minix/servers/vfs/request.c

1064 lines
27 KiB
C
Raw Normal View History

/* This file contains the wrapper functions for issueing a request
* and receiving response from FS processes.
* Each function builds a request message according to the request
* parameter, calls the most low-level fs_sendrec and copies
* back the response.
* The low-level fs_sendrec handles the recovery mechanism from
* a dead driver and reissues the request.
*
* Sep 2006 (Balazs Gerofi)
*/
#include "fs.h"
#include <string.h>
2007-01-05 17:36:55 +01:00
#include <sys/stat.h>
#include <sys/statfs.h>
#include <minix/callnr.h>
#include <minix/com.h>
#include <minix/keymap.h>
#include <minix/const.h>
#include <minix/endpoint.h>
#include <minix/u64.h>
#include <unistd.h>
#include <minix/vfsif.h>
#include "fproc.h"
#include "vmnt.h"
#include "vnode.h"
#include "param.h"
Mostly bugfixes of bugs triggered by the test set. bugfixes: SYSTEM: . removed rc->p_priv->s_flags = 0; for the priv struct shared by all user processes in get_priv(). this should only be done once. doing a SYS_PRIV_USER in sys_privctl() caused the flags of all user processes to be reset, so they were no longer PREEMPTIBLE. this happened when RS executed a policy script. (this broke test1 in the test set) VFS/MFS: . chown can change the mode of a file, and chmod arguments are only part of the full file mode so the full filemode is slightly magic. changed these calls so that the final modes are returned to VFS, so that the vnode can be kept up-to-date. (this broke test11 in the test set) MFS: . lookup() checked for sizeof(string) instead of sizeof(user_path), truncating long path names (caught by test 23) . truncate functions neglected to update ctime (this broke test16) VFS: . corner case of an empty filename lookup caused fields of a request not to be filled in in the lookup functions, not making it clear that the lookup had failed, causing messages to garbage processes, causing strange failures. (caught by test 30) . trust v_size in vnode when doing reads or writes on non-special files, truncating i/o where necessary; this is necessary for pipes, as MFS can't tell when a pipe has been truncated without it being told explicitly each time. when the last reader/writer on a pipe closes, tell FS about the new size using truncate_vn(). (this broke test 25, among others) . permission check for chdir() had disappeared; added a forbidden() call (caught by test 23) new code, shouldn't change anything: . introduced RTS_SET, RTS_UNSET, and RTS_ISSET macro's, and their LOCK variants. These macros set and clear the p_rts_flags field, causing a lot of duplicated logic like old_flags = rp->p_rts_flags; /* save value of the flags */ rp->p_rts_flags &= ~NO_PRIV; if (old_flags != 0 && rp->p_rts_flags == 0) lock_enqueue(rp); to change into the simpler RTS_LOCK_UNSET(rp, NO_PRIV); so the macros take care of calling dequeue() and enqueue() (or lock_*()), as the case may be). This makes the code a bit more readable and a bit less fragile. . removed return code from do_clocktick in CLOCK as it currently never replies . removed some debug code from VFS . fixed grant debug message in device.c preemptive checks, tests, changes: . added return code checks of receive() to SYSTEM and CLOCK . O_TRUNC should never arrive at MFS (added sanity check and removed O_TRUNC code) . user_path declared with PATH_MAX+1 to let it be null-terminated . checks in MFS to see if strings passed by VFS are null-terminated IS: . static irq name table thrown out
2007-02-01 18:50:02 +01:00
FORWARD _PROTOTYPE(int fs_sendrec_f, (char *file, int line, endpoint_t fs_e, message *reqm));
#define fs_sendrec(e, m) fs_sendrec_f(__FILE__, __LINE__, (e), (m))
/*===========================================================================*
* req_getnode *
*===========================================================================*/
Mostly bugfixes of bugs triggered by the test set. bugfixes: SYSTEM: . removed rc->p_priv->s_flags = 0; for the priv struct shared by all user processes in get_priv(). this should only be done once. doing a SYS_PRIV_USER in sys_privctl() caused the flags of all user processes to be reset, so they were no longer PREEMPTIBLE. this happened when RS executed a policy script. (this broke test1 in the test set) VFS/MFS: . chown can change the mode of a file, and chmod arguments are only part of the full file mode so the full filemode is slightly magic. changed these calls so that the final modes are returned to VFS, so that the vnode can be kept up-to-date. (this broke test11 in the test set) MFS: . lookup() checked for sizeof(string) instead of sizeof(user_path), truncating long path names (caught by test 23) . truncate functions neglected to update ctime (this broke test16) VFS: . corner case of an empty filename lookup caused fields of a request not to be filled in in the lookup functions, not making it clear that the lookup had failed, causing messages to garbage processes, causing strange failures. (caught by test 30) . trust v_size in vnode when doing reads or writes on non-special files, truncating i/o where necessary; this is necessary for pipes, as MFS can't tell when a pipe has been truncated without it being told explicitly each time. when the last reader/writer on a pipe closes, tell FS about the new size using truncate_vn(). (this broke test 25, among others) . permission check for chdir() had disappeared; added a forbidden() call (caught by test 23) new code, shouldn't change anything: . introduced RTS_SET, RTS_UNSET, and RTS_ISSET macro's, and their LOCK variants. These macros set and clear the p_rts_flags field, causing a lot of duplicated logic like old_flags = rp->p_rts_flags; /* save value of the flags */ rp->p_rts_flags &= ~NO_PRIV; if (old_flags != 0 && rp->p_rts_flags == 0) lock_enqueue(rp); to change into the simpler RTS_LOCK_UNSET(rp, NO_PRIV); so the macros take care of calling dequeue() and enqueue() (or lock_*()), as the case may be). This makes the code a bit more readable and a bit less fragile. . removed return code from do_clocktick in CLOCK as it currently never replies . removed some debug code from VFS . fixed grant debug message in device.c preemptive checks, tests, changes: . added return code checks of receive() to SYSTEM and CLOCK . O_TRUNC should never arrive at MFS (added sanity check and removed O_TRUNC code) . user_path declared with PATH_MAX+1 to let it be null-terminated . checks in MFS to see if strings passed by VFS are null-terminated IS: . static irq name table thrown out
2007-02-01 18:50:02 +01:00
PUBLIC int req_getnode_f(file, line, req, res)
char *file;
int line;
node_req_t *req;
node_details_t *res;
{
int r;
message m;
/* Fill in request message */
m.m_type = REQ_GETNODE;
m.REQ_INODE_NR = req->inode_nr;
/* Send/rec request */
if ((r = fs_sendrec(req->fs_e, &m)) != OK) return r;
/* Fill in response structure */
res->fs_e = m.m_source;
res->inode_nr = m.RES_INODE_NR;
res->fmode = m.RES_MODE;
res->fsize = m.RES_FILE_SIZE;
res->dev = m.RES_DEV;
2007-01-05 17:36:55 +01:00
res->uid = m.RES_UID;
res->gid = m.RES_GID;
return OK;
}
/*===========================================================================*
* req_putnode *
*===========================================================================*/
2007-01-05 17:36:55 +01:00
PUBLIC int req_putnode(fs_e, inode_nr, count)
int fs_e;
ino_t inode_nr;
int count;
{
message m;
/* Fill in request message */
m.m_type = REQ_PUTNODE;
2007-01-05 17:36:55 +01:00
m.REQ_INODE_NR = inode_nr;
m.REQ_COUNT = count;
/* Send/rec request */
2007-01-05 17:36:55 +01:00
return fs_sendrec(fs_e, &m);
}
/*===========================================================================*
* req_open *
*===========================================================================*/
int req_open(req, res)
open_req_t *req;
node_details_t *res;
{
int r;
message m;
/* Fill in request message */
m.m_type = REQ_OPEN;
m.REQ_INODE_NR = req->inode_nr;
m.REQ_FLAGS = req->oflags;
m.REQ_MODE = req->omode;
m.REQ_UID = req->uid;
m.REQ_GID = req->gid;
m.REQ_PATH = req->lastc;
m.REQ_PATH_LEN = strlen(req->lastc) + 1;
/* Send/rec request */
if ((r = fs_sendrec(req->fs_e, &m)) != OK) return r;
/* Fill in response structure */
res->fs_e = m.m_source;
res->inode_nr = m.RES_INODE_NR;
res->fmode = m.RES_MODE;
res->fsize = m.RES_FILE_SIZE;
res->dev = m.RES_DEV;
res->inode_index = m.RES_INODE_INDEX;
/* For exec */
res->uid = m.RES_UID;
res->gid = m.RES_GID;
res->ctime = m.RES_CTIME;
return OK;
}
2007-01-05 17:36:55 +01:00
/*===========================================================================*
* req_create *
