* THE SOFTWARE.
*/
#include "qemu-common.h"
-#include "qemu-timer.h"
-#include "qemu-char.h"
-#include "qemu-log.h"
-#include "block_int.h"
-#include "module.h"
+#include "qemu/timer.h"
+#include "qemu/log.h"
+#include "block/block_int.h"
+#include "qemu/module.h"
#include "trace.h"
-#include "thread-pool.h"
-#include "iov.h"
+#include "block/thread-pool.h"
+#include "qemu/iov.h"
#include "raw-aio.h"
#if defined(__APPLE__) && (__MACH__)
#ifdef CONFIG_FIEMAP
#include <linux/fiemap.h>
#endif
+#ifdef CONFIG_FALLOCATE_PUNCH_HOLE
+#include <linux/falloc.h>
+#endif
#if defined (__FreeBSD__) || defined(__FreeBSD_kernel__)
#include <sys/disk.h>
#include <sys/cdio.h>
#ifdef CONFIG_XFS
bool is_xfs : 1;
#endif
+ bool has_discard : 1;
} BDRVRawState;
typedef struct BDRVRawReopenState {
void *aio_ioctl_buf;
};
int aio_niov;
- size_t aio_nbytes;
+ uint64_t aio_nbytes;
#define aio_ioctl_cmd aio_nbytes /* for QEMU_AIO_IOCTL */
off_t aio_offset;
int aio_type;
}
#endif
-static int raw_open_common(BlockDriverState *bs, const char *filename,
+static QemuOptsList raw_runtime_opts = {
+ .name = "raw",
+ .head = QTAILQ_HEAD_INITIALIZER(raw_runtime_opts.head),
+ .desc = {
+ {
+ .name = "filename",
+ .type = QEMU_OPT_STRING,
+ .help = "File name of the image",
+ },
+ { /* end of list */ }
+ },
+};
+
+static int raw_open_common(BlockDriverState *bs, QDict *options,
int bdrv_flags, int open_flags)
{
BDRVRawState *s = bs->opaque;
+ QemuOpts *opts;
+ Error *local_err = NULL;
+ const char *filename;
int fd, ret;
+ opts = qemu_opts_create_nofail(&raw_runtime_opts);
+ qemu_opts_absorb_qdict(opts, options, &local_err);
+ if (error_is_set(&local_err)) {
+ qerror_report_err(local_err);
+ error_free(local_err);
+ ret = -EINVAL;
+ goto fail;
+ }
+
+ filename = qemu_opt_get(opts, "filename");
+
ret = raw_normalize_devicepath(&filename);
if (ret != 0) {
- return ret;
+ goto fail;
}
s->open_flags = open_flags;
fd = qemu_open(filename, s->open_flags, 0644);
if (fd < 0) {
ret = -errno;
- if (ret == -EROFS)
+ if (ret == -EROFS) {
ret = -EACCES;
- return ret;
+ }
+ goto fail;
}
s->fd = fd;
#ifdef CONFIG_LINUX_AIO
if (raw_set_aio(&s->aio_ctx, &s->use_aio, bdrv_flags)) {
qemu_close(fd);
- return -errno;
+ ret = -errno;
+ goto fail;
}
#endif
+ s->has_discard = 1;
#ifdef CONFIG_XFS
if (platform_test_xfs_fd(s->fd)) {
s->is_xfs = 1;
}
#endif
- return 0;
+ ret = 0;
+fail:
+ qemu_opts_del(opts);
+ return ret;
}
-static int raw_open(BlockDriverState *bs, const char *filename, int flags)
+static int raw_open(BlockDriverState *bs, QDict *options, int flags,
+ Error **errp)
{
BDRVRawState *s = bs->opaque;
s->type = FTYPE_FILE;
- return raw_open_common(bs, filename, flags, 0);
+ return raw_open_common(bs, options, flags, 0);
}
static int raw_reopen_prepare(BDRVReopenState *state,
}
#endif
+ if (s->type == FTYPE_FD || s->type == FTYPE_CD) {
+ raw_s->open_flags |= O_NONBLOCK;
+ }
+
raw_parse_flags(state->flags, &raw_s->open_flags);
raw_s->fd = -1;
- int fcntl_flags = O_APPEND | O_ASYNC | O_NONBLOCK;
+ int fcntl_flags = O_APPEND | O_NONBLOCK;
#ifdef O_NOATIME
fcntl_flags |= O_NOATIME;
#endif
+#ifdef O_ASYNC
+ /* Not all operating systems have O_ASYNC, and those that don't
+ * will not let us track the state into raw_s->open_flags (typically
+ * you achieve the same effect with an ioctl, for example I_SETSIG
+ * on Solaris). But we do not use O_ASYNC, so that's fine.
