* [PULL 1/6] io: Add helper for setting socket send buffer size
2025-05-29 22:02 [PULL 0/6] NBD patches through 2025-05-29 Eric Blake
@ 2025-05-29 22:02 ` Eric Blake
2025-05-29 22:02 ` [PULL 2/6] nbd: Set unix socket send buffer on macOS Eric Blake
` (5 subsequent siblings)
6 siblings, 0 replies; 8+ messages in thread
From: Eric Blake @ 2025-05-29 22:02 UTC (permalink / raw)
To: qemu-devel; +Cc: Nir Soffer, Daniel P. Berrangé
From: Nir Soffer <nirsof@gmail.com>
Testing reading and writing from qemu-nbd using a unix domain socket
shows that the platform default send buffer size is too low, leading to
poor performance and hight cpu usage.
Add a helper for setting socket send buffer size to be used in NBD code.
It can also be used in other contexts.
We don't need a helper for receive buffer size since it is not used with
unix domain sockets. This is documented for Linux, and not documented
for macOS.
Failing to set the socket buffer size is not a fatal error, but the
caller may want to warn about the failure.
Signed-off-by: Nir Soffer <nirsof@gmail.com>
Message-ID: <20250517201154.88456-2-nirsof@gmail.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
---
include/io/channel-socket.h | 13 +++++++++++++
io/channel-socket.c | 11 +++++++++++
2 files changed, 24 insertions(+)
diff --git a/include/io/channel-socket.h b/include/io/channel-socket.h
index ab15577d384..a88cf8b3a9f 100644
--- a/include/io/channel-socket.h
+++ b/include/io/channel-socket.h
@@ -261,5 +261,18 @@ QIOChannelSocket *
qio_channel_socket_accept(QIOChannelSocket *ioc,
Error **errp);
+/**
+ * qio_channel_socket_set_send_buffer:
+ * @ioc: the socket channel object
+ * @size: buffer size
+ * @errp: pointer to a NULL-initialized error object
+ *
+ * Set the underlying socket send buffer size.
+ *
+ * Retruns: 0 on success, or -1 on error.
+ */
+int qio_channel_socket_set_send_buffer(QIOChannelSocket *ioc,
+ size_t size,
+ Error **errp);
#endif /* QIO_CHANNEL_SOCKET_H */
diff --git a/io/channel-socket.c b/io/channel-socket.c
index 088b49ffdb0..3b7ca924ff3 100644
--- a/io/channel-socket.c
+++ b/io/channel-socket.c
@@ -78,6 +78,17 @@ qio_channel_socket_new(void)
return sioc;
}
+int qio_channel_socket_set_send_buffer(QIOChannelSocket *ioc,
+ size_t size,
+ Error **errp)
+{
+ if (setsockopt(ioc->fd, SOL_SOCKET, SO_SNDBUF, &size, sizeof(size)) < 0) {
+ error_setg_errno(errp, errno, "Unable to set socket send buffer size");
+ return -1;
+ }
+
+ return 0;
+}
static int
qio_channel_socket_set_fd(QIOChannelSocket *sioc,
--
2.49.0
^ permalink raw reply related [flat|nested] 8+ messages in thread* [PULL 2/6] nbd: Set unix socket send buffer on macOS
2025-05-29 22:02 [PULL 0/6] NBD patches through 2025-05-29 Eric Blake
2025-05-29 22:02 ` [PULL 1/6] io: Add helper for setting socket send buffer size Eric Blake
@ 2025-05-29 22:02 ` Eric Blake
2025-05-29 22:02 ` [PULL 3/6] nbd: Set unix socket send buffer on Linux Eric Blake
` (4 subsequent siblings)
6 siblings, 0 replies; 8+ messages in thread
From: Eric Blake @ 2025-05-29 22:02 UTC (permalink / raw)
To: qemu-devel
Cc: Nir Soffer, Daniel P. Berrangé, Vladimir Sementsov-Ogievskiy,
open list:Network Block Dev...
