Greg Thelen [Fri, 17 Aug 2018 22:46:47 +0000 (15:46 -0700)]
mm/page-writeback.c: update stale account_page_redirty() comment
Commit 93f78d882865 ("writeback: move backing_dev_info->bdi_stat[] into
bdi_writeback") replaced BDI_DIRTIED with WB_DIRTIED in
account_page_redirty(). Update comment to track that change.
Shakeel Butt [Fri, 17 Aug 2018 22:46:44 +0000 (15:46 -0700)]
fs, mm: account buffer_head to kmemcg
The buffer_head can consume a significant amount of system memory and is
directly related to the amount of page cache. In our production
environment we have observed that a lot of machines are spending a
significant amount of memory as buffer_head and can not be left as
system memory overhead.
Charging buffer_head is not as simple as adding __GFP_ACCOUNT to the
allocation. The buffer_heads can be allocated in a memcg different from
the memcg of the page for which buffer_heads are being allocated. One
concrete example is memory reclaim. The reclaim can trigger I/O of
pages of any memcg on the system. So, the right way to charge
buffer_head is to extract the memcg from the page for which buffer_heads
are being allocated and then use targeted memcg charging API.
Shakeel Butt [Fri, 17 Aug 2018 22:46:39 +0000 (15:46 -0700)]
fs: fsnotify: account fsnotify metadata to kmemcg
Patch series "Directed kmem charging", v8.
The Linux kernel's memory cgroup allows limiting the memory usage of the
jobs running on the system to provide isolation between the jobs. All
the kernel memory allocated in the context of the job and marked with
__GFP_ACCOUNT will also be included in the memory usage and be limited
by the job's limit.
The kernel memory can only be charged to the memcg of the process in
whose context kernel memory was allocated. However there are cases
where the allocated kernel memory should be charged to the memcg
different from the current processes's memcg. This patch series
contains two such concrete use-cases i.e. fsnotify and buffer_head.
The fsnotify event objects can consume a lot of system memory for large
or unlimited queues if there is either no or slow listener. The events
are allocated in the context of the event producer. However they should
be charged to the event consumer. Similarly the buffer_head objects can
be allocated in a memcg different from the memcg of the page for which
buffer_head objects are being allocated.
To solve this issue, this patch series introduces mechanism to charge
kernel memory to a given memcg. In case of fsnotify events, the memcg
of the consumer can be used for charging and for buffer_head, the memcg
of the page can be charged. For directed charging, the caller can use
the scope API memalloc_[un]use_memcg() to specify the memcg to charge
for all the __GFP_ACCOUNT allocations within the scope.
This patch (of 2):
A lot of memory can be consumed by the events generated for the huge or
unlimited queues if there is either no or slow listener. This can cause
system level memory pressure or OOMs. So, it's better to account the
fsnotify kmem caches to the memcg of the listener.
However the listener can be in a different memcg than the memcg of the
producer and these allocations happen in the context of the event
producer. This patch introduces remote memcg charging API which the
producer can use to charge the allocations to the memcg of the listener.
There are seven fsnotify kmem caches and among them allocations from
dnotify_struct_cache, dnotify_mark_cache, fanotify_mark_cache and
inotify_inode_mark_cachep happens in the context of syscall from the
listener. So, SLAB_ACCOUNT is enough for these caches.
The objects from fsnotify_mark_connector_cachep are not accounted as
they are small compared to the notification mark or events and it is
unclear whom to account connector to since it is shared by all events
attached to the inode.
The allocations from the event caches happen in the context of the event
producer. For such caches we will need to remote charge the allocations
to the listener's memcg. Thus we save the memcg reference in the
fsnotify_group structure of the listener.
This patch has also moved the members of fsnotify_group to keep the size
same, at least for 64 bit build, even with additional member by filling
the holes.
mm: provide a fallback for PAGE_KERNEL_EXEC for architectures
Some architectures just don't have PAGE_KERNEL_EXEC. The mm/nommu.c and
mm/vmalloc.c code have been using PAGE_KERNEL as a fallback for years.
Move this fallback to asm-generic.
mm: provide a fallback for PAGE_KERNEL_RO for architectures
Some architectures do not define certain PAGE_KERNEL_* flags, this is
either because:
a) The way to implement some of these flags is *not yet ported*, or
b) The architecture *has no way* to describe them
Over time we have accumulated a few PAGE_KERNEL_* fallback workarounds
for architectures in the kernel which do not define them using
*relatively safe* equivalents. Move these scattered fallback hacks into
asm-generic.
We start off with PAGE_KERNEL_RO using PAGE_KERNEL as a fallback. This
has been in place on the firmware loader for years. Move the fallback
into the respective asm-generic header.
Oscar Salvador [Fri, 17 Aug 2018 22:46:22 +0000 (15:46 -0700)]
mm/memory_hotplug.c: make register_mem_sect_under_node() a callback of walk_memory_range()
link_mem_sections() and walk_memory_range() share most of the code, so
we can use convert link_mem_sections() into a dummy function that calls
walk_memory_range() with a callback to register_mem_sect_under_node().
This patch converts register_mem_sect_under_node() in order to match a
walk_memory_range's callback, getting rid of the check_nid argument and
checking instead if the system is still boothing, since we only have to
check for the nid if the system is in such state.
When hotplugging memory, it is possible that two calls are being made to
register_mem_sect_under_node().
One comes from __add_section()->hotplug_memory_register() and the other
from add_memory_resource()->link_mem_sections() if we had to register a
new node.
In case we had to register a new node, hotplug_memory_register() will
only handle/allocate the memory_block's since
register_mem_sect_under_node() will return right away because the node
it is not online yet.
I think it is better if we leave hotplug_memory_register() to
handle/allocate only memory_block's and make link_mem_sections() to call
register_mem_sect_under_node().
So this patch removes the call to register_mem_sect_under_node() from
hotplug_memory_register(), and moves the call to link_mem_sections() out
of the condition, so it will always be called. In this way we only have
one place where the memory sections are registered.
Oscar Salvador [Fri, 17 Aug 2018 22:46:15 +0000 (15:46 -0700)]
mm/memory_hotplug.c: make add_memory_resource use __try_online_node
This is a small cleanup for the memhotplug code. A lot more could be
done, but it is better to start somewhere. I tried to unify/remove
duplicated code.
The following is what this patchset does:
1) add_memory_resource() has code to allocate a node in case it was
offline. Since try_online_node has some code for that as well, I just
made add_memory_resource() to use that so we can remove duplicated
code.. This is better explained in patch 1/4.
2) register_mem_sect_under_node() will be called only from
link_mem_sections()
3) Make register_mem_sect_under_node() a callback of
walk_memory_range()
4) Drop unnecessary checks from register_mem_sect_under_node()
I have done some tests and I could not see anything broken because of
this patchset.
add_memory_resource() contains code to allocate a new node in case it is
necessary. Since try_online_node() also has some code for this purpose,
let us make use of that and remove duplicate code.
This introduces __try_online_node(), which is called by
add_memory_resource() and try_online_node(). __try_online_node() has
two new parameters, start_addr of the node, and if the node should be
onlined and registered right away. This is always wanted if we are
calling from do_cpu_up(), but not when we are calling from memhotplug
code. Nothing changes from the point of view of the users of
try_online_node(), since try_online_node passes start_addr=0 and
online_node=true to __try_online_node().
mm: workingset: make shadow_lru_isolate() use locking suffix
shadow_lru_isolate() disables interrupts and acquires a lock. It could
use spin_lock_irq() instead. It also uses local_irq_enable() while it
could use spin_unlock_irq()/xa_unlock_irq().
Use proper suffix for lock/unlock in order to enable/disable interrupts
during release/acquire of a lock.
mm: workingset: remove local_irq_disable() from count_shadow_nodes()
Patch series "mm: use irq locking suffix instead local_irq_disable()".
A small series which avoids using local_irq_disable()/local_irq_enable()
but instead does spin_lock_irq()/spin_unlock_irq() so it is within the
context of the lock which it belongs to. Patch #1 is a cleanup where
local_irq_.*() remained after the lock was removed.
