refactor: simplify the vmm design, single responsibility. But using it should with...
[lunaix-os.git] / lunaix-os / kernel / sched.c
index ed2cc74a2b02b813d9578d81af5eb36b88774d6c..a93c9bb5e37a87802df17a761039793bca0cb607 100644 (file)
@@ -1,17 +1,24 @@
-#include <lunaix/process.h>
-#include <lunaix/sched.h>
-#include <lunaix/mm/vmm.h>
-#include <hal/cpu.h>
 #include <arch/x86/interrupts.h>
+#include <arch/x86/tss.h>
+
 #include <hal/apic.h>
+#include <hal/cpu.h>
 
+#include <lunaix/mm/kalloc.h>
+#include <lunaix/mm/pmm.h>
+#include <lunaix/mm/vmm.h>
+#include <lunaix/process.h>
+#include <lunaix/sched.h>
+#include <lunaix/signal.h>
 #include <lunaix/spike.h>
 #include <lunaix/status.h>
+#include <lunaix/syscall.h>
 #include <lunaix/syslog.h>
 
 #define MAX_PROCESS 512
 
-struct proc_info* __current;
+volatile struct proc_info* __current;
+
 struct proc_info dummy;
 
 extern void __proc_table;
@@ -20,27 +27,46 @@ struct scheduler sched_ctx;
 
 LOG_MODULE("SCHED")
 
-void sched_init() {
+void
+sched_init()
+{
     size_t pg_size = ROUNDUP(sizeof(struct proc_info) * MAX_PROCESS, 0x1000);
-    assert_msg(
-        vmm_alloc_pages(KERNEL_PID, &__proc_table, pg_size, PG_PREM_RW, PP_FGPERSIST), 
-        "Fail to allocate proc table"
-    );
-    
-    sched_ctx = (struct scheduler) {
-        ._procs = (struct proc_info*) &__proc_table,
-        .ptable_len = 0,
-        .procs_index = 0
-    };
-
-    __current = &dummy;
+
+    for (size_t i = 0; i <= pg_size; i += 4096) {
+        uintptr_t pa = pmm_alloc_page(KERNEL_PID, PP_FGPERSIST);
+        vmm_set_mapping(PD_REFERENCED, &__proc_table + i, pa, PG_PREM_RW);
+    }
+
+    sched_ctx = (struct scheduler){ ._procs = (struct proc_info*)&__proc_table,
+                                    .ptable_len = 0,
+                                    .procs_index = 0 };
 }
 
-void schedule() {
+void
+run(struct proc_info* proc)
+{
+    if (!(__current->state & ~PROC_RUNNING)) {
+        __current->state = PROC_STOPPED;
+    }
+    proc->state = PROC_RUNNING;
+
+    // FIXME: 这里还是得再考虑一下。
+    // tss_update_esp(__current->intr_ctx.esp);
+    apic_done_servicing();
+
+    asm volatile("pushl %0\n"
+                 "jmp switch_to\n" ::"r"(proc)); // kernel/asm/x86/interrupt.S
+}
+
+void
+schedule()
+{
     if (!sched_ctx.ptable_len) {
         return;
     }
 
+    // 上下文切换相当的敏感!我们不希望任何的中断打乱栈的顺序……
+    cpu_disable_interrupt();
     struct proc_info* next;
     int prev_ptr = sched_ctx.procs_index;
     int ptr = prev_ptr;
@@ -48,73 +74,210 @@ void schedule() {
     do {
         ptr = (ptr + 1) % sched_ctx.ptable_len;
         next = &sched_ctx._procs[ptr];
-    } while((next->state != PROC_STOPPED && next->state != PROC_CREATED) && ptr != prev_ptr);
-    
+    } while (next->state != PROC_STOPPED && ptr != prev_ptr);
+
     sched_ctx.procs_index = ptr;
-    
-    __current->state = PROC_STOPPED;
-    next->state = PROC_RUNNING;
-    
-    __current = next;
 
-    cpu_lcr3(__current->page_table);
+    run(next);
+}
 
-    apic_done_servicing();
+static void
+proc_timer_callback(struct proc_info* proc)
+{
+    proc->timer = NULL;
+    proc->state = PROC_STOPPED;
+}
+
+__DEFINE_LXSYSCALL1(unsigned int, sleep, unsigned int, seconds)
+{
+    // FIXME: sleep的实现或许需要改一下。专门绑一个计时器好像没有必要……
+    if (!seconds) {
+        return 0;
+    }
+
+    if (__current->timer) {
+        return __current->timer->counter / timer_context()->running_frequency;
+    }
+
+    struct lx_timer* timer =
+      timer_run_second(seconds, proc_timer_callback, __current, 0);
+    __current->timer = timer;
+    __current->intr_ctx.registers.eax = seconds;
+    __current->state = PROC_BLOCKED;
+    schedule();
+}
+
+__DEFINE_LXSYSCALL1(void, exit, int, status)
+{
+    terminate_proc(status);
+}
+
+__DEFINE_LXSYSCALL(void, yield)
+{
+    schedule();
+}
+
+pid_t
+_wait(pid_t wpid, int* status, int options);
+
+__DEFINE_LXSYSCALL1(pid_t, wait, int*, status)
+{
+    return _wait(-1, status, 0);
+}
+
+__DEFINE_LXSYSCALL3(pid_t, waitpid, pid_t, pid, int*, status, int, options)
+{
+    return _wait(pid, status, options);
+}
+
+pid_t
+_wait(pid_t wpid, int* status, int options)
+{
+    pid_t cur = __current->pid;
+    int status_flags = 0;
+    struct proc_info *proc, *n;
+    if (llist_empty(&__current->children)) {
+        return -1;
+    }
 
