minix/lib/libsys/safecopies.c

338 lines
7.1 KiB
C
Raw Normal View History

/* Library functions to maintain internal data copying tables.
*
* April 21 2006: Initial version (Ben Gras)
*
*/
#include <lib.h>
#include <errno.h>
#include <minix/sysutil.h>
#include <assert.h>
#include <stdlib.h>
#include <minix/syslib.h>
#include <minix/safecopies.h>
#include <minix/com.h>
#include <string.h>
#define ACCESS_CHECK(a) { \
make vfs & filesystems use failable copying Change the kernel to add features to vircopy and safecopies so that transparent copy fixing won't happen to avoid deadlocks, and such copies fail with EFAULT. Transparently making copying work from filesystems (as normally done by the kernel & VM when copying fails because of missing/readonly memory) is problematic as it can happen that, for file-mapped ranges, that that same filesystem that is blocked on the copy request is needed to satisfy the memory range, leading to deadlock. Dito for VFS itself, if done with a blocking call. This change makes the copying done from a filesystem fail in such cases with EFAULT by VFS adding the CPF_TRY flag to the grants. If a FS call fails with EFAULT, VFS will then request the range to be made available to VM after the FS is unblocked, allowing it to be used to satisfy the range if need be in another VFS thread. Similarly, for datacopies that VFS itself does, it uses the failable vircopy variant and callers use a wrapper that talk to VM if necessary to get the copy to work. . kernel: add CPF_TRY flag to safecopies . kernel: only request writable ranges to VM for the target buffer when copying fails . do copying in VFS TRY-first . some fixes in VM to build SANITYCHECK mode . add regression test for the cases where - a FS system call needs memory mapped in a process that the FS itself must map. - such a range covers more than one file-mapped region. . add 'try' mode to vircopy, physcopy . add flags field to copy kernel call messages . if CP_FLAG_TRY is set, do not transparently try to fix memory ranges . for use by VFS when accessing user buffers to avoid deadlock . remove some obsolete backwards compatability assignments . VFS: let thread scheduling work for VM requests too Allows VFS to make calls to VM while suspending and resuming the currently running thread. Does currently not work for the main thread. . VM: add fix memory range call for use by VFS Change-Id: I295794269cea51a3163519a9cfe5901301d90b32
2014-01-16 14:22:13 +01:00
if((a) & ~(CPF_READ|CPF_WRITE|CPF_TRY)) { \
errno = EINVAL; \
return -1; \
} \
}
#define GID_CHECK(gid) { \
if(!GRANT_VALID(gid) || (gid) < 0 || (gid) >= ngrants) {\
errno = EINVAL; \
return -1; \
} \
}
#define GID_CHECK_USED(gid) { \
GID_CHECK(gid); \
if(!(grants[gid].cp_flags & CPF_USED)) { \
errno = EINVAL; \
return -1; \
} \
}
#define CLICK_ALIGNMENT_CHECK(addr, bytes) { \
if(((vir_bytes)(addr) % CLICK_SIZE != 0) \
|| ((vir_bytes)(bytes) % CLICK_SIZE != 0)) { \
return EINVAL; \
} \
}
Initialization protocol for system services. SYSLIB CHANGES: - SEF framework now supports a new SEF Init request type from RS. 3 different callbacks are available (init_fresh, init_lu, init_restart) to specify initialization code when a service starts fresh, starts after a live update, or restarts. SYSTEM SERVICE CHANGES: - Initialization code for system services is now enclosed in a callback SEF will automatically call at init time. The return code of the callback will tell RS whether the initialization completed successfully. - Each init callback can access information passed by RS to initialize. As of now, each system service has access to the public entries of RS's system process table to gather all the information required to initialize. This design eliminates many existing or potential races at boot time and provides a uniform initialization interface to system services. The same interface will be reused for the upcoming publish/subscribe model to handle dynamic registration / deregistration of system services. VM CHANGES: - Uniform privilege management for all system services. Every service uses the same call mask format. For boot services, VM copies the call mask from init data. For dynamic services, VM still receives the call mask via rs_set_priv call that will be soon replaced by the upcoming publish/subscribe model. RS CHANGES: - The system process table has been reorganized and split into private entries and public entries. Only the latter ones are exposed to system services. - VM call masks are now entirely configured in rs/table.c - RS has now its own slot in the system process table. Only kernel tasks and user processes not included in the boot image are now left out from the system process table. - RS implements the initialization protocol for system services. - For services in the boot image, RS blocks till initialization is complete and panics when failure is reported back. Services are initialized in their order of appearance in the boot image priv table and RS blocks to implements synchronous initialization for every system service having the flag SF_SYNCH_BOOT set. - For services started dynamically, the initialization protocol is implemented as though it were the first ping for the service. In this case, if the system service fails to report back (or reports failure), RS brings the service down rather than trying to restart it.
