qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: "Andreas Färber" <afaerber@suse.de>
To: Stefan Hajnoczi <stefanha@redhat.com>, qemu-devel@nongnu.org
Cc: Paolo Bonzini <pbonzini@redhat.com>,
	Markus Armbruster <armbru@redhat.com>
Subject: Re: [Qemu-devel] [PATCH v3 6/7] qdev-monitor-test: add device_add leak test cases
Date: Wed, 06 Nov 2013 16:38:32 +0100	[thread overview]
Message-ID: <527A6278.9030308@suse.de> (raw)
In-Reply-To: <1383141276-19230-7-git-send-email-stefanha@redhat.com>

Am 30.10.2013 14:54, schrieb Stefan Hajnoczi:
> Ensure that the device_add error code path deletes device objects.
> Failure to do so not only leaks the objects but can also keep other
> objects (like drive or netdev) alive due to qdev properties holding
> references.
> 
> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
> ---
>  tests/Makefile            |  2 ++
>  tests/qdev-monitor-test.c | 81 +++++++++++++++++++++++++++++++++++++++++++++++
>  2 files changed, 83 insertions(+)
>  create mode 100644 tests/qdev-monitor-test.c
> 
> diff --git a/tests/Makefile b/tests/Makefile
> index 7863123..2771f92 100644
> --- a/tests/Makefile
> +++ b/tests/Makefile
> @@ -68,6 +68,7 @@ check-qtest-i386-y += tests/rtc-test$(EXESUF)
>  check-qtest-i386-y += tests/i440fx-test$(EXESUF)
>  check-qtest-i386-y += tests/fw_cfg-test$(EXESUF)
>  check-qtest-i386-y += tests/blockdev-test$(EXESUF)
> +check-qtest-i386-y += tests/qdev-monitor-test$(EXESUF)
>  check-qtest-x86_64-y = $(check-qtest-i386-y)
>  gcov-files-i386-y += i386-softmmu/hw/mc146818rtc.c

Same comment here about adding to the list.

>  gcov-files-x86_64-y = $(subst i386-softmmu/,x86_64-softmmu/,$(gcov-files-i386-y))
> @@ -176,6 +177,7 @@ tests/tmp105-test$(EXESUF): tests/tmp105-test.o $(libqos-omap-obj-y)
>  tests/i440fx-test$(EXESUF): tests/i440fx-test.o $(libqos-pc-obj-y)
>  tests/fw_cfg-test$(EXESUF): tests/fw_cfg-test.o $(libqos-pc-obj-y)
>  tests/blockdev-test$(EXESUF): tests/blockdev-test.o $(libqos-pc-obj-y)
> +tests/qdev-monitor-test$(EXESUF): tests/qdev-monitor-test.o $(libqos-pc-obj-y)
>  tests/qemu-iotests/socket_scm_helper$(EXESUF): tests/qemu-iotests/socket_scm_helper.o
>  
>  # QTest rules
> diff --git a/tests/qdev-monitor-test.c b/tests/qdev-monitor-test.c
> new file mode 100644
> index 0000000..d3d6d39
> --- /dev/null
> +++ b/tests/qdev-monitor-test.c
> @@ -0,0 +1,81 @@
> +/*
> + * qdev-monitor.c test cases
> + *
> + * Copyright (C) 2013 Red Hat Inc.
> + *
> + * Authors:
> + *  Stefan Hajnoczi <stefanha@redhat.com>
> + *
> + * This work is licensed under the terms of the GNU LGPL, version 2.1 or later.
> + * See the COPYING.LIB file in the top-level directory.
> + */
> +
> +#include <string.h>
> +#include <glib.h>
> +#include "libqtest.h"
> +#include "qapi/qmp/qjson.h"
> +
> +static void test_device_add(void)
> +{
> +    QDict *response;
> +    QDict *error;
> +
> +    qtest_start("-drive if=none,id=drive0");
> +
> +    /* Make device_add fail.  If this leaks the virtio-blk-pci device then a
> +     * reference to drive0 will also be held (via qdev properties).
> +     */
> +    response = qmp("{\"execute\": \"device_add\","
> +                   " \"arguments\": {"
> +                   "   \"driver\": \"virtio-blk-pci\","
> +                   "   \"drive\": \"drive0\""
> +                   "}}");
> +    g_assert(response);
> +    error = qdict_get_qdict(response, "error");
> +    g_assert(!strcmp(qdict_get_try_str(error, "class") ?: "",
> +                     "GenericError"));
> +    g_assert(!strcmp(qdict_get_try_str(error, "desc") ?: "",
> +                     "Device initialization failed."));
> +    QDECREF(response);
> +
> +    /* Delete the drive */
> +    response = qmp("{\"execute\": \"human-monitor-command\","
> +                   " \"arguments\": {"
> +                   "   \"command-line\": \"drive_del drive0\""
> +                   "}}");
> +    g_assert(response);
> +    g_assert(!strcmp(qdict_get_try_str(response, "return") ?: "(null)", ""));
> +    QDECREF(response);
> +
> +    /* Try to re-add the drive.  This fails with duplicate IDs if a leaked
> +     * virtio-blk-pci exists that holds a reference to the old drive0.
> +     */
> +    response = qmp("{\"execute\": \"human-monitor-command\","
> +                   " \"arguments\": {"
> +                   "   \"command-line\": \"drive_add pci-addr=auto if=none,id=drive0\""
> +                   "}}");
> +    g_assert(response);
> +    g_assert(!strcmp(qdict_get_try_str(response, "return") ?: "",
> +                     "OK\r\n"));
> +    QDECREF(response);
> +
> +    qtest_end();
> +}
> +
> +int main(int argc, char **argv)
> +{
> +    const char *arch = qtest_get_arch();
> +
> +    /* Check architecture */
> +    if (strcmp(arch, "i386") && strcmp(arch, "x86_64")) {
> +        g_test_message("Skipping test for non-x86\n");
> +        return 0;
> +    }
> +
> +    /* Run the tests */
> +    g_test_init(&argc, &argv, NULL);
> +
> +    qtest_add_func("/qdev/device_add", test_device_add);

