headers
Del Merritt | 2 Aug 2006 22:40
Picon
Favicon

Re: V1_0_DEVEL: libusb.so: undefined reference to `usbi_parse_device_descriptor'

Tim Roberts wrote on 5/01/2006:
> Timo Juhani Lindfors wrote:
>
>   
>> Hi,
>>
>> I tried to build V1_0_DEVEL brach from cvs to port my scripts to use
>> the new API and isochronous transfers. However, the build failed with
>> the following messages on a debian sarge system. Please let me know if
>> you can't reproduce the problem so that I can provide more info.
>>  
>>
>>     
>
> The cvs archive is no longer being maintained.  The libusb source is now
> in svn.
>
> However, when I did my build a month ago, this problem also existed
> there.  It looked like there was a partial checkin;
> usbi_parse_device_descriptor was supposed to be moved out of its old
> location into linux.c, but it didn't get added.  I just copied it from
> its old location into linux.c, and all was well.
>
> I filed a bug report on it, but I haven't followed up; it may have been
> fixed by this time.
>   

Would you mind telling me the bug number?  I don't seem to have "its old 
location" in my newly-populated svn code.  At what revision did it 
move/disappear?  I also don't see this rectified in Chris Frey's
(Continue reading)

Permalink | Reply | 3 comments |
headers
Tim Roberts | 2 Aug 2006 23:31

Re: V1_0_DEVEL: libusb.so: undefined reference to `usbi_parse_device_descriptor'

Del Merritt wrote:

> Tim Roberts wrote on 5/01/2006:
>
>> Timo Juhani Lindfors wrote:
>>  
>>
>>> Hi,
>>>
>>> I tried to build V1_0_DEVEL brach from cvs to port my scripts to use
>>> the new API and isochronous transfers. However, the build failed with
>>> the following messages on a debian sarge system. Please let me know if
>>> you can't reproduce the problem so that I can provide more info.  
>>
>>
>> The cvs archive is no longer being maintained.  The libusb source is now
>> in svn.
>>
>> However, when I did my build a month ago, this problem also existed
>> there.  It looked like there was a partial checkin;
>> usbi_parse_device_descriptor was supposed to be moved out of its old
>> location into linux.c, but it didn't get added.  I just copied it from
>> its old location into linux.c, and all was well.
>>
>> I filed a bug report on it, but I haven't followed up; it may have been
>> fixed by this time.
>>   
>
>
> Would you mind telling me the bug number?  I don't seem to have "its
(Continue reading)

Permalink | Reply | 2 comments |
headers
Del Merritt | 3 Aug 2006 17:34
Picon
Favicon

Re: V1_0_DEVEL: libusb.so: undefined reference to `usbi_parse_device_descriptor'

Tim Roberts wrote:
> Del Merritt wrote:
>
>   
>> Tim Roberts wrote on 5/01/2006:
>>
>>     
>>> Timo Juhani Lindfors wrote:
>>>  
>>>
>>>       
>>>> Hi,
>>>>
>>>> I tried to build V1_0_DEVEL brach from cvs to port my scripts to use
>>>> the new API and isochronous transfers. However, the build failed with
>>>> the following messages on a debian sarge system. Please let me know if
>>>> you can't reproduce the problem so that I can provide more info.  
>>>>         
>>> The cvs archive is no longer being maintained.  The libusb source is now
>>> in svn.
>>>
>>> However, when I did my build a month ago, this problem also existed
>>> there.  It looked like there was a partial checkin;
>>> usbi_parse_device_descriptor was supposed to be moved out of its old
>>> location into linux.c, but it didn't get added.  I just copied it from
>>> its old location into linux.c, and all was well.
>>>
>>> I filed a bug report on it, but I haven't followed up; it may have been
>>> fixed by this time.
>>>
(Continue reading)

Permalink | Reply | 1 comment |
headers
Johannes Erdfelt | 3 Aug 2006 19:30

Re: V1_0_DEVEL: libusb.so: undefined reference to `usbi_parse_device_descriptor'

On Thu, Aug 03, 2006, Del Merritt <del <at> alum.mit.edu> wrote:
> Johannes may be it for "active" developers, and he's been quiet on this 
> list since the beginning of May.

I'd say Chris Frey has been more active than I am lately.

Chris, give me your SourceForge username and I'll add you to the list.

> There are six developers listed on SF, and I presume they all have write
> access:
> 
>     Brad Hards
>     Richard Tobin
>     Nathan Hjelm
>     Hans Ulrich Niedermann
>     Johannes Erdfelt
>     John Reynolds
> 
> See the list yourself at:
> 
>     http://sourceforge.net/project/memberlist.php?group_id=1674
> 
> Johannes is himself the only project admin, and thus the only person who 
> can bless new developers.
> 
> Johannes, do you have your ears on? What are your project criteria for 
> grating write access to this project? And, like, how's it going? We had 
> some email traffic some years back when I was looking in to dents...

