Pull up following revision(s) (requested by yurix in ticket #561):
sys/arch/arm/ti/am3_prcm.c: revision 1.15
am335x: fix timer clock offset
Closes PR 60910
Fix by Elias Howell.
Pull up following revision(s) (requested by joe in ticket #1405):
sys/dev/pci/if_vmx.c: revision 1.19
fix wrong range from computing the MIN of tx and rx queues.
Pull up following revision(s) (requested by joe in ticket #560):
sys/dev/pci/if_vmx.c: revision 1.19
fix wrong range from computing the MIN of tx and rx queues.
Pull up following revision(s) (requested by rin in ticket #1402):
sys/arch/alpha/alpha/vm_machdep.c: revision 1.124
sys/arch/alpha/alpha/machdep.c: revision 1.386
alpha: Align kernel stack pointer to 16-byte boundaries
as ABI requires. With this fix, full ATF run successfully completed
without any regression.
PR port-alpha/60865 NetBSD-current/alpha panics running getcwd test
OK thorpej@
Pull up following revision(s) (requested by rin in ticket #557):
sys/arch/alpha/alpha/vm_machdep.c: revision 1.124
sys/arch/alpha/alpha/machdep.c: revision 1.386
alpha: Align kernel stack pointer to 16-byte boundaries
as ABI requires. With this fix, full ATF run successfully completed
without any regression.
PR port-alpha/60865 NetBSD-current/alpha panics running getcwd test
OK thorpej@
Pull up following revision(s) (requested by riastradh in ticket #1401):
libexec/ld.elf_so/expand.c: revision 1.8
libexec/ld.elf_so/paths.c: revision 1.43
lib/libc/misc/stack_protector.c: revision 1.10
lib/libc/include/__sysctl.h: revision 1.1
ld.elf_so: Use __sysctl() syscall stub rather than sysctl() wrapper.
This avoids pulling in some unnecessary junk from libc, saving on a
couple kilobytes of space and some relative relocations at runtime.
PR lib/59387: ld.elf_so: needlessly includes libc/gen/sysctl.c user.*
Pull up following revision(s) (requested by riastradh in ticket #554):
sys/kern/vfs_vnode.c: revision 1.159
vdrain: Recompute target after each iteration.
This way if sysctl -w kern.maxvnodes=N fails, the vdrain task doesn't
perpetually keep trying to attain the unattainable.
Doesn't fix everything about racy access to desiredvnodes, but it
should help with the major symptom.
PR kern/60603: Sisyphean vdrain task
Pull up following revision(s) (requested by riastradh in ticket #1400):
sys/arch/sparc/sparc/machdep.c: revision 1.342
sparc: Load the signal handler _before_ resetting it, not after.
Should fix SA_RESETHAND.
PR port-sparc/59323: t_sigaction:sigaction_resethand test is failing
Pull up following revision(s) (requested by riastradh in ticket #1399):
sys/kern/sys_futex.c: revision 1.26
futex(2): Avoid returning early on timeout.
Rounding in the arithmetic leading into cv_timedwait_sig, and any
skew between the timecounter used by clock_gettime and the hardclock
timer used to wake cv_timedwait_sig, can lead cv_timedwait_sig to
wake up before the deadline as observable by clock_gettime.
futex(FUTEX_WAIT) is not supposed to do that, so ignore when
cv_timedwait_sig returns EWOULDBLOCK -- we'll notice the deadline has
passed in the next iteration anyway, if it has actually passed.
While here, make sure that we never pass less than 1 tick to
cv_timedwait_sig -- that turns it into cv_wait_sig, to wait
indefinitely with no timeout.
With this change, I have not seen any failures as reported in:
[8 lines not shown]
Pull up following revision(s) (requested by riastradh in ticket #1398):
sys/kern/sys_futex.c: revision 1.23
sys/kern/sys_futex.c: revision 1.24
sys/kern/sys_futex.c: revision 1.25
tests/lib/libc/sys/t_futex_ops.c: revision 1.11
tests/lib/libc/sys/t_futex_ops.c: revision 1.12
tests/lib/libc/sys/t_futex_ops.c: revision 1.13
tests/lib/libc/sys/t_futex_ops.c: revision 1.14
sys/sys/futex.h: revision 1.6
sys/sys/futex.h: revision 1.7
sys/futex.h: Nix trailing whitespace.
No functional change intended.
t_futex_ops: Spruce up diagnostics.
- Use RL to report errno for any syscalls that set it.
[45 lines not shown]
Pull up following revision(s) (requested by riastradh in ticket #1397):
sys/uvm/uvm_map.c: revision 1.432
sys/sys/proc.h: revision 1.374
sys/kern/kern_proc.c: revision 1.284
sys/kern/kern_proc.c: revision 1.285
sys/kern/kern_proc.c: revision 1.286
(all via patch)
kern: New function proc_find_reflocked.
This is like proc_find_locked but returns with p_reflock held rather
than p_lock. Will be used to replace proc_find_locked in order to
fix:
PR kern/60568: panic locking against myself (p->p_lock) in NFS from
sysctl_vmproc
No modules currently use proc_find_locked or proc_find_reflocked, so
[32 lines not shown]
Pull up following revision(s) (requested by riastradh in ticket #553):
sys/uvm/uvm_map.c: revision 1.432
sys/sys/proc.h: revision 1.374
sys/kern/kern_proc.c: revision 1.284
sys/kern/kern_proc.c: revision 1.285
sys/kern/kern_proc.c: revision 1.286
kern: New function proc_find_reflocked.
This is like proc_find_locked but returns with p_reflock held rather
than p_lock. Will be used to replace proc_find_locked in order to
fix:
PR kern/60568: panic locking against myself (p->p_lock) in NFS from
sysctl_vmproc
No modules currently use proc_find_locked or proc_find_reflocked, so
there should be no module ABI compatibility issues here.
[31 lines not shown]
Pull up following revision(s) (requested by riastradh in ticket #551):
sys/nfs/nfs_vnops.c: revision 1.328
sys/nfs/nfs_vnops.c: revision 1.329
nfs_vnops.c: Sort includes.
No functional change intended. Preparation for:
PR kern/60554: nfs: integer shift UB in pathconf(_PC_FILESIZEBITS)
nfs: Avoid UB in integer shift amounts.
Just use ilog2 explicitly here, instead of writing out a loop with
shifts and increments to compute it more slowly and buggily. Note
the confusing fencepost adjustment. Clamp _PC_FILESIZEBITS at the
actual size of off_t since we can't go beyond that.
PR kern/60554: nfs: integer shift UB in pathconf(_PC_FILESIZEBITS)
sun4v: Force the hypervisor interrupt api to use version 3.
This way the sun4v_intr_* functions in intr.c will not be using the
deprecated version 1 hypervisor interrupt calls.
Verified working on a T4 ldom with version 3.
