[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [PATCH v3 1/5] block/io: fix bdrv_co_block_status_above
From: |
Eric Blake |
Subject: |
Re: [PATCH v3 1/5] block/io: fix bdrv_co_block_status_above |
Date: |
Wed, 20 May 2020 09:38:30 -0500 |
User-agent: |
Mozilla/5.0 (X11; Linux x86_64; rv:68.0) Gecko/20100101 Thunderbird/68.7.0 |
On 5/20/20 3:35 AM, Vladimir Sementsov-Ogievskiy wrote:
bdrv_co_block_status_above has several design problems with handling
short backing files:
1. With want_zeros=true, it may return ret with BDRV_BLOCK_ZERO but
without BDRV_BLOCK_ALLOCATED flag, when actually short backing file
which produces these after-EOF zeros is inside requested backing
sequence.
2. With want_zero=false, it may return pnum=0 prior to actual EOF,
because of EOF of short backing file.
Fix these things, making logic about short backing files clearer.
With fixed bdrv_block_status_above we also have to improve is_zero in
qcow2 code, otherwise iotest 154 will fail, because with this patch we
stop to merge zeros of different types (produced by fully unallocated
in the whole backing chain regions vs produced by short backing files).
Note also, that this patch leaves for another day the general problem
around block-status: misuse of BDRV_BLOCK_ALLOCATED as is-fs-allocated
vs go-to-backing.
Signed-off-by: Vladimir Sementsov-Ogievskiy <address@hidden>
---
block/io.c | 39 ++++++++++++++++++++++++++++++---------
block/qcow2.c | 16 ++++++++++++++--
2 files changed, 44 insertions(+), 11 deletions(-)
Reviewed-by: Eric Blake <address@hidden>
+++ b/block/qcow2.c
@@ -3827,8 +3827,20 @@ static bool is_zero(BlockDriverState *bs, int64_t
offset, int64_t bytes)
if (!bytes) {
return true;
}
- res = bdrv_block_status_above(bs, NULL, offset, bytes, &nr, NULL, NULL);
- return res >= 0 && (res & BDRV_BLOCK_ZERO) && nr == bytes;
+
+ /*
+ * bdrv_block_status_above doesn't merge different types of zeros, for
+ * example, zeros which comes from the region which is unallocated in
s/comes/come/
+ * the whole backing chain, and zeros which comes because of short backing
and again, also s/of/of a/
+ * file. So, we need a loop.
+ */
+ do {
+ res = bdrv_block_status_above(bs, NULL, offset, bytes, &nr, NULL,
NULL);
+ offset += nr;
+ bytes -= nr;
+ } while (res >= 0 && (res & BDRV_BLOCK_ZERO) && nr && bytes);
+
+ return res >= 0 && (res & BDRV_BLOCK_ZERO) && bytes == 0;
}
static coroutine_fn int qcow2_co_pwrite_zeroes(BlockDriverState *bs,
--
Eric Blake, Principal Software Engineer
Red Hat, Inc. +1-919-301-3226
Virtualization: qemu.org | libvirt.org
- [PATCH v3 0/5] fix & merge block_status_above and is_allocated_above, Vladimir Sementsov-Ogievskiy, 2020/05/20
- [PATCH v3 1/5] block/io: fix bdrv_co_block_status_above, Vladimir Sementsov-Ogievskiy, 2020/05/20
- Re: [PATCH v3 1/5] block/io: fix bdrv_co_block_status_above,
Eric Blake <=
- [PATCH v3 2/5] block/io: bdrv_common_block_status_above: support include_base, Vladimir Sementsov-Ogievskiy, 2020/05/20
- [PATCH v3 3/5] block/io: bdrv_common_block_status_above: support bs == base, Vladimir Sementsov-Ogievskiy, 2020/05/20
- [PATCH v3 4/5] block/io: fix bdrv_is_allocated_above, Vladimir Sementsov-Ogievskiy, 2020/05/20
- [PATCH v3 5/5] iotests: add commit top->base cases to 274, Vladimir Sementsov-Ogievskiy, 2020/05/20