[Qemu-devel] [PATCHv2] block: add native support for NFS

[Peter Lieven]

This patch adds native support for accessing images on NFS shares without
the requirement to actually mount the entire NFS share on the host.

NFS Images can simply be specified by an url of the form:
nfs://<host>/<export>/<filename>

For example:
qemu-img create -f qcow2 nfs://10.0.0.1/qemu-images/test.qcow2

You need libnfs from Ronnie Sahlberg available at:
   git://github.com/sahlberg/libnfs.git
for this to work.

During configure it is automatically probed for libnfs and support
is enabled on-the-fly. You can forbid or enforce libnfs support
with --disable-libnfs or --enable-libnfs respectively.

Due to NFS restrictions you might need to execute your binaries
as root, allow them to open priviledged ports (<1024) or specify
insecure option on the NFS server.

Signed-off-by: Peter Lieven <address@hidden>
---
v1->v2:
 - fixed block/Makefile.objs [Ronnie]
 - do not always register a read handler [Ronnie]
 - add support for reading beyond EOF [Fam]
 - fixed struct and paramter naming [Fam]
 - fixed overlong lines and whitespace errors [Fam]
 - return return status from libnfs whereever possible [Fam]
 - added comment why we set allocated_file_size to -ENOTSUP after write [Fam]
 - avoid segfault when parsing filname [Fam]
 - remove unused close_bh from NFSClient [Fam]
 - avoid dividing and mutliplying total_size by BDRV_SECTOR_SIZE in 
nfs_file_create [Fam]

 MAINTAINERS         |    5 +
 block/Makefile.objs |    1 +
 block/nfs.c         |  419 +++++++++++++++++++++++++++++++++++++++++++++++++++
 configure           |   38 +++++
 4 files changed, 463 insertions(+)
 create mode 100644 block/nfs.c

diff --git a/MAINTAINERS b/MAINTAINERS
index c19133f..f53d184 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -899,6 +899,11 @@ M: Peter Lieven <address@hidden>
 S: Supported
 F: block/iscsi.c
 
