[Bug 226961] [wsp] [patch] Add support for right & middle click with integrated button
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=226961 Mark Linimon changed: What|Removed |Added Keywords||patch -- You are receiving this mail because: You are the assignee for the bug. ___ freebsd-usb@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-usb To unsubscribe, send any mail to "freebsd-usb-unsubscr...@freebsd.org"
[Bug 226961] [wsp] [patch] Add support for right & middle click with integrated button
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=226961 Hans Petter Selasky changed: What|Removed |Added CC||hsela...@freebsd.org Assignee|freebsd-usb@FreeBSD.org |hsela...@freebsd.org -- You are receiving this mail because: You are the assignee for the bug. ___ freebsd-usb@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-usb To unsubscribe, send any mail to "freebsd-usb-unsubscr...@freebsd.org"
[Bug 226968] IRQ storm on cpu0 timer when holding down key on USB keyboard
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=226968 Bug ID: 226968 Summary: IRQ storm on cpu0 timer when holding down key on USB keyboard Product: Base System Version: CURRENT Hardware: Any OS: Any Status: New Severity: Affects Some People Priority: --- Component: usb Assignee: freebsd-usb@FreeBSD.org Reporter: johal...@gmail.com I noticed this when playing a first person shooter with keyboard controls. To verify, do the following: Open a terminal window (I used xfce4-terminal) Run $ systat -vmstat On the USB keyboard, hold down some key like 'a'. All is normal at this stage. Without releasing 'a', press and release another key, like 'b'. See the irqs on cpu0:timer go up to several hundred thousand. Release 'a' and it will go back to normal. USB debug information (hw.usb.xhci.debug=16) does not show abnormal amounts of interrupts. With this bug, playing a fps-like game with keyboard controls is impossible since the whole game lags during irq storm. -- You are receiving this mail because: You are the assignee for the bug. ___ freebsd-usb@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-usb To unsubscribe, send any mail to "freebsd-usb-unsubscr...@freebsd.org"
[Bug 226968] IRQ storm on cpu0 timer when holding down key on USB keyboard
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=226968 --- Comment #1 from Johannes Lundberg --- I should mention that I'm using a EVDEV enabled kernel. I have not tried this on a non-EVDEV kernel. -- You are receiving this mail because: You are the assignee for the bug. ___ freebsd-usb@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-usb To unsubscribe, send any mail to "freebsd-usb-unsubscr...@freebsd.org"
[Bug 226968] IRQ storm on cpu0 timer when holding down key on USB keyboard
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=226968 --- Comment #2 from Johannes Lundberg --- Also, the laptop's built-in PS/2 keyboard does not show this behavior, only the external USB keyboard (tried two different keyboards). -- You are receiving this mail because: You are the assignee for the bug. ___ freebsd-usb@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-usb To unsubscribe, send any mail to "freebsd-usb-unsubscr...@freebsd.org"