*===========================================================================*/
int req_create(fs_e, inode_nr, omode, uid, gid, path, res)
int fs_e;
ino_t inode_nr;
int omode;
uid_t uid;
gid_t gid;
char *path;
node_details_t *res;
{
int r;
message m;
/* Fill in request message */
m.m_type = REQ_CREATE;
m.REQ_INODE_NR = inode_nr;
m.REQ_MODE = omode;
m.REQ_UID = uid;
m.REQ_GID = gid;
m.REQ_PATH = path;
m.REQ_PATH_LEN = strlen(path) + 1;
/* Send/rec request */
if ((r = fs_sendrec(fs_e, &m)) != OK) return r;
/* Fill in response structure */
res->fs_e = m.m_source;
res->inode_nr = m.RES_INODE_NR;
res->fmode = m.RES_MODE;
res->fsize = m.RES_FILE_SIZE;
res->dev = m.RES_DEV;
res->inode_index = m.RES_INODE_INDEX;
/* For exec */
res->uid = m.RES_UID;
res->gid = m.RES_GID;
res->ctime = m.RES_CTIME;
return OK;
}
/*===========================================================================*
* req_readwrite *
*===========================================================================*/
int req_readwrite(req, res)
readwrite_req_t *req;
readwrite_res_t *res;
{
int r;
message m;
if (ex64hi(req->pos) != 0)
panic(__FILE__, "req_readwrite: pos too large", NO_NUM);
/* Fill in request message */
m.m_type = req->rw_flag == READING ? REQ_READ : REQ_WRITE;
m.REQ_FD_INODE_NR = req->inode_nr;
m.REQ_FD_WHO_E = req->user_e;
m.REQ_FD_SEG = req->seg;
m.REQ_FD_POS = ex64lo(req->pos);
m.REQ_FD_NBYTES = req->num_of_bytes;
m.REQ_FD_USER_ADDR = req->user_addr;
m.REQ_FD_INODE_INDEX = req->inode_index;
/* Send/rec request */
if ((r = fs_sendrec(req->fs_e, &m)) != OK) return r;
/* Fill in response structure */
res->new_pos = cvul64(m.RES_FD_POS);
res->cum_io = m.RES_FD_CUM_IO;
return OK;
}
/*===========================================================================*
* req_pipe *
*===========================================================================*/
PUBLIC int req_pipe(req, res)
pipe_req_t *req;
node_details_t *res;
{
int r;
message m;
/* Fill in request message */
m.m_type = REQ_PIPE;
m.REQ_UID = req->uid;
m.REQ_GID = req->gid;
/* Send/rec request */
if ((r = fs_sendrec(req->fs_e, &m)) != OK) return r;
/* Fill in response structure */
res->fs_e = m.m_source;
res->inode_nr = m.RES_INODE_NR;
res->fmode = m.RES_MODE;
res->fsize = m.RES_FILE_SIZE;
res->dev = m.RES_DEV;
res->inode_index = m.RES_INODE_INDEX;
return OK;
}
/*===========================================================================*
* req_clone_opcl *
*===========================================================================*/
PUBLIC int req_clone_opcl(req, res)
clone_opcl_req_t *req;
node_details_t *res;
{
int r;
message m;
/* Fill in request message */
m.m_type = REQ_CLONE_OPCL;
m.REQ_DEV = req->dev;
/* Send/rec request */
if ((r = fs_sendrec(req->fs_e, &m)) != OK) return r;
/* Fill in response structure */
res->fs_e = m.m_source;
res->inode_nr = m.RES_INODE_NR;
res->fmode = m.RES_MODE;
res->fsize = m.RES_FILE_SIZE;
res->dev = m.RES_DEV;
res->inode_index = m.RES_INODE_INDEX;
return OK;
}
/*===========================================================================*
* req_ftrunc *
*===========================================================================*/
PUBLIC int req_ftrunc(req)
ftrunc_req_t *req;
{
message m;
/* Fill in request message */
m.m_type = REQ_FTRUNC;
m.REQ_FD_INODE_NR = req->inode_nr;
m.REQ_FD_START = req->start;
m.REQ_FD_END = req->end;
/* Send/rec request */
return fs_sendrec(req->fs_e, &m);
}
/*===========================================================================*
* req_chmod *
*===========================================================================*/
Mostly bugfixes of bugs triggered by the test set. bugfixes: SYSTEM: . removed rc->p_priv->s_flags = 0; for the priv struct shared by all user processes in get_priv(). this should only be done once. doing a SYS_PRIV_USER in sys_privctl() caused the flags of all user processes to be reset, so they were no longer PREEMPTIBLE. this happened when RS executed a policy script. (this broke test1 in the test set) VFS/MFS: . chown can change the mode of a file, and chmod arguments are only part of the full file mode so the full filemode is slightly magic. changed these calls so that the final modes are returned to VFS, so that the vnode can be kept up-to-date. (this broke test11 in the test set) MFS: . lookup() checked for sizeof(string) instead of sizeof(user_path), truncating long path names (caught by test 23) . truncate functions neglected to update ctime (this broke test16) VFS: . corner case of an empty filename lookup caused fields of a request not to be filled in in the lookup functions, not making it clear that the lookup had failed, causing messages to garbage processes, causing strange failures. (caught by test 30) . trust v_size in vnode when doing reads or writes on non-special files, truncating i/o where necessary; this is necessary for pipes, as MFS can't tell when a pipe has been truncated without it being told explicitly each time. when the last reader/writer on a pipe closes, tell FS about the new size using truncate_vn(). (this broke test 25, among others) . permission check for chdir() had disappeared; added a forbidden() call (caught by test 23) new code, shouldn't change anything: . introduced RTS_SET, RTS_UNSET, and RTS_ISSET macro's, and their LOCK variants. These macros set and clear the p_rts_flags field, causing a lot of duplicated logic like old_flags = rp->p_rts_flags; /* save value of the flags */ rp->p_rts_flags &= ~NO_PRIV; if (old_flags != 0 && rp->p_rts_flags == 0) lock_enqueue(rp); to change into the simpler RTS_LOCK_UNSET(rp, NO_PRIV); so the macros take care of calling dequeue() and enqueue() (or lock_*()), as the case may be). This makes the code a bit more readable and a bit less fragile. . removed return code from do_clocktick in CLOCK as it currently never replies . removed some debug code from VFS . fixed grant debug message in device.c preemptive checks, tests, changes: . added return code checks of receive() to SYSTEM and CLOCK . O_TRUNC should never arrive at MFS (added sanity check and removed O_TRUNC code) . user_path declared with PATH_MAX+1 to let it be null-terminated . checks in MFS to see if strings passed by VFS are null-terminated IS: . static irq name table thrown out
2007-02-01 18:50:02 +01:00
PUBLIC int req_chmod(req, ch_mode)
chmod_req_t *req;
Mostly bugfixes of bugs triggered by the test set. bugfixes: SYSTEM: . removed rc->p_priv->s_flags = 0; for the priv struct shared by all user processes in get_priv(). this should only be done once. doing a SYS_PRIV_USER in sys_privctl() caused the flags of all user processes to be reset, so they were no longer PREEMPTIBLE. this happened when RS executed a policy script. (this broke test1 in the test set) VFS/MFS: . chown can change the mode of a file, and chmod arguments are only part of the full file mode so the full filemode is slightly magic. changed these calls so that the final modes are returned to VFS, so that the vnode can be kept up-to-date. (this broke test11 in the test set) MFS: . lookup() checked for sizeof(string) instead of sizeof(user_path), truncating long path names (caught by test 23) . truncate functions neglected to update ctime (this broke test16) VFS: . corner case of an empty filename lookup caused fields of a request not to be filled in in the lookup functions, not making it clear that the lookup had failed, causing messages to garbage processes, causing strange failures. (caught by test 30) . trust v_size in vnode when doing reads or writes on non-special files, truncating i/o where necessary; this is necessary for pipes, as MFS can't tell when a pipe has been truncated without it being told explicitly each time. when the last reader/writer on a pipe closes, tell FS about the new size using truncate_vn(). (this broke test 25, among others) . permission check for chdir() had disappeared; added a forbidden() call (caught by test 23) new code, shouldn't change anything: . introduced RTS_SET, RTS_UNSET, and RTS_ISSET macro's, and their LOCK variants. These macros set and clear the p_rts_flags field, causing a lot of duplicated logic like old_flags = rp->p_rts_flags; /* save value of the flags */ rp->p_rts_flags &= ~NO_PRIV; if (old_flags != 0 && rp->p_rts_flags == 0) lock_enqueue(rp); to change into the simpler RTS_LOCK_UNSET(rp, NO_PRIV); so the macros take care of calling dequeue() and enqueue() (or lock_*()), as the case may be). This makes the code a bit more readable and a bit less fragile. . removed return code from do_clocktick in CLOCK as it currently never replies . removed some debug code from VFS . fixed grant debug message in device.c preemptive checks, tests, changes: . added return code checks of receive() to SYSTEM and CLOCK . O_TRUNC should never arrive at MFS (added sanity check and removed O_TRUNC code) . user_path declared with PATH_MAX+1 to let it be null-terminated . checks in MFS to see if strings passed by VFS are null-terminated IS: . static irq name table thrown out