From: Nir Soffer <nirsof@gmail.com>
On macOS we need to increase unix socket buffers size on the client and
server to get good performance. We set socket buffers on macOS after
connecting or accepting a client connection.
Testing shows that setting socket receive buffer size (SO_RCVBUF) has no
effect on performance, so we set only the send buffer size (SO_SNDBUF).
It seems to work like Linux but not documented.
Testing shows that optimal buffer size is 512k to 4 MiB, depending on
the test case. The difference is very small, so I chose 2 MiB.
I tested reading from qemu-nbd and writing to qemu-nbd with qemu-img and
computing a blkhash with nbdcopy and blksum.
To focus on NBD communication and get less noisy results, I tested
reading and writing to null-co driver. I added a read-pattern option to
the null-co driver to return data full of 0xff:
NULL="json:{'driver': 'raw', 'file': {'driver': 'null-co', 'size': '10g', 'read-pattern': 255}}"
For testing buffer size I added an environment variable for setting the
socket buffer size.
Read from qemu-nbd via qemu-img convert. In this test buffer size of 2m
is optimal (12.6 times faster).
qemu-nbd -r -t -e 0 -f raw -k /tmp/nbd.sock "$NULL" &
qemu-img convert -f raw -O raw -W -n "nbd+unix:///?socket=/tmp/nbd.sock" "$NULL"
| buffer size | time | user | system |
|-------------|---------|---------|---------|
| default | 13.361 | 2.653 | 5.702 |
| 65536 | 2.283 | 0.204 | 1.318 |
| 131072 | 1.673 | 0.062 | 1.008 |
| 262144 | 1.592 | 0.053 | 0.952 |
| 524288 | 1.496 | 0.049 | 0.887 |
| 1048576 | 1.234 | 0.047 | 0.738 |
| 2097152 | 1.060 | 0.080 | 0.602 |
| 4194304 | 1.061 | 0.076 | 0.604 |
Write to qemu-nbd with qemu-img convert. In this test buffer size of 2m
is optimal (9.2 times faster).
qemu-nbd -t -e 0 -f raw -k /tmp/nbd.sock "$NULL" &
qemu-img convert -f raw -O raw -W -n "$NULL" "nbd+unix:///?socket=/tmp/nbd.sock"
| buffer size | time | user | system |
|-------------|---------|---------|---------|
| default | 8.063 | 2.522 | 4.184 |
| 65536 | 1.472 | 0.430 | 0.867 |
| 131072 | 1.071 | 0.297 | 0.654 |
| 262144 | 1.012 | 0.239 | 0.587 |
| 524288 | 0.970 | 0.201 | 0.514 |
| 1048576 | 0.895 | 0.184 | 0.454 |
| 2097152 | 0.877 | 0.174 | 0.440 |
| 4194304 | 0.944 | 0.231 | 0.535 |
Compute a blkhash with nbdcopy, using 4 NBD connections and 256k request
size. In this test buffer size of 4m is optimal (5.1 times faster).
qemu-nbd -r -t -e 0 -f raw -k /tmp/nbd.sock "$NULL" &
nbdcopy --blkhash "nbd+unix:///?socket=/tmp/nbd.sock" null:
| buffer size | time | user | system |
|-------------|---------|---------|---------|
| default | 8.624 | 5.727 | 6.507 |
| 65536 | 2.563 | 4.760 | 2.498 |
| 131072 | 1.903 | 4.559 | 2.093 |
| 262144 | 1.759 | 4.513 | 1.935 |
| 524288 | 1.729 | 4.489 | 1.924 |
| 1048576 | 1.696 | 4.479 | 1.884 |
| 2097152 | 1.710 | 4.480 | 1.763 |
| 4194304 | 1.687 | 4.479 | 1.712 |
Compute a blkhash with blksum, using 1 NBD connection and 256k read
size. In this test buffer size of 512k is optimal (10.3 times faster).
