mirror of
https://github.com/torvalds/linux.git
synced 2024-11-15 00:21:59 +00:00
67850b7bdc
of Peter Zijlstra was encountering with ptrace in his freezer rewrite I identified some cleanups to ptrace_stop that make sense on their own and move make resolving the other problems much simpler. The biggest issue is the habbit of the ptrace code to change task->__state from the tracer to suppress TASK_WAKEKILL from waking up the tracee. No other code in the kernel does that and it is straight forward to update signal_wake_up and friends to make that unnecessary. Peter's task freezer sets frozen tasks to a new state TASK_FROZEN and then it stores them by calling "wake_up_state(t, TASK_FROZEN)" relying on the fact that all stopped states except the special stop states can tolerate spurious wake up and recover their state. The state of stopped and traced tasked is changed to be stored in task->jobctl as well as in task->__state. This makes it possible for the freezer to recover tasks in these special states, as well as serving as a general cleanup. With a little more work in that direction I believe TASK_STOPPED can learn to tolerate spurious wake ups and become an ordinary stop state. The TASK_TRACED state has to remain a special state as the registers for a process are only reliably available when the process is stopped in the scheduler. Fundamentally ptrace needs acess to the saved register values of a task. There are bunch of semi-random ptrace related cleanups that were found while looking at these issues. One cleanup that deserves to be called out is from commit57b6de08b5
("ptrace: Admit ptrace_stop can generate spuriuos SIGTRAPs"). This makes a change that is technically user space visible, in the handling of what happens to a tracee when a tracer dies unexpectedly. According to our testing and our understanding of userspace nothing cares that spurious SIGTRAPs can be generated in that case. The entire discussion can be found at: https://lkml.kernel.org/r/87a6bv6dl6.fsf_-_@email.froward.int.ebiederm.org Eric W. Biederman (11): signal: Rename send_signal send_signal_locked signal: Replace __group_send_sig_info with send_signal_locked ptrace/um: Replace PT_DTRACE with TIF_SINGLESTEP ptrace/xtensa: Replace PT_SINGLESTEP with TIF_SINGLESTEP ptrace: Remove arch_ptrace_attach signal: Use lockdep_assert_held instead of assert_spin_locked ptrace: Reimplement PTRACE_KILL by always sending SIGKILL ptrace: Document that wait_task_inactive can't fail ptrace: Admit ptrace_stop can generate spuriuos SIGTRAPs ptrace: Don't change __state ptrace: Always take siglock in ptrace_resume Peter Zijlstra (1): sched,signal,ptrace: Rework TASK_TRACED, TASK_STOPPED state arch/ia64/include/asm/ptrace.h | 4 -- arch/ia64/kernel/ptrace.c | 57 ---------------- arch/um/include/asm/thread_info.h | 2 + arch/um/kernel/exec.c | 2 +- arch/um/kernel/process.c | 2 +- arch/um/kernel/ptrace.c | 8 +-- arch/um/kernel/signal.c | 4 +- arch/x86/kernel/step.c | 3 +- arch/xtensa/kernel/ptrace.c | 4 +- arch/xtensa/kernel/signal.c | 4 +- drivers/tty/tty_jobctrl.c | 4 +- include/linux/ptrace.h | 7 -- include/linux/sched.h | 10 ++- include/linux/sched/jobctl.h | 8 +++ include/linux/sched/signal.h | 20 ++++-- include/linux/signal.h | 3 +- kernel/ptrace.c | 87 ++++++++--------------- kernel/sched/core.c | 5 +- kernel/signal.c | 140 +++++++++++++++++--------------------- kernel/time/posix-cpu-timers.c | 6 +- 20 files changed, 140 insertions(+), 240 deletions(-) Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com> -----BEGIN PGP SIGNATURE----- iQIzBAABCgAdFiEEgjlraLDcwBA2B+6cC/v6Eiajj0AFAmKaXaYACgkQC/v6Eiaj j0CgoA/+JncSQ6PY2D5Jh1apvHzmnRsFXzr3DRvtv/CVx4oIebOXRQFyVDeD5tRn TmMgB29HpBlHRDLojlmlZRGAld1HR/aPEW9j8W1D3Sy/ZFO5L8lQitv9aDHO9Ntw 