Skip to contentSkip to conversation

block: Generic file creation fallback

PATCH In review Max Reitz 3 participants 13 messages Jan 22, 2020 – Feb 19, 2020 Block layer

lore
Share & tools
lore

Max Reitz

author
#0 lore raw

v2 cover block: Generic file creation fallback

Hi,

As version 1, this series adds a fallback path for creating files (on the protocol layer) if the protocol driver does not support file creation, but the file already exists.

Branch: https://github.com/XanClic/qemu.git skip-proto-create-v2
Branch: https://git.xanclic.moe/XanClic/qemu.git skip-proto-create-v2

v2:
- Drop blk_truncate_for_formatting(): It doesn’t make sense to introduce this function any more after 26536c7fc25917d1bd13781f81fe3ab039643bff (“block: Do not truncate file node when formatting”), because we’d only use it in bdrv_create_file_fallback().
Thus, it makes more sense to create special helper functions specifically for bdrv_create_file_fallback().

- Thus, dropped patches 2 and 3.

- And changed patch 4 to include those helper functions.

- Rebased, which was a bit of a pain.

git-backport-diff against v1:

Key:
[----] : patches are identical
[####] : number of functional differences between upstream/downstream patch [down] : patch is downstream-only
The flags [FC] indicate (F)unctional and (C)ontextual differences, respectively

001/5:[----] [-C] 'block/nbd: Fix hang in .bdrv_close()' 002/5:[0080] [FC] 'block: Generic file creation fallback' 003/5:[----] [--] 'file-posix: Drop hdev_co_create_opts()' 004/5:[----] [--] 'iscsi: Drop iscsi_co_create_opts()' 005/5:[----] [-C] 'iotests: Add test for image creation fallback'

Max Reitz (5):
block/nbd: Fix hang in .bdrv_close()
block: Generic file creation fallback
file-posix: Drop hdev_co_create_opts()
iscsi: Drop iscsi_co_create_opts()
iotests: Add test for image creation fallback

 block.c                    | 159 ++++++++++++++++++++++++++++++++++---
 block/file-posix.c         |  67 ----------------
 block/iscsi.c              |  56 -------------
 block/nbd.c                |  14 +++-
 tests/qemu-iotests/259     |  61 ++++++++++++++
 tests/qemu-iotests/259.out |  14 ++++
 tests/qemu-iotests/group   |   1 +
 7 files changed, 236 insertions(+), 136 deletions(-)
 create mode 100755 tests/qemu-iotests/259
 create mode 100644 tests/qemu-iotests/259.out
Signature · 1 line
2.24.1
author
#1 lore raw

PATCH v2 1/5 block/nbd: Fix hang in .bdrv_close()

When nbd_close() is called from a coroutine, the connection_co never gets to run, and thus nbd_teardown_connection() hangs.

This is because aio_co_enter() only puts the connection_co into the main coroutine's wake-up queue, so this main coroutine needs to yield and wait for connection_co to terminate.

---
 block/nbd.c | 14 +++++++++++++-
 1 file changed, 13 insertions(+), 1 deletion(-)
block/nbd.c +13 -1
index d085554f21..6d3b22f844 100644
@@ -70,6 +70,7 @@ typedef struct BDRVNBDState {
     CoMutex send_mutex;
     CoQueue free_sema;
     Coroutine *connection_co;
+    Coroutine *teardown_co;
     QemuCoSleepState *connection_co_sleep_ns_state;
     bool drained;
     bool wait_drained_end;
@@ -203,7 +204,15 @@ static void nbd_teardown_connection(BlockDriverState *bs)
             qemu_co_sleep_wake(s->connection_co_sleep_ns_state);
         }
     }
-    BDRV_POLL_WHILE(bs, s->connection_co);
+    if (qemu_in_coroutine()) {
+        s->teardown_co = qemu_coroutine_self();
+        /* connection_co resumes us when it terminates */
+        qemu_coroutine_yield();
+        s->teardown_co = NULL;
+    } else {
+        BDRV_POLL_WHILE(bs, s->connection_co);
+    }
+    assert(!s->connection_co);
 }
 
 static bool nbd_client_connecting(BDRVNBDState *s)
@@ -395,6 +404,9 @@ static coroutine_fn void nbd_connection_entry(void *opaque)
         s->ioc = NULL;
     }
 
+    if (s->teardown_co) {
+        aio_co_wake(s->teardown_co);
+    }
     aio_wait_kick();
 }
 
Signature · 1 line
2.24.1
Reviewed-by on v2 1/5
#2 lore raw

Re: [PATCH v2 1/5] block/nbd: Fix hang in .bdrv_close()

Quoted 7 lines · Max Reitz

When nbd_close() is called from a coroutine, the connection_co never gets to run, and thus nbd_teardown_connection() hangs.

This is because aio_co_enter() only puts the connection_co into the main coroutine's wake-up queue, so this main coroutine needs to yield and wait for connection_co to terminate.

