From: Takashi Sakamoto <o-takashi@sakamocchi.jp>
To: linux1394-devel@lists.sourceforge.net
Cc: linux-kernel@vger.kernel.org
Subject: Re: [PATCH 3/9] firewire: cdev: add context analysis annotations for isochronous context
Date: Mon, 2 Nov 2026 19:40:42 +0900 [thread overview]
Message-ID: <20261102104042.GA1197999@sakamocchi.jp> (raw)
In-Reply-To: <20261102041913.1187814-4-o-takashi@sakamocchi.jp>
On Mon, Nov 02, 2026 at 01:19:07PM +0900, Takashi Sakamoto wrote:
> A client can operate on a single isochronous context at a time. A mutex
> is used to ensure exclusive access to the context.
>
> Use guard annotations to describe the synchronization requirements for
> accessing the context.
>
> Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
> ---
> drivers/firewire/core-cdev.c | 24 +++++++++++++++++++-----
> 1 file changed, 19 insertions(+), 5 deletions(-)
>
> diff --git a/drivers/firewire/core-cdev.c b/drivers/firewire/core-cdev.c
> index 512afa589e2e..418b72cb8835 100644
> --- a/drivers/firewire/core-cdev.c
> +++ b/drivers/firewire/core-cdev.c
> @@ -62,7 +62,7 @@ struct client {
> wait_queue_head_t tx_flush_wait;
> u64 bus_reset_closure;
>
> - struct fw_iso_context *iso_context;
> + struct fw_iso_context *iso_context __pt_guarded_by(iso_context_mutex);
> struct mutex iso_context_mutex;
> u64 iso_closure;
> struct fw_iso_buffer buffer;
> @@ -1129,8 +1129,11 @@ static int ioctl_create_iso_context(struct client *client, union ioctl_arg *arg)
> static int ioctl_set_iso_channels(struct client *client, union ioctl_arg *arg)
> {
> struct fw_cdev_set_iso_channels *a = &arg->set_iso_channels;
> - struct fw_iso_context *ctx = client->iso_context;
> + struct fw_iso_context *ctx;
>
> + guard(mutex)(&client->iso_context_mutex);
> +
> + ctx = client->iso_context;
> if (ctx == NULL || a->handle != 0)
> return -EINVAL;
>
> @@ -1149,12 +1152,15 @@ static int ioctl_queue_iso(struct client *client, union ioctl_arg *arg)
> {
> struct fw_cdev_queue_iso *a = &arg->queue_iso;
> struct fw_cdev_iso_packet __user *p, *end, *next;
> - struct fw_iso_context *ctx = client->iso_context;
> + struct fw_iso_context *ctx;
> unsigned long payload, buffer_end, transmit_header_bytes = 0;
> u32 control;
> int count;
> DEFINE_RAW_FLEX(struct fw_iso_packet, u, header, 64);
>
> + guard(mutex)(&client->iso_context_mutex);
Sashiko reports that get_user()/copy_from_user() under mutex locking can
cause A-B/B-A deadlock between the above mutex and mm->mmap_lock.
https://sashiko.dev/#/patchset/20261102041913.1187814-1-o-takashi%40sakamocchi.jp
Let me rework for the series.
> + ctx = client->iso_context;
> if (ctx == NULL || a->handle != 0)
> return -EINVAL;
>
> @@ -1252,6 +1258,8 @@ static int ioctl_start_iso(struct client *client, union ioctl_arg *arg)
> FW_CDEV_ISO_CONTEXT_MATCH_TAG3 != FW_ISO_CONTEXT_MATCH_TAG3 ||
> FW_CDEV_ISO_CONTEXT_MATCH_ALL_TAGS != FW_ISO_CONTEXT_MATCH_ALL_TAGS);
>
> + guard(mutex)(&client->iso_context_mutex);
> +
> if (client->iso_context == NULL || a->handle != 0)
> return -EINVAL;
>
> @@ -1267,6 +1275,8 @@ static int ioctl_stop_iso(struct client *client, union ioctl_arg *arg)
> {
> struct fw_cdev_stop_iso *a = &arg->stop_iso;
>
> + guard(mutex)(&client->iso_context_mutex);
> +
> if (client->iso_context == NULL || a->handle != 0)
> return -EINVAL;
>
> @@ -1277,6 +1287,8 @@ static int ioctl_flush_iso(struct client *client, union ioctl_arg *arg)
> {
> struct fw_cdev_flush_iso *a = &arg->flush_iso;
>
> + guard(mutex)(&client->iso_context_mutex);
> +
> if (client->iso_context == NULL || a->handle != 0)
> return -EINVAL;
>
> @@ -1937,8 +1949,10 @@ static int fw_device_op_release(struct inode *inode, struct file *file)
> scoped_guard(mutex, &client->device->client_list_mutex)
> list_del(&client->link);
>
> - if (client->iso_context)
> - fw_iso_context_destroy(client->iso_context);
> + scoped_guard(mutex, &client->iso_context_mutex) {
> + if (client->iso_context)
> + fw_iso_context_destroy(client->iso_context);
> + }
> mutex_destroy(&client->iso_context_mutex);
>
> if (client->buffer.pages)
> --
> 2.53.0
Regards
Takashi Sakamoto
next prev parent reply other threads:[~2026-11-02 10:40 UTC|newest]
Thread overview: 11+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-11-02 4:19 [PATCH 0/9] firewire: cdev: add annotations for compiler-based context analysis Takashi Sakamoto
2026-11-02 4:19 ` [PATCH 1/9] firewire: core: add context analysis annotations for address handler Takashi Sakamoto
2026-11-02 4:19 ` [PATCH 2/9] firewire: cdev: add context analysis annotations for phy_receiver_list Takashi Sakamoto
2026-11-02 4:19 ` [PATCH 3/9] firewire: cdev: add context analysis annotations for isochronous context Takashi Sakamoto
2026-11-02 10:40 ` Takashi Sakamoto [this message]
2026-11-02 4:19 ` [PATCH 4/9] firewire: cdev: annotate data length in inbound_transaction_resource Takashi Sakamoto
2026-11-02 4:19 ` [PATCH 5/9] firewire: core: add context analysis annotation to queue_bus_reset_event() Takashi Sakamoto
2026-11-02 4:19 ` [PATCH 6/9] firewire: cdev: add context analysis annotations for client shutdown flag and event list Takashi Sakamoto
2026-11-02 4:19 ` [PATCH 7/9] firewire: cdev: use cleanup function for outbound transaction event allocation Takashi Sakamoto
2026-11-02 4:19 ` [PATCH 8/9] firewire: cdev: use cleanup function for descriptor_resource allocation Takashi Sakamoto
2026-11-02 4:19 ` [PATCH 9/9] firewire: cdev: enable compiler-based context analysis Takashi Sakamoto
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=20261102104042.GA1197999@sakamocchi.jp \
--to=o-takashi@sakamocchi.jp \
--cc=linux-kernel@vger.kernel.org \
--cc=linux1394-devel@lists.sourceforge.net \
/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
all inboxes | Powered by JetHome®