This patch (of 2):
In 0c7c1bed7e13 ("mm: make counting of list_lru_one::nr_items lockless")
the
spin_lock(&nlru->lock);
statement was replaced with
rcu_read_lock();
in __list_lru_count_one(). The comment in count_shadow_nodes() says
that the local_irq_disable() is required because the lock must be
acquired with disabled interrupts and (spin_lock()) does not do so.
Since the lock is replaced with rcu_read_lock() the local_irq_disable()
is no longer needed. The code path is
Michal Hocko [Fri, 17 Aug 2018 22:46:01 +0000 (15:46 -0700)]
mm: drop VM_BUG_ON from __get_free_pages
There is no real reason to blow up just because the caller doesn't know
that __get_free_pages cannot return highmem pages. Simply fix that up
silently. Even if we have some confused users such a fixup will not be
harmful.
Huang Ying [Fri, 17 Aug 2018 22:45:57 +0000 (15:45 -0700)]
mm, hugetlbfs: pass fault address to cow handler
This is to take better advantage of the general huge page copying
optimization. Where, the target subpage will be copied last to avoid
the cache lines of target subpage to be evicted when copying other
subpages. This works better if the address of the target subpage is
available when copying huge page. So hugetlbfs page fault handlers are
changed to pass that information to hugetlb_cow(). This will benefit
workloads which don't access the begin of the hugetlbfs huge page after
the page fault under heavy cache contention.
Huang Ying [Fri, 17 Aug 2018 22:45:53 +0000 (15:45 -0700)]
mm, hugetlbfs: rename address to haddr in hugetlb_cow()
To take better advantage of general huge page copying optimization, the
target subpage address will be passed to hugetlb_cow(), then
copy_user_huge_page(). So we will use both target subpage address and
huge page size aligned address in hugetlb_cow(). To distinguish between
them, "haddr" is used for huge page size aligned address to be
consistent with Transparent Huge Page naming convention.
Now, only huge page size aligned address is used in hugetlb_cow(), so
the "address" is renamed to "haddr" in hugetlb_cow() in this patch.
Next patch will use target subpage address in hugetlb_cow() too.
The patch is just code cleanup without any functionality changes.
Huang Ying [Fri, 17 Aug 2018 22:45:49 +0000 (15:45 -0700)]
mm, huge page: copy target sub-page last when copy huge page
Huge page helps to reduce TLB miss rate, but it has higher cache
footprint, sometimes this may cause some issue. For example, when
copying huge page on x86_64 platform, the cache footprint is 4M. But on
a Xeon E5 v3 2699 CPU, there are 18 cores, 36 threads, and only 45M LLC
(last level cache). That is, in average, there are 2.5M LLC for each
core and 1.25M LLC for each thread.
If the cache contention is heavy when copying the huge page, and we copy
the huge page from the begin to the end, it is possible that the begin
of huge page is evicted from the cache after we finishing copying the
end of the huge page. And it is possible for the application to access
the begin of the huge page after copying the huge page.
In c79b57e462b5d ("mm: hugetlb: clear target sub-page last when clearing
huge page"), to keep the cache lines of the target subpage hot, the
order to clear the subpages in the huge page in clear_huge_page() is
changed to clearing the subpage which is furthest from the target
subpage firstly, and the target subpage last. The similar order
changing helps huge page copying too. That is implemented in this
patch. Because we have put the order algorithm into a separate
function, the implementation is quite simple.
The patch is a generic optimization which should benefit quite some
workloads, not for a specific use case. To demonstrate the performance
benefit of the patch, we tested it with vm-scalability run on
transparent huge page.
With this patch, the throughput increases ~16.6% in vm-scalability
anon-cow-seq test case with 36 processes on a 2 socket Xeon E5 v3 2699
system (36 cores, 72 threads). The test case set
/sys/kernel/mm/transparent_hugepage/enabled to be always, mmap() a big
anonymous memory area and populate it, then forked 36 child processes,
each writes to the anonymous memory area from the begin to the end, so
cause copy on write. For each child process, other child processes
could be seen as other workloads which generate heavy cache pressure.
At the same time, the IPC (instruction per cycle) increased from 0.63 to
0.78, and the time spent in user space is reduced ~7.2%.
Huang Ying [Fri, 17 Aug 2018 22:45:46 +0000 (15:45 -0700)]
mm, clear_huge_page: move order algorithm into a separate function
Patch series "mm, huge page: Copy target sub-page last when copy huge
page", v2.
Huge page helps to reduce TLB miss rate, but it has higher cache
footprint, sometimes this may cause some issue. For example, when
copying huge page on x86_64 platform, the cache footprint is 4M. But on
a Xeon E5 v3 2699 CPU, there are 18 cores, 36 threads, and only 45M LLC
(last level cache). That is, in average, there are 2.5M LLC for each
core and 1.25M LLC for each thread.
If the cache contention is heavy when copying the huge page, and we copy
the huge page from the begin to the end, it is possible that the begin
of huge page is evicted from the cache after we finishing copying the
end of the huge page. And it is possible for the application to access
the begin of the huge page after copying the huge page.
In c79b57e462b5d ("mm: hugetlb: clear target sub-page last when clearing
huge page"), to keep the cache lines of the target subpage hot, the
order to clear the subpages in the huge page in clear_huge_page() is
changed to clearing the subpage which is furthest from the target
subpage firstly, and the target subpage last. The similar order
changing helps huge page copying too. That is implemented in this
patchset.
The patchset is a generic optimization which should benefit quite some
workloads, not for a specific use case. To demonstrate the performance
benefit of the patchset, we have tested it with vm-scalability run on
transparent huge page.
With this patchset, the throughput increases ~16.6% in vm-scalability
anon-cow-seq test case with 36 processes on a 2 socket Xeon E5 v3 2699
system (36 cores, 72 threads). The test case set
/sys/kernel/mm/transparent_hugepage/enabled to be always, mmap() a big
anonymous memory area and populate it, then forked 36 child processes,
each writes to the anonymous memory area from the begin to the end, so
cause copy on write. For each child process, other child processes
could be seen as other workloads which generate heavy cache pressure.
At the same time, the IPC (instruction per cycle) increased from 0.63 to
0.78, and the time spent in user space is reduced ~7.2%.
This patch (of 4):
In c79b57e462b5d ("mm: hugetlb: clear target sub-page last when clearing
huge page"), to keep the cache lines of the target subpage hot, the
order to clear the subpages in the huge page in clear_huge_page() is
changed to clearing the subpage which is furthest from the target
subpage firstly, and the target subpage last. This optimization could
be applied to copying huge page too with the same order algorithm. To
avoid code duplication and reduce maintenance overhead, in this patch,
the order algorithm is moved out of clear_huge_page() into a separate
function: process_huge_page(). So that we can use it for copying huge
page too.
This will change the direct calls to clear_user_highpage() into the
indirect calls. But with the proper inline support of the compilers,
the indirect call will be optimized to be the direct call. Our tests
show no performance change with the patch.
This patch is a code cleanup without functionality change.
Jens Axboe [Fri, 17 Aug 2018 22:45:36 +0000 (15:45 -0700)]
mpage: mpage_readpages() should submit IO as read-ahead
a_ops->readpages() is only ever used for read-ahead, yet we don't flag
the IO being submitted as such. Fix that up. Any file system that uses
mpage_readpages() as its ->readpages() implementation will now get this
right.
Since we're passing in whether the IO is read-ahead or not, we don't
need to pass in the 'gfp' separately, as it is dependent on the IO being
read-ahead. Kill off that member.
Add some documentation notes on ->readpages() being purely for
read-ahead.
Jens Axboe [Fri, 17 Aug 2018 22:45:32 +0000 (15:45 -0700)]
mpage: add argument structure for do_mpage_readpage()
Patch series "Submit ->readpages() IO as read-ahead", v4.