Quoted 5 lines
---
  block/nbd.c | 14 +++++++++++++-
  1 file changed, 13 insertions(+), 1 deletion(-)

I'm assuming it is better to take the entire series through somewhere other than my NBD tree, since the remaining patches touch more than NBD?

Signature · 3 lines
Eric Blake, Principal Software Engineer
Red Hat, Inc.           +1-919-301-3226
Virtualization:  qemu.org | libvirt.org
Reviewed-by on v2 1/5
#3 lore raw

Re: [PATCH v2 1/5] block/nbd: Fix hang in .bdrv_close()

In reply to Max Reitz
Quoted 6 lines · Max Reitz

When nbd_close() is called from a coroutine, the connection_co never gets to run, and thus nbd_teardown_connection() hangs.

This is because aio_co_enter() only puts the connection_co into the main coroutine's wake-up queue, so this main coroutine needs to yield and wait for connection_co to terminate.

Took me a while to understand this back then, good that I explained my thoughts in the review comment, although the commit message is alright when you already understand qemu co-routines I guess.

Best regards,
Maxim Levitsky

Quoted 46 lines
---
 block/nbd.c | 14 +++++++++++++-
 1 file changed, 13 insertions(+), 1 deletion(-)
block/nbd.c +13 -1
index d085554f21..6d3b22f844 100644
@@ -70,6 +70,7 @@ typedef struct BDRVNBDState {
     CoMutex send_mutex;
     CoQueue free_sema;
     Coroutine *connection_co;
+    Coroutine *teardown_co;
     QemuCoSleepState *connection_co_sleep_ns_state;
     bool drained;
     bool wait_drained_end;
@@ -203,7 +204,15 @@ static void nbd_teardown_connection(BlockDriverState *bs)
             qemu_co_sleep_wake(s->connection_co_sleep_ns_state);
         }
     }
-    BDRV_POLL_WHILE(bs, s->connection_co);
+    if (qemu_in_coroutine()) {
+        s->teardown_co = qemu_coroutine_self();
+        /* connection_co resumes us when it terminates */
+        qemu_coroutine_yield();
+        s->teardown_co = NULL;
+    } else {
+        BDRV_POLL_WHILE(bs, s->connection_co);
+    }
+    assert(!s->connection_co);
 }
 
 static bool nbd_client_connecting(BDRVNBDState *s)
@@ -395,6 +404,9 @@ static coroutine_fn void nbd_connection_entry(void *opaque)
         s->ioc = NULL;
     }
 
+    if (s->teardown_co) {
+        aio_co_wake(s->teardown_co);
+    }
     aio_wait_kick();
 }
 
author
#4 lore raw

PATCH v2 2/5 block: Generic file creation fallback

In reply to Max Reitz

If a protocol driver does not support image creation, we can see whether maybe the file exists already. If so, just truncating it will be sufficient.

---
 block.c | 159 +++++++++++++++++++++++++++++++++++++++++++++++++++-----
 1 file changed, 147 insertions(+), 12 deletions(-)
block.c +147 -12
index 99ce26d64d..e167eca04b 100644
@@ -532,20 +532,139 @@ out:
     return ret;
 }
 
-int bdrv_create_file(const char *filename, QemuOpts *opts, Error **errp)
+/**
+ * Helper function for bdrv_create_file_fallback(): Resize @blk to at
+ * least the given @minimum_size.
+ *
+ * On success, return @blk's actual length.
+ * Otherwise, return -errno.
+ */
+static int64_t create_file_fallback_truncate(BlockBackend *blk,
+                                             int64_t minimum_size, Error **errp)
 {
-    BlockDriver *drv;
+    Error *local_err = NULL;
+    int64_t size;
+    int ret;
+
+    ret = blk_truncate(blk, minimum_size, false, PREALLOC_MODE_OFF, &local_err);
+    if (ret < 0 && ret != -ENOTSUP) {
+        error_propagate(errp, local_err);
+        return ret;
+    }
+
+    size = blk_getlength(blk);
+    if (size < 0) {
+        error_free(local_err);
+        error_setg_errno(errp, -size,
+                         "Failed to inquire the new image file's length");
+        return size;
+    }
+
+    if (size < minimum_size) {
+        /* Need to grow the image, but we failed to do that */
+        error_propagate(errp, local_err);
+        return -ENOTSUP;
+    }
+
+    error_free(local_err);
+    local_err = NULL;
+
+    return size;
+}
+
+/**
+ * Helper function for bdrv_create_file_fallback(): Zero the first
+ * sector to remove any potentially pre-existing image header.
+ */
+static int create_file_fallback_zero_first_sector(BlockBackend *blk,
+                                                  int64_t current_size,
+                                                  Error **errp)
+{
+    int64_t bytes_to_clear;
+    int ret;
+
+    bytes_to_clear = MIN(current_size, BDRV_SECTOR_SIZE);
+    if (bytes_to_clear) {
+        ret = blk_pwrite_zeroes(blk, 0, bytes_to_clear, BDRV_REQ_MAY_UNMAP);
+        if (ret < 0) {
+            error_setg_errno(errp, -ret,
+                             "Failed to clear the new image's first sector");
+            return ret;
+        }
+    }
+
+    return 0;
+}
+
+static int bdrv_create_file_fallback(const char *filename, BlockDriver *drv,
+                                     QemuOpts *opts, Error **errp)
+{
+    BlockBackend *blk;
+    QDict *options = qdict_new();
+    int64_t size = 0;
+    char *buf = NULL;
+    PreallocMode prealloc;
     Error *local_err = NULL;
     int ret;
 