qemu-nbd -r -t -e 0 -f raw -k /tmp/nbd.sock "$NULL" &
blksum "nbd+unix:///?socket=/tmp/nbd.sock"
| buffer size | time | user | system |
|-------------|---------|---------|---------|
| default | 13.085 | 5.664 | 6.461 |
| 65536 | 3.299 | 5.106 | 2.515 |
| 131072 | 2.396 | 4.989 | 2.069 |
| 262144 | 1.607 | 4.724 | 1.555 |
| 524288 | 1.271 | 4.528 | 1.224 |
| 1048576 | 1.294 | 4.565 | 1.333 |
| 2097152 | 1.299 | 4.569 | 1.344 |
| 4194304 | 1.291 | 4.559 | 1.327 |
Signed-off-by: Nir Soffer <nirsof@gmail.com>
Message-ID: <20250517201154.88456-3-nirsof@gmail.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
---
nbd/nbd-internal.h | 5 +++++
nbd/client-connection.c | 3 +++
nbd/common.c | 25 +++++++++++++++++++++++++
nbd/server.c | 2 ++
4 files changed, 35 insertions(+)
diff --git a/nbd/nbd-internal.h b/nbd/nbd-internal.h
index 715d92d6efc..6bafeef5ddc 100644
--- a/nbd/nbd-internal.h
+++ b/nbd/nbd-internal.h
@@ -74,4 +74,9 @@ static inline int nbd_write(QIOChannel *ioc, const void *buffer, size_t size,
int nbd_drop(QIOChannel *ioc, size_t size, Error **errp);
+/* nbd_set_socket_send_buffer
+ * Set the socket send buffer size for optimal performance.
+ */
+void nbd_set_socket_send_buffer(QIOChannelSocket *sioc);
+
#endif
diff --git a/nbd/client-connection.c b/nbd/client-connection.c
index b11e266807d..79ea97e4cc1 100644
--- a/nbd/client-connection.c
+++ b/nbd/client-connection.c
@@ -31,6 +31,8 @@
#include "qapi/clone-visitor.h"
#include "qemu/coroutine.h"
+#include "nbd/nbd-internal.h"
+
struct NBDClientConnection {
/* Initialization constants, never change */
SocketAddress *saddr; /* address to connect to */
@@ -140,6 +142,7 @@ static int nbd_connect(QIOChannelSocket *sioc, SocketAddress *addr,
return ret;
}
+ nbd_set_socket_send_buffer(sioc);
qio_channel_set_delay(QIO_CHANNEL(sioc), false);
if (!info) {
diff --git a/nbd/common.c b/nbd/common.c
index 589a748cfe6..9436e9d1d14 100644
--- a/nbd/common.c
+++ b/nbd/common.c
@@ -18,6 +18,9 @@
#include "qemu/osdep.h"
#include "trace.h"
+#include "io/channel-socket.h"
+#include "qapi/error.h"
+#include "qemu/units.h"
#include "nbd-internal.h"
/* Discard length bytes from channel. Return -errno on failure and 0 on
@@ -264,3 +267,25 @@ const char *nbd_mode_lookup(NBDMode mode)
return "<unknown>";
}
}
+
+/*
+ * Testing shows that 2m send buffer is optimal. Changing the receive buffer
+ * size has no effect on performance.