The only caller of ->readpages() is from read-ahead, yet we don't submit
IO flagged with REQ_RAHEAD. This means we don't see it in blktrace, for
instance, which is a shame. Additionally, it's preventing further
functional changes in the block layer for deadling with read-ahead more
intelligently. We already make assumptions about ->readpages() just
being for read-ahead in the mpage implementation, using
readahead_gfp_mask(mapping) as out GFP mask of choice.
This small series fixes up mpage_readpages() to submit with REQ_RAHEAD,
which takes care of file systems using mpage_readpages(). The first
patch is a prep patch, that makes do_mpage_readpage() take an argument
structure.
This patch (of 4):
We're currently passing 8 arguments to this function, clean it up a bit
by packing the arguments in an args structure we pass to it.
Yang Shi [Fri, 17 Aug 2018 22:45:29 +0000 (15:45 -0700)]
mm: thp: inc counter for collapsed shmem THP
/sys/kernel/mm/transparent_hugepage/khugepaged/pages_collapsed is used
to record the counter of collapsed THP, but it just gets inc'ed in
anonymous THP collapse path, do this for shmem THP collapse too.
Yang Shi [Fri, 17 Aug 2018 22:45:26 +0000 (15:45 -0700)]
mm: thp: register mm for khugepaged when merging vma for shmem
When merging anonymous page vma, if the size of the vma can fit in at
least one hugepage, the mm will be registered for khugepaged for
collapsing THP in the future.
But it skips shmem vmas. Do so for shmem also, but not for file-private
mappings when merging a vma in order to increase the odds of collapsing
a hugepage via khugepaged.
hugepage_vma_check() sounds like a good fit to do the check. And move
the definition of it before khugepaged_enter_vma_merge() to avoid a
build error.
Arnd Bergmann [Fri, 17 Aug 2018 22:45:09 +0000 (15:45 -0700)]
shmem: use monotonic time for i_generation
get_seconds() is deprecated because it will lead to a 32-bit overflow in
2038 or 2106. We don't need the i_generation to be strictly monotonic
anyway, and other file systems like ext4 and xfs just use prandom_u32(),
so let's use the same one here.
If this is considered too slow, we could also use ktime_get_seconds() or
ktime_get_real_seconds() to keep the previous behavior. Both of these
return a time64_t and are not deprecated, but only return a unique value
once per second, and are predictable.
Vlastimil Babka [Fri, 17 Aug 2018 22:45:05 +0000 (15:45 -0700)]
mm, page_alloc: actually ignore mempolicies for high priority allocations
__alloc_pages_slowpath() has for a long time contained code to ignore
node restrictions from memory policies for high priority allocations.
The current code that resets the zonelist iterator however does
effectively nothing after commit 7810e6781e0f ("mm, page_alloc: do not
break __GFP_THISNODE by zonelist reset") removed a buggy zonelist reset.
Even before that commit, mempolicy restrictions were still not ignored,
as they are passed in ac->nodemask which is untouched by the code.
We can either remove the code, or make it work as intended. Since
ac->nodemask can be set from task's mempolicy via alloc_pages_current()
and thus also alloc_pages(), it may indeed affect kernel allocations,
and it makes sense to ignore it to allow progress for high priority
allocations.
Thus, this patch resets ac->nodemask to NULL in such cases. This
assumes all callers can handle it (i.e. there are no guarantees as in
the case of __GFP_THISNODE) which seems to be the case. The same
assumption is already present in check_retry_cpuset() for some time.
The expected effect is that high priority kernel allocations in the
context of userspace tasks (e.g. OOM victims) restricted by mempolicies
will have higher chance to succeed if they are restricted to nodes with
depleted memory, while there are other nodes with free memory left.
It's not a new intention, but for the first time the code will match the
intention, AFAICS. It was intended by commit 183f6371aac2 ("mm: ignore
mempolicies when using ALLOC_NO_WATERMARK") in v3.6 but I think it never
really worked, as mempolicy restriction was already encoded in nodemask,
not zonelist, at that time.
So originally that was for ALLOC_NO_WATERMARK only. Then it was
adjusted by e46e7b77c909 ("mm, page_alloc: recalculate the preferred
zoneref if the context can ignore memory policies") and cd04ae1e2dc8
("mm, oom: do not rely on TIF_MEMDIE for memory reserves access") to the
current state. So even GFP_ATOMIC would now ignore mempolicies after
the initial attempts fail - if the code worked as people thought it
does.
Christian Hansen [Fri, 17 Aug 2018 22:45:02 +0000 (15:45 -0700)]
tools/vm/page-types.c: add support for idle page tracking
Add a flag which causes page-types to use the kernels's idle page
tracking to mark pages idle. As the tool already prints the idle flag
if set, subsequent runs will show which pages have been accessed since
last run.
Christian Hansen [Fri, 17 Aug 2018 22:44:59 +0000 (15:44 -0700)]
tools/vm/page-types.c: include shared map counts
Add a new flag that will read kpagecount for each PFN and print out the
number of times the page is mapped along with the flags in the listing
view.
This information is useful in understanding and optimizing memory usage.
Identifying pages which are not shared allows us to focus on adjusting
the memory layout or access patterns for the sole owning process.
Knowing the number of processes that share a page tells us how many
other times we must make the same adjustments or how many processes to
potentially disable.
Yang Shi [Fri, 17 Aug 2018 22:44:55 +0000 (15:44 -0700)]
thp: use mm_file_counter to determine update which rss counter
Since commit eca56ff906bd ("mm, shmem: add internal shmem resident
memory accounting"), MM_SHMEMPAGES is added to separate the shmem
accounting from regular files. So, all shmem pages should be accounted
to MM_SHMEMPAGES instead of MM_FILEPAGES.
And, normal 4K shmem pages have been accounted to MM_SHMEMPAGES, so
shmem thp pages should be not treated differently. Account them to
MM_SHMEMPAGES via mm_counter_file() since shmem pages are swap backed to
keep consistent with normal 4K shmem pages.
This will not change the rss counter of processes since shmem pages are
still a part of it.
The /proc/pid/status and /proc/pid/statm counters will however be more
accurate wrt shmem usage, as originally intended. And as eca56ff906bd
("mm, shmem: add internal shmem resident memory accounting") mentioned,
oom also could report more accurate "shmem-rss".
Pavel Tatashin [Fri, 17 Aug 2018 22:44:52 +0000 (15:44 -0700)]
mm: skip invalid pages block at a time in zero_resv_unresv()
The role of zero_resv_unavail() is to make sure that every struct page
that is allocated but is not backed by memory that is accessible by
kernel is zeroed and not in some uninitialized state.
Since struct pages are allocated in blocks (2M pages in x86 case), we
can skip pageblock_nr_pages at a time, when the first one is found to be
invalid.
This optimization may help since now on x86 every hole in e820 maps is
marked as reserved in memblock, and thus will go through this function.
This function is called before sched_clock() is initialized, so I used
my x86 early boot clock patches to measure the performance improvement.
With 1T hole on i7-8700 currently we would take 0.606918s of boot time,
but with this optimization 0.001103s.
Souptick Joarder [Fri, 17 Aug 2018 22:44:47 +0000 (15:44 -0700)]
mm: convert return type of handle_mm_fault() caller to vm_fault_t
Use new return type vm_fault_t for fault handler. For now, this is just
documenting that the function returns a VM_FAULT value rather than an
errno. Once all instances are converted, vm_fault_t will become a
distinct type.
Ref-> commit 1c8f422059ae ("mm: change return type to vm_fault_t")
In this patch all the caller of handle_mm_fault() are changed to return
vm_fault_t type.
Vlastimil Babka [Fri, 17 Aug 2018 22:44:44 +0000 (15:44 -0700)]
mm, slub: restore the original intention of prefetch_freepointer()
In SLUB, prefetch_freepointer() is used when allocating an object from
cache's freelist, to make sure the next object in the list is cache-hot,
since it's probable it will be allocated soon.
Commit 2482ddec670f ("mm: add SLUB free list pointer obfuscation") has
unintentionally changed the prefetch in a way where the prefetch is
turned to a real fetch, and only the next->next pointer is prefetched.
