Explicitly state that WARN*() should be used only for recoverable kernel
issues/bugs and that it should not be used for any kind of invalid
external inputs or transient conditions.
Motivation: it's a very useful capability to be able to understand if a
particular kernel splat means a kernel bug or simply an invalid user-space
program. For the former one wants to notify kernel developers, while
notifying kernel developers for the latter is annoying. Even a kernel
developer may not know what to do with a WARNING in an unfamiliar
subsystem. This is especially critical for any automated testing systems
that may use panic_on_warn and mail kernel developers.
The clear separation also serves as an additional documentation: is it a
condition that must never occur because of additional checks/logic
elsewhere? or is it simply a check for invalid inputs or unfortunate
conditions?
Use of pr_err() for user messages also leads to better error messages.
"Something is wrong in file foo on line X" is not particularly useful
message for end user. pr_err() forces developers to write more meaningful
error messages for user.
As of now we are almost there. We are doing systematic kernel testing
with panic_on_warn and are not seeing massive amounts of false positives.
But every now and then another WARN on ENOMEM or invalid inputs pops up
and leads to a lengthy argument each time. The goal of this change is to
officially document the rules.
Link: http://lkml.kernel.org/r/20180620103716.61636-1-dvyukov@gmail.com
Signed-off-by: Dmitry Vyukov <dvyukov@google.com>
Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This is preparation for allowing CRASH_CORE to be enabled for any
architecture.
swapper_pg_dir is always either an array or a macro expanding to NULL.
In the latter case, VMCOREINFO_SYMBOL() won't work, as it tries to take
the address of the given symbol:
#define VMCOREINFO_SYMBOL(name) \
vmcoreinfo_append_str("SYMBOL(%s)=%lx\n", #name, (unsigned long)&name)
Instead, use VMCOREINFO_SYMBOL_ARRAY(), which uses the value:
#define VMCOREINFO_SYMBOL_ARRAY(name) \
vmcoreinfo_append_str("SYMBOL(%s)=%lx\n", #name, (unsigned long)name)
This is the same thing for the array case but isn't an error for the macro
case.
Link: http://lkml.kernel.org/r/c05f9781ec204f40fc96f95086e7b6de6a3eb2c3.1532563124.git.osandov@fb.com
Signed-off-by: Omar Sandoval <osandov@fb.com>
Cc: Alexey Dobriyan <adobriyan@gmail.com>
Cc: Bhupesh Sharma <bhsharma@redhat.com>
Cc: Eric Biederman <ebiederm@xmission.com>
Cc: James Morse <james.morse@arm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
There's a theoretical race condition that will cause /proc/kcore to miss
a memory hotplug event:
CPU0 CPU1
// hotplug event 1
kcore_need_update = 1
open_kcore() open_kcore()
kcore_update_ram() kcore_update_ram()
// Walk RAM // Walk RAM
__kcore_update_ram() __kcore_update_ram()
kcore_need_update = 0
// hotplug event 2
kcore_need_update = 1
kcore_need_update = 0
Note that CPU1 set up the RAM kcore entries with the state after hotplug
event 1 but cleared the flag for hotplug event 2. The RAM entries will
therefore be stale until there is another hotplug event.
This is an extremely unlikely sequence of events, but the fix makes the
synchronization saner, anyways: we serialize the entire update sequence,
which means that whoever clears the flag will always succeed in replacing
the kcore list.
Link: http://lkml.kernel.org/r/6106c509998779730c12400c1b996425df7d7089.1531953780.git.osandov@fb.com
Signed-off-by: Omar Sandoval <osandov@fb.com>
Cc: Alexey Dobriyan <adobriyan@gmail.com>
Cc: Bhupesh Sharma <bhsharma@redhat.com>
Cc: Eric Biederman <ebiederm@xmission.com>
Cc: James Morse <james.morse@arm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Patch series "/proc/kcore improvements", v4.
This series makes a few improvements to /proc/kcore. It fixes a couple of
small issues in v3 but is otherwise the same. Patches 1, 2, and 3 are
prep patches. Patch 4 is a fix/cleanup. Patch 5 is another prep patch.
Patches 6 and 7 are optimizations to ->read(). Patch 8 makes it possible
to enable CRASH_CORE on any architecture, which is needed for patch 9.
Patch 9 adds vmcoreinfo to /proc/kcore.
