Shakeel Butt [Tue, 6 Feb 2018 23:40:59 +0000 (15:40 -0800)]
vfs: remove might_sleep() from clear_inode()
Commit 7994e6f72543 ("vfs: Move waiting for inode writeback from
end_writeback() to evict_inode()") removed inode_sync_wait() from
end_writeback() and commit dbd5768f87ff ("vfs: Rename end_writeback() to
clear_inode()") renamed end_writeback() to clear_inode().
After these patches there is no sleeping operation in clear_inode().
So, remove might_sleep() from it.
Jonathan Haws [Tue, 6 Feb 2018 23:40:52 +0000 (15:40 -0800)]
ipc/mqueue.c: have RT tasks queue in by priority in wq_add()
Previous behavior added tasks to the work queue using the static_prio
value instead of the dynamic priority value in prio. This caused RT tasks
to be added to the work queue in a FIFO manner rather than by priority.
Normal tasks were handled by priority.
This fix utilizes the dynamic priority of the task to ensure that both RT
and normal tasks are added to the work queue in priority order. Utilizing
the dynamic priority (prio) rather than the base priority (normal_prio)
was chosen to ensure that if a task had a boosted priority when it was
added to the work queue, it would be woken sooner to to ensure that it
releases any other locks it may be holding in a more timely manner. It is
understood that the task could have a lower priority when it wakes than
when it was added to the queue in this (unlikely) case.
As described in the title, this patch fixes <ipc>id_ds inconsistency when
<ipc>ctl_stat executes concurrently with some ds-changing function, e.g.
shmat, msgsnd or whatever.
For instance, if shmctl(IPC_STAT) is running concurrently
with shmat, following data structure can be returned:
{... shm_lpid = 0, shm_nattch = 1, ...}
Similarly to type mismatch checks, new GCC 8.x and Clang also changed for
ABI for returns_nonnull checks. While we can update our code to conform
the new ABI it's more reasonable to just remove it. Because it's just
dead code, we don't have any single user of returns_nonnull attribute in
the whole kernel.
And AFAIU the advantage that this attribute could bring would be mitigated
by -fno-delete-null-pointer-checks cflag that we use to build the kernel.
So it's unlikely we will have a lot of returns_nonnull attribute in
future.
Andrey Ryabinin [Tue, 6 Feb 2018 23:40:42 +0000 (15:40 -0800)]
lib/ubsan: add type mismatch handler for new GCC/Clang
UBSAN=y fails to build with new GCC/clang:
arch/x86/kernel/head64.o: In function `sanitize_boot_params':
arch/x86/include/asm/bootparam_utils.h:37: undefined reference to `__ubsan_handle_type_mismatch_v1'
because Clang and GCC 8 slightly changed ABI for 'type mismatch' errors.
Compiler now uses new __ubsan_handle_type_mismatch_v1() function with
slightly modified 'struct type_mismatch_data'.
Let's add new 'struct type_mismatch_data_common' which is independent from
compiler's layout of 'struct type_mismatch_data'. And make
__ubsan_handle_type_mismatch[_v1]() functions transform compiler-dependent
type mismatch data to our internal representation. This way, we can
support both old and new compilers with minimal amount of change.
Dmitry Vyukov [Tue, 6 Feb 2018 23:40:28 +0000 (15:40 -0800)]
kcov: detect double association with a single task
Currently KCOV_ENABLE does not check if the current task is already
associated with another kcov descriptor. As the result it is possible
to associate a single task with more than one kcov descriptor, which
later leads to a memory leak of the old descriptor. This relation is
really meant to be one-to-one (task has only one back link).
Arnd Bergmann [Tue, 6 Feb 2018 23:40:21 +0000 (15:40 -0800)]
pps: parport: use timespec64 instead of timespec
getnstimeofday() is deprecated, so I'm converting this to use
ktime_get_real_ts64() as a safe replacement. I considered using
ktime_get_real() instead, but since the algorithm here depends on the
exact timing, I decided to introduce fewer changes and leave the code
that determines the nanoseconds since the last seconds wrap untouched.
It's not entirely clear to me whether we should also change the time
base to CLOCK_BOOTTIME or CLOCK_TAI. With boottime, we would be
independent of changes due to settimeofday() and only see the speed
adjustment from the upstream clock source, with the downside of having
the signal be at an arbirary offset from the start of the UTC second
signal. With CLOCK_TAI, we would use the same offset from the UTC
second as before and still suffer from settimeofday() adjustments, but
would be less confused during leap seconds.
Both boottime and tai only offer usable (i.e. avoiding ktime_t to
timespec64 conversion) interfaces for ktime_t though, so either way,
changing it wouldn't take significantly more work. CLOCK_MONOTONIC
could be used with ktime_get_ts64(), but would lose synchronization
across a suspend/resume cycle, which seems worse.
Markus Elfring [Tue, 6 Feb 2018 23:39:51 +0000 (15:39 -0800)]
rapidio: improve a size determination in five functions
Replace the specification of data structures by pointer dereferences as
the parameter for the operator "sizeof" to make the corresponding size
determination a bit safer according to the Linux coding style
convention.
This issue was detected by using the Coccinelle software.
When creating a file inside a directory that has the setgid flag set, give
the new file the group ID of the parent, and also the setgid flag if it is
a directory itself.
