https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=235257
Christos Margiolis changed:
What|Removed |Added
CC||chris...@freebsd.org
---
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=235257
--- Comment #29 from Max ---
Hello,
FYI: The symptoms of the issue still keep occurring on the latest
*FreeBSD-13.0-CURRENT*
Could you please let me know whether you found something on your side?
Thanks
Max
--
You are receiving this
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=235257
--- Comment #28 from Max ---
Hi
devd seems to be the only process getting to sys/kern/subr_bus.c:devread() and
it is pretty active there.
>>>grep devread /var/log/messages |grep -oe 'PID [0-9]*, comm
>>>[a-zA-Z0-9]*'|sort |uniq
>>>PID
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=235257
--- Comment #27 from Hans Petter Selasky ---
Try to put the print inside this function instead.
static int
devread(struct cdev *dev, struct uio *uio, int ioflag)
{
--HPS
--
You are receiving this mail because:
You are the assignee for
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=235257
--- Comment #26 from Max ---
The first one who is getting to devfs_read_f is openrc.
devd under openrc seems to use the following configuration which I previously
corrected:
>>> sbin/devd/devmatch-openrc.conf
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=235257
--- Comment #25 from Hans Petter Selasky ---
In the devfs read routine for the events, try to print who is reading the
events:
printf("PID=%d %s\n", curthread->td_proc->p_pid, curthread->td_proc->p_comm);
Maybe some program is draining
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=235257
--- Comment #24 from Max ---
I am able to trace event through from sys/dev/usb/usb_hub.c:usb_notify_addq to
sys/kern/subrc_bus.c:devctl_queue_data_f where it is supposed to invoke
pgsigio() but still I don't see devd triggering module
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=235257
--- Comment #23 from Hans Petter Selasky ---
There are a couple of functions in the kernel, where you might want to add a
print:
sys/kern/subr_bus.c:devctl_queue_data(char *data)
sys/kern/subr_bus.c:devctl_queue_data_f(data,
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=235257
--- Comment #22 from Max ---
How can I verify nomatch events are really queued by USB hub controller in the
kernel for devd?
I see devd registers descriptor for /dev/devctl with select() in
devd.cc:event_loop()...
Thanks,
Max
--
You
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=235257
--- Comment #21 from Max ---
Hi
So any ideas why nomatch evens are not picked up by devd in the daemon mode
after the system starts up?
Even if I restart devd as daemon instead of in foreground and detach /
attach device, there is no
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=235257
--- Comment #20 from Hans Petter Selasky ---
Hi,
The kernel will buffer all nomatch events until devd is loaded or the first
client starts reading them.
--HPS
--
You are receiving this mail because:
You are the assignee for the bug.
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=235257
--- Comment #19 from Max ---
Hello
Still on startup module is not autoloaded. It only works when devd is started
in foreground and then USB device is attached. On startup according to syslog
XHCI controller detects device attachment
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=235257
--- Comment #18 from Max ---
Hi
The latter is related to OpenRC integration where devd configuration for OpenRC
is:
>>>action "sh /etc/rc.devd devmatch start.'?'$_ '?'$_";
Replacing with
>>> action "/etc/rc.d/devmatch quietstart '?'$_";
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=235257
--- Comment #17 from Max ---
And I see why. devmatch binary is not in the execution when launched by rc.
>>>Executing 'sh /etc/rc.devd devmatch start.'?'$' at bus=0 hubaddr=3 port=4
>>>devaddr=9 interface=0 >>>ugen=ugen0.9
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=235257
--- Comment #16 from Max ---
Although on startup I see devd service is starting but devmatch nomatch action
is not getting triggered...
Thanks,
Max
--
You are receiving this mail because:
You are the assignee for the bug.
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=235257
--- Comment #15 from Max ---
Hi
1. It seems /sbin/service does not recognize 'quietstart' option
>>>sudo /sbin/service devmatch quietstart
>>> * devmatch: unknown function `quietstart'
Are there any plans to patch /sbin/service?