+    size = qemu_opt_get_size_del(opts, BLOCK_OPT_SIZE, 0);
+    buf = qemu_opt_get_del(opts, BLOCK_OPT_PREALLOC);
+    prealloc = qapi_enum_parse(&PreallocMode_lookup, buf,
+                               PREALLOC_MODE_OFF, &local_err);
+    g_free(buf);
+    if (local_err) {
+        error_propagate(errp, local_err);
+        return -EINVAL;
+    }
+
+    if (prealloc != PREALLOC_MODE_OFF) {
+        error_setg(errp, "Unsupported preallocation mode '%s'",
+                   PreallocMode_str(prealloc));
+        return -ENOTSUP;
+    }
+
+    qdict_put_str(options, "driver", drv->format_name);
+
+    blk = blk_new_open(filename, NULL, options,
+                       BDRV_O_RDWR | BDRV_O_RESIZE, errp);
+    if (!blk) {
+        error_prepend(errp, "Protocol driver '%s' does not support image "
+                      "creation, and opening the image failed: ",
+                      drv->format_name);
+        return -EINVAL;
+    }
+
+    size = create_file_fallback_truncate(blk, size, errp);
+    if (size < 0) {
+        ret = size;
+        goto out;
+    }
+
+    ret = create_file_fallback_zero_first_sector(blk, size, errp);
+    if (ret < 0) {
+        goto out;
+    }
+
+    ret = 0;
+out:
+    blk_unref(blk);
+    return ret;
+}
+
+int bdrv_create_file(const char *filename, QemuOpts *opts, Error **errp)
+{
+    BlockDriver *drv;
+
     drv = bdrv_find_protocol(filename, true, errp);
     if (drv == NULL) {
         return -ENOENT;
     }
 
-    ret = bdrv_create(drv, filename, opts, &local_err);
-    error_propagate(errp, local_err);
-    return ret;
+    if (drv->bdrv_co_create_opts) {
+        return bdrv_create(drv, filename, opts, errp);
+    } else {
+        return bdrv_create_file_fallback(filename, drv, opts, errp);
+    }
 }
 
 /**
@@ -1422,6 +1541,24 @@ QemuOptsList bdrv_runtime_opts = {
     },
 };
 
+static QemuOptsList fallback_create_opts = {
+    .name = "fallback-create-opts",
+    .head = QTAILQ_HEAD_INITIALIZER(fallback_create_opts.head),
+    .desc = {
+        {
+            .name = BLOCK_OPT_SIZE,
+            .type = QEMU_OPT_SIZE,
+            .help = "Virtual disk size"
+        },
+        {
+            .name = BLOCK_OPT_PREALLOC,
+            .type = QEMU_OPT_STRING,
+            .help = "Preallocation mode (allowed values: off)"
+        },
+        { /* end of list */ }
+    }
+};
+
 /*
  * Common part for opening disk images and files
  *
@@ -5749,15 +5886,13 @@ void bdrv_img_create(const char *filename, const char *fmt,
         return;
     }
 
-    if (!proto_drv->create_opts) {
-        error_setg(errp, "Protocol driver '%s' does not support image creation",
-                   proto_drv->format_name);
-        return;
-    }
-
     /* Create parameter list */
     create_opts = qemu_opts_append(create_opts, drv->create_opts);
-    create_opts = qemu_opts_append(create_opts, proto_drv->create_opts);
+    if (proto_drv->create_opts) {
+        create_opts = qemu_opts_append(create_opts, proto_drv->create_opts);
+    } else {
+        create_opts = qemu_opts_append(create_opts, &fallback_create_opts);
+    }
 
     opts = qemu_opts_create(create_opts, NULL, 0, &error_abort);
 
Signature · 1 line
2.24.1
#5 lore raw
Quoted 39 lines · Max Reitz

If a protocol driver does not support image creation, we can see whether maybe the file exists already. If so, just truncating it will be sufficient.