Arnd Bergmann [Tue, 6 Feb 2018 23:39:21 +0000 (15:39 -0800)]
nilfs2: use time64_t internally
The superblock and segment timestamps are used only internally in nilfs2
and can be read out using sysfs.
Since we are using the old 'get_seconds()' interface and store the data
as timestamps, the behavior differs slightly between 64-bit and 32-bit
kernels, the latter will show incorrect timestamps after 2038 in sysfs,
and presumably fail completely in 2106 as comparisons go wrong.
This changes nilfs2 to use time64_t with ktime_get_real_seconds() to
handle timestamps, making the behavior consistent and correct on both
32-bit and 64-bit machines.
The on-disk format already uses 64-bit timestamps, so nothing changes
there.
Alexey Dobriyan [Tue, 6 Feb 2018 23:39:13 +0000 (15:39 -0800)]
elf: fix NT_FILE integer overflow
If vm.max_map_count bumped above 2^26 (67+ mil) and system has enough RAM
to allocate all the VMAs (~12.8 GB on Fedora 27 with 200-byte VMAs), then
it should be possible to overflow 32-bit "size", pass paranoia check,
allocate very little vmalloc space and oops while writing into vmalloc
guard page...
But I didn't test this, only coredump of regular process.
Joe Perches [Tue, 6 Feb 2018 23:39:06 +0000 (15:39 -0800)]
checkpatch: avoid some false positives for TABSTOP declaration test
Using an open bracket after what seems to be a declaration can also be a
function definition and declaration argument line continuation so remove
the open bracket from the possible declaration/definition matching.
Joe Perches [Tue, 6 Feb 2018 23:39:03 +0000 (15:39 -0800)]
checkpatch: exclude drivers/staging from if with unnecessary parentheses test
Greg KH doesn't like this test so exclude the staging directory from the
implied --strict only test unless --strict is actually used on the
command-line.
Allow lines with URL to exceed the 80 char limit for improved interaction
in adaption to ongoing but undocumented practice.
$ git grep -E '://\S{77}.*' -- '*.[ch]'
As per RFC3986 [1], the URL format allows for alphanum, +, - and .
characters in the scheme before the separator :// as long as it starts
with a letter (e.g. https, git, f.-+).
Recognition of URIs without more context information is prone to false
positives and thus currently left out of the heuristics.
$rawline is used in the check as comments are removed from $line.
Pravin Shedge [Tue, 6 Feb 2018 23:38:42 +0000 (15:38 -0800)]
lib/test_sort.c: add module unload support
test_sort.c performs array-based and linked list sort test. Code allows
to compile either as a loadable modules or builtin into the kernel.
Current code is not allow to unload the test_sort.ko module after
successful completion.
This patch adds support to unload the "test_sort.ko" module by adding
module_exit support.
Previous patch was implemented auto unload support by returning -EAGAIN
from module_init() function on successful case, but this approach is not
ideal.
The auto-unload might seem like a nice optimization, but it encourages
inconsistent behaviour. And behaviour that is different from all other
normal modules.
Vincent Legoll [Tue, 6 Feb 2018 23:38:38 +0000 (15:38 -0800)]
lib/: make RUNTIME_TESTS a menuconfig to ease disabling it all
No need to get into the submenu to disable all related config entries.
This makes it easier to disable all RUNTIME_TESTS config options without
entering the submenu. It will also enable one to see that en/dis-abled
state from the outside menu.
This is only intended to change menuconfig UI, not change the config
dependencies.
Clement Courbet [Tue, 6 Feb 2018 23:38:34 +0000 (15:38 -0800)]
lib: optimize cpumask_next_and()
We've measured that we spend ~0.6% of sys cpu time in cpumask_next_and().
It's essentially a joined iteration in search for a non-zero bit, which is
currently implemented as a lookup join (find a nonzero bit on the lhs,
lookup the rhs to see if it's set there).
Implement a direct join (find a nonzero bit on the incrementally built
join). Also add generic bitmap benchmarks in the new `test_find_bit`
module for new function (see `find_next_and_bit` in [2] and [3] below).
For cpumask_next_and, direct benchmarking shows that it's 1.17x to 14x
faster with a geometric mean of 2.1 on 32 CPUs [1]. No impact on memory
usage. Note that on Arm, the new pure-C implementation still outperforms
the old one that uses a mix of C and asm (`find_next_bit`) [3].
[1] Approximate benchmark code:
```
unsigned long src1p[nr_cpumask_longs] = {pattern1};
unsigned long src2p[nr_cpumask_longs] = {pattern2};
for (/*a bunch of repetitions*/) {
for (int n = -1; n <= nr_cpu_ids; ++n) {
asm volatile("" : "+rm"(src1p)); // prevent any optimization
asm volatile("" : "+rm"(src2p));
unsigned long result = cpumask_next_and(n, src1p, src2p);
asm volatile("" : "+rm"(result));
}
}
```
Yury Norov [Tue, 6 Feb 2018 23:38:31 +0000 (15:38 -0800)]
lib/find_bit_benchmark.c: improvements
As suggested in review comments:
* printk: align numbers using whitespaces instead of tabs;
* return error value from init() to avoid calling rmmod if testing again;
* use ktime_get instead of get_cycles as some arches don't support it;
lib/stackdepot.c: use a non-instrumented version of memcmp()
stackdepot used to call memcmp(), which compiler tools normally
instrument, therefore every lookup used to unnecessarily call instrumented
code. This is somewhat ok in the case of KASAN, but under KMSAN a lot of
time was spent in the instrumentation.