+ */
+ assert((s->open_flags & O_ASYNC) == 0);
+#endif
+
if ((raw_s->open_flags & ~fcntl_flags) == (s->open_flags & ~fcntl_flags)) {
/* dup the original fd */
/* TODO: use qemu fcntl wrapper */
#endif
*/
-/*
- * Check if all memory in this vector is sector aligned.
- */
-static int qiov_is_aligned(BlockDriverState *bs, QEMUIOVector *qiov)
-{
- int i;
-
- for (i = 0; i < qiov->niov; i++) {
- if ((uintptr_t) qiov->iov[i].iov_base % bs->buffer_alignment) {
- return 0;
- }
- }
-
- return 1;
-}
-
static ssize_t handle_aiocb_ioctl(RawPosixAIOData *aiocb)
{
int ret;
return -errno;
}
- /*
- * This looks weird, but the aio code only considers a request
- * successful if it has written the full number of bytes.
- *
- * Now we overload aio_nbytes as aio_ioctl_cmd for the ioctl command,
- * so in fact we return the ioctl command here to make posix_aio_read()
- * happy..
- */
- return aiocb->aio_nbytes;
+ return 0;
}
static ssize_t handle_aiocb_flush(RawPosixAIOData *aiocb)
return nbytes;
}
+#ifdef CONFIG_XFS
+static int xfs_discard(BDRVRawState *s, int64_t offset, uint64_t bytes)
+{
+ struct xfs_flock64 fl;
+
+ memset(&fl, 0, sizeof(fl));
+ fl.l_whence = SEEK_SET;
+ fl.l_start = offset;
+ fl.l_len = bytes;
+
+ if (xfsctl(NULL, s->fd, XFS_IOC_UNRESVSP64, &fl) < 0) {
+ DEBUG_BLOCK_PRINT("cannot punch hole (%s)\n", strerror(errno));
+ return -errno;
+ }
+
+ return 0;
+}
+#endif
+
+static ssize_t handle_aiocb_discard(RawPosixAIOData *aiocb)
+{
+ int ret = -EOPNOTSUPP;
+ BDRVRawState *s = aiocb->bs->opaque;
+
+ if (s->has_discard == 0) {
+ return 0;
+ }
+
+ if (aiocb->aio_type & QEMU_AIO_BLKDEV) {
+#ifdef BLKDISCARD
+ do {
+ uint64_t range[2] = { aiocb->aio_offset, aiocb->aio_nbytes };
+ if (ioctl(aiocb->aio_fildes, BLKDISCARD, range) == 0) {
+ return 0;
+ }
+ } while (errno == EINTR);
+
+ ret = -errno;
+#endif
+ } else {
+#ifdef CONFIG_XFS
+ if (s->is_xfs) {
+ return xfs_discard(s, aiocb->aio_offset, aiocb->aio_nbytes);
+ }
+#endif
+
+#ifdef CONFIG_FALLOCATE_PUNCH_HOLE
+ do {
+ if (fallocate(s->fd, FALLOC_FL_PUNCH_HOLE | FALLOC_FL_KEEP_SIZE,
+ aiocb->aio_offset, aiocb->aio_nbytes) == 0) {
+ return 0;
+ }
+ } while (errno == EINTR);
+
+ ret = -errno;
+#endif
+ }
+
+ if (ret == -ENODEV || ret == -ENOSYS || ret == -EOPNOTSUPP ||
+ ret == -ENOTTY) {
+ s->has_discard = 0;
+ ret = 0;
+ }
+ return ret;
+}
+
static int aio_worker(void *arg)
{
RawPosixAIOData *aiocb = arg;
case QEMU_AIO_IOCTL:
ret = handle_aiocb_ioctl(aiocb);
break;
+ case QEMU_AIO_DISCARD:
+ ret = handle_aiocb_discard(aiocb);
+ break;
default:
fprintf(stderr, "invalid aio request (0x%x)\n", aiocb->aio_type);