In case there is not a stream of allocations that would benefit from
prefetching, the extra real fetch might add a useless cache miss to the
allocation. Restore the previous behavior.
NeilBrown [Fri, 17 Aug 2018 22:44:41 +0000 (15:44 -0700)]
fs/seq_file.c: simplify seq_file iteration code and interface
The documentation for seq_file suggests that it is necessary to be able
to move the iterator to a given offset, however that is not the case.
If the iterator is stored in the private data and is stable from one
read() syscall to the next, it is only necessary to support first/next
interactions. Implementing this in a client is a little clumsy.
- if ->start() is given a pos of zero, it should go to start of
sequence.
- if ->start() is given the name pos that was given to the most recent
next() or start(), it should restore the iterator to state just
before that last call
- if ->start is given another number, it should set the iterator one
beyond the start just before the last ->start or ->next call.
Also, the documentation says that the implementation can interpret the
pos however it likes (other than zero meaning start), but seq_file
increments the pos sometimes which does impose on the implementation.
This patch simplifies the interface for first/next iteration and
simplifies the code, while maintaining complete backward compatability.
Now:
- if ->start() is given a pos of zero, it should return an iterator
placed at the start of the sequence
- if ->start() is given a non-zero pos, it should return the iterator
in the same state it was after the last ->start or ->next.
This is particularly useful for interators which walk the multiple
chains in a hash table, e.g. using rhashtable_walk*. See
fs/gfs2/glock.c and drivers/staging/lustre/lustre/llite/vvp_dev.c
A large part of achieving this is to *always* call ->next after ->show
has successfully stored all of an entry in the buffer. Never just
increment the index instead. Also:
- always pass &m->index to ->start() and ->next(), never a temp
variable
- don't clear ->from when ->count is zero, as ->from is dead when
->count is zero.
Some ->next functions do not increment *pos when they return NULL. To
maintain compatability with this, we still need to increment m->index in
one place, if ->next didn't increment it. Note that such ->next
functions are buggy and should be fixed. A simple demonstration is
dd if=/proc/swaps bs=1000 skip=1
Choose any block size larger than the size of /proc/swaps. This will
always show the whole last line of /proc/swaps.
This patch doesn't work around buggy next() functions for this case.
Colin Ian King [Fri, 17 Aug 2018 22:44:31 +0000 (15:44 -0700)]
ocfs2: make several functions and variables static (and some const)
There are a variety of functions and variables that are local to the
source and do not need to be in global scope, so make them static. Also
make a couple of char arrays static const.
Cleans up sparse warnings:
symbol 'o2hb_heartbeat_mode_desc' was not declared. Should it be static?
symbol 'o2hb_heartbeat_mode' was not declared. Should it be static?
symbol 'o2hb_dependent_users' was not declared. Should it be static?
symbol 'o2hb_region_dec_user' was not declared. Should it be static?
symbol 'o2nm_fence_method_desc' was not declared. Should it be static?
symbol 'lockdep_keys' was not declared. Should it be static?
Nick Desaulniers [Fri, 17 Aug 2018 22:44:21 +0000 (15:44 -0700)]
sh: prefer _THIS_IP_ to current_text_addr
As part of the effort to reduce the code duplication between _THIS_IP_
and current_text_addr(), let's consolidate callers of
current_text_addr() to use _THIS_IP_.
Kees Cook [Fri, 17 Aug 2018 22:44:14 +0000 (15:44 -0700)]
ntfs: mft: remove VLA usage
In the quest to remove all stack VLA usage from the kernel[1], this
allocates the maximum size stack buffer. Existing checks already
require that blocksize >= NTFS_BLOCK_SIZE and mft_record_size <=
PAGE_SIZE, so max_bhs can be at most PAGE_SIZE / NTFS_BLOCK_SIZE.
Sanity checks are added for robustness.
Kees Cook [Fri, 17 Aug 2018 22:44:11 +0000 (15:44 -0700)]
ntfs: decompress: remove VLA usage
In the quest to remove all stack VLA usage from the kernel[1], this
moves the stack buffer used during decompression to be allocated
externally.
The existing "dest_max_index" used in the VLA is bounded by cb_max_page.
cb_max_page is bounded by max_page, and max_page is bounded by nr_pages.
Since nr_pages is used for the "pages" allocation, it can similarly be
used for the "completed_pages" allocation and passed into the
decompression function. The error paths are updated to free the new
allocation.
Kees Cook [Fri, 17 Aug 2018 22:44:07 +0000 (15:44 -0700)]
ntfs: aops: remove VLA usage
In the quest to remove all stack VLA usage from the kernel[1], this uses
the maximum size needed on the stack and adds a sanity check for
robustness: index.block_size cannot be larger than PAGE_SIZE nor less
than NTFS_BLOCK_SIZE.
fs/ntfs/aops.c: don't disable interrupts during kmap_atomic()
ntfs_end_buffer_async_read() disables interrupts around kmap_atomic().
This is a leftover from the old kmap_atomic() implementation which
relied on fixed mapping slots, so the caller had to make sure that the
same slot could not be reused from an interrupting context.
kmap_atomic() was changed to dynamic slots long ago and commit 1ec9c5ddc17a ("include/linux/highmem.h: remove the second argument of
k[un]map_atomic()") removed the slot assignements, but the callers were
not checked for now redundant interrupt disabling.
Jeremy Cline [Fri, 17 Aug 2018 22:44:01 +0000 (15:44 -0700)]
scripts: add Python 3 compatibility to spdxcheck.py
"dict.has_key(key)" on dictionaries has been replaced with "key in
dict". Additionally, when run under Python 3 some files don't decode
with the default encoding (tested with UTF-8). To handle that, don't
open the file in text mode and decode text line-by-line, ignoring
encoding errors.
This remains compatible with Python 2 and should have no functional
change.
Arnd Bergmann [Fri, 17 Aug 2018 22:43:54 +0000 (15:43 -0700)]
fs/hpfs: extend gmt_to_local() conversion to 64-bit times
The VFS timestamps are all 64-bit now, the only missing piece for hpfs
is the internal conversion function. One interesting bit about hpfs is
that it can already deal with moving the 136 year window of its
timestamps to support a much wider range than other file systems with
32-bit timestamps. It also treats the timestamps as 'unsigned' on
64-bit architectures (but signed on 32-bit, because time_t always around
to negative numbers in 2038).
Changing the conversion to use time64_t makes 32-bit architectures
behave the same way as 64-bit. For completeness, this also adds a
clamp_t call for each conversion, so we don't wrap the timestamps but
instead stay within the [0..U32_MAX] range of the on-disk timestamps.
Arnd Bergmann [Fri, 17 Aug 2018 22:43:47 +0000 (15:43 -0700)]
fs/ufs: use ktime_get_real_seconds for sb and cg timestamps
get_seconds() is deprecated because of the 32-bit overflow and will be
removed. All callers in ufs also truncate to a 32-bit number, so
nothing changes during the conversion, but this should be harmless as
the superblock and cylinder group timestamps are not visible to user
space, except for checking the fs-dirty state, wich works fine across
the overflow.
This moves the call to get_seconds() into a new inline function, with a
comment explaining the constraints, while converting it to
ktime_get_real_seconds().
Arnd Bergmann [Fri, 17 Aug 2018 22:43:44 +0000 (15:43 -0700)]
firewire: use 64-bit time_t based interfaces
32-bit CLOCK_REALTIME timestamps overflow in year 2038, so all such
interfaces are deprecated now. For the FW_CDEV_IOC_GET_CYCLE_TIMER2
ioctl, we already support 64-bit timestamps, but the implementation
still uses timespec.
This changes the code to use timespec64 instead with the appropriate
accessor functions.
