x86 selected neither fork primitive, so vfork() was not available on this
architecture at all.
fork.S takes the register snapshot and hands it to x86_fork(), which allocates
the child, copies the used part of the caller's stack, and starts it. There is
one entry point for both primitives, because the snapshot is the same for
either.
Unlike the register-passing architectures, cdecl puts the flag on the stack, so
up_fork() loads it from 4(%esp). That slot is also the stack pointer the
caller had: it pushed the argument, then `call' pushed the return address. So
the low end of the region that x86_fork() copies is unchanged.
POSIX fork() is not provided. It needs an address environment that can be
duplicated and this architecture has none, so CONFIG_ARCH_HAVE_FORK is never
set here. x86_fork.c makes that a build error rather than a silent omission.
Verified under QEMU with qemu-i486:nsh. ostest runs to the end and reports
"Child 5 ran and exited before the parent resumed", with fork() correctly
absent.
Assisted-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
Signed-off-by: Marco Casaroli <marco.casaroli@gmail.com>
In a kernel build vfork() is reached through a system call, so the return
address and stack pointer the architecture's entry point can see for itself
are the kernel's, not the caller's. A child built from those resumes at a
kernel address, which is why x86_64 selected the fork family only for the
flat build.
x86_64_syscall() now publishes the caller's frame in xcp.sregs for the
duration of the stub call, and x86_64_fork() builds the child from it:
x86_64_fork_syscall() when xcp.sregs is non-NULL, so that the child
returns from the very same `syscall' instruction as
the parent, in user mode, on its own stack;
x86_64_fork_direct() otherwise, which is the flat build and any kernel
thread that calls the entry point as a plain
function.
The discriminator is xcp.sregs rather than TCB_FLAG_SYSCALL, which arm64 and
RISC-V use: that flag also defers signal actions, x86_64 has never raised it,
and its kernel-build signal path does not survive being made to -- a
pre-existing problem that does not belong to this work.
Two properties of SYSCALL/SYSRET shape the child's frame. The instruction
leaves the caller's RIP and RFLAGS in RCX and R11 rather than on a stack, so
they are moved into the RIP and RFLAGS slots of the interrupt frame the child
is resumed from; and the hardware never records the caller's CS and SS at all,
SYSRETQ reconstructing them from IA32_STAR, so the child's are filled in with
the user code and data selectors at RPL 3. The frame is therefore not copied
wholesale: the extended state and the general registers are inherited, while
the segment registers and the thread pointer stay as up_initial_state() left
them, the child's stack being a fresh allocation the parent's FS base does not
describe.
x86_64_fork_relocfp() is new and is not optional here. A function returns
with `leave', which feeds the frame pointer into the stack pointer, so
relocating only the RBP the child resumes with gets it exactly one frame:
the next return loads a saved RBP still pointing into the parent's stack.
With that in place ARCH_X86_64 can select ARCH_HAVE_VFORK unconditionally.
Build-verified on qemu-intel64:knsh_romfs and qemu-intel64:ostest. NuttX on
qemu-intel64 requires tsc-deadline and pcid, which TCG does not implement, so
it cannot be run on this host.
Assisted-by: Claude Code:claude-opus-5
Signed-off-by: Marco Casaroli <marco.casaroli@gmail.com>
In a kernel or protected build vfork() is reached through a system call, so
the return address and stack pointer that the entry point in
arm64_fork_func.S can snapshot for itself belong to the kernel-side stub, not
to the caller. A child built from that snapshot resumes at a kernel address
on a kernel stack. This is why arm64 selected the fork family only for the
flat build.
Record what the caller was actually doing instead. arm64_sync_exc passes the
exception frame to dispatch_syscall() in x7 -- x0-x6 carry the call number and
its six parameters, so x7 is free -- and dispatch_syscall() stores it in
xcp.sregs, mirroring what riscv_swint.c does.
arm64_fork() then chooses where the caller's registers live:
arm64_fork_syscall() when TCB_FLAG_SYSCALL is set, rebuilding the child
from xcp.sregs so that it returns from the very same
SVC as the parent;
arm64_fork_direct() otherwise, which is the flat build and any kernel
thread that calls the entry point as a plain function.