2010-01-08 02:20:42 +01:00
#define NR_STATIC_GRANTS 2
2012-03-25 20:25:53 +02:00
static cp_grant_t static_grants[NR_STATIC_GRANTS];
static cp_grant_t *grants = NULL;
static int ngrants = 0;
2012-03-25 20:25:53 +02:00
static void
cpf_grow(void)
{
2006-06-30 16:40:29 +02:00
/* Grow the grants table if possible. */
cp_grant_t *new_grants;
cp_grant_id_t g;
int new_size;
Initialization protocol for system services. SYSLIB CHANGES: - SEF framework now supports a new SEF Init request type from RS. 3 different callbacks are available (init_fresh, init_lu, init_restart) to specify initialization code when a service starts fresh, starts after a live update, or restarts. SYSTEM SERVICE CHANGES: - Initialization code for system services is now enclosed in a callback SEF will automatically call at init time. The return code of the callback will tell RS whether the initialization completed successfully. - Each init callback can access information passed by RS to initialize. As of now, each system service has access to the public entries of RS's system process table to gather all the information required to initialize. This design eliminates many existing or potential races at boot time and provides a uniform initialization interface to system services. The same interface will be reused for the upcoming publish/subscribe model to handle dynamic registration / deregistration of system services. VM CHANGES: - Uniform privilege management for all system services. Every service uses the same call mask format. For boot services, VM copies the call mask from init data. For dynamic services, VM still receives the call mask via rs_set_priv call that will be soon replaced by the upcoming publish/subscribe model. RS CHANGES: - The system process table has been reorganized and split into private entries and public entries. Only the latter ones are exposed to system services. - VM call masks are now entirely configured in rs/table.c - RS has now its own slot in the system process table. Only kernel tasks and user processes not included in the boot image are now left out from the system process table. - RS implements the initialization protocol for system services. - For services in the boot image, RS blocks till initialization is complete and panics when failure is reported back. Services are initialized in their order of appearance in the boot image priv table and RS blocks to implements synchronous initialization for every system service having the flag SF_SYNCH_BOOT set. - For services started dynamically, the initialization protocol is implemented as though it were the first ping for the service. In this case, if the system service fails to report back (or reports failure), RS brings the service down rather than trying to restart it.
2010-01-08 02:20:42 +01:00
if(!ngrants) {
/* Use statically allocated grants the first time. */
new_size = NR_STATIC_GRANTS;
new_grants = static_grants;
}
else {
new_size = (1+ngrants)*2;
assert(new_size > ngrants);
Initialization protocol for system services. SYSLIB CHANGES: - SEF framework now supports a new SEF Init request type from RS. 3 different callbacks are available (init_fresh, init_lu, init_restart) to specify initialization code when a service starts fresh, starts after a live update, or restarts. SYSTEM SERVICE CHANGES: - Initialization code for system services is now enclosed in a callback SEF will automatically call at init time. The return code of the callback will tell RS whether the initialization completed successfully. - Each init callback can access information passed by RS to initialize. As of now, each system service has access to the public entries of RS's system process table to gather all the information required to initialize. This design eliminates many existing or potential races at boot time and provides a uniform initialization interface to system services. The same interface will be reused for the upcoming publish/subscribe model to handle dynamic registration / deregistration of system services. VM CHANGES: - Uniform privilege management for all system services. Every service uses the same call mask format. For boot services, VM copies the call mask from init data. For dynamic services, VM still receives the call mask via rs_set_priv call that will be soon replaced by the upcoming publish/subscribe model. RS CHANGES: - The system process table has been reorganized and split into private entries and public entries. Only the latter ones are exposed to system services. - VM call masks are now entirely configured in rs/table.c - RS has now its own slot in the system process table. Only kernel tasks and user processes not included in the boot image are now left out from the system process table. - RS implements the initialization protocol for system services. - For services in the boot image, RS blocks till initialization is complete and panics when failure is reported back. Services are initialized in their order of appearance in the boot image priv table and RS blocks to implements synchronous initialization for every system service having the flag SF_SYNCH_BOOT set. - For services started dynamically, the initialization protocol is implemented as though it were the first ping for the service. In this case, if the system service fails to report back (or reports failure), RS brings the service down rather than trying to restart it.