ret = -EINVAL;
BlockDriverCompletionFunc *cb, void *opaque, int type)
{
RawPosixAIOData *acb = g_slice_new(RawPosixAIOData);
+ ThreadPool *pool;
acb->bs = bs;
acb->aio_type = type;
acb->aio_offset = sector_num * 512;
trace_paio_submit(acb, opaque, sector_num, nb_sectors, type);
- return thread_pool_submit_aio(aio_worker, acb, cb, opaque);
-}
-
-static BlockDriverAIOCB *paio_ioctl(BlockDriverState *bs, int fd,
- unsigned long int req, void *buf,
- BlockDriverCompletionFunc *cb, void *opaque)
-{
- RawPosixAIOData *acb = g_slice_new(RawPosixAIOData);
-
- acb->bs = bs;
- acb->aio_type = QEMU_AIO_IOCTL;
- acb->aio_fildes = fd;
- acb->aio_offset = 0;
- acb->aio_ioctl_buf = buf;
- acb->aio_ioctl_cmd = req;
-
- return thread_pool_submit_aio(aio_worker, acb, cb, opaque);
+ pool = aio_get_thread_pool(bdrv_get_aio_context(bs));
+ return thread_pool_submit_aio(pool, aio_worker, acb, cb, opaque);
}
static BlockDriverAIOCB *raw_aio_submit(BlockDriverState *bs,
* driver that it needs to copy the buffer.
*/
if ((bs->open_flags & BDRV_O_NOCACHE)) {
- if (!qiov_is_aligned(bs, qiov)) {
+ if (!bdrv_qiov_is_aligned(bs, qiov)) {
type |= QEMU_AIO_MISALIGNED;
#ifdef CONFIG_LINUX_AIO
} else if (s->use_aio) {
return (int64_t)st.st_blocks * 512;
}
-static int raw_create(const char *filename, QEMUOptionParameter *options)
+static int raw_create(const char *filename, QEMUOptionParameter *options,
+ Error **errp)
{
int fd;
int result = 0;
* 'nb_sectors' is the max value 'pnum' should be set to. If nb_sectors goes
* beyond the end of the disk image it will be clamped.
*/
-static int coroutine_fn raw_co_is_allocated(BlockDriverState *bs,
+static int64_t coroutine_fn raw_co_get_block_status(BlockDriverState *bs,
int64_t sector_num,
int nb_sectors, int *pnum)
{
off_t start, data, hole;
- int ret;
+ int64_t ret;
ret = fd_open(bs);
if (ret < 0) {
}
start = sector_num * BDRV_SECTOR_SIZE;
+ ret = BDRV_BLOCK_DATA | BDRV_BLOCK_OFFSET_VALID | start;
#ifdef CONFIG_FIEMAP
if (ioctl(s->fd, FS_IOC_FIEMAP, &f) == -1) {
/* Assume everything is allocated. */
*pnum = nb_sectors;
- return 1;
+ return ret;
}
if (f.fm.fm_mapped_extents == 0) {
} else {
data = f.fe.fe_logical;
hole = f.fe.fe_logical + f.fe.fe_length;
+ if (f.fe.fe_flags & FIEMAP_EXTENT_UNWRITTEN) {
+ ret |= BDRV_BLOCK_ZERO;
+ }
}
#elif defined SEEK_HOLE && defined SEEK_DATA
/* Most likely EINVAL. Assume everything is allocated. */
*pnum = nb_sectors;
- return 1;
+ return ret;
}
if (hole > start) {
}
}
#else
- *pnum = nb_sectors;
- return 1;
+ data = 0;
+ hole = start + nb_sectors * BDRV_SECTOR_SIZE;
#endif
if (data <= start) {
