* [PATCH v1 0/2] usb: gadget: dummy_hcd: Switch to hrtimers for better performance @ 2024-04-11 15:22 Marcello Sylvester Bauer 2024-04-11 15:22 ` Marcello Sylvester Bauer 2024-04-11 15:22 ` Marcello Sylvester Bauer 0 siblings, 2 replies; 7+ messages in thread From: Marcello Sylvester Bauer @ 2024-04-11 15:22 UTC (permalink / raw) To: linux-kernel, linux-usb Cc: Greg Kroah-Hartman, Marcello Sylvester Bauer, Matthias Stoeckl, Uwe Kleine-Koenig, Marcello Sylvester Bauer About 10 years ago, there was a discussion about the poor performance of dummy_hcd due to the use of jiffies instead of hrtimers [1]. There was a proposal to switch to hrtimers, but the patch never made it into the mainline, and I cannot figure out why. This patch set brings the idea back to the mailing list with the intention of improving the overall performance of this driver. [1](https://linux-usb.vger.kernel.narkive.com/Fdt5oBR8/dummy-hcd-performance-correctness) Marcello Sylvester Bauer (2): usb: gadget: dummy_hcd: Switch to hrtimer transfer scheduler usb: gadget: dummy_hcd: Set transfer interval to 1 microframe drivers/usb/gadget/udc/dummy_hcd.c | 37 ++++++++++++++++++------------ 1 file changed, 22 insertions(+), 15 deletions(-) -- 2.44.0 ^ permalink raw reply [flat|nested] 7+ messages in thread
* [PATCH v1 1/2] usb: gadget: dummy_hcd: Switch to hrtimer transfer scheduler @ 2024-04-11 15:22 ` Marcello Sylvester Bauer 0 siblings, 0 replies; 7+ messages in thread From: Marcello Sylvester Bauer @ 2024-04-11 14:51 UTC (permalink / raw) To: Greg Kroah-Hartman Cc: Marcello Sylvester Bauer, Marcello Sylvester Bauer, Matthias Stöckl, linux-usb, linux-kernel From: Marcello Sylvester Bauer <sylv@sylv.io> The dummy_hcd transfer scheduler assumes that the internal kernel timer frequency is set to 1000Hz to give a polling interval of 1ms. Reducing the timer frequency will result in an anti-proportional reduction in transfer performance. Switch to a hrtimer to decouple this association. Signed-off-by: Marcello Sylvester Bauer <marcello.bauer@9elements.com> Signed-off-by: Marcello Sylvester Bauer <sylv@sylv.io> --- drivers/usb/gadget/udc/dummy_hcd.c | 35 +++++++++++++++++------------- 1 file changed, 20 insertions(+), 15 deletions(-) diff --git a/drivers/usb/gadget/udc/dummy_hcd.c b/drivers/usb/gadget/udc/dummy_hcd.c index 0953e1b5c030..dab559d8ee8c 100644 --- a/drivers/usb/gadget/udc/dummy_hcd.c +++ b/drivers/usb/gadget/udc/dummy_hcd.c @@ -30,7 +30,7 @@ #include <linux/slab.h> #include <linux/errno.h> #include <linux/init.h> -#include <linux/timer.h> +#include <linux/hrtimer.h> #include <linux/list.h> #include <linux/interrupt.h> #include <linux/platform_device.h> @@ -240,7 +240,7 @@ enum dummy_rh_state { struct dummy_hcd { struct dummy *dum; enum dummy_rh_state rh_state; - struct timer_list timer; + struct hrtimer timer; u32 port_status; u32 old_status; unsigned long re_timeout; @@ -1301,8 +1301,8 @@ static int dummy_urb_enqueue( urb->error_count = 1; /* mark as a new urb */ /* kick the scheduler, it'll do the rest */ - if (!timer_pending(&dum_hcd->timer)) - mod_timer(&dum_hcd->timer, jiffies + 1); + if (!hrtimer_active(&dum_hcd->timer)) + hrtimer_start(&dum_hcd->timer, ms_to_ktime(1), HRTIMER_MODE_REL); done: spin_unlock_irqrestore(&dum_hcd->dum->lock, flags); @@ -1323,7 +1323,7 @@ static int dummy_urb_dequeue(struct usb_hcd *hcd, struct urb *urb, int status) rc = usb_hcd_check_unlink_urb(hcd, urb, status); if (!rc && dum_hcd->rh_state != DUMMY_RH_RUNNING && !list_empty(&dum_hcd->urbp_list)) - mod_timer(&dum_hcd->timer, jiffies); + hrtimer_start(&dum_hcd->timer, ns_to_ktime(0), HRTIMER_MODE_REL); spin_unlock_irqrestore(&dum_hcd->dum->lock, flags); return rc; @@ -1777,7 +1777,7 @@ static int handle_control_request(struct dummy_hcd *dum_hcd, struct urb *urb, * drivers except that the callbacks are invoked from soft interrupt * context. */ -static void dummy_timer(struct timer_list *t) +static enum hrtimer_restart dummy_timer(struct hrtimer *t) { struct dummy_hcd *dum_hcd = from_timer(dum_hcd, t, timer); struct dummy *dum = dum_hcd->dum; @@ -1808,8 +1808,6 @@ static void dummy_timer(struct timer_list *t) break; } - /* FIXME if HZ != 1000 this will probably misbehave ... */ - /* look at each urb queued by the host side driver */ spin_lock_irqsave(&dum->lock, flags); @@ -1817,7 +1815,7 @@ static void dummy_timer(struct timer_list *t) dev_err(dummy_dev(dum_hcd), "timer fired with no URBs pending?\n"); spin_unlock_irqrestore(&dum->lock, flags); - return; + return HRTIMER_NORESTART; } dum_hcd->next_frame_urbp = NULL; @@ -1995,10 +1993,12 @@ static void dummy_timer(struct timer_list *t) dum_hcd->udev = NULL; } else if (dum_hcd->rh_state == DUMMY_RH_RUNNING) { /* want a 1 msec delay here */ - mod_timer(&dum_hcd->timer, jiffies + msecs_to_jiffies(1)); + hrtimer_start(&dum_hcd->timer, ms_to_ktime(1), HRTIMER_MODE_REL); } spin_unlock_irqrestore(&dum->lock, flags); + + return HRTIMER_NORESTART; } /*-------------------------------------------------------------------------*/ @@ -2387,7 +2387,7 @@ static int dummy_bus_resume(struct usb_hcd *hcd) dum_hcd->rh_state = DUMMY_RH_RUNNING; set_link_state(dum_hcd); if (!list_empty(&dum_hcd->urbp_list)) - mod_timer(&dum_hcd->timer, jiffies); + hrtimer_start(&dum_hcd->timer, ns_to_ktime(0), HRTIMER_MODE_REL); hcd->state = HC_STATE_RUNNING; } spin_unlock_irq(&dum_hcd->dum->lock); @@ -2465,7 +2465,8 @@ static DEVICE_ATTR_RO(urbs); static int dummy_start_ss(struct dummy_hcd *dum_hcd) { - timer_setup(&dum_hcd->timer, dummy_timer, 0); + hrtimer_init(&dum_hcd->timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); + dum_hcd->timer.function = dummy_timer; dum_hcd->rh_state = DUMMY_RH_RUNNING; dum_hcd->stream_en_ep = 0; INIT_LIST_HEAD(&dum_hcd->urbp_list); @@ -2494,7 +2495,8 @@ static int dummy_start(struct usb_hcd *hcd) return dummy_start_ss(dum_hcd); spin_lock_init(&dum_hcd->dum->lock); - timer_setup(&dum_hcd->timer, dummy_timer, 0); + hrtimer_init(&dum_hcd->timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); + dum_hcd->timer.function = dummy_timer; dum_hcd->rh_state = DUMMY_RH_RUNNING; INIT_LIST_HEAD(&dum_hcd->urbp_list); @@ -2513,8 +2515,11 @@ static int dummy_start(struct usb_hcd *hcd) static void dummy_stop(struct usb_hcd *hcd) { - device_remove_file(dummy_dev(hcd_to_dummy_hcd(hcd)), &dev_attr_urbs); - dev_info(dummy_dev(hcd_to_dummy_hcd(hcd)), "stopped\n"); + struct dummy_hcd *dum_hcd = hcd_to_dummy_hcd(hcd); + + hrtimer_cancel(&dum_hcd->timer); + device_remove_file(dummy_dev(dum_hcd), &dev_attr_urbs); + dev_info(dummy_dev(dum_hcd), "stopped\n"); } /*-------------------------------------------------------------------------*/ -- 2.44.0 ^ permalink raw reply related [flat|nested] 7+ messages in thread
