The following changes since commit 22d96eac64877c4d96f9928babb6f2fcc68faacf: Merge remote-tracking branch 'remotes/stefanha/tags/tracing-pull-request' into staging (2019-04-29 19:11:15 +0100) are available in the Git repository at: git://repo.or.cz/qemu/kevin.git tags/for-upstream for you to fetch changes up to 54277a2aab876aba7b55c7e88e2b372691849741: block/qed: add missed coroutine_fn markers (2019-04-30 15:29:00 +0200) ---------------------------------------------------------------- Block layer patches: - iotests: Fix output of qemu-io related tests - Don't ignore bdrv_set_aio_context() for nodes with bs->drv = NUL - vmdk: Set vmdk parent backing_format to vmdk - qcow2: Preallocation fixes (especially for external data files) - Add linear-buffer-based APIs (as wrappers around qiov-based ones) - Various code cleanups and small corner case fixes ---------------------------------------------------------------- Alberto Garcia (3): qcow2: Fix error handling in the compression code commit: Make base read-only if there is an early failure iotests: Check that images are in read-only mode after block-commit Eric Blake (2): cutils: Fix size_to_str() on 32-bit platforms qemu-img: Saner printing of large file sizes Kevin Wolf (5): block: Fix AioContext switch for bs->drv == NULL qcow2: Avoid COW during metadata preallocation qcow2: Add errp to preallocate_co() qcow2: Fix full preallocation with external data file qcow2: Fix qcow2_make_empty() with external data file Max Reitz (2): iotests: Perform the correct test in 082 qemu-img: Make create hint at protocol options Sam Eiderman (1): vmdk: Set vmdk parent backing_format to vmdk Stefano Garzarella (2): block/vhdx: Remove redundant IEC binary prefixes definition block/vhdx: Use IEC binary prefixes for size constants Thomas Huth (1): tests/qemu-iotests: Fix output of qemu-io related tests Vladimir Sementsov-Ogievskiy (10): block: introduce byte-based io helpers block/qcow2: use buffer-based io block/qcow: use buffer-based io block/qed: use buffer-based io block/parallels: use buffer-based io block/backup: use buffer-based io block/commit: use buffer-based io block/stream: use buffer-based io qemu-img: use buffer-based io block/qed: add missed coroutine_fn markers Zhengui li (1): vpc: unlock Coroutine lock to make IO submit Concurrently block/qed.h | 28 ++++--- block/vhdx.h | 16 ++-- include/block/block_int.h | 16 ++++ include/sysemu/block-backend.h | 19 +++++ block.c | 12 +-- block/backup.c | 14 ++-- block/commit.c | 8 +- block/parallels.c | 14 ++-- block/qapi.c | 49 +++--------- block/qcow.c | 19 ++--- block/qcow2.c | 80 ++++++++++---------- block/qed-check.c | 4 +- block/qed-table.c | 45 ++++++----- block/qed.c | 11 ++- block/stream.c | 4 +- block/vhdx-log.c | 2 +- block/vhdx.c | 7 +- block/vmdk.c | 2 + block/vpc.c | 4 + qemu-img.c | 26 ++++--- util/cutils.c | 2 +- tests/qemu-iotests/026.out | 168 ++++++++++++++++++++--------------------- tests/qemu-iotests/043.out | 6 +- tests/qemu-iotests/053.out | 2 +- tests/qemu-iotests/059.out | 10 +-- tests/qemu-iotests/060.out | 16 ++-- tests/qemu-iotests/061.out | 12 +-- tests/qemu-iotests/069.out | 2 +- tests/qemu-iotests/070.out | 4 +- tests/qemu-iotests/075.out | 14 ++-- tests/qemu-iotests/076.out | 6 +- tests/qemu-iotests/078.out | 12 +-- tests/qemu-iotests/080.out | 40 +++++----- tests/qemu-iotests/081.out | 2 +- tests/qemu-iotests/082 | 5 +- tests/qemu-iotests/082.out | 51 ++++++++----- tests/qemu-iotests/084.out | 8 +- tests/qemu-iotests/088.out | 12 +-- tests/qemu-iotests/089.out | 2 +- tests/qemu-iotests/095.out | 4 +- tests/qemu-iotests/103.out | 14 ++-- tests/qemu-iotests/104.out | 6 +- tests/qemu-iotests/110.out | 6 +- tests/qemu-iotests/114.out | 4 +- tests/qemu-iotests/116.out | 14 ++-- tests/qemu-iotests/126.out | 4 +- tests/qemu-iotests/130.out | 10 +-- tests/qemu-iotests/131.out | 2 +- tests/qemu-iotests/133.out | 30 ++++---- tests/qemu-iotests/137.out | 28 +++---- tests/qemu-iotests/140.out | 2 +- tests/qemu-iotests/143.out | 2 +- tests/qemu-iotests/153.out | 32 ++++---- tests/qemu-iotests/187.out | 6 +- tests/qemu-iotests/188.out | 2 +- tests/qemu-iotests/191.out | 8 +- tests/qemu-iotests/195.out | 4 +- tests/qemu-iotests/197.out | 2 +- tests/qemu-iotests/198.out | 4 +- tests/qemu-iotests/205 | 2 +- tests/qemu-iotests/206.out | 10 +-- tests/qemu-iotests/207.out | 12 +-- tests/qemu-iotests/210.out | 8 +- tests/qemu-iotests/211.out | 10 +-- tests/qemu-iotests/212.out | 10 +-- tests/qemu-iotests/213.out | 10 +-- tests/qemu-iotests/215.out | 2 +- tests/qemu-iotests/226.out | 16 ++-- tests/qemu-iotests/233.out | 4 +- tests/qemu-iotests/237.out | 22 +++--- tests/qemu-iotests/242.out | 10 +-- tests/qemu-iotests/244.out | 10 +-- tests/qemu-iotests/249 | 115 ++++++++++++++++++++++++++++ tests/qemu-iotests/249.out | 35 +++++++++ tests/qemu-iotests/group | 1 + 75 files changed, 696 insertions(+), 519 deletions(-) create mode 100755 tests/qemu-iotests/249 create mode 100644 tests/qemu-iotests/249.out From: Thomas Huth <thuth@redhat.com> One of the recent commits changed the way qemu-io prints out its errors and warnings - they are now prefixed with the program name. We've got to adapt the iotests accordingly to prevent that they are failing. Fixes: 99e98d7c9fc1a1639fad ("qemu-io: Use error_[gs]et_progname()") Signed-off-by: Thomas Huth <thuth@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- tests/qemu-iotests/026.out | 168 ++++++++++++++++++------------------- tests/qemu-iotests/060.out | 6 +- tests/qemu-iotests/069.out | 2 +- tests/qemu-iotests/070.out | 2 +- tests/qemu-iotests/075.out | 14 ++-- tests/qemu-iotests/076.out | 6 +- tests/qemu-iotests/078.out | 12 +-- tests/qemu-iotests/080.out | 40 ++++----- tests/qemu-iotests/081.out | 2 +- tests/qemu-iotests/088.out | 12 +-- tests/qemu-iotests/103.out | 14 ++-- tests/qemu-iotests/114.out | 2 +- tests/qemu-iotests/116.out | 14 ++-- tests/qemu-iotests/131.out | 2 +- tests/qemu-iotests/133.out | 30 +++---- tests/qemu-iotests/137.out | 28 +++---- tests/qemu-iotests/140.out | 2 +- tests/qemu-iotests/143.out | 2 +- tests/qemu-iotests/153.out | 30 +++---- tests/qemu-iotests/187.out | 6 +- tests/qemu-iotests/188.out | 2 +- tests/qemu-iotests/197.out | 2 +- tests/qemu-iotests/205 | 2 +- tests/qemu-iotests/215.out | 2 +- tests/qemu-iotests/226.out | 16 ++-- tests/qemu-iotests/244.out | 10 +-- 26 files changed, 214 insertions(+), 214 deletions(-) diff --git a/tests/qemu-iotests/026.out b/tests/qemu-iotests/026.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/026.out +++ b/tests/qemu-iotests/026.out @@ -XXX,XX +XXX,XX @@ No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: l1_update; errno: 5; imm: off; once: off; write -Failed to flush the L2 table cache: Input/output error -Failed to flush the refcount block cache: Input/output error +qemu-io: Failed to flush the L2 table cache: Input/output error +qemu-io: Failed to flush the refcount block cache: Input/output error write failed: Input/output error 1 leaked clusters were found on the image. @@ -XXX,XX +XXX,XX @@ This means waste of disk space, but no harm to data. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: l1_update; errno: 5; imm: off; once: off; write -b -Failed to flush the L2 table cache: Input/output error -Failed to flush the refcount block cache: Input/output error +qemu-io: Failed to flush the L2 table cache: Input/output error +qemu-io: Failed to flush the refcount block cache: Input/output error write failed: Input/output error 1 leaked clusters were found on the image. @@ -XXX,XX +XXX,XX @@ No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: l1_update; errno: 28; imm: off; once: off; write -Failed to flush the L2 table cache: No space left on device -Failed to flush the refcount block cache: No space left on device +qemu-io: Failed to flush the L2 table cache: No space left on device +qemu-io: Failed to flush the refcount block cache: No space left on device write failed: No space left on device 1 leaked clusters were found on the image. @@ -XXX,XX +XXX,XX @@ This means waste of disk space, but no harm to data. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: l1_update; errno: 28; imm: off; once: off; write -b -Failed to flush the L2 table cache: No space left on device -Failed to flush the refcount block cache: No space left on device +qemu-io: Failed to flush the L2 table cache: No space left on device +qemu-io: Failed to flush the refcount block cache: No space left on device write failed: No space left on device 1 leaked clusters were found on the image. @@ -XXX,XX +XXX,XX @@ No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: l2_update; errno: 5; imm: off; once: off; write -Failed to flush the L2 table cache: Input/output error -Failed to flush the refcount block cache: Input/output error +qemu-io: Failed to flush the L2 table cache: Input/output error +qemu-io: Failed to flush the refcount block cache: Input/output error write failed: Input/output error 127 leaked clusters were found on the image. @@ -XXX,XX +XXX,XX @@ This means waste of disk space, but no harm to data. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: l2_update; errno: 5; imm: off; once: off; write -b -Failed to flush the L2 table cache: Input/output error -Failed to flush the refcount block cache: Input/output error +qemu-io: Failed to flush the L2 table cache: Input/output error +qemu-io: Failed to flush the refcount block cache: Input/output error write failed: Input/output error 127 leaked clusters were found on the image. @@ -XXX,XX +XXX,XX @@ No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: l2_update; errno: 28; imm: off; once: off; write -Failed to flush the L2 table cache: No space left on device -Failed to flush the refcount block cache: No space left on device +qemu-io: Failed to flush the L2 table cache: No space left on device +qemu-io: Failed to flush the refcount block cache: No space left on device write failed: No space left on device 127 leaked clusters were found on the image. @@ -XXX,XX +XXX,XX @@ This means waste of disk space, but no harm to data. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: l2_update; errno: 28; imm: off; once: off; write -b -Failed to flush the L2 table cache: No space left on device -Failed to flush the refcount block cache: No space left on device +qemu-io: Failed to flush the L2 table cache: No space left on device +qemu-io: Failed to flush the refcount block cache: No space left on device write failed: No space left on device 127 leaked clusters were found on the image. @@ -XXX,XX +XXX,XX @@ No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: l2_alloc_write; errno: 5; imm: off; once: off; write -Failed to flush the L2 table cache: Input/output error -Failed to flush the refcount block cache: Input/output error +qemu-io: Failed to flush the L2 table cache: Input/output error +qemu-io: Failed to flush the refcount block cache: Input/output error write failed: Input/output error No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: l2_alloc_write; errno: 5; imm: off; once: off; write -b -Failed to flush the L2 table cache: Input/output error -Failed to flush the refcount block cache: Input/output error +qemu-io: Failed to flush the L2 table cache: Input/output error +qemu-io: Failed to flush the refcount block cache: Input/output error write failed: Input/output error 1 leaked clusters were found on the image. @@ -XXX,XX +XXX,XX @@ No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: l2_alloc_write; errno: 28; imm: off; once: off; write -Failed to flush the L2 table cache: No space left on device -Failed to flush the refcount block cache: No space left on device +qemu-io: Failed to flush the L2 table cache: No space left on device +qemu-io: Failed to flush the refcount block cache: No space left on device write failed: No space left on device No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: l2_alloc_write; errno: 28; imm: off; once: off; write -b -Failed to flush the L2 table cache: No space left on device -Failed to flush the refcount block cache: No space left on device +qemu-io: Failed to flush the L2 table cache: No space left on device +qemu-io: Failed to flush the refcount block cache: No space left on device write failed: No space left on device 1 leaked clusters were found on the image. @@ -XXX,XX +XXX,XX @@ No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: write_aio; errno: 5; imm: off; once: off; write -Failed to flush the L2 table cache: Input/output error -Failed to flush the refcount block cache: Input/output error +qemu-io: Failed to flush the L2 table cache: Input/output error +qemu-io: Failed to flush the refcount block cache: Input/output error write failed: Input/output error No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: write_aio; errno: 5; imm: off; once: off; write -b -Failed to flush the L2 table cache: Input/output error -Failed to flush the refcount block cache: Input/output error +qemu-io: Failed to flush the L2 table cache: Input/output error +qemu-io: Failed to flush the refcount block cache: Input/output error write failed: Input/output error No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 @@ -XXX,XX +XXX,XX @@ No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: write_aio; errno: 28; imm: off; once: off; write -Failed to flush the L2 table cache: No space left on device -Failed to flush the refcount block cache: No space left on device +qemu-io: Failed to flush the L2 table cache: No space left on device +qemu-io: Failed to flush the refcount block cache: No space left on device write failed: No space left on device No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: write_aio; errno: 28; imm: off; once: off; write -b -Failed to flush the L2 table cache: No space left on device -Failed to flush the refcount block cache: No space left on device +qemu-io: Failed to flush the L2 table cache: No space left on device +qemu-io: Failed to flush the refcount block cache: No space left on device write failed: No space left on device No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 @@ -XXX,XX +XXX,XX @@ No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: refblock_load; errno: 5; imm: off; once: off; write -Failed to flush the L2 table cache: Input/output error -Failed to flush the refcount block cache: Input/output error +qemu-io: Failed to flush the L2 table cache: Input/output error +qemu-io: Failed to flush the refcount block cache: Input/output error write failed: Input/output error No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: refblock_load; errno: 5; imm: off; once: off; write -b -Failed to flush the L2 table cache: Input/output error -Failed to flush the refcount block cache: Input/output error +qemu-io: Failed to flush the L2 table cache: Input/output error +qemu-io: Failed to flush the refcount block cache: Input/output error write failed: Input/output error No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 @@ -XXX,XX +XXX,XX @@ No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: refblock_load; errno: 28; imm: off; once: off; write -Failed to flush the L2 table cache: No space left on device -Failed to flush the refcount block cache: No space left on device +qemu-io: Failed to flush the L2 table cache: No space left on device +qemu-io: Failed to flush the refcount block cache: No space left on device write failed: No space left on device No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: refblock_load; errno: 28; imm: off; once: off; write -b -Failed to flush the L2 table cache: No space left on device -Failed to flush the refcount block cache: No space left on device +qemu-io: Failed to flush the L2 table cache: No space left on device +qemu-io: Failed to flush the refcount block cache: No space left on device write failed: No space left on device No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 @@ -XXX,XX +XXX,XX @@ No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: refblock_update_part; errno: 5; imm: off; once: off; write -Failed to flush the L2 table cache: Input/output error -Failed to flush the refcount block cache: Input/output error +qemu-io: Failed to flush the L2 table cache: Input/output error +qemu-io: Failed to flush the refcount block cache: Input/output error write failed: Input/output error No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: refblock_update_part; errno: 5; imm: off; once: off; write -b -Failed to flush the L2 table cache: Input/output error -Failed to flush the refcount block cache: Input/output error +qemu-io: Failed to flush the L2 table cache: Input/output error +qemu-io: Failed to flush the refcount block cache: Input/output error write failed: Input/output error No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 @@ -XXX,XX +XXX,XX @@ No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: refblock_update_part; errno: 28; imm: off; once: off; write -Failed to flush the L2 table cache: No space left on device -Failed to flush the refcount block cache: No space left on device +qemu-io: Failed to flush the L2 table cache: No space left on device +qemu-io: Failed to flush the refcount block cache: No space left on device write failed: No space left on device No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: refblock_update_part; errno: 28; imm: off; once: off; write -b -Failed to flush the L2 table cache: No space left on device -Failed to flush the refcount block cache: No space left on device +qemu-io: Failed to flush the L2 table cache: No space left on device +qemu-io: Failed to flush the refcount block cache: No space left on device write failed: No space left on device No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 @@ -XXX,XX +XXX,XX @@ No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: refblock_alloc; errno: 5; imm: off; once: off; write -Failed to flush the L2 table cache: Input/output error -Failed to flush the refcount block cache: Input/output error +qemu-io: Failed to flush the L2 table cache: Input/output error +qemu-io: Failed to flush the refcount block cache: Input/output error write failed: Input/output error No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: refblock_alloc; errno: 5; imm: off; once: off; write -b -Failed to flush the L2 table cache: Input/output error -Failed to flush the refcount block cache: Input/output error +qemu-io: Failed to flush the L2 table cache: Input/output error +qemu-io: Failed to flush the refcount block cache: Input/output error write failed: Input/output error No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 @@ -XXX,XX +XXX,XX @@ No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: refblock_alloc; errno: 28; imm: off; once: off; write -Failed to flush the L2 table cache: No space left on device -Failed to flush the refcount block cache: No space left on device +qemu-io: Failed to flush the L2 table cache: No space left on device +qemu-io: Failed to flush the refcount block cache: No space left on device write failed: No space left on device No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: refblock_alloc; errno: 28; imm: off; once: off; write -b -Failed to flush the L2 table cache: No space left on device -Failed to flush the refcount block cache: No space left on device +qemu-io: Failed to flush the L2 table cache: No space left on device +qemu-io: Failed to flush the refcount block cache: No space left on device write failed: No space left on device No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 @@ -XXX,XX +XXX,XX @@ No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: refblock_alloc_hookup; errno: 28; imm: off; once: off; write -Failed to flush the L2 table cache: No space left on device -Failed to flush the refcount block cache: No space left on device +qemu-io: Failed to flush the L2 table cache: No space left on device +qemu-io: Failed to flush the refcount block cache: No space left on device write failed: No space left on device 55 leaked clusters were found on the image. @@ -XXX,XX +XXX,XX @@ This means waste of disk space, but no harm to data. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: refblock_alloc_hookup; errno: 28; imm: off; once: off; write -b -Failed to flush the L2 table cache: No space left on device -Failed to flush the refcount block cache: No space left on device +qemu-io: Failed to flush the L2 table cache: No space left on device +qemu-io: Failed to flush the refcount block cache: No space left on device write failed: No space left on device 251 leaked clusters were found on the image. @@ -XXX,XX +XXX,XX @@ No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: refblock_alloc_write; errno: 28; imm: off; once: off; write -Failed to flush the L2 table cache: No space left on device -Failed to flush the refcount block cache: No space left on device +qemu-io: Failed to flush the L2 table cache: No space left on device +qemu-io: Failed to flush the refcount block cache: No space left on device write failed: No space left on device No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: refblock_alloc_write; errno: 28; imm: off; once: off; write -b -Failed to flush the L2 table cache: No space left on device -Failed to flush the refcount block cache: No space left on device +qemu-io: Failed to flush the L2 table cache: No space left on device +qemu-io: Failed to flush the refcount block cache: No space left on device write failed: No space left on device No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 @@ -XXX,XX +XXX,XX @@ No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: refblock_alloc_write_blocks; errno: 28; imm: off; once: off; write -Failed to flush the L2 table cache: No space left on device -Failed to flush the refcount block cache: No space left on device +qemu-io: Failed to flush the L2 table cache: No space left on device +qemu-io: Failed to flush the refcount block cache: No space left on device write failed: No space left on device 10 leaked clusters were found on the image. @@ -XXX,XX +XXX,XX @@ This means waste of disk space, but no harm to data. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: refblock_alloc_write_blocks; errno: 28; imm: off; once: off; write -b -Failed to flush the L2 table cache: No space left on device -Failed to flush the refcount block cache: No space left on device +qemu-io: Failed to flush the L2 table cache: No space left on device +qemu-io: Failed to flush the refcount block cache: No space left on device write failed: No space left on device 23 leaked clusters were found on the image. @@ -XXX,XX +XXX,XX @@ No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: refblock_alloc_write_table; errno: 28; imm: off; once: off; write -Failed to flush the L2 table cache: No space left on device -Failed to flush the refcount block cache: No space left on device +qemu-io: Failed to flush the L2 table cache: No space left on device +qemu-io: Failed to flush the refcount block cache: No space left on device write failed: No space left on device 10 leaked clusters were found on the image. @@ -XXX,XX +XXX,XX @@ This means waste of disk space, but no harm to data. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: refblock_alloc_write_table; errno: 28; imm: off; once: off; write -b -Failed to flush the L2 table cache: No space left on device -Failed to flush the refcount block cache: No space left on device +qemu-io: Failed to flush the L2 table cache: No space left on device +qemu-io: Failed to flush the refcount block cache: No space left on device write failed: No space left on device 23 leaked clusters were found on the image. @@ -XXX,XX +XXX,XX @@ No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: refblock_alloc_switch_table; errno: 28; imm: off; once: off; write -Failed to flush the L2 table cache: No space left on device -Failed to flush the refcount block cache: No space left on device +qemu-io: Failed to flush the L2 table cache: No space left on device +qemu-io: Failed to flush the refcount block cache: No space left on device write failed: No space left on device 10 leaked clusters were found on the image. @@ -XXX,XX +XXX,XX @@ This means waste of disk space, but no harm to data. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: refblock_alloc_switch_table; errno: 28; imm: off; once: off; write -b -Failed to flush the L2 table cache: No space left on device -Failed to flush the refcount block cache: No space left on device +qemu-io: Failed to flush the L2 table cache: No space left on device +qemu-io: Failed to flush the refcount block cache: No space left on device write failed: No space left on device 23 leaked clusters were found on the image. @@ -XXX,XX +XXX,XX @@ No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: l1_grow_write_table; errno: 5; imm: off; once: off -Failed to flush the L2 table cache: Input/output error -Failed to flush the refcount block cache: Input/output error +qemu-io: Failed to flush the L2 table cache: Input/output error +qemu-io: Failed to flush the refcount block cache: Input/output error write failed: Input/output error No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 @@ -XXX,XX +XXX,XX @@ No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: l1_grow_write_table; errno: 28; imm: off; once: off -Failed to flush the L2 table cache: No space left on device -Failed to flush the refcount block cache: No space left on device +qemu-io: Failed to flush the L2 table cache: No space left on device +qemu-io: Failed to flush the refcount block cache: No space left on device write failed: No space left on device No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 @@ -XXX,XX +XXX,XX @@ No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: l1_grow_activate_table; errno: 5; imm: off; once: off -Failed to flush the L2 table cache: Input/output error -Failed to flush the refcount block cache: Input/output error +qemu-io: Failed to flush the L2 table cache: Input/output error +qemu-io: Failed to flush the refcount block cache: Input/output error write failed: Input/output error 96 leaked clusters were found on the image. @@ -XXX,XX +XXX,XX @@ No errors were found on the image. Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824 Event: l1_grow_activate_table; errno: 28; imm: off; once: off -Failed to flush the L2 table cache: No space left on device -Failed to flush the refcount block cache: No space left on device +qemu-io: Failed to flush the L2 table cache: No space left on device +qemu-io: Failed to flush the refcount block cache: No space left on device write failed: No space left on device 96 leaked clusters were found on the image. diff --git a/tests/qemu-iotests/060.out b/tests/qemu-iotests/060.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/060.out +++ b/tests/qemu-iotests/060.out @@ -XXX,XX +XXX,XX @@ Format specific information: lazy refcounts: false refcount bits: 16 corrupt: true -can't open device TEST_DIR/t.IMGFMT: IMGFMT: Image is corrupt; cannot be opened read/write +qemu-io: can't open device TEST_DIR/t.IMGFMT: IMGFMT: Image is corrupt; cannot be opened read/write no file open, try 'help open' read 512/512 bytes at offset 0 512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) @@ -XXX,XX +XXX,XX @@ No errors were found on the image. === Testing zero refcount table size === Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 -can't open device TEST_DIR/t.IMGFMT: Image does not contain a reference count table +qemu-io: can't open device TEST_DIR/t.IMGFMT: Image does not contain a reference count table ERROR cluster 0 refcount=0 reference=1 ERROR cluster 3 refcount=0 reference=1 Rebuilding refcount structure @@ -XXX,XX +XXX,XX @@ Can't get refcount for cluster 2: Input/output error Can't get refcount for cluster 3: Input/output error Rebuilding refcount structure Repairing cluster 1 refcount=1 reference=0 -can't open device TEST_DIR/t.IMGFMT: Could not repair dirty image: Input/output error +qemu-io: can't open device TEST_DIR/t.IMGFMT: Could not repair dirty image: Input/output error --- Repairing --- Leaked cluster 1 refcount=1 reference=0 Repairing cluster 1 refcount=1 reference=0 diff --git a/tests/qemu-iotests/069.out b/tests/qemu-iotests/069.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/069.out +++ b/tests/qemu-iotests/069.out @@ -XXX,XX +XXX,XX @@ QA output created by 069 Formatting 'TEST_DIR/t.IMGFMT.base', fmt=IMGFMT size=131072 Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=131072 backing_file=TEST_DIR/t.IMGFMT.base -can't open device TEST_DIR/t.IMGFMT: Could not open backing file: Could not open 'TEST_DIR/t.IMGFMT.base': No such file or directory +qemu-io: can't open device TEST_DIR/t.IMGFMT: Could not open backing file: Could not open 'TEST_DIR/t.IMGFMT.base': No such file or directory *** done diff --git a/tests/qemu-iotests/070.out b/tests/qemu-iotests/070.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/070.out +++ b/tests/qemu-iotests/070.out @@ -XXX,XX +XXX,XX @@ QA output created by 070 === Verify open image read-only fails, due to dirty log === -can't open device TEST_DIR/iotest-dirtylog-10G-4M.vhdx: VHDX image file 'TEST_DIR/iotest-dirtylog-10G-4M.vhdx' opened read-only, but contains a log that needs to be replayed +qemu-io: can't open device TEST_DIR/iotest-dirtylog-10G-4M.vhdx: VHDX image file 'TEST_DIR/iotest-dirtylog-10G-4M.vhdx' opened read-only, but contains a log that needs to be replayed To replay the log, run: qemu-img check -r all 'TEST_DIR/iotest-dirtylog-10G-4M.vhdx' === Verify open image replays log === diff --git a/tests/qemu-iotests/075.out b/tests/qemu-iotests/075.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/075.out +++ b/tests/qemu-iotests/075.out @@ -XXX,XX +XXX,XX @@ read 512/512 bytes at offset 1048064 512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) == block_size must be a multiple of 512 == -can't open device TEST_DIR/simple-pattern.cloop: block_size 513 must be a multiple of 512 +qemu-io: can't open device TEST_DIR/simple-pattern.cloop: block_size 513 must be a multiple of 512 == block_size cannot be zero == -can't open device TEST_DIR/simple-pattern.cloop: block_size cannot be zero +qemu-io: can't open device TEST_DIR/simple-pattern.cloop: block_size cannot be zero == huge block_size === -can't open device TEST_DIR/simple-pattern.cloop: block_size 4294966784 must be 64 MB or less +qemu-io: can't open device TEST_DIR/simple-pattern.cloop: block_size 4294966784 must be 64 MB or less == offsets_size overflow === -can't open device TEST_DIR/simple-pattern.cloop: n_blocks 4294967295 must be 536870911 or less +qemu-io: can't open device TEST_DIR/simple-pattern.cloop: n_blocks 4294967295 must be 536870911 or less == refuse images that require too many offsets === -can't open device TEST_DIR/simple-pattern.cloop: image requires too many offsets, try increasing block size +qemu-io: can't open device TEST_DIR/simple-pattern.cloop: image requires too many offsets, try increasing block size == refuse images with non-monotonically increasing offsets == -can't open device TEST_DIR/simple-pattern.cloop: offsets not monotonically increasing at index 1, image file is corrupt +qemu-io: can't open device TEST_DIR/simple-pattern.cloop: offsets not monotonically increasing at index 1, image file is corrupt == refuse images with invalid compressed block size == -can't open device TEST_DIR/simple-pattern.cloop: invalid compressed block size at index 1, image file is corrupt +qemu-io: can't open device TEST_DIR/simple-pattern.cloop: invalid compressed block size at index 1, image file is corrupt *** done diff --git a/tests/qemu-iotests/076.out b/tests/qemu-iotests/076.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/076.out +++ b/tests/qemu-iotests/076.out @@ -XXX,XX +XXX,XX @@ read 65536/65536 bytes at offset 0 64 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) == Negative catalog size == -can't open device TEST_DIR/parallels-v1: Catalog too large +qemu-io: can't open device TEST_DIR/parallels-v1: Catalog too large == Overflow in catalog allocation == -can't open device TEST_DIR/parallels-v1: Catalog too large +qemu-io: can't open device TEST_DIR/parallels-v1: Catalog too large == Zero sectors per track == -can't open device TEST_DIR/parallels-v1: Invalid image: Zero sectors per track +qemu-io: can't open device TEST_DIR/parallels-v1: Invalid image: Zero sectors per track == Read from a valid v2 image == read 65536/65536 bytes at offset 0 diff --git a/tests/qemu-iotests/078.out b/tests/qemu-iotests/078.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/078.out +++ b/tests/qemu-iotests/078.out @@ -XXX,XX +XXX,XX @@ read 512/512 bytes at offset 0 512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) == Negative catalog size == -can't open device TEST_DIR/empty.bochs: Catalog size is too large +qemu-io: can't open device TEST_DIR/empty.bochs: Catalog size is too large == Overflow for catalog size * sizeof(uint32_t) == -can't open device TEST_DIR/empty.bochs: Catalog size is too large +qemu-io: can't open device TEST_DIR/empty.bochs: Catalog size is too large == Too small catalog bitmap for image size == -can't open device TEST_DIR/empty.bochs: Catalog size is too small for this disk size -can't open device TEST_DIR/empty.bochs: Catalog size is too small for this disk size +qemu-io: can't open device TEST_DIR/empty.bochs: Catalog size is too small for this disk size +qemu-io: can't open device TEST_DIR/empty.bochs: Catalog size is too small for this disk size == Negative extent size == -can't open device TEST_DIR/empty.bochs: Extent size 2147483648 is too large +qemu-io: can't open device TEST_DIR/empty.bochs: Extent size 2147483648 is too large == Zero extent size == -can't open device TEST_DIR/empty.bochs: Extent size must be at least 512 +qemu-io: can't open device TEST_DIR/empty.bochs: Extent size must be at least 512 *** done diff --git a/tests/qemu-iotests/080.out b/tests/qemu-iotests/080.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/080.out +++ b/tests/qemu-iotests/080.out @@ -XXX,XX +XXX,XX @@ QA output created by 080 == Huge header size == Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 -can't open device TEST_DIR/t.qcow2: qcow2 header exceeds cluster size -can't open device TEST_DIR/t.qcow2: qcow2 header exceeds cluster size +qemu-io: can't open device TEST_DIR/t.qcow2: qcow2 header exceeds cluster size +qemu-io: can't open device TEST_DIR/t.qcow2: qcow2 header exceeds cluster size == Huge unknown header extension == Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 -can't open device TEST_DIR/t.qcow2: Invalid backing file offset -can't open device TEST_DIR/t.qcow2: Header extension too large -can't open device TEST_DIR/t.qcow2: Header extension too large +qemu-io: can't open device TEST_DIR/t.qcow2: Invalid backing file offset +qemu-io: can't open device TEST_DIR/t.qcow2: Header extension too large +qemu-io: can't open device TEST_DIR/t.qcow2: Header extension too large == Huge refcount table size == Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 -can't open device TEST_DIR/t.qcow2: Reference count table too large -can't open device TEST_DIR/t.qcow2: Reference count table too large +qemu-io: can't open device TEST_DIR/t.qcow2: Reference count table too large +qemu-io: can't open device TEST_DIR/t.qcow2: Reference count table too large == Misaligned refcount table == Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 -can't open device TEST_DIR/t.qcow2: Reference count table offset invalid +qemu-io: can't open device TEST_DIR/t.qcow2: Reference count table offset invalid == Huge refcount offset == Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 -can't open device TEST_DIR/t.qcow2: Reference count table offset invalid +qemu-io: can't open device TEST_DIR/t.qcow2: Reference count table offset invalid == Invalid snapshot table == Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 -can't open device TEST_DIR/t.qcow2: Snapshot table too large -can't open device TEST_DIR/t.qcow2: Snapshot table too large -can't open device TEST_DIR/t.qcow2: Snapshot table offset invalid -can't open device TEST_DIR/t.qcow2: Snapshot table offset invalid +qemu-io: can't open device TEST_DIR/t.qcow2: Snapshot table too large +qemu-io: can't open device TEST_DIR/t.qcow2: Snapshot table too large +qemu-io: can't open device TEST_DIR/t.qcow2: Snapshot table offset invalid +qemu-io: can't open device TEST_DIR/t.qcow2: Snapshot table offset invalid == Hitting snapshot table size limit == Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 @@ -XXX,XX +XXX,XX @@ read 512/512 bytes at offset 0 == Invalid L1 table == Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 -can't open device TEST_DIR/t.qcow2: Active L1 table too large -can't open device TEST_DIR/t.qcow2: Active L1 table too large -can't open device TEST_DIR/t.qcow2: Active L1 table offset invalid -can't open device TEST_DIR/t.qcow2: Active L1 table offset invalid +qemu-io: can't open device TEST_DIR/t.qcow2: Active L1 table too large +qemu-io: can't open device TEST_DIR/t.qcow2: Active L1 table too large +qemu-io: can't open device TEST_DIR/t.qcow2: Active L1 table offset invalid +qemu-io: can't open device TEST_DIR/t.qcow2: Active L1 table offset invalid == Invalid L1 table (with internal snapshot in the image) == Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 @@ -XXX,XX +XXX,XX @@ qemu-img: Could not open 'TEST_DIR/t.IMGFMT': L1 table is too small == Invalid backing file size == Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 -can't open device TEST_DIR/t.qcow2: Backing file name too long +qemu-io: can't open device TEST_DIR/t.qcow2: Backing file name too long == Invalid L2 entry (huge physical offset) == Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 @@ -XXX,XX +XXX,XX @@ wrote 512/512 bytes at offset 0 qemu-img: Failed to load snapshot: Snapshot L1 table offset invalid qemu-img: Snapshot L1 table offset invalid qemu-img: Failed to turn zero into data clusters: Invalid argument -Failed to flush the refcount block cache: Invalid argument +qemu-io: Failed to flush the refcount block cache: Invalid argument write failed: Invalid argument qemu-img: Snapshot L1 table offset invalid qemu-img: Could not apply snapshot 'test': Failed to load snapshot: Invalid argument @@ -XXX,XX +XXX,XX @@ wrote 512/512 bytes at offset 0 qemu-img: Failed to load snapshot: Snapshot L1 table too large qemu-img: Snapshot L1 table too large qemu-img: Failed to turn zero into data clusters: File too large -Failed to flush the refcount block cache: File too large +qemu-io: Failed to flush the refcount block cache: File too large write failed: File too large qemu-img: Snapshot L1 table too large qemu-img: Could not apply snapshot 'test': Failed to load snapshot: File too large diff --git a/tests/qemu-iotests/081.out b/tests/qemu-iotests/081.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/081.out +++ b/tests/qemu-iotests/081.out @@ -XXX,XX +XXX,XX @@ read 10485760/10485760 bytes at offset 0 10 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) == checking the blkverify mode with invalid settings == -can't open: blkverify=on can only be set if there are exactly two files and vote-threshold is 2 +qemu-io: can't open: blkverify=on can only be set if there are exactly two files and vote-threshold is 2 == dynamically adding a child to a quorum == Testing: diff --git a/tests/qemu-iotests/088.out b/tests/qemu-iotests/088.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/088.out +++ b/tests/qemu-iotests/088.out @@ -XXX,XX +XXX,XX @@ QA output created by 088 == Invalid block size == Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 -can't open device TEST_DIR/t.vpc: Invalid block size 0 -can't open device TEST_DIR/t.vpc: Invalid block size 0 -can't open device TEST_DIR/t.vpc: Invalid block size 128 -can't open device TEST_DIR/t.vpc: Invalid block size 128 -can't open device TEST_DIR/t.vpc: Invalid block size 305419896 -can't open device TEST_DIR/t.vpc: Invalid block size 305419896 +qemu-io: can't open device TEST_DIR/t.vpc: Invalid block size 0 +qemu-io: can't open device TEST_DIR/t.vpc: Invalid block size 0 +qemu-io: can't open device TEST_DIR/t.vpc: Invalid block size 128 +qemu-io: can't open device TEST_DIR/t.vpc: Invalid block size 128 +qemu-io: can't open device TEST_DIR/t.vpc: Invalid block size 305419896 +qemu-io: can't open device TEST_DIR/t.vpc: Invalid block size 305419896 *** done diff --git a/tests/qemu-iotests/103.out b/tests/qemu-iotests/103.