Andy Shevchenko [Tue, 6 Feb 2018 23:38:20 +0000 (15:38 -0800)]
include/linux/bitmap.h: make bitmap_fill() and bitmap_zero() consistent
Behaviour of bitmap_fill() differs from bitmap_zero() in a way how bits
behind bitmap are handed. bitmap_zero() clears entire bitmap by unsigned
long boundary, while bitmap_fill() mimics bitmap_set().
Here we change bitmap_fill() behaviour to be consistent with bitmap_zero()
and add a note to documentation.
The change might reveal some bugs in the code where unused bits are
handled differently and in such cases bitmap_set() has to be used.
Andy Shevchenko [Tue, 6 Feb 2018 23:38:17 +0000 (15:38 -0800)]
lib/test_bitmap.c: clean up test_zero_fill_copy() test case and rename
Since we have separate explicit test cases for bitmap_zero() /
bitmap_clear() and bitmap_fill() / bitmap_set(), clean up
test_zero_fill_copy() to only test bitmap_copy() functionality and thus
rename a function to reflect the changes.
While here, replace bitmap_fill() by bitmap_set() with proper values.
Yury Norov [Tue, 6 Feb 2018 23:38:06 +0000 (15:38 -0800)]
bitmap: replace bitmap_{from,to}_u32array
with bitmap_{from,to}_arr32 over the kernel. Additionally to it:
* __check_eq_bitmap() now takes single nbits argument.
* __check_eq_u32_array is not used in new test but may be used in
future. So I don't remove it here, but annotate as __used.
Yury Norov [Tue, 6 Feb 2018 23:38:02 +0000 (15:38 -0800)]
bitmap: new bitmap_copy_safe and bitmap_{from,to}_arr32
This patchset replaces bitmap_{to,from}_u32array with more simple and
standard looking copy-like functions.
bitmap_from_u32array() takes 4 arguments (bitmap_to_u32array is similar):
- unsigned long *bitmap, which is destination;
- unsigned int nbits, the length of destination bitmap, in bits;
- const u32 *buf, the source; and
- unsigned int nwords, the length of source buffer in ints.
In description to the function it is detailed like:
* copy min(nbits, 32*nwords) bits from @buf to @bitmap, remaining
* bits between nword and nbits in @bitmap (if any) are cleared.
Having two size arguments looks unneeded and potentially dangerous.
It is unneeded because normally user of copy-like function should take
care of the size of destination and make it big enough to fit source
data.
And it is dangerous because function may hide possible error if user
doesn't provide big enough bitmap, and data becomes silently dropped.
That's why all copy-like functions have 1 argument for size of copying
data, and I don't see any reason to make bitmap_from_u32array()
different.
One exception that comes in mind is strncpy() which also provides size
of destination in arguments, but it's strongly argued by the possibility
of taking broken strings in source. This is not the case of
bitmap_{from,to}_u32array().
There is no many real users of bitmap_{from,to}_u32array(), and they all
very clearly provide size of destination matched with the size of
source, so additional functionality is not used in fact. Like this:
bitmap_from_u32array(to->link_modes.supported,
__ETHTOOL_LINK_MODE_MASK_NBITS,
link_usettings.link_modes.supported,
__ETHTOOL_LINK_MODE_MASK_NU32);
Where:
#define __ETHTOOL_LINK_MODE_MASK_NU32 \
DIV_ROUND_UP(__ETHTOOL_LINK_MODE_MASK_NBITS, 32)
In this patch, bitmap_copy_safe and bitmap_{from,to}_arr32 are introduced.
'Safe' in bitmap_copy_safe() stands for clearing unused bits in bitmap
beyond last bit till the end of last word. It is useful for hardening
API when bitmap is assumed to be exposed to userspace.
bitmap_{from,to}_arr32 functions are replacements for
bitmap_{from,to}_u32array. They don't take unneeded nwords argument, and
so simpler in implementation and understanding.
This patch suggests optimization for 32-bit systems - aliasing
bitmap_{from,to}_arr32 to bitmap_copy_safe.
Other possible optimization is aliasing 64-bit LE bitmap_{from,to}_arr32 to
more generic function(s). But I didn't end up with the function that would
be helpful by itself, and can be used to alias 64-bit LE
bitmap_{from,to}_arr32, like bitmap_copy_safe() does. So I preferred to
leave things as is.
The following patch switches kernel to new API and introduces test for it.
Discussion is here: https://lkml.org/lkml/2017/11/15/592
This reverts commit 92266d6ef60c ("async: simplify lowest_in_progress()")
which was simply wrong: In the case where domain is NULL, we now use the
wrong offsetof() in the list_first_entry macro, so we don't actually
fetch the ->cookie value, but rather the eight bytes located
sizeof(struct list_head) further into the struct async_entry.
On 64 bit, that's the data member, while on 32 bit, that's a u64 built
from func and data in some order.
I think the bug happens to be harmless in practice: It obviously only
affects callers which pass a NULL domain, and AFAICT the only such
caller is
and the ASYNC_COOKIE_MAX means that in practice we end up waiting for
the async_global_pending list to be empty - but it would break if
somebody happened to pass (void*)-1 as the data element to
async_schedule, and of course also if somebody ever does a
async_synchronize_cookie_domain(, NULL) with a "finite" cookie value.