* [PATCH v1 1/2] usb: gadget: dummy_hcd: Switch to hrtimer transfer scheduler @ 2024-04-11 15:22 ` Marcello Sylvester Bauer 0 siblings, 0 replies; 7+ messages in thread From: Marcello Sylvester Bauer @ 2024-04-11 15:22 UTC (permalink / raw) To: linux-kernel, linux-usb, Greg Kroah-Hartman Cc: Marcello Sylvester Bauer, Matthias Stoeckl, Uwe Kleine-Koenig, Marcello Sylvester Bauer The dummy_hcd transfer scheduler assumes that the internal kernel timer frequency is set to 1000Hz to give a polling interval of 1ms. Reducing the timer frequency will result in an anti-proportional reduction in transfer performance. Switch to a hrtimer to decouple this association. Signed-off-by: Marcello Sylvester Bauer <marcello.bauer@9elements.com> Signed-off-by: Marcello Sylvester Bauer <sylv@sylv.io> --- drivers/usb/gadget/udc/dummy_hcd.c | 35 +++++++++++++++++------------- 1 file changed, 20 insertions(+), 15 deletions(-) diff --git a/drivers/usb/gadget/udc/dummy_hcd.c b/drivers/usb/gadget/udc/dummy_hcd.c index 0953e1b5c030..dab559d8ee8c 100644 --- a/drivers/usb/gadget/udc/dummy_hcd.c +++ b/drivers/usb/gadget/udc/dummy_hcd.c @@ -30,7 +30,7 @@ #include <linux/slab.h> #include <linux/errno.h> #include <linux/init.h> -#include <linux/timer.h> +#include <linux/hrtimer.h> #include <linux/list.h> #include <linux/interrupt.h> #include <linux/platform_device.h> @@ -240,7 +240,7 @@ enum dummy_rh_state { struct dummy_hcd { struct dummy *dum; enum dummy_rh_state rh_state; - struct timer_list timer; + struct hrtimer timer; u32 port_status; u32 old_status; unsigned long re_timeout; @@ -1301,8 +1301,8 @@ static int dummy_urb_enqueue( urb->error_count = 1; /* mark as a new urb */ /* kick the scheduler, it'll do the rest */ - if (!timer_pending(&dum_hcd->timer)) - mod_timer(&dum_hcd->timer, jiffies + 1); + if (!hrtimer_active(&dum_hcd->timer)) + hrtimer_start(&dum_hcd->timer, ms_to_ktime(1), HRTIMER_MODE_REL); done: spin_unlock_irqrestore(&dum_hcd->dum->lock, flags); @@ -1323,7 +1323,7 @@ static int dummy_urb_dequeue(struct usb_hcd *hcd, struct urb *urb, int status) rc = usb_hcd_check_unlink_urb(hcd, urb, status); if (!rc && dum_hcd->rh_state != DUMMY_RH_RUNNING && !list_empty(&dum_hcd->urbp_list)) - mod_timer(&dum_hcd->timer, jiffies); + hrtimer_start(&dum_hcd->timer, ns_to_ktime(0), HRTIMER_MODE_REL); spin_unlock_irqrestore(&dum_hcd->dum->lock, flags); return rc; @@ -1777,7 +1777,7 @@ static int handle_control_request(struct dummy_hcd *dum_hcd, struct urb *urb, * drivers except that the callbacks are invoked from soft interrupt * context. */ -static void dummy_timer(struct timer_list *t) +static enum hrtimer_restart dummy_timer(struct hrtimer *t) { struct dummy_hcd *dum_hcd = from_timer(dum_hcd, t, timer); struct dummy *dum = dum_hcd->dum; @@ -1808,8 +1808,6 @@ static void dummy_timer(struct timer_list *t) break; } - /* FIXME if HZ != 1000 this will probably misbehave ... */ - /* look at each urb queued by the host side driver */ spin_lock_irqsave(&dum->lock, flags); @@ -1817,7 +1815,7 @@ static void dummy_timer(struct timer_list *t) dev_err(dummy_dev(dum_hcd), "timer fired with no URBs pending?