What about naming these tests /qmp/...?

Again, test itself looks great on a brief look.

Andreas

> +
> +    return g_test_run();
> +}
> 


-- 
SUSE LINUX Products GmbH, Maxfeldstr. 5, 90409 Nürnberg, Germany
GF: Jeff Hawn, Jennifer Guild, Felix Imendörffer; HRB 16746 AG Nürnberg

  parent reply	other threads:[~2013-11-06 17:04 UTC|newest]

Thread overview: 23+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-10-30 13:54 [Qemu-devel] [PATCH v3 0/7] qdev and blockdev refcount leak fixes Stefan Hajnoczi
2013-10-30 13:54 ` [Qemu-devel] [PATCH v3 1/7] blockdev: fix drive_init() opts and bs_opts leaks Stefan Hajnoczi
2013-11-05 11:48   ` Eric Blake
2013-10-30 13:54 ` [Qemu-devel] [PATCH v3 2/7] qdev: unref qdev when device_add fails Stefan Hajnoczi
2013-11-05 11:50   ` Eric Blake
2013-11-05 15:28     ` Stefan Hajnoczi
2013-10-30 13:54 ` [Qemu-devel] [PATCH v3 3/7] libqtest: rename qmp() to qmp_discard_response() Stefan Hajnoczi
2013-11-05 11:52   ` Eric Blake
2013-11-06 15:24   ` Andreas Färber
2013-10-30 13:54 ` [Qemu-devel] [PATCH v3 4/7] libqtest: add qmp(fmt, ...) -> QDict* function Stefan Hajnoczi
2013-11-05 11:56   ` Eric Blake
2013-11-06 15:32   ` Andreas Färber
2013-10-30 13:54 ` [Qemu-devel] [PATCH v3 5/7] blockdev-test: add test case for drive_add duplicate IDs Stefan Hajnoczi
2013-11-05 11:57   ` Eric Blake
2013-11-06 15:36   ` Andreas Färber
2013-10-30 13:54 ` [Qemu-devel] [PATCH v3 6/7] qdev-monitor-test: add device_add leak test cases Stefan Hajnoczi
2013-11-05 11:58   ` Eric Blake
2013-11-06 15:38   ` Andreas Färber [this message]
2013-10-30 13:54 ` [Qemu-devel] [PATCH v3 7/7] qdev: drop misleading qdev_free() function Stefan Hajnoczi
2013-11-05 12:06   ` Eric Blake
2013-10-30 15:44 ` [Qemu-devel] [PATCH v3 0/7] qdev and blockdev refcount leak fixes Andreas Färber
2013-10-31  9:22   ` Stefan Hajnoczi
2013-11-06 10:30 ` 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=527A6278.9030308@suse.de \
    --to=afaerber@suse.de \
    --cc=armbru@redhat.com \
    --cc=pbonzini@redhat.com \
    --cc=qemu-devel@nongnu.org \
    --cc=stefanha@redhat.com \
    /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).