---
 block.c | 159 +++++++++++++++++++++++++++++++++++++++++++++++++++-----
 1 file changed, 147 insertions(+), 12 deletions(-)
block.c +19 -2
index 99ce26d64d..e167eca04b 100644
@@ -532,20 +532,139 @@ out:
     return ret;
 }
 
-int bdrv_create_file(const char *filename, QemuOpts *opts, Error **errp)
+/**
+ * Helper function for bdrv_create_file_fallback(): Resize @blk to at
+ * least the given @minimum_size.
+ *
+ * On success, return @blk's actual length.
+ * Otherwise, return -errno.
+ */
+static int64_t create_file_fallback_truncate(BlockBackend *blk,
+                                             int64_t minimum_size, Error **errp)
 {
-    BlockDriver *drv;
+    Error *local_err = NULL;
+    int64_t size;
+    int ret;
+
+    ret = blk_truncate(blk, minimum_size, false, PREALLOC_MODE_OFF, &local_err);
+    if (ret < 0 && ret != -ENOTSUP) {
+        error_propagate(errp, local_err);
+        return ret;
+    }
+

...

block.c out: · line 567

Quoted 13 lines
block.c +13 -0
+    size = blk_getlength(blk);
+    if (size < 0) {
+        error_free(local_err);
+        error_setg_errno(errp, -size,
+                         "Failed to inquire the new image file's length");
+        return size;
+    }
+
+    if (size < minimum_size) {
+        /* Need to grow the image, but we failed to do that */
+        error_propagate(errp, local_err);
+        return -ENOTSUP;
+    }

Very minor nitpick:

The above code basically handles case when truncate is not supported,
by trying to see file size is large enough anyway.
If truncate succeed this is a bit redundant, but doesn't hurt to be honest.
If truncate is not supported, it also works, but I am thinking that
maybe its is better to create a generic truncate failback instead
that will 'work' when asked for same size resize or smaller that existing size + exact=false.

Currently when the above failback doesn't work (meaning that we indeed have to enlarge the file) we get this error message because truncate is not supported.

[root@fedora31vm ~/qemu]# qemu-img create -f raw nvme://0000:03:00.0 10000M Formatting 'nvme://0000:03:00.0', fmt=raw size=10485760000
qemu-img: nvme://0000:03:00.0: Image format driver does not support resize

If we had generic truncate failback, it could maybe be smarter about this and say something like 'Can increase the size of the image'

But if you feel like that is not important, I don't have any issue to keep this as is, since the code does work.

Quoted 74 lines
block.c +71 -0
+    error_free(local_err);
+    local_err = NULL;
+
+    return size;
+}
+
+/**
+ * Helper function for bdrv_create_file_fallback(): Zero the first
+ * sector to remove any potentially pre-existing image header.
+ */
+static int create_file_fallback_zero_first_sector(BlockBackend *blk,
+                                                  int64_t current_size,
+                                                  Error **errp)
+{
+    int64_t bytes_to_clear;
+    int ret;
+
+    bytes_to_clear = MIN(current_size, BDRV_SECTOR_SIZE);
+    if (bytes_to_clear) {
+        ret = blk_pwrite_zeroes(blk, 0, bytes_to_clear, BDRV_REQ_MAY_UNMAP);
+        if (ret < 0) {
+            error_setg_errno(errp, -ret,
+                             "Failed to clear the new image's first sector");
+            return ret;
+        }
+    }
+
+    return 0;
+}
+
+static int bdrv_create_file_fallback(const char *filename, BlockDriver *drv,
+                                     QemuOpts *opts, Error **errp)
+{
+    BlockBackend *blk;
+    QDict *options = qdict_new();
+    int64_t size = 0;
+    char *buf = NULL;
+    PreallocMode prealloc;
     Error *local_err = NULL;
     int ret;
 
+    size = qemu_opt_get_size_del(opts, BLOCK_OPT_SIZE, 0);
+    buf = qemu_opt_get_del(opts, BLOCK_OPT_PREALLOC);
+    prealloc = qapi_enum_parse(&PreallocMode_lookup, buf,
+                               PREALLOC_MODE_OFF, &local_err);
+    g_free(buf);
+    if (local_err) {
+        error_propagate(errp, local_err);
+        return -EINVAL;
+    }
+
+    if (prealloc != PREALLOC_MODE_OFF) {
+        error_setg(errp, "Unsupported preallocation mode '%s'",
+                   PreallocMode_str(prealloc));
+        return -ENOTSUP;
+    }
+
+    qdict_put_str(options, "driver", drv->format_name);
+
+    blk = blk_new_open(filename, NULL, options,
+                       BDRV_O_RDWR | BDRV_O_RESIZE, errp);
+    if (!blk) {
+        error_prepend(errp, "Protocol driver '%s' does not support image "
+                      "creation, and opening the image failed: ",
+                      drv->format_name);
+        return -EINVAL;
+    }
+
+    size = create_file_fallback_truncate(blk, size, errp);
+    if (size < 0) {
+        ret = size;
+        goto out;
+    }
+

block.c out: · line 652

Quoted 10 lines
block.c +10 -0
+    ret = create_file_fallback_zero_first_sector(blk, size, errp);
+    if (ret < 0) {
+        goto out;
+    }
+
+    ret = 0;
+out:
+    blk_unref(blk);
+    return ret;
+}

Looks all right, very good code.