[Bug 226968] IRQ storm on cpu0 timer when holding down key on USB keyboard
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=226968
--- Comment #3 from Hans Petter Selasky ---
Hi,
I have been discussing some patches with Bruce Evans on this topic. The problem
is the keyboard repeat rate is zero, which cause the timer to trigger
instantly. This issue fell off my radar. Here is the last patch WIP:
Index: sys/arm/samsung/exynos/chrome_kb.c
===
--- sys/arm/samsung/exynos/chrome_kb.c
+++ sys/arm/samsung/exynos/chrome_kb.c
@@ -514,17 +514,8 @@
if (!KBD_HAS_DEVICE(kbd)) {
return (0);
}
-if (((int *)arg)[1] < 0) {
-return (EINVAL);
-}
-if (((int *)arg)[0] < 0) {
-return (EINVAL);
-}
-if (((int *)arg)[0] < 200)/* fastest possible value */
-kbd->kb_delay1 = 200;
-else
-kbd->kb_delay1 = ((int *)arg)[0];
-kbd->kb_delay2 = ((int *)arg)[1];
+kbd->kb_delay1 = KEYBOARD_REPEAT_GET(arg,0,200,1000);
+kbd->kb_delay2 = KEYBOARD_REPEAT_GET(arg,1,34,504);
return (0);
case KDSETRAD:/* set keyboard repeat rate (old
Index: sys/arm/versatile/pl050.c
===
--- sys/arm/versatile/pl050.c
+++ sys/arm/versatile/pl050.c
@@ -418,17 +418,8 @@
if (!KBD_HAS_DEVICE(kbd)) {
return (0);
}
-if (((int *)arg)[1] < 0) {
-return (EINVAL);
-}
-if (((int *)arg)[0] < 0) {
-return (EINVAL);
-}
-if (((int *)arg)[0] < 200)/* fastest possible value */
-kbd->kb_delay1 = 200;
-else
-kbd->kb_delay1 = ((int *)arg)[0];
-kbd->kb_delay2 = ((int *)arg)[1];
+kbd->kb_delay1 = KEYBOARD_REPEAT_GET(arg,0,200,1000);
+kbd->kb_delay2 = KEYBOARD_REPEAT_GET(arg,1,34,504);
return (0);
#if defined(COMPAT_FREEBSD6) || defined(COMPAT_FREEBSD5) || \
Index: sys/dev/adb/adb_kbd.c
===
--- sys/dev/adb/adb_kbd.c
+++ sys/dev/adb/adb_kbd.c
@@ -780,16 +780,8 @@
case KDSETREPEAT:
if (!KBD_HAS_DEVICE(kbd))
return 0;
-if (((int *)data)[1] < 0)
-return EINVAL;
-if (((int *)data)[0] < 0)
-return EINVAL;
-else if (((int *)data)[0] == 0) /* fastest possible value */
-kbd->kb_delay1 = 200;
-else
-kbd->kb_delay1 = ((int *)data)[0];
-kbd->kb_delay2 = ((int *)data)[1];
-
+kbd->kb_delay1 = KEYBOARD_REPEAT_GET(data,0,200,1000);
+kbd->kb_delay2 = KEYBOARD_REPEAT_GET(data,1,34,504);
break;
case KDSETRAD:
Index: sys/dev/gpio/gpiokeys.c
===
--- sys/dev/gpio/gpiokeys.c
+++ sys/dev/gpio/gpiokeys.c
@@ -822,17 +822,8 @@
if (!KBD_HAS_DEVICE(kbd)) {
return (0);
}
-if (((int *)arg)[1] < 0) {
-return (EINVAL);
-}
-if (((int *)arg)[0] < 0) {
-return (EINVAL);
-}
-if (((int *)arg)[0] < 200)/* fastest possible value */
-kbd->kb_delay1 = 200;
-else
-kbd->kb_delay1 = ((int *)arg)[0];
-kbd->kb_delay2 = ((int *)arg)[1];
+kbd->kb_delay1 = KEYBOARD_REPEAT_GET(arg,0,200,1000);
+kbd->kb_delay2 = KEYBOARD_REPEAT_GET(arg,1,34,504);
return (0);
#if defined(COMPAT_FREEBSD6) || defined(COMPAT_FREEBSD5) || \
Index: sys/dev/usb/input/ukbd.c
===
--- sys/dev/usb/input/ukbd.c
+++ sys/dev/usb/input/ukbd.c
@@ -1946,14 +1946,8 @@
if (!KBD_HAS_DEVICE(kbd)) {
return (0);
}
-/*
- * Convert negative, zero and tiny args to the same limits
- * as atkbd. We could support delays of 1 msec, but
- * anything much shorter than the shortest atkbd value
- * of 250.34 is almost unusable as well as incompatible.
