xen: branch for v6.6-rc1

-----BEGIN PGP SIGNATURE-----
 
 iHUEABYIAB0WIQRTLbB6QfY48x44uB6AXGG7T9hjvgUCZOdp9QAKCRCAXGG7T9hj
 vii+AQDgY+/PhND0pS40yKHPZpfbe1dGdNlbuD0ktumHQ3a5nAEAu6WNFEEG8NPx
 Unopygv7irUdytK4RY+DPgB7FKRCuA4=
 =XrQ0
 -----END PGP SIGNATURE-----

Merge tag 'for-linus-6.6-rc1-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/xen/tip

Pull xen updates from Juergen Gross:

 - a bunch of minor cleanups

 - a patch adding irqfd support for virtio backends running as user
   daemon supporting Xen guests

* tag 'for-linus-6.6-rc1-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/xen/tip:
  xen: privcmd: Add support for irqfd
  xen/xenbus: Avoid a lockdep warning when adding a watch
  xen: Fix one kernel-doc comment
  xen: xenbus: Use helper function IS_ERR_OR_NULL()
  xen: Switch to use kmemdup() helper
  xen-pciback: Remove unused function declarations
  x86/xen: Make virt_to_pfn() a static inline
  xen: remove a confusing comment on auto-translated guest I/O
  xen/evtchn: Remove unused function declaration xen_set_affinity_evtchn()

arch/x86/include/asm/xen/page.h

@@ -295,7 +295,10 @@ static inline unsigned long bfn_to_local_pfn(unsigned long mfn)
 
 /* VIRT <-> MACHINE conversion */
 #define virt_to_machine(v)	(phys_to_machine(XPADDR(__pa(v))))
-#define virt_to_pfn(v)          (PFN_DOWN(__pa(v)))
+static inline unsigned long virt_to_pfn(const void *v)
+{
+	return PFN_DOWN(__pa(v));
+}
 #define virt_to_mfn(v)		(pfn_to_mfn(virt_to_pfn(v)))
 #define mfn_to_virt(m)		(__va(mfn_to_pfn(m) << PAGE_SHIFT))
 

arch/x86/xen/enlighten_pv.c

@@ -523,7 +523,7 @@ static void __init xen_load_gdt_boot(const struct desc_ptr *dtr)
 	BUG_ON(size > PAGE_SIZE);
 	BUG_ON(va & ~PAGE_MASK);
 
-	pfn = virt_to_pfn(va);
+	pfn = virt_to_pfn((void *)va);
 	mfn = pfn_to_mfn(pfn);
 
 	pte = pfn_pte(pfn, PAGE_KERNEL_RO);

arch/x86/xen/mmu_pv.c

@@ -2202,13 +2202,13 @@ static void xen_zap_pfn_range(unsigned long vaddr, unsigned int order,
 		mcs = __xen_mc_entry(0);
 
 		if (in_frames)
-			in_frames[i] = virt_to_mfn(vaddr);
+			in_frames[i] = virt_to_mfn((void *)vaddr);
 
 		MULTI_update_va_mapping(mcs.mc, vaddr, VOID_PTE, 0);
-		__set_phys_to_machine(virt_to_pfn(vaddr), INVALID_P2M_ENTRY);
+		__set_phys_to_machine(virt_to_pfn((void *)vaddr), INVALID_P2M_ENTRY);
 
 		if (out_frames)
-			out_frames[i] = virt_to_pfn(vaddr);
+			out_frames[i] = virt_to_pfn((void *)vaddr);
 	}
 	xen_mc_issue(0);
 }
@@ -2250,7 +2250,7 @@ static void xen_remap_exchanged_ptes(unsigned long vaddr, int order,
 		MULTI_update_va_mapping(mcs.mc, vaddr,
 				mfn_pte(mfn, PAGE_KERNEL), flags);
 
-		set_phys_to_machine(virt_to_pfn(vaddr), mfn);
+		set_phys_to_machine(virt_to_pfn((void *)vaddr), mfn);
 	}
 
 	xen_mc_issue(0);
@@ -2310,12 +2310,6 @@ int xen_create_contiguous_region(phys_addr_t pstart, unsigned int order,
 	int            success;
 	unsigned long vstart = (unsigned long)phys_to_virt(pstart);
 
-	/*
-	 * Currently an auto-translated guest will not perform I/O, nor will
-	 * it require PAE page directories below 4GB. Therefore any calls to
-	 * this function are redundant and can be ignored.
-	 */
-
 	if (unlikely(order > MAX_CONTIG_ORDER))
 		return -ENOMEM;
 
