sanchayanmaity/minix
1
0
Fork 0
Code Issues Pull requests Packages Projects Releases Wiki Activity

VFS: change locking to ease concurrent FSes

Browse source 
This patch uses stricter locking for REQ_LINK, REQ_MKDIR, REQ_MKNOD,
REQ_RENAME, REQ_RMDIR, REQ_SLINK and REQ_UNLINK. For all requests, VFS
locks the directory in which we add or remove an inode with VNODE_WRITE.
I.e., the operations have exclusive access to that directory.

Furthermore, REQ_CHOWN, REQ_CHMOD, and REQ_FTRUNC now lock the vmnt
VMNT_READ; VMNT_WRITE was unnecessary.
This commit is contained in:
Thomas Veerman 2012-12-21 15:30:37 +00:00
parent 93103f497f
commit 23c5f56e32
4 changed files with 50 additions and 36 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

54
servers/vfs/link.c
View file

@ -56,7 +56,7 @@ int do_link()
/* Does the final directory of 'name2' exist? */ /* Does the final directory of 'name2' exist? */
lookup_init(&resolve, fullpath, PATH_NOFLAGS, &vmp2, &dirp); lookup_init(&resolve, fullpath, PATH_NOFLAGS, &vmp2, &dirp);
resolve.l_vmnt_lock = VMNT_READ; resolve.l_vmnt_lock = VMNT_READ;
resolve.l_vnode_lock = VNODE_READ; resolve.l_vnode_lock = VNODE_WRITE;
if (fetch_name(vname2, vname2_length, fullpath) != OK) if (fetch_name(vname2, vname2_length, fullpath) != OK)
r = err_code; r = err_code;
else if ((dirp = last_dir(&resolve, fp)) == NULL) else if ((dirp = last_dir(&resolve, fp)) == NULL)
@ -98,11 +98,11 @@ int do_unlink()
* may be used by the superuser to do dangerous things; rmdir() may not. * may be used by the superuser to do dangerous things; rmdir() may not.
* The syscall might provide 'name' embedded in the message. * The syscall might provide 'name' embedded in the message.
*/ */
struct vnode *dirp, *vp; struct vnode *dirp, *dirp_l, *vp;
struct vmnt *vmp, *vmp2; struct vmnt *vmp, *vmp2;
int r; int r;
char fullpath[PATH_MAX]; char fullpath[PATH_MAX];
struct lookup resolve; struct lookup resolve, stickycheck;
vir_bytes vname; vir_bytes vname;
size_t vname_length; size_t vname_length;
@ -114,13 +114,12 @@ int do_unlink()
return(err_code); return(err_code);
} }
lookup_init(&resolve, fullpath, PATH_RET_SYMLINK, &vmp, &dirp); lookup_init(&resolve, fullpath, PATH_RET_SYMLINK, &vmp, &dirp_l);
resolve.l_vmnt_lock = VMNT_WRITE; resolve.l_vmnt_lock = VMNT_WRITE;
resolve.l_vnode_lock = VNODE_READ; resolve.l_vnode_lock = VNODE_WRITE;
/* Get the last directory in the path. */ /* Get the last directory in the path. */
if ((dirp = last_dir(&resolve, fp)) == NULL) return(err_code); if ((dirp = last_dir(&resolve, fp)) == NULL) return(err_code);
assert(vmp != NULL); /* We must have locked the vmnt */
/* Make sure that the object is a directory */ /* Make sure that the object is a directory */
if (!S_ISDIR(dirp->v_mode)) { if (!S_ISDIR(dirp->v_mode)) {
@ -142,12 +141,10 @@ int do_unlink()
user is allowed to unlink */ user is allowed to unlink */
if ((dirp->v_mode & S_ISVTX) == S_ISVTX) { if ((dirp->v_mode & S_ISVTX) == S_ISVTX) {
/* Look up inode of file to unlink to retrieve owner */ /* Look up inode of file to unlink to retrieve owner */
resolve.l_flags = PATH_RET_SYMLINK; lookup_init(&stickycheck, resolve.l_path, PATH_RET_SYMLINK, &vmp2, &vp);
resolve.l_vmp = &vmp2; /* Shouldn't actually get locked */ stickycheck.l_vmnt_lock = VMNT_READ;
resolve.l_vmnt_lock = VMNT_READ; stickycheck.l_vnode_lock = VNODE_READ;
resolve.l_vnode = &vp; vp = advance(dirp, &stickycheck, fp);
resolve.l_vnode_lock = VNODE_READ;
vp = advance(dirp, &resolve, fp);
assert(vmp2 == NULL); assert(vmp2 == NULL);
if (vp != NULL) { if (vp != NULL) {
if (vp->v_uid != fp->fp_effuid && fp->fp_effuid != SU_UID) if (vp->v_uid != fp->fp_effuid && fp->fp_effuid != SU_UID)
@ -164,7 +161,6 @@ int do_unlink()
} }
} }
assert(vmp != NULL);
upgrade_vmnt_lock(vmp); upgrade_vmnt_lock(vmp);
if (job_call_nr == UNLINK) if (job_call_nr == UNLINK)
@ -184,11 +180,11 @@ int do_rename()
{ {
/* Perform the rename(name1, name2) system call. */ /* Perform the rename(name1, name2) system call. */
int r = OK, r1; int r = OK, r1;
struct vnode *old_dirp, *new_dirp = NULL, *vp; struct vnode *old_dirp, *new_dirp, *new_dirp_l, *vp;
struct vmnt *oldvmp, *newvmp, *vmp2; struct vmnt *oldvmp, *newvmp, *vmp2;
char old_name[PATH_MAX]; char old_name[PATH_MAX];
char fullpath[PATH_MAX]; char fullpath[PATH_MAX];
struct lookup resolve; struct lookup resolve, stickycheck;
