Merge tag 'uml-for-linux-6.17-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/uml/linux

Pull uml updates from Johannes Berg:
 "Mostly cleanups, except:

   - dynamic addition of vfio passthrough devices

   - implementation of HAVE_SYSCALL_TRACEPOINTS"

* tag 'uml-for-linux-6.17-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/uml/linux:
  um: Replace __ASSEMBLY__ with __ASSEMBLER__ in the usermode headers
  um: Stop tracking stub's PID via userspace_pid[]
  um: Remove the pid parameter of handle_trap()
  um: Use err consistently in userspace()
  um: vfio: Support adding devices via mconsole
  um: rtc: Avoid shadowing err in uml_rtc_start()
  um: Avoid redefining ARCH_HAS_CACHE_LINE_SIZE
  um: Make mm_list and mm_list_lock static
  um: Make unscheduled_userspace_iterations static
  um: Re-evaluate thread flags repeatedly
  um: simplify syscall header files
  um/ptrace: Implement HAVE_SYSCALL_TRACEPOINTS
  um/x86: Add system call table to header file
  um: virt-pci: Switch to msi_create_parent_irq_domain()
  um: virtio_pcidev: Rename UM_PCI_STAT_WAITING

arch/um/Kconfig

@@ -6,6 +6,7 @@ config UML
 	bool
 	default y
 	select ARCH_WANTS_DYNAMIC_TASK_STRUCT
+	select ARCH_HAS_CACHE_LINE_SIZE
 	select ARCH_HAS_CPU_FINALIZE_INIT
 	select ARCH_HAS_FORTIFY_SOURCE
 	select ARCH_HAS_GCOV_PROFILE_ALL
@@ -35,6 +36,7 @@ config UML
 	select HAVE_RUST
 	select ARCH_HAS_UBSAN
 	select HAVE_ARCH_TRACEHOOK
+	select HAVE_SYSCALL_TRACEPOINTS
 	select THREAD_INFO_IN_TASK
 
 config MMU
@@ -82,9 +84,6 @@ config NR_CPUS
 	range 1 1
 	default 1
 
-config ARCH_HAS_CACHE_LINE_SIZE
-	def_bool y
-
 source "arch/$(HEADER_ARCH)/um/Kconfig"
 
 config MAY_HAVE_RUNTIME_DEPS

arch/um/drivers/Kconfig

@@ -160,6 +160,7 @@ config UML_RTC
 config UML_PCI
 	bool
 	select FORCE_PCI
+	select IRQ_MSI_LIB
 	select UML_IOMEM_EMULATION
 	select UML_DMA_EMULATION
 	select PCI_MSI

arch/um/drivers/rtc_user.c

@@ -28,7 +28,7 @@ int uml_rtc_start(bool timetravel)
 	int err;
 
 	if (timetravel) {
-		int err = os_pipe(uml_rtc_irq_fds, 1, 1);
+		err = os_pipe(uml_rtc_irq_fds, 1, 1);
 		if (err)
 			goto fail;
 	} else {

arch/um/drivers/vfio_kern.c

@@ -16,6 +16,7 @@
 #include <init.h>
 #include <os.h>
 
+#include "mconsole_kern.h"
 #include "virt-pci.h"
 #include "vfio_user.h"
 
@@ -60,6 +61,7 @@ static LIST_HEAD(uml_vfio_groups);
 static DEFINE_MUTEX(uml_vfio_groups_mtx);
 
 static LIST_HEAD(uml_vfio_devices);
+static DEFINE_MUTEX(uml_vfio_devices_mtx);
 
 static int uml_vfio_set_container(int group_fd)
 {
@@ -581,32 +583,44 @@ static struct uml_vfio_device *uml_vfio_find_device(const char *device)
 	return NULL;
 }
 
-static int uml_vfio_cmdline_set(const char *device, const struct kernel_param *kp)
+static struct uml_vfio_device *uml_vfio_add_device(const char *device)
 {
 	struct uml_vfio_device *dev;
 	int fd;
 
