pointer comparison is unsigned, when returning -errno will be converted
to a large positive number, can not enter the error handling branch,
therefore, the error code is returned directly and the sem is returned
through the parameters.
Signed-off-by: zhanghongyu <zhanghongyu@xiaomi.com>
Implement the bzero function as an alternative to macro expansion.
and support gcc FORTIFY_SOURCE features for nuttx libc
This function will use gcc's function
__builtin_dynamic_object_size and __builtin_object_size
Its function is to obtain the size of the object through compilation,
so as to judge whether there are out-of-bounds operations in commonly used functions.
It should be noted that the option -O2 and above is required to enable this function
Signed-off-by: zhanghongyu <zhanghongyu@xiaomi.com>
In the gcc-riscv compiler, __builtin_ffs will call ffs, and calling __builtin_ffs in ffs will cause recursion
Change ffs to an inline function, and compile ffs implemented by nuttx by default.
Only call the implementation of nuttx when the compiler cannot provide an ffs implementation.
Signed-off-by: yinshengkai <yinshengkai@xiaomi.com>
This commit adds support for reallocarray function. The functionality
is the same as for standard realloc, but the function also checks for
multiplication overflow and fails safely in case it occurs.
Signed-off-by: Michal Lenc <michallenc@seznam.cz>
The current macro guard for TLS related symbols is
not enough, only if !defined(CONFIG_DISABLE_PTHREAD)
and CONFIG_TLS_NELEM > 0, the TLS related symbols
will enabled.
So if CONFIG_DISABLE_PTHREAD is not defined, but
TLS is disabled, the reference to TLS related symbols
will cause build error (undefined reference to xxx).
Signed-off-by: Huang Qi <huangqi3@xiaomi.com>
Don't include the build timestamp into final binary when the symbol
CONFIG_LIBC_UNAME_DISABLE_TIMESTAMP is defined.
Signed-off-by: Alan C. Assis <acassis@gmail.com>
This fixes calls like memmem("hello", 5, "lo", 2);
Also zero-length needle is deemed to exist at beginning of haystack.
This behavior matches memmem() on Linux, FreeBSD, NetBSD and OpenBSD.
Signed-off-by: Juha Niskanen <juha.niskanen@haltian.com>
The maximum startup parameters have been checked accordingly in nxtask_setup_stackargs(),
let us save argument counter to avoid limit check.
Signed-off-by: chao an <anchao@lixiang.com>
I recently imported NuttX version 6.0 (and nsh) into a Microchip
Studio project [1] on Windows to figure out what was going wrong with
the avr32dev1 build. I also briefly checked NuttX version 10.
I worked with the assumption that the avr32 (avr32dev1) specific
changes to the codebase were minimal across NuttX releases.
For the initial proof of concept I used Microchip Studio version 7.0
(with the recent Microchip's ASF updates). I use avr32-gcc (4.4.7)
hosted here [2] for building NuttX for avr32dev1 on GNU/Linux.
Even with the Microchip Studio project, I had initial debug problems
with just stepping through the code a line at a time. I had to bring
in crt0, a trampoline stub and the linker file from one of my older
projects to really build on the suspicion I had with the linker file.
Perhaps an older version of avr32-gcc did something differently. I am
not sure about this. I used avr32-objdump to see the output sections
of the generated elf file. I just had to tweak the linker script to
ensure correct linking of the sections.
With those changes, I was able to inspect the UART sections within
NuttX Microchip Studio project.
Second important change: the transmit pin: I had to reassign the pin
to see the nsh console.
These are the currently assigned UART pins:
RX: PA_24 -> Physical IC pin 59
TX: PB_02 -> Physical IC pin 24
For the avr32dev1 board, they are pins: J1 (berg pin 28) and J2 (berg
pin 10).
In addition, the PR fixes silly compilation problems with avr32dev1.
