pcpu_down(pcpu->cpu_id);
+ break;
+ case 1:
+ ret = xen_pcpu_up(pcpu->cpu_id);
+ break;
+ default:
+ ret = -EINVAL;
+ }
+
+ if (ret >= 0)
+ ret = count;
+ return ret;
+}
+static DEVICE_ATTR(online, S_IRUGO | S_IWUSR, show_online, store_online);
+
+static bool xen_pcpu_online(uint32_t flags)
+{
+ return !!(flags & XEN_PCPU_FLAGS_ONLINE);
+}
+
+static void pcpu_online_status(struct xenpf_pcpuinfo *info,
+ struct pcpu *pcpu)
+{
+ if (xen_pcpu_online(info->flags) &&
+ !xen_pcpu_online(pcpu->flags)) {
+ /* the pcpu is onlined */
+ pcpu->flags |= XEN_PCPU_FLAGS_ONLINE;
+ kobject_uevent(&pcpu->dev.kobj, KOBJ_ONLINE);
+ } else if (!xen_pcpu_online(info->flags) &&
+ xen_pcpu_online(pcpu->flags)) {
+ /* The pcpu is offlined */
+ pcpu->flags &= ~XEN_PCPU_FLAGS_ONLINE;
+ kobject_uevent(&pcpu->dev.kobj, KOBJ_OFFLINE);
+ }
+}
+
+static struct pcpu *get_pcpu(uint32_t cpu_id)
+{
+ struct pcpu *pcpu;
+
+ list_for_each_entry(pcpu, &xen_pcpus, list) {
+ if (pcpu->cpu_id == cpu_id)
+ return pcpu;
+ }
+
+ return NULL;
+}
+
+static void pcpu_release(struct device *dev)
+{
+ struct pcpu *pcpu = container_of(dev, struct pcpu, dev);
+
+ list_del(&pcpu->list);
+ kfree(pcpu);
+}
+
+static void unregister_and_remove_pcpu(struct pcpu *pcpu)
+{
+ struct device *dev;
+
+ if (!pcpu)
+ return;
+
+ dev = &pcpu->dev;
+ if (dev->id)
+ device_remove_file(dev, &dev_attr_online);
+
+ /* pcpu remove would be implicitly done */
+ device_unregister(dev);
+}
+
+static int register_pcpu(struct pcpu *pcpu)
+{
+ struct device *dev;
+ int err = -EINVAL;
+
+ if (!pcpu)
+ return err;
+
+ dev = &pcpu->dev;
+ dev->bus = &xen_pcpu_subsys;
+ dev->id = pcpu->cpu_id;
+ dev->release = pcpu_release;
+
+ err = device_register(dev);
+ if (err) {
+ pcpu_release(dev);
+ return err;
+ }
+
+ /*
+ * Xen never offline cpu0 due to several restrictions
+ * and assumptions. This basically doesn't add a sys control
+ * to user, one cannot attempt to offline BSP.