Maybe the "harmless in practice" means this isn't -stable material. But
I'm not completely confident my quick git grep'ing is enough, and there
might be affected code in one of the earlier kernels that has since been
removed, so I'll leave the decision to the stable guys.
Kees Cook [Tue, 6 Feb 2018 23:37:45 +0000 (15:37 -0800)]
Makefile: introduce CONFIG_CC_STACKPROTECTOR_AUTO
Nearly all modern compilers support a stack-protector option, and nearly
all modern distributions enable the kernel stack-protector, so enabling
this by default in kernel builds would make sense. However, Kconfig does
not have knowledge of available compiler features, so it isn't safe to
force on, as this would unconditionally break builds for the compilers or
architectures that don't have support. Instead, this introduces a new
option, CONFIG_CC_STACKPROTECTOR_AUTO, which attempts to discover the best
possible stack-protector available, and will allow builds to proceed even
if the compiler doesn't support any stack-protector.
This option is made the default so that kernels built with modern
compilers will be protected-by-default against stack buffer overflows,
avoiding things like the recent BlueBorne attack. Selection of a specific
stack-protector option remains available, including disabling it.
Additionally, tiny.config is adjusted to use CC_STACKPROTECTOR_NONE, since
that's the option with the least code size (and it used to be the default,
so we have to explicitly choose it there now).
Kees Cook [Tue, 6 Feb 2018 23:37:41 +0000 (15:37 -0800)]
Makefile: move stack-protector availability out of Kconfig
Various portions of the kernel, especially per-architecture pieces,
need to know if the compiler is building with the stack protector.
This was done in the arch/Kconfig with 'select', but this doesn't
allow a way to do auto-detected compiler support. In preparation for
creating an on-if-available default, move the logic for the definition of
CONFIG_CC_STACKPROTECTOR into the Makefile.
Kees Cook [Tue, 6 Feb 2018 23:37:38 +0000 (15:37 -0800)]
Makefile: move stack-protector compiler breakage test earlier
In order to make stack-protector failures warn instead of unconditionally
breaking the build, this moves the compiler output sanity-check earlier,
and sets a flag for later testing. Future patches can choose to warn or
fail, depending on the flag value.
Alexey Dobriyan [Tue, 6 Feb 2018 23:37:28 +0000 (15:37 -0800)]
proc: spread likely/unlikely a bit
use_pde() is used at every open/read/write/... of every random /proc
file. Negative refcount happens only if PDE is being deleted by module
(read: never). So it gets "likely".
unuse_pde() gets "unlikely" for the same reason.
close_pdeo() gets unlikely as the completion is filled only if there is a
race between PDE removal and close() (read: never ever).
It even saves code on x86_64 defconfig:
add/remove: 0/0 grow/shrink: 1/2 up/down: 2/-20 (-18)
Function old new delta
close_pdeo 183 185 +2
proc_reg_get_unmapped_area 119 111 -8
proc_reg_poll 85 73 -12
struct proc_dir_entry became bit messy over years:
* move 16-bit ->mode_t before namelen to get rid of padding
* make ->in_use first field: it seems to be most used resulting in
smaller code on x86_64 (defconfig):
* move writeable fields together to a first cacheline (on x86_64),
those include
* ->in_use: reference count, taken every open/read/write/close etc
* ->count: reference count, taken at readdir on every entry
* ->pde_openers: tracks (nearly) every open, dirtied
* ->pde_unload_lock: spinlock protecting ->pde_openers
* ->proc_iops, ->proc_fops, ->data: writeonce fields,
used right together with previous group.
* other rarely written fields go into 1st/2nd and 2nd/3rd cacheline on
32-bit and 64-bit respectively.
Additionally on 32-bit, ->subdir, ->subdir_node, ->namelen, ->name go
fully into 2nd cacheline, separated from writeable fields. They are all
used during lookup.
Heiko Carstens [Tue, 6 Feb 2018 23:37:13 +0000 (15:37 -0800)]
fs/proc/kcore.c: use probe_kernel_read() instead of memcpy()
Commit df04abfd181a ("fs/proc/kcore.c: Add bounce buffer for ktext
data") added a bounce buffer to avoid hardened usercopy checks. Copying
to the bounce buffer was implemented with a simple memcpy() assuming
that it is always valid to read from kernel memory iff the
kern_addr_valid() check passed.
A simple, but pointless, test case like "dd if=/proc/kcore of=/dev/null"
now can easily crash the kernel, since the former execption handling on
invalid kernel addresses now doesn't work anymore.
Also adding a kern_addr_valid() implementation wouldn't help here. Most
architectures simply return 1 here, while a couple implemented a page
table walk to figure out if something is mapped at the address in
question.
With DEBUG_PAGEALLOC active mappings are established and removed all the
time, so that relying on the result of kern_addr_valid() before
executing the memcpy() also doesn't work.
Therefore simply use probe_kernel_read() to copy to the bounce buffer.
This also allows to simplify read_kcore().
At least on s390 this fixes the observed crashes and doesn't introduce
warnings that were removed with df04abfd181a ("fs/proc/kcore.c: Add
bounce buffer for ktext data"), even though the generic
probe_kernel_read() implementation uses uaccess functions.