The stack copy and the relocation of pointers into it are shared by both
paths in arm64_fork_stack() and arm64_fork_reloc().
With that in place ARCH_ARM64 can select ARCH_HAVE_VFORK unconditionally.
Verified on qemu-armv8a:knsh (BUILD_KERNEL), qemu-armv8a:nsh (BUILD_FLAT) and
qemu-armv8a:citest_smp under qemu-system-aarch64: ostest's vfork_test passes
on all three, and it was absent from knsh before the change. The protected
configurations are build-verified only (fvp-armv8r:pnsh), there being no
emulator for them here.
Assisted-by: Claude Code:claude-opus-5
Signed-off-by: Marco Casaroli <marco.casaroli@gmail.com>
ARCH_ARM has selected ARCH_HAVE_VFORK only "if !BUILD_KERNEL" since the
fork()/vfork() split. That condition was deliberate but temporary: it was
added because the fork family had never worked on a 32-bit ARM kernel build --
the entry point in fork.S snapshots the kernel-side stub rather than the
caller, so a child resumes at a kernel address -- and said in as many words
that "arch/arm takes the condition off again in the patch that adds its
saved-syscall-frame path".
That patch is the one before this. arm_syscall() records the caller's
exception frame in xcp.sregs and arm_fork() builds the child from it, so the
condition has nothing left to protect against.
Cortex-M is unaffected either way -- BUILD_KERNEL depends on ARCH_USE_MMU,
which it does not have -- so the only configurations this changes are the
MMU-capable ARM ports, which are exactly the ones the previous commit fixed.
Verified on qemu-armv7a:knsh under qemu-system-arm: ostest's vfork_test
passes, where before the change vfork() was absent.
Assisted-by: Claude Code:claude-opus-5
Signed-off-by: Marco Casaroli <marco.casaroli@gmail.com>
NuttX implemented fork() and vfork() as the same function. Both were libc
wrappers around a single up_fork() syscall; vfork() differed only by a
trailing waitpid(). Underneath, the child joined the parent's address
environment -- the same addrenv_join() that pthread_create() uses -- and got
a private copy of the stack. So the child shared .data, .bss and the heap
with its parent and ran concurrently with it.
That is not fork(). It is vfork()-with-a-private-stack under fork()'s name,
and the history says so: today's fork() is NuttX's old vfork(), renamed in
c33d1c9c97 (2023) without any change of behaviour. The failure was silent --
a program written against POSIX fork() compiled, ran, and had its child's
writes land in the parent's variables.
Separate them into two primitives, chosen by which function the caller
called rather than by what the hardware happens to be:
fork() child gets its own copy of the parent's memory at the same
virtual addresses; runs concurrently. Only where an address
environment can be duplicated -- elsewhere it is not declared at
all, so calling it is a build error naming the function.
vfork() child shares the parent's memory; parent suspended until the
child _exit()s or exec()s. Implementable everywhere.
Below libc there is still one syscall. up_fork() gains a bool saying which
primitive the caller used, since the per-architecture register snapshot is
the same for both, and passes it to nxtask_setup_fork(), which is the single
place the memory semantics are decided. The argument arrives in the first
argument register and is never touched: each architecture's snapshot takes
some other call-clobbered register for its scratch, so the flag is simply
still there when the C worker is called.
The vfork() parent suspension moves out of libc into nxtask_start_fork(),
released from nxsched_release_tcb() by nxtask_resume_vfork(). Two things
follow: the parent is resumed at exec(), since exec_swap() has already handed
the child's pid to the loaded program by the time the vfork stub exits, and
vfork() no longer depends on CONFIG_SCHED_WAITPID.
