minix/servers/procfs/pid.c
Ben Gras 50e2064049 No more intel/minix segments.
This commit removes all traces of Minix segments (the text/data/stack
memory map abstraction in the kernel) and significance of Intel segments
(hardware segments like CS, DS that add offsets to all addressing before
page table translation). This ultimately simplifies the memory layout
and addressing and makes the same layout possible on non-Intel
architectures.

There are only two types of addresses in the world now: virtual
and physical; even the kernel and processes have the same virtual
address space. Kernel and user processes can be distinguished at a
glance as processes won't use 0xF0000000 and above.

No static pre-allocated memory sizes exist any more.

Changes to booting:
        . The pre_init.c leaves the kernel and modules exactly as
          they were left by the bootloader in physical memory
        . The kernel starts running using physical addressing,
          loaded at a fixed location given in its linker script by the
          bootloader.  All code and data in this phase are linked to
          this fixed low location.
        . It makes a bootstrap pagetable to map itself to a
          fixed high location (also in linker script) and jumps to
          the high address. All code and data then use this high addressing.
        . All code/data symbols linked at the low addresses is prefixed by
          an objcopy step with __k_unpaged_*, so that that code cannot
          reference highly-linked symbols (which aren't valid yet) or vice
          versa (symbols that aren't valid any more).
        . The two addressing modes are separated in the linker script by
          collecting the unpaged_*.o objects and linking them with low
          addresses, and linking the rest high. Some objects are linked
          twice, once low and once high.
        . The bootstrap phase passes a lot of information (e.g. free memory
          list, physical location of the modules, etc.) using the kinfo
          struct.
        . After this bootstrap the low-linked part is freed.
        . The kernel maps in VM into the bootstrap page table so that VM can
          begin executing. Its first job is to make page tables for all other
          boot processes. So VM runs before RS, and RS gets a fully dynamic,
          VM-managed address space. VM gets its privilege info from RS as usual
          but that happens after RS starts running.
        . Both the kernel loading VM and VM organizing boot processes happen
	  using the libexec logic. This removes the last reason for VM to
	  still know much about exec() and vm/exec.c is gone.

Further Implementation:
        . All segments are based at 0 and have a 4 GB limit.
        . The kernel is mapped in at the top of the virtual address
          space so as not to constrain the user processes.
        . Processes do not use segments from the LDT at all; there are
          no segments in the LDT any more, so no LLDT is needed.
        . The Minix segments T/D/S are gone and so none of the
          user-space or in-kernel copy functions use them. The copy
          functions use a process endpoint of NONE to realize it's
          a physical address, virtual otherwise.
        . The umap call only makes sense to translate a virtual address
          to a physical address now.
        . Segments-related calls like newmap and alloc_segments are gone.
        . All segments-related translation in VM is gone (vir2map etc).
        . Initialization in VM is simpler as no moving around is necessary.
        . VM and all other boot processes can be linked wherever they wish
          and will be mapped in at the right location by the kernel and VM
          respectively.

Other changes:
        . The multiboot code is less special: it does not use mb_print
          for its diagnostics any more but uses printf() as normal, saving
          the output into the diagnostics buffer, only printing to the
          screen using the direct print functions if a panic() occurs.
        . The multiboot code uses the flexible 'free memory map list'
          style to receive the list of free memory if available.
        . The kernel determines the memory layout of the processes to
          a degree: it tells VM where the kernel starts and ends and
          where the kernel wants the top of the process to be. VM then
          uses this entire range, i.e. the stack is right at the top,
          and mmap()ped bits of memory are placed below that downwards,
          and the break grows upwards.