vir_bytes vname1, vname2; vir_bytes vname1, vname2;
size_t vname1_length, vname2_length; size_t vname1_length, vname2_length;
@ -200,7 +196,7 @@ int do_rename()
lookup_init(&resolve, fullpath, PATH_RET_SYMLINK, &oldvmp, &old_dirp); lookup_init(&resolve, fullpath, PATH_RET_SYMLINK, &oldvmp, &old_dirp);
/* Do not yet request exclusive lock on vmnt to prevent deadlocks later on */ /* Do not yet request exclusive lock on vmnt to prevent deadlocks later on */
resolve.l_vmnt_lock = VMNT_WRITE; resolve.l_vmnt_lock = VMNT_WRITE;
resolve.l_vnode_lock = VNODE_READ; resolve.l_vnode_lock = VNODE_WRITE;
/* See if 'name1' (existing file) exists. Get dir and file inodes. */ /* See if 'name1' (existing file) exists. Get dir and file inodes. */
if (fetch_name(vname1, vname1_length, fullpath) != OK) return(err_code); if (fetch_name(vname1, vname1_length, fullpath) != OK) return(err_code);
@ -210,10 +206,10 @@ int do_rename()
user is allowed to rename */ user is allowed to rename */
if ((old_dirp->v_mode & S_ISVTX) == S_ISVTX) { if ((old_dirp->v_mode & S_ISVTX) == S_ISVTX) {
/* Look up inode of file to unlink to retrieve owner */ /* Look up inode of file to unlink to retrieve owner */
lookup_init(&resolve, resolve.l_path, PATH_RET_SYMLINK, &vmp2, &vp); lookup_init(&stickycheck, resolve.l_path, PATH_RET_SYMLINK, &vmp2, &vp);
resolve.l_vmnt_lock = VMNT_READ; stickycheck.l_vmnt_lock = VMNT_READ;
resolve.l_vnode_lock = VNODE_READ; stickycheck.l_vnode_lock = VNODE_READ;
vp = advance(old_dirp, &resolve, fp); vp = advance(old_dirp, &stickycheck, fp);
assert(vmp2 == NULL); assert(vmp2 == NULL);
if (vp != NULL) { if (vp != NULL) {
if(vp->v_uid != fp->fp_effuid && fp->fp_effuid != SU_UID) if(vp->v_uid != fp->fp_effuid && fp->fp_effuid != SU_UID)
@ -240,12 +236,19 @@ int do_rename()
strlcpy(old_name, fullpath, PATH_MAX); strlcpy(old_name, fullpath, PATH_MAX);
/* See if 'name2' (new name) exists. Get dir inode */ /* See if 'name2' (new name) exists. Get dir inode */
lookup_init(&resolve, fullpath, PATH_RET_SYMLINK, &newvmp, &new_dirp); lookup_init(&resolve, fullpath, PATH_RET_SYMLINK, &newvmp, &new_dirp_l);
resolve.l_vmnt_lock = VMNT_READ; resolve.l_vmnt_lock = VMNT_READ;
resolve.l_vnode_lock = VNODE_READ; resolve.l_vnode_lock = VNODE_WRITE;
if (fetch_name(vname2, vname2_length, fullpath) != OK) r = err_code; if (fetch_name(vname2, vname2_length, fullpath) != OK) r = err_code;
else if ((new_dirp = last_dir(&resolve, fp)) == NULL) r = err_code; else if ((new_dirp = last_dir(&resolve, fp)) == NULL) r = err_code;
/* We used a separate vnode pointer to see whether we obtained a lock on the
* new_dirp vnode. If the new directory and old directory are the same, then
* the VNODE_WRITE lock on new_dirp will fail. In that case, new_dirp_l will
* be NULL, but new_dirp will not.
*/
if (new_dirp == old_dirp) assert(new_dirp_l == NULL);
if (r != OK) { if (r != OK) {
unlock_vnode(old_dirp); unlock_vnode(old_dirp);
unlock_vmnt(oldvmp); unlock_vmnt(oldvmp);
@ -265,9 +268,10 @@ int do_rename()
r = req_rename(old_dirp->v_fs_e, old_dirp->v_inode_nr, old_name, r = req_rename(old_dirp->v_fs_e, old_dirp->v_inode_nr, old_name,
new_dirp->v_inode_nr, fullpath); new_dirp->v_inode_nr, fullpath);
} }
unlock_vnode(old_dirp); unlock_vnode(old_dirp);
unlock_vnode(new_dirp);
unlock_vmnt(oldvmp); unlock_vmnt(oldvmp);
if (new_dirp_l) unlock_vnode(new_dirp_l);
if (newvmp) unlock_vmnt(newvmp); if (newvmp) unlock_vmnt(newvmp);
put_vnode(old_dirp); put_vnode(old_dirp);
@ -299,7 +303,7 @@ int do_truncate()
vname_length = job_m_in.m2_i1; vname_length = job_m_in.m2_i1;
lookup_init(&resolve, fullpath, PATH_NOFLAGS, &vmp, &vp); lookup_init(&resolve, fullpath, PATH_NOFLAGS, &vmp, &vp);
resolve.l_vmnt_lock = VMNT_EXCL; resolve.l_vmnt_lock = VMNT_READ;
resolve.l_vnode_lock = VNODE_WRITE; resolve.l_vnode_lock = VNODE_WRITE;
length = (off_t) job_m_in.flength; length = (off_t) job_m_in.flength;
@ -404,7 +408,7 @@ int do_slink()
lookup_init(&resolve, fullpath, PATH_NOFLAGS, &vmp, &vp); lookup_init(&resolve, fullpath, PATH_NOFLAGS, &vmp, &vp);
resolve.l_vmnt_lock = VMNT_WRITE; resolve.l_vmnt_lock = VMNT_WRITE;
resolve.l_vnode_lock = VNODE_READ; resolve.l_vnode_lock = VNODE_WRITE;
vname1 = (vir_bytes) job_m_in.name1; vname1 = (vir_bytes) job_m_in.name1;
vname1_length = job_m_in.name1_length; vname1_length = job_m_in.name1_length;

