minix/drivers/log/log.c
Cristiano Giuffrida cb176df60f New RS and new signal handling for system processes.
UPDATING INFO:
20100317:
        /usr/src/etc/system.conf updated to ignore default kernel calls: copy
        it (or merge it) to /etc/system.conf.
        The hello driver (/dev/hello) added to the distribution:
        # cd /usr/src/commands/scripts && make clean install
        # cd /dev && MAKEDEV hello

KERNEL CHANGES:
- Generic signal handling support. The kernel no longer assumes PM as a signal
manager for every process. The signal manager of a given process can now be
specified in its privilege slot. When a signal has to be delivered, the kernel
performs the lookup and forwards the signal to the appropriate signal manager.
PM is the default signal manager for user processes, RS is the default signal
manager for system processes. To enable ptrace()ing for system processes, it
is sufficient to change the default signal manager to PM. This will temporarily
disable crash recovery, though.
- sys_exit() is now split into sys_exit() (i.e. exit() for system processes,
which generates a self-termination signal), and sys_clear() (i.e. used by PM
to ask the kernel to clear a process slot when a process exits).
- Added a new kernel call (i.e. sys_update()) to swap two process slots and
implement live update.

PM CHANGES:
- Posix signal handling is no longer allowed for system processes. System
signals are split into two fixed categories: termination and non-termination
signals. When a non-termination signaled is processed, PM transforms the signal
into an IPC message and delivers the message to the system process. When a
termination signal is processed, PM terminates the process.
- PM no longer assumes itself as the signal manager for system processes. It now
makes sure that every system signal goes through the kernel before being
actually processes. The kernel will then dispatch the signal to the appropriate
signal manager which may or may not be PM.

SYSLIB CHANGES:
- Simplified SEF init and LU callbacks.
- Added additional predefined SEF callbacks to debug crash recovery and
live update.
- Fixed a temporary ack in the SEF init protocol. SEF init reply is now
completely synchronous.
- Added SEF signal event type to provide a uniform interface for system
processes to deal with signals. A sef_cb_signal_handler() callback is
available for system processes to handle every received signal. A
sef_cb_signal_manager() callback is used by signal managers to process
system signals on behalf of the kernel.
- Fixed a few bugs with memory mapping and DS.

VM CHANGES:
- Page faults and memory requests coming from the kernel are now implemented
using signals.
- Added a new VM call to swap two process slots and implement live update.
- The call is used by RS at update time and in turn invokes the kernel call
sys_update().

RS CHANGES:
- RS has been reworked with a better functional decomposition.
- Better kernel call masks. com.h now defines the set of very basic kernel calls
every system service is allowed to use. This makes system.conf simpler and
easier to maintain. In addition, this guarantees a higher level of isolation
for system libraries that use one or more kernel calls internally (e.g. printf).
- RS is the default signal manager for system processes. By default, RS
intercepts every signal delivered to every system process. This makes crash
recovery possible before bringing PM and friends in the loop.
- RS now supports fast rollback when something goes wrong while initializing
the new version during a live update.
- Live update is now implemented by keeping the two versions side-by-side and
swapping the process slots when the old version is ready to update.
- Crash recovery is now implemented by keeping the two versions side-by-side
and cleaning up the old version only when the recovery process is complete.

DS CHANGES:
- Fixed a bug when the process doing ds_publish() or ds_delete() is not known
by DS.
- Fixed the completely broken support for strings. String publishing is now
implemented in the system library and simply wraps publishing of memory ranges.
Ideally, we should adopt a similar approach for other data types as well.
- Test suite fixed.

DRIVER CHANGES:
- The hello driver has been added to the Minix distribution to demonstrate basic
live update and crash recovery functionalities.
- Other drivers have been adapted to conform the new SEF interface.
2010-03-17 01:15:29 +00:00

