During early boot, when the scheduler hasn't really been fully set up,
we really can't do blocking allocations because with certain (dubious)
configurations the "might_resched()" calls can actually result in
scheduling events.
We could just make such users always use GFP_ATOMIC, but quite often the
code that does the allocation isn't really aware of the fact that the
scheduler isn't up yet, and forcing that kind of random knowledge on the
initialization code is just annoying and not good for anybody.
And we actually have a the 'gfp_allowed_mask' exactly for this reason:
it's just that the kernel init sequence happens to set it to allow
blocking allocations much too early.
So move the 'gfp_allowed_mask' initialization from 'start_kernel()'
(which is some of the earliest init code, and runs with preemption
disabled for good reasons) into 'kernel_init()'. kernel_init() is run
in the newly created thread that will become the 'init' process, as
opposed to the early startup code that runs within the context of what
will be the first idle thread.
So by the time we reach 'kernel_init()', we know that the scheduler must
be at least limping along, because we've already scheduled from the idle
thread into the init thread.
Reported-by: Steven Rostedt <rostedt@goodmis.org>
Cc: David Rientjes <rientjes@google.com>
Cc: stable@vger.kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Merge reason: We are going to queue up a dependent patch:
"perf tools: Move parse event automated tests to separated object"
That depends on:
commit e7c72d8
perf tools: Add 'G' and 'H' modifiers to event parsing
Conflicts:
tools/perf/builtin-stat.c
Conflicted with the recent 'perf_target' patches when checking the
result of perf_evsel open routines to see if a retry is needed to cope
with older kernels where the exclude guest/host perf_event_attr bits
were not used.
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Add a new internal Kconfig option UIDGID_CONVERTED that is true when the selected
Kconfig options have been converted to be user namespace safe, and guard
USER_NS and guard the UIDGID_STRICT_TYPE_CHECK options with it.
This keeps innocent kernel users from having the choice to enable
the user namespace in the cases where it is known not to work.
Most of the rest of the conversions are simple and straight forward but
their sheer number means it is good not to count on having them all done
and reviwed before thinking of merging this code.
Acked-by: Serge Hallyn <serge.hallyn@canonical.com>
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Pull the v3.5 RCU tree from Paul E. McKenney:
1) A set of improvements and fixes to the RCU_FAST_NO_HZ feature
(with more on the way for 3.6). Posted to LKML:
https://lkml.org/lkml/2012/4/23/324 (commits 1-3 and 5),
https://lkml.org/lkml/2012/4/16/611 (commit 4),
https://lkml.org/lkml/2012/4/30/390 (commit 6), and
https://lkml.org/lkml/2012/5/4/410 (commit 7, combined with
the other commits for the convenience of the tester).
2) Changes to make rcu_barrier() avoid disrupting execution of CPUs
that have no RCU callbacks. Posted to LKML:
https://lkml.org/lkml/2012/4/23/322.
3) A couple of commits that improve the efficiency of the interaction
between preemptible RCU and the scheduler, these two being all
that survived an abortive attempt to allow preemptible RCU's
__rcu_read_lock() to be inlined. The full set was posted to
LKML at https://lkml.org/lkml/2012/4/14/143, and the first and
third patches of that set remain.
4) Lai Jiangshan's algorithmic implementation of SRCU, which includes
call_srcu() and srcu_barrier(). A major feature of this new
implementation is that synchronize_srcu() no longer disturbs
the execution of other CPUs. This work is based on earlier
implementations by Peter Zijlstra and Paul E. McKenney. Posted to
LKML: https://lkml.org/lkml/2012/2/22/82.
5) A number of miscellaneous bug fixes and improvements which were
posted to LKML at: https://lkml.org/lkml/2012/4/23/353 with
subsequent updates posted to LKML.
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Currently, we'll try mounting any device who's major device number is
UNNAMED_MAJOR as NFS root. This would happen for non-NFS devices as
well (such as 9p devices) but it wouldn't cause any issues since
mounting the device as NFS would fail quickly and the code proceeded to
doing the proper mount:
[ 101.522716] VFS: Unable to mount root fs via NFS, trying floppy.
[ 101.534499] VFS: Mounted root (9p filesystem) on device 0:18.