This patch (of 9):
kclist_add() is only called at init time, so there's no point in grabbing
any locks. We're also going to replace the rwlock with a rwsem, which we
don't want to try grabbing during early boot.
While we're here, mark kclist_add() with __init so that we'll get a
warning if it's called from non-init code.
Link: http://lkml.kernel.org/r/98208db1faf167aa8b08eebfa968d95c70527739.1531953780.git.osandov@fb.com
Signed-off-by: Omar Sandoval <osandov@fb.com>
Reviewed-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Bhupesh Sharma <bhsharma@redhat.com>
Tested-by: Bhupesh Sharma <bhsharma@redhat.com>
Cc: Alexey Dobriyan <adobriyan@gmail.com>
Cc: Bhupesh Sharma <bhsharma@redhat.com>
Cc: Eric Biederman <ebiederm@xmission.com>
Cc: James Morse <james.morse@arm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Code checks if write is done by current to its own attributes.
For that get/put pair is unnecessary as it can be done under RCU.
Note: rcu_read_unlock() can be done even earlier since pointer to a task
is not dereferenced. It depends if /proc code should look scary or not:
rcu_read_lock();
task = pid_task(...);
rcu_read_unlock();
if (!task)
return -ESRCH;
if (task != current)
return -EACCESS:
P.S.: rename "length" variable. Code like this
length = -EINVAL;
should not exist.
Link: http://lkml.kernel.org/r/20180627200218.GF18113@avx2
Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Reviewed-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Currently, NODEMASK_ALLOC allocates a nodemask_t with kmalloc when
NODES_SHIFT is higher than 8, otherwise it declares it within the stack.
The comment says that the reasoning behind this, is that nodemask_t will
be 256 bytes when NODES_SHIFT is higher than 8, but this is not true. For
example, NODES_SHIFT = 9 will give us a 64 bytes nodemask_t. Let us fix
up the comment for that.
Another thing is that it might make sense to let values lower than
128bytes be allocated in the stack. Although this all depends on the
depth of the stack (and this changes from function to function), I think
that 64 bytes is something we can easily afford. So we could even bump
the limit by 1 (from > 8 to > 9).
Link: http://lkml.kernel.org/r/20180820085516.9687-1-osalvador@techadventures.net
Signed-off-by: Oscar Salvador <osalvador@suse.de>
Reviewed-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The call to strlcpy in backing_dev_store is incorrect. It should take
the size of the destination buffer instead of the size of the source
buffer. Additionally, ignore the newline character (\n) when reading
the new file_name buffer. This makes it possible to set the backing_dev
as follows:
echo /dev/sdX > /sys/block/zram0/backing_dev
The reason it worked before was the fact that strlcpy() copies 'len - 1'
bytes, which is strlen(buf) - 1 in our case, so it accidentally didn't
copy the trailing new line symbol. Which also means that "echo -n
/dev/sdX" most likely was broken.
Signed-off-by: Peter Kalauskas <peskal@google.com>
Link: http://lkml.kernel.org/r/20180813061623.GC64836@rodete-desktop-imager.corp.google.com
Acked-by: Minchan Kim <minchan@kernel.org>
Reviewed-by: Sergey Senozhatsky <sergey.senozhatsky@gmail.com>
Cc: <stable@vger.kernel.org> [4.14+]
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Currently, percpu memory only exposes allocation and utilization
information via debugfs. This more or less is only really useful for
understanding the fragmentation and allocation information at a per-chunk
level with a few global counters. This is also gated behind a config.
BPF and cgroup, for example, have seen an increase in use causing
increased use of percpu memory. Let's make it easier for someone to
identify how much memory is being used.
This patch adds the "Percpu" stat to meminfo to more easily look up how
much percpu memory is in use. This number includes the cost for all
allocated backing pages and not just insight at the per a unit, per chunk
level. Metadata is excluded. I think excluding metadata is fair because
the backing memory scales with the numbere of cpus and can quickly
outweigh the metadata. It also makes this calculation light.
Link: http://lkml.kernel.org/r/20180807184723.74919-1-dennisszhou@gmail.com
Signed-off-by: Dennis Zhou <dennisszhou@gmail.com>
Acked-by: Tejun Heo <tj@kernel.org>
Acked-by: Roman Gushchin <guro@fb.com>
Reviewed-by: Andrew Morton <akpm@linux-foundation.org>
Acked-by: David Rientjes <rientjes@google.com>
Acked-by: Vlastimil Babka <vbabka@suse.cz>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Christoph Lameter <cl@linux.com>
Cc: Alexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
For some workloads an intervention from the OOM killer can be painful.