Quoted 56 lines
block.c +28 -9
+
+int bdrv_create_file(const char *filename, QemuOpts *opts, Error **errp)
+{
+    BlockDriver *drv;
+
     drv = bdrv_find_protocol(filename, true, errp);
     if (drv == NULL) {
         return -ENOENT;
     }
 
-    ret = bdrv_create(drv, filename, opts, &local_err);
-    error_propagate(errp, local_err);
-    return ret;
+    if (drv->bdrv_co_create_opts) {
+        return bdrv_create(drv, filename, opts, errp);
+    } else {
+        return bdrv_create_file_fallback(filename, drv, opts, errp);
+    }
 }
 
 /**
@@ -1422,6 +1541,24 @@ QemuOptsList bdrv_runtime_opts = {
     },
 };
 
+static QemuOptsList fallback_create_opts = {
+    .name = "fallback-create-opts",
+    .head = QTAILQ_HEAD_INITIALIZER(fallback_create_opts.head),
+    .desc = {
+        {
+            .name = BLOCK_OPT_SIZE,
+            .type = QEMU_OPT_SIZE,
+            .help = "Virtual disk size"
+        },
+        {
+            .name = BLOCK_OPT_PREALLOC,
+            .type = QEMU_OPT_STRING,
+            .help = "Preallocation mode (allowed values: off)"
+        },
+        { /* end of list */ }
+    }
+};
+
 /*
  * Common part for opening disk images and files
  *
@@ -5749,15 +5886,13 @@ void bdrv_img_create(const char *filename, const char *fmt,
         return;
     }
 
-    if (!proto_drv->create_opts) {
-        error_setg(errp, "Protocol driver '%s' does not support image creation",
-                   proto_drv->format_name);
-        return;
-    }
-

block.c void bdrv_img_create(const char *filename, const char *fmt, · line 5897

Quoted 10 lines
block.c +5 -1
     /* Create parameter list */
     create_opts = qemu_opts_append(create_opts, drv->create_opts);
-    create_opts = qemu_opts_append(create_opts, proto_drv->create_opts);
+    if (proto_drv->create_opts) {
+        create_opts = qemu_opts_append(create_opts, proto_drv->create_opts);
+    } else {
+        create_opts = qemu_opts_append(create_opts, &fallback_create_opts);
+    }
 
     opts = qemu_opts_create(create_opts, NULL, 0, &error_abort);

I also tested the code a bit on the qemu virtual nvme drive. Thanks for the patch series,

Best regards,
Maxim Levitsky

author
#6 lore raw

PATCH v2 3/5 file-posix: Drop hdev_co_create_opts()

In reply to Max Reitz

The generic fallback implementation effectively does the same.

---
 block/file-posix.c | 67 ----------------------------------------------
 1 file changed, 67 deletions(-)
block/file-posix.c +0 -67
index 1b805bd938..fd29372599 100644
@@ -3418,67 +3418,6 @@ static coroutine_fn int hdev_co_pwrite_zeroes(BlockDriverState *bs,
     return raw_do_pwrite_zeroes(bs, offset, bytes, flags, true);
 }
 
-static int coroutine_fn hdev_co_create_opts(const char *filename, QemuOpts *opts,
-                                            Error **errp)
-{
-    int fd;
-    int ret = 0;
-    struct stat stat_buf;
-    int64_t total_size = 0;
-    bool has_prefix;
-
-    /* This function is used by both protocol block drivers and therefore either
-     * of these prefixes may be given.
-     * The return value has to be stored somewhere, otherwise this is an error
-     * due to -Werror=unused-value. */
-    has_prefix =
-        strstart(filename, "host_device:", &filename) ||
-        strstart(filename, "host_cdrom:" , &filename);
-
-    (void)has_prefix;
-
-    ret = raw_normalize_devicepath(&filename, errp);
-    if (ret < 0) {
-        return ret;
-    }
-
-    /* Read out options */
-    total_size = ROUND_UP(qemu_opt_get_size_del(opts, BLOCK_OPT_SIZE, 0),
-                          BDRV_SECTOR_SIZE);
-
-    fd = qemu_open(filename, O_WRONLY | O_BINARY);
-    if (fd < 0) {
-        ret = -errno;
-        error_setg_errno(errp, -ret, "Could not open device");
-        return ret;
-    }
-
-    if (fstat(fd, &stat_buf) < 0) {
-        ret = -errno;
-        error_setg_errno(errp, -ret, "Could not stat device");
-    } else if (!S_ISBLK(stat_buf.st_mode) && !S_ISCHR(stat_buf.st_mode)) {
-        error_setg(errp,
-                   "The given file is neither a block nor a character device");
-        ret = -ENODEV;
-    } else if (lseek(fd, 0, SEEK_END) < total_size) {
-        error_setg(errp, "Device is too small");
-        ret = -ENOSPC;
-    }
-
-    if (!ret && total_size) {
-        uint8_t buf[BDRV_SECTOR_SIZE] = { 0 };
-        int64_t zero_size = MIN(BDRV_SECTOR_SIZE, total_size);
-        if (lseek(fd, 0, SEEK_SET) == -1) {
-            ret = -errno;
-        } else {
-            ret = qemu_write_full(fd, buf, zero_size);
-            ret = ret == zero_size ? 0 : -errno;
-        }
-    }
-    qemu_close(fd);
-    return ret;
-}
-
 static BlockDriver bdrv_host_device = {
     .format_name        = "host_device",
     .protocol_name        = "host_device",
@@ -3491,8 +3430,6 @@ static BlockDriver bdrv_host_device = {
     .bdrv_reopen_prepare = raw_reopen_prepare,
     .bdrv_reopen_commit  = raw_reopen_commit,
     .bdrv_reopen_abort   = raw_reopen_abort,
-    .bdrv_co_create_opts = hdev_co_create_opts,
-    .create_opts         = &raw_create_opts,
     .mutable_opts        = mutable_opts,
     .bdrv_co_invalidate_cache = raw_co_invalidate_cache,
     .bdrv_co_pwrite_zeroes = hdev_co_pwrite_zeroes,
@@ -3619,8 +3556,6 @@ static BlockDriver bdrv_host_cdrom = {
     .bdrv_reopen_prepare = raw_reopen_prepare,
     .bdrv_reopen_commit  = raw_reopen_commit,
     .bdrv_reopen_abort   = raw_reopen_abort,
-    .bdrv_co_create_opts = hdev_co_create_opts,
-    .create_opts         = &raw_create_opts,
     .mutable_opts        = mutable_opts,
     .bdrv_co_invalidate_cache = raw_co_invalidate_cache,
 