Other Consequences:
        . Every process gets its own page table as address spaces
          can't be separated any more by segments.
        . As all segments are 0-based, there is no distinction between
          virtual and linear addresses, nor between userspace and
          kernel addresses.
        . Less work is done when context switching, leading to a net
          performance increase. (8% faster on my machine for 'make servers'.)
	. The layout and configuration of the GDT makes sysenter and syscall
	  possible.
2012-07-15 22:30:15 +02:00

381 lines
10 KiB
C

/* ProcFS - pid.c - by Alen Stojanov and David van Moolenbroek */
#include "inc.h"
#include <sys/mman.h>
#include <minix/vm.h>
#define S_FRAME_SIZE 4096 /* use malloc if larger than this */
static char s_frame[S_FRAME_SIZE]; /* static storage for process frame */
static char *frame; /* pointer to process frame buffer */
static void pid_psinfo(int slot);
static void pid_cmdline(int slot);
static void pid_environ(int slot);
static void pid_map(int slot);
/* The files that are dynamically created in each PID directory. The data field
* contains each file's read function. Subdirectories are not yet supported.
*/
struct file pid_files[] = {
{ "psinfo", REG_ALL_MODE, (data_t) pid_psinfo },
{ "cmdline", REG_ALL_MODE, (data_t) pid_cmdline },
{ "environ", REG_ALL_MODE, (data_t) pid_environ },
{ "map", REG_ALL_MODE, (data_t) pid_map },
{ NULL, 0, (data_t) NULL }
};
/*===========================================================================*
* is_zombie *
*===========================================================================*/
static int is_zombie(int slot)
{
/* Is the given slot a zombie process?
*/
return (slot >= NR_TASKS &&
(mproc[slot - NR_TASKS].mp_flags & (TRACE_ZOMBIE | ZOMBIE)));
}
/*===========================================================================*
* pid_psinfo *
*===========================================================================*/
static void pid_psinfo(int i)
{
/* Print information used by ps(1) and top(1).
*/
int pi, task, state, type, p_state, f_state;
char name[PROC_NAME_LEN+1], *p;
struct vm_usage_info vui;
pid_t ppid;
pi = i - NR_TASKS;
task = proc[i].p_nr < 0;
/* Get the name of the process. Spaces would mess up the format.. */
if (task || mproc[i].mp_name[0] == 0)
strncpy(name, proc[i].p_name, sizeof(name) - 1);
else
strncpy(name, mproc[pi].mp_name, sizeof(name) - 1);
name[sizeof(name) - 1] = 0;
if ((p = strchr(name, ' ')) != NULL)
p[0] = 0;
/* Get the type of the process. */
if (task)
type = TYPE_TASK;
else if (mproc[i].mp_flags & PRIV_PROC)
type = TYPE_SYSTEM;
else
type = TYPE_USER;
/* Get the state of the process. */
if (!task) {
if (is_zombie(i))
state = STATE_ZOMBIE; /* zombie */
else if (mproc[pi].mp_flags & STOPPED)
state = STATE_STOP; /* stopped (traced) */
else if (proc[i].p_rts_flags == 0)
state = STATE_RUN; /* in run-queue */
else if (fp_is_blocked(&fproc[pi]) ||
(mproc[pi].mp_flags & (WAITING | PAUSED | SIGSUSPENDED)))
state = STATE_SLEEP; /* sleeping */
else
state = STATE_WAIT; /* waiting */
} else {
if (proc[i].p_rts_flags == 0)
state = STATE_RUN; /* in run-queue */
else
state = STATE_WAIT; /* other i.e. waiting */
}
/* We assume that even if a process has become a zombie, its kernel
* proc entry still contains the old (but valid) information. Currently
* this is true, but in the future we may have to filter some fields.
*/
buf_printf("%d %c %d %s %c %d %d %lu %lu %lu %lu",
PSINFO_VERSION, /* information version */