+ */
+#if defined(__APPLE__)
+#define UNIX_STREAM_SOCKET_SEND_BUFFER_SIZE (2 * MiB)
+#endif
+
+void nbd_set_socket_send_buffer(QIOChannelSocket *sioc)
+{
+#ifdef UNIX_STREAM_SOCKET_SEND_BUFFER_SIZE
+ if (sioc->localAddr.ss_family == AF_UNIX) {
+ size_t size = UNIX_STREAM_SOCKET_SEND_BUFFER_SIZE;
+ Error *errp = NULL;
+
+ if (qio_channel_socket_set_send_buffer(sioc, size, &errp) < 0) {
+ warn_report_err(errp);
+ }
+ }
+#endif /* UNIX_STREAM_SOCKET_SEND_BUFFER_SIZE */
+}
diff --git a/nbd/server.c b/nbd/server.c
index 2076fb2666b..d242be98115 100644
--- a/nbd/server.c
+++ b/nbd/server.c
@@ -3291,6 +3291,8 @@ void nbd_client_new(QIOChannelSocket *sioc,
client->close_fn = close_fn;
client->owner = owner;
+ nbd_set_socket_send_buffer(sioc);
+
co = qemu_coroutine_create(nbd_co_client_start, client);
qemu_coroutine_enter(co);
}
--
2.49.0
^ permalink raw reply related [flat|nested] 8+ messages in thread* [PULL 3/6] nbd: Set unix socket send buffer on Linux
2025-05-29 22:02 [PULL 0/6] NBD patches through 2025-05-29 Eric Blake
2025-05-29 22:02 ` [PULL 1/6] io: Add helper for setting socket send buffer size Eric Blake
2025-05-29 22:02 ` [PULL 2/6] nbd: Set unix socket send buffer on macOS Eric Blake
@ 2025-05-29 22:02 ` Eric Blake
2025-05-29 22:02 ` [PULL 4/6] iotests: Use disk_usage in more places Eric Blake
` (3 subsequent siblings)
6 siblings, 0 replies; 8+ messages in thread
From: Eric Blake @ 2025-05-29 22:02 UTC (permalink / raw)
To: qemu-devel
Cc: Nir Soffer, Daniel P. Berrangé, Vladimir Sementsov-Ogievskiy,
open list:Network Block Dev...
From: Nir Soffer <nirsof@gmail.com>
Like macOS we have similar issue on Linux. For TCP socket the send
buffer size is 2626560 bytes (~2.5 MiB) and we get good performance.
However for unix socket the default and maximum buffer size is 212992
bytes (208 KiB) and we see poor performance when using one NBD
connection, up to 4 times slower than macOS on the same machine.
Tracing shows that for every 2 MiB payload (qemu uses 2 MiB io size), we
do 1 recvmsg call with TCP socket, and 10 recvmsg calls with unix
socket.
Fixing this issue requires changing the maximum send buffer size (the
receive buffer size is ignored). This can be done using:
$ cat /etc/sysctl.d/net-mem-max.conf
net.core.wmem_max = 2097152
$ sudo sysctl -p /etc/sysctl.d/net-mem-max.conf
With this we can set the socket buffer size to 2 MiB. With the defaults
the value requested by qemu is clipped to the maximum size and has no
effect.
I tested on 2 machines:
- Fedora 42 VM on MacBook Pro M2 Max
- Dell PowerEdge R640 (Intel(R) Xeon(R) Gold 6230 CPU @ 2.10GHz)
On the older Dell machine we see very little improvement, up to 1.03
higher throughput. On the M2 machine we see up to 2.67 times higher
throughput. The following results are from the M2 machine.
Reading from qemu-nbd with qemu-img convert. In this test buffer size of
4m is optimal (2.28 times faster).
| buffer size | time | user | system |
|-------------|---------|---------|---------|
| default | 4.292 | 0.243 | 1.604 |
| 524288 | 2.167 | 0.058 | 1.288 |
| 1048576 | 2.041 | 0.060 | 1.238 |
| 2097152 | 1.884 | 0.060 | 1.191 |
| 4194304 | 1.881 | 0.054 | 1.196 |
Writing to qemu-nbd with qemu-img convert. In this test buffer size of
1m is optimal (2.67 times faster).
| buffer size | time | user | system |
|-------------|---------|---------|---------|
| default | 3.113 | 0.334 | 1.094 |
| 524288 | 1.173 | 0.179 | 0.654 |
| 1048576 | 1.164 | 0.164 | 0.670 |
| 2097152 | 1.227 | 0.197 | 0.663 |
| 4194304 | 1.227 | 0.198 | 0.666 |
Computing a blkhash with nbdcopy. In this test buffer size of 512k is
optimal (1.19 times faster).
| buffer size | time | user | system |
|-------------|---------|---------|---------|
| default | 2.140 | 4.483 | 2.681 |
| 524288 | 1.794 | 4.467 | 2.572 |
| 1048576 | 1.807 | 4.447 | 2.644 |
| 2097152 | 1.822 | 4.461 | 2.698 |
| 4194304 | 1.827 | 4.465 | 2.700 |
Computing a blkhash with blksum. In this test buffer size of 4m is
optimal (2.65 times faster).