4lZvlhS1M0KhATudVVBqSPixiG6CnV5SsGmixqdOyg7xcXSY6G1l2nB7Zk9I3Tat ZlmhuZ6R5Z5qsm4MEq0vUSrnsHiGxYrpk6uQOaVz8Wkv8ZFmbutt6XgxF0tsyZNn mHSmWSiZzIgBjTlaibEmxi8urYJTPj3vGBeJQVYHblFwLFi6+Oy7bDxQbWjQvaZh DsgWPScfBF4Jm0+8hhCiSYpvPp8XnZuklb4LNCeok/VFr+KfSmpJTIhn00kagQ1u vxQDqLws8YLW4qsfGydfx9uUIFCbQE/V2VDYk5J3Re3gkUNDOOR1A56hPniKv6VB 2aqGO2Fl0RdBbUa3JF+XI5Pwq5y1WrqR93EUvj+5+u5W9rZL/8WLBHBMEz6gbmfD DhwFE0y8TG2WRlWJVEDRId+5zo3di/YvasH0vJZ5HbrxhS2RE/yIGAd+kKGx/lZO qWDJC7IHvFJ7Mw5KugacyF0SHeNdloyBM7KZW6HeXmgKn9IMJBpmwib92uUkRZJx D8j/bHHqD/zsgQ39nO+c4M0MmhO/DsPLG/dnGKrRCu7v1tmEnkY= =ZUuO -----END PGP SIGNATURE----- Merge tag 'ptrace_stop-cleanup-for-v5.19' of git://git.kernel.org/pub/scm/linux/kernel/git/ebiederm/user-namespace Pull ptrace_stop cleanups from Eric Biederman: "While looking at the ptrace problems with PREEMPT_RT and the problems Peter Zijlstra was encountering with ptrace in his freezer rewrite I identified some cleanups to ptrace_stop that make sense on their own and move make resolving the other problems much simpler. The biggest issue is the habit of the ptrace code to change task->__state from the tracer to suppress TASK_WAKEKILL from waking up the tracee. No other code in the kernel does that and it is straight forward to update signal_wake_up and friends to make that unnecessary. Peter's task freezer sets frozen tasks to a new state TASK_FROZEN and then it stores them by calling "wake_up_state(t, TASK_FROZEN)" relying on the fact that all stopped states except the special stop states can tolerate spurious wake up and recover their state. The state of stopped and traced tasked is changed to be stored in task->jobctl as well as in task->__state. This makes it possible for the freezer to recover tasks in these special states, as well as serving as a general cleanup. With a little more work in that direction I believe TASK_STOPPED can learn to tolerate spurious wake ups and become an ordinary stop state. The TASK_TRACED state has to remain a special state as the registers for a process are only reliably available when the process is stopped in the scheduler. Fundamentally ptrace needs acess to the saved register values of a task. There are bunch of semi-random ptrace related cleanups that were found while looking at these issues. One cleanup that deserves to be called out is from commit57b6de08b5
("ptrace: Admit ptrace_stop can generate spuriuos SIGTRAPs"). This makes a change that is technically user space visible, in the handling of what happens to a tracee when a tracer dies unexpectedly. According to our testing and our understanding of userspace nothing cares that spurious SIGTRAPs can be generated in that case" * tag 'ptrace_stop-cleanup-for-v5.19' of git://git.kernel.org/pub/scm/linux/kernel/git/ebiederm/user-namespace: sched,signal,ptrace: Rework TASK_TRACED, TASK_STOPPED state ptrace: Always take siglock in ptrace_resume ptrace: Don't change __state ptrace: Admit ptrace_stop can generate spuriuos SIGTRAPs ptrace: Document that wait_task_inactive can't fail ptrace: Reimplement PTRACE_KILL by always sending SIGKILL signal: Use lockdep_assert_held instead of assert_spin_locked ptrace: Remove arch_ptrace_attach ptrace/xtensa: Replace PT_SINGLESTEP with TIF_SINGLESTEP ptrace/um: Replace PT_DTRACE with TIF_SINGLESTEP signal: Replace __group_send_sig_info with send_signal_locked signal: Rename send_signal send_signal_locked
403 lines
8.6 KiB
C
403 lines
8.6 KiB
C
// SPDX-License-Identifier: GPL-2.0
|
|
/*
|
|
* Copyright (C) 2015 Anton Ivanov (aivanov@{brocade.com,kot-begemot.co.uk})
|
|
* Copyright (C) 2015 Thomas Meyer (thomas@m3y3r.de)
|
|
* Copyright (C) 2000 - 2007 Jeff Dike (jdike@{addtoit,linux.intel}.com)
|
|
* Copyright 2003 PathScale, Inc.