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/103.out +++ b/tests/qemu-iotests/103.out @@ -XXX,XX +XXX,XX @@ wrote 65536/65536 bytes at offset 0 === Testing invalid option combinations === -can't open device TEST_DIR/t.IMGFMT: cache-size, l2-cache-size and refcount-cache-size may not be set at the same time -can't open device TEST_DIR/t.IMGFMT: l2-cache-size may not exceed cache-size -can't open device TEST_DIR/t.IMGFMT: refcount-cache-size may not exceed cache-size -can't open device TEST_DIR/t.IMGFMT: cache-size, l2-cache-size and refcount-cache-size may not be set at the same time -can't open device TEST_DIR/t.IMGFMT: L2 cache entry size must be a power of two between 512 and the cluster size (65536) -can't open device TEST_DIR/t.IMGFMT: L2 cache entry size must be a power of two between 512 and the cluster size (65536) -can't open device TEST_DIR/t.IMGFMT: L2 cache entry size must be a power of two between 512 and the cluster size (65536) +qemu-io: can't open device TEST_DIR/t.IMGFMT: cache-size, l2-cache-size and refcount-cache-size may not be set at the same time +qemu-io: can't open device TEST_DIR/t.IMGFMT: l2-cache-size may not exceed cache-size +qemu-io: can't open device TEST_DIR/t.IMGFMT: refcount-cache-size may not exceed cache-size +qemu-io: can't open device TEST_DIR/t.IMGFMT: cache-size, l2-cache-size and refcount-cache-size may not be set at the same time +qemu-io: can't open device TEST_DIR/t.IMGFMT: L2 cache entry size must be a power of two between 512 and the cluster size (65536) +qemu-io: can't open device TEST_DIR/t.IMGFMT: L2 cache entry size must be a power of two between 512 and the cluster size (65536) +qemu-io: can't open device TEST_DIR/t.IMGFMT: L2 cache entry size must be a power of two between 512 and the cluster size (65536) === Testing valid option combinations === diff --git a/tests/qemu-iotests/114.out b/tests/qemu-iotests/114.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/114.out +++ b/tests/qemu-iotests/114.out @@ -XXX,XX +XXX,XX @@ virtual size: 64M (67108864 bytes) cluster_size: 65536 backing file: TEST_DIR/t.IMGFMT.base backing file format: foo -can't open device TEST_DIR/t.qcow2: Could not open backing file: Unknown driver 'foo' +qemu-io: can't open device TEST_DIR/t.qcow2: Could not open backing file: Unknown driver 'foo' no file open, try 'help open' read 4096/4096 bytes at offset 0 4 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) diff --git a/tests/qemu-iotests/116.out b/tests/qemu-iotests/116.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/116.out +++ b/tests/qemu-iotests/116.out @@ -XXX,XX +XXX,XX @@ QA output created by 116 == truncated header cluster == Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=134217728 -can't open device TEST_DIR/t.qed: Could not open 'TEST_DIR/t.qed': Invalid argument +qemu-io: can't open device TEST_DIR/t.qed: Could not open 'TEST_DIR/t.qed': Invalid argument == invalid header magic == Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=134217728 -can't open device TEST_DIR/t.qed: Image not in QED format +qemu-io: can't open device TEST_DIR/t.qed: Image not in QED format == invalid cluster size == Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=134217728 -can't open device TEST_DIR/t.qed: Could not open 'TEST_DIR/t.qed': Invalid argument +qemu-io: can't open device TEST_DIR/t.qed: Could not open 'TEST_DIR/t.qed': Invalid argument == invalid table size == Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=134217728 -can't open device TEST_DIR/t.qed: Could not open 'TEST_DIR/t.qed': Invalid argument +qemu-io: can't open device TEST_DIR/t.qed: Could not open 'TEST_DIR/t.qed': Invalid argument == invalid header size == Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=134217728 -can't open device TEST_DIR/t.qed: Could not open 'TEST_DIR/t.qed': Invalid argument +qemu-io: can't open device TEST_DIR/t.qed: Could not open 'TEST_DIR/t.qed': Invalid argument == invalid L1 table offset == Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=134217728 -can't open device TEST_DIR/t.qed: Could not open 'TEST_DIR/t.qed': Invalid argument +qemu-io: can't open device TEST_DIR/t.qed: Could not open 'TEST_DIR/t.qed': Invalid argument == invalid image size == Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=134217728 -can't open device TEST_DIR/t.qed: Could not open 'TEST_DIR/t.qed': Invalid argument +qemu-io: can't open device TEST_DIR/t.qed: Could not open 'TEST_DIR/t.qed': Invalid argument *** done diff --git a/tests/qemu-iotests/131.out b/tests/qemu-iotests/131.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/131.out +++ b/tests/qemu-iotests/131.out @@ -XXX,XX +XXX,XX @@ read 32768/32768 bytes at offset 163840 read 32768/32768 bytes at offset 0 32 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) == Corrupt image == -can't open device TEST_DIR/t.parallels: parallels: Image was not closed correctly; cannot be opened read/write +qemu-io: can't open device TEST_DIR/t.parallels: parallels: Image was not closed correctly; cannot be opened read/write ERROR image was not closed correctly 1 errors were found on the image. diff --git a/tests/qemu-iotests/133.out b/tests/qemu-iotests/133.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/133.out +++ b/tests/qemu-iotests/133.out @@ -XXX,XX +XXX,XX @@ Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 backing_file=TEST_DIR/t === Check that node-name can't be changed === -Cannot change the option 'node-name' -Cannot change the option 'node-name' -Cannot change the option 'node-name' +qemu-io: Cannot change the option 'node-name' +qemu-io: Cannot change the option 'node-name' +qemu-io: Cannot change the option 'node-name' === Check that unchanged node-name is okay === === Check that driver can't be changed === -Cannot change the option 'driver' -Cannot change the option 'driver' -Cannot change the option 'driver' +qemu-io: Cannot change the option 'driver' +qemu-io: Cannot change the option 'driver' +qemu-io: Cannot change the option 'driver' === Check that unchanged driver is okay === @@ -XXX,XX +XXX,XX @@ format name: null-co === Check that mixing -c/-r/-w and their corresponding options is forbidden === -Cannot set both -r/-w and 'read-only' -Cannot set both -r/-w and 'read-only' -Cannot set both -c and the cache options -Cannot set both -c and the cache options -Cannot set both -c and the cache options +qemu-io: Cannot set both -r/-w and 'read-only' +qemu-io: Cannot set both -r/-w and 'read-only' +qemu-io: Cannot set both -c and the cache options +qemu-io: Cannot set both -c and the cache options +qemu-io: Cannot set both -c and the cache options === Check that invalid options are handled correctly === -Parameter 'read-only' expects 'on' or 'off' -Parameter 'cache.no-flush' expects 'on' or 'off' -Parameter 'cache.direct' expects 'on' or 'off' -Parameter 'auto-read-only' expects 'on' or 'off' +qemu-io: Parameter 'read-only' expects 'on' or 'off' +qemu-io: Parameter 'cache.no-flush' expects 'on' or 'off' +qemu-io: Parameter 'cache.direct' expects 'on' or 'off' +qemu-io: Parameter 'auto-read-only' expects 'on' or 'off' *** done diff --git a/tests/qemu-iotests/137.out b/tests/qemu-iotests/137.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/137.out +++ b/tests/qemu-iotests/137.out @@ -XXX,XX +XXX,XX @@ read 33554432/33554432 bytes at offset 0 === Try setting some invalid values === -Parameter 'lazy-refcounts' expects 'on' or 'off' -cache-size, l2-cache-size and refcount-cache-size may not be set at the same time -l2-cache-size may not exceed cache-size -refcount-cache-size may not exceed cache-size -L2 cache entry size must be a power of two between 512 and the cluster size (65536) -L2 cache entry size must be a power of two between 512 and the cluster size (65536) -Refcount cache size too big -Conflicting values for qcow2 options 'overlap-check' ('constant') and 'overlap-check.template' ('all') -Unsupported value 'blubb' for qcow2 option 'overlap-check'. Allowed are any of the following: none, constant, cached, all -Unsupported value 'blubb' for qcow2 option 'overlap-check'. Allowed are any of the following: none, constant, cached, all -Cache clean interval too big +qemu-io: Parameter 'lazy-refcounts' expects 'on' or 'off' +qemu-io: cache-size, l2-cache-size and refcount-cache-size may not be set at the same time +qemu-io: l2-cache-size may not exceed cache-size +qemu-io: refcount-cache-size may not exceed cache-size +qemu-io: L2 cache entry size must be a power of two between 512 and the cluster size (65536) +qemu-io: L2 cache entry size must be a power of two between 512 and the cluster size (65536) +qemu-io: Refcount cache size too big +qemu-io: Conflicting values for qcow2 options 'overlap-check' ('constant') and 'overlap-check.template' ('all') +qemu-io: Unsupported value 'blubb' for qcow2 option 'overlap-check'. Allowed are any of the following: none, constant, cached, all +qemu-io: Unsupported value 'blubb' for qcow2 option 'overlap-check'. Allowed are any of the following: none, constant, cached, all +qemu-io: Cache clean interval too big Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=36028797018963968 -L2 cache size too big +qemu-io: L2 cache size too big Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 === Test transaction semantics === -Unsupported value 'blubb' for qcow2 option 'overlap-check'. Allowed are any of the following: none, constant, cached, all +qemu-io: Unsupported value 'blubb' for qcow2 option 'overlap-check'. Allowed are any of the following: none, constant, cached, all wrote 512/512 bytes at offset 0 512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) ./common.rc: Killed ( if [ "${VALGRIND_QEMU}" == "y" ]; then @@ -XXX,XX +XXX,XX @@ incompatible_features 0x0 Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 wrote 65536/65536 bytes at offset 0 64 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -Parameter 'lazy-refcounts' expects 'on' or 'off' +qemu-io: Parameter 'lazy-refcounts' expects 'on' or 'off' qcow2: Marking image as corrupt: Preventing invalid write on metadata (overlaps with qcow2_header); further corruption events will be suppressed write failed: Input/output error *** done diff --git a/tests/qemu-iotests/140.out b/tests/qemu-iotests/140.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/140.out +++ b/tests/qemu-iotests/140.out @@ -XXX,XX +XXX,XX @@ wrote 65536/65536 bytes at offset 0 read 65536/65536 bytes at offset 0 64 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) {"return": {}} -can't open device nbd+unix:///drv?socket=TEST_DIR/nbd: Requested export not available +qemu-io: can't open device nbd+unix:///drv?socket=TEST_DIR/nbd: Requested export not available server reported: export 'drv' not present {"return": {}} {"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "SHUTDOWN", "data": {"guest": false, "reason": "host-qmp-quit"}} diff --git a/tests/qemu-iotests/143.out b/tests/qemu-iotests/143.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/143.out +++ b/tests/qemu-iotests/143.out @@ -XXX,XX +XXX,XX @@ QA output created by 143 {"return": {}} {"return": {}} -can't open device nbd+unix:///no_such_export?socket=TEST_DIR/nbd: Requested export not available +qemu-io: can't open device nbd+unix:///no_such_export?socket=TEST_DIR/nbd: Requested export not available server reported: export 'no_such_export' not present {"return": {}} {"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "SHUTDOWN", "data": {"guest": false, "reason": "host-qmp-quit"}} diff --git a/tests/qemu-iotests/153.out b/tests/qemu-iotests/153.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/153.out +++ b/tests/qemu-iotests/153.out @@ -XXX,XX +XXX,XX @@ Is another process using the image [TEST_DIR/t.qcow2]? == Running utility commands == _qemu_io_wrapper -c read 0 512 TEST_DIR/t.qcow2 -can't open device TEST_DIR/t.qcow2: Failed to get "write" lock +qemu-io: can't open device TEST_DIR/t.qcow2: Failed to get "write" lock Is another process using the image [TEST_DIR/t.qcow2]? _qemu_io_wrapper -r -c read 0 512 TEST_DIR/t.qcow2 -can't open device TEST_DIR/t.qcow2: Failed to get shared "write" lock +qemu-io: can't open device TEST_DIR/t.qcow2: Failed to get shared "write" lock Is another process using the image [TEST_DIR/t.qcow2]? _qemu_io_wrapper -c open TEST_DIR/t.qcow2 -c read 0 512 -can't open device TEST_DIR/t.qcow2: Failed to get "write" lock +qemu-io: can't open device TEST_DIR/t.qcow2: Failed to get "write" lock Is another process using the image [TEST_DIR/t.qcow2]? no file open, try 'help open' _qemu_io_wrapper -c open -r TEST_DIR/t.qcow2 -c read 0 512 -can't open device TEST_DIR/t.qcow2: Failed to get shared "write" lock +qemu-io: can't open device TEST_DIR/t.qcow2: Failed to get shared "write" lock Is another process using the image [TEST_DIR/t.qcow2]? no file open, try 'help open' @@ -XXX,XX +XXX,XX @@ file format: IMGFMT == Running utility commands -U == _qemu_io_wrapper -U -c read 0 512 TEST_DIR/t.qcow2 -can't open device TEST_DIR/t.qcow2: force-share=on can only be used with read-only images +qemu-io: can't open device TEST_DIR/t.qcow2: force-share=on can only be used with read-only images _qemu_io_wrapper -U -r -c read 0 512 TEST_DIR/t.qcow2 _qemu_io_wrapper -c open -U TEST_DIR/t.qcow2 -c read 0 512 -can't open device TEST_DIR/t.qcow2: force-share=on can only be used with read-only images +qemu-io: can't open device TEST_DIR/t.qcow2: force-share=on can only be used with read-only images no file open, try 'help open' _qemu_io_wrapper -c open -r -U TEST_DIR/t.qcow2 -c read 0 512 @@ -XXX,XX +XXX,XX @@ Is another process using the image [TEST_DIR/t.qcow2]? == Running utility commands == _qemu_io_wrapper -c read 0 512 TEST_DIR/t.qcow2 -can't open device TEST_DIR/t.qcow2: Failed to get "write" lock +qemu-io: can't open device TEST_DIR/t.qcow2: Failed to get "write" lock Is another process using the image [TEST_DIR/t.qcow2]? _qemu_io_wrapper -r -c read 0 512 TEST_DIR/t.qcow2 _qemu_io_wrapper -c open TEST_DIR/t.qcow2 -c read 0 512 -can't open device TEST_DIR/t.qcow2: Failed to get "write" lock +qemu-io: can't open device TEST_DIR/t.qcow2: Failed to get "write" lock Is another process using the image [TEST_DIR/t.qcow2]? no file open, try 'help open' @@ -XXX,XX +XXX,XX @@ file format: IMGFMT == Running utility commands -U == _qemu_io_wrapper -U -c read 0 512 TEST_DIR/t.qcow2 -can't open device TEST_DIR/t.qcow2: force-share=on can only be used with read-only images +qemu-io: can't open device TEST_DIR/t.qcow2: force-share=on can only be used with read-only images _qemu_io_wrapper -U -r -c read 0 512 TEST_DIR/t.qcow2 _qemu_io_wrapper -c open -U TEST_DIR/t.qcow2 -c read 0 512 -can't open device TEST_DIR/t.qcow2: force-share=on can only be used with read-only images +qemu-io: can't open device TEST_DIR/t.qcow2: force-share=on can only be used with read-only images no file open, try 'help open' _qemu_io_wrapper -c open -r -U TEST_DIR/t.qcow2 -c read 0 512 @@ -XXX,XX +XXX,XX @@ file format: IMGFMT == Running utility commands -U == _qemu_io_wrapper -U -c read 0 512 TEST_DIR/t.qcow2 -can't open device TEST_DIR/t.qcow2: force-share=on can only be used with read-only images +qemu-io: can't open device TEST_DIR/t.qcow2: force-share=on can only be used with read-only images _qemu_io_wrapper -U -r -c read 0 512 TEST_DIR/t.qcow2 _qemu_io_wrapper -c open -U TEST_DIR/t.qcow2 -c read 0 512 -can't open device TEST_DIR/t.qcow2: force-share=on can only be used with read-only images +qemu-io: can't open device TEST_DIR/t.qcow2: force-share=on can only be used with read-only images no file open, try 'help open' _qemu_io_wrapper -c open -r -U TEST_DIR/t.qcow2 -c read 0 512 @@ -XXX,XX +XXX,XX @@ Adding drive {"return": "OKrn"} _qemu_io_wrapper TEST_DIR/t.qcow2 -c write 0 512 -can't open device TEST_DIR/t.qcow2: Failed to get "write" lock +qemu-io: can't open device TEST_DIR/t.qcow2: Failed to get "write" lock Is another process using the image [TEST_DIR/t.qcow2]? Creating overlay with qemu-img when the guest is running should be allowed @@ -XXX,XX +XXX,XX @@ _qemu_img_wrapper info TEST_DIR/t.qcow2 {"return": ""} _qemu_io_wrapper TEST_DIR/t.qcow2 -c write 0 512 -can't open device TEST_DIR/t.qcow2: Failed to get "write" lock +qemu-io: can't open device TEST_DIR/t.qcow2: Failed to get "write" lock Is another process using the image [TEST_DIR/t.qcow2]? Closing the other {"return": ""} @@ -XXX,XX +XXX,XX @@ qemu-img: --force-share/-U conflicts with image options No conflict: Conflict: --U conflicts with image options +qemu-io: -U conflicts with image options *** done diff --git a/tests/qemu-iotests/187.out b/tests/qemu-iotests/187.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/187.out +++ b/tests/qemu-iotests/187.out @@ -XXX,XX +XXX,XX @@ Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 Start from read-only -Block node is read-only +qemu-io: Block node is read-only wrote 65536/65536 bytes at offset 0 64 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -Block node is read-only +qemu-io: Block node is read-only Start from read-write wrote 65536/65536 bytes at offset 0 64 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -Block node is read-only +qemu-io: Block node is read-only wrote 65536/65536 bytes at offset 0 64 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) *** done diff --git a/tests/qemu-iotests/188.out b/tests/qemu-iotests/188.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/188.out +++ b/tests/qemu-iotests/188.out @@ -XXX,XX +XXX,XX @@ read 16777216/16777216 bytes at offset 0 16 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) == verify open failure with wrong password == -can't open: Invalid password, cannot unlock any keyslot +qemu-io: can't open: Invalid password, cannot unlock any keyslot *** done diff --git a/tests/qemu-iotests/197.out b/tests/qemu-iotests/197.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/197.out +++ b/tests/qemu-iotests/197.out @@ -XXX,XX +XXX,XX @@ read 2147483136/2147483136 bytes at offset 1024 2 GiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) read 1024/1024 bytes at offset 3221226496 1 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -can't open device TEST_DIR/t.wrap.qcow2: Can't use copy-on-read on read-only device +qemu-io: can't open device TEST_DIR/t.wrap.qcow2: Can't use copy-on-read on read-only device 2 GiB (0x80010000) bytes allocated at offset 0 bytes (0x0) 1023.938 MiB (0x3fff0000) bytes not allocated at offset 2 GiB (0x80010000) 64 KiB (0x10000) bytes allocated at offset 3 GiB (0xc0000000) diff --git a/tests/qemu-iotests/205 b/tests/qemu-iotests/205 index XXXXXXX..XXXXXXX 100755 --- a/tests/qemu-iotests/205 +++ b/tests/qemu-iotests/205 @@ -XXX,XX +XXX,XX @@ class TestNbdServerRemove(iotests.QMPTestCase): def assertConnectFailed(self, qemu_io_output): self.assertEqual(filter_qemu_io(qemu_io_output).strip(), - "can't open device " + nbd_uri + + "qemu-io: can't open device " + nbd_uri + ": Requested export not available\n" "server reported: export 'exp' not present") diff --git a/tests/qemu-iotests/215.out b/tests/qemu-iotests/215.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/215.out +++ b/tests/qemu-iotests/215.out @@ -XXX,XX +XXX,XX @@ read 2147483136/2147483136 bytes at offset 1024 2 GiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) read 1024/1024 bytes at offset 3221226496 1 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -can't open device TEST_DIR/t.wrap.qcow2: Block node is read-only +qemu-io: can't open device TEST_DIR/t.wrap.qcow2: Block node is read-only 2 GiB (0x80010000) bytes allocated at offset 0 bytes (0x0) 1023.938 MiB (0x3fff0000) bytes not allocated at offset 2 GiB (0x80010000) 64 KiB (0x10000) bytes allocated at offset 3 GiB (0xc0000000) diff --git a/tests/qemu-iotests/226.out b/tests/qemu-iotests/226.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/226.out +++ b/tests/qemu-iotests/226.out @@ -XXX,XX +XXX,XX @@ QA output created by 226 === Testing with driver:file === == Testing RO == -can't open: A regular file was expected by the 'file' driver, but something else was given -warning: Opening a character device as a file using the 'file' driver is deprecated +qemu-io: can't open: A regular file was expected by the 'file' driver, but something else was given +qemu-io: warning: Opening a character device as a file using the 'file' driver is deprecated == Testing RW == -can't open: Could not open 'TEST_DIR/t.IMGFMT': Is a directory -warning: Opening a character device as a file using the 'file' driver is deprecated +qemu-io: can't open: Could not open 'TEST_DIR/t.IMGFMT': Is a directory +qemu-io: warning: Opening a character device as a file using the 'file' driver is deprecated === Testing with driver:host_device === == Testing RO == -can't open: 'host_device' driver expects either a character or block device +qemu-io: can't open: 'host_device' driver expects either a character or block device == Testing RW == -can't open: Could not open 'TEST_DIR/t.IMGFMT': Is a directory +qemu-io: can't open: Could not open 'TEST_DIR/t.IMGFMT': Is a directory === Testing with driver:host_cdrom === == Testing RO == -can't open: 'host_cdrom' driver expects either a character or block device +qemu-io: can't open: 'host_cdrom' driver expects either a character or block device == Testing RW == -can't open: Could not open 'TEST_DIR/t.IMGFMT': Is a directory +qemu-io: can't open: Could not open 'TEST_DIR/t.IMGFMT': Is a directory *** done diff --git a/tests/qemu-iotests/244.out b/tests/qemu-iotests/244.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/244.out +++ b/tests/qemu-iotests/244.out @@ -XXX,XX +XXX,XX @@ read 65536/65536 bytes at offset 0 64 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) read 65536/65536 bytes at offset 0 64 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -can't open device TEST_DIR/t.qcow2: Could not open 'inexistent': No such file or directory +qemu-io: can't open device TEST_DIR/t.qcow2: Could not open 'inexistent': No such file or directory no file open, try 'help open' Data file required, but without data file name in the image: -can't open device TEST_DIR/t.qcow2: 'data-file' is required for this image +qemu-io: can't open device TEST_DIR/t.qcow2: 'data-file' is required for this image no file open, try 'help open' read 65536/65536 bytes at offset 0 64 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) -can't open device TEST_DIR/t.qcow2: Could not open 'inexistent': No such file or directory +qemu-io: can't open device TEST_DIR/t.qcow2: Could not open 'inexistent': No such file or directory no file open, try 'help open' Setting data-file for an image with internal data: Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 -can't open device TEST_DIR/t.qcow2: 'data-file' can only be set for images with an external data file +qemu-io: can't open device TEST_DIR/t.qcow2: 'data-file' can only be set for images with an external data file no file open, try 'help open' -can't open device TEST_DIR/t.qcow2: Could not open 'inexistent': No such file or directory +qemu-io: can't open device TEST_DIR/t.qcow2: Could not open 'inexistent': No such file or directory no file open, try 'help open' === Conflicting features === -- 2.20.1 Even for block nodes with bs->drv == NULL, we can't just ignore a bdrv_set_aio_context() call. Leaving the node in its old context can mean that it's still in an iothread context in bdrv_close_all() during shutdown, resulting in an attempted unlock of the AioContext lock which we don't hold. This is an example stack trace of a related crash: #0 0x00007ffff59da57f in raise () at /lib64/libc.so.6 #1 0x00007ffff59c4895 in abort () at /lib64/libc.so.6 #2 0x0000555555b97b1e in error_exit (err=<optimized out>, msg=msg@entry=0x555555d386d0 <__func__.19059> "qemu_mutex_unlock_impl") at util/qemu-thread-posix.c:36 #3 0x0000555555b97f7f in qemu_mutex_unlock_impl (mutex=mutex@entry=0x5555568002f0, file=file@entry=0x555555d378df "util/async.c", line=line@entry=507) at util/qemu-thread-posix.c:97 #4 0x0000555555b92f55 in aio_context_release (ctx=ctx@entry=0x555556800290) at util/async.c:507 #5 0x0000555555b05cf8 in bdrv_prwv_co (child=child@entry=0x7fffc80012f0, offset=offset@entry=131072, qiov=qiov@entry=0x7fffffffd4f0, is_write=is_write@entry=true, flags=flags@entry=0) at block/io.c:833 #6 0x0000555555b060a9 in bdrv_pwritev (qiov=0x7fffffffd4f0, offset=131072, child=0x7fffc80012f0) at block/io.c:990 #7 0x0000555555b060a9 in bdrv_pwrite (child=0x7fffc80012f0, offset=131072, buf=<optimized out>, bytes=<optimized out>) at block/io.c:990 #8 0x0000555555ae172b in qcow2_cache_entry_flush (bs=bs@entry=0x555556810680, c=c@entry=0x5555568cc740, i=i@entry=0) at block/qcow2-cache.c:51 #9 0x0000555555ae18dd in qcow2_cache_write (bs=bs@entry=0x555556810680, c=0x5555568cc740) at block/qcow2-cache.c:248 #10 0x0000555555ae15de in qcow2_cache_flush (bs=0x555556810680, c=<optimized out>) at block/qcow2-cache.c:259 #11 0x0000555555ae16b1 in qcow2_cache_flush_dependency (c=0x5555568a1700, c=0x5555568a1700, bs=0x555556810680) at block/qcow2-cache.c:194 #12 0x0000555555ae16b1 in qcow2_cache_entry_flush (bs=bs@entry=0x555556810680, c=c@entry=0x5555568a1700, i=i@entry=0) at block/qcow2-cache.c:194 #13 0x0000555555ae18dd in qcow2_cache_write (bs=bs@entry=0x555556810680, c=0x5555568a1700) at block/qcow2-cache.c:248 #14 0x0000555555ae15de in qcow2_cache_flush (bs=bs@entry=0x555556810680, c=<optimized out>) at block/qcow2-cache.c:259 #15 0x0000555555ad242c in qcow2_inactivate (bs=bs@entry=0x555556810680) at block/qcow2.c:2124 #16 0x0000555555ad2590 in qcow2_close (bs=0x555556810680) at block/qcow2.c:2153 #17 0x0000555555ab0c62 in bdrv_close (bs=0x555556810680) at block.c:3358 #18 0x0000555555ab0c62 in bdrv_delete (bs=0x555556810680) at block.c:3542 #19 0x0000555555ab0c62 in bdrv_unref (bs=0x555556810680) at block.c:4598 #20 0x0000555555af4d72 in blk_remove_bs (blk=blk@entry=0x5555568103d0) at block/block-backend.c:785 #21 0x0000555555af4dbb in blk_remove_all_bs () at block/block-backend.c:483 #22 0x0000555555aae02f in bdrv_close_all () at block.c:3412 #23 0x00005555557f9796 in main (argc=<optimized out>, argv=<optimized out>, envp=<optimized out>) at vl.c:4776 The reproducer I used is a qcow2 image on gluster volume, where the virtual disk size (4 GB) is larger than the gluster volume size (64M), so we can easily trigger an ENOSPC. This backend is assigned to a virtio-blk device using an iothread, and then from the guest a 'dd if=/dev/zero of=/dev/vda bs=1G count=1' causes the VM to stop because of an I/O error. qemu_gluster_co_flush_to_disk() sets bs->drv = NULL on error, so when virtio-blk stops the dataplane, the block nodes stay in the iothread AioContext. A 'quit' monitor command issued from this paused state crashes the process. Fixes: https://bugzilla.redhat.com/show_bug.cgi?id=1631227 Cc: qemu-stable@nongnu.org Signed-off-by: Kevin Wolf <kwolf@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> Reviewed-by: Max Reitz <mreitz@redhat.com> Reviewed-by: Stefano Garzarella <sgarzare@redhat.com> --- block.c | 12 ++---------- 1 file changed, 2 insertions(+), 10 deletions(-) diff --git a/block.c b/block.c index XXXXXXX..XXXXXXX 100644 --- a/block.c +++ b/block.c @@ -XXX,XX +XXX,XX @@ void bdrv_detach_aio_context(BlockDriverState *bs) BdrvAioNotifier *baf, *baf_tmp; BdrvChild *child; - if (!bs->drv) { - return; - } - assert(!bs->walking_aio_notifiers); bs->walking_aio_notifiers = true; QLIST_FOREACH_SAFE(baf, &bs->aio_notifiers, list, baf_tmp) { @@ -XXX,XX +XXX,XX @@ void bdrv_detach_aio_context(BlockDriverState *bs) */ bs->walking_aio_notifiers = false; - if (bs->drv->bdrv_detach_aio_context) { + if (bs->drv && bs->drv->bdrv_detach_aio_context) { bs->drv->bdrv_detach_aio_context(bs); } QLIST_FOREACH(child, &bs->children, next) { @@ -XXX,XX +XXX,XX @@ void bdrv_attach_aio_context(BlockDriverState *bs, BdrvAioNotifier *ban, *ban_tmp; BdrvChild *child; - if (!bs->drv) { - return; - } - if (bs->quiesce_counter) { aio_disable_external(new_context); } @@ -XXX,XX +XXX,XX @@ void bdrv_attach_aio_context(BlockDriverState *bs, QLIST_FOREACH(child, &bs->children, next) { bdrv_attach_aio_context(child->bs, new_context); } - if (bs->drv->bdrv_attach_aio_context) { + if (bs->drv && bs->drv->bdrv_attach_aio_context) { bs->drv->bdrv_attach_aio_context(bs, new_context); } -- 2.20.1 From: Zhengui li <lizhengui@huawei.com> Concurrent IO becomes serial IO because of the qemu Coroutine lock, which reduce IO performance severely. So unlock Coroutine lock before bdrv_co_pwritev and bdrv_co_preadv to fix it. Signed-off-by: Zhengui li <lizhengui@huawei.com> Reviewed-by: Paolo Bonzini <pbonzini@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block/vpc.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/block/vpc.c b/block/vpc.c index XXXXXXX..XXXXXXX 100644 --- a/block/vpc.c +++ b/block/vpc.c @@ -XXX,XX +XXX,XX @@ vpc_co_preadv(BlockDriverState *bs, uint64_t offset, uint64_t bytes, qemu_iovec_reset(&local_qiov); qemu_iovec_concat(&local_qiov, qiov, bytes_done, n_bytes); + qemu_co_mutex_unlock(&s->lock); ret = bdrv_co_preadv(bs->file, image_offset, n_bytes, &local_qiov, 0); + qemu_co_mutex_lock(&s->lock); if (ret < 0) { goto fail; } @@ -XXX,XX +XXX,XX @@ vpc_co_pwritev(BlockDriverState *bs, uint64_t offset, uint64_t bytes, qemu_iovec_reset(&local_qiov); qemu_iovec_concat(&local_qiov, qiov, bytes_done, n_bytes); + qemu_co_mutex_unlock(&s->lock); ret = bdrv_co_pwritev(bs->file, image_offset, n_bytes, &local_qiov, 0); + qemu_co_mutex_lock(&s->lock); if (ret < 0) { goto fail; } -- 2.20.1 From: Sam Eiderman <shmuel.eiderman@oracle.com> Commit b69864e5a ("vmdk: Support version=3 in VMDK descriptor files") fixed the probe function to correctly guess vmdk descriptors with version=3. This solves the issue where vmdk snapshot with parent vmdk descriptor containing "version=3" would be treated as raw instead vmdk. In the future case where a new vmdk version is introduced, we will again experience this issue, even if the user will provide "-f vmdk" it will only apply to the tip image and not to the underlying "misprobed" parent image. The code in vmdk.c already assumes that the backing file of vmdk must be vmdk (see vmdk_is_cid_valid which returns 0 if backing file is not vmdk). So let's make it official by supplying the backing_format as vmdk. Reviewed-by: Mark Kanda <mark.kanda@oracle.com> Reviewed-By: Liran Alon <liran.alon@oracle.com> Reviewed-by: Arbel Moshe <arbel.moshe@oracle.com> Signed-off-by: Shmuel Eiderman <shmuel.eiderman@oracle.com> Reviewed-by: Eric Blake <eblake@redhat.com> Reviewed-by: Fam Zheng <fam@euphon.net> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block/vmdk.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/block/vmdk.c b/block/vmdk.c index XXXXXXX..XXXXXXX 100644 --- a/block/vmdk.c +++ b/block/vmdk.c @@ -XXX,XX +XXX,XX @@ static int vmdk_parent_open(BlockDriverState *bs) pstrcpy(bs->auto_backing_file, end_name - p_name + 1, p_name); pstrcpy(bs->backing_file, sizeof(bs->backing_file), bs->auto_backing_file); + pstrcpy(bs->backing_format, sizeof(bs->backing_format), + "vmdk"); } out: -- 2.20.1 From: Stefano Garzarella <sgarzare@redhat.com> IEC binary prefixes are already defined in "qemu/units.h", so we can remove redundant definitions in "block/vhdx.h". Signed-off-by: Stefano Garzarella <sgarzare@redhat.com> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: John Snow <jsnow@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block/vhdx.h | 6 +----- block/vhdx.c | 3 ++- 2 files changed, 3 insertions(+), 6 deletions(-) diff --git a/block/vhdx.h b/block/vhdx.h index XXXXXXX..XXXXXXX 100644 --- a/block/vhdx.h +++ b/block/vhdx.h @@ -XXX,XX +XXX,XX @@ #ifndef BLOCK_VHDX_H #define BLOCK_VHDX_H - -#define KiB (1 * 1024) -#define MiB (KiB * 1024) -#define GiB (MiB * 1024) -#define TiB ((uint64_t) GiB * 1024) +#include "qemu/units.h" #define DEFAULT_LOG_SIZE 1048576 /* 1MiB */ /* Structures and fields present in the VHDX file */ diff --git a/block/vhdx.c b/block/vhdx.c index XXXXXXX..XXXXXXX 100644 --- a/block/vhdx.c +++ b/block/vhdx.c @@ -XXX,XX +XXX,XX @@ static int coroutine_fn vhdx_co_create(BlockdevCreateOptions *opts, return -EINVAL; } if (block_size > VHDX_BLOCK_SIZE_MAX) { - error_setg(errp, "Block size must not exceed %d", VHDX_BLOCK_SIZE_MAX); + error_setg(errp, "Block size must not exceed %" PRId64, + VHDX_BLOCK_SIZE_MAX); return -EINVAL; } -- 2.20.1 From: Stefano Garzarella <sgarzare@redhat.com> Using IEC binary prefixes in order to make the code more readable, with the exception of DEFAULT_LOG_SIZE because it's passed to stringify(). Signed-off-by: Stefano Garzarella <sgarzare@redhat.com> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: John Snow <jsnow@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block/vhdx.h | 10 ++++++---- block/vhdx-log.c | 2 +- block/vhdx.c | 4 ++-- 3 files changed, 9 insertions(+), 7 deletions(-) diff --git a/block/vhdx.h b/block/vhdx.h index XXXXXXX..XXXXXXX 100644 --- a/block/vhdx.h +++ b/block/vhdx.h @@ -XXX,XX +XXX,XX @@ #include "qemu/units.h" #define DEFAULT_LOG_SIZE 1048576 /* 1MiB */ +/* Note: can't use 1 * MiB, because it's passed to stringify() */ + /* Structures and fields present in the VHDX file */ /* The header section has the following blocks, @@ -XXX,XX +XXX,XX @@ * 0.........64KB...........128KB........192KB..........256KB................1MB */ -#define VHDX_HEADER_BLOCK_SIZE (64 * 1024) +#define VHDX_HEADER_BLOCK_SIZE (64 * KiB) #define VHDX_FILE_ID_OFFSET 0 #define VHDX_HEADER1_OFFSET (VHDX_HEADER_BLOCK_SIZE * 1) @@ -XXX,XX +XXX,XX @@ typedef struct QEMU_PACKED MSGUID { #define guid_eq(a, b) \ (memcmp(&(a), &(b), sizeof(MSGUID)) == 0) -#define VHDX_HEADER_SIZE (4 * 1024) /* although the vhdx_header struct in disk +#define VHDX_HEADER_SIZE (4 * KiB) /* although the vhdx_header struct in disk is only 582 bytes, for purposes of crc the header is the first 4KB of the 64KB block */ @@ -XXX,XX +XXX,XX @@ typedef struct QEMU_PACKED VHDXRegionTableEntry { /* ---- LOG ENTRY STRUCTURES ---- */ -#define VHDX_LOG_MIN_SIZE (1024 * 1024) -#define VHDX_LOG_SECTOR_SIZE 4096 +#define VHDX_LOG_MIN_SIZE (1 * MiB) +#define VHDX_LOG_SECTOR_SIZE (4 * KiB) #define VHDX_LOG_HDR_SIZE 64 #define VHDX_LOG_SIGNATURE 0x65676f6c typedef struct QEMU_PACKED VHDXLogEntryHeader { diff --git a/block/vhdx-log.c b/block/vhdx-log.c index XXXXXXX..XXXXXXX 100644 --- a/block/vhdx-log.c +++ b/block/vhdx-log.c @@ -XXX,XX +XXX,XX @@ static int vhdx_log_flush(BlockDriverState *bs, BDRVVHDXState *s, } if (file_length < desc_entries->hdr.last_file_offset) { new_file_size = desc_entries->hdr.last_file_offset; - if (new_file_size % (1024*1024)) { + if (new_file_size % (1 * MiB)) { /* round up to nearest 1MB boundary */ new_file_size = QEMU_ALIGN_UP(new_file_size, MiB); if (new_file_size > INT64_MAX) { diff --git a/block/vhdx.c b/block/vhdx.c index XXXXXXX..XXXXXXX 100644 --- a/block/vhdx.c +++ b/block/vhdx.c @@ -XXX,XX +XXX,XX @@ static int vhdx_allocate_block(BlockDriverState *bs, BDRVVHDXState *s, *new_offset = current_len; /* per the spec, the address for a block is in units of 1MB */ - *new_offset = ROUND_UP(*new_offset, 1024 * 1024); + *new_offset = ROUND_UP(*new_offset, 1 * MiB); if (*new_offset > INT64_MAX) { return -EINVAL; } @@ -XXX,XX +XXX,XX @@ static coroutine_fn int vhdx_co_writev(BlockDriverState *bs, int64_t sector_num, case PAYLOAD_BLOCK_FULLY_PRESENT: /* if the file offset address is in the header zone, * there is a problem */ - if (sinfo.file_offset < (1024 * 1024)) { + if (sinfo.file_offset < (1 * MiB)) { ret = -EFAULT; goto error_bat_restore; } -- 2.20.1 From: Eric Blake <eblake@redhat.com> When extracting a human-readable size formatter, we changed 'uint64_t div' pre-patch to 'unsigned long div' post-patch. Which breaks on 32-bit platforms, resulting in 'inf' instead of intended values larger than 999GB. Fixes: 22951aaa CC: qemu-stable@nongnu.org Reported-by: Max Reitz <mreitz@redhat.com> Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Max Reitz <mreitz@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- util/cutils.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/util/cutils.c b/util/cutils.c index XXXXXXX..XXXXXXX 100644 --- a/util/cutils.c +++ b/util/cutils.c @@ -XXX,XX +XXX,XX @@ const char *qemu_ether_ntoa(const MACAddr *mac) char *size_to_str(uint64_t val) { static const char *suffixes[] = { "", "Ki", "Mi", "Gi", "Ti", "Pi", "Ei" }; - unsigned long div; + uint64_t div; int i; /* -- 2.20.1 From: Eric Blake <eblake@redhat.com> Disk sizes close to INT64_MAX cause overflow, for some pretty ridiculous output: $ ./nbdkit -U - memory size=$((2**63 - 512)) --run 'qemu-img info $nbd' image: nbd+unix://?socket=/tmp/nbdkitHSAzNz/socket file format: raw virtual size: -8388607T (9223372036854775296 bytes) disk size: unavailable But there's no reason to have two separate implementations of integer to human-readable abbreviation, where one has overflow and stops at 'T', while the other avoids overflow and goes all the way to 'E'. With this patch, the output now claims 8EiB instead of -8388607T, which really is the correct rounding of largest file size supported by qemu (we could go 511 bytes larger if we used byte-accurate sizing instead of rounding up to the next sector boundary, but that wouldn't change the human-readable result). Quite a few iotests need updates to expected output to match. Reported-by: Richard W.M. Jones <rjones@redhat.com> Signed-off-by: Eric Blake <eblake@redhat.com> Tested-by: Richard W.M. Jones <rjones@redhat.com> Reviewed-by: Alberto Garcia <berto@igalia.com> Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> Tested-by: Max Reitz <mreitz@redhat.com> --- block/qapi.c | 49 +++++++++----------------------------- tests/qemu-iotests/043.out | 6 ++--- tests/qemu-iotests/053.out | 2 +- tests/qemu-iotests/059.out | 10 ++++---- tests/qemu-iotests/060.out | 10 ++++---- tests/qemu-iotests/061.out | 12 +++++----- tests/qemu-iotests/070.out | 2 +- tests/qemu-iotests/082.out | 26 ++++++++++---------- tests/qemu-iotests/084.out | 8 +++---- tests/qemu-iotests/089.out | 2 +- tests/qemu-iotests/095.out | 4 ++-- tests/qemu-iotests/104.out | 6 ++--- tests/qemu-iotests/110.out | 6 ++--- tests/qemu-iotests/114.out | 2 +- tests/qemu-iotests/126.out | 4 ++-- tests/qemu-iotests/130.out | 10 ++++---- tests/qemu-iotests/153.out | 2 +- tests/qemu-iotests/191.out | 8 +++---- tests/qemu-iotests/195.out | 4 ++-- tests/qemu-iotests/198.out | 4 ++-- tests/qemu-iotests/206.out | 10 ++++---- tests/qemu-iotests/207.out | 12 +++++----- tests/qemu-iotests/210.out | 8 +++---- tests/qemu-iotests/211.out | 10 ++++---- tests/qemu-iotests/212.out | 10 ++++---- tests/qemu-iotests/213.out | 10 ++++---- tests/qemu-iotests/233.out | 4 ++-- tests/qemu-iotests/237.out | 22 ++++++++--------- tests/qemu-iotests/242.out | 10 ++++---- 29 files changed, 123 insertions(+), 150 deletions(-) diff --git a/block/qapi.c b/block/qapi.c index XXXXXXX..XXXXXXX 100644 --- a/block/qapi.c +++ b/block/qapi.c @@ -XXX,XX +XXX,XX @@ BlockStatsList *qmp_query_blockstats(bool has_query_nodes, return head; } -#define NB_SUFFIXES 4 - -static char *get_human_readable_size(char *buf, int buf_size, int64_t size) -{ - static const char suffixes[NB_SUFFIXES] = {'K', 'M', 'G', 'T'}; - int64_t base; - int i; - - if (size <= 999) { - snprintf(buf, buf_size, "%" PRId64, size); - } else { - base = 1024; - for (i = 0; i < NB_SUFFIXES; i++) { - if (size < (10 * base)) { - snprintf(buf, buf_size, "%0.1f%c", - (double)size / base, - suffixes[i]); - break; - } else if (size < (1000 * base) || i == (NB_SUFFIXES - 1)) { - snprintf(buf, buf_size, "%" PRId64 "%c", - ((size + (base >> 1)) / base), - suffixes[i]); - break; - } - base = base * 1024; - } - } - return buf; -} - void bdrv_snapshot_dump(QEMUSnapshotInfo *sn) { - char buf1[128], date_buf[128], clock_buf[128]; + char date_buf[128], clock_buf[128]; struct tm tm; time_t ti; int64_t secs; + char *sizing = NULL; if (!sn) { qemu_printf("%-10s%-20s%7s%20s%15s", @@ -XXX,XX +XXX,XX @@ void bdrv_snapshot_dump(QEMUSnapshotInfo *sn) (int)((secs / 60) % 60), (int)(secs % 60), (int)((sn->vm_clock_nsec / 1000000) % 1000)); + sizing = size_to_str(sn->vm_state_size); qemu_printf("%-10s%-20s%7s%20s%15s", sn->id_str, sn->name, - get_human_readable_size(buf1, sizeof(buf1), - sn->vm_state_size), + sizing, date_buf, clock_buf); } + g_free(sizing); } static void dump_qdict(int indentation, QDict *dict); @@ -XXX,XX +XXX,XX @@ void bdrv_image_info_specific_dump(ImageInfoSpecific *info_spec) void bdrv_image_info_dump(ImageInfo *info) { - char size_buf[128], dsize_buf[128]; + char *size_buf, *dsize_buf; if (!info->has_actual_size) { - snprintf(dsize_buf, sizeof(dsize_buf), "unavailable"); + dsize_buf = g_strdup("unavailable"); } else { - get_human_readable_size(dsize_buf, sizeof(dsize_buf), - info->actual_size); + dsize_buf = size_to_str(info->actual_size); } - get_human_readable_size(size_buf, sizeof(size_buf), info->virtual_size); + size_buf = size_to_str(info->virtual_size); qemu_printf("image: %s\n" "file format: %s\n" "virtual size: %s (%" PRId64 " bytes)\n" @@ -XXX,XX +XXX,XX @@ void bdrv_image_info_dump(ImageInfo *info) info->filename, info->format, size_buf, info->virtual_size, dsize_buf); + g_free(size_buf); + g_free(dsize_buf); if (info->has_encrypted && info->encrypted) { qemu_printf("encrypted: yes\n"); diff --git a/tests/qemu-iotests/043.out b/tests/qemu-iotests/043.