| buffer size | time | user | system |
|-------------|---------|---------|---------|
| default | 3.582 | 4.595 | 2.392 |
| 524288 | 1.499 | 4.384 | 1.482 |
| 1048576 | 1.377 | 4.381 | 1.345 |
| 2097152 | 1.388 | 4.389 | 1.354 |
| 4194304 | 1.352 | 4.395 | 1.302 |
Signed-off-by: Nir Soffer <nirsof@gmail.com>
Message-ID: <20250517201154.88456-4-nirsof@gmail.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
---
nbd/common.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/nbd/common.c b/nbd/common.c
index 9436e9d1d14..2a133a66c39 100644
--- a/nbd/common.c
+++ b/nbd/common.c
@@ -271,8 +271,9 @@ const char *nbd_mode_lookup(NBDMode mode)
/*
* Testing shows that 2m send buffer is optimal. Changing the receive buffer
* size has no effect on performance.
+ * On Linux we need to increase net.core.wmem_max to make this effective.
*/
-#if defined(__APPLE__)
+#if defined(__APPLE__) || defined(__linux__)
#define UNIX_STREAM_SOCKET_SEND_BUFFER_SIZE (2 * MiB)
#endif
--
2.49.0
^ permalink raw reply related [flat|nested] 8+ messages in thread* [PULL 4/6] iotests: Use disk_usage in more places
2025-05-29 22:02 [PULL 0/6] NBD patches through 2025-05-29 Eric Blake
` (2 preceding siblings ...)
2025-05-29 22:02 ` [PULL 3/6] nbd: Set unix socket send buffer on Linux Eric Blake
@ 2025-05-29 22:02 ` Eric Blake
2025-05-29 22:02 ` [PULL 5/6] iotests: Improve mirror-sparse on ext4 and xfs Eric Blake
` (2 subsequent siblings)
6 siblings, 0 replies; 8+ messages in thread
From: Eric Blake @ 2025-05-29 22:02 UTC (permalink / raw)
To: qemu-devel
Cc: Fiona Ebner, Kevin Wolf, Hanna Reitz, open list:Block layer core
Commit be9bac07 added a utility disk_usage function, but there are
a couple of other tests that could also use it.
Signed-off-by: Eric Blake <eblake@redhat.com>
Message-ID: <20250523163041.2548675-6-eblake@redhat.com>
Reviewed-by: Fiona Ebner <f.ebner@proxmox.com>
Tested-by: Fiona Ebner <f.ebner@proxmox.com>
---
tests/qemu-iotests/125 | 2 +-
tests/qemu-iotests/308 | 4 ++--
2 files changed, 3 insertions(+), 3 deletions(-)
diff --git a/tests/qemu-iotests/125 b/tests/qemu-iotests/125
index 46279d6b382..708e7c5ba21 100755
--- a/tests/qemu-iotests/125
+++ b/tests/qemu-iotests/125
@@ -35,7 +35,7 @@ trap "_cleanup; exit \$status" 0 1 2 3 15
get_image_size_on_host()
{
- echo $(($(stat -c '%b * %B' "$TEST_IMG_FILE")))
+ disk_usage "$TEST_IMG_FILE"
}
# get standard environment and filters
diff --git a/tests/qemu-iotests/308 b/tests/qemu-iotests/308
index ea81dc496a0..437a9014da5 100755
--- a/tests/qemu-iotests/308
+++ b/tests/qemu-iotests/308
@@ -290,7 +290,7 @@ echo '--- Try growing non-growable export ---'
# Get the current size so we can write beyond the EOF
orig_len=$(get_proto_len "$EXT_MP" "$TEST_IMG")
-orig_disk_usage=$(stat -c '%b' "$TEST_IMG")
+orig_disk_usage=$(disk_usage "$TEST_IMG")
# Should fail (exports are non-growable by default)
# (Note that qemu-io can never write beyond the EOF, so we have to use
@@ -312,7 +312,7 @@ else
echo 'OK: Post-truncate image size is as expected'
fi
-new_disk_usage=$(stat -c '%b' "$TEST_IMG")
+new_disk_usage=$(disk_usage "$TEST_IMG")
if [ "$new_disk_usage" -gt "$orig_disk_usage" ]; then
echo 'OK: Disk usage grew with fallocate'
else
--
2.49.0
^ permalink raw reply related [flat|nested] 8+ messages in thread* [PULL 5/6] iotests: Improve mirror-sparse on ext4 and xfs
2025-05-29 22:02 [PULL 0/6] NBD patches through 2025-05-29 Eric Blake
` (3 preceding siblings ...)
