decompress_bunzip2: handle concatenated .bz2 files

function                                             old     new   delta
unpack_bz2_stream                                    207     307    +100
start_bunzip                                         199     209     +10
------------------------------------------------------------------------------
(add/remove: 0/0 grow/shrink: 2/0 up/down: 110/0)             Total: 110 bytes

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

5 files changed, 81 insertions, 30 deletions

diff --git a/archival/libunarchive/decompress_bunzip2.c b/archival/libunarchive/decompress_bunzip2.c
index 22015683c..549c8b19f 100644
--- a/archival/libunarchive/decompress_bunzip2.c
+++ b/archival/libunarchive/decompress_bunzip2.c
@@ -44,7 +44,7 @@
 #define RETVAL_LAST_BLOCK               (-1)
 #define RETVAL_NOT_BZIP_DATA            (-2)
 #define RETVAL_UNEXPECTED_INPUT_EOF     (-3)
-#define RETVAL_SHORT_WRITE              (-4)
+//#define RETVAL_SHORT_WRITE              (-4)
 #define RETVAL_DATA_ERROR               (-5)
 #define RETVAL_OUT_OF_MEMORY            (-6)
 #define RETVAL_OBSOLETE_INPUT           (-7)
@@ -584,8 +584,8 @@ int FAST_FUNC read_bunzip(bunzip_data *bd, char *outbuf, int len)
 /* Because bunzip2 is used for help text unpacking, and because bb_show_usage()
    should work for NOFORK applets too, we must be extremely careful to not leak
    any allocations! */
-int FAST_FUNC start_bunzip(bunzip_data **bdp, int in_fd, const unsigned char *inbuf,
-                                                int len)
+int FAST_FUNC start_bunzip(bunzip_data **bdp, int in_fd,
+                const void *inbuf, int len)
 {
         bunzip_data *bd;
         unsigned i;
@@ -606,9 +606,11 @@ int FAST_FUNC start_bunzip(bunzip_data **bdp, int in_fd, const unsigned char *in
         if (-1 == in_fd) {
                 /* in this case, bd->inbuf is read-only */
                 bd->inbuf = (void*)inbuf; /* cast away const-ness */
-                bd->inbufCount = len;
-        } else
+        } else {
                 bd->inbuf = (unsigned char *)(bd + 1);
+                memcpy(bd->inbuf, inbuf, len);
+        }
+        bd->inbufCount = len;
 
         /* Init the CRC32 table (big endian) */
         crc32_filltable(bd->crc32Table, 1);
@@ -652,37 +654,59 @@ IF_DESKTOP(long long) int FAST_FUNC
 unpack_bz2_stream(int src_fd, int dst_fd)
 {
         IF_DESKTOP(long long total_written = 0;)
+                bunzip_data *bd;
         char *outbuf;
-        bunzip_data *bd;
         int i;
+        unsigned len;
 
         outbuf = xmalloc(IOBUF_SIZE);
-        i = start_bunzip(&bd, src_fd, NULL, 0);
-        if (!i) {
-                for (;;) {
-                        i = read_bunzip(bd, outbuf, IOBUF_SIZE);
-                        if (i <= 0) break;
-                        if (i != full_write(dst_fd, outbuf, i)) {
-                                i = RETVAL_SHORT_WRITE;
-                                break;
+        len = 0;
+        while (1) { /* "Process one BZ... stream" loop */
+
+                i = start_bunzip(&bd, src_fd, outbuf + 2, len);
+
+                if (i == 0) {
+                        while (1) { /* "Produce some output bytes" loop */
+                                i = read_bunzip(bd, outbuf, IOBUF_SIZE);
+                                if (i <= 0)
+                                        break;
+                                if (i != full_write(dst_fd, outbuf, i)) {
+                                        bb_error_msg("short write");
+                                        goto release_mem;
+                                }
+                                IF_DESKTOP(total_written += i;)
                         }
-                        IF_DESKTOP(total_written += i;)
                 }
-        }
 