/* On a data extent, compute sectors to the end of the extent. */
*pnum = MIN(nb_sectors, (hole - start) / BDRV_SECTOR_SIZE);
- return 1;
} else {
/* On a hole, compute sectors to the beginning of the next extent. */
*pnum = MIN(nb_sectors, (data - start) / BDRV_SECTOR_SIZE);
- return 0;
+ ret &= ~BDRV_BLOCK_DATA;
+ ret |= BDRV_BLOCK_ZERO;
}
-}
-#ifdef CONFIG_XFS
-static int xfs_discard(BDRVRawState *s, int64_t sector_num, int nb_sectors)
-{
- struct xfs_flock64 fl;
-
- memset(&fl, 0, sizeof(fl));
- fl.l_whence = SEEK_SET;
- fl.l_start = sector_num << 9;
- fl.l_len = (int64_t)nb_sectors << 9;
-
- if (xfsctl(NULL, s->fd, XFS_IOC_UNRESVSP64, &fl) < 0) {
- DEBUG_BLOCK_PRINT("cannot punch hole (%s)\n", strerror(errno));
- return -errno;
- }
-
- return 0;
+ return ret;
}
-#endif
-static coroutine_fn int raw_co_discard(BlockDriverState *bs,
- int64_t sector_num, int nb_sectors)
+static coroutine_fn BlockDriverAIOCB *raw_aio_discard(BlockDriverState *bs,
+ int64_t sector_num, int nb_sectors,
+ BlockDriverCompletionFunc *cb, void *opaque)
{
-#ifdef CONFIG_XFS
BDRVRawState *s = bs->opaque;
- if (s->is_xfs) {
- return xfs_discard(s, sector_num, nb_sectors);
- }
-#endif
-
- return 0;
+ return paio_submit(bs, s->fd, sector_num, NULL, nb_sectors,
+ cb, opaque, QEMU_AIO_DISCARD);
}
static QEMUOptionParameter raw_create_options[] = {
.format_name = "file",
.protocol_name = "file",
.instance_size = sizeof(BDRVRawState),
+ .bdrv_needs_filename = true,
.bdrv_probe = NULL, /* no probe for protocols */
.bdrv_file_open = raw_open,
.bdrv_reopen_prepare = raw_reopen_prepare,
.bdrv_reopen_abort = raw_reopen_abort,
.bdrv_close = raw_close,
.bdrv_create = raw_create,
- .bdrv_co_discard = raw_co_discard,
- .bdrv_co_is_allocated = raw_co_is_allocated,
+ .bdrv_has_zero_init = bdrv_has_zero_init_1,
+ .bdrv_co_get_block_status = raw_co_get_block_status,
.bdrv_aio_readv = raw_aio_readv,
.bdrv_aio_writev = raw_aio_writev,
.bdrv_aio_flush = raw_aio_flush,
+ .bdrv_aio_discard = raw_aio_discard,
.bdrv_truncate = raw_truncate,
.bdrv_getlength = raw_getlength,
return 0;
}
-static int hdev_open(BlockDriverState *bs, const char *filename, int flags)
+static int check_hdev_writable(BDRVRawState *s)
+{
+#if defined(BLKROGET)
+ /* Linux block devices can be configured "read-only" using blockdev(8).
+ * This is independent of device node permissions and therefore open(2)
+ * with O_RDWR succeeds. Actual writes fail with EPERM.
+ *
+ * bdrv_open() is supposed to fail if the disk is read-only. Explicitly
+ * check for read-only block devices so that Linux block devices behave
+ * properly.
