Fix info leaks.

branches: 1.19.4;
Rename min/max -> uimin/uimax for better honesty.

These functions are defined on unsigned int. The generic name
min/max should not silently truncate to 32 bits on 64-bit systems.
This is purely a name change -- no functional change intended.

HOWEVER! Some subsystems have

#define min(a, b) ((a) < (b) ? (a) : (b))
#define max(a, b) ((a) > (b) ? (a) : (b))

even though our standard name for that is MIN/MAX. Although these
may invite multiple evaluation bugs, these do _not_ cause integer
truncation.

To avoid `fixing' these cases, I first changed the name in libkern,
and then compile-tested every file where min/max occurred in order to
confirm that it failed -- and thus confirm that nothing shadowed
min/max -- before changing it.

I have left a handful of bootloaders that are too annoying to
compile-test, and some dead code:

cobalt ews4800mips hp300 hppa ia64 luna68k vax
acorn32/if_ie.c (not included in any kernels)
macppc/if_gm.c (superseded by gem(4))

It should be easy to fix the fallout once identified -- this way of
doing things fails safe, and the goal here, after all, is to _avoid_
silent integer truncations, not introduce them.

Maybe one day we can reintroduce min/max as type-generic things that
never silently truncate. But we should avoid doing that for a while,
so that existing code has a chance to be detected by the compiler for
conversion to uimin/uimax without changing the semantics until we can
properly audit it all. (Who knows, maybe in some cases integer
truncation is actually intended!)

branches: 1.18.2; 1.18.4;
Fail, don't panic, on bad dirents from file system.

Controllable via puffs from userland.

From Ilja Van Sprundel.

branches: 1.17.6;
avoid the shadow

we have space for one more.

fix size computation

branches: 1.14.2;
convert strcpy to a bounded copy to avoid compiler warnings, although the
reclen test prevents overflow.

branches: 1.13.8; 1.13.12; 1.13.14; 1.13.22; 1.13.28; 1.13.30; 1.13.32; 1.13.34; 1.13.38;
Change the vnode locking protocol of VOP_GETATTR() to request at least
a shared lock. Make all calls outside of file systems respect it.

The calls from file systems need review.

No objections from tech-kern.

recent versions of linux (which we now claim to be) supply a BSD-style
d_type value in getdents() results, after the d_name field.
make our emulation do the same. fixes part of PR 43695.

Clean up vnode lock operations pass 2:

VOP_UNLOCK(vp, flags) -> VOP_UNLOCK(vp): Remove the unneeded flags argument.

Welcome to 5.99.32.

Discussed on tech-kern.

branches: 1.10.2;
When implementing "read directory", when there are too many empty entries
in a row, and we need to try to read the next block, and have passed a
non-NULL cookie pointer to VOP_READDIR, ensure that we free the cookie
buffer before re-doing VOP_READDIR, so that we don't leak memory.
This fix is similar to nfs_serv.c revisions 1.115 + 1.124.

This should fix the long-standing problem observed by e.g. using Linux-
emulated programs to take backup of servers, which is one of the problems
which were reported in PR#42661.

Thanks to pooka@ for the hints for traversing the VOP* layer.

branches: 1.9.2;
Make compat linux/linux32 getdents(2) fail with ENOTDIR instead of
EINVAL when file descriptor does not refer to a directory.

branches: 1.8.2;
s/syscallcarg/syscallarg/ in comments.

Make the emulations, exec formats, coredump, NFS, and the NFS server
into modules. By and large this commit:

- shuffles header files and ifdefs
- splits code out where necessary to be modular
- adds module glue for each of the components
- adds/replaces hooks for things that can be installed at runtime

branches: 1.6.2; 1.6.4; 1.6.8;
Make linux32 getdents(2) do its own job instead of calling the
corresponding compat linux function, where struct dirent members types
differs.

branches: 1.5.6; 1.5.10; 1.5.12; 1.5.16;
Add readdir syscall.

Convert all the system call entry points from:
int foo(struct lwp *l, void *v, register_t *retval)
to:
int foo(struct lwp *l, const struct foo_args *uap, register_t *retval)
Fixup compat code to not write into 'uap' and (in some cases) to actually
pass a correctly formatted 'uap' structure with the right name to the
next routine.
A few 'compat' routines that just call standard ones have been deleted.
All the 'compat' code compiles (along with the kernels required to test
build it).
98% done by automated scripts.

branches: 1.3.4;
ANSIfy most of the function definitions in sys/compat (but not ndis).
All by the magic of sed ...

branches: 1.2.18; 1.2.20; 1.2.26; 1.2.28;
Merge newlock2 to head.

branches: 1.1.2; 1.1.10; 1.1.16; 1.1.20; 1.1.22;
Add initial (but unfinished) COMPAT_LINUX32 for amd64. This is good enough so
that the i386 license manager part of amd64 version of Fluent works.

While I'm here, add SysV IPC to COMPAT_LINUX/amd64

- Complete exit_group() implementation. We now maintain a list of Linux
threads in a processes and kill them properly. The code is a bit too
complicated, but I could not find a simplier way of dealing with it

- Change getpid() and getppid() semantics to match what Linux does,
and implement gettid(). In the Linux kernel, threads are implemnted
as plain old processes. A thread group is just a set of processes,
with the parent called leader. Thread ID, which are returned by gettid(),
are just the PID of the plain old processes, and getpid() returns the
PID of the thread group leader.

- Remove struct linux32_emuldata. COMPAT_LINUX32 uses a lot of COMPAT_LINUX
code, where a struct linux_emuldata is assumed. By having distinct emuldata
structure with different sizes and layouts, we caused kernel memory
corruptions.

- Fix setprioriry() and getpriority()

Thanks to Nicolas Joly for tracking down the problem and providing me the
hardware to fix them.

branches: 1.1.2; 1.1.4; 1.1.10; 1.1.14; 1.1.16;
Add initial (but unfinished) COMPAT_LINUX32 for amd64. This is good enough so
that the i386 license manager part of amd64 version of Fluent works.

While I'm here, add SysV IPC to COMPAT_LINUX/amd64

add support COMPAT_LINUX32 for aarch64

branches: 1.1.2; 1.1.10; 1.1.16; 1.1.22;
Add initial (but unfinished) COMPAT_LINUX32 for amd64. This is good enough so
that the i386 license manager part of amd64 version of Fluent works.

While I'm here, add SysV IPC to COMPAT_LINUX/amd64

Allow syscall_establish() to install new syscalls when the existing
entry-point is either sys_nomodule or sys_nosys. Update the
makesyscalls.sh script to create a const array of bits to allow
syscall_disestablish() to properly restore the original entry-point.
Update all the initializers of struct emul to initialize the pointer
to the bit array struct emul.

XXX Regen of all files created by makesyscalls.sh will come soon,
XXX followed by a kernel version bump (since struct emul is being
XXX modified).

This commit should address PR kern/45781 and also removes the need
for the work-around for that PR in file

sys/arch/usermode/modules/syscallemu/syscallemu.c

branches: 1.23.2;
Remove an element from struct emul: e_tracesig

e_tracesig used to be implemented for Darwin compat. Nowadays the Darwin
compatiblity layer is gone and there are no other users.

This functionality isn't used where it shall be used in the existing
codebase.

If we want to emulate debugging interfaces in compat layers we would need
to implement that from scratch anyway. We would need to be bug compatible
with other OSes too.

Proposed on tech-kern@.

Welcome to NetBSD 8.99.16!

Sponsored by <The NetBSD Foundation>

branches: 1.22.2;
remove struct emul's e_fault.

It used to be used by COMPAT_IRIX for the purpose of overriding
uvm_fault (only implemented in MIPS), now removed.

Ride 8.99.12 version bump.

Remove COMPAT_SA / KERN_SA. Welcome to 6.99.3!
Approved by core@.

branches: 1.20.8; 1.20.12;
many changes for COMPAT_LINUX:
- update the linux syscall table for each platform.
- support new-style (NPTL) linux pthreads on all platforms.
clone() with CLONE_THREAD uses 1 process with many LWPs
instead of separate processes.
- move the contents of sys__lwp_setprivate() into a new
lwp_setprivate() and use that everywhere.
- update linux_release[] and linux32_release[] to "2.6.18".
- adjust placement of emul fork/exec/exit hooks as needed
and adjust other emul code to match.
- convert all struct emul definitions to use named initializers.
- change the pid allocator to allow multiple pids to refer to the same proc.
- remove a few fields from struct proc that are no longer needed.
- disable the non-functional "vdso" code in linux32/amd64,
glibc works fine without it.
- fix a race in the futex code where we could miss a wakeup after
a requeue operation.
- redo futex locking to be a little more efficient.

branches: 1.19.2; 1.19.4;
Initialise struct emul members by name (it is readable now and one can search
them in the tree).

ANSIfy another 1261 function definitions.
The only ones left in sys are beyond by sed script!
(or in sys/dist or sys/external)
Mostly they have function pointer parameters.

branches: 1.17.2;
kill MALLOC and FREE macros.

Make the emulations, exec formats, coredump, NFS, and the NFS server
into modules. By and large this commit:

- shuffles header files and ifdefs
- splits code out where necessary to be modular
- adds module glue for each of the components
- adds/replaces hooks for things that can be installed at runtime

branches: 1.15.2;
Merge wrstuden-revivesa into HEAD.

Compile NetBSD/amd64 kernels with "-Wextra". Patches contributed by
Juan RP in PR port-amd64/39266.

branches: 1.13.2; 1.13.6;
Remove clause 3 and 4 from TNF licenses

branches: 1.12.2;
Remove a duplicate mutex_enter call in linux32_e_proc_init().
Fix a kernel panic for processes created with FORK_SHAREVM.

Network protocol interrupts can now block on locks, so merge the globals
proclist_mutex and proclist_lock into a single adaptive mutex (proc_lock).
Implications:

- Inspecting process state requires thread context, so signals can no longer
be sent from a hardware interrupt handler. Signal activity must be
deferred to a soft interrupt or kthread.

- As the proc state locking is simplified, it's now safe to take exit()
and wait() out from under kernel_lock.

- The system spends less time at IPL_SCHED, and there is less lock activity.

Catch up with linux.

branches: 1.9.12; 1.9.14;
ANSIfy most of the function definitions in sys/compat (but not ndis).
All by the magic of sed ...

Remove all the __P

branches: 1.7.2; 1.7.4;
machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h

branches: 1.6.2; 1.6.14; 1.6.16; 1.6.20;
more caddr_t lossage.

branches: 1.5.2;
Fix COMPAT_LINUX32.

Merge newlock2 to head.

branches: 1.3.2; 1.3.4;
1) Complete Linux exit_group() emulation
Members of the thread group must die without reporting to the parent and
without going to zombie stage. We do that by reparenting to init before
catching a SIGKILL. The parent will not see the child death.

The thread group leader must report the exit status, even if it exits
because of another thread calling exit_group(). We do that by storing the
exit status in struct linux_emuldata_shared, and the exit hook has the
duty of setting struct proc's p_xstat for the thread group leader.

2) For exit/fork/exec hooks, move the NPTL specific code to separate functions
that are shared between COMPAT_LINUX and COMPAT_LINUX32

3) Fix LINUX_CLONE_PARENT_SETTID semantics

- Complete exit_group() implementation. We now maintain a list of Linux
threads in a processes and kill them properly. The code is a bit too
complicated, but I could not find a simplier way of dealing with it

- Change getpid() and getppid() semantics to match what Linux does,
and implement gettid(). In the Linux kernel, threads are implemnted
as plain old processes. A thread group is just a set of processes,
with the parent called leader. Thread ID, which are returned by gettid(),
are just the PID of the plain old processes, and getpid() returns the
PID of the thread group leader.

- Remove struct linux32_emuldata. COMPAT_LINUX32 uses a lot of COMPAT_LINUX
code, where a struct linux_emuldata is assumed. By having distinct emuldata
structure with different sizes and layouts, we caused kernel memory
corruptions.

- Fix setprioriry() and getpriority()

Thanks to Nicolas Joly for tracking down the problem and providing me the
hardware to fix them.

branches: 1.1.2; 1.1.4; 1.1.10; 1.1.14; 1.1.16;
Add initial (but unfinished) COMPAT_LINUX32 for amd64. This is good enough so
that the i386 license manager part of amd64 version of Fluent works.