I have tested the nsh build with my avr32dev1 boards. I used Atmel ICE
to program one of them (flash at 0x80000000) and dfu-programmer to
test my other board (flash at 0x80002000). The other RS-232 parameters
are the same as they were.
References:
[1]: https://github.com/ramangopalan/nuttx_avr32dev1
[2]: https://github.com/ramangopalan/avr32-gnu-toolchain-linux_x86_64
The lowest bit of the thumb instruction is 1 by default, which is used to distinguish arm instructions and thumb instructions.
Fixed the problem of misalignment of symbol table when performing binary search
In arm, the lowest bit of the instruction is 1, which is a thumb instruction, and 0, which is an arm instruction.
The nm command was used in mkallsym.sh before, and the result it will return will set the lowest bit of the thumb instruction to 0. There will be a one-byte deviation during binary search, so mkallsyms.py will also set the lowest bit to 0 according to the previous format.
```sh
arm-none-eabi-nm -Cn nuttx | grep hello
0801c384 T hello_main
arm-none-eabi-objdump nuttx -t |grep hello
0801c384 g F .text 0000004c hello_main
arm-none-eabi-readelf nuttx -s |grep hello
4558: 0801c385 76 FUNC GLOBAL DEFAULT 1 hello_main
```
However, in the following case, when you need to find the function address according to the symbol name and execute the corresponding function, the lowest address obtained is 0. It will follow the arm instruction, causing an exception.
```c
void sym_test(void)
{
printf("call sym_test\n");
}
int main(int argc, FAR char *argv[])
{
FAR void *addr = sym_test;
printf("sym_test:%p %pS\n",addr, addr);
printf("sym_test - 1: %pS\n", (char *)addr - 1);
printf("sym_test + 1: %pS\n", (char *)addr + 1);
size_t size;
void (*func)(void);
const struct symtab_s *sym = allsyms_findbyname("sym_test", &size);
printf("sym_test:%p %pS\n",sym, sym);
func = sym->sym_value;
func();
return 0;
}
```
Therefore, you need to change mkallsyms.py back to the correct result and correct the binary search.
Signed-off-by: yinshengkai <yinshengkai@xiaomi.com>
As pointed out in #11322 there is a hardware design issue in RISC-V that
affects RV64 relocations. The problem is with how address bits are loaded
into registers via lui / auipc and sign extension.
If the hi20 relocation value happens to have its 32-bit sign bit set, i.e.
value is 0x80000000 (but not negative! i.e. negative in 64-bit format) the
relocation will fail, as the address is erroneously sign extended:
0x00000000_80000000 becomes 0xffffffff_80000000 which is not correct.
Also, make sure the correct opcode is used with PCREL_HI20, it expects
AUIPC (not LUI). The C compiler will never emit such code but when hand-
writing assembly code this can happen.
There is a problem with the current elf loader for risc-v: when a pair of
PCREL_HI20 / LO12 relocations are encountered, it is assumed that these
will follow each other immediately, as follows:
label:
auipc a0, %pcrel_hi(symbol) // R_RISCV_PCREL_HI20
load/store a0, %pcrel_lo(label)(a0) // R_RISCV_PCREL_LO12_I/S
With this assumption, the hi/lo relocations are both done when a hi20
relocation entry is encountered, first to the current instruction (addr)
and to the next instruction (addr + 4).
However, this assumption is wrong. There is nothing in the elf relocation
specification[1] that mandates this. Thus, the hi/lo relocation always
needs to first fixup the hi-part, and when the lo-part is encountered, it
needs to find the corresponding hi relocation entry, via the given "label".
This necessitates (re-)visiting the relocation entries for the current
section as well as looking for "label" in the symbol table.
The NuttX elf loader does not allow such operations to be done in the
machine specific part, so this patch fixes the relocation issue by
introducing an architecture specific cache for the hi20 relocation and
symbol table entries. When a lo12 relocation is encountered, the cache
can be consulted to find the hi20 part.