Releasing there requires one fix in nxtask_exit(). It raises rtcb->lockcount
directly rather than through sched_lock() while it tears the TCB down, so the
nxsem_post() that wakes the vfork() parent leaves it queued where a blocked
task collects while pre-emption is off -- g_pendingtasks, or g_readytorun on
SMP -- and the matching raw lockcount-- does not publish it the way
sched_unlock() would, leaving the parent stranded with nothing to move it on.
The fix mirrors sched_unlock() for each case: nxsched_merge_pending(), or
nxsched_deliver_task() under CONFIG_SMP. Both are no-ops while pre-emption is
still disabled, and up_exit() re-reads this_task() afterwards, so a change of
the ready-to-run head is honoured. Without it vfork() deadlocks wherever no
other task happens to call sched_unlock() afterwards -- rv-virt:nsh64 and
rv-virt:pnsh64, where NSH is blocked in waitpid() holding the lock, and
qemu-armv8a:citest_smp, which hangs the moment the vfork() test runs.
fork() is built on a new addrenv_fork(), backed by an up_addrenv_fork() hook
that duplicates an address environment into freshly allocated pages mapped at
the same virtual addresses -- unlike up_addrenv_clone(), which copies only
the representation and leaves both pointing at the same page tables. The
child then adopts the parent's stack geometry rather than being given a
relocated copy: a pointer to a stack local taken before fork() must name the
same object in the child that it named in the parent, and the parent's stack
is already in the duplicate, with its contents, at the parent's address.
No architecture implements up_addrenv_fork() yet, so this commit leaves
fork() unavailable everywhere. That is the intended state. It withdraws
fork() from ARCH_ARM, flat ARCH_ARM64, ARCH_RISCV, ARCH_SIM and ARCH_X86_64,
where until now it named the sharing primitive; per-architecture patches
restore it, with POSIX semantics, as up_addrenv_fork() lands. In the
meantime the sharing primitive is still there under the name that describes
it: vfork() for a child that runs a program, pthread_create() for a second
flow of control that shares memory, posix_spawn() for both at once.
Kconfig: ARCH_HAVE_VFORK inherits ARCH_HAVE_FORK's select lines, conditions
included, so no configuration gains machinery; ARCH_HAVE_FORK is redefined to
mean "can provide POSIX fork() semantics" and now depends on ARCH_ADDRENV.
There is one deliberate departure from "verbatim". ARCH_ARM selected the
fork family unconditionally, BUILD_KERNEL included, and that has never
worked: on a kernel build the architecture's fork entry point sees the
kernel's return address and stack pointer rather than the caller's, so the
child resumes at a kernel address. On qemu-armv7a:knsh master faults in
ostest's fork case with "Child did not run" and then a data abort; without
the condition this change faults the same way through vfork(). ARCH_ARM64
and ARCH_X86_64 already carried "if !BUILD_KERNEL" for exactly this reason --
ARM was the outlier. Conditioning it turns a runtime fault into an honest
absence, which is the whole point of the change; arch/arm takes the condition
off again in the patch that adds its saved-syscall-frame path. Only the
MMU-capable ARM ports are affected, since Cortex-M cannot build BUILD_KERNEL
at all.
Also fixes two latent syntax errors found on the way: a missing comma in
riscv_fork.c and mips_fork.c, both in *_FRAMEPOINTER && !SAVE_GP branches
that are never compiled today.
BREAKING CHANGE: fork() is withdrawn from every architecture. It is no
longer declared in unistd.h, so code that calls it fails to build with an error
naming the function, and the sharing behaviour it used to have is gone rather
than renamed. CONFIG_ARCH_HAVE_FORK no longer means "fork() exists"; it means
"this configuration can provide POSIX fork() semantics", and no architecture
selects it yet.
Quick fix, chosen by why the call was made:
to run a program vfork() + exec*(), or better posix_spawn()
a second flow of control that pthread_create()
shares the caller's memory
a genuinely independent copy keep fork(), and wait for the per-arch patch
of the process that implements up_addrenv_fork() and selects
CONFIG_ARCH_HAVE_FORK
Out-of-tree code that tests CONFIG_ARCH_HAVE_FORK to decide whether a
fork-then-exec path is available wants CONFIG_ARCH_HAVE_VFORK instead, which is
selected in exactly the places CONFIG_ARCH_HAVE_FORK used to be. The full
migration guide is Documentation/guides/fork_vfork_migration.rst.