2025-05-29 22:02 ` [PULL 4/6] iotests: Use disk_usage in more places Eric Blake
@ 2025-05-29 22:02 ` Eric Blake
2025-05-29 22:02 ` [PULL 6/6] iotests: Filter out ZFS in several tests Eric Blake
2025-05-31 23:05 ` [PULL 0/6] NBD patches through 2025-05-29 Stefan Hajnoczi
6 siblings, 0 replies; 8+ messages in thread
From: Eric Blake @ 2025-05-29 22:02 UTC (permalink / raw)
To: qemu-devel
Cc: Fiona Ebner, Markus Armbruster, Kevin Wolf, Hanna Reitz,
open list:Block layer core
Fiona reported that an ext4 filesystem on top of LVM can sometimes
report over-allocation to du (based on the heuristics the filesystem
is making while observing the contents being mirrored); even though
the contents and actual size matched, about 50% of the time the size
reported by disk_usage was too large by 4k, failing the test. In
auditing other iotests, this is a common problem we've had to deal
with.
Meanwhile, Markus reported that an xfs filesystem reports disk usage
at a default granularity of 1M (so the sparse file occupies 3M, since
it has just over 2M data).
Reported-by: Fiona Ebner <f.ebner@proxmox.com>
Reported-by: Markus Armbruster <armbru@redhat.com>
Fixes: c0ddcb2c ("tests: Add iotest mirror-sparse for recent patches")
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Fiona Ebner <f.ebner@proxmox.com>
Tested-by: Fiona Ebner <f.ebner@proxmox.com>
Message-ID: <20250523163041.2548675-7-eblake@redhat.com>
[eblake: Also fix xfs issue]
Signed-off-by: Eric Blake <eblake@redhat.com>
---
tests/qemu-iotests/tests/mirror-sparse | 8 +++++---
1 file changed, 5 insertions(+), 3 deletions(-)
diff --git a/tests/qemu-iotests/tests/mirror-sparse b/tests/qemu-iotests/tests/mirror-sparse
index 8c52a4e2448..11418c08713 100755
--- a/tests/qemu-iotests/tests/mirror-sparse
+++ b/tests/qemu-iotests/tests/mirror-sparse
@@ -96,13 +96,15 @@ _send_qemu_cmd $h1 '{"execute": "blockdev-del", "arguments":
{"node-name": "dst"}}' 'return' \
| _filter_block_job_offset | _filter_block_job_len
$QEMU_IMG compare -U -f $IMGFMT -F raw $TEST_IMG.base $TEST_IMG
+# Some filesystems can fudge allocations for various reasons; rather
+# than expecting precise 2M and 20M images, it is better to allow for slop.
result=$(disk_usage $TEST_IMG)
-if test $result -lt $((3*1024*1024)); then
+if test $result -lt $((4*1024*1024)); then
actual=sparse
-elif test $result = $((20*1024*1024)); then
+elif test $result -gt $((19*1024*1024)); then
actual=full
else
- actual=unknown
+ actual="unexpected size ($result)"
fi
echo "Destination is $actual; expected $expected"
}
--
2.49.0
^ permalink raw reply related [flat|nested] 8+ messages in thread* [PULL 6/6] iotests: Filter out ZFS in several tests
2025-05-29 22:02 [PULL 0/6] NBD patches through 2025-05-29 Eric Blake
` (4 preceding siblings ...)
