/* * Copyright (c) 1986 Regents of the University of California. * All rights reserved. The Berkeley software License Agreement * specifies the terms and conditions for redistribution. * * @(#)kern_exec.c 1.8 (2.11BSD) 1999/9/6 */ #include "param.h" #include "../machine/reg.h" #include "../machine/seg.h" #include "systm.h" #include "map.h" #include "user.h" #include "proc.h" #include "buf.h" #include "inode.h" #include "acct.h" #include "namei.h" #include "fs.h" #include "mount.h" #include "file.h" #include "text.h" #include "signalvar.h" extern char sigprop[]; /* XXX */ /* * exec system call, with and without environments. */ struct execa { char *fname; char **argp; char **envp; }; execv() { ((struct execa *)u.u_ap)->envp = NULL; execve(); } execve() { int nc; register char *cp; register struct buf *bp; struct execa *uap = (struct execa *)u.u_ap; int na, ne, ucp, ap; register int cc; unsigned len; int indir, uid, gid; char *sharg; struct inode *ip; memaddr bno; char cfname[MAXCOMLEN + 1]; #define SHSIZE 32 char cfarg[SHSIZE]; union { char ex_shell[SHSIZE]; /* #! and name of interpreter */ struct exec ex_exec; } exdata; struct nameidata nd; register struct nameidata *ndp = &nd; int resid, error; NDINIT(ndp, LOOKUP, FOLLOW, UIO_USERSPACE, uap->fname); if ((ip = namei(ndp)) == NULL) return; bno = 0; bp = 0; indir = 0; uid = u.u_uid; gid = u.u_groups[0]; if (ip->i_fs->fs_flags & MNT_NOEXEC) { u.u_error = EACCES; goto bad; } if ((ip->i_fs->fs_flags & MNT_NOSUID) == 0) { if (ip->i_mode & ISUID) uid = ip->i_uid; if (ip->i_mode & ISGID) gid = ip->i_gid; } again: if (access(ip, IEXEC)) goto bad; if ((u.u_procp->p_flag & P_TRACED) && access(ip, IREAD)) goto bad; if ((ip->i_mode & IFMT) != IFREG || (ip->i_mode & (IEXEC|(IEXEC>>3)|(IEXEC>>6))) == 0) { u.u_error = EACCES; goto bad; } /* * Read in first few bytes of file for segment sizes, magic number: * 407 = plain executable * 410 = RO text * 411 = separated I/D * 405 = text overlay * 430 = auto-overlay (nonseparate) * 431 = auto-overlay (separate) * Also an ASCII line beginning with #! is * the file name of a ``shell'' and arguments may be prepended * to the argument list if given here. * * SHELL NAMES ARE LIMITED IN LENGTH. * * ONLY ONE ARGUMENT MAY BE PASSED TO THE SHELL FROM * THE ASCII LINE. */ exdata.ex_shell[0] = '\0'; /* for zero length files */ u.u_error = rdwri(UIO_READ, ip, &exdata, sizeof(exdata), (off_t)0, UIO_SYSSPACE, IO_UNIT, &resid); if (u.u_error) goto bad; if (resid > sizeof(exdata) - sizeof(exdata.ex_exec) && exdata.ex_shell[0] != '#') { u.u_error = ENOEXEC; goto bad; } switch((int)exdata.ex_exec.a_magic) { case A_MAGIC1: case A_MAGIC2: case A_MAGIC3: case A_MAGIC4: case A_MAGIC5: case A_MAGIC6: break; default: if (exdata.ex_shell[0] != '#' || exdata.ex_shell[1] != '!' || indir) { u.u_error = ENOEXEC; goto bad; } /* * If setuid/gid scripts were to be disallowed this is where it would * have to be done. * u.u_uid = uid; * u.u_gid = u_groups[0]; */ cp = &exdata.ex_shell[2]; /* skip "#!" */ while (cp < &exdata.ex_shell[SHSIZE]) { if (*cp == '\t') *cp = ' '; else if (*cp == '\n') { *cp = '\0'; break; } cp++; } if (*cp != '\0') { u.u_error = ENOEXEC; goto bad; } cp = &exdata.ex_shell[2]; while (*cp == ' ') cp++; ndp->ni_dirp = cp; while (*cp && *cp != ' ') cp++; cfarg[0] = '\0'; if (*cp) { *cp++ = '\0'; while (*cp == ' ') cp++; if (*cp) bcopy((caddr_t)cp, (caddr_t)cfarg, SHSIZE); } indir = 1; iput(ip); ndp->ni_nameiop = LOOKUP | FOLLOW; ndp->ni_segflg = UIO_SYSSPACE; ip = namei(ndp); if (ip == NULL) return; bcopy((caddr_t)ndp->ni_dent.d_name, (caddr_t)cfname, MAXCOMLEN); cfname[MAXCOMLEN] = '\0'; goto again; } /* * Collect arguments on "file" in swap space. */ na = 0; ne = 0; nc = 0; cc = 0; bno = malloc(swapmap, ctod((int)btoc(NCARGS + MAXBSIZE))); if (bno == 0) { swkill(u.u_procp, "exec"); goto bad; } /* * Copy arguments into file in argdev area. */ if (uap->argp) for (;;) { ap = NULL; sharg = NULL; if (indir && na == 0) { sharg = cfname; ap = (int)sharg; uap->argp++; /* ignore argv[0] */ } else if (indir && (na == 1 && cfarg[0])) { sharg = cfarg; ap = (int)sharg; } else if (indir && (na == 1 || na == 2 && cfarg[0])) ap = (int)uap->fname; else if (uap->argp) { ap = fuword((caddr_t)uap->argp); uap->argp++; } if (ap == NULL && uap->envp) { uap->argp = NULL; if ((ap = fuword((caddr_t)uap->envp)) != NULL) uap->envp++, ne++; } if (ap == NULL) break; na++; if (ap == -1) { u.u_error = EFAULT; break; } do { if (cc <= 0) { /* * We depend on NCARGS being a multiple of * CLSIZE*NBPG. This way we need only check * overflow before each buffer allocation. */ if (nc >= NCARGS-1) { error = E2BIG; break; } if (bp) { mapout(bp); bdwrite(bp); } cc = CLSIZE*NBPG; bp = getblk(swapdev, dbtofsb(clrnd(bno)) + lblkno(nc)); cp = mapin(bp); } if (sharg) { error = copystr(sharg, cp, (unsigned)cc, &len); sharg += len; } else { error = copyinstr((caddr_t)ap, cp, (unsigned)cc, &len); ap += len; } cp += len; nc += len; cc -= len; } while (error == ENOENT); if (error) { u.u_error = error; if (bp) { mapout(bp); bp->b_flags |= B_AGE; bp->b_flags &= ~B_DELWRI; brelse(bp); } bp = 0; goto badarg; } } if (bp) { mapout(bp); bdwrite(bp); } bp = 0; nc = (nc + NBPW-1) & ~(NBPW-1); getxfile(ip, &exdata.ex_exec, nc + (na+4)*NBPW, uid, gid); if (u.u_error) { badarg: for (cc = 0;cc < nc; cc += CLSIZE * NBPG) { daddr_t blkno; blkno = dbtofsb(clrnd(bno)) + lblkno(cc); if (incore(swapdev,blkno)) { bp = bread(swapdev,blkno); bp->b_flags |= B_AGE; /* throw away */ bp->b_flags &= ~B_DELWRI; /* cancel io */ brelse(bp); bp = 0; } } goto bad; } iput(ip); ip = NULL; /* * Copy back arglist. */ ucp = -nc - NBPW; ap = ucp - na*NBPW - 3*NBPW; u.u_ar0[R6] = ap; (void) suword((caddr_t)ap, na-ne); nc = 0; cc = 0; for (;;) { ap += NBPW; if (na == ne) { (void) suword((caddr_t)ap, 0); ap += NBPW; } if (--na < 0) break; (void) suword((caddr_t)ap, ucp); do { if (cc <= 0) { if (bp) { mapout(bp); brelse(bp); } cc = CLSIZE*NBPG; bp = bread(swapdev, dbtofsb(clrnd(bno)) + lblkno(nc)); bp->b_flags |= B_AGE; /* throw away */ bp->b_flags &= ~B_DELWRI; /* cancel io */ cp = mapin(bp); } error = copyoutstr(cp, (caddr_t)ucp, (unsigned)cc, &len); ucp += len; cp += len; nc += len; cc -= len; } while (error == ENOENT); if (error == EFAULT) panic("exec: EFAULT"); } (void) suword((caddr_t)ap, 0); (void) suword((caddr_t)(-NBPW), 0); if (bp) { mapout(bp); bp->b_flags |= B_AGE; brelse(bp); bp = NULL; } execsigs(u.u_procp); for (cp = u.u_pofile, cc = 0; cc <= u.u_lastfile; cc++, cp++) { if (*cp & UF_EXCLOSE) { (void)closef(u.u_ofile[cc]); u.u_ofile[cc] = NULL; *cp = 0; } } while (u.u_lastfile >= 0 && u.u_ofile[u.u_lastfile] == NULL) u.u_lastfile--; /* * inline expansion of setregs(), found * in ../pdp/machdep.c * * setregs(exdata.ex_exec.a_entry); */ u.u_ar0[PC] = exdata.ex_exec.a_entry & ~01; u.u_fps.u_fpsr = 0; /* * Remember file name for accounting. */ u.u_acflag &= ~AFORK; if (indir) bcopy((caddr_t)cfname, (caddr_t)u.u_comm, MAXCOMLEN); else bcopy((caddr_t)ndp->ni_dent.d_name, (caddr_t)u.u_comm, MAXCOMLEN); bad: if (bp) { mapout(bp); bp->b_flags |= B_AGE; brelse(bp); } if (bno) mfree(swapmap, ctod((int)btoc(NCARGS + MAXBSIZE)), bno); if (ip) iput(ip); } /* * Reset signals for an exec of the specified process. In 4.4 this function * was in kern_sig.c but since in 2.11 kern_sig and kern_exec will likely be * in different overlays placing this here potentially saves a kernel overlay * switch. */ void execsigs(p) register struct proc *p; { register int nc; unsigned long mask; /* * Reset caught signals. Held signals remain held * through p_sigmask (unless they were caught, * and are now ignored by default). */ while (p->p_sigcatch) { nc = ffs(p->p_sigcatch); mask = sigmask(nc); p->p_sigcatch &= ~mask; if (sigprop[nc] & SA_IGNORE) { if (nc != SIGCONT) p->p_sigignore |= mask; p->p_sig &= ~mask; } u.u_signal[nc] = SIG_DFL; } /* * Reset stack state to the user stack (disable the alternate stack). */ u.u_sigstk.ss_flags = SA_DISABLE; u.u_sigstk.ss_size = 0; u.u_sigstk.ss_base = 0; u.u_psflags = 0; } /* * Read in and set up memory for executed file. * u.u_error set on error */ getxfile(ip, ep, nargc, uid, gid) struct inode *ip; register struct exec *ep; int nargc, uid, gid; { struct u_ovd sovdata; long lsize; off_t offset; u_int ds, ts, ss; u_int ovhead[NOVL + 1]; int sep, overlay, ovflag, ovmax, resid; overlay = sep = ovflag = 0; switch(ep->a_magic) { case A_MAGIC1: lsize = (long)ep->a_data + ep->a_text; ep->a_data = (u_int)lsize; if (lsize != ep->a_data) { /* check overflow */ u.u_error = ENOMEM; return; } ep->a_text = 0; break; case A_MAGIC3: sep++; break; case A_MAGIC4: overlay++; break; case A_MAGIC5: ovflag++; break; case A_MAGIC6: sep++; ovflag++; break; } if (ip->i_text && (ip->i_text->x_flag & XTRC)) { u.u_error = ETXTBSY; return; } if (ep->a_text != 0 && (ip->i_flag&ITEXT) == 0 && ip->i_count != 1) { register struct file *fp; for (fp = file; fp < fileNFILE; fp++) { if (fp->f_type == DTYPE_INODE && fp->f_count > 0 && (struct inode *)fp->f_data == ip && (fp->f_flag&FWRITE)) { u.u_error = ETXTBSY; return; } } } /* * find text and data sizes try; them out for possible * overflow of max sizes */ ts = btoc(ep->a_text); lsize = (long)ep->a_data + ep->a_bss; if (lsize != (u_int)lsize) { u.u_error = ENOMEM; return; } ds = btoc(lsize); ss = SSIZE + btoc(nargc); /* * if auto overlay get second header */ sovdata = u.u_ovdata; u.u_ovdata.uo_ovbase = 0; u.u_ovdata.uo_curov = 0; if (ovflag) { u.u_error = rdwri(UIO_READ, ip, ovhead, sizeof(ovhead), (off_t)sizeof(struct exec), UIO_SYSSPACE, IO_UNIT, &resid); if (resid != 0) u.u_error = ENOEXEC; if (u.u_error) { u.u_ovdata = sovdata; return; } /* set beginning of overlay segment */ u.u_ovdata.uo_ovbase = ctos(ts); /* 0th entry is max size of the overlays */ ovmax = btoc(ovhead[0]); /* set max number of segm. registers to be used */ u.u_ovdata.uo_nseg = ctos(ovmax); /* set base of data space */ u.u_ovdata.uo_dbase = stoc(u.u_ovdata.uo_ovbase + u.u_ovdata.uo_nseg); /* * Set up a table of offsets to each of the overlay * segements. The ith overlay runs from ov_offst[i-1] * to ov_offst[i]. */ u.u_ovdata.uo_ov_offst[0] = ts; { register int t, i; /* check if any overlay is larger than ovmax */ for (i = 1; i <= NOVL; i++) { if ((t = btoc(ovhead[i])) > ovmax) { u.u_error = ENOEXEC; u.u_ovdata = sovdata; return; } u.u_ovdata.uo_ov_offst[i] = t + u.u_ovdata.uo_ov_offst[i - 1]; } } } if (overlay) { if (u.u_sep == 0 && ctos(ts) != ctos(u.u_tsize) || nargc) { u.u_error = ENOMEM; return; } ds = u.u_dsize; ss = u.u_ssize; sep = u.u_sep; xfree(); xalloc(ip,ep); u.u_ar0[PC] = ep->a_entry & ~01; } else { if (estabur(ts, ds, ss, sep, RO)) { u.u_ovdata = sovdata; return; } /* * allocate and clear core at this point, committed * to the new image */ u.u_prof.pr_scale = 0; if (u.u_procp->p_flag & SVFORK) endvfork(); else xfree(); expand(ds, S_DATA); { register u_int numc, startc; startc = btoc(ep->a_data); /* clear BSS only */ if (startc != 0) startc--; numc = ds - startc; clear(u.u_procp->p_daddr + startc, numc); } expand(ss, S_STACK); clear(u.u_procp->p_saddr, ss); xalloc(ip, ep); /* * read in data segment */ estabur((u_int)0, ds, (u_int)0, 0, RO); offset = sizeof(struct exec); if (ovflag) { offset += sizeof(ovhead); offset += (((long)u.u_ovdata.uo_ov_offst[NOVL]) << 6); } else offset += ep->a_text; rdwri(UIO_READ, ip, (caddr_t) 0, ep->a_data, offset, UIO_USERSPACE, IO_UNIT, (int *)0); /* * set SUID/SGID protections, if no tracing */ if ((u.u_procp->p_flag & P_TRACED)==0) { u.u_uid = uid; u.u_procp->p_uid = uid; u.u_groups[0] = gid; } else psignal(u.u_procp, SIGTRAP); u.u_svuid = u.u_uid; u.u_svgid = u.u_groups[0]; u.u_acflag &= ~ASUGID; /* start fresh setuid/gid priv use */ } u.u_tsize = ts; u.u_dsize = ds; u.u_ssize = ss; u.u_sep = sep; estabur(ts, ds, ss, sep, RO); }