type, /* process type */
(int) proc[i].p_endpoint, /* process endpoint */
name, /* process name */
state, /* process state letter */
(int) P_BLOCKEDON(&proc[i]), /* endpt blocked on, or NONE */
(int) proc[i].p_priority, /* process priority */
(long) proc[i].p_user_time, /* user time */
(long) proc[i].p_sys_time, /* system time */
ex64hi(proc[i].p_cycles), /* execution cycles */
ex64lo(proc[i].p_cycles)
);
/* If the process is not a kernel task, we add some extra info. */
if (!task) {
memset(&vui, 0, sizeof(vui));
if (!is_zombie(i)) {
/* We don't care if this fails. */
(void) vm_info_usage(proc[i].p_endpoint, &vui);
}
if (mproc[pi].mp_flags & PAUSED)
p_state = PSTATE_PAUSED;
else if (mproc[pi].mp_flags & WAITING)
p_state = PSTATE_WAITING;
else if (mproc[pi].mp_flags & SIGSUSPENDED)
p_state = PSTATE_SIGSUSP;
else
p_state = '-';
if (mproc[pi].mp_parent == pi)
ppid = NO_PID;
else
ppid = mproc[mproc[pi].mp_parent].mp_pid;
switch (fproc[pi].fp_blocked_on) {
case FP_BLOCKED_ON_NONE: f_state = FSTATE_NONE; break;
case FP_BLOCKED_ON_PIPE: f_state = FSTATE_PIPE; break;
case FP_BLOCKED_ON_LOCK: f_state = FSTATE_LOCK; break;
case FP_BLOCKED_ON_POPEN: f_state = FSTATE_POPEN; break;
case FP_BLOCKED_ON_SELECT: f_state = FSTATE_SELECT; break;
case FP_BLOCKED_ON_DOPEN: f_state = FSTATE_DOPEN; break;
case FP_BLOCKED_ON_OTHER: f_state = FSTATE_TASK; break;
default: f_state = FSTATE_UNKNOWN;
}
buf_printf(" %lu %lu %lu %c %d %u %u %u %d %c %d %u",
vui.vui_total, /* total memory */
vui.vui_common, /* common memory */
vui.vui_shared, /* shared memory */
p_state, /* sleep state */
ppid, /* parent PID */
mproc[pi].mp_realuid, /* real UID */
mproc[pi].mp_effuid, /* effective UID */
mproc[pi].mp_procgrp, /* process group */
mproc[pi].mp_nice, /* nice value */
f_state, /* VFS block state */
(int) (fproc[pi].fp_blocked_on == FP_BLOCKED_ON_OTHER)
? fproc[pi].fp_task : NONE, /* block proc */
fproc[pi].fp_tty /* controlling tty */
);
}
/* always add kernel cycles */
buf_printf(" %lu %lu %lu %lu",
ex64hi(proc[i].p_kipc_cycles),
ex64lo(proc[i].p_kipc_cycles),
ex64hi(proc[i].p_kcall_cycles),
ex64lo(proc[i].p_kcall_cycles));
/* Newline at the end of the file. */
buf_printf("\n");
}
/*===========================================================================*
* put_frame *
*===========================================================================*/
static void put_frame(void)
{
/* If we allocated memory dynamically during a call to get_frame(),
* free it up here.
*/
if (frame != s_frame)
free(frame);
}
/*===========================================================================*
* get_frame *
*===========================================================================*/
static int get_frame(int slot, vir_bytes *basep, vir_bytes *sizep,
size_t *nargsp)
{
/* Get the execution frame from the top of the given process's stack.
* It may be very large, in which case we temporarily allocate memory
* for it (up to a certain size).
*/
vir_bytes base, size;
size_t nargs;
if (proc[slot].p_nr < 0 || is_zombie(slot))
return FALSE;
/* Get the frame base address and size. Limit the size to whatever we
* can handle. If our static buffer is not sufficiently large to store
* the entire frame, allocate memory dynamically. It is then later
* freed by put_frame().
*/
base = mproc[slot - NR_TASKS].mp_frame_addr;
size = mproc[slot - NR_TASKS].mp_frame_len;
if (size < sizeof(size_t)) return FALSE;
if (size > ARG_MAX) size = ARG_MAX;
if (size > sizeof(s_frame)) {
frame = malloc(size);