+ */
+ if (dev->id) {
+ err = device_create_file(dev, &dev_attr_online);
+ if (err) {
+ device_unregister(dev);
+ return err;
+ }
+ }
+
+ return 0;
+}
+
+static struct pcpu *create_and_register_pcpu(struct xenpf_pcpuinfo *info)
+{
+ struct pcpu *pcpu;
+ int err;
+
+ if (info->flags & XEN_PCPU_FLAGS_INVALID)
+ return ERR_PTR(-ENODEV);
+
+ pcpu = kzalloc(sizeof(struct pcpu), GFP_KERNEL);
+ if (!pcpu)
+ return ERR_PTR(-ENOMEM);
+
+ INIT_LIST_HEAD(&pcpu->list);
+ pcpu->cpu_id = info->xen_cpuid;
+ pcpu->flags = info->flags;
+
+ /* Need hold on xen_pcpu_lock before pcpu list manipulations */
+ list_add_tail(&pcpu->list, &xen_pcpus);
+
+ err = register_pcpu(pcpu);
+ if (err) {
+ pr_warning(XEN_PCPU "Failed to register pcpu%u\n",
+ info->xen_cpuid);
+ return ERR_PTR(-ENOENT);
+ }
+
+ return pcpu;
+}
+
+/*
+ * Caller should hold the xen_pcpu_lock
+ */
+static int sync_pcpu(uint32_t cpu, uint32_t *max_cpu)
+{
+ int ret;
+ struct pcpu *pcpu = NULL;
+ struct xenpf_pcpuinfo *info;
+ struct xen_platform_op op = {
+ .cmd = XENPF_get_cpuinfo,
+ .interface_version = XENPF_INTERFACE_VERSION,
+ .u.pcpu_info.xen_cpuid = cpu,
+ };
+
+ ret = HYPERVISOR_dom0_op(&op);
+ if (ret)
+ return ret;
+
+ info = &op.u.pcpu_info;
+ if (max_cpu)
+ *max_cpu = info->max_present;
+
+ pcpu = get_pcpu(cpu);
+
+ /*
+ * Only those at cpu present map has its sys interface.
+ */
+ if (info->flags & XEN_PCPU_FLAGS_INVALID) {
+ if (pcpu)
+ unregister_and_remove_pcpu(pcpu);
+ return 0;
+ }
+
+ if (!pcpu) {
+ pcpu = create_and_register_pcpu(info);
+ if (IS_ERR_OR_NULL(pcpu))
+ return -ENODEV;
+ } else
+ pcpu_online_status(info, pcpu);
+
+ return 0;
+}
+
+/*
+ * Sync dom0's pcpu information with xen hypervisor's
+ */
+static int xen_sync_pcpus(void)
+{
+ /*
+ * Boot cpu always have cpu_id 0 in xen
+ */
+ uint32_t cpu = 0, max_cpu = 0;
+ int err = 0;
+ struct pcpu *pcpu, *tmp;
+
+ mutex_lock(&xen_pcpu_lock);
+
+ while (!err && (cpu <= max_cpu)) {
+ err = sync_pcpu(cpu, &max_cpu);
+ cpu++;
+ }
+
+ if (err)
+ list_for_each_entry_safe(pcpu, tmp, &xen_pcpus, list)
+ unregister_and_remove_pcpu(pcpu);
+
+ mutex_unlock(&xen_pcpu_lock);
+
+ return err;
+}
+
+static void xen_pcpu_work_fn(struct work_struct *work)
+{
+ xen_sync_pcpus();
+}
+static DECLARE_WORK(xen_pcpu_work, xen_pcpu_work_fn);
+
+static irqreturn_t xen_pcpu_interrupt(int irq, void *dev_id)
+{
+ schedule_work(&xen_pcpu_work);
+ return IRQ_HANDLED;
+}
+
+static int __init xen_pcpu_init(void)
+{
+ int irq, ret;
+
+ if (!xen_initial_domain())
+ return -ENODEV;
+
+ irq = bind_virq_to_irqhandler(VIRQ_PCPU_STATE, 0,
+ xen_pcpu_interrupt, 0,
+ "xen-pcpu", NULL);
+ if (irq < 0) {
+ pr_warning(XEN_PCPU "Failed to bind pcpu virq\n");
+ return irq;
+ }
+
+ ret = subsys_system_register(&xen_pcpu_subsys, NULL);
+ if (ret) {
+ pr_warning(XEN_PCPU "Failed to register pcpu subsys\n");
+ goto err1;
+ }
+
+ ret = xen_sync_pcpus();
+ if (ret) {
+ pr_warning(XEN_PCPU "Failed to sync pcpu info\n");