I don't completely ignore the list, but it feels like I have effectively
(Continue reading)

Permalink | Reply | 0 comments |
headers
Konrad.Jankowski | 3 Aug 2006 19:59

[PATCH] FreeBSD fix for usb_control_msg()

I don't know how Net- or Open- handle this, but
FreeBSD returns the number of transmited bytes in
req.ucr_actlen, not in req.ucr_request.wLength.
(/usr/include/dev/usb/usb.h:544:
        int     ucr_actlen;             /* actual length transferred */

So then:
--- bsd.c       Sat Mar  4 03:52:46 2006
+++ ../../bsd.c Thu Aug  3 19:19:15 2006
 <at>  <at>  -476,8 +476,11  <at>  <at> 
   if (ret < 0)
     USB_ERROR_STR(-errno, "error sending control message: %s",
                   strerror(errno));
-
+#ifdef __FreeBSD_kernel__
+  return req.ucr_actlen;
+#else
   return UGETW(req.ucr_request.wLength);
+#endif
 }

 int usb_os_find_busses(struct usb_bus **busses)

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys -- and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
Permalink | Reply | 2 comments |
headers
Rene Ladan | 4 Aug 2006 09:39
Picon

Re: [PATCH] FreeBSD fix for usb_control_msg()

Konrad.Jankowski <at> master.ether.net schreef:
> I don't know how Net- or Open- handle this, but
> FreeBSD returns the number of transmited bytes in
> req.ucr_actlen, not in req.ucr_request.wLength.
> (/usr/include/dev/usb/usb.h:544:
>         int     ucr_actlen;             /* actual length transferred */
> 
> So then:
> --- bsd.c       Sat Mar  4 03:52:46 2006
> +++ ../../bsd.c Thu Aug  3 19:19:15 2006
>  <at>  <at>  -476,8 +476,11  <at>  <at> 
>    if (ret < 0)
>      USB_ERROR_STR(-errno, "error sending control message: %s",
>                    strerror(errno));
> -
> +#ifdef __FreeBSD_kernel__
> +  return req.ucr_actlen;
> +#else
>    return UGETW(req.ucr_request.wLength);
> +#endif
>  }
>  
>  int usb_os_find_busses(struct usb_bus **busses)
>
This seems correct, wLength is a uWord defined in usb_device_request_t
(usb/usb.h:112).
usb_device_request_t is included in usb_ctl_request (usb/usb.h:542),
which is the structure in which ucr_actlen (int) is defined (usb/usb.h:546)

I am cc'ing freebsd-usb <at>  since  I'm not sure about this patch...
(Continue reading)

Permalink | Reply | 0 comments |
headers
gully | 5 Aug 2006 06:56
Favicon

LibUsb

Hello Everyone,
I've been trying for quite a while to connect to a (SBIG) CCD camera
device under Red Hat Linux 4, Kernel version 2.6.9.  I was temporarily
able to connect to this device, but got the error message below:

usb_bulk_read() from Com pipe error: -110/4

I ran depmod -a in the /sbin directory, and then my USB device failed to
reconnect stably (it will connect but requires USB recycling after each
successful image acquisition).  Is this a Libusb problem or a
Hotplug/FXLoad problem?  Any suggestions?  Quick fixes?  Re-install with
different parameters?

Any help whatsoever is tremendously appreciated!

Regards,
Michael Gully

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys -- and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
Permalink | Reply | 2 comments |
headers
Peter Stuge | 6 Aug 2006 03:59

Re: LibUsb

On Sat, Aug 05, 2006 at 12:56:00AM -0400, gully <at> bu.edu wrote:
> Hello Everyone,
> I've been trying for quite a while to connect to a (SBIG) CCD camera
> device under Red Hat Linux 4, Kernel version 2.6.9.  I was
> temporarily able to connect to this device, but got the error message
> below:

2.6.9 is broken in many ways. Try upgrading to a recent kernel, e.g.
2.6.17.6.

//Peter

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys -- and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
Permalink | Reply | 0 comments |
headers
KShots | 6 Aug 2006 04:58

Several errors, but partial success?


I'm a little flustered here... Here's what I'm trying to do.

I'm trying to reverse-engineer a couple of models of electronic CD
organizers (KDS CDM-751 and CDS-USA) I have here. I'm bothering because they
do not work on any of the operating systems that the machine they're plugged
into are running (Windows XP x64, gentoo linux). I took one look at libusb
and decided it should do exactly what I need to make this happen... and it's
cross-platform. Perfect.

Using another machine with the software required to use the organizer, I
captured some traffic going across the USB connection with a USB monitor
tool. Using this, I isolated the commands to do an eject / insert on any
given bay. The commands are all done using a control message to the devices.
I had all the required information for these messages... except the
requestType. I'll get to that later.

Now at this point, I've isolated the commands I need to send, and I'm ready
to go... time to learn how to use libusb! First place I look:
/usr/share/docs/libusb - bingo, there's docs there. But wait... they're
horrible! No more than a single line of description for each function, no
explanation for any function parameters, and worst of all, google can't seem
to dig much up either. Ok, fine... I'll do this by trial and error and
looking at the source itself.

Looking down the USB bus, I've managed to isolate the vendor and product IDs
for my organizers. Using this, I use usb_open() to obtain a device... and
assume it worked (I saw nothing to indicate error conditions in the docs for
that function). Next, I call usb_claim_interface on the given handle. This
goes through without complaint.
(Continue reading)

Permalink | Reply | 13 comments |
headers
KShots | 6 Aug 2006 05:10

Unique device identification - possible?


I have multiple USB devices plugged in (same vendor / product ID) that must
be communicated with separately, and need to be remembered separately
regardless of which USB port they're plugged into. Is there anything that
can accomplish this?

Thanks
--

-- 
View this message in context: http://www.nabble.com/Unique-device-identification---possible--tf2057782.html#a5669702
Sent from the LibUSB Dev forum at Nabble.com.

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys -- and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
Permalink | Reply | 5 comments |

Gmane