@@ -2327,7 +2321,7 @@ int xen_create_contiguous_region(phys_addr_t pstart, unsigned int order,
 	xen_zap_pfn_range(vstart, order, in_frames, NULL);
 
 	/* 2. Get a new contiguous memory extent. */
-	out_frame = virt_to_pfn(vstart);
+	out_frame = virt_to_pfn((void *)vstart);
 	success = xen_exchange_memory(1UL << order, 0, in_frames,
 				      1, order, &out_frame,
 				      address_bits);
@@ -2360,7 +2354,7 @@ void xen_destroy_contiguous_region(phys_addr_t pstart, unsigned int order)
 	spin_lock_irqsave(&xen_reservation_lock, flags);
 
 	/* 1. Find start MFN of contiguous extent. */
-	in_frame = virt_to_mfn(vstart);
+	in_frame = virt_to_mfn((void *)vstart);
 
 	/* 2. Zap current PTEs. */
 	xen_zap_pfn_range(vstart, order, NULL, out_frames);

arch/x86/xen/setup.c

@@ -340,7 +340,7 @@ static void __init xen_do_set_identity_and_remap_chunk(
 
 	WARN_ON(size == 0);
 
-	mfn_save = virt_to_mfn(buf);
+	mfn_save = virt_to_mfn((void *)buf);
 
 	for (ident_pfn_iter = start_pfn, remap_pfn_iter = remap_pfn;
 	     ident_pfn_iter < ident_end_pfn;
@@ -503,7 +503,7 @@ void __init xen_remap_memory(void)
 	unsigned long pfn_s = ~0UL;
 	unsigned long len = 0;
 
-	mfn_save = virt_to_mfn(buf);
+	mfn_save = virt_to_mfn((void *)buf);
 
 	while (xen_remap_mfn != INVALID_P2M_ENTRY) {
 		/* Map the remap information */

drivers/xen/Kconfig

@@ -269,6 +269,13 @@ config XEN_PRIVCMD
 	  disaggregated Xen setups this driver might be needed for other
 	  domains, too.
 
+config XEN_PRIVCMD_IRQFD
+	bool "Xen irqfd support"
+	depends on XEN_PRIVCMD && XEN_VIRTIO && EVENTFD
+	help
+	  Using the irqfd mechanism a virtio backend running in a daemon can
+	  speed up interrupt injection into a guest.
+
 config XEN_ACPI_PROCESSOR
 	tristate "Xen ACPI processor"
 	depends on XEN && XEN_PV_DOM0 && X86 && ACPI_PROCESSOR && CPU_FREQ

drivers/xen/grant-table.c

@@ -1044,7 +1044,7 @@ EXPORT_SYMBOL_GPL(gnttab_pages_clear_private);
 
 /**
  * gnttab_free_pages - free pages allocated by gnttab_alloc_pages()
- * @nr_pages; number of pages to free
+ * @nr_pages: number of pages to free
  * @pages: the pages
  */
 void gnttab_free_pages(int nr_pages, struct page **pages)

drivers/xen/privcmd.c

@@ -9,11 +9,16 @@
 
 #define pr_fmt(fmt) "xen:" KBUILD_MODNAME ": " fmt
 
+#include <linux/eventfd.h>
+#include <linux/file.h>
 #include <linux/kernel.h>
 #include <linux/module.h>
+#include <linux/mutex.h>
+#include <linux/poll.h>
 #include <linux/sched.h>
 #include <linux/slab.h>
 #include <linux/string.h>
+#include <linux/workqueue.h>
 #include <linux/errno.h>
 #include <linux/mm.h>
 #include <linux/mman.h>
@@ -833,6 +838,263 @@ out:
 	return rc;
 }
 