-    asm volatile ("pushl %0\n jmp soft_iret\n"::"r"(&__current->intr_ctx): "memory");
+    wpid = wpid ? wpid : -__current->pgid;
+    cpu_enable_interrupt();
+repeat:
+    llist_for_each(proc, n, &__current->children, siblings)
+    {
+        if (!~wpid || proc->pid == wpid || proc->pgid == -wpid) {
+            if (proc->state == PROC_TERMNAT && !options) {
+                status_flags |= PROCTERM;
+                goto done;
+            }
+            if (proc->state == PROC_STOPPED && (options & WUNTRACED)) {
+                status_flags |= PROCSTOP;
+                goto done;
+            }
+        }
+    }
+    if ((options & WNOHANG)) {
+        return 0;
+    }
+    // 放弃当前的运行机会
+    sched_yield();
+    goto repeat;
+
+done:
+    cpu_disable_interrupt();
+    *status = (proc->exit_code & 0xffff) | status_flags;
+    return destroy_process(proc->pid);
 }
 
-pid_t alloc_pid() {
+pid_t
+alloc_pid()
+{
     pid_t i = 0;
-    for (; i < sched_ctx.ptable_len && sched_ctx._procs[i].state != PROC_DESTROY; i++);
+    for (;
+         i < sched_ctx.ptable_len && sched_ctx._procs[i].state != PROC_DESTROY;
+         i++)
+        ;
 
     if (i == MAX_PROCESS) {
-        __current->k_status = LXPROCFULL;
-        return -1;
+        panick("Panic in Ponyville shimmer!");
     }
-    return i + 1;
+    return i;
 }
 
-void push_process(struct proc_info* process) {
-    int index = process->pid - 1;
+void
+push_process(struct proc_info* process)
+{
+    int index = process->pid;
     if (index < 0 || index > sched_ctx.ptable_len) {
         __current->k_status = LXINVLDPID;
         return;
     }
-    
+
     if (index == sched_ctx.ptable_len) {
         sched_ctx.ptable_len++;
     }
-    
-    process->parent = __current->pid;
-    process->state = PROC_CREATED;
 
     sched_ctx._procs[index] = *process;
+
+    process = &sched_ctx._procs[index];
+
+    // make sure the reference is relative to process table
+    llist_init_head(&process->children);
+    llist_init_head(&process->grp_member);
+
+    // every process is the child of first process (pid=1)
+    if (process->parent) {
+        llist_append(&process->parent->children, &process->siblings);
+    } else {
+        process->parent = &sched_ctx._procs[0];
+    }
+
+    process->state = PROC_STOPPED;
 }
 
-void destroy_process(pid_t pid) {
-    int index = pid - 1;
-    if (index < 0 || index > sched_ctx.ptable_len) {
+// from <kernel/process.c>
+extern void
+__del_pagetable(pid_t pid, uintptr_t mount_point);
+
+pid_t
+destroy_process(pid_t pid)
+{
+    int index = pid;
+    if (index <= 0 || index > sched_ctx.ptable_len) {
         __current->k_status = LXINVLDPID;
         return;
     }
+    struct proc_info* proc = &sched_ctx._procs[index];
+    proc->state = PROC_DESTROY;
+    llist_delete(&proc->siblings);
 
-    sched_ctx._procs[index].state = PROC_DESTROY;
+    if (proc->mm.regions) {
+        struct mm_region *pos, *n;
+        llist_for_each(pos, n, &proc->mm.regions->head, head)
+        {
+            lxfree(pos);
+        }
+    }
+
+    vmm_mount_pd(PD_MOUNT_2, proc->page_table);
+
+    __del_pagetable(pid, PD_MOUNT_2);
+
+    vmm_unmount_pd(PD_MOUNT_2);
 
-    // TODO: recycle the physical pages used by page tables
+    return pid;
 }
 
-void terminate_process(int exit_code) {
+void
+terminate_proc(int exit_code)
+{
     __current->state = PROC_TERMNAT;
     __current->exit_code = exit_code;
 
     schedule();
 }
 
-struct proc_info* get_process(pid_t pid) {
-    int index = pid - 1;
+struct proc_info*
+get_process(pid_t pid)
+{
+    int index = pid;
     if (index < 0 || index > sched_ctx.ptable_len) {
         return NULL;
     }
     return &sched_ctx._procs[index];
+}
+
+int
+orphaned_proc(pid_t pid)
+{
+    if (!pid)
+        return 0;
+    if (pid >= sched_ctx.ptable_len)
+        return 0;
+    struct proc_info* proc = &sched_ctx._procs[pid];
+    struct proc_info* parent = proc->parent;
+
+    // 如果其父进程的状态是terminated 或 destroy中的一种
+    // 或者其父进程是在该进程之后创建的,那么该进程为孤儿进程
+    return (parent->state & PROC_TERMMASK) || parent->created > proc->created;
 }
\ No newline at end of file