On Thu, 29 Sep 2011, gene heskett wrote: > On Thursday, September 29, 2011 10:43:52 AM Alan Stern did opine: > > > On Wed, 28 Sep 2011, gene heskett wrote: > > > Greets all; > > > > > > I have an old Labtec 'webcam' that I bought 6 or 7 years ago, and > > > which was worthless then because its data stream could overpower the > > > usb drivers & stall. > > > > > > Then I built an ASUS M2N_SLI Deluxe machine with 4Gb of ram, an AMD > > > Phenom X4 running at 2.1Ghz, figuring that would, or should be, > > > enough horsepower. > > > > > > Now, while the images have better color, I can still only run it in > > > 320x240 mode. 640x480 will freeze within the first 5 seconds. > > > According to the files motion saves, 320x240 isn't quite 16k a frame, > > > and 640x480 is about 43k a frame. > > > > > > dmesg shows this: > > > > > > gspca: probing 046d:08a2 > > > zc3xx: probe sensor -> 000e > > > zc3xx: Find Sensor PAS202B > > > input: zc3xx as /devices/pci0000:00/0000:00:02.0/usb2/2-6/input/input8 > > > gspca: video0 created > > > ohci_hcd 0000:00:02.0: leak ed f3eb2000 (#81) state 2 > > > > That shouldn't happen. Can you provide a usbmon trace showing what > > happens when the driver is loaded and you run your test? > > I am not at all familiar with usbmon, perhaps you can instruct me how to do that? > I installed it, but am getting a universal error message: > usbmon -i 2 > usbmon: Can't find usbmon in /proc/devices > > and it indeed is not there. > Running it as root, on pclos, kernel 2.6.38.8-pclos3.pae.bfs Instructions for usbmon are in the kernel source file Documentation/usb/usbmon.txt. If you use the text interface then you don't need to install any usbmon packages. > > > gspca: bandwidth not wide enough - trying again > > > ohci_hcd 0000:00:02.0: leak ed f3eb2240 (#81) state 2 > > > gspca: bandwidth not wide enough - trying again > > > ohci_hcd 0000:00:02.0: leak ed f3eb2280 (#81) state 2 > > > gspca: bandwidth not wide enough - trying again > > > ohci_hcd 0000:00:02.0: leak ed f3eb22c0 (#81) state 2 > > > gspca: bandwidth not wide enough - trying again > > > ohci_hcd 0000:00:02.0: leak ed f3eb2300 (#81) state 2 > > > gspca: bandwidth not wide enough - trying again > > > ohci_hcd 0000:00:02.0: leak ed f3eb2340 (#81) state 2 > > > gspca: bandwidth not wide enough - trying again > > > > > > Is this z3xx driver still suffering from not enough system cajones to > > > keep up? > > > > It's probably not system resources but rather the limitations of > > full-speed USB. What does "lsusb -v" have to say about this webcam? > > And do you have any other USB devices plugged into the same controller > > (the lsusb listing will show them)? > > > > Alan Stern > > First, thank you Alan. > > My usb tree looks like a weeping willow, with stuff all > over the house. Some is in the basement on a boosted cable. ;) > First, an lsusb to demo that: > > [root@coyote /]# lsusb > Bus 001 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub > Bus 001 Device 002: ID 0409:005a NEC Corp. HighSpeed Hub > Bus 001 Device 004: ID 04b8:0856 Seiko Epson Corp. Stylus SX515W > (Miss identified, is actually an NX515.) > Bus 001 Device 008: ID 05e3:0608 Genesys Logic, Inc. USB-2.0 4-Port HUB > Bus 001 Device 017: ID 03eb:3301 Atmel Corp. at43301 4-Port Hub > Bus 001 Device 010: ID 03eb:3301 Atmel Corp. at43301 4-Port Hub > Bus 001 Device 011: ID 0403:6001 Future Technology Devices International, Ltd FT232 USB-Serial (UART) IC > Bus 001 Device 012: ID 05e3:0608 Genesys Logic, Inc. USB-2.0 4-Port HUB > Bus 001 Device 013: ID 0a5c:4500 Broadcom Corp. BCM2046B1 USB 2.0 Hub (part of BCM2046 Bluetooth) > Bus 001 Device 014: ID 0a5c:4502 Broadcom Corp. Keyboard (Boot Interface Subclass) > Bus 001 Device 015: ID 0a5c:4503 Broadcom Corp. Mouse (Boot Interface Subclass) > Bus 001 Device 016: ID 0a5c:2148 Broadcom Corp. BCM92046DG-CL1ROM Bluetooth 2.1 Adapter > Bus 001 Device 018: ID 0409:0059 NEC Corp. HighSpeed Hub > Bus 001 Device 019: ID 0409:0059 NEC Corp. HighSpeed Hub > Bus 001 Device 020: ID 0403:6001 Future Technology Devices International, Ltd FT232 USB-Serial (UART) IC > Bus 001 Device 021: ID 04f9:0033 Brother Industries, Ltd > Bus 002 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub > Bus 002 Device 006: ID 046d:08a2 Logitech, Inc. Labtec Webcam Pro > Bus 002 Device 003: ID 046d:c517 Logitech, Inc. LX710 Cordless Desktop Laser > Bus 002 Device 004: ID 04b8:010f Seiko Epson Corp. Perfection 1250 > Bus 002 Device 005: ID 050d:0751 Belkin Components Okay. The stuff attached to bus 1 doesn't matter, because it's using a different USB controller. The other devices attached to bus 2 might cause a problem. Can you try running a test with the Logitech LX710, the Epson Perfection 1250, and the Belkin thing (a hub?) all unplugged? > Then the lsusb -v for the camera, which ATM is plugged into a motherboard > breakout cable: ... That looks okay. The highest resolution setting will push the limit, but as long as no other devices are plugged in, it should work. Full speed USB is limited to about 1100 bytes of periodic data per frame, and the webcam's highest resolution uses 1023 bytes for video, 8 bytes for something else (I don't know what it is), and 32 bytes for audio. That ought to fit within the limit. Even with some other devices, the less-than-highest resolution setting should be okay. That's why I need to see the usbmon trace. One other thing: After you mount /sys/kernel/debug (as described in the usbmon instructions), send a copy of /sys/kernel/debug/usb/devices. > Thanks again Alan. > > I also tried this on a much older machine loaded with Ubuntu-10.04 LTS, > It cannot even do the 320x240 at normal speed, and the color is terrible. > Very overexposed. Plugged into the front panel of an HP Pavilion with > a 'Taho'(sp?) motherboard. 1Ghz athlon, very slow, non-interleaved memory. > USB could be 1.1 on that box. It's not the box; it's the webcam. Or maybe an intermediate hub or extension cable, if the webcam is plugged into one. Alan Stern -- To unsubscribe from this list: send the line "unsubscribe linux-usb" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
