NetBSD/lib/libkvm/kvm_proc.c
1997-06-20 05:18:22 +00:00

841 lines
20 KiB
C

/* $NetBSD: kvm_proc.c,v 1.17 1997/06/20 05:18:22 mikel Exp $ */
/*-
* Copyright (c) 1994, 1995 Charles M. Hannum. All rights reserved.
* Copyright (c) 1989, 1992, 1993
* The Regents of the University of California. All rights reserved.
*
* This code is derived from software developed by the Computer Systems
* Engineering group at Lawrence Berkeley Laboratory under DARPA contract
* BG 91-66 and contributed to Berkeley.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* 3. All advertising materials mentioning features or use of this software
* must display the following acknowledgement:
* This product includes software developed by the University of
* California, Berkeley and its contributors.
* 4. Neither the name of the University nor the names of its contributors
* may be used to endorse or promote products derived from this software
* without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*/
#if defined(LIBC_SCCS) && !defined(lint)
#if 0
static char sccsid[] = "@(#)kvm_proc.c 8.3 (Berkeley) 9/23/93";
#else
static char *rcsid = "$NetBSD: kvm_proc.c,v 1.17 1997/06/20 05:18:22 mikel Exp $";
#endif
#endif /* LIBC_SCCS and not lint */
/*
* Proc traversal interface for kvm. ps and w are (probably) the exclusive
* users of this code, so we've factored it out into a separate module.
* Thus, we keep this grunge out of the other kvm applications (i.e.,
* most other applications are interested only in open/close/read/nlist).
*/
#include <sys/param.h>
#include <sys/user.h>
#include <sys/proc.h>
#include <sys/exec.h>
#include <sys/stat.h>
#include <sys/ioctl.h>
#include <sys/tty.h>
#include <stdlib.h>
#include <string.h>
#include <unistd.h>
#include <nlist.h>
#include <kvm.h>
#include <vm/vm.h>
#include <vm/vm_param.h>
#include <vm/swap_pager.h>
#include <sys/sysctl.h>
#include <limits.h>
#include <db.h>
#include <paths.h>
#include "kvm_private.h"
#define KREAD(kd, addr, obj) \
(kvm_read(kd, addr, (char *)(obj), sizeof(*obj)) != sizeof(*obj))
int _kvm_readfromcore __P((kvm_t *, u_long, u_long));
int _kvm_readfrompager __P((kvm_t *, struct vm_object *, u_long));
ssize_t kvm_uread __P((kvm_t *, const struct proc *, u_long, char *,
size_t));
static char **kvm_argv __P((kvm_t *, const struct proc *, u_long, int,
int));
static int kvm_deadprocs __P((kvm_t *, int, int, u_long, u_long, int));
static char **kvm_doargv __P((kvm_t *, const struct kinfo_proc *, int,
void (*)(struct ps_strings *, u_long *, int *)));
static int kvm_proclist __P((kvm_t *, int, int, struct proc *,
struct kinfo_proc *, int));
static int proc_verify __P((kvm_t *, u_long, const struct proc *));
static void ps_str_a __P((struct ps_strings *, u_long *, int *));
static void ps_str_e __P((struct ps_strings *, u_long *, int *));
char *
_kvm_uread(kd, p, va, cnt)
kvm_t *kd;
const struct proc *p;
u_long va;
u_long *cnt;
{
register u_long addr, head;
register u_long offset;
struct vm_map_entry vme;
struct vm_object vmo;
int rv;
if (kd->swapspc == 0) {
kd->swapspc = (char *)_kvm_malloc(kd, kd->nbpg);
if (kd->swapspc == 0)
return (0);
}
/*
* Look through the address map for the memory object
* that corresponds to the given virtual address.
* The header just has the entire valid range.
*/
head = (u_long)&p->p_vmspace->vm_map.header;
addr = head;
while (1) {
if (KREAD(kd, addr, &vme))
return (0);
if (va >= vme.start && va < vme.end &&
vme.object.vm_object != 0)
break;
addr = (u_long)vme.next;
if (addr == head)
return (0);
}
/*
* We found the right object -- follow shadow links.