While looking into this I'm also wondering if kern_addr_valid() could be
completely removed...(?)
: This patch breaks criu. It was a bug in criu. And this bug is on a minor
: path, which works when memfd_create() isn't available. It is a reason why
: I ask to not backport this patch to stable kernels.
:
: In CRIU this bug can be triggered, only if this patch will be backported
: to a kernel which version is lower than v3.16.
Colin Ian King [Tue, 6 Feb 2018 23:36:48 +0000 (15:36 -0800)]
kasan: remove redundant initialization of variable 'real_size'
Variable real_size is initialized with a value that is never read, it is
re-assigned a new value later on, hence the initialization is redundant
and can be removed.
Cleans up clang warning:
lib/test_kasan.c:422:21: warning: Value stored to 'real_size' during its initialization is never read
Dmitry Vyukov [Tue, 6 Feb 2018 23:36:34 +0000 (15:36 -0800)]
kasan: unify code between kasan_slab_free() and kasan_poison_kfree()
Both of these functions deal with freeing of slab objects.
However, kasan_poison_kfree() mishandles SLAB_TYPESAFE_BY_RCU
(must also not poison such objects) and does not detect double-frees.
Unify code between these functions.
This solves both of the problems and allows to add more common code
(e.g. detection of invalid frees).
Dmitry Vyukov [Tue, 6 Feb 2018 23:36:30 +0000 (15:36 -0800)]
kasan: detect invalid frees for large mempool objects
Detect frees of pointers into middle of mempool objects.
I did a one-off test, but it turned out to be very tricky, so I reverted
it. First, mempool does not call kasan_poison_kfree() unless allocation
function fails. I stubbed an allocation function to fail on second and
subsequent allocations. But then mempool stopped to call
kasan_poison_kfree() at all, because it does it only when allocation
function is mempool_kmalloc(). We could support this special failing
test allocation function in mempool, but it also can't live with kasan
tests, because these are in a module.
Dmitry Vyukov [Tue, 6 Feb 2018 23:36:23 +0000 (15:36 -0800)]
kasan: detect invalid frees for large objects
Patch series "kasan: detect invalid frees".
KASAN detects double-frees, but does not detect invalid-frees (when a
pointer into a middle of heap object is passed to free). We recently had
a very unpleasant case in crypto code which freed an inner object inside
of a heap allocation. This left unnoticed during free, but totally
corrupted heap and later lead to a bunch of random crashes all over kernel
code.
Detect invalid frees.
This patch (of 5):
Detect frees of pointers into middle of large heap objects.
I dropped const from kasan_kfree_large() because it starts propagating
through a bunch of functions in kasan_report.c, slab/slub nearest_obj(),
all of their local variables, fixup_red_left(), etc.
kasan: add functions for unpoisoning stack variables
As a code-size optimization, LLVM builds since r279383 may bulk-manipulate
the shadow region when (un)poisoning large memory blocks. This requires
new callbacks that simply do an uninstrumented memset().
This fixes linking the Clang-built kernel when using KASAN.
Paul Lawrence [Tue, 6 Feb 2018 23:36:11 +0000 (15:36 -0800)]
kasan: support alloca() poisoning
clang's AddressSanitizer implementation adds redzones on either side of
alloca()ed buffers. These redzones are 32-byte aligned and at least 32
bytes long.
__asan_alloca_poison() is passed the size and address of the allocated
buffer, *excluding* the redzones on either side. The left redzone will
always be to the immediate left of this buffer; but AddressSanitizer may
need to add padding between the end of the buffer and the right redzone.
If there are any 8-byte chunks inside this padding, we should poison
those too.
__asan_allocas_unpoison() is just passed the top and bottom of the dynamic
stack area, so unpoisoning is simpler.
Andrey Ryabinin [Tue, 6 Feb 2018 23:36:08 +0000 (15:36 -0800)]
kasan/Makefile: support LLVM style asan parameters
LLVM doesn't understand GCC-style paramters ("--param asan-foo=bar"), thus
we currently we don't use inline/globals/stack instrumentation when
building the kernel with clang.
Add support for LLVM-style parameters ("-mllvm -asan-foo=bar") to enable
all KASAN features.
Paul Lawrence [Tue, 6 Feb 2018 23:36:03 +0000 (15:36 -0800)]
kasan: add compiler support for clang
Patch series "kasan: support alloca, LLVM", v4.
This patch (of 5):
For now we can hard-code ASAN ABI level 5, since historical clang builds
can't build the kernel anyway. We also need to emulate gcc's
__SANITIZE_ADDRESS__ flag, or memset() calls won't be instrumented.
kasan: don't emit builtin calls when sanitization is off
With KASAN enabled the kernel has two different memset() functions, one
with KASAN checks (memset) and one without (__memset). KASAN uses some
macro tricks to use the proper version where required. For example
memset() calls in mm/slub.c are without KASAN checks, since they operate
on poisoned slab object metadata.
The issue is that clang emits memset() calls even when there is no
memset() in the source code. They get linked with improper memset()
implementation and the kernel fails to boot due to a huge amount of KASAN
reports during early boot stages.
The solution is to add -fno-builtin flag for files with KASAN_SANITIZE :=
n marker.