Killing a random task can bring the workload into an inconsistent state.
Historically, there are two common solutions for this
problem:
1) enabling panic_on_oom,
2) using a userspace daemon to monitor OOMs and kill
all outstanding processes.
Both approaches have their downsides: rebooting on each OOM is an obvious
waste of capacity, and handling all in userspace is tricky and requires a
userspace agent, which will monitor all cgroups for OOMs.
In most cases an in-kernel after-OOM cleaning-up mechanism can eliminate
the necessity of enabling panic_on_oom. Also, it can simplify the cgroup
management for userspace applications.
This commit introduces a new knob for cgroup v2 memory controller:
memory.oom.group. The knob determines whether the cgroup should be
treated as an indivisible workload by the OOM killer. If set, all tasks
belonging to the cgroup or to its descendants (if the memory cgroup is not
a leaf cgroup) are killed together or not at all.
To determine which cgroup has to be killed, we do traverse the cgroup
hierarchy from the victim task's cgroup up to the OOMing cgroup (or root)
and looking for the highest-level cgroup with memory.oom.group set.
Tasks with the OOM protection (oom_score_adj set to -1000) are treated as
an exception and are never killed.
This patch doesn't change the OOM victim selection algorithm.
Link: http://lkml.kernel.org/r/20180802003201.817-4-guro@fb.com
Signed-off-by: Roman Gushchin <guro@fb.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Cc: David Rientjes <rientjes@google.com>
Cc: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Cc: Tejun Heo <tj@kernel.org>
Cc: Vladimir Davydov <vdavydov.dev@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Patch series "introduce memory.oom.group", v2.
This is a tiny implementation of cgroup-aware OOM killer, which adds an
ability to kill a cgroup as a single unit and so guarantee the integrity
of the workload.
Although it has only a limited functionality in comparison to what now
resides in the mm tree (it doesn't change the victim task selection
algorithm, doesn't look at memory stas on cgroup level, etc), it's also
much simpler and more straightforward. So, hopefully, we can avoid having
long debates here, as we had with the full implementation.
As it doesn't prevent any futher development, and implements an useful and
complete feature, it looks as a sane way forward.
This patch (of 2):
oom_kill_process() consists of two logical parts: the first one is
responsible for considering task's children as a potential victim and
printing the debug information. The second half is responsible for
sending SIGKILL to all tasks sharing the mm struct with the given victim.
This commit splits oom_kill_process() with an intention to re-use the the
second half: __oom_kill_process().
The cgroup-aware OOM killer will kill multiple tasks belonging to the
victim cgroup. We don't need to print the debug information for the each
task, as well as play with task selection (considering task's children),
so we can't use the existing oom_kill_process().
Link: http://lkml.kernel.org/r/20171130152824.1591-2-guro@fb.com
Link: http://lkml.kernel.org/r/20180802003201.817-3-guro@fb.com
Signed-off-by: Roman Gushchin <guro@fb.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: David Rientjes <rientjes@google.com>
Cc: Vladimir Davydov <vdavydov.dev@gmail.com>
Cc: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Cc: David Rientjes <rientjes@google.com>
Cc: Tejun Heo <tj@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Currently, whenever a new node is created/re-used from the memhotplug
path, we call free_area_init_node()->free_area_init_core(). But there is
some code that we do not really need to run when we are coming from such
path.
free_area_init_core() performs the following actions:
1) Initializes pgdat internals, such as spinlock, waitqueues and more.
2) Account # nr_all_pages and # nr_kernel_pages. These values are used later on
when creating hash tables.
3) Account number of managed_pages per zone, substracting dma_reserved and
memmap pages.
4) Initializes some fields of the zone structure data
5) Calls init_currently_empty_zone to initialize all the freelists
6) Calls memmap_init to initialize all pages belonging to certain zone
When called from memhotplug path, free_area_init_core() only performs
actions #1 and #4.
Action #2 is pointless as the zones do not have any pages since either the
node was freed, or we are re-using it, eitherway all zones belonging to
this node should have 0 pages. For the same reason, action #3 results
always in manages_pages being 0.