@@ -3753,8 +3688,6 @@ static BlockDriver bdrv_host_cdrom = {
     .bdrv_reopen_prepare = raw_reopen_prepare,
     .bdrv_reopen_commit  = raw_reopen_commit,
     .bdrv_reopen_abort   = raw_reopen_abort,
-    .bdrv_co_create_opts = hdev_co_create_opts,
-    .create_opts        = &raw_create_opts,
     .mutable_opts       = mutable_opts,
 
     .bdrv_co_preadv         = raw_co_preadv,
Signature · 1 line
2.24.1
author
#7 lore raw

PATCH v2 4/5 iscsi: Drop iscsi_co_create_opts()

In reply to Max Reitz

The generic fallback implementation effectively does the same.

---
 block/iscsi.c | 56 ---------------------------------------------------
 1 file changed, 56 deletions(-)
block/iscsi.c +0 -56
index 2aea7e3f13..68562e108f 100644
@@ -2163,58 +2163,6 @@ static int coroutine_fn iscsi_co_truncate(BlockDriverState *bs, int64_t offset,
     return 0;
 }
 
-static int coroutine_fn iscsi_co_create_opts(const char *filename, QemuOpts *opts,
-                                             Error **errp)
-{
-    int ret = 0;
-    int64_t total_size = 0;
-    BlockDriverState *bs;
-    IscsiLun *iscsilun = NULL;
-    QDict *bs_options;
-    Error *local_err = NULL;
-
-    bs = bdrv_new();
-
-    /* Read out options */
-    total_size = DIV_ROUND_UP(qemu_opt_get_size_del(opts, BLOCK_OPT_SIZE, 0),
-                              BDRV_SECTOR_SIZE);
-    bs->opaque = g_new0(struct IscsiLun, 1);
-    iscsilun = bs->opaque;
-
-    bs_options = qdict_new();
-    iscsi_parse_filename(filename, bs_options, &local_err);
-    if (local_err) {
-        error_propagate(errp, local_err);
-        ret = -EINVAL;
-    } else {
-        ret = iscsi_open(bs, bs_options, 0, NULL);
-    }
-    qobject_unref(bs_options);
-
-    if (ret != 0) {
-        goto out;
-    }
-    iscsi_detach_aio_context(bs);
-    if (iscsilun->type != TYPE_DISK) {
-        ret = -ENODEV;
-        goto out;
-    }
-    if (bs->total_sectors < total_size) {
-        ret = -ENOSPC;
-        goto out;
-    }
-
-    ret = 0;
-out:
-    if (iscsilun->iscsi != NULL) {
-        iscsi_destroy_context(iscsilun->iscsi);
-    }
-    g_free(bs->opaque);
-    bs->opaque = NULL;
-    bdrv_unref(bs);
-    return ret;
-}
-
 static int iscsi_get_info(BlockDriverState *bs, BlockDriverInfo *bdi)
 {
     IscsiLun *iscsilun = bs->opaque;
@@ -2485,8 +2433,6 @@ static BlockDriver bdrv_iscsi = {
     .bdrv_parse_filename    = iscsi_parse_filename,
     .bdrv_file_open         = iscsi_open,
     .bdrv_close             = iscsi_close,
-    .bdrv_co_create_opts    = iscsi_co_create_opts,
-    .create_opts            = &iscsi_create_opts,
     .bdrv_reopen_prepare    = iscsi_reopen_prepare,
     .bdrv_reopen_commit     = iscsi_reopen_commit,
     .bdrv_co_invalidate_cache = iscsi_co_invalidate_cache,
@@ -2524,8 +2470,6 @@ static BlockDriver bdrv_iser = {
     .bdrv_parse_filename    = iscsi_parse_filename,
     .bdrv_file_open         = iscsi_open,
     .bdrv_close             = iscsi_close,
-    .bdrv_co_create_opts    = iscsi_co_create_opts,
-    .create_opts            = &iscsi_create_opts,
     .bdrv_reopen_prepare    = iscsi_reopen_prepare,
     .bdrv_reopen_commit     = iscsi_reopen_commit,
     .bdrv_co_invalidate_cache  = iscsi_co_invalidate_cache,
Signature · 1 line
2.24.1
author
#8 lore raw

