Folks, in the discussion about preempt count consistency accross kernel configurations: https://lore.kernel.org/r/20200914204209.256266093@xxxxxxxxxxxxx/ Linus clearly requested that code in drivers and libraries which changes behaviour based on execution context should either be split up so that e.g. task context invocations and BH invocations have different interfaces or if that's not possible the context information has to be provided by the caller which knows in which context it is executing. This includes conditional locking, allocation mode (GFP_*) decisions and avoidance of code paths which might sleep. In the long run, usage of 'preemptible, in_*irq etc.' should be banned from driver code completely. The usage of such constructs in USB is rather limited. Most of it is in debug code and (misleading) comments. But of course there are also a few few bugs including one unfixable. With the following series applied, USB is clean. Thanks, tglx --- atm/usbatm.c | 2 core/buffer.c | 6 +- core/hcd-pci.c | 6 +- core/hcd.c | 21 ++++---- core/hub.c | 3 - core/message.c | 35 +++++++++----- core/usb.c | 4 - gadget/udc/core.c | 2 gadget/udc/dummy_hcd.c | 5 +- gadget/udc/pxa27x_udc.c | 19 ++++--- host/ehci-fsl.c | 9 +-- host/ehci-pmcmsp.c | 15 +++--- host/isp1362.h | 5 +- host/ohci-at91.c | 11 +++- host/ohci-omap.c | 7 +- host/ohci-pxa27x.c | 11 ++-- host/ohci-s3c2410.c | 12 ++--- host/xhci-mem.c | 2 host/xhci.c | 6 -- misc/sisusbvga/Kconfig | 2 serial/digi_acceleport.c | 7 +- serial/keyspan_pda.c | 112 ++++++++++++++++++++--------------------------- usbip/usbip_common.c | 5 -- 23 files changed, 156 insertions(+), 151 deletions(-)