-        /* Check CRC and release memory */
-
-        if (i == RETVAL_LAST_BLOCK) {
+                if (i != RETVAL_LAST_BLOCK) {
+                        bb_error_msg("bunzip error %d", i);
+                        break;
+                }
                 if (bd->headerCRC != bd->totalCRC) {
                         bb_error_msg("CRC error");
-                } else {
-                        i = RETVAL_OK;
+                        break;
                 }
-        } else if (i == RETVAL_SHORT_WRITE) {
-                bb_error_msg("short write");
-        } else {
-                bb_error_msg("bunzip error %d", i);
+
+                /* Successfully unpacked one BZ stream */
+                i = RETVAL_OK;
+
+                /* Do we have "BZ..." after last processed byte?
+                 * pbzip2 (parallelized bzip2) produces such files.
+                 */
+                len = bd->inbufCount - bd->inbufPos;
+                memcpy(outbuf, &bd->inbuf[bd->inbufPos], len);
+                if (len < 2) {
+                        if (safe_read(src_fd, outbuf + len, 2 - len) != 2 - len)
+                                break;
+                        len = 2;
+                }
+                if (*(uint16_t*)outbuf != BZIP2_MAGIC) /* "BZ"? */
+                        break;
+                dealloc_bunzip(bd);
+                len -= 2;
         }
+
+ release_mem:
         dealloc_bunzip(bd);
         free(outbuf);
 
diff --git a/include/unarchive.h b/include/unarchive.h
index b55af6d9d..11d8c77a0 100644
--- a/include/unarchive.h
+++ b/include/unarchive.h
@@ -193,7 +193,7 @@ extern const llist_t *find_list_entry2(const llist_t *list, const char *filename
 
 /* A bit of bunzip2 internals are exposed for compressed help support: */
 typedef struct bunzip_data bunzip_data;
-int start_bunzip(bunzip_data **bdp, int in_fd, const unsigned char *inbuf, int len) FAST_FUNC;
+int start_bunzip(bunzip_data **bdp, int in_fd, const void *inbuf, int len) FAST_FUNC;
 int read_bunzip(bunzip_data *bd, char *outbuf, int len) FAST_FUNC;
 void dealloc_bunzip(bunzip_data *bd) FAST_FUNC;
 
diff --git a/libbb/appletlib.c b/libbb/appletlib.c
index b32ff8808..fab996ca0 100644
--- a/libbb/appletlib.c
+++ b/libbb/appletlib.c
@@ -75,7 +75,7 @@ static const char *unpack_usage_messages(void)
 
         i = start_bunzip(&bd,
                         /* src_fd: */ -1,
-                        /* inbuf:  */ (void *)packed_usage,
+                        /* inbuf:  */ packed_usage,
                         /* len:    */ sizeof(packed_usage));
         /* read_bunzip can longjmp to start_bunzip, and ultimately
          * end up here with i != 0 on read data errors! Not trivial */
diff --git a/miscutils/bbconfig.c b/miscutils/bbconfig.c
index 286077168..1cb42edeb 100644
--- a/miscutils/bbconfig.c
+++ b/miscutils/bbconfig.c
@@ -15,7 +15,7 @@ int bbconfig_main(int argc UNUSED_PARAM, char **argv UNUSED_PARAM)
         bunzip_data *bd;
         int i = start_bunzip(&bd,
                         /* src_fd: */ -1,
-                        /* inbuf:  */ (void *)bbconfig_config_bz2,
+                        /* inbuf:  */ bbconfig_config_bz2,
                         /* len:    */ sizeof(bbconfig_config_bz2));
         /* read_bunzip can longjmp to start_bunzip, and ultimately
          * end up here with i != 0 on read data errors! Not trivial */
diff --git a/testsuite/bunzip2.tests b/testsuite/bunzip2.tests
index 827aee867..fcfce1a31 100755
--- a/testsuite/bunzip2.tests
+++ b/testsuite/bunzip2.tests
@@ -463,6 +463,24 @@ $ECHO -ne "\x40\xa0\x00\x8b\x12\xe8\xfb\xb7\x27\xaa\xd3\x36\x0c\xfc\xe1\x40"
 $ECHO -ne "\x01\xff\x8b\xb9\x22\x9c\x28\x48\x5f\xa5\xca\xf3\x80"
 }
 