+	guard(mutex)(&uml_vfio_devices_mtx);
+
 	if (uml_vfio_container.fd < 0) {
 		fd = uml_vfio_user_open_container();
 		if (fd < 0)
-			return fd;
+			return ERR_PTR(fd);
 		uml_vfio_container.fd = fd;
 	}
 
 	if (uml_vfio_find_device(device))
-		return -EEXIST;
+		return ERR_PTR(-EEXIST);
 
 	dev = kzalloc(sizeof(*dev), GFP_KERNEL);
 	if (!dev)
-		return -ENOMEM;
+		return ERR_PTR(-ENOMEM);
 
 	dev->name = kstrdup(device, GFP_KERNEL);
 	if (!dev->name) {
 		kfree(dev);
-		return -ENOMEM;
+		return ERR_PTR(-ENOMEM);
 	}
 
 	list_add_tail(&dev->list, &uml_vfio_devices);
+	return dev;
+}
+
+static int uml_vfio_cmdline_set(const char *device, const struct kernel_param *kp)
+{
+	struct uml_vfio_device *dev;
+
+	dev = uml_vfio_add_device(device);
+	if (IS_ERR(dev))
+		return PTR_ERR(dev);
 	return 0;
 }
 
@@ -629,6 +643,42 @@ __uml_help(uml_vfio_cmdline_param_ops,
 "    through multiple PCI devices to UML.\n\n"
 );
 
+static int uml_vfio_mc_config(char *str, char **error_out)
+{
+	struct uml_vfio_device *dev;
+
+	if (*str != '=') {
+		*error_out = "Invalid config";
+		return -EINVAL;
+	}
+	str += 1;
+
+	dev = uml_vfio_add_device(str);
+	if (IS_ERR(dev))
+		return PTR_ERR(dev);
+	uml_vfio_open_device(dev);
+	return 0;
+}
+
+static int uml_vfio_mc_id(char **str, int *start_out, int *end_out)
+{
+	return -EOPNOTSUPP;
+}
+
+static int uml_vfio_mc_remove(int n, char **error_out)
+{
+	return -EOPNOTSUPP;
+}
+
+static struct mc_device uml_vfio_mc = {
+	.list           = LIST_HEAD_INIT(uml_vfio_mc.list),
+	.name           = "vfio_uml.device",
+	.config         = uml_vfio_mc_config,
+	.get_config     = NULL,
+	.id             = uml_vfio_mc_id,
+	.remove         = uml_vfio_mc_remove,
+};
+
 static int __init uml_vfio_init(void)
 {
 	struct uml_vfio_device *dev, *n;
@@ -639,6 +689,8 @@ static int __init uml_vfio_init(void)
 	list_for_each_entry_safe(dev, n, &uml_vfio_devices, list)
 		uml_vfio_open_device(dev);
 
+	mconsole_register_dev(&uml_vfio_mc);
+
 	return 0;
 }
 late_initcall(uml_vfio_init);

arch/um/drivers/virt-pci.c

@@ -7,6 +7,7 @@
 #include <linux/pci.h>
 #include <linux/logic_iomem.h>
 #include <linux/of_platform.h>
+#include <linux/irqchip/irq-msi-lib.h>
 #include <linux/irqdomain.h>
 #include <linux/msi.h>
 #include <linux/unaligned.h>
@@ -29,7 +30,6 @@ static struct um_pci_device *um_pci_platform_device;
 static struct um_pci_device_reg um_pci_devices[MAX_DEVICES];
 static struct fwnode_handle *um_pci_fwnode;
 static struct irq_domain *um_pci_inner_domain;
-static struct irq_domain *um_pci_msi_domain;
 static unsigned long um_pci_msi_used[BITS_TO_LONGS(MAX_MSI_VECTORS)];
 
 static unsigned long um_pci_cfgspace_read(void *priv, unsigned int offset,
@@ -400,21 +400,24 @@ static void um_pci_inner_domain_free(struct irq_domain *domain,
 }
 
 static const struct irq_domain_ops um_pci_inner_domain_ops = {
+	.select = msi_lib_irq_domain_select,
 	.alloc = um_pci_inner_domain_alloc,
 	.free = um_pci_inner_domain_free,
 };
 
-static struct irq_chip um_pci_msi_irq_chip = {
-	.name = "UM virtual PCIe MSI",
-	.irq_mask = pci_msi_mask_irq,
-	.irq_unmask = pci_msi_unmask_irq,
-};
+#define UM_PCI_MSI_FLAGS_REQUIRED (MSI_FLAG_USE_DEF_DOM_OPS		| \
+				   MSI_FLAG_USE_DEF_CHIP_OPS		| \
+				   MSI_FLAG_NO_AFFINITY)
+#define UM_PCI_MSI_FLAGS_SUPPORTED (MSI_GENERIC_FLAGS_MASK		| \
+				    MSI_FLAG_PCI_MSIX)
 
