Modify the kernel to use only atomic_xx and atomic64_xx interfaces,
avoiding the use of sizeof or typeof to determine the type of
atomic operations, thereby simplifying the kernel's atomic
interface operations.
Signed-off-by: zhangyuan29 <zhangyuan29@xiaomi.com>
Most tools used for compliance and SBOM generation use SPDX identifiers
This change brings us a step closer to an easy SBOM generation.
Signed-off-by: Alin Jerpelea <alin.jerpelea@sony.com>
The problem has been inherited from the original libc readv/writev
implementation. However, now it's exposed in more situations because
this implemenation is used to back read/write as well.
I expect this fixes the regressions observed on the Espressif CI.
https://github.com/apache/nuttx/pull/13498#issuecomment-2448031197
Note that, even with this fix, these "compat" readv/writev
implementations are still inheritedly broken. (E.g. consider that
a data boundary happens to match one of iovec boundaries) However,
this fix is enough for read/write, where iovcnt is always 1.
This patch fixed userspace headers conflict. Architecture-related definition and API should not be exposed to users.
Signed-off-by: ouyangxiangzhen <ouyangxiangzhen@xiaomi.com>
currently, nuttx implements readv/writev on the top of read/write.
while it might work for the simplest cases, it's broken by design.
for example, it's impossible to make it work correctly for files
which need to preserve data boundaries without allocating a single
contiguous buffer. (udp socket, some character devices, etc)
this change is a start of the migration to a better design.
that is, implement read/write on the top of readv/writev.
to avoid a single huge change, following things will NOT be done in
this commit:
* fix actual bugs caused by the original readv-based-on-read design.
(cf. https://github.com/apache/nuttx/pull/12674)
* adapt filesystems/drivers to actually benefit from the new interface.
(except a few trivial examples)
* eventually retire the old interface.
* retire read/write syscalls. implement them in libc instead.
* pread/pwrite/preadv/pwritev (except the introduction of struct uio,
which is a preparation to back these variations with the new
interface.)
The man manual describes that poll only has EFAULT, EINTR, EINVAL, and ENOMEM return values.
If a file returns an error, the POLLERR event should be set and OK should be returned
https://man7.org/linux/man-pages/man2/poll.2.html
When using libuv to poll the socket, the socket poll returned an EBUSY error, causing libuv to abort.
The expected logic should be to return OK, allowing libuv to notify the event listener that the POLLERR event occurred.
Signed-off-by: yinshengkai <yinshengkai@xiaomi.com>
Signed-off-by: ligd <liguiding1@xiaomi.com>
timeout is a type of int
in MSEC2TICK, timeout may be multiplied by USEC_PER_MSEC
and the result may be out of range of int
Besides, this will induce following ERROR while running simulator:
vfs/fs_poll.c:498:38: runtime error: signed integer overflow:
19768268 * 1000 cannot be represented in type 'int'
Signed-off-by: zhengyu9 <zhengyu9@xiaomi.com>
Signed-off-by: ligd <liguiding1@xiaomi.com>
Summary:
Reference https://man7.org/linux/man-pages/man2/open.2.html
EISDIR pathname refers to a directory and the access requested
involved writing (that is, O_WRONLY or O_RDWR is set).
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
Close operation on file should lead to IN_CLOSE_WRITE or
IN_CLOSE_NOWRITE notifications. This commits adds the notification
support. Notifying on close is a little bit trickier as a lower layer
may not have the full file path after successful close and inode release.
Calling notification before close is not a solution since close might
not end successfully.
The solution is to obtain and buffer the path before calling close
and then pass the buffered path to the notify_close. This required the
change in notify_close function arguments: filep is no longer
required, path and oflags are passed instead.
Signed-off-by: Michal Lenc <michallenc@seznam.cz>
IN_MODIFY event should occur on file modification, which includes
truncate. This is consistent with the inotify usage on Linux.
Signed-off-by: Michal Lenc <michallenc@seznam.cz>
Summary:
Fixed the problem of releasing the bucket prematurely in multi-threaded flock scenarios.
A thread setlk
B thread setlk_wait
A thread releases lock but fails to determine if nwaiter causes the bucket to be released prematurely
post B thread causes crash due to heap use after free
https://github.com/apache/nuttx/issues/13821
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
Example:
When executing "df -h" on Core A to view mount information, this
process will traverse inode nodes, thereby holding the inode_lock.
Since the inode type of the mount point may be rpmsgfs, it will fetch statfs
information from another Core B.
Meanwhile, rcS on Core B needs to obtain file information from Core A,
which will be achieved by fetching stat information through rpmsgfs.
When this message arrives at Core A, a deadlock can occur between Core A's
rptun ap and nsh task.
However, both of these places involve read operations only, thus a reader-writer lock
can be utilized to prevent such a deadlock.
Signed-off-by: dongjiuzhu1 <dongjiuzhu1@xiaomi.com>
Summary:
1.Add configuration to allocate memory from the specified section
2.Replace all memory operations (kmm_) in the vfs with
fs_heap_. When FS_HEAPSIZE > 0, memory is requested for the file system by specifying a configured heap location. By default (i.e. FS_HEAPSIZE=0) fs_heap_ is equivalent to kmm_
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
Summary:
1.Modified the i_crefs from int16_t to atomic_int
2.Modified the i_crefs add, delete, read, and initialize interfaces to atomic operations
The purpose of this change is to avoid deadlock in cross-core scenarios, where A Core blocks B Core’s request for a write operation to A Core when A Core requests a read operation to B Core.
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
Using the ts/tick conversion functions provided in clock.h
Do this caused we want speed up the time calculation, so change:
clock_time2ticks, clock_ticks2time, clock_timespec_add,
clock_timespec_compare, clock_timespec_subtract... to MACRO
Signed-off-by: ligd <liguiding1@xiaomi.com>
For some case, need large files from tmpfs or shmfs, will lead to high
pressure on memory fragments, add an optional fs_heap with independent
heap will benifit for memory fragments issue.
Signed-off-by: buxiasen <buxiasen@xiaomi.com>
Summary:
Indicate whether the file is currently locked by adding a new field locked to filep.
0 - Unlocked
1 - Locked
The status of the filep at close is used to determine whether to continue with the following procedure.
Optimizing performance:
Before
Time taken to close the file: 33984 nsec
After
Time taken to close the file: 23744 nsec
Improvement of about 10 msec
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