2007-02-01 18:50:02 +01:00
int *ch_mode;
{
message m;
Mostly bugfixes of bugs triggered by the test set. bugfixes: SYSTEM: . removed rc->p_priv->s_flags = 0; for the priv struct shared by all user processes in get_priv(). this should only be done once. doing a SYS_PRIV_USER in sys_privctl() caused the flags of all user processes to be reset, so they were no longer PREEMPTIBLE. this happened when RS executed a policy script. (this broke test1 in the test set) VFS/MFS: . chown can change the mode of a file, and chmod arguments are only part of the full file mode so the full filemode is slightly magic. changed these calls so that the final modes are returned to VFS, so that the vnode can be kept up-to-date. (this broke test11 in the test set) MFS: . lookup() checked for sizeof(string) instead of sizeof(user_path), truncating long path names (caught by test 23) . truncate functions neglected to update ctime (this broke test16) VFS: . corner case of an empty filename lookup caused fields of a request not to be filled in in the lookup functions, not making it clear that the lookup had failed, causing messages to garbage processes, causing strange failures. (caught by test 30) . trust v_size in vnode when doing reads or writes on non-special files, truncating i/o where necessary; this is necessary for pipes, as MFS can't tell when a pipe has been truncated without it being told explicitly each time. when the last reader/writer on a pipe closes, tell FS about the new size using truncate_vn(). (this broke test 25, among others) . permission check for chdir() had disappeared; added a forbidden() call (caught by test 23) new code, shouldn't change anything: . introduced RTS_SET, RTS_UNSET, and RTS_ISSET macro's, and their LOCK variants. These macros set and clear the p_rts_flags field, causing a lot of duplicated logic like old_flags = rp->p_rts_flags; /* save value of the flags */ rp->p_rts_flags &= ~NO_PRIV; if (old_flags != 0 && rp->p_rts_flags == 0) lock_enqueue(rp); to change into the simpler RTS_LOCK_UNSET(rp, NO_PRIV); so the macros take care of calling dequeue() and enqueue() (or lock_*()), as the case may be). This makes the code a bit more readable and a bit less fragile. . removed return code from do_clocktick in CLOCK as it currently never replies . removed some debug code from VFS . fixed grant debug message in device.c preemptive checks, tests, changes: . added return code checks of receive() to SYSTEM and CLOCK . O_TRUNC should never arrive at MFS (added sanity check and removed O_TRUNC code) . user_path declared with PATH_MAX+1 to let it be null-terminated . checks in MFS to see if strings passed by VFS are null-terminated IS: . static irq name table thrown out
2007-02-01 18:50:02 +01:00
int r;
/* Fill in request message */
m.m_type = REQ_CHMOD;
m.REQ_INODE_NR = req->inode_nr;
m.REQ_MODE = req->rmode;
m.REQ_UID = req->uid;
m.REQ_GID = req->gid;
/* Send/rec request */
Mostly bugfixes of bugs triggered by the test set. bugfixes: SYSTEM: . removed rc->p_priv->s_flags = 0; for the priv struct shared by all user processes in get_priv(). this should only be done once. doing a SYS_PRIV_USER in sys_privctl() caused the flags of all user processes to be reset, so they were no longer PREEMPTIBLE. this happened when RS executed a policy script. (this broke test1 in the test set) VFS/MFS: . chown can change the mode of a file, and chmod arguments are only part of the full file mode so the full filemode is slightly magic. changed these calls so that the final modes are returned to VFS, so that the vnode can be kept up-to-date. (this broke test11 in the test set) MFS: . lookup() checked for sizeof(string) instead of sizeof(user_path), truncating long path names (caught by test 23) . truncate functions neglected to update ctime (this broke test16) VFS: . corner case of an empty filename lookup caused fields of a request not to be filled in in the lookup functions, not making it clear that the lookup had failed, causing messages to garbage processes, causing strange failures. (caught by test 30) . trust v_size in vnode when doing reads or writes on non-special files, truncating i/o where necessary; this is necessary for pipes, as MFS can't tell when a pipe has been truncated without it being told explicitly each time. when the last reader/writer on a pipe closes, tell FS about the new size using truncate_vn(). (this broke test 25, among others) . permission check for chdir() had disappeared; added a forbidden() call (caught by test 23) new code, shouldn't change anything: . introduced RTS_SET, RTS_UNSET, and RTS_ISSET macro's, and their LOCK variants. These macros set and clear the p_rts_flags field, causing a lot of duplicated logic like old_flags = rp->p_rts_flags; /* save value of the flags */ rp->p_rts_flags &= ~NO_PRIV; if (old_flags != 0 && rp->p_rts_flags == 0) lock_enqueue(rp); to change into the simpler RTS_LOCK_UNSET(rp, NO_PRIV); so the macros take care of calling dequeue() and enqueue() (or lock_*()), as the case may be). This makes the code a bit more readable and a bit less fragile. . removed return code from do_clocktick in CLOCK as it currently never replies . removed some debug code from VFS . fixed grant debug message in device.c preemptive checks, tests, changes: . added return code checks of receive() to SYSTEM and CLOCK . O_TRUNC should never arrive at MFS (added sanity check and removed O_TRUNC code) . user_path declared with PATH_MAX+1 to let it be null-terminated . checks in MFS to see if strings passed by VFS are null-terminated IS: . static irq name table thrown out
2007-02-01 18:50:02 +01:00
r = fs_sendrec(req->fs_e, &m);
/* Copy back actual mode. */
if(ch_mode) *ch_mode = m.RES_MODE;
return r;
}
/*===========================================================================*
* req_chown *
*===========================================================================*/
Mostly bugfixes of bugs triggered by the test set. bugfixes: SYSTEM: . removed rc->p_priv->s_flags = 0; for the priv struct shared by all user processes in get_priv(). this should only be done once. doing a SYS_PRIV_USER in sys_privctl() caused the flags of all user processes to be reset, so they were no longer PREEMPTIBLE. this happened when RS executed a policy script. (this broke test1 in the test set) VFS/MFS: . chown can change the mode of a file, and chmod arguments are only part of the full file mode so the full filemode is slightly magic. changed these calls so that the final modes are returned to VFS, so that the vnode can be kept up-to-date. (this broke test11 in the test set) MFS: . lookup() checked for sizeof(string) instead of sizeof(user_path), truncating long path names (caught by test 23) . truncate functions neglected to update ctime (this broke test16) VFS: . corner case of an empty filename lookup caused fields of a request not to be filled in in the lookup functions, not making it clear that the lookup had failed, causing messages to garbage processes, causing strange failures. (caught by test 30) . trust v_size in vnode when doing reads or writes on non-special files, truncating i/o where necessary; this is necessary for pipes, as MFS can't tell when a pipe has been truncated without it being told explicitly each time. when the last reader/writer on a pipe closes, tell FS about the new size using truncate_vn(). (this broke test 25, among others) . permission check for chdir() had disappeared; added a forbidden() call (caught by test 23) new code, shouldn't change anything: . introduced RTS_SET, RTS_UNSET, and RTS_ISSET macro's, and their LOCK variants. These macros set and clear the p_rts_flags field, causing a lot of duplicated logic like old_flags = rp->p_rts_flags; /* save value of the flags */ rp->p_rts_flags &= ~NO_PRIV; if (old_flags != 0 && rp->p_rts_flags == 0) lock_enqueue(rp); to change into the simpler RTS_LOCK_UNSET(rp, NO_PRIV); so the macros take care of calling dequeue() and enqueue() (or lock_*()), as the case may be). This makes the code a bit more readable and a bit less fragile. . removed return code from do_clocktick in CLOCK as it currently never replies . removed some debug code from VFS . fixed grant debug message in device.c preemptive checks, tests, changes: . added return code checks of receive() to SYSTEM and CLOCK . O_TRUNC should never arrive at MFS (added sanity check and removed O_TRUNC code) . user_path declared with PATH_MAX+1 to let it be null-terminated . checks in MFS to see if strings passed by VFS are null-terminated IS: . static irq name table thrown out