+ */
+ struct stat st;
+ int readonly = 0;
+
+ if (fstat(s->fd, &st)) {
+ return -errno;
+ }
+
+ if (!S_ISBLK(st.st_mode)) {
+ return 0;
+ }
+
+ if (ioctl(s->fd, BLKROGET, &readonly) < 0) {
+ return -errno;
+ }
+
+ if (readonly) {
+ return -EACCES;
+ }
+#endif /* defined(BLKROGET) */
+ return 0;
+}
+
+static int hdev_open(BlockDriverState *bs, QDict *options, int flags,
+ Error **errp)
{
BDRVRawState *s = bs->opaque;
+ int ret;
+ const char *filename = qdict_get_str(options, "filename");
#if defined(__APPLE__) && defined(__MACH__)
if (strstart(filename, "/dev/cdrom", NULL)) {
qemu_close(fd);
}
filename = bsdPath;
+ qdict_put(options, "filename", qstring_from_str(filename));
}
if ( mediaIterator )
}
#endif
- return raw_open_common(bs, filename, flags, 0);
+ ret = raw_open_common(bs, options, flags, 0);
+ if (ret < 0) {
+ return ret;
+ }
+
+ if (flags & BDRV_O_RDWR) {
+ ret = check_hdev_writable(s);
+ if (ret < 0) {
+ raw_close(bs);
+ return ret;
+ }
+ }
+
+ return ret;
}
#if defined(__linux__)
BlockDriverCompletionFunc *cb, void *opaque)
{
BDRVRawState *s = bs->opaque;
+ RawPosixAIOData *acb;
+ ThreadPool *pool;
if (fd_open(bs) < 0)
return NULL;
- return paio_ioctl(bs, s->fd, req, buf, cb, opaque);
+
+ acb = g_slice_new(RawPosixAIOData);
+ acb->bs = bs;
+ acb->aio_type = QEMU_AIO_IOCTL;
+ acb->aio_fildes = s->fd;
+ acb->aio_offset = 0;
+ acb->aio_ioctl_buf = buf;
+ acb->aio_ioctl_cmd = req;
+ pool = aio_get_thread_pool(bdrv_get_aio_context(bs));
+ return thread_pool_submit_aio(pool, aio_worker, acb, cb, opaque);
}
#elif defined(__FreeBSD__) || defined(__FreeBSD_kernel__)
#endif /* !linux && !FreeBSD */
-static int hdev_create(const char *filename, QEMUOptionParameter *options)
+static coroutine_fn BlockDriverAIOCB *hdev_aio_discard(BlockDriverState *bs,
+ int64_t sector_num, int nb_sectors,
+ BlockDriverCompletionFunc *cb, void *opaque)
+{
+ BDRVRawState *s = bs->opaque;
+
+ if (fd_open(bs) < 0) {
+ return NULL;
+ }
+ return paio_submit(bs, s->fd, sector_num, NULL, nb_sectors,
+ cb, opaque, QEMU_AIO_DISCARD|QEMU_AIO_BLKDEV);
+}
+
+static int hdev_create(const char *filename, QEMUOptionParameter *options,
+ Error **errp)
{
int fd;
int ret = 0;
return ret;
}
-static int hdev_has_zero_init(BlockDriverState *bs)
-{
- return 0;
-}
-
static BlockDriver bdrv_host_device = {
.format_name = "host_device",
.protocol_name = "host_device",
.instance_size = sizeof(BDRVRawState),
+ .bdrv_needs_filename = true,
.bdrv_probe_device = hdev_probe_device,
.bdrv_file_open = hdev_open,
.bdrv_close = raw_close,
+ .bdrv_reopen_prepare = raw_reopen_prepare,
+ .bdrv_reopen_commit = raw_reopen_commit,
+ .bdrv_reopen_abort = raw_reopen_abort,
.bdrv_create = hdev_create,
.create_options = raw_create_options,
- .bdrv_has_zero_init = hdev_has_zero_init,
.bdrv_aio_readv = raw_aio_readv,
.bdrv_aio_writev = raw_aio_writev,
.bdrv_aio_flush = raw_aio_flush,
+ .bdrv_aio_discard = hdev_aio_discard,
.bdrv_truncate = raw_truncate,
.bdrv_getlength = raw_getlength,
};
#ifdef __linux__
-static int floppy_open(BlockDriverState *bs, const char *filename, int flags)
+static int floppy_open(BlockDriverState *bs, QDict *options, int flags,
+ Error **errp)
{
BDRVRawState *s = bs->opaque;
int ret;
s->type = FTYPE_FD;
/* open will not fail even if no floppy is inserted, so add O_NONBLOCK */
- ret = raw_open_common(bs, filename, flags, O_NONBLOCK);
+ ret = raw_open_common(bs, options, flags, O_NONBLOCK);