Dave Jiang [Fri, 17 Aug 2018 22:43:40 +0000 (15:43 -0700)]
dax: remove VM_MIXEDMAP for fsdax and device dax
This patch is reworked from an earlier patch that Dan has posted:
https://patchwork.kernel.org/patch/10131727/
VM_MIXEDMAP is used by dax to direct mm paths like vm_normal_page() that
the memory page it is dealing with is not typical memory from the linear
map. The get_user_pages_fast() path, since it does not resolve the vma,
is already using {pte,pmd}_devmap() as a stand-in for VM_MIXEDMAP, so we
use that as a VM_MIXEDMAP replacement in some locations. In the cases
where there is no pte to consult we fallback to using vma_is_dax() to
detect the VM_MIXEDMAP special case.
Now that we have explicit driver pfn_t-flag opt-in/opt-out for
get_user_pages() support for DAX we can stop setting VM_MIXEDMAP. This
also means we no longer need to worry about safely manipulating vm_flags
in a future where we support dynamically changing the dax mode of a
file.
DAX should also now be supported with madvise_behavior(), vma_merge(),
and copy_page_range().
This patch has been tested against ndctl unit test. It has also been
tested against xfstests commit: 625515d using fake pmem created by
memmap and no additional issues have been observed.
Passing an enum into FIELD_GET() produces a long but harmless warning on
newer compilers:
from include/linux/linkage.h:7,
from include/linux/kernel.h:7,
from include/linux/skbuff.h:17,
from include/linux/if_ether.h:23,
from include/linux/etherdevice.h:25,
from drivers/net/wireless/intel/iwlwifi/mvm/rxmq.c:63:
drivers/net/wireless/intel/iwlwifi/mvm/rxmq.c: In function 'iwl_mvm_rx_mpdu_mq':
include/linux/bitfield.h:56:20: error: enum constant in boolean context [-Werror=int-in-bool-context]
BUILD_BUG_ON_MSG(!(_mask), _pfx "mask is zero"); \
^
...
include/linux/bitfield.h:103:3: note: in expansion of macro '__BF_FIELD_CHECK'
__BF_FIELD_CHECK(_mask, _reg, 0U, "FIELD_GET: "); \
^~~~~~~~~~~~~~~~
drivers/net/wireless/intel/iwlwifi/mvm/rxmq.c:1025:21: note: in expansion of macro 'FIELD_GET'
le16_encode_bits(FIELD_GET(IWL_RX_HE_PHY_SIBG_SYM_OR_USER_NUM_MASK,
The problem here is that the caller has no idea how the macro gets
expanding, leading to a false-positive. It can be trivially avoided by
doing a comparison against zero.
This only recently started appearing as the iwlwifi driver was patched
to use FIELD_GET.
Daniel Borkmann [Fri, 17 Aug 2018 21:26:14 +0000 (23:26 +0200)]
bpf: fix redirect to map under tail calls
Commits 109980b894e9 ("bpf: don't select potentially stale ri->map
from buggy xdp progs") and 7c3001313396 ("bpf: fix ri->map_owner
pointer on bpf_prog_realloc") tried to mitigate that buggy programs
using bpf_redirect_map() helper call do not leave stale maps behind.
Idea was to add a map_owner cookie into the per CPU struct redirect_info
which was set to prog->aux by the prog making the helper call as a
proof that the map is not stale since the prog is implicitly holding
a reference to it. This owner cookie could later on get compared with
the program calling into BPF whether they match and therefore the
redirect could proceed with processing the map safely.
In (obvious) hindsight, this approach breaks down when tail calls are
involved since the original caller's prog->aux pointer does not have
to match the one from one of the progs out of the tail call chain,
and therefore the xdp buffer will be dropped instead of redirected.
A way around that would be to fix the issue differently (which also
allows to remove related work in fast path at the same time): once
the life-time of a redirect map has come to its end we use it's map
free callback where we need to wait on synchronize_rcu() for current
outstanding xdp buffers and remove such a map pointer from the
redirect info if found to be present. At that time no program is
using this map anymore so we simply invalidate the map pointers to
NULL iff they previously pointed to that instance while making sure
that the redirect path only reads out the map once.
Fixes: 97f91a7cf04f ("bpf: add bpf_redirect_map helper routine") Fixes: 109980b894e9 ("bpf: don't select potentially stale ri->map from buggy xdp progs") Reported-by: Sebastiano Miano <[email protected]> Signed-off-by: Daniel Borkmann <[email protected]> Acked-by: John Fastabend <[email protected]> Signed-off-by: Alexei Starovoitov <[email protected]>
Linus Torvalds [Fri, 17 Aug 2018 19:44:48 +0000 (12:44 -0700)]
Merge tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma
Pull rdma updates from Jason Gunthorpe:
"This has been a large cycle for RDMA, with several major patch series
reworking parts of the core code.
- Rework the so-called 'gid cache' and internal APIs to use a kref'd
pointer to a struct instead of copying, push this upwards into the
callers and add more stuff to the struct. The new design avoids
some ugly races the old one suffered with. This is part of the
namespace enablement work as the new struct is learning to be
namespace aware.
- Various uapi cleanups, moving more stuff to include/uapi and fixing
some long standing bugs that have recently been discovered.
- Driver updates for mlx5, mlx4 i40iw, rxe, cxgb4, hfi1, usnic,
pvrdma, and hns
- Provide max_send_sge and max_recv_sge attributes to better support
HW where these values are asymmetric.
- mlx5 user API 'devx' allows sending commands directly to the device
FW, instead of trying to cram every wild and niche feature into the
common API. Sort of like what GPU does.
- Major write() and ioctl() API rework to cleanly support PCI device
hot unplug and advance the ioctl conversion work
- Sparse and compile warning cleanups
- Add 'const' to the ib_poll_cq() signature, and permit a NULL
'bad_wr', which is the common use case
- Various patches to avoid high order allocations across the stack
- SRQ support for cxgb4, hns and qedr
- Changes to IPoIB to better follow the netdev model for working with
struct net_device liftime"
* tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma: (312 commits)
Revert "net/smc: Replace ib_query_gid with rdma_get_gid_attr"
RDMA/hns: Fix usage of bitmap allocation functions return values
IB/core: Change filter function return type from int to bool
IB/core: Update GID entries for netdevice whose mac address changes
IB/core: Add default GIDs of the bond master netdev
IB/core: Consider adding default GIDs of bond device
IB/core: Delete lower netdevice default GID entries in bonding scenario
IB/core: Avoid confusing del_netdev_default_ips
IB/core: Add comment for change upper netevent handling
qedr: Add user space support for SRQ
qedr: Add support for kernel mode SRQ's
qedr: Add wrapping generic structure for qpidr and adjust idr routines.
IB/mlx5: Fix leaking stack memory to userspace
Update the e-mail address of Bart Van Assche
IB/ucm: Fix compiling ucm.c
IB/uverbs: Do not check for device disassociation during ioctl
IB/uverbs: Remove struct uverbs_root_spec and all supporting code
IB/uverbs: Use uverbs_api to unmarshal ioctl commands
IB/uverbs: Use uverbs_alloc for allocations
IB/uverbs: Add a simple allocator to uverbs_attr_bundle
...
The bpf selftest test_cgroup_storage failed in one of
our production test servers.
# sudo ./test_cgroup_storage
Failed to create map: Operation not permitted
It turns out this is due to insufficient locked memory
with system default 16KB.
Similar to other self tests, let us arm the process
with unlimited locked memory. With this change,
the test passed.
# sudo ./test_cgroup_storage
test_cgroup_storage:PASS
Linus Torvalds [Fri, 17 Aug 2018 19:13:15 +0000 (12:13 -0700)]
Merge tag 'drm-next-2018-08-17-1' of git://anongit.freedesktop.org/drm/drm
Pull drm msm support for adreno a6xx from Dave Airlie:
"This is the support for new Qualcomm Snapdragon SoCs with the A6xx
core. Userspace support is in mesa now"
* tag 'drm-next-2018-08-17-1' of git://anongit.freedesktop.org/drm/drm:
drm/msm: a6xx: fix spelling mistake: "initalization" -> "initialization"
drm/msm: Add A6XX device support
drm/msm: update generated headers
drm/msm/adreno: Load the firmware before bringing up the hardware
drm/msm: Add a helper function to parse clock names
Linus Torvalds [Fri, 17 Aug 2018 19:10:22 +0000 (12:10 -0700)]
Merge tag 'drm-next-2018-08-17' of git://anongit.freedesktop.org/drm/drm
Pull drm fixes from Dave Airlie:
"First round of fixes for -rc1. I'll follow this up with the msm new hw
support pull request.