4
servers/vfs/open.c
View file

@ -547,7 +547,7 @@ int do_mknod()
lookup_init(&resolve, fullpath, PATH_NOFLAGS, &vmp, &vp); lookup_init(&resolve, fullpath, PATH_NOFLAGS, &vmp, &vp);
resolve.l_vmnt_lock = VMNT_WRITE; resolve.l_vmnt_lock = VMNT_WRITE;
resolve.l_vnode_lock = VNODE_READ; resolve.l_vnode_lock = VNODE_WRITE;
/* Only the super_user may make nodes other than fifos. */ /* Only the super_user may make nodes other than fifos. */
if (!super_user && (!S_ISFIFO(mode_bits) && !S_ISSOCK(mode_bits))) { if (!super_user && (!S_ISFIFO(mode_bits) && !S_ISSOCK(mode_bits))) {
@ -595,7 +595,7 @@ int do_mkdir()
lookup_init(&resolve, fullpath, PATH_NOFLAGS, &vmp, &vp); lookup_init(&resolve, fullpath, PATH_NOFLAGS, &vmp, &vp);
resolve.l_vmnt_lock = VMNT_WRITE; resolve.l_vmnt_lock = VMNT_WRITE;
resolve.l_vnode_lock = VNODE_READ; resolve.l_vnode_lock = VNODE_WRITE;
if (fetch_name(vname1, vname1_length, fullpath) != OK) return(err_code); if (fetch_name(vname1, vname1_length, fullpath) != OK) return(err_code);
bits = I_DIRECTORY | (dirmode & RWX_MODES & fp->fp_umask); bits = I_DIRECTORY | (dirmode & RWX_MODES & fp->fp_umask);