*/
offset = va - vme.start + vme.offset;
addr = (u_long)vme.object.vm_object;
while (1) {
/* Try reading the page from core first. */
if ((rv = _kvm_readfromcore(kd, addr, offset)))
break;
if (KREAD(kd, addr, &vmo))
return (0);
/* If there is a pager here, see if it has the page. */
if (vmo.pager != 0 &&
(rv = _kvm_readfrompager(kd, &vmo, offset)))
break;
/* Move down the shadow chain. */
addr = (u_long)vmo.shadow;
if (addr == 0)
return (0);
offset += vmo.shadow_offset;
}
if (rv == -1)
return (0);
/* Found the page. */
offset %= kd->nbpg;
*cnt = kd->nbpg - offset;
return (&kd->swapspc[offset]);
}
#define vm_page_hash(kd, object, offset) \
(((u_long)object + (u_long)(offset / kd->nbpg)) & kd->vm_page_hash_mask)
int
_kvm_coreinit(kd)
kvm_t *kd;
{
struct nlist nlist[3];
nlist[0].n_name = "_vm_page_buckets";
nlist[1].n_name = "_vm_page_hash_mask";
nlist[2].n_name = 0;
if (kvm_nlist(kd, nlist) != 0)
return (-1);
if (KREAD(kd, nlist[0].n_value, &kd->vm_page_buckets) ||
KREAD(kd, nlist[1].n_value, &kd->vm_page_hash_mask))
return (-1);
return (0);
}
int
_kvm_readfromcore(kd, object, offset)
kvm_t *kd;
u_long object, offset;
{
u_long addr;
struct pglist bucket;
struct vm_page mem;
off_t seekpoint;
if (kd->vm_page_buckets == 0 &&
_kvm_coreinit(kd))
return (-1);
addr = (u_long)&kd->vm_page_buckets[vm_page_hash(kd, object, offset)];
if (KREAD(kd, addr, &bucket))
return (-1);
addr = (u_long)bucket.tqh_first;
offset &= ~(kd->nbpg -1);
while (1) {
if (addr == 0)
return (0);
if (KREAD(kd, addr, &mem))
return (-1);
if ((u_long)mem.object == object &&
(u_long)mem.offset == offset)
break;
addr = (u_long)mem.hashq.tqe_next;
}
seekpoint = mem.phys_addr;
if (lseek(kd->pmfd, seekpoint, 0) == -1)
return (-1);
if (read(kd->pmfd, kd->swapspc, kd->nbpg) != kd->nbpg)
return (-1);
return (1);
}
int
_kvm_readfrompager(kd, vmop, offset)
kvm_t *kd;
struct vm_object *vmop;
u_long offset;
{
u_long addr;
struct pager_struct pager;
struct swpager swap;
int ix;
struct swblock swb;
off_t seekpoint;
/* Read in the pager info and make sure it's a swap device. */
addr = (u_long)vmop->pager;
if (KREAD(kd, addr, &pager) || pager.pg_type != PG_SWAP)
return (-1);
/* Read in the swap_pager private data. */
addr = (u_long)pager.pg_data;
if (KREAD(kd, addr, &swap))
return (-1);
/*
* Calculate the paging offset, and make sure it's within the
* bounds of the pager.