This just has three sets of fixes, some for msm before the new hw, a
bunch of AMD fixes (includiing some required firmware changes for new
hw), and a set of i915 (+gvt) fixes"
* tag 'drm-next-2018-08-17' of git://anongit.freedesktop.org/drm/drm: (30 commits)
drm/amdgpu: Use kvmalloc for allocating UVD/VCE/VCN BO backup memory
drm/i915: set DP Main Stream Attribute for color range on DDI platforms
drm/i915/selftests: Hold rpm for unparking
drm/i915: Restore user forcewake domains across suspend
drm/i915: Unmask user interrupts writes into HWSP on snb/ivb/vlv/hsw
drm/i915/gvt: fix memory leak in intel_vgpu_ioctl()
drm/i915/gvt: Off by one in intel_vgpu_write_fence()
drm/i915/kvmgt: Fix potential Spectre v1
drm/i915/gvt: return error on cmd access
drm/i915/gvt: initialize dmabuf mutex in vgpu_create
drm/i915/gvt: fix cleanup sequence in intel_gvt_clean_device
drm/amd/display: Guard against null crtc in CRC IRQ
drm/amd/display: Pass connector id when executing VBIOS CT
drm/amd/display: Check if clock source in use before disabling
drm/amd/display: Allow clock sharing b/w HDMI and DVI
drm/amd/display: Fix warning observed in mode change on Vega
drm/amd/display: fix single link DVI has no display
drm/amdgpu/vce: VCE entity initialization relies on ring initializtion
drm/amdgpu/uvd: UVD entity initialization relys on ring initialization
drm/amdgpu:add VCN booting with firmware loaded by PSP
...
Linus Torvalds [Fri, 17 Aug 2018 18:48:04 +0000 (11:48 -0700)]
Merge tag 'arm64-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux
Pull arm64 fixes from Will Deacon:
"A couple of arm64 fixes
- Fix boot on Hikey-960 by avoiding an IPI with interrupts disabled
- Fix address truncation in pfn_valid() implementation"
* tag 'arm64-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux:
arm64: mm: check for upper PAGE_SHIFT bits in pfn_valid()
arm64: Avoid calling stop_machine() when patching jump labels
Linus Torvalds [Fri, 17 Aug 2018 18:32:50 +0000 (11:32 -0700)]
Merge tag 'powerpc-4.19-1' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux
Pull powerpc updates from Michael Ellerman:
"Notable changes:
- A fix for a bug in our page table fragment allocator, where a page
table page could be freed and reallocated for something else while
still in use, leading to memory corruption etc. The fix reuses
pt_mm in struct page (x86 only) for a powerpc only refcount.
- Fixes to our pkey support. Several are user-visible changes, but
bring us in to line with x86 behaviour and/or fix outright bugs.
Thanks to Florian Weimer for reporting many of these.
- A series to improve the hvc driver & related OPAL console code,
which have been seen to cause hardlockups at times. The hvc driver
changes in particular have been in linux-next for ~month.
- Increase our MAX_PHYSMEM_BITS to 128TB when SPARSEMEM_VMEMMAP=y.
- Remove Power8 DD1 and Power9 DD1 support, neither chip should be in
use anywhere other than as a paper weight.
- An optimised memcmp implementation using Power7-or-later VMX
instructions
- Support for barrier_nospec on some NXP CPUs.
- Support for flushing the count cache on context switch on some IBM
CPUs (controlled by firmware), as a Spectre v2 mitigation.
- A series to enhance the information we print on unhandled signals
to bring it into line with other arches, including showing the
offending VMA and dumping the instructions around the fault.
Thanks to: Aaro Koskinen, Akshay Adiga, Alastair D'Silva, Alexey
Kardashevskiy, Alexey Spirkov, Alistair Popple, Andrew Donnellan,
Aneesh Kumar K.V, Anju T Sudhakar, Arnd Bergmann, Bartosz Golaszewski,
Benjamin Herrenschmidt, Bharat Bhushan, Bjoern Noetel, Boqun Feng,
Breno Leitao, Bryant G. Ly, Camelia Groza, Christophe Leroy, Christoph
Hellwig, Cyril Bur, Dan Carpenter, Daniel Klamt, Darren Stevens, Dave
Young, David Gibson, Diana Craciun, Finn Thain, Florian Weimer,
Frederic Barrat, Gautham R. Shenoy, Geert Uytterhoeven, Geoff Levand,
Guenter Roeck, Gustavo Romero, Haren Myneni, Hari Bathini, Joel
Stanley, Jonathan Neuschäfer, Kees Cook, Madhavan Srinivasan, Mahesh
Salgaonkar, Markus Elfring, Mathieu Malaterre, Mauro S. M. Rodrigues,
Michael Hanselmann, Michael Neuling, Michael Schmitz, Mukesh Ojha,
Murilo Opsfelder Araujo, Nicholas Piggin, Parth Y Shah, Paul
Mackerras, Paul Menzel, Ram Pai, Randy Dunlap, Rashmica Gupta, Reza
Arbab, Rodrigo R. Galvao, Russell Currey, Sam Bobroff, Scott Wood,
Shilpasri G Bhat, Simon Guo, Souptick Joarder, Stan Johnson, Thiago
Jung Bauermann, Tyrel Datwyler, Vaibhav Jain, Vasant Hegde, Venkat
Rao, zhong jiang"
* tag 'powerpc-4.19-1' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux: (234 commits)
powerpc/mm/book3s/radix: Add mapping statistics
powerpc/uaccess: Enable get_user(u64, *p) on 32-bit
powerpc/mm/hash: Remove unnecessary do { } while(0) loop
powerpc/64s: move machine check SLB flushing to mm/slb.c
powerpc/powernv/idle: Fix build error
powerpc/mm/tlbflush: update the mmu_gather page size while iterating address range
powerpc/mm: remove warning about ‘type’ being set
powerpc/32: Include setup.h header file to fix warnings
powerpc: Move `path` variable inside DEBUG_PROM
powerpc/powermac: Make some functions static
powerpc/powermac: Remove variable x that's never read
cxl: remove a dead branch
powerpc/powermac: Add missing include of header pmac.h
powerpc/kexec: Use common error handling code in setup_new_fdt()
powerpc/xmon: Add address lookup for percpu symbols
powerpc/mm: remove huge_pte_offset_and_shift() prototype
powerpc/lib: Use patch_site to patch copy_32 functions once cache is enabled
powerpc/pseries: Fix endianness while restoring of r3 in MCE handler.
powerpc/fadump: merge adjacent memory ranges to reduce PT_LOAD segements
powerpc/fadump: handle crash memory ranges array index overflow
...
Linus Torvalds [Fri, 17 Aug 2018 17:51:22 +0000 (10:51 -0700)]
Merge tag 'modules-for-v4.19' of git://git.kernel.org/pub/scm/linux/kernel/git/jeyu/linux
Pull modules updates from Jessica Yu:
"Summary of modules changes for the 4.19 merge window:
- Fix modules kallsyms for livepatch. Livepatch modules can have
SHN_UNDEF symbols in their module symbol tables for later symbol
resolution, but kallsyms shouldn't be returning these symbols
- Some code cleanups and minor reshuffling in load_module() were done
to log the module name when module signature verification fails"
* tag 'modules-for-v4.19' of git://git.kernel.org/pub/scm/linux/kernel/git/jeyu/linux:
kernel/module: Use kmemdup to replace kmalloc+memcpy
ARM: module: fix modsign build error
modsign: log module name in the event of an error
module: replace VMLINUX_SYMBOL_STR() with __stringify() or string literal
module: print sensible error code
module: setup load info before module_sig_check()
module: make it clear when we're handling the module copy in info->hdr
module: exclude SHN_UNDEF symbols from kallsyms api
x86/speculation/l1tf: Exempt zeroed PTEs from inversion
It turns out that we should *not* invert all not-present mappings,
because the all zeroes case is obviously special.
clear_page() does not undergo the XOR logic to invert the address bits,
i.e. PTE, PMD and PUD entries that have not been individually written
will have val=0 and so will trigger __pte_needs_invert(). As a result,
{pte,pmd,pud}_pfn() will return the wrong PFN value, i.e. all ones
(adjusted by the max PFN mask) instead of zero. A zeroed entry is ok
because the page at physical address 0 is reserved early in boot
specifically to mitigate L1TF, so explicitly exempt them from the
inversion when reading the PFN.
