In the virtualized environment, each time an SGI is sent, the value of IGROUPR0 needs to be read once. Since the GIC Redistributor is a purely emulated device, each read of IGROUPR0 will cause a VM exit, causing serious performance degradation. This patch replaces the read with the value previously set in `gicv3_cpuif_init`, and we assume that this value has not been modified after initialization.
Signed-off-by: ouyangxiangzhen <ouyangxiangzhen@xiaomi.com>
Usage example for merging the factory partition and reserve
partition into the merge partition:
register_merge_blockdriver("/dev/merge", "/dev/factory", "/dev/reserve",
NULL)
Signed-off-by: wanggang26 <wanggang26@xiaomi.com>
This is a combination of 2 commits:
H5 Kconfig ethernet options added.
- This commit cleans up unecessary comments in the Kconfig and adds the the Ethernet/MAC menu.
The menu is copied from the H7 menu, since the peripheral IP is identical.
Remove trailing whitespaces.
Update suggested mask
reason:
some arch do not support issuing interrupts to the local CPU.
This commit fixes the regression from https://github.com/apache/nuttx/pull/14663
Signed-off-by: hujun5 <hujun5@xiaomi.com>
PLL1FRACN was being set improperly. stm32h5xxx_rcc.c does not shift the value provided by board.h. So it was being set wrong. The defintions in stm32h5xxx_rcc.h shift the FRACN value and are now used by board.h. Also, board.h was not setting PLL1P properly. PLL1P can not have odd divisors. Therefore a value of 0 was invalid. Set it to a value of 1 (divide by 2), then adjust PLL1N to 31 and PLL1FRAC1 to 2048 to actually set SYSCLK to 250MHz.
The write should return even in case of O_NONBLOCK if at least some
bytes were written.
The previous state where always all bytes were written was breaking a
common combination with poll, because poll would signal POLLOUT and some
bytes would really be consumed but write could still block afterwards.
That would prevent from execution returning to the poll loop again.
None the less it is also the standard C library behavior for the write
function.
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>
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>
Made fixes to issues from CI. Nxstyle and defconfig syntax.
This is a combination of 6 commits.
Adding STM32H5 arch files. With comments addressed.
Created stm32h5 directory to add support for the H5 chip, and used a Nucleo-H563ZI dev board during development. The goal was to get a working nutshell through the STLink connector on the board.
Remove board/docs changes for PR update.
Squash commits into one for PR guideline conformity.
trying to fix build issues
Fix format from review
Nucleo-H563ZI support for NSH.
Created stm32h5 directory to add support for the H5 chip, and used a Nucleo-H563ZI dev board during development. The goal was to get a working nutshell through the STLink connector on the board.
Fix switch default case placement.
NXstyle fixes
Renaming files
rename stm32h5_gpio.x files
rename h5 hsi48 files
Rename h5 idle file
rename stm32h5_irq.c
Rename some rcc functions and stm32h5_rcc.c
rename stm32h5_rcc.h
Rename stm32h5_pwr.x
lowputc renames
timerisr renames
uart renamed
rename serial file
rename start
Turn off the defines that enable DMA on serial
remove DMA Kconfig options
Remove H5 documentation. Will add in a future PR.
Fix styling and defconfig improper syntax.
reason:
The old implementation of the SMP call, even when using the "no wait" parameter,
could still result in waiting, if invoking it within a critical section
may lead to deadlocks. Therefore, in order to implement a truly asynchronous SMP
call strategy, we have added nxsched_smp_call_async.
Signed-off-by: hujun5 <hujun5@xiaomi.com>
Created stm32h5 directory to add support for the H5 chip, and used a Nucleo-H563ZI dev board during development. The goal was to get a working nutshell through the STLink connector on the board.
Remove board/docs changes for PR update.
Squash commits into one for PR guideline conformity.
trying to fix build issues
Fix format from review
The reason for removing math.h is that undefining __GLIBC__ does not take effect. By default, sim will use the toolchain's math library and undef __GLIBC__ in the source file.
Signed-off-by: cuiziwei <cuiziwei@xiaomi.com>
fixed this error for msys2
====================================================================================
Configuration/Tool: nucleo-l152re/nsh,CONFIG_ARM_TOOLCHAIN_GNU_EABI
2024-11-05 06:12:26
------------------------------------------------------------------------------------
Cleaning...
Configuring...
make[3]: warning: jobserver unavailable: using -j1. Add '+' to parent make rule.
make[3]: warning: jobserver unavailable: using -j1. Add '+' to parent make rule.
make[3]: warning: jobserver unavailable: using -j1. Add '+' to parent make rule.
make[3]: warning: jobserver unavailable: using -j1. Add '+' to parent make rule.
make[3]: warning: jobserver unavailable: using -j1. Add '+' to parent make rule.
make[3]: warning: jobserver unavailable: using -j1. Add '+' to parent make rule.
make[3]: warning: jobserver unavailable: using -j1. Add '+' to parent make rule.
make[3]: warning: jobserver unavailable: using -j1. Add '+' to parent make rule.
make[3]: warning: jobserver unavailable: using -j1. Add '+' to parent make rule.
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>
use the dynamic library .so of the binding compilation environment,
and then modify the dynamic loader of elf at runtime to run sim elf across system versions
1.add post-build for sim build
2.collect dynamic .so and dynamic interpreter
3.add execution script for the entire package sim elf
change text-segment to 0x50000000
because patchelf will modify the .dynamic section in elf,
the upward alignment of elf will be less than 0x3fffffff,
which will cause asan shadow memory overlap error
Signed-off-by: xuxin19 <xuxin19@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.
Some chips have different memory addressing spaces for the same
region. This is true, for instance, for ESP32-S3: the same memory
region can be accessed using the data bus or the data bus using
different address ranges. The instruction bus, however, requires
word-aligned access. That being said, it is recommended to use the
data bus while copying sections to the text heap to avoid any
illegal access using the instruction bus address which will be
later used to run the program.
Fixes psock_socket: ERROR: socket address family unsupported: 2
When using the minimal CONFIG_NET configuration typically used for
SocketCAN
Update net/inet/inet.h
Co-authored-by: Xiang Xiao <xiaoxiang781216@gmail.com>
Update net/inet/inet.h
Co-authored-by: Xiang Xiao <xiaoxiang781216@gmail.com>