*/
offset += vmop->paging_offset;
ix = offset / dbtob(swap.sw_bsize);
#if 0
if (swap.sw_blocks == 0 || ix >= swap.sw_nblocks)
return (-1);
#else
if (swap.sw_blocks == 0 || ix >= swap.sw_nblocks) {
int i;
printf("BUG BUG BUG BUG:\n");
printf("object %p offset %lx pgoffset %lx ",
vmop, offset - vmop->paging_offset,
(u_long)vmop->paging_offset);
printf("pager %p swpager %p\n",
vmop->pager, pager.pg_data);
printf("osize %lx bsize %x blocks %p nblocks %x\n",
(u_long)swap.sw_osize, swap.sw_bsize, swap.sw_blocks,
swap.sw_nblocks);
for (ix = 0; ix < swap.sw_nblocks; ix++) {
addr = (u_long)&swap.sw_blocks[ix];
if (KREAD(kd, addr, &swb))
return (0);
printf("sw_blocks[%d]: block %x mask %x\n", ix,
swb.swb_block, swb.swb_mask);
}
return (-1);
}
#endif
/* Read in the swap records. */
addr = (u_long)&swap.sw_blocks[ix];
if (KREAD(kd, addr, &swb))
return (-1);
/* Calculate offset within pager. */
offset %= dbtob(swap.sw_bsize);
/* Check that the page is actually present. */
if ((swb.swb_mask & (1 << (offset / kd->nbpg))) == 0)
return (0);
if (!ISALIVE(kd))
return (-1);
/* Calculate the physical address and read the page. */
seekpoint = dbtob(swb.swb_block) + (offset & ~(kd->nbpg -1));
if (lseek(kd->swfd, seekpoint, 0) == -1)
return (-1);
if (read(kd->swfd, kd->swapspc, kd->nbpg) != kd->nbpg)
return (-1);
return (1);
}
/*
* Read proc's from memory file into buffer bp, which has space to hold
* at most maxcnt procs.
*/
static int
kvm_proclist(kd, what, arg, p, bp, maxcnt)
kvm_t *kd;
int what, arg;
struct proc *p;
struct kinfo_proc *bp;
int maxcnt;
{
register int cnt = 0;
struct eproc eproc;
struct pgrp pgrp;
struct session sess;
struct tty tty;
struct proc proc;
for (; cnt < maxcnt && p != NULL; p = proc.p_list.le_next) {
if (KREAD(kd, (u_long)p, &proc)) {
_kvm_err(kd, kd->program, "can't read proc at %x", p);
return (-1);
}
if (KREAD(kd, (u_long)proc.p_cred, &eproc.e_pcred) == 0)
KREAD(kd, (u_long)eproc.e_pcred.pc_ucred,
&eproc.e_ucred);
switch(what) {
case KERN_PROC_PID:
if (proc.p_pid != (pid_t)arg)
continue;
break;
case KERN_PROC_UID:
if (eproc.e_ucred.cr_uid != (uid_t)arg)
continue;
break;
case KERN_PROC_RUID:
if (eproc.e_pcred.p_ruid != (uid_t)arg)
continue;
break;
}
/*
* We're going to add another proc to the set. If this
* will overflow the buffer, assume the reason is because
* nprocs (or the proc list) is corrupt and declare an error.
*/
if (cnt >= maxcnt) {
_kvm_err(kd, kd->program, "nprocs corrupt");
return (-1);
}
/*
* gather eproc
*/
eproc.e_paddr = p;
if (KREAD(kd, (u_long)proc.p_pgrp, &pgrp)) {
_kvm_err(kd, kd->program, "can't read pgrp at %x",
proc.p_pgrp);
return (-1);
}
eproc.e_sess = pgrp.pg_session;
eproc.e_pgid = pgrp.pg_id;
eproc.e_jobc = pgrp.pg_jobc;
if (KREAD(kd, (u_long)pgrp.pg_session, &sess)) {
_kvm_err(kd, kd->program, "can't read session at %x",
pgrp.pg_session);
return (-1);
}
if ((proc.p_flag & P_CONTROLT) && sess.s_ttyp != NULL) {
if (KREAD(kd, (u_long)sess.s_ttyp, &tty)) {
_kvm_err(kd, kd->program,