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/043.out +++ b/tests/qemu-iotests/043.out @@ -XXX,XX +XXX,XX @@ Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=134217728 backing_file=TEST_DIR/ == finite chain of length 3 (human) == image: TEST_DIR/t.IMGFMT file format: IMGFMT -virtual size: 128M (134217728 bytes) +virtual size: 128 MiB (134217728 bytes) cluster_size: 65536 backing file: TEST_DIR/t.IMGFMT.2.base image: TEST_DIR/t.IMGFMT.2.base file format: IMGFMT -virtual size: 128M (134217728 bytes) +virtual size: 128 MiB (134217728 bytes) cluster_size: 65536 backing file: TEST_DIR/t.IMGFMT.1.base image: TEST_DIR/t.IMGFMT.1.base file format: IMGFMT -virtual size: 128M (134217728 bytes) +virtual size: 128 MiB (134217728 bytes) cluster_size: 65536 == finite chain of length 3 (json) == diff --git a/tests/qemu-iotests/053.out b/tests/qemu-iotests/053.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/053.out +++ b/tests/qemu-iotests/053.out @@ -XXX,XX +XXX,XX @@ wrote 512/512 bytes at offset 0 No errors were found on the image. == Checking compressed image virtual disk size == -virtual size: 512 (512 bytes) +virtual size: 512 B (512 bytes) == Verifying the compressed image == read 512/512 bytes at offset 0 diff --git a/tests/qemu-iotests/059.out b/tests/qemu-iotests/059.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/059.out +++ b/tests/qemu-iotests/059.out @@ -XXX,XX +XXX,XX @@ can't open device TEST_DIR/t.vmdk: L1 size too big Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=2147483648 subformat=monolithicFlat image: TEST_DIR/t.IMGFMT file format: IMGFMT -virtual size: 2.0G (2147483648 bytes) +virtual size: 2 GiB (2147483648 bytes) === Testing monolithicFlat with zeroed_grain === qemu-img: TEST_DIR/t.IMGFMT: Flat image can't enable zeroed grain @@ -XXX,XX +XXX,XX @@ Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=2147483648 subformat=monolithicF Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1073741824000 subformat=twoGbMaxExtentFlat image: TEST_DIR/t.vmdk file format: vmdk -virtual size: 1.0T (1073741824000 bytes) -disk size: 16K +virtual size: 0.977 TiB (1073741824000 bytes) +disk size: 16 KiB Format specific information: cid: XXXXXXXX parent cid: XXXXXXXX @@ -XXX,XX +XXX,XX @@ can't open: Cannot use relative extent paths with VMDK descriptor file 'json:{"i === Testing version 3 === image: TEST_DIR/iotest-version3.IMGFMT file format: IMGFMT -virtual size: 16G (17179869184 bytes) +virtual size: 16 GiB (17179869184 bytes) cluster_size: 65536 read 512/512 bytes at offset 0 512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) @@ -XXX,XX +XXX,XX @@ read 512/512 bytes at offset 64931328 Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=4398046511104 subformat=monolithicFlat image: TEST_DIR/t.IMGFMT file format: IMGFMT -virtual size: 4.0T (4398046511104 bytes) +virtual size: 4 TiB (4398046511104 bytes) wrote 512/512 bytes at offset 966367641600 512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) e100000000: 0a 0a 0a 0a 0a 0a 0a 0a 0a 0a 0a 0a 0a 0a 0a 0a ................ diff --git a/tests/qemu-iotests/060.out b/tests/qemu-iotests/060.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/060.out +++ b/tests/qemu-iotests/060.out @@ -XXX,XX +XXX,XX @@ write failed: Input/output error incompatible_features 0x2 image: TEST_DIR/t.IMGFMT file format: IMGFMT -virtual size: 64M (67108864 bytes) +virtual size: 64 MiB (67108864 bytes) cluster_size: 65536 Format specific information: compat: 1.1 @@ -XXX,XX +XXX,XX @@ Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 qcow2: Marking image as corrupt: Refblock at 0xffffff00000000 is not covered by the refcount structures; further corruption events will be suppressed qemu-img: Failed to discard unused refblocks: Input/output error --- Checking and retrying --- -virtual size: 64M (67108864 bytes) +virtual size: 64 MiB (67108864 bytes) No errors were found on the image. Image resized. -virtual size: 32M (33554432 bytes) +virtual size: 32 MiB (33554432 bytes) === Discarding a non-covered in-bounds refblock === @@ -XXX,XX +XXX,XX @@ Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 qcow2: Marking image as corrupt: Refblock at 0x1000000000 is not covered by the refcount structures; further corruption events will be suppressed qemu-img: Failed to discard unused refblocks: Input/output error --- Checking and retrying --- -virtual size: 64M (67108864 bytes) +virtual size: 64 MiB (67108864 bytes) No errors were found on the image. Image resized. -virtual size: 32M (33554432 bytes) +virtual size: 32 MiB (33554432 bytes) === Discarding a refblock covered by an unaligned refblock === diff --git a/tests/qemu-iotests/061.out b/tests/qemu-iotests/061.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/061.out +++ b/tests/qemu-iotests/061.out @@ -XXX,XX +XXX,XX @@ Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 data_file=TEST_DIR/t.IM qemu-img: Cannot downgrade an image with a data file image: TEST_DIR/t.IMGFMT file format: IMGFMT -virtual size: 64M (67108864 bytes) +virtual size: 64 MiB (67108864 bytes) cluster_size: 65536 Format specific information: compat: 1.1 @@ -XXX,XX +XXX,XX @@ Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 data_file=TEST_DIR/t.IM qemu-img: Could not open 'TEST_DIR/t.IMGFMT': Could not open 'foo': No such file or directory image: TEST_DIR/t.IMGFMT file format: IMGFMT -virtual size: 64M (67108864 bytes) +virtual size: 64 MiB (67108864 bytes) cluster_size: 65536 Format specific information: compat: 1.1 @@ -XXX,XX +XXX,XX @@ Format specific information: qemu-img: Could not open 'TEST_DIR/t.IMGFMT': 'data-file' is required for this image image: TEST_DIR/t.IMGFMT file format: IMGFMT -virtual size: 64M (67108864 bytes) +virtual size: 64 MiB (67108864 bytes) cluster_size: 65536 Format specific information: compat: 1.1 @@ -XXX,XX +XXX,XX @@ Format specific information: Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 data_file=TEST_DIR/t.IMGFMT.data data_file_raw=on image: TEST_DIR/t.IMGFMT file format: IMGFMT -virtual size: 64M (67108864 bytes) +virtual size: 64 MiB (67108864 bytes) cluster_size: 65536 Format specific information: compat: 1.1 @@ -XXX,XX +XXX,XX @@ Format specific information: No errors were found on the image. image: TEST_DIR/t.IMGFMT file format: IMGFMT -virtual size: 64M (67108864 bytes) +virtual size: 64 MiB (67108864 bytes) cluster_size: 65536 Format specific information: compat: 1.1 @@ -XXX,XX +XXX,XX @@ No errors were found on the image. qemu-img: data-file-raw cannot be set on existing images image: TEST_DIR/t.IMGFMT file format: IMGFMT -virtual size: 64M (67108864 bytes) +virtual size: 64 MiB (67108864 bytes) cluster_size: 65536 Format specific information: compat: 1.1 diff --git a/tests/qemu-iotests/070.out b/tests/qemu-iotests/070.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/070.out +++ b/tests/qemu-iotests/070.out @@ -XXX,XX +XXX,XX @@ read 18874368/18874368 bytes at offset 0 === Verify image created by Disk2VHD can be opened === image: TEST_DIR/test-disk2vhd.IMGFMT file format: IMGFMT -virtual size: 256M (268435456 bytes) +virtual size: 256 MiB (268435456 bytes) cluster_size: 2097152 *** done diff --git a/tests/qemu-iotests/082.out b/tests/qemu-iotests/082.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/082.out +++ b/tests/qemu-iotests/082.out @@ -XXX,XX +XXX,XX @@ Testing: create -f foo -f qcow2 TEST_DIR/t.qcow2 128M Formatting 'TEST_DIR/t.qcow2', fmt=qcow2 size=134217728 cluster_size=65536 lazy_refcounts=off refcount_bits=16 image: TEST_DIR/t.IMGFMT file format: IMGFMT -virtual size: 128M (134217728 bytes) +virtual size: 128 MiB (134217728 bytes) cluster_size: 65536 Testing: create -f qcow2 -o cluster_size=4k -o lazy_refcounts=on TEST_DIR/t.qcow2 128M Formatting 'TEST_DIR/t.qcow2', fmt=qcow2 size=134217728 cluster_size=4096 lazy_refcounts=on refcount_bits=16 image: TEST_DIR/t.IMGFMT file format: IMGFMT -virtual size: 128M (134217728 bytes) +virtual size: 128 MiB (134217728 bytes) cluster_size: 4096 Format specific information: compat: 1.1 @@ -XXX,XX +XXX,XX @@ Testing: create -f qcow2 -o cluster_size=4k -o lazy_refcounts=on -o cluster_size Formatting 'TEST_DIR/t.qcow2', fmt=qcow2 size=134217728 cluster_size=8192 lazy_refcounts=on refcount_bits=16 image: TEST_DIR/t.IMGFMT file format: IMGFMT -virtual size: 128M (134217728 bytes) +virtual size: 128 MiB (134217728 bytes) cluster_size: 8192 Format specific information: compat: 1.1 @@ -XXX,XX +XXX,XX @@ Testing: create -f qcow2 -o cluster_size=4k,cluster_size=8k TEST_DIR/t.qcow2 128 Formatting 'TEST_DIR/t.qcow2', fmt=qcow2 size=134217728 cluster_size=8192 lazy_refcounts=off refcount_bits=16 image: TEST_DIR/t.IMGFMT file format: IMGFMT -virtual size: 128M (134217728 bytes) +virtual size: 128 MiB (134217728 bytes) cluster_size: 8192 === create: help for -o === @@ -XXX,XX +XXX,XX @@ Formatting 'TEST_DIR/t.qcow2', fmt=qcow2 size=134217728 cluster_size=65536 lazy_ Testing: convert -f foo -f qcow2 TEST_DIR/t.qcow2 TEST_DIR/t.qcow2.base image: TEST_DIR/t.IMGFMT.base file format: raw -virtual size: 128M (134217728 bytes) +virtual size: 128 MiB (134217728 bytes) Testing: convert -O foo -O qcow2 TEST_DIR/t.qcow2 TEST_DIR/t.qcow2.base image: TEST_DIR/t.IMGFMT.base file format: IMGFMT -virtual size: 128M (134217728 bytes) +virtual size: 128 MiB (134217728 bytes) cluster_size: 65536 Testing: convert -O qcow2 -o cluster_size=4k -o lazy_refcounts=on TEST_DIR/t.qcow2 TEST_DIR/t.qcow2.base image: TEST_DIR/t.IMGFMT.base file format: IMGFMT -virtual size: 128M (134217728 bytes) +virtual size: 128 MiB (134217728 bytes) cluster_size: 4096 Format specific information: compat: 1.1 @@ -XXX,XX +XXX,XX @@ Format specific information: Testing: convert -O qcow2 -o cluster_size=4k -o lazy_refcounts=on -o cluster_size=8k TEST_DIR/t.qcow2 TEST_DIR/t.qcow2.base image: TEST_DIR/t.IMGFMT.base file format: IMGFMT -virtual size: 128M (134217728 bytes) +virtual size: 128 MiB (134217728 bytes) cluster_size: 8192 Format specific information: compat: 1.1 @@ -XXX,XX +XXX,XX @@ Format specific information: Testing: convert -O qcow2 -o cluster_size=4k,cluster_size=8k TEST_DIR/t.qcow2 TEST_DIR/t.qcow2.base image: TEST_DIR/t.IMGFMT.base file format: IMGFMT -virtual size: 128M (134217728 bytes) +virtual size: 128 MiB (134217728 bytes) cluster_size: 8192 === convert: help for -o === @@ -XXX,XX +XXX,XX @@ qemu-img: Cannot enable copy offloading when -c is used Testing: amend -f foo -f qcow2 -o lazy_refcounts=on TEST_DIR/t.qcow2 image: TEST_DIR/t.IMGFMT file format: IMGFMT -virtual size: 128M (134217728 bytes) +virtual size: 128 MiB (134217728 bytes) cluster_size: 65536 Format specific information: compat: 1.1 @@ -XXX,XX +XXX,XX @@ Format specific information: Testing: amend -f qcow2 -o size=130M -o lazy_refcounts=off TEST_DIR/t.qcow2 image: TEST_DIR/t.IMGFMT file format: IMGFMT -virtual size: 130M (136314880 bytes) +virtual size: 130 MiB (136314880 bytes) cluster_size: 65536 Format specific information: compat: 1.1 @@ -XXX,XX +XXX,XX @@ Format specific information: Testing: amend -f qcow2 -o size=8M -o lazy_refcounts=on -o size=132M TEST_DIR/t.qcow2 image: TEST_DIR/t.IMGFMT file format: IMGFMT -virtual size: 132M (138412032 bytes) +virtual size: 132 MiB (138412032 bytes) cluster_size: 65536 Format specific information: compat: 1.1 @@ -XXX,XX +XXX,XX @@ Format specific information: Testing: amend -f qcow2 -o size=4M,size=148M TEST_DIR/t.qcow2 image: TEST_DIR/t.IMGFMT file format: IMGFMT -virtual size: 148M (155189248 bytes) +virtual size: 148 MiB (155189248 bytes) cluster_size: 65536 === amend: help for -o === diff --git a/tests/qemu-iotests/084.out b/tests/qemu-iotests/084.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/084.out +++ b/tests/qemu-iotests/084.out @@ -XXX,XX +XXX,XX @@ QA output created by 084 Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 image: TEST_DIR/t.IMGFMT file format: IMGFMT -virtual size: 64M (67108864 bytes) +virtual size: 64 MiB (67108864 bytes) cluster_size: 1048576 disk image file size in bytes: 67109888 @@ -XXX,XX +XXX,XX @@ disk image file size in bytes: 67109888 Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 image: TEST_DIR/t.IMGFMT file format: IMGFMT -virtual size: 64M (67108864 bytes) +virtual size: 64 MiB (67108864 bytes) cluster_size: 1048576 disk image file size in bytes: 1024 Test 1: Maximum size (512 TB - 128 MB): image: TEST_DIR/t.IMGFMT file format: IMGFMT -virtual size: 512T (562949819203584 bytes) +virtual size: 512 TiB (562949819203584 bytes) cluster_size: 1048576 Test 2: Size too large (512 TB - 128 MB + 64 kB) @@ -XXX,XX +XXX,XX @@ qemu-img: Could not open 'TEST_DIR/t.IMGFMT': unsupported VDI image (too many bl Test 5: Valid Image: 64MB, Blocks In Image 64, Block Size 1MB image: TEST_DIR/t.IMGFMT file format: IMGFMT -virtual size: 64M (67108864 bytes) +virtual size: 64 MiB (67108864 bytes) cluster_size: 1048576 Test 6: Block Size != 1MB; too small test (1MB - 1) diff --git a/tests/qemu-iotests/089.out b/tests/qemu-iotests/089.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/089.out +++ b/tests/qemu-iotests/089.out @@ -XXX,XX +XXX,XX @@ read failed: Input/output error image: TEST_DIR/t.IMGFMT file format: IMGFMT -virtual size: 64M (67108864 bytes) +virtual size: 64 MiB (67108864 bytes) cluster_size: 65536 === Testing option merging === diff --git a/tests/qemu-iotests/095.out b/tests/qemu-iotests/095.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/095.out +++ b/tests/qemu-iotests/095.out @@ -XXX,XX +XXX,XX @@ Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=104857600 backing_file=TEST_DIR/ === Base image info before commit and resize === image: TEST_DIR/t.IMGFMT.base file format: IMGFMT -virtual size: 5.0M (5242880 bytes) +virtual size: 5 MiB (5242880 bytes) === Running QEMU Live Commit Test === @@ -XXX,XX +XXX,XX @@ virtual size: 5.0M (5242880 bytes) === Base image info after commit and resize === image: TEST_DIR/t.IMGFMT.base file format: IMGFMT -virtual size: 100M (104857600 bytes) +virtual size: 100 MiB (104857600 bytes) *** done diff --git a/tests/qemu-iotests/104.out b/tests/qemu-iotests/104.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/104.out +++ b/tests/qemu-iotests/104.out @@ -XXX,XX +XXX,XX @@ QA output created by 104 Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1024 image: TEST_DIR/t.IMGFMT file format: IMGFMT -virtual size: 1.0K (1024 bytes) +virtual size: 1 KiB (1024 bytes) Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1234 image: TEST_DIR/t.IMGFMT file format: IMGFMT -virtual size: 1.5K (1536 bytes) -***done +virtual size: 1.5 KiB (1536 bytes) +*** done diff --git a/tests/qemu-iotests/110.out b/tests/qemu-iotests/110.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/110.out +++ b/tests/qemu-iotests/110.out @@ -XXX,XX +XXX,XX @@ Formatting 'TEST_DIR/t.IMGFMT.base', fmt=IMGFMT size=67108864 Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 backing_file=t.IMGFMT.base image: TEST_DIR/t.IMGFMT file format: IMGFMT -virtual size: 64M (67108864 bytes) +virtual size: 64 MiB (67108864 bytes) backing file: t.IMGFMT.base (actual path: TEST_DIR/t.IMGFMT.base) === Non-reconstructable filename === image: json:{"driver": "IMGFMT", "file": {"set-state.0.event": "read_aio", "image": {"driver": "file", "filename": "TEST_DIR/t.IMGFMT"}, "driver": "blkdebug", "set-state.0.new_state": 42}} file format: IMGFMT -virtual size: 64M (67108864 bytes) +virtual size: 64 MiB (67108864 bytes) backing file: t.IMGFMT.base (actual path: TEST_DIR/t.IMGFMT.base) === Backing name is always relative to the backed image === @@ -XXX,XX +XXX,XX @@ Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 backing_file=t.IMGFMT.b image: json:{"driver": "IMGFMT", "file": {"children": [{"driver": "file", "filename": "TEST_DIR/t.IMGFMT"}, {"driver": "file", "filename": "TEST_DIR/t.IMGFMT.copy"}], "driver": "quorum", "vote-threshold": 1}} file format: IMGFMT -virtual size: 64M (67108864 bytes) +virtual size: 64 MiB (67108864 bytes) backing file: t.IMGFMT.base (cannot determine actual path) *** done diff --git a/tests/qemu-iotests/114.out b/tests/qemu-iotests/114.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/114.out +++ b/tests/qemu-iotests/114.out @@ -XXX,XX +XXX,XX @@ Formatting 'TEST_DIR/t.IMGFMT.base', fmt=IMGFMT size=67108864 Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 backing_file=TEST_DIR/t.IMGFMT.base image: TEST_DIR/t.IMGFMT file format: IMGFMT -virtual size: 64M (67108864 bytes) +virtual size: 64 MiB (67108864 bytes) cluster_size: 65536 backing file: TEST_DIR/t.IMGFMT.base backing file format: foo diff --git a/tests/qemu-iotests/126.out b/tests/qemu-iotests/126.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/126.out +++ b/tests/qemu-iotests/126.out @@ -XXX,XX +XXX,XX @@ Formatting 'TEST_DIR/image:base.IMGFMT', fmt=IMGFMT size=67108864 Formatting 'TEST_DIR/image:top.IMGFMT', fmt=IMGFMT size=67108864 backing_file=./image:base.IMGFMT image: TEST_DIR/image:top.IMGFMT file format: IMGFMT -virtual size: 64M (67108864 bytes) +virtual size: 64 MiB (67108864 bytes) backing file: ./image:base.IMGFMT (actual path: TEST_DIR/./image:base.IMGFMT) Formatting 'base.IMGFMT', fmt=IMGFMT size=67108864 Formatting 'file:image:top.IMGFMT', fmt=IMGFMT size=67108864 backing_file=base.IMGFMT image: ./image:top.IMGFMT file format: IMGFMT -virtual size: 64M (67108864 bytes) +virtual size: 64 MiB (67108864 bytes) backing file: base.IMGFMT (actual path: ./base.IMGFMT) *** done diff --git a/tests/qemu-iotests/130.out b/tests/qemu-iotests/130.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/130.out +++ b/tests/qemu-iotests/130.out @@ -XXX,XX +XXX,XX @@ Formatting 'TEST_DIR/t.IMGFMT.base', fmt=IMGFMT size=67108864 Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 image: TEST_DIR/t.IMGFMT file format: IMGFMT -virtual size: 64M (67108864 bytes) +virtual size: 64 MiB (67108864 bytes) === HMP commit === @@ -XXX,XX +XXX,XX @@ QEMU X.Y.Z monitor - type 'help' for more information (qemu) image: TEST_DIR/t.IMGFMT file format: IMGFMT -virtual size: 64M (67108864 bytes) +virtual size: 64 MiB (67108864 bytes) Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 backing_file=TEST_DIR/t.IMGFMT.orig backing_fmt=raw QEMU X.Y.Z monitor - type 'help' for more information (qemu) commit testdisk (qemu) image: TEST_DIR/t.IMGFMT file format: IMGFMT -virtual size: 64M (67108864 bytes) +virtual size: 64 MiB (67108864 bytes) backing file: TEST_DIR/t.IMGFMT.orig backing file format: raw @@ -XXX,XX +XXX,XX @@ wrote 4096/4096 bytes at offset 0 4 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) image: TEST_DIR/t.IMGFMT file format: IMGFMT -virtual size: 64M (67108864 bytes) +virtual size: 64 MiB (67108864 bytes) Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=67108864 backing_file=TEST_DIR/t.IMGFMT.orig backing_fmt=raw wrote 4096/4096 bytes at offset 0 4 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) image: TEST_DIR/t.IMGFMT file format: IMGFMT -virtual size: 64M (67108864 bytes) +virtual size: 64 MiB (67108864 bytes) backing file: TEST_DIR/t.IMGFMT.orig backing file format: raw *** done diff --git a/tests/qemu-iotests/153.out b/tests/qemu-iotests/153.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/153.out +++ b/tests/qemu-iotests/153.out @@ -XXX,XX +XXX,XX @@ _qemu_io_wrapper TEST_DIR/t.qcow2 -c write 0 512 No conflict: image: null-co:// file format: null-co -virtual size: 1.0G (1073741824 bytes) +virtual size: 1 GiB (1073741824 bytes) disk size: unavailable Conflict: diff --git a/tests/qemu-iotests/191.out b/tests/qemu-iotests/191.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/191.out +++ b/tests/qemu-iotests/191.out @@ -XXX,XX +XXX,XX @@ wrote 65536/65536 bytes at offset 1048576 } image: TEST_DIR/t.IMGFMT file format: IMGFMT -virtual size: 64M (67108864 bytes) +virtual size: 64 MiB (67108864 bytes) cluster_size: 65536 backing file: TEST_DIR/t.IMGFMT.base backing file format: IMGFMT image: TEST_DIR/t.IMGFMT.ovl2 file format: IMGFMT -virtual size: 64M (67108864 bytes) +virtual size: 64 MiB (67108864 bytes) cluster_size: 65536 backing file: TEST_DIR/t.IMGFMT.base backing file format: IMGFMT @@ -XXX,XX +XXX,XX @@ wrote 65536/65536 bytes at offset 1048576 } image: TEST_DIR/t.IMGFMT file format: IMGFMT -virtual size: 64M (67108864 bytes) +virtual size: 64 MiB (67108864 bytes) cluster_size: 65536 backing file: TEST_DIR/t.IMGFMT.base backing file format: IMGFMT image: TEST_DIR/t.IMGFMT.ovl2 file format: IMGFMT -virtual size: 64M (67108864 bytes) +virtual size: 64 MiB (67108864 bytes) cluster_size: 65536 backing file: TEST_DIR/t.IMGFMT.base backing file format: IMGFMT diff --git a/tests/qemu-iotests/195.out b/tests/qemu-iotests/195.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/195.out +++ b/tests/qemu-iotests/195.out @@ -XXX,XX +XXX,XX @@ Testing: -drive if=none,file=TEST_DIR/t.IMGFMT,backing.node-name=mid image: TEST_DIR/t.IMGFMT.mid file format: IMGFMT -virtual size: 64M (67108864 bytes) +virtual size: 64 MiB (67108864 bytes) cluster_size: 65536 backing file: /dev/null backing file format: IMGFMT @@ -XXX,XX +XXX,XX @@ Testing: -drive if=none,file=TEST_DIR/t.IMGFMT,node-name=top image: TEST_DIR/t.IMGFMT file format: IMGFMT -virtual size: 64M (67108864 bytes) +virtual size: 64 MiB (67108864 bytes) cluster_size: 65536 backing file: /dev/null backing file format: IMGFMT diff --git a/tests/qemu-iotests/198.out b/tests/qemu-iotests/198.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/198.out +++ b/tests/qemu-iotests/198.out @@ -XXX,XX +XXX,XX @@ read 16777216/16777216 bytes at offset 0 == checking image base == image: json:{"encrypt.key-secret": "sec0", "driver": "IMGFMT", "file": {"driver": "file", "filename": "TEST_DIR/t.IMGFMT.base"}} file format: IMGFMT -virtual size: 16M (16777216 bytes) +virtual size: 16 MiB (16777216 bytes) Format specific information: encrypt: ivgen alg: plain64 @@ -XXX,XX +XXX,XX @@ Format specific information: == checking image layer == image: json:{"encrypt.key-secret": "sec1", "driver": "IMGFMT", "file": {"driver": "file", "filename": "TEST_DIR/t.IMGFMT"}} file format: IMGFMT -virtual size: 16M (16777216 bytes) +virtual size: 16 MiB (16777216 bytes) backing file: TEST_DIR/t.IMGFMT.base Format specific information: encrypt: diff --git a/tests/qemu-iotests/206.out b/tests/qemu-iotests/206.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/206.out +++ b/tests/qemu-iotests/206.out @@ -XXX,XX +XXX,XX @@ image: TEST_IMG file format: IMGFMT -virtual size: 128M (134217728 bytes) +virtual size: 128 MiB (134217728 bytes) cluster_size: 65536 Format specific information: compat: 1.1 @@ -XXX,XX +XXX,XX @@ Format specific information: image: TEST_IMG file format: IMGFMT -virtual size: 64M (67108864 bytes) +virtual size: 64 MiB (67108864 bytes) cluster_size: 65536 Format specific information: compat: 1.1 @@ -XXX,XX +XXX,XX @@ Format specific information: image: TEST_IMG file format: IMGFMT -virtual size: 32M (33554432 bytes) +virtual size: 32 MiB (33554432 bytes) cluster_size: 2097152 Format specific information: compat: 1.1 @@ -XXX,XX +XXX,XX @@ Format specific information: image: TEST_IMG file format: IMGFMT -virtual size: 32M (33554432 bytes) +virtual size: 32 MiB (33554432 bytes) cluster_size: 512 backing file: TEST_IMG.base backing file format: IMGFMT @@ -XXX,XX +XXX,XX @@ Format specific information: image: TEST_IMG file format: IMGFMT -virtual size: 32M (33554432 bytes) +virtual size: 32 MiB (33554432 bytes) encrypted: yes cluster_size: 65536 Format specific information: diff --git a/tests/qemu-iotests/207.out b/tests/qemu-iotests/207.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/207.out +++ b/tests/qemu-iotests/207.out @@ -XXX,XX +XXX,XX @@ image: json:{"driver": "IMGFMT", "file": {"server.host": "127.0.0.1", "server.port": "22", "driver": "ssh", "path": "TEST_IMG"}} file format: IMGFMT -virtual size: 4.0M (4194304 bytes) +virtual size: 4 MiB (4194304 bytes) image: TEST_IMG file format: IMGFMT -virtual size: 4.0M (4194304 bytes) +virtual size: 4 MiB (4194304 bytes) === Test host-key-check options === @@ -XXX,XX +XXX,XX @@ virtual size: 4.0M (4194304 bytes) image: json:{"driver": "IMGFMT", "file": {"server.host": "127.0.0.1", "server.port": "22", "driver": "ssh", "path": "TEST_IMG"}} file format: IMGFMT -virtual size: 8.0M (8388608 bytes) +virtual size: 8 MiB (8388608 bytes) {"execute": "blockdev-create", "arguments": {"job-id": "job0", "options": {"driver": "ssh", "location": {"host-key-check": {"mode": "known_hosts"}, "path": "TEST_DIR/PID-t.img", "server": {"host": "127.0.0.1", "port": "22"}}, "size": 4194304}}} {"return": {}} @@ -XXX,XX +XXX,XX @@ virtual size: 8.0M (8388608 bytes) image: json:{"driver": "IMGFMT", "file": {"server.host": "127.0.0.1", "server.port": "22", "driver": "ssh", "path": "TEST_IMG"}} file format: IMGFMT -virtual size: 4.0M (4194304 bytes) +virtual size: 4 MiB (4194304 bytes) {"execute": "blockdev-create", "arguments": {"job-id": "job0", "options": {"driver": "ssh", "location": {"host-key-check": {"hash": "wrong", "mode": "hash", "type": "md5"}, "path": "TEST_DIR/PID-t.img", "server": {"host": "127.0.0.1", "port": "22"}}, "size": 2097152}}} {"return": {}} @@ -XXX,XX +XXX,XX @@ Job failed: remote host key does not match host_key_check 'wrong' image: json:{"driver": "IMGFMT", "file": {"server.host": "127.0.0.1", "server.port": "22", "driver": "ssh", "path": "TEST_IMG"}} file format: IMGFMT -virtual size: 8.0M (8388608 bytes) +virtual size: 8 MiB (8388608 bytes) {"execute": "blockdev-create", "arguments": {"job-id": "job0", "options": {"driver": "ssh", "location": {"host-key-check": {"hash": "wrong", "mode": "hash", "type": "sha1"}, "path": "TEST_DIR/PID-t.img", "server": {"host": "127.0.0.1", "port": "22"}}, "size": 2097152}}} {"return": {}} @@ -XXX,XX +XXX,XX @@ Job failed: remote host key does not match host_key_check 'wrong' image: json:{"driver": "IMGFMT", "file": {"server.host": "127.0.0.1", "server.port": "22", "driver": "ssh", "path": "TEST_IMG"}} file format: IMGFMT -virtual size: 4.0M (4194304 bytes) +virtual size: 4 MiB (4194304 bytes) === Invalid path and user === diff --git a/tests/qemu-iotests/210.out b/tests/qemu-iotests/210.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/210.out +++ b/tests/qemu-iotests/210.out @@ -XXX,XX +XXX,XX @@ image: json:{"driver": "IMGFMT", "file": {"driver": "file", "filename": "TEST_IMG"}, "key-secret": "keysec0"} file format: IMGFMT -virtual size: 128M (134217728 bytes) +virtual size: 128 MiB (134217728 bytes) encrypted: yes Format specific information: ivgen alg: plain64 @@ -XXX,XX +XXX,XX @@ Format specific information: image: json:{"driver": "IMGFMT", "file": {"driver": "file", "filename": "TEST_IMG"}, "key-secret": "keysec0"} file format: IMGFMT -virtual size: 64M (67108864 bytes) +virtual size: 64 MiB (67108864 bytes) encrypted: yes Format specific information: ivgen alg: plain64 @@ -XXX,XX +XXX,XX @@ Job failed: Cannot find device=this doesn't exist nor node_name=this doesn't exi image: json:{"driver": "IMGFMT", "file": {"driver": "file", "filename": "TEST_IMG"}, "key-secret": "keysec0"} file format: IMGFMT -virtual size: 0 (0 bytes) +virtual size: 0 B (0 bytes) encrypted: yes Format specific information: ivgen alg: plain64 @@ -XXX,XX +XXX,XX @@ Job failed: The requested file size is too large {"error": {"class": "GenericError", "desc": "Parameter 'size' expects a >0 size"}} image: json:{"driver": "IMGFMT", "file": {"driver": "file", "filename": "TEST_IMG"}, "key-secret": "keysec0"} file format: IMGFMT -virtual size: 0 (0 bytes) +virtual size: 0 B (0 bytes) encrypted: yes Format specific information: ivgen alg: plain64 diff --git a/tests/qemu-iotests/211.out b/tests/qemu-iotests/211.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/211.out +++ b/tests/qemu-iotests/211.out @@ -XXX,XX +XXX,XX @@ image: TEST_IMG file format: IMGFMT -virtual size: 128M (134217728 bytes) +virtual size: 128 MiB (134217728 bytes) cluster_size: 1048576 [{ "start": 0, "length": 134217728, "depth": 0, "zero": true, "data": false}] @@ -XXX,XX +XXX,XX @@ cluster_size: 1048576 image: TEST_IMG file format: IMGFMT -virtual size: 64M (67108864 bytes) +virtual size: 64 MiB (67108864 bytes) cluster_size: 1048576 [{ "start": 0, "length": 67108864, "depth": 0, "zero": true, "data": false}] @@ -XXX,XX +XXX,XX @@ cluster_size: 1048576 image: TEST_IMG file format: IMGFMT -virtual size: 32M (33554432 bytes) +virtual size: 32 MiB (33554432 bytes) cluster_size: 1048576 [{ "start": 0, "length": 3072, "depth": 0, "zero": false, "data": true, "offset": 1024}, @@ -XXX,XX +XXX,XX @@ Job failed: Cannot find device=this doesn't exist nor node_name=this doesn't exi image: TEST_IMG file format: IMGFMT -virtual size: 0 (0 bytes) +virtual size: 0 B (0 bytes) cluster_size: 1048576 === Maximum size === @@ -XXX,XX +XXX,XX @@ cluster_size: 1048576 image: TEST_IMG file format: IMGFMT -virtual size: 512T (562949819203584 bytes) +virtual size: 512 TiB (562949819203584 bytes) cluster_size: 1048576 === Invalid sizes === diff --git a/tests/qemu-iotests/212.out b/tests/qemu-iotests/212.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/212.out +++ b/tests/qemu-iotests/212.out @@ -XXX,XX +XXX,XX @@ image: TEST_IMG file format: IMGFMT -virtual size: 128M (134217728 bytes) +virtual size: 128 MiB (134217728 bytes) === Successful image creation (explicit defaults) === @@ -XXX,XX +XXX,XX @@ virtual size: 128M (134217728 bytes) image: TEST_IMG file format: IMGFMT -virtual size: 64M (67108864 bytes) +virtual size: 64 MiB (67108864 bytes) === Successful image creation (with non-default options) === @@ -XXX,XX +XXX,XX @@ virtual size: 64M (67108864 bytes) image: TEST_IMG file format: IMGFMT -virtual size: 32M (33554432 bytes) +virtual size: 32 MiB (33554432 bytes) === Invalid BlockdevRef === @@ -XXX,XX +XXX,XX @@ Job failed: Cannot find device=this doesn't exist nor node_name=this doesn't exi image: TEST_IMG file format: IMGFMT -virtual size: 0 (0 bytes) +virtual size: 0 B (0 bytes) === Maximum size === @@ -XXX,XX +XXX,XX @@ virtual size: 0 (0 bytes) image: TEST_IMG file format: IMGFMT -virtual size: 4096T (4503599627369984 bytes) +virtual size: 4 PiB (4503599627369984 bytes) === Invalid sizes === diff --git a/tests/qemu-iotests/213.out b/tests/qemu-iotests/213.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/213.out +++ b/tests/qemu-iotests/213.out @@ -XXX,XX +XXX,XX @@ image: TEST_IMG file format: IMGFMT -virtual size: 128M (134217728 bytes) +virtual size: 128 MiB (134217728 bytes) cluster_size: 8388608 === Successful image creation (explicit defaults) === @@ -XXX,XX +XXX,XX @@ cluster_size: 8388608 image: TEST_IMG file format: IMGFMT -virtual size: 64M (67108864 bytes) +virtual size: 64 MiB (67108864 bytes) cluster_size: 8388608 === Successful image creation (with non-default options) === @@ -XXX,XX +XXX,XX @@ cluster_size: 8388608 image: TEST_IMG file format: IMGFMT -virtual size: 32M (33554432 bytes) +virtual size: 32 MiB (33554432 bytes) cluster_size: 268435456 === Invalid BlockdevRef === @@ -XXX,XX +XXX,XX @@ Job failed: Cannot find device=this doesn't exist nor node_name=this doesn't exi image: TEST_IMG file format: IMGFMT -virtual size: 0 (0 bytes) +virtual size: 0 B (0 bytes) cluster_size: 8388608 === Maximum size === @@ -XXX,XX +XXX,XX @@ cluster_size: 8388608 image: TEST_IMG file format: IMGFMT -virtual size: 64T (70368744177664 bytes) +virtual size: 64 TiB (70368744177664 bytes) cluster_size: 67108864 === Invalid sizes === diff --git a/tests/qemu-iotests/233.out b/tests/qemu-iotests/233.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/233.out +++ b/tests/qemu-iotests/233.out @@ -XXX,XX +XXX,XX @@ server reported: Option 0x8 not permitted before TLS == check TLS works == image: nbd://127.0.0.1:PORT file format: nbd -virtual size: 64M (67108864 bytes) +virtual size: 64 MiB (67108864 bytes) disk size: unavailable image: nbd://127.0.0.1:PORT file format: nbd -virtual size: 64M (67108864 bytes) +virtual size: 64 MiB (67108864 bytes) disk size: unavailable exports available: 1 export: '' diff --git a/tests/qemu-iotests/237.out b/tests/qemu-iotests/237.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/237.out +++ b/tests/qemu-iotests/237.out @@ -XXX,XX +XXX,XX @@ image: TEST_IMG file format: IMGFMT -virtual size: 5.0G (5368709120 bytes) +virtual size: 5 GiB (5368709120 bytes) cluster_size: 65536 Format specific information: cid: XXXXXXXXXX @@ -XXX,XX +XXX,XX @@ Format specific information: image: TEST_IMG file format: IMGFMT -virtual size: 64M (67108864 bytes) +virtual size: 64 MiB (67108864 bytes) cluster_size: 65536 Format specific information: cid: XXXXXXXXXX @@ -XXX,XX +XXX,XX @@ Format specific information: image: TEST_IMG file format: IMGFMT -virtual size: 32M (33554432 bytes) +virtual size: 32 MiB (33554432 bytes) cluster_size: 65536 Format specific information: cid: XXXXXXXXXX @@ -XXX,XX +XXX,XX @@ Job failed: List of extents contains unused extents image: TEST_IMG file format: IMGFMT -virtual size: 512 (512 bytes) +virtual size: 512 B (512 bytes) Format specific information: cid: XXXXXXXXXX parent cid: XXXXXXXXXX @@ -XXX,XX +XXX,XX @@ Format specific information: image: TEST_IMG file format: IMGFMT -virtual size: 512 (512 bytes) +virtual size: 512 B (512 bytes) cluster_size: 65536 Format specific information: cid: XXXXXXXXXX @@ -XXX,XX +XXX,XX @@ Format specific information: image: TEST_IMG file format: IMGFMT -virtual size: 1.0G (1073741824 bytes) +virtual size: 1 GiB (1073741824 bytes) Format specific information: cid: XXXXXXXXXX parent cid: XXXXXXXXXX @@ -XXX,XX +XXX,XX @@ Format specific information: image: TEST_IMG file format: IMGFMT -virtual size: 1.0G (1073741824 bytes) +virtual size: 1 GiB (1073741824 bytes) cluster_size: 65536 Format specific information: cid: XXXXXXXXXX @@ -XXX,XX +XXX,XX @@ Format specific information: image: TEST_IMG file format: IMGFMT -virtual size: 2.0G (2147483648 bytes) +virtual size: 2 GiB (2147483648 bytes) Format specific information: cid: XXXXXXXXXX parent cid: XXXXXXXXXX @@ -XXX,XX +XXX,XX @@ Format specific information: image: TEST_IMG file format: IMGFMT -virtual size: 2.0G (2147483648 bytes) +virtual size: 2 GiB (2147483648 bytes) cluster_size: 65536 Format specific information: cid: XXXXXXXXXX @@ -XXX,XX +XXX,XX @@ Format specific information: image: TEST_IMG file format: IMGFMT -virtual size: 5.0G (5368709120 bytes) +virtual size: 5 GiB (5368709120 bytes) Format specific information: cid: XXXXXXXXXX parent cid: XXXXXXXXXX @@ -XXX,XX +XXX,XX @@ Format specific information: image: TEST_IMG file format: IMGFMT -virtual size: 5.0G (5368709120 bytes) +virtual size: 5 GiB (5368709120 bytes) cluster_size: 65536 Format specific information: cid: XXXXXXXXXX diff --git a/tests/qemu-iotests/242.out b/tests/qemu-iotests/242.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/242.out +++ b/tests/qemu-iotests/242.out @@ -XXX,XX +XXX,XX @@ qemu-img info dump: image: TEST_IMG file format: IMGFMT -virtual size: 1.0M (1048576 bytes) +virtual size: 1 MiB (1048576 bytes) cluster_size: 65536 Format specific information: compat: 1.1 @@ -XXX,XX +XXX,XX @@ qemu-img info dump: image: TEST_IMG file format: IMGFMT -virtual size: 1.0M (1048576 bytes) +virtual size: 1 MiB (1048576 bytes) cluster_size: 65536 Format specific information: compat: 1.1 @@ -XXX,XX +XXX,XX @@ qemu-img info dump: image: TEST_IMG file format: IMGFMT -virtual size: 1.0M (1048576 bytes) +virtual size: 1 MiB (1048576 bytes) cluster_size: 65536 Format specific information: compat: 1.1 @@ -XXX,XX +XXX,XX @@ qemu-img info dump: image: TEST_IMG file format: IMGFMT -virtual size: 1.0M (1048576 bytes) +virtual size: 1 MiB (1048576 bytes) cluster_size: 65536 Format specific information: compat: 1.1 @@ -XXX,XX +XXX,XX @@ Unset the unknown bitmap flag '0x4' in the bitmap directory entry: image: TEST_IMG file format: IMGFMT -virtual size: 1.0M (1048576 bytes) +virtual size: 1 MiB (1048576 bytes) cluster_size: 65536 Format specific information: compat: 1.1 -- 2.20.1 Limiting the allocation to INT_MAX bytes isn't particularly clever because it means that the final cluster will be a partial cluster which will be completed through a COW operation. This results in unnecessary data read and write requests which lead to an unwanted non-sparse filesystem block for metadata preallocation. Align the maximum allocation size down to the cluster size to avoid this situation. Cc: qemu-stable@nongnu.org Signed-off-by: Kevin Wolf <kwolf@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> --- block/qcow2.