+#ifdef CONFIG_XEN_PRIVCMD_IRQFD
+/* Irqfd support */
+static struct workqueue_struct *irqfd_cleanup_wq;
+static DEFINE_MUTEX(irqfds_lock);
+static LIST_HEAD(irqfds_list);
+
+struct privcmd_kernel_irqfd {
+	struct xen_dm_op_buf xbufs;
+	domid_t dom;
+	bool error;
+	struct eventfd_ctx *eventfd;
+	struct work_struct shutdown;
+	wait_queue_entry_t wait;
+	struct list_head list;
+	poll_table pt;
+};
+
+static void irqfd_deactivate(struct privcmd_kernel_irqfd *kirqfd)
+{
+	lockdep_assert_held(&irqfds_lock);
+
+	list_del_init(&kirqfd->list);
+	queue_work(irqfd_cleanup_wq, &kirqfd->shutdown);
+}
+
+static void irqfd_shutdown(struct work_struct *work)
+{
+	struct privcmd_kernel_irqfd *kirqfd =
+		container_of(work, struct privcmd_kernel_irqfd, shutdown);
+	u64 cnt;
+
+	eventfd_ctx_remove_wait_queue(kirqfd->eventfd, &kirqfd->wait, &cnt);
+	eventfd_ctx_put(kirqfd->eventfd);
+	kfree(kirqfd);
+}
+
+static void irqfd_inject(struct privcmd_kernel_irqfd *kirqfd)
+{
+	u64 cnt;
+	long rc;
+
+	eventfd_ctx_do_read(kirqfd->eventfd, &cnt);
+
+	xen_preemptible_hcall_begin();
+	rc = HYPERVISOR_dm_op(kirqfd->dom, 1, &kirqfd->xbufs);
+	xen_preemptible_hcall_end();
+
+	/* Don't repeat the error message for consecutive failures */
+	if (rc && !kirqfd->error) {
+		pr_err("Failed to configure irq for guest domain: %d\n",
+		       kirqfd->dom);
+	}
+
+	kirqfd->error = rc;
+}
+
+static int
+irqfd_wakeup(wait_queue_entry_t *wait, unsigned int mode, int sync, void *key)
+{
+	struct privcmd_kernel_irqfd *kirqfd =
+		container_of(wait, struct privcmd_kernel_irqfd, wait);
+	__poll_t flags = key_to_poll(key);
+
+	if (flags & EPOLLIN)
+		irqfd_inject(kirqfd);
+
+	if (flags & EPOLLHUP) {
+		mutex_lock(&irqfds_lock);
+		irqfd_deactivate(kirqfd);
+		mutex_unlock(&irqfds_lock);
+	}
+
+	return 0;
+}
+
+static void
+irqfd_poll_func(struct file *file, wait_queue_head_t *wqh, poll_table *pt)
+{
+	struct privcmd_kernel_irqfd *kirqfd =
+		container_of(pt, struct privcmd_kernel_irqfd, pt);
+
+	add_wait_queue_priority(wqh, &kirqfd->wait);
+}
+
+static int privcmd_irqfd_assign(struct privcmd_irqfd *irqfd)
+{
+	struct privcmd_kernel_irqfd *kirqfd, *tmp;
+	__poll_t events;
+	struct fd f;
+	void *dm_op;
+	int ret;
+
+	kirqfd = kzalloc(sizeof(*kirqfd) + irqfd->size, GFP_KERNEL);
+	if (!kirqfd)
+		return -ENOMEM;
+	dm_op = kirqfd + 1;
+
+	if (copy_from_user(dm_op, irqfd->dm_op, irqfd->size)) {
+		ret = -EFAULT;
+		goto error_kfree;
+	}
+
+	kirqfd->xbufs.size = irqfd->size;
+	set_xen_guest_handle(kirqfd->xbufs.h, dm_op);
+	kirqfd->dom = irqfd->dom;
+	INIT_WORK(&kirqfd->shutdown, irqfd_shutdown);
+
+	f = fdget(irqfd->fd);
+	if (!f.file) {
+		ret = -EBADF;
+		goto error_kfree;
+	}
+
+	kirqfd->eventfd = eventfd_ctx_fileget(f.file);
+	if (IS_ERR(kirqfd->eventfd)) {
+		ret = PTR_ERR(kirqfd->eventfd);
+		goto error_fd_put;
+	}
+
+	/*
+	 * Install our own custom wake-up handling so we are notified via a
+	 * callback whenever someone signals the underlying eventfd.
+	 */
+	init_waitqueue_func_entry(&kirqfd->wait, irqfd_wakeup);
+	init_poll_funcptr(&kirqfd->pt, irqfd_poll_func);
+
+	mutex_lock(&irqfds_lock);
+
+	list_for_each_entry(tmp, &irqfds_list, list) {
+		if (kirqfd->eventfd == tmp->eventfd) {
+			ret = -EBUSY;
+			mutex_unlock(&irqfds_lock);
+			goto error_eventfd;
+		}
+	}
+
+	list_add_tail(&kirqfd->list, &irqfds_list);
+	mutex_unlock(&irqfds_lock);
+
+	/*
+	 * Check if there was an event already pending on the eventfd before we
+	 * registered, and trigger it as if we didn't miss it.
+	 */
+	events = vfs_poll(f.file, &kirqfd->pt);
+	if (events & EPOLLIN)
+		irqfd_inject(kirqfd);
+
+	/*
+	 * Do not drop the file until the kirqfd is fully initialized, otherwise
+	 * we might race against the EPOLLHUP.
+	 */
+	fdput(f);
+	return 0;
+
+error_eventfd:
+	eventfd_ctx_put(kirqfd->eventfd);
+
+error_fd_put:
+	fdput(f);
+
+error_kfree:
+	kfree(kirqfd);
+	return ret;
+}
+
+static int privcmd_irqfd_deassign(struct privcmd_irqfd *irqfd)
+{
+	struct privcmd_kernel_irqfd *kirqfd;
+	struct eventfd_ctx *eventfd;
+
+	eventfd = eventfd_ctx_fdget(irqfd->fd);
+	if (IS_ERR(eventfd))
+		return PTR_ERR(eventfd);
+
+	mutex_lock(&irqfds_lock);
+
+	list_for_each_entry(kirqfd, &irqfds_list, list) {
+		if (kirqfd->eventfd == eventfd) {
+			irqfd_deactivate(kirqfd);
+			break;
+		}
+	}
+
+	mutex_unlock(&irqfds_lock);
+
+	eventfd_ctx_put(eventfd);
+
+	/*
+	 * Block until we know all outstanding shutdown jobs have completed so
+	 * that we guarantee there will not be any more interrupts once this
+	 * deassign function returns.
+	 */
+	flush_workqueue(irqfd_cleanup_wq);
+
+	return 0;
+}
+
+static long privcmd_ioctl_irqfd(struct file *file, void __user *udata)
+{
+	struct privcmd_data *data = file->private_data;
+	struct privcmd_irqfd irqfd;
+
+	if (copy_from_user(&irqfd, udata, sizeof(irqfd)))
+		return -EFAULT;
+
+	/* No other flags should be set */
+	if (irqfd.flags & ~PRIVCMD_IRQFD_FLAG_DEASSIGN)
+		return -EINVAL;
+
+	/* If restriction is in place, check the domid matches */
+	if (data->domid != DOMID_INVALID && data->domid != irqfd.dom)
+		return -EPERM;
+
+	if (irqfd.flags & PRIVCMD_IRQFD_FLAG_DEASSIGN)
+		return privcmd_irqfd_deassign(&irqfd);
+
+	return privcmd_irqfd_assign(&irqfd);
+}
+
+static int privcmd_irqfd_init(void)
+{
+	irqfd_cleanup_wq = alloc_workqueue("privcmd-irqfd-cleanup", 0, 0);
+	if (!irqfd_cleanup_wq)
+		return -ENOMEM;
+
+	return 0;
+}
+
+static void privcmd_irqfd_exit(void)
+{
+	struct privcmd_kernel_irqfd *kirqfd, *tmp;
+
+	mutex_lock(&irqfds_lock);
+
+	list_for_each_entry_safe(kirqfd, tmp, &irqfds_list, list)
+		irqfd_deactivate(kirqfd);
+
+	mutex_unlock(&irqfds_lock);
+
+	destroy_workqueue(irqfd_cleanup_wq);
+}
+#else
+static inline long privcmd_ioctl_irqfd(struct file *file, void __user *udata)
+{
+	return -EOPNOTSUPP;
+}
+
+static inline int privcmd_irqfd_init(void)
+{
+	return 0;
+}
+
+static inline void privcmd_irqfd_exit(void)
+{
+}
+#endif /* CONFIG_XEN_PRIVCMD_IRQFD */
+
 static long privcmd_ioctl(struct file *file,
 			  unsigned int cmd, unsigned long data)
 {
@@ -868,6 +1130,10 @@ static long privcmd_ioctl(struct file *file,
 		ret = privcmd_ioctl_mmap_resource(file, udata);
 		break;
 
