From: "Philippe Mathieu-Daudé" <philmd@linaro.org>
To: Albert Esteve <aesteve@redhat.com>, qemu-devel@nongnu.org
Cc: Fam Zheng <fam@euphon.net>, "Michael S. Tsirkin" <mst@redhat.com>,
cohuck@redhat.com, kraxel@redhat.com, marcandre.lureau@gmail.com
Subject: Re: [PATCH v6 2/3] virtio-dmabuf: introduce virtio-dmabuf
Date: Wed, 6 Sep 2023 16:21:31 +0200 [thread overview]
Message-ID: <c53f4cae-7014-066a-de3d-09f47b27dbc6@linaro.org> (raw)
In-Reply-To: <20230906111549.357178-3-aesteve@redhat.com>
On 6/9/23 13:15, Albert Esteve wrote:
> This API manages objects (in this iteration,
> dmabuf fds) that can be shared along different
> virtio devices, associated to a UUID.
>
> The API allows the different devices to add,
> remove and/or retrieve the objects by simply
> invoking the public functions that reside in the
> virtio-dmabuf file.
>
> For vhost backends, the API stores the pointer
> to the backend holding the object.
>
> Suggested-by: Gerd Hoffmann <kraxel@redhat.com>
> Signed-off-by: Albert Esteve <aesteve@redhat.com>
> ---
> MAINTAINERS | 7 ++
> hw/display/meson.build | 1 +
> hw/display/virtio-dmabuf.c | 134 +++++++++++++++++++++++++++++
> include/hw/virtio/virtio-dmabuf.h | 103 ++++++++++++++++++++++
> tests/unit/meson.build | 1 +
> tests/unit/test-virtio-dmabuf.c | 137 ++++++++++++++++++++++++++++++
> 6 files changed, 383 insertions(+)
> create mode 100644 hw/display/virtio-dmabuf.c
> create mode 100644 include/hw/virtio/virtio-dmabuf.h
> create mode 100644 tests/unit/test-virtio-dmabuf.c
>
> diff --git a/MAINTAINERS b/MAINTAINERS
> index 3b29568ed4..fb0f7b823f 100644
> --- a/MAINTAINERS
> +++ b/MAINTAINERS
> @@ -2150,6 +2150,13 @@ T: git https://gitlab.com/cohuck/qemu.git s390-next
> T: git https://github.com/borntraeger/qemu.git s390-next
> L: qemu-s390x@nongnu.org
>
> +virtio-dmabuf
> +M: Albert Esteve <aesteve@redhat.com>
> +S: Supported
> +F: hw/display/virtio-dmabuf.c
> +F: include/hw/virtio/virtio-dmabuf.h
> +F: tests/unit/test-virtio-dmabuf.c
> +
> virtiofs
> M: Stefan Hajnoczi <stefanha@redhat.com>
> S: Supported
> diff --git a/hw/display/meson.build b/hw/display/meson.build
> index 413ba4ab24..05619c6968 100644
> --- a/hw/display/meson.build
> +++ b/hw/display/meson.build
> @@ -37,6 +37,7 @@ system_ss.add(when: 'CONFIG_MACFB', if_true: files('macfb.c'))
> system_ss.add(when: 'CONFIG_NEXTCUBE', if_true: files('next-fb.c'))
>
> system_ss.add(when: 'CONFIG_VGA', if_true: files('vga.c'))
> +system_ss.add(when: 'CONFIG_VIRTIO', if_true: files('virtio-dmabuf.c'))
>
> if (config_all_devices.has_key('CONFIG_VGA_CIRRUS') or
> config_all_devices.has_key('CONFIG_VGA_PCI') or
> diff --git a/hw/display/virtio-dmabuf.c b/hw/display/virtio-dmabuf.c
> new file mode 100644
> index 0000000000..268ffe81ec
> --- /dev/null
> +++ b/hw/display/virtio-dmabuf.c
> @@ -0,0 +1,134 @@
> +/*
> + * Virtio Shared dma-buf
> + *
> + * Copyright Red Hat, Inc. 2023
> + *
> + * Authors:
> + * Albert Esteve <aesteve@redhat.com>
> + *
> + * This work is licensed under the terms of the GNU GPL, version 2 or later.
> + * See the COPYING file in the top-level directory.
> + */
> +
> +#include "hw/virtio/virtio-dmabuf.h"
> +
> +
> +static GMutex lock;
> +static GHashTable *resource_uuids;
> +
> +/*
> + * uuid_equal_func: wrapper for UUID is_equal function to
> + * satisfy g_hash_table_new expected parameters signatures.