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/block/qcow2.c b/block/qcow2.c index XXXXXXX..XXXXXXX 100644 --- a/block/qcow2.c +++ b/block/qcow2.c @@ -XXX,XX +XXX,XX @@ static int coroutine_fn preallocate_co(BlockDriverState *bs, uint64_t offset, bytes = new_length - offset; while (bytes) { - cur_bytes = MIN(bytes, INT_MAX); + cur_bytes = MIN(bytes, QEMU_ALIGN_DOWN(INT_MAX, s->cluster_size)); ret = qcow2_alloc_cluster_offset(bs, offset, &cur_bytes, &host_offset, &meta); if (ret < 0) { -- 2.20.1 We'll add a bdrv_co_truncate() call in the next patch which can return an Error that we don't want to discard. So add an errp parameter to preallocate_co(). Cc: qemu-stable@nongnu.org Signed-off-by: Kevin Wolf <kwolf@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> --- block/qcow2.c | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-) diff --git a/block/qcow2.c b/block/qcow2.c index XXXXXXX..XXXXXXX 100644 --- a/block/qcow2.c +++ b/block/qcow2.c @@ -XXX,XX +XXX,XX @@ static int qcow2_set_up_encryption(BlockDriverState *bs, * Returns: 0 on success, -errno on failure. */ static int coroutine_fn preallocate_co(BlockDriverState *bs, uint64_t offset, - uint64_t new_length) + uint64_t new_length, Error **errp) { BDRVQcow2State *s = bs->opaque; uint64_t bytes; @@ -XXX,XX +XXX,XX @@ static int coroutine_fn preallocate_co(BlockDriverState *bs, uint64_t offset, ret = qcow2_alloc_cluster_offset(bs, offset, &cur_bytes, &host_offset, &meta); if (ret < 0) { + error_setg_errno(errp, -ret, "Allocating clusters failed"); return ret; } @@ -XXX,XX +XXX,XX @@ static int coroutine_fn preallocate_co(BlockDriverState *bs, uint64_t offset, ret = qcow2_alloc_cluster_link_l2(bs, meta); if (ret < 0) { + error_setg_errno(errp, -ret, "Mapping clusters failed"); qcow2_free_any_clusters(bs, meta->alloc_offset, meta->nb_clusters, QCOW2_DISCARD_NEVER); return ret; @@ -XXX,XX +XXX,XX @@ static int coroutine_fn preallocate_co(BlockDriverState *bs, uint64_t offset, ret = bdrv_pwrite(s->data_file, (host_offset + cur_bytes) - 1, &data, 1); if (ret < 0) { + error_setg_errno(errp, -ret, "Writing to EOF failed"); return ret; } } @@ -XXX,XX +XXX,XX @@ static int coroutine_fn qcow2_co_truncate(BlockDriverState *bs, int64_t offset, break; case PREALLOC_MODE_METADATA: - ret = preallocate_co(bs, old_length, offset); + ret = preallocate_co(bs, old_length, offset, errp); if (ret < 0) { - error_setg_errno(errp, -ret, "Preallocation failed"); goto fail; } break; @@ -XXX,XX +XXX,XX @@ static int coroutine_fn qcow2_co_truncate(BlockDriverState *bs, int64_t offset, /* With a data file, preallocation means just allocating the metadata * and forwarding the truncate request to the data file */ if (has_data_file(bs)) { - ret = preallocate_co(bs, old_length, offset); + ret = preallocate_co(bs, old_length, offset, errp); if (ret < 0) { - error_setg_errno(errp, -ret, "Preallocation failed"); goto fail; } break; -- 2.20.1 preallocate_co() already gave the data file the full size without forwarding the requested preallocation mode to the protocol. When bdrv_co_truncate() was called later with the preallocation mode, the file didn't actually grow any more, so the data file stayed unallocated even if full preallocation was requested. Pass the right preallocation mode to preallocate_co() and remove the second bdrv_co_truncate() to fix this. As a side effect, the ugly one-byte write in preallocate_co() is replaced with a truncate call, now leaving the last block unallocated on the protocol level as it should be. Cc: qemu-stable@nongnu.org Signed-off-by: Kevin Wolf <kwolf@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> --- block/qcow2.c | 41 +++++++++++++++++++++++------------------ 1 file changed, 23 insertions(+), 18 deletions(-) diff --git a/block/qcow2.c b/block/qcow2.c index XXXXXXX..XXXXXXX 100644 --- a/block/qcow2.c +++ b/block/qcow2.c @@ -XXX,XX +XXX,XX @@ static int qcow2_set_up_encryption(BlockDriverState *bs, * Returns: 0 on success, -errno on failure. */ static int coroutine_fn preallocate_co(BlockDriverState *bs, uint64_t offset, - uint64_t new_length, Error **errp) + uint64_t new_length, PreallocMode mode, + Error **errp) { BDRVQcow2State *s = bs->opaque; uint64_t bytes; uint64_t host_offset = 0; + int64_t file_length; unsigned int cur_bytes; int ret; QCowL2Meta *meta; @@ -XXX,XX +XXX,XX @@ static int coroutine_fn preallocate_co(BlockDriverState *bs, uint64_t offset, * all of the allocated clusters (otherwise we get failing reads after * EOF). Extend the image to the last allocated sector. */ - if (host_offset != 0) { - uint8_t data = 0; - ret = bdrv_pwrite(s->data_file, (host_offset + cur_bytes) - 1, - &data, 1); + file_length = bdrv_getlength(s->data_file->bs); + if (file_length < 0) { + error_setg_errno(errp, -file_length, "Could not get file size"); + return file_length; + } + + if (host_offset + cur_bytes > file_length) { + if (mode == PREALLOC_MODE_METADATA) { + mode = PREALLOC_MODE_OFF; + } + ret = bdrv_co_truncate(s->data_file, host_offset + cur_bytes, mode, + errp); if (ret < 0) { - error_setg_errno(errp, -ret, "Writing to EOF failed"); return ret; } } @@ -XXX,XX +XXX,XX @@ static int coroutine_fn qcow2_co_truncate(BlockDriverState *bs, int64_t offset, switch (prealloc) { case PREALLOC_MODE_OFF: + if (has_data_file(bs)) { + ret = bdrv_co_truncate(s->data_file, offset, prealloc, errp); + if (ret < 0) { + goto fail; + } + } break; case PREALLOC_MODE_METADATA: - ret = preallocate_co(bs, old_length, offset, errp); + ret = preallocate_co(bs, old_length, offset, prealloc, errp); if (ret < 0) { goto fail; } @@ -XXX,XX +XXX,XX @@ static int coroutine_fn qcow2_co_truncate(BlockDriverState *bs, int64_t offset, /* With a data file, preallocation means just allocating the metadata * and forwarding the truncate request to the data file */ if (has_data_file(bs)) { - ret = preallocate_co(bs, old_length, offset, errp); + ret = preallocate_co(bs, old_length, offset, prealloc, errp); if (ret < 0) { goto fail; } @@ -XXX,XX +XXX,XX @@ static int coroutine_fn qcow2_co_truncate(BlockDriverState *bs, int64_t offset, bs->total_sectors = offset / BDRV_SECTOR_SIZE; - if (has_data_file(bs)) { - if (prealloc == PREALLOC_MODE_METADATA) { - prealloc = PREALLOC_MODE_OFF; - } - ret = bdrv_co_truncate(s->data_file, offset, prealloc, errp); - if (ret < 0) { - goto fail; - } - } - /* write updated header.size */ offset = cpu_to_be64(offset); ret = bdrv_pwrite_sync(bs->file, offsetof(QCowHeader, size), -- 2.20.1 From: Max Reitz <mreitz@redhat.com> In the "amend" section of 082, we perform a single "convert" test (namely "convert -o help"). That does not make sense, especially because we have done exactly that "convert" test earlier in 082 already. Replacing "convert" by "amend" yields an error, which is correct because there is no point in "amend" having a default format. The user has to either specify the format, or give a file for qemu-img to probe. Signed-off-by: Max Reitz <mreitz@redhat.com> Reviewed-by: John Snow <jsnow@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- tests/qemu-iotests/082 | 5 ++++- tests/qemu-iotests/082.out | 5 ++--- 2 files changed, 6 insertions(+), 4 deletions(-) diff --git a/tests/qemu-iotests/082 b/tests/qemu-iotests/082 index XXXXXXX..XXXXXXX 100755 --- a/tests/qemu-iotests/082 +++ b/tests/qemu-iotests/082 @@ -XXX,XX +XXX,XX @@ run_qemu_img amend -f $IMGFMT -o backing_file="$TEST_IMG" -o ,, -o help "$TEST_I # Leave out everything that isn't needed run_qemu_img amend -f $IMGFMT -o help -run_qemu_img convert -o help + +# amend requires specifying either a format explicitly, or a file +# which it can probe +run_qemu_img amend -o help # Try help option for a format that does not support amendment run_qemu_img amend -f bochs -o help diff --git a/tests/qemu-iotests/082.out b/tests/qemu-iotests/082.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/082.out +++ b/tests/qemu-iotests/082.out @@ -XXX,XX +XXX,XX @@ Creation options for 'qcow2': Note that not all of these options may be amendable. -Testing: convert -o help -Supported options: - size=<size> - Virtual disk size +Testing: amend -o help +qemu-img: Expecting one image file name Testing: amend -f bochs -o help qemu-img: Format driver 'bochs' does not support option amendment -- 2.20.1 From: Max Reitz <mreitz@redhat.com> qemu-img create allows giving just a format and "-o help" to get a list of the options supported by that format. Users may not realize that the protocol level may offer even more options, which they only get to see by specifying a filename. This patch adds a note to hint at that fact. Signed-off-by: Max Reitz <mreitz@redhat.com> Reviewed-by: John Snow <jsnow@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- qemu-img.c | 13 ++++++++++++- tests/qemu-iotests/082.out | 20 ++++++++++++++++---- 2 files changed, 28 insertions(+), 5 deletions(-) diff --git a/qemu-img.c b/qemu-img.c index XXXXXXX..XXXXXXX 100644 --- a/qemu-img.c +++ b/qemu-img.c @@ -XXX,XX +XXX,XX @@ static int print_block_option_help(const char *filename, const char *fmt) create_opts = qemu_opts_append(create_opts, proto_drv->create_opts); } - printf("Supported options:\n"); + if (filename) { + printf("Supported options:\n"); + } else { + printf("Supported %s options:\n", fmt); + } qemu_opts_print_help(create_opts, false); qemu_opts_free(create_opts); + + if (!filename) { + printf("\n" + "The protocol level may support further options.\n" + "Specify the target filename to include those options.\n"); + } + return 0; } diff --git a/tests/qemu-iotests/082.out b/tests/qemu-iotests/082.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/082.out +++ b/tests/qemu-iotests/082.out @@ -XXX,XX +XXX,XX @@ Testing: create -f qcow2 -o backing_file=TEST_DIR/t.qcow2 -o ,, -o help TEST_DIR qemu-img: Invalid option list: ,, Testing: create -f qcow2 -o help -Supported options: +Supported qcow2 options: backing_file=<str> - File name of a base image backing_fmt=<str> - Image format of the base image cluster_size=<size> - qcow2 cluster size @@ -XXX,XX +XXX,XX @@ Supported options: refcount_bits=<num> - Width of a reference count entry in bits size=<size> - Virtual disk size +The protocol level may support further options. +Specify the target filename to include those options. + Testing: create -o help -Supported options: +Supported raw options: size=<size> - Virtual disk size +The protocol level may support further options. +Specify the target filename to include those options. + Testing: create -f bochs -o help qemu-img: Format driver 'bochs' does not support image creation @@ -XXX,XX +XXX,XX @@ Testing: convert -O qcow2 -o backing_file=TEST_DIR/t.qcow2 -o ,, -o help TEST_DI qemu-img: Invalid option list: ,, Testing: convert -O qcow2 -o help -Supported options: +Supported qcow2 options: backing_file=<str> - File name of a base image backing_fmt=<str> - Image format of the base image cluster_size=<size> - qcow2 cluster size @@ -XXX,XX +XXX,XX @@ Supported options: refcount_bits=<num> - Width of a reference count entry in bits size=<size> - Virtual disk size +The protocol level may support further options. +Specify the target filename to include those options. + Testing: convert -o help -Supported options: +Supported raw options: size=<size> - Virtual disk size +The protocol level may support further options. +Specify the target filename to include those options. + Testing: convert -O bochs -o help qemu-img: Format driver 'bochs' does not support image creation -- 2.20.1 make_completely_empty() is an optimisated path for bdrv_make_empty() where completely new metadata is created inside the image file instead of going through all clusters and discarding them. For an external data file, however, we actually need to do discard operations on the data file; just overwriting the qcow2 file doesn't get rid of the data. The necessary slow path with an explicit discard operation already exists for other cases. Use it for external data files, too. Cc: qemu-stable@nongnu.org Signed-off-by: Kevin Wolf <kwolf@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> --- block/qcow2.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/block/qcow2.c b/block/qcow2.c index XXXXXXX..XXXXXXX 100644 --- a/block/qcow2.c +++ b/block/qcow2.c @@ -XXX,XX +XXX,XX @@ static int qcow2_make_empty(BlockDriverState *bs) if (s->qcow_version >= 3 && !s->snapshots && !s->nb_bitmaps && 3 + l1_clusters <= s->refcount_block_size && - s->crypt_method_header != QCOW_CRYPT_LUKS) { + s->crypt_method_header != QCOW_CRYPT_LUKS && + !has_data_file(bs)) { /* The following function only works for qcow2 v3 images (it * requires the dirty flag) and only as long as there are no * features that reserve extra clusters (such as snapshots, * LUKS header, or persistent bitmaps), because it completely * empties the image. Furthermore, the L1 table and three * additional clusters (image header, refcount table, one - * refcount block) have to fit inside one refcount block. */ + * refcount block) have to fit inside one refcount block. It + * only resets the image file, i.e. does not work with an + * external data file. */ return make_completely_empty(bs); } -- 2.20.1 From: Alberto Garcia <berto@igalia.com> This patch fixes a few things in the way error codes are handled in the qcow2 compression code: a) qcow2_co_pwritev_compressed() expects qcow2_co_compress() to only return -1 or -2 on failure, but this is not correct. Since the change from qcow2_compress() to qcow2_co_compress() in commit ceb029cd6feccf9f7607 the new code can also return -EINVAL (although there does not seem to exist any code path that would cause that error in the current implementation). b) -1 and -2 are ad-hoc error codes defined in qcow2_compress(). This patch replaces them with standard constants from errno.h. c) Both qcow2_compress() and qcow2_co_do_compress() return a negative value on failure, but qcow2_co_pwritev_compressed() stores the value in an unsigned data type. Signed-off-by: Alberto Garcia <berto@igalia.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block/qcow2.c | 18 +++++++++--------- 1 file changed, 9 insertions(+), 9 deletions(-) diff --git a/block/qcow2.c b/block/qcow2.c index XXXXXXX..XXXXXXX 100644 --- a/block/qcow2.c +++ b/block/qcow2.c @@ -XXX,XX +XXX,XX @@ fail: * @src - source buffer, @src_size bytes * * Returns: compressed size on success - * -1 destination buffer is not enough to store compressed data - * -2 on any other error + * -ENOMEM destination buffer is not enough to store compressed data + * -EIO on any other error */ static ssize_t qcow2_compress(void *dest, size_t dest_size, const void *src, size_t src_size) @@ -XXX,XX +XXX,XX @@ static ssize_t qcow2_compress(void *dest, size_t dest_size, ret = deflateInit2(&strm, Z_DEFAULT_COMPRESSION, Z_DEFLATED, -12, 9, Z_DEFAULT_STRATEGY); if (ret != Z_OK) { - return -2; + return -EIO; } /* strm.next_in is not const in old zlib versions, such as those used on @@ -XXX,XX +XXX,XX @@ static ssize_t qcow2_compress(void *dest, size_t dest_size, if (ret == Z_STREAM_END) { ret = dest_size - strm.avail_out; } else { - ret = (ret == Z_OK ? -1 : -2); + ret = (ret == Z_OK ? -ENOMEM : -EIO); } deflateEnd(&strm); @@ -XXX,XX +XXX,XX @@ qcow2_co_pwritev_compressed(BlockDriverState *bs, uint64_t offset, BDRVQcow2State *s = bs->opaque; QEMUIOVector hd_qiov; int ret; - size_t out_len; + ssize_t out_len; uint8_t *buf, *out_buf; uint64_t cluster_offset; @@ -XXX,XX +XXX,XX @@ qcow2_co_pwritev_compressed(BlockDriverState *bs, uint64_t offset, out_len = qcow2_co_compress(bs, out_buf, s->cluster_size - 1, buf, s->cluster_size); - if (out_len == -2) { - ret = -EINVAL; - goto fail; - } else if (out_len == -1) { + if (out_len == -ENOMEM) { /* could not compress: write normal cluster */ ret = qcow2_co_pwritev(bs, offset, bytes, qiov, 0); if (ret < 0) { goto fail; } goto success; + } else if (out_len < 0) { + ret = -EINVAL; + goto fail; } qemu_co_mutex_lock(&s->lock); -- 2.20.1 From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Stefano Garzarella <sgarzare@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- include/block/block_int.h | 16 ++++++++++++++++ include/sysemu/block-backend.h | 19 +++++++++++++++++++ 2 files changed, 35 insertions(+) diff --git a/include/block/block_int.h b/include/block/block_int.h index XXXXXXX..XXXXXXX 100644 --- a/include/block/block_int.h +++ b/include/block/block_int.h @@ -XXX,XX +XXX,XX @@ int coroutine_fn bdrv_co_pwritev(BdrvChild *child, int64_t offset, unsigned int bytes, QEMUIOVector *qiov, BdrvRequestFlags flags); +static inline int coroutine_fn bdrv_co_pread(BdrvChild *child, + int64_t offset, unsigned int bytes, void *buf, BdrvRequestFlags flags) +{ + QEMUIOVector qiov = QEMU_IOVEC_INIT_BUF(qiov, buf, bytes); + + return bdrv_co_preadv(child, offset, bytes, &qiov, flags); +} + +static inline int coroutine_fn bdrv_co_pwrite(BdrvChild *child, + int64_t offset, unsigned int bytes, void *buf, BdrvRequestFlags flags) +{ + QEMUIOVector qiov = QEMU_IOVEC_INIT_BUF(qiov, buf, bytes); + + return bdrv_co_pwritev(child, offset, bytes, &qiov, flags); +} + extern unsigned int bdrv_drain_all_count; void bdrv_apply_subtree_drain(BdrvChild *child, BlockDriverState *new_parent); void bdrv_unapply_subtree_drain(BdrvChild *child, BlockDriverState *old_parent); diff --git a/include/sysemu/block-backend.h b/include/sysemu/block-backend.h index XXXXXXX..XXXXXXX 100644 --- a/include/sysemu/block-backend.h +++ b/include/sysemu/block-backend.h @@ -XXX,XX +XXX,XX @@ int coroutine_fn blk_co_preadv(BlockBackend *blk, int64_t offset, int coroutine_fn blk_co_pwritev(BlockBackend *blk, int64_t offset, unsigned int bytes, QEMUIOVector *qiov, BdrvRequestFlags flags); + +static inline int coroutine_fn blk_co_pread(BlockBackend *blk, int64_t offset, + unsigned int bytes, void *buf, + BdrvRequestFlags flags) +{ + QEMUIOVector qiov = QEMU_IOVEC_INIT_BUF(qiov, buf, bytes); + + return blk_co_preadv(blk, offset, bytes, &qiov, flags); +} + +static inline int coroutine_fn blk_co_pwrite(BlockBackend *blk, int64_t offset, + unsigned int bytes, void *buf, + BdrvRequestFlags flags) +{ + QEMUIOVector qiov = QEMU_IOVEC_INIT_BUF(qiov, buf, bytes); + + return blk_co_pwritev(blk, offset, bytes, &qiov, flags); +} + int blk_pwrite_zeroes(BlockBackend *blk, int64_t offset, int bytes, BdrvRequestFlags flags); BlockAIOCB *blk_aio_pwrite_zeroes(BlockBackend *blk, int64_t offset, -- 2.20.1 From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Stefano Garzarella <sgarzare@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block/qcow2.c | 9 ++------- 1 file changed, 2 insertions(+), 7 deletions(-) diff --git a/block/qcow2.c b/block/qcow2.c index XXXXXXX..XXXXXXX 100644 --- a/block/qcow2.c +++ b/block/qcow2.c @@ -XXX,XX +XXX,XX @@ qcow2_co_pwritev_compressed(BlockDriverState *bs, uint64_t offset, uint64_t bytes, QEMUIOVector *qiov) { BDRVQcow2State *s = bs->opaque; - QEMUIOVector hd_qiov; int ret; ssize_t out_len; uint8_t *buf, *out_buf; @@ -XXX,XX +XXX,XX @@ qcow2_co_pwritev_compressed(BlockDriverState *bs, uint64_t offset, goto fail; } - qemu_iovec_init_buf(&hd_qiov, out_buf, out_len); - BLKDBG_EVENT(s->data_file, BLKDBG_WRITE_COMPRESSED); - ret = bdrv_co_pwritev(s->data_file, cluster_offset, out_len, &hd_qiov, 0); + ret = bdrv_co_pwrite(s->data_file, cluster_offset, out_len, out_buf, 0); if (ret < 0) { goto fail; } @@ -XXX,XX +XXX,XX @@ qcow2_co_preadv_compressed(BlockDriverState *bs, int ret = 0, csize, nb_csectors; uint64_t coffset; uint8_t *buf, *out_buf; - QEMUIOVector local_qiov; int offset_in_cluster = offset_into_cluster(s, offset); coffset = file_cluster_offset & s->cluster_offset_mask; @@ -XXX,XX +XXX,XX @@ qcow2_co_preadv_compressed(BlockDriverState *bs, if (!buf) { return -ENOMEM; } - qemu_iovec_init_buf(&local_qiov, buf, csize); out_buf = qemu_blockalign(bs, s->cluster_size); BLKDBG_EVENT(bs->file, BLKDBG_READ_COMPRESSED); - ret = bdrv_co_preadv(bs->file, coffset, csize, &local_qiov, 0); + ret = bdrv_co_pread(bs->file, coffset, csize, buf, 0); if (ret < 0) { goto fail; } -- 2.20.1 From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Stefano Garzarella <sgarzare@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block/qcow.c | 19 ++++++------------- 1 file changed, 6 insertions(+), 13 deletions(-) diff --git a/block/qcow.c b/block/qcow.c index XXXXXXX..XXXXXXX 100644 --- a/block/qcow.c +++ b/block/qcow.c @@ -XXX,XX +XXX,XX @@ static coroutine_fn int qcow_co_preadv(BlockDriverState *bs, uint64_t offset, int offset_in_cluster; int ret = 0, n; uint64_t cluster_offset; - QEMUIOVector hd_qiov; uint8_t *buf; void *orig_buf; @@ -XXX,XX +XXX,XX @@ static coroutine_fn int qcow_co_preadv(BlockDriverState *bs, uint64_t offset, if (!cluster_offset) { if (bs->backing) { /* read from the base image */ - qemu_iovec_init_buf(&hd_qiov, buf, n); qemu_co_mutex_unlock(&s->lock); /* qcow2 emits this on bs->file instead of bs->backing */ BLKDBG_EVENT(bs->file, BLKDBG_READ_BACKING_AIO); - ret = bdrv_co_preadv(bs->backing, offset, n, &hd_qiov, 0); + ret = bdrv_co_pread(bs->backing, offset, n, buf, 0); qemu_co_mutex_lock(&s->lock); if (ret < 0) { break; @@ -XXX,XX +XXX,XX @@ static coroutine_fn int qcow_co_preadv(BlockDriverState *bs, uint64_t offset, ret = -EIO; break; } - qemu_iovec_init_buf(&hd_qiov, buf, n); qemu_co_mutex_unlock(&s->lock); BLKDBG_EVENT(bs->file, BLKDBG_READ_AIO); - ret = bdrv_co_preadv(bs->file, cluster_offset + offset_in_cluster, - n, &hd_qiov, 0); + ret = bdrv_co_pread(bs->file, cluster_offset + offset_in_cluster, + n, buf, 0); qemu_co_mutex_lock(&s->lock); if (ret < 0) { break; @@ -XXX,XX +XXX,XX @@ static coroutine_fn int qcow_co_pwritev(BlockDriverState *bs, uint64_t offset, int offset_in_cluster; uint64_t cluster_offset; int ret = 0, n; - QEMUIOVector hd_qiov; uint8_t *buf; void *orig_buf; @@ -XXX,XX +XXX,XX @@ static coroutine_fn int qcow_co_pwritev(BlockDriverState *bs, uint64_t offset, } } - qemu_iovec_init_buf(&hd_qiov, buf, n); qemu_co_mutex_unlock(&s->lock); BLKDBG_EVENT(bs->file, BLKDBG_WRITE_AIO); - ret = bdrv_co_pwritev(bs->file, cluster_offset + offset_in_cluster, - n, &hd_qiov, 0); + ret = bdrv_co_pwrite(bs->file, cluster_offset + offset_in_cluster, + n, buf, 0); qemu_co_mutex_lock(&s->lock); if (ret < 0) { break; @@ -XXX,XX +XXX,XX @@ qcow_co_pwritev_compressed(BlockDriverState *bs, uint64_t offset, uint64_t bytes, QEMUIOVector *qiov) { BDRVQcowState *s = bs->opaque; - QEMUIOVector hd_qiov; z_stream strm; int ret, out_len; uint8_t *buf, *out_buf; @@ -XXX,XX +XXX,XX @@ qcow_co_pwritev_compressed(BlockDriverState *bs, uint64_t offset, } cluster_offset &= s->cluster_offset_mask; - qemu_iovec_init_buf(&hd_qiov, out_buf, out_len); BLKDBG_EVENT(bs->file, BLKDBG_WRITE_COMPRESSED); - ret = bdrv_co_pwritev(bs->file, cluster_offset, out_len, &hd_qiov, 0); + ret = bdrv_co_pwrite(bs->file, cluster_offset, out_len, out_buf, 0); if (ret < 0) { goto fail; } -- 2.20.1 From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Move to _co_ versions of io functions qed_read_table() and qed_write_table(), as we use qemu_co_mutex_unlock() anyway. Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Stefano Garzarella <sgarzare@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block/qed-table.c | 12 +++++------- block/qed.c | 6 ++---- 2 files changed, 7 insertions(+), 11 deletions(-) diff --git a/block/qed-table.c b/block/qed-table.c index XXXXXXX..XXXXXXX 100644 --- a/block/qed-table.c +++ b/block/qed-table.c @@ -XXX,XX +XXX,XX @@ /* Called with table_lock held. */ static int qed_read_table(BDRVQEDState *s, uint64_t offset, QEDTable *table) { - QEMUIOVector qiov = QEMU_IOVEC_INIT_BUF( - qiov, table->offsets, s->header.cluster_size * s->header.table_size); + unsigned int bytes = s->header.cluster_size * s->header.table_size; + int noffsets; int i, ret; trace_qed_read_table(s, offset, table); qemu_co_mutex_unlock(&s->table_lock); - ret = bdrv_preadv(s->bs->file, offset, &qiov); + ret = bdrv_co_pread(s->bs->file, offset, bytes, table->offsets, 0); qemu_co_mutex_lock(&s->table_lock); if (ret < 0) { goto out; } /* Byteswap offsets */ - noffsets = qiov.size / sizeof(uint64_t); + noffsets = bytes / sizeof(uint64_t); for (i = 0; i < noffsets; i++) { table->offsets[i] = le64_to_cpu(table->offsets[i]); } @@ -XXX,XX +XXX,XX @@ static int qed_write_table(BDRVQEDState *s, uint64_t offset, QEDTable *table, unsigned int sector_mask = BDRV_SECTOR_SIZE / sizeof(uint64_t) - 1; unsigned int start, end, i; QEDTable *new_table; - QEMUIOVector qiov; size_t len_bytes; int ret; @@ -XXX,XX +XXX,XX @@ static int qed_write_table(BDRVQEDState *s, uint64_t offset, QEDTable *table, len_bytes = (end - start) * sizeof(uint64_t); new_table = qemu_blockalign(s->bs, len_bytes); - qemu_iovec_init_buf(&qiov, new_table->offsets, len_bytes); /* Byteswap table */ for (i = start; i < end; i++) { @@ -XXX,XX +XXX,XX @@ static int qed_write_table(BDRVQEDState *s, uint64_t offset, QEDTable *table, offset += start * sizeof(uint64_t); qemu_co_mutex_unlock(&s->table_lock); - ret = bdrv_pwritev(s->bs->file, offset, &qiov); + ret = bdrv_co_pwrite(s->bs->file, offset, len_bytes, new_table->offsets, 0); qemu_co_mutex_lock(&s->table_lock); trace_qed_write_table_cb(s, table, flush, ret); if (ret < 0) { diff --git a/block/qed.c b/block/qed.c index XXXXXXX..XXXXXXX 100644 --- a/block/qed.c +++ b/block/qed.c @@ -XXX,XX +XXX,XX @@ static int coroutine_fn qed_write_header(BDRVQEDState *s) int nsectors = DIV_ROUND_UP(sizeof(QEDHeader), BDRV_SECTOR_SIZE); size_t len = nsectors * BDRV_SECTOR_SIZE; uint8_t *buf; - QEMUIOVector qiov; int ret; assert(s->allocating_acb || s->allocating_write_reqs_plugged); buf = qemu_blockalign(s->bs, len); - qemu_iovec_init_buf(&qiov, buf, len); - ret = bdrv_co_preadv(s->bs->file, 0, qiov.size, &qiov, 0); + ret = bdrv_co_pread(s->bs->file, 0, len, buf, 0); if (ret < 0) { goto out; } @@ -XXX,XX +XXX,XX @@ static int coroutine_fn qed_write_header(BDRVQEDState *s) /* Update header */ qed_header_cpu_to_le(&s->header, (QEDHeader *) buf); - ret = bdrv_co_pwritev(s->bs->file, 0, qiov.size, &qiov, 0); + ret = bdrv_co_pwrite(s->bs->file, 0, len, buf, 0); if (ret < 0) { goto out; } -- 2.20.1 From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Stefano Garzarella <sgarzare@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block/parallels.c | 14 ++++++-------- 1 file changed, 6 insertions(+), 8 deletions(-) diff --git a/block/parallels.c b/block/parallels.c index XXXXXXX..XXXXXXX 100644 --- a/block/parallels.c +++ b/block/parallels.c @@ -XXX,XX +XXX,XX @@ static int64_t allocate_clusters(BlockDriverState *bs, int64_t sector_num, if (bs->backing) { int64_t nb_cow_sectors = to_allocate * s->tracks; int64_t nb_cow_bytes = nb_cow_sectors << BDRV_SECTOR_BITS; - QEMUIOVector qiov = - QEMU_IOVEC_INIT_BUF(qiov, qemu_blockalign(bs, nb_cow_bytes), - nb_cow_bytes); + void *buf = qemu_blockalign(bs, nb_cow_bytes); - ret = bdrv_co_preadv(bs->backing, idx * s->tracks * BDRV_SECTOR_SIZE, - nb_cow_bytes, &qiov, 0); + ret = bdrv_co_pread(bs->backing, idx * s->tracks * BDRV_SECTOR_SIZE, + nb_cow_bytes, buf, 0); if (ret < 0) { - qemu_vfree(qemu_iovec_buf(&qiov)); + qemu_vfree(buf); return ret; } ret = bdrv_co_pwritev(bs->file, s->data_end * BDRV_SECTOR_SIZE, - nb_cow_bytes, &qiov, 0); - qemu_vfree(qemu_iovec_buf(&qiov)); + nb_cow_bytes, buf, 0); + qemu_vfree(buf); if (ret < 0) { return ret; } -- 2.20.1 From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Stefano Garzarella <sgarzare@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block/backup.c | 14 ++++++-------- 1 file changed, 6 insertions(+), 8 deletions(-) diff --git a/block/backup.c b/block/backup.c index XXXXXXX..XXXXXXX 100644 --- a/block/backup.c +++ b/block/backup.c @@ -XXX,XX +XXX,XX @@ static int coroutine_fn backup_cow_with_bounce_buffer(BackupBlockJob *job, void **bounce_buffer) { int ret; - QEMUIOVector qiov; BlockBackend *blk = job->common.blk; int nbytes; int read_flags = is_write_notifier ? BDRV_REQ_NO_SERIALISING : 0; @@ -XXX,XX +XXX,XX @@ static int coroutine_fn backup_cow_with_bounce_buffer(BackupBlockJob *job, if (!*bounce_buffer) { *bounce_buffer = blk_blockalign(blk, job->cluster_size); } - qemu_iovec_init_buf(&qiov, *bounce_buffer, nbytes); - ret = blk_co_preadv(blk, start, qiov.size, &qiov, read_flags); + ret = blk_co_pread(blk, start, nbytes, *bounce_buffer, read_flags); if (ret < 0) { trace_backup_do_cow_read_fail(job, start, ret); if (error_is_read) { @@ -XXX,XX +XXX,XX @@ static int coroutine_fn backup_cow_with_bounce_buffer(BackupBlockJob *job, goto fail; } - if (qemu_iovec_is_zero(&qiov)) { + if (buffer_is_zero(*bounce_buffer, nbytes)) { ret = blk_co_pwrite_zeroes(job->target, start, - qiov.size, write_flags | BDRV_REQ_MAY_UNMAP); + nbytes, write_flags | BDRV_REQ_MAY_UNMAP); } else { - ret = blk_co_pwritev(job->target, start, - qiov.size, &qiov, write_flags | - (job->compress ? BDRV_REQ_WRITE_COMPRESSED : 0)); + ret = blk_co_pwrite(job->target, start, + nbytes, *bounce_buffer, write_flags | + (job->compress ? BDRV_REQ_WRITE_COMPRESSED : 0)); } if (ret < 0) { trace_backup_do_cow_write_fail(job, start, ret); -- 2.20.1 From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Stefano Garzarella <sgarzare@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block/commit.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/block/commit.c b/block/commit.c index XXXXXXX..XXXXXXX 100644 --- a/block/commit.c +++ b/block/commit.c @@ -XXX,XX +XXX,XX @@ static int coroutine_fn commit_populate(BlockBackend *bs, BlockBackend *base, void *buf) { int ret = 0; - QEMUIOVector qiov = QEMU_IOVEC_INIT_BUF(qiov, buf, bytes); assert(bytes < SIZE_MAX); - ret = blk_co_preadv(bs, offset, qiov.size, &qiov, 0); + ret = blk_co_pread(bs, offset, bytes, buf, 0); if (ret < 0) { return ret; } - ret = blk_co_pwritev(base, offset, qiov.size, &qiov, 0); + ret = blk_co_pwrite(base, offset, bytes, buf, 0); if (ret < 0) { return ret; } -- 2.20.1 From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Stefano Garzarella <sgarzare@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block/stream.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) diff --git a/block/stream.c b/block/stream.c index XXXXXXX..XXXXXXX 100644 --- a/block/stream.c +++ b/block/stream.c @@ -XXX,XX +XXX,XX @@ static int coroutine_fn stream_populate(BlockBackend *blk, int64_t offset, uint64_t bytes, void *buf) { - QEMUIOVector qiov = QEMU_IOVEC_INIT_BUF(qiov, buf, bytes); - assert(bytes < SIZE_MAX); /* Copy-on-read the unallocated clusters */ - return blk_co_preadv(blk, offset, qiov.size, &qiov, BDRV_REQ_COPY_ON_READ); + return blk_co_pread(blk, offset, bytes, buf, BDRV_REQ_COPY_ON_READ); } static void stream_abort(Job *job) -- 2.20.1 From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Stefano Garzarella <sgarzare@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- qemu-img.c | 13 ++++--------- 1 file changed, 4 insertions(+), 9 deletions(-) diff --git a/qemu-img.c b/qemu-img.c index XXXXXXX..XXXXXXX 100644 --- a/qemu-img.c +++ b/qemu-img.c @@ -XXX,XX +XXX,XX @@ static int coroutine_fn convert_co_read(ImgConvertState *s, int64_t sector_num, int nb_sectors, uint8_t *buf) { int n, ret; - QEMUIOVector qiov; assert(nb_sectors <= s->buf_sectors); while (nb_sectors > 0) { @@ -XXX,XX +XXX,XX @@ static int coroutine_fn convert_co_read(ImgConvertState *s, int64_t sector_num, bs_sectors = s->src_sectors[src_cur]; n = MIN(nb_sectors, bs_sectors - (sector_num - src_cur_offset)); - qemu_iovec_init_buf(&qiov, buf, n << BDRV_SECTOR_BITS); - ret = blk_co_preadv( + ret = blk_co_pread( blk, (sector_num - src_cur_offset) << BDRV_SECTOR_BITS, - n << BDRV_SECTOR_BITS, &qiov, 0); + n << BDRV_SECTOR_BITS, buf, 0); if (ret < 0) { return ret; } @@ -XXX,XX +XXX,XX @@ static int coroutine_fn convert_co_write(ImgConvertState *s, int64_t sector_num, enum ImgConvertBlockStatus status) { int ret; - QEMUIOVector qiov; while (nb_sectors > 0) { int n = nb_sectors; @@ -XXX,XX +XXX,XX @@ static int coroutine_fn convert_co_write(ImgConvertState *s, int64_t sector_num, (s->compressed && !buffer_is_zero(buf, n * BDRV_SECTOR_SIZE))) { - qemu_iovec_init_buf(&qiov, buf, n << BDRV_SECTOR_BITS); - - ret = blk_co_pwritev(s->target, sector_num << BDRV_SECTOR_BITS, - n << BDRV_SECTOR_BITS, &qiov, flags); + ret = blk_co_pwrite(s->target, sector_num << BDRV_SECTOR_BITS, + n << BDRV_SECTOR_BITS, buf, flags); if (ret < 0) { return ret; } -- 2.20.1 From: Alberto Garcia <berto@igalia.com> You can reproduce this by passing an invalid filter-node-name (like "1234") to block-commit. In this case the base image is put in read-write mode but is never reset back to read-only. Signed-off-by: Alberto Garcia <berto@igalia.com> Reviewed-by: Max Reitz <mreitz@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block/commit.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/block/commit.c b/block/commit.c index XXXXXXX..XXXXXXX 100644 --- a/block/commit.c +++ b/block/commit.c @@ -XXX,XX +XXX,XX @@ fail: if (s->top) { blk_unref(s->top); } + if (s->base_read_only) { + bdrv_reopen_set_read_only(base, true, NULL); + } job_early_fail(&s->common.job); /* commit_top_bs has to be replaced after deleting the block job, * otherwise this would fail because of lack of permissions. */ -- 2.20.1 From: Alberto Garcia <berto@igalia.com> This tests the fix from the previous patch. Signed-off-by: Alberto Garcia <berto@igalia.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- tests/qemu-iotests/249 | 115 +++++++++++++++++++++++++++++++++++++ tests/qemu-iotests/249.out | 35 +++++++++++ tests/qemu-iotests/group | 1 + 3 files changed, 151 insertions(+) create mode 100755 tests/qemu-iotests/249 create mode 100644 tests/qemu-iotests/249.out diff --git a/tests/qemu-iotests/249 b/tests/qemu-iotests/249 new file mode 100755 index XXXXXXX..XXXXXXX --- /dev/null +++ b/tests/qemu-iotests/249 @@ -XXX,XX +XXX,XX @@ +#!/usr/bin/env bash +# +# Test that a backing image is put back in read-only mode after +# block-commit (both when it fails and when it succeeds). +# +# Copyright (C) 2019 Igalia, S.L. +# +# 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 +owner=berto@igalia.com + +seq="$(basename $0)" +echo "QA output created by $seq" + +status=1 # failure is the default! + +_cleanup() +{ + _cleanup_test_img + rm -f "$TEST_IMG.base" + rm -f "$TEST_IMG.int" +} +trap "_cleanup; exit \$status" 0 1 2 3 15 + +# get standard environment, filters and checks +. ./common.rc +. ./common.filter +. ./common.qemu + +# Any format implementing BlockDriver.bdrv_change_backing_file +_supported_fmt qcow2 qed +_supported_proto file +_supported_os Linux + +IMG_SIZE=1M + +# Create the images: base <- int <- active +TEST_IMG="$TEST_IMG.base" _make_test_img $IMG_SIZE | _filter_imgfmt +TEST_IMG="$TEST_IMG.int" _make_test_img -b "$TEST_IMG.base" | _filter_imgfmt +_make_test_img -b "$TEST_IMG.int" | _filter_imgfmt + +# Launch QEMU with these two drives: +# none0: base (read-only) +# none1: base <- int <- active +_launch_qemu -drive if=none,file="${TEST_IMG}.base",node-name=base,read-only=on \ + -drive if=none,file="${TEST_IMG}",backing.node-name=int,backing.backing=base + +_send_qemu_cmd $QEMU_HANDLE \ + "{ 'execute': 'qmp_capabilities' }" \ + 'return' + +echo +echo '=== Send a write command to a drive opened in read-only mode (1)' +echo +_send_qemu_cmd $QEMU_HANDLE \ + "{ 'execute': 'human-monitor-command', + 'arguments': {'command-line': 'qemu-io none0 \"aio_write 0 2k\"'}}" \ + 'return' + +echo +echo '=== Run block-commit on base using an invalid filter node name' +echo +_send_qemu_cmd $QEMU_HANDLE \ + "{ 'execute': 'block-commit', + 'arguments': {'job-id': 'job0', 'device': 'none1', 'top-node': 'int', + 'filter-node-name': '1234'}}" \ + 'error' + +echo +echo '=== Send a write command to a drive opened in read-only mode (2)' +echo +_send_qemu_cmd $QEMU_HANDLE \ + "{ 'execute': 'human-monitor-command', + 'arguments': {'command-line': 'qemu-io none0 \"aio_write 0 2k\"'}}" \ + 'return' + +echo +echo '=== Run block-commit on base using the default filter node name' +echo +_send_qemu_cmd $QEMU_HANDLE \ + "{ 'execute': 'block-commit', + 'arguments': {'job-id': 'job0', 'device': 'none1', 'top-node': 'int'}}" \ + 'return' + +# Wait for block-commit to finish +_send_qemu_cmd $QEMU_HANDLE '' \ + '"status": "null"' + +echo +echo '=== Send a write command to a drive opened in read-only mode (3)' +echo +_send_qemu_cmd $QEMU_HANDLE \ + "{ 'execute': 'human-monitor-command', + 'arguments': {'command-line': 'qemu-io none0 \"aio_write 0 2k\"'}}" \ + 'return' + +_cleanup_qemu + +# success, all done +echo "*** done" +rm -f $seq.full +status=0 diff --git a/tests/qemu-iotests/249.out b/tests/qemu-iotests/249.out new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/tests/qemu-iotests/249.out @@ -XXX,XX +XXX,XX @@ +QA output created by 249 +Formatting 'TEST_DIR/t.IMGFMT.base', fmt=IMGFMT size=1048576 +Formatting 'TEST_DIR/t.IMGFMT.int', fmt=IMGFMT size=1048576 backing_file=TEST_DIR/t.IMGFMT.base +Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=1048576 backing_file=TEST_DIR/t.IMGFMT.int +{"return": {}} + +=== Send a write command to a drive opened in read-only mode (1) + +{"return": "Block node is read-onlyrn"} + +=== Run block-commit on base using an invalid filter node name + +{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "created", "id": "job0"}} +{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "null", "id": "job0"}} +{"error": {"class": "GenericError", "desc": "Invalid node name"}} + +=== Send a write command to a drive opened in read-only mode (2) + +{"return": "Block node is read-onlyrn"} + +=== Run block-commit on base using the default filter node name + +{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "created", "id": "job0"}} +{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "running", "id": "job0"}} +{"return": {}} +{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "waiting", "id": "job0"}} +{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "pending", "id": "job0"}} +{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "BLOCK_JOB_COMPLETED", "data": {"device": "job0", "len": 1048576, "offset": 1048576, "speed": 0, "type": "commit"}} +{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "concluded", "id": "job0"}} +{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "null", "id": "job0"}} + +=== Send a write command to a drive opened in read-only mode (3) + +{"return": "Block node is read-onlyrn"} +*** done diff --git a/tests/qemu-iotests/group b/tests/qemu-iotests/group index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/group +++ b/tests/qemu-iotests/group @@ -XXX,XX +XXX,XX @@ 246 rw auto quick 247 rw auto quick 248 rw auto quick +249 rw auto quick -- 2.20.1 From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> qed_read_table and qed_write_table use coroutine-only interfaces but are not marked coroutine_fn. Happily, they are called only from coroutine context, so we only need to add missed markers. Reported-by: Kevin Wolf <kwolf@redhat.com> Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block/qed.h | 28 ++++++++++++++++------------ block/qed-check.c | 4 ++-- block/qed-table.c | 33 ++++++++++++++++++++------------- block/qed.c | 5 +++-- 4 files changed, 41 insertions(+), 29 deletions(-) diff --git a/block/qed.h b/block/qed.h index XXXXXXX..XXXXXXX 100644 --- a/block/qed.h +++ b/block/qed.h @@ -XXX,XX +XXX,XX @@ void qed_commit_l2_cache_entry(L2TableCache *l2_cache, CachedL2Table *l2_table); /** * Table I/O functions */ -int qed_read_l1_table_sync(BDRVQEDState *s); -int qed_write_l1_table(BDRVQEDState *s, unsigned int index, unsigned int n); -int qed_write_l1_table_sync(BDRVQEDState *s, unsigned int index, - unsigned int n); -int qed_read_l2_table_sync(BDRVQEDState *s, QEDRequest *request, - uint64_t offset); -int qed_read_l2_table(BDRVQEDState *s, QEDRequest *request, uint64_t offset); -int qed_write_l2_table(BDRVQEDState *s, QEDRequest *request, - unsigned int index, unsigned int n, bool flush); -int qed_write_l2_table_sync(BDRVQEDState *s, QEDRequest *request, - unsigned int index, unsigned int n, bool flush); +int coroutine_fn qed_read_l1_table_sync(BDRVQEDState *s); +int coroutine_fn qed_write_l1_table(BDRVQEDState *s, unsigned int index, + unsigned int n); +int coroutine_fn qed_write_l1_table_sync(BDRVQEDState *s, unsigned int index, + unsigned int n); +int coroutine_fn qed_read_l2_table_sync(BDRVQEDState *s, QEDRequest *request, + uint64_t offset); +int coroutine_fn qed_read_l2_table(BDRVQEDState *s, QEDRequest *request, + uint64_t offset); +int coroutine_fn qed_write_l2_table(BDRVQEDState *s, QEDRequest *request, + unsigned int index, unsigned int n, + bool flush); +int coroutine_fn qed_write_l2_table_sync(BDRVQEDState *s, QEDRequest *request, + unsigned int index, unsigned int n, + bool flush); /** * Cluster functions @@ -XXX,XX +XXX,XX @@ int coroutine_fn qed_find_cluster(BDRVQEDState *s, QEDRequest *request, /** * Consistency check */ -int qed_check(BDRVQEDState *s, BdrvCheckResult *result, bool fix); +int coroutine_fn qed_check(BDRVQEDState *s, BdrvCheckResult *result, bool fix); QEDTable *qed_alloc_table(BDRVQEDState *s); diff --git a/block/qed-check.c b/block/qed-check.c index XXXXXXX..XXXXXXX 100644 --- a/block/qed-check.c +++ b/block/qed-check.c @@ -XXX,XX +XXX,XX @@ static unsigned int qed_check_l2_table(QEDCheck *check, QEDTable *table) /** * Descend tables and check each cluster is referenced once only */ -static int qed_check_l1_table(QEDCheck *check, QEDTable *table) +static int coroutine_fn qed_check_l1_table(QEDCheck *check, QEDTable *table) { BDRVQEDState *s = check->s; unsigned int i, num_invalid_l1 = 0; @@ -XXX,XX +XXX,XX @@ static void qed_check_mark_clean(BDRVQEDState *s, BdrvCheckResult *result) } /* Called with table_lock held. */ -int qed_check(BDRVQEDState *s, BdrvCheckResult *result, bool fix) +int coroutine_fn qed_check(BDRVQEDState *s, BdrvCheckResult *result, bool fix) { QEDCheck check = { .s = s, diff --git a/block/qed-table.c b/block/qed-table.c index XXXXXXX..XXXXXXX 100644 --- a/block/qed-table.c +++ b/block/qed-table.c @@ -XXX,XX +XXX,XX @@ #include "qemu/bswap.h" /* Called with table_lock held. */ -static int qed_read_table(BDRVQEDState *s, uint64_t offset, QEDTable *table) +static int coroutine_fn qed_read_table(BDRVQEDState *s, uint64_t offset, + QEDTable *table) { unsigned int bytes = s->header.cluster_size * s->header.table_size; @@ -XXX,XX +XXX,XX @@ out: * * Called with table_lock held. */ -static int qed_write_table(BDRVQEDState *s, uint64_t offset, QEDTable *table, - unsigned int index, unsigned int n, bool flush) +static int coroutine_fn qed_write_table(BDRVQEDState *s, uint64_t offset, + QEDTable *table, unsigned int index, + unsigned int n, bool flush) { unsigned int sector_mask = BDRV_SECTOR_SIZE / sizeof(uint64_t) - 1; unsigned int start, end, i; @@ -XXX,XX +XXX,XX @@ out: return ret; } -int qed_read_l1_table_sync(BDRVQEDState *s) +int coroutine_fn qed_read_l1_table_sync(BDRVQEDState *s) { return qed_read_table(s, s->header.l1_table_offset, s->l1_table); } /* Called with table_lock held. */ -int qed_write_l1_table(BDRVQEDState *s, unsigned int index, unsigned int n) +int coroutine_fn qed_write_l1_table(BDRVQEDState *s, unsigned int index, + unsigned int n) { BLKDBG_EVENT(s->bs->file, BLKDBG_L1_UPDATE); return qed_write_table(s, s->header.l1_table_offset, s->l1_table, index, n, false); } -int qed_write_l1_table_sync(BDRVQEDState *s, unsigned int index, - unsigned int n) +int coroutine_fn qed_write_l1_table_sync(BDRVQEDState *s, unsigned int index, + unsigned int n) { return qed_write_l1_table(s, index, n); } /* Called with table_lock held. */ -int qed_read_l2_table(BDRVQEDState *s, QEDRequest *request, uint64_t