- */
-kbd->kb_delay1 = imax(((int *)arg)[0], 250);
-kbd->kb_delay2 = imax(((int *)arg)[1], 34);
+kbd->kb_delay1 = KEYBOARD_REPEAT_GET(arg,0,200,1000);
+kbd->kb_delay2 = KEYBOARD_REPEAT_GET(arg,1,34,504);
#ifdef EVDEV_SUPPORT
if (sc->sc_evdev != NULL)
evdev_push_repeats(sc->sc_evdev, kbd);
Index: sys/sys/kbio.h
===
--- sys/sys/kbio.h
+++ sys/sys/kbio.h
@@ -87,6 +87,15 @@
intkb_repeat[2];
};
typedef struct keyboard_repeat keyboard_repeat_t;
+
+#defineKEYBOARD_REPEAT_GET(p,i,min,max) ({\
+const struct keyboard_repeat *__ptr =\
+(const struct keyboard_repeat *)(p);\
+__ptr->kb_repeat[i] < (min) ? (int)(min) :\
+__ptr->kb_repeat[i] > (max) ? (int)(max) :\
+__ptr->kb_repeat[i];\
+})
+
#define
[Bug 226968] IRQ storm on cpu0 timer when holding down key on USB keyboard
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=226968 --- Comment #4 from Johannes Lundberg --- Hi I tried a variant of this patch a few weeks ago and it did help a bit to reduce interrupts but the problem described in this report still remains, even with the patch you provided. -- You are receiving this mail because: You are the assignee for the bug. ___ freebsd-usb@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-usb To unsubscribe, send any mail to "freebsd-usb-unsubscr...@freebsd.org"
[Bug 226968] IRQ storm on cpu0 timer when holding down key on USB keyboard
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=226968 --- Comment #5 from Hans Petter Selasky --- Can you try to identify the timer causing this? --HPS -- You are receiving this mail because: You are the assignee for the bug. ___ freebsd-usb@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-usb To unsubscribe, send any mail to "freebsd-usb-unsubscr...@freebsd.org"
[Bug 225794] VM images for 12.0-CURRENT have problem with USB 3.0 ports
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=225794 --- Comment #38 from Andriy Gapon --- (In reply to David.Boyd49 from comment #37) That's interesting and puzzling. When you rebooted to multi-user, did you still have the patched kernel? If so, then maybe when you booted the hardware (or VirtualBox) was already wedged? Because single-user vs multi-user mode should not have any effect on CAM and USB. So, I'd like to double-check with you if the patch helps. And, if yes, could you please repeat the single user mode test again? -- You are receiving this mail because: You are on the CC list for the bug. ___ freebsd-usb@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-usb To unsubscribe, send any mail to "freebsd-usb-unsubscr...@freebsd.org"
[Bug 226968] IRQ storm on cpu0 timer when holding down key on USB keyboard
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=226968
--- Comment #6 from Johannes Lundberg ---
Here output from
# dtrace -n 'profile-997hz /arg0/ { @[func(arg0)]=count(); }'
run for about 10 seconds for each state
First idle state
-- SNIP --
linux_common.ko`linux_kernver 1
linux.ko`linux_gettimeofday 1
linux.ko`linux_sendsig1
kernel`ithread_loop 2
kernel`witness_checkorder 2
kernel`witness_warn 2
linux.ko`0x820c2f60 2
kernel`_isitmyx 3
kernel`witness_unlock 4
kernel`vt_fb_bitblt_bitmap 11
kernel`spinlock_exit 11
kernel`cpu_idle 37
kernel`acpi_cpu_idle 26778
And then with cpu0:timer firing at 800k/s
-- SNIP --
kernel`pagezero 1
linux.ko`0x820c2f60 1
kernel`vt_fb_bitblt_text 2
kernel`vt_fb_bitblt_bitmap 11
kernel`lock_mtx 16
kernel`lock_delay21
kernel`unlock_mtx31
kernel`cpu_idle 55
kernel`callout_when 57
kernel`witness_lock 77
kernel`callout_reset_sbt_on 85
kernel`spinlock_enter99
kernel`__mtx_lock_spin_flags103
kernel`softclock_call_cc118
kernel`ukbd_timeout 119
kernel`witness_checkorder 150
kernel`__mtx_assert 155
kernel`usbd_in_polling_mode 167
kernel`__mtx_lock_flags 179
kernel`__mtx_unlock_flags 180
kernel`ukbd_interrupt 298
kernel`memcpy 313
kernel`witness_unlock 318