\n"); spin_unlock_irqrestore(&dum->lock, flags); - return; + return HRTIMER_NORESTART; } dum_hcd->next_frame_urbp = NULL; @@ -1995,10 +1993,12 @@ static void dummy_timer(struct timer_list *t) dum_hcd->udev = NULL; } else if (dum_hcd->rh_state == DUMMY_RH_RUNNING) { /* want a 1 msec delay here */ - mod_timer(&dum_hcd->timer, jiffies + msecs_to_jiffies(1)); + hrtimer_start(&dum_hcd->timer, ms_to_ktime(1), HRTIMER_MODE_REL); } spin_unlock_irqrestore(&dum->lock, flags); + + return HRTIMER_NORESTART; } /*-------------------------------------------------------------------------*/ @@ -2387,7 +2387,7 @@ static int dummy_bus_resume(struct usb_hcd *hcd) dum_hcd->rh_state = DUMMY_RH_RUNNING; set_link_state(dum_hcd); if (!list_empty(&dum_hcd->urbp_list)) - mod_timer(&dum_hcd->timer, jiffies); + hrtimer_start(&dum_hcd->timer, ns_to_ktime(0), HRTIMER_MODE_REL); hcd->state = HC_STATE_RUNNING; } spin_unlock_irq(&dum_hcd->dum->lock); @@ -2465,7 +2465,8 @@ static DEVICE_ATTR_RO(urbs); static int dummy_start_ss(struct dummy_hcd *dum_hcd) { - timer_setup(&dum_hcd->timer, dummy_timer, 0); + hrtimer_init(&dum_hcd->timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); + dum_hcd->timer.function = dummy_timer; dum_hcd->rh_state = DUMMY_RH_RUNNING; dum_hcd->stream_en_ep = 0; INIT_LIST_HEAD(&dum_hcd->urbp_list); @@ -2494,7 +2495,8 @@ static int dummy_start(struct usb_hcd *hcd) return dummy_start_ss(dum_hcd); spin_lock_init(&dum_hcd->dum->lock); - timer_setup(&dum_hcd->timer, dummy_timer, 0); + hrtimer_init(&dum_hcd->timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); + dum_hcd->timer.function = dummy_timer; dum_hcd->rh_state = DUMMY_RH_RUNNING; INIT_LIST_HEAD(&dum_hcd->urbp_list); @@ -2513,8 +2515,11 @@ static int dummy_start(struct usb_hcd *hcd) static void dummy_stop(struct usb_hcd *hcd) { - device_remove_file(dummy_dev(hcd_to_dummy_hcd(hcd)), &dev_attr_urbs); - dev_info(dummy_dev(hcd_to_dummy_hcd(hcd)), "stopped\n"); + struct dummy_hcd *dum_hcd = hcd_to_dummy_hcd(hcd); + + hrtimer_cancel(&dum_hcd->timer); + device_remove_file(dummy_dev(dum_hcd), &dev_attr_urbs); + dev_info(dummy_dev(dum_hcd), "stopped\n"); } /*-------------------------------------------------------------------------*/ -- 2.44.0 ^ permalink raw reply related [flat|nested] 7+ messages in thread
* Re: [PATCH v1 1/2] usb: gadget: dummy_hcd: Switch to hrtimer transfer scheduler 2024-04-11 15:22 ` Marcello Sylvester Bauer (?) @ 2024-04-11 16:30 ` Alan Stern -1 siblings, 0 replies; 7+ messages in thread From: Alan Stern @ 2024-04-11 16:30 UTC (permalink / raw) To: Marcello Sylvester Bauer Cc: linux-kernel, linux-usb, Greg Kroah-Hartman, Marcello Sylvester Bauer, Matthias Stoeckl, Uwe Kleine-Koenig On Thu, Apr 11, 2024 at 05:22:10PM +0200, Marcello Sylvester Bauer wrote: > The dummy_hcd transfer scheduler assumes that the internal kernel timer > frequency is set to 1000Hz to give a polling interval of 1ms. Reducing > the timer frequency will result in an anti-proportional reduction in > transfer performance. Switch to a hrtimer to decouple this association. > > Signed-off-by: Marcello Sylvester Bauer <marcello.bauer@9elements.com> > Signed-off-by: Marcello Sylvester Bauer <sylv@sylv.io> > --- Reviewed-by: Alan Stern <stern@rowland.harvard.edu> ^ permalink raw reply [flat|nested] 7+ messages in thread