"can't read tty at %x", sess.s_ttyp);
return (-1);
}
eproc.e_tdev = tty.t_dev;
eproc.e_tsess = tty.t_session;
if (tty.t_pgrp != NULL) {
if (KREAD(kd, (u_long)tty.t_pgrp, &pgrp)) {
_kvm_err(kd, kd->program,
"can't read tpgrp at &x",
tty.t_pgrp);
return (-1);
}
eproc.e_tpgid = pgrp.pg_id;
} else
eproc.e_tpgid = -1;
} else
eproc.e_tdev = NODEV;
eproc.e_flag = sess.s_ttyvp ? EPROC_CTTY : 0;
if (sess.s_leader == p)
eproc.e_flag |= EPROC_SLEADER;
if (proc.p_wmesg)
(void)kvm_read(kd, (u_long)proc.p_wmesg,
eproc.e_wmesg, WMESGLEN);
(void)kvm_read(kd, (u_long)proc.p_vmspace,
(char *)&eproc.e_vm, sizeof(eproc.e_vm));
eproc.e_xsize = eproc.e_xrssize = 0;
eproc.e_xccount = eproc.e_xswrss = 0;
switch (what) {
case KERN_PROC_PGRP:
if (eproc.e_pgid != (pid_t)arg)
continue;
break;
case KERN_PROC_TTY:
if ((proc.p_flag & P_CONTROLT) == 0 ||
eproc.e_tdev != (dev_t)arg)
continue;
break;
}
bcopy(&proc, &bp->kp_proc, sizeof(proc));
bcopy(&eproc, &bp->kp_eproc, sizeof(eproc));
++bp;
++cnt;
}
return (cnt);
}
/*
* Build proc info array by reading in proc list from a crash dump.
* Return number of procs read. maxcnt is the max we will read.
*/
static int
kvm_deadprocs(kd, what, arg, a_allproc, a_zombproc, maxcnt)
kvm_t *kd;
int what, arg;
u_long a_allproc;
u_long a_zombproc;
int maxcnt;
{
register struct kinfo_proc *bp = kd->procbase;
register int acnt, zcnt;
struct proc *p;
if (KREAD(kd, a_allproc, &p)) {
_kvm_err(kd, kd->program, "cannot read allproc");
return (-1);
}
acnt = kvm_proclist(kd, what, arg, p, bp, maxcnt);
if (acnt < 0)
return (acnt);
if (KREAD(kd, a_zombproc, &p)) {
_kvm_err(kd, kd->program, "cannot read zombproc");
return (-1);
}
zcnt = kvm_proclist(kd, what, arg, p, bp + acnt, maxcnt - acnt);
if (zcnt < 0)
zcnt = 0;
return (acnt + zcnt);
}
struct kinfo_proc *
kvm_getprocs(kd, op, arg, cnt)
kvm_t *kd;
int op, arg;
int *cnt;
{
size_t size;
int mib[4], st, nprocs;
if (kd->procbase != 0) {
free((void *)kd->procbase);
/*
* Clear this pointer in case this call fails. Otherwise,
* kvm_close() will free it again.
*/
kd->procbase = 0;
}
if (ISALIVE(kd)) {
size = 0;
mib[0] = CTL_KERN;
mib[1] = KERN_PROC;
mib[2] = op;
mib[3] = arg;
st = sysctl(mib, 4, NULL, &size, NULL, 0);
if (st == -1) {
_kvm_syserr(kd, kd->program, "kvm_getprocs");
return (0);
}
kd->procbase = (struct kinfo_proc *)_kvm_malloc(kd, size);
if (kd->procbase == 0)
return (0);
st = sysctl(mib, 4, kd->procbase, &size, NULL, 0);
if (st == -1) {
_kvm_syserr(kd, kd->program, "kvm_getprocs");
return (0);
}
if (size % sizeof(struct kinfo_proc) != 0) {
_kvm_err(kd, kd->program,
"proc size mismatch (%d total, %d chunks)",
size, sizeof(struct kinfo_proc));
return (0);
}
nprocs = size / sizeof(struct kinfo_proc);
} else {
struct nlist nl[4], *p;
nl[0].n_name = "_nprocs";
nl[1].n_name = "_allproc";
nl[2].n_name = "_zombproc";
nl[3].n_name = 0;
if (kvm_nlist(kd, nl) != 0) {
for (p = nl; p->n_type != 0; ++p)
;
_kvm_err(kd, kd->program,
"%s: no such symbol", p->n_name);
return (0);
}