|
|
*/
|
|
|
|
#include <linux/stddef.h>
|
|
#include <linux/err.h>
|
|
#include <linux/hardirq.h>
|
|
#include <linux/mm.h>
|
|
#include <linux/module.h>
|
|
#include <linux/personality.h>
|
|
#include <linux/proc_fs.h>
|
|
#include <linux/ptrace.h>
|
|
#include <linux/random.h>
|
|
#include <linux/slab.h>
|
|
#include <linux/sched.h>
|
|
#include <linux/sched/debug.h>
|
|
#include <linux/sched/task.h>
|
|
#include <linux/sched/task_stack.h>
|
|
#include <linux/seq_file.h>
|
|
#include <linux/tick.h>
|
|
#include <linux/threads.h>
|
|
#include <linux/resume_user_mode.h>
|
|
#include <asm/current.h>
|
|
#include <asm/mmu_context.h>
|
|
#include <linux/uaccess.h>
|
|
#include <as-layout.h>
|
|
#include <kern_util.h>
|
|
#include <os.h>
|
|
#include <skas.h>
|
|
#include <registers.h>
|
|
#include <linux/time-internal.h>
|
|
|
|
/*
|
|
* This is a per-cpu array. A processor only modifies its entry and it only
|
|
* cares about its entry, so it's OK if another processor is modifying its
|
|
* entry.
|
|
*/
|
|
struct cpu_task cpu_tasks[NR_CPUS] = { [0 ... NR_CPUS - 1] = { -1, NULL } };
|
|
|
|
static inline int external_pid(void)
|
|
{
|
|
/* FIXME: Need to look up userspace_pid by cpu */
|
|
return userspace_pid[0];
|
|
}
|
|
|
|
int pid_to_processor_id(int pid)
|
|
{
|
|
int i;
|
|
|
|
for (i = 0; i < ncpus; i++) {
|
|
if (cpu_tasks[i].pid == pid)
|
|
return i;
|
|
}
|
|
return -1;
|
|
}
|
|
|
|
void free_stack(unsigned long stack, int order)
|
|
{
|
|
free_pages(stack, order);
|
|
}
|
|
|
|
unsigned long alloc_stack(int order, int atomic)
|
|
{
|
|
unsigned long page;
|
|
gfp_t flags = GFP_KERNEL;
|
|
|
|
if (atomic)
|
|
flags = GFP_ATOMIC;
|
|
page = __get_free_pages(flags, order);
|
|
|
|
return page;
|
|
}
|
|
|
|
static inline void set_current(struct task_struct *task)
|
|
{
|
|
cpu_tasks[task_thread_info(task)->cpu] = ((struct cpu_task)
|
|
{ external_pid(), task });
|
|
}
|
|
|
|
extern void arch_switch_to(struct task_struct *to);
|
|
|
|
void *__switch_to(struct task_struct *from, struct task_struct *to)
|
|
{
|
|
to->thread.prev_sched = from;
|
|
set_current(to);
|
|
|
|
switch_threads(&from->thread.switch_buf, &to->thread.switch_buf);
|
|
arch_switch_to(current);
|
|
|
|
return current->thread.prev_sched;
|
|
}
|
|
|
|
void interrupt_end(void)
|
|
{
|
|
struct pt_regs *regs = ¤t->thread.regs;
|
|
|
|
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);
|
|
}
|
|
|
|
int get_current_pid(void)
|
|
{
|
|
return task_pid_nr(current);
|
|
}
|
|
|
|
/*
|
|
* This is called magically, by its address being stuffed in a jmp_buf
|
|
* and being longjmp-d to.