+NFS
+M: Peter Lieven <address@hidden>
+S: Maintained
+F: block/nfs.c
+
 SSH
 M: Richard W.M. Jones <address@hidden>
 S: Supported
diff --git a/block/Makefile.objs b/block/Makefile.objs
index f43ecbc..aa8eaf9 100644
--- a/block/Makefile.objs
+++ b/block/Makefile.objs
@@ -12,6 +12,7 @@ block-obj-$(CONFIG_LINUX_AIO) += linux-aio.o
 ifeq ($(CONFIG_POSIX),y)
 block-obj-y += nbd.o sheepdog.o
 block-obj-$(CONFIG_LIBISCSI) += iscsi.o
+block-obj-$(CONFIG_LIBNFS) += nfs.o
 block-obj-$(CONFIG_CURL) += curl.o
 block-obj-$(CONFIG_RBD) += rbd.o
 block-obj-$(CONFIG_GLUSTERFS) += gluster.o
diff --git a/block/nfs.c b/block/nfs.c
new file mode 100644
index 0000000..006b8cc
--- /dev/null
+++ b/block/nfs.c
@@ -0,0 +1,419 @@
+/*
+ * QEMU Block driver for native access to files on NFS shares
+ *
+ * Copyright (c) 2013 Peter Lieven <address@hidden>
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to 
deal
+ * in the Software without restriction, including without limitation the rights
+ * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+ * copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING 
FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+ * THE SOFTWARE.
+ */
+
+#include "config-host.h"
+
+#include <poll.h>
+#include "qemu-common.h"
+#include "qemu/config-file.h"
+#include "qemu/error-report.h"
+#include "block/block_int.h"
+#include "trace.h"
+#include "qemu/iov.h"
+#include "sysemu/sysemu.h"
+
+#include <nfsc/libnfs-zdr.h>
+#include <nfsc/libnfs.h>
+#include <nfsc/libnfs-raw.h>
+#include <nfsc/libnfs-raw-mount.h>
+
+typedef struct nfsclient {
+    struct nfs_context *context;
+    struct nfsfh *fh;
+    int events;
+    bool has_zero_init;
+    int64_t allocated_file_size;
+} NFSClient;
+
+typedef struct NFSTask {
+    int status;
+    int complete;
+    QEMUIOVector *iov;
+    Coroutine *co;
+    QEMUBH *bh;
+} NFSTask;
+
+static void nfs_process_read(void *arg);
+static void nfs_process_write(void *arg);
+
+static void nfs_set_events(NFSClient *client)
+{
+    int ev = nfs_which_events(client->context);
+    if (ev != client->events) {
+        qemu_aio_set_fd_handler(nfs_get_fd(client->context),
+                      (ev & POLLIN) ? nfs_process_read : NULL,
+                      (ev & POLLOUT) ? nfs_process_write : NULL,
+                      client);
+
+    }
+    client->events = ev;
+}
+
+static void nfs_process_read(void *arg)
+{
+    NFSClient *client = arg;
+    nfs_service(client->context, POLLIN);
+    nfs_set_events(client);
+}
+
+static void nfs_process_write(void *arg)
+{
+    NFSClient *client = arg;
+    nfs_service(client->context, POLLOUT);
+    nfs_set_events(client);
+}
+
+static void nfs_co_init_task(NFSClient *client, NFSTask *Task)
+{
+    *Task = (NFSTask) {
+        .co         = qemu_coroutine_self(),
+    };
+}
+
+static void nfs_co_generic_bh_cb(void *opaque)
+{
+    NFSTask *Task = opaque;
+    qemu_bh_delete(Task->bh);
+    qemu_coroutine_enter(Task->co, NULL);
+}
+
+static void
+nfs_co_generic_cb(int status, struct nfs_context *nfs, void *data,
+                  void *private_data)
+{
+    NFSTask *Task = private_data;
+    Task->complete = 1;
+    Task->status = status;
+    if (Task->status > 0 && Task->iov) {
+        if (Task->status <= Task->iov->size) {
+            qemu_iovec_from_buf(Task->iov, 0, data, Task->status);
+        } else {
+            Task->status = -EIO;
+        }
+    }
+    if (Task->co) {
+        Task->bh = qemu_bh_new(nfs_co_generic_bh_cb, Task);
+        qemu_bh_schedule(Task->bh);
+    }
+}
+
+static int coroutine_fn nfs_co_readv(BlockDriverState *bs,
+                                     int64_t sector_num, int nb_sectors,
+                                     QEMUIOVector *iov)
+{
+    NFSClient *client = bs->opaque;
+    NFSTask task;
+
+    nfs_co_init_task(client, &task);
+    task.iov = iov;
+
+    if (nfs_pread_async(client->context, client->fh,
+                        sector_num * BDRV_SECTOR_SIZE,
+                        nb_sectors * BDRV_SECTOR_SIZE,
+                        nfs_co_generic_cb, &task) != 0) {
+        return -EIO;
+    }
+
+    while (!task.complete) {
+        nfs_set_events(client);
+        qemu_coroutine_yield();
+    }
+
+    if (task.status < 0) {
+        return task.status;
+    }
+
+    return 0;
+}
+
+static int coroutine_fn nfs_co_writev(BlockDriverState *bs,
+                                        int64_t sector_num, int nb_sectors,
+                                        QEMUIOVector *iov)
+{
+    NFSClient *client = bs->opaque;
+    NFSTask task;
+    char *buf = NULL;
+
+    nfs_co_init_task(client, &task);
+
+    buf = g_malloc(nb_sectors * BDRV_SECTOR_SIZE);
+    qemu_iovec_to_buf(iov, 0, buf, nb_sectors * BDRV_SECTOR_SIZE);
+
+    if (nfs_pwrite_async(client->context, client->fh,
+                         sector_num * BDRV_SECTOR_SIZE,
+                         nb_sectors * BDRV_SECTOR_SIZE,
+                         buf, nfs_co_generic_cb, &task) != 0) {
+        g_free(buf);
+        return -EIO;
+    }
+
+    while (!task.complete) {
+        nfs_set_events(client);
+        qemu_coroutine_yield();
+    }
+
+    g_free(buf);
+
+    if (task.status != nb_sectors * BDRV_SECTOR_SIZE) {
+        return task.status < 0 ? task.status : -EIO;
+    }
+
+    bs->total_sectors = MAX(bs->total_sectors, sector_num + nb_sectors);
+    /* set to -ENOTSUP since bdrv_allocated_file_size is only used
+     * in qemu-img open. So we can use the cached value for allocate
+     * filesize obtained from fstat at open time */
+    client->allocated_file_size = -ENOTSUP;
+    return 0;
+}
+
+static int coroutine_fn nfs_co_flush(BlockDriverState *bs)
+{
+    NFSClient *client = bs->opaque;
+    NFSTask task;
+
+    nfs_co_init_task(client, &task);
+
+    if (nfs_fsync_async(client->context, client->fh, nfs_co_generic_cb,
+                        &task) != 0) {
+        return -EIO;
+    }
+
+    while (!task.complete) {
+        nfs_set_events(client);
+        qemu_coroutine_yield();
+    }
+
+    return task.status;
+}
+
+static QemuOptsList runtime_opts = {
+    .name = "nfs",
+    .head = QTAILQ_HEAD_INITIALIZER(runtime_opts.head),
+    .desc = {
+        {
+            .name = "filename",
+            .type = QEMU_OPT_STRING,
+            .help = "URL to the NFS file",
+        },
+        { /* end of list */ }
+    },
+};
+