504 lines
14 KiB
C

/* This file contains a driver for:
* /dev/klog - system log device
*
* Changes:
* 21 July 2005 - Support for diagnostic messages (Jorrit N. Herder)
* 7 July 2005 - Created (Ben Gras)
*/
#include "log.h"
#include <sys/time.h>
#include <sys/select.h>
#include <minix/endpoint.h>
#define LOG_DEBUG 0 /* enable/ disable debugging */
#define NR_DEVS 1 /* number of minor devices */
#define MINOR_KLOG 0 /* /dev/klog */
#define LOGINC(n, i) do { (n) = (((n) + (i)) % LOG_SIZE); } while(0)
PUBLIC struct logdevice logdevices[NR_DEVS];
PRIVATE struct device log_geom[NR_DEVS]; /* base and size of devices */
PRIVATE int log_device = -1; /* current device */
FORWARD _PROTOTYPE( char *log_name, (void) );
FORWARD _PROTOTYPE( struct device *log_prepare, (int device) );
FORWARD _PROTOTYPE( int log_transfer, (int proc_nr, int opcode, u64_t position,
iovec_t *iov, unsigned nr_req) );
FORWARD _PROTOTYPE( int log_do_open, (struct driver *dp, message *m_ptr) );
FORWARD _PROTOTYPE( int log_cancel, (struct driver *dp, message *m_ptr) );
FORWARD _PROTOTYPE( int log_select, (struct driver *dp, message *m_ptr) );
FORWARD _PROTOTYPE( int log_other, (struct driver *dp, message *m_ptr) );
FORWARD _PROTOTYPE( void log_geometry, (struct partition *entry) );
FORWARD _PROTOTYPE( int subread, (struct logdevice *log, int count, int proc_nr, vir_bytes user_vir, size_t) );
/* Entry points to this driver. */
PRIVATE struct driver log_dtab = {
log_name, /* current device's name */
log_do_open, /* open or mount */
do_nop, /* nothing on a close */
nop_ioctl, /* ioctl nop */
log_prepare, /* prepare for I/O on a given minor device */
log_transfer, /* do the I/O */
nop_cleanup, /* no need to clean up */
log_geometry, /* geometry */
nop_alarm, /* no alarm */
log_cancel, /* CANCEL request */
log_select, /* DEV_SELECT request */
log_other, /* Unrecognized messages */
NULL /* HW int */
};
extern int device_caller;
/* SEF functions and variables. */
FORWARD _PROTOTYPE( void sef_local_startup, (void) );
FORWARD _PROTOTYPE( int sef_cb_init_fresh, (int type, sef_init_info_t *info) );
EXTERN _PROTOTYPE( int sef_cb_lu_prepare, (int state) );
EXTERN _PROTOTYPE( int sef_cb_lu_state_isvalid, (int state) );
EXTERN _PROTOTYPE( void sef_cb_lu_state_dump, (int state) );
FORWARD _PROTOTYPE( void sef_cb_signal_handler, (int signo) );
/*===========================================================================*
* main *
*===========================================================================*/
PUBLIC int main(void)
{
/* SEF local startup. */
sef_local_startup();
/* Call the generic receive loop. */
driver_task(&log_dtab, DRIVER_ASYN);
return(OK);
}
/*===========================================================================*
* sef_local_startup *
*===========================================================================*/
PRIVATE void sef_local_startup()
{
/* Register init callbacks. */
sef_setcb_init_fresh(sef_cb_init_fresh);
sef_setcb_init_lu(sef_cb_init_fresh);
sef_setcb_init_restart(sef_cb_init_fresh);
/* Register live update callbacks. */
sef_setcb_lu_prepare(sef_cb_lu_prepare);
sef_setcb_lu_state_isvalid(sef_cb_lu_state_isvalid);
sef_setcb_lu_state_dump(sef_cb_lu_state_dump);
/* Register signal callbacks. */
sef_setcb_signal_handler(sef_cb_signal_handler);
/* Let SEF perform startup. */
sef_startup();
}
/*===========================================================================*
* sef_cb_init_fresh *
*===========================================================================*/
PRIVATE int sef_cb_init_fresh(int type, sef_init_info_t *info)
{
/* Initialize the log driver. */
int i;
/* Initialize log devices. */
for(i = 0; i < NR_DEVS; i++) {
log_geom[i].dv_size = cvul64(LOG_SIZE);
log_geom[i].dv_base = cvul64((long)logdevices[i].log_buffer);
logdevices[i].log_size = logdevices[i].log_read =
logdevices[i].log_write =
logdevices[i].log_select_alerted =
logdevices[i].log_selected =
logdevices[i].log_select_ready_ops = 0;
logdevices[i].log_proc_nr = 0;
logdevices[i].log_revive_alerted = 0;
}
return(OK);
}
/*===========================================================================*
* sef_cb_signal_handler *
*===========================================================================*/
PRIVATE void sef_cb_signal_handler(int signo)
{
/* Only check for a pending message from the kernel, ignore anything else. */
if (signo != SIGKMESS) return;
do_new_kmess(SYSTEM);
}
/*===========================================================================*