offset) +int coroutine_fn qed_read_l2_table(BDRVQEDState *s, QEDRequest *request, + uint64_t offset) { int ret; @@ -XXX,XX +XXX,XX @@ int qed_read_l2_table(BDRVQEDState *s, QEDRequest *request, uint64_t offset) return ret; } -int qed_read_l2_table_sync(BDRVQEDState *s, QEDRequest *request, uint64_t offset) +int coroutine_fn qed_read_l2_table_sync(BDRVQEDState *s, QEDRequest *request, + uint64_t offset) { return qed_read_l2_table(s, request, offset); } /* Called with table_lock held. */ -int qed_write_l2_table(BDRVQEDState *s, QEDRequest *request, - unsigned int index, unsigned int n, bool flush) +int coroutine_fn qed_write_l2_table(BDRVQEDState *s, QEDRequest *request, + unsigned int index, unsigned int n, + bool flush) { BLKDBG_EVENT(s->bs->file, BLKDBG_L2_UPDATE); return qed_write_table(s, request->l2_table->offset, request->l2_table->table, index, n, flush); } -int qed_write_l2_table_sync(BDRVQEDState *s, QEDRequest *request, - unsigned int index, unsigned int n, bool flush) +int coroutine_fn qed_write_l2_table_sync(BDRVQEDState *s, QEDRequest *request, + unsigned int index, unsigned int n, + bool flush) { return qed_write_l2_table(s, request, index, n, flush); } diff --git a/block/qed.c b/block/qed.c index XXXXXXX..XXXXXXX 100644 --- a/block/qed.c +++ b/block/qed.c @@ -XXX,XX +XXX,XX @@ static void coroutine_fn bdrv_qed_co_invalidate_cache(BlockDriverState *bs, } } -static int bdrv_qed_co_check(BlockDriverState *bs, BdrvCheckResult *result, - BdrvCheckMode fix) +static int coroutine_fn bdrv_qed_co_check(BlockDriverState *bs, + BdrvCheckResult *result, + BdrvCheckMode fix) { BDRVQEDState *s = bs->opaque; int ret; -- 2.20.1 The following changes since commit ccdf06c1db192152ac70a1dd974c624f566cb7d4: Open 6.1 development tree (2021-04-30 11:15:40 +0100) are available in the Git repository at: git://repo.or.cz/qemu/kevin.git tags/for-upstream for you to fetch changes up to 68bf7336533faa6aa90fdd4558edddbf5d8ef814: vhost-user-blk: Fail gracefully on too large queue size (2021-04-30 12:27:48 +0200) ---------------------------------------------------------------- Block layer patches - Fix permission update order problems with block graph changes - qemu-img convert: Unshare write permission for source - vhost-user-blk: Fail gracefully on too large queue size ---------------------------------------------------------------- Kevin Wolf (3): block: Add BDRV_O_NO_SHARE for blk_new_open() qemu-img convert: Unshare write permission for source vhost-user-blk: Fail gracefully on too large queue size Vladimir Sementsov-Ogievskiy (36): tests/test-bdrv-graph-mod: add test_parallel_exclusive_write tests/test-bdrv-graph-mod: add test_parallel_perm_update tests/test-bdrv-graph-mod: add test_append_greedy_filter block: bdrv_append(): don't consume reference block: BdrvChildClass: add .get_parent_aio_context handler block: drop ctx argument from bdrv_root_attach_child block: make bdrv_reopen_{prepare,commit,abort} private util: add transactions.c block: bdrv_refresh_perms: check for parents permissions conflict block: refactor bdrv_child* permission functions block: rewrite bdrv_child_try_set_perm() using bdrv_refresh_perms() block: inline bdrv_child_*() permission functions calls block: use topological sort for permission update block: add bdrv_drv_set_perm transaction action block: add bdrv_list_* permission update functions block: add bdrv_replace_child_safe() transaction action block: fix bdrv_replace_node_common block: add bdrv_attach_child_common() transaction action block: add bdrv_attach_child_noperm() transaction action block: split out bdrv_replace_node_noperm() block: adapt bdrv_append() for inserting filters block: add bdrv_remove_filter_or_cow transaction action block: introduce bdrv_drop_filter() block/backup-top: drop .active block: drop ignore_children for permission update functions block: make bdrv_unset_inherits_from to be a transaction action block: make bdrv_refresh_limits() to be a transaction action block: add bdrv_set_backing_noperm() transaction action block: bdrv_reopen_multiple(): move bdrv_flush to separate pre-prepare block: bdrv_reopen_multiple: refresh permissions on updated graph block: drop unused permission update functions block: inline bdrv_check_perm_common() block: inline bdrv_replace_child() block: refactor bdrv_child_set_perm_safe() transaction action block: rename bdrv_replace_child_safe() to bdrv_replace_child() block: refactor bdrv_node_check_perm() include/block/block.h | 14 +- include/block/block_int.h | 8 +- include/qemu/transactions.h | 63 ++ block.c | 1329 ++++++++++++++++++++------------- block/backup-top.c | 48 +- block/block-backend.c | 30 +- block/commit.c | 1 + block/file-posix.c | 91 +-- block/io.c | 31 +- block/mirror.c | 3 - blockdev.c | 4 - blockjob.c | 11 +- hw/block/vhost-user-blk.c | 5 + qemu-img.c | 2 +- tests/unit/test-bdrv-drain.c | 2 +- tests/unit/test-bdrv-graph-mod.c | 209 +++++- util/transactions.c | 96 +++ MAINTAINERS | 6 + tests/qemu-iotests/245 | 2 +- tests/qemu-iotests/283.out | 2 +- tests/qemu-iotests/tests/qsd-jobs.out | 2 +- util/meson.build | 1 + 22 files changed, 1280 insertions(+), 680 deletions(-) create mode 100644 include/qemu/transactions.h create mode 100644 util/transactions.c From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Add the test that shows that concept of ignore_children is incomplete. Actually, when we want to update something, ignoring permission of some existing BdrvChild, we should ignore also the propagated effect of this child to the other children. But that's not done. Better approach (update permissions on already updated graph) will be implemented later. Now the test fails, so it's added with -d argument to not break make check. Test fails with "Conflicts with use by fl1 as 'backing', which does not allow 'write' on base" because when updating permissions we can ignore original top->fl1 BdrvChild. But we don't ignore exclusive write permission in fl1->base BdrvChild, which is propagated. Correct thing to do is make graph change first and then do permission update from the top node. To run test do ./test-bdrv-graph-mod -d -p /bdrv-graph-mod/parallel-exclusive-write from <build-directory>/tests. Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Kevin Wolf <kwolf@redhat.com> Message-Id: <20210428151804.439460-2-vsementsov@virtuozzo.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- tests/unit/test-bdrv-graph-mod.c | 70 +++++++++++++++++++++++++++++++- 1 file changed, 69 insertions(+), 1 deletion(-) diff --git a/tests/unit/test-bdrv-graph-mod.c b/tests/unit/test-bdrv-graph-mod.c index XXXXXXX..XXXXXXX 100644 --- a/tests/unit/test-bdrv-graph-mod.c +++ b/tests/unit/test-bdrv-graph-mod.c @@ -XXX,XX +XXX,XX @@ /* * Block node graph modifications tests * - * Copyright (c) 2019 Virtuozzo International GmbH. All rights reserved. + * Copyright (c) 2019-2021 Virtuozzo International GmbH. All rights reserved. * * 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 @@ -XXX,XX +XXX,XX @@ static BlockDriver bdrv_no_perm = { .bdrv_child_perm = no_perm_default_perms, }; +static void exclusive_write_perms(BlockDriverState *bs, BdrvChild *c, + BdrvChildRole role, + BlockReopenQueue *reopen_queue, + uint64_t perm, uint64_t shared, + uint64_t *nperm, uint64_t *nshared) +{ + *nperm = BLK_PERM_WRITE; + *nshared = BLK_PERM_ALL & ~BLK_PERM_WRITE; +} + +static BlockDriver bdrv_exclusive_writer = { + .format_name = "exclusive-writer", + .bdrv_child_perm = exclusive_write_perms, +}; + static BlockDriverState *no_perm_node(const char *name) { return bdrv_new_open_driver(&bdrv_no_perm, name, BDRV_O_RDWR, &error_abort); @@ -XXX,XX +XXX,XX @@ static BlockDriverState *pass_through_node(const char *name) BDRV_O_RDWR, &error_abort); } +static BlockDriverState *exclusive_writer_node(const char *name) +{ + return bdrv_new_open_driver(&bdrv_exclusive_writer, name, + BDRV_O_RDWR, &error_abort); +} + /* * test_update_perm_tree * @@ -XXX,XX +XXX,XX @@ static void test_should_update_child(void) blk_unref(root); } +/* + * test_parallel_exclusive_write + * + * Check that when we replace node, old permissions of the node being removed + * doesn't break the replacement. + */ +static void test_parallel_exclusive_write(void) +{ + BlockDriverState *top = exclusive_writer_node("top"); + BlockDriverState *base = no_perm_node("base"); + BlockDriverState *fl1 = pass_through_node("fl1"); + BlockDriverState *fl2 = pass_through_node("fl2"); + + /* + * bdrv_attach_child() eats child bs reference, so we need two @base + * references for two filters: + */ + bdrv_ref(base); + + bdrv_attach_child(top, fl1, "backing", &child_of_bds, BDRV_CHILD_DATA, + &error_abort); + bdrv_attach_child(fl1, base, "backing", &child_of_bds, BDRV_CHILD_FILTERED, + &error_abort); + bdrv_attach_child(fl2, base, "backing", &child_of_bds, BDRV_CHILD_FILTERED, + &error_abort); + + bdrv_replace_node(fl1, fl2, &error_abort); + + bdrv_unref(fl2); + bdrv_unref(top); +} + int main(int argc, char *argv[]) { + int i; + bool debug = false; + + for (i = 1; i < argc; i++) { + if (!strcmp(argv[i], "-d")) { + debug = true; + break; + } + } + bdrv_init(); qemu_init_main_loop(&error_abort); @@ -XXX,XX +XXX,XX @@ int main(int argc, char *argv[]) g_test_add_func("/bdrv-graph-mod/should-update-child", test_should_update_child); + if (debug) { + g_test_add_func("/bdrv-graph-mod/parallel-exclusive-write", + test_parallel_exclusive_write); + } + return g_test_run(); } -- 2.30.2 From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Add test to show that simple DFS recursion order is not correct for permission update. Correct order is topological-sort order, which will be introduced later. Consider the block driver which has two filter children: one active with exclusive write access and one inactive with no specific permissions. And, these two children has a common base child, like this: ┌─────┐ ┌──────┐ │ fl2 │ ◀── │ top │ └─────┘ └──────┘ │ │ │ │ w │ ▼ │ ┌──────┐ │ │ fl1 │ │ └──────┘ │ │ │ │ w │ ▼ │ ┌──────┐ └───────▶ │ base │ └──────┘ So, exclusive write is propagated. Assume, we want to make fl2 active instead of fl1. So, we set some option for top driver and do permission update. If permission update (remember, it's DFS) goes first through top->fl1->base branch it will succeed: it firstly drop exclusive write permissions and than apply them for another BdrvChildren. But if permission update goes first through top->fl2->base branch it will fail, as when we try to update fl2->base child, old not yet updated fl1->base child will be in conflict. Now test fails, so it runs only with -d flag. To run do ./test-bdrv-graph-mod -d -p /bdrv-graph-mod/parallel-perm-update from <build-directory>/tests. Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Kevin Wolf <kwolf@redhat.com> Message-Id: <20210428151804.439460-3-vsementsov@virtuozzo.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- tests/unit/test-bdrv-graph-mod.c | 116 +++++++++++++++++++++++++++++++ 1 file changed, 116 insertions(+) diff --git a/tests/unit/test-bdrv-graph-mod.c b/tests/unit/test-bdrv-graph-mod.c index XXXXXXX..XXXXXXX 100644 --- a/tests/unit/test-bdrv-graph-mod.c +++ b/tests/unit/test-bdrv-graph-mod.c @@ -XXX,XX +XXX,XX @@ static void test_parallel_exclusive_write(void) bdrv_unref(top); } +static void write_to_file_perms(BlockDriverState *bs, BdrvChild *c, + BdrvChildRole role, + BlockReopenQueue *reopen_queue, + uint64_t perm, uint64_t shared, + uint64_t *nperm, uint64_t *nshared) +{ + if (bs->file && c == bs->file) { + *nperm = BLK_PERM_WRITE; + *nshared = BLK_PERM_ALL & ~BLK_PERM_WRITE; + } else { + *nperm = 0; + *nshared = BLK_PERM_ALL; + } +} + +static BlockDriver bdrv_write_to_file = { + .format_name = "tricky-perm", + .bdrv_child_perm = write_to_file_perms, +}; + + +/* + * The following test shows that topological-sort order is required for + * permission update, simple DFS is not enough. + * + * Consider the block driver which has two filter children: one active + * with exclusive write access and one inactive with no specific + * permissions. + * + * And, these two children has a common base child, like this: + * + * ┌─────┐ ┌──────┐ + * │ fl2 │ ◀── │ top │ + * └─────┘ └──────┘ + * │ │ + * │ │ w + * │ ▼ + * │ ┌──────┐ + * │ │ fl1 │ + * │ └──────┘ + * │ │ + * │ │ w + * │ ▼ + * │ ┌──────┐ + * └───────▶ │ base │ + * └──────┘ + * + * So, exclusive write is propagated. + * + * Assume, we want to make fl2 active instead of fl1. + * So, we set some option for top driver and do permission update. + * + * With simple DFS, if permission update goes first through + * top->fl1->base branch it will succeed: it firstly drop exclusive write + * permissions and than apply them for another BdrvChildren. + * But if permission update goes first through top->fl2->base branch it + * will fail, as when we try to update fl2->base child, old not yet + * updated fl1->base child will be in conflict. + * + * With topological-sort order we always update parents before children, so fl1 + * and fl2 are both updated when we update base and there is no conflict. + */ +static void test_parallel_perm_update(void) +{ + BlockDriverState *top = no_perm_node("top"); + BlockDriverState *tricky = + bdrv_new_open_driver(&bdrv_write_to_file, "tricky", BDRV_O_RDWR, + &error_abort); + BlockDriverState *base = no_perm_node("base"); + BlockDriverState *fl1 = pass_through_node("fl1"); + BlockDriverState *fl2 = pass_through_node("fl2"); + BdrvChild *c_fl1, *c_fl2; + + /* + * bdrv_attach_child() eats child bs reference, so we need two @base + * references for two filters: + */ + bdrv_ref(base); + + bdrv_attach_child(top, tricky, "file", &child_of_bds, BDRV_CHILD_DATA, + &error_abort); + c_fl1 = bdrv_attach_child(tricky, fl1, "first", &child_of_bds, + BDRV_CHILD_FILTERED, &error_abort); + c_fl2 = bdrv_attach_child(tricky, fl2, "second", &child_of_bds, + BDRV_CHILD_FILTERED, &error_abort); + bdrv_attach_child(fl1, base, "backing", &child_of_bds, BDRV_CHILD_FILTERED, + &error_abort); + bdrv_attach_child(fl2, base, "backing", &child_of_bds, BDRV_CHILD_FILTERED, + &error_abort); + + /* Select fl1 as first child to be active */ + tricky->file = c_fl1; + bdrv_child_refresh_perms(top, top->children.lh_first, &error_abort); + + assert(c_fl1->perm & BLK_PERM_WRITE); + assert(!(c_fl2->perm & BLK_PERM_WRITE)); + + /* Now, try to switch active child and update permissions */ + tricky->file = c_fl2; + bdrv_child_refresh_perms(top, top->children.lh_first, &error_abort); + + assert(c_fl2->perm & BLK_PERM_WRITE); + assert(!(c_fl1->perm & BLK_PERM_WRITE)); + + /* Switch once more, to not care about real child order in the list */ + tricky->file = c_fl1; + bdrv_child_refresh_perms(top, top->children.lh_first, &error_abort); + + assert(c_fl1->perm & BLK_PERM_WRITE); + assert(!(c_fl2->perm & BLK_PERM_WRITE)); + + bdrv_unref(top); +} + int main(int argc, char *argv[]) { int i; @@ -XXX,XX +XXX,XX @@ int main(int argc, char *argv[]) if (debug) { g_test_add_func("/bdrv-graph-mod/parallel-exclusive-write", test_parallel_exclusive_write); + g_test_add_func("/bdrv-graph-mod/parallel-perm-update", + test_parallel_perm_update); } return g_test_run(); -- 2.30.2 From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> bdrv_append() is not quite good for inserting filters: it does extra permission update in intermediate state, where filter get it filtered child but is not yet replace it in a backing chain. Some filters (for example backup-top) may want permissions even when have no parents. And described intermediate state becomes invalid. That's (half a) reason, why we need "inactive" state for backup-top filter. bdrv_append() will be improved later, now let's add a unit test. Now test fails, so it runs only with -d flag. To run do ./test-bdrv-graph-mod -d -p /bdrv-graph-mod/append-greedy-filter from <build-directory>/tests. Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Kevin Wolf <kwolf@redhat.com> Message-Id: <20210428151804.439460-4-vsementsov@virtuozzo.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- tests/unit/test-bdrv-graph-mod.c | 33 ++++++++++++++++++++++++++++++++ 1 file changed, 33 insertions(+) diff --git a/tests/unit/test-bdrv-graph-mod.c b/tests/unit/test-bdrv-graph-mod.c index XXXXXXX..XXXXXXX 100644 --- a/tests/unit/test-bdrv-graph-mod.c +++ b/tests/unit/test-bdrv-graph-mod.c @@ -XXX,XX +XXX,XX @@ static void test_parallel_perm_update(void) bdrv_unref(top); } +/* + * It's possible that filter required permissions allows to insert it to backing + * chain, like: + * + * 1. [top] -> [filter] -> [base] + * + * but doesn't allow to add it as a branch: + * + * 2. [filter] --\ + * v + * [top] -> [base] + * + * So, inserting such filter should do all graph modifications and only then + * update permissions. If we try to go through intermediate state [2] and update + * permissions on it we'll fail. + * + * Let's check that bdrv_append() can append such a filter. + */ +static void test_append_greedy_filter(void) +{ + BlockDriverState *top = exclusive_writer_node("top"); + BlockDriverState *base = no_perm_node("base"); + BlockDriverState *fl = exclusive_writer_node("fl1"); + + bdrv_attach_child(top, base, "backing", &child_of_bds, BDRV_CHILD_COW, + &error_abort); + + bdrv_append(fl, base, &error_abort); + bdrv_unref(top); +} + int main(int argc, char *argv[]) { int i; @@ -XXX,XX +XXX,XX @@ int main(int argc, char *argv[]) test_parallel_exclusive_write); g_test_add_func("/bdrv-graph-mod/parallel-perm-update", test_parallel_perm_update); + g_test_add_func("/bdrv-graph-mod/append-greedy-filter", + test_append_greedy_filter); } return g_test_run(); -- 2.30.2 From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> We have too much comments for this feature. It seems better just don't do it. Most of real users (tests don't count) have to create additional reference. Drop also comment in external_snapshot_prepare: - bdrv_append doesn't "remove" old bs in common sense, it sounds strange - the fact that bdrv_append can fail is obvious from the context - the fact that we must rollback all changes in transaction abort is known (it's the direct role of abort) Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Kevin Wolf <kwolf@redhat.com> Message-Id: <20210428151804.439460-5-vsementsov@virtuozzo.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block.c | 25 +++---------------------- block/backup-top.c | 1 - block/commit.c | 1 + block/mirror.c | 3 --- blockdev.c | 4 ---- tests/unit/test-bdrv-drain.c | 2 +- tests/unit/test-bdrv-graph-mod.c | 3 +++ 7 files changed, 8 insertions(+), 31 deletions(-) diff --git a/block.c b/block.c index XXXXXXX..XXXXXXX 100644 --- a/block.c +++ b/block.c @@ -XXX,XX +XXX,XX @@ static BlockDriverState *bdrv_append_temp_snapshot(BlockDriverState *bs, goto out; } - /* bdrv_append() consumes a strong reference to bs_snapshot - * (i.e. it will call bdrv_unref() on it) even on error, so in - * order to be able to return one, we have to increase - * bs_snapshot's refcount here */ - bdrv_ref(bs_snapshot); ret = bdrv_append(bs_snapshot, bs, errp); if (ret < 0) { bs_snapshot = NULL; @@ -XXX,XX +XXX,XX @@ int bdrv_replace_node(BlockDriverState *from, BlockDriverState *to, * bs_new must not be attached to a BlockBackend. * * This function does not create any image files. - * - * bdrv_append() takes ownership of a bs_new reference and unrefs it because - * that's what the callers commonly need. bs_new will be referenced by the old - * parents of bs_top after bdrv_append() returns. If the caller needs to keep a - * reference of its own, it must call bdrv_ref(). */ int bdrv_append(BlockDriverState *bs_new, BlockDriverState *bs_top, Error **errp) { int ret = bdrv_set_backing_hd(bs_new, bs_top, errp); if (ret < 0) { - goto out; + return ret; } ret = bdrv_replace_node(bs_top, bs_new, errp); if (ret < 0) { bdrv_set_backing_hd(bs_new, NULL, &error_abort); - goto out; + return ret; } - ret = 0; - -out: - /* - * bs_new is now referenced by its new parents, we don't need the - * additional reference any more. - */ - bdrv_unref(bs_new); - - return ret; + return 0; } static void bdrv_delete(BlockDriverState *bs) diff --git a/block/backup-top.c b/block/backup-top.c index XXXXXXX..XXXXXXX 100644 --- a/block/backup-top.c +++ b/block/backup-top.c @@ -XXX,XX +XXX,XX @@ BlockDriverState *bdrv_backup_top_append(BlockDriverState *source, bdrv_drained_begin(source); - bdrv_ref(top); ret = bdrv_append(top, source, errp); if (ret < 0) { error_prepend(errp, "Cannot append backup-top filter: "); diff --git a/block/commit.c b/block/commit.c index XXXXXXX..XXXXXXX 100644 --- a/block/commit.c +++ b/block/commit.c @@ -XXX,XX +XXX,XX @@ void commit_start(const char *job_id, BlockDriverState *bs, commit_top_bs->total_sectors = top->total_sectors; ret = bdrv_append(commit_top_bs, top, errp); + bdrv_unref(commit_top_bs); /* referenced by new parents or failed */ if (ret < 0) { commit_top_bs = NULL; goto fail; diff --git a/block/mirror.c b/block/mirror.c index XXXXXXX..XXXXXXX 100644 --- a/block/mirror.c +++ b/block/mirror.c @@ -XXX,XX +XXX,XX @@ static BlockJob *mirror_start_job( bs_opaque->is_commit = target_is_backing; - /* bdrv_append takes ownership of the mirror_top_bs reference, need to keep - * it alive until block_job_create() succeeds even if bs has no parent. */ - bdrv_ref(mirror_top_bs); bdrv_drained_begin(bs); ret = bdrv_append(mirror_top_bs, bs, errp); bdrv_drained_end(bs); diff --git a/blockdev.c b/blockdev.c index XXXXXXX..XXXXXXX 100644 --- a/blockdev.c +++ b/blockdev.c @@ -XXX,XX +XXX,XX @@ static void external_snapshot_prepare(BlkActionState *common, goto out; } - /* This removes our old bs and adds the new bs. This is an operation that - * can fail, so we need to do it in .prepare; undoing it for abort is - * always possible. */ - bdrv_ref(state->new_bs); ret = bdrv_append(state->new_bs, state->old_bs, errp); if (ret < 0) { goto out; diff --git a/tests/unit/test-bdrv-drain.c b/tests/unit/test-bdrv-drain.c index XXXXXXX..XXXXXXX 100644 --- a/tests/unit/test-bdrv-drain.c +++ b/tests/unit/test-bdrv-drain.c @@ -XXX,XX +XXX,XX @@ static void test_append_to_drained(void) g_assert_cmpint(base_s->drain_count, ==, 1); g_assert_cmpint(base->in_flight, ==, 0); - /* Takes ownership of overlay, so we don't have to unref it later */ bdrv_append(overlay, base, &error_abort); g_assert_cmpint(base->in_flight, ==, 0); g_assert_cmpint(overlay->in_flight, ==, 0); @@ -XXX,XX +XXX,XX @@ static void test_append_to_drained(void) g_assert_cmpint(overlay->quiesce_counter, ==, 0); g_assert_cmpint(overlay_s->drain_count, ==, 0); + bdrv_unref(overlay); bdrv_unref(base); blk_unref(blk); } diff --git a/tests/unit/test-bdrv-graph-mod.c b/tests/unit/test-bdrv-graph-mod.c index XXXXXXX..XXXXXXX 100644 --- a/tests/unit/test-bdrv-graph-mod.c +++ b/tests/unit/test-bdrv-graph-mod.c @@ -XXX,XX +XXX,XX @@ static void test_update_perm_tree(void) ret = bdrv_append(filter, bs, NULL); g_assert_cmpint(ret, <, 0); + bdrv_unref(filter); blk_unref(root); } @@ -XXX,XX +XXX,XX @@ static void test_should_update_child(void) bdrv_append(filter, bs, &error_abort); g_assert(target->backing->bs == bs); + bdrv_unref(filter); bdrv_unref(bs); blk_unref(root); } @@ -XXX,XX +XXX,XX @@ static void test_append_greedy_filter(void) &error_abort); bdrv_append(fl, base, &error_abort); + bdrv_unref(fl); bdrv_unref(top); } -- 2.30.2 From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Add new handler to get aio context and implement it in all child classes. Add corresponding public interface to be used soon. Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Alberto Garcia <berto@igalia.com> Reviewed-by: Kevin Wolf <kwolf@redhat.com> Message-Id: <20210428151804.439460-6-vsementsov@virtuozzo.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- include/block/block.h | 2 ++ include/block/block_int.h | 2 ++ block.c | 13 +++++++++++++ block/block-backend.c | 9 +++++++++ blockjob.c | 8 ++++++++ 5 files changed, 34 insertions(+) diff --git a/include/block/block.h b/include/block/block.h index XXXXXXX..XXXXXXX 100644 --- a/include/block/block.h +++ b/include/block/block.h @@ -XXX,XX +XXX,XX @@ bool bdrv_child_can_set_aio_context(BdrvChild *c, AioContext *ctx, GSList **ignore, Error **errp); bool bdrv_can_set_aio_context(BlockDriverState *bs, AioContext *ctx, GSList **ignore, Error **errp); +AioContext *bdrv_child_get_parent_aio_context(BdrvChild *c); + int bdrv_probe_blocksizes(BlockDriverState *bs, BlockSizes *bsz); int bdrv_probe_geometry(BlockDriverState *bs, HDGeometry *geo); diff --git a/include/block/block_int.h b/include/block/block_int.h index XXXXXXX..XXXXXXX 100644 --- a/include/block/block_int.h +++ b/include/block/block_int.h @@ -XXX,XX +XXX,XX @@ struct BdrvChildClass { bool (*can_set_aio_ctx)(BdrvChild *child, AioContext *ctx, GSList **ignore, Error **errp); void (*set_aio_ctx)(BdrvChild *child, AioContext *ctx, GSList **ignore); + + AioContext *(*get_parent_aio_context)(BdrvChild *child); }; extern const BdrvChildClass child_of_bds; diff --git a/block.c b/block.c index XXXXXXX..XXXXXXX 100644 --- a/block.c +++ b/block.c @@ -XXX,XX +XXX,XX @@ static int bdrv_child_cb_update_filename(BdrvChild *c, BlockDriverState *base, return 0; } +static AioContext *bdrv_child_cb_get_parent_aio_context(BdrvChild *c) +{ + BlockDriverState *bs = c->opaque; + + return bdrv_get_aio_context(bs); +} + const BdrvChildClass child_of_bds = { .parent_is_bds = true, .get_parent_desc = bdrv_child_get_parent_desc, @@ -XXX,XX +XXX,XX @@ const BdrvChildClass child_of_bds = { .can_set_aio_ctx = bdrv_child_cb_can_set_aio_ctx, .set_aio_ctx = bdrv_child_cb_set_aio_ctx, .update_filename = bdrv_child_cb_update_filename, + .get_parent_aio_context = bdrv_child_cb_get_parent_aio_context, }; +AioContext *bdrv_child_get_parent_aio_context(BdrvChild *c) +{ + return c->klass->get_parent_aio_context(c); +} + static int bdrv_open_flags(BlockDriverState *bs, int flags) { int open_flags = flags; diff --git a/block/block-backend.c b/block/block-backend.c index XXXXXXX..XXXXXXX 100644 --- a/block/block-backend.c +++ b/block/block-backend.c @@ -XXX,XX +XXX,XX @@ static void blk_root_detach(BdrvChild *child) } } +static AioContext *blk_root_get_parent_aio_context(BdrvChild *c) +{ + BlockBackend *blk = c->opaque; + + return blk_get_aio_context(blk); +} + static const BdrvChildClass child_root = { .inherit_options = blk_root_inherit_options, @@ -XXX,XX +XXX,XX @@ static const BdrvChildClass child_root = { .can_set_aio_ctx = blk_root_can_set_aio_ctx, .set_aio_ctx = blk_root_set_aio_ctx, + + .get_parent_aio_context = blk_root_get_parent_aio_context, }; /* diff --git a/blockjob.c b/blockjob.c index XXXXXXX..XXXXXXX 100644 --- a/blockjob.c +++ b/blockjob.c @@ -XXX,XX +XXX,XX @@ static void child_job_set_aio_ctx(BdrvChild *c, AioContext *ctx, job->job.aio_context = ctx; } +static AioContext *child_job_get_parent_aio_context(BdrvChild *c) +{ + BlockJob *job = c->opaque; + + return job->job.aio_context; +} + static const BdrvChildClass child_job = { .get_parent_desc = child_job_get_parent_desc, .drained_begin = child_job_drained_begin, @@ -XXX,XX +XXX,XX @@ static const BdrvChildClass child_job = { .can_set_aio_ctx = child_job_can_set_aio_ctx, .set_aio_ctx = child_job_set_aio_ctx, .stay_at_node = true, + .get_parent_aio_context = child_job_get_parent_aio_context, }; void block_job_remove_all_bdrv(BlockJob *job) -- 2.30.2 From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Passing parent aio context is redundant, as child_class and parent opaque pointer are enough to retrieve it. Drop the argument and use new bdrv_child_get_parent_aio_context() interface. Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Alberto Garcia <berto@igalia.com> Reviewed-by: Kevin Wolf <kwolf@redhat.com> Message-Id: <20210428151804.439460-7-vsementsov@virtuozzo.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- include/block/block_int.h | 1 - block.c | 8 +++++--- block/block-backend.c | 4 ++-- blockjob.c | 3 +-- 4 files changed, 8 insertions(+), 8 deletions(-) diff --git a/include/block/block_int.h b/include/block/block_int.h index XXXXXXX..XXXXXXX 100644 --- a/include/block/block_int.h +++ b/include/block/block_int.h @@ -XXX,XX +XXX,XX @@ BdrvChild *bdrv_root_attach_child(BlockDriverState *child_bs, const char *child_name, const BdrvChildClass *child_class, BdrvChildRole child_role, - AioContext *ctx, uint64_t perm, uint64_t shared_perm, void *opaque, Error **errp); void bdrv_root_unref_child(BdrvChild *child); diff --git a/block.c b/block.c index XXXXXXX..XXXXXXX 100644 --- a/block.c +++ b/block.c @@ -XXX,XX +XXX,XX @@ BdrvChild *bdrv_root_attach_child(BlockDriverState *child_bs, const char *child_name, const BdrvChildClass *child_class, BdrvChildRole child_role, - AioContext *ctx, uint64_t perm, uint64_t shared_perm, void *opaque, Error **errp) { BdrvChild *child; Error *local_err = NULL; int ret; + AioContext *ctx; ret = bdrv_check_update_perm(child_bs, NULL, perm, shared_perm, NULL, errp); if (ret < 0) { @@ -XXX,XX +XXX,XX @@ BdrvChild *bdrv_root_attach_child(BlockDriverState *child_bs, .opaque = opaque, }; + ctx = bdrv_child_get_parent_aio_context(child); + /* If the AioContexts don't match, first try to move the subtree of * child_bs into the AioContext of the new parent. If this doesn't work, * try moving the parent into the AioContext of child_bs instead. */ @@ -XXX,XX +XXX,XX @@ BdrvChild *bdrv_attach_child(BlockDriverState *parent_bs, perm, shared_perm, &perm, &shared_perm); child = bdrv_root_attach_child(child_bs, child_name, child_class, - child_role, bdrv_get_aio_context(parent_bs), - perm, shared_perm, parent_bs, errp); + child_role, perm, shared_perm, parent_bs, + errp); if (child == NULL) { return NULL; } diff --git a/block/block-backend.c b/block/block-backend.c index XXXXXXX..XXXXXXX 100644 --- a/block/block-backend.c +++ b/block/block-backend.c @@ -XXX,XX +XXX,XX @@ BlockBackend *blk_new_open(const char *filename, const char *reference, blk->root = bdrv_root_attach_child(bs, "root", &child_root, BDRV_CHILD_FILTERED | BDRV_CHILD_PRIMARY, - blk->ctx, perm, BLK_PERM_ALL, blk, errp); + perm, BLK_PERM_ALL, blk, errp); if (!blk->root) { blk_unref(blk); return NULL; @@ -XXX,XX +XXX,XX @@ int blk_insert_bs(BlockBackend *blk, BlockDriverState *bs, Error **errp) bdrv_ref(bs); blk->root = bdrv_root_attach_child(bs, "root", &child_root, BDRV_CHILD_FILTERED | BDRV_CHILD_PRIMARY, - blk->ctx, blk->perm, blk->shared_perm, + blk->perm, blk->shared_perm, blk, errp); if (blk->root == NULL) { return -EPERM; diff --git a/blockjob.c b/blockjob.c index XXXXXXX..XXXXXXX 100644 --- a/blockjob.c +++ b/blockjob.c @@ -XXX,XX +XXX,XX @@ int block_job_add_bdrv(BlockJob *job, const char *name, BlockDriverState *bs, if (need_context_ops && job->job.aio_context != qemu_get_aio_context()) { aio_context_release(job->job.aio_context); } - c = bdrv_root_attach_child(bs, name, &child_job, 0, - job->job.aio_context, perm, shared_perm, job, + c = bdrv_root_attach_child(bs, name, &child_job, 0, perm, shared_perm, job, errp); if (need_context_ops && job->job.aio_context != qemu_get_aio_context()) { aio_context_acquire(job->job.aio_context); -- 2.30.2 From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> These functions are called only from bdrv_reopen_multiple() in block.c. No reason to publish them. Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Alberto Garcia <berto@igalia.com> Reviewed-by: Kevin Wolf <kwolf@redhat.com> Message-Id: <20210428151804.439460-8-vsementsov@virtuozzo.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- include/block/block.h | 4 ---- block.c | 13 +++++++++---- 2 files changed, 9 insertions(+), 8 deletions(-) diff --git a/include/block/block.h b/include/block/block.h index XXXXXXX..XXXXXXX 100644 --- a/include/block/block.h +++ b/include/block/block.h @@ -XXX,XX +XXX,XX @@ BlockReopenQueue *bdrv_reopen_queue(BlockReopenQueue *bs_queue, int bdrv_reopen_multiple(BlockReopenQueue *bs_queue, Error **errp); int bdrv_reopen_set_read_only(BlockDriverState *bs, bool read_only, Error **errp); -int bdrv_reopen_prepare(BDRVReopenState *reopen_state, - BlockReopenQueue *queue, Error **errp); -void bdrv_reopen_commit(BDRVReopenState *reopen_state); -void bdrv_reopen_abort(BDRVReopenState *reopen_state); int bdrv_pwrite_zeroes(BdrvChild *child, int64_t offset, int64_t bytes, BdrvRequestFlags flags); int bdrv_make_zero(BdrvChild *child, BdrvRequestFlags flags); diff --git a/block.c b/block.c index XXXXXXX..XXXXXXX 100644 --- a/block.c +++ b/block.c @@ -XXX,XX +XXX,XX @@ static BlockDriverState *bdrv_open_inherit(const char *filename, BdrvChildRole child_role, Error **errp); +static int bdrv_reopen_prepare(BDRVReopenState *reopen_state, BlockReopenQueue + *queue, Error **errp); +static void bdrv_reopen_commit(BDRVReopenState *reopen_state); +static void bdrv_reopen_abort(BDRVReopenState *reopen_state); + /* If non-zero, use only whitelisted block drivers */ static int use_bdrv_whitelist; @@ -XXX,XX +XXX,XX @@ static int bdrv_reopen_parse_backing(BDRVReopenState *reopen_state, * commit() for any other BDS that have been left in a prepare() state * */ -int bdrv_reopen_prepare(BDRVReopenState *reopen_state, BlockReopenQueue *queue, - Error **errp) +static int bdrv_reopen_prepare(BDRVReopenState *reopen_state, + BlockReopenQueue *queue, Error **errp) { int ret = -1; int old_flags; @@ -XXX,XX +XXX,XX @@ error: * makes them final by swapping the staging BlockDriverState contents into * the active BlockDriverState contents. */ -void bdrv_reopen_commit(BDRVReopenState *reopen_state) +static void bdrv_reopen_commit(BDRVReopenState *reopen_state) { BlockDriver *drv; BlockDriverState *bs; @@ -XXX,XX +XXX,XX @@ void bdrv_reopen_commit(BDRVReopenState *reopen_state) * Abort the reopen, and delete and free the staged changes in * reopen_state */ -void bdrv_reopen_abort(BDRVReopenState *reopen_state) +static void bdrv_reopen_abort(BDRVReopenState *reopen_state) { BlockDriver *drv; -- 2.30.2 From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Add simple transaction API to use in further update of block graph operations. Supposed usage is: - "prepare" is main function of the action and it should make the main effect of the action to be visible for the following actions, keeping possibility of roll-back, saving necessary things in action state, which is prepended to the action list (to do that, prepare func should call tran_add()). So, driver struct doesn't include "prepare" field, as it is supposed to be called directly. - commit/rollback is supposed to be called for the list of action states, to commit/rollback all the actions in reverse order - When possible "commit" should not make visible effect for other actions, which make possible transparent logical interaction between actions. Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Kevin Wolf <kwolf@redhat.com> Message-Id: <20210428151804.439460-9-vsementsov@virtuozzo.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- include/qemu/transactions.h | 63 ++++++++++++++++++++++++ util/transactions.c | 96 +++++++++++++++++++++++++++++++++++++ MAINTAINERS | 6 +++ util/meson.build | 1 + 4 files changed, 166 insertions(+) create mode 100644 include/qemu/transactions.h create mode 100644 util/transactions.c diff --git a/include/qemu/transactions.h b/include/qemu/transactions.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/include/qemu/transactions.h @@ -XXX,XX +XXX,XX @@ +/* + * Simple transactions API + * + * Copyright (c) 2021 Virtuozzo International GmbH. + * + * Author: + * Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> + * + * 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/>. + * + * + * = Generic transaction API = + * + * The intended usage is the following: you create "prepare" functions, which + * represents the actions. They will usually have Transaction* argument, and + * call tran_add() to register finalization callbacks. For finalization + * callbacks, prepare corresponding TransactionActionDrv structures. + * + * Then, when you need to make a transaction, create an empty Transaction by + * tran_create(), call your "prepare" functions on it, and finally call + * tran_abort() or tran_commit() to finalize the transaction by corresponding + * finalization actions in reverse order. + */ + +#ifndef QEMU_TRANSACTIONS_H +#define QEMU_TRANSACTIONS_H + +#include <gmodule.h> + +typedef struct TransactionActionDrv { + void (*abort)(void *opaque); + void (*commit)(void *opaque); + void (*clean)(void *opaque); +} TransactionActionDrv; + +typedef struct Transaction Transaction; + +Transaction *tran_new(void); +void tran_add(Transaction *tran, TransactionActionDrv *drv, void *opaque); +void tran_abort(Transaction *tran); +void tran_commit(Transaction *tran); + +static inline void tran_finalize(Transaction *tran, int ret) +{ + if (ret < 0) { + tran_abort(tran); + } else { + tran_commit(tran); + } +} + +#endif /* QEMU_TRANSACTIONS_H */ diff --git a/util/transactions.c b/util/transactions.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/util/transactions.c @@ -XXX,XX +XXX,XX @@ +/* + * Simple transactions API + * + * Copyright (c) 2021 Virtuozzo International GmbH. + * + * Author: + * Sementsov-Ogievskiy Vladimir <vsementsov@virtuozzo.com> + * + * 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/>. + */ + +#include "qemu/osdep.h" + +#include "qemu/transactions.h" +#include "qemu/queue.h" + +typedef struct TransactionAction { + TransactionActionDrv *drv; + void *opaque; + QSLIST_ENTRY(TransactionAction) entry; +} TransactionAction; + +struct Transaction { + QSLIST_HEAD(, TransactionAction) actions; +}; + +Transaction *tran_new(void) +{ + Transaction *tran = g_new(Transaction, 1); + + QSLIST_INIT(&tran->actions); + + return tran; +} + +void tran_add(Transaction *tran, TransactionActionDrv *drv, void *opaque) +{ + TransactionAction *act; + + act = g_new(TransactionAction, 1); + *act = (TransactionAction) { + .drv = drv, + .opaque = opaque + }; + + QSLIST_INSERT_HEAD(&tran->actions, act, entry); +} + +void tran_abort(Transaction *tran) +{ + TransactionAction *act, *next; + + QSLIST_FOREACH_SAFE(act, &tran->actions, entry, next) { + if (act->drv->abort) { + act->drv->abort(act->opaque); + } + + if (act->drv->clean) { + act->drv->clean(act->opaque); + } + + g_free(act); + } + + g_free(tran); +} + +void tran_commit(Transaction *tran) +{ + TransactionAction *act, *next; + + QSLIST_FOREACH_SAFE(act, &tran->actions, entry, next) { + if (act->drv->commit) { + act->drv->commit(act->opaque); + } + + if (act->drv->clean) { + act->drv->clean(act->opaque); + } + + g_free(act); + } + + g_free(tran); +} diff --git a/MAINTAINERS b/MAINTAINERS index XXXXXXX..XXXXXXX 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -XXX,XX +XXX,XX @@ M: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> S: Maintained F: scripts/simplebench/ +Transactions helper +M: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> +S: Maintained +F: include/qemu/transactions.h +F: util/transactions.c + QAPI M: Markus Armbruster <armbru@redhat.com> M: Michael Roth <michael.roth@amd.com> diff --git a/util/meson.build b/util/meson.build index XXXXXXX..XXXXXXX 100644 --- a/util/meson.build +++ b/util/meson.build @@ -XXX,XX +XXX,XX @@ util_ss.add(files('qsp.c')) util_ss.add(files('range.c')) util_ss.add(files('stats64.c')) util_ss.add(files('systemd.c')) +util_ss.add(files('transactions.c')) util_ss.add(when: 'CONFIG_POSIX', if_true: files('drm.c')) util_ss.add(files('guest-random.c')) util_ss.add(files('yank.c')) -- 2.30.2 From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Add additional check that node parents do not interfere with each other. This should not hurt existing callers and allows in further patch use bdrv_refresh_perms() to update a subtree of changed BdrvChild (check that change is correct). New check will substitute bdrv_check_update_perm() in following permissions refactoring, so keep error messages the same to avoid unit test result changes. Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Alberto Garcia <berto@igalia.com> Reviewed-by: Kevin Wolf <kwolf@redhat.com> Message-Id: <20210428151804.439460-10-vsementsov@virtuozzo.