> + */
> +static int uuid_equal_func(const void *lhv, const void *rhv)
> +{
> + return qemu_uuid_is_equal(lhv, rhv);
> +}
> +
> +static bool virtio_add_resource(QemuUUID *uuid, VirtioSharedObject *value)
> +{
> + bool result;
> + g_mutex_lock(&lock);
> + if (resource_uuids == NULL) {
> + resource_uuids = g_hash_table_new_full(
> + qemu_uuid_hash, uuid_equal_func, NULL, g_free);
> + }
> + if (g_hash_table_lookup(resource_uuids, uuid) != NULL) {
> + g_mutex_unlock(&lock);
> + return false;
> + }
> + result = g_hash_table_insert(resource_uuids, uuid, value);
> + g_mutex_unlock(&lock);
> +
> + return result;
> +}
Alternatively same logic, but simpler / safer:
static bool virtio_add_resource(...)
{
bool result = false;
g_mutex_lock(&lock);
if (resource_uuids == NULL) {
resource_uuids = g_hash_table_new_full(qemu_uuid_hash,
uuid_equal_func,
NULL,
g_free);
}
if (g_hash_table_lookup(resource_uuids, uuid) == NULL) {
result = g_hash_table_insert(resource_uuids, uuid, value);
}
g_mutex_unlock(&lock);
return result;
}
> +static VirtioSharedObject *get_shared_object(const QemuUUID *uuid)
> +{
> + g_mutex_lock(&lock);
> + if (resource_uuids == NULL) {
> + g_mutex_unlock(&lock);
> + return NULL;
> + }
> + gpointer lookup_res = g_hash_table_lookup(resource_uuids, uuid);
> + g_mutex_unlock(&lock);
> + return (VirtioSharedObject*) lookup_res;
> +}
Similarly:
static VirtioSharedObject *get_shared_object(const QemuUUID *uuid)
{
gpointer lookup_res = NULL;
g_mutex_lock(&lock);
if (resource_uuids != NULL) {
lookup_res = g_hash_table_lookup(resource_uuids, uuid);
}
g_mutex_unlock(&lock);
return (VirtioSharedObject *)lookup_res;
}
> +int main(int argc, char **argv)
> +{
> + g_test_init(&argc, &argv, NULL);
> + g_test_add_func("/virtio-dmabuf/add_rm_res", test_add_remove_resources);
> + g_test_add_func("/virtio-dmabuf/add_rm_dev", test_add_remove_dev);
Mis-indent.
> + g_test_add_func("/virtio-dmabuf/rm_invalid_res",
> + test_remove_invalid_resource);
> + g_test_add_func("/virtio-dmabuf/add_invalid_res",
> + test_add_invalid_resource);
> + g_test_add_func("/virtio-dmabuf/free_res", test_free_resources);
> +
> + return g_test_run();
> +}
Thanks for updating, LGTM!
next prev parent reply other threads:[~2023-09-06 14:22 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-09-06 11:15 [PATCH v6 0/3] Virtio shared dma-buf Albert Esteve
2023-09-06 11:15 ` [PATCH v6 1/3] uuid: add a hash function Albert Esteve
2023-09-06 11:15 ` [PATCH v6 2/3] virtio-dmabuf: introduce virtio-dmabuf Albert Esteve
2023-09-06 12:03 ` Albert Esteve
2023-09-06 14:21 ` Philippe Mathieu-Daudé [this message]
2023-09-06 15:34 ` Albert Esteve
2023-09-06 11:15 ` [PATCH v6 3/3] vhost-user: add shared_object msg Albert Esteve
2023-09-06 14:27 ` Philippe Mathieu-Daudé
2023-09-06 14:33 ` Albert Esteve
2023-09-06 14:43 ` Philippe Mathieu-Daudé
2023-09-06 16:01 ` Albert Esteve
2023-09-06 16:18 ` Albert Esteve
2023-09-07 6:19 ` Philippe Mathieu-Daudé
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=c53f4cae-7014-066a-de3d-09f47b27dbc6@linaro.org \
--to=philmd@linaro.org \
--cc=aesteve@redhat.com \
--cc=cohuck@redhat.com \
--cc=fam@euphon.net \
--cc=kraxel@redhat.com \
--cc=marcandre.lureau@gmail.com \
--cc=mst@redhat.com \
--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).