|
|
*/
|
|
void new_thread_handler(void)
|
|
{
|
|
int (*fn)(void *), n;
|
|
void *arg;
|
|
|
|
if (current->thread.prev_sched != NULL)
|
|
schedule_tail(current->thread.prev_sched);
|
|
current->thread.prev_sched = NULL;
|
|
|
|
fn = current->thread.request.u.thread.proc;
|
|
arg = current->thread.request.u.thread.arg;
|
|
|
|
/*
|
|
* callback returns only if the kernel thread execs a process
|
|
*/
|
|
n = fn(arg);
|
|
userspace(¤t->thread.regs.regs, current_thread_info()->aux_fp_regs);
|
|
}
|
|
|
|
/* Called magically, see new_thread_handler above */
|
|
void fork_handler(void)
|
|
{
|
|
force_flush_all();
|
|
|
|
schedule_tail(current->thread.prev_sched);
|
|
|
|
/*
|
|
* XXX: if interrupt_end() calls schedule, this call to
|
|
* arch_switch_to isn't needed. We could want to apply this to
|
|
* improve performance. -bb
|
|
*/
|
|
arch_switch_to(current);
|
|
|
|
current->thread.prev_sched = NULL;
|
|
|
|
userspace(¤t->thread.regs.regs, current_thread_info()->aux_fp_regs);
|
|
}
|
|
|
|
int copy_thread(struct task_struct * p, const struct kernel_clone_args *args)
|
|
{
|
|
unsigned long clone_flags = args->flags;
|
|
unsigned long sp = args->stack;
|
|
unsigned long tls = args->tls;
|
|
void (*handler)(void);
|
|
int ret = 0;
|
|
|
|
p->thread = (struct thread_struct) INIT_THREAD;
|
|
|
|
if (!args->fn) {
|
|
memcpy(&p->thread.regs.regs, current_pt_regs(),
|
|
sizeof(p->thread.regs.regs));
|
|
PT_REGS_SET_SYSCALL_RETURN(&p->thread.regs, 0);
|
|
if (sp != 0)
|
|
REGS_SP(p->thread.regs.regs.gp) = sp;
|
|
|
|
handler = fork_handler;
|
|
|
|
arch_copy_thread(¤t->thread.arch, &p->thread.arch);
|
|
} else {
|
|
get_safe_registers(p->thread.regs.regs.gp, p->thread.regs.regs.fp);
|
|
p->thread.request.u.thread.proc = args->fn;
|
|
p->thread.request.u.thread.arg = args->fn_arg;
|
|
handler = new_thread_handler;
|
|
}
|
|
|
|
new_thread(task_stack_page(p), &p->thread.switch_buf, handler);
|
|
|
|
if (!args->fn) {
|
|
clear_flushed_tls(p);
|
|
|
|
/*
|
|
* Set a new TLS for the child thread?
|
|
*/
|
|
if (clone_flags & CLONE_SETTLS)
|
|
ret = arch_set_tls(p, tls);
|
|
}
|
|
|
|
return ret;
|
|
}
|
|
|
|
void initial_thread_cb(void (*proc)(void *), void *arg)
|
|
{
|
|
int save_kmalloc_ok = kmalloc_ok;
|
|
|
|
kmalloc_ok = 0;
|
|
initial_thread_cb_skas(proc, arg);
|
|
kmalloc_ok = save_kmalloc_ok;
|
|
}
|
|
|
|
void um_idle_sleep(void)
|
|
{
|
|
if (time_travel_mode != TT_MODE_OFF)
|
|
time_travel_sleep();
|
|
else
|
|
os_idle_sleep();
|
|
}
|
|
|
|
void arch_cpu_idle(void)
|
|
{
|
|
cpu_tasks[current_thread_info()->cpu].pid = os_getpid();
|
|
um_idle_sleep();
|
|
raw_local_irq_enable();
|
|
}
|
|
|
|
int __cant_sleep(void) {
|
|
return in_atomic() || irqs_disabled() || in_interrupt();
|
|
/* Is in_interrupt() really needed? */
|
|
}
|
|
|
|
int user_context(unsigned long sp)
|
|
{
|
|
unsigned long stack;
|
|
|
|
stack = sp & (PAGE_MASK << CONFIG_KERNEL_STACK_ORDER);
|
|
return stack != (unsigned long) current_thread_info();
|
|
}
|
|
|
|
extern exitcall_t __uml_exitcall_begin, __uml_exitcall_end;
|
|
|
|
void do_uml_exitcalls(void)
|
|
{
|
|
exitcall_t *call;
|
|
|
|
call = &__uml_exitcall_end;
|
|
while (--call >= &__uml_exitcall_begin)
|
|
(*call)();
|
|
}
|
|
|
|
char *uml_strdup(const char *string)
|
|
{
|
|
return kstrdup(string, GFP_KERNEL);
|
|
}
|
|
EXPORT_SYMBOL(uml_strdup);
|
|
|
|
int copy_to_user_proc(void __user *to, void *from, int size)
|
|
{
|
|
return copy_to_user(to, from, size);
|
|
}
|
|
|
|