PATCH v2 5/5 iotests: Add test for image creation fallback

In reply to Max Reitz
---
 tests/qemu-iotests/259     | 61 ++++++++++++++++++++++++++++++++++++++
 tests/qemu-iotests/259.out | 14 +++++++++
 tests/qemu-iotests/group   |  1 +
 3 files changed, 76 insertions(+)
 create mode 100755 tests/qemu-iotests/259
 create mode 100644 tests/qemu-iotests/259.out
tests/qemu-iotests/259 new file +61 -0
new file mode 100755
index 0000000000..22b4c10241
@@ -0,0 +1,61 @@
+#!/usr/bin/env bash
+#
+# Test generic image creation fallback (by using NBD)
+#
+# Copyright (C) 2019 Red Hat, Inc.
+#
+# This program is free software; you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 2 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+#
+
+# creator
[email protected]
+
+seq=$(basename $0)
+echo "QA output created by $seq"
+
+status=1	# failure is the default!
+
+_cleanup()
+{
+    _cleanup_test_img
+}
+trap "_cleanup; exit \$status" 0 1 2 3 15
+
+# get standard environment, filters and checks
+. ./common.rc
+. ./common.filter
+
+_supported_fmt raw
+_supported_proto nbd
+_supported_os Linux
+
+
+_make_test_img 64M
+
+echo
+echo '--- Testing creation ---'
+
+$QEMU_IMG create -f qcow2 "$TEST_IMG" 64M | _filter_img_create
+$QEMU_IMG info "$TEST_IMG" | _filter_img_info
+
+echo
+echo '--- Testing creation for which the node would need to grow ---'
+
+$QEMU_IMG create -f qcow2 -o preallocation=metadata "$TEST_IMG" 64M 2>&1 \
+    | _filter_img_create
+
+# success, all done
+echo "*** done"
+rm -f $seq.full
+status=0
tests/qemu-iotests/259.out new file +14 -0
new file mode 100644
index 0000000000..ffed19c2a0
@@ -0,0 +1,14 @@
+QA output created by 259
+Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864
+
+--- Testing creation ---
+Formatting 'TEST_DIR/t.IMGFMT', fmt=qcow2 size=67108864
+image: TEST_DIR/t.IMGFMT
+file format: qcow2
+virtual size: 64 MiB (67108864 bytes)
+disk size: unavailable
+
+--- Testing creation for which the node would need to grow ---
+qemu-img: TEST_DIR/t.IMGFMT: Could not resize image: Image format driver does not support resize
+Formatting 'TEST_DIR/t.IMGFMT', fmt=qcow2 size=67108864 preallocation=metadata
+*** done
tests/qemu-iotests/group +1 -0
index 1904223020..ec47c2216a 100644
@@ -273,6 +273,7 @@
 256 rw auto quick
 257 rw
 258 rw quick
+259 rw auto quick
 260 rw quick
 261 rw
 262 rw quick migration
Signature · 1 line
2.24.1
Reviewed-by on v2 5/5
#9 lore raw

Re: [PATCH v2 5/5] iotests: Add test for image creation fallback

tests/qemu-iotests/259.out

Quoted 23 lines · Max Reitz

---

tests/qemu-iotests/259 new file +19 -0
+
+_make_test_img 64M
+
+echo
+echo '--- Testing creation ---'
+
+$QEMU_IMG create -f qcow2 "$TEST_IMG" 64M | _filter_img_create
+$QEMU_IMG info "$TEST_IMG" | _filter_img_info
+
+echo
+echo '--- Testing creation for which the node would need to grow ---'
+
+$QEMU_IMG create -f qcow2 -o preallocation=metadata "$TEST_IMG" 64M 2>&1 \
+    | _filter_img_create
+
+# success, all done
+echo "*** done"
+rm -f $seq.full
+status=0
tests/qemu-iotests/259.out +0 -0
Signature · 3 lines
Eric Blake, Principal Software Engineer
Red Hat, Inc.           +1-919-301-3226
Virtualization:  qemu.org | libvirt.org
Reviewed-by on v2 5/5
#10 lore raw