2007-02-01 18:50:02 +01:00
PUBLIC int req_chown(req, ch_mode)
chown_req_t *req;
Mostly bugfixes of bugs triggered by the test set. bugfixes: SYSTEM: . removed rc->p_priv->s_flags = 0; for the priv struct shared by all user processes in get_priv(). this should only be done once. doing a SYS_PRIV_USER in sys_privctl() caused the flags of all user processes to be reset, so they were no longer PREEMPTIBLE. this happened when RS executed a policy script. (this broke test1 in the test set) VFS/MFS: . chown can change the mode of a file, and chmod arguments are only part of the full file mode so the full filemode is slightly magic. changed these calls so that the final modes are returned to VFS, so that the vnode can be kept up-to-date. (this broke test11 in the test set) MFS: . lookup() checked for sizeof(string) instead of sizeof(user_path), truncating long path names (caught by test 23) . truncate functions neglected to update ctime (this broke test16) VFS: . corner case of an empty filename lookup caused fields of a request not to be filled in in the lookup functions, not making it clear that the lookup had failed, causing messages to garbage processes, causing strange failures. (caught by test 30) . trust v_size in vnode when doing reads or writes on non-special files, truncating i/o where necessary; this is necessary for pipes, as MFS can't tell when a pipe has been truncated without it being told explicitly each time. when the last reader/writer on a pipe closes, tell FS about the new size using truncate_vn(). (this broke test 25, among others) . permission check for chdir() had disappeared; added a forbidden() call (caught by test 23) new code, shouldn't change anything: . introduced RTS_SET, RTS_UNSET, and RTS_ISSET macro's, and their LOCK variants. These macros set and clear the p_rts_flags field, causing a lot of duplicated logic like old_flags = rp->p_rts_flags; /* save value of the flags */ rp->p_rts_flags &= ~NO_PRIV; if (old_flags != 0 && rp->p_rts_flags == 0) lock_enqueue(rp); to change into the simpler RTS_LOCK_UNSET(rp, NO_PRIV); so the macros take care of calling dequeue() and enqueue() (or lock_*()), as the case may be). This makes the code a bit more readable and a bit less fragile. . removed return code from do_clocktick in CLOCK as it currently never replies . removed some debug code from VFS . fixed grant debug message in device.c preemptive checks, tests, changes: . added return code checks of receive() to SYSTEM and CLOCK . O_TRUNC should never arrive at MFS (added sanity check and removed O_TRUNC code) . user_path declared with PATH_MAX+1 to let it be null-terminated . checks in MFS to see if strings passed by VFS are null-terminated IS: . static irq name table thrown out
2007-02-01 18:50:02 +01:00
int *ch_mode;
{
message m;
Mostly bugfixes of bugs triggered by the test set. bugfixes: SYSTEM: . removed rc->p_priv->s_flags = 0; for the priv struct shared by all user processes in get_priv(). this should only be done once. doing a SYS_PRIV_USER in sys_privctl() caused the flags of all user processes to be reset, so they were no longer PREEMPTIBLE. this happened when RS executed a policy script. (this broke test1 in the test set) VFS/MFS: . chown can change the mode of a file, and chmod arguments are only part of the full file mode so the full filemode is slightly magic. changed these calls so that the final modes are returned to VFS, so that the vnode can be kept up-to-date. (this broke test11 in the test set) MFS: . lookup() checked for sizeof(string) instead of sizeof(user_path), truncating long path names (caught by test 23) . truncate functions neglected to update ctime (this broke test16) VFS: . corner case of an empty filename lookup caused fields of a request not to be filled in in the lookup functions, not making it clear that the lookup had failed, causing messages to garbage processes, causing strange failures. (caught by test 30) . trust v_size in vnode when doing reads or writes on non-special files, truncating i/o where necessary; this is necessary for pipes, as MFS can't tell when a pipe has been truncated without it being told explicitly each time. when the last reader/writer on a pipe closes, tell FS about the new size using truncate_vn(). (this broke test 25, among others) . permission check for chdir() had disappeared; added a forbidden() call (caught by test 23) new code, shouldn't change anything: . introduced RTS_SET, RTS_UNSET, and RTS_ISSET macro's, and their LOCK variants. These macros set and clear the p_rts_flags field, causing a lot of duplicated logic like old_flags = rp->p_rts_flags; /* save value of the flags */ rp->p_rts_flags &= ~NO_PRIV; if (old_flags != 0 && rp->p_rts_flags == 0) lock_enqueue(rp); to change into the simpler RTS_LOCK_UNSET(rp, NO_PRIV); so the macros take care of calling dequeue() and enqueue() (or lock_*()), as the case may be). This makes the code a bit more readable and a bit less fragile. . removed return code from do_clocktick in CLOCK as it currently never replies . removed some debug code from VFS . fixed grant debug message in device.c preemptive checks, tests, changes: . added return code checks of receive() to SYSTEM and CLOCK . O_TRUNC should never arrive at MFS (added sanity check and removed O_TRUNC code) . user_path declared with PATH_MAX+1 to let it be null-terminated . checks in MFS to see if strings passed by VFS are null-terminated IS: . static irq name table thrown out
2007-02-01 18:50:02 +01:00
int r;
/* Fill in request message */
m.m_type = REQ_CHOWN;
m.REQ_INODE_NR = req->inode_nr;
m.REQ_UID = req->uid;
m.REQ_GID = req->gid;
m.REQ_NEW_UID = req->newuid;
m.REQ_NEW_GID = req->newgid;
/* Send/rec request */
Mostly bugfixes of bugs triggered by the test set. bugfixes: SYSTEM: . removed rc->p_priv->s_flags = 0; for the priv struct shared by all user processes in get_priv(). this should only be done once. doing a SYS_PRIV_USER in sys_privctl() caused the flags of all user processes to be reset, so they were no longer PREEMPTIBLE. this happened when RS executed a policy script. (this broke test1 in the test set) VFS/MFS: . chown can change the mode of a file, and chmod arguments are only part of the full file mode so the full filemode is slightly magic. changed these calls so that the final modes are returned to VFS, so that the vnode can be kept up-to-date. (this broke test11 in the test set) MFS: . lookup() checked for sizeof(string) instead of sizeof(user_path), truncating long path names (caught by test 23) . truncate functions neglected to update ctime (this broke test16) VFS: . corner case of an empty filename lookup caused fields of a request not to be filled in in the lookup functions, not making it clear that the lookup had failed, causing messages to garbage processes, causing strange failures. (caught by test 30) . trust v_size in vnode when doing reads or writes on non-special files, truncating i/o where necessary; this is necessary for pipes, as MFS can't tell when a pipe has been truncated without it being told explicitly each time. when the last reader/writer on a pipe closes, tell FS about the new size using truncate_vn(). (this broke test 25, among others) . permission check for chdir() had disappeared; added a forbidden() call (caught by test 23) new code, shouldn't change anything: . introduced RTS_SET, RTS_UNSET, and RTS_ISSET macro's, and their LOCK variants. These macros set and clear the p_rts_flags field, causing a lot of duplicated logic like old_flags = rp->p_rts_flags; /* save value of the flags */ rp->p_rts_flags &= ~NO_PRIV; if (old_flags != 0 && rp->p_rts_flags == 0) lock_enqueue(rp); to change into the simpler RTS_LOCK_UNSET(rp, NO_PRIV); so the macros take care of calling dequeue() and enqueue() (or lock_*()), as the case may be). This makes the code a bit more readable and a bit less fragile. . removed return code from do_clocktick in CLOCK as it currently never replies . removed some debug code from VFS . fixed grant debug message in device.c preemptive checks, tests, changes: . added return code checks of receive() to SYSTEM and CLOCK . O_TRUNC should never arrive at MFS (added sanity check and removed O_TRUNC code) . user_path declared with PATH_MAX+1 to let it be null-terminated . checks in MFS to see if strings passed by VFS are null-terminated IS: . static irq name table thrown out
2007-02-01 18:50:02 +01:00
r = fs_sendrec(req->fs_e, &m);