-static struct msi_domain_info um_pci_msi_domain_info = {
-	.flags	= MSI_FLAG_USE_DEF_DOM_OPS |
-		  MSI_FLAG_USE_DEF_CHIP_OPS |
-		  MSI_FLAG_PCI_MSIX,
-	.chip	= &um_pci_msi_irq_chip,
+static const struct msi_parent_ops um_pci_msi_parent_ops = {
+	.required_flags		= UM_PCI_MSI_FLAGS_REQUIRED,
+	.supported_flags	= UM_PCI_MSI_FLAGS_SUPPORTED,
+	.bus_select_token	= DOMAIN_BUS_NEXUS,
+	.bus_select_mask	= MATCH_PCI_MSI,
+	.prefix			= "UM-virtual-",
+	.init_dev_msi_info	= msi_lib_init_dev_msi_info,
 };
 
 static struct resource busn_resource = {
@@ -559,17 +562,14 @@ static int __init um_pci_init(void)
 		goto free;
 	}
 
-	um_pci_inner_domain = irq_domain_create_linear(um_pci_fwnode, MAX_MSI_VECTORS,
-						       &um_pci_inner_domain_ops, NULL);
-	if (!um_pci_inner_domain) {
-		err = -ENOMEM;
-		goto free;
-	}
+	struct irq_domain_info info = {
+		.fwnode		= um_pci_fwnode,
+		.ops		= &um_pci_inner_domain_ops,
+		.size		= MAX_MSI_VECTORS,
+	};
 
-	um_pci_msi_domain = pci_msi_create_irq_domain(um_pci_fwnode,
-						      &um_pci_msi_domain_info,
-						      um_pci_inner_domain);
-	if (!um_pci_msi_domain) {
+	um_pci_inner_domain = msi_create_parent_irq_domain(&info, &um_pci_msi_parent_ops);
+	if (!um_pci_inner_domain) {
 		err = -ENOMEM;
 		goto free;
 	}
@@ -611,7 +611,6 @@ device_initcall(um_pci_init);
 
 static void __exit um_pci_exit(void)
 {
-	irq_domain_remove(um_pci_msi_domain);
 	irq_domain_remove(um_pci_inner_domain);
 	pci_free_resource_list(&bridge->windows);
 	pci_free_host_bridge(bridge);

arch/um/drivers/virtio_pcidev.c

@@ -42,7 +42,7 @@ struct virtio_pcidev_device {
 	void *extra_ptrs[VIRTIO_PCIDEV_WRITE_BUFS + 1];
 	DECLARE_BITMAP(used_bufs, VIRTIO_PCIDEV_WRITE_BUFS);
 
-#define UM_PCI_STAT_WAITING	0
+#define VIRTIO_PCIDEV_STAT_WAITING	0
 	unsigned long status;
 
 	bool platform;
@@ -172,7 +172,7 @@ static int virtio_pcidev_send_cmd(struct virtio_pcidev_device *dev,
 	}
 
 	/* kick and poll for getting a response on the queue */
-	set_bit(UM_PCI_STAT_WAITING, &dev->status);
+	set_bit(VIRTIO_PCIDEV_STAT_WAITING, &dev->status);
 	virtqueue_kick(dev->cmd_vq);
 	ret = 0;
 
@@ -193,7 +193,7 @@ static int virtio_pcidev_send_cmd(struct virtio_pcidev_device *dev,
 		}
 		udelay(1);
 	}
-	clear_bit(UM_PCI_STAT_WAITING, &dev->status);
+	clear_bit(VIRTIO_PCIDEV_STAT_WAITING, &dev->status);
 
 	if (bounce_out)
 		memcpy(out, buf->data, out_size);
@@ -439,7 +439,7 @@ static void virtio_pcidev_cmd_vq_cb(struct virtqueue *vq)
 	void *cmd;
 	int len;
 
-	if (test_bit(UM_PCI_STAT_WAITING, &dev->status))
+	if (test_bit(VIRTIO_PCIDEV_STAT_WAITING, &dev->status))
 		return;
 
 	while ((cmd = virtqueue_get_buf(vq, &len)))

arch/um/include/asm/cpufeature.h

@@ -4,7 +4,7 @@
 
 #include <asm/processor.h>
 
-#if defined(__KERNEL__) && !defined(__ASSEMBLY__)
+#if defined(__KERNEL__) && !defined(__ASSEMBLER__)
 
 #include <asm/asm.h>
 #include <linux/bitops.h>
@@ -137,5 +137,5 @@ t_no:
 #define CPU_FEATURE_TYPEVAL		boot_cpu_data.x86_vendor, boot_cpu_data.x86, \
 					boot_cpu_data.x86_model
 
