G15Daemon

Logged In: YES
user_id=356931

the changes necessary are in libg15 version 1.0.1,
released yesterday. Feel free to re-open this bug
if the problem recurs.

Logged In: YES
user_id=1312539

This Tracker item was closed automatically by the system. It was
previously set to a Pending status, and the original submitter
did not respond within 14 days (the time period specified by
the administrator of this Tracker).

Logged In: YES
user_id=356931

Thanks very much for testing, and everyone for reporting
this issue. I'm glad the patch helped. I'll commit the
changes the patch provides to libg15 svn in the next few
days, and release a new version of libg15 on the g15tools
site shortly after that. I'd like you to test one more
change, if you wouldn't mind, as I'd like to get that error
rate down to zero. I'll post the patch here tomorrow.

Logged In: YES
user_id=822473

I have added the patch in the files dir of libg15 and added :
src_unpack() {
unpack ${A}
cd "${S}"
epatch ${FILESDIR}/libg15-bulk-transfer.diff
}
then ebuilded digest and remerged libg15 :
Oct 15 12:04:34 ar g15daemon[17987]: g15daemon 1.2.1 loaded
Oct 15 12:04:34 ar usb 1-8.1.3.4: usbfs: usb_submit_urb
returned -28
Oct 15 12:04:36 ar usb 1-8.1.3.4: usbfs: usb_submit_urb
returned -28
Oct 15 12:05:37 ar usb 1-8.1.3.4: usbfs: usb_submit_urb
returned -28
Oct 15 12:06:38 ar usb 1-8.1.3.4: usbfs: usb_submit_urb
returned -28
Oct 15 12:07:39 ar usb 1-8.1.3.4: usbfs: usb_submit_urb
returned -28

Which means it's much much better as I only get one error
per minutes which don't use all CPU.

Thank you very much :)

Logged In: YES
user_id=356931

Thanks. The patch is at the bottom of this page -
libg15-bulk-transfer.diff, it should be applied against
libg15.c in libg15. It simply changes the transfer type
from interrupt to bulk, which might reduce the load on the
bus if the keyboard can't process the requests fast enough
for some reason (such as other devices on the bus).

Logged In: YES
user_id=822473

As soon as I understand what patch should be taken were I'll
test with great pleasure :-)

Logged In: YES
user_id=1550003

I just reported the bug here because I took care of it over
at the Gentoo Bugzilla, unfortunately I don't have a G15
myself. Another g15 tester I asked also couldn't get a hub
for him, so I hope gregoire can test the patch and post the
results.

Logged In: YES
user_id=356931

I'm unable to reproduce this unfortunately. Is someone
able to try the attached patch against libg15, and report
whether it improves the situation?

Logged In: YES
user_id=822473

I don't think so : if I plug my g15 to one of my usb port on
my usb monitor I get into trouble, and on the monitor there
is only the keyboard and the cards readers (not in use).
I don't think an unused cards reader need lots of usb
traffic, no ?

Logged In: YES
user_id=356931

This appears to be a bug in either libg15 or libusb. I'll
try and reproduce this here, and fix it in libg15 if
possible. Thanks for bringing it to my attention.

To clarify, it only appears when there is a lot of usb
traffic on the bus, correct?