+ goto err2;
+ }
+
+ return 0;
+
+err2:
+ bus_unregister(&xen_pcpu_subsys);
+err1:
+ unbind_from_irqhandler(irq, NULL);
+ return ret;
+}
+arch_initcall(xen_pcpu_init);
diff --git a/drivers/xen/platform-pci.c b/drivers/xen/platform-pci.c
index 2389e58..d4c50d6 100644
--- a/drivers/xen/platform-pci.c
+++ b/drivers/xen/platform-pci.c
@@ -101,6 +101,19 @@ static int platform_pci_resume(struct pci_dev *pdev)
return 0;
}
+static void __devinit prepare_shared_info(void)
+{
+#ifdef CONFIG_KEXEC
+ unsigned long addr;
+ struct shared_info *hvm_shared_info;
+
+ addr = alloc_xen_mmio(PAGE_SIZE);
+ hvm_shared_info = ioremap(addr, PAGE_SIZE);
+ memset(hvm_shared_info, 0, PAGE_SIZE);
+ xen_hvm_prepare_kexec(hvm_shared_info, addr >> PAGE_SHIFT);
+#endif
+}
+
static int __devinit platform_pci_init(struct pci_dev *pdev,
const struct pci_device_id *ent)
{
@@ -109,6 +122,9 @@ static int __devinit platform_pci_init(struct pci_dev *pdev,
long mmio_addr, mmio_len;
unsigned int max_nr_gframes;
+ if (!xen_domain())
+ return -ENODEV;
+
i = pci_enable_device(pdev);
if (i)
return i;
@@ -135,6 +151,8 @@ static int __devinit platform_pci_init(struct pci_dev *pdev,
platform_mmio = mmio_addr;
platform_mmiolen = mmio_len;
+ prepare_shared_info();
+
if (!xen_have_vector_callback) {
ret = xen_allocate_irq(pdev);
if (ret) {
diff --git a/drivers/xen/xen-acpi-processor.c b/drivers/xen/xen-acpi-processor.c
index 7ff2569..b590ee0 100644
--- a/drivers/xen/xen-acpi-processor.c
+++ b/drivers/xen/xen-acpi-processor.c
@@ -520,15 +520,18 @@ static int __init xen_acpi_processor_init(void)
if (!pr_backup) {
pr_backup = kzalloc(sizeof(struct acpi_processor), GFP_KERNEL);
- memcpy(pr_backup, _pr, sizeof(struct acpi_processor));
+ if (pr_backup)
+ memcpy(pr_backup, _pr, sizeof(struct acpi_processor));
}
(void)upload_pm_data(_pr);
}
rc = check_acpi_ids(pr_backup);
- if (rc)
- goto err_unregister;
kfree(pr_backup);
+ pr_backup = NULL;
+
+ if (rc)
+ goto err_unregister;
return 0;
err_unregister:
diff --git a/drivers/xen/xenbus/xenbus_xs.c b/drivers/xen/xenbus/xenbus_xs.c
index d1c217b..bce15cf 100644
--- a/drivers/xen/xenbus/xenbus_xs.c
+++ b/drivers/xen/xenbus/xenbus_xs.c
@@ -618,6 +618,23 @@ static struct xenbus_watch *find_watch(const char *token)
return NULL;
}
+static void xs_reset_watches(void)
+{
+ int err, supported = 0;
+
+ if (!xen_hvm_domain())
+ return;
+
+ err = xenbus_scanf(XBT_NIL, "control",
+ "platform-feature-xs_reset_watches", "%d", &supported);
+ if (err != 1 || !supported)
+ return;
+
+ err = xs_error(xs_single(XBT_NIL, XS_RESET_WATCHES, "", NULL));
+ if (err && err != -EEXIST)
+ printk(KERN_WARNING "xs_reset_watches failed: %d\n", err);
+}
+
/* Register callback to watch this node. */
int register_xenbus_watch(struct xenbus_watch *watch)
{
@@ -900,5 +917,8 @@ int xs_init(void)
if (IS_ERR(task))
return PTR_ERR(task);
+ /* shutdown watches for kexec boot */
+ xs_reset_watches();
+
return 0;
}
OpenPOWER on IntegriCloud