Every flow_offload entry is added into the table twice. Because of this,
rhashtable_free_and_destroy can't be used, since it would call kfree for
each flow_offload object twice.
This patch cleans up the flowtable via nf_flow_table_iterate() to
schedule removal of entries by setting on the dying bit, then there is
an explicitly invocation of the garbage collector to release resources.
Cong Wang [Mon, 5 Feb 2018 22:41:45 +0000 (14:41 -0800)]
netfilter: xt_RATEEST: acquire xt_rateest_mutex for hash insert
rateest_hash is supposed to be protected by xt_rateest_mutex,
and, as suggested by Eric, lookup and insert should be atomic,
so we should acquire the xt_rateest_mutex once for both.
So introduce a non-locking helper for internal use and keep the
locking one for external.
Linus Torvalds [Tue, 6 Feb 2018 23:30:52 +0000 (15:30 -0800)]
Merge tag 'platform-drivers-x86-v4.16-1' of git://git.infradead.org/linux-platform-drivers-x86
Pull x86 platform-driver updates from Darren Hart:
"New model support added for Dell, Ideapad, Acer, Asus, Thinkpad, and
GPD laptops. Improvements to the common intel-vbtn driver, including
tablet mode, rotate, and front button support. Intel CPU support added
for Cannonlake and platform support for Dollar Cove power button.
Overhaul of the mellanox platform driver, creating a new
platform/mellanox directory for the newly multi-architecture regmap
interface.
Significant Intel PMC update with CannonLake support, Coffeelake
update, CPUID enumeration, module support, new read64 API, refactoring
and cleanups.
Revert the apple-gmux iGP IO lock, addressing reported issues with
non-binary drivers, leaving Nvidia binary driver users to comment out
conflicting code.
Miscellaneous fixes and cleanups"
* tag 'platform-drivers-x86-v4.16-1' of git://git.infradead.org/linux-platform-drivers-x86: (81 commits)
platform/x86: mlx-platform: Fix an ERR_PTR vs NULL issue
platform/x86: intel_pmc_core: Special case for Coffeelake
platform/x86: intel_pmc_core: Add CannonLake PCH support
x86/cpu: Add Cannonlake to Intel family
platform/x86: intel_pmc_core: Read base address from LPIT
ACPI / LPIT: Export lpit_read_residency_count_address()
platform/x86: intel-vbtn: Replace License by SDPX identifier
platform/x86: intel-vbtn: Remove redundant inclusions
platform/x86: intel-vbtn: Support tablet mode switch
platform/x86: dell-laptop: Allocate buffer on heap rather than globally
platform/x86: intel_pmc_core: Remove unused header file
platform/x86: mlx-platform: Add hotplug device unregister to error path
platform/x86: mlx-platform: fix module aliases
platform/mellanox: mlxreg-hotplug: Add check for negative adapter number
platform/x86: mlx-platform: Add IO access verification callbacks
platform/x86: mlx-platform: Document pdev_hotplug field
platform/x86: mlx-platform: Allow compilation for 32 bit arch
platform/mellanox: mlxreg-hotplug: Enable building for ARM
platform/mellanox: mlxreg-hotplug: Modify to use a regmap interface
platform/mellanox: Group create/destroy with attribute functions
...
Linus Torvalds [Tue, 6 Feb 2018 23:04:58 +0000 (15:04 -0800)]
Merge branch 'next' of git://git.kernel.org/pub/scm/linux/kernel/git/rzhang/linux
Pull thermal management updates from Zhang Rui:
- fix a race condition issue in power allocator governor (Yi Zeng).
- add support for AP806 and CP110 in armada thermal driver, together
with several improvements (Baruch Siach, Miquel Raynal)
- add support for r8z7743 in rcar thermal driver (Biju Das)
- convert thermal core to use new hwmon API to avoid warning (Fabio
Estevam)
- small fixes and cleanups in thermal core and x86_pkg_thermal,
int3400_thermal, hisi_thermal, mtk_thermal and imx_thermal drivers
(Pravin Shedge, Geert Uytterhoeven, Alexey Khoroshilov, Brian Bian,
Matthias Brugger, Nicolin Chen, Uwe Kleine-König)
* 'next' of git://git.kernel.org/pub/scm/linux/kernel/git/rzhang/linux: (25 commits)
thermal: thermal_hwmon: Convert to hwmon_device_register_with_info()
thermal/x86 pkg temp: Remove debugfs_create_u32() casts
thermal: int3400_thermal: fix error handling in int3400_thermal_probe()
thermal/drivers/hisi: Remove bogus const from function return type
thermal: armada: Give meaningful names to the thermal zones
thermal: armada: Wait sensors validity before exiting the init callback
thermal: armada: Change sensors trim default value
thermal: armada: Update Kconfig and module description
thermal: armada: Add support for Armada CP110
thermal: armada: Add support for Armada AP806
thermal: armada: Use real status register name
thermal: armada: Clarify control registers accesses
thermal: armada: Simplify the check of the validity bit
thermal: armada: Use msleep for long delays
dt-bindings: thermal: Describe Armada AP806 and CP110
dt-bindings: thermal: rcar: Add device tree support for r8a7743
thermal: mtk: Cleanup unused defines
thermal: imx: update to new formula according to NXP AN5215
thermal: imx: use consistent style to write temperatures
thermal: imx: improve comments describing algorithm for temp calculation
...