if (frame == NULL)
return FALSE;
}
else frame = s_frame;
/* Copy in the complete process frame. */
if (sys_datacopy(proc[slot].p_endpoint, base,
SELF, (vir_bytes) frame, (phys_bytes) size) != OK) {
put_frame();
return FALSE;
}
frame[size] = 0; /* terminate any last string */
nargs = * (size_t *) frame;
if (nargs < 1 || sizeof(size_t) + sizeof(char *) * (nargs + 1) > size) {
put_frame();
return FALSE;
}
*basep = base;
*sizep = size;
*nargsp = nargs;
/* The caller now has to called put_frame() to clean up. */
return TRUE;
}
/*===========================================================================*
* pid_cmdline *
*===========================================================================*/
static void pid_cmdline(int slot)
{
/* Dump the process's command line as it is contained in the process
* itself. Each argument is terminated with a null character.
*/
vir_bytes base, size, ptr;
size_t i, len, nargs;
char **argv;
if (!get_frame(slot, &base, &size, &nargs))
return;
argv = (char **) &frame[sizeof(size_t)];
for (i = 0; i < nargs; i++) {
ptr = (vir_bytes) argv[i] - base;
/* Check for bad pointers. */
if ((long) ptr < 0L || ptr >= size)
break;
len = strlen(&frame[ptr]) + 1;
buf_append(&frame[ptr], len);
}
put_frame();
}
/*===========================================================================*
* pid_environ *
*===========================================================================*/
static void pid_environ(int slot)
{
/* Dump the process's initial environment as it is contained in the
* process itself. Each entry is terminated with a null character.
*/
vir_bytes base, size, ptr;
size_t nargs, off, len;
char **envp;
if (!get_frame(slot, &base, &size, &nargs))
return;
off = sizeof(size_t) + sizeof(char *) * (nargs + 1);
envp = (char **) &frame[off];
for (;;) {
/* Make sure there is no buffer overrun. */
if (off + sizeof(char *) > size)
break;
ptr = (vir_bytes) *envp;
/* Stop at the terminating NULL pointer. */
if (ptr == 0L)
break;
ptr -= base;
/* Check for bad pointers. */
if ((long) ptr < 0L || ptr >= size)
break;
len = strlen(&frame[ptr]) + 1;
buf_append(&frame[ptr], len);
off += sizeof(char *);
envp++;
}
put_frame();
}
/*===========================================================================*
* dump_regions *
*===========================================================================*/
static int dump_regions(int slot)
{
/* Print the virtual memory regions of a process.
*/
struct vm_region_info vri[MAX_VRI_COUNT];
vir_bytes next;
int i, r, count;
count = 0;
next = 0;
do {
r = vm_info_region(proc[slot].p_endpoint, vri, MAX_VRI_COUNT,
&next);
if (r < 0)
return r;
if (r == 0)
break;
for (i = 0; i < r; i++) {
buf_printf("%08lx-%08lx %c%c%c %c\n",
vri[i].vri_addr, vri[i].vri_addr + vri[i].vri_length,
(vri[i].vri_prot & PROT_READ) ? 'r' : '-',
(vri[i].vri_prot & PROT_WRITE) ? 'w' : '-',
(vri[i].vri_prot & PROT_EXEC) ? 'x' : '-',
(vri[i].vri_flags & MAP_IPC_SHARED) ? 's' : 'p');
count++;
}
} while (r == MAX_VRI_COUNT);
return count;
}
/*===========================================================================*
* pid_map *
*===========================================================================*/
static void pid_map(int slot)
{
/* Print a memory map of the process. Obtain the information from VM if
* possible; otherwise fall back on segments from the kernel.
*/
/* Zombies have no memory. */
if (is_zombie(slot))
return;
/* Kernel tasks also have no memory. */
if (proc[slot].p_nr >= 0) {
if (dump_regions(slot) != 0)
return;
}
}