565f13088f
Change the kernel to add features to vircopy and safecopies so that transparent copy fixing won't happen to avoid deadlocks, and such copies fail with EFAULT. Transparently making copying work from filesystems (as normally done by the kernel & VM when copying fails because of missing/readonly memory) is problematic as it can happen that, for file-mapped ranges, that that same filesystem that is blocked on the copy request is needed to satisfy the memory range, leading to deadlock. Dito for VFS itself, if done with a blocking call. This change makes the copying done from a filesystem fail in such cases with EFAULT by VFS adding the CPF_TRY flag to the grants. If a FS call fails with EFAULT, VFS will then request the range to be made available to VM after the FS is unblocked, allowing it to be used to satisfy the range if need be in another VFS thread. Similarly, for datacopies that VFS itself does, it uses the failable vircopy variant and callers use a wrapper that talk to VM if necessary to get the copy to work. . kernel: add CPF_TRY flag to safecopies . kernel: only request writable ranges to VM for the target buffer when copying fails . do copying in VFS TRY-first . some fixes in VM to build SANITYCHECK mode . add regression test for the cases where - a FS system call needs memory mapped in a process that the FS itself must map. - such a range covers more than one file-mapped region. . add 'try' mode to vircopy, physcopy . add flags field to copy kernel call messages . if CP_FLAG_TRY is set, do not transparently try to fix memory ranges . for use by VFS when accessing user buffers to avoid deadlock . remove some obsolete backwards compatability assignments . VFS: let thread scheduling work for VM requests too Allows VFS to make calls to VM while suspending and resuming the currently running thread. Does currently not work for the main thread. . VM: add fix memory range call for use by VFS Change-Id: I295794269cea51a3163519a9cfe5901301d90b32
169 lines
4.8 KiB
C
169 lines
4.8 KiB
C
#include "inc.h"
|
|
#include <assert.h>
|
|
#include <sys/exec.h>
|
|
#include <libexec.h>
|
|
#include <minix/param.h>
|
|
#include <machine/vmparam.h>
|
|
|
|
static int do_exec(int proc_e, char *exec, size_t exec_len, char *progname,
|
|
char *frame, int frame_len, vir_bytes ps_str);
|
|
static int exec_restart(int proc_e, int result, vir_bytes pc, vir_bytes ps_str);
|
|
static int read_seg(struct exec_info *execi, off_t off,
|
|
vir_bytes seg_addr, size_t seg_bytes);
|
|
|
|
/* Array of loaders for different object formats */
|
|
static struct exec_loaders {
|
|
libexec_exec_loadfunc_t load_object;
|
|
} const exec_loaders[] = {
|
|
{ libexec_load_elf },
|
|
{ NULL }
|
|
};
|
|
|
|
int srv_execve(int proc_e, char *exec, size_t exec_len, char **argv,
|
|
char **envp)
|
|
{
|
|
size_t frame_size = 0; /* Size of the new initial stack. */
|
|
int argc = 0; /* Argument count. */
|
|
int envc = 0; /* Environment count */
|
|
char overflow = 0; /* No overflow yet. */
|
|
char *frame;
|
|
struct ps_strings *psp;
|
|
int vsp = 0; /* (virtual) Stack pointer in new address space. */
|
|
|
|
char *progname;
|
|
int r;
|
|
|
|
minix_stack_params(argv[0], argv, envp, &frame_size, &overflow,
|
|
&argc, &envc);
|
|
|
|
/* The party is off if there is an overflow. */
|
|
if (overflow) {
|
|
errno = E2BIG;
|
|
return -1;
|
|
}
|
|
|
|
/* Allocate space for the stack frame. */
|
|
if ((frame = (char *) sbrk(frame_size)) == (char *) -1) {
|
|
errno = E2BIG;
|
|
return -1;
|
|
}
|
|
|
|
minix_stack_fill(argv[0], argc, argv, envc, envp, frame_size, frame,
|
|
&vsp, &psp);
|
|
|
|
(progname=strrchr(argv[0], '/')) ? progname++ : (progname=argv[0]);
