usb: gadget: workaround storage command size issues
Try to workaround issues with bad SCSI implementations
by ignoring the command size error.
Signed-off-by: Felipe Balbi <felipe.balbi@nokia.com>
Acked-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
diff --git a/drivers/usb/gadget/file_storage.c b/drivers/usb/gadget/file_storage.c
index 0c632d2..e0f616f 100644
--- a/drivers/usb/gadget/file_storage.c
+++ b/drivers/usb/gadget/file_storage.c
@@ -2676,11 +2676,24 @@
/* Verify the length of the command itself */
if (cmnd_size != fsg->cmnd_size) {
- /* Special case workaround: MS-Windows issues REQUEST SENSE
- * with cbw->Length == 12 (it should be 6). */
- if (fsg->cmnd[0] == SC_REQUEST_SENSE && fsg->cmnd_size == 12)
+ /* Special case workaround: There are plenty of buggy SCSI
+ * implementations. Many have issues with cbw->Length
+ * field passing a wrong command size. For those cases we
+ * always try to work around the problem by using the length
+ * sent by the host side provided it is at least as large
+ * as the correct command length.
+ * Examples of such cases would be MS-Windows, which issues
+ * REQUEST SENSE with cbw->Length == 12 where it should
+ * be 6, and xbox360 issuing INQUIRY, TEST UNIT READY and
+ * REQUEST SENSE with cbw->Length == 10 where it should
+ * be 6 as well.
+ */
+ if (cmnd_size <= fsg->cmnd_size) {
+ DBG(fsg, "%s is buggy! Expected length %d "
+ "but we got %d\n", name,
+ cmnd_size, fsg->cmnd_size);
cmnd_size = fsg->cmnd_size;
- else {
+ } else {
fsg->phase_error = 1;
return -EINVAL;
}