Manifested as an unexpected mprotect(..., PROT_NONE) failure when called
on a VMA that has VM_PFNMAP and was mmap'd to as something other than
PROT_NONE but never used. mprotect() sends the PROT_NONE request down
prot_none_walk(), which walks the PTEs to check the PFNs.
prot_none_pte_entry() gets the bogus PFN from pte_pfn() and returns
-EACCES because it thinks mprotect() is trying to adjust a high MMIO
address.
[ This is a very modified version of Sean's original patch, but all
credit goes to Sean for doing this and also pointing out that
sometimes the __pte_needs_invert() function only gets the protection
bits, not the full eventual pte. But zero remains special even in
just protection bits, so that's ok. - Linus ]
Linus Torvalds [Fri, 17 Aug 2018 16:52:15 +0000 (09:52 -0700)]
Merge tag 'for-4.19/dm-changes' of git://git.kernel.org/pub/scm/linux/kernel/git/device-mapper/linux-dm
Pull device mapper updates from Mike Snitzer:
- A couple stable fixes for the DM writecache target.
- A stable fix for the DM cache target that fixes the potential for
data corruption after an unclean shutdown of a cache device using
writeback mode.
- Update DM integrity target to allow the metadata to be stored on a
separate device from data.
- Fix DM kcopyd and the snapshot target to cond_resched() where
appropriate and be more efficient with processing completed work.
- A few fixes and improvements for DM crypt.
- Add DM delay target feature to configure delay of flushes independent
of writes.
- Update DM thin-provisioning target to include metadata_low_watermark
threshold in pool status.
- Fix stale DM thin-provisioning Documentation.
* tag 'for-4.19/dm-changes' of git://git.kernel.org/pub/scm/linux/kernel/git/device-mapper/linux-dm: (26 commits)
dm writecache: fix a crash due to reading past end of dirty_bitmap
dm crypt: don't decrease device limits
dm cache metadata: set dirty on all cache blocks after a crash
dm snapshot: remove stale FIXME in snapshot_map()
dm snapshot: improve performance by switching out_of_order_list to rbtree
dm kcopyd: avoid softlockup in run_complete_job
dm cache metadata: save in-core policy_hint_size to on-disk superblock
dm thin: stop no_space_timeout worker when switching to write-mode
dm kcopyd: return void from dm_kcopyd_copy()
dm thin: include metadata_low_watermark threshold in pool status
dm writecache: report start_sector in status line
dm crypt: convert essiv from ahash to shash
dm crypt: use wake_up_process() instead of a wait queue
dm integrity: recalculate checksums on creation
dm integrity: flush journal on suspend when using separate metadata device
dm integrity: use version 2 for separate metadata
dm integrity: allow separate metadata device
dm integrity: add ic->start in get_data_sector()
dm integrity: report provided data sectors in the status
dm integrity: implement fair range locks
...
Linus Torvalds [Fri, 17 Aug 2018 16:41:28 +0000 (09:41 -0700)]
Merge tag 'fsnotify_for_v4.19-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/jack/linux-fs
Pull fsnotify updates from Jan Kara:
"fsnotify cleanups from Amir and a small inotify improvement"
* tag 'fsnotify_for_v4.19-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/jack/linux-fs:
inotify: Add flag IN_MASK_CREATE for inotify_add_watch()
fanotify: factor out helpers to add/remove mark
fsnotify: add helper to get mask from connector
fsnotify: let connector point to an abstract object
fsnotify: pass connp and object type to fsnotify_add_mark()
fsnotify: use typedef fsnotify_connp_t for brevity
Linus Torvalds [Fri, 17 Aug 2018 16:38:39 +0000 (09:38 -0700)]
Merge tag 'for_v4.19-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/jack/linux-fs
Pull UDF and ext2 update from Jan Kara.
* tag 'for_v4.19-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/jack/linux-fs:
ext2: use ktime_get_real_seconds for timestamps
udf: convert inode stamps to timespec64
Greg Hackmann [Wed, 15 Aug 2018 19:51:21 +0000 (12:51 -0700)]
arm64: mm: check for upper PAGE_SHIFT bits in pfn_valid()
ARM64's pfn_valid() shifts away the upper PAGE_SHIFT bits of the input
before seeing if the PFN is valid. This leads to false positives when
some of the upper bits are set, but the lower bits match a valid PFN.
For example, the following userspace code looks up a bogus entry in
/proc/kpageflags:
int pagemap = open("/proc/self/pagemap", O_RDONLY);
int pageflags = open("/proc/kpageflags", O_RDONLY);
uint64_t pfn, val;
On ARM64 this causes the userspace process to crash with SIGSEGV rather
than reading (1 << KPF_NOPAGE). kpageflags_read() treats the offset as
valid, and stable_page_flags() will try to access an address between the
user and kernel address ranges.
Will Deacon [Thu, 16 Aug 2018 10:45:50 +0000 (11:45 +0100)]
arm64: Avoid calling stop_machine() when patching jump labels
Patching a jump label involves patching a single instruction at a time,
swizzling between a branch and a NOP. The architecture treats these
instructions specially, so a concurrently executing CPU is guaranteed to
see either the NOP or the branch, rather than an amalgamation of the two
instruction encodings.
However, in order to guarantee that the new instruction is visible, it
is necessary to send an IPI to the concurrently executing CPU so that it
discards any previously fetched instructions from its pipeline. This
operation therefore cannot be completed from a context with IRQs
disabled, but this is exactly what happens on the jump label path where
the hotplug lock is held and irqs are subsequently disabled by
stop_machine_cpuslocked(). This results in a deadlock during boot on
Hikey-960.
Due to the architectural guarantees around patching NOPs and branches,
we don't actually need to stop_machine() at all on the jump label path,
so we can avoid the deadlock by using the "nosync" variant of our
instruction patching routine.
Dave Airlie [Fri, 17 Aug 2018 00:46:45 +0000 (10:46 +1000)]
Merge tag 'drm-msm-next-2018-08-10' of git://people.freedesktop.org/~robclark/linux into drm-next
An optional follow-on PR for 4.19, on top of previous -fixes PR, which
brings in a6xx support.
These patches have been on list since earlier in the year (mostly
waiting for userspace). They have been in linux-next since earlier in
the week, now that we have freedreno userspace working on a6xx[1][2].
So far glmark2, Chromium/ChromiumOS, gnome-shell, glamor, xonotic,
etc, are working. And a healthy chuck of deqp works, and I've been
busy fixing things. The needed libdrm changes (no new uapi changes
needed) are already on master, and the 2nd branch is rebased on that.
Dave Airlie [Thu, 16 Aug 2018 23:26:51 +0000 (09:26 +1000)]
Merge tag 'drm-intel-next-fixes-2018-08-16-1' of git://anongit.freedesktop.org/drm/drm-intel into drm-next
Fixes for:
- DP full color range.