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block.c | 63 ++++++++++++++++++++++++++++++++++++++++++++++++--------- 1 file changed, 54 insertions(+), 9 deletions(-) diff --git a/block.c b/block.c index XXXXXXX..XXXXXXX 100644 --- a/block.c +++ b/block.c @@ -XXX,XX +XXX,XX @@ bool bdrv_is_writable(BlockDriverState *bs) return bdrv_is_writable_after_reopen(bs, NULL); } +static char *bdrv_child_user_desc(BdrvChild *c) +{ + if (c->klass->get_parent_desc) { + return c->klass->get_parent_desc(c); + } + + return g_strdup("another user"); +} + +static bool bdrv_a_allow_b(BdrvChild *a, BdrvChild *b, Error **errp) +{ + g_autofree char *user = NULL; + g_autofree char *perm_names = NULL; + + if ((b->perm & a->shared_perm) == b->perm) { + return true; + } + + perm_names = bdrv_perm_names(b->perm & ~a->shared_perm); + user = bdrv_child_user_desc(a); + error_setg(errp, "Conflicts with use by %s as '%s', which does not " + "allow '%s' on %s", + user, a->name, perm_names, bdrv_get_node_name(b->bs)); + + return false; +} + +static bool bdrv_parent_perms_conflict(BlockDriverState *bs, Error **errp) +{ + BdrvChild *a, *b; + + /* + * During the loop we'll look at each pair twice. That's correct because + * bdrv_a_allow_b() is asymmetric and we should check each pair in both + * directions. + */ + QLIST_FOREACH(a, &bs->parents, next_parent) { + QLIST_FOREACH(b, &bs->parents, next_parent) { + if (a == b) { + continue; + } + + if (!bdrv_a_allow_b(a, b, errp)) { + return true; + } + } + } + + return false; +} + static void bdrv_child_perm(BlockDriverState *bs, BlockDriverState *child_bs, BdrvChild *c, BdrvChildRole role, BlockReopenQueue *reopen_queue, @@ -XXX,XX +XXX,XX @@ void bdrv_get_cumulative_perm(BlockDriverState *bs, uint64_t *perm, *shared_perm = cumulative_shared_perms; } -static char *bdrv_child_user_desc(BdrvChild *c) -{ - if (c->klass->get_parent_desc) { - return c->klass->get_parent_desc(c); - } - - return g_strdup("another user"); -} - char *bdrv_perm_names(uint64_t perm) { struct perm_name { @@ -XXX,XX +XXX,XX @@ static int bdrv_refresh_perms(BlockDriverState *bs, Error **errp) int ret; uint64_t perm, shared_perm; + if (bdrv_parent_perms_conflict(bs, errp)) { + return -EPERM; + } bdrv_get_cumulative_perm(bs, &perm, &shared_perm); ret = bdrv_check_perm(bs, NULL, perm, shared_perm, NULL, errp); if (ret < 0) { -- 2.30.2 From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Split out non-recursive parts, and refactor as block graph transaction action. Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Kevin Wolf <kwolf@redhat.com> Message-Id: <20210428151804.439460-11-vsementsov@virtuozzo.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block.c | 79 ++++++++++++++++++++++++++++++++++++++++++--------------- 1 file changed, 59 insertions(+), 20 deletions(-) diff --git a/block.c b/block.c index XXXXXXX..XXXXXXX 100644 --- a/block.c +++ b/block.c @@ -XXX,XX +XXX,XX @@ #include "qemu/timer.h" #include "qemu/cutils.h" #include "qemu/id.h" +#include "qemu/transactions.h" #include "block/coroutines.h" #ifdef CONFIG_BSD @@ -XXX,XX +XXX,XX @@ static void bdrv_child_perm(BlockDriverState *bs, BlockDriverState *child_bs, } } +static void bdrv_child_set_perm_commit(void *opaque) +{ + BdrvChild *c = opaque; + + c->has_backup_perm = false; +} + +static void bdrv_child_set_perm_abort(void *opaque) +{ + BdrvChild *c = opaque; + /* + * We may have child->has_backup_perm unset at this point, as in case of + * _check_ stage of permission update failure we may _check_ not the whole + * subtree. Still, _abort_ is called on the whole subtree anyway. + */ + if (c->has_backup_perm) { + c->perm = c->backup_perm; + c->shared_perm = c->backup_shared_perm; + c->has_backup_perm = false; + } +} + +static TransactionActionDrv bdrv_child_set_pem_drv = { + .abort = bdrv_child_set_perm_abort, + .commit = bdrv_child_set_perm_commit, +}; + +/* + * With tran=NULL needs to be followed by direct call to either + * bdrv_child_set_perm_commit() or bdrv_child_set_perm_abort(). + * + * With non-NULL tran needs to be followed by tran_abort() or tran_commit() + * instead. + */ +static void bdrv_child_set_perm_safe(BdrvChild *c, uint64_t perm, + uint64_t shared, Transaction *tran) +{ + if (!c->has_backup_perm) { + c->has_backup_perm = true; + c->backup_perm = c->perm; + c->backup_shared_perm = c->shared_perm; + } + /* + * Note: it's OK if c->has_backup_perm was already set, as we can find the + * same c twice during check_perm procedure + */ + + c->perm = perm; + c->shared_perm = shared; + + if (tran) { + tran_add(tran, &bdrv_child_set_pem_drv, c); + } +} + /* * Check whether permissions on this node can be changed in a way that * @cumulative_perms and @cumulative_shared_perms are the new cumulative @@ -XXX,XX +XXX,XX @@ static int bdrv_child_check_perm(BdrvChild *c, BlockReopenQueue *q, return ret; } - if (!c->has_backup_perm) { - c->has_backup_perm = true; - c->backup_perm = c->perm; - c->backup_shared_perm = c->shared_perm; - } - /* - * Note: it's OK if c->has_backup_perm was already set, as we can find the - * same child twice during check_perm procedure - */ - - c->perm = perm; - c->shared_perm = shared; + bdrv_child_set_perm_safe(c, perm, shared, NULL); return 0; } static void bdrv_child_set_perm(BdrvChild *c) { - c->has_backup_perm = false; - + bdrv_child_set_perm_commit(c); bdrv_set_perm(c->bs); } static void bdrv_child_abort_perm_update(BdrvChild *c) { - if (c->has_backup_perm) { - c->perm = c->backup_perm; - c->shared_perm = c->backup_shared_perm; - c->has_backup_perm = false; - } - + bdrv_child_set_perm_abort(c); bdrv_abort_perm_update(c->bs); } -- 2.30.2 From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> We are going to drop recursive bdrv_child_* functions, so stop use them in bdrv_child_try_set_perm() as a first step. Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Kevin Wolf <kwolf@redhat.com> Message-Id: <20210428151804.439460-12-vsementsov@virtuozzo.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block.c | 14 ++++++++------ 1 file changed, 8 insertions(+), 6 deletions(-) diff --git a/block.c b/block.c index XXXXXXX..XXXXXXX 100644 --- a/block.c +++ b/block.c @@ -XXX,XX +XXX,XX @@ int bdrv_child_try_set_perm(BdrvChild *c, uint64_t perm, uint64_t shared, Error **errp) { Error *local_err = NULL; + Transaction *tran = tran_new(); int ret; - ret = bdrv_child_check_perm(c, NULL, perm, shared, NULL, &local_err); + bdrv_child_set_perm_safe(c, perm, shared, tran); + + ret = bdrv_refresh_perms(c->bs, &local_err); + + tran_finalize(tran, ret); + if (ret < 0) { - bdrv_child_abort_perm_update(c); if ((perm & ~c->perm) || (c->shared_perm & ~shared)) { /* tighten permissions */ error_propagate(errp, local_err); @@ -XXX,XX +XXX,XX @@ int bdrv_child_try_set_perm(BdrvChild *c, uint64_t perm, uint64_t shared, error_free(local_err); ret = 0; } - return ret; } - bdrv_child_set_perm(c); - - return 0; + return ret; } int bdrv_child_refresh_perms(BlockDriverState *bs, BdrvChild *c, Error **errp) -- 2.30.2 From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Each of them has only one caller. Open-coding simplifies further pemission-update system changes. Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Alberto Garcia <berto@igalia.com> Reviewed-by: Kevin Wolf <kwolf@redhat.com> Message-Id: <20210428151804.439460-13-vsementsov@virtuozzo.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block.c | 59 +++++++++++++++++---------------------------------------- 1 file changed, 17 insertions(+), 42 deletions(-) diff --git a/block.c b/block.c index XXXXXXX..XXXXXXX 100644 --- a/block.c +++ b/block.c @@ -XXX,XX +XXX,XX @@ static int bdrv_fill_options(QDict **options, const char *filename, return 0; } -static int bdrv_child_check_perm(BdrvChild *c, BlockReopenQueue *q, - uint64_t perm, uint64_t shared, - GSList *ignore_children, Error **errp); -static void bdrv_child_abort_perm_update(BdrvChild *c); -static void bdrv_child_set_perm(BdrvChild *c); +static int bdrv_check_update_perm(BlockDriverState *bs, BlockReopenQueue *q, + uint64_t new_used_perm, + uint64_t new_shared_perm, + GSList *ignore_children, + Error **errp); typedef struct BlockReopenQueueEntry { bool prepared; @@ -XXX,XX +XXX,XX @@ static int bdrv_check_perm(BlockDriverState *bs, BlockReopenQueue *q, /* Check all children */ QLIST_FOREACH(c, &bs->children, next) { uint64_t cur_perm, cur_shared; + GSList *cur_ignore_children; bdrv_child_perm(bs, c->bs, c, c->role, q, cumulative_perms, cumulative_shared_perms, &cur_perm, &cur_shared); - ret = bdrv_child_check_perm(c, q, cur_perm, cur_shared, ignore_children, - errp); + + cur_ignore_children = g_slist_prepend(g_slist_copy(ignore_children), c); + ret = bdrv_check_update_perm(c->bs, q, cur_perm, cur_shared, + cur_ignore_children, errp); + g_slist_free(cur_ignore_children); if (ret < 0) { return ret; } + + bdrv_child_set_perm_safe(c, cur_perm, cur_shared, NULL); } return 0; @@ -XXX,XX +XXX,XX @@ static void bdrv_abort_perm_update(BlockDriverState *bs) } QLIST_FOREACH(c, &bs->children, next) { - bdrv_child_abort_perm_update(c); + bdrv_child_set_perm_abort(c); + bdrv_abort_perm_update(c->bs); } } @@ -XXX,XX +XXX,XX @@ static void bdrv_set_perm(BlockDriverState *bs) /* Update all children */ QLIST_FOREACH(c, &bs->children, next) { - bdrv_child_set_perm(c); + bdrv_child_set_perm_commit(c); + bdrv_set_perm(c->bs); } } @@ -XXX,XX +XXX,XX @@ static int bdrv_check_update_perm(BlockDriverState *bs, BlockReopenQueue *q, ignore_children, errp); } -/* Needs to be followed by a call to either bdrv_child_set_perm() or - * bdrv_child_abort_perm_update(). */ -static int bdrv_child_check_perm(BdrvChild *c, BlockReopenQueue *q, - uint64_t perm, uint64_t shared, - GSList *ignore_children, Error **errp) -{ - int ret; - - ignore_children = g_slist_prepend(g_slist_copy(ignore_children), c); - ret = bdrv_check_update_perm(c->bs, q, perm, shared, ignore_children, errp); - g_slist_free(ignore_children); - - if (ret < 0) { - return ret; - } - - bdrv_child_set_perm_safe(c, perm, shared, NULL); - - return 0; -} - -static void bdrv_child_set_perm(BdrvChild *c) -{ - bdrv_child_set_perm_commit(c); - bdrv_set_perm(c->bs); -} - -static void bdrv_child_abort_perm_update(BdrvChild *c) -{ - bdrv_child_set_perm_abort(c); - bdrv_abort_perm_update(c->bs); -} - static int bdrv_refresh_perms(BlockDriverState *bs, Error **errp) { int ret; -- 2.30.2 From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Rewrite bdrv_check_perm(), bdrv_abort_perm_update() and bdrv_set_perm() to update nodes in topological sort order instead of simple DFS. With topologically sorted nodes, we update a node only when all its parents already updated. With DFS it's not so. Consider the following example: A -+ | | | v | B | | v | C<-+ A is parent for B and C, B is parent for C. Obviously, to update permissions, we should go in order A B C, so, when we update C, all parent permissions already updated. But with current approach (simple recursion) we can update in sequence A C B C (C is updated twice). On first update of C, we consider old B permissions, so doing wrong thing. If it succeed, all is OK, on second C update we will finish with correct graph. But if the wrong thing failed, we break the whole process for no reason (it's possible that updated B permission will be less strict, but we will never check it). Also new approach gives a way to simultaneously and correctly update several nodes, we just need to run bdrv_topological_dfs() several times to add all nodes and their subtrees into one topologically sorted list (next patch will update bdrv_replace_node() in this manner). Test test_parallel_perm_update() is now passing, so move it out of debugging "if". We also need to support ignore_children in bdrv_parent_perms_conflict() For test 283 order of conflicting parents check is changed. Note also that in bdrv_check_perm() we don't check for parents conflict at root bs, as we may be in the middle of permission update in bdrv_reopen_multiple(). bdrv_reopen_multiple() will be updated soon. Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Kevin Wolf <kwolf@redhat.com> Message-Id: <20210428151804.439460-14-vsementsov@virtuozzo.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block.c | 116 +++++++++++++++++++++++++------ tests/unit/test-bdrv-graph-mod.c | 4 +- tests/qemu-iotests/283.out | 2 +- 3 files changed, 99 insertions(+), 23 deletions(-) diff --git a/block.c b/block.c index XXXXXXX..XXXXXXX 100644 --- a/block.c +++ b/block.c @@ -XXX,XX +XXX,XX @@ static bool bdrv_a_allow_b(BdrvChild *a, BdrvChild *b, Error **errp) return false; } -static bool bdrv_parent_perms_conflict(BlockDriverState *bs, Error **errp) +static bool bdrv_parent_perms_conflict(BlockDriverState *bs, + GSList *ignore_children, + Error **errp) { BdrvChild *a, *b; @@ -XXX,XX +XXX,XX @@ static bool bdrv_parent_perms_conflict(BlockDriverState *bs, Error **errp) * directions. */ QLIST_FOREACH(a, &bs->parents, next_parent) { + if (g_slist_find(ignore_children, a)) { + continue; + } + QLIST_FOREACH(b, &bs->parents, next_parent) { - if (a == b) { + if (a == b || g_slist_find(ignore_children, b)) { continue; } @@ -XXX,XX +XXX,XX @@ static void bdrv_child_perm(BlockDriverState *bs, BlockDriverState *child_bs, } } +/* + * Adds the whole subtree of @bs (including @bs itself) to the @list (except for + * nodes that are already in the @list, of course) so that final list is + * topologically sorted. Return the result (GSList @list object is updated, so + * don't use old reference after function call). + * + * On function start @list must be already topologically sorted and for any node + * in the @list the whole subtree of the node must be in the @list as well. The + * simplest way to satisfy this criteria: use only result of + * bdrv_topological_dfs() or NULL as @list parameter. + */ +static GSList *bdrv_topological_dfs(GSList *list, GHashTable *found, + BlockDriverState *bs) +{ + BdrvChild *child; + g_autoptr(GHashTable) local_found = NULL; + + if (!found) { + assert(!list); + found = local_found = g_hash_table_new(NULL, NULL); + } + + if (g_hash_table_contains(found, bs)) { + return list; + } + g_hash_table_add(found, bs); + + QLIST_FOREACH(child, &bs->children, next) { + list = bdrv_topological_dfs(list, found, child->bs); + } + + return g_slist_prepend(list, bs); +} + static void bdrv_child_set_perm_commit(void *opaque) { BdrvChild *c = opaque; @@ -XXX,XX +XXX,XX @@ static void bdrv_child_set_perm_safe(BdrvChild *c, uint64_t perm, * A call to this function must always be followed by a call to bdrv_set_perm() * or bdrv_abort_perm_update(). */ -static int bdrv_check_perm(BlockDriverState *bs, BlockReopenQueue *q, - uint64_t cumulative_perms, - uint64_t cumulative_shared_perms, - GSList *ignore_children, Error **errp) +static int bdrv_node_check_perm(BlockDriverState *bs, BlockReopenQueue *q, + uint64_t cumulative_perms, + uint64_t cumulative_shared_perms, + GSList *ignore_children, Error **errp) { BlockDriver *drv = bs->drv; BdrvChild *c; @@ -XXX,XX +XXX,XX @@ static int bdrv_check_perm(BlockDriverState *bs, BlockReopenQueue *q, /* Check all children */ QLIST_FOREACH(c, &bs->children, next) { uint64_t cur_perm, cur_shared; - GSList *cur_ignore_children; bdrv_child_perm(bs, c->bs, c, c->role, q, cumulative_perms, cumulative_shared_perms, &cur_perm, &cur_shared); + bdrv_child_set_perm_safe(c, cur_perm, cur_shared, NULL); + } + + return 0; +} + +static int bdrv_check_perm(BlockDriverState *bs, BlockReopenQueue *q, + uint64_t cumulative_perms, + uint64_t cumulative_shared_perms, + GSList *ignore_children, Error **errp) +{ + int ret; + BlockDriverState *root = bs; + g_autoptr(GSList) list = bdrv_topological_dfs(NULL, NULL, root); - cur_ignore_children = g_slist_prepend(g_slist_copy(ignore_children), c); - ret = bdrv_check_update_perm(c->bs, q, cur_perm, cur_shared, - cur_ignore_children, errp); - g_slist_free(cur_ignore_children); + for ( ; list; list = list->next) { + bs = list->data; + + if (bs != root) { + if (bdrv_parent_perms_conflict(bs, ignore_children, errp)) { + return -EINVAL; + } + + bdrv_get_cumulative_perm(bs, &cumulative_perms, + &cumulative_shared_perms); + } + + ret = bdrv_node_check_perm(bs, q, cumulative_perms, + cumulative_shared_perms, + ignore_children, errp); if (ret < 0) { return ret; } - - bdrv_child_set_perm_safe(c, cur_perm, cur_shared, NULL); } return 0; @@ -XXX,XX +XXX,XX @@ static int bdrv_check_perm(BlockDriverState *bs, BlockReopenQueue *q, * Notifies drivers that after a previous bdrv_check_perm() call, the * permission update is not performed and any preparations made for it (e.g. * taken file locks) need to be undone. - * - * This function recursively notifies all child nodes. */ -static void bdrv_abort_perm_update(BlockDriverState *bs) +static void bdrv_node_abort_perm_update(BlockDriverState *bs) { BlockDriver *drv = bs->drv; BdrvChild *c; @@ -XXX,XX +XXX,XX @@ static void bdrv_abort_perm_update(BlockDriverState *bs) QLIST_FOREACH(c, &bs->children, next) { bdrv_child_set_perm_abort(c); - bdrv_abort_perm_update(c->bs); } } -static void bdrv_set_perm(BlockDriverState *bs) +static void bdrv_abort_perm_update(BlockDriverState *bs) +{ + g_autoptr(GSList) list = bdrv_topological_dfs(NULL, NULL, bs); + + for ( ; list; list = list->next) { + bdrv_node_abort_perm_update((BlockDriverState *)list->data); + } +} + +static void bdrv_node_set_perm(BlockDriverState *bs) { uint64_t cumulative_perms, cumulative_shared_perms; BlockDriver *drv = bs->drv; @@ -XXX,XX +XXX,XX @@ static void bdrv_set_perm(BlockDriverState *bs) /* Update all children */ QLIST_FOREACH(c, &bs->children, next) { bdrv_child_set_perm_commit(c); - bdrv_set_perm(c->bs); + } +} + +static void bdrv_set_perm(BlockDriverState *bs) +{ + g_autoptr(GSList) list = bdrv_topological_dfs(NULL, NULL, bs); + + for ( ; list; list = list->next) { + bdrv_node_set_perm((BlockDriverState *)list->data); } } @@ -XXX,XX +XXX,XX @@ static int bdrv_refresh_perms(BlockDriverState *bs, Error **errp) int ret; uint64_t perm, shared_perm; - if (bdrv_parent_perms_conflict(bs, errp)) { + if (bdrv_parent_perms_conflict(bs, NULL, errp)) { return -EPERM; } bdrv_get_cumulative_perm(bs, &perm, &shared_perm); diff --git a/tests/unit/test-bdrv-graph-mod.c b/tests/unit/test-bdrv-graph-mod.c index XXXXXXX..XXXXXXX 100644 --- a/tests/unit/test-bdrv-graph-mod.c +++ b/tests/unit/test-bdrv-graph-mod.c @@ -XXX,XX +XXX,XX @@ int main(int argc, char *argv[]) g_test_add_func("/bdrv-graph-mod/update-perm-tree", test_update_perm_tree); g_test_add_func("/bdrv-graph-mod/should-update-child", test_should_update_child); + g_test_add_func("/bdrv-graph-mod/parallel-perm-update", + test_parallel_perm_update); if (debug) { g_test_add_func("/bdrv-graph-mod/parallel-exclusive-write", test_parallel_exclusive_write); - g_test_add_func("/bdrv-graph-mod/parallel-perm-update", - test_parallel_perm_update); g_test_add_func("/bdrv-graph-mod/append-greedy-filter", test_append_greedy_filter); } diff --git a/tests/qemu-iotests/283.out b/tests/qemu-iotests/283.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/283.out +++ b/tests/qemu-iotests/283.out @@ -XXX,XX +XXX,XX @@ {"execute": "blockdev-add", "arguments": {"driver": "blkdebug", "image": "base", "node-name": "other", "take-child-perms": ["write"]}} {"return": {}} {"execute": "blockdev-backup", "arguments": {"device": "source", "sync": "full", "target": "target"}} -{"error": {"class": "GenericError", "desc": "Cannot set permissions for backup-top filter: Conflicts with use by other as 'image', which uses 'write' on base"}} +{"error": {"class": "GenericError", "desc": "Cannot set permissions for backup-top filter: Conflicts with use by source as 'image', which does not allow 'write' on base"}} === backup-top should be gone after job-finalize === -- 2.30.2 From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Refactor calling driver callbacks to a separate transaction action to be used later. Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Kevin Wolf <kwolf@redhat.com> Message-Id: <20210428151804.439460-15-vsementsov@virtuozzo.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block.c | 70 ++++++++++++++++++++++++++++++++++++++++++++------------- 1 file changed, 54 insertions(+), 16 deletions(-) diff --git a/block.c b/block.c index XXXXXXX..XXXXXXX 100644 --- a/block.c +++ b/block.c @@ -XXX,XX +XXX,XX @@ static void bdrv_child_set_perm_safe(BdrvChild *c, uint64_t perm, } } +static void bdrv_drv_set_perm_commit(void *opaque) +{ + BlockDriverState *bs = opaque; + uint64_t cumulative_perms, cumulative_shared_perms; + + if (bs->drv->bdrv_set_perm) { + bdrv_get_cumulative_perm(bs, &cumulative_perms, + &cumulative_shared_perms); + bs->drv->bdrv_set_perm(bs, cumulative_perms, cumulative_shared_perms); + } +} + +static void bdrv_drv_set_perm_abort(void *opaque) +{ + BlockDriverState *bs = opaque; + + if (bs->drv->bdrv_abort_perm_update) { + bs->drv->bdrv_abort_perm_update(bs); + } +} + +TransactionActionDrv bdrv_drv_set_perm_drv = { + .abort = bdrv_drv_set_perm_abort, + .commit = bdrv_drv_set_perm_commit, +}; + +static int bdrv_drv_set_perm(BlockDriverState *bs, uint64_t perm, + uint64_t shared_perm, Transaction *tran, + Error **errp) +{ + if (!bs->drv) { + return 0; + } + + if (bs->drv->bdrv_check_perm) { + int ret = bs->drv->bdrv_check_perm(bs, perm, shared_perm, errp); + if (ret < 0) { + return ret; + } + } + + if (tran) { + tran_add(tran, &bdrv_drv_set_perm_drv, bs); + } + + return 0; +} + /* * Check whether permissions on this node can be changed in a way that * @cumulative_perms and @cumulative_shared_perms are the new cumulative @@ -XXX,XX +XXX,XX @@ static int bdrv_node_check_perm(BlockDriverState *bs, BlockReopenQueue *q, return 0; } - if (drv->bdrv_check_perm) { - ret = drv->bdrv_check_perm(bs, cumulative_perms, - cumulative_shared_perms, errp); - if (ret < 0) { - return ret; - } + ret = bdrv_drv_set_perm(bs, cumulative_perms, cumulative_shared_perms, NULL, + errp); + if (ret < 0) { + return ret; } /* Drivers that never have children can omit .bdrv_child_perm() */ @@ -XXX,XX +XXX,XX @@ static void bdrv_node_abort_perm_update(BlockDriverState *bs) return; } - if (drv->bdrv_abort_perm_update) { - drv->bdrv_abort_perm_update(bs); - } + bdrv_drv_set_perm_abort(bs); QLIST_FOREACH(c, &bs->children, next) { bdrv_child_set_perm_abort(c); @@ -XXX,XX +XXX,XX @@ static void bdrv_abort_perm_update(BlockDriverState *bs) static void bdrv_node_set_perm(BlockDriverState *bs) { - uint64_t cumulative_perms, cumulative_shared_perms; BlockDriver *drv = bs->drv; BdrvChild *c; @@ -XXX,XX +XXX,XX @@ static void bdrv_node_set_perm(BlockDriverState *bs) return; } - bdrv_get_cumulative_perm(bs, &cumulative_perms, &cumulative_shared_perms); - - /* Update this node */ - if (drv->bdrv_set_perm) { - drv->bdrv_set_perm(bs, cumulative_perms, cumulative_shared_perms); - } + bdrv_drv_set_perm_commit(bs); /* Drivers that never have children can omit .bdrv_child_perm() */ if (!drv->bdrv_child_perm) { -- 2.30.2 From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Add new interface, allowing use of existing node list. It will be used to fix bdrv_replace_node() in the further commit. Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Kevin Wolf <kwolf@redhat.com> Message-Id: <20210428151804.439460-16-vsementsov@virtuozzo.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block.c | 106 +++++++++++++++++++++++++++++++++++++------------------- 1 file changed, 71 insertions(+), 35 deletions(-) diff --git a/block.c b/block.c index XXXXXXX..XXXXXXX 100644 --- a/block.c +++ b/block.c @@ -XXX,XX +XXX,XX @@ static int bdrv_drv_set_perm(BlockDriverState *bs, uint64_t perm, static int bdrv_node_check_perm(BlockDriverState *bs, BlockReopenQueue *q, uint64_t cumulative_perms, uint64_t cumulative_shared_perms, - GSList *ignore_children, Error **errp) + GSList *ignore_children, + Transaction *tran, Error **errp) { BlockDriver *drv = bs->drv; BdrvChild *c; @@ -XXX,XX +XXX,XX @@ static int bdrv_node_check_perm(BlockDriverState *bs, BlockReopenQueue *q, return 0; } - ret = bdrv_drv_set_perm(bs, cumulative_perms, cumulative_shared_perms, NULL, + ret = bdrv_drv_set_perm(bs, cumulative_perms, cumulative_shared_perms, tran, errp); if (ret < 0) { return ret; @@ -XXX,XX +XXX,XX @@ static int bdrv_node_check_perm(BlockDriverState *bs, BlockReopenQueue *q, bdrv_child_perm(bs, c->bs, c, c->role, q, cumulative_perms, cumulative_shared_perms, &cur_perm, &cur_shared); - bdrv_child_set_perm_safe(c, cur_perm, cur_shared, NULL); + bdrv_child_set_perm_safe(c, cur_perm, cur_shared, tran); } return 0; } -static int bdrv_check_perm(BlockDriverState *bs, BlockReopenQueue *q, - uint64_t cumulative_perms, - uint64_t cumulative_shared_perms, - GSList *ignore_children, Error **errp) +/* + * If use_cumulative_perms is true, use cumulative_perms and + * cumulative_shared_perms for first element of the list. Otherwise just refresh + * all permissions. + */ +static int bdrv_check_perm_common(GSList *list, BlockReopenQueue *q, + bool use_cumulative_perms, + uint64_t cumulative_perms, + uint64_t cumulative_shared_perms, + GSList *ignore_children, + Transaction *tran, Error **errp) { int ret; - BlockDriverState *root = bs; - g_autoptr(GSList) list = bdrv_topological_dfs(NULL, NULL, root); + BlockDriverState *bs; - for ( ; list; list = list->next) { + if (use_cumulative_perms) { bs = list->data; - if (bs != root) { - if (bdrv_parent_perms_conflict(bs, ignore_children, errp)) { - return -EINVAL; - } + ret = bdrv_node_check_perm(bs, q, cumulative_perms, + cumulative_shared_perms, + ignore_children, tran, errp); + if (ret < 0) { + return ret; + } - bdrv_get_cumulative_perm(bs, &cumulative_perms, - &cumulative_shared_perms); + list = list->next; + } + + for ( ; list; list = list->next) { + bs = list->data; + + if (bdrv_parent_perms_conflict(bs, ignore_children, errp)) { + return -EINVAL; } + bdrv_get_cumulative_perm(bs, &cumulative_perms, + &cumulative_shared_perms); + ret = bdrv_node_check_perm(bs, q, cumulative_perms, cumulative_shared_perms, - ignore_children, errp); + ignore_children, tran, errp); if (ret < 0) { return ret; } @@ -XXX,XX +XXX,XX @@ static int bdrv_check_perm(BlockDriverState *bs, BlockReopenQueue *q, return 0; } +static int bdrv_check_perm(BlockDriverState *bs, BlockReopenQueue *q, + uint64_t cumulative_perms, + uint64_t cumulative_shared_perms, + GSList *ignore_children, Error **errp) +{ + g_autoptr(GSList) list = bdrv_topological_dfs(NULL, NULL, bs); + return bdrv_check_perm_common(list, q, true, cumulative_perms, + cumulative_shared_perms, ignore_children, + NULL, errp); +} + +static int bdrv_list_refresh_perms(GSList *list, BlockReopenQueue *q, + Transaction *tran, Error **errp) +{ + return bdrv_check_perm_common(list, q, false, 0, 0, NULL, tran, errp); +} + /* * Notifies drivers that after a previous bdrv_check_perm() call, the * permission update is not performed and any preparations made for it (e.g. @@ -XXX,XX +XXX,XX @@ static void bdrv_node_abort_perm_update(BlockDriverState *bs) } } -static void bdrv_abort_perm_update(BlockDriverState *bs) +static void bdrv_list_abort_perm_update(GSList *list) { - g_autoptr(GSList) list = bdrv_topological_dfs(NULL, NULL, bs); - for ( ; list; list = list->next) { bdrv_node_abort_perm_update((BlockDriverState *)list->data); } } +static void bdrv_abort_perm_update(BlockDriverState *bs) +{ + g_autoptr(GSList) list = bdrv_topological_dfs(NULL, NULL, bs); + return bdrv_list_abort_perm_update(list); +} + static void bdrv_node_set_perm(BlockDriverState *bs) { BlockDriver *drv = bs->drv; @@ -XXX,XX +XXX,XX @@ static void bdrv_node_set_perm(BlockDriverState *bs) } } -static void bdrv_set_perm(BlockDriverState *bs) +static void bdrv_list_set_perm(GSList *list) { - g_autoptr(GSList) list = bdrv_topological_dfs(NULL, NULL, bs); - for ( ; list; list = list->next) { bdrv_node_set_perm((BlockDriverState *)list->data); } } +static void bdrv_set_perm(BlockDriverState *bs) +{ + g_autoptr(GSList) list = bdrv_topological_dfs(NULL, NULL, bs); + return bdrv_list_set_perm(list); +} + void bdrv_get_cumulative_perm(BlockDriverState *bs, uint64_t *perm, uint64_t *shared_perm) { @@ -XXX,XX +XXX,XX @@ static int bdrv_check_update_perm(BlockDriverState *bs, BlockReopenQueue *q, static int bdrv_refresh_perms(BlockDriverState *bs, Error **errp) { int ret; - uint64_t perm, shared_perm; + Transaction *tran = tran_new(); + g_autoptr(GSList) list = bdrv_topological_dfs(NULL, NULL, bs); - if (bdrv_parent_perms_conflict(bs, NULL, errp)) { - return -EPERM; - } - bdrv_get_cumulative_perm(bs, &perm, &shared_perm); - ret = bdrv_check_perm(bs, NULL, perm, shared_perm, NULL, errp); - if (ret < 0) { - bdrv_abort_perm_update(bs); - return ret; - } - bdrv_set_perm(bs); + ret = bdrv_list_refresh_perms(list, NULL, tran, errp); + tran_finalize(tran, ret); - return 0; + return ret; } int bdrv_child_try_set_perm(BdrvChild *c, uint64_t perm, uint64_t shared, -- 2.30.2 From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> To be used in the following commit. Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Kevin Wolf <kwolf@redhat.com> Message-Id: <20210428151804.439460-17-vsementsov@virtuozzo.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block.c | 54 ++++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 54 insertions(+) diff --git a/block.c b/block.c index XXXXXXX..XXXXXXX 100644 --- a/block.c +++ b/block.c @@ -XXX,XX +XXX,XX @@ static BlockDriverState *bdrv_open_inherit(const char *filename, BdrvChildRole child_role, Error **errp); +static void bdrv_replace_child_noperm(BdrvChild *child, + BlockDriverState *new_bs); + static int bdrv_reopen_prepare(BDRVReopenState *reopen_state, BlockReopenQueue *queue, Error **errp); static void bdrv_reopen_commit(BDRVReopenState *reopen_state); @@ -XXX,XX +XXX,XX @@ static int bdrv_drv_set_perm(BlockDriverState *bs, uint64_t perm, return 0; } +typedef struct BdrvReplaceChildState { + BdrvChild *child; + BlockDriverState *old_bs; +} BdrvReplaceChildState; + +static void bdrv_replace_child_commit(void *opaque) +{ + BdrvReplaceChildState *s = opaque; + + bdrv_unref(s->old_bs); +} + +static void bdrv_replace_child_abort(void *opaque) +{ + BdrvReplaceChildState *s = opaque; + BlockDriverState *new_bs = s->child->bs; + + /* old_bs reference is transparently moved from @s to @s->child */ + bdrv_replace_child_noperm(s->child, s->old_bs); + bdrv_unref(new_bs); +} + +static TransactionActionDrv bdrv_replace_child_drv = { + .commit = bdrv_replace_child_commit, + .abort = bdrv_replace_child_abort, + .clean = g_free, +}; + +/* + * bdrv_replace_child_safe + * + * Note: real unref of old_bs is done only on commit. + */ +__attribute__((unused)) +static void bdrv_replace_child_safe(BdrvChild *child, BlockDriverState *new_bs, + Transaction *tran) +{ + BdrvReplaceChildState *s = g_new(BdrvReplaceChildState, 1); + *s = (BdrvReplaceChildState) { + .child = child, + .old_bs = child->bs, + }; + tran_add(tran, &bdrv_replace_child_drv, s); + + if (new_bs) { + bdrv_ref(new_bs); + } + bdrv_replace_child_noperm(child, new_bs); + /* old_bs reference is transparently moved from @child to @s */ +} + /* * Check whether permissions on this node can be changed in a way that * @cumulative_perms and @cumulative_shared_perms are the new cumulative -- 2.30.2 From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> inore_children thing doesn't help to track all propagated permissions of children we want to ignore. The simplest way to correctly update permissions is update graph first and then do permission update. In this case we just referesh permissions for the whole subgraph (in topological-sort defined order) and everything is correctly calculated automatically without any ignore_children. So, refactor bdrv_replace_node_common to first do graph update and then refresh the permissions. Test test_parallel_exclusive_write() now pass, so move it out of debugging "if". Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Kevin Wolf <kwolf@redhat.com> Message-Id: <20210428151804.439460-18-vsementsov@virtuozzo.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block.c | 43 +++++++++++++------------------- tests/unit/test-bdrv-graph-mod.c | 4 +-- 2 files changed, 20 insertions(+), 27 deletions(-) diff --git a/block.c b/block.c index XXXXXXX..XXXXXXX 100644 --- a/block.c +++ b/block.c @@ -XXX,XX +XXX,XX @@ static TransactionActionDrv bdrv_replace_child_drv = { * * Note: real unref of old_bs is done only on commit. */ -__attribute__((unused)) static void bdrv_replace_child_safe(BdrvChild *child, BlockDriverState *new_bs, Transaction *tran) { @@ -XXX,XX +XXX,XX @@ static int bdrv_replace_node_common(BlockDriverState *from, bool auto_skip, Error **errp) { BdrvChild *c, *next; - GSList *list = NULL, *p; - uint64_t perm = 0, shared = BLK_PERM_ALL; + Transaction *tran = tran_new(); + g_autoptr(GHashTable) found = NULL; + g_autoptr(GSList) refresh_list = NULL; int ret; /* Make sure that @from doesn't go away until we have successfully attached @@ -XXX,XX +XXX,XX @@ static int bdrv_replace_node_common(BlockDriverState *from, assert(bdrv_get_aio_context(from) == bdrv_get_aio_context(to)); bdrv_drained_begin(from); - /* Put all parents into @list and calculate their cumulative permissions */ + /* + * Do the replacement without permission update. + * Replacement may influence the permissions, we should calculate new + * permissions based on new graph. If we fail, we'll roll-back the + * replacement. + */ QLIST_FOREACH_SAFE(c, &from->parents, next_parent, next) { assert(c->bs == from); if (!should_update_child(c, to)) { @@ -XXX,XX +XXX,XX @@ static int bdrv_replace_node_common(BlockDriverState *from, c->name, from->node_name); goto out; } - list = g_slist_prepend(list, c); - perm |= c->perm; - shared &= c->shared_perm; + bdrv_replace_child_safe(c, to, tran); } - /* Check whether the required permissions can be granted on @to, ignoring - * all BdrvChild in @list so that they can't block themselves. */ - ret = bdrv_check_update_perm(to, NULL, perm, shared, list, errp); - if (ret < 0) { - bdrv_abort_perm_update(to); - goto out; - } + found = g_hash_table_new(NULL, NULL); - /* Now actually perform the change. We performed the permission check for - * all elements of @list at once, so set the permissions all at once at the - * very end. */ - for (p = list; p != NULL; p = p->next) { - c = p->data; + refresh_list = bdrv_topological_dfs(refresh_list, found, to); + refresh_list = bdrv_topological_dfs(refresh_list, found, from); - bdrv_ref(to); - bdrv_replace_child_noperm(c, to); - bdrv_unref(from); + ret = bdrv_list_refresh_perms(refresh_list, NULL, tran, errp); + if (ret < 0) { + goto out; } - bdrv_set_perm(to); - ret = 0; out: - g_slist_free(list); + tran_finalize(tran, ret); + bdrv_drained_end(from); bdrv_unref(from); diff --git a/tests/unit/test-bdrv-graph-mod.c b/tests/unit/test-bdrv-graph-mod.c index XXXXXXX..XXXXXXX 100644 --- a/tests/unit/test-bdrv-graph-mod.c +++ b/tests/unit/test-bdrv-graph-mod.c @@ -XXX,XX +XXX,XX @@ int main(int argc, char *argv[]) test_should_update_child); g_test_add_func("/bdrv-graph-mod/parallel-perm-update", test_parallel_perm_update); + g_test_add_func("/bdrv-graph-mod/parallel-exclusive-write", + test_parallel_exclusive_write); if (debug) { - g_test_add_func("/bdrv-graph-mod/parallel-exclusive-write", - test_parallel_exclusive_write); g_test_add_func("/bdrv-graph-mod/append-greedy-filter", test_append_greedy_filter); } -- 2.30.2 From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Split out no-perm part of bdrv_root_attach_child() into separate transaction action. bdrv_root_attach_child() now moves to new permission update paradigm: first update graph relations then update permissions. qsd-jobs test output updated. Seems now permission update goes in another order. Still, the test comment say that we only want to check that command doesn't crash, and it's still so. Error message is a bit misleading as it looks like job was added first. But actually in new paradigm of graph update we can't distinguish such things. We should update the error message, but let's not do it now. Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Message-Id: <20210428151804.439460-19-vsementsov@virtuozzo.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block.c | 190 ++++++++++++++++++-------- tests/qemu-iotests/tests/qsd-jobs.out | 2 +- 2 files changed, 137 insertions(+), 55 deletions(-) diff --git a/block.c b/block.c index XXXXXXX..XXXXXXX 100644 --- a/block.c +++ b/block.c @@ -XXX,XX +XXX,XX @@ static void bdrv_replace_child(BdrvChild *child, BlockDriverState *new_bs) } } -/* - * This function steals the reference to child_bs from the caller. - * That reference is later dropped by bdrv_root_unref_child(). - * - * On failure NULL is returned, errp is set and the reference to - * child_bs is also dropped. - * - * The caller must hold the AioContext lock @child_bs, but not that of @ctx - * (unless @child_bs is already in @ctx). - */ -BdrvChild *bdrv_root_attach_child(BlockDriverState *child_bs, - const char *child_name, - const BdrvChildClass *child_class, - BdrvChildRole child_role, - uint64_t perm, uint64_t shared_perm, - void *opaque, Error **errp) +static void bdrv_remove_empty_child(BdrvChild *child) { - BdrvChild *child; - Error *local_err = NULL; - int ret; - AioContext *ctx; + assert(!child->bs); + QLIST_SAFE_REMOVE(child, next); + g_free(child->name); + g_free(child); +} - ret = bdrv_check_update_perm(child_bs, NULL, perm, shared_perm, NULL, errp); - if (ret < 0) { - bdrv_abort_perm_update(child_bs); - bdrv_unref(child_bs); - return NULL; +typedef struct BdrvAttachChildCommonState { + BdrvChild **child; + AioContext *old_parent_ctx; + AioContext *old_child_ctx; +} BdrvAttachChildCommonState; + +static void bdrv_attach_child_common_abort(void *opaque) +{ + BdrvAttachChildCommonState *s = opaque; + BdrvChild *child = *s->child; + BlockDriverState *bs = child->bs; + + bdrv_replace_child_noperm(child, NULL); + + if (bdrv_get_aio_context(bs) != s->old_child_ctx) { + bdrv_try_set_aio_context(bs, s->old_child_ctx, &error_abort); } - child = g_new(BdrvChild, 1); - *child = (BdrvChild) { + if (bdrv_child_get_parent_aio_context(child) != s->old_parent_ctx) { + GSList *ignore = g_slist_prepend(NULL, child); + + child->klass->can_set_aio_ctx(child, s->old_parent_ctx, &ignore, + &error_abort); + g_slist_free(ignore); + ignore = g_slist_prepend(NULL, child); + child->klass->set_aio_ctx(child, s->old_parent_ctx, &ignore); + + g_slist_free(ignore); + } + + bdrv_unref(bs); + bdrv_remove_empty_child(child); + *s->child = NULL; +} + +static TransactionActionDrv bdrv_attach_child_common_drv = { + .abort = bdrv_attach_child_common_abort, + .clean = g_free, +}; + +/* + * Common part of attaching bdrv child to bs or to blk or to job + */ +static int bdrv_attach_child_common(BlockDriverState *child_bs, + const char *child_name, + const BdrvChildClass *child_class, + BdrvChildRole child_role, + uint64_t perm, uint64_t shared_perm, + void *opaque, BdrvChild **child, + Transaction *tran, Error **errp) +{ + BdrvChild *new_child; + AioContext *parent_ctx; + AioContext *child_ctx = bdrv_get_aio_context(child_bs); + + assert(child); + assert(*child == NULL); + + new_child = g_new(BdrvChild, 1); + *new_child = (BdrvChild) { .bs = NULL, .name = g_strdup(child_name), .klass = child_class, @@ -XXX,XX +XXX,XX @@ BdrvChild *bdrv_root_attach_child(BlockDriverState *child_bs, .opaque = opaque, }; - ctx = bdrv_child_get_parent_aio_context(child); - - /* If the AioContexts don't match, first try to move the subtree of + /* + * If the AioContexts don't match, first try to move the subtree of * child_bs into the AioContext of the new parent. If this doesn't work, - * try moving the parent into the AioContext of child_bs instead. */ - if (bdrv_get_aio_context(child_bs) != ctx) { - ret = bdrv_try_set_aio_context(child_bs, ctx, &local_err); + * try moving the parent into the AioContext of child_bs instead. + */ + parent_ctx = bdrv_child_get_parent_aio_context(new_child); + if (child_ctx != parent_ctx) { + Error *local_err = NULL; + int ret = bdrv_try_set_aio_context(child_bs, parent_ctx, &local_err); + if (ret < 0 && child_class->can_set_aio_ctx) { - GSList *ignore = g_slist_prepend(NULL, child); - ctx = bdrv_get_aio_context(child_bs); - if (child_class->can_set_aio_ctx(child, ctx, &ignore, NULL)) { + GSList *ignore = g_slist_prepend(NULL, new_child); + if (child_class->can_set_aio_ctx(new_child, child_ctx, &ignore, + NULL)) + { error_free(local_err); ret = 0; g_slist_free(ignore); - ignore = g_slist_prepend(NULL, child); - child_class->set_aio_ctx(child, ctx, &ignore); + ignore = g_slist_prepend(NULL, new_child); + child_class->set_aio_ctx(new_child, child_ctx, &ignore); } g_slist_free(ignore); } + if (ret < 0) { error_propagate(errp, local_err); - g_free(child); - bdrv_abort_perm_update(child_bs); - bdrv_unref(child_bs); - return NULL; + bdrv_remove_empty_child(new_child); + return ret; } } - /* This performs the matching bdrv_set_perm() for the above check. */ - bdrv_replace_child(child, child_bs); + bdrv_ref(child_bs); + bdrv_replace_child_noperm(new_child, child_bs); + *child = new_child; + + BdrvAttachChildCommonState *s = g_new(BdrvAttachChildCommonState, 1); + *s = (BdrvAttachChildCommonState) { + .child = child, + .old_parent_ctx = parent_ctx, + .old_child_ctx = child_ctx, + }; + tran_add(tran, &bdrv_attach_child_common_drv, s); + + return 0; +} + +static void bdrv_detach_child(BdrvChild *child) +{ + bdrv_replace_child(child, NULL); + bdrv_remove_empty_child(child); +} + +/* + * This function steals the reference to child_bs from the caller. + * That reference is later dropped by bdrv_root_unref_child(). + * + * On failure NULL is returned, errp is set and the reference to + * child_bs is also dropped. + * + * The caller must hold the AioContext lock @child_bs, but not that of @ctx + * (unless @child_bs is already in @ctx). + */ +BdrvChild *bdrv_root_attach_child(BlockDriverState *child_bs, + const char *child_name, + const BdrvChildClass *child_class, + BdrvChildRole child_role, + uint64_t perm, uint64_t shared_perm, + void *opaque, Error **errp) +{ + int ret; + BdrvChild *child = NULL; + Transaction *tran = tran_new(); + + ret = bdrv_attach_child_common(child_bs, child_name, child_class, + child_role, perm, shared_perm, opaque, + &child, tran, errp); + if (ret < 0) { + bdrv_unref(child_bs); + return NULL; + } + + ret = bdrv_refresh_perms(child_bs, errp); + tran_finalize(tran, ret); + + bdrv_unref(child_bs); return child; } @@ -XXX,XX +XXX,XX @@ BdrvChild *bdrv_attach_child(BlockDriverState *parent_bs, return child; } -static void bdrv_detach_child(BdrvChild *child) -{ - QLIST_SAFE_REMOVE(child, next); - - bdrv_replace_child(child, NULL); - - g_free(child->name); - g_free(child); -} - /* Callers must ensure that child->frozen is false. */ void bdrv_root_unref_child(BdrvChild *child) { diff --git a/tests/qemu-iotests/tests/qsd-jobs.out b/tests/qemu-iotests/tests/qsd-jobs.