| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <87h6n38l0n.fsf@intel.com>
Date: Fri, 06 Oct 2023 15:05:12 -0700
From: Vinicius Costa Gomes <vinicius.gomes@...el.com>
To: Xabier Marquiegui <reibax@...il.com>, netdev@...r.kernel.org
Cc: richardcochran@...il.com, tglx@...utronix.de, jstultz@...gle.com,
horms@...nel.org, chrony-dev@...ony.tuxfamily.org, mlichvar@...hat.com,
reibax@...il.com, ntp-lists@...tcorallo.com, alex.maftei@....com,
davem@...emloft.net, rrameshbabu@...dia.com, shuah@...nel.org
Subject: Re: [PATCH net-next v4 4/6] ptp: support event queue reader channel
masks
Xabier Marquiegui <reibax@...il.com> writes:
> On systems with multiple timestamp event channels, some readers might
> want to receive only a subset of those channels.
>
> This patch adds the necessary modifications to support timestamp event
> channel filtering, including two IOCTL operations:
>
> - Clear all channels
> - Enable one channel
>
> The mask modification operations will be applied exclusively on the
> event queue assigned to the file descriptor used on the IOCTL operation,
> so the typical procedure to have a reader receiving only a subset of the
> enabled channels would be:
>
> - Open device file
> - ioctl: clear all channels
> - ioctl: enable one channel
> - start reading
>
> Calling the enable one channel ioctl more than once will result in
> multiple enabled channels.
>
> Signed-off-by: Xabier Marquiegui <reibax@...il.com>
> Suggested-by: Richard Cochran <richardcochran@...il.com>
> Suggested-by: Vinicius Costa Gomes <vinicius.gomes@...el.com>
> ---
> v4:
> - split modifications in different patches for improved organization
> - filter modifications exclusive to currently open instance for
> simplicity and security
> - expand mask to 2048 channels
> - remove unnecessary tests
> v3: https://lore.kernel.org/netdev/20230928133544.3642650-4-reibax@gmail.com/
> - filter application by object id, aided by process id
> - friendlier testptp implementation of event queue channel filters
> v2: https://lore.kernel.org/netdev/20230912220217.2008895-3-reibax@gmail.com/
> - fix testptp compilation error: unknown type name 'pid_t'
> - rename mask variable for easier code traceability
> - more detailed commit message with two examples
> v1: https://lore.kernel.org/netdev/20230906104754.1324412-4-reibax@gmail.com/
> ---
> drivers/ptp/ptp_chardev.c | 24 ++++++++++++++++++++++++
> drivers/ptp/ptp_clock.c | 12 ++++++++++--
> drivers/ptp/ptp_private.h | 3 +++
> include/uapi/linux/ptp_clock.h | 2 ++
> 4 files changed, 39 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/ptp/ptp_chardev.c b/drivers/ptp/ptp_chardev.c
> index abe94bb80cf6..dbbe551a044f 100644
> --- a/drivers/ptp/ptp_chardev.c
> +++ b/drivers/ptp/ptp_chardev.c
> @@ -110,6 +110,10 @@ int ptp_open(struct posix_clock_context *pccontext, fmode_t fmode)
> queue = kzalloc(sizeof(*queue), GFP_KERNEL);
> if (!queue)
> return -EINVAL;
> + queue->mask = bitmap_alloc(PTP_MAX_CHANNELS, GFP_KERNEL);
> + if (!queue->mask)
> + return -EINVAL;
> + bitmap_set(queue->mask, 0, PTP_MAX_CHANNELS);
> spin_lock_init(&queue->lock);
> list_add_tail(&queue->qlist, &ptp->tsevqs);
> pccontext->private_clkdata = queue;
> @@ -126,6 +130,7 @@ int ptp_release(struct posix_clock_context *pccontext)
> spin_lock_irqsave(&queue->lock, flags);
> list_del(&queue->qlist);
> spin_unlock_irqrestore(&queue->lock, flags);
> + bitmap_free(queue->mask);
> kfree(queue);
> }
> return 0;
> @@ -141,6 +146,7 @@ long ptp_ioctl(struct posix_clock_context *pccontext, unsigned int cmd,
> struct system_device_crosststamp xtstamp;
> struct ptp_clock_info *ops = ptp->info;
> struct ptp_sys_offset *sysoff = NULL;
> + struct timestamp_event_queue *tsevq;
> struct ptp_system_timestamp sts;
> struct ptp_clock_request req;
> struct ptp_clock_caps caps;
> @@ -150,6 +156,8 @@ long ptp_ioctl(struct posix_clock_context *pccontext, unsigned int cmd,
> struct timespec64 ts;
> int enable, err = 0;
>
> + tsevq = pccontext->private_clkdata;
> +
> switch (cmd) {
>
> case PTP_CLOCK_GETCAPS:
> @@ -448,6 +456,22 @@ long ptp_ioctl(struct posix_clock_context *pccontext, unsigned int cmd,
> mutex_unlock(&ptp->pincfg_mux);
> break;
>
> + case PTP_MASK_CLEAR_ALL:
> + bitmap_clear(tsevq->mask, 0, PTP_MAX_CHANNELS);
> + break;
> +
> + case PTP_MASK_EN_SINGLE:
> + if (copy_from_user(&i, (void __user *)arg, sizeof(i))) {
> + err = -EFAULT;
> + break;
> + }
> + if (i >= PTP_MAX_CHANNELS) {
> + err = -EFAULT;
> + break;
> + }
> + set_bit(i, tsevq->mask);
> + break;
> +
> default:
> err = -ENOTTY;
> break;
> diff --git a/drivers/ptp/ptp_clock.c b/drivers/ptp/ptp_clock.c
> index 74f1ce2dbccb..ed16d9787ce9 100644
> --- a/drivers/ptp/ptp_clock.c
> +++ b/drivers/ptp/ptp_clock.c
> @@ -183,6 +183,7 @@ static void ptp_clock_release(struct device *dev)
> spin_lock_irqsave(&tsevq->lock, flags);
> list_del(&tsevq->qlist);
> spin_unlock_irqrestore(&tsevq->lock, flags);
> + bitmap_free(tsevq->mask);
> kfree(tsevq);
> ida_free(&ptp_clocks_map, ptp->index);
> kfree(ptp);
> @@ -243,6 +244,10 @@ struct ptp_clock *ptp_clock_register(struct ptp_clock_info *info,
> if (!queue)
> goto no_memory_queue;
> list_add_tail(&queue->qlist, &ptp->tsevqs);
> + queue->mask = bitmap_alloc(PTP_MAX_CHANNELS, GFP_KERNEL);
> + if (!queue->mask)
> + goto no_memory_bitmap;
> + bitmap_set(queue->mask, 0, PTP_MAX_CHANNELS);
Sorry that I only noticed a (possible) change in behavior now.
Before this series, when there was a single queue, events where
accumulated until the application reads the fd associated with the PTP
device. i.e. it doesn't matter when the application calls open().
AFter this series events, are only accumulated after the queue
associated with that fd is created, i.e. after open(). Events that
happened before open() are lost (is this true? are we leaking them?).
Is this a desired/wanted change? Is it possible that we have
applications that depend on the "old" behavior?
> spin_lock_init(&queue->lock);
> mutex_init(&ptp->pincfg_mux);
> mutex_init(&ptp->n_vclocks_mux);
> @@ -346,6 +351,8 @@ struct ptp_clock *ptp_clock_register(struct ptp_clock_info *info,
> kworker_err:
> mutex_destroy(&ptp->pincfg_mux);
> mutex_destroy(&ptp->n_vclocks_mux);
> + bitmap_free(queue->mask);
> +no_memory_bitmap:
> list_del(&queue->qlist);
> kfree(queue);
> no_memory_queue:
> @@ -400,9 +407,10 @@ void ptp_clock_event(struct ptp_clock *ptp, struct ptp_clock_event *event)
> break;
>
> case PTP_CLOCK_EXTTS:
> - /* Enqueue timestamp on all queues */
> + /* Enqueue timestamp on selected queues */
> list_for_each_entry(tsevq, &ptp->tsevqs, qlist) {
> - enqueue_external_timestamp(tsevq, event);
> + if (test_bit((unsigned int)event->index, tsevq->mask))
> + enqueue_external_timestamp(tsevq, event);
> }
> wake_up_interruptible(&ptp->tsev_wq);
> break;
> diff --git a/drivers/ptp/ptp_private.h b/drivers/ptp/ptp_private.h
> index 9d5f3d95058e..ad4ce1b25c86 100644
> --- a/drivers/ptp/ptp_private.h
> +++ b/drivers/ptp/ptp_private.h
> @@ -16,10 +16,12 @@
> #include <linux/ptp_clock_kernel.h>
> #include <linux/time.h>
> #include <linux/list.h>
> +#include <linux/bitmap.h>
>
> #define PTP_MAX_TIMESTAMPS 128
> #define PTP_BUF_TIMESTAMPS 30
> #define PTP_DEFAULT_MAX_VCLOCKS 20
> +#define PTP_MAX_CHANNELS 2048
>
> struct timestamp_event_queue {
> struct ptp_extts_event buf[PTP_MAX_TIMESTAMPS];
> @@ -27,6 +29,7 @@ struct timestamp_event_queue {
> int tail;
> spinlock_t lock;
> struct list_head qlist;
> + unsigned long *mask;
> };
>
> struct ptp_clock {
> diff --git a/include/uapi/linux/ptp_clock.h b/include/uapi/linux/ptp_clock.h
> index 05cc35fc94ac..da700999cad4 100644
> --- a/include/uapi/linux/ptp_clock.h
> +++ b/include/uapi/linux/ptp_clock.h
> @@ -224,6 +224,8 @@ struct ptp_pin_desc {
> _IOWR(PTP_CLK_MAGIC, 17, struct ptp_sys_offset_precise)
> #define PTP_SYS_OFFSET_EXTENDED2 \
> _IOWR(PTP_CLK_MAGIC, 18, struct ptp_sys_offset_extended)
> +#define PTP_MASK_CLEAR_ALL _IO(PTP_CLK_MAGIC, 19)
> +#define PTP_MASK_EN_SINGLE _IOW(PTP_CLK_MAGIC, 20, unsigned int)
>
> struct ptp_extts_event {
> struct ptp_clock_time t; /* Time event occured. */
> --
> 2.34.1
>
--
Vinicius
Powered by blists - more mailing lists