* [PATCH v1 2/2] usb: gadget: dummy_hcd: Set transfer interval to 1 microframe @ 2024-04-11 15:22 ` Marcello Sylvester Bauer 0 siblings, 0 replies; 7+ messages in thread From: Marcello Sylvester Bauer @ 2024-04-11 14:51 UTC (permalink / raw) To: Greg Kroah-Hartman Cc: Marcello Sylvester Bauer, Marcello Sylvester Bauer, Matthias Stöckl, linux-usb, linux-kernel From: Marcello Sylvester Bauer <sylv@sylv.io> Currently, the transfer polling interval is set to 1ms, which is the frame rate of full-speed and low-speed USB. The USB 2.0 specification introduces microframes (125 microseconds) to improve the timing precision of data transfers. Reducing the transfer interval to 1 microframe increases data throughput for high-speed and super-speed USB communication Signed-off-by: Marcello Sylvester Bauer <marcello.bauer@9elements.com> Signed-off-by: Marcello Sylvester Bauer <sylv@sylv.io> --- drivers/usb/gadget/udc/dummy_hcd.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/usb/gadget/udc/dummy_hcd.c b/drivers/usb/gadget/udc/dummy_hcd.c index dab559d8ee8c..f37b0d8386c1 100644 --- a/drivers/usb/gadget/udc/dummy_hcd.c +++ b/drivers/usb/gadget/udc/dummy_hcd.c @@ -50,6 +50,8 @@ #define POWER_BUDGET 500 /* in mA; use 8 for low-power port testing */ #define POWER_BUDGET_3 900 /* in mA */ +#define DUMMY_TIMER_INT_NSECS 125000 /* 1 microframe */ + static const char driver_name[] = "dummy_hcd"; static const char driver_desc[] = "USB Host+Gadget Emulator"; @@ -1302,7 +1304,7 @@ static int dummy_urb_enqueue( /* kick the scheduler, it'll do the rest */ if (!hrtimer_active(&dum_hcd->timer)) - hrtimer_start(&dum_hcd->timer, ms_to_ktime(1), HRTIMER_MODE_REL); + hrtimer_start(&dum_hcd->timer, ns_to_ktime(DUMMY_TIMER_INT_NSECS), HRTIMER_MODE_REL); done: spin_unlock_irqrestore(&dum_hcd->dum->lock, flags); @@ -1993,7 +1995,7 @@ static enum hrtimer_restart dummy_timer(struct hrtimer *t) dum_hcd->udev = NULL; } else if (dum_hcd->rh_state == DUMMY_RH_RUNNING) { /* want a 1 msec delay here */ - hrtimer_start(&dum_hcd->timer, ms_to_ktime(1), HRTIMER_MODE_REL); + hrtimer_start(&dum_hcd->timer, ns_to_ktime(DUMMY_TIMER_INT_NSECS), HRTIMER_MODE_REL); } spin_unlock_irqrestore(&dum->lock, flags); -- 2.44.0 ^ permalink raw reply related [flat|nested] 7+ messages in thread
* [PATCH v1 2/2] usb: gadget: dummy_hcd: Set transfer interval to 1 microframe @ 2024-04-11 15:22 ` Marcello Sylvester Bauer 0 siblings, 0 replies; 7+ messages in thread From: Marcello Sylvester Bauer @ 2024-04-11 15:22 UTC (permalink / raw) To: linux-kernel, linux-usb, Greg Kroah-Hartman Cc: Marcello Sylvester Bauer, Matthias Stoeckl, Uwe Kleine-Koenig, Marcello Sylvester Bauer Currently, the transfer polling interval is set to 1ms, which is the frame rate of full-speed and low-speed USB. The USB 2.0 specification introduces microframes (125 microseconds) to improve the timing precision of data transfers. Reducing the transfer interval to 1 microframe increases data throughput for high-speed and super-speed USB communication Signed-off-by: Marcello Sylvester Bauer <marcello.bauer@9elements.com> Signed-off-by: Marcello Sylvester Bauer <sylv@sylv.io> --- drivers/usb/gadget/udc/dummy_hcd.