On Fri, May 18, 2018 at 11:57:01AM +0000, guido@xxxxxxxxxxxxxxxxxx wrote: > > Zitat von Greg KH <gregkh@xxxxxxxxxxxxxxxxxxx>: > > > On Thu, May 17, 2018 at 07:03:24PM +0200, Guido Kiener wrote: > > > The working group "VISA for Linux" of the IVI Foundation > > > www.ivifoundation.org specifies common rules, shared libraries and > > > drivers to implement the specification of "VPP-4.3: The VISA Library" > > > on Linux to be compatible with implementations on other operating systems. > > > > > > The USBTMC protocol is part of the "VISA Library" that is used by many > > > popular T&M applications. > > > > > > An initial implementation for Linux based on libusb has been created. > > > While functional it has some drawbacks: > > > - Performance > > > - Requires root privileges to reclaim devices already claimed by > > > the usbtmc driver > > > > > > The following collaborative patches meet the requirements of the IVI > > > Foundation to implement the library based on the usbtmc driver. > > > > > > Improvements in the data transfer rate of over 130 MByte/s for > > > usb 3.x connections have been measured. > > > > Why is the libusb version "slower"? Last I checked, we could reach > > line-speeds for USB quite easily from userspace with no problem. If you > > keep the endpoints full you should be just fine. Unless you have a > > horrid protocol that doesn't allow for that :) > > Sorry, the wording is misleading. The performance of libusb is pretty well, > whereas the bandwidth of the current usbtmc driver is slow. > > From my point of view the main advantage of the new usbtmc driver in > contrast to libusb is: > - Multiple applications can share access to the same instruments. > - The driver handles SRQ conflicts. > - usbtmc driver simplifies definition of udev rules > - usbtmc driver simplifies development of applications using T&M instruments. Ok, yes, those are all great reasons to use a kernel driver instead of a userspace implementation. Put that in the documentation somewhere next time. thanks, greg k-h -- 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
