writeout=immediate implies the after pwritev we do a sync_file_range.
Reviewed-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
/* FsContext flag values */
#define PATHNAME_FSCONTEXT 0x1
+/* export flags */
+#define V9FS_IMMEDIATE_WRITEOUT 0x1
+
typedef struct FsContext
{
int flags;
char *fs_root;
SecModel fs_sm;
uid_t uid;
+ int export_flags;
struct xattr_operations **xops;
/* fs driver specific data */
void *private;
const char *fstype = qemu_opt_get(opts, "fstype");
const char *path = qemu_opt_get(opts, "path");
const char *sec_model = qemu_opt_get(opts, "security_model");
+ const char *writeout = qemu_opt_get(opts, "writeout");
+
if (!fsdev_id) {
fprintf(stderr, "fsdev: No id specified\n");
fsle->fse.path = g_strdup(path);
fsle->fse.security_model = g_strdup(sec_model);
fsle->fse.ops = FsTypes[i].ops;
-
+ fsle->fse.export_flags = 0;
+ if (writeout) {
+ if (!strcmp(writeout, "immediate")) {
+ fsle->fse.export_flags = V9FS_IMMEDIATE_WRITEOUT;
+ }
+ }
QTAILQ_INSERT_TAIL(&fstype_entries, fsle, next);
return 0;
-
}
FsTypeEntry *get_fsdev_fsentry(char *id)
char *fsdev_id;
char *path;
char *security_model;
+ int export_flags;
FileOperations *ops;
} FsTypeEntry;
exit(1);
}
+ s->ctx.export_flags = fse->export_flags;
s->ctx.fs_root = g_strdup(fse->path);
len = strlen(conf->tag);
if (len > MAX_TAG_LEN) {
static ssize_t handle_pwritev(FsContext *ctx, int fd, const struct iovec *iov,
int iovcnt, off_t offset)
{
+ ssize_t ret;
#ifdef CONFIG_PREADV
- return pwritev(fd, iov, iovcnt, offset);
+ ret = pwritev(fd, iov, iovcnt, offset);
#else
int err = lseek(fd, offset, SEEK_SET);
if (err == -1) {
return err;
} else {
- return writev(fd, iov, iovcnt);
+ ret = writev(fd, iov, iovcnt);
}
#endif
+#ifdef CONFIG_SYNC_FILE_RANGE
+ if (ret > 0 && ctx->export_flags & V9FS_IMMEDIATE_WRITEOUT) {
+ /*
+ * Initiate a writeback. This is not a data integrity sync.
+ * We want to ensure that we don't leave dirty pages in the cache
+ * after write when writeout=immediate is sepcified.
+ */
+ sync_file_range(fd, offset, ret,
+ SYNC_FILE_RANGE_WAIT_BEFORE | SYNC_FILE_RANGE_WRITE);
+ }
+#endif
+ return ret;
}
static int handle_chmod(FsContext *fs_ctx, V9fsPath *fs_path, FsCred *credp)
static ssize_t local_pwritev(FsContext *ctx, int fd, const struct iovec *iov,
int iovcnt, off_t offset)
{
+ ssize_t ret
+;
#ifdef CONFIG_PREADV
- return pwritev(fd, iov, iovcnt, offset);
+ ret = pwritev(fd, iov, iovcnt, offset);
#else
int err = lseek(fd, offset, SEEK_SET);
if (err == -1) {
return err;
} else {
- return writev(fd, iov, iovcnt);
+ ret = writev(fd, iov, iovcnt);
}
#endif
+#ifdef CONFIG_SYNC_FILE_RANGE
+ if (ret > 0 && ctx->export_flags & V9FS_IMMEDIATE_WRITEOUT) {
+ /*
+ * Initiate a writeback. This is not a data integrity sync.
+ * We want to ensure that we don't leave dirty pages in the cache
+ * after write when writeout=immediate is sepcified.
+ */
+ sync_file_range(fd, offset, ret,
+ SYNC_FILE_RANGE_WAIT_BEFORE | SYNC_FILE_RANGE_WRITE);
+ }
+#endif
+ return ret;
}
static int local_chmod(FsContext *fs_ctx, V9fsPath *fs_path, FsCred *credp)
credp->fc_rdev = -1;
}
+static int get_dotl_openflags(V9fsState *s, int oflags)
+{
+ int flags;
+ /*
+ * Filter the client open flags
+ */
+ flags = oflags & ~(O_NOCTTY | O_ASYNC | O_CREAT);
+ /*
+ * Ignore direct disk access hint until the server supports it.