-#endif /* defined(__KERNEL__) && !defined(__ASSEMBLY__) */
+#endif /* defined(__KERNEL__) && !defined(__ASSEMBLER__) */
 #endif /* _ASM_UM_CPUFEATURE_H */

arch/um/include/asm/current.h

@@ -5,7 +5,7 @@
 #include <linux/compiler.h>
 #include <linux/threads.h>
 
-#ifndef __ASSEMBLY__
+#ifndef __ASSEMBLER__
 
 struct task_struct;
 extern struct task_struct *cpu_tasks[NR_CPUS];
@@ -18,6 +18,6 @@ static __always_inline struct task_struct *get_current(void)
 
 #define current get_current()
 
-#endif /* __ASSEMBLY__ */
+#endif /* __ASSEMBLER__ */
 
 #endif /* __ASM_CURRENT_H */

arch/um/include/asm/mmu_context.h

@@ -16,11 +16,6 @@
 #define activate_mm activate_mm
 static inline void activate_mm(struct mm_struct *old, struct mm_struct *new)
 {
-	/*
-	 * This is called by fs/exec.c and sys_unshare()
-	 * when the new ->mm is used for the first time.
-	 */
-	__switch_mm(&new->context.id);
 }
 
 static inline void switch_mm(struct mm_struct *prev, struct mm_struct *next, 
@@ -28,11 +23,9 @@ static inline void switch_mm(struct mm_struct *prev, struct mm_struct *next,
 {
 	unsigned cpu = smp_processor_id();
 
-	if(prev != next){
+	if (prev != next) {
 		cpumask_clear_cpu(cpu, mm_cpumask(prev));
 		cpumask_set_cpu(cpu, mm_cpumask(next));
-		if(next != &init_mm)
-			__switch_mm(&next->context.id);
 	}
 }
 

arch/um/include/asm/page.h

@@ -11,7 +11,7 @@
 
 #include <vdso/page.h>
 
-#ifndef __ASSEMBLY__
+#ifndef __ASSEMBLER__
 
 struct page;
 
@@ -94,7 +94,7 @@ extern unsigned long uml_physmem;
 #include <asm-generic/memory_model.h>
 #include <asm-generic/getorder.h>
 
-#endif	/* __ASSEMBLY__ */
+#endif	/* __ASSEMBLER__ */
 
 #ifdef CONFIG_X86_32
 #define __HAVE_ARCH_GATE_AREA 1

arch/um/include/asm/ptrace-generic.h

@@ -6,7 +6,7 @@
 #ifndef __UM_PTRACE_GENERIC_H
 #define __UM_PTRACE_GENERIC_H
 
-#ifndef __ASSEMBLY__
+#ifndef __ASSEMBLER__
 
 #include <sysdep/ptrace.h>
 

arch/um/include/asm/thread_info.h

@@ -9,7 +9,7 @@
 #define THREAD_SIZE_ORDER CONFIG_KERNEL_STACK_ORDER
 #define THREAD_SIZE ((1 << CONFIG_KERNEL_STACK_ORDER) * PAGE_SIZE)
 
-#ifndef __ASSEMBLY__
+#ifndef __ASSEMBLER__
 
 #include <asm/types.h>
 #include <asm/page.h>
@@ -43,6 +43,8 @@ struct thread_info {
 #define TIF_NOTIFY_RESUME	8
 #define TIF_SECCOMP		9	/* secure computing */
 #define TIF_SINGLESTEP		10	/* single stepping userspace */
+#define TIF_SYSCALL_TRACEPOINT	11	/* syscall tracepoint instrumentation */
+
 
 #define _TIF_SYSCALL_TRACE	(1 << TIF_SYSCALL_TRACE)
 #define _TIF_SIGPENDING		(1 << TIF_SIGPENDING)
@@ -50,7 +52,11 @@ struct thread_info {
 #define _TIF_NOTIFY_SIGNAL	(1 << TIF_NOTIFY_SIGNAL)
 #define _TIF_MEMDIE		(1 << TIF_MEMDIE)
 #define _TIF_SYSCALL_AUDIT	(1 << TIF_SYSCALL_AUDIT)
+#define _TIF_NOTIFY_RESUME	(1 << TIF_NOTIFY_RESUME)
 #define _TIF_SECCOMP		(1 << TIF_SECCOMP)
 #define _TIF_SINGLESTEP		(1 << TIF_SINGLESTEP)
 
+#define _TIF_WORK_MASK		(_TIF_NEED_RESCHED | _TIF_SIGPENDING | _TIF_NOTIFY_SIGNAL | \
+				 _TIF_NOTIFY_RESUME)
+
 #endif

arch/um/include/shared/as-layout.h

@@ -26,7 +26,7 @@
 #define STUB_DATA_PAGES 2 /* must be a power of two */
 #define STUB_END (STUB_DATA + STUB_DATA_PAGES * UM_KERN_PAGE_SIZE)
 