Assisted-by: Claude Code:claude-opus-5
Signed-off-by: Marco Casaroli <marco.casaroli@gmail.com>
794c325947 ("arch/x64:Syscall support is enabled by default", 2025-05-27)
changed nine guards from CONFIG_LIB_SYSCALL to CONFIG_ARCH_HAVE_SYSCALL but
never added the Kconfig symbol. Only ARCH_HAVE_SYSCALL_HOOKS exists in tree;
ARCH_HAVE_SYSCALL itself is defined nowhere, so it is always unset and since
that commit x86_64 has had no x86_64_syscall_entry(), no x86_64_syscall(), no
IA32_LSTAR/IA32_STAR programming and no syscall stub layer in any
configuration:
arch/x86_64/src/common/Make.defs:37 x86_64_syscall.c not compiled
arch/x86_64/src/common/CMakeLists.txt:41 likewise
arch/x86_64/include/irq.h:86
arch/x86_64/src/intel64/intel64_cpu.c:247, :386
arch/x86_64/src/intel64/intel64_head.S:83, :351, :538
arch/x86_64/src/intel64/intel64_saveusercontext.S:108
qemu-intel64:knsh_romfs and qemu-intel64:knsh_romfs_pci are the two
CONFIG_BUILD_KERNEL configurations in tree, and neither can have worked in
that time. They still link -- nothing references the missing pieces, so
libstubs.a is simply never pulled in -- and then die the first time user code
executes SYSCALL.
Define the symbol with the condition the code had before that commit, which
is LIB_SYSCALL: every protected and every kernel build needs the interface,
and a flat build is left exactly as it is today.
Impact: restores the system call interface for CONFIG_BUILD_KERNEL and
CONFIG_BUILD_PROTECTED on x86_64. CONFIG_BUILD_FLAT is unaffected --
qemu-intel64:nsh still builds with ARCH_HAVE_SYSCALL unset.
Assisted-by: Claude:claude-opus-5
Signed-off-by: Marco Casaroli <marco.casaroli@gmail.com>
While attempting to create architecture-specific implementation
of up_udelay, it was discovered that the overriding function is not
included in the final binary, the weak implementation was used instead.
Further investigation and experimentation showed that the linker
only overrides the weak implementation with the custom one if
the custom one is present in a .c source file that contains at least
one other function that is called from somewhere. Some additional
testing revealed that at least one other already present up_udelay
override (rv32m1-vega:nsh) is affected by this.
In a short mailing list discussion it was determined that this
is a likely result of using static libraries during the build process
and it was suggested to introduce configuration option that will
exclude weak implementations of the function from the build altogether.
This patch does that.
This patch does not enable this configuration option for any existing
board/chip because doing so would change its behaviour and needs
to be tested by users of the hardware.
Also changed is the static assertion in sched/clock/clock_delay.c
to not prevent building the code when architecture declares that
it does not use BOARD_LOOPSPERMSEC to determine required loop count.
BOARD_LOOPSPERMSEC is made undefined in such case.
Patch was tested by building breadxavr:nsh (identical binary by SHA256),
rv32m1-vega:nsh (identical text section) and rv-virt:nsh (text section
differs because of different ordering of functions in the binary, ostest
passed though.)
Signed-off-by: Kerogit <kr.git@kerogit.eu>
Implement up_backtrace() for TriCore by walking the CSA chain and
collecting return addresses from upper CSA A11 register.
Signed-off-by: LukeKun <donghaokun@lixiang.com>
Changes the default value for CONFIG_BOARD_LOOPSPERMSEC to -1, which is
invalid. All boards that forget to configure this value will encounter a
static assertion at compile time, enforcing that configurations
upstreamed to NuttX have calibrated values for correct delay timings.
