commit 6b802ab9fd967e2478a29a583521a13fb91aef57
parent 5aa63dbe055d7a655ea14626abdf051ff29e057a
Author: Xan Phung <xan.phung@gmail.com>
Date: Sun, 1 Jun 2025 22:09:11 +1000
tar: man page update, and more robust & descriptive errors
1. Manual page has been updated to reflect previous change to tar.
2. Error messages are more detailed & secured against unterminated
header strings. Previously, sanitize() assumed numeric fields were
terminated by space or null, but did not check termination is
actually present.
3. Default mode of the blank header in archive() is now 0600. For
all file headers, this is immediately set to the file's mode. Only
L-headers keep the 0600 value (which are not user visible). GNU
tar uses 0644, but 0600 was chosen for symmetry with sbase
unarchive(), which also uses 0600 as the initial mode.
Diffstat:
| M | tar.1 | | | 14 | ++++++++++---- |
| M | tar.c | | | 42 | +++++++++++++++++++++++------------------- |
2 files changed, 33 insertions(+), 23 deletions(-)
diff --git a/tar.1 b/tar.1
@@ -6,16 +6,18 @@
.Nd create, list or extract a tape archive
.Sh SYNOPSIS
.Nm
+.Cm x | Cm t | Fl x | Fl t
.Op Fl C Ar dir
.Op Fl J | Fl Z | Fl a | Fl j | Fl z
-.Fl x Op Fl m | Fl t
+.Op Fl m
+.Op Fl p
.Op Fl f Ar file
.Op Ar file ...
.Nm
-.Op Fl C Ar dir
+.Cm c | Fl c Op Fl C Ar dir
.Op Fl J | Fl Z | Fl a | Fl j | Fl z
.Op Fl h
-.Fl c Ar path ...
+.Ar path ...
.Op Fl f Ar file
.Sh DESCRIPTION
.Nm
@@ -66,4 +68,8 @@ The
utility is compliant with the UStar (Uniform Standard Tape ARchive)
format defined in the
.St -p1003.1-88
-specification.
+specification. For long file paths (>99 bytes), the UStar, 'L' and 'x'
+header formats are supported for reading (to a maximum size of PATH_MAX
+or 255 bytes, depending on format), and the 'L' format is supported for
+writing (with unlimited path size). Link targets are limited to the
+UStar maximum of 100 bytes.
diff --git a/tar.c b/tar.c
@@ -186,15 +186,15 @@ static void
putoctal(char *dst, unsigned num, int size)
{
if (snprintf(dst, size, "%.*o", size - 1, num) >= size)
- eprintf("snprintf: input number too large\n");
+ eprintf("putoctal: input number '%o' too large\n", num);
}
static int
archive(const char *path)
{
static const struct header blank = {
- "././@LongLink", "0000000" , "0000000", "0000000", "00000000000",
- "00000000000" , " ", AREG , "" , "ustar", "00",
+ "././@LongLink", "0000600", "0000000", "0000000", "00000000000",
+ "00000000000" , " ", AREG , "" , "ustar", "00",
};
char b[BLKSIZ + BLKSIZ], *p;
struct header *h = (struct header *)b;
@@ -221,8 +221,8 @@ archive(const char *path)
h->type = 'L';
putoctal(h->size, n, sizeof(h->size));
putoctal(h->chksum, chksum(h), sizeof(h->chksum));
-
ewrite(tarfd, (char *)h, BLKSIZ);
+
for (p = (char *)path; n > 0; n -= BLKSIZ, p += BLKSIZ) {
if (n < BLKSIZ) {
p = memcpy(h--, p, n);
@@ -285,7 +285,7 @@ unarchive(char *fname, ssize_t l, char b[BLKSIZ])
int fd = -1, lnk = h->type == SYMLINK;
if (!mflag && ((mtime = strtol(h->mtime, &p, 8)) < 0 || *p != '\0'))
- eprintf("strtol %s: invalid number\n", h->mtime);
+ eprintf("strtol %s: invalid mtime\n", h->mtime);
if (strcmp(fname, ".") && strcmp(fname, "./") && remove(fname) < 0)
if (errno != ENOENT) weprintf("remove %s:", fname);
@@ -298,7 +298,7 @@ unarchive(char *fname, ssize_t l, char b[BLKSIZ])
case AREG:
case RESERVED:
if ((mode = strtol(h->mode, &p, 8)) < 0 || *p != '\0')
- eprintf("strtol %s: invalid number\n", h->mode);