* log_name *
*===========================================================================*/
PRIVATE char *log_name()
{
/* Return a name for the current device. */
static char name[] = "log";
return name;
}
/*===========================================================================*
* log_prepare *
*===========================================================================*/
PRIVATE struct device *log_prepare(device)
int device;
{
/* Prepare for I/O on a device: check if the minor device number is ok. */
if (device < 0 || device >= NR_DEVS) return(NIL_DEV);
log_device = device;
return(&log_geom[device]);
}
/*===========================================================================*
* subwrite *
*===========================================================================*/
PRIVATE int
subwrite(struct logdevice *log, int count, int proc_nr,
vir_bytes user_vir, size_t offset)
{
int d, r;
char *buf;
message m;
if (log->log_write + count > LOG_SIZE)
count = LOG_SIZE - log->log_write;
buf = log->log_buffer + log->log_write;
if(proc_nr == SELF) {
memcpy(buf, (char *) user_vir, count);
}
else {
if((r=sys_safecopyfrom(proc_nr, user_vir, offset,
(vir_bytes)buf, count, D)) != OK)
return r;
}
LOGINC(log->log_write, count);
log->log_size += count;
if(log->log_size > LOG_SIZE) {
int overflow;
overflow = log->log_size - LOG_SIZE;
log->log_size -= overflow;
LOGINC(log->log_read, overflow);
}
if(log->log_size > 0 && log->log_proc_nr && !log->log_revive_alerted) {
/* Someone who was suspended on read can now
* be revived.
*/
log->log_status = subread(log, log->log_iosize,
log->log_proc_nr, log->log_user_vir_g,
log->log_user_vir_offset);
m.m_type = DEV_REVIVE;
m.REP_ENDPT = log->log_proc_nr;
m.REP_STATUS = log->log_status;
m.REP_IO_GRANT = log->log_user_vir_g;
r= send(log->log_source, &m);
if (r != OK)
{
printf("log`subwrite: send to %d failed: %d\n",
log->log_source, r);
}
log->log_proc_nr = 0;
}
if(log->log_size > 0)
log->log_select_ready_ops |= SEL_RD;
if(log->log_size > 0 && log->log_selected &&
!(log->log_select_alerted)) {
/* Someone(s) who was/were select()ing can now
* be awoken. If there was a blocking read (above),
* this can only happen if the blocking read didn't
* swallow all the data (log_size > 0).
*/
if(log->log_selected & SEL_RD) {
d= log-logdevices;
m.m_type = DEV_SEL_REPL2;
m.DEV_SEL_OPS = log->log_select_ready_ops;
m.DEV_MINOR = d;
#if LOG_DEBUG
printf("select sending DEV_SEL_REPL2\n");
#endif
r= send(log->log_select_proc, &m);
if (r != OK)
{
printf(
"log`subwrite: send to %d failed: %d\n",
log->log_select_proc, r);
}
log->log_selected &= ~log->log_select_ready_ops;
}
}
return count;
}
/*===========================================================================*
* log_append *
*===========================================================================*/
PUBLIC void
log_append(char *buf, int count)
{
int w = 0, skip = 0;
if(count < 1) return;
if(count > LOG_SIZE) skip = count - LOG_SIZE;
count -= skip;
buf += skip;
w = subwrite(&logdevices[0], count, SELF, (vir_bytes) buf,0);
if(w > 0 && w < count)
subwrite(&logdevices[0], count-w, SELF, (vir_bytes) buf+w,0);
return;
}
/*===========================================================================*
* subread *
*===========================================================================*/
PRIVATE int
subread(struct logdevice *log, int count, int proc_nr,
vir_bytes user_vir, size_t offset)
{
char *buf;
int r;
if (count > log->log_size)
count = log->log_size;
if (log->log_read + count > LOG_SIZE)
count = LOG_SIZE - log->log_read;
buf = log->log_buffer + log->log_read;
if((r=sys_safecopyto(proc_nr, user_vir, offset,
(vir_bytes)buf, count, D)) != OK)
return r;
LOGINC(log->log_read, count);
log->log_size -= count;
return count;
}
/*===========================================================================*
* log_transfer *
*===========================================================================*/
PRIVATE int log_transfer(proc_nr, opcode, position, iov, nr_req)
int proc_nr; /* process doing the request */
int opcode; /* DEV_GATHER_S or DEV_SCATTER_S */
u64_t position; /* offset on device to read or write */
iovec_t *iov; /* pointer to read or write request vector */
unsigned nr_req; /* length of request vector */
{
/* Read or write one the driver's minor devices. */
unsigned count;
vir_bytes user_vir;
struct device *dv;
int accumulated_read = 0;
struct logdevice *log;
size_t vir_offset = 0;
if(log_device < 0 || log_device >= NR_DEVS)