- selftest for gem_object
- forcewake on suspend
- GPU reset
This also include accumulated fixes from GVT:
- Fix an error code in gvt_dma_map_page() (Dan)
- Fix off by one error in intel_vgpu_write_fence() (Dan)
- Fix potential Spectre v1 (Gustavo)
- Fix workload free in vgpu release (Henry)
- Fix cleanup sequence in intel_gvt_clean_device (Henry)
- dmabuf mutex init place fix (Henry)
- possible memory leak in intel_vgpu_ioctl() err path (Yi)
- return error on cmd access check failure (Yan)
Dave Airlie [Thu, 16 Aug 2018 23:26:07 +0000 (09:26 +1000)]
Merge branch 'drm-next-4.19' of git://people.freedesktop.org/~agd5f/linux into drm-next
Fixes for 4.19:
- Add VCN PSP FW loading for RV (this is required on upcoming parts)
- Fix scheduler setup ordering for VCE and UVD
- Few misc display fixes
====================
Batch of various fixes related to BPF sockmap and ULP, including
adding module alias to restrict module requests, races and memory
leaks in sockmap code. For details please refer to the individual
patches. Thanks!
====================
Daniel Borkmann [Thu, 16 Aug 2018 19:49:10 +0000 (21:49 +0200)]
bpf, sockmap: fix sock_map_ctx_update_elem race with exist/noexist
The current code in sock_map_ctx_update_elem() allows for BPF_EXIST
and BPF_NOEXIST map update flags. While on array-like maps this approach
is rather uncommon, e.g. bpf_fd_array_map_update_elem() and others
enforce map update flags to be BPF_ANY such that xchg() can be used
directly, the current implementation in sock map does not guarantee
that such operation with BPF_EXIST / BPF_NOEXIST is atomic.
The initial test does a READ_ONCE(stab->sock_map[i]) to fetch the
socket from the slot which is then tested for NULL / non-NULL. However
later after __sock_map_ctx_update_elem(), the actual update is done
through osock = xchg(&stab->sock_map[i], sock). Problem is that in
the meantime a different CPU could have updated / deleted a socket
on that specific slot and thus flag contraints won't hold anymore.
I've been thinking whether best would be to just break UAPI and do
an enforcement of BPF_ANY to check if someone actually complains,
however trouble is that already in BPF kselftest we use BPF_NOEXIST
for the map update, and therefore it might have been copied into
applications already. The fix to keep the current behavior intact
would be to add a map lock similar to the sock hash bucket lock only
for covering the whole map.
Daniel Borkmann [Thu, 16 Aug 2018 19:49:09 +0000 (21:49 +0200)]
bpf, sockmap: fix map elem deletion race with smap_stop_sock
The smap_start_sock() and smap_stop_sock() are each protected under
the sock->sk_callback_lock from their call-sites except in the case
of sock_map_delete_elem() where we drop the old socket from the map
slot. This is racy because the same sock could be part of multiple
sock maps, so we run smap_stop_sock() in parallel, and given at that
point psock->strp_enabled might be true on both CPUs, we might for
example wrongly restore the sk->sk_data_ready / sk->sk_write_space.
Therefore, hold the sock->sk_callback_lock as well on delete. Looks
like 2f857d04601a ("bpf: sockmap, remove STRPARSER map_flags and add
multi-map support") had this right, but later on e9db4ef6bf4c ("bpf:
sockhash fix omitted bucket lock in sock_close") removed it again
from delete leaving this smap_stop_sock() instance unprotected.
Daniel Borkmann [Thu, 16 Aug 2018 19:49:08 +0000 (21:49 +0200)]
bpf, sockmap: fix leakage of smap_psock_map_entry
While working on sockmap I noticed that we do not always kfree the
struct smap_psock_map_entry list elements which track psocks attached
to maps. In the case of sock_hash_ctx_update_elem(), these map entries
are allocated outside of __sock_map_ctx_update_elem() with their
linkage to the socket hash table filled. In the case of sock array,
the map entries are allocated inside of __sock_map_ctx_update_elem()
and added with their linkage to the psock->maps. Both additions are
under psock->maps_lock each.
Now, we drop these elements from their psock->maps list in a few
occasions: i) in sock array via smap_list_map_remove() when an entry
is either deleted from the map from user space, or updated via
user space or BPF program where we drop the old socket at that map
slot, or the sock array is freed via sock_map_free() and drops all
its elements; ii) for sock hash via smap_list_hash_remove() in exactly
the same occasions as just described for sock array; iii) in the
bpf_tcp_close() where we remove the elements from the list via
psock_map_pop() and iterate over them dropping themselves from either
sock array or sock hash; and last but not least iv) once again in
smap_gc_work() which is a callback for deferring the work once the
psock refcount hit zero and thus the socket is being destroyed.
Problem is that the only case where we kfree() the list entry is
in case iv), which at that point should have an empty list in
normal cases. So in cases from i) to iii) we unlink the elements
without freeing where they go out of reach from us. Hence fix is
to properly kfree() them as well to stop the leakage. Given these
are all handled under psock->maps_lock there is no need for deferred
RCU freeing.
I later also ran with kmemleak detector and it confirmed the finding
as well where in the state before the fix the object goes unreferenced
while after the patch no kmemleak report related to BPF showed up.
Daniel Borkmann [Thu, 16 Aug 2018 19:49:07 +0000 (21:49 +0200)]
tcp, ulp: fix leftover icsk_ulp_ops preventing sock from reattach
I found that in BPF sockmap programs once we either delete a socket
from the map or we updated a map slot and the old socket was purged
from the map that these socket can never get reattached into a map
even though their related psock has been dropped entirely at that
point.
Reason is that tcp_cleanup_ulp() leaves the old icsk->icsk_ulp_ops
intact, so that on the next tcp_set_ulp_id() the kernel returns an
-EEXIST thinking there is still some active ULP attached.
BPF sockmap is the only one that has this issue as the other user,
kTLS, only calls tcp_cleanup_ulp() from tcp_v4_destroy_sock() whereas
sockmap semantics allow dropping the socket from the map with all
related psock state being cleaned up.
Daniel Borkmann [Thu, 16 Aug 2018 19:49:06 +0000 (21:49 +0200)]
tcp, ulp: add alias for all ulp modules
Lets not turn the TCP ULP lookup into an arbitrary module loader as
we only intend to load ULP modules through this mechanism, not other
unrelated kernel modules:
Fix it by adding module alias to TCP ULP modules, so probing module
via request_module() will be limited to tcp-ulp-[name]. The existing
modules like kTLS will load fine given tcp-ulp-tls alias, but others
will fail to load:
Jason Gunthorpe [Thu, 16 Aug 2018 20:13:03 +0000 (14:13 -0600)]
Merge branch 'linus/master' into rdma.git for-next
rdma.git merge resolution for the 4.19 merge window
Conflicts:
drivers/infiniband/core/rdma_core.c
- Use the rdma code and revise with the new spelling for
atomic_fetch_add_unless
drivers/nvme/host/rdma.c
- Replace max_sge with max_send_sge in new blk code
drivers/nvme/target/rdma.c
- Use the blk code and revise to use NULL for ib_post_recv when
appropriate
- Replace max_sge with max_recv_sge in new blk code
net/rds/ib_send.c
- Use the net code and revise to use NULL for ib_post_recv when
appropriate
Yonghong Song [Tue, 14 Aug 2018 18:01:12 +0000 (11:01 -0700)]
bpf: fix a rcu usage warning in bpf_prog_array_copy_core()
Commit 394e40a29788 ("bpf: extend bpf_prog_array to store pointers
to the cgroup storage") refactored the bpf_prog_array_copy_core()
to accommodate new structure bpf_prog_array_item which contains
bpf_prog array itself.
In the old code, we had
perf_event_query_prog_array():
mutex_lock(...)
bpf_prog_array_copy_call():
prog = rcu_dereference_check(array, 1)->progs
bpf_prog_array_copy_core(prog, ...)
mutex_unlock(...)
With the above commit, we had
perf_event_query_prog_array():
mutex_lock(...)
bpf_prog_array_copy_call():
bpf_prog_array_copy_core(array, ...):
item = rcu_dereference(array)->items;
...
mutex_unlock(...)
The new code will trigger a lockdep rcu checking warning.
The fix is to change rcu_dereference() to rcu_dereference_check()
to prevent such a warning.
projects / linux.git / log