Commit 6829a048102a ("NFS: Retry mounting NFSROOT") introduced retries
when mounting NFS root, which means that now we don't immediately fail
and instead it takes an additional 90+ seconds until we stop retrying,
which has revealed the issue this patch fixes.
This meant that it would take an additional 90 seconds to boot when
we're not using a device type which gets detected in order before NFS.
This patch modifies the NFS type check to require device type to be
'Root_NFS' instead of requiring the device to have an UNNAMED_MAJOR
major. This makes boot process cleaner since we now won't go through
the NFS mounting code at all when the device isn't an NFS root
("/dev/nfs").
Signed-off-by: Sasha Levin <levinsasha928@gmail.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The core branch is behind driver commits that we want to build
on for 3.5, hence I'm pulling in a later -rc.
Linux 3.4-rc5
Conflicts:
Documentation/feature-removal-schedule.txt
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Add a 3rd arg, named "doing", to unknown-options callbacks invoked
from parse_args(). The arg is passed as:
"Booting kernel" from start_kernel(),
initcall_level_names[i] from do_initcall_level(),
mod->name from load_module(), via parse_args(), parse_one()
parse_args() already has the "name" parameter, which is renamed to
"doing" to better reflect current uses 1,2 above. parse_args() passes
it to an altered parse_one(), which now passes it down into the
unknown option handler callbacks.
The mod->name will be needed to handle dyndbg for loadable modules,
since params passed by modprobe are not qualified (they do not have a
"$modname." prefix), and by the time the unknown-param callback is
called, the module name is not otherwise available.
Minor tweaks:
Add param-name to parse_one's pr_debug(), current message doesnt
identify the param being handled, add it.
Add a pr_info to print current level and level_name of the initcall,
and number of registered initcalls at that level. This adds 7 lines
to dmesg output, like:
initlevel:6=device, 172 registered initcalls
Drop "parameters" from initcall_level_names[], its unhelpful in the
pr_info() added above. This array is passed into parse_args() by
do_initcall_level().
CC: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Jim Cromie <jim.cromie@gmail.com>
Acked-by: Jason Baron <jbaron@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Commit 026cee0086 had the side-effect of dropping the '=' from
the unknown boot arguments that are passed to init as environment
variables. This is because parse_args() puts a NUL in the string
where the '=' was when it passes the "param" and "val" pointers
to the parsing subfunctions. Previously, unknown_bootoption() was
the last parse_args() subfunction to run, and it carefully put back
the '=' character. Now the ignore_unknown_bootoption() is the last
one to run, and it wasn't doing the necessary repair, so the
envp params ended up with the embedded NUL and were no longer
seen as valid environment variables by init.
Tested-by: Woody Suwalski <terraluna977@gmail.com>
Acked-by: Pawel Moll <pawel.moll@arm.com>
Signed-off-by: Chris Metcalf <cmetcalf@tilera.com>
Commit #0209f649 (rcu: limit rcu_node leaf-level fanout) set an upper
limit of 16 on the leaf-level fanout for the rcu_node tree. This was
needed to reduce lock contention that was induced by the synchronization
of scheduling-clock interrupts, which was in turn needed to improve
energy efficiency for moderate-sized lightly loaded servers.
However, reducing the leaf-level fanout means that there are more
leaf-level rcu_node structures in the tree, which in turn means that
RCU's grace-period initialization incurs more cache misses. This is
not a problem on moderate-sized servers with only a few tens of CPUs,
but becomes a major source of real-time latency spikes on systems with
many hundreds of CPUs. In addition, the workloads running on these large
systems tend to be CPU-bound, which eliminates the energy-efficiency
advantages of synchronizing scheduling-clock interrupts. Therefore,
these systems need maximal values for the rcu_node leaf-level fanout.
This commit addresses this problem by introducing a new kernel parameter
named RCU_FANOUT_LEAF that directly controls the leaf-level fanout.
This parameter defaults to 16 to handle the common case of a moderate
sized lightly loaded servers, but may be set higher on larger systems.
Reported-by: Mike Galbraith <efault@gmx.de>
Reported-by: Dimitri Sivanich <sivanich@sgi.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Linus Torvalds
Arnaldo Carvalho de Melo
Eric W. Biederman
Ingo Molnar
Thomas Gleixner
Sasha Levin
Greg Kroah-Hartman
Jens Axboe
Jim Cromie
Robert Richter
Chris Metcalf
Paul E. McKenney