/* Return new mode to caller. */
if(ch_mode) *ch_mode = m.RES_MODE;
return r;
}
/*===========================================================================*
* req_access *
*===========================================================================*/
PUBLIC int req_access(req)
access_req_t *req;
{
message m;
/* Fill in request message */
m.m_type = REQ_ACCESS;
m.REQ_INODE_NR = req->inode_nr;
m.REQ_MODE = req->amode;
m.REQ_UID = req->uid;
m.REQ_GID = req->gid;
/* Send/rec request */
return fs_sendrec(req->fs_e, &m);
}
/*===========================================================================*
* req_mknod *
*===========================================================================*/
PUBLIC int req_mknod(req)
mknod_req_t *req;
{
message m;
/* Fill in request message */
m.m_type = REQ_MKNOD;
m.REQ_INODE_NR = req->inode_nr;
m.REQ_MODE = req->rmode;
m.REQ_DEV = req->dev;
m.REQ_UID = req->uid;
m.REQ_GID = req->gid;
m.REQ_PATH = req->lastc;
m.REQ_PATH_LEN = strlen(req->lastc) + 1;
/* Send/rec request */
return fs_sendrec(req->fs_e, &m);
}
/*===========================================================================*
* req_mkdir *
*===========================================================================*/
PUBLIC int req_mkdir(req)
mkdir_req_t *req;
{
message m;
/* Fill in request message */
m.m_type = REQ_MKDIR;
m.REQ_INODE_NR = req->d_inode_nr;
m.REQ_MODE = req->rmode;
m.REQ_UID = req->uid;
m.REQ_GID = req->gid;
m.REQ_PATH = req->lastc;
m.REQ_PATH_LEN = strlen(req->lastc) + 1;
/* Send/rec request */
return fs_sendrec(req->fs_e, &m);
}
/*===========================================================================*
* req_inhibread *
*===========================================================================*/
PUBLIC int req_inhibread(req)
node_req_t *req;
{
message m;
/* Fill in request message */
m.m_type = REQ_INHIBREAD;
m.REQ_INODE_NR = req->inode_nr;
/* Send/rec request */
return fs_sendrec(req->fs_e, &m);
}
/*===========================================================================*
* req_stat *
*===========================================================================*/
2007-01-05 17:36:55 +01:00
PUBLIC int req_stat(fs_e, inode_nr, who_e, buf, pos)
int fs_e;
ino_t inode_nr;
int who_e;
char *buf;
int pos;
{
2007-01-05 17:36:55 +01:00
cp_grant_id_t gid;
int r;
message m;
struct stat sb;
2007-01-05 17:36:55 +01:00
if (pos != 0)
{
gid= cpf_grant_direct(fs_e, (vir_bytes)&sb, sizeof(struct stat),
CPF_WRITE);
}
else
{
gid= cpf_grant_magic(fs_e, who_e, (vir_bytes)buf, sizeof(struct stat),
CPF_WRITE);
}
if (gid < 0)
return gid;
2007-01-05 17:36:55 +01:00
/* Fill in request message */
m.m_type = REQ_STAT;
m.REQ_INODE_NR = inode_nr;
m.REQ_GRANT = gid;
2007-01-05 17:36:55 +01:00
/* Send/rec request */
r= fs_sendrec(fs_e, &m);
2007-01-05 17:36:55 +01:00
cpf_revoke(gid);
2007-01-05 17:36:55 +01:00
if (r == OK && pos != 0)
{
sb.st_size -= pos;
r= sys_vircopy(SELF, D, (vir_bytes)&sb, who_e, D, (vir_bytes)buf,
sizeof(struct stat));
}
2007-01-05 17:36:55 +01:00
return r;
}
/*===========================================================================*
* req_fstatfs *
*===========================================================================*/
2007-01-05 17:36:55 +01:00
PUBLIC int req_fstatfs(fs_e, inode_nr, who_e, buf)
int fs_e;
ino_t inode_nr;
int who_e;
char *buf;
{
2007-01-05 17:36:55 +01:00
int r;
cp_grant_id_t gid;
message m;
2007-01-05 17:36:55 +01:00
gid= cpf_grant_magic(fs_e, who_e, (vir_bytes)buf, sizeof(struct statfs),
CPF_WRITE);
if (gid < 0)
return gid;
2007-01-05 17:36:55 +01:00
/* Fill in request message */
m.m_type = REQ_FSTATFS;
m.REQ_INODE_NR = inode_nr;
m.REQ_GRANT = gid;
/* Send/rec request */
r= fs_sendrec(fs_e, &m);
cpf_revoke(gid);
return r;
}
/*===========================================================================*
* req_unlink *
*===========================================================================*/
PUBLIC int req_unlink(req)
unlink_req_t *req;
{
message m;
/* Fill in request message */
m.m_type = REQ_UNLINK;
m.REQ_INODE_NR = req->d_inode_nr;
m.REQ_UID = req->uid;
m.REQ_GID = req->gid;
m.REQ_PATH = req->lastc;
m.REQ_PATH_LEN = strlen(req->lastc) + 1;
/* Send/rec request */
return fs_sendrec(req->fs_e, &m);
}
/*===========================================================================*
* req_rmdir *
*===========================================================================*/
PUBLIC int req_rmdir(req)
unlink_req_t *req;
{
message m;
/* Fill in request message */
m.m_type = REQ_RMDIR;
m.REQ_INODE_NR = req->d_inode_nr;
m.REQ_UID = req->uid;
m.REQ_GID = req->gid;
m.REQ_PATH = req->lastc;
m.REQ_PATH_LEN = strlen(req->lastc) + 1;
/* Send/rec request */
return fs_sendrec(req->fs_e, &m);
}
/*===========================================================================*
* req_utime *
*===========================================================================*/
PUBLIC int req_utime(req)
utime_req_t *req;
{
message m;
/* Fill in request message */
m.m_type = REQ_UTIME;
m.REQ_INODE_NR = req->inode_nr;
m.REQ_UID = req->uid;
m.REQ_GID = req->gid;
m.REQ_ACTIME = req->actime;
m.REQ_MODTIME = req->modtime;
/* Send/rec request */
return fs_sendrec(req->fs_e, &m);
}
/*===========================================================================*
* req_stime *
*===========================================================================*/
PUBLIC int req_stime(fs_e, boottime)
endpoint_t fs_e;
time_t boottime;
{
message m;
/* Fill in request message */
m.m_type = REQ_STIME;
m.REQ_BOOTTIME = boottime;
/* Send/rec request */
return fs_sendrec(fs_e, &m);
}
/*===========================================================================*
* req_sync *
*===========================================================================*/
PUBLIC int req_sync(fs_e)
endpoint_t fs_e;
{
message m;
/* Fill in request message */
m.m_type = REQ_SYNC;
/* Send/rec request */
return fs_sendrec(fs_e, &m);
}
/*===========================================================================*
* req_link *
*===========================================================================*/
PUBLIC int req_link(req)
link_req_t *req;
{
message m;
/* Fill in request message */
m.m_type = REQ_LINK;
m.REQ_LINKED_FILE = req->linked_file;
m.REQ_LINK_PARENT = req->link_parent;
m.REQ_UID = req->uid;
m.REQ_GID = req->gid;
m.REQ_PATH = req->lastc;
m.REQ_PATH_LEN = strlen(req->lastc) + 1;
/* Send/rec request */
return fs_sendrec(req->fs_e, &m);
}
/*===========================================================================*
* req_slink *
*===========================================================================*/
PUBLIC int req_slink(req)
slink_req_t *req;
{
message m;
/* Fill in request message */
m.m_type = REQ_SLINK;
m.REQ_INODE_NR = req->parent_dir;
m.REQ_UID = req->uid;
m.REQ_GID = req->gid;
m.REQ_PATH = req->lastc;
m.REQ_PATH_LEN = strlen(req->lastc) + 1;
m.REQ_WHO_E = req->who_e;
m.REQ_USER_ADDR = req->path_addr;
m.REQ_SLENGTH = req->path_length;
/* Send/rec request */
return fs_sendrec(req->fs_e, &m);
}
/*===========================================================================*
* req_rdlink *
*===========================================================================*/
PUBLIC int req_rdlink(req)
rdlink_req_t *req;
{
message m;
/* Fill in request message */
m.m_type = REQ_RDLINK;
m.REQ_INODE_NR = req->inode_nr;
m.REQ_UID = req->uid;
m.REQ_GID = req->gid;
m.REQ_WHO_E = req->who_e;
m.REQ_USER_ADDR = req->path_buffer;
m.REQ_SLENGTH = req->max_length;
/* Send/rec request */
return fs_sendrec(req->fs_e, &m);
}
/*===========================================================================*
* req_rename *
*===========================================================================*/
PUBLIC int req_rename(req)
rename_req_t *req;
{
message m;
/* Fill in request message */
m.m_type = REQ_RENAME;
m.REQ_OLD_DIR = req->old_dir;
m.REQ_NEW_DIR = req->new_dir;
m.REQ_UID = req->uid;
m.REQ_GID = req->gid;
m.REQ_PATH = req->old_name;
m.REQ_PATH_LEN = strlen(req->old_name) + 1;
m.REQ_USER_ADDR = req->new_name;
m.REQ_SLENGTH = strlen(req->new_name) + 1;
/* Send/rec request */
return fs_sendrec(req->fs_e, &m);
}
/*===========================================================================*
* req_mountpoint *
*===========================================================================*/
PUBLIC int req_mountpoint(req, res)
mountpoint_req_t *req;
node_details_t *res;
{
int r;
message m;
/* Fill in request message */
m.m_type = REQ_MOUNTPOINT;
m.REQ_INODE_NR = req->inode_nr;
m.REQ_UID = req->uid;
m.REQ_GID = req->gid;
/* Send/rec request */
if ((r = fs_sendrec(req->fs_e, &m)) != OK) return r;
/* Fill in response structure */
res->fs_e = m.m_source;