Boards which implement ALARM_ARCH or TIMER_ARCH do not rely on the
busy-loop delay implementation and thus only have a run-time check to
ensure proper delay-timings (in the case where delays are used before
the alarm/timer driver is registered).
Some boards already in the NuttX upstream do not have calibrated values,
but there are no users who currently own the board to submit a
calibrated value for the configurations to use. In this scenario, the
value is temporarily set to 0 and a warning is displayed so that users
of these boards are informed of the calibration process.
Signed-off-by: Matteo Golin <matteo.golin@gmail.com>
Enable SYSCALL_HOOKS architecture capability for TriCore platform.
This allows runtime interception and hooking of system calls on
TriCore architecture, enabling profiling, tracing, and debugging features.
Signed-off-by: hujun5 <hujun5@xiaomi.com>
Add ARCH_HAVE_PERF_EVENTS_USER_ACCESS capability to allow applications
to directly access hardware perf counters via perf_gettime() from
userspace, enabling performance monitoring and profiling without syscalls.
Signed-off-by: Yanfeng Liu <p-liuyanfeng9@xiaomi.com>
dd ARCH_PERF_COUNT_BITWIDTH configuration to support architectures
with different perf counter bit widths (TriCore 31-bit, others 32-bit),
enabling proper overflow correction across diverse platforms.
Signed-off-by: zhangyuan29 <zhangyuan29@xiaomi.com>
Add default values (0x0) for ARCH_TEXT_VBASE, ARCH_DATA_VBASE, and
ARCH_HEAP_VBASE to resolve cmake configuration warnings when these
values are not explicitly set during initialization.
Signed-off-by: hujun5 <hujun5@xiaomi.com>
Expand ARCH_KERNEL_STACK configuration to support both BUILD_KERNEL and
BUILD_PROTECTED modes. Change Kconfig dependency from "depends on BUILD_KERNEL"
to "depends on BUILD_KERNEL || BUILD_PROTECTED". Update pthread_create.c to
remove redundant BUILD_KERNEL check when allocating kernel stacks with
CONFIG_ARCH_KERNEL_STACK enabled. Enables protected builds to use separate
kernel and user stacks for processes.
Signed-off-by: hujun5 <hujun5@xiaomi.com>
Modify ARM64 architecture configuration to disable fork support in protected
build mode (BUILD_PROTECTED). Update ARCH_HAVE_FORK condition from
"!BUILD_KERNEL" to "!BUILD_KERNEL && !BUILD_PROTECTED", ensuring fork is
only available in flat/monolithic builds, not in protected kernel builds.
Signed-off-by: hujun5 <hujun5@xiaomi.com>
Add new CONFIG_NCPUS configuration option to support both SMP and AMP modes.
Replace SMP-specific CONFIG_SMP_NCPUS checks with generic CONFIG_NCPUS in ARM
boot code (ARMv7-A, ARMv7-R, ARMv8-R). CONFIG_NCPUS defaults to SMP_NCPUS when
SMP is enabled, or 1 otherwise, enabling multi-CPU support regardless of SMP mode.
Signed-off-by: hujun5 <hujun5@xiaomi.com>
tricore csa is not continuous. when assert prints information, we need to handle the regs specially in order to dump all the registers.
Signed-off-by: zhangyu117 <zhangyu117@xiaomi.com>
Add condition checking in Kconfig source statement, for example:
source "arch/arm/Kconfig" is updated to:
if ARCH_ARM
source "arch/arm/Kconfig"
endif
Signed-off-by: Chengdong Wang <wangchengdong@lixiang.com>
In the toolchain, ARCH_TOOLCHAIN_XXX is used to select the compiler to be used. If clang is selected here, subsequent errors will occur
Signed-off-by: wangmingrong1 <wangmingrong1@xiaomi.com>
1. Add hw_tags.c, which will call arm64_mte to implement tagging of memory blocks by operating registers
2. It has been able to run normally on the default NX memory allocator, excluding mempool and tlsf
3. On more complex configurations, memory tests such as memstress can run normally without system crashes
Signed-off-by: wangmingrong1 <wangmingrong1@xiaomi.com>
Some code paths in drivers/serial/serial.c load head and tail values
of receive and transmit circular buffers with interrupts enabled,
making it possible that the interrupt handler changes the value.