Linus Torvalds [Tue, 6 Feb 2018 19:27:48 +0000 (11:27 -0800)]
Merge tag 'media/v4.16-2' of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-media
Pull media updates from Mauro Carvalho Chehab:
- videobuf2 was moved to a media/common dir, as it is now used by the
DVB subsystem too
- Digital TV core memory mapped support interface
- new sensor driver: ov7740
- several improvements at ddbridge driver
- new V4L2 driver: IPU3 CIO2 CSI-2 receiver unit, found on some Intel
SoCs
- new tuner driver: tda18250
- finally got rid of all LIRC staging drivers
- as we don't have old lirc drivers anymore, restruct the lirc device
code
- add support for UVC metadata
- add a new staging driver for NVIDIA Tegra Video Decoder Engine
- DVB kAPI headers moved to include/media
- synchronize the kAPI and uAPI for the DVB subsystem, removing the gap
for non-legacy APIs
- reduce the kAPI gap for V4L2
- lots of other driver enhancements, cleanups, etc.
* tag 'media/v4.16-2' of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-media: (407 commits)
media: v4l2-compat-ioctl32.c: make ctrl_is_pointer work for subdevs
media: v4l2-compat-ioctl32.c: refactor compat ioctl32 logic
media: v4l2-compat-ioctl32.c: don't copy back the result for certain errors
media: v4l2-compat-ioctl32.c: drop pr_info for unknown buffer type
media: v4l2-compat-ioctl32.c: copy clip list in put_v4l2_window32
media: v4l2-compat-ioctl32.c: fix ctrl_is_pointer
media: v4l2-compat-ioctl32.c: copy m.userptr in put_v4l2_plane32
media: v4l2-compat-ioctl32.c: avoid sizeof(type)
media: v4l2-compat-ioctl32.c: move 'helper' functions to __get/put_v4l2_format32
media: v4l2-compat-ioctl32.c: fix the indentation
media: v4l2-compat-ioctl32.c: add missing VIDIOC_PREPARE_BUF
media: v4l2-ioctl.c: don't copy back the result for -ENOTTY
media: v4l2-ioctl.c: use check_fmt for enum/g/s/try_fmt
media: vivid: fix module load error when enabling fb and no_error_inj=1
media: dvb_demux: improve debug messages
media: dvb_demux: Better handle discontinuity errors
media: cxusb, dib0700: ignore XC2028_I2C_FLUSH
media: ts2020: avoid integer overflows on 32 bit machines
media: i2c: ov7740: use gpio/consumer.h instead of gpio.h
media: entity: Add a nop variant of media_entity_cleanup
...
Linus Torvalds [Tue, 6 Feb 2018 19:09:45 +0000 (11:09 -0800)]
Merge tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma
Pull more rdma updates from Doug Ledford:
"Items of note:
- two patches fix a regression in the 4.15 kernel. The 4.14 kernel
worked fine with NVMe over Fabrics and mlx5 adapters. That broke in
4.15. The fix is here.
- one of the patches (the endian notation patch from Lijun) looks
like a lot of lines of change, but it's mostly mechanical in
nature. It amounts to the biggest chunk of change in it (it's about
2/3rds of the overall pull request).
Summary:
- Clean up some function signatures in rxe for clarity
- Tidy the RDMA netlink header to remove unimplemented constants
- bnxt_re driver fixes, one is a regression this window.
- Minor hns driver fixes
- Various fixes from Dan Carpenter and his tool
- Fix IRQ cleanup race in HFI1
- HF1 performance optimizations and a fix to report counters in the right units
- Fix for an IPoIB startup sequence race with the external manager
- Oops fix for the new kabi path
- Endian cleanups for hns
- Fix for mlx5 related to the new automatic affinity support"
* tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma: (38 commits)
net/mlx5: increase async EQ to avoid EQ overrun
mlx5: fix mlx5_get_vector_affinity to start from completion vector 0
RDMA/hns: Fix the endian problem for hns
IB/uverbs: Use the standard kConfig format for experimental
IB: Update references to libibverbs
IB/hfi1: Add 16B rcvhdr trace support
IB/hfi1: Convert kzalloc_node and kcalloc to use kcalloc_node
IB/core: Avoid a potential OOPs for an unused optional parameter
IB/core: Map iWarp AH type to undefined in rdma_ah_find_type
IB/ipoib: Fix for potential no-carrier state
IB/hfi1: Show fault stats in both TX and RX directions
IB/hfi1: Remove blind constants from 16B update
IB/hfi1: Convert PortXmitWait/PortVLXmitWait counters to flit times
IB/hfi1: Do not override given pcie_pset value
IB/hfi1: Optimize process_receive_ib()
IB/hfi1: Remove unnecessary fecn and becn fields
IB/hfi1: Look up ibport using a pointer in receive path
IB/hfi1: Optimize packet type comparison using 9B and bypass code paths
IB/hfi1: Compute BTH only for RDMA_WRITE_LAST/SEND_LAST packet
IB/hfi1: Remove dependence on qp->s_hdrwords
...
Linus Torvalds [Tue, 6 Feb 2018 18:41:33 +0000 (10:41 -0800)]
Merge tag 'libnvdimm-for-4.16' of git://git.kernel.org/pub/scm/linux/kernel/git/nvdimm/nvdimm
Pull libnvdimm updates from Ross Zwisler:
- Require struct page by default for filesystem DAX to remove a number
of surprising failure cases. This includes failures with direct I/O,
gdb and fork(2).
