Wireshark-bugs: [Wireshark-bugs] [Bug 9630] USB Mass Storage dissector doesn't support the RealT
Tyson Key
changed
bug 9630
| What |
Removed |
Added |
| CC |
|
tyson.key@gmail.com
|
Comment # 2
on bug 9630
from Tyson Key
Thanks for offering to investigate this.
To be honest, although the chipset maker claims that the device conforms to the
USB Mass Storage Class specs, it seems that they're doing just enough
non-standard stuff to be awkward...
I haven't seen any standalone devices with these chips "in the wild", either.
(Mine is integrated into a Lenovo G505 laptop).
I haven't gotten around to generating traces under Linux yet, since the USBPCap
filter driver seems to be conflicting with VirtualBox's USB filter driver (so I
can't connect devices to virtual machines), unfortunately. (And I'd have to
find, and burn a Linux distribution ISO to a DVD, and hope that it boots on
this hardware).
I can try to supply some more USBPCap traces, if they're of use, though.
Tyson.
(In reply to comment #1)
> Ok, I can handle it. Wireshark is able to dissect payload per device or
> class/subclass/protocol or VendorId/ProductID. However existing dissectors
> (usbaudio, usbcom, usbhid, usbhub, usbvideo, usbms) must be updated to use
> those tables. I have a plan to do that, so let start with "usbms". If I
> understand your device is standard usbms? (+ something new)? If payload will
> be "Decode As" usbms, than you will have correctly dissected packets?
> Otherwise it need new device-specific dissector.
>
> I see that USB payload in your trace file is not correctly dissected. For
> example in packet 1 is correct "GET DESCRIPTOR Request DEVICE", so in frame
> 2 should be "GET DESCRIPTOR Response DEVICE" instead of "URB_CONTROL in". I
> will check that.
>
> By the way: do you have trace file while using this device under Linux? It
> might be helpful.
You are receiving this mail because:
- You are watching all bug changes.
