minix/kernel/clock.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

247 lines
7.8 KiB
C

/* This file contains the clock task, which handles time related functions.
* Important events that are handled by the CLOCK include setting and
* monitoring alarm timers and deciding when to (re)schedule processes.
* The CLOCK offers a direct interface to kernel processes. System services
* can access its services through system calls, such as sys_setalarm(). The
* CLOCK task thus is hidden from the outside world.
*
* Changes:
* Aug 18, 2006 removed direct hardware access etc, MinixPPC (Ingmar Alting)
* Oct 08, 2005 reordering and comment editing (A. S. Woodhull)
* Mar 18, 2004 clock interface moved to SYSTEM task (Jorrit N. Herder)
* Sep 30, 2004 source code documentation updated (Jorrit N. Herder)
* Sep 24, 2004 redesigned alarm timers (Jorrit N. Herder)
*
* Clock task is notified by the clock's interrupt handler when a timer
* has expired.
*
* In addition to the main clock_task() entry point, which starts the main
* loop, there are several other minor entry points:
* clock_stop: called just before MINIX shutdown
* get_uptime: get realtime since boot in clock ticks
* set_timer: set a watchdog timer (+)
* reset_timer: reset a watchdog timer (+)
* read_clock: read the counter of channel 0 of the 8253A timer
*
* (+) The CLOCK task keeps tracks of watchdog timers for the entire kernel.
* It is crucial that watchdog functions not block, or the CLOCK task may
* be blocked. Do not send() a message when the receiver is not expecting it.
* Instead, notify(), which always returns, should be used.
*/
#include "kernel.h"
#include "proc.h"
#include <minix/endpoint.h>
#include <assert.h>
#include "clock.h"
#ifdef USE_WATCHDOG
#include "watchdog.h"
#endif
/* Function prototype for PRIVATE functions.
*/
static void load_update(void);
/* The CLOCK's timers queue. The functions in <timers.h> operate on this.
* Each system process possesses a single synchronous alarm timer. If other
* kernel parts want to use additional timers, they must declare their own
* persistent (static) timer structure, which can be passed to the clock
* via (re)set_timer().
* When a timer expires its watchdog function is run by the CLOCK task.
*/
static timer_t *clock_timers; /* queue of CLOCK timers */
static clock_t next_timeout; /* realtime that next timer expires */
/* The time is incremented by the interrupt handler on each clock tick.
*/
static clock_t realtime = 0; /* real time clock */
/*
* The boot processos timer interrupt handler. In addition to non-boot cpus it
* keeps real time and notifies the clock task if need be
*/
int timer_int_handler(void)
{
/* Update user and system accounting times. Charge the current process
* for user time. If the current process is not billable, that is, if a
* non-user process is running, charge the billable process for system
* time as well. Thus the unbillable process' user time is the billable
* user's system time.
*/
struct proc * p, * billp;
/* FIXME watchdog for slave cpus! */
#ifdef USE_WATCHDOG
/*
* we need to know whether local timer ticks are happening or whether
* the kernel is locked up. We don't care about overflows as we only
* need to know that it's still ticking or not
*/
watchdog_local_timer_ticks++;
#endif
if (cpu_is_bsp(cpuid))
realtime++;
/* Update user and system accounting times. Charge the current process
* for user time. If the current process is not billable, that is, if a
* non-user process is running, charge the billable process for system
* time as well. Thus the unbillable process' user time is the billable
* user's system time.
*/
p = get_cpulocal_var(proc_ptr);
billp = get_cpulocal_var(bill_ptr);
p->p_user_time++;
if (! (priv(p)->s_flags & BILLABLE)) {
billp->p_sys_time++;
}
/* Decrement virtual timers, if applicable. We decrement both the
* virtual and the profile timer of the current process, and if the
* current process is not billable, the timer of the billed process as
* well. If any of the timers expire, do_clocktick() will send out
* signals.
*/
if ((p->p_misc_flags & MF_VIRT_TIMER)){
p->p_virt_left--;
}
if ((p->p_misc_flags & MF_PROF_TIMER)){
p->p_prof_left--;
}
if (! (priv(p)->s_flags & BILLABLE) &&
(billp->p_misc_flags & MF_PROF_TIMER)){
billp->p_prof_left--;
}
/*
* Check if a process-virtual timer expired. Check current process, but
* also bill_ptr - one process's user time is another's system time, and
* the profile timer decreases for both!
*/
vtimer_check(p);
if (p != billp)
vtimer_check(billp);
/* Update load average. */
load_update();
if (cpu_is_bsp(cpuid)) {
/* if a timer expired, notify the clock task */
if ((next_timeout <= realtime)) {
tmrs_exptimers(&clock_timers, realtime, NULL);
next_timeout = (clock_timers == NULL) ?
TMR_NEVER : clock_timers->tmr_exp_time;
}
#ifdef DEBUG_SERIAL
if (kinfo.do_serial_debug)
do_ser_debug();
#endif
}
return(1); /* reenable interrupts */
}
/*===========================================================================*
* get_uptime *
*===========================================================================*/
clock_t get_uptime(void)
{
/* Get and return the current clock uptime in ticks. */
return(realtime);
}
/*===========================================================================*
* set_timer *
*===========================================================================*/
void set_timer(tp, exp_time, watchdog)
struct timer *tp; /* pointer to timer structure */
clock_t exp_time; /* expiration realtime */
tmr_func_t watchdog; /* watchdog to be called */
{
/* Insert the new timer in the active timers list. Always update the
* next timeout time by setting it to the front of the active list.
*/
tmrs_settimer(&clock_timers, tp, exp_time, watchdog, NULL);
next_timeout = clock_timers->tmr_exp_time;
}
/*===========================================================================*
* reset_timer *
*===========================================================================*/
void reset_timer(tp)
struct timer *tp; /* pointer to timer structure */
{
/* The timer pointed to by 'tp' is no longer needed. Remove it from both the
* active and expired lists. Always update the next timeout time by setting
* it to the front of the active list.
*/
tmrs_clrtimer(&clock_timers, tp, NULL);
next_timeout = (clock_timers == NULL) ?
TMR_NEVER : clock_timers->tmr_exp_time;
}
/*===========================================================================*
* load_update *
*===========================================================================*/
static void load_update(void)
{
u16_t slot;
int enqueued = 0, q;
struct proc *p;
struct proc **rdy_head;
/* Load average data is stored as a list of numbers in a circular
* buffer. Each slot accumulates _LOAD_UNIT_SECS of samples of
* the number of runnable processes. Computations can then
* be made of the load average over variable periods, in the
* user library (see getloadavg(3)).
*/
slot = (realtime / system_hz / _LOAD_UNIT_SECS) % _LOAD_HISTORY;
if(slot != kloadinfo.proc_last_slot) {
kloadinfo.proc_load_history[slot] = 0;
kloadinfo.proc_last_slot = slot;
}
rdy_head = get_cpulocal_var(run_q_head);
/* Cumulation. How many processes are ready now? */
for(q = 0; q < NR_SCHED_QUEUES; q++) {
for(p = rdy_head[q]; p != NULL; p = p->p_nextready) {
enqueued++;
}
}
kloadinfo.proc_load_history[slot] += enqueued;
/* Up-to-dateness. */
kloadinfo.last_clock = realtime;
}
int boot_cpu_init_timer(unsigned freq)
{
if (init_local_timer(freq))
return -1;
if (register_local_timer_handler(
(irq_handler_t) timer_int_handler))
return -1;
return 0;
}
int app_cpu_init_timer(unsigned freq)
{
if (init_local_timer(freq))
return -1;
return 0;
}