While I'm here, add SysV IPC to COMPAT_LINUX/amd64

- LINUX_GO_RT0_SIGNATURE for compat_linux is no longer enabled by default.
it is now defined in each arch same as the other *_SIGNATURE definitions.
- add new LINUX32_GO_RT0_SIGNATURE for compat_linux32, and is defined in amd64 and aarch64.

add support COMPAT_LINUX32 for aarch64

many changes for COMPAT_LINUX:
- update the linux syscall table for each platform.
- support new-style (NPTL) linux pthreads on all platforms.
clone() with CLONE_THREAD uses 1 process with many LWPs
instead of separate processes.
- move the contents of sys__lwp_setprivate() into a new
lwp_setprivate() and use that everywhere.
- update linux_release[] and linux32_release[] to "2.6.18".
- adjust placement of emul fork/exec/exit hooks as needed
and adjust other emul code to match.
- convert all struct emul definitions to use named initializers.
- change the pid allocator to allow multiple pids to refer to the same proc.
- remove a few fields from struct proc that are no longer needed.
- disable the non-functional "vdso" code in linux32/amd64,
glibc works fine without it.
- fix a race in the futex code where we could miss a wakeup after
a requeue operation.
- redo futex locking to be a little more efficient.

branches: 1.6.2; 1.6.4;
Change u_long to vaddr_t/vsize_t in exec code where appropriate (mostly
involves setregs and vmcmds). Should result in no code differences.

Make the emulations, exec formats, coredump, NFS, and the NFS server
into modules. By and large this commit:

- shuffles header files and ifdefs
- splits code out where necessary to be modular
- adds module glue for each of the components
- adds/replaces hooks for things that can be installed at runtime

branches: 1.4.12; 1.4.16; 1.4.22; 1.4.24;
Remove all the __P

branches: 1.3.4; 1.3.28; 1.3.30; 1.3.36; 1.3.40;
1) Complete Linux exit_group() emulation
Members of the thread group must die without reporting to the parent and
without going to zombie stage. We do that by reparenting to init before
catching a SIGKILL. The parent will not see the child death.

The thread group leader must report the exit status, even if it exits
because of another thread calling exit_group(). We do that by storing the
exit status in struct linux_emuldata_shared, and the exit hook has the
duty of setting struct proc's p_xstat for the thread group leader.

2) For exit/fork/exec hooks, move the NPTL specific code to separate functions
that are shared between COMPAT_LINUX and COMPAT_LINUX32

3) Fix LINUX_CLONE_PARENT_SETTID semantics

Add a new signature test for linux probe function. We look for a .debuglink
section, which is specific to the Linux dynamic interpeter (yes, Linux
can execute it as a stand alone program)

branches: 1.1.2; 1.1.4; 1.1.10; 1.1.16;
Add initial (but unfinished) COMPAT_LINUX32 for amd64. This is good enough so
that the i386 license manager part of amd64 version of Fluent works.

While I'm here, add SysV IPC to COMPAT_LINUX/amd64

- LINUX_GO_RT0_SIGNATURE for compat_linux is no longer enabled by default.
it is now defined in each arch same as the other *_SIGNATURE definitions.
- add new LINUX32_GO_RT0_SIGNATURE for compat_linux32, and is defined in amd64 and aarch64.

add support COMPAT_LINUX32 for aarch64

Extra arguments for AT_HWCAP and AT_PLATFORM are now passed when executing linux32 binaries.

AT_SYSINFO (vsyscall) feature seems to be broken, so leave it commented out.

sys/compat: Memset zero before copyout.

Just in case of uninitialized padding which would lead to kernel
stack disclosure. If the compiler can prove the memset redundant
then it can optimize it away; otherwise better safe than sorry.

branches: 1.19.8; 1.19.20;
The way the kernel tries to prevent a userland process from allocating page
zero is hugely flawed. It is easy to demonstrate that one can trick UVM
into chosing a NULL hint after the user_va0_disable check from uvm_map.
Such a bypass allows kernel NULL pointer dereferences to be exploitable on
architectures with a shared userland<->kernel VA, like amd64.

Fix this by increasing the limit of the vm space made available for
userland processes. This way, UVM will never chose a NULL hint, since it
would be outside of the vm space.

The user_va0_disable sysctl still controls this feature.

Zero-fill the ELF auxiliary vectors. Otherwise, on 64bit systems, the
padding between a_v and a_type contains kernel garbage, therefore
exposed to userland.

Original report by uebayasi@

branches: 1.17.4; 1.17.6;
Use Aux32Info not AuxInfo. Noticed by Maxime Villard.

Fix wrong KASSERTs. Do not compare size vs. entries count.

Increase LINUX32_ELF_AUX_ENTRIES to avoid overrun in linux32/. Also,
add comments and KASSERTs to make sure people don't forget to increase
XX_AUX_ENTRIES's when adding vectors.

Reported by martin@ (CV), with suggestions from chs@.

ok martin@ chs@

account for the 16 bytes of AT_RANDOM data in the stack setup. fixes PR 48518.
use cprng_strong32() instead of random(). add AT_RANDOM support for linux32.

branches: 1.13.6; 1.13.10;
Add a hook for freeing an ep_emul_arg. Add a wrapper routine
(exec_free_emul_arg) to call the hook and then clear the ep_emul_arg
and ep_emul_arg_free members in the exec_package.
Change users/accessors to use these routines.
Approved by releng.

branches: 1.12.8; 1.12.12;
always supply an auxiliary vector for linux ELF processes.
static executables (such as newer versions of /sbin/ldconfig)
require this to work properly. since static executables
also don't have a PT_PHDR entry, use the same heuristic as
linux does to provide a value for AT_PHDR in this case.

many changes for COMPAT_LINUX:
- update the linux syscall table for each platform.
- support new-style (NPTL) linux pthreads on all platforms.
clone() with CLONE_THREAD uses 1 process with many LWPs
instead of separate processes.
- move the contents of sys__lwp_setprivate() into a new
lwp_setprivate() and use that everywhere.
- update linux_release[] and linux32_release[] to "2.6.18".
- adjust placement of emul fork/exec/exit hooks as needed
and adjust other emul code to match.
- convert all struct emul definitions to use named initializers.
- change the pid allocator to allow multiple pids to refer to the same proc.
- remove a few fields from struct proc that are no longer needed.
- disable the non-functional "vdso" code in linux32/amd64,
glibc works fine without it.
- fix a race in the futex code where we could miss a wakeup after
a requeue operation.
- redo futex locking to be a little more efficient.

This patch fixes the NX regression issue observed on amd64 kernels, where
per-page execution right was disabled (therefore leading to the inability
of the kernel to detect fraudulent use of memory mappings marked as not
being executable).

- replace cpu_feature and ci_feature_flags variables by cpu_feature and
ci_feat_val arrays. This makes it cleaner and brings kernel code closer
to the design of cpuctl(8). A warning will be raised for each CPU that
does not expose the same features as the Boot Processor (BP).

- the blacklist of CPU features is now a macro defined in the
specialreg.h header, instead of hardcoding it inside MD initialization
code; fix comments.

- replace checks against CPUID_TSC with the cpu_hascounter() function.

- clean up the code in init_x86_64(), as cpu_feature variables are set
inside cpu_probe().

- use cpu_init_msrs() for i386. It will be eventually used later for NX
feature under i386 PAE kernels.

- remove code that checks for CPUID_NOX in amd64 mptramp.S, this is already
performed by cpu_hatch() through cpu_init_msrs().

- remove cpu_signature and feature_flags members from struct mpbios_proc
(they were never used).

This patch was tested with i386 MONOLITHIC, XEN3PAE_DOM0 and XEN3_DOM0 under
a native i386 host, and amd64 GENERIC, XEN3_DOM0 via QEMU virtual machines.

XXX Should kernel rev be bumped?

XXX A similar patch should be pulled-up for NetBSD-5, hopefully tomorrow.

branches: 1.9.2; 1.9.4;
ansify function definitions

branches: 1.8.8; 1.8.14;
Remove clause 3 and 4 from TNF licenses

branches: 1.7.28; 1.7.30; 1.7.32;
Change the way that emulations locate files within the emulation root to
avoid having to allocate space in the 'stackgap'
- which is very LWP unfriendly.
The additional code for non-emulation namei() is trivial, the reduction for
the emulations is massive.
The vnode for a processes emulation root is saved in the cwdi structure
during process exec.
If the emulation root the TRYEMULROOT flag are set, namei() will do an initial
search for absolute pathnames in the emulation root, if that fails it will
retry from the normal root.
".." at the emulation root will always go to the real root, even in the middle
of paths and when expanding symlinks.
Absolute symlinks found using absolute paths in the emulation root will be
relative to the emulation root (so /usr/lib/xxx.so -> /lib/xxx.so links
inside the emulation root don't need changing).
If the root of the emulation would be returned (for an emulation lookup), then
the real root is returned instead (matching the behaviour of emul_lookup,
but being a cheap comparison here) so that programs that scan "../.."
looking for the root dircetory don't loop forever.
The target for symbolic links is no longer mangled (it used to get the
CHECK_ALT_xxx() treatment, so could get /emul/xxx prepended).
CHECK_ALT_xxx() are no more. Most of the change is deleting them, and adding
TRYEMULROOT to the flags to NDINIT().
A lot of the emulation system call stubs could now be deleted.

remove all the double (and triple) casts used to convert 32bit userspace
pointers to and from 64bit kernel pointers. Instead use the defines
NETBSD32PTR64(p32) to read a 32bit pointer and (the new) NETBSD32PTR32(p32,p64)
to write a 32bit pointer throughout.
The 32bit pointer is now a struct to enforce the above.
amd64 (with linux emul) and sparc64 will both compile (when the arch stuff
goes in soon), and amd64 still runs some i386 binaries.

branches: 1.5.6; 1.5.12; 1.5.16; 1.5.18; 1.5.20;
Add a new signature test for linux probe function. We look for a .debuglink
section, which is specific to the Linux dynamic interpeter (yes, Linux
can execute it as a stand alone program)

Use the LWP cached credentials where sane.

branches: 1.3.4;
include kauth.h
workaround a gcc4 warning.

branches: 1.2.2;
integrate kauth.

branches: 1.1.2; 1.1.4; 1.1.6; 1.1.8; 1.1.10;
Add initial (but unfinished) COMPAT_LINUX32 for amd64. This is good enough so
that the i386 license manager part of amd64 version of Fluent works.

While I'm here, add SysV IPC to COMPAT_LINUX/amd64

- Use LINUX32_F_xxx instead of LINUX_F_xxx for compat_linux32.
aarch64 and arm have different values.
- Compile linux32_sys_fadvise64() only if the syscall is required.

Fix fadvise64 syscalls. Unlike our, linux fadvise syscall do not
return error code; call do_posix_fadvise().

branches: 1.9.14; 1.9.18; 1.9.28;
Add fadvise64 and fadvise64_64 syscalls to compat_linux and compat_linux32.

branches: 1.8.6; 1.8.8;
s/syscallcarg/syscallarg/ in comments.

Make the emulations, exec formats, coredump, NFS, and the NFS server
into modules. By and large this commit:

- shuffles header files and ifdefs
- splits code out where necessary to be modular
- adds module glue for each of the components
- adds/replaces hooks for things that can be installed at runtime

branches: 1.6.6; 1.6.10; 1.6.12; 1.6.16; 1.6.18;
Use the same function to implement linux32_sys_fcntl() and linux_sys_fcntl64().
A rather excessive amount of 'appending 64' to names seems to have happened
when linux won support for files > 4G.
technically the linux32_sys_fcntl() function should error the 64bit file
locking requests - but life is too short.

Fix the linux32 emulations of fcntl() locking.
The 64bit linux emulation versions can't be used because the lock structure
alignment and field sizes all differ.
Since there need to be 4 different versions of the linux struct flock, and
amd64 kernel needs 3 of them compiled in, rather than replicating the same
code block twice more, move the body of the code into a few #defines
that can be expanded with the correct types in the linux[32]_sys_fcntl[64]()
functions.
Should fix problems running progams like skype running under linux32
emulation on amd64.

Convert all the system call entry points from:
int foo(struct lwp *l, void *v, register_t *retval)
to:
int foo(struct lwp *l, const struct foo_args *uap, register_t *retval)
Fixup compat code to not write into 'uap' and (in some cases) to actually
pass a correctly formatted 'uap' structure with the right name to the
next routine.
A few 'compat' routines that just call standard ones have been deleted.
All the 'compat' code compiles (along with the kernels required to test
build it).
98% done by automated scripts.

branches: 1.3.4;
ANSIfy most of the function definitions in sys/compat (but not ndis).
All by the magic of sed ...

branches: 1.2.18; 1.2.20; 1.2.26; 1.2.28;
Merge newlock2 to head.

branches: 1.1.2; 1.1.10; 1.1.16; 1.1.20; 1.1.22;
Add initial (but unfinished) COMPAT_LINUX32 for amd64. This is good enough so
that the i386 license manager part of amd64 version of Fluent works.

While I'm here, add SysV IPC to COMPAT_LINUX/amd64

- Use LINUX32_F_xxx instead of LINUX_F_xxx for compat_linux32.
aarch64 and arm have different values.
- Compile linux32_sys_fadvise64() only if the syscall is required.

Add the timerfd syscalls to COMPAT_LINUX and COMPAT_LINUX32.

branches: 1.14.8;
Put the printf under DEBUG_LINUX.

branches: 1.13.38; 1.13.50; 1.13.56; 1.13.64; 1.13.68;
add the video ioctls so that the 32 bit skype works with video
more commits to come

Make the emulations, exec formats, coredump, NFS, and the NFS server
into modules. By and large this commit:

- shuffles header files and ifdefs
- splits code out where necessary to be modular
- adds module glue for each of the components
- adds/replaces hooks for things that can be installed at runtime

branches: 1.11.2; 1.11.4;
an attempt at linux32_ioctl_socket().

branches: 1.10.6; 1.10.10; 1.10.12; 1.10.14;
Add ossaudio support for COMPAT_LINUX32, with help from mrg. With this
change, audio now works on amd64 with native firefox, nspluginwrapper, and
the 32-bit linux flash binaries.

Convert all the system call entry points from:
int foo(struct lwp *l, void *v, register_t *retval)
to:
int foo(struct lwp *l, const struct foo_args *uap, register_t *retval)
Fixup compat code to not write into 'uap' and (in some cases) to actually
pass a correctly formatted 'uap' structure with the right name to the
next routine.
A few 'compat' routines that just call standard ones have been deleted.
All the 'compat' code compiles (along with the kernels required to test
build it).
98% done by automated scripts.

branches: 1.8.4;
ANSIfy most of the function definitions in sys/compat (but not ndis).
All by the magic of sed ...

branches: 1.7.8; 1.7.10; 1.7.16; 1.7.18;
Change all the NETBSD32PTR64(SCARG(uap, xxx))) to SCARG_P32(uap, xxx).

remove all the double (and triple) casts used to convert 32bit userspace
pointers to and from 64bit kernel pointers. Instead use the defines
NETBSD32PTR64(p32) to read a 32bit pointer and (the new) NETBSD32PTR32(p32,p64)
to write a 32bit pointer throughout.
The 32bit pointer is now a struct to enforce the above.
amd64 (with linux emul) and sparc64 will both compile (when the arch stuff
goes in soon), and amd64 still runs some i386 binaries.

branches: 1.5.2; 1.5.4; 1.5.6;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.

branches: 1.4.4; 1.4.6; 1.4.10;
Jumbo COMPAT_LINUX/COMPAT_LINUX32 bugfix, with the help of Nicolas Joly
- Fix shmat return value on amd64: it uses no black magic with retval[0]
- Fix integer overflows in sysinfo
- Implement sysinfo, mmap2, sched_getparam, sched_getscheduler, mremap,
and madvise in COMPAT_LINUX32
- Fix improper types used in setgroups16/getgroups16
- Implement mmap2 for COMPAT_LINUX32
- Ifdef debug messages by DEBUG_LINUX

branches: 1.3.2; 1.3.4;
Add support for socket ioctl.

branches: 1.2.2; 1.2.4; 1.2.10; 1.2.16;
- Move the BSD to Linux and Linux to BSD termio/termios conversion functions
to a header where they can be shared between COMPAT_LINUX and COMPAT_LINUX32
- Add termios ioctl emulation to COMPAT_LINUX32
- Add the getcwd system call to COMPAT_LINUX32/amd64

That makes Linux's bash working with COMPAT_LINUX32.

Add initial (but unfinished) COMPAT_LINUX32 for amd64. This is good enough so
that the i386 license manager part of amd64 version of Fluent works.

While I'm here, add SysV IPC to COMPAT_LINUX/amd64

add support COMPAT_LINUX32 for aarch64

an attempt at linux32_ioctl_socket().

branches: 1.4.2; 1.4.4;
Remove clause 3 and 4 from TNF licenses

branches: 1.3.6; 1.3.8; 1.3.10;
Convert all the system call entry points from:
int foo(struct lwp *l, void *v, register_t *retval)
to:
int foo(struct lwp *l, const struct foo_args *uap, register_t *retval)
Fixup compat code to not write into 'uap' and (in some cases) to actually
pass a correctly formatted 'uap' structure with the right name to the
next routine.
A few 'compat' routines that just call standard ones have been deleted.
All the 'compat' code compiles (along with the kernels required to test
build it).
98% done by automated scripts.

branches: 1.2.4;
Remove all the __P

branches: 1.1.2; 1.1.10; 1.1.16; 1.1.22; 1.1.46; 1.1.48; 1.1.54; 1.1.58;
- Move the BSD to Linux and Linux to BSD termio/termios conversion functions
to a header where they can be shared between COMPAT_LINUX and COMPAT_LINUX32
- Add termios ioctl emulation to COMPAT_LINUX32
- Add the getcwd system call to COMPAT_LINUX32/amd64

That makes Linux's bash working with COMPAT_LINUX32.

Make linux32_msqid_ds l_msg_[src]time members use the right type,
linux32_time_t.

Follow up commit to fix breakage that somehow passed by the compiler.

Add SYSVMSG support for linux32.

branches: 1.1.2; 1.1.4; 1.1.8; 1.1.22;
Add compat linux32 ipc(2) syscall support, for shared memory and
semaphores.

for sysv ipc stat operations, explicitly copy the exported parts
instead of the whole ds structure.

besides triggering a recently added assert in netbsd32, this stops
exposing kernel addresses.

copy the mode clamping to 0777 from sem to shm and msg.


while here, make sure that the compat callers to sysv_ipc clear
the contents of the compat structure before setting the result
members to ensure padding bytes are cleared.

don't set/copy _sem_base, _msg_first, _msg_last or _shm_internal.
even if used, which seems very dodgy, they leak KVAs as well.
possibly this may affect linux binaries, in particular, the
comments around _shm_internal ("XXX Oh well.") may mean apps
rely upon these but hopefully not -- the comments date back to
rev 1.1 in 1995.

the _key, _seq and _msg_cbytes members are exported as before as
i found multiple consumers of these (no less than ipcs(1), and
they appear to be useful for debugging and more.


XXX: the naming of compat functions have too many styles. there
are at least 3 different ones changed here.

branches: 1.11.34; 1.11.38; 1.11.46; 1.11.52; 1.11.60;
Fix build without sysvipc configured, from Wolfgang Stukenbrock in PR 43376,
adjusted for current.

branches: 1.10.2; 1.10.4;
Remove extra semicolon.

Make compat linux/linux32 msgrcv_msgarg type member of type
long/netbsd32_long to match linux definition.

Make linux32 msgsnd/msgrcv use the netbsd32 syscalls instead of native
ones, to ensure that msgbuf mtype member will be of the right type.

Kill noisy debug printf.

Follow up commit to fix breakage that somehow passed by the compiler.

Add SYSVMSG support for linux32.

branches: 1.4.2;
Add IPC_64 support to all semctl(2) commands, following corresponding
compat linux change.

Add IPC_64 support for all shmctl(2) commands, and reduce diffs with
compat linux version.

branches: 1.2.2; 1.2.8;
In linux32_shmat() don't try to copyout a 64-bit pointer where userspace
is expecting a 32-bit result.

Fixes nspluginwrapper on amd64.

branches: 1.1.2; 1.1.4; 1.1.8; 1.1.10;
Add compat linux32 ipc(2) syscall support, for shared memory and
semaphores.

add support COMPAT_LINUX32 for aarch64

many changes for COMPAT_LINUX:
- update the linux syscall table for each platform.
- support new-style (NPTL) linux pthreads on all platforms.
clone() with CLONE_THREAD uses 1 process with many LWPs
instead of separate processes.
- move the contents of sys__lwp_setprivate() into a new
lwp_setprivate() and use that everywhere.
- update linux_release[] and linux32_release[] to "2.6.18".
- adjust placement of emul fork/exec/exit hooks as needed
and adjust other emul code to match.
- convert all struct emul definitions to use named initializers.
- change the pid allocator to allow multiple pids to refer to the same proc.
- remove a few fields from struct proc that are no longer needed.
- disable the non-functional "vdso" code in linux32/amd64,
glibc works fine without it.
- fix a race in the futex code where we could miss a wakeup after
a requeue operation.
- redo futex locking to be a little more efficient.

branches: 1.2.16; 1.2.36; 1.2.38;
Remove all the __P

branches: 1.1.2; 1.1.10; 1.1.16; 1.1.22; 1.1.46; 1.1.48; 1.1.54; 1.1.58;
Add initial (but unfinished) COMPAT_LINUX32 for amd64. This is good enough so
that the i386 license manager part of amd64 version of Fluent works.

While I'm here, add SysV IPC to COMPAT_LINUX/amd64

add support COMPAT_LINUX32 for aarch64

Add preadv(2) and pwritev(2) system calls to COMPAT_LINUX and COMPAT_LINUX32.

Add the eventfd system calls to COMPAT_LINUX and COMPAT_LINUX32.

Remove redundant linux32_to_native_timespec() prototype.

branches: 1.30.2;
Always clear FUTEX_PRIVATE_FLAG for Linux processes. NetBSD-native futexes
exist in different namespace depending on FUTEX_PRIVATE_FLAG. This appears
not to be the case in Linux, and some futex users will mix private and non-
private ops on the same futex object. Provide a convenience wrapper that
puts this logic in one place witn a comment explaining why.

While here, move the Linux futex wrapper out of its own file and plop
it in linux_misc.c, which is where it lives in the linux32 module.

Oops, fix a silly mistake in the Linux futex syscall stubs -- we also
copy in the timeout for FUTEX_WAIT_BITSET.

Add a NetBSD native futex implementation, mostly written by riastradh@.
Map the COMPAT_LINUX futex calls to the native ones.

Hum, don't forget the 'pid' argument, otherwise we're not gonna go very
far.

branches: 1.26.6; 1.26.14; 1.26.18;
Return value pointers should be the last argument.

branches: 1.25.2; 1.25.4; 1.25.6;
Add ppoll(2) for compat linux32.

Do not uselessly include <sys/malloc.h>.

branches: 1.23.4;
implement the *at() syscalls.
bring the unimplemented syscall list up to date.

branches: 1.22.10; 1.22.14;
add sigtimedwait support

branches: 1.21.8;
personality() now interprets its parameter as having
the base personality type in the low byte and
various flags in the upper bytes. for now just mask off
the flags to make sure the base type is one we accept.
store the current personality in the emuldata so that
we can return the expected value for PER_QUERY.

implement the following syscalls for linux32:
truncate64
ftruncate64
profil
ioperm
iopl
setdomainname
modify_ldt
statfs64
fstatfs64

note that iopl(), ioperm() and modify_ldt() just call
the respective 64-bit handlers, which don't do anything yet.

many changes for COMPAT_LINUX:
- update the linux syscall table for each platform.
- support new-style (NPTL) linux pthreads on all platforms.
clone() with CLONE_THREAD uses 1 process with many LWPs
instead of separate processes.
- move the contents of sys__lwp_setprivate() into a new
lwp_setprivate() and use that everywhere.
- update linux_release[] and linux32_release[] to "2.6.18".
- adjust placement of emul fork/exec/exit hooks as needed
and adjust other emul code to match.
- convert all struct emul definitions to use named initializers.
- change the pid allocator to allow multiple pids to refer to the same proc.
- remove a few fields from struct proc that are no longer needed.
- disable the non-functional "vdso" code in linux32/amd64,
glibc works fine without it.
- fix a race in the futex code where we could miss a wakeup after
a requeue operation.
- redo futex locking to be a little more efficient.

branches: 1.18.2; 1.18.4;
Update personality(2) to match Linux definition where the argument is an
unsigned long.

Add fstatfs syscall.

branches: 1.16.2;
Add basic support for linux32 personality(2) syscall.

Make the emulations, exec formats, coredump, NFS, and the NFS server
into modules. By and large this commit:

- shuffles header files and ifdefs
- splits code out where necessary to be modular
- adds module glue for each of the components
- adds/replaces hooks for things that can be installed at runtime

Add missing sys/syscall.h include for SYS_ptrace definition.

Remove LKMs and switch to the module framework, pass 1.

Proposed on tech-kern@.

branches: 1.12.2; 1.12.4; 1.12.6;
linux32_sys_ptrace() fixes.
- Add opt_ptrace.h include to activate syscall.
- Convert to use sys_ptrace() instead of netbsd32_ptrace().
- Do not use arch specific call for now.

an attempt at ptrace32

branches: 1.10.2; 1.10.4;
Remove clause 3 and 4 from TNF licenses

branches: 1.9.6; 1.9.8; 1.9.10;
Convert all the system call entry points from:
int foo(struct lwp *l, void *v, register_t *retval)
to:
int foo(struct lwp *l, const struct foo_args *uap, register_t *retval)
Fixup compat code to not write into 'uap' and (in some cases) to actually
pass a correctly formatted 'uap' structure with the right name to the
next routine.
A few 'compat' routines that just call standard ones have been deleted.
All the 'compat' code compiles (along with the kernels required to test
build it).
98% done by automated scripts.

branches: 1.8.4;
ANSIfy most of the function definitions in sys/compat (but not ndis).
All by the magic of sed ...

branches: 1.7.2; 1.7.8; 1.7.10; 1.7.18;
Rework compat stat() and statvfs() code so that it no longer uses the stackgap.

Change the way that emulations locate files within the emulation root to
avoid having to allocate space in the 'stackgap'
- which is very LWP unfriendly.
The additional code for non-emulation namei() is trivial, the reduction for
the emulations is massive.
The vnode for a processes emulation root is saved in the cwdi structure
during process exec.
If the emulation root the TRYEMULROOT flag are set, namei() will do an initial
search for absolute pathnames in the emulation root, if that fails it will
retry from the normal root.
".." at the emulation root will always go to the real root, even in the middle
of paths and when expanding symlinks.
Absolute symlinks found using absolute paths in the emulation root will be
relative to the emulation root (so /usr/lib/xxx.so -> /lib/xxx.so links
inside the emulation root don't need changing).
If the root of the emulation would be returned (for an emulation lookup), then
the real root is returned instead (matching the behaviour of emul_lookup,
but being a cheap comparison here) so that programs that scan "../.."
looking for the root dircetory don't loop forever.
The target for symbolic links is no longer mangled (it used to get the
CHECK_ALT_xxx() treatment, so could get /emul/xxx prepended).
CHECK_ALT_xxx() are no more. Most of the change is deleting them, and adding
TRYEMULROOT to the flags to NDINIT().
A lot of the emulation system call stubs could now be deleted.

branches: 1.5.4;
- Add compat amd64 linux32 statfs support, adapted from PR/35956.
- Move bsd_to_linux_statfs() function to its own file to be shared
between both linux compats.

ok by manu.

branches: 1.4.4; 1.4.8;
This file was not used anymore

branches: 1.3.4; 1.3.6;
Use the LWP cached credentials where sane.

branches: 1.2.6;
integrate kauth.

branches: 1.1.2; 1.1.4; 1.1.6; 1.1.8; 1.1.10;
Add initial (but unfinished) COMPAT_LINUX32 for amd64. This is good enough so
that the i386 license manager part of amd64 version of Fluent works.

While I'm here, add SysV IPC to COMPAT_LINUX/amd64

Make the emulations, exec formats, coredump, NFS, and the NFS server
into modules. By and large this commit:

- shuffles header files and ifdefs
- splits code out where necessary to be modular
- adds module glue for each of the components
- adds/replaces hooks for things that can be installed at runtime

branches: 1.7.4; 1.7.6; 1.7.10; 1.7.12;
Be sure to not include linux_mmap.h before linux_syscallargs.h.
Otherwise wrong syscallarg macro may used for linux_sys_mmap_args
definition.
This makes mmap2 works again (at least to launch oo2 binary).

branches: 1.6.6;
Convert all the system call entry points from:
int foo(struct lwp *l, void *v, register_t *retval)
to:
int foo(struct lwp *l, const struct foo_args *uap, register_t *retval)
Fixup compat code to not write into 'uap' and (in some cases) to actually
pass a correctly formatted 'uap' structure with the right name to the
next routine.
A few 'compat' routines that just call standard ones have been deleted.
All the 'compat' code compiles (along with the kernels required to test
build it).
98% done by automated scripts.

branches: 1.5.4;
ANSIfy most of the function definitions in sys/compat (but not ndis).
All by the magic of sed ...

branches: 1.4.4;
Remove extra return

branches: 1.3.18; 1.3.20; 1.3.24;
Merge newlock2 to head.

Jumbo COMPAT_LINUX/COMPAT_LINUX32 bugfix, with the help of Nicolas Joly
- Fix shmat return value on amd64: it uses no black magic with retval[0]
- Fix integer overflows in sysinfo
- Implement sysinfo, mmap2, sched_getparam, sched_getscheduler, mremap,
and madvise in COMPAT_LINUX32
- Fix improper types used in setgroups16/getgroups16
- Implement mmap2 for COMPAT_LINUX32
- Ifdef debug messages by DEBUG_LINUX

branches: 1.1.2; 1.1.4; 1.1.10; 1.1.16; 1.1.20; 1.1.22;
Add initial (but unfinished) COMPAT_LINUX32 for amd64. This is good enough so
that the i386 license manager part of amd64 version of Fluent works.

While I'm here, add SysV IPC to COMPAT_LINUX/amd64

The default name of the `exec_setup_stack' function for linux32 should be `linux32_exec_setup_stack',
and it should be defined in each arch.
declaration of linux32_exec_setup_stack() in linux32/amd64 has been removed because it does not exist.

NFC.

Deal with having to teardown sysctl entries in multiple sub-trees.

(This used to work, but I broke it recently.)

Use the module subsystem's ability to process SYSCTL_SETUP() entries to
automate installation of sysctl nodes.

Note that there are still a number of device and pseudo-device modules
that create entries tied to individual device units, rather than to the
module itself. These are not changed.

Merge the [pgoyette-compat] branch

branches: 1.11.2; 1.11.4;

Finally (hopefully) return completely to the state where compat_linux is
enabled by default.

While here make it possible to enable this if already enabled, or disable
if already disabled, without error. Some minor KNF.

Finish the previous change - as well as setting the default for these
compat modules back to enabled, stop forcing it to switch back to
disabled any time the module is init'd (which is truly barbaric behaviour.)

Unbreak Linux emulation by default. While not loading it automatically
is somewhat sensible, breaking functionality of GENERIC is not and has
been objected to on a regular base.

Remove compat_linux32 from the autoload list and add a enable/disable
sysctl, like compat_linux.

Update dependencies - compat_linux32 requires compat_netbsd32_sysvipc

branches: 1.6.4;
1) On 64bit systems, don't add the 32bit execsw[] to the global exec array.
exec_elf32 works on 32bit systems only, and will crash 32bit binaries on
64bit systems.
2) Now that exec_elf32 is dormant, we can give the native ELF loaders the
highest priority.

Binaries will load faster now (system boot, compilation, etc.).

With the help of njloy@. Discussed a bit on tech-kern@, no disagreement.

branches: 1.5.2;
c99 initializers for struct execsw

exec modules need to be of the exec kind

branches: 1.3.8; 1.3.18; 1.3.22;
many changes for COMPAT_LINUX:
- update the linux syscall table for each platform.
- support new-style (NPTL) linux pthreads on all platforms.
clone() with CLONE_THREAD uses 1 process with many LWPs
instead of separate processes.
- move the contents of sys__lwp_setprivate() into a new
lwp_setprivate() and use that everywhere.
- update linux_release[] and linux32_release[] to "2.6.18".
- adjust placement of emul fork/exec/exit hooks as needed
and adjust other emul code to match.
- convert all struct emul definitions to use named initializers.
- change the pid allocator to allow multiple pids to refer to the same proc.
- remove a few fields from struct proc that are no longer needed.
- disable the non-functional "vdso" code in linux32/amd64,
glibc works fine without it.
- fix a race in the futex code where we could miss a wakeup after
a requeue operation.
- redo futex locking to be a little more efficient.

branches: 1.2.4; 1.2.6; 1.2.8; 1.2.12; 1.2.14; 1.2.16;
PR port-amd64/40091 compat_linux32 module loading = panic due to missing syms

Make the emulations, exec formats, coredump, NFS, and the NFS server
into modules. By and large this commit:

- shuffles header files and ifdefs
- splits code out where necessary to be modular
- adds module glue for each of the components
- adds/replaces hooks for things that can be installed at runtime

add prlimit64(2) syscall to COMPAT_LINUX and COMPAT_LINUX32

sys/compat: Memset zero before copyout.

Just in case of uninitialized padding which would lead to kernel
stack disclosure. If the compiler can prove the memset redundant
then it can optimize it away; otherwise better safe than sorry.

branches: 1.11.30; 1.11.42;
LINUX_RLIM_INFINITY doesn't fit into the value range for linux32, so
introduce a variant with correct value.

branches: 1.10.16; 1.10.26;
Make the emulations, exec formats, coredump, NFS, and the NFS server
into modules. By and large this commit:

- shuffles header files and ifdefs
- splits code out where necessary to be modular
- adds module glue for each of the components
- adds/replaces hooks for things that can be installed at runtime

branches: 1.9.2;
Kill a few linux32 syscalls identical to netbsd32 ones.
No functional changes expected.

branches: 1.8.6; 1.8.10; 1.8.12; 1.8.16;
Convert all the system call entry points from:
int foo(struct lwp *l, void *v, register_t *retval)
to:
int foo(struct lwp *l, const struct foo_args *uap, register_t *retval)
Fixup compat code to not write into 'uap' and (in some cases) to actually
pass a correctly formatted 'uap' structure with the right name to the
next routine.
A few 'compat' routines that just call standard ones have been deleted.
All the 'compat' code compiles (along with the kernels required to test
build it).
98% done by automated scripts.

branches: 1.7.4;
ANSIfy most of the function definitions in sys/compat (but not ndis).
All by the magic of sed ...

branches: 1.6.6; 1.6.8; 1.6.16;
There is no need to use the stackgap for get/setrlimit.

Change all the NETBSD32PTR64(SCARG(uap, xxx))) to SCARG_P32(uap, xxx).

branches: 1.4.2; 1.4.4; 1.4.6;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.

branches: 1.3.2;
Merge newlock2 to head.

branches: 1.2.4; 1.2.6;
- Complete exit_group() implementation. We now maintain a list of Linux
threads in a processes and kill them properly. The code is a bit too
complicated, but I could not find a simplier way of dealing with it

- Change getpid() and getppid() semantics to match what Linux does,
and implement gettid(). In the Linux kernel, threads are implemnted
as plain old processes. A thread group is just a set of processes,
with the parent called leader. Thread ID, which are returned by gettid(),
are just the PID of the plain old processes, and getpid() returns the
PID of the thread group leader.

- Remove struct linux32_emuldata. COMPAT_LINUX32 uses a lot of COMPAT_LINUX
code, where a struct linux_emuldata is assumed. By having distinct emuldata
structure with different sizes and layouts, we caused kernel memory
corruptions.

- Fix setprioriry() and getpriority()

Thanks to Nicolas Joly for tracking down the problem and providing me the
hardware to fix them.

branches: 1.1.2; 1.1.4; 1.1.10; 1.1.14; 1.1.16;
Add initial (but unfinished) COMPAT_LINUX32 for amd64. This is good enough so
that the i386 license manager part of amd64 version of Fluent works.

While I'm here, add SysV IPC to COMPAT_LINUX/amd64

many changes for COMPAT_LINUX:
- update the linux syscall table for each platform.
- support new-style (NPTL) linux pthreads on all platforms.
clone() with CLONE_THREAD uses 1 process with many LWPs
instead of separate processes.
- move the contents of sys__lwp_setprivate() into a new
lwp_setprivate() and use that everywhere.
- update linux_release[] and linux32_release[] to "2.6.18".
- adjust placement of emul fork/exec/exit hooks as needed
and adjust other emul code to match.
- convert all struct emul definitions to use named initializers.
- change the pid allocator to allow multiple pids to refer to the same proc.
- remove a few fields from struct proc that are no longer needed.
- disable the non-functional "vdso" code in linux32/amd64,
glibc works fine without it.
- fix a race in the futex code where we could miss a wakeup after
a requeue operation.
- redo futex locking to be a little more efficient.

branches: 1.8.6; 1.8.8;
Add sched_get_priority_{min,max} syscalls.

Make the emulations, exec formats, coredump, NFS, and the NFS server
into modules. By and large this commit:

- shuffles header files and ifdefs
- splits code out where necessary to be modular
- adds module glue for each of the components
- adds/replaces hooks for things that can be installed at runtime

branches: 1.6.6; 1.6.10; 1.6.12; 1.6.16; 1.6.18;
Convert all the system call entry points from:
int foo(struct lwp *l, void *v, register_t *retval)
to:
int foo(struct lwp *l, const struct foo_args *uap, register_t *retval)
Fixup compat code to not write into 'uap' and (in some cases) to actually
pass a correctly formatted 'uap' structure with the right name to the
next routine.
A few 'compat' routines that just call standard ones have been deleted.
All the 'compat' code compiles (along with the kernels required to test
build it).
98% done by automated scripts.

branches: 1.5.4;
ANSIfy most of the function definitions in sys/compat (but not ndis).
All by the magic of sed ...

branches: 1.4.2;
Remove remaining 64-bit compat linux syscalls linux_sys_xxx() (except
for ones without arguments), and replace them by their 32-bit
equivalent linux32_sys_xxxx().

branches: 1.3.18; 1.3.20; 1.3.24; 1.3.26;
Merge newlock2 to head.

Jumbo COMPAT_LINUX/COMPAT_LINUX32 bugfix, with the help of Nicolas Joly
- Fix shmat return value on amd64: it uses no black magic with retval[0]
- Fix integer overflows in sysinfo
- Implement sysinfo, mmap2, sched_getparam, sched_getscheduler, mremap,
and madvise in COMPAT_LINUX32
- Fix improper types used in setgroups16/getgroups16
- Implement mmap2 for COMPAT_LINUX32
- Ifdef debug messages by DEBUG_LINUX

branches: 1.1.2; 1.1.4; 1.1.10; 1.1.16; 1.1.20; 1.1.22;
Add initial (but unfinished) COMPAT_LINUX32 for amd64. This is good enough so
that the i386 license manager part of amd64 version of Fluent works.

While I'm here, add SysV IPC to COMPAT_LINUX/amd64

Add the POSIX timer syscalls (timer_create(), timer_settime(), timer_gettime(),
timer_getoverrun(), and timer_delete()) to COMPAT_LINUX and COMPAT_LINUX32.

G/C some unused (and unnecessary) definitions.

branches: 1.1.8; 1.1.66;
add sigtimedwait support

branches: 1.1.2; 1.1.4; 1.1.8; 1.1.22;
Add compat linux32 ipc(2) syscall support, for shared memory and
semaphores.

branches: 1.1.2; 1.1.4; 1.1.8; 1.1.22;
Add compat linux32 ipc(2) syscall support, for shared memory and
semaphores.

Add routines to convert from linux32 to native sigevent_t.

branches: 1.1.2;
file linux32_sigevent.h was initially added on branch thorpej-futex.

add support COMPAT_LINUX32 for aarch64

branches: 1.1.8; 1.1.66;
add sigtimedwait support

Remove #error, fixes the build. I don't understand why this would not work
on big endian. We use this construct to convert pointers everywhere else.

Use "stack_t" instead of "struct sigaltstack", as the former is the
newer standardized name. NFC.

Add routines to convert from linux32 to native sigevent_t.

sys/compat: Memset zero before copyout.

Just in case of uninitialized padding which would lead to kernel
stack disclosure. If the compiler can prove the memset redundant
then it can optimize it away; otherwise better safe than sorry.

branches: 1.20.8;
Fix info leaks in sigaltstack.

branches: 1.19.4; 1.19.8;
ignore signal 64 for now (go uses it)
XXX: pullup-8

branches: 1.18.10;
fix inconsistent names

branches: 1.17.4; 1.17.6; 1.17.12;
implement the *at() syscalls.
bring the unimplemented syscall list up to date.

branches: 1.16.2; 1.16.4;
Fix rt_sigtimedwait():
It is wishful thinking that:
1. declaring a 32 bit syscall with 64 bit pointers
2. passing a struct with 32 bit pointers to a 64 bit function
is going to work.

branches: 1.15.4; 1.15.6;
remove incorrect assertions (the len passed is the size of the bsd structure
not the linux one)

add sigtimedwait support

branches: 1.13.12;
Add rt_queueinfo(2) support.

Add siggetmask/sigsetmask syscalls.

bzero -> memset

branches: 1.10.2;
add some debugging.

branches: 1.9.2;
Add rt_sigpending syscall.

branches: 1.8.2; 1.8.4; 1.8.6; 1.8.8;
Merge proc::p_mutex and proc::p_smutex into a single adaptive mutex, since
we no longer need to guard against access from hardware interrupt handlers.

Additionally, if cloning a process with CLONE_SIGHAND, arrange to have the
child process share the parent's lock so that signal state may be kept in
sync. Partially addresses PR kern/37437.

branches: 1.7.6; 1.7.8;
Convert all the system call entry points from:
int foo(struct lwp *l, void *v, register_t *retval)
to:
int foo(struct lwp *l, const struct foo_args *uap, register_t *retval)
Fixup compat code to not write into 'uap' and (in some cases) to actually
pass a correctly formatted 'uap' structure with the right name to the
next routine.
A few 'compat' routines that just call standard ones have been deleted.
All the 'compat' code compiles (along with the kernels required to test
build it).
98% done by automated scripts.

use __KERNEL_RCSID()

branches: 1.5.2; 1.5.4;
ANSIfy most of the function definitions in sys/compat (but not ndis).
All by the magic of sed ...

branches: 1.4.8; 1.4.10; 1.4.18;
Change all the NETBSD32PTR64(SCARG(uap, xxx))) to SCARG_P32(uap, xxx).

remove all the double (and triple) casts used to convert 32bit userspace
pointers to and from 64bit kernel pointers. Instead use the defines
NETBSD32PTR64(p32) to read a 32bit pointer and (the new) NETBSD32PTR32(p32,p64)
to write a 32bit pointer throughout.
The 32bit pointer is now a struct to enforce the above.
amd64 (with linux emul) and sparc64 will both compile (when the arch stuff
goes in soon), and amd64 still runs some i386 binaries.

branches: 1.2.2; 1.2.6; 1.2.8; 1.2.10;
Fix COMPAT_LINUX32.

branches: 1.1.2; 1.1.10; 1.1.16; 1.1.22;
Add initial (but unfinished) COMPAT_LINUX32 for amd64. This is good enough so
that the i386 license manager part of amd64 version of Fluent works.

While I'm here, add SysV IPC to COMPAT_LINUX/amd64

Add pselect6(2) system call to COMPAT_LINUX32

add support COMPAT_LINUX32 for aarch64

Use "stack_t" instead of "struct sigaltstack", as the former is the
newer standardized name. NFC.

add sigtimedwait support

branches: 1.2.50;
Remove all the __P

branches: 1.1.2; 1.1.10; 1.1.16; 1.1.22; 1.1.46; 1.1.48; 1.1.54; 1.1.58;
Add initial (but unfinished) COMPAT_LINUX32 for amd64. This is good enough so
that the i386 license manager part of amd64 version of Fluent works.

While I'm here, add SysV IPC to COMPAT_LINUX/amd64

s/reqest/request/, s/requst/request/ and s/reuqest/request/ in comments.

Don't accept negative value.

Reported-by: syzbot+e71a77402d6668f1868d@syzkaller.appspotmail.com

branches: 1.30.4;
Zero out the ifreq struct for SIOCGIFCONF to avoid up to 127 bytes of stack
disclosure. From Andy Nguyen, many thanks! This is the compat code part
pointed out by ozaki-r@

branches: 1.29.2;
Fix pserialize enter/exit pairs in linux_getifconf and linux32_getifconf

Tested by using a small linux program that uses ioctl(SIOCGIFCONF).
Probably fix PR kern/53259 as well.

branches: 1.28.2;
Fix and make consistent of usages of psz/psref in ifconf variants

branches: 1.27.6;
Use if_acquire and if_release instead of using psref API directly

- Provide if_release for consistency to if_acquire
- Use if_acquire and if_release for ifp iterations
- Make ifnet_psref_class static

branches: 1.26.2;
Apply pserialize and psref to struct ifaddr and its variants

This change makes struct ifaddr and its variants (in_ifaddr and in6_ifaddr)
MP-safe by using pserialize and psref. At this moment, pserialize_perform
and psref_target_destroy are disabled because (1) we don't need them
because of softnet_lock (2) they cause a deadlock because of softnet_lock.
So we'll enable them when we remove softnet_lock in the future.

branches: 1.25.2;
Switch the address list of intefaces to pslist(9)

As usual, we leave the old list to avoid breaking kvm(3) users.

Fix psref isn't released in a case of IFADDR_EMPTY

Use curlwp_bind and curlwp_bindx instead of open-coding LP_BOUND

Protect if_byindex by pserialize

Protect ifnet list with psz and psref

The change ensures that ifnet objects in the ifnet list aren't freed during
list iterations by using pserialize(9) and psref(9).

Note that the change adds a pslist(9) for ifnet but doesn't remove the
original ifnet list (ifnet_list) to avoid breaking kvm(3) users. We
shouldn't use the original list in the kernel anymore.

Unused inits (harmless).

Found by Brainy.

branches: 1.19.2;
Tweak ifconf variants

The tweaks make the code intention clear and make further changes easy.

No functional change.

branches: 1.18.2; 1.18.4; 1.18.8;
Replace open-coded access (and boundary checking) of ifindex2ifnet with
if_byindex() function.

branches: 1.17.2; 1.17.10;
Use copyin/copyout and linux-specific ifreq structures (they are the same
as the netbsd ones, but this disconnects them)

branches: 1.16.2;
Do not read past end of array. Found by gcc -03

branches: 1.15.12; 1.15.16;
Add missing semicolons after NETBSD32TOx_UAP macro calls.

Add SIOCGIFMTU support.

Do not give native requests to netbsd32_ioctl(), use the compat
netbsd32 equivalent instead.

Return the result of copyout. Reminded by Niolas Joly.

Provide SIOCGIFNAME.

Make the emulations, exec formats, coredump, NFS, and the NFS server
into modules. By and large this commit:

- shuffles header files and ifdefs
- splits code out where necessary to be modular
- adds module glue for each of the components
- adds/replaces hooks for things that can be installed at runtime

branches: 1.9.2; 1.9.4; 1.9.8;
Add linux32_getifconf, following recent linux_getifconf addition.

branches: 1.8.2;
Add SIOCGIFHWADDR support.
Tested by Cem Kayali.

Fix SIOCGIFCONF ioctl under compat linux32.

an attempt at linux32_ioctl_socket().

branches: 1.5.6; 1.5.10; 1.5.12; 1.5.14;
Convert all the system call entry points from:
int foo(struct lwp *l, void *v, register_t *retval)
to:
int foo(struct lwp *l, const struct foo_args *uap, register_t *retval)
Fixup compat code to not write into 'uap' and (in some cases) to actually
pass a correctly formatted 'uap' structure with the right name to the
next routine.
A few 'compat' routines that just call standard ones have been deleted.
All the 'compat' code compiles (along with the kernels required to test
build it).
98% done by automated scripts.

branches: 1.4.4;
ANSIfy most of the function definitions in sys/compat (but not ndis).
All by the magic of sed ...

branches: 1.3.18; 1.3.20; 1.3.28;
Merge newlock2 to head.

Jumbo COMPAT_LINUX/COMPAT_LINUX32 bugfix, with the help of Nicolas Joly
- Fix shmat return value on amd64: it uses no black magic with retval[0]
- Fix integer overflows in sysinfo
- Implement sysinfo, mmap2, sched_getparam, sched_getscheduler, mremap,
and madvise in COMPAT_LINUX32
- Fix improper types used in setgroups16/getgroups16
- Implement mmap2 for COMPAT_LINUX32
- Ifdef debug messages by DEBUG_LINUX

branches: 1.1.2; 1.1.4; 1.1.10; 1.1.16; 1.1.20; 1.1.22;
Add initial (but unfinished) COMPAT_LINUX32 for amd64. This is good enough so
that the i386 license manager part of amd64 version of Fluent works.

While I'm here, add SysV IPC to COMPAT_LINUX/amd64

If SCARG(uap, what) = 0, copyin() will copy (size_t)-1 bytes, and it's not
a good idea; but not proven harmful.

With the help of njoly@

branches: 1.8.2; 1.8.12;
prefix message with linux32

branches: 1.7.16; 1.7.20; 1.7.22; 1.7.26; 1.7.28;
Make the emulations, exec formats, coredump, NFS, and the NFS server
into modules. By and large this commit:

- shuffles header files and ifdefs
- splits code out where necessary to be modular
- adds module glue for each of the components
- adds/replaces hooks for things that can be installed at runtime

branches: 1.6.6; 1.6.10; 1.6.16; 1.6.18; 1.6.20; 1.6.26; 1.6.30;
Convert all the system call entry points from:
int foo(struct lwp *l, void *v, register_t *retval)
to:
int foo(struct lwp *l, const struct foo_args *uap, register_t *retval)
Fixup compat code to not write into 'uap' and (in some cases) to actually
pass a correctly formatted 'uap' structure with the right name to the
next routine.
A few 'compat' routines that just call standard ones have been deleted.
All the 'compat' code compiles (along with the kernels required to test
build it).
98% done by automated scripts.

branches: 1.5.4;
ANSIfy most of the function definitions in sys/compat (but not ndis).
All by the magic of sed ...

branches: 1.4.2; 1.4.10;
Changes to make ktrace LKM friendly and reduce ifdef KTRACE. Proposed
on tech-kern.

branches: 1.3.2; 1.3.6;
(Ab)use a KTR_USER trace entry for the linux socket command and arguments.

Change all the NETBSD32PTR64(SCARG(uap, xxx))) to SCARG_P32(uap, xxx).

branches: 1.1.2; 1.1.10; 1.1.16; 1.1.22; 1.1.28; 1.1.32; 1.1.34; 1.1.36;
Add initial (but unfinished) COMPAT_LINUX32 for amd64. This is good enough so
that the i386 license manager part of amd64 version of Fluent works.

While I'm here, add SysV IPC to COMPAT_LINUX/amd64

Convert all the system call entry points from:
int foo(struct lwp *l, void *v, register_t *retval)
to:
int foo(struct lwp *l, const struct foo_args *uap, register_t *retval)
Fixup compat code to not write into 'uap' and (in some cases) to actually
pass a correctly formatted 'uap' structure with the right name to the
next routine.
A few 'compat' routines that just call standard ones have been deleted.
All the 'compat' code compiles (along with the kernels required to test
build it).
98% done by automated scripts.

branches: 1.2.4;
Remove all the __P

branches: 1.1.2; 1.1.10; 1.1.16; 1.1.22; 1.1.46; 1.1.48; 1.1.54; 1.1.58;
Add initial (but unfinished) COMPAT_LINUX32 for amd64. This is good enough so
that the i386 license manager part of amd64 version of Fluent works.

While I'm here, add SysV IPC to COMPAT_LINUX/amd64

Use copyin/copyout and linux-specific ifreq structures (they are the same
as the netbsd ones, but this disconnects them)

branches: 1.3.12; 1.3.22;
Provide SIOCGIFNAME.

branches: 1.2.2; 1.2.6; 1.2.14; 1.2.16;
Add linux32_getifconf, following recent linux_getifconf addition.

branches: 1.1.2; 1.1.4; 1.1.6;
Add SIOCGIFHWADDR support.
Tested by Cem Kayali.

- Add compat_linux statx(2) syscall.
- The AT_EMPTY_PATH processing from the modification of
sys/compat/linux/common/linux_file64.c r1.63 has been separated, and made
common to linux_statat(), so that it can be used not only by
linux32_sys_fstatat64() but also by other *statat() variants.

implement the *at() syscalls.
bring the unimplemented syscall list up to date.

branches: 1.16.12; 1.16.22; 1.16.26;
Add stat/lstat/fstat syscalls.

Cleanup (no functional changes).
Kill some unneeded variables and return stattement.
Rename linux32_from_stat() to better bsd_to_linux32_stat64().
Fix some types.

Fix inverted lst_ino/__lst_ino assignment in linux32_from_stat().

bzero -> memset

branches: 1.12.4;
Make the emulations, exec formats, coredump, NFS, and the NFS server
into modules. By and large this commit:

- shuffles header files and ifdefs
- splits code out where necessary to be modular
- adds module glue for each of the components
- adds/replaces hooks for things that can be installed at runtime

branches: 1.11.4; 1.11.6; 1.11.10; 1.11.12; 1.11.14;
Catch up with descriptor handling changes. See kern_descrip.c revision
1.173 for details.

branches: 1.10.6;
Convert all the system call entry points from:
int foo(struct lwp *l, void *v, register_t *retval)
to:
int foo(struct lwp *l, const struct foo_args *uap, register_t *retval)
Fixup compat code to not write into 'uap' and (in some cases) to actually
pass a correctly formatted 'uap' structure with the right name to the
next routine.
A few 'compat' routines that just call standard ones have been deleted.
All the 'compat' code compiles (along with the kernels required to test
build it).
98% done by automated scripts.

branches: 1.9.4;
ANSIfy most of the function definitions in sys/compat (but not ndis).
All by the magic of sed ...

branches: 1.8.6; 1.8.8; 1.8.16;
Change the way that emulations locate files within the emulation root to
avoid having to allocate space in the 'stackgap'
- which is very LWP unfriendly.
The additional code for non-emulation namei() is trivial, the reduction for
the emulations is massive.
The vnode for a processes emulation root is saved in the cwdi structure
during process exec.
If the emulation root the TRYEMULROOT flag are set, namei() will do an initial
search for absolute pathnames in the emulation root, if that fails it will
retry from the normal root.
".." at the emulation root will always go to the real root, even in the middle
of paths and when expanding symlinks.
Absolute symlinks found using absolute paths in the emulation root will be
relative to the emulation root (so /usr/lib/xxx.so -> /lib/xxx.so links
inside the emulation root don't need changing).
If the root of the emulation would be returned (for an emulation lookup), then
the real root is returned instead (matching the behaviour of emul_lookup,
but being a cheap comparison here) so that programs that scan "../.."
looking for the root dircetory don't loop forever.
The target for symbolic links is no longer mangled (it used to get the
CHECK_ALT_xxx() treatment, so could get /emul/xxx prepended).
CHECK_ALT_xxx() are no more. Most of the change is deleting them, and adding
TRYEMULROOT to the flags to NDINIT().
A lot of the emulation system call stubs could now be deleted.

Change all the NETBSD32PTR64(SCARG(uap, xxx))) to SCARG_P32(uap, xxx).

remove all the double (and triple) casts used to convert 32bit userspace
pointers to and from 64bit kernel pointers. Instead use the defines
NETBSD32PTR64(p32) to read a 32bit pointer and (the new) NETBSD32PTR32(p32,p64)
to write a 32bit pointer throughout.
The 32bit pointer is now a struct to enforce the above.
amd64 (with linux emul) and sparc64 will both compile (when the arch stuff
goes in soon), and amd64 still runs some i386 binaries.

branches: 1.5.2; 1.5.4;
Fix amd64 build for compat sys_stat() changes.
This file isn't included in amd64:GENERIC

Update all the compat stuff to not use the 'stackgap' for processing
sys_stat() and friends, instead use do_sys_stat() and do_sys_fstat()
that write the answer into a kernel buffer (on stack) that can be
converted to the correct form and written the userspace.
I've test compiled a few kernels, and tested i386 netbsd1.6 ls.
Given I think I've fixed some bugs, it might be 50-50 with new ones.

branches: 1.3.2;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.

branches: 1.2.2;
Merge newlock2 to head.

branches: 1.1.2; 1.1.10; 1.1.16; 1.1.20; 1.1.22;
Add initial (but unfinished) COMPAT_LINUX32 for amd64. This is good enough so
that the i386 license manager part of amd64 version of Fluent works.

While I'm here, add SysV IPC to COMPAT_LINUX/amd64

Deal with having to teardown sysctl entries in multiple sub-trees.

(This used to work, but I broke it recently.)

Use the module subsystem's ability to process SYSCTL_SETUP() entries to
automate installation of sysctl nodes.

Note that there are still a number of device and pseudo-device modules
that create entries tied to individual device units, rather than to the
module itself. These are not changed.

branches: 1.17.4;
Remove compat_linux32 from the autoload list and add a enable/disable
sysctl, like compat_linux.

Get rid of all sysc_init_field uses - initialize fields directly in C99
notation.

branches: 1.15.2;
Ensure that the top level sysctl nodes (kern, vfs, net, ...) exist before
the sysctl link sets are processed, and remove redundancy.

Shaves >13kB off of an amd64 GENERIC, not to mention >1k duplicate
lines of code.

claim to be linux 3.11.6 (opensuse 13.1).

branches: 1.13.8; 1.13.18; 1.13.22;
many changes for COMPAT_LINUX:
- update the linux syscall table for each platform.
- support new-style (NPTL) linux pthreads on all platforms.
clone() with CLONE_THREAD uses 1 process with many LWPs
instead of separate processes.
- move the contents of sys__lwp_setprivate() into a new
lwp_setprivate() and use that everywhere.
- update linux_release[] and linux32_release[] to "2.6.18".
- adjust placement of emul fork/exec/exit hooks as needed
and adjust other emul code to match.
- convert all struct emul definitions to use named initializers.
- change the pid allocator to allow multiple pids to refer to the same proc.
- remove a few fields from struct proc that are no longer needed.
- disable the non-functional "vdso" code in linux32/amd64,
glibc works fine without it.
- fix a race in the futex code where we could miss a wakeup after
a requeue operation.
- redo futex locking to be a little more efficient.

branches: 1.12.4; 1.12.6;
Make sysctl(2) fails with ENOTDIR (instead of EINVAL) for invalid mib
vector length.

Make the emulations, exec formats, coredump, NFS, and the NFS server
into modules. By and large this commit:

- shuffles header files and ifdefs
- splits code out where necessary to be modular
- adds module glue for each of the components
- adds/replaces hooks for things that can be installed at runtime

Remove LKMs and switch to the module framework, pass 1.

Proposed on tech-kern@.

branches: 1.9.6; 1.9.10; 1.9.12; 1.9.16; 1.9.18;
Patch up sysctl locking:

- Lock processes, credentials, filehead etc correctly.
- Acquire a read hold on sysctl_treelock if only doing a query.
- Don't wire down the output buffer. It doesn't work correctly and the code
regularly does long term sleeps with it held - it's not worth it.
- Don't hold locks other than sysctl_lock while doing copyout().
- Drop sysctl_lock while doing copyout / allocating memory in a few places.
- Don't take kernel_lock for sysctl.
- Fix a number of bugs spotted along the way

Convert all the system call entry points from:
int foo(struct lwp *l, void *v, register_t *retval)
to:
int foo(struct lwp *l, const struct foo_args *uap, register_t *retval)
Fixup compat code to not write into 'uap' and (in some cases) to actually
pass a correctly formatted 'uap' structure with the right name to the
next routine.
A few 'compat' routines that just call standard ones have been deleted.
All the 'compat' code compiles (along with the kernels required to test
build it).
98% done by automated scripts.

branches: 1.7.4;
ANSIfy most of the function definitions in sys/compat (but not ndis).
All by the magic of sed ...

branches: 1.6.2; 1.6.8; 1.6.10;
Changes to make ktrace LKM friendly and reduce ifdef KTRACE. Proposed
on tech-kern.

branches: 1.5.4; 1.5.8;
Change all the NETBSD32PTR64(SCARG(uap, xxx))) to SCARG_P32(uap, xxx).

branches: 1.4.2; 1.4.6; 1.4.8; 1.4.10;
Merge newlock2 to head.

Add sysctl tracing to emulations.
While we are there, fix a bug in FreeBSD sysctl emulation: use copyin for
moving data to the kernel

branches: 1.2.2; 1.2.4; 1.2.6;
Fix linux32_sys_sysctl, it was horribly broken.

branches: 1.1.2; 1.1.4; 1.1.10; 1.1.16;
Add initial (but unfinished) COMPAT_LINUX32 for amd64. This is good enough so
that the i386 license manager part of amd64 version of Fluent works.

While I'm here, add SysV IPC to COMPAT_LINUX/amd64

Deal with having to teardown sysctl entries in multiple sub-trees.

(This used to work, but I broke it recently.)

Use the module subsystem's ability to process SYSCTL_SETUP() entries to
automate installation of sysctl nodes.

Note that there are still a number of device and pseudo-device modules
that create entries tied to individual device units, rather than to the
module itself. These are not changed.

- Cleanup for dynamic sysctl:
- Remove unused *_NAMES macros for sysctl.
- Remove unused *_MAXID for sysctls.
- Move CTL_MACHDEP sysctl definitions for m68k into m68k/include/cpu.h and
use them on all m68k machines.

branches: 1.3.2; 1.3.4;
Remove compat_linux32 from the autoload list and add a enable/disable
sysctl, like compat_linux.

branches: 1.2.26;
Make the emulations, exec formats, coredump, NFS, and the NFS server
into modules. By and large this commit:

- shuffles header files and ifdefs
- splits code out where necessary to be modular
- adds module glue for each of the components
- adds/replaces hooks for things that can be installed at runtime

branches: 1.1.2; 1.1.10; 1.1.16; 1.1.22; 1.1.72; 1.1.76; 1.1.82; 1.1.86;
Add initial (but unfinished) COMPAT_LINUX32 for amd64. This is good enough so
that the i386 license manager part of amd64 version of Fluent works.

While I'm here, add SysV IPC to COMPAT_LINUX/amd64

add support COMPAT_LINUX32 for aarch64

Counter tweaks:

- Don't need to count anonpages+filepages any more; clean+unknown+dirty for
each kind of page can be summed to get the totals.

- Track the number of free pages with a counter so that it's one less thing
for the allocator to do, which opens up further options there.

- Remove cpu_count_sync_one(). It has no users and doesn't save a whole lot.
For the cheap option, give cpu_count_sync() a boolean parameter indicating
that a cached value is okay, and rate limit the updates for cached values
to hz.

uvm_availmem(): give it a boolean argument to specify whether a recent
cached value will do, or if the very latest total must be fetched. It can
be called thousands of times a second and fetching the totals impacts not
only the calling LWP but other CPUs doing unrelated activity in the VM
system.

- Only increment nprocs when we're creating a new process, not just
when allocating a PID.
- Per above, proc_free_pid() no longer decrements nprocs. It's now done
in proc_free() right after proc_free_pid().
- Ensure nprocs is accessed using atomics everywhere.

branches: 1.10.6;
Rename uvm_free() -> uvm_availmem().

uvmexp.free -> uvm_free()

Fix info leaks in sysinfo().

branches: 1.7.42; 1.7.54; 1.7.60; 1.7.68; 1.7.72;
Make the emulations, exec formats, coredump, NFS, and the NFS server
into modules. By and large this commit:

- shuffles header files and ifdefs
- splits code out where necessary to be modular
- adds module glue for each of the components
- adds/replaces hooks for things that can be installed at runtime

branches: 1.6.6; 1.6.10; 1.6.12; 1.6.16; 1.6.18;
Convert all the system call entry points from:
int foo(struct lwp *l, void *v, register_t *retval)
to:
int foo(struct lwp *l, const struct foo_args *uap, register_t *retval)
Fixup compat code to not write into 'uap' and (in some cases) to actually
pass a correctly formatted 'uap' structure with the right name to the
next routine.
A few 'compat' routines that just call standard ones have been deleted.
All the 'compat' code compiles (along with the kernels required to test
build it).
98% done by automated scripts.

branches: 1.5.4;
ANSIfy most of the function definitions in sys/compat (but not ndis).
All by the magic of sed ...

branches: 1.4.4; 1.4.6; 1.4.14; 1.4.16;
<sys/proc.h> is needed to get nprocs... fixes the build on amd64.

branches: 1.3.4;
Change all the NETBSD32PTR64(SCARG(uap, xxx))) to SCARG_P32(uap, xxx).

branches: 1.2.2; 1.2.6; 1.2.8; 1.2.10;
Merge newlock2 to head.

branches: 1.1.2; 1.1.6; 1.1.10;
Jumbo COMPAT_LINUX/COMPAT_LINUX32 bugfix, with the help of Nicolas Joly
- Fix shmat return value on amd64: it uses no black magic with retval[0]
- Fix integer overflows in sysinfo
- Implement sysinfo, mmap2, sched_getparam, sched_getscheduler, mremap,
and madvise in COMPAT_LINUX32
- Fix improper types used in setgroups16/getgroups16
- Implement mmap2 for COMPAT_LINUX32
- Ifdef debug messages by DEBUG_LINUX

The changes on compat/linux/common/linux_termios.c should also
be applied to compat/linux32/common/linux_termios.c, and reduce diff.

- remove duplicate fd_putfile(). This will be done after "goto out;".
- catch up from compat/linux/common/linux_termios.c r1.26 TIOCSPTLCK
- catch up from compat/linux/common/linux_termios.c r1.37 (TCOOFF/TCOON) of the TCXONC
- catch up from compat/linux/common/linux_termios.c r1.39 FIOCLEX and FIONCLEX

Add a default case, don't call sys_ioctl() with an uninitialized 'com'
argument.

branches: 1.14.42; 1.14.54; 1.14.60; 1.14.68; 1.14.72;
Make the emulations, exec formats, coredump, NFS, and the NFS server
into modules. By and large this commit:

- shuffles header files and ifdefs
- splits code out where necessary to be modular
- adds module glue for each of the components
- adds/replaces hooks for things that can be installed at runtime

branches: 1.13.2; 1.13.4;
make LINUX_TCSBRK work. The code is mostly copied from
sys/compat/ibcs2/ibcs2_ioctl.c.

branches: 1.12.2; 1.12.4;
Remove clause 3 and 4 from TNF licenses

branches: 1.11.2; 1.11.4;
fdputfile -> fd_putfile

Catch up with descriptor handling changes. See kern_descrip.c revision
1.173 for details.

branches: 1.9.6;
Convert all the system call entry points from:
int foo(struct lwp *l, void *v, register_t *retval)
to:
int foo(struct lwp *l, const struct foo_args *uap, register_t *retval)
Fixup compat code to not write into 'uap' and (in some cases) to actually
pass a correctly formatted 'uap' structure with the right name to the
next routine.
A few 'compat' routines that just call standard ones have been deleted.
All the 'compat' code compiles (along with the kernels required to test
build it).
98% done by automated scripts.

branches: 1.8.8; 1.8.16; 1.8.20;
This code is already set for directly calling a files ioctl fn with a
kernel buffer - so why not use it when getting the slave pty number.

Sync with recent compat linux termios changes.

Fix TIOCGPTN ioctl.
Do not use sysioctl() directly, but netbsd32_ioctl() instead.

Change all the NETBSD32PTR64(SCARG(uap, xxx))) to SCARG_P32(uap, xxx).

remove all the double (and triple) casts used to convert 32bit userspace
pointers to and from 64bit kernel pointers. Instead use the defines
NETBSD32PTR64(p32) to read a 32bit pointer and (the new) NETBSD32PTR32(p32,p64)
to write a 32bit pointer throughout.
The 32bit pointer is now a struct to enforce the above.
amd64 (with linux emul) and sparc64 will both compile (when the arch stuff
goes in soon), and amd64 still runs some i386 binaries.

branches: 1.3.2; 1.3.4; 1.3.6;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.

branches: 1.2.4; 1.2.10;
netbsd32_sys_ioctl was called with improper syscall argument structure, this
caused any X11 program to fail because FIONREAD ioctl always returned an
error.

branches: 1.1.2; 1.1.4; 1.1.10; 1.1.16;
- Move the BSD to Linux and Linux to BSD termio/termios conversion functions
to a header where they can be shared between COMPAT_LINUX and COMPAT_LINUX32
- Add termios ioctl emulation to COMPAT_LINUX32
- Add the getcwd system call to COMPAT_LINUX32/amd64

That makes Linux's bash working with COMPAT_LINUX32.

add support COMPAT_LINUX32 for aarch64

branches: 1.1.2; 1.1.10; 1.1.16; 1.1.22;
- Move the BSD to Linux and Linux to BSD termio/termios conversion functions
to a header where they can be shared between COMPAT_LINUX and COMPAT_LINUX32
- Add termios ioctl emulation to COMPAT_LINUX32
- Add the getcwd system call to COMPAT_LINUX32/amd64

That makes Linux's bash working with COMPAT_LINUX32.

Add the timerfd syscalls to COMPAT_LINUX and COMPAT_LINUX32.

Add the POSIX timer syscalls (timer_create(), timer_settime(), timer_gettime(),
timer_getoverrun(), and timer_delete()) to COMPAT_LINUX and COMPAT_LINUX32.

sys/compat: Memset zero before copyout.

Just in case of uninitialized padding which would lead to kernel
stack disclosure. If the compiler can prove the memset redundant
then it can optimize it away; otherwise better safe than sorry.

branches: 1.37.30; 1.37.42; 1.37.50;
kernel portion of clock_nanosleep()

branches: 1.36.10;
add sigtimedwait support

branches: 1.35.8;
Add CLOCK_MONOTONIC support for compat linux/linux32
clock_nanosleep(2).

many changes for COMPAT_LINUX:
- update the linux syscall table for each platform.
- support new-style (NPTL) linux pthreads on all platforms.
clone() with CLONE_THREAD uses 1 process with many LWPs
instead of separate processes.
- move the contents of sys__lwp_setprivate() into a new
lwp_setprivate() and use that everywhere.
- update linux_release[] and linux32_release[] to "2.6.18".
- adjust placement of emul fork/exec/exit hooks as needed
and adjust other emul code to match.
- convert all struct emul definitions to use named initializers.
- change the pid allocator to allow multiple pids to refer to the same proc.
- remove a few fields from struct proc that are no longer needed.
- disable the non-functional "vdso" code in linux32/amd64,
glibc works fine without it.
- fix a race in the futex code where we could miss a wakeup after
a requeue operation.
- redo futex locking to be a little more efficient.

Do not use 0 for pointers, but NULL.

Add a new clock_gettime1() function that holds most of the
clock_gettime syscall code (except for the copyout). Adjust all
corresponding syscalls to make use of it.

Move most clock_getres syscall code, except for coypout call, to a new
clock_getres1() function which can be used by emulations. Adjust all
clock_getres syscalls to now make of use it.

Convert linux/linux32/netbsd32 clock_settime syscalls, to use the
common clock_settime1() function.

branches: 1.29.2; 1.29.4;
Kill unreachable return statement.

branches: 1.28.2;
Update some syscalls that now needs compat50 timeval structure.

Fix stime(2) inverted copyin arguments.

s/syscallcarg/syscallarg/ in comments.

Add clock_nanosleep syscall.

Remove unused timeval_to_clock_t function, superseeded by CONVTCK
macro.

Make the emulations, exec formats, coredump, NFS, and the NFS server
into modules. By and large this commit:

- shuffles header files and ifdefs
- splits code out where necessary to be modular
- adds module glue for each of the components
- adds/replaces hooks for things that can be installed at runtime

Fix nanosleep(2) on 64-bit archs. Do not call native nanosleep
syscall, but rather a modified version the will take care of `struct
timespec' conversions.

branches: 1.21.2; 1.21.4;
make this compile again.

fix broken linux32_sys_times, copied from linux_sys_times.

branches: 1.19.2; 1.19.4; 1.19.8;
Merge proc::p_mutex and proc::p_smutex into a single adaptive mutex, since
we no longer need to guard against access from hardware interrupt handlers.

Additionally, if cloning a process with CLONE_SIGHAND, arrange to have the
child process share the parent's lock so that signal state may be kept in
sync. Partially addresses PR kern/37437.

branches: 1.18.2;
Add compat linux32 sys_clock_{getres,gettime,settime} syscalls.

Make rusage collection per-LWP and collate in the appropriate places.
cloned threads need a little bit more work but the locking needs to
be fixed first.

branches: 1.16.6;
Convert all the system call entry points from:
int foo(struct lwp *l, void *v, register_t *retval)
to:
int foo(struct lwp *l, const struct foo_args *uap, register_t *retval)
Fixup compat code to not write into 'uap' and (in some cases) to actually
pass a correctly formatted 'uap' structure with the right name to the
next routine.
A few 'compat' routines that just call standard ones have been deleted.
All the 'compat' code compiles (along with the kernels required to test
build it).
98% done by automated scripts.

branches: 1.15.4;
ANSIfy most of the function definitions in sys/compat (but not ndis).
All by the magic of sed ...

branches: 1.14.2;
Refactor time modification checks and place them in the secmodel code.

okay christos@

branches: 1.13.6; 1.13.8; 1.13.14;
Get rusage directly, not via 64bit linux call and stackgap.

Change the compat sys_[fl]utime code to not use the stackgap.

Change the way that emulations locate files within the emulation root to
avoid having to allocate space in the 'stackgap'
- which is very LWP unfriendly.
The additional code for non-emulation namei() is trivial, the reduction for
the emulations is massive.
The vnode for a processes emulation root is saved in the cwdi structure
during process exec.
If the emulation root the TRYEMULROOT flag are set, namei() will do an initial
search for absolute pathnames in the emulation root, if that fails it will
retry from the normal root.
".." at the emulation root will always go to the real root, even in the middle
of paths and when expanding symlinks.
Absolute symlinks found using absolute paths in the emulation root will be
relative to the emulation root (so /usr/lib/xxx.so -> /lib/xxx.so links
inside the emulation root don't need changing).
If the root of the emulation would be returned (for an emulation lookup), then
the real root is returned instead (matching the behaviour of emul_lookup,
but being a cheap comparison here) so that programs that scan "../.."
looking for the root dircetory don't loop forever.
The target for symbolic links is no longer mangled (it used to get the
CHECK_ALT_xxx() treatment, so could get /emul/xxx prepended).
CHECK_ALT_xxx() are no more. Most of the change is deleting them, and adding
TRYEMULROOT to the flags to NDINIT().
A lot of the emulation system call stubs could now be deleted.

Change all the NETBSD32PTR64(SCARG(uap, xxx))) to SCARG_P32(uap, xxx).

remove all the double (and triple) casts used to convert 32bit userspace
pointers to and from 64bit kernel pointers. Instead use the defines
NETBSD32PTR64(p32) to read a 32bit pointer and (the new) NETBSD32PTR32(p32,p64)
to write a 32bit pointer throughout.
The 32bit pointer is now a struct to enforce the above.
amd64 (with linux emul) and sparc64 will both compile (when the arch stuff
goes in soon), and amd64 still runs some i386 binaries.

branches: 1.8.2; 1.8.4; 1.8.6;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.

branches: 1.7.2;
Merge newlock2 to head.

Clean some KAUTH_GENERIC_ISSUSER usage in compat code.

branches: 1.5.4; 1.5.6; 1.5.8;
Use the LWP cached credentials where sane.

branches: 1.4.4;
include kauth.h
workaround a gcc4 warning.

branches: 1.3.2;
integrate kauth.

branches: 1.2.2; 1.2.4; 1.2.6; 1.2.8;
Fix compat_linux32 time(2) emulation

branches: 1.1.2;
Add initial (but unfinished) COMPAT_LINUX32 for amd64. This is good enough so
that the i386 license manager part of amd64 version of Fluent works.

