qemu-kvm/kvm-qemu-img-make-is_allocated_sectors-more-efficient.patch
Miroslav Rezanina 0daf0004a7 * Mon Feb 14 2022 Miroslav Rezanina <mrezanin@redhat.com> - 6.2.0-8
- kvm-numa-Enable-numa-for-SGX-EPC-sections.patch [bz#2033708]
- kvm-numa-Support-SGX-numa-in-the-monitor-and-Libvirt-int.patch [bz#2033708]
- kvm-doc-Add-the-SGX-numa-description.patch [bz#2033708]
- kvm-Enable-SGX-RH-Only.patch [bz#2033708]
- kvm-qapi-Cleanup-SGX-related-comments-and-restore-sectio.patch [bz#2033708]
- kvm-block-io-Update-BSC-only-if-want_zero-is-true.patch [bz#2041461]
- kvm-iotests-block-status-cache-New-test.patch [bz#2041461]
- kvm-iotests-Test-qemu-img-convert-of-zeroed-data-cluster.patch [bz#1882917]
- kvm-qemu-img-make-is_allocated_sectors-more-efficient.patch [bz#1882917]
- kvm-block-backend-prevent-dangling-BDS-pointers-across-a.patch [bz#2040123]
- kvm-iotests-stream-error-on-reset-New-test.patch [bz#2040123]
- kvm-hw-arm-smmuv3-Fix-device-reset.patch [bz#2042481]
- Resolves: bz#2033708
  ([Intel 9.0 Feat] qemu-kvm: SGX 1.5 (SGX1 + Flexible Launch Control) support)
- Resolves: bz#2041461
  (Inconsistent block status reply in qemu-nbd)
- Resolves: bz#1882917
  (the target image size is incorrect when converting a badly fragmented file)
- Resolves: bz#2040123
  (Qemu core dumped when do block-stream to a snapshot node on non-enough space storage)
- Resolves: bz#2042481
  ([aarch64] Launch guest with "default-bus-bypass-iommu=off,iommu=smmuv3" and "iommu_platform=on", guest hangs after system_reset)
2022-02-14 06:48:51 -05:00

109 lines
4.6 KiB
Diff

From a221f5a8ed02690687e6709c49ae0e1e01c5f466 Mon Sep 17 00:00:00 2001
From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Date: Fri, 17 Dec 2021 17:46:54 +0100
Subject: [PATCH 09/12] qemu-img: make is_allocated_sectors() more efficient
RH-Author: Kevin Wolf <kwolf@redhat.com>
RH-MergeRequest: 70: qemu-img convert: Fix sparseness of output image
RH-Commit: [2/2] cc05aa4ac506b57ff9b430c007618cdf1485a03f (kmwolf/centos-qemu-kvm)
RH-Bugzilla: 1882917
RH-Acked-by: Eric Blake <eblake@redhat.com>
RH-Acked-by: Stefan Hajnoczi <stefanha@redhat.com>
RH-Acked-by: Hanna Reitz <hreitz@redhat.com>
Consider the case when the whole buffer is zero and end is unaligned.
If i <= tail, we return 1 and do one unaligned WRITE, RMW happens.
If i > tail, we do on aligned WRITE_ZERO (or skip if target is zeroed)
and again one unaligned WRITE, RMW happens.
Let's do better: don't fragment the whole-zero buffer and report it as
ZERO: in case of zeroed target we just do nothing and avoid RMW. If
target is not zeroes, one unaligned WRITE_ZERO should not be much worse
than one unaligned WRITE.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20211217164654.1184218-3-vsementsov@virtuozzo.com>
Tested-by: Peter Lieven <pl@kamp.de>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
(cherry picked from commit 96054c76ff2db74165385a69f234c57a6bbc941e)
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
qemu-img.c | 23 +++++++++++++++++++----
tests/qemu-iotests/122.out | 8 ++------
2 files changed, 21 insertions(+), 10 deletions(-)
diff --git a/qemu-img.c b/qemu-img.c
index f036a1d428..d7ddfcc528 100644
--- a/qemu-img.c
+++ b/qemu-img.c
@@ -1171,19 +1171,34 @@ static int is_allocated_sectors(const uint8_t *buf, int n, int *pnum,
}
}
+ if (i == n) {
+ /*
+ * The whole buf is the same.
+ * No reason to split it into chunks, so return now.
+ */
+ *pnum = i;
+ return !is_zero;
+ }
+
tail = (sector_num + i) & (alignment - 1);
if (tail) {
if (is_zero && i <= tail) {
- /* treat unallocated areas which only consist
- * of a small tail as allocated. */
+ /*
+ * For sure next sector after i is data, and it will rewrite this
+ * tail anyway due to RMW. So, let's just write data now.
+ */
is_zero = false;
}
if (!is_zero) {
- /* align up end offset of allocated areas. */
+ /* If possible, align up end offset of allocated areas. */
i += alignment - tail;
i = MIN(i, n);
} else {
- /* align down end offset of zero areas. */
+ /*
+ * For sure next sector after i is data, and it will rewrite this
+ * tail anyway due to RMW. Better is avoid RMW and write zeroes up
+ * to aligned bound.
+ */
i -= tail;
}
}
diff --git a/tests/qemu-iotests/122.out b/tests/qemu-iotests/122.out
index 69b8e8b803..e18766e167 100644
--- a/tests/qemu-iotests/122.out
+++ b/tests/qemu-iotests/122.out
@@ -201,9 +201,7 @@ convert -S 4k
{ "start": 8192, "length": 4096, "depth": 0, "present": true, "zero": false, "data": true, "offset": OFFSET},
{ "start": 12288, "length": 4096, "depth": 0, "present": false, "zero": true, "data": false},
{ "start": 16384, "length": 4096, "depth": 0, "present": true, "zero": false, "data": true, "offset": OFFSET},
-{ "start": 20480, "length": 46080, "depth": 0, "present": false, "zero": true, "data": false},
-{ "start": 66560, "length": 1024, "depth": 0, "present": true, "zero": false, "data": true, "offset": OFFSET},
-{ "start": 67584, "length": 67041280, "depth": 0, "present": false, "zero": true, "data": false}]
+{ "start": 20480, "length": 67088384, "depth": 0, "present": false, "zero": true, "data": false}]
convert -c -S 4k
[{ "start": 0, "length": 1024, "depth": 0, "present": true, "zero": false, "data": true},
@@ -215,9 +213,7 @@ convert -c -S 4k
convert -S 8k
[{ "start": 0, "length": 24576, "depth": 0, "present": true, "zero": false, "data": true, "offset": OFFSET},
-{ "start": 24576, "length": 41984, "depth": 0, "present": false, "zero": true, "data": false},
-{ "start": 66560, "length": 1024, "depth": 0, "present": true, "zero": false, "data": true, "offset": OFFSET},
-{ "start": 67584, "length": 67041280, "depth": 0, "present": false, "zero": true, "data": false}]
+{ "start": 24576, "length": 67084288, "depth": 0, "present": false, "zero": true, "data": false}]
convert -c -S 8k
[{ "start": 0, "length": 1024, "depth": 0, "present": true, "zero": false, "data": true},
--
2.27.0