qemu-kvm/kvm-iotests-stream-error-on-reset-New-test.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

197 lines
7.4 KiB
Diff

From 300f912d4a5afe4ecca9c68a71429fbc9966ec34 Mon Sep 17 00:00:00 2001
From: Hanna Reitz <hreitz@redhat.com>
Date: Tue, 11 Jan 2022 15:36:13 +0000
Subject: [PATCH 11/12] iotests/stream-error-on-reset: New test
RH-Author: Hanna Reitz <hreitz@redhat.com>
RH-MergeRequest: 71: block-backend: prevent dangling BDS pointers across aio_poll()
RH-Commit: [2/2] 3167f31b91eb433f338564201f4ef336e39f7f7d (hreitz/qemu-kvm-c-9-s)
RH-Bugzilla: 2040123
RH-Acked-by: Kevin Wolf <kwolf@redhat.com>
RH-Acked-by: Stefan Hajnoczi <stefanha@redhat.com>
RH-Acked-by: Emanuele Giuseppe Esposito <eesposit@redhat.com>
Test the following scenario:
- Simple stream block in two-layer backing chain (base and top)
- The job is drained via blk_drain(), then an error occurs while the job
settles the ongoing request
- And so the job completes while in blk_drain()
This was reported as a segfault, but is fixed by "block-backend: prevent
dangling BDS pointers across aio_poll()".
Buglink: https://bugzilla.redhat.com/show_bug.cgi?id=2036178
Signed-off-by: Hanna Reitz <hreitz@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20220111153613.25453-3-stefanha@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
(cherry picked from commit 2ca1d5d6b91f8a52a5c651f660b2f58c94bf97ba)
Signed-off-by: Hanna Reitz <hreitz@redhat.com>
---
.../qemu-iotests/tests/stream-error-on-reset | 140 ++++++++++++++++++
.../tests/stream-error-on-reset.out | 5 +
2 files changed, 145 insertions(+)
create mode 100755 tests/qemu-iotests/tests/stream-error-on-reset
create mode 100644 tests/qemu-iotests/tests/stream-error-on-reset.out
diff --git a/tests/qemu-iotests/tests/stream-error-on-reset b/tests/qemu-iotests/tests/stream-error-on-reset
new file mode 100755
index 0000000000..7eaedb24d7
--- /dev/null
+++ b/tests/qemu-iotests/tests/stream-error-on-reset
@@ -0,0 +1,140 @@
+#!/usr/bin/env python3
+# group: rw quick
+#
+# Test what happens when a stream job completes in a blk_drain().
+#
+# Copyright (C) 2022 Red Hat, Inc.
+#
+# 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/>.
+#
+
+import os
+import iotests
+from iotests import imgfmt, qemu_img_create, qemu_io_silent, QMPTestCase
+
+
+image_size = 1 * 1024 * 1024
+data_size = 64 * 1024
+base = os.path.join(iotests.test_dir, 'base.img')
+top = os.path.join(iotests.test_dir, 'top.img')
+
+
+# We want to test completing a stream job in a blk_drain().
+#
+# The blk_drain() we are going to use is a virtio-scsi device resetting,
+# which we can trigger by resetting the system.
+#
+# In order to have the block job complete on drain, we (1) throttle its
+# base image so we can start the drain after it has begun, but before it
+# completes, and (2) make it encounter an I/O error on the ensuing write.
+# (If it completes regularly, the completion happens after the drain for
+# some reason.)
+
+class TestStreamErrorOnReset(QMPTestCase):
+ def setUp(self) -> None:
+ """
+ Create two images:
+ - base image {base} with {data_size} bytes allocated
+ - top image {top} without any data allocated
+
+ And the following VM configuration:
+ - base image throttled to {data_size}
+ - top image with a blkdebug configuration so the first write access
+ to it will result in an error
+ - top image is attached to a virtio-scsi device
+ """
+ assert qemu_img_create('-f', imgfmt, base, str(image_size)) == 0
+ assert qemu_io_silent('-c', f'write 0 {data_size}', base) == 0
+ assert qemu_img_create('-f', imgfmt, top, str(image_size)) == 0
+
+ self.vm = iotests.VM()
+ self.vm.add_args('-accel', 'tcg') # Make throttling work properly
+ self.vm.add_object(self.vm.qmp_to_opts({
+ 'qom-type': 'throttle-group',
+ 'id': 'thrgr',
+ 'x-bps-total': str(data_size)
+ }))
+ self.vm.add_blockdev(self.vm.qmp_to_opts({
+ 'driver': imgfmt,
+ 'node-name': 'base',
+ 'file': {
+ 'driver': 'throttle',
+ 'throttle-group': 'thrgr',
+ 'file': {
+ 'driver': 'file',
+ 'filename': base
+ }
+ }
+ }))
+ self.vm.add_blockdev(self.vm.qmp_to_opts({
+ 'driver': imgfmt,
+ 'node-name': 'top',
+ 'file': {
+ 'driver': 'blkdebug',
+ 'node-name': 'top-blkdebug',
+ 'inject-error': [{
+ 'event': 'pwritev',
+ 'immediately': 'true',
+ 'once': 'true'
+ }],
+ 'image': {
+ 'driver': 'file',
+ 'filename': top
+ }
+ },
+ 'backing': 'base'
+ }))
+ self.vm.add_device(self.vm.qmp_to_opts({
+ 'driver': 'virtio-scsi',
+ 'id': 'vscsi'
+ }))
+ self.vm.add_device(self.vm.qmp_to_opts({
+ 'driver': 'scsi-hd',
+ 'bus': 'vscsi.0',
+ 'drive': 'top'
+ }))
+ self.vm.launch()
+
+ def tearDown(self) -> None:
+ self.vm.shutdown()
+ os.remove(top)
+ os.remove(base)
+
+ def test_stream_error_on_reset(self) -> None:
+ # Launch a stream job, which will take at least a second to
+ # complete, because the base image is throttled (so we can
+ # get in between it having started and it having completed)
+ res = self.vm.qmp('block-stream', job_id='stream', device='top')
+ self.assert_qmp(res, 'return', {})
+
+ while True:
+ ev = self.vm.event_wait('JOB_STATUS_CHANGE')
+ if ev['data']['status'] == 'running':
+ # Once the stream job is running, reset the system, which
+ # forces the virtio-scsi device to be reset, thus draining
+ # the stream job, and making it complete. Completing
+ # inside of that drain should not result in a segfault.
+ res = self.vm.qmp('system_reset')
+ self.assert_qmp(res, 'return', {})
+ elif ev['data']['status'] == 'null':
+ # The test is done once the job is gone
+ break
+
+
+if __name__ == '__main__':
+ # Passes with any format with backing file support, but qed and
+ # qcow1 do not seem to exercise the used-to-be problematic code
+ # path, so there is no point in having them in this list
+ iotests.main(supported_fmts=['qcow2', 'vmdk'],
+ supported_protocols=['file'])
diff --git a/tests/qemu-iotests/tests/stream-error-on-reset.out b/tests/qemu-iotests/tests/stream-error-on-reset.out
new file mode 100644
index 0000000000..ae1213e6f8
--- /dev/null
+++ b/tests/qemu-iotests/tests/stream-error-on-reset.out
@@ -0,0 +1,5 @@
+.
+----------------------------------------------------------------------
+Ran 1 tests
+
+OK
--
2.27.0