+static void nfs_file_close(BlockDriverState *bs)
+{
+    NFSClient *client = bs->opaque;
+    if (client->context) {
+        if (client->fh) {
+            nfs_close(client->context, client->fh);
+        }
+        qemu_aio_set_fd_handler(nfs_get_fd(client->context), NULL, NULL, NULL);
+        nfs_destroy_context(client->context);
+    }
+    memset(client, 0, sizeof(NFSClient));
+}
+
+
+static int nfs_file_open_common(BlockDriverState *bs, QDict *options, int 
flags,
+                         int open_flags, Error **errp)
+{
+    NFSClient *client = bs->opaque;
+    const char *filename;
+    int ret = 0;
+    QemuOpts *opts;
+    Error *local_err = NULL;
+    char *server = NULL, *path = NULL, *file = NULL, *strp;
+    struct stat st;
+
+    opts = qemu_opts_create_nofail(&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");
+
+    client->context = nfs_init_context();
+
+    if (client->context == NULL) {
+        error_setg(errp, "Failed to init NFS context");
+        ret = -EINVAL;
+        goto fail;
+    }
+
+    if (strlen(filename) <= 6) {
+        error_setg(errp, "Invalid server in URL");
+        ret = -EINVAL;
+        goto fail;
+    }
+
+    server = g_strdup(filename + 6);
+    if (server[0] == '/' || server[0] == '\0') {
+        error_setg(errp, "Invalid server in URL");
+        ret = -EINVAL;
+        goto fail;
+    }
+    strp = strchr(server, '/');
+    if (strp == NULL) {
+        error_setg(errp, "Invalid URL specified.\n");
+        ret = -EINVAL;
+        goto fail;
+    }
+    path = g_strdup(strp);
+    *strp = 0;
+    strp = strrchr(path, '/');
+    if (strp == NULL) {
+        error_setg(errp, "Invalid URL specified.\n");
+        ret = -EINVAL;
+        goto fail;
+    }
+    file = g_strdup(strp);
+    *strp = 0;
+
+    ret = nfs_mount(client->context, server, path);
+    if (ret < 0) {
+        error_setg(errp, "Failed to mount nfs share: %s",
+                    nfs_get_error(client->context));
+        goto fail;
+    }
+
+    if (open_flags & O_CREAT) {
+        ret = nfs_creat(client->context, file, 0600, &client->fh);
+        if (ret < 0) {
+            error_setg(errp, "Failed to create file: %s",
+                             nfs_get_error(client->context));
+            goto fail;
+        }
+    } else {
+        open_flags = (flags & BDRV_O_RDWR) ? O_RDWR : O_RDONLY;
+        ret = nfs_open(client->context, file, open_flags, &client->fh);
+        if (ret < 0) {
+            error_setg(errp, "Failed to open file : %s",
+                       nfs_get_error(client->context));
+            goto fail;
+        }
+    }
+
+    ret = nfs_fstat(client->context, client->fh, &st);
+    if (ret < 0) {
+        error_setg(errp, "Failed to fstat file: %s",
+                   nfs_get_error(client->context));
+        goto fail;
+    }
+
+    /* TODO allow reading beyond EOF */
+    bs->total_sectors = DIV_ROUND_UP(st.st_size, BDRV_SECTOR_SIZE);
+    client->has_zero_init = S_ISREG(st.st_mode);
+    client->allocated_file_size = st.st_blocks * st.st_blksize;
+    goto out;
+fail:
+    nfs_file_close(bs);
+out:
+    g_free(server);
+    g_free(path);
+    g_free(file);
+    return ret;
+}
+
+static int nfs_file_open(BlockDriverState *bs, QDict *options, int flags,
+                         Error **errp) {
+    return nfs_file_open_common(bs, options, flags, 0, errp);
+}
+
+static int nfs_file_create(const char *filename, QEMUOptionParameter *options,
+                         Error **errp)
+{
+    int ret = 0;
+    int64_t total_size = 0;
+    BlockDriverState *bs;
+    NFSClient *client = NULL;
+    QDict *bs_options;
+
+    bs = bdrv_new("");
+
+    /* Read out options */
+    while (options && options->name) {
+        if (!strcmp(options->name, "size")) {
+            total_size = options->value.n;
+        }
+        options++;
+    }
+
+    bs->opaque = g_malloc0(sizeof(NFSClient));
+    client = bs->opaque;
+
+    bs_options = qdict_new();
+    qdict_put(bs_options, "filename", qstring_from_str(filename));
+    ret = nfs_file_open_common(bs, bs_options, 0, O_CREAT, NULL);
+    QDECREF(bs_options);
+    if (ret < 0) {
+        goto out;
+    }
+    ret = nfs_ftruncate(client->context, client->fh, total_size);
+out:
+    nfs_file_close(bs);
+    g_free(bs->opaque);
+    bs->opaque = NULL;
+    bdrv_unref(bs);
+    return ret;
+}
+
+static int nfs_has_zero_init(BlockDriverState *bs)
+{
+    NFSClient *client = bs->opaque;
+    return client->has_zero_init;
+}
+
+static int64_t nfs_get_allocated_file_size(BlockDriverState *bs)
+{
+    NFSClient *client = bs->opaque;
+    return client->allocated_file_size;
+}
+
+static BlockDriver bdrv_nfs = {
+    .format_name     = "nfs",
+    .protocol_name   = "nfs",
+
+    .instance_size   = sizeof(NFSClient),
+    .bdrv_needs_filename = true,
+    .bdrv_has_zero_init = nfs_has_zero_init,
+    .bdrv_get_allocated_file_size = nfs_get_allocated_file_size,
+
+    .bdrv_file_open  = nfs_file_open,
+    .bdrv_close      = nfs_file_close,
+    .bdrv_create     = nfs_file_create,
+
+    .bdrv_co_readv         = nfs_co_readv,
+    .bdrv_co_writev        = nfs_co_writev,
+    .bdrv_co_flush_to_disk = nfs_co_flush,
+};
+
+static void nfs_block_init(void)
+{
+    bdrv_register(&bdrv_nfs);
+}
+
+block_init(nfs_block_init);
diff --git a/configure b/configure
index 8144d9f..6c9e47a 100755
--- a/configure
+++ b/configure
@@ -250,6 +250,7 @@ vss_win32_sdk=""
 win_sdk="no"
 want_tools="yes"
 libiscsi=""
+libnfs=""
 coroutine=""
 coroutine_pool=""
 seccomp=""
@@ -833,6 +834,10 @@ for opt do
   ;;
   --enable-libiscsi) libiscsi="yes"
   ;;
+  --disable-libnfs) libnfs="no"
+  ;;
+  --enable-libnfs) libnfs="yes"
+  ;;
   --enable-profiler) profiler="yes"
   ;;
   --disable-cocoa) cocoa="no"
@@ -1202,6 +1207,8 @@ echo "  --enable-spice           enable spice"
 echo "  --enable-rbd             enable building the rados block device (rbd)"
 echo "  --disable-libiscsi       disable iscsi support"
 echo "  --enable-libiscsi        enable iscsi support"
+echo "  --disable-libnfs         disable nfs support"
+echo "  --enable-libnfs          enable nfs support"
 echo "  --disable-smartcard-nss  disable smartcard nss support"
 echo "  --enable-smartcard-nss   enable smartcard nss support"
 echo "  --disable-libusb         disable libusb (for usb passthrough)"
@@ -3050,6 +3057,32 @@ EOF
   fi
 fi
 