res->inode_nr = m.RES_INODE_NR;
res->fmode = m.RES_MODE;
res->fsize = m.RES_FILE_SIZE;
return OK;
}
/*===========================================================================*
* req_readsuper *
*===========================================================================*/
PUBLIC int req_readsuper(req, res)
readsuper_req_t *req;
readsuper_res_t *res;
{
int r;
message m;
/* Fill in request message */
m.m_type = REQ_READSUPER;
m.REQ_READONLY = req->readonly;
m.REQ_BOOTTIME = req->boottime;
m.REQ_DRIVER_E = req->driver_e;
m.REQ_DEV = req->dev;
m.REQ_SLINK_STORAGE = req->slink_storage;
m.REQ_ISROOT = req->isroot;
/* Send/rec request */
if ((r = fs_sendrec(req->fs_e, &m)) != OK) return r;
/* Fill in response structure */
res->fs_e = m.m_source;
res->inode_nr = m.RES_INODE_NR;
res->fmode = m.RES_MODE;
res->fsize = m.RES_FILE_SIZE;
res->blocksize = m.RES_BLOCKSIZE;
res->maxsize = m.RES_MAXSIZE;
return OK;
}
/*===========================================================================*
* req_unmount *
*===========================================================================*/
PUBLIC int req_unmount(fs_e)
endpoint_t fs_e;
{
message m;
/* Fill in request message */
m.m_type = REQ_UNMOUNT;
/* Send/rec request */
return fs_sendrec(fs_e, &m);
}
/*===========================================================================*
* req_trunc *
*===========================================================================*/
PUBLIC int req_trunc(req)
trunc_req_t *req;
{
message m;
/* Fill in request message */
m.m_type = REQ_TRUNC;
m.REQ_FD_INODE_NR = req->inode_nr;
m.REQ_UID = req->uid;
m.REQ_GID = req->gid;
m.REQ_LENGTH = req->length;
/* Send/rec request */
return fs_sendrec(req->fs_e, &m);
}
/*===========================================================================*
* req_newdriver *
*===========================================================================*/
PUBLIC int req_newdriver(fs_e, dev, driver_e)
endpoint_t fs_e;
Dev_t dev;
endpoint_t driver_e;
{
/* Note: this is the only request function that doesn't use the
* fs_sendrec internal routine, since we want to avoid the dead
* driver recovery mechanism here. This function is actually called
* during the recovery.
*/
message m;
int r;
/* Fill in request message */
m.m_type = REQ_NEW_DRIVER;
m.REQ_DEV = dev;
m.REQ_DRIVER_E = driver_e;
/* Issue request */
if ((r = sendrec(fs_e, &m)) != OK) {
printf("VFSreq_newdriver: error sending message to %d: %d\n", fs_e, r);
return r;
}
return OK;
}
/*===========================================================================*
* req_lookup *
*===========================================================================*/
PUBLIC int req_lookup(req, res)
lookup_req_t *req;
lookup_res_t *res;
{
int r;
message m;
/* Fill in request message */
m.m_type = REQ_LOOKUP;
m.REQ_PATH = req->path;
m.REQ_PATH_LEN = strlen(req->path) + 1;
m.REQ_USER_ADDR = req->lastc;
m.REQ_FLAGS = req->flags;
m.REQ_INODE_NR = req->start_dir;
m.REQ_CHROOT_NR = req->root_dir;
m.REQ_UID = req->uid;
m.REQ_GID = req->gid;
m.REQ_SYMLOOP = req->symloop;
/* Send/rec request */
r = fs_sendrec(req->fs_e, &m);
/* Fill in response according to the return value */
res->fs_e = m.m_source;
switch (r) {
case OK:
2007-01-05 17:36:55 +01:00
default:
res->inode_nr = m.RES_INODE_NR;
res->fmode = m.RES_MODE;
res->fsize = m.RES_FILE_SIZE;
res->dev = m.RES_DEV;
2007-01-05 17:36:55 +01:00
res->uid= m.RES_UID;
res->gid= m.RES_GID;
res->char_processed = m.RES_OFFSET; /* For ENOENT */
break;
case EENTERMOUNT:
res->inode_nr = m.RES_INODE_NR;
case ELEAVEMOUNT:
case ESYMLINK:
res->char_processed = m.RES_OFFSET;
res->symloop = m.RES_SYMLOOP;
break;
}
return r;
}
/*===========================================================================*
* req_breadwrite *
*===========================================================================*/
int req_breadwrite(req, res)
breadwrite_req_t *req;
readwrite_res_t *res;
{
int r;
message m;
/* Fill in request message */
m.m_type = req->rw_flag == READING ? REQ_BREAD : REQ_BWRITE;
m.REQ_XFD_BDEV = req->dev;
m.REQ_XFD_BLOCK_SIZE = req->blocksize;
m.REQ_XFD_WHO_E = req->user_e;
m.REQ_XFD_POS_LO = ex64lo(req->pos);
m.REQ_XFD_POS_HI = ex64hi(req->pos);
m.REQ_XFD_NBYTES = req->num_of_bytes;
m.REQ_XFD_USER_ADDR = req->user_addr;
/* Send/rec request */
if ((r = fs_sendrec(req->fs_e, &m)) != OK) return r;
/* Fill in response structure */
res->new_pos = make64(m.RES_XFD_POS_LO, m.RES_XFD_POS_HI);
res->cum_io = m.RES_XFD_CUM_IO;
return OK;
}
PUBLIC int req_getdents(fs_e, inode_nr, pos, gid, size, pos_change)
endpoint_t fs_e;
ino_t inode_nr;
off_t pos;
cp_grant_id_t gid;
size_t size;
off_t *pos_change;
{
int r;
message m;
m.m_type= REQ_GETDENTS;
m.REQ_GDE_INODE= inode_nr;
m.REQ_GDE_GRANT= gid;
m.REQ_GDE_SIZE= size;
m.REQ_GDE_POS= pos;
r = fs_sendrec(fs_e, &m);
*pos_change= m.RES_GDE_POS_CHANGE;
return r;
}
/*===========================================================================*
* req_flush *
*===========================================================================*/
PUBLIC int req_flush(fs_e, dev)
endpoint_t fs_e;
dev_t dev;
{
message m;
/* Fill in request message */
m.m_type = REQ_FLUSH;
m.REQ_DEV = dev;
/* Send/rec request */
return fs_sendrec(fs_e, &m);
}
#if 0
/* Wrapper pattern: */
/*===========================================================================*
* req_ *
*===========================================================================*/
PUBLIC int req_(req, res)
_req_t *req;
_t *res;
{
int r;
message m;
/* Fill in request message */
/* Send/rec request */
if ((r = fs_sendrec(req->fs_e, &m)) != OK) return r;
/* Fill in response structure */
return OK;
}
#endif
/*===========================================================================*
* fs_sendrec *
*===========================================================================*/
Mostly bugfixes of bugs triggered by the test set. bugfixes: SYSTEM: . removed rc->p_priv->s_flags = 0; for the priv struct shared by all user processes in get_priv(). this should only be done once. doing a SYS_PRIV_USER in sys_privctl() caused the flags of all user processes to be reset, so they were no longer PREEMPTIBLE. this happened when RS executed a policy script. (this broke test1 in the test set) VFS/MFS: . chown can change the mode of a file, and chmod arguments are only part of the full file mode so the full filemode is slightly magic. changed these calls so that the final modes are returned to VFS, so that the vnode can be kept up-to-date. (this broke test11 in the test set) MFS: . lookup() checked for sizeof(string) instead of sizeof(user_path), truncating long path names (caught by test 23) . truncate functions neglected to update ctime (this broke test16) VFS: . corner case of an empty filename lookup caused fields of a request not to be filled in in the lookup functions, not making it clear that the lookup had failed, causing messages to garbage processes, causing strange failures. (caught by test 30) . trust v_size in vnode when doing reads or writes on non-special files, truncating i/o where necessary; this is necessary for pipes, as MFS can't tell when a pipe has been truncated without it being told explicitly each time. when the last reader/writer on a pipe closes, tell FS about the new size using truncate_vn(). (this broke test 25, among others) . permission check for chdir() had disappeared; added a forbidden() call (caught by test 23) new code, shouldn't change anything: . introduced RTS_SET, RTS_UNSET, and RTS_ISSET macro's, and their LOCK variants. These macros set and clear the p_rts_flags field, causing a lot of duplicated logic like old_flags = rp->p_rts_flags; /* save value of the flags */ rp->p_rts_flags &= ~NO_PRIV; if (old_flags != 0 && rp->p_rts_flags == 0) lock_enqueue(rp); to change into the simpler RTS_LOCK_UNSET(rp, NO_PRIV); so the macros take care of calling dequeue() and enqueue() (or lock_*()), as the case may be). This makes the code a bit more readable and a bit less fragile. . removed return code from do_clocktick in CLOCK as it currently never replies . removed some debug code from VFS . fixed grant debug message in device.c preemptive checks, tests, changes: . added return code checks of receive() to SYSTEM and CLOCK . O_TRUNC should never arrive at MFS (added sanity check and removed O_TRUNC code) . user_path declared with PATH_MAX+1 to let it be null-terminated . checks in MFS to see if strings passed by VFS are null-terminated IS: . static irq name table thrown out
2007-02-01 18:50:02 +01:00
PRIVATE int fs_sendrec_f(char *file, int line, endpoint_t fs_e, message *reqm)
{
/* This is the low level function that sends requests to FS processes.
