Skip to contentSkip to conversation

blockdev: loosen restrictions on drive-backup source node

PATCH Superseded John Snow 2 participants 4 messages May 10, 2019 – May 21, 2019

v1 of 2 Superseded Latest is v2 →

lore
Share & tools
lore

John Snow

author
#0 lore raw

PATCH 1/1 blockdev: loosen restrictions on drive-backup source node

We mandate that the source node must be a root node; but there's no reason I am aware of that it needs to be restricted to such. In some cases, we need to make sure that there's a medium present, but in the general case we can allow the backup job itself to do the graph checking.

This patch helps improve the error message when you try to backup from the same node more than once, which is reflected in the change to test 056.

For backups with bitmaps, it will also show a better error message that the bitmap is in use instead of giving you something cryptic like "need a root node."

---
 blockdev.c             | 6 +++++-
 tests/qemu-iotests/056 | 2 +-
 2 files changed, 6 insertions(+), 2 deletions(-)
blockdev.c +5 -1
index 79fbac8450..27cb72f7aa 100644
@@ -3450,7 +3450,7 @@ static BlockJob *do_drive_backup(DriveBackup *backup, JobTxn *txn,
         backup->compress = false;
     }
 
-    bs = qmp_get_root_bs(backup->device, errp);
+    bs = bdrv_lookup_bs(backup->device, backup->device, errp);
     if (!bs) {
         return NULL;
     }
@@ -3459,6 +3459,10 @@ static BlockJob *do_drive_backup(DriveBackup *backup, JobTxn *txn,
     aio_context_acquire(aio_context);
 
     if (!backup->has_format) {
+        if (!bs->drv) {
+            error_setg(errp, "Device has no medium");
+            return NULL;
+        }
         backup->format = backup->mode == NEW_IMAGE_MODE_EXISTING ?
                          NULL : (char*) bs->drv->format_name;
     }
tests/qemu-iotests/056 +1 -1
index 3df323984d..f40fc11a09 100755
@@ -214,7 +214,7 @@ class BackupTest(iotests.QMPTestCase):
         res = self.vm.qmp('query-block-jobs')
         self.assert_qmp(res, 'return[0]/status', 'concluded')
         # Leave zombie job un-dismissed, observe a failure:
-        res = self.qmp_backup_and_wait(serror='Need a root block node',
+        res = self.qmp_backup_and_wait(serror="Node 'drive0' is busy: block device is in use by block job: backup",
                                        device='drive0', format=iotests.imgfmt,
                                        sync='full', target=self.dest_img,
                                        auto_dismiss=False)
Signature · 1 line
2.20.1
author
#1 lore raw
Quoted 34 lines · John Snow

We mandate that the source node must be a root node; but there's no reason I am aware of that it needs to be restricted to such. In some cases, we need to make sure that there's a medium present, but in the general case we can allow the backup job itself to do the graph checking.

This patch helps improve the error message when you try to backup from the same node more than once, which is reflected in the change to test 056.

For backups with bitmaps, it will also show a better error message that the bitmap is in use instead of giving you something cryptic like "need a root node."

---
 blockdev.c             | 6 +++++-
 tests/qemu-iotests/056 | 2 +-
 2 files changed, 6 insertions(+), 2 deletions(-)
blockdev.c +1 -1
index 79fbac8450..27cb72f7aa 100644
@@ -3450,7 +3450,7 @@ static BlockJob *do_drive_backup(DriveBackup *backup, JobTxn *txn,
         backup->compress = false;
     }
 
-    bs = qmp_get_root_bs(backup->device, errp);
+    bs = bdrv_lookup_bs(backup->device, backup->device, errp);
     if (!bs) {
         return NULL;
     }

blockdev.c static BlockJob *do_drive_backup(DriveBackup *backup, JobTxn *txn, · line 3465

Quoted 7 lines
blockdev.c +4 -0
@@ -3459,6 +3459,10 @@ static BlockJob *do_drive_backup(DriveBackup *backup, JobTxn *txn,
     aio_context_acquire(aio_context);
 
     if (!backup->has_format) {
+        if (!bs->drv) {
+            error_setg(errp, "Device has no medium");
+            return NULL;
+        }

Pinging my own patch with a review comment. It is weird that I shuffled the error checking down below a conditional, but it's the only case where we directly need do access bs->drv now.

Otherwise, block/backup already checks for this in its own routine and I felt like it was best to let the job handle if it had the right type of arguments instead of splitting that out up here.

Still, it probably looks weird to see the "Device has no medium" error in a conditional here, so if this patch looks okay otherwise, I can send a v2 with that error checking shuffled back up to top-level to maintain some consistency with how the error checking used to be handled.

Quoted 17 lines
blockdev.c +0 -0
         backup->format = backup->mode == NEW_IMAGE_MODE_EXISTING ?
                          NULL : (char*) bs->drv->format_name;
     }
tests/qemu-iotests/056 +1 -1
index 3df323984d..f40fc11a09 100755
@@ -214,7 +214,7 @@ class BackupTest(iotests.QMPTestCase):
         res = self.vm.qmp('query-block-jobs')
         self.assert_qmp(res, 'return[0]/status', 'concluded')
         # Leave zombie job un-dismissed, observe a failure:
-        res = self.qmp_backup_and_wait(serror='Need a root block node',
+        res = self.qmp_backup_and_wait(serror="Node 'drive0' is busy: block device is in use by block job: backup",
                                        device='drive0', format=iotests.imgfmt,
                                        sync='full', target=self.dest_img,
                                        auto_dismiss=False)
#2 lore raw
Quoted 39 lines · John Snow
Quoted 36 lines · John Snow

We mandate that the source node must be a root node; but there's no reason I am aware of that it needs to be restricted to such. In some cases, we need to make sure that there's a medium present, but in the general case we can allow the backup job itself to do the graph checking.

