thread sleeps in wdcwait() - check current lwp rather than relying
on global ATACH_TH_RUN channel flag
should fix the hang part of the problem reported in
http://mail-index.netbsd.org/netbsd-users/2020/03/12/msg024249.html
thanks to Paul Ripke for providing extensive debugging info
This bug was reported by UBSan runs.
lib/libusbhid/parse.c:246:20
Can result in left shift changes signedness bit as a side effect positive number
can go negative, cast it to unsigned for the operation and silence the issue.
Reviewed by: kamil@
This bug was reported by UBSan runs.
lib/libusbhid/usage.c:247:27
lib/libusbhid/usage.c:244:28
lib/libusbhid/usage.c:235:13
Can result in left shift changes signedness bit as a side effect positive number
can go negative, cast it to unsigned for the operation and silence the issue.
Reviewed by: kamil@
This bug was reported by UBSan runs.
lib/libusbhid/data.c:58:25
lib/libusbhid/data.c:91:7
lib/libusbhid/data.c:92:7
Can result in left shift changes signedness bit as a side effect positive number
can go negative, cast it to unsigned for the operation and silence the issue.
Reviewed by: kamil@
and getcwd():
- push vnode locking back as far as possible.
- do most lookups directly in the namecache, avoiding vnode locks & refs.
- don't block new refs to vnodes across VOP_INACTIVE().
- get shared locks for VOP_LOOKUP() if the file system supports it.
- correct lock types for VOP_ACCESS() / VOP_GETATTR() in a few places.
Possible future enhancements:
- make the lookups lockless.
- support dotdot lookups by being lockless and inferring absence of chroot.
- maybe make it work for layered file systems.
- avoid vnode references at the root & cwd.
identifier uniquely identifies an LWP across the entire system, and will
be used in future improvements in user-space synchronization primitives.
(Test disabled and libc stub not included intentionally so as to avoid
multiple libc version bumps.)
to avoid stealing back the LWP, remember the target CPU and avoid it.
- sched_preempted(): only choose a totally idle CPU; in the interests of
overall throughput it's better to make use of SMT / slow CPUs.
/netbsd/modules respectively instead of /netbsd and
/stand/<arch>/<version>/modules. This is only supported for x86,
and is turned off by default. To try it, add KERNEL_DIR=yes in your
/mk.conf and install a system from that build.
kamil