https://bugzilla.kernel.org/show_bug.cgi?id=215740 --- Comment #36 from Ferry Toth (ferry.toth@xxxxxxxxxxxx) --- Your latest patch gives me: root@yuna:~# journalctl -k | grep "hub buffer" kernel: usb usb1: hub buffer at 6e1cda0, status at 6e1cda8 kernel: usb usb2: hub buffer at 6e1ce20, status at 6e1ce28 kernel: usb 1-1: hub buffer at 8001380, status at 8001388 The first 2 lines are for the root hubs. In my case on Intel Edison-Arduino the distribution is Yocto generated. The platform is especially suited to attach custom hardware, iio devices, arduino shields and the like, as you have direct access to gpios, i2c, spi, hsu. So CONFIG_DMA_API_DEBUG was enabled to help debug device drivers. I have to admit I never realised it to be an expensive option. And certainly didn't expect to find bugs in DMA_API_DEBUG itself or kmalloc(). I'll turn that into an optional thing with a comment on possible side effects. -- You may reply to this email to add a comment. You are receiving this mail because: You are watching the assignee of the bug.