Re: [PATCH v2 5/5] iotests: Add test for image creation fallback

In reply to Max Reitz
Quoted 101 lines · Max Reitz
---
 tests/qemu-iotests/259     | 61 ++++++++++++++++++++++++++++++++++++++
 tests/qemu-iotests/259.out | 14 +++++++++
 tests/qemu-iotests/group   |  1 +
 3 files changed, 76 insertions(+)
 create mode 100755 tests/qemu-iotests/259
 create mode 100644 tests/qemu-iotests/259.out
tests/qemu-iotests/259 new file +61 -0
new file mode 100755
index 0000000000..22b4c10241
@@ -0,0 +1,61 @@
+#!/usr/bin/env bash
+#
+# Test generic image creation fallback (by using NBD)
+#
+# Copyright (C) 2019 Red Hat, Inc.
+#
+# This program is free software; you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 2 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+#
+
+# creator
[email protected]
+
+seq=$(basename $0)
+echo "QA output created by $seq"
+
+status=1	# failure is the default!
+
+_cleanup()
+{
+    _cleanup_test_img
+}
+trap "_cleanup; exit \$status" 0 1 2 3 15
+
+# get standard environment, filters and checks
+. ./common.rc
+. ./common.filter
+
+_supported_fmt raw
+_supported_proto nbd
+_supported_os Linux
+
+
+_make_test_img 64M
+
+echo
+echo '--- Testing creation ---'
+
+$QEMU_IMG create -f qcow2 "$TEST_IMG" 64M | _filter_img_create
+$QEMU_IMG info "$TEST_IMG" | _filter_img_info
+
+echo
+echo '--- Testing creation for which the node would need to grow ---'
+
+$QEMU_IMG create -f qcow2 -o preallocation=metadata "$TEST_IMG" 64M 2>&1 \
+    | _filter_img_create
+
+# success, all done
+echo "*** done"
+rm -f $seq.full
+status=0
tests/qemu-iotests/259.out new file +14 -0
new file mode 100644
index 0000000000..ffed19c2a0
@@ -0,0 +1,14 @@
+QA output created by 259
+Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864
+
+--- Testing creation ---
+Formatting 'TEST_DIR/t.IMGFMT', fmt=qcow2 size=67108864
+image: TEST_DIR/t.IMGFMT
+file format: qcow2
+virtual size: 64 MiB (67108864 bytes)
+disk size: unavailable
+
+--- Testing creation for which the node would need to grow ---
+qemu-img: TEST_DIR/t.IMGFMT: Could not resize image: Image format driver does not support resize
+Formatting 'TEST_DIR/t.IMGFMT', fmt=qcow2 size=67108864 preallocation=metadata
+*** done
tests/qemu-iotests/group +0 -0
index 1904223020..ec47c2216a 100644

tests/qemu-iotests/group line 279

Quoted 7 lines
tests/qemu-iotests/group +1 -0
@@ -273,6 +273,7 @@
 256 rw auto quick
 257 rw
 258 rw quick
+259 rw auto quick
 260 rw quick
 261 rw
 262 rw quick migration

Very minor nitpick: maybe add a note that nbd doesn't support resize / truncate. It might start supporting it one day.
I didn't notice it first.

Best regards,
Maxim Levitsky

author
#11 lore raw
In reply to Max Reitz
Quoted 19 lines · Max Reitz

Hi,

As version 1, this series adds a fallback path for creating files (on the protocol layer) if the protocol driver does not support file creation, but the file already exists.

Branch: https://github.com/XanClic/qemu.git skip-proto-create-v2
Branch: https://git.xanclic.moe/XanClic/qemu.git skip-proto-create-v2

v2:
- Drop blk_truncate_for_formatting(): It doesn’t make sense to introduce this function any more after 26536c7fc25917d1bd13781f81fe3ab039643bff (“block: Do not truncate file node when formatting”), because we’d only use it in bdrv_create_file_fallback().
Thus, it makes more sense to create special helper functions specifically for bdrv_create_file_fallback().

Quoted 5 lines

- Thus, dropped patches 2 and 3.

- And changed patch 4 to include those helper functions.

- Rebased, which was a bit of a pain.

Thanks for the reviews, added a note to the new test why the second case is expected to fail (as requested by Maxim), and applied the series to my block branch:

https://git.xanclic.moe/XanClic/qemu/commits/branch/block

Max

#12 lore raw
Quoted 26 lines · Max Reitz
Quoted 19 lines · Max Reitz

Hi,

As version 1, this series adds a fallback path for creating files (on the protocol layer) if the protocol driver does not support file creation, but the file already exists.

Branch: https://github.com/XanClic/qemu.git skip-proto-create-v2
Branch: https://git.xanclic.moe/XanClic/qemu.git skip-proto-create-v2

v2:
- Drop blk_truncate_for_formatting(): It doesn’t make sense to introduce this function any more after 26536c7fc25917d1bd13781f81fe3ab039643bff (“block: Do not truncate file node when formatting”), because we’d only use it in bdrv_create_file_fallback().
Thus, it makes more sense to create special helper functions specifically for bdrv_create_file_fallback().

Quoted 5 lines

- Thus, dropped patches 2 and 3.

- And changed patch 4 to include those helper functions.

- Rebased, which was a bit of a pain.

Quoted 8 lines

Thanks for the reviews, added a note to the new test why the second case is expected to fail (as requested by Maxim), and applied the series to my block branch:

https://git.xanclic.moe/XanClic/qemu/commits/branch/block

Max

Thank you too!
Best regards,
Maxim Levitsky