if (ret)
return ret;
.format_name = "host_floppy",
.protocol_name = "host_floppy",
.instance_size = sizeof(BDRVRawState),
+ .bdrv_needs_filename = true,
.bdrv_probe_device = floppy_probe_device,
.bdrv_file_open = floppy_open,
.bdrv_close = raw_close,
+ .bdrv_reopen_prepare = raw_reopen_prepare,
+ .bdrv_reopen_commit = raw_reopen_commit,
+ .bdrv_reopen_abort = raw_reopen_abort,
.bdrv_create = hdev_create,
.create_options = raw_create_options,
- .bdrv_has_zero_init = hdev_has_zero_init,
.bdrv_aio_readv = raw_aio_readv,
.bdrv_aio_writev = raw_aio_writev,
.bdrv_eject = floppy_eject,
};
-static int cdrom_open(BlockDriverState *bs, const char *filename, int flags)
+static int cdrom_open(BlockDriverState *bs, QDict *options, int flags,
+ Error **errp)
{
BDRVRawState *s = bs->opaque;
s->type = FTYPE_CD;
/* open will not fail even if no CD is inserted, so add O_NONBLOCK */
- return raw_open_common(bs, filename, flags, O_NONBLOCK);
+ return raw_open_common(bs, options, flags, O_NONBLOCK);
}
static int cdrom_probe_device(const char *filename)
.format_name = "host_cdrom",
.protocol_name = "host_cdrom",
.instance_size = sizeof(BDRVRawState),
+ .bdrv_needs_filename = true,
.bdrv_probe_device = cdrom_probe_device,
.bdrv_file_open = cdrom_open,
.bdrv_close = raw_close,
+ .bdrv_reopen_prepare = raw_reopen_prepare,
+ .bdrv_reopen_commit = raw_reopen_commit,
+ .bdrv_reopen_abort = raw_reopen_abort,
.bdrv_create = hdev_create,
.create_options = raw_create_options,
- .bdrv_has_zero_init = hdev_has_zero_init,
.bdrv_aio_readv = raw_aio_readv,
.bdrv_aio_writev = raw_aio_writev,
#endif /* __linux__ */
#if defined (__FreeBSD__) || defined(__FreeBSD_kernel__)
-static int cdrom_open(BlockDriverState *bs, const char *filename, int flags)
+static int cdrom_open(BlockDriverState *bs, QDict *options, int flags)
{
BDRVRawState *s = bs->opaque;
int ret;
s->type = FTYPE_CD;
- ret = raw_open_common(bs, filename, flags, 0);
+ ret = raw_open_common(bs, options, flags, 0);
if (ret)
return ret;
.format_name = "host_cdrom",
.protocol_name = "host_cdrom",
.instance_size = sizeof(BDRVRawState),
+ .bdrv_needs_filename = true,
.bdrv_probe_device = cdrom_probe_device,
.bdrv_file_open = cdrom_open,
.bdrv_close = raw_close,
+ .bdrv_reopen_prepare = raw_reopen_prepare,
+ .bdrv_reopen_commit = raw_reopen_commit,
+ .bdrv_reopen_abort = raw_reopen_abort,
.bdrv_create = hdev_create,
.create_options = raw_create_options,
- .bdrv_has_zero_init = hdev_has_zero_init,
.bdrv_aio_readv = raw_aio_readv,
.bdrv_aio_writev = raw_aio_writev,
};
#endif /* __FreeBSD__ */
+#ifdef CONFIG_LINUX_AIO
+/**
+ * Return the file descriptor for Linux AIO
+ *
+ * This function is a layering violation and should be removed when it becomes
+ * possible to call the block layer outside the global mutex. It allows the
+ * caller to hijack the file descriptor so I/O can be performed outside the
+ * block layer.
+ */
+int raw_get_aio_fd(BlockDriverState *bs)
+{
+ BDRVRawState *s;
+
+ if (!bs->drv) {
+ return -ENOMEDIUM;
+ }
+
+ if (bs->drv == bdrv_find_format("raw")) {
+ bs = bs->file;
+ }
+
+ /* raw-posix has several protocols so just check for raw_aio_readv */
+ if (bs->drv->bdrv_aio_readv != raw_aio_readv) {
+ return -ENOTSUP;
+ }
+
+ s = bs->opaque;
+ if (!s->use_aio) {
+ return -ENOTSUP;
+ }
+ return s->fd;
+}
+#endif /* CONFIG_LINUX_AIO */
+
static void bdrv_file_init(void)
{
/*
projects / qemu.git / blobdiff