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/usb/gadget/udc/dummy_hcd.c b/drivers/usb/gadget/udc/dummy_hcd.c index dab559d8ee8c..f37b0d8386c1 100644 --- a/drivers/usb/gadget/udc/dummy_hcd.c +++ b/drivers/usb/gadget/udc/dummy_hcd.c @@ -50,6 +50,8 @@ #define POWER_BUDGET 500 /* in mA; use 8 for low-power port testing */ #define POWER_BUDGET_3 900 /* in mA */ +#define DUMMY_TIMER_INT_NSECS 125000 /* 1 microframe */ + static const char driver_name[] = "dummy_hcd"; static const char driver_desc[] = "USB Host+Gadget Emulator"; @@ -1302,7 +1304,7 @@ static int dummy_urb_enqueue( /* kick the scheduler, it'll do the rest */ if (!hrtimer_active(&dum_hcd->timer)) - hrtimer_start(&dum_hcd->timer, ms_to_ktime(1), HRTIMER_MODE_REL); + hrtimer_start(&dum_hcd->timer, ns_to_ktime(DUMMY_TIMER_INT_NSECS), HRTIMER_MODE_REL); done: spin_unlock_irqrestore(&dum_hcd->dum->lock, flags); @@ -1993,7 +1995,7 @@ static enum hrtimer_restart dummy_timer(struct hrtimer *t) dum_hcd->udev = NULL; } else if (dum_hcd->rh_state == DUMMY_RH_RUNNING) { /* want a 1 msec delay here */ - hrtimer_start(&dum_hcd->timer, ms_to_ktime(1), HRTIMER_MODE_REL); + hrtimer_start(&dum_hcd->timer, ns_to_ktime(DUMMY_TIMER_INT_NSECS), HRTIMER_MODE_REL); } spin_unlock_irqrestore(&dum->lock, flags); -- 2.44.0 ^ permalink raw reply related [flat|nested] 7+ messages in thread
* Re: [PATCH v1 2/2] usb: gadget: dummy_hcd: Set transfer interval to 1 microframe 2024-04-11 15:22 ` Marcello Sylvester Bauer (?) @ 2024-04-11 16:21 ` Alan Stern -1 siblings, 0 replies; 7+ messages in thread From: Alan Stern @ 2024-04-11 16:21 UTC (permalink / raw) To: Marcello Sylvester Bauer Cc: linux-kernel, linux-usb, Greg Kroah-Hartman, Marcello Sylvester Bauer, Matthias Stoeckl, Uwe Kleine-Koenig On Thu, Apr 11, 2024 at 05:22:11PM +0200, Marcello Sylvester Bauer wrote: > Currently, the transfer polling interval is set to 1ms, which is the > frame rate of full-speed and low-speed USB. The USB 2.0 specification > introduces microframes (125 microseconds) to improve the timing > precision of data transfers. > > Reducing the transfer interval to 1 microframe increases data throughput > for high-speed and super-speed USB communication > > Signed-off-by: Marcello Sylvester Bauer <marcello.bauer@9elements.com> > Signed-off-by: Marcello Sylvester Bauer <sylv@sylv.io> > --- > drivers/usb/gadget/udc/dummy_hcd.c | 6 ++++-- > 1 file changed, 4 insertions(+), 2 deletions(-) > > diff --git a/drivers/usb/gadget/udc/dummy_hcd.c b/drivers/usb/gadget/udc/dummy_hcd.c > index dab559d8ee8c..f37b0d8386c1 100644 > --- a/drivers/usb/gadget/udc/dummy_hcd.c > +++ b/drivers/usb/gadget/udc/dummy_hcd.c > @@ -50,6 +50,8 @@ > #define POWER_BUDGET 500 /* in mA; use 8 for low-power port testing */ > #define POWER_BUDGET_3 900 /* in mA */ > > +#define DUMMY_TIMER_INT_NSECS 125000 /* 1 microframe */ If the is_super_speed and is_high_speed module parameters are both false then the timer should run at 1-ms intervals. So this needs to be something that can be adjusted at runtime, not a constant. Alan Stern ^ permalink raw reply [flat|nested] 7+ messages in thread
end of thread, other threads:[~2024-04-11 16:30 UTC | newest] Thread overview: 7+ messages (download: mbox.gz follow: Atom feed -- links below jump to the message on this page -- 2024-04-11 15:22 [PATCH v1 0/2] usb: gadget: dummy_hcd: Switch to hrtimers for better performance Marcello Sylvester Bauer 2024-04-11 14:51 ` [PATCH v1 1/2] usb: gadget: dummy_hcd: Switch to hrtimer transfer scheduler Marcello Sylvester Bauer 2024-04-11 15:22 ` Marcello Sylvester Bauer 2024-04-11 16:30 ` Alan Stern 2024-04-11 14:51 ` [PATCH v1 2/2] usb: gadget: dummy_hcd: Set transfer interval to 1 microframe Marcello Sylvester Bauer 2024-04-11 15:22 ` Marcello Sylvester Bauer 2024-04-11 16:21 ` Alan Stern
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.