+ eprintf("strtol %s: invalid mode\n", h->mode);
fd = open(fname, O_WRONLY | O_TRUNC | O_CREAT, 0600);
if (fd < 0)
eprintf("open %s:", fname);
@@ -313,7 +313,7 @@ unarchive(char *fname, ssize_t l, char b[BLKSIZ])
break;
case DIRECTORY:
if ((mode = strtol(h->mode, &p, 8)) < 0 || *p != '\0')
- eprintf("strtol %s: invalid number\n", h->mode);
+ eprintf("strtol %s: invalid mode\n", h->mode);
if (mkdir(fname, (mode_t)mode) < 0 && errno != EEXIST)
eprintf("mkdir %s:", fname);
pushdirtime(fname, mtime);
@@ -321,18 +321,18 @@ unarchive(char *fname, ssize_t l, char b[BLKSIZ])
case CHARDEV:
case BLOCKDEV:
if ((mode = strtol(h->mode, &p, 8)) < 0 || *p != '\0')
- eprintf("strtol %s: invalid number\n", h->mode);
+ eprintf("strtol %s: invalid mode\n", h->mode);
if ((major = strtol(h->major, &p, 8)) < 0 || *p != '\0')
- eprintf("strtol %s: invalid number\n", h->major);
+ eprintf("strtol %s: invalid major device\n", h->major);
if ((minor = strtol(h->minor, &p, 8)) < 0 || *p != '\0')
- eprintf("strtol %s: invalid number\n", h->minor);
+ eprintf("strtol %s: invalid minor device\n", h->minor);
type = (h->type == CHARDEV) ? S_IFCHR : S_IFBLK;
if (mknod(fname, type | mode, makedev(major, minor)) < 0)
eprintf("mknod %s:", fname);
break;
case FIFO:
if ((mode = strtol(h->mode, &p, 8)) < 0 || *p != '\0')
- eprintf("strtol %s: invalid number\n", h->mode);
+ eprintf("strtol %s: invalid mode\n", h->mode);
if (mknod(fname, S_IFIFO | mode, 0) < 0)
eprintf("mknod %s:", fname);
break;
@@ -341,9 +341,9 @@ unarchive(char *fname, ssize_t l, char b[BLKSIZ])
}
if ((uid = strtol(h->uid, &p, 8)) < 0 || *p != '\0')
- eprintf("strtol %s: invalid number\n", h->uid);
+ eprintf("strtol %s: invalid uid\n", h->uid);
if ((gid = strtol(h->gid, &p, 8)) < 0 || *p != '\0')
- eprintf("strtol %s: invalid number\n", h->gid);
+ eprintf("strtol %s: invalid gid\n", h->gid);
if (fd != -1) {
for (; l > 0; l -= BLKSIZ)
@@ -404,7 +404,7 @@ c(int dirfd, const char *name, struct stat *st, void *data, struct recursor *r)
static void
sanitize(struct header *h)
{
- size_t i, j;
+ size_t i, j, l;
struct {
char *f;
size_t l;
@@ -423,10 +423,14 @@ sanitize(struct header *h)
* NULs as per the ustar specification. Patch all of them to
* use NULs so we can perform string operations on them. */
for (i = 0; i < LEN(fields); i++){
- for (j = 0; j < fields[i].l && fields[i].f[j] == ' '; j++);
- for (; j < fields[i].l; j++)
+ j = 0, l = fields[i].l - 1;
+ for (; j < l && fields[i].f[j] == ' '; j++);
+ for (; j <= l; j++)
if (fields[i].f[j] == ' ')
fields[i].f[j] = '\0';
+ if (fields[i].f[l])
+ eprintf("numeric field #%d (%.*s) is not null or space terminated\n",
+ i, l+1, fields[i].f);
}
}
@@ -434,7 +438,7 @@ static void
chktar(struct header *h)
{
const char *reason;
- char tmp[sizeof h->chksum], *err = "";
+ char tmp[sizeof h->chksum], *err;
long sum, i;
if (h->prefix[0] == '\0' && h->name[0] == '\0') {
@@ -482,13 +486,13 @@ xt(int argc, char *argv[], int mode)
if ((size = strtol(h->size, &p, 8)) < 0 || *p != '\0')
eprintf("strtol %s: invalid size\n", h->size);
- /* Long file path is read direcly into fname*/
+ /* Long file path is read directly into fname*/
if (h->type == 'L' || h->type == 'x' || h->type == 'g') {
/* Read header only up to size of fname buffer */
for (q = fname; q < fname+size; q += BLKSIZ) {
if (q + BLKSIZ >= fname + l)
- eprintf("name exceeds buffer: %s\n", fname);
+ eprintf("name exceeds buffer: %.*s\n", q-fname, fname);
eread(tarfd, q, BLKSIZ);
}