2025-05-29 22:02 ` [PULL 5/6] iotests: Improve mirror-sparse on ext4 and xfs Eric Blake
@ 2025-05-29 22:02 ` Eric Blake
2025-05-31 23:05 ` [PULL 0/6] NBD patches through 2025-05-29 Stefan Hajnoczi
6 siblings, 0 replies; 8+ messages in thread
From: Eric Blake @ 2025-05-29 22:02 UTC (permalink / raw)
To: qemu-devel
Cc: Fiona Ebner, Kevin Wolf, Hanna Reitz, open list:Block layer core
Fiona reported that ZFS makes sparse file testing awkward, since:
- it has asynchronous allocation (not even 'fsync $file' makes du see
the desired size; it takes the slower 'fsync -f $file' which is not
appropriate for the tests)
- for tests of fully allocated files, ZFS with compression enabled
still reports smaller disk usage
Add a new _require_disk_usage that quickly probes whether an attempt
to create a sparse 5M file shows as less than 1M usage, while the same
file with -o preallocation=full shows as more than 4M usage without
sync, which should filter out ZFS behavior. Then use it in various
affected tests.
This does not add the new filter on all tests that Fiona is seeing ZFS
failures on, but only those where I could quickly spot that there is
at least one place where the test depends on the output of 'du -b' or
'stat -c %b'.
Reported-by: Fiona Ebner <f.ebner@proxmox.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
Message-ID: <20250523163041.2548675-8-eblake@redhat.com>
Reviewed-by: Fiona Ebner <f.ebner@proxmox.com>
Tested-by: Fiona Ebner <f.ebner@proxmox.com>
---
tests/qemu-iotests/common.rc | 30 +++++++++++++++++++++
tests/qemu-iotests/106 | 1 +
tests/qemu-iotests/175 | 1 +
tests/qemu-iotests/221 | 1 +
tests/qemu-iotests/253 | 1 +
tests/qemu-iotests/308 | 1 +
tests/qemu-iotests/tests/mirror-sparse | 1 +
tests/qemu-iotests/tests/write-zeroes-unmap | 1 +
8 files changed, 37 insertions(+)
diff --git a/tests/qemu-iotests/common.rc b/tests/qemu-iotests/common.rc
index 237f746af88..e977cb4eb61 100644
--- a/tests/qemu-iotests/common.rc
+++ b/tests/qemu-iotests/common.rc
@@ -996,6 +996,36 @@ _require_large_file()
rm "$FILENAME"
}
+# Check whether disk_usage can be reliably used.
+_require_disk_usage()
+{
+ local unusable=false
+ # ZFS triggers known failures on this front; it does not immediately
+ # allocate files, and then aggressively compresses writes even when full
+ # allocation was requested.
+ if [ -z "$TEST_IMG_FILE" ]; then
+ FILENAME="$TEST_IMG"
+ else
+ FILENAME="$TEST_IMG_FILE"
+ fi
+ if [ -e "FILENAME" ]; then
+ echo "unwilling to overwrite existing file"
+ exit 1
+ fi
+ $QEMU_IMG create -f raw "$FILENAME" 5M > /dev/null
+ if [ $(disk_usage "$FILENAME") -gt $((1024*1024)) ]; then
+ unusable=true
+ fi
+ $QEMU_IMG create -f raw -o preallocation=full "$FILENAME" 5M > /dev/null
+ if [ $(disk_usage "$FILENAME") -lt $((4*1024*1024)) ]; then
+ unusable=true
+ fi
+ rm -f "$FILENAME"
+ if $unusable; then
+ _notrun "file system on $TEST_DIR does not handle sparse files nicely"
+ fi
+}
+
# Check that a set of devices is available in the QEMU binary
#
_require_devices()
diff --git a/tests/qemu-iotests/106 b/tests/qemu-iotests/106
index ae0fc466910..55548439aad 100755
--- a/tests/qemu-iotests/106