Action #5 and #6 are performed later on when onlining the pages:
online_pages()->move_pfn_range_to_zone()->init_currently_empty_zone()
online_pages()->move_pfn_range_to_zone()->memmap_init_zone()
This patch does two things:
First, moves the node/zone initializtion to their own function, so it
allows us to create a small version of free_area_init_core, where we only
perform:
1) Initialization of pgdat internals, such as spinlock, waitqueues and more
4) Initialization of some fields of the zone structure data
These two functions are: pgdat_init_internals() and zone_init_internals().
The second thing this patch does, is to introduce
free_area_init_core_hotplug(), the memhotplug version of
free_area_init_core():
Currently, we call free_area_init_node() from the memhotplug path. In
there, we set some pgdat's fields, and call calculate_node_totalpages().
calculate_node_totalpages() calculates the # of pages the node has.
Since the node is either new, or we are re-using it, the zones belonging
to this node should not have any pages, so there is no point to calculate
this now.
Actually, we re-set these values to 0 later on with the calls to:
reset_node_managed_pages()
reset_node_present_pages()
The # of pages per node and the # of pages per zone will be calculated when
onlining the pages:
online_pages()->move_pfn_range()->move_pfn_range_to_zone()->resize_zone_range()
online_pages()->move_pfn_range()->move_pfn_range_to_zone()->resize_pgdat_range()
Also, since free_area_init_core/free_area_init_node will now only get called during early init, let us replace
__paginginit with __init, so their code gets freed up.
[osalvador@techadventures.net: fix section usage]
Link: http://lkml.kernel.org/r/20180731101752.GA473@techadventures.net
[osalvador@suse.de: v6]
Link: http://lkml.kernel.org/r/20180801122348.21588-6-osalvador@techadventures.net
Link: http://lkml.kernel.org/r/20180730101757.28058-5-osalvador@techadventures.net
Signed-off-by: Oscar Salvador <osalvador@suse.de>
Reviewed-by: Pavel Tatashin <pasha.tatashin@oracle.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Acked-by: Vlastimil Babka <vbabka@suse.cz>
Cc: Pasha Tatashin <Pavel.Tatashin@microsoft.com>
Cc: Aaron Lu <aaron.lu@intel.com>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: David Hildenbrand <david@redhat.com>
Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Cc: Mel Gorman <mgorman@techsingularity.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Currently cgroup-v1's memcg_stat_show traverses the memcg tree ~17 times
to collect the stats while cgroup-v2's memory_stat_show traverses the
memcg tree thrice. On a large machine, a couple thousand memcgs is very
normal and if the churn is high and memcgs stick around during to several
reasons, tens of thousands of nodes in memcg tree can exist. This patch
has refactored and shared the stat collection code between cgroup-v1 and
cgroup-v2 and has reduced the tree traversal to just one.
I ran a simple benchmark which reads the root_mem_cgroup's stat file
1000 times in the presense of 2500 memcgs on cgroup-v1. The results are:
Without the patch:
$ time ./read-root-stat-1000-times
real 0m1.663s
user 0m0.000s
sys 0m1.660s
With the patch:
$ time ./read-root-stat-1000-times
real 0m0.468s
user 0m0.000s
sys 0m0.467s
Link: http://lkml.kernel.org/r/20180724224635.143944-1-shakeelb@google.com
Signed-off-by: Shakeel Butt <shakeelb@google.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Vladimir Davydov <vdavydov.dev@gmail.com>
Cc: Greg Thelen <gthelen@google.com>
Cc: Bruce Merry <bmerry@ska.ac.za>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The Kconfig text for CONFIG_PAGE_POISONING doesn't mention that it has to
be enabled explicitly. This updates the documentation for that and adds a
note about CONFIG_PAGE_POISONING to the "page_poison" command line docs.
While here, change description of CONFIG_PAGE_POISONING_ZERO too, as it's
not "random" data, but rather the fixed debugging value that would be used
when not zeroing. Additionally removes a stray "bool" in the Kconfig.
Link: http://lkml.kernel.org/r/20180725223832.GA43733@beast
Signed-off-by: Kees Cook <keescook@chromium.org>
Reviewed-by: Andrew Morton <akpm@linux-foundation.org>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Laura Abbott <labbott@redhat.com>
Cc: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The /proc/pid/smaps_rollup file is currently implemented via the
m_start/m_next/m_stop seq_file iterators shared with the other maps files,
that iterate over vma's. However, the rollup file doesn't print anything
for each vma, only accumulate the stats.
