qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Stefan Hajnoczi <stefanha@redhat.com>
To: qemu-devel@nongnu.org
Cc: Peter Maydell <peter.maydell@linaro.org>,
	Stefan Hajnoczi <stefanha@redhat.com>
Subject: [Qemu-devel] [PULL 14/23] blockdev: add x-blockdev-set-iothread testing command
Date: Mon, 18 Dec 2017 14:35:21 +0000	[thread overview]
Message-ID: <20171218143530.12082-15-stefanha@redhat.com> (raw)
In-Reply-To: <20171218143530.12082-1-stefanha@redhat.com>

Currently there is no easy way for iotests to ensure that a BDS is bound
to a particular IOThread.  Normally the virtio-blk device calls
blk_set_aio_context() when dataplane is enabled during guest driver
initialization.  This never happens in iotests since -machine
accel=qtest means there is no guest activity (including device driver
initialization).

This patch adds a QMP command to explicitly assign IOThreads in test
cases.  See qapi/block-core.json for a description of the command.

Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-id: 20171206144550.22295-9-stefanha@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
 qapi/block-core.json | 36 ++++++++++++++++++++++++++++++++++++
 blockdev.c           | 41 +++++++++++++++++++++++++++++++++++++++++
 2 files changed, 77 insertions(+)

diff --git a/qapi/block-core.json b/qapi/block-core.json
index dd763dcf87..741d6c4367 100644
--- a/qapi/block-core.json
+++ b/qapi/block-core.json
@@ -3949,3 +3949,39 @@
   'data' : { 'parent': 'str',
              '*child': 'str',
              '*node': 'str' } }
+
+##
+# @x-blockdev-set-iothread:
+#
+# Move @node and its children into the @iothread.  If @iothread is null then
+# move @node and its children into the main loop.
+#
+# The node must not be attached to a BlockBackend.
+#
+# @node-name: the name of the block driver node
+#
+# @iothread: the name of the IOThread object or null for the main loop
+#
+# Note: this command is experimental and intended for test cases that need
+# control over IOThreads only.
+#
+# Since: 2.12
+#
+# Example:
+#
+# 1. Move a node into an IOThread
+# -> { "execute": "x-blockdev-set-iothread",
+#      "arguments": { "node-name": "disk1",
+#                     "iothread": "iothread0" } }
+# <- { "return": {} }
+#
+# 2. Move a node into the main loop
+# -> { "execute": "x-blockdev-set-iothread",
+#      "arguments": { "node-name": "disk1",
+#                     "iothread": null } }
+# <- { "return": {} }
+#
+##
+{ 'command': 'x-blockdev-set-iothread',
+  'data' : { 'node-name': 'str',
+             'iothread': 'StrOrNull' } }
diff --git a/blockdev.c b/blockdev.c
index e865ae4873..f75c01f664 100644
--- a/blockdev.c
+++ b/blockdev.c
@@ -45,6 +45,7 @@
 #include "qapi/qmp/qerror.h"
 #include "qapi/qobject-output-visitor.h"
 #include "sysemu/sysemu.h"
+#include "sysemu/iothread.h"
 #include "block/block_int.h"
 #include "qmp-commands.h"
 #include "block/trace.h"
@@ -4129,6 +4130,46 @@ BlockJobInfoList *qmp_query_block_jobs(Error **errp)
     return head;
 }
 
+void qmp_x_blockdev_set_iothread(const char *node_name, StrOrNull *iothread,
+                                 Error **errp)
+{
+    AioContext *old_context;
+    AioContext *new_context;
+    BlockDriverState *bs;
+
+    bs = bdrv_find_node(node_name);
+    if (!bs) {
+        error_setg(errp, "Cannot find node %s", node_name);
+        return;
+    }
+
+    /* If we want to allow more extreme test scenarios this guard could be
+     * removed.  For now it protects against accidents. */
+    if (bdrv_has_blk(bs)) {
+        error_setg(errp, "Node %s is in use", node_name);
+        return;
+    }
+
+    if (iothread->type == QTYPE_QSTRING) {
+        IOThread *obj = iothread_by_id(iothread->u.s);
+        if (!obj) {
+            error_setg(errp, "Cannot find iothread %s", iothread->u.s);
+            return;
+        }
+
+        new_context = iothread_get_aio_context(obj);
+    } else {
+        new_context = qemu_get_aio_context();
+    }
+
+    old_context = bdrv_get_aio_context(bs);
+    aio_context_acquire(old_context);
+
+    bdrv_set_aio_context(bs, new_context);
+
+    aio_context_release(old_context);
+}
+
 QemuOptsList qemu_common_drive_opts = {
     .name = "drive",
     .head = QTAILQ_HEAD_INITIALIZER(qemu_common_drive_opts.head),
-- 
2.14.3

  parent reply	other threads:[~2017-12-18 14:36 UTC|newest]