-#ifndef __ASSEMBLY__
+#ifndef __ASSEMBLER__
 
 #include <sysdep/ptrace.h>
 

arch/um/include/shared/skas/mm_id.h

@@ -19,8 +19,6 @@ struct mm_id {
 	int syscall_fd_map[STUB_MAX_FDS];
 };
 
-void __switch_mm(struct mm_id *mm_idp);
-
 void notify_mm_kill(int pid);
 
 #endif

arch/um/include/shared/skas/skas.h

@@ -9,7 +9,6 @@
 #include <sysdep/ptrace.h>
 
 extern int using_seccomp;
-extern int userspace_pid[];
 
 extern void new_thread_handler(void);
 extern void handle_syscall(struct uml_pt_regs *regs);

arch/um/kernel/exec.c

@@ -26,8 +26,6 @@ void flush_thread(void)
 
 	get_safe_registers(current_pt_regs()->regs.gp,
 			   current_pt_regs()->regs.fp);
-
-	__switch_mm(&current->mm->context.id);
 }
 
 void start_thread(struct pt_regs *regs, unsigned long eip, unsigned long esp)

arch/um/kernel/process.c

@@ -82,14 +82,18 @@ struct task_struct *__switch_to(struct task_struct *from, struct task_struct *to
 void interrupt_end(void)
 {
 	struct pt_regs *regs = &current->thread.regs;
+	unsigned long thread_flags;
 
-	if (need_resched())
-		schedule();
-	if (test_thread_flag(TIF_SIGPENDING) ||
-	    test_thread_flag(TIF_NOTIFY_SIGNAL))
-		do_signal(regs);
-	if (test_thread_flag(TIF_NOTIFY_RESUME))
-		resume_user_mode_work(regs);
+	thread_flags = read_thread_flags();
+	while (thread_flags & _TIF_WORK_MASK) {
+		if (thread_flags & _TIF_NEED_RESCHED)
+			schedule();
+		if (thread_flags & (_TIF_SIGPENDING | _TIF_NOTIFY_SIGNAL))
+			do_signal(regs);
+		if (thread_flags & _TIF_NOTIFY_RESUME)
+			resume_user_mode_work(regs);
+		thread_flags = read_thread_flags();
+	}
 }
 
 int get_current_pid(void)

arch/um/kernel/ptrace.c

@@ -9,6 +9,9 @@
 #include <linux/uaccess.h>
 #include <asm/ptrace-abi.h>
 
+#define CREATE_TRACE_POINTS
+#include <trace/events/syscalls.h>
+
 void user_enable_single_step(struct task_struct *child)
 {
 	set_tsk_thread_flag(child, TIF_SINGLESTEP);
@@ -126,6 +129,9 @@ int syscall_trace_enter(struct pt_regs *regs)
 			    UPT_SYSCALL_ARG3(&regs->regs),
 			    UPT_SYSCALL_ARG4(&regs->regs));
 
+	if (test_thread_flag(TIF_SYSCALL_TRACEPOINT))
+		trace_sys_enter(regs, UPT_SYSCALL_NR(&regs->regs));
+
 	if (!test_thread_flag(TIF_SYSCALL_TRACE))
 		return 0;
 
@@ -142,6 +148,9 @@ void syscall_trace_leave(struct pt_regs *regs)
 	if (test_thread_flag(TIF_SINGLESTEP))
 		send_sigtrap(&regs->regs, 0);
 
+	if (test_thread_flag(TIF_SYSCALL_TRACEPOINT))
+		trace_sys_exit(regs, PT_REGS_SYSCALL_RET(regs));
+
 	if (!test_thread_flag(TIF_SYSCALL_TRACE))
 		return;
 

arch/um/kernel/skas/mmu.c

@@ -20,8 +20,8 @@
 /* Ensure the stub_data struct covers the allocated area */
 static_assert(sizeof(struct stub_data) == STUB_DATA_PAGES * UM_KERN_PAGE_SIZE);
 
-spinlock_t mm_list_lock;
-struct list_head mm_list;
+static spinlock_t mm_list_lock;
+static struct list_head mm_list;
 
 int init_new_context(struct task_struct *task, struct mm_struct *mm)
 {

arch/um/kernel/skas/process.c

@@ -26,8 +26,6 @@ static int __init start_kernel_proc(void *unused)
 	return 0;
 }
 
-extern int userspace_pid[];
-
 static char cpu0_irqstack[THREAD_SIZE] __aligned(THREAD_SIZE);
 
 int __init start_uml(void)