On Wed, Nov 15, 2017 at 12:12:16PM +0000, Billy Araujo wrote: > Hi all, > > I built a kernel/rootfs for altera SoCkit amd uses the default SoCkit > device tree and I plug when in plug in a USB stick it uses the dwc2 > driver. What kernel version are you using here? > When doing cat /proc/interrupts I get normal amount of interrupts - > everything seems ok. > > However, when I connect the USB stick to a USB hub and then to the > same USB port I get a great amount of interrupts approx. 8000 per > second. Anyone know why this different behaviour? > > root@cyclone5:~# cat /proc/interrupts > > CPU0 > > 16: 53922 GIC 29 Edge twd > > 17: 0 GIC 199 Level timer0 > > 18: 0 GIC 136 Level ffe01000.pdma > > 26: 0 GIC 190 Level ffc04000.i2c > > 27: 0 GIC 191 Level ffc05000.i2c > > 29: 24718 GIC 171 Level dw-mci > > 40: 1138 GIC 194 Level serial > > 41: 3946968 GIC 160 Level ffb40000.usb, ffb40000.usb, > dwc2_hsotg:usb1 > > > > root@cyclone5:~# [ 651.854021] usb 1-1: new high-speed USB device > number 7 using dwc2 > > [ 652.065190] usb 1-1: New USB device found, idVendor=05e3, idProduct=0608 > > [ 652.071869] usb 1-1: New USB device strings: Mfr=0, Product=1, SerialNumber=0 > > [ 652.079007] usb 1-1: Product: USB2.0 Hub > > [ 652.088468] hub 1-1:1.0: USB hub found > > [ 652.092830] hub 1-1:1.0: 4 ports detected > > [ 652.374026] usb 1-1.1: new high-speed USB device number 8 using dwc2 > > [ 652.475473] usb 1-1.1: New USB device found, idVendor=05e3, idProduct=0608 > > [ 652.482331] usb 1-1.1: New USB device strings: Mfr=0, Product=1, > SerialNumber=0 > > [ 652.489639] usb 1-1.1: Product: USB2.0 Hub > > [ 652.499214] hub 1-1.1:1.0: USB hub found > > [ 652.503762] hub 1-1.1:1.0: 4 ports detected > > > I have seen several threads with stating this issue but haven't found > a clear answer. Why is this an "issue"? Is the device not working properly? Is the device not seen correctly? USB is a "constantly asking the device for data" type of protocol, lots of interrupts is a normal thing. Heck, we know of some machines where you plug a USB keyboard into them and it starts to take 30% of the CPU time up just to handle the interrupts. That's not USB's fault, or the kernels, it's the horrible hardware implementation that the board has on it, where it is up to the CPU to do most of the work. Odds are that is what is happening here with your platform. 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
