Scroll to navigation

VM_MAP_STACK(9) Kernel Developer's Manual VM_MAP_STACK(9)


vm_map_stack, vm_map_growstackmanage process stacks


#include <sys/param.h>
#include <vm/vm.h>
#include <vm/vm_map.h>
vm_map_stack(vm_map_t map, vm_offset_t addrbos, vm_size_t max_ssize, vm_prot_t prot, vm_prot_t max, int cow);
vm_map_growstack(struct proc *p, vm_offset_t addr);


The vm_map_stack() function maps a process stack for a new process image. The stack is mapped addrbos in map, with a maximum size of max_ssize. Copy-on-write flags passed in cow are also applied to the new mapping. Protection bits are supplied by prot and max.
It is typically called by execve(2).
The vm_map_growstack() function is responsible for growing a stack for the process p to the desired address addr, similar to the legacy sbrk(2) call.


The vm_map_stack() function calls vm_map_insert(9) to create its mappings.
The vm_map_growstack() function acquires the Giant lock, and the process lock on p, for the duration of the call.


The vm_map_stack() function returns KERN_SUCCESS if the mapping was allocated successfully.
Otherwise, if mapping the stack would exceed the process's VMEM resource limit, or if the specified bottom-of-stack address is out of range for the map, or if there is already a mapping at the address which would result, or if max_ssize could not be accommodated within the current mapping, KERN_NO_SPACE is returned.
Other possible return values for this function are documented in vm_map_insert(9).
The vm_map_growstack() function returns KERN_SUCCESS if addr is already mapped, or if the stack was grown successfully.
It also returns KERN_SUCCESS if addr is outside the stack range; this is done in order to preserve compatibility with the deprecated grow() function previously located in the file vm_machdep.c.


vm_map(9), vm_map_insert(9)


This manual page was written by Bruce M Simpson ⟨⟩.
July 19, 2003 Debian