- Add support for the new Platform Capabilities Structure added to the
NFIT in ACPI 6.2a. This new table tells us whether the platform
supports flushing of CPU and memory controller caches on unexpected
power loss events.
- Revamp vmem_altmap and dev_pagemap handling to clean up code and
better support future future PCI P2P uses.
- Deprecate the ND_IOCTL_SMART_THRESHOLD command whose payload has
become out-of-sync with recent versions of the NVDIMM_FAMILY_INTEL
spec, and instead rely on the generic ND_CMD_CALL approach used by
the two other IOCTL families, NVDIMM_FAMILY_{HPE,MSFT}.
- Enhance nfit_test so we can test some of the new things added in
version 1.6 of the DSM specification. This includes testing firmware
download and simulating the Last Shutdown State (LSS) status.
* tag 'libnvdimm-for-4.16' of git://git.kernel.org/pub/scm/linux/kernel/git/nvdimm/nvdimm: (37 commits)
libnvdimm, namespace: remove redundant initialization of 'nd_mapping'
acpi, nfit: fix register dimm error handling
libnvdimm, namespace: make min namespace size 4K
tools/testing/nvdimm: force nfit_test to depend on instrumented modules
libnvdimm/nfit_test: adding support for unit testing enable LSS status
libnvdimm/nfit_test: add firmware download emulation
nfit-test: Add platform cap support from ACPI 6.2a to test
libnvdimm: expose platform persistence attribute for nd_region
acpi: nfit: add persistent memory control flag for nd_region
acpi: nfit: Add support for detect platform CPU cache flush on power loss
device-dax: Fix trailing semicolon
libnvdimm, btt: fix uninitialized err_lock
dax: require 'struct page' by default for filesystem dax
ext2: auto disable dax instead of failing mount
ext4: auto disable dax instead of failing mount
mm, dax: introduce pfn_t_special()
mm: Fix devm_memremap_pages() collision handling
mm: Fix memory size alignment in devm_memremap_pages_release()
memremap: merge find_dev_pagemap into get_dev_pagemap
memremap: change devm_memremap_pages interface to use struct dev_pagemap
...
Linus Torvalds [Tue, 6 Feb 2018 17:59:40 +0000 (09:59 -0800)]
Merge tag 'pci-v4.16-changes' of git://git.kernel.org/pub/scm/linux/kernel/git/helgaas/pci
Pull PCI updates from Bjorn Helgaas:
- skip AER driver error recovery callbacks for correctable errors
reported via ACPI APEI, as we already do for errors reported via the
native path (Tyler Baicar)
- rework Tegra config space mapping to increase space available for
endpoints (Vidya Sagar)
- simplify Tegra driver by using bus->sysdata (Manikanta Maddireddy)
- remove PCI_REASSIGN_ALL_BUS usage on Tegra (Manikanta Maddireddy)
- add support for Global Fabric Manager Server (GFMS) event to
Microsemi Switchtec switch driver (Logan Gunthorpe)
- add IDs for Switchtec PSX 24xG3 and PSX 48xG3 (Kelvin Cao)
* tag 'pci-v4.16-changes' of git://git.kernel.org/pub/scm/linux/kernel/git/helgaas/pci: (140 commits)
PCI: cadence: Add EndPoint Controller driver for Cadence PCIe controller
dt-bindings: PCI: cadence: Add DT bindings for Cadence PCIe endpoint controller
PCI: endpoint: Fix EPF device name to support multi-function devices
PCI: endpoint: Add the function number as argument to EPC ops
PCI: cadence: Add host driver for Cadence PCIe controller
dt-bindings: PCI: cadence: Add DT bindings for Cadence PCIe host controller
PCI: Add vendor ID for Cadence
PCI: Add generic function to probe PCI host controllers
PCI: generic: fix missing call of pci_free_resource_list()
PCI: OF: Add generic function to parse and allocate PCI resources
PCI: Regroup all PCI related entries into drivers/pci/Makefile
PCI/DPC: Reformat DPC register definitions
PCI/DPC: Add and use DPC Status register field definitions
PCI/DPC: Squash dpc_rp_pio_get_info() into dpc_process_rp_pio_error()
PCI/DPC: Remove unnecessary RP PIO register structs
PCI/DPC: Push dpc->rp_pio_status assignment into dpc_rp_pio_get_info()
PCI/DPC: Squash dpc_rp_pio_print_error() into dpc_rp_pio_get_info()
PCI/DPC: Make RP PIO log size check more generic
PCI/DPC: Rename local "status" to "dpc_status"
PCI/DPC: Squash dpc_rp_pio_print_tlp_header() into dpc_rp_pio_print_error()
...
Suresh Reddy [Tue, 6 Feb 2018 13:52:42 +0000 (08:52 -0500)]
be2net: Handle transmit completion errors in Lancer
If the driver receives a TX CQE with status as 0x1 or 0x9 or 0xb,
the completion indexes should not be used. The driver must stop
consuming CQEs from this TXQ/CQ. The TXQ from this point on-wards
to be in a bad state. Driver should destroy and recreate the TXQ.
projects / linux.git / log