While I'm here, add SysV IPC to COMPAT_LINUX/amd64

Add pselect6(2) system call to COMPAT_LINUX32

add support COMPAT_LINUX32 for aarch64

- Add compat_linux statx(2) syscall.
- The AT_EMPTY_PATH processing from the modification of
sys/compat/linux/common/linux_file64.c r1.63 has been separated, and made
common to linux_statat(), so that it can be used not only by
linux32_sys_fstatat64() but also by other *statat() variants.

add sigtimedwait support

branches: 1.14.8;
implement the following syscalls for linux32:
truncate64
ftruncate64
profil
ioperm
iopl
setdomainname
modify_ldt
statfs64
fstatfs64

note that iopl(), ioperm() and modify_ldt() just call
the respective 64-bit handlers, which don't do anything yet.

recent versions of linux (which we now claim to be) supply a BSD-style
d_type value in getdents() results, after the d_name field.
make our emulation do the same. fixes part of PR 43695.

many changes for COMPAT_LINUX:
- update the linux syscall table for each platform.
- support new-style (NPTL) linux pthreads on all platforms.
clone() with CLONE_THREAD uses 1 process with many LWPs
instead of separate processes.
- move the contents of sys__lwp_setprivate() into a new
lwp_setprivate() and use that everywhere.
- update linux_release[] and linux32_release[] to "2.6.18".
- adjust placement of emul fork/exec/exit hooks as needed
and adjust other emul code to match.
- convert all struct emul definitions to use named initializers.
- change the pid allocator to allow multiple pids to refer to the same proc.
- remove a few fields from struct proc that are no longer needed.
- disable the non-functional "vdso" code in linux32/amd64,
glibc works fine without it.
- fix a race in the futex code where we could miss a wakeup after
a requeue operation.
- redo futex locking to be a little more efficient.

branches: 1.11.4; 1.11.6;
Update some syscalls that now needs compat50 timeval structure.

Restore struct sysctl padding member name, which was damaged during
__unused removal on arguments.

Add getres{uid,gid} syscalls.

branches: 1.8.2;
Make linux32 getdents(2) do its own job instead of calling the
corresponding compat linux function, where struct dirent members types
differs.

branches: 1.7.4; 1.7.6; 1.7.10;
Add compat linux32 sys_clock_{getres,gettime,settime} syscalls.

branches: 1.6.6;
compat linux (and linux32) uid16 functions cleanup and fixes.

- Move uid16 functions to their own file linux_uid16.c, included by
needed archs (arm, i386 and m68k).
- Add new MI types linux_{u,g}id16_t.
- Add macros to handle linux_uid16_t and uid_t conversions.
- Add linux_sys_getres{uid,gid}16 syscalls, to fix an overflow with
bad sizes given to copyout when linux_sys_getres{uid,gid} are used.
- Update arm syscall table to use more uid16 functions.

Add old_uname syscall.

branches: 1.4.10; 1.4.16; 1.4.18; 1.4.22;
- Add compat amd64 linux32 statfs support, adapted from PR/35956.
- Move bsd_to_linux_statfs() function to its own file to be shared
between both linux compats.

ok by manu.

branches: 1.3.2; 1.3.4; 1.3.8; 1.3.10;
__unused removal on arguments; approved by core.

branches: 1.2.2;
Jumbo COMPAT_LINUX/COMPAT_LINUX32 bugfix, with the help of Nicolas Joly
- Fix shmat return value on amd64: it uses no black magic with retval[0]
- Fix integer overflows in sysinfo
- Implement sysinfo, mmap2, sched_getparam, sched_getscheduler, mremap,
and madvise in COMPAT_LINUX32
- Fix improper types used in setgroups16/getgroups16
- Implement mmap2 for COMPAT_LINUX32
- Ifdef debug messages by DEBUG_LINUX

branches: 1.1.2; 1.1.4; 1.1.10; 1.1.16; 1.1.20; 1.1.22;
Add initial (but unfinished) COMPAT_LINUX32 for amd64. This is good enough so
that the i386 license manager part of amd64 version of Fluent works.

While I'm here, add SysV IPC to COMPAT_LINUX/amd64

Make the emulations, exec formats, coredump, NFS, and the NFS server
into modules. By and large this commit:

- shuffles header files and ifdefs
- splits code out where necessary to be modular
- adds module glue for each of the components
- adds/replaces hooks for things that can be installed at runtime

branches: 1.1.2; 1.1.4; 1.1.6; 1.1.12; 1.1.14; 1.1.18; 1.1.24; 1.1.26;
compat linux (and linux32) uid16 functions cleanup and fixes.

- Move uid16 functions to their own file linux_uid16.c, included by
needed archs (arm, i386 and m68k).
- Add new MI types linux_{u,g}id16_t.
- Add macros to handle linux_uid16_t and uid_t conversions.
- Add linux_sys_getres{uid,gid}16 syscalls, to fix an overflow with
bad sizes given to copyout when linux_sys_getres{uid,gid} are used.
- Update arm syscall table to use more uid16 functions.

Add pselect6(2) system call to COMPAT_LINUX32

add support COMPAT_LINUX32 for aarch64

add dummy implementation of linux fallocate() which just returns EOPNOTSUPP;
this is needed so that glibc falls back to emulation and apps behaving
properly, since EOPNOTSUPP is a documented and expected return code, but
ENOSYS is not

right now there are no filesystems in NetBSD tree supporting the fallocate
VOP, so no point trying to map this to a native call

supposed to help with problem reported in
https://mail-index.netbsd.org/tech-kern/2019/11/03/msg025641.html

Avoid signed integer overflow when convering linux timeval to timespec

Linux accepts garbage as timeout and attempts to set it to something
meaningful. Instead of checking for valid ranges of usec, just convert
the type safely, regardless of what is inside it.

branches: 1.40.4;
Refactor pipe1() and correct a bug in sys_pipe2() (SYS_pipe2)

sys_pipe2() returns two integers (values), the 2nd one is a copy of the 2nd
file descriptor that lands in fildes[2]. This is a side effect of reusing
the code for sys_pipe() (SYS_pipe) and not cleaning it up.

The first returned value is (on success) 0.

Introduced a small refactoring in pipe1() that it does not operate over
retval[], but on an array int[2]. A user sets retval[] for pipe() when
desired and needed.

This refactoring touches compat code: netbsd32, linux, linux32.

Before the changes on NetBSD/amd64:

$ ktruss -i ./a.out
[...]
15131 1 a.out pipe2(0x7f7fff2e62b8, 0) = 0, 4
[...]

After the changes:

$ ktruss -i ./a.out
[...]
782 1 a.out pipe2(0x7f7fff97e850, 0) = 0
[...]

There should not be a visible change for current users.

Sponsored by <The NetBSD Foundation>

Cleanup pipe(2) flags, now that native handle them.

Simplify dup3 emulation to call dodup() directly instead of
sys_dup2()+fd_set_exclose(). While here, add some error conditions.

Fix pread/pwrite syscalls which need a 64bit offset argument.

branches: 1.36.2;
implement the *at() syscalls.
bring the unimplemented syscall list up to date.

branches: 1.35.4; 1.35.14; 1.35.18;
implement pipe2 and dup3

branches: 1.34.4; 1.34.6;
- selcommon/pollcommon: drop redundant l argument.
- Use cached curlwp->l_fd, instead of p->p_fd.
- Inline selscan/pollscan.

Make nice(2) return EPERM on error, not EACCES (from sys_setpriority).

Move the internal poll/select related API's to use timespec instead
of timeval (rides the uvm bump).

ansify function definitions

branches: 1.30.2;
Update some syscalls that now needs compat50 timeval structure.

s/syscallcarg/syscallarg/ in comments.

Cleanup; kill netbsd32_caddr_t use.

Small setres{uid,gid} cleanup; there is no need to use a specific
handling for -1 uid values.

Add getres{uid,gid} syscalls.

Make the emulations, exec formats, coredump, NFS, and the NFS server
into modules. By and large this commit:

- shuffles header files and ifdefs
- splits code out where necessary to be modular
- adds module glue for each of the components
- adds/replaces hooks for things that can be installed at runtime

branches: 1.24.2;
Compat linux/linux32 nice(2) fix. The syscall argument is an increment
to be added to the current nice value, not an absolute value to be
set.

Kill a few linux32 syscalls identical to netbsd32 ones.
No functional changes expected.

branches: 1.22.4; 1.22.6; 1.22.10;
Make linux32_sys_nice call sys_setpriority directly instead of
netbsd32_setpriority.

branches: 1.21.6;
pread/pwrite fixes: use SCARG_P32 to access pointers, and update
syscalls definitions.

Impletemt pread() and pwrite().

compat linux (and linux32) uid16 functions cleanup and fixes.

- Move uid16 functions to their own file linux_uid16.c, included by
needed archs (arm, i386 and m68k).
- Add new MI types linux_{u,g}id16_t.
- Add macros to handle linux_uid16_t and uid_t conversions.
- Add linux_sys_getres{uid,gid}16 syscalls, to fix an overflow with
bad sizes given to copyout when linux_sys_getres{uid,gid} are used.
- Update arm syscall table to use more uid16 functions.

- Fix setfsuid to return uid on success.
- Replace getfsuid, which never existed on Linux, by correct setfsgid
syscall.

ok by christos and dsl.

Convert all the system call entry points from:
int foo(struct lwp *l, void *v, register_t *retval)
to:
int foo(struct lwp *l, const struct foo_args *uap, register_t *retval)
Fixup compat code to not write into 'uap' and (in some cases) to actually
pass a correctly formatted 'uap' structure with the right name to the
next routine.
A few 'compat' routines that just call standard ones have been deleted.
All the 'compat' code compiles (along with the kernels required to test
build it).
98% done by automated scripts.

s/netbsd32_void */netbsd32_caddr_t/

branches: 1.15.4;
ANSIfy most of the function definitions in sys/compat (but not ndis).
All by the magic of sed ...

branches: 1.14.2;
Update, for linux_sys_readlink() removal.

Remove remaining 64-bit compat linux syscalls linux_sys_xxx() (except
for ones without arguments), and replace them by their 32-bit
equivalent linux32_sys_xxxx().

branches: 1.12.2;
compat linux32 syscalls cleanup.

With stackgap and CHECK_ALT_xxx removal, some linux32 and netbsd32
syscalls are now identical.
To avoid code duplication, remove the linux32 definition and use the
netbsd32 one (no functional change).

Add missing chown's syscalls family members.
This makes chown(1)/chgrp(1) works under compat linux32.

Problem reported and fix tested by Arto Huusko. Thanks.

branches: 1.10.6; 1.10.8; 1.10.12;
Remove a spare 'timeval' structure - makes this version of this function
match some of the other copies lurking in the world of compat.

Change the way that emulations locate files within the emulation root to
avoid having to allocate space in the 'stackgap'
- which is very LWP unfriendly.
The additional code for non-emulation namei() is trivial, the reduction for
the emulations is massive.
The vnode for a processes emulation root is saved in the cwdi structure
during process exec.
If the emulation root the TRYEMULROOT flag are set, namei() will do an initial
search for absolute pathnames in the emulation root, if that fails it will
retry from the normal root.
".." at the emulation root will always go to the real root, even in the middle
of paths and when expanding symlinks.
Absolute symlinks found using absolute paths in the emulation root will be
relative to the emulation root (so /usr/lib/xxx.so -> /lib/xxx.so links
inside the emulation root don't need changing).
If the root of the emulation would be returned (for an emulation lookup), then
the real root is returned instead (matching the behaviour of emul_lookup,
but being a cheap comparison here) so that programs that scan "../.."
looking for the root dircetory don't loop forever.
The target for symbolic links is no longer mangled (it used to get the
CHECK_ALT_xxx() treatment, so could get /emul/xxx prepended).
CHECK_ALT_xxx() are no more. Most of the change is deleting them, and adding
TRYEMULROOT to the flags to NDINIT().
A lot of the emulation system call stubs could now be deleted.

Fix select() behaviour with NULL timeout, to blocks indefinitely
instead of returning immediately.
Reported by Edgar Fu� on port-amd64@.

Change all the NETBSD32PTR64(SCARG(uap, xxx))) to SCARG_P32(uap, xxx).

branches: 1.6.2; 1.6.4; 1.6.6;
Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.

branches: 1.5.2;
Merge newlock2 to head.

branches: 1.4.4;
Jumbo COMPAT_LINUX/COMPAT_LINUX32 bugfix, with the help of Nicolas Joly
- Fix shmat return value on amd64: it uses no black magic with retval[0]
- Fix integer overflows in sysinfo
- Implement sysinfo, mmap2, sched_getparam, sched_getscheduler, mremap,
and madvise in COMPAT_LINUX32
- Fix improper types used in setgroups16/getgroups16
- Implement mmap2 for COMPAT_LINUX32
- Ifdef debug messages by DEBUG_LINUX