* It also handles driver recovery mechanism and reissuing the
* request which failed due to a dead driver.
*/
int r, old_driver_e, new_driver_e;
message origm, m;
struct vmnt *vmp;
Mostly bugfixes of bugs triggered by the test set. bugfixes: SYSTEM: . removed rc->p_priv->s_flags = 0; for the priv struct shared by all user processes in get_priv(). this should only be done once. doing a SYS_PRIV_USER in sys_privctl() caused the flags of all user processes to be reset, so they were no longer PREEMPTIBLE. this happened when RS executed a policy script. (this broke test1 in the test set) VFS/MFS: . chown can change the mode of a file, and chmod arguments are only part of the full file mode so the full filemode is slightly magic. changed these calls so that the final modes are returned to VFS, so that the vnode can be kept up-to-date. (this broke test11 in the test set) MFS: . lookup() checked for sizeof(string) instead of sizeof(user_path), truncating long path names (caught by test 23) . truncate functions neglected to update ctime (this broke test16) VFS: . corner case of an empty filename lookup caused fields of a request not to be filled in in the lookup functions, not making it clear that the lookup had failed, causing messages to garbage processes, causing strange failures. (caught by test 30) . trust v_size in vnode when doing reads or writes on non-special files, truncating i/o where necessary; this is necessary for pipes, as MFS can't tell when a pipe has been truncated without it being told explicitly each time. when the last reader/writer on a pipe closes, tell FS about the new size using truncate_vn(). (this broke test 25, among others) . permission check for chdir() had disappeared; added a forbidden() call (caught by test 23) new code, shouldn't change anything: . introduced RTS_SET, RTS_UNSET, and RTS_ISSET macro's, and their LOCK variants. These macros set and clear the p_rts_flags field, causing a lot of duplicated logic like old_flags = rp->p_rts_flags; /* save value of the flags */ rp->p_rts_flags &= ~NO_PRIV; if (old_flags != 0 && rp->p_rts_flags == 0) lock_enqueue(rp); to change into the simpler RTS_LOCK_UNSET(rp, NO_PRIV); so the macros take care of calling dequeue() and enqueue() (or lock_*()), as the case may be). This makes the code a bit more readable and a bit less fragile. . removed return code from do_clocktick in CLOCK as it currently never replies . removed some debug code from VFS . fixed grant debug message in device.c preemptive checks, tests, changes: . added return code checks of receive() to SYSTEM and CLOCK . O_TRUNC should never arrive at MFS (added sanity check and removed O_TRUNC code) . user_path declared with PATH_MAX+1 to let it be null-terminated . checks in MFS to see if strings passed by VFS are null-terminated IS: . static irq name table thrown out
2007-02-01 18:50:02 +01:00
/* Make a copy of the request so that we can load it back in
* case of a dead driver */
origm = *reqm;
for (;;) {
/* Do the actual send, receive */
if (OK != (r=sendrec(fs_e, reqm))) {
Mostly bugfixes of bugs triggered by the test set. bugfixes: SYSTEM: . removed rc->p_priv->s_flags = 0; for the priv struct shared by all user processes in get_priv(). this should only be done once. doing a SYS_PRIV_USER in sys_privctl() caused the flags of all user processes to be reset, so they were no longer PREEMPTIBLE. this happened when RS executed a policy script. (this broke test1 in the test set) VFS/MFS: . chown can change the mode of a file, and chmod arguments are only part of the full file mode so the full filemode is slightly magic. changed these calls so that the final modes are returned to VFS, so that the vnode can be kept up-to-date. (this broke test11 in the test set) MFS: . lookup() checked for sizeof(string) instead of sizeof(user_path), truncating long path names (caught by test 23) . truncate functions neglected to update ctime (this broke test16) VFS: . corner case of an empty filename lookup caused fields of a request not to be filled in in the lookup functions, not making it clear that the lookup had failed, causing messages to garbage processes, causing strange failures. (caught by test 30) . trust v_size in vnode when doing reads or writes on non-special files, truncating i/o where necessary; this is necessary for pipes, as MFS can't tell when a pipe has been truncated without it being told explicitly each time. when the last reader/writer on a pipe closes, tell FS about the new size using truncate_vn(). (this broke test 25, among others) . permission check for chdir() had disappeared; added a forbidden() call (caught by test 23) new code, shouldn't change anything: . introduced RTS_SET, RTS_UNSET, and RTS_ISSET macro's, and their LOCK variants. These macros set and clear the p_rts_flags field, causing a lot of duplicated logic like old_flags = rp->p_rts_flags; /* save value of the flags */ rp->p_rts_flags &= ~NO_PRIV; if (old_flags != 0 && rp->p_rts_flags == 0) lock_enqueue(rp); to change into the simpler RTS_LOCK_UNSET(rp, NO_PRIV); so the macros take care of calling dequeue() and enqueue() (or lock_*()), as the case may be). This makes the code a bit more readable and a bit less fragile. . removed return code from do_clocktick in CLOCK as it currently never replies . removed some debug code from VFS . fixed grant debug message in device.c preemptive checks, tests, changes: . added return code checks of receive() to SYSTEM and CLOCK . O_TRUNC should never arrive at MFS (added sanity check and removed O_TRUNC code) . user_path declared with PATH_MAX+1 to let it be null-terminated . checks in MFS to see if strings passed by VFS are null-terminated IS: . static irq name table thrown out
2007-02-01 18:50:02 +01:00
printf("VFS:fs_sendrec:%s:%d: error sending message. FS_e: %d req_nr: %d err: %d\n",
file, line, fs_e, reqm->m_type, r);
}
if(r == OK) {
/* Sendrec was okay */
break;
}
/* Dead driver */
if (r == EDEADSRCDST || r == EDSTDIED || r == ESRCDIED) {
old_driver_e = NONE;
Mostly bugfixes of bugs triggered by the test set. bugfixes: SYSTEM: . removed rc->p_priv->s_flags = 0; for the priv struct shared by all user processes in get_priv(). this should only be done once. doing a SYS_PRIV_USER in sys_privctl() caused the flags of all user processes to be reset, so they were no longer PREEMPTIBLE. this happened when RS executed a policy script. (this broke test1 in the test set) VFS/MFS: . chown can change the mode of a file, and chmod arguments are only part of the full file mode so the full filemode is slightly magic. changed these calls so that the final modes are returned to VFS, so that the vnode can be kept up-to-date. (this broke test11 in the test set) MFS: . lookup() checked for sizeof(string) instead of sizeof(user_path), truncating long path names (caught by test 23) . truncate functions neglected to update ctime (this broke test16) VFS: . corner case of an empty filename lookup caused fields of a request not to be filled in in the lookup functions, not making it clear that the lookup had failed, causing messages to garbage processes, causing strange failures. (caught by test 30) . trust v_size in vnode when doing reads or writes on non-special files, truncating i/o where necessary; this is necessary for pipes, as MFS can't tell when a pipe has been truncated without it being told explicitly each time. when the last reader/writer on a pipe closes, tell FS about the new size using truncate_vn(). (this broke test 25, among others) . permission check for chdir() had disappeared; added a forbidden() call (caught by test 23) new code, shouldn't change anything: . introduced RTS_SET, RTS_UNSET, and RTS_ISSET macro's, and their LOCK variants. These macros set and clear the p_rts_flags field, causing a lot of duplicated logic like old_flags = rp->p_rts_flags; /* save value of the flags */ rp->p_rts_flags &= ~NO_PRIV; if (old_flags != 0 && rp->p_rts_flags == 0) lock_enqueue(rp); to change into the simpler RTS_LOCK_UNSET(rp, NO_PRIV); so the macros take care of calling dequeue() and enqueue() (or lock_*()), as the case may be). This makes the code a bit more readable and a bit less fragile. . removed return code from do_clocktick in CLOCK as it currently never replies . removed some debug code from VFS . fixed grant debug message in device.c preemptive checks, tests, changes: . added return code checks of receive() to SYSTEM and CLOCK . O_TRUNC should never arrive at MFS (added sanity check and removed O_TRUNC code) . user_path declared with PATH_MAX+1 to let it be null-terminated . checks in MFS to see if strings passed by VFS are null-terminated IS: . static irq name table thrown out