2010-01-08 02:20:42 +01:00
/* Allocate a block of new size. */
if(!(new_grants=malloc(new_size * sizeof(grants[0])))) {
return;
}
}
/* Copy old block to new block. */
if(grants && ngrants > 0)
memcpy(new_grants, grants, ngrants * sizeof(grants[0]));
/* Make sure new slots are marked unused (CPF_USED is clear). */
for(g = ngrants; g < new_size; g++)
new_grants[g].cp_flags = 0;
/* Inform kernel about new size (and possibly new location). */
2006-06-23 17:35:05 +02:00
if((sys_setgrant(new_grants, new_size))) {
Initialization protocol for system services. SYSLIB CHANGES: - SEF framework now supports a new SEF Init request type from RS. 3 different callbacks are available (init_fresh, init_lu, init_restart) to specify initialization code when a service starts fresh, starts after a live update, or restarts. SYSTEM SERVICE CHANGES: - Initialization code for system services is now enclosed in a callback SEF will automatically call at init time. The return code of the callback will tell RS whether the initialization completed successfully. - Each init callback can access information passed by RS to initialize. As of now, each system service has access to the public entries of RS's system process table to gather all the information required to initialize. This design eliminates many existing or potential races at boot time and provides a uniform initialization interface to system services. The same interface will be reused for the upcoming publish/subscribe model to handle dynamic registration / deregistration of system services. VM CHANGES: - Uniform privilege management for all system services. Every service uses the same call mask format. For boot services, VM copies the call mask from init data. For dynamic services, VM still receives the call mask via rs_set_priv call that will be soon replaced by the upcoming publish/subscribe model. RS CHANGES: - The system process table has been reorganized and split into private entries and public entries. Only the latter ones are exposed to system services. - VM call masks are now entirely configured in rs/table.c - RS has now its own slot in the system process table. Only kernel tasks and user processes not included in the boot image are now left out from the system process table. - RS implements the initialization protocol for system services. - For services in the boot image, RS blocks till initialization is complete and panics when failure is reported back. Services are initialized in their order of appearance in the boot image priv table and RS blocks to implements synchronous initialization for every system service having the flag SF_SYNCH_BOOT set. - For services started dynamically, the initialization protocol is implemented as though it were the first ping for the service. In this case, if the system service fails to report back (or reports failure), RS brings the service down rather than trying to restart it.
2010-01-08 02:20:42 +01:00
if(new_grants != static_grants) free(new_grants);
return; /* Failed - don't grow then. */
}
/* Update internal data. */
Initialization protocol for system services. SYSLIB CHANGES: - SEF framework now supports a new SEF Init request type from RS. 3 different callbacks are available (init_fresh, init_lu, init_restart) to specify initialization code when a service starts fresh, starts after a live update, or restarts. SYSTEM SERVICE CHANGES: - Initialization code for system services is now enclosed in a callback SEF will automatically call at init time. The return code of the callback will tell RS whether the initialization completed successfully. - Each init callback can access information passed by RS to initialize. As of now, each system service has access to the public entries of RS's system process table to gather all the information required to initialize. This design eliminates many existing or potential races at boot time and provides a uniform initialization interface to system services. The same interface will be reused for the upcoming publish/subscribe model to handle dynamic registration / deregistration of system services. VM CHANGES: - Uniform privilege management for all system services. Every service uses the same call mask format. For boot services, VM copies the call mask from init data. For dynamic services, VM still receives the call mask via rs_set_priv call that will be soon replaced by the upcoming publish/subscribe model. RS CHANGES: - The system process table has been reorganized and split into private entries and public entries. Only the latter ones are exposed to system services. - VM call masks are now entirely configured in rs/table.c - RS has now its own slot in the system process table. Only kernel tasks and user processes not included in the boot image are now left out from the system process table. - RS implements the initialization protocol for system services. - For services in the boot image, RS blocks till initialization is complete and panics when failure is reported back. Services are initialized in their order of appearance in the boot image priv table and RS blocks to implements synchronous initialization for every system service having the flag SF_SYNCH_BOOT set. - For services started dynamically, the initialization protocol is implemented as though it were the first ping for the service. In this case, if the system service fails to report back (or reports failure), RS brings the service down rather than trying to restart it.