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/tests/qsd-jobs.out +++ b/tests/qemu-iotests/tests/qsd-jobs.out @@ -XXX,XX +XXX,XX @@ QMP_VERSION {"return": {}} {"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "created", "id": "job0"}} {"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "null", "id": "job0"}} -{"error": {"class": "GenericError", "desc": "Conflicts with use by a block device as 'root', which uses 'write' on fmt_base"}} +{"error": {"class": "GenericError", "desc": "Conflicts with use by stream job 'job0' as 'intermediate node', which does not allow 'write' on fmt_base"}} {"return": {}} {"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "BLOCK_EXPORT_DELETED", "data": {"id": "export1"}} *** done -- 2.30.2 From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Split no-perm part of bdrv_attach_child as separate transaction action. It will be used in later commits. Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Kevin Wolf <kwolf@redhat.com> Message-Id: <20210428151804.439460-20-vsementsov@virtuozzo.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block.c | 71 ++++++++++++++++++++++++++++++++++++++++++++++----------- 1 file changed, 58 insertions(+), 13 deletions(-) diff --git a/block.c b/block.c index XXXXXXX..XXXXXXX 100644 --- a/block.c +++ b/block.c @@ -XXX,XX +XXX,XX @@ static BlockDriverState *bdrv_open_inherit(const char *filename, static void bdrv_replace_child_noperm(BdrvChild *child, BlockDriverState *new_bs); +static int bdrv_attach_child_noperm(BlockDriverState *parent_bs, + BlockDriverState *child_bs, + const char *child_name, + const BdrvChildClass *child_class, + BdrvChildRole child_role, + BdrvChild **child, + Transaction *tran, + Error **errp); static int bdrv_reopen_prepare(BDRVReopenState *reopen_state, BlockReopenQueue *queue, Error **errp); @@ -XXX,XX +XXX,XX @@ static int bdrv_attach_child_common(BlockDriverState *child_bs, return 0; } +static int bdrv_attach_child_noperm(BlockDriverState *parent_bs, + BlockDriverState *child_bs, + const char *child_name, + const BdrvChildClass *child_class, + BdrvChildRole child_role, + BdrvChild **child, + Transaction *tran, + Error **errp) +{ + int ret; + uint64_t perm, shared_perm; + + assert(parent_bs->drv); + + bdrv_get_cumulative_perm(parent_bs, &perm, &shared_perm); + bdrv_child_perm(parent_bs, child_bs, NULL, child_role, NULL, + perm, shared_perm, &perm, &shared_perm); + + ret = bdrv_attach_child_common(child_bs, child_name, child_class, + child_role, perm, shared_perm, parent_bs, + child, tran, errp); + if (ret < 0) { + return ret; + } + + QLIST_INSERT_HEAD(&parent_bs->children, *child, next); + /* + * child is removed in bdrv_attach_child_common_abort(), so don't care to + * abort this change separately. + */ + + return 0; +} + static void bdrv_detach_child(BdrvChild *child) { bdrv_replace_child(child, NULL); @@ -XXX,XX +XXX,XX @@ BdrvChild *bdrv_attach_child(BlockDriverState *parent_bs, BdrvChildRole child_role, Error **errp) { - BdrvChild *child; - uint64_t perm, shared_perm; - - bdrv_get_cumulative_perm(parent_bs, &perm, &shared_perm); + int ret; + BdrvChild *child = NULL; + Transaction *tran = tran_new(); - assert(parent_bs->drv); - bdrv_child_perm(parent_bs, child_bs, NULL, child_role, NULL, - perm, shared_perm, &perm, &shared_perm); + ret = bdrv_attach_child_noperm(parent_bs, child_bs, child_name, child_class, + child_role, &child, tran, errp); + if (ret < 0) { + goto out; + } - child = bdrv_root_attach_child(child_bs, child_name, child_class, - child_role, perm, shared_perm, parent_bs, - errp); - if (child == NULL) { - return NULL; + ret = bdrv_refresh_perms(parent_bs, errp); + if (ret < 0) { + goto out; } - QLIST_INSERT_HEAD(&parent_bs->children, child, next); +out: + tran_finalize(tran, ret); + + bdrv_unref(child_bs); + return child; } -- 2.30.2 From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Split part of bdrv_replace_node_common() to be used separately. Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Kevin Wolf <kwolf@redhat.com> Message-Id: <20210428151804.439460-21-vsementsov@virtuozzo.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block.c | 50 +++++++++++++++++++++++++++++++------------------- 1 file changed, 31 insertions(+), 19 deletions(-) diff --git a/block.c b/block.c index XXXXXXX..XXXXXXX 100644 --- a/block.c +++ b/block.c @@ -XXX,XX +XXX,XX @@ static bool should_update_child(BdrvChild *c, BlockDriverState *to) return ret; } +static int bdrv_replace_node_noperm(BlockDriverState *from, + BlockDriverState *to, + bool auto_skip, Transaction *tran, + Error **errp) +{ + BdrvChild *c, *next; + + QLIST_FOREACH_SAFE(c, &from->parents, next_parent, next) { + assert(c->bs == from); + if (!should_update_child(c, to)) { + if (auto_skip) { + continue; + } + error_setg(errp, "Should not change '%s' link to '%s'", + c->name, from->node_name); + return -EINVAL; + } + if (c->frozen) { + error_setg(errp, "Cannot change '%s' link to '%s'", + c->name, from->node_name); + return -EPERM; + } + bdrv_replace_child_safe(c, to, tran); + } + + return 0; +} + /* * With auto_skip=true bdrv_replace_node_common skips updating from parents * if it creates a parent-child relation loop or if parent is block-job. @@ -XXX,XX +XXX,XX @@ static int bdrv_replace_node_common(BlockDriverState *from, BlockDriverState *to, bool auto_skip, Error **errp) { - BdrvChild *c, *next; Transaction *tran = tran_new(); g_autoptr(GHashTable) found = NULL; g_autoptr(GSList) refresh_list = NULL; @@ -XXX,XX +XXX,XX @@ static int bdrv_replace_node_common(BlockDriverState *from, * permissions based on new graph. If we fail, we'll roll-back the * replacement. */ - QLIST_FOREACH_SAFE(c, &from->parents, next_parent, next) { - assert(c->bs == from); - if (!should_update_child(c, to)) { - if (auto_skip) { - continue; - } - ret = -EINVAL; - error_setg(errp, "Should not change '%s' link to '%s'", - c->name, from->node_name); - goto out; - } - if (c->frozen) { - ret = -EPERM; - error_setg(errp, "Cannot change '%s' link to '%s'", - c->name, from->node_name); - goto out; - } - bdrv_replace_child_safe(c, to, tran); + ret = bdrv_replace_node_noperm(from, to, auto_skip, tran, errp); + if (ret < 0) { + goto out; } found = g_hash_table_new(NULL, NULL); -- 2.30.2 From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> bdrv_append is not very good for inserting filters: it does extra permission update as part of bdrv_set_backing_hd(). During this update filter may conflict with other parents of top_bs. Instead, let's first do all graph modifications and after it update permissions. append-greedy-filter test-case in test-bdrv-graph-mod is now works, so move it out of debug option. Note: bdrv_append() is still only works for backing-child based filters. It's something to improve later. Note2: we use the fact that bdrv_append() is used to append new nodes, without backing child, so we don't need frozen check and inherits_from logic from bdrv_set_backing_hd(). Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Kevin Wolf <kwolf@redhat.com> Message-Id: <20210428151804.439460-22-vsementsov@virtuozzo.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block.c | 27 ++++++++++++++++++++------- tests/unit/test-bdrv-graph-mod.c | 17 ++--------------- 2 files changed, 22 insertions(+), 22 deletions(-) diff --git a/block.c b/block.c index XXXXXXX..XXXXXXX 100644 --- a/block.c +++ b/block.c @@ -XXX,XX +XXX,XX @@ int bdrv_replace_node(BlockDriverState *from, BlockDriverState *to, * This will modify the BlockDriverState fields, and swap contents * between bs_new and bs_top. Both bs_new and bs_top are modified. * - * bs_new must not be attached to a BlockBackend. + * bs_new must not be attached to a BlockBackend and must not have backing + * child. * * This function does not create any image files. */ int bdrv_append(BlockDriverState *bs_new, BlockDriverState *bs_top, Error **errp) { - int ret = bdrv_set_backing_hd(bs_new, bs_top, errp); + int ret; + Transaction *tran = tran_new(); + + assert(!bs_new->backing); + + ret = bdrv_attach_child_noperm(bs_new, bs_top, "backing", + &child_of_bds, bdrv_backing_role(bs_new), + &bs_new->backing, tran, errp); if (ret < 0) { - return ret; + goto out; } - ret = bdrv_replace_node(bs_top, bs_new, errp); + ret = bdrv_replace_node_noperm(bs_top, bs_new, true, tran, errp); if (ret < 0) { - bdrv_set_backing_hd(bs_new, NULL, &error_abort); - return ret; + goto out; } - return 0; + ret = bdrv_refresh_perms(bs_new, errp); +out: + tran_finalize(tran, ret); + + bdrv_refresh_limits(bs_top, NULL); + + return ret; } static void bdrv_delete(BlockDriverState *bs) diff --git a/tests/unit/test-bdrv-graph-mod.c b/tests/unit/test-bdrv-graph-mod.c index XXXXXXX..XXXXXXX 100644 --- a/tests/unit/test-bdrv-graph-mod.c +++ b/tests/unit/test-bdrv-graph-mod.c @@ -XXX,XX +XXX,XX @@ static void test_append_greedy_filter(void) int main(int argc, char *argv[]) { - int i; - bool debug = false; - - for (i = 1; i < argc; i++) { - if (!strcmp(argv[i], "-d")) { - debug = true; - break; - } - } - bdrv_init(); qemu_init_main_loop(&error_abort); @@ -XXX,XX +XXX,XX @@ int main(int argc, char *argv[]) test_parallel_perm_update); g_test_add_func("/bdrv-graph-mod/parallel-exclusive-write", test_parallel_exclusive_write); - - if (debug) { - g_test_add_func("/bdrv-graph-mod/append-greedy-filter", - test_append_greedy_filter); - } + g_test_add_func("/bdrv-graph-mod/append-greedy-filter", + test_append_greedy_filter); return g_test_run(); } -- 2.30.2 From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Kevin Wolf <kwolf@redhat.com> Message-Id: <20210428151804.439460-23-vsementsov@virtuozzo.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block.c | 84 +++++++++++++++++++++++++++++++++++++++++++++++++++++++-- 1 file changed, 82 insertions(+), 2 deletions(-) diff --git a/block.c b/block.c index XXXXXXX..XXXXXXX 100644 --- a/block.c +++ b/block.c @@ -XXX,XX +XXX,XX @@ static void bdrv_replace_child(BdrvChild *child, BlockDriverState *new_bs) } } +static void bdrv_child_free(void *opaque) +{ + BdrvChild *c = opaque; + + g_free(c->name); + g_free(c); +} + static void bdrv_remove_empty_child(BdrvChild *child) { assert(!child->bs); QLIST_SAFE_REMOVE(child, next); - g_free(child->name); - g_free(child); + bdrv_child_free(child); } typedef struct BdrvAttachChildCommonState { @@ -XXX,XX +XXX,XX @@ static bool should_update_child(BdrvChild *c, BlockDriverState *to) return ret; } +typedef struct BdrvRemoveFilterOrCowChild { + BdrvChild *child; + bool is_backing; +} BdrvRemoveFilterOrCowChild; + +static void bdrv_remove_filter_or_cow_child_abort(void *opaque) +{ + BdrvRemoveFilterOrCowChild *s = opaque; + BlockDriverState *parent_bs = s->child->opaque; + + QLIST_INSERT_HEAD(&parent_bs->children, s->child, next); + if (s->is_backing) { + parent_bs->backing = s->child; + } else { + parent_bs->file = s->child; + } + + /* + * We don't have to restore child->bs here to undo bdrv_replace_child() + * because that function is transactionable and it registered own completion + * entries in @tran, so .abort() for bdrv_replace_child_safe() will be + * called automatically. + */ +} + +static void bdrv_remove_filter_or_cow_child_commit(void *opaque) +{ + BdrvRemoveFilterOrCowChild *s = opaque; + + bdrv_child_free(s->child); +} + +static TransactionActionDrv bdrv_remove_filter_or_cow_child_drv = { + .abort = bdrv_remove_filter_or_cow_child_abort, + .commit = bdrv_remove_filter_or_cow_child_commit, + .clean = g_free, +}; + +/* + * A function to remove backing-chain child of @bs if exists: cow child for + * format nodes (always .backing) and filter child for filters (may be .file or + * .backing) + */ +__attribute__((unused)) +static void bdrv_remove_filter_or_cow_child(BlockDriverState *bs, + Transaction *tran) +{ + BdrvRemoveFilterOrCowChild *s; + BdrvChild *child = bdrv_filter_or_cow_child(bs); + + if (!child) { + return; + } + + if (child->bs) { + bdrv_replace_child_safe(child, NULL, tran); + } + + s = g_new(BdrvRemoveFilterOrCowChild, 1); + *s = (BdrvRemoveFilterOrCowChild) { + .child = child, + .is_backing = (child == bs->backing), + }; + tran_add(tran, &bdrv_remove_filter_or_cow_child_drv, s); + + QLIST_SAFE_REMOVE(child, next); + if (s->is_backing) { + bs->backing = NULL; + } else { + bs->file = NULL; + } +} + static int bdrv_replace_node_noperm(BlockDriverState *from, BlockDriverState *to, bool auto_skip, Transaction *tran, -- 2.30.2 From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Using bdrv_replace_node() for removing filter is not good enough: it keeps child reference of the filter, which may conflict with original top node during permission update. Instead let's create new interface, which will do all graph modifications first and then update permissions. Let's modify bdrv_replace_node_common(), allowing it additionally drop backing chain child link pointing to new node. This is quite appropriate for bdrv_drop_intermediate() and makes possible to add new bdrv_drop_filter() as a simple wrapper. Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Kevin Wolf <kwolf@redhat.com> Message-Id: <20210428151804.439460-24-vsementsov@virtuozzo.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- include/block/block.h | 1 + block.c | 43 +++++++++++++++++++++++++++++++++++++++---- 2 files changed, 40 insertions(+), 4 deletions(-) diff --git a/include/block/block.h b/include/block/block.h index XXXXXXX..XXXXXXX 100644 --- a/include/block/block.h +++ b/include/block/block.h @@ -XXX,XX +XXX,XX @@ int bdrv_replace_node(BlockDriverState *from, BlockDriverState *to, Error **errp); BlockDriverState *bdrv_insert_node(BlockDriverState *bs, QDict *node_options, int flags, Error **errp); +int bdrv_drop_filter(BlockDriverState *bs, Error **errp); int bdrv_parse_aio(const char *mode, int *flags); int bdrv_parse_cache_mode(const char *mode, int *flags, bool *writethrough); diff --git a/block.c b/block.c index XXXXXXX..XXXXXXX 100644 --- a/block.c +++ b/block.c @@ -XXX,XX +XXX,XX @@ static TransactionActionDrv bdrv_remove_filter_or_cow_child_drv = { * format nodes (always .backing) and filter child for filters (may be .file or * .backing) */ -__attribute__((unused)) static void bdrv_remove_filter_or_cow_child(BlockDriverState *bs, Transaction *tran) { @@ -XXX,XX +XXX,XX @@ static int bdrv_replace_node_noperm(BlockDriverState *from, * * With auto_skip=false the error is returned if from has a parent which should * not be updated. + * + * With @detach_subchain=true @to must be in a backing chain of @from. In this + * case backing link of the cow-parent of @to is removed. */ static int bdrv_replace_node_common(BlockDriverState *from, BlockDriverState *to, - bool auto_skip, Error **errp) + bool auto_skip, bool detach_subchain, + Error **errp) { Transaction *tran = tran_new(); g_autoptr(GHashTable) found = NULL; g_autoptr(GSList) refresh_list = NULL; + BlockDriverState *to_cow_parent; int ret; + if (detach_subchain) { + assert(bdrv_chain_contains(from, to)); + assert(from != to); + for (to_cow_parent = from; + bdrv_filter_or_cow_bs(to_cow_parent) != to; + to_cow_parent = bdrv_filter_or_cow_bs(to_cow_parent)) + { + ; + } + } + /* Make sure that @from doesn't go away until we have successfully attached * all of its parents to @to. */ bdrv_ref(from); @@ -XXX,XX +XXX,XX @@ static int bdrv_replace_node_common(BlockDriverState *from, goto out; } + if (detach_subchain) { + bdrv_remove_filter_or_cow_child(to_cow_parent, tran); + } + found = g_hash_table_new(NULL, NULL); refresh_list = bdrv_topological_dfs(refresh_list, found, to); @@ -XXX,XX +XXX,XX @@ out: int bdrv_replace_node(BlockDriverState *from, BlockDriverState *to, Error **errp) { - return bdrv_replace_node_common(from, to, true, errp); + return bdrv_replace_node_common(from, to, true, false, errp); +} + +int bdrv_drop_filter(BlockDriverState *bs, Error **errp) +{ + return bdrv_replace_node_common(bs, bdrv_filter_or_cow_bs(bs), true, true, + errp); } /* @@ -XXX,XX +XXX,XX @@ int bdrv_drop_intermediate(BlockDriverState *top, BlockDriverState *base, updated_children = g_slist_prepend(updated_children, c); } - bdrv_replace_node_common(top, base, false, &local_err); + /* + * It seems correct to pass detach_subchain=true here, but it triggers + * one more yet not fixed bug, when due to nested aio_poll loop we switch to + * another drained section, which modify the graph (for example, removing + * the child, which we keep in updated_children list). So, it's a TODO. + * + * Note, bug triggered if pass detach_subchain=true here and run + * test-bdrv-drain. test_drop_intermediate_poll() test-case will crash. + * That's a FIXME. + */ + bdrv_replace_node_common(top, base, false, false, &local_err); if (local_err) { error_report_err(local_err); goto exit; -- 2.30.2 From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> We don't need this workaround anymore: bdrv_append is already smart enough and we can use new bdrv_drop_filter(). This commit efficiently reverts also recent 705dde27c6c53b73, which checked .active on io path. Still it said that the problem should be theoretical. And the logic of filter removement is changed anyway. Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Kevin Wolf <kwolf@redhat.com> Message-Id: <20210428151804.439460-25-vsementsov@virtuozzo.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block/backup-top.c | 47 +------------------------------------- tests/qemu-iotests/283.out | 2 +- 2 files changed, 2 insertions(+), 47 deletions(-) diff --git a/block/backup-top.c b/block/backup-top.c index XXXXXXX..XXXXXXX 100644 --- a/block/backup-top.c +++ b/block/backup-top.c @@ -XXX,XX +XXX,XX @@ typedef struct BDRVBackupTopState { BlockCopyState *bcs; BdrvChild *target; - bool active; int64_t cluster_size; } BDRVBackupTopState; @@ -XXX,XX +XXX,XX @@ static coroutine_fn int backup_top_co_preadv( BlockDriverState *bs, uint64_t offset, uint64_t bytes, QEMUIOVector *qiov, int flags) { - BDRVBackupTopState *s = bs->opaque; - - if (!s->active) { - return -EIO; - } - return bdrv_co_preadv(bs->backing, offset, bytes, qiov, flags); } @@ -XXX,XX +XXX,XX @@ static coroutine_fn int backup_top_cbw(BlockDriverState *bs, uint64_t offset, BDRVBackupTopState *s = bs->opaque; uint64_t off, end; - if (!s->active) { - return -EIO; - } - if (flags & BDRV_REQ_WRITE_UNCHANGED) { return 0; } @@ -XXX,XX +XXX,XX @@ static void backup_top_child_perm(BlockDriverState *bs, BdrvChild *c, uint64_t perm, uint64_t shared, uint64_t *nperm, uint64_t *nshared) { - BDRVBackupTopState *s = bs->opaque; - - if (!s->active) { - /* - * The filter node may be in process of bdrv_append(), which firstly do - * bdrv_set_backing_hd() and then bdrv_replace_node(). This means that - * we can't unshare BLK_PERM_WRITE during bdrv_append() operation. So, - * let's require nothing during bdrv_append() and refresh permissions - * after it (see bdrv_backup_top_append()). - */ - *nperm = 0; - *nshared = BLK_PERM_ALL; - return; - } - if (!(role & BDRV_CHILD_FILTERED)) { /* * Target child @@ -XXX,XX +XXX,XX @@ BlockDriverState *bdrv_backup_top_append(BlockDriverState *source, } appended = true; - /* - * bdrv_append() finished successfully, now we can require permissions - * we want. - */ - state->active = true; - ret = bdrv_child_refresh_perms(top, top->backing, errp); - if (ret < 0) { - error_prepend(errp, "Cannot set permissions for backup-top filter: "); - goto fail; - } - state->cluster_size = cluster_size; state->bcs = block_copy_state_new(top->backing, state->target, cluster_size, perf->use_copy_range, @@ -XXX,XX +XXX,XX @@ BlockDriverState *bdrv_backup_top_append(BlockDriverState *source, fail: if (appended) { - state->active = false; bdrv_backup_top_drop(top); } else { bdrv_unref(top); @@ -XXX,XX +XXX,XX @@ void bdrv_backup_top_drop(BlockDriverState *bs) { BDRVBackupTopState *s = bs->opaque; - bdrv_drained_begin(bs); + bdrv_drop_filter(bs, &error_abort); block_copy_state_free(s->bcs); - s->active = false; - bdrv_child_refresh_perms(bs, bs->backing, &error_abort); - bdrv_replace_node(bs, bs->backing->bs, &error_abort); - bdrv_set_backing_hd(bs, NULL, &error_abort); - - bdrv_drained_end(bs); - bdrv_unref(bs); } diff --git a/tests/qemu-iotests/283.out b/tests/qemu-iotests/283.out index XXXXXXX..XXXXXXX 100644 --- a/tests/qemu-iotests/283.out +++ b/tests/qemu-iotests/283.out @@ -XXX,XX +XXX,XX @@ {"execute": "blockdev-add", "arguments": {"driver": "blkdebug", "image": "base", "node-name": "other", "take-child-perms": ["write"]}} {"return": {}} {"execute": "blockdev-backup", "arguments": {"device": "source", "sync": "full", "target": "target"}} -{"error": {"class": "GenericError", "desc": "Cannot set permissions for backup-top filter: Conflicts with use by source as 'image', which does not allow 'write' on base"}} +{"error": {"class": "GenericError", "desc": "Cannot append backup-top filter: Conflicts with use by source as 'image', which does not allow 'write' on base"}} === backup-top should be gone after job-finalize === -- 2.30.2 From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> This argument is always NULL. Drop it. Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Kevin Wolf <kwolf@redhat.com> Message-Id: <20210428151804.439460-26-vsementsov@virtuozzo.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block.c | 38 +++++++++++--------------------------- 1 file changed, 11 insertions(+), 27 deletions(-) diff --git a/block.c b/block.c index XXXXXXX..XXXXXXX 100644 --- a/block.c +++ b/block.c @@ -XXX,XX +XXX,XX @@ static int bdrv_fill_options(QDict **options, const char *filename, static int bdrv_check_update_perm(BlockDriverState *bs, BlockReopenQueue *q, uint64_t new_used_perm, uint64_t new_shared_perm, - GSList *ignore_children, Error **errp); typedef struct BlockReopenQueueEntry { @@ -XXX,XX +XXX,XX @@ static bool bdrv_a_allow_b(BdrvChild *a, BdrvChild *b, Error **errp) return false; } -static bool bdrv_parent_perms_conflict(BlockDriverState *bs, - GSList *ignore_children, - Error **errp) +static bool bdrv_parent_perms_conflict(BlockDriverState *bs, Error **errp) { BdrvChild *a, *b; @@ -XXX,XX +XXX,XX @@ static bool bdrv_parent_perms_conflict(BlockDriverState *bs, * directions. */ QLIST_FOREACH(a, &bs->parents, next_parent) { - if (g_slist_find(ignore_children, a)) { - continue; - } - QLIST_FOREACH(b, &bs->parents, next_parent) { - if (a == b || g_slist_find(ignore_children, b)) { + if (a == b) { continue; } @@ -XXX,XX +XXX,XX @@ static void bdrv_replace_child_safe(BdrvChild *child, BlockDriverState *new_bs, static int bdrv_node_check_perm(BlockDriverState *bs, BlockReopenQueue *q, uint64_t cumulative_perms, uint64_t cumulative_shared_perms, - GSList *ignore_children, Transaction *tran, Error **errp) { BlockDriver *drv = bs->drv; @@ -XXX,XX +XXX,XX @@ static int bdrv_check_perm_common(GSList *list, BlockReopenQueue *q, bool use_cumulative_perms, uint64_t cumulative_perms, uint64_t cumulative_shared_perms, - GSList *ignore_children, Transaction *tran, Error **errp) { int ret; @@ -XXX,XX +XXX,XX @@ static int bdrv_check_perm_common(GSList *list, BlockReopenQueue *q, ret = bdrv_node_check_perm(bs, q, cumulative_perms, cumulative_shared_perms, - ignore_children, tran, errp); + tran, errp); if (ret < 0) { return ret; } @@ -XXX,XX +XXX,XX @@ static int bdrv_check_perm_common(GSList *list, BlockReopenQueue *q, for ( ; list; list = list->next) { bs = list->data; - if (bdrv_parent_perms_conflict(bs, ignore_children, errp)) { + if (bdrv_parent_perms_conflict(bs, errp)) { return -EINVAL; } @@ -XXX,XX +XXX,XX @@ static int bdrv_check_perm_common(GSList *list, BlockReopenQueue *q, ret = bdrv_node_check_perm(bs, q, cumulative_perms, cumulative_shared_perms, - ignore_children, tran, errp); + tran, errp); if (ret < 0) { return ret; } @@ -XXX,XX +XXX,XX @@ static int bdrv_check_perm_common(GSList *list, BlockReopenQueue *q, static int bdrv_check_perm(BlockDriverState *bs, BlockReopenQueue *q, uint64_t cumulative_perms, - uint64_t cumulative_shared_perms, - GSList *ignore_children, Error **errp) + uint64_t cumulative_shared_perms, Error **errp) { g_autoptr(GSList) list = bdrv_topological_dfs(NULL, NULL, bs); return bdrv_check_perm_common(list, q, true, cumulative_perms, - cumulative_shared_perms, ignore_children, - NULL, errp); + cumulative_shared_perms, NULL, errp); } static int bdrv_list_refresh_perms(GSList *list, BlockReopenQueue *q, Transaction *tran, Error **errp) { - return bdrv_check_perm_common(list, q, false, 0, 0, NULL, tran, errp); + return bdrv_check_perm_common(list, q, false, 0, 0, tran, errp); } /* @@ -XXX,XX +XXX,XX @@ char *bdrv_perm_names(uint64_t perm) static int bdrv_check_update_perm(BlockDriverState *bs, BlockReopenQueue *q, uint64_t new_used_perm, uint64_t new_shared_perm, - GSList *ignore_children, Error **errp) { BdrvChild *c; @@ -XXX,XX +XXX,XX @@ static int bdrv_check_update_perm(BlockDriverState *bs, BlockReopenQueue *q, assert(new_shared_perm & BLK_PERM_WRITE_UNCHANGED); QLIST_FOREACH(c, &bs->parents, next_parent) { - if (g_slist_find(ignore_children, c)) { - continue; - } - if ((new_used_perm & c->shared_perm) != new_used_perm) { char *user = bdrv_child_user_desc(c); char *perm_names = bdrv_perm_names(new_used_perm & ~c->shared_perm); @@ -XXX,XX +XXX,XX @@ static int bdrv_check_update_perm(BlockDriverState *bs, BlockReopenQueue *q, } return bdrv_check_perm(bs, q, cumulative_perms, cumulative_shared_perms, - ignore_children, errp); + errp); } static int bdrv_refresh_perms(BlockDriverState *bs, Error **errp) @@ -XXX,XX +XXX,XX @@ int bdrv_reopen_multiple(BlockReopenQueue *bs_queue, Error **errp) QTAILQ_FOREACH(bs_entry, bs_queue, entry) { BDRVReopenState *state = &bs_entry->state; ret = bdrv_check_perm(state->bs, bs_queue, state->perm, - state->shared_perm, NULL, errp); + state->shared_perm, errp); if (ret < 0) { goto cleanup_perm; } @@ -XXX,XX +XXX,XX @@ int bdrv_reopen_multiple(BlockReopenQueue *bs_queue, Error **errp) bs_queue, state->perm, state->shared_perm, &nperm, &nshared); ret = bdrv_check_update_perm(state->new_backing_bs, NULL, - nperm, nshared, NULL, errp); + nperm, nshared, errp); if (ret < 0) { goto cleanup_perm; } -- 2.30.2 From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> To be used in the further commit. Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Kevin Wolf <kwolf@redhat.com> Message-Id: <20210428151804.439460-27-vsementsov@virtuozzo.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block.c | 46 ++++++++++++++++++++++++++++++++++++++++++---- 1 file changed, 42 insertions(+), 4 deletions(-) diff --git a/block.c b/block.c index XXXXXXX..XXXXXXX 100644 --- a/block.c +++ b/block.c @@ -XXX,XX +XXX,XX @@ void bdrv_root_unref_child(BdrvChild *child) bdrv_unref(child_bs); } +typedef struct BdrvSetInheritsFrom { + BlockDriverState *bs; + BlockDriverState *old_inherits_from; +} BdrvSetInheritsFrom; + +static void bdrv_set_inherits_from_abort(void *opaque) +{ + BdrvSetInheritsFrom *s = opaque; + + s->bs->inherits_from = s->old_inherits_from; +} + +static TransactionActionDrv bdrv_set_inherits_from_drv = { + .abort = bdrv_set_inherits_from_abort, + .clean = g_free, +}; + +/* @tran is allowed to be NULL. In this case no rollback is possible */ +static void bdrv_set_inherits_from(BlockDriverState *bs, + BlockDriverState *new_inherits_from, + Transaction *tran) +{ + if (tran) { + BdrvSetInheritsFrom *s = g_new(BdrvSetInheritsFrom, 1); + + *s = (BdrvSetInheritsFrom) { + .bs = bs, + .old_inherits_from = bs->inherits_from, + }; + + tran_add(tran, &bdrv_set_inherits_from_drv, s); + } + + bs->inherits_from = new_inherits_from; +} + /** * Clear all inherits_from pointers from children and grandchildren of * @root that point to @root, where necessary. + * @tran is allowed to be NULL. In this case no rollback is possible */ -static void bdrv_unset_inherits_from(BlockDriverState *root, BdrvChild *child) +static void bdrv_unset_inherits_from(BlockDriverState *root, BdrvChild *child, + Transaction *tran) { BdrvChild *c; @@ -XXX,XX +XXX,XX @@ static void bdrv_unset_inherits_from(BlockDriverState *root, BdrvChild *child) } } if (c == NULL) { - child->bs->inherits_from = NULL; + bdrv_set_inherits_from(child->bs, NULL, tran); } } QLIST_FOREACH(c, &child->bs->children, next) { - bdrv_unset_inherits_from(root, c); + bdrv_unset_inherits_from(root, c, tran); } } @@ -XXX,XX +XXX,XX @@ void bdrv_unref_child(BlockDriverState *parent, BdrvChild *child) return; } - bdrv_unset_inherits_from(parent, child); + bdrv_unset_inherits_from(parent, child, NULL); bdrv_root_unref_child(child); } -- 2.30.2 From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> To be used in further commit. Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Kevin Wolf <kwolf@redhat.com> Message-Id: <20210428151804.439460-28-vsementsov@virtuozzo.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- include/block/block.h | 3 ++- block.c | 9 ++++----- block/io.c | 31 +++++++++++++++++++++++++++++-- 3 files changed, 35 insertions(+), 8 deletions(-) diff --git a/include/block/block.h b/include/block/block.h index XXXXXXX..XXXXXXX 100644 --- a/include/block/block.h +++ b/include/block/block.h @@ -XXX,XX +XXX,XX @@ #include "block/dirty-bitmap.h" #include "block/blockjob.h" #include "qemu/hbitmap.h" +#include "qemu/transactions.h" /* * generated_co_wrapper @@ -XXX,XX +XXX,XX @@ int64_t bdrv_get_allocated_file_size(BlockDriverState *bs); BlockMeasureInfo *bdrv_measure(BlockDriver *drv, QemuOpts *opts, BlockDriverState *in_bs, Error **errp); void bdrv_get_geometry(BlockDriverState *bs, uint64_t *nb_sectors_ptr); -void bdrv_refresh_limits(BlockDriverState *bs, Error **errp); +void bdrv_refresh_limits(BlockDriverState *bs, Transaction *tran, Error **errp); int bdrv_commit(BlockDriverState *bs); int bdrv_make_empty(BdrvChild *c, Error **errp); int bdrv_change_backing_file(BlockDriverState *bs, const char *backing_file, diff --git a/block.c b/block.c index XXXXXXX..XXXXXXX 100644 --- a/block.c +++ b/block.c @@ -XXX,XX +XXX,XX @@ #include "qemu/timer.h" #include "qemu/cutils.h" #include "qemu/id.h" -#include "qemu/transactions.h" #include "block/coroutines.h" #ifdef CONFIG_BSD @@ -XXX,XX +XXX,XX @@ static int bdrv_open_driver(BlockDriverState *bs, BlockDriver *drv, return ret; } - bdrv_refresh_limits(bs, &local_err); + bdrv_refresh_limits(bs, NULL, &local_err); if (local_err) { error_propagate(errp, local_err); return -EINVAL; @@ -XXX,XX +XXX,XX @@ int bdrv_set_backing_hd(BlockDriverState *bs, BlockDriverState *backing_hd, } out: - bdrv_refresh_limits(bs, NULL); + bdrv_refresh_limits(bs, NULL, NULL); return ret; } @@ -XXX,XX +XXX,XX @@ static void bdrv_reopen_commit(BDRVReopenState *reopen_state) bdrv_set_backing_hd(bs, reopen_state->new_backing_bs, &error_abort); } - bdrv_refresh_limits(bs, NULL); + bdrv_refresh_limits(bs, NULL, NULL); } /* @@ -XXX,XX +XXX,XX @@ int bdrv_append(BlockDriverState *bs_new, BlockDriverState *bs_top, out: tran_finalize(tran, ret); - bdrv_refresh_limits(bs_top, NULL); + bdrv_refresh_limits(bs_top, NULL, NULL); return ret; } diff --git a/block/io.c b/block/io.c index XXXXXXX..XXXXXXX 100644 --- a/block/io.c +++ b/block/io.c @@ -XXX,XX +XXX,XX @@ static void bdrv_merge_limits(BlockLimits *dst, const BlockLimits *src) dst->max_iov = MIN_NON_ZERO(dst->max_iov, src->max_iov); } -void bdrv_refresh_limits(BlockDriverState *bs, Error **errp) +typedef struct BdrvRefreshLimitsState { + BlockDriverState *bs; + BlockLimits old_bl; +} BdrvRefreshLimitsState; + +static void bdrv_refresh_limits_abort(void *opaque) +{ + BdrvRefreshLimitsState *s = opaque; + + s->bs->bl = s->old_bl; +} + +static TransactionActionDrv bdrv_refresh_limits_drv = { + .abort = bdrv_refresh_limits_abort, + .clean = g_free, +}; + +/* @tran is allowed to be NULL, in this case no rollback is possible. */ +void bdrv_refresh_limits(BlockDriverState *bs, Transaction *tran, Error **errp) { ERRP_GUARD(); BlockDriver *drv = bs->drv; BdrvChild *c; bool have_limits; + if (tran) { + BdrvRefreshLimitsState *s = g_new(BdrvRefreshLimitsState, 1); + *s = (BdrvRefreshLimitsState) { + .bs = bs, + .old_bl = bs->bl, + }; + tran_add(tran, &bdrv_refresh_limits_drv, s); + } + memset(&bs->bl, 0, sizeof(bs->bl)); if (!drv) { @@ -XXX,XX +XXX,XX @@ void bdrv_refresh_limits(BlockDriverState *bs, Error **errp) QLIST_FOREACH(c, &bs->children, next) { if (c->role & (BDRV_CHILD_DATA | BDRV_CHILD_FILTERED | BDRV_CHILD_COW)) { - bdrv_refresh_limits(c->bs, errp); + bdrv_refresh_limits(c->bs, tran, errp); if (*errp) { return; } -- 2.30.2 From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Split out no-perm part of bdrv_set_backing_hd() as a separate transaction action. Note the in case of existing BdrvChild we reuse it, not recreate, just to do less actions. We don't need to create extra reference to backing_hd as we don't lose it in bdrv_attach_child(). Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Kevin Wolf <kwolf@redhat.com> Message-Id: <20210428151804.439460-29-vsementsov@virtuozzo.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block.c | 54 +++++++++++++++++++++++++++++++++++++----------------- 1 file changed, 37 insertions(+), 17 deletions(-) diff --git a/block.c b/block.c index XXXXXXX..XXXXXXX 100644 --- a/block.c +++ b/block.c @@ -XXX,XX +XXX,XX @@ static int bdrv_attach_child_noperm(BlockDriverState *parent_bs, BdrvChild **child, Transaction *tran, Error **errp); +static void bdrv_remove_filter_or_cow_child(BlockDriverState *bs, + Transaction *tran); static int bdrv_reopen_prepare(BDRVReopenState *reopen_state, BlockReopenQueue *queue, Error **errp); @@ -XXX,XX +XXX,XX @@ static BdrvChildRole bdrv_backing_role(BlockDriverState *bs) * Sets the bs->backing link of a BDS. A new reference is created; callers * which don't need their own reference any more must call bdrv_unref(). */ -int bdrv_set_backing_hd(BlockDriverState *bs, BlockDriverState *backing_hd, - Error **errp) +static int bdrv_set_backing_noperm(BlockDriverState *bs, + BlockDriverState *backing_hd, + Transaction *tran, Error **errp) { int ret = 0; bool update_inherits_from = bdrv_chain_contains(bs, backing_hd) && @@ -XXX,XX +XXX,XX @@ int bdrv_set_backing_hd(BlockDriverState *bs, BlockDriverState *backing_hd, return -EPERM; } - if (backing_hd) { - bdrv_ref(backing_hd); - } - if (bs->backing) { /* Cannot be frozen, we checked that above */ - bdrv_unref_child(bs, bs->backing); - bs->backing = NULL; + bdrv_unset_inherits_from(bs, bs->backing, tran); + bdrv_remove_filter_or_cow_child(bs, tran); } if (!backing_hd) { goto out; } - bs->backing = bdrv_attach_child(bs, backing_hd, "backing", &child_of_bds, - bdrv_backing_role(bs), errp); - if (!bs->backing) { - ret = -EPERM; - goto out; + ret = bdrv_attach_child_noperm(bs, backing_hd, "backing", + &child_of_bds, bdrv_backing_role(bs), + &bs->backing, tran, errp); + if (ret < 0) { + return ret; } - /* If backing_hd was already part of bs's backing chain, and + + /* + * If backing_hd was already part of bs's backing chain, and * inherits_from pointed recursively to bs then let's update it to - * point directly to bs (else it will become NULL). */ + * point directly to bs (else it will become NULL). + */ if (update_inherits_from) { - backing_hd->inherits_from = bs; + bdrv_set_inherits_from(backing_hd, bs, tran); } out: - bdrv_refresh_limits(bs, NULL, NULL); + bdrv_refresh_limits(bs, tran, NULL); + + return 0; +} + +int bdrv_set_backing_hd(BlockDriverState *bs, BlockDriverState *backing_hd, + Error **errp) +{ + int ret; + Transaction *tran = tran_new(); + + ret = bdrv_set_backing_noperm(bs, backing_hd, tran, errp); + if (ret < 0) { + goto out; + } + + ret = bdrv_refresh_perms(bs, errp); +out: + tran_finalize(tran, ret); return ret; } -- 2.30.2 From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> During reopen we may add backing bs from other aio context, which may lead to changing original context of top bs. We are going to move graph modification to prepare stage. So, it will be possible that bdrv_flush() in bdrv_reopen_prepare called on bs in non-original aio context, which we didn't aquire which leads to crash. To avoid this problem move bdrv_flush() to be a separate reopen stage before bdrv_reopen_prepare(). This doesn't seem correct to acquire only one aio context and not all contexts participating in reopen. But it's not obvious how to do it correctly, keeping in mind: 1. rules of bdrv_set_aio_context_ignore() that requires new_context lock not being held 2. possible deadlocks because of holding all (or several?) AioContext locks Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Kevin Wolf <kwolf@redhat.com> Message-Id: <20210428151804.439460-30-vsementsov@virtuozzo.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block.c | 14 ++++++++------ 1 file changed, 8 insertions(+), 6 deletions(-) diff --git a/block.c b/block.c index XXXXXXX..XXXXXXX 100644 --- a/block.c +++ b/block.c @@ -XXX,XX +XXX,XX @@ int bdrv_reopen_multiple(BlockReopenQueue *bs_queue, Error **errp) assert(bs_queue != NULL); + QTAILQ_FOREACH(bs_entry, bs_queue, entry) { + ret = bdrv_flush(bs_entry->state.bs); + if (ret < 0) { + error_setg_errno(errp, -ret, "Error flushing drive"); + goto cleanup; + } + } + QTAILQ_FOREACH(bs_entry, bs_queue, entry) { assert(bs_entry->state.bs->quiesce_counter > 0); if (bdrv_reopen_prepare(&bs_entry->state, bs_queue, errp)) { @@ -XXX,XX +XXX,XX @@ static int bdrv_reopen_prepare(BDRVReopenState *reopen_state, bdrv_reopen_perm(queue, reopen_state->bs, &reopen_state->perm, &reopen_state->shared_perm); - ret = bdrv_flush(reopen_state->bs); - if (ret) { - error_setg_errno(errp, -ret, "Error flushing drive"); - goto error; - } - if (drv->bdrv_reopen_prepare) { /* * If a driver-specific option is missing, it means that we -- 2.30.2 From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Move bdrv_reopen_multiple to new paradigm of permission update: first update graph relations, then do refresh the permissions. We have to modify reopen process in file-posix driver: with new scheme we don't have prepared permissions in raw_reopen_prepare(), so we should reconfigure fd in raw_check_perm(). Still this seems more native and simple anyway. Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Kevin Wolf <kwolf@redhat.com> Message-Id: <20210428151804.439460-31-vsementsov@virtuozzo.