if (KREAD(kd, nl[0].n_value, &nprocs)) {
_kvm_err(kd, kd->program, "can't read nprocs");
return (0);
}
size = nprocs * sizeof(struct kinfo_proc);
kd->procbase = (struct kinfo_proc *)_kvm_malloc(kd, size);
if (kd->procbase == 0)
return (0);
nprocs = kvm_deadprocs(kd, op, arg, nl[1].n_value,
nl[2].n_value, nprocs);
#ifdef notdef
size = nprocs * sizeof(struct kinfo_proc);
(void)realloc(kd->procbase, size);
#endif
}
*cnt = nprocs;
return (kd->procbase);
}
void
_kvm_freeprocs(kd)
kvm_t *kd;
{
if (kd->procbase) {
free(kd->procbase);
kd->procbase = 0;
}
}
void *
_kvm_realloc(kd, p, n)
kvm_t *kd;
void *p;
size_t n;
{
void *np = (void *)realloc(p, n);
if (np == 0)
_kvm_err(kd, kd->program, "out of memory");
return (np);
}
#ifndef MAX
#define MAX(a, b) ((a) > (b) ? (a) : (b))
#endif
/*
* Read in an argument vector from the user address space of process p.
* addr if the user-space base address of narg null-terminated contiguous
* strings. This is used to read in both the command arguments and
* environment strings. Read at most maxcnt characters of strings.
*/
static char **
kvm_argv(kd, p, addr, narg, maxcnt)
kvm_t *kd;
const struct proc *p;
register u_long addr;
register int narg;
register int maxcnt;
{
register char *np, *cp, *ep, *ap;
register u_long oaddr = -1;
register int len, cc;
register char **argv;
/*
* Check that there aren't an unreasonable number of agruments,
* and that the address is in user space.
*/
if (narg > ARG_MAX || addr < VM_MIN_ADDRESS || addr >= VM_MAXUSER_ADDRESS)
return (0);
if (kd->argv == 0) {
/*
* Try to avoid reallocs.
*/
kd->argc = MAX(narg + 1, 32);
kd->argv = (char **)_kvm_malloc(kd, kd->argc *
sizeof(*kd->argv));
if (kd->argv == 0)
return (0);
} else if (narg + 1 > kd->argc) {
kd->argc = MAX(2 * kd->argc, narg + 1);
kd->argv = (char **)_kvm_realloc(kd, kd->argv, kd->argc *
sizeof(*kd->argv));
if (kd->argv == 0)
return (0);
}
if (kd->argspc == 0) {
kd->argspc = (char *)_kvm_malloc(kd, kd->nbpg);
if (kd->argspc == 0)
return (0);
kd->arglen = kd->nbpg;
}
if (kd->argbuf == 0) {
kd->argbuf = (char *)_kvm_malloc(kd, kd->nbpg);
if (kd->argbuf == 0)
return (0);
}
cc = sizeof(char *) * narg;
if (kvm_uread(kd, p, addr, (char *)kd->argv, cc) != cc)
return (0);
ap = np = kd->argspc;
argv = kd->argv;
len = 0;
/*
* Loop over pages, filling in the argument vector.
*/
while (argv < kd->argv + narg && *argv != 0) {
addr = (u_long)*argv & ~(kd->nbpg - 1);
if (addr != oaddr) {
if (kvm_uread(kd, p, addr, kd->argbuf, kd->nbpg) !=
kd->nbpg)
return (0);
oaddr = addr;
}
addr = (u_long)*argv & (kd->nbpg - 1);
cp = kd->argbuf + addr;
cc = kd->nbpg - addr;
if (maxcnt > 0 && cc > maxcnt - len)
cc = maxcnt - len;;
ep = memchr(cp, '\0', cc);
if (ep != 0)
cc = ep - cp + 1;
if (len + cc > kd->arglen) {
register int off;
register char **pp;
register char *op = kd->argspc;
kd->arglen *= 2;
kd->argspc = (char *)_kvm_realloc(kd, kd->argspc,
kd->arglen);
if (kd->argspc == 0)
return (0);
/*
* Adjust argv pointers in case realloc moved
* the string space.