2007-02-01 18:50:02 +01:00
/* Find old driver by endpoint */
for (vmp = &vmnt[0]; vmp < &vmnt[NR_MNTS]; ++vmp) {
Mostly bugfixes of bugs triggered by the test set. bugfixes: SYSTEM: . removed rc->p_priv->s_flags = 0; for the priv struct shared by all user processes in get_priv(). this should only be done once. doing a SYS_PRIV_USER in sys_privctl() caused the flags of all user processes to be reset, so they were no longer PREEMPTIBLE. this happened when RS executed a policy script. (this broke test1 in the test set) VFS/MFS: . chown can change the mode of a file, and chmod arguments are only part of the full file mode so the full filemode is slightly magic. changed these calls so that the final modes are returned to VFS, so that the vnode can be kept up-to-date. (this broke test11 in the test set) MFS: . lookup() checked for sizeof(string) instead of sizeof(user_path), truncating long path names (caught by test 23) . truncate functions neglected to update ctime (this broke test16) VFS: . corner case of an empty filename lookup caused fields of a request not to be filled in in the lookup functions, not making it clear that the lookup had failed, causing messages to garbage processes, causing strange failures. (caught by test 30) . trust v_size in vnode when doing reads or writes on non-special files, truncating i/o where necessary; this is necessary for pipes, as MFS can't tell when a pipe has been truncated without it being told explicitly each time. when the last reader/writer on a pipe closes, tell FS about the new size using truncate_vn(). (this broke test 25, among others) . permission check for chdir() had disappeared; added a forbidden() call (caught by test 23) new code, shouldn't change anything: . introduced RTS_SET, RTS_UNSET, and RTS_ISSET macro's, and their LOCK variants. These macros set and clear the p_rts_flags field, causing a lot of duplicated logic like old_flags = rp->p_rts_flags; /* save value of the flags */ rp->p_rts_flags &= ~NO_PRIV; if (old_flags != 0 && rp->p_rts_flags == 0) lock_enqueue(rp); to change into the simpler RTS_LOCK_UNSET(rp, NO_PRIV); so the macros take care of calling dequeue() and enqueue() (or lock_*()), as the case may be). This makes the code a bit more readable and a bit less fragile. . removed return code from do_clocktick in CLOCK as it currently never replies . removed some debug code from VFS . fixed grant debug message in device.c preemptive checks, tests, changes: . added return code checks of receive() to SYSTEM and CLOCK . O_TRUNC should never arrive at MFS (added sanity check and removed O_TRUNC code) . user_path declared with PATH_MAX+1 to let it be null-terminated . checks in MFS to see if strings passed by VFS are null-terminated IS: . static irq name table thrown out
2007-02-01 18:50:02 +01:00
if (vmp->m_fs_e == fs_e) { /* found FS */
old_driver_e = vmp->m_driver_e;
dmap_unmap_by_endpt(old_driver_e); /* unmap driver */
break;
}
}
/* No FS ?? */
Mostly bugfixes of bugs triggered by the test set. bugfixes: SYSTEM: . removed rc->p_priv->s_flags = 0; for the priv struct shared by all user processes in get_priv(). this should only be done once. doing a SYS_PRIV_USER in sys_privctl() caused the flags of all user processes to be reset, so they were no longer PREEMPTIBLE. this happened when RS executed a policy script. (this broke test1 in the test set) VFS/MFS: . chown can change the mode of a file, and chmod arguments are only part of the full file mode so the full filemode is slightly magic. changed these calls so that the final modes are returned to VFS, so that the vnode can be kept up-to-date. (this broke test11 in the test set) MFS: . lookup() checked for sizeof(string) instead of sizeof(user_path), truncating long path names (caught by test 23) . truncate functions neglected to update ctime (this broke test16) VFS: . corner case of an empty filename lookup caused fields of a request not to be filled in in the lookup functions, not making it clear that the lookup had failed, causing messages to garbage processes, causing strange failures. (caught by test 30) . trust v_size in vnode when doing reads or writes on non-special files, truncating i/o where necessary; this is necessary for pipes, as MFS can't tell when a pipe has been truncated without it being told explicitly each time. when the last reader/writer on a pipe closes, tell FS about the new size using truncate_vn(). (this broke test 25, among others) . permission check for chdir() had disappeared; added a forbidden() call (caught by test 23) new code, shouldn't change anything: . introduced RTS_SET, RTS_UNSET, and RTS_ISSET macro's, and their LOCK variants. These macros set and clear the p_rts_flags field, causing a lot of duplicated logic like old_flags = rp->p_rts_flags; /* save value of the flags */ rp->p_rts_flags &= ~NO_PRIV; if (old_flags != 0 && rp->p_rts_flags == 0) lock_enqueue(rp); to change into the simpler RTS_LOCK_UNSET(rp, NO_PRIV); so the macros take care of calling dequeue() and enqueue() (or lock_*()), as the case may be). This makes the code a bit more readable and a bit less fragile. . removed return code from do_clocktick in CLOCK as it currently never replies . removed some debug code from VFS . fixed grant debug message in device.c preemptive checks, tests, changes: . added return code checks of receive() to SYSTEM and CLOCK . O_TRUNC should never arrive at MFS (added sanity check and removed O_TRUNC code) . user_path declared with PATH_MAX+1 to let it be null-terminated . checks in MFS to see if strings passed by VFS are null-terminated IS: . static irq name table thrown out
2007-02-01 18:50:02 +01:00
if (old_driver_e == NONE)
panic(__FILE__, "VFSdead_driver: couldn't find FS\n", fs_e);
/* Wait for a new driver. */
for (;;) {
new_driver_e = 0;
printf("VFSdead_driver: waiting for new driver\n");
r = receive(RS_PROC_NR, &m);
if (r != OK) {
panic(__FILE__, "VFSdead_driver: unable to receive from RS",
r);
}
if (m.m_type == DEVCTL) {
/* Map new driver */
r = fs_devctl(m.ctl_req, m.dev_nr, m.driver_nr,
m.dev_style, m.m_force);
if (m.ctl_req == DEV_MAP && r == OK) {
new_driver_e = m.driver_nr;
printf("VFSdead_driver: new driver endpoint: %d\n",
new_driver_e);
}
}
else {
panic(__FILE__, "VFSdead_driver: got message from RS, type",
m.m_type);
}
m.m_type = r;
if ((r = send(RS_PROC_NR, &m)) != OK) {
panic(__FILE__, "VFSdead_driver: unable to send to RS",
r);
}
/* New driver is ready */
if (new_driver_e) break;
}
/* Copy back original request */
*reqm = origm;
continue;
}
printf("fs_sendrec: unhandled error %d sending to %d\n", r, fs_e);
panic(__FILE__, "fs_sendrec: unhandled error", NO_NUM);
}
/* Return message type */
return reqm->m_type;
}