kernel`spinlock_exit 11745
kernel`acpi_cpu_idle 45519
And some procstat -ak during irq storm
root@jd2:/usr/home/johannes # procstat -ak|grep 100023
12 100023 intrswi4: clock (0) mi_switch ithread_loop
fork_exit fork_trampoline
root@jd2:/usr/home/johannes # procstat -ak | grep 100023
12 100023 intrswi4: clock (0)
root@jd2:/usr/home/johannes # procstat -ak | grep 100023
12 100023 intrswi4: clock (0) callout_reset_sbt_on
ukbd_timeout softclock_call_cc softclock intr_event_execute_handlers
ithread_loop fork_exit fork_trampoline
root@jd2:/usr/home/johannes # procstat -ak | grep 100023
12 100023 intrswi4: clock (0)
root@jd2:/usr/home/johannes # procstat -ak | grep 100023
12 100023 intrswi4: clock (0) ukbd_timeout
softclock_call_cc softclock intr_event_execute_handlers ithread_loop fork_exit
fork_trampoline
root@jd2:/usr/home/johannes # procstat -ak | grep 100023
12 100023 intrswi4: clock (0)
root@jd2:/usr/home/johannes # procstat -ak | grep 100023
12 100023 intrswi4: clock (0) ukbd_interrupt
ukbd_timeout softclock_call_cc softclock intr_event_execute_handlers
ithread_loop fork_exit fork_trampoline
root@jd2:/usr/home/johannes # procstat -ak | grep 100023
12 100023 intrswi4: clock (0) callout_reset_sbt_on
ukbd_timeout softclock_call_cc softclock intr_event_execute_handlers
ithread_loop fork_exit fork_trampoline
root@jd2:/usr/home/johannes # procstat -ak | grep 100023
12 100023 intrswi4: clock (0) callout_reset_sbt_on
ukbd_timeout softclock_call_cc softclock intr_event_execute_handlers
ithread_loop fork_exit fork_trampoline
root@jd2:/usr/home/johannes # procstat -ak | grep 100023
12 100023 intrswi4: clock (0)
root@jd2:/usr/home/johannes # procstat -ak | grep 100023
12 100023 intrswi4: clock (0) callout_reset_sbt_on
ukbd_timeout
[Bug 225794] VM images for 12.0-CURRENT have problem with USB 3.0 ports
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=225794 --- Comment #39 from david.boy...@twc.com --- Andriy, Yes, the patch disabling dareprobe(s) was/is applied to the kernel in both single-user and multi-user mode. In single-user mode the error condition in the dmesg output only appears if the device is attached after the "camcontrol debug -c -p -I -P all" statement is issued. Just booting into single-user mode, but without issuing the debug command, allows access to the USB flash drive (gpart show da0, mount/unmount, fsck, etc.). As stated, with the patch disabling dareprobe(s), booting into multi-user mode allows access as above. -- You are receiving this mail because: You are on the CC list for the bug. ___ freebsd-usb@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-usb To unsubscribe, send any mail to "freebsd-usb-unsubscr...@freebsd.org"
[Bug 225794] VM images for 12.0-CURRENT have problem with USB 3.0 ports
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=225794 Warner Losh changed: What|Removed |Added CC||i...@freebsd.org --- Comment #40 from Warner Losh --- What revision of FreeBSD were the latest tests done by? I've fixed some locking bugs ending with r331435... dareprobe suggests this may help, but I'm unsure. -- You are receiving this mail because: You are on the CC list for the bug. ___ freebsd-usb@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-usb To unsubscribe, send any mail to "freebsd-usb-unsubscr...@freebsd.org"
[Bug 226968] IRQ storm on cpu0 timer when holding down key on USB keyboard
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=226968 --- Comment #7 from Hans Petter Selasky --- Can you try to trace the "sbt" argument passed to: kernel`callout_reset_sbt_on 85 ? It might be some invalid timeout. --HPS -- You are receiving this mail because: You are the assignee for the bug. ___ freebsd-usb@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-usb To unsubscribe, send any mail to "freebsd-usb-unsubscr...@freebsd.org"
