update seamless uncompression code

This change makes "tar tf hello_world.txz" work without
adding special-casing for ".txