Pull vfs update from Al Viro:
- big one - consolidation of descriptor-related logics; almost all of
that is moved to fs/file.c
(BTW, I'm seriously tempted to rename the result to fd.c. As it is,
we have a situation when file_table.c is about handling of struct
file and file.c is about handling of descriptor tables; the reasons
are historical - file_table.c used to be about a static array of
struct file we used to have way back).
A lot of stray ends got cleaned up and converted to saner primitives,
disgusting mess in android/binder.c is still disgusting, but at least
doesn't poke so much in descriptor table guts anymore. A bunch of
relatively minor races got fixed in process, plus an ext4 struct file
leak.
- related thing - fget_light() partially unuglified; see fdget() in
there (and yes, it generates the code as good as we used to have).
- also related - bits of Cyrill's procfs stuff that got entangled into
that work; _not_ all of it, just the initial move to fs/proc/fd.c and
switch of fdinfo to seq_file.
- Alex's fs/coredump.c spiltoff - the same story, had been easier to
take that commit than mess with conflicts. The rest is a separate
pile, this was just a mechanical code movement.
- a few misc patches all over the place. Not all for this cycle,
there'll be more (and quite a few currently sit in akpm's tree)."
Fix up trivial conflicts in the android binder driver, and some fairly
simple conflicts due to two different changes to the sock_alloc_file()
interface ("take descriptor handling from sock_alloc_file() to callers"
vs "net: Providing protocol type via system.sockprotoname xattr of
/proc/PID/fd entries" adding a dentry name to the socket)
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs: (72 commits)
MAX_LFS_FILESIZE should be a loff_t
compat: fs: Generic compat_sys_sendfile implementation
fs: push rcu_barrier() from deactivate_locked_super() to filesystems
btrfs: reada_extent doesn't need kref for refcount
coredump: move core dump functionality into its own file
coredump: prevent double-free on an error path in core dumper
usb/gadget: fix misannotations
fcntl: fix misannotations
ceph: don't abuse d_delete() on failure exits
hypfs: ->d_parent is never NULL or negative
vfs: delete surplus inode NULL check
switch simple cases of fget_light to fdget
new helpers: fdget()/fdput()
switch o2hb_region_dev_write() to fget_light()
proc_map_files_readdir(): don't bother with grabbing files
make get_file() return its argument
vhost_set_vring(): turn pollstart/pollstop into bool
switch prctl_set_mm_exe_file() to fget_light()
switch xfs_find_handle() to fget_light()
switch xfs_swapext() to fget_light()
...
Pull user namespace changes from Eric Biederman:
"This is a mostly modest set of changes to enable basic user namespace
support. This allows the code to code to compile with user namespaces
enabled and removes the assumption there is only the initial user
namespace. Everything is converted except for the most complex of the
filesystems: autofs4, 9p, afs, ceph, cifs, coda, fuse, gfs2, ncpfs,
nfs, ocfs2 and xfs as those patches need a bit more review.
The strategy is to push kuid_t and kgid_t values are far down into
subsystems and filesystems as reasonable. Leaving the make_kuid and
from_kuid operations to happen at the edge of userspace, as the values
come off the disk, and as the values come in from the network.
Letting compile type incompatible compile errors (present when user
namespaces are enabled) guide me to find the issues.
The most tricky areas have been the places where we had an implicit
union of uid and gid values and were storing them in an unsigned int.
Those places were converted into explicit unions. I made certain to
handle those places with simple trivial patches.
Out of that work I discovered we have generic interfaces for storing
quota by projid. I had never heard of the project identifiers before.
Adding full user namespace support for project identifiers accounts
for most of the code size growth in my git tree.
Ultimately there will be work to relax privlige checks from
"capable(FOO)" to "ns_capable(user_ns, FOO)" where it is safe allowing
root in a user names to do those things that today we only forbid to
non-root users because it will confuse suid root applications.
While I was pushing kuid_t and kgid_t changes deep into the audit code
I made a few other cleanups. I capitalized on the fact we process
netlink messages in the context of the message sender. I removed
usage of NETLINK_CRED, and started directly using current->tty.
Some of these patches have also made it into maintainer trees, with no
problems from identical code from different trees showing up in
linux-next.
After reading through all of this code I feel like I might be able to
win a game of kernel trivial pursuit."
Fix up some fairly trivial conflicts in netfilter uid/git logging code.
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ebiederm/user-namespace: (107 commits)
userns: Convert the ufs filesystem to use kuid/kgid where appropriate
userns: Convert the udf filesystem to use kuid/kgid where appropriate
userns: Convert ubifs to use kuid/kgid
userns: Convert squashfs to use kuid/kgid where appropriate
userns: Convert reiserfs to use kuid and kgid where appropriate
userns: Convert jfs to use kuid/kgid where appropriate
userns: Convert jffs2 to use kuid and kgid where appropriate
userns: Convert hpfs to use kuid and kgid where appropriate
userns: Convert btrfs to use kuid/kgid where appropriate
userns: Convert bfs to use kuid/kgid where appropriate
userns: Convert affs to use kuid/kgid wherwe appropriate
userns: On alpha modify linux_to_osf_stat to use convert from kuids and kgids
userns: On ia64 deal with current_uid and current_gid being kuid and kgid
userns: On ppc convert current_uid from a kuid before printing.
userns: Convert s390 getting uid and gid system calls to use kuid and kgid
userns: Convert s390 hypfs to use kuid and kgid where appropriate
userns: Convert binder ipc to use kuids
userns: Teach security_path_chown to take kuids and kgids
userns: Add user namespace support to IMA
userns: Convert EVM to deal with kuids and kgids in it's hmac computation
...
Pull s390 updates from Martin Schwidefsky:
"The main new feature is machine support for System zEC12 including
transactional memory, runtime instrumentation, support for scm block
devices via eadm subchannels, and support for CEX4 crypto cards.
In addition there are some nice improvements: bpf jit compiler, arch
backend for cmpxchg_double, relative exception table entries, dasd
partition detection independent from the dasd driver ioctls, and cpu
cache information in /proc/cpuinfo and /sys/device/cpu.
And last but not least a series of cleanup patches from Heiko."
Fix up trivial add-add conflict in arch/s390/Kconfig due to commit
b952741c80 ("cputime: Generalize CONFIG_VIRT_CPU_ACCOUNTING")
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux: (76 commits)
s390: update defconfig
s390/jump label,nss: let shared kernel support depend on !JUMP_LABEL
s390/disassembler: fix decoding of risblg instruction
s390/bpf,jit: add support for BPF_S_ANC_ALU_XOR_X instruction
s390/traps: move call to print_modules() out of show_regs()
s390/mm: mark free_initrd_mem() as __init
s390/dasd: check count address during online setting
drivers/s390/char/monreader.c: fix error return code
s390/cmpxchg,percpu: implement cmpxchg_double()
s390/percpu: implement this_cpu_add_return()
s390/percpu: implement this_cpu_xchg()
s390/kexec: remove CONFIG_KEXEC
s390/irq: use designated initializers for irq class array
s390: add uninitialized_var() to suppress false positive compiler warnings
s390/crashdump: move fill_cpu_elf_notes() prototype to header file
s390/process: add missing header include
s390/ptrace: add missing ifdef
s390/ipl,decrompressor: disable branch profiling
s390/perf_events: compile only for CONFIG_64BIT
s390/tape: remove even more tape block leftovers
...
Pull scheduler changes from Ingo Molnar:
"Continued quest to clean up and enhance the cputime code by Frederic
Weisbecker, in preparation for future tickless kernel features.
Other than that, smallish changes."
Fix up trivial conflicts due to additions next to each other in arch/{x86/}Kconfig
* 'sched-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (24 commits)
cputime: Make finegrained irqtime accounting generally available
cputime: Gather time/stats accounting config options into a single menu
ia64: Reuse system and user vtime accounting functions on task switch
ia64: Consolidate user vtime accounting
vtime: Consolidate system/idle context detection
cputime: Use a proper subsystem naming for vtime related APIs
sched: cpu_power: enable ARCH_POWER
sched/nohz: Clean up select_nohz_load_balancer()
sched: Fix load avg vs. cpu-hotplug
sched: Remove __ARCH_WANT_INTERRUPTS_ON_CTXSW
sched: Fix nohz_idle_balance()
sched: Remove useless code in yield_to()
sched: Add time unit suffix to sched sysctl knobs
sched/debug: Limit sd->*_idx range on sysctl
sched: Remove AFFINE_WAKEUPS feature flag
s390: Remove leftover account_tick_vtime() header
cputime: Consolidate vtime handling on context switch
sched: Move cputime code to its own file
cputime: Generalize CONFIG_VIRT_CPU_ACCOUNTING
tile: Remove SD_PREFER_LOCAL leftover
...
We can have either shared kernel or jump label support, but not both.
If a kernel gets IPL'ed from an NSS it's not possible to patch the
text segment, since it's read-only.
Therefore any static branches cannot be updated. So we need to make
sure that shared kernel support is disabled if jump label support
is enabled.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Add support for new BPF_S_ANC_ALU_XOR_X instruction which got added
with ffe06c17 "filter: add XOR operation".
s390 version of 4bfaddf1 "x86 bpf_jit: support BPF_S_ANC_ALU_XOR_X instruction".
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Same as 0fa0e2f0 "x86: Move call to print_modules() out of show_regs()".
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Same as 0d26d1d8 "x86/mm: Mark free_initrd_mem() as __init".
In addition also add the __init annotation to setup_zero_pages().
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
The generic variant has a local_irq_save/restore pair which is quite
expensive. It is sufficient to disable preemption, which is a no-op
with !CONFIG_PREEMPT and then use the regular xchg macro.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Since "Kconfig: split the s390 base menu" CONFIG_KEXEC gets always selected.
Therefore there is no point in keeping CONFIG_KEXEC anywhere.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Use designated initializers for the irq class array in irq.c so
it's always guaranteed that the order of elements is equal to
their corresponding parts in irq.h.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Get rid of these:
arch/s390/kernel/smp.c:134:19: warning: ‘status’ may be used uninitialized in this function [-Wuninitialized]
arch/s390/mm/pgtable.c:641:10: warning: ‘table’ may be used uninitialized in this function [-Wuninitialized]
arch/s390/mm/pgtable.c:644:12: warning: ‘page’ may be used uninitialized in this function [-Wuninitialized]
drivers/s390/cio/cio.c:1037:14: warning: ‘schid’ may be used uninitialized in this function [-Wuninitialized]
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Move fill_cpu_elf_notes() prototype to header file.
This way we get compile errors if e.g. the number of function
parameters get changed.
Otherwise it's possible to change just the definition and everything
else still compiles fine, but the result is broken code.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Add appropriate header file:
arch/s390/kernel/process.c:327:15: warning: symbol 'arch_align_stack' was not declared.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
if (MACHINE_HAS_TE) translates to if (0) on !CONFIG_64BIT however the
compiler still warns about invalid shifts within non-reachable code.
So add an explicit ifdef to get rid of this warning:
arch/s390/kernel/ptrace.c: In function ‘update_per_regs’:
arch/s390/kernel/ptrace.c:63:4: warning: left shift count >= width of type [enabled by default]
arch/s390/kernel/ptrace.c:65:4: warning: left shift count >= width of type [enabled by default]
Reported-by: Fengguang Wu <fengguang.wu@intel.com>
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
The early kernel decrompressing code can't call into the kernel
in order to profile branches. Fixes this link error:
arch/s390/boot/compressed/misc.o: In function `decompress_kernel':
misc.c:(.text+0x9a6): undefined reference to `ftrace_likely_update'
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
The whole hardware support is only available in zArch mode.
Fixes also this compile warning:
arch/s390/kernel/perf_cpum_cf.c: In function ‘cpumf_pmu_init’:
arch/s390/kernel/perf_cpum_cf.c:670:2: warning: left shift count >= width of type [enabled by default]
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Use for_each_set_bit() to simplify the code.
spatch with a semantic match is used to found this.
(http://coccinelle.lip6.fr/)
Signed-off-by: Wei Yongjun <yongjun_wei@trendmicro.com.cn>
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Change return value of appldata_asm() to -EOPNOTSUPP in case of an
error. The return value was only used internally and not passed to
user space.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Returning -ENOSYS on kexec_load() is a bad idea since user space cannot
tell if the system call is not implmented or if it failed.
Use -EOPNOTSUPP in case somebody tries a kexec_load on a NSS image based
kernel instead.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Change -ENOSYS to -EOPNOTSUPP. Return value is used only internally.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Change return code handling of the stsi() function:
In case function code 0 was specified the return value is the
current configuration level (already shifted). That way all
the code that actually copied the stsi_0() function can go
away.
Otherwise the return value is 0 (success) or negative to
indicate an error (currently only -EOPNOTSUPP).
Also stsi() is no longer an inline function. The function is
not performance critical, but every caller would generate an
exception table entry for this function.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
This is the s390 port of 70627654 "x86, extable: Switch to relative
exception table entries".
Reduces the size of our exception tables by 50% on 64 bit builds.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Remove the spinlock from struct scm_device. drvdata and attributes
are guarded via device_lock.
Signed-off-by: Sebastian Ott <sebott@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
The cpu polarization member is the only per cpu state that is not part
of the pcpu structure. So add it there and have everything in one place.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
The maximum nesting of the cpu topology is evaluated when /proc/sysinfo
is the first time read. This happens without a lock and a concurrent
reader on a different cpu can see and use an invalid intermediate value.
Besides the fact that this race is quite unlikely the worst thing that
could happen is that /proc/sysinfo would contain bogus information about
the machine's cpu topology.
Nevertheless this should be fixed. So move the detection code to the
early machine detection code and since now the value is early available
use it in the topology code as well.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
This patch introduces the __get_user_pages_fast() function on s390,
which will be needed with CONFIG_TRANSPARENT_HUGEPAGE and futex.
Signed-off-by: Gerald Schaefer <gerald.schaefer@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Add a couple of missing fields that were introduced with z196.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
The current proc implementation of the /proc/sysinfo file writes all
informations contained in all system information blocks to a single
page.
This is done by calling sprintf all the time in the expectation that
everything will fit into a single page. This however is not necessarily
true if the configuration of a machine is very large.
So convert /proc/sysinfo to avoid writing into random memory regions.
For readability reasons a couple of lines are longer than 80 characters.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Any change to sysinfo.h causes a whole kernel recompile since sysinfo.h is
included by topology.h, which again is used nearly everywhere.
So remove that include and add a forward declaration instead.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Shorten the code for addressing mode initialization. Also add missing
__init annotations, since this code is only used during kernel initialization.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Renaming the globally visible variable "user_mode" to "addressing_mode" in
order to fix a name clash was not a good idea. (Commit 37fe1d73 "s390/mm:
rename user_mode variable to addressing_mode")
Looking at the code after a couple of weeks one thinks: addressing mode of
what?
So rename the variable again. This time to s390_user_mode. Which hopefully
makes more sense.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Change the default addressing mode so that user space runs in primary space
and the kernel runs in home space.
In addition remove the "switch_amode" kernel parameter so all users who
already specified they want the new default behaviour will stay in the
"switched" mode instead of in the opposite they intended.
If there is a need to switch addressing modes, this can be done with the
"user_mode" kernel parameter: user_mode=home
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
When a sigp instruction is issued it may store a status. This status is
currently stored in a per cpu field of the target cpu.
If multiple cpus issue a sigp instruction with the same target cpu
and a status is stored the result is not necessarily as expected.
Currently this is not an issue:
- on cpu hotplug no sigps, except "restart" and "sense" are sent to the
target cpu.
- on external call we don't look at the status if it is stored
- on sense running the condition code "status stored" is sufficient to
tell if a cpu is running or not
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Add a line for each cpu cache to /proc/cpuinfo.
Since we only have information of private cpu caches in sysfs we
add a line for each cpu cache in /proc/cpuinfo which will also
contain information about shared caches.
For a z196 machine /proc/cpuinfo now looks like:
vendor_id : IBM/S390
bogomips per cpu: 14367.00
features : esan3 zarch stfle msa ldisp eimm dfp etf3eh highgprs
cache0 : level=1 type=Data scope=Private size=64K line_size=256 associativity=4
cache1 : level=1 type=Instruction scope=Private size=128K line_size=256 associativity=8
cache2 : level=2 type=Unified scope=Private size=1536K line_size=256 associativity=12
cache3 : level=3 type=Unified scope=Shared size=24576K line_size=256 associativity=12
cache4 : level=4 type=Unified scope=Shared size=196608K line_size=256 associativity=24
processor 0: version = FF, identification = 000123, machine = 2817
processor 1: version = FF, identification = 100123, machine = 2817
processor 2: version = FF, identification = 200123, machine = 2817
processor 3: version = FF, identification = 200123, machine = 2817
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Allow user-space processes to use transactional execution (TX).
If the TX facility is available user space programs can use
transactions for fine-grained serialization based on the data
objects that are referenced during a transaction. This is
useful for lockless data structures and speculative compiler
optimizations.
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Allow user-space threads to use runtime instrumentation (RI). To enable RI
for a thread there is a new s390 specific system call, sys_s390_runtime_instr,
that takes as parameter a realtime signal number. If the RI facility is
available the system call sets up a control block for the calling thread with
the appropriate permissions for the thread to modify the control block.
The user-space thread can then use the store and modify RI instructions to
alter the control block and start/stop the instrumentation via RION/RIOFF.
If the user specified program buffer runs full RI triggers an external
interrupt. The external interrupt is translated to a real-time signal that
is delivered to the thread that enabled RI on that CPU. The number of
the real-time signal is the number specified in the RI system call. So,
user-space can select any available real-time signal number in case the
application itself uses real-time signals for other purposes.
The kernel saves the RI control blocks on task switch only if the running
thread was enabled for RI. Therefore, the performance impact on task switch
should be negligible if RI is not used.
RI is only enabled for user-space mode and is disabled for the supervisor
state.
Reviewed-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Jan Glauber <jang@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Add support for EADM interrupt statistics in /proc/interrupts.
Signed-off-by: Sebastian Ott <sebott@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
This driver allows usage of EADM subchannels. EADM subchannels
act as a communication vehicle for SCM increments.
Signed-off-by: Sebastian Ott <sebott@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Detect an scm change notification in store event information.
Update affected scm devices and notify their drivers.
Signed-off-by: Sebastian Ott <sebott@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Bus driver to manage Storage Class Memory.
Signed-off-by: Sebastian Ott <sebott@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Add structures to be used by the eadm subchannel driver.
Signed-off-by: Sebastian Ott <sebott@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Add the eadm facility bits to the css characteristics and move
them to a new header.
Signed-off-by: Sebastian Ott <sebott@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