+ */
+ flags &= ~O_DIRECT;
+ return flags;
+}
+
void v9fs_string_init(V9fsString *str)
{
str->data = NULL;
err = offset;
} else {
if (s->proto_version == V9FS_PROTO_2000L) {
- flags = mode;
- flags &= ~(O_NOCTTY | O_ASYNC | O_CREAT);
- /* Ignore direct disk access hint until the server supports it. */
- flags &= ~O_DIRECT;
+ flags = get_dotl_openflags(s, mode);
} else {
flags = omode_to_uflags(mode);
}
goto out_nofid;
}
- /* Ignore direct disk access hint until the server supports it. */
- flags &= ~O_DIRECT;
+ flags = get_dotl_openflags(pdu->s, flags);
err = v9fs_co_open2(pdu, fidp, &name, gid,
flags | O_CREAT, mode, &stbuf);
if (err < 0) {
}, {
.name = "security_model",
.type = QEMU_OPT_STRING,
+ }, {
+ .name = "writeout",
+ .type = QEMU_OPT_STRING,
},
{ /*End of list */ }
},
}, {
.name = "security_model",
.type = QEMU_OPT_STRING,
+ }, {
+ .name = "writeout",
+ .type = QEMU_OPT_STRING,
},
{ /*End of list */ }
DEFHEADING(File system options:)
DEF("fsdev", HAS_ARG, QEMU_OPTION_fsdev,
- "-fsdev local,id=id,path=path,security_model=[mapped|passthrough|none]\n",
+ "-fsdev local,id=id,path=path,security_model=[mapped|passthrough|none]\n"
+ " [,writeout=immediate]\n",
QEMU_ARCH_ALL)
STEXI
Options to each backend are described below.
-@item -fsdev local ,id=@var{id} ,path=@var{path} ,security_model=@var{security_model}
+@item -fsdev local ,id=@var{id} ,path=@var{path} ,security_model=@var{security_model}[,writeout=@var{writeout}]
Create a file-system-"device" for local-filesystem.
@option{security_model} specifies the security model to be followed.
@option{security_model} is required.
+@option{writeout} specifies whether to skip the host page cache.
+@option{writeout} is an optional argument.
+
@end table
ETEXI
DEFHEADING(Virtual File system pass-through options:)
DEF("virtfs", HAS_ARG, QEMU_OPTION_virtfs,
- "-virtfs local,path=path,mount_tag=tag,security_model=[mapped|passthrough|none]\n",
+ "-virtfs local,path=path,mount_tag=tag,security_model=[mapped|passthrough|none]\n"
+ " [,writeout=immediate]\n",
QEMU_ARCH_ALL)
STEXI
Options to each backend are described below.
-@item -virtfs local ,path=@var{path} ,mount_tag=@var{mount_tag} ,security_model=@var{security_model}
+@item -virtfs local ,path=@var{path} ,mount_tag=@var{mount_tag} ,security_model=@var{security_model}[,writeout=@var{writeout}]
Create a Virtual file-system-pass through for local-filesystem.
@option{security_model} specifies the security model to be followed.
@option{security_model} is required.
-
@option{mount_tag} specifies the tag with which the exported file is mounted.
@option{mount_tag} is required.
+@option{writeout} specifies whether to skip the host page cache.
+@option{writeout} is an optional argument.
+
@end table
ETEXI
case QEMU_OPTION_virtfs: {
QemuOpts *fsdev;
QemuOpts *device;
+ const char *writeout;
olist = qemu_find_opts("virtfs");
if (!olist) {
qemu_opt_get(opts, "mount_tag"));
exit(1);
}
+
+ writeout = qemu_opt_get(opts, "writeout");
+ if (writeout) {
+#ifdef CONFIG_SYNC_FILE_RANGE
+ qemu_opt_set(fsdev, "writeout", writeout);
+#else
+ fprintf(stderr, "writeout=immediate not supported on "
+ "this platform\n");
+ exit(1);
+#endif
+ }
qemu_opt_set(fsdev, "fstype", qemu_opt_get(opts, "fstype"));
qemu_opt_set(fsdev, "path", qemu_opt_get(opts, "path"));
qemu_opt_set(fsdev, "security_model",