2010-01-08 02:20:42 +01:00
if(grants && ngrants > 0 && grants != static_grants) free(grants);
grants = new_grants;
ngrants = new_size;
}
2012-03-25 20:25:53 +02:00
static cp_grant_id_t
cpf_new_grantslot(void)
{
/* Find a new, free grant slot in the grant table, grow it if
* necessary. If no free slot is found and the grow failed,
* return -1. Otherwise, return grant slot number.
*/
cp_grant_id_t g;
/* Find free slot. */
2006-06-27 16:15:47 +02:00
for(g = 0; g < ngrants && (grants[g].cp_flags & CPF_USED); g++)
;
assert(g <= ngrants);
/* No free slot found? */
2006-06-27 16:15:47 +02:00
if(g == ngrants) {
cpf_grow();
2006-06-27 16:15:47 +02:00
assert(g <= ngrants); /* ngrants can't shrink. */
if(g == ngrants) {
/* ngrants hasn't increased. */
errno = ENOSPC;
return -1;
}
}
/* Basic sanity checks - if we get this far, g must be a valid,
* free slot.
*/
assert(GRANT_VALID(g));
assert(g >= 0);
assert(g < ngrants);
assert(!(grants[g].cp_flags & CPF_USED));
return g;
}
2012-03-25 20:25:53 +02:00
cp_grant_id_t
cpf_grant_direct(endpoint_t who_to, vir_bytes addr, size_t bytes, int access)
{
cp_grant_id_t g;
int r;
/* Get new slot to put new grant in. */
if((g = cpf_new_grantslot()) < 0)
return(GRANT_INVALID);
assert(GRANT_VALID(g));
assert(g >= 0);
assert(g < ngrants);
assert(!(grants[g].cp_flags & CPF_USED));
if((r=cpf_setgrant_direct(g, who_to, addr, bytes, access)) < 0) {
cpf_revoke(g);
return(GRANT_INVALID);
}
return g;
}
2012-03-25 20:25:53 +02:00
cp_grant_id_t
cpf_grant_indirect(endpoint_t who_to, endpoint_t who_from, cp_grant_id_t gr)
{
/* Grant process A access into process B. B has granted us access as grant
* id 'gr'.
*/
cp_grant_id_t g;
int r;
/* Obtain new slot. */
if((g = cpf_new_grantslot()) < 0)
return -1;
/* Basic sanity checks. */
assert(GRANT_VALID(g));
assert(g >= 0);
assert(g < ngrants);
assert(!(grants[g].cp_flags & CPF_USED));
/* Fill in new slot data. */
if((r=cpf_setgrant_indirect(g, who_to, who_from, gr)) < 0) {
cpf_revoke(g);
return GRANT_INVALID;
}
return g;
}
2012-03-25 20:25:53 +02:00
cp_grant_id_t
cpf_grant_magic(endpoint_t who_to, endpoint_t who_from,
vir_bytes addr, size_t bytes, int access)
{
/* Grant process A access into process B. Not everyone can do this. */
cp_grant_id_t g;
int r;
ACCESS_CHECK(access);
/* Obtain new slot. */
if((g = cpf_new_grantslot()) < 0)
return -1;
/* Basic sanity checks. */
assert(GRANT_VALID(g));
assert(g >= 0);
assert(g < ngrants);
assert(!(grants[g].cp_flags & CPF_USED));
if((r=cpf_setgrant_magic(g, who_to, who_from, addr,
bytes, access)) < 0) {
cpf_revoke(g);
return -1;
}
return g;
}
2012-03-25 20:25:53 +02:00
int
cpf_revoke(cp_grant_id_t g)
{
/* Revoke previously granted access, identified by grant id. */
GID_CHECK_USED(g);
/* Make grant invalid by setting flags to 0, clearing CPF_USED.