24
servers/vfs/path.c
View file

@ -168,7 +168,7 @@ struct fproc *rfp;
size_t len; size_t len;
char *cp; char *cp;
char dir_entry[NAME_MAX+1]; char dir_entry[NAME_MAX+1];
struct vnode *start_dir, *res_vp, *sym_vp, *loop_start; struct vnode *start_dir, *res_vp, *sym_vp, *sym_vp_l, *loop_start;
struct vmnt *sym_vmp = NULL; struct vmnt *sym_vmp = NULL;
int r, symloop = 0, ret_on_symlink = 0; int r, symloop = 0, ret_on_symlink = 0;
struct lookup symlink; struct lookup symlink;
@ -256,12 +256,18 @@ struct fproc *rfp;
strlcpy(resolve->l_path, dir_entry, NAME_MAX + 1); strlcpy(resolve->l_path, dir_entry, NAME_MAX + 1);
/* Look up the directory entry, but do not follow the symlink when it /* Look up the directory entry, but do not follow the symlink when it
* is one. * is one. Note: depending on the previous advance, we might not be
* able to lock the resulting vnode. For example, when we look up "./."
* and request a VNODE_WRITE lock on the result, then the previous
* advance has "./" locked. The next advance to "." will try to lock
* the same vnode with a VNODE_READ lock, and fail. When that happens,
* sym_vp_l will be NULL and we must not unlock the vnode. If we would
* unlock, we actually unlock the vnode locked by the previous advance.
*/ */
lookup_init(&symlink, resolve->l_path, lookup_init(&symlink, resolve->l_path,
resolve->l_flags|PATH_RET_SYMLINK, &sym_vmp, &sym_vp); resolve->l_flags|PATH_RET_SYMLINK, &sym_vmp, &sym_vp_l);
symlink.l_vnode_lock = VNODE_READ;
symlink.l_vmnt_lock = VMNT_READ; symlink.l_vmnt_lock = VMNT_READ;
symlink.l_vnode_lock = VNODE_READ;
sym_vp = advance(res_vp, &symlink, rfp); sym_vp = advance(res_vp, &symlink, rfp);
if (sym_vp == NULL) break; if (sym_vp == NULL) break;
@ -291,7 +297,8 @@ struct fproc *rfp;
resolve->l_path[r] = '\0'; resolve->l_path[r] = '\0';
if (strrchr(resolve->l_path, '/') != NULL) { if (strrchr(resolve->l_path, '/') != NULL) {
unlock_vnode(sym_vp); if (sym_vp_l != NULL)
unlock_vnode(sym_vp);
unlock_vmnt(*resolve->l_vmp); unlock_vmnt(*resolve->l_vmp);
if (sym_vmp != NULL) if (sym_vmp != NULL)
unlock_vmnt(sym_vmp); unlock_vmnt(sym_vmp);
@ -323,7 +330,8 @@ struct fproc *rfp;
assert(sym_vmp != NULL); assert(sym_vmp != NULL);
/* Unlock final file, it might have wrong lock types */ /* Unlock final file, it might have wrong lock types */
unlock_vnode(sym_vp); if (sym_vp_l != NULL)
unlock_vnode(sym_vp);
unlock_vmnt(sym_vmp); unlock_vmnt(sym_vmp);
put_vnode(sym_vp); put_vnode(sym_vp);
sym_vp = NULL; sym_vp = NULL;
@ -357,7 +365,9 @@ struct fproc *rfp;
} }
if (sym_vp != NULL) { if (sym_vp != NULL) {
unlock_vnode(sym_vp); if (sym_vp_l != NULL) {
unlock_vnode(sym_vp);
}
if (sym_vmp != NULL) { if (sym_vmp != NULL) {
unlock_vmnt(sym_vmp); unlock_vmnt(sym_vmp);
} }

4
servers/vfs/protect.c
View file

@ -47,7 +47,7 @@ int do_chmod()
new_mode = (mode_t) job_m_in.mode; new_mode = (mode_t) job_m_in.mode;
lookup_init(&resolve, fullpath, PATH_NOFLAGS, &vmp, &vp); lookup_init(&resolve, fullpath, PATH_NOFLAGS, &vmp, &vp);
resolve.l_vmnt_lock = VMNT_WRITE; resolve.l_vmnt_lock = VMNT_READ;
resolve.l_vnode_lock = VNODE_WRITE; resolve.l_vnode_lock = VNODE_WRITE;
if (job_call_nr == CHMOD) { if (job_call_nr == CHMOD) {
@ -123,7 +123,7 @@ int do_chown()
gid = job_m_in.group; gid = job_m_in.group;
lookup_init(&resolve, fullpath, PATH_NOFLAGS, &vmp, &vp); lookup_init(&resolve, fullpath, PATH_NOFLAGS, &vmp, &vp);
resolve.l_vmnt_lock = VMNT_WRITE; resolve.l_vmnt_lock = VMNT_READ;
resolve.l_vnode_lock = VNODE_WRITE; resolve.l_vnode_lock = VNODE_WRITE;
if (job_call_nr == CHOWN) { if (job_call_nr == CHOWN) {