This patch helps improve the error message when you try to backup from the same node more than once, which is reflected in the change to test 056.

For backups with bitmaps, it will also show a better error message that the bitmap is in use instead of giving you something cryptic like "need a root node."

---
  blockdev.c             | 6 +++++-
  tests/qemu-iotests/056 | 2 +-
  2 files changed, 6 insertions(+), 2 deletions(-)
blockdev.c +1 -1
index 79fbac8450..27cb72f7aa 100644
@@ -3450,7 +3450,7 @@ static BlockJob *do_drive_backup(DriveBackup *backup, JobTxn *txn,
          backup->compress = false;
      }
 
-    bs = qmp_get_root_bs(backup->device, errp);
+    bs = bdrv_lookup_bs(backup->device, backup->device, errp);
      if (!bs) {
          return NULL;
      }
@@ -3459,6 +3459,10 @@ static BlockJob *do_drive_backup(DriveBackup *backup, JobTxn *txn,
      aio_context_acquire(aio_context);
 
Quoted 9 lines

blockdev.c static BlockJob *do_drive_backup(DriveBackup *backup, JobTxn *txn, · line 3465

Quoted 5 lines
blockdev.c +4 -0
      if (!backup->has_format) {
+        if (!bs->drv) {
+            error_setg(errp, "Device has no medium");
+            return NULL;
+        }

Pinging my own patch with a review comment. It is weird that I shuffled the error checking down below a conditional, but it's the only case where we directly need do access bs->drv now.

not the only: it's accessed in following

     if (backup->mode != NEW_IMAGE_MODE_EXISTING) {
         assert(backup->format);
         if (source) {

as well.

Quoted 28 lines

Otherwise, block/backup already checks for this in its own routine and I felt like it was best to let the job handle if it had the right type of arguments instead of splitting that out up here.

Still, it probably looks weird to see the "Device has no medium" error in a conditional here, so if this patch looks okay otherwise, I can send a v2 with that error checking shuffled back up to top-level to maintain some consistency with how the error checking used to be handled.

Quoted 17 lines
blockdev.c +0 -0
          backup->format = backup->mode == NEW_IMAGE_MODE_EXISTING ?
                           NULL : (char*) bs->drv->format_name;
      }
tests/qemu-iotests/056 +1 -1
index 3df323984d..f40fc11a09 100755
@@ -214,7 +214,7 @@ class BackupTest(iotests.QMPTestCase):
          res = self.vm.qmp('query-block-jobs')
          self.assert_qmp(res, 'return[0]/status', 'concluded')
          # Leave zombie job un-dismissed, observe a failure:
-        res = self.qmp_backup_and_wait(serror='Need a root block node',
+        res = self.qmp_backup_and_wait(serror="Node 'drive0' is busy: block device is in use by block job: backup",
                                         device='drive0', format=iotests.imgfmt,
                                         sync='full', target=self.dest_img,
                                         auto_dismiss=False)
Signature · 2 lines
Best regards,
Vladimir
author
#3 lore raw
Quoted 50 lines · Vladimir Sementsov-Ogievskiy
Quoted 39 lines · John Snow
Quoted 36 lines · John Snow

We mandate that the source node must be a root node; but there's no reason I am aware of that it needs to be restricted to such. In some cases, we need to make sure that there's a medium present, but in the general case we can allow the backup job itself to do the graph checking.

This patch helps improve the error message when you try to backup from the same node more than once, which is reflected in the change to test 056.

For backups with bitmaps, it will also show a better error message that the bitmap is in use instead of giving you something cryptic like "need a root node."

---
  blockdev.c             | 6 +++++-
  tests/qemu-iotests/056 | 2 +-
  2 files changed, 6 insertions(+), 2 deletions(-)
blockdev.c +1 -1
index 79fbac8450..27cb72f7aa 100644
@@ -3450,7 +3450,7 @@ static BlockJob *do_drive_backup(DriveBackup *backup, JobTxn *txn,
          backup->compress = false;
      }
 
-    bs = qmp_get_root_bs(backup->device, errp);
+    bs = bdrv_lookup_bs(backup->device, backup->device, errp);
      if (!bs) {
          return NULL;
      }
@@ -3459,6 +3459,10 @@ static BlockJob *do_drive_backup(DriveBackup *backup, JobTxn *txn,
      aio_context_acquire(aio_context);
 
Quoted 9 lines

blockdev.c static BlockJob *do_drive_backup(DriveBackup *backup, JobTxn *txn, · line 3465

Quoted 5 lines
blockdev.c +4 -0
      if (!backup->has_format) {
+        if (!bs->drv) {
+            error_setg(errp, "Device has no medium");
+            return NULL;
+        }

Pinging my own patch with a review comment. It is weird that I shuffled the error checking down below a conditional, but it's the only case where we directly need do access bs->drv now.

Quoted 6 lines

not the only: it's accessed in following

     if (backup->mode != NEW_IMAGE_MODE_EXISTING) {
         assert(backup->format);
         if (source) {

as well.

Ah, of course. Thanks.