+++ b/tests/qemu-iotests/106
@@ -40,6 +40,7 @@ trap "_cleanup; exit \$status" 0 1 2 3 15
_supported_fmt raw
_supported_proto file fuse
_supported_os Linux
+_require_disk_usage
# in kB
CREATION_SIZE=128
diff --git a/tests/qemu-iotests/175 b/tests/qemu-iotests/175
index f74f053b719..bbbf550a5af 100755
--- a/tests/qemu-iotests/175
+++ b/tests/qemu-iotests/175
@@ -77,6 +77,7 @@ _supported_os Linux
_default_cache_mode none
_supported_cache_modes none directsync
+_require_disk_usage
size=$((1 * 1024 * 1024))
diff --git a/tests/qemu-iotests/221 b/tests/qemu-iotests/221
index c463fd4b113..eba00b80adb 100755
--- a/tests/qemu-iotests/221
+++ b/tests/qemu-iotests/221
@@ -41,6 +41,7 @@ _supported_os Linux
_default_cache_mode writeback
_supported_cache_modes writeback writethrough unsafe
+_require_disk_usage
echo
echo "=== Check mapping of unaligned raw image ==="
diff --git a/tests/qemu-iotests/253 b/tests/qemu-iotests/253
index 35039d20a89..6da85e6a113 100755
--- a/tests/qemu-iotests/253
+++ b/tests/qemu-iotests/253
@@ -41,6 +41,7 @@ _supported_os Linux
_default_cache_mode none
_supported_cache_modes none directsync
+_require_disk_usage
echo
echo "=== Check mapping of unaligned raw image ==="
diff --git a/tests/qemu-iotests/308 b/tests/qemu-iotests/308
index 437a9014da5..6eced3aefb9 100755
--- a/tests/qemu-iotests/308
+++ b/tests/qemu-iotests/308
@@ -51,6 +51,7 @@ _unsupported_fmt vpc
_supported_proto file # We create the FUSE export manually
_supported_os Linux # We need /dev/urandom
+_require_disk_usage
# $1: Export ID
# $2: Options (beyond the node-name and ID)
diff --git a/tests/qemu-iotests/tests/mirror-sparse b/tests/qemu-iotests/tests/mirror-sparse
index 11418c08713..cfcaa600ab4 100755
--- a/tests/qemu-iotests/tests/mirror-sparse
+++ b/tests/qemu-iotests/tests/mirror-sparse
@@ -40,6 +40,7 @@ cd ..
_supported_fmt qcow2 raw # Format of the source. dst is always raw file
_supported_proto file
_supported_os Linux
+_require_disk_usage
echo
echo "=== Initial image setup ==="
diff --git a/tests/qemu-iotests/tests/write-zeroes-unmap b/tests/qemu-iotests/tests/write-zeroes-unmap
index 7cfeeaf8391..f90fb8e8d27 100755
--- a/tests/qemu-iotests/tests/write-zeroes-unmap
+++ b/tests/qemu-iotests/tests/write-zeroes-unmap
@@ -32,6 +32,7 @@ cd ..
_supported_fmt raw
_supported_proto file
_supported_os Linux
+_require_disk_usage
create_test_image() {
_make_test_img -f $IMGFMT 1m
--
2.49.0
^ permalink raw reply related [flat|nested] 8+ messages in thread* Re: [PULL 0/6] NBD patches through 2025-05-29
2025-05-29 22:02 [PULL 0/6] NBD patches through 2025-05-29 Eric Blake
` (5 preceding siblings ...)
2025-05-29 22:02 ` [PULL 6/6] iotests: Filter out ZFS in several tests Eric Blake
@ 2025-05-31 23:05 ` Stefan Hajnoczi
6 siblings, 0 replies; 8+ messages in thread
From: Stefan Hajnoczi @ 2025-05-31 23:05 UTC (permalink / raw)
To: Eric Blake; +Cc: qemu-devel
[-- Attachment #1: Type: text/plain, Size: 116 bytes --]
Applied, thanks.
Please update the changelog at https://wiki.qemu.org/ChangeLog/10.1 for any user-visible changes.
[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]
^ permalink raw reply [flat|nested] 8+ messages in thread