+pbzip_4m_zeros() {
+$ECHO -ne "\x42\x5a\x68\x31\x31\x41\x59\x26\x53\x59\x63\xe3\xec\xa2\x00\x06"
+$ECHO -ne "\xe4\xc1\x00\xc0\x00\x02\x00\x00\x08\x20\x00\x30\xcc\x09\xaa\x69"
+$ECHO -ne "\x94\xa1\x36\xa9\x28\x4f\x17\x72\x45\x38\x50\x90\x63\xe3\xec\xa2"
+$ECHO -ne "\x42\x5a\x68\x31\x31\x41\x59\x26\x53\x59\x63\xe3\xec\xa2\x00\x06"
+$ECHO -ne "\xe4\xc1\x00\xc0\x00\x02\x00\x00\x08\x20\x00\x30\xcc\x09\xaa\x69"
+$ECHO -ne "\x94\xa1\x36\xa9\x28\x4f\x17\x72\x45\x38\x50\x90\x63\xe3\xec\xa2"
+$ECHO -ne "\x42\x5a\x68\x31\x31\x41\x59\x26\x53\x59\x63\xe3\xec\xa2\x00\x06"
+$ECHO -ne "\xe4\xc1\x00\xc0\x00\x02\x00\x00\x08\x20\x00\x30\xcc\x09\xaa\x69"
+$ECHO -ne "\x94\xa1\x36\xa9\x28\x4f\x17\x72\x45\x38\x50\x90\x63\xe3\xec\xa2"
+$ECHO -ne "\x42\x5a\x68\x31\x31\x41\x59\x26\x53\x59\x63\xe3\xec\xa2\x00\x06"
+$ECHO -ne "\xe4\xc1\x00\xc0\x00\x02\x00\x00\x08\x20\x00\x30\xcc\x09\xaa\x69"
+$ECHO -ne "\x94\xa1\x36\xa9\x28\x4f\x17\x72\x45\x38\x50\x90\x63\xe3\xec\xa2"
+$ECHO -ne "\x42\x5a\x68\x31\x31\x41\x59\x26\x53\x59\xc9\xb5\x21\xef\x00\x04"
+$ECHO -ne "\x8d\x40\x20\xc0\x00\x01\x00\x00\x08\x20\x00\x30\xcc\x05\x29\xa6"
+$ECHO -ne "\x4a\x11\xb1\x4a\x11\xe2\xee\x48\xa7\x0a\x12\x19\x36\xa4\x3d\xe0"
+}
+
 prep() {
     rm -f t*
     hello_$ext >t1.$ext
@@ -520,9 +538,18 @@ if test "${0##*/}" = "bunzip2.tests"; then
     if test1_bz2 | ${bb}bunzip2 >/dev/null \
         && test "`test1_bz2 | ${bb}bunzip2 | md5sum`" = "61bbeee4be9c6f110a71447f584fda7b  -"
     then
-        echo "PASS: $unpack: test bz2 file"
+        echo "PASS: $unpack: test_bz2 file"
+    else
+        echo "FAIL: $unpack: test_bz2 file"
+        FAILCOUNT=$((FAILCOUNT + 1))
+    fi
+
+    if pbzip_4m_zeros | ${bb}bunzip2 >/dev/null \
+        && test "`pbzip_4m_zeros | ${bb}bunzip2 | md5sum`" = "b5cfa9d6c8febd618f91ac2843d50a1c  -"
+    then
+        echo "PASS: $unpack: pbzip_4m_zeros file"
     else
-        echo "FAIL: $unpack: test bz2 file"
+        echo "FAIL: $unpack: pbzip_4m_zeros file"
         FAILCOUNT=$((FAILCOUNT + 1))
     fi
 fi