+	case IOCTL_PRIVCMD_IRQFD:
+		ret = privcmd_ioctl_irqfd(file, udata);
+		break;
+
 	default:
 		break;
 	}
@@ -992,15 +1258,27 @@ static int __init privcmd_init(void)
 	err = misc_register(&xen_privcmdbuf_dev);
 	if (err != 0) {
 		pr_err("Could not register Xen hypercall-buf device\n");
-		misc_deregister(&privcmd_dev);
-		return err;
+		goto err_privcmdbuf;
+	}
+
+	err = privcmd_irqfd_init();
+	if (err != 0) {
+		pr_err("irqfd init failed\n");
+		goto err_irqfd;
 	}
 
 	return 0;
+
+err_irqfd:
+	misc_deregister(&xen_privcmdbuf_dev);
+err_privcmdbuf:
+	misc_deregister(&privcmd_dev);
+	return err;
 }
 
 static void __exit privcmd_exit(void)
 {
+	privcmd_irqfd_exit();
 	misc_deregister(&privcmd_dev);
 	misc_deregister(&xen_privcmdbuf_dev);
 }

drivers/xen/xen-acpi-processor.c

@@ -473,11 +473,8 @@ static int xen_upload_processor_pm_data(void)
 		if (!_pr)
 			continue;
 