return EIO;
/* Get minor device number and check for /dev/null. */
dv = &log_geom[log_device];
log = &logdevices[log_device];
while (nr_req > 0) {
/* How much to transfer and where to / from. */
count = iov->iov_size;
user_vir = iov->iov_addr;
switch (log_device) {
case MINOR_KLOG:
if (opcode == DEV_GATHER_S) {
if (log->log_proc_nr || count < 1) {
/* There's already someone hanging to read, or
* no real I/O requested.
*/
return(OK);
}
if (!log->log_size) {
if(accumulated_read)
return OK;
/* No data available; let caller block. */
log->log_proc_nr = proc_nr;
log->log_iosize = count;
log->log_user_vir_g = user_vir;
log->log_user_vir_offset = 0;
log->log_revive_alerted = 0;
/* Device_caller is a global in drivers library. */
log->log_source = device_caller;
#if LOG_DEBUG
printf("blocked %d (%d)\n",
log->log_source, log->log_proc_nr);
#endif
return(EDONTREPLY);
}
count = subread(log, count, proc_nr, user_vir, vir_offset);
if(count < 0) {
return count;
}
accumulated_read += count;
} else {
count = subwrite(log, count, proc_nr, user_vir, vir_offset);
if(count < 0)
return count;
}
break;
/* Unknown (illegal) minor device. */
default:
return(EINVAL);
}
/* Book the number of bytes transferred. */
vir_offset += count;
if ((iov->iov_size -= count) == 0) { iov++; nr_req--; vir_offset = 0; }
}
return(OK);
}
/*============================================================================*
* log_do_open *
*============================================================================*/
PRIVATE int log_do_open(dp, m_ptr)
struct driver *dp;
message *m_ptr;
{
if (log_prepare(m_ptr->DEVICE) == NIL_DEV) return(ENXIO);
return(OK);
}
/*============================================================================*
* log_geometry *
*============================================================================*/
PRIVATE void log_geometry(entry)
struct partition *entry;
{
/* take a page from the fake memory device geometry */
entry->heads = 64;
entry->sectors = 32;
entry->cylinders = div64u(log_geom[log_device].dv_size, SECTOR_SIZE) /
(entry->heads * entry->sectors);
}
/*============================================================================*
* log_cancel *
*============================================================================*/
PRIVATE int log_cancel(dp, m_ptr)
struct driver *dp;
message *m_ptr;
{
int d;
d = m_ptr->TTY_LINE;
if(d < 0 || d >= NR_DEVS)
return EINVAL;
logdevices[d].log_proc_nr = 0;
logdevices[d].log_revive_alerted = 0;
return(OK);
}
/*============================================================================*
* log_other *
*============================================================================*/
PRIVATE int log_other(dp, m_ptr)
struct driver *dp;
message *m_ptr;
{
int r;
/* This function gets messages that the generic driver doesn't
* understand.
*/
if (is_notify(m_ptr->m_type)) {
switch (_ENDPOINT_P(m_ptr->m_source)) {
case TTY_PROC_NR:
do_new_kmess(m_ptr->m_source);
r = EDONTREPLY;
break;
default:
r = EINVAL;
break;
}
return r;
}
switch(m_ptr->m_type) {
case DIAGNOSTICS_OLD: {
r = do_diagnostics(m_ptr, 0);
break;
}
case ASYN_DIAGNOSTICS_OLD:
case DIAGNOSTICS_S_OLD:
r = do_diagnostics(m_ptr, 1);
break;
case DEV_STATUS: {
printf("log_other: unexpected DEV_STATUS request\n");
r = EDONTREPLY;
break;
}
default:
r = EINVAL;
break;
}
return r;
}
/*============================================================================*
* log_select *
*============================================================================*/
PRIVATE int log_select(dp, m_ptr)
struct driver *dp;
message *m_ptr;
{
int d, ready_ops = 0, ops = 0;
d = m_ptr->TTY_LINE;
if(d < 0 || d >= NR_DEVS) {
#if LOG_DEBUG
printf("line %d? EINVAL\n", d);
#endif
return EINVAL;
}
ops = m_ptr->IO_ENDPT & (SEL_RD|SEL_WR|SEL_ERR);
/* Read blocks when there is no log. */
if((m_ptr->IO_ENDPT & SEL_RD) && logdevices[d].log_size > 0) {
#if LOG_DEBUG
printf("log can read; size %d\n", logdevices[d].log_size);
#endif
ready_ops |= SEL_RD; /* writes never block */
}
/* Write never blocks. */
if(m_ptr->IO_ENDPT & SEL_WR) ready_ops |= SEL_WR;
/* Enable select calback if no operations were
* ready to go, but operations were requested,
* and notify was enabled.
*/
if((m_ptr->IO_ENDPT & SEL_NOTIFY) && ops && !ready_ops) {
logdevices[d].log_selected |= ops;
logdevices[d].log_select_proc = m_ptr->m_source;
#if LOG_DEBUG
printf("log setting selector.\n");
#endif
}
#if LOG_DEBUG
printf("log returning ops %d\n", ready_ops);
#endif
return(ready_ops);
}