* This invalidates the grant.
*/
grants[g].cp_flags = 0;
return 0;
}
2012-03-25 20:25:53 +02:00
int
cpf_lookup(cp_grant_id_t g, endpoint_t *granter, endpoint_t *grantee)
{
/* First check slot validity, and if it's in use currently. */
GID_CHECK_USED(g);
if(grants[g].cp_flags & CPF_DIRECT) {
if(granter) *granter = SELF;
if(grantee) *grantee = grants[g].cp_u.cp_direct.cp_who_to;
} else if(grants[g].cp_flags & CPF_MAGIC) {
if(granter) *granter = grants[g].cp_u.cp_magic.cp_who_from;
if(grantee) *grantee = grants[g].cp_u.cp_magic.cp_who_to;
} else return -1;
return 0;
}
2012-03-25 20:25:53 +02:00
int
cpf_getgrants(cp_grant_id_t *grant_ids, int n)
{
int i;
for(i = 0; i < n; i++) {
if((grant_ids[i] = cpf_new_grantslot()) < 0)
break;
2006-06-27 16:15:47 +02:00
grants[grant_ids[i]].cp_flags = CPF_USED;
}
/* return however many grants were assigned. */
return i;
}
2012-03-25 20:25:53 +02:00
int
cpf_setgrant_direct(gid, who, addr, bytes, access)
cp_grant_id_t gid;
endpoint_t who;
vir_bytes addr;
size_t bytes;
int access;
{
GID_CHECK(gid);
ACCESS_CHECK(access);
/* Fill in new slot data. */
grants[gid].cp_flags = access | CPF_DIRECT | CPF_USED | CPF_VALID;
grants[gid].cp_u.cp_direct.cp_who_to = who;
grants[gid].cp_u.cp_direct.cp_start = addr;
grants[gid].cp_u.cp_direct.cp_len = bytes;
return 0;
}
2012-03-25 20:25:53 +02:00
int
cpf_setgrant_indirect(gid, who_to, who_from, his_gid)
cp_grant_id_t gid;
endpoint_t who_to, who_from;
cp_grant_id_t his_gid;
{
GID_CHECK(gid);
/* Fill in new slot data. */
grants[gid].cp_flags = CPF_USED | CPF_INDIRECT | CPF_VALID;
grants[gid].cp_u.cp_indirect.cp_who_to = who_to;
grants[gid].cp_u.cp_indirect.cp_who_from = who_from;
grants[gid].cp_u.cp_indirect.cp_grant = his_gid;
return 0;
}
2012-03-25 20:25:53 +02:00
int
cpf_setgrant_magic(gid, who_to, who_from, addr, bytes, access)
cp_grant_id_t gid;
endpoint_t who_to, who_from;
vir_bytes addr;
size_t bytes;
int access;
{
GID_CHECK(gid);
ACCESS_CHECK(access);
/* Fill in new slot data. */
grants[gid].cp_flags = CPF_USED | CPF_MAGIC | CPF_VALID | access;
grants[gid].cp_u.cp_magic.cp_who_to = who_to;
grants[gid].cp_u.cp_magic.cp_who_from = who_from;
grants[gid].cp_u.cp_magic.cp_start = addr;
grants[gid].cp_u.cp_magic.cp_len = bytes;
return 0;
}
2012-03-25 20:25:53 +02:00
int
cpf_setgrant_disable(gid)
cp_grant_id_t gid;
{
GID_CHECK(gid);
/* Grant is now no longer valid, but still in use. */
grants[gid].cp_flags = CPF_USED;
return 0;
}
2012-03-25 20:25:53 +02:00
void
cpf_reload(void)
{
/* Inform the kernel about the location of the grant table. This is needed
* after a fork.
*/
if (grants)
sys_setgrant(grants, ngrants); /* Do we need error checking? */
}