int copy_from_user_proc(void *to, void __user *from, int size)
|
|
{
|
|
return copy_from_user(to, from, size);
|
|
}
|
|
|
|
int clear_user_proc(void __user *buf, int size)
|
|
{
|
|
return clear_user(buf, size);
|
|
}
|
|
|
|
static atomic_t using_sysemu = ATOMIC_INIT(0);
|
|
int sysemu_supported;
|
|
|
|
void set_using_sysemu(int value)
|
|
{
|
|
if (value > sysemu_supported)
|
|
return;
|
|
atomic_set(&using_sysemu, value);
|
|
}
|
|
|
|
int get_using_sysemu(void)
|
|
{
|
|
return atomic_read(&using_sysemu);
|
|
}
|
|
|
|
static int sysemu_proc_show(struct seq_file *m, void *v)
|
|
{
|
|
seq_printf(m, "%d\n", get_using_sysemu());
|
|
return 0;
|
|
}
|
|
|
|
static int sysemu_proc_open(struct inode *inode, struct file *file)
|
|
{
|
|
return single_open(file, sysemu_proc_show, NULL);
|
|
}
|
|
|
|
static ssize_t sysemu_proc_write(struct file *file, const char __user *buf,
|
|
size_t count, loff_t *pos)
|
|
{
|
|
char tmp[2];
|
|
|
|
if (copy_from_user(tmp, buf, 1))
|
|
return -EFAULT;
|
|
|
|
if (tmp[0] >= '0' && tmp[0] <= '2')
|
|
set_using_sysemu(tmp[0] - '0');
|
|
/* We use the first char, but pretend to write everything */
|
|
return count;
|
|
}
|
|
|
|
static const struct proc_ops sysemu_proc_ops = {
|
|
.proc_open = sysemu_proc_open,
|
|
.proc_read = seq_read,
|
|
.proc_lseek = seq_lseek,
|
|
.proc_release = single_release,
|
|
.proc_write = sysemu_proc_write,
|
|
};
|
|
|
|
int __init make_proc_sysemu(void)
|
|
{
|
|
struct proc_dir_entry *ent;
|
|
if (!sysemu_supported)
|
|
return 0;
|
|
|
|
ent = proc_create("sysemu", 0600, NULL, &sysemu_proc_ops);
|
|
|
|
if (ent == NULL)
|
|
{
|
|
printk(KERN_WARNING "Failed to register /proc/sysemu\n");
|
|
return 0;
|
|
}
|
|
|
|
return 0;
|
|
}
|
|
|
|
late_initcall(make_proc_sysemu);
|
|
|
|
int singlestepping(void * t)
|
|
{
|
|
struct task_struct *task = t ? t : current;
|
|
|
|
if (!test_thread_flag(TIF_SINGLESTEP))
|
|
return 0;
|
|
|
|
if (task->thread.singlestep_syscall)
|
|
return 1;
|
|
|
|
return 2;
|
|
}
|
|
|
|
/*
|
|
* Only x86 and x86_64 have an arch_align_stack().
|
|
* All other arches have "#define arch_align_stack(x) (x)"
|
|
* in their asm/exec.h
|
|
* As this is included in UML from asm-um/system-generic.h,
|
|
* we can use it to behave as the subarch does.
|
|
*/
|
|
#ifndef arch_align_stack
|
|
unsigned long arch_align_stack(unsigned long sp)
|
|
{
|
|
if (!(current->personality & ADDR_NO_RANDOMIZE) && randomize_va_space)
|
|
sp -= get_random_int() % 8192;
|
|
return sp & ~0xf;
|
|
}
|
|
#endif
|
|
|
|
unsigned long __get_wchan(struct task_struct *p)
|
|
{
|
|
unsigned long stack_page, sp, ip;
|
|
bool seen_sched = 0;
|
|
|
|
stack_page = (unsigned long) task_stack_page(p);
|
|
/* Bail if the process has no kernel stack for some reason */
|
|
if (stack_page == 0)
|
|
return 0;
|
|
|
|
sp = p->thread.switch_buf->JB_SP;
|
|
/*
|
|
* Bail if the stack pointer is below the bottom of the kernel
|
|
* stack for some reason
|
|
*/
|
|
if (sp < stack_page)
|
|
return 0;
|
|
|
|
while (sp < stack_page + THREAD_SIZE) {
|
|
ip = *((unsigned long *) sp);
|
|
if (in_sched_functions(ip))
|
|
/* Ignore everything until we're above the scheduler */
|
|
seen_sched = 1;
|
|
else if (kernel_text_address(ip) && seen_sched)
|
|
return ip;
|
|
|
|
sp += sizeof(unsigned long);
|
|
}
|
|
|
|
return 0;
|
|
}
|
|
|
|
int elf_core_copy_fpregs(struct task_struct *t, elf_fpregset_t *fpu)
|
|
{
|
|
int cpu = current_thread_info()->cpu;
|
|
|
|
return save_i387_registers(userspace_pid[cpu], (unsigned long *) fpu);
|
|
}
|
|
|