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- include/block/block.h | 3 +- block.c | 187 ++++++++++++------------------------------ block/file-posix.c | 91 +++++++------------- 3 files changed, 84 insertions(+), 197 deletions(-) diff --git a/include/block/block.h b/include/block/block.h index XXXXXXX..XXXXXXX 100644 --- a/include/block/block.h +++ b/include/block/block.h @@ -XXX,XX +XXX,XX @@ typedef struct BDRVReopenState { BlockdevDetectZeroesOptions detect_zeroes; bool backing_missing; bool replace_backing_bs; /* new_backing_bs is ignored if this is false */ - BlockDriverState *new_backing_bs; /* If NULL then detach the current bs */ - uint64_t perm, shared_perm; + BlockDriverState *old_backing_bs; /* keep pointer for permissions update */ QDict *options; QDict *explicit_options; void *opaque; diff --git a/block.c b/block.c index XXXXXXX..XXXXXXX 100644 --- a/block.c +++ b/block.c @@ -XXX,XX +XXX,XX @@ static int bdrv_attach_child_noperm(BlockDriverState *parent_bs, static void bdrv_remove_filter_or_cow_child(BlockDriverState *bs, Transaction *tran); -static int bdrv_reopen_prepare(BDRVReopenState *reopen_state, BlockReopenQueue - *queue, Error **errp); +static int bdrv_reopen_prepare(BDRVReopenState *reopen_state, + BlockReopenQueue *queue, + Transaction *set_backings_tran, Error **errp); static void bdrv_reopen_commit(BDRVReopenState *reopen_state); static void bdrv_reopen_abort(BDRVReopenState *reopen_state); @@ -XXX,XX +XXX,XX @@ static void bdrv_list_abort_perm_update(GSList *list) } } +__attribute__((unused)) static void bdrv_abort_perm_update(BlockDriverState *bs) { g_autoptr(GSList) list = bdrv_topological_dfs(NULL, NULL, bs); @@ -XXX,XX +XXX,XX @@ char *bdrv_perm_names(uint64_t perm) * * Needs to be followed by a call to either bdrv_set_perm() or * bdrv_abort_perm_update(). */ +__attribute__((unused)) static int bdrv_check_update_perm(BlockDriverState *bs, BlockReopenQueue *q, uint64_t new_used_perm, uint64_t new_shared_perm, @@ -XXX,XX +XXX,XX @@ static BlockReopenQueue *bdrv_reopen_queue_child(BlockReopenQueue *bs_queue, bs_entry->state.explicit_options = explicit_options; bs_entry->state.flags = flags; - /* This needs to be overwritten in bdrv_reopen_prepare() */ - bs_entry->state.perm = UINT64_MAX; - bs_entry->state.shared_perm = 0; - /* * If keep_old_opts is false then it means that unspecified * options must be reset to their original value. We don't allow @@ -XXX,XX +XXX,XX @@ int bdrv_reopen_multiple(BlockReopenQueue *bs_queue, Error **errp) { int ret = -1; BlockReopenQueueEntry *bs_entry, *next; + Transaction *tran = tran_new(); + g_autoptr(GHashTable) found = NULL; + g_autoptr(GSList) refresh_list = NULL; assert(bs_queue != NULL); @@ -XXX,XX +XXX,XX @@ int bdrv_reopen_multiple(BlockReopenQueue *bs_queue, Error **errp) QTAILQ_FOREACH(bs_entry, bs_queue, entry) { assert(bs_entry->state.bs->quiesce_counter > 0); - if (bdrv_reopen_prepare(&bs_entry->state, bs_queue, errp)) { - goto cleanup; + ret = bdrv_reopen_prepare(&bs_entry->state, bs_queue, tran, errp); + if (ret < 0) { + goto abort; } bs_entry->prepared = true; } + found = g_hash_table_new(NULL, NULL); QTAILQ_FOREACH(bs_entry, bs_queue, entry) { BDRVReopenState *state = &bs_entry->state; - ret = bdrv_check_perm(state->bs, bs_queue, state->perm, - state->shared_perm, errp); - if (ret < 0) { - goto cleanup_perm; - } - /* Check if new_backing_bs would accept the new permissions */ - if (state->replace_backing_bs && state->new_backing_bs) { - uint64_t nperm, nshared; - bdrv_child_perm(state->bs, state->new_backing_bs, - NULL, bdrv_backing_role(state->bs), - bs_queue, state->perm, state->shared_perm, - &nperm, &nshared); - ret = bdrv_check_update_perm(state->new_backing_bs, NULL, - nperm, nshared, errp); - if (ret < 0) { - goto cleanup_perm; - } + + refresh_list = bdrv_topological_dfs(refresh_list, found, state->bs); + if (state->old_backing_bs) { + refresh_list = bdrv_topological_dfs(refresh_list, found, + state->old_backing_bs); } - bs_entry->perms_checked = true; + } + + /* + * Note that file-posix driver rely on permission update done during reopen + * (even if no permission changed), because it wants "new" permissions for + * reconfiguring the fd and that's why it does it in raw_check_perm(), not + * in raw_reopen_prepare() which is called with "old" permissions. + */ + ret = bdrv_list_refresh_perms(refresh_list, bs_queue, tran, errp); + if (ret < 0) { + goto abort; } /* @@ -XXX,XX +XXX,XX @@ int bdrv_reopen_multiple(BlockReopenQueue *bs_queue, Error **errp) bdrv_reopen_commit(&bs_entry->state); } - ret = 0; -cleanup_perm: - QTAILQ_FOREACH_SAFE(bs_entry, bs_queue, entry, next) { - BDRVReopenState *state = &bs_entry->state; - - if (!bs_entry->perms_checked) { - continue; - } - - if (ret == 0) { - uint64_t perm, shared; + tran_commit(tran); - bdrv_get_cumulative_perm(state->bs, &perm, &shared); - assert(perm == state->perm); - assert(shared == state->shared_perm); + QTAILQ_FOREACH_REVERSE(bs_entry, bs_queue, entry) { + BlockDriverState *bs = bs_entry->state.bs; - bdrv_set_perm(state->bs); - } else { - bdrv_abort_perm_update(state->bs); - if (state->replace_backing_bs && state->new_backing_bs) { - bdrv_abort_perm_update(state->new_backing_bs); - } + if (bs->drv->bdrv_reopen_commit_post) { + bs->drv->bdrv_reopen_commit_post(&bs_entry->state); } } - if (ret == 0) { - QTAILQ_FOREACH_REVERSE(bs_entry, bs_queue, entry) { - BlockDriverState *bs = bs_entry->state.bs; + ret = 0; + goto cleanup; - if (bs->drv->bdrv_reopen_commit_post) - bs->drv->bdrv_reopen_commit_post(&bs_entry->state); +abort: + tran_abort(tran); + QTAILQ_FOREACH_SAFE(bs_entry, bs_queue, entry, next) { + if (bs_entry->prepared) { + bdrv_reopen_abort(&bs_entry->state); } + qobject_unref(bs_entry->state.explicit_options); + qobject_unref(bs_entry->state.options); } + cleanup: QTAILQ_FOREACH_SAFE(bs_entry, bs_queue, entry, next) { - if (ret) { - if (bs_entry->prepared) { - bdrv_reopen_abort(&bs_entry->state); - } - qobject_unref(bs_entry->state.explicit_options); - qobject_unref(bs_entry->state.options); - } - if (bs_entry->state.new_backing_bs) { - bdrv_unref(bs_entry->state.new_backing_bs); - } g_free(bs_entry); } g_free(bs_queue); @@ -XXX,XX +XXX,XX @@ int bdrv_reopen_set_read_only(BlockDriverState *bs, bool read_only, return ret; } -static BlockReopenQueueEntry *find_parent_in_reopen_queue(BlockReopenQueue *q, - BdrvChild *c) -{ - BlockReopenQueueEntry *entry; - - QTAILQ_FOREACH(entry, q, entry) { - BlockDriverState *bs = entry->state.bs; - BdrvChild *child; - - QLIST_FOREACH(child, &bs->children, next) { - if (child == c) { - return entry; - } - } - } - - return NULL; -} - -static void bdrv_reopen_perm(BlockReopenQueue *q, BlockDriverState *bs, - uint64_t *perm, uint64_t *shared) -{ - BdrvChild *c; - BlockReopenQueueEntry *parent; - uint64_t cumulative_perms = 0; - uint64_t cumulative_shared_perms = BLK_PERM_ALL; - - QLIST_FOREACH(c, &bs->parents, next_parent) { - parent = find_parent_in_reopen_queue(q, c); - if (!parent) { - cumulative_perms |= c->perm; - cumulative_shared_perms &= c->shared_perm; - } else { - uint64_t nperm, nshared; - - bdrv_child_perm(parent->state.bs, bs, c, c->role, q, - parent->state.perm, parent->state.shared_perm, - &nperm, &nshared); - - cumulative_perms |= nperm; - cumulative_shared_perms &= nshared; - } - } - *perm = cumulative_perms; - *shared = cumulative_shared_perms; -} - static bool bdrv_reopen_can_attach(BlockDriverState *parent, BdrvChild *child, BlockDriverState *new_child, @@ -XXX,XX +XXX,XX @@ static bool bdrv_reopen_can_attach(BlockDriverState *parent, * Return 0 on success, otherwise return < 0 and set @errp. */ static int bdrv_reopen_parse_backing(BDRVReopenState *reopen_state, + Transaction *set_backings_tran, Error **errp) { BlockDriverState *bs = reopen_state->bs; @@ -XXX,XX +XXX,XX @@ static int bdrv_reopen_parse_backing(BDRVReopenState *reopen_state, /* If we want to replace the backing file we need some extra checks */ if (new_backing_bs != bdrv_filter_or_cow_bs(overlay_bs)) { + int ret; + /* Check for implicit nodes between bs and its backing file */ if (bs != overlay_bs) { error_setg(errp, "Cannot change backing link if '%s' has " @@ -XXX,XX +XXX,XX @@ static int bdrv_reopen_parse_backing(BDRVReopenState *reopen_state, return -EPERM; } reopen_state->replace_backing_bs = true; - if (new_backing_bs) { - bdrv_ref(new_backing_bs); - reopen_state->new_backing_bs = new_backing_bs; + reopen_state->old_backing_bs = bs->backing ? bs->backing->bs : NULL; + ret = bdrv_set_backing_noperm(bs, new_backing_bs, set_backings_tran, + errp); + if (ret < 0) { + return ret; } } @@ -XXX,XX +XXX,XX @@ static int bdrv_reopen_parse_backing(BDRVReopenState *reopen_state, * */ static int bdrv_reopen_prepare(BDRVReopenState *reopen_state, - BlockReopenQueue *queue, Error **errp) + BlockReopenQueue *queue, + Transaction *set_backings_tran, Error **errp) { int ret = -1; int old_flags; @@ -XXX,XX +XXX,XX @@ static int bdrv_reopen_prepare(BDRVReopenState *reopen_state, goto error; } - /* Calculate required permissions after reopening */ - bdrv_reopen_perm(queue, reopen_state->bs, - &reopen_state->perm, &reopen_state->shared_perm); - if (drv->bdrv_reopen_prepare) { /* * If a driver-specific option is missing, it means that we @@ -XXX,XX +XXX,XX @@ static int bdrv_reopen_prepare(BDRVReopenState *reopen_state, * either a reference to an existing node (using its node name) * or NULL to simply detach the current backing file. */ - ret = bdrv_reopen_parse_backing(reopen_state, errp); + ret = bdrv_reopen_parse_backing(reopen_state, set_backings_tran, errp); if (ret < 0) { goto error; } @@ -XXX,XX +XXX,XX @@ static void bdrv_reopen_commit(BDRVReopenState *reopen_state) qdict_del(bs->explicit_options, child->name); qdict_del(bs->options, child->name); } - - /* - * Change the backing file if a new one was specified. We do this - * after updating bs->options, so bdrv_refresh_filename() (called - * from bdrv_set_backing_hd()) has the new values. - */ - if (reopen_state->replace_backing_bs) { - BlockDriverState *old_backing_bs = child_bs(bs->backing); - assert(!old_backing_bs || !old_backing_bs->implicit); - /* Abort the permission update on the backing bs we're detaching */ - if (old_backing_bs) { - bdrv_abort_perm_update(old_backing_bs); - } - bdrv_set_backing_hd(bs, reopen_state->new_backing_bs, &error_abort); - } - bdrv_refresh_limits(bs, NULL, NULL); } diff --git a/block/file-posix.c b/block/file-posix.c index XXXXXXX..XXXXXXX 100644 --- a/block/file-posix.c +++ b/block/file-posix.c @@ -XXX,XX +XXX,XX @@ typedef struct BDRVRawState { } BDRVRawState; typedef struct BDRVRawReopenState { - int fd; int open_flags; bool drop_cache; bool check_cache_dropped; @@ -XXX,XX +XXX,XX @@ static int raw_reopen_prepare(BDRVReopenState *state, BDRVRawReopenState *rs; QemuOpts *opts; int ret; - Error *local_err = NULL; assert(state != NULL); assert(state->bs != NULL); @@ -XXX,XX +XXX,XX @@ static int raw_reopen_prepare(BDRVReopenState *state, * bdrv_reopen_prepare() will detect changes and complain. */ qemu_opts_to_qdict(opts, state->options); - rs->fd = raw_reconfigure_getfd(state->bs, state->flags, &rs->open_flags, - state->perm, true, &local_err); - if (local_err) { - error_propagate(errp, local_err); - ret = -1; - goto out; - } - - /* Fail already reopen_prepare() if we can't get a working O_DIRECT - * alignment with the new fd. */ - if (rs->fd != -1) { - raw_probe_alignment(state->bs, rs->fd, &local_err); - if (local_err) { - error_propagate(errp, local_err); - ret = -EINVAL; - goto out_fd; - } - } + /* + * As part of reopen prepare we also want to create new fd by + * raw_reconfigure_getfd(). But it wants updated "perm", when in + * bdrv_reopen_multiple() .bdrv_reopen_prepare() callback called prior to + * permission update. Happily, permission update is always a part (a seprate + * stage) of bdrv_reopen_multiple() so we can rely on this fact and + * reconfigure fd in raw_check_perm(). + */ s->reopen_state = state; ret = 0; -out_fd: - if (ret < 0) { - qemu_close(rs->fd); - rs->fd = -1; - } + out: qemu_opts_del(opts); return ret; @@ -XXX,XX +XXX,XX @@ static void raw_reopen_commit(BDRVReopenState *state) s->drop_cache = rs->drop_cache; s->check_cache_dropped = rs->check_cache_dropped; s->open_flags = rs->open_flags; - - qemu_close(s->fd); - s->fd = rs->fd; - g_free(state->opaque); state->opaque = NULL; @@ -XXX,XX +XXX,XX @@ static void raw_reopen_abort(BDRVReopenState *state) return; } - if (rs->fd >= 0) { - qemu_close(rs->fd); - rs->fd = -1; - } g_free(state->opaque); state->opaque = NULL; @@ -XXX,XX +XXX,XX @@ static int raw_check_perm(BlockDriverState *bs, uint64_t perm, uint64_t shared, Error **errp) { BDRVRawState *s = bs->opaque; - BDRVRawReopenState *rs = NULL; + int input_flags = s->reopen_state ? s->reopen_state->flags : bs->open_flags; int open_flags; int ret; - if (s->perm_change_fd) { + /* We may need a new fd if auto-read-only switches the mode */ + ret = raw_reconfigure_getfd(bs, input_flags, &open_flags, perm, + false, errp); + if (ret < 0) { + return ret; + } else if (ret != s->fd) { + Error *local_err = NULL; + /* - * In the context of reopen, this function may be called several times - * (directly and recursively while change permissions of the parent). - * This is even true for children that don't inherit from the original - * reopen node, so s->reopen_state is not set. - * - * Ignore all but the first call. + * Fail already check_perm() if we can't get a working O_DIRECT + * alignment with the new fd. */ - return 0; - } - - if (s->reopen_state) { - /* We already have a new file descriptor to set permissions for */ - assert(s->reopen_state->perm == perm); - assert(s->reopen_state->shared_perm == shared); - rs = s->reopen_state->opaque; - s->perm_change_fd = rs->fd; - s->perm_change_flags = rs->open_flags; - } else { - /* We may need a new fd if auto-read-only switches the mode */ - ret = raw_reconfigure_getfd(bs, bs->open_flags, &open_flags, perm, - false, errp); - if (ret < 0) { - return ret; - } else if (ret != s->fd) { - s->perm_change_fd = ret; - s->perm_change_flags = open_flags; + raw_probe_alignment(bs, ret, &local_err); + if (local_err) { + error_propagate(errp, local_err); + return -EINVAL; } + + s->perm_change_fd = ret; + s->perm_change_flags = open_flags; } /* Prepare permissions on old fd to avoid conflicts between old and new, @@ -XXX,XX +XXX,XX @@ static int raw_check_perm(BlockDriverState *bs, uint64_t perm, uint64_t shared, return 0; fail: - if (s->perm_change_fd && !s->reopen_state) { + if (s->perm_change_fd) { qemu_close(s->perm_change_fd); } s->perm_change_fd = 0; @@ -XXX,XX +XXX,XX @@ static void raw_abort_perm_update(BlockDriverState *bs) /* For reopen, .bdrv_reopen_abort is called afterwards and will close * the file descriptor. */ - if (s->perm_change_fd && !s->reopen_state) { + if (s->perm_change_fd) { qemu_close(s->perm_change_fd); } s->perm_change_fd = 0; -- 2.30.2 From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Kevin Wolf <kwolf@redhat.com> Message-Id: <20210428151804.439460-32-vsementsov@virtuozzo.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block.c | 103 -------------------------------------------------------- 1 file changed, 103 deletions(-) diff --git a/block.c b/block.c index XXXXXXX..XXXXXXX 100644 --- a/block.c +++ b/block.c @@ -XXX,XX +XXX,XX @@ static int bdrv_fill_options(QDict **options, const char *filename, return 0; } -static int bdrv_check_update_perm(BlockDriverState *bs, BlockReopenQueue *q, - uint64_t new_used_perm, - uint64_t new_shared_perm, - Error **errp); - typedef struct BlockReopenQueueEntry { bool prepared; bool perms_checked; @@ -XXX,XX +XXX,XX @@ static int bdrv_check_perm_common(GSList *list, BlockReopenQueue *q, return 0; } -static int bdrv_check_perm(BlockDriverState *bs, BlockReopenQueue *q, - uint64_t cumulative_perms, - uint64_t cumulative_shared_perms, Error **errp) -{ - g_autoptr(GSList) list = bdrv_topological_dfs(NULL, NULL, bs); - return bdrv_check_perm_common(list, q, true, cumulative_perms, - cumulative_shared_perms, NULL, errp); -} - static int bdrv_list_refresh_perms(GSList *list, BlockReopenQueue *q, Transaction *tran, Error **errp) { return bdrv_check_perm_common(list, q, false, 0, 0, tran, errp); } -/* - * Notifies drivers that after a previous bdrv_check_perm() call, the - * permission update is not performed and any preparations made for it (e.g. - * taken file locks) need to be undone. - */ -static void bdrv_node_abort_perm_update(BlockDriverState *bs) -{ - BlockDriver *drv = bs->drv; - BdrvChild *c; - - if (!drv) { - return; - } - - bdrv_drv_set_perm_abort(bs); - - QLIST_FOREACH(c, &bs->children, next) { - bdrv_child_set_perm_abort(c); - } -} - -static void bdrv_list_abort_perm_update(GSList *list) -{ - for ( ; list; list = list->next) { - bdrv_node_abort_perm_update((BlockDriverState *)list->data); - } -} - -__attribute__((unused)) -static void bdrv_abort_perm_update(BlockDriverState *bs) -{ - g_autoptr(GSList) list = bdrv_topological_dfs(NULL, NULL, bs); - return bdrv_list_abort_perm_update(list); -} - static void bdrv_node_set_perm(BlockDriverState *bs) { BlockDriver *drv = bs->drv; @@ -XXX,XX +XXX,XX @@ char *bdrv_perm_names(uint64_t perm) return g_string_free(result, FALSE); } -/* - * Checks whether a new reference to @bs can be added if the new user requires - * @new_used_perm/@new_shared_perm as its permissions. If @ignore_children is - * set, the BdrvChild objects in this list are ignored in the calculations; - * this allows checking permission updates for an existing reference. - * - * Needs to be followed by a call to either bdrv_set_perm() or - * bdrv_abort_perm_update(). */ -__attribute__((unused)) -static int bdrv_check_update_perm(BlockDriverState *bs, BlockReopenQueue *q, - uint64_t new_used_perm, - uint64_t new_shared_perm, - Error **errp) -{ - BdrvChild *c; - uint64_t cumulative_perms = new_used_perm; - uint64_t cumulative_shared_perms = new_shared_perm; - - - /* There is no reason why anyone couldn't tolerate write_unchanged */ - assert(new_shared_perm & BLK_PERM_WRITE_UNCHANGED); - - QLIST_FOREACH(c, &bs->parents, next_parent) { - if ((new_used_perm & c->shared_perm) != new_used_perm) { - char *user = bdrv_child_user_desc(c); - char *perm_names = bdrv_perm_names(new_used_perm & ~c->shared_perm); - - error_setg(errp, "Conflicts with use by %s as '%s', which does not " - "allow '%s' on %s", - user, c->name, perm_names, bdrv_get_node_name(c->bs)); - g_free(user); - g_free(perm_names); - return -EPERM; - } - - if ((c->perm & new_shared_perm) != c->perm) { - char *user = bdrv_child_user_desc(c); - char *perm_names = bdrv_perm_names(c->perm & ~new_shared_perm); - - error_setg(errp, "Conflicts with use by %s as '%s', which uses " - "'%s' on %s", - user, c->name, perm_names, bdrv_get_node_name(c->bs)); - g_free(user); - g_free(perm_names); - return -EPERM; - } - - cumulative_perms |= c->perm; - cumulative_shared_perms &= c->shared_perm; - } - - return bdrv_check_perm(bs, q, cumulative_perms, cumulative_shared_perms, - errp); -} static int bdrv_refresh_perms(BlockDriverState *bs, Error **errp) { -- 2.30.2 From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> bdrv_check_perm_common() has only one caller, so no more sense in "common". Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Kevin Wolf <kwolf@redhat.com> Message-Id: <20210428151804.439460-33-vsementsov@virtuozzo.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block.c | 32 +++----------------------------- 1 file changed, 3 insertions(+), 29 deletions(-) diff --git a/block.c b/block.c index XXXXXXX..XXXXXXX 100644 --- a/block.c +++ b/block.c @@ -XXX,XX +XXX,XX @@ static int bdrv_node_check_perm(BlockDriverState *bs, BlockReopenQueue *q, return 0; } -/* - * If use_cumulative_perms is true, use cumulative_perms and - * cumulative_shared_perms for first element of the list. Otherwise just refresh - * all permissions. - */ -static int bdrv_check_perm_common(GSList *list, BlockReopenQueue *q, - bool use_cumulative_perms, - uint64_t cumulative_perms, - uint64_t cumulative_shared_perms, - Transaction *tran, Error **errp) +static int bdrv_list_refresh_perms(GSList *list, BlockReopenQueue *q, + Transaction *tran, Error **errp) { int ret; + uint64_t cumulative_perms, cumulative_shared_perms; BlockDriverState *bs; - if (use_cumulative_perms) { - bs = list->data; - - ret = bdrv_node_check_perm(bs, q, cumulative_perms, - cumulative_shared_perms, - tran, errp); - if (ret < 0) { - return ret; - } - - list = list->next; - } - for ( ; list; list = list->next) { bs = list->data; @@ -XXX,XX +XXX,XX @@ static int bdrv_check_perm_common(GSList *list, BlockReopenQueue *q, return 0; } -static int bdrv_list_refresh_perms(GSList *list, BlockReopenQueue *q, - Transaction *tran, Error **errp) -{ - return bdrv_check_perm_common(list, q, false, 0, 0, tran, errp); -} - static void bdrv_node_set_perm(BlockDriverState *bs) { BlockDriver *drv = bs->drv; -- 2.30.2 From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> bdrv_replace_child() has only one caller, the second argument is unused. Inline it now. This triggers deletion of some more unused interfaces. Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Kevin Wolf <kwolf@redhat.com> Message-Id: <20210428151804.439460-34-vsementsov@virtuozzo.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block.c | 101 ++++++++++---------------------------------------------- 1 file changed, 18 insertions(+), 83 deletions(-) diff --git a/block.c b/block.c index XXXXXXX..XXXXXXX 100644 --- a/block.c +++ b/block.c @@ -XXX,XX +XXX,XX @@ static int bdrv_list_refresh_perms(GSList *list, BlockReopenQueue *q, return 0; } -static void bdrv_node_set_perm(BlockDriverState *bs) -{ - BlockDriver *drv = bs->drv; - BdrvChild *c; - - if (!drv) { - return; - } - - bdrv_drv_set_perm_commit(bs); - - /* Drivers that never have children can omit .bdrv_child_perm() */ - if (!drv->bdrv_child_perm) { - assert(QLIST_EMPTY(&bs->children)); - return; - } - - /* Update all children */ - QLIST_FOREACH(c, &bs->children, next) { - bdrv_child_set_perm_commit(c); - } -} - -static void bdrv_list_set_perm(GSList *list) -{ - for ( ; list; list = list->next) { - bdrv_node_set_perm((BlockDriverState *)list->data); - } -} - -static void bdrv_set_perm(BlockDriverState *bs) -{ - g_autoptr(GSList) list = bdrv_topological_dfs(NULL, NULL, bs); - return bdrv_list_set_perm(list); -} - void bdrv_get_cumulative_perm(BlockDriverState *bs, uint64_t *perm, uint64_t *shared_perm) { @@ -XXX,XX +XXX,XX @@ static void bdrv_replace_child_noperm(BdrvChild *child, } } -/* - * Updates @child to change its reference to point to @new_bs, including - * checking and applying the necessary permission updates both to the old node - * and to @new_bs. - * - * NULL is passed as @new_bs for removing the reference before freeing @child. - * - * If @new_bs is not NULL, bdrv_check_perm() must be called beforehand, as this - * function uses bdrv_set_perm() to update the permissions according to the new - * reference that @new_bs gets. - * - * Callers must ensure that child->frozen is false. - */ -static void bdrv_replace_child(BdrvChild *child, BlockDriverState *new_bs) -{ - BlockDriverState *old_bs = child->bs; - - /* Asserts that child->frozen == false */ - bdrv_replace_child_noperm(child, new_bs); - - /* - * Start with the new node's permissions. If @new_bs is a (direct - * or indirect) child of @old_bs, we must complete the permission - * update on @new_bs before we loosen the restrictions on @old_bs. - * Otherwise, bdrv_check_perm() on @old_bs would re-initiate - * updating the permissions of @new_bs, and thus not purely loosen - * restrictions. - */ - if (new_bs) { - bdrv_set_perm(new_bs); - } - - if (old_bs) { - /* - * Update permissions for old node. We're just taking a parent away, so - * we're loosening restrictions. Errors of permission update are not - * fatal in this case, ignore them. - */ - bdrv_refresh_perms(old_bs, NULL); - - /* When the parent requiring a non-default AioContext is removed, the - * node moves back to the main AioContext */ - bdrv_try_set_aio_context(old_bs, qemu_get_aio_context(), NULL); - } -} - static void bdrv_child_free(void *opaque) { BdrvChild *c = opaque; @@ -XXX,XX +XXX,XX @@ static int bdrv_attach_child_noperm(BlockDriverState *parent_bs, static void bdrv_detach_child(BdrvChild *child) { - bdrv_replace_child(child, NULL); + BlockDriverState *old_bs = child->bs; + + bdrv_replace_child_noperm(child, NULL); bdrv_remove_empty_child(child); + + if (old_bs) { + /* + * Update permissions for old node. We're just taking a parent away, so + * we're loosening restrictions. Errors of permission update are not + * fatal in this case, ignore them. + */ + bdrv_refresh_perms(old_bs, NULL); + + /* + * When the parent requiring a non-default AioContext is removed, the + * node moves back to the main AioContext + */ + bdrv_try_set_aio_context(old_bs, qemu_get_aio_context(), NULL); + } } /* -- 2.30.2 From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Old interfaces dropped, nobody directly calls bdrv_child_set_perm_abort() and bdrv_child_set_perm_commit(), so we can use personal state structure for the action and stop exploiting BdrvChild structure. Also, drop "_safe" suffix which is redundant now. Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Kevin Wolf <kwolf@redhat.com> Message-Id: <20210428151804.439460-35-vsementsov@virtuozzo.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- include/block/block_int.h | 5 ---- block.c | 63 ++++++++++++++------------------------- 2 files changed, 22 insertions(+), 46 deletions(-) diff --git a/include/block/block_int.h b/include/block/block_int.h index XXXXXXX..XXXXXXX 100644 --- a/include/block/block_int.h +++ b/include/block/block_int.h @@ -XXX,XX +XXX,XX @@ struct BdrvChild { */ uint64_t shared_perm; - /* backup of permissions during permission update procedure */ - bool has_backup_perm; - uint64_t backup_perm; - uint64_t backup_shared_perm; - /* * This link is frozen: the child can neither be replaced nor * detached from the parent. diff --git a/block.c b/block.c index XXXXXXX..XXXXXXX 100644 --- a/block.c +++ b/block.c @@ -XXX,XX +XXX,XX @@ static GSList *bdrv_topological_dfs(GSList *list, GHashTable *found, return g_slist_prepend(list, bs); } -static void bdrv_child_set_perm_commit(void *opaque) -{ - BdrvChild *c = opaque; - - c->has_backup_perm = false; -} +typedef struct BdrvChildSetPermState { + BdrvChild *child; + uint64_t old_perm; + uint64_t old_shared_perm; +} BdrvChildSetPermState; static void bdrv_child_set_perm_abort(void *opaque) { - BdrvChild *c = opaque; - /* - * We may have child->has_backup_perm unset at this point, as in case of - * _check_ stage of permission update failure we may _check_ not the whole - * subtree. Still, _abort_ is called on the whole subtree anyway. - */ - if (c->has_backup_perm) { - c->perm = c->backup_perm; - c->shared_perm = c->backup_shared_perm; - c->has_backup_perm = false; - } + BdrvChildSetPermState *s = opaque; + + s->child->perm = s->old_perm; + s->child->shared_perm = s->old_shared_perm; } static TransactionActionDrv bdrv_child_set_pem_drv = { .abort = bdrv_child_set_perm_abort, - .commit = bdrv_child_set_perm_commit, + .clean = g_free, }; -/* - * With tran=NULL needs to be followed by direct call to either - * bdrv_child_set_perm_commit() or bdrv_child_set_perm_abort(). - * - * With non-NULL tran needs to be followed by tran_abort() or tran_commit() - * instead. - */ -static void bdrv_child_set_perm_safe(BdrvChild *c, uint64_t perm, - uint64_t shared, Transaction *tran) +static void bdrv_child_set_perm(BdrvChild *c, uint64_t perm, + uint64_t shared, Transaction *tran) { - if (!c->has_backup_perm) { - c->has_backup_perm = true; - c->backup_perm = c->perm; - c->backup_shared_perm = c->shared_perm; - } - /* - * Note: it's OK if c->has_backup_perm was already set, as we can find the - * same c twice during check_perm procedure - */ + BdrvChildSetPermState *s = g_new(BdrvChildSetPermState, 1); + + *s = (BdrvChildSetPermState) { + .child = c, + .old_perm = c->perm, + .old_shared_perm = c->shared_perm, + }; c->perm = perm; c->shared_perm = shared; - if (tran) { - tran_add(tran, &bdrv_child_set_pem_drv, c); - } + tran_add(tran, &bdrv_child_set_pem_drv, s); } static void bdrv_drv_set_perm_commit(void *opaque) @@ -XXX,XX +XXX,XX @@ static int bdrv_node_check_perm(BlockDriverState *bs, BlockReopenQueue *q, bdrv_child_perm(bs, c->bs, c, c->role, q, cumulative_perms, cumulative_shared_perms, &cur_perm, &cur_shared); - bdrv_child_set_perm_safe(c, cur_perm, cur_shared, tran); + bdrv_child_set_perm(c, cur_perm, cur_shared, tran); } return 0; @@ -XXX,XX +XXX,XX @@ int bdrv_child_try_set_perm(BdrvChild *c, uint64_t perm, uint64_t shared, Transaction *tran = tran_new(); int ret; - bdrv_child_set_perm_safe(c, perm, shared, tran); + bdrv_child_set_perm(c, perm, shared, tran); ret = bdrv_refresh_perms(c->bs, &local_err); -- 2.30.2 From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> We don't have bdrv_replace_child(), so it's time for bdrv_replace_child_safe() to take its place. Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Kevin Wolf <kwolf@redhat.com> Message-Id: <20210428151804.439460-36-vsementsov@virtuozzo.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/block.c b/block.c index XXXXXXX..XXXXXXX 100644 --- a/block.c +++ b/block.c @@ -XXX,XX +XXX,XX @@ static TransactionActionDrv bdrv_replace_child_drv = { }; /* - * bdrv_replace_child_safe + * bdrv_replace_child * * Note: real unref of old_bs is done only on commit. */ -static void bdrv_replace_child_safe(BdrvChild *child, BlockDriverState *new_bs, - Transaction *tran) +static void bdrv_replace_child(BdrvChild *child, BlockDriverState *new_bs, + Transaction *tran) { BdrvReplaceChildState *s = g_new(BdrvReplaceChildState, 1); *s = (BdrvReplaceChildState) { @@ -XXX,XX +XXX,XX @@ static void bdrv_remove_filter_or_cow_child(BlockDriverState *bs, } if (child->bs) { - bdrv_replace_child_safe(child, NULL, tran); + bdrv_replace_child(child, NULL, tran); } s = g_new(BdrvRemoveFilterOrCowChild, 1); @@ -XXX,XX +XXX,XX @@ static int bdrv_replace_node_noperm(BlockDriverState *from, c->name, from->node_name); return -EPERM; } - bdrv_replace_child_safe(c, to, tran); + bdrv_replace_child(c, to, tran); } return 0; -- 2.30.2 From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Now, bdrv_node_check_perm() is called only with fresh cumulative permissions, so its actually "refresh_perm". Move permission calculation to the function. Also, drop unreachable error message and rewrite the remaining one to be more generic (as now we don't know which node is added and which was already here). Add also Virtuozzo copyright, as big work is done at this point. Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Kevin Wolf <kwolf@redhat.com> Message-Id: <20210428151804.439460-37-vsementsov@virtuozzo.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- block.c | 38 +++++++++++--------------------------- tests/qemu-iotests/245 | 2 +- 2 files changed, 12 insertions(+), 28 deletions(-) diff --git a/block.c b/block.c index XXXXXXX..XXXXXXX 100644 --- a/block.c +++ b/block.c @@ -XXX,XX +XXX,XX @@ * QEMU System Emulator block driver * * Copyright (c) 2003 Fabrice Bellard + * Copyright (c) 2020 Virtuozzo International GmbH. * * Permission is hereby granted, free of charge, to any person obtaining a copy * of this software and associated documentation files (the "Software"), to deal @@ -XXX,XX +XXX,XX @@ static void bdrv_replace_child(BdrvChild *child, BlockDriverState *new_bs, } /* - * Check whether permissions on this node can be changed in a way that - * @cumulative_perms and @cumulative_shared_perms are the new cumulative - * permissions of all its parents. This involves checking whether all necessary - * permission changes to child nodes can be performed. - * - * A call to this function must always be followed by a call to bdrv_set_perm() - * or bdrv_abort_perm_update(). + * Refresh permissions in @bs subtree. The function is intended to be called + * after some graph modification that was done without permission update. */ -static int bdrv_node_check_perm(BlockDriverState *bs, BlockReopenQueue *q, - uint64_t cumulative_perms, - uint64_t cumulative_shared_perms, - Transaction *tran, Error **errp) +static int bdrv_node_refresh_perm(BlockDriverState *bs, BlockReopenQueue *q, + Transaction *tran, Error **errp) { BlockDriver *drv = bs->drv; BdrvChild *c; int ret; + uint64_t cumulative_perms, cumulative_shared_perms; + + bdrv_get_cumulative_perm(bs, &cumulative_perms, &cumulative_shared_perms); /* Write permissions never work with read-only images */ if ((cumulative_perms & (BLK_PERM_WRITE | BLK_PERM_WRITE_UNCHANGED)) && @@ -XXX,XX +XXX,XX @@ static int bdrv_node_check_perm(BlockDriverState *bs, BlockReopenQueue *q, if (!bdrv_is_writable_after_reopen(bs, NULL)) { error_setg(errp, "Block node is read-only"); } else { - uint64_t current_perms, current_shared; - bdrv_get_cumulative_perm(bs, &current_perms, &current_shared); - if (current_perms & (BLK_PERM_WRITE | BLK_PERM_WRITE_UNCHANGED)) { - error_setg(errp, "Cannot make block node read-only, there is " - "a writer on it"); - } else { - error_setg(errp, "Cannot make block node read-only and create " - "a writer on it"); - } + error_setg(errp, "Read-only block node '%s' cannot support " + "read-write users", bdrv_get_node_name(bs)); } return -EPERM; @@ -XXX,XX +XXX,XX @@ static int bdrv_list_refresh_perms(GSList *list, BlockReopenQueue *q, Transaction *tran, Error **errp) { int ret; - uint64_t cumulative_perms, cumulative_shared_perms; BlockDriverState *bs; for ( ; list; list = list->next) { @@ -XXX,XX +XXX,XX @@ static int bdrv_list_refresh_perms(GSList *list, BlockReopenQueue *q, return -EINVAL; } - bdrv_get_cumulative_perm(bs, &cumulative_perms, - &cumulative_shared_perms); - - ret = bdrv_node_check_perm(bs, q, cumulative_perms, - cumulative_shared_perms, - tran, errp); + ret = bdrv_node_refresh_perm(bs, q, tran, errp); if (ret < 0) { return ret; } diff --git a/tests/qemu-iotests/245 b/tests/qemu-iotests/245 index XXXXXXX..XXXXXXX 100755 --- a/tests/qemu-iotests/245 +++ b/tests/qemu-iotests/245 @@ -XXX,XX +XXX,XX @@ class TestBlockdevReopen(iotests.QMPTestCase): # We can't reopen hd1 to read-only, as block-stream requires it to be # read-write self.reopen(opts['backing'], {'read-only': True}, - "Cannot make block node read-only, there is a writer on it") + "Read-only block node 'hd1' cannot support read-write users") # We can't remove hd2 while the stream job is ongoing opts['backing']['backing'] = None -- 2.30.2 Normally, blk_new_open() just shares all permissions. This was fine originally when permissions only protected against uses in the same process because no other part of the code would actually get to access the block nodes opened with blk_new_open(). However, since we use it for file locking now, unsharing permissions becomes desirable. Add a new BDRV_O_NO_SHARE flag that is used in blk_new_open() to unshare any permissions that can be unshared. Signed-off-by: Kevin Wolf <kwolf@redhat.com> Message-Id: <20210422164344.283389-2-kwolf@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- include/block/block.h | 1 + block/block-backend.c | 19 +++++++++++++------ 2 files changed, 14 insertions(+), 6 deletions(-) diff --git a/include/block/block.h b/include/block/block.h index XXXXXXX..XXXXXXX 100644 --- a/include/block/block.h +++ b/include/block/block.h @@ -XXX,XX +XXX,XX @@ typedef struct HDGeometry { uint32_t cylinders; } HDGeometry; +#define BDRV_O_NO_SHARE 0x0001 /* don't share permissions */ #define BDRV_O_RDWR 0x0002 #define BDRV_O_RESIZE 0x0004 /* request permission for resizing the node */ #define BDRV_O_SNAPSHOT 0x0008 /* open the file read only and save writes in a snapshot */ diff --git a/block/block-backend.c b/block/block-backend.c index XXXXXXX..XXXXXXX 100644 --- a/block/block-backend.c +++ b/block/block-backend.c @@ -XXX,XX +XXX,XX @@ BlockBackend *blk_new_open(const char *filename, const char *reference, BlockBackend *blk; BlockDriverState *bs; uint64_t perm = 0; + uint64_t shared = BLK_PERM_ALL; - /* blk_new_open() is mainly used in .bdrv_create implementations and the - * tools where sharing isn't a concern because the BDS stays private, so we - * just request permission according to the flags. + /* + * blk_new_open() is mainly used in .bdrv_create implementations and the + * tools where sharing isn't a major concern because the BDS stays private + * and the file is generally not supposed to be used by a second process, + * so we just request permission according to the flags. * * The exceptions are xen_disk and blockdev_init(); in these cases, the * caller of blk_new_open() doesn't make use of the permissions, but they * shouldn't hurt either. We can still share everything here because the - * guest devices will add their own blockers if they can't share. */ + * guest devices will add their own blockers if they can't share. + */ if ((flags & BDRV_O_NO_IO) == 0) { perm |= BLK_PERM_CONSISTENT_READ; if (flags & BDRV_O_RDWR) { @@ -XXX,XX +XXX,XX @@ BlockBackend *blk_new_open(const char *filename, const char *reference, if (flags & BDRV_O_RESIZE) { perm |= BLK_PERM_RESIZE; } + if (flags & BDRV_O_NO_SHARE) { + shared = BLK_PERM_CONSISTENT_READ | BLK_PERM_WRITE_UNCHANGED; + } - blk = blk_new(qemu_get_aio_context(), perm, BLK_PERM_ALL); + blk = blk_new(qemu_get_aio_context(), perm, shared); bs = bdrv_open(filename, reference, options, flags, errp); if (!bs) { blk_unref(blk); @@ -XXX,XX +XXX,XX @@ BlockBackend *blk_new_open(const char *filename, const char *reference, blk->root = bdrv_root_attach_child(bs, "root", &child_root, BDRV_CHILD_FILTERED | BDRV_CHILD_PRIMARY, - perm, BLK_PERM_ALL, blk, errp); + perm, shared, blk, errp); if (!blk->root) { blk_unref(blk); return NULL; -- 2.30.2 For a successful conversion of an image, we must make sure that its content doesn't change during the conversion. A special case of this is using the same image file both as the source and as the destination. If both input and output format are raw, the operation would just be useless work, with other formats it is a sure way to destroy the image. This will now fail because the image file can't be opened a second time for the output when opening it for the input has already acquired file locks to unshare BLK_PERM_WRITE. Nevertheless, if there is some reason in a special case why it is actually okay to allow writes to the image while it is being converted, -U can still be used to force sharing all permissions. Note that for most image formats, BLK_PERM_WRITE would already be unshared by the format driver, so this only really makes a difference for raw source images (but any output format). Reported-by: Xueqiang Wei <xuwei@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> Message-Id: <20210422164344.283389-3-kwolf@redhat.com> Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- qemu-img.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/qemu-img.c b/qemu-img.c index XXXXXXX..XXXXXXX 100644 --- a/qemu-img.c +++ b/qemu-img.c @@ -XXX,XX +XXX,XX @@ static void set_rate_limit(BlockBackend *blk, int64_t rate_limit) static int img_convert(int argc, char **argv) { - int c, bs_i, flags, src_flags = 0; + int c, bs_i, flags, src_flags = BDRV_O_NO_SHARE; const char *fmt = NULL, *out_fmt = NULL, *cache = "unsafe", *src_cache = BDRV_DEFAULT_CACHE, *out_baseimg = NULL, *out_filename, *out_baseimg_param, *snapshot_name = NULL; -- 2.30.2 virtio_add_queue() aborts when queue_size > VIRTQUEUE_MAX_SIZE, so vhost_user_blk_device_realize() should check this before calling it. Simple reproducer: qemu-system-x86_64 \ -chardev null,id=foo \ -device vhost-user-blk-pci,queue-size=4096,chardev=foo Fixes: https://bugzilla.redhat.com/show_bug.cgi?id=1935014 Signed-off-by: Kevin Wolf <kwolf@redhat.com> Message-Id: <20210413165654.50810-1-kwolf@redhat.com> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Raphael Norwitz <raphael.norwitz@nutanix.com> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com> Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com> --- hw/block/vhost-user-blk.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/hw/block/vhost-user-blk.c b/hw/block/vhost-user-blk.c index XXXXXXX..XXXXXXX 100644 --- a/hw/block/vhost-user-blk.c +++ b/hw/block/vhost-user-blk.c @@ -XXX,XX +XXX,XX @@ static void vhost_user_blk_device_realize(DeviceState *dev, Error **errp) error_setg(errp, "vhost-user-blk: queue size must be non-zero"); return; } + if (s->queue_size > VIRTQUEUE_MAX_SIZE) { + error_setg(errp, "vhost-user-blk: queue size must not exceed %d", + VIRTQUEUE_MAX_SIZE); + return; + } if (!vhost_user_init(&s->vhost_user, &s->chardev, errp)) { return; -- 2.30.2