*/
off = kd->argspc - op;
for (pp = kd->argv; pp < argv; pp++)
*pp += off;
ap += off;
np += off;
}
memcpy(np, cp, cc);
np += cc;
len += cc;
if (ep != 0) {
*argv++ = ap;
ap = np;
} else
*argv += cc;
if (maxcnt > 0 && len >= maxcnt) {
/*
* We're stopping prematurely. Terminate the
* current string.
*/
if (ep == 0) {
*np = '\0';
*argv++ = ap;
}
break;
}
}
/* Make sure argv is terminated. */
*argv = 0;
return (kd->argv);
}
static void
ps_str_a(p, addr, n)
struct ps_strings *p;
u_long *addr;
int *n;
{
*addr = (u_long)p->ps_argvstr;
*n = p->ps_nargvstr;
}
static void
ps_str_e(p, addr, n)
struct ps_strings *p;
u_long *addr;
int *n;
{
*addr = (u_long)p->ps_envstr;
*n = p->ps_nenvstr;
}
/*
* Determine if the proc indicated by p is still active.
* This test is not 100% foolproof in theory, but chances of
* being wrong are very low.
*/
static int
proc_verify(kd, kernp, p)
kvm_t *kd;
u_long kernp;
const struct proc *p;
{
struct proc kernproc;
/*
* Just read in the whole proc. It's not that big relative
* to the cost of the read system call.
*/
if (kvm_read(kd, kernp, (char *)&kernproc, sizeof(kernproc)) !=
sizeof(kernproc))
return (0);
return (p->p_pid == kernproc.p_pid &&
(kernproc.p_stat != SZOMB || p->p_stat == SZOMB));
}
static char **
kvm_doargv(kd, kp, nchr, info)
kvm_t *kd;
const struct kinfo_proc *kp;
int nchr;
void (*info)(struct ps_strings *, u_long *, int *);
{
register const struct proc *p = &kp->kp_proc;
register char **ap;
u_long addr;
int cnt;
struct ps_strings arginfo;
/*
* Pointers are stored at the top of the user stack.
*/
if (p->p_stat == SZOMB ||
kvm_uread(kd, p, USRSTACK - sizeof(arginfo), (char *)&arginfo,
sizeof(arginfo)) != sizeof(arginfo))
return (0);
(*info)(&arginfo, &addr, &cnt);
if (cnt == 0)
return (0);
ap = kvm_argv(kd, p, addr, cnt, nchr);
/*
* For live kernels, make sure this process didn't go away.
*/
if (ap != 0 && ISALIVE(kd) &&
!proc_verify(kd, (u_long)kp->kp_eproc.e_paddr, p))
ap = 0;
return (ap);
}
/*
* Get the command args. This code is now machine independent.
*/
char **
kvm_getargv(kd, kp, nchr)
kvm_t *kd;
const struct kinfo_proc *kp;
int nchr;
{
return (kvm_doargv(kd, kp, nchr, ps_str_a));
}
char **
kvm_getenvv(kd, kp, nchr)
kvm_t *kd;
const struct kinfo_proc *kp;
int nchr;
{
return (kvm_doargv(kd, kp, nchr, ps_str_e));
}
/*
* Read from user space. The user context is given by p.
*/
ssize_t
kvm_uread(kd, p, uva, buf, len)
kvm_t *kd;
register const struct proc *p;
register u_long uva;
register char *buf;
register size_t len;
{
register char *cp;
cp = buf;
while (len > 0) {
register int cc;
register char *dp;
u_long cnt;
dp = _kvm_uread(kd, p, uva, &cnt);
if (dp == 0) {
_kvm_err(kd, 0, "invalid address (%x)", uva);
return (0);
}
cc = MIN(cnt, len);
bcopy(dp, cp, cc);
cp += cc;
uva += cc;
len -= cc;
}
return (ssize_t)(cp - buf);
}