There are some issues with the current code as reported in [1] - the
accumulated stats can get skewed if seq_file start()/stop() op is called
multiple times, if show() is called multiple times, and after seeks to
non-zero position.
Patch [1] fixed those within existing design, but I believe it is
fundamentally wrong to expose the vma iterators to the seq_file mechanism
when smaps_rollup shows logically a single set of values for the whole
address space.
This patch thus refactors the code to provide a single "value" at offset
0, with vma iteration to gather the stats done internally. This fixes the
situations where results are skewed, and simplifies the code, especially
in show_smap(), at the expense of somewhat less code reuse.
[1] https://marc.info/?l=linux-mm&m=151927723128134&w=2
[vbabka@suse.c: use seq_file infrastructure]
Link: http://lkml.kernel.org/r/bf4525b0-fd5b-4c4c-2cb3-adee3dd95a48@suse.cz
Link: http://lkml.kernel.org/r/20180723111933.15443-5-vbabka@suse.cz
Signed-off-by: Vlastimil Babka <vbabka@suse.cz>
Reported-by: Daniel Colascione <dancol@google.com>
Reviewed-by: Alexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Patch series "cleanups and refactor of /proc/pid/smaps*".
The recent regression in /proc/pid/smaps made me look more into the code.
Especially the issues with smaps_rollup reported in [1] as explained in
Patch 4, which fixes them by refactoring the code. Patches 2 and 3 are
preparations for that. Patch 1 is me realizing that there's a lot of
boilerplate left from times where we tried (unsuccessfuly) to mark thread
stacks in the output.
Originally I had also plans to rework the translation from
/proc/pid/*maps* file offsets to the internal structures. Now the offset
means "vma number", which is not really stable (vma's can come and go
between read() calls) and there's an extra caching of last vma's address.
My idea was that offsets would be interpreted directly as addresses, which
would also allow meaningful seeks (see the ugly seek_to_smaps_entry() in
tools/testing/selftests/vm/mlock2.h). However loff_t is (signed) long
long so that might be insufficient somewhere for the unsigned long
addresses.
So the result is fixed issues with skewed /proc/pid/smaps_rollup results,
simpler smaps code, and a lot of unused code removed.
[1] https://marc.info/?l=linux-mm&m=151927723128134&w=2
This patch (of 4):
Commit b76437579d ("procfs: mark thread stack correctly in
proc/<pid>/maps") introduced differences between /proc/PID/maps and
/proc/PID/task/TID/maps to mark thread stacks properly, and this was
also done for smaps and numa_maps. However it didn't work properly and
was ultimately removed by commit b18cb64ead ("fs/proc: Stop trying to
report thread stacks").
Now the is_pid parameter for the related show_*() functions is unused
and we can remove it together with wrapper functions and ops structures
that differ for PID and TID cases only in this parameter.
Link: http://lkml.kernel.org/r/20180723111933.15443-2-vbabka@suse.cz
Signed-off-by: Vlastimil Babka <vbabka@suse.cz>
Reviewed-by: Alexey Dobriyan <adobriyan@gmail.com>
Cc: Daniel Colascione <dancol@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Andrew has noticed some inconsistencies in oom_reap_task_mm. Notably
- Undocumented return value.
- comment "failed to reap part..." is misleading - sounds like it's
referring to something which happened in the past, is in fact
referring to something which might happen in the future.
- fails to call trace_finish_task_reaping() in one case
- code duplication.
- Increases mmap_sem hold time a little by moving
trace_finish_task_reaping() inside the locked region. So sue me ;)
- Sharing the finish: path means that the trace event won't
distinguish between the two sources of finishing.
Add a short explanation for the return value and fix the rest by
reorganizing the function a bit to have unified function exit paths.
Link: http://lkml.kernel.org/r/20180724141747.GP28386@dhcp22.suse.cz
Suggested-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Michal Hocko <mhocko@suse.com>
Reviewed-by: Andrew Morton <akpm@linux-foundation.org>
Cc: Tetsuo Handa <penguin-kernel@i-love.sakura.ne.jp>
Cc: David Rientjes <rientjes@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