-		if (!pr_backup) {
-			pr_backup = kzalloc(sizeof(struct acpi_processor), GFP_KERNEL);
-			if (pr_backup)
-				memcpy(pr_backup, _pr, sizeof(struct acpi_processor));
-		}
+		if (!pr_backup)
+			pr_backup = kmemdup(_pr, sizeof(*_pr), GFP_KERNEL);
 		(void)upload_pm_data(_pr);
 	}
 

drivers/xen/xen-pciback/conf_space_quirks.h

@@ -21,8 +21,6 @@ struct xen_pcibk_config_quirk {
 int xen_pcibk_config_quirks_add_field(struct pci_dev *dev, struct config_field
 				    *field);
 
-int xen_pcibk_config_quirks_remove_field(struct pci_dev *dev, int reg);
-
 int xen_pcibk_config_quirks_init(struct pci_dev *dev);
 
 void xen_pcibk_config_field_free(struct config_field *field);

drivers/xen/xen-pciback/pciback.h

@@ -201,6 +201,3 @@ static inline void xen_pcibk_lateeoi(struct xen_pcibk_device *pdev,
 int xen_pcibk_xenbus_register(void);
 void xen_pcibk_xenbus_unregister(void);
 #endif
-
-/* Handles shared IRQs that can to device domain and control domain. */
-void xen_pcibk_irq_handler(struct pci_dev *dev, int reset);

drivers/xen/xenbus/xenbus_probe_frontend.c

@@ -429,7 +429,7 @@ static void xenbus_check_frontend(char *class, char *dev)
 		printk(KERN_DEBUG "XENBUS: frontend %s %s\n",
 				frontend, xenbus_strstate(fe_state));
 		backend = xenbus_read(XBT_NIL, frontend, "backend", NULL);
-		if (!backend || IS_ERR(backend))
+		if (IS_ERR_OR_NULL(backend))
 			goto out;
 		err = xenbus_scanf(XBT_NIL, backend, "state", "%i", &be_state);
 		if (err == 1)

drivers/xen/xenbus/xenbus_xs.c

@@ -840,8 +840,8 @@ void xs_suspend(void)
 {
 	xs_suspend_enter();
 
-	down_write(&xs_watch_rwsem);
 	mutex_lock(&xs_response_mutex);
+	down_write(&xs_watch_rwsem);
 }
 
 void xs_resume(void)
@@ -866,8 +866,8 @@ void xs_resume(void)
 
 void xs_suspend_cancel(void)
 {
-	mutex_unlock(&xs_response_mutex);
 	up_write(&xs_watch_rwsem);
+	mutex_unlock(&xs_response_mutex);
 
 	xs_suspend_exit();
 }

include/uapi/xen/privcmd.h

@@ -98,6 +98,18 @@ struct privcmd_mmap_resource {
 	__u64 addr;
 };
 
+/* For privcmd_irqfd::flags */
+#define PRIVCMD_IRQFD_FLAG_DEASSIGN (1 << 0)
+
+struct privcmd_irqfd {
+	void __user *dm_op;
+	__u32 size; /* Size of structure pointed by dm_op */
+	__u32 fd;
+	__u32 flags;
+	domid_t dom;
+	__u8 pad[2];
+};
+
 /*
  * @cmd: IOCTL_PRIVCMD_HYPERCALL
  * @arg: &privcmd_hypercall_t
@@ -125,5 +137,7 @@ struct privcmd_mmap_resource {
 	_IOC(_IOC_NONE, 'P', 6, sizeof(domid_t))
 #define IOCTL_PRIVCMD_MMAP_RESOURCE				\
 	_IOC(_IOC_NONE, 'P', 7, sizeof(struct privcmd_mmap_resource))
+#define IOCTL_PRIVCMD_IRQFD					\
+	_IOC(_IOC_NONE, 'P', 8, sizeof(struct privcmd_irqfd))
 
 #endif /* __LINUX_PUBLIC_PRIVCMD_H__ */

include/xen/events.h

@@ -75,7 +75,6 @@ void evtchn_put(evtchn_port_t evtchn);
 
 void xen_send_IPI_one(unsigned int cpu, enum ipi_vector vector);
 void rebind_evtchn_irq(evtchn_port_t evtchn, int irq);
-int xen_set_affinity_evtchn(struct irq_desc *desc, unsigned int tcpu);
 
 static inline void notify_remote_via_evtchn(evtchn_port_t port)
 {