On Fri, 13 Jun 2003, Benjamin Milde wrote: >Hi, >first of all: I repoduced the illegal byte sequence error with a 64-byte buffer. This >error comes only if I submit the buffer AFTER the output request, >the data comes first after I wait a while and only then I turn OFF the camera because >the program hangs (and I'm to lazy to kill the java process). >So FIRST after this the strange data and error comes in:
Ok...so. I'm assuming you're using a UHCI controller, and that's where the error is coming from. That's the most likely case here. So when you get an EILSEQ error (Illegal Byte Sequence) that really means that your TD (it's a UHCI thing) got a CRC error. Now that's normally not a good thing, because it means you're getting hardware errors. If it's really a hardware error (and chances are it is) then it may be your cable. I've seen too-long cable lengths really destroy the USB protocol - it just completely screws everything up. So check that first. If you're sure the cable's ok, and you're sure the device and host system are ok, then it might be a Linux bug, but I think that isn't likely. An EILSEQ error is generated by the hardware, _not_ the software. The software only looks at the flags that the hardware has set, and acts on those. So I really would be suprised if it was getting caused by Linux. And it's very, very unlikely that anything above the Linux kernel is causing this, because if so, then the kernel should have seen that userspace is passing it something bogus and vetoed it (which it obviously didn't). Another completely different thought is, that it's possible your device is where the problem is, but this only happens because you didn't do something that it was expecting to happen...so if you duplicate what Windoze does, and everything is cool, that's probably the answer. Anyway, I'm 99% sure that javax.usb can't do anything about the EILSEQ error. >So the solution must be somethere else, and I don't exclude a bug either in javaxusb >or in the kernel. >I can try the 2.5 series, I heard there is much done on the usb-task. Well, the 2.4 USB stuff isn't nearly as good as the 2.5 stuff, but I think you should be able to make your device work with 2.4. It isn't that bad. >Another possibility is, that the device really sends nothing out at this time >and I must do something before, like a null-byte packet to the other >bulk-out-endpoint. >Perhaps you see something of that sort in the usblog. I think that is it. See my other note for my suggestions. -- Dan Streetman [EMAIL PROTECTED] --------------------- 186,272 miles per second: It isn't just a good idea, it's the law! ------------------------------------------------------- This SF.NET email is sponsored by: eBay Great deals on office technology -- on eBay now! Click here: http://adfarm.mediaplex.com/ad/ck/711-11697-6916-5 _______________________________________________ javax-usb-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/javax-usb-devel