[1] https://github.com/riscv-non-isa/riscv-elf-psabi-doc/blob/master/riscv-elf.adoc
we can use uart to debug nuttx,like debugger:
1. read/write memory
2. Use watchpoint,breakpoint,single step.
use up_debugpoint api
3. Ctrl+c to stop, continue, or single step.
hold uart send and receive
4. register a panic event, when crash or assert/panic, we use uart to
debug.
Signed-off-by: anjiahao <anjiahao@xiaomi.com>
Add registration function instrumentation API,
which can achieve instrumentation of entering and
exiting functions through the compiler's functionality.
We can use CONFIG_ARCH_INSTRUMENT_ALL to add instrumentation for all
source, or add '-finstrument-functions' to CFLAGS for Part of the
source.
Notice:
1. use CONFIG_ARCH_INSTRUMENT_ALL must mark _start or entry noinstrument_function,
becuase bss not set.
2. Make sure your callbacks are not instrumented recursively.
use instrument_register to register entry function and exit function.
They will be called by the instrumented function
Signed-off-by: anjiahao <anjiahao@xiaomi.com>
Binary nibble to/from ascii hex conversion was buggy on both
lib_slcdencode and lib_slcddecode libraries.
This bug caused the slcd library to fail to decode 5-byte sequence command
which have 'count' argument value bigger than 0x9.
Signed-off-by: Federico Braghiroli <federico.braghiroli@gmail.com>
Adds support for POSIX interface open_memstream() that allows writing
to dynamic memory buffer stream. The stream is dynamically reallocated
as the buffer grows with initial size set to zero.
The caller has to free the buffer after the stream is closed.
The implementation uses fopencookie() for custom stream operations and
callbacks.
Signed-off-by: Michal Lenc <michallenc@seznam.cz>
This moves all the public POSIX semaphore functions into libc and with
this most of the user-space logic is also moved; namely cancel point and
errno handling.
This also removes the need for the _SEM_XX macros used to differentiate
which API is used per user-/kernel mode. Such macros are henceforth
unnecessary.
PR #11165 causes an unnecessary regression; task_delete no longer works,
if the deleted task is from another group.
The logic that prevents this comes from:
nxnotify_cancellation() ->
tls_get_info_pid() ->
nxsched_get_stackinfo()
Which checks for permissions, which does not make sense in this case since
it is the kernel asking for the stack information.
Fix this by partially reverting 11165 and implementing a direct path for
the kernel to query for any tasks TLS.
This moves task / thread cancel point logic from the NuttX kernel into
libc, while the data needed by the cancel point logic is moved to TLS.
The change is an enabler to move user-space APIs to libc as well, for
a coherent user/kernel separation.
Add a minimal implementation to suppress warnings when building
application code shared with other operating systems.
For example:
When building with a c++ compiler and GCC 12.2.0, the following warning is emitted:
nuttx/include/spawn.h:178:40: warning: statement has no effect [-Wunused-value]
178 | #define posix_spawnattr_destroy(attr) (0)
Add support for POSIX interface fmemopen(). This interface open a memory
buffer as a stream and permits access to this buffer specified by mode.
This allows I/O operations to be performed on the memory buffer.
The implementation uses fopencookie() for custom stream operations and
callbacks.