2. I
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=235257
--- Comment #14 from Warner Losh ---
As we discovered on IRC, this is caused by an out of date /usr/sbin/service
file, I think. The patch, if it works, confirms that's the case.
--
You are receiving this mail because:
You are the
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=235257
--- Comment #13 from Max ---
Hi
Here is output of devd upon attaching uaudio device:
>>>setting *=+uaudio1 at bus=0 hubaddr=4 port=1 devaddr=13 interface=0
>>>ugen=ugen0.13 vendor=0x041e >>>product=0x323c devclass=0x00 devsubclass=0x00
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=235257
--- Comment #12 from Hans Petter Selasky ---
You need to specify arguments for devmatch. Devd will print these arguments.
Can you try to kill devd. Run it in the foreground and attach your USB audio
device. Also make sure devd is
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=235257
--- Comment #11 from Max ---
Hi
This worked:
>>>sudo /etc/rc.d/devmatch start '? at bus=0 hubaddr=2 port=4 devaddr=4
>>>interface=0 ugen=ugen0.4 >>>vendor=0x046d product=0xc069 devclass=0x00
>>>devsubclass=0x00 devproto=0x00 sernum=""
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=235257
--- Comment #10 from Hans Petter Selasky ---
Created attachment 201550
--> https://bugs.freebsd.org/bugzilla/attachment.cgi?id=201550=edit
Fix for (/etc/devd/devmatch.conf)
Can you test this patch?
--
You are receiving this mail
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=235257
--- Comment #9 from Hans Petter Selasky ---
I found it:
Try these two commands:
service devmatch quietstart '? at bus=0 hubaddr=2 port=4 devaddr=4 interface=0
ugen=ugen0.4 vendor=0x046d product=0xc069 devclass=0x00 devsubclass=0x00
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=235257
Hans Petter Selasky changed:
What|Removed |Added
Status|Open|In Progress
--
You are
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=235257
--- Comment #8 from Hans Petter Selasky ---
I guess I need to update my system to reproduce this issue. I'm running older
binaries which don't exhibit this problem yet.
Warner: Any input on this issue?
--HPS
--
You are receiving this
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=235257
--- Comment #7 from Max ---
Yes I do
>>>[~]$ tail -1 /etc/rc.conf
>>>devmatch_enable="YES"
The /etc/devd/devmatch.conf is also in-place:
>>>[~]$ cat /etc/devd/devmatch.conf| head -13
>>#
>>># $FreeBSD$
>>>#
>>
>>>#
>>># Example devd
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=235257
--- Comment #6 from Hans Petter Selasky ---
And you have devmatch_enable="YES" in /etc/rc.conf ?
--HPS
--
You are receiving this mail because:
You are the assignee for the bug.
___
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=235257
--- Comment #5 from Max ---
Hi
I ran the following and it didn't any of the error conditions when reading
linker.hints
>>>devmatch -a -h /boot/kernel/linker.hints
This was the resulting output:
>>> read_hints() sb.st_size 323176
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=235257
--- Comment #4 from Hans Petter Selasky ---
Hi,
The USB descriptors are all OK and should match USB audio.
Can you invoke devmatch manually:
devmatch -a -h /boot/kernel/linker.hints
Typical errors:
1) Due to "struct stat" changes:
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=235257
Max changed:
What|Removed |Added
CC||maxstec...@gmail.com
--- Comment #3 from
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=235257
--- Comment #2 from Hans Petter Selasky ---
The output from "usbconfig -d X.Y dump_device_desc dump_curr_config_desc" would
also be useful.
--
You are receiving this mail because:
You are the assignee for the bug.
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=235257
--- Comment #1 from Hans Petter Selasky ---
> Could you please help me understand how devmatch is supposed to work for such
> cases? Do I need to configure anything else?
The mask field should select which fields are target for
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=235257
Kubilay Kocak changed:
What|Removed |Added
Status|New |Open
Flags|
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=235257
Bug ID: 235257
Summary: snd_uaudio not getting loaded automatically after
migration to devmatch
Product: Base System
Version: CURRENT
Hardware: amd64
33 matches
Mail list logo