Thread overview: 26+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-12-18 14:35 [Qemu-devel] [PULL 00/23] Block patches Stefan Hajnoczi
2017-12-18 14:35 ` [Qemu-devel] [PULL 01/23] coroutine: simplify co_aio_sleep_ns() prototype Stefan Hajnoczi
2017-12-18 14:35 ` [Qemu-devel] [PULL 02/23] hw/block/nvme: Convert to realize Stefan Hajnoczi
2017-12-18 14:35 ` [Qemu-devel] [PULL 03/23] hw/block: Fix the return type Stefan Hajnoczi
2017-12-18 14:35 ` [Qemu-devel] [PULL 04/23] hw/block: Use errp directly rather than local_err Stefan Hajnoczi
2017-12-18 14:35 ` [Qemu-devel] [PULL 05/23] dev-storage: Fix the unusual function name Stefan Hajnoczi
2017-12-18 14:35 ` [Qemu-devel] [PULL 06/23] qdev: drop unused #include "sysemu/iothread.h" Stefan Hajnoczi
2017-12-18 14:35 ` [Qemu-devel] [PULL 07/23] blockdev: hold AioContext for bdrv_unref() in external_snapshot_clean() Stefan Hajnoczi
2017-12-18 14:35 ` [Qemu-devel] [PULL 08/23] block: don't keep AioContext acquired after external_snapshot_prepare() Stefan Hajnoczi
2017-12-18 14:35 ` [Qemu-devel] [PULL 09/23] block: don't keep AioContext acquired after drive_backup_prepare() Stefan Hajnoczi
2017-12-18 14:35 ` [Qemu-devel] [PULL 10/23] block: don't keep AioContext acquired after blockdev_backup_prepare() Stefan Hajnoczi
2017-12-18 14:35 ` [Qemu-devel] [PULL 11/23] block: don't keep AioContext acquired after internal_snapshot_prepare() Stefan Hajnoczi
2017-12-18 14:35 ` [Qemu-devel] [PULL 12/23] block: drop unused BlockDirtyBitmapState->aio_context field Stefan Hajnoczi
2017-12-18 14:35 ` [Qemu-devel] [PULL 13/23] iothread: add iothread_by_id() API Stefan Hajnoczi
2017-12-18 14:35 ` Stefan Hajnoczi [this message]
2017-12-18 14:35 ` [Qemu-devel] [PULL 15/23] qemu-iotests: add 202 external snapshots IOThread test Stefan Hajnoczi
2017-12-18 14:35 ` [Qemu-devel] [PULL 16/23] virtio-blk: make queue size configurable Stefan Hajnoczi
2017-12-18 14:35 ` [Qemu-devel] [PULL 17/23] virtio-blk: reject configs with logical block size > physical block size Stefan Hajnoczi
2017-12-18 14:35 ` [Qemu-devel] [PULL 18/23] block: avoid recursive AioContext acquire in bdrv_inactivate_all() Stefan Hajnoczi
2017-12-18 14:35 ` [Qemu-devel] [PULL 19/23] docs: mark nested AioContext locking as a legacy API Stefan Hajnoczi
2017-12-18 14:35 ` [Qemu-devel] [PULL 20/23] blockdev: add x-blockdev-set-iothread force boolean Stefan Hajnoczi
2017-12-18 14:35 ` [Qemu-devel] [PULL 21/23] iotests: add VM.add_object() Stefan Hajnoczi
2017-12-18 14:35 ` [Qemu-devel] [PULL 22/23] iothread: fix iothread_stop() race condition Stefan Hajnoczi
2017-12-18 14:35 ` [Qemu-devel] [PULL 23/23] qemu-iotests: add 203 savevm with IOThreads test Stefan Hajnoczi
2017-12-19  0:15 ` [Qemu-devel] [PULL 00/23] Block patches Peter Maydell
2017-12-19  9:15   ` Stefan Hajnoczi

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20171218143530.12082-15-stefanha@redhat.com \
    --to=stefanha@redhat.com \
    --cc=peter.maydell@linaro.org \
    --cc=qemu-devel@nongnu.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).