On Thu, 12 Aug 2004, Norbert Preining wrote:
> BIngo, you are a genius! I found what is causing me the problems: Only
> when a specific device is plugged in, the hangs occur (in kernels after
> 2.6.7-mm5).
>
> It is a Benq radio mouse with emtpy device descriptor:
> T: Bus=04 Lev=01 Prnt=01 Port=00 Cnt=01 Dev#= 3 Spd=1.5 MxCh= 0
> D: Ver= 1.00 Cls=00(>ifc ) Sub=00 Prot=00 MxPS= 8 #Cfgs= 1
> P: Vendor=0d62 ProdID=1000 Rev= 2.00
> C:* #Ifs= 1 Cfg#= 1 Atr=a0 MxPwr=100mA
> I: If#= 0 Alt= 0 #EPs= 1 Cls=03(HID ) Sub=01 Prot=02 Driver=usbhid
> E: Ad=81(I) Atr=03(Int.) MxPS= 8 Ivl=10ms
>
>
> In all cases, with 2.6.7-mm5 and any later I get the following syslog
> messages when I plug in the mouse or it is plugged when usb sys is
> starting:
> vmunix: drivers/usb/input/hid-core.c: ctrl urb status -2 received
> vmunix: drivers/usb/input/hid-core.c: timeout initializing reports
> vmunix:
> kernel: input: USB HID v1.10 Mouse [0d62:1000] on usb-0000:02:09.0-1
Do you mean with 2.6.7-mm5 and any _earlier_? With kernels after that
your trace below shows khubd hanging in hid_init_reports before the last
two messages are written.
> For the newer kernels the process khubd in fact is in D state:
> khubd D 00000000 0 1199 1 1477 348 (L-TLB)
> df965d04 00000046 c0c5f400 00000000 1d800000 00800000 df5501fc df55020c
> 00001726 9ea049b3 000000f4 df8ee7d0 df8ee978 0008de11 df965d18 00002710
> df965d64 c02da922 c0d2fc00 e08d38fd 00000282 c0388ea4 c040d398 0008de11
> Call Trace:
> [<c02da922>] schedule_timeout+0x60/0xb3
> [<e08d38fd>] unlink1+0x11/0x18 [usbcore]
> [<c01232a1>] process_timeout+0x0/0x5
> [<e08c8b28>] hid_wait_io+0x88/0xcc [usbhid]
> [<c0118fd7>] default_wake_function+0x0/0xc
> [<c0118fd7>] default_wake_function+0x0/0xc
> [<e08c8d08>] hid_init_reports+0xd5/0x17c [usbhid]
> I hope this helps you ...
>
> Bottom line for this, I would like to use this mouse, but with the
> new kernels it isn't possible.
>
> ANother interesting point is the following, but I think this is just
> a consequence of the above: When I do NOT have plugged in the mouse
> while booting, the Synaptics TOuchpad is recognized, while when it is plugged
> in during boot, it is not recognized ... strange, at least to me!
This probably _is_ a consequence of the problem with the mouse. Once
khubd hangs, no new devices will be recognized.
Below is a test patch for you to try. It will provide a lot of extra
information in you system log, which may help to pin down where the
problem occurs.
Alan Stern
===== drivers/usb/core/hcd.c 1.152 vs edited =====
--- 1.152/drivers/usb/core/hcd.c Tue Aug 3 10:17:59 2004
+++ edited/drivers/usb/core/hcd.c Thu Aug 12 12:17:05 2004
@@ -1146,6 +1146,7 @@
wake_up (&usb_kill_urb_queue);
usb_put_urb (urb);
}
+ printk(KERN_INFO "%s: urb=%p\n", __FUNCTION__, urb);
return status;
}
@@ -1185,6 +1186,7 @@
if (value != 0)
dev_dbg (hcd->self.controller, "dequeue %p --> %d\n",
urb, value);
+ printk(KERN_INFO "%s: urb=%p\n", __FUNCTION__, urb);
return value;
}
@@ -1506,6 +1508,7 @@
}
/* pass ownership to the completion handler */
+ printk(KERN_INFO "%s: urb=%p\n", __FUNCTION__, urb);
urb->complete (urb, regs);
atomic_dec (&urb->use_count);
if (unlikely (urb->reject))
===== drivers/usb/input/hid-core.c 1.119 vs edited =====
--- 1.119/drivers/usb/input/hid-core.c Fri Jul 23 09:29:18 2004
+++ edited/drivers/usb/input/hid-core.c Thu Aug 12 12:13:02 2004
@@ -25,7 +25,7 @@
#include <asm/byteorder.h>
#include <linux/input.h>
-#undef DEBUG
+#define DEBUG
#undef DEBUG_DATA
#include <linux/usb.h>
@@ -1070,7 +1070,7 @@
hid->urbout->transfer_buffer_length = ((report->size - 1) >> 3) + 1 +
(report->id > 0);
hid->urbout->dev = hid->dev;
- dbg("submitting out urb");
+ dbg("submitting out urb=%p", hid->urbout);
if (usb_submit_urb(hid->urbout, GFP_ATOMIC)) {
err("usb_submit_urb(out) failed");
@@ -1116,7 +1116,7 @@
hid->cr->wIndex = cpu_to_le16(hid->ifnum);
hid->cr->wLength = cpu_to_le16(len);
- dbg("submitting ctrl urb: %s wValue=0x%04x wIndex=0x%04x wLength=%u",
+ dbg("submitting ctrl urb=%p: %s wValue=0x%04x wIndex=0x%04x wLength=%u",
hid->urbctrl,
hid->cr->bRequest == HID_REQ_SET_REPORT ? "Set_Report" : "Get_Report",
hid->cr->wValue, hid->cr->wIndex, hid->cr->wLength);
@@ -1137,6 +1137,7 @@
struct hid_device *hid = urb->context;
unsigned long flags;
+ dbg("complete output urb=%p\n", urb);
if (urb->status)
warn("output irq status %d received", urb->status);
@@ -1166,6 +1167,7 @@
struct hid_device *hid = urb->context;
unsigned long flags;
+ dbg("complete ctrl urb=%p\n", urb);
if (urb->status)
warn("ctrl urb status %d received", urb->status);
@@ -1334,9 +1336,13 @@
while (ret) {
err |= ret;
if (test_bit(HID_CTRL_RUNNING, &hid->iofl))
+{dbg("unlink ctrl urb=%p\n", hid->urbctrl);
usb_unlink_urb(hid->urbctrl);
+}
if (test_bit(HID_OUT_RUNNING, &hid->iofl))
+{dbg("unlink out urb=%p\n", hid->urbout);
usb_unlink_urb(hid->urbout);
+}
ret = hid_wait_io(hid);
}
-------------------------------------------------------
SF.Net email is sponsored by Shop4tech.com-Lowest price on Blank Media
100pk Sonic DVD-R 4x for only $29 -100pk Sonic DVD+R for only $33
Save 50% off Retail on Ink & Toner - Free Shipping and Free Gift.
http://www.shop4tech.com/z/Inkjet_Cartridges/9_108_r285
_______________________________________________
[EMAIL PROTECTED]
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