Signed-off-by: Michal Lenc <michallenc@seznam.cz>
CPP: nuttx-names.in-> nuttx-names.dat
LD: nuttx
/usr/lib/gcc/x86_64-pc-msys/13.2.0/../../../../x86_64-pc-msys/bin/ld: nuttx.rel:/d/a/nuttx_windows/nuttx_windows/nuttxspace/nuttx/libs/libc/stdlib/lib_exit.c:48: multiple definition of `__dso_handle'; /usr/lib/gcc/x86_64-pc-msys/13.2.0/crtbegin.o:cygming-crtbeg:(.data+0x0): first defined here
CC: icmp/icmp_input.c modlib/modlib_load.c: In function 'modlib_elfsize':
modlib/modlib_load.c:87:30: error: cast to pointer from integer of different size [-Werror=int-to-pointer-cast]
87 | textaddr = (FAR void *)phdr->p_vaddr;
| ^
cc1: all warnings being treated as errors
Signed-off-by: dongjiuzhu1 <dongjiuzhu1@xiaomi.com>
modlib/modlib_symbols.c: In function ‘modlib_symcallback’:
modlib/modlib_symbols.c:215:13: warning: implicit declaration of function ‘modlib_depend’; did you mean ‘modlib_read’? [-Wimplicit-function-declaration]
215 | ret = modlib_depend(exportinfo->modp, modp);
| ^~~~~~~~~~~~~
| modlib_read
Signed-off-by: dongjiuzhu1 <dongjiuzhu1@xiaomi.com>
CP: nuttx/include/nuttx/config.h
modlib/modlib_init.c:53: warning: "modlib_dumpbuffer" redefined
53 | # define modlib_dumpbuffer(m,b,n) binfodumpbuffer(m,b,n)
|
In file included from modlib/modlib_init.c:36:
nuttx/include/nuttx/lib/modlib.h:64: note: this is the location of the previous definition
64 | # define modlib_dumpbuffer(m,b,n) sinfodumpbuffer(m,b,n)
|
Signed-off-by: dongjiuzhu1 <dongjiuzhu1@xiaomi.com>
__dtoa is not used because currently NuttX uses other
function called __dtoa_engine() to do the same thing
Signed-off-by: Xiang Xiao <xiaoxiang@xiaomi.com>
This commit adds support for custom stream via fopencookie function.
The function allows the programmer the create his own custom stream
for IO operations and hook his custom functions to it.
This is a non POSIX interface defined in Standard C library and implemented
according to it. The only difference is in usage of off_t instead of
off64_t. Programmer can use 64 bits offset if CONFIG_FS_LARGEFILE is
enabled. In that case off_t is defined as int64_t (int32_t otherwise).
Field fs_fd is removed from file_struct and fs_cookie is used instead
as a shared variable for file descriptor or user defined cookie.
The interface will be useful for future fmemopen implementation.
Signed-off-by: Michal Lenc <michallenc@seznam.cz>
both functions aren't suitable to be put into libc,
because they call the kernel internal functions directly.
Signed-off-by: Xiang Xiao <xiaoxiang@xiaomi.com>
https://man7.org/linux/man-pages/man2/memfd_create.2.html:
The name supplied in name is used as a filename and will be
displayed as the target of the corresponding symbolic link in the
directory /proc/self/fd/. The displayed name is always prefixed
with memfd: and serves only for debugging purposes. Names do not
affect the behavior of the file descriptor, and as such multiple
files can have the same name without any side effects.
Signed-off-by: Xiang Xiao <xiaoxiang@xiaomi.com>
For programs with the dependencies logic in pthread_once callback , using global locks may cause deadlock:
task A
pthread_once()
|
|-> nxrmutex_lock(&g_lock);
-> init_routine(); // callback to wait task B
task B
pthread_once()
|
->nxrmutex_lock(&g_lock); // Deadlock
->init_routine(); // hold resource to wake up task A
Signed-off-by: hujun5 <hujun5@xiaomi.com>
Domain name has the similar layout as file path, so it's too small
to use 32 bytes as the default value, and better to has the same
default value as PATH_MAX which is 255.
Signed-off-by: Xiang Xiao <xiaoxiang@xiaomi.com>
The ipv6 address filled the cache, and the ipv4 address did not have a
place to store it, causing the resolution to fail. so if IPV6 has already
filled the buffer, rewrite ipv4 DNS results from half of the buffer.
Signed-off-by: zhanghongyu <zhanghongyu@xiaomi.com>
