Add PDP_ENDIAN definition, required for makefs (v7fs).

(jdc)

Change 0x1229's product name from 82557 to 8255x.
82550, 82551, 82558 and 82559 share the same product id.
This change makes "pcictl dump" happy.

(msaitoh)

regen.

(msaitoh)

Change Intel 0x1229 product name from 82557 to 8255x

(msaitoh)

Reduce priority of syslog message if getpeername returns ENOTCONN.
PR bin/18934 by Greg A. Woods (with supplied fix).

(lukem)

don't delete signal from the debugger.

(christos)

remove 3 empty directories

(christos)

Remove code that was used to avoid register spills. setcontext(2) can change
the registers, so re-fetching will produce the wrong result for trace_exit().

(christos)

implement splraise/spllower

(jmcneill)

if pmap_page_activate fails, print the returned and expected addresses

(jmcneill)

pcb->errno to pcb->pcb_errno

(reinoud)

static + __dead

(joerg)

Try to preserve errno over traps and on context switches

(reinoud)

Rename err to nerrno

(reinoud)

Implement thunk_seterrno()

(reinoud)

Sprinkle one more static and a bunch of consts.

(joerg)

static + __dead

(joerg)

ANSIfy + static + __dead

(joerg)

ANSIfy + static + __dead

(joerg)

ANSIfy + static + __dead

(joerg)

ANSIfy + static + __dead

(joerg)

static + __dead

(joerg)

ANSIfy + static + __dead

(joerg)

ANSIfy + static + __dead

(joerg)

static + __dead

(joerg)

static + __dead

(joerg)

static + __dead

(joerg)

Use static, change deep if / else if / ... chains to switch(), apply
__dead, ANSIfy.

(joerg)

static + __dead

(joerg)

-thunk_makecontext_trapframe2go(ucontext_t *ucp, void *func, void *trapframe)
+thunk_makecontext_1(ucontext_t *ucp, void *func, void *arg)

(reinoud)

Space around | (style)

(reinoud)

PR# kern/45021: Please support /emul/linux/proc/version

Add /proc/version for procfs with -o linux. The version reported depends
on the emulation type of the calling process:

$ cat /proc/version
NetBSD version 5.99.55 (netbsd@localhost) (gcc version 4.1.3 20080704 prerelease (NetBSD nb2 20081120)) NetBSD 5.99.55 (GENERIC) #39: Sun Sep  4 09:10:05 EDT 2011

$ /emul/linux/bin/cat /proc/version
Linux version 2.6.18 (linux@localhost) (gcc version 4.1.3 20080704 prerelease (NetBSD nb2 20081120)) #0 Wed Mar 3 03:03:03 PST 2010

$ /emul/linux32/bin/cat /proc/version
Linux version 2.6.18 (linux32@localhost) (gcc version 4.1.3 20080704 prerelease (NetBSD nb2 20081120)) #0 Wed Mar 3 03:03:03 PST 2010

(jmcneill)

adjust sets for new zoneinfo code

(christos)

Don't attach GPIO framework if NGPIO = 0

(sborrill)

forgot to change one usage call.

(christos)

Split sigget into sigget() and siggetinfo(). When a signal comes from the
debugger (l->l_sigpendset == NULL), using siggetinfo() try to fetch the
siginfo information from l->l_sigpend and then from p->p_sigpend if it
was not found. This allows us to pass siginfo information for traps from
the debugger.

(christos)

getlabelusesmbr: initialize 'oldlenp' parameter to sysctlbyname

(jmcneill)

Disable gpio(4) which can now be built and used as modules.

(mbalmer)

mem_access_handler: panic if we receive sigsegv or sigbus with SI_NOINFO,
and when "peeing outside the box" print the offending va

(jmcneill)

Preset ih_pil.

(nakayama)

ih_pil is an interrupt level, not an interrupt mask.

(nakayama)

mention new tzcode and tzdata

(christos)

merge changes.

(christos)

Notable changes: Russia is now in perpetual summer.

Status:

Vendor Tag: ADO
Release Tags: TZDATA2011I

(christos)

Update current version of BIND and tz.

(taca)

bring in the 2011i changes.

(christos)

Protect new functions with #ifndef __LIBC12_SOURCE__ so that they are not
visibile in the libc12 compat environment.

(christos)

Use a mutex to protect updates to gnt_entries.  Fixes a KASSERT panic
with an MP kernel.

(mhitch)

Now that pr45327 is fixed, remove BROKEN_SIGINFO code.

(jmcneill)

PR/45327: Jared McNeill: ptrace: siginfo doesn't work with traced processes
When saving the signal in p->p_xstat, clear it from the pending mask, but
don't remove it from the siginfo queue, so that next time the debugger
delivers it, the original information is found.
When posting a signal from the debugger l->l_sigpendset is not set, so we
use the process pending signal and add it back to the process pending set.

(christos)

add a workaround for kern/45327

(jmcneill)

unbreak build on netbsd

(jmcneill)

Let NetBSD/usermode build & link on a Linux host:
- Need to add options CPU_HOST=i386 or CPU_HOST=amd64 for membar & atomic_ops
- Don't use MAP_* and PROT_* with thunk API since the kernel and host
  might not be the same; add THUNK_MAP_* and THUNK_PROT_* and translate them
- Add thunk_posix_memalign
- allocate mem_uvm with thunk_posix_memalign instead of thunk_malloc
- Fix thunk_mmap callers to always pass either THUNK_MAP_PRIVATE or
  THUNK_MAP_SHARED
- mkstemp on Linux requires exactly 6 "X" characters at the end of the
  template string, so add an X

(jmcneill)

Regenerate for Kingston DT101 G2

(mlelstv)

Correct vendor code for Kingston DT101 G2

(mlelstv)

Add SIGILL handler, doing nothing than printing out details.
Move the address check to be done later so pmap and uvm have their go first.

Also some minor cosmetic fixes

(reinoud)

Get this compiling (but not quite linking yet) on Linux

(jmcneill)

Instead of doing #include </usr/include/machine/..>, use the headers
that ship with the kernel sources as #include "../../$host/include/$hdr",
and auto-generate these ugly headers with a script.

(jmcneill)

Make sure that we initialize all accounting bytes.

(christos)

- minor KNF
- for system processes (0, 1) where fd_lastfile == -1, fail early.
- instead of printing the pcb connection address, print the pathname if
  possible for unix domain sockets.

(christos)

Mention that fstat -f will not list open unix sockets on that filesystem.

(christos)

if emulating the syscall, toggle insyscall as the child won't get stopped a
second time

(jmcneill)

move call of urkelvisor_init to main instead of pmap_bootstrap

(jmcneill)

Add usermode kernel supervisor, can be disabled with 'no options URKELVISOR'

(jmcneill)

Oops forgot to commit this one

(reinoud)

minor knf, no functional change.

(christos)

leave the original errno as it was. don't report EFBIG on all mmap failures.

(christos)

We need to process SA_STOP signals immediately, and not deliver them to
the process. Instead of re-structuring the code to do that, call issignal()
like before in that case. (tail -F /file^Zfg should not get interrupted).

(christos)

Instead of declaring our own err() which is different than the standard one,
and using it incorrectly in a few places because of confusion (does it print
errno or not?), declare two versions following the standard ones xerrx and,
xerr, and use those as appropriate, implementing them using them vwarn and
vwarnx.

(christos)

Add quirks for Kingston DT 101 to stop them giving HBA errors.

(asau)

Regenerate for Kingston Data Travel 101.

(asau)

Add two Kingston Data Travel 101 models.

(asau)

add stvii driver for gdium's onboard microcontroller
while there add comments on new(ly added) drivers

(macallan)

Report vmem(9) errors out-of-band so that we can use vmem(9) to manage
ranges that include the least and the greatest vmem_addr_t.  Update
vmem(9) uses throughout the kernel.  Slightly expand on the tests in
subr_vmem.c, which still pass.  I've been running a kernel with this
patch without any trouble.

(dyoung)

add new voyager driver, follow related changes from LOONGSON
now this builds again
thanks he for noticing

(macallan)

Add a target, subr_vmem, for building the vmem(9) test program.

(dyoung)

In the tags target, move x86 sources above xen includes, for more
helpful tags.

(dyoung)

Update vmem(9).  Changes to the implementation and usage are following,
soon.

(dyoung)

- use PT_STRINGS
- use __arraycount()

(christos)

Add support for PTRACE_FORK.
- add a field in struct proc to save the forker/forkee pid, and a flag.
- add 3 new ptrace calls: PT_GET_PROCESS_STATE, PT_GET_EVENT_MASK,
  PT_SET_EVENT_MASK
Add a PT_STRINGS constant so that we don't hard-code the list of ptrace
subcalls in other programs (kdump).

(christos)

add support for PRACE_FORK.

(christos)

Add support for PTRACE_FORK. NB: This does not (yet) work for vfork(), because:
1. When we vfork() PL_PPWAIT is set, and that makes us do regular disposition
  of the TRAP signal, and not indirect through the debugger.
2. The parent needs to keep running, so that the debugger can release it.
  Unfortunately, with vfork() we cannot release the parent because it will
  eventually core-dump since the parent and the child cannot run on the
  same address space.

(christos)

If the process is traced, resulting from a PTRACE_FORK inherited setting,
stop it right now.

XXX[1]: Cannot make this MI, because I cannot wrap child_return because there
is MD code that checks fun == child_return. I think it is better to have an
mi child_return() and add a cpu_child_return()?
XXX[2]: Why do we need to stop so early? Perhaps stopping just after exec
is better?

(christos)

meta_oodate should ignore makeDependfile

(sjg)

Revert

(reinoud)

Update trapframe and add system call switchframe

(reinoud)

-thunk_makecontext_trapframe2go(ucontext_t *ucp, void *func, void *trapframe)
+thunk_makecontext_1(ucontext_t *ucp, void (*func)(void), void *arg)

Create a more general prototype for makecontext() with one variable.

(reinoud)

Add note on how to posibly create a signal and send it to a userland program

(reinoud)

Bump SYMTAB_SPACE so that it fits again.

(he)

trylockowner is not needed anymore.

(christos)

- use the correct buffer size
- remove obsolete comment

(christos)

Disable the debug_fh code for now

(reinoud)

Terminate line with a '.', not the word 'done', for consistency with other
startup scripts.

(mbalmer)

Make sure error is not used unitialized when sc->sc_attach_busy is 0.

(mbalmer)

Add support for some fxp devices from FreeBSD and OpenBSD.
{Free,Open}BSD say that we have to do some work to make fxp
stable.

(msaitoh)

regen.

(msaitoh)

Add some fxp devices.

(msaitoh)

Clarify the history and add proper attribution.

(fair)

Make strnlen(3) accessible in kernel. ok christos@.

(jym)

fix the build for rumpserver.

(christos)

Allocate a color compatible VA for the page we are entering.

(matt)

Allocate a color compatible page to the VA we are going to use.

(matt)

Print syscall name on syscall exit (when using SYSCALL_DEBUG).
XXX this should use kernhist.

(matt)

Always make sure that kte_entry is aligned appropriately so register_t can
be stored there.  (not really a problem right now since time_t is now 64bit
bit and it just works but it's better to be explicit).

(matt)

Use a alternative dedicated signal stack for the sigsegv errors; the userland
stack like that of ld.elf_so can be too short to allow for such deep
stacktraces.

(reinoud)

Implement thunk_sigaltstack() so an alternative signal stack can be used

(reinoud)

Use the new UVM_KMF_COLORMATCH flag to get a congruent mappings of the user
buffer so we can use unmanaged mappings (pmap_kenter_pa/pmap_kremove).

(matt)

pcc cpp(1) does not understand the '-undef' and '-traditional' long form
arguments, use $(CC) -E instead of $(CPP) for these

(plunky)

Oops used the wrong page flag for referenced and modified check

(reinoud)

redo vfs_lookup.c rev.1.126.
when crossing a mount point, don't keep the parent vnode locked.
ie. don't lock a vnode while holding another vnode which belongs to a
different filesystem.  otherwise we propagate slowness (or deadness) of a
filesystem to another via vnode lock chain.

(yamt)

Hack register setup for __i386__. This now seems to start up the code a lot
further.

(reinoud)

Prevent thunk_mmap() to mmap a page outside the box; just as a precaution.

(reinoud)

Add bus_dma overrides. From dyoung

(christos)

Enhance debugging capabilities of NetBSD/usermode by allowing `trap.c' to
write a system image of the memory it traps. Its extremely dirty but at least
you can look at the data and code thats giving the trouble. Its memory or
state is NOT saved in the coredump of the NetBSD/usermode program.

(reinoud)

It seems mvsata_bio_ready() is actually needed, and it seems
to work better if we always poll for completion for the
duration of this function.

(jakllsch)

We occasionally will try to set up the multicast filter in a context that
kmem(9) is unusable (such as transmit timeout). Thus, put the filter table
on the stack instead, all 528 bytes of it.

(jakllsch)

Initialize MVGBE_MTU to 0, as suggested by the Orion databook.

If the platform firmware did not initialize this to 0 for us,
we were unable to transmit frames.

(jakllsch)

a driver for the gdium's ST7 microcontroller
so far it only reports AC and lid switch events to sysmon

(macallan)

some i2c devices we're going to deal with need a little delay after each
write. Also make sure the gpios we're using are outputs.

(macallan)

static + __dead

(joerg)

static + __dead

(joerg)

static + __dead

(joerg)

update set list for pcc-20110901 (libpcc is not required)

(plunky)

update build machinery for pcc-20110901

(plunky)

fix conflicts created by new import

(plunky)

Update pcc snapshot to 1.1.0.DEVEL 20110901

Continual development has happened since the previous 0.9.9
snapshot of 20100603 and the full commit log is too long to
include here, but can be seen at

http://pcc.ludd.ltu.se/fisheye/changelog/~date=2011-09-01T06%3A00%3A00/pcc

Many bugs were reported and fixed, along with reworks of the compiler
internals in several areas, notable milestones were

* more builtins, libpcc is no longer required on NetBSD
* improved gcc compatibility (eg gnu-inline semantics)
* NetBSD/VAX kernel compiles and boots
* using cpp as backend for lint now fully working
* PIC support seems complete
* NetBSD/i386 kernel can be compiled and booted
* better float/double/complex support
* amd64 support on OSX
* reworked Makefiles
* pcc 1.0.0 was release on April 1 2011
* several xasm improvements
* FreeBSD/amd64 system can be compiled with very few changes
* DragonFly fixes from downstream
* rewritten #pragma support
* TLS support added
* many regression tests
* rewrite of internal attribute handling

Status:

Vendor Tag: ragge
Release Tags: pcc-20110901

(plunky)

Make linux_sys_getdents64 fails with ENOTDIR instead of EINVAL, when fd
does not refer to a directory.

(njoly)

undo previous

(christos)

Simplify, --analyze gives the same settings as default

(joerg)

Update LLVM/Clang snapshot to r138875. Primarily to get the rewritten
temp file handling fixing boost-jam build.

(joerg)

fix typo.

(christos)

explain NULL casts

(christos)

Check for v_type before v_rdev because it is cheaper and safer.

(christos)

add a couple of asserts

(christos)

In pmap_unwire() fix an obvious editor mishap - enable kernel preemption
before returning, not the other way around.
Might fix PR kern/45137.

(martin)

Minor nit: avoid kpreempt_disable() if cold (as we return early and do
not reenable it; actually makes no difference but looks more balanced
this way)

(martin)

fix %cr3 init. from mhitch@, tested by riz@ & mhitch@

(cherry)

reinstate NULL cast by request, where the NULL was being passed as a vararg

(plunky)

Use UVM_KMF_COLORMATCH and switch to pmap_kenter_pa for vmapbuf/vunmapbuf.

(matt)

Forward some UVM from matt-nb5-mips64.  Add UVM_KMF_COLORMATCH flag.
When uvm_map gets passed UVM_FLAG_COLORMATCH, the align argument contains
the color of the starting address to be allocated (0..colormask).
When uvm_km_alloc is passed UVM_KMF_COLORMATCH (which can only be used with
UVM_KMF_VAONLY), the align argument contain the color of the starting address
to be allocated.
Change uvm_pagermapin to use this.  When mapping user pages in the kernel,
if colormatch is used with the color of the starting user page then the kernel
mapping will be congruent with the existing user mappings.

(matt)

add a small file forgotten in previous

(agc)

add bit arithmetic operations to eawk (implementation based on the gawk
man page). this provides:

and(x, y) (in C, x & y)
or(x, y) (in C, x | y)
xor(x, y) (in C, x ^ y)
lshift(x, n) (in C, x << n)
rshift(x, n) (in C, x >> n)
compl(x) (in C, ~x)

operations, with exactly the same calling conventions as gawk.

(agc)

especially on LE platforms, we need to be careful to set the size of
the format specifier in awk printf() for integer output.  this was
being done for 'd' and 'i' formats, but not for 'o', 'x', 'X' and 'u'
formats.  rectify that.

allows hex patterns like 0xffffff00 to be printed

(agc)

only attach the i2c bus if the gpio pins we're using are actually
configured as gpio
while there remove some accidentially left in debug code

(macallan)

For PT_SYSCALLEMU, no need to stop again on syscall exit.

(jmcneill)

Document PT_SYSCALLEMU

(jmcneill)

PR# kern/45312: ptrace: PT_SETREGS can't alter system calls

Add a new PT_SYSCALLEMU request that cancels the current syscall, for
use with PT_SYSCALL.

(jmcneill)

sigispending: simplify a little.

(rmind)

Refactor to fix some signedness problems.

(matt)

put back the byte swapping code, and remove the double byte swapping.

(christos)

genfs_do_directio: acquire the lock of page owner for now and fix PR/45177.
Will be revisited to avoid locking dance and be more efficient, e.g. we can
use unmanaged-mapping by allocating with colouring in mind.

(rmind)

Remove erroneous seq=ntohs(seq) introduced in rev. 1.74, which caused
icmp_seq numbers to increment by 256 instead of 1 on LE platforms.

(riz)

fix shadow warnings.
XXX: might be a good idea to kill global 'cpu' and 'version' variables.

(christos)

Add override members to x86_bus_dma_tag.

(dyoung)

for multi-cpu machines, cpu_hatch() straddles the init of
__stack_chk_guard, so ensure stack protection is disabled
on x86 architectures (I don't know about other archs)

(plunky)

extend __clang__ hack for __PCC__ also
(CMSG_SPACE() is not a constant)

(plunky)

Add stub implementations of bus_dma_tag_create()/bus_dma_tag_destroy().

(dyoung)

Add some definitions for bus_dma(9) overrides.

(dyoung)

fix reversed test.

(christos)

NULL does not need a cast

(plunky)

on xen we don't initialize memory segment maps, so no sparse dumps for now.

(christos)

turn kernel output green

(macallan)

add voyager, voyagerfb and drivers for Gdium's rtc and temperature sensor

(macallan)

Regen for dup3.

(njoly)

Add dup3 syscall support.

(njoly)