|
|
|
|
r = do_exec(proc_e, exec, exec_len, progname, frame, frame_size,
|
|
vsp + ((char *)psp - frame));
|
|
|
|
/* Failure, return the memory used for the frame and exit. */
|
|
(void) sbrk(-frame_size);
|
|
|
|
return r;
|
|
}
|
|
|
|
|
|
static int do_exec(int proc_e, char *exec, size_t exec_len, char *progname,
|
|
char *frame, int frame_len, vir_bytes ps_str)
|
|
{
|
|
int r;
|
|
vir_bytes vsp;
|
|
struct exec_info execi;
|
|
int i;
|
|
|
|
memset(&execi, 0, sizeof(execi));
|
|
|
|
execi.stack_high = kinfo.user_sp;
|
|
execi.stack_size = DEFAULT_STACK_LIMIT;
|
|
execi.proc_e = proc_e;
|
|
execi.hdr = exec;
|
|
execi.filesize = execi.hdr_len = exec_len;
|
|
strncpy(execi.progname, progname, PROC_NAME_LEN-1);
|
|
execi.progname[PROC_NAME_LEN-1] = '\0';
|
|
execi.frame_len = frame_len;
|
|
|
|
/* callback functions and data */
|
|
execi.copymem = read_seg;
|
|
execi.clearproc = libexec_clearproc_vm_procctl;
|
|
execi.clearmem = libexec_clear_sys_memset;
|
|
execi.allocmem_prealloc_cleared = libexec_alloc_mmap_prealloc_cleared;
|
|
execi.allocmem_prealloc_junk = libexec_alloc_mmap_prealloc_junk;
|
|
execi.allocmem_ondemand = libexec_alloc_mmap_ondemand;
|
|
|
|
for(i = 0; exec_loaders[i].load_object != NULL; i++) {
|
|
r = (*exec_loaders[i].load_object)(&execi);
|
|
/* Loaded successfully, so no need to try other loaders */
|
|
if (r == OK) break;
|
|
}
|
|
|
|
/* No exec loader could load the object */
|
|
if (r != OK) {
|
|
printf("RS: do_exec: loading error %d\n", r);
|
|
return r;
|
|
}
|
|
|
|
/* Inform PM */
|
|
if((r = libexec_pm_newexec(execi.proc_e, &execi)) != OK)
|
|
return r;
|
|
|
|
/* Patch up stack and copy it from RS to new core image. */
|
|
vsp = execi.stack_high - frame_len;
|
|
r = sys_datacopy(SELF, (vir_bytes) frame,
|
|
proc_e, (vir_bytes) vsp, (phys_bytes)frame_len);
|
|
if (r != OK) {
|
|
printf("do_exec: copying out new stack failed: %d\n", r);
|
|
exec_restart(proc_e, r, execi.pc, ps_str);
|
|
return r;
|
|
}
|
|
|
|
return exec_restart(proc_e, OK, execi.pc, ps_str);
|
|
}
|
|
|
|
/*===========================================================================*
|
|
* exec_restart *
|
|
*===========================================================================*/
|
|
static int exec_restart(int proc_e, int result, vir_bytes pc, vir_bytes ps_str)
|
|
{
|
|
int r;
|
|
message m;
|
|
|
|
memset(&m, 0, sizeof(m));
|
|
m.m_type = PM_EXEC_RESTART;
|
|
m.PM_EXEC_RESTART_ENDPT = proc_e;
|
|
m.PM_EXEC_RESTART_RESULT = result;
|
|
m.PM_EXEC_RESTART_PC = (void *)pc;
|
|
m.PM_EXEC_RESTART_PS_STR = (void *)ps_str;
|
|
|
|
r = ipc_sendrec(PM_PROC_NR, &m);
|
|
if (r != OK)
|
|
return r;
|
|
|
|
return m.m_type;
|
|
}
|
|
|
|
/*===========================================================================*
|
|
* read_seg *
|
|
*===========================================================================*/
|
|
static int read_seg(
|
|
struct exec_info *execi, /* various data needed for exec */
|
|
off_t off, /* offset in file */
|
|
vir_bytes seg_addr, /* address to load segment */
|
|
size_t seg_bytes /* how much is to be transferred? */
|
|
)
|
|
{
|
|
/*
|
|
* The byte count on read is usually smaller than the segment count, because
|
|
* a segment is padded out to a click multiple, and the data segment is only
|
|
* partially initialized.
|
|
*/
|
|
|
|
int r;
|
|
|
|
if (off+seg_bytes > execi->hdr_len) return ENOEXEC;
|
|
if((r= sys_datacopy(SELF, ((vir_bytes)execi->hdr)+off,
|
|
execi->proc_e, seg_addr, seg_bytes)) != OK) {
|
|
printf("RS: exec read_seg: copy 0x%x bytes into %i at 0x%08lx failed: %i\n",
|
|
seg_bytes, execi->proc_e, seg_addr, r);
|
|
}
|
|
return r;
|
|
}
|