As noted in the code, this is safe as long as the load itself is atomic.
That is not true for 8bit architectures which fetch the 16-bit values
using two load instructions. If interrupt handler runs between those
two instructions and changes the value, the read returns corrupted data.
This patch introduces CONFIG_ARCH_LDST_16BIT_NOT_ATOMIC configuration
option which is automatically selected for AVR architecture. Based
on this option, head and tail values are reduced to 8-bit length
so the read remains atomic.
Patch was tested by building on rv-virt:nsh - disassembly of functions
from serial.c showed no difference which is correct as Risc-V does
not need to protect reads of these values. There should be no impact
for architectures that do not set the new configuration option.
It was also tested by by custom echo application running on AVR128DA28.
Signed-off-by: Kerogit <kr.git@kerogit.eu>
Optimize crc32 standard(poly:0x04C11DB7) and crc32
castagnoli(poly:0x1EDC6F41) with arm crc32 extension instructions.
For example, crc32 standard caculates(lookup crc32 table) 1812 bytes data,
reduced the time from 118 us to 14 us through optimization.
Performance improved ~700%
Signed-off-by: Jinliang Li <lijinliang1@lixiang.com>
reason:
svc call may trigger hardfault
Background
The origin of this issue is our desire to eliminate the function of storing
"regs" in g_current_regs and instead utilize (*running_task)->xcp.regs for storage.
The benefits of this approach include faster storage speed and
avoiding multiple accesses to g_current_regs during context switching,
thus ensuring that whether returning from an interrupt or an exception,
we consistently use this_task()->xcp.regs
Issue Encountered
However, when storing registers, we must ensure that (running_task)->xcp.regs is invalid
so that it can be safely overwritten.
According to the existing logic, the only scenario where (running_task)->xcp.regs
is valid is during restore_context. We must accurately identify this scenario.
Initially, we used the condition (running_task)==NULL for this purpose, but we deemed
this approach unsatisfactory as it did not align well with the actual logic.
(running_task) should not be NULL. Consequently, we adopted other arch-specific methods for judgment,
but due to special logic in some arch, the judgment was not accurate, leading to this issue.
Solution:
For armv6-m, we haven't found a more suitable solution, so we are sticking with (*running_task)==NULL.
For armv7-m/armv8-m, by removing support for primask, we can achieve accurate judgment.
PRIMASK is a design in armv6-m, that's why arm introduce BASEPRI from armv7-m.
It's wrong to provide this option for armv7-m/armv8-m arch.
Signed-off-by: hujun5 <hujun5@xiaomi.com>
Implement hartid<->cpuid mapping for RISC-V. This is necessary for some
platforms which cannot use 1:1 mapping between logical and physical CPU /
core IDs. One example is MPFS where hart0 cannot be used for NuttX SMP as
it is a less capable "monitor" core (E51) compared to the application
cores hart1...3 (E54).
Why not just use a generic offset then? We also need the physical hart ID
for many things:
- Communication between harts (IPI)
- External interrupt acknowledgment (interrupt claim for specific CPU)
- Communication to SBI
Thus, create procedures that can do this translation:
- The default mapping is still logical=physical.
- Another flavor is to use the existing CONFIG_ARCH_RV_HARTID_BASE config
variable, which is just a simple offset
- The final flavor is to overload hartid<->cpuid on a per chip basis (no
example for this is provided yet)
This implements empty hooks to the arch/chip layer, which can implement
an optional translation between logical<->physical CPU/core id.
The default mapping is 1:1 i.e. logical=physical.
Summary:
Add the default CPU frequency configuration.
Signed-off-by: wangming9 <wangming9@xiaomi.com>
Signed-off-by: lipengfei28 <lipengfei28@xiaomi.com>