+##########################################
+# Do we have libnfs
+if test "$libnfs" != "no" ; then
+  cat > $TMPC << EOF
+#include <nfsc/libnfs-zdr.h>
+#include <nfsc/libnfs.h>
+#include <nfsc/libnfs-raw.h>
+#include <nfsc/libnfs-raw-mount.h>
+int main(void) {
+    nfs_init_context();
+    nfs_pread_async(0,0,0,0,0,0);
+    nfs_pwrite_async(0,0,0,0,0,0,0);
+    nfs_fstat(0,0,0);
+    return 0;
+    }
+EOF
+  if compile_prog "-Werror" "-lnfs" ; then
+    libnfs="yes"
+    LIBS="$LIBS -lnfs"
+  else
+    if test "$libnfs" = "yes" ; then
+      feature_not_found "libnfs"
+    fi
+    libnfs="no"
+  fi
+fi
 
 ##########################################
 # Do we need libm
@@ -3777,6 +3810,7 @@ echo "libusb            $libusb"
 echo "usb net redir     $usb_redir"
 echo "GLX support       $glx"
 echo "libiscsi support  $libiscsi"
+echo "libnfs support    $libnfs"
 echo "build guest agent $guest_agent"
 echo "QGA VSS support   $guest_agent_with_vss"
 echo "seccomp support   $seccomp"
@@ -4107,6 +4141,10 @@ if test "$libiscsi" = "yes" ; then
   echo "CONFIG_LIBISCSI=y" >> $config_host_mak
 fi
 
+if test "$libnfs" = "yes" ; then
+  echo "CONFIG_LIBNFS=y" >> $config_host_mak
+fi
+
 if test "$seccomp" = "yes"; then
   echo "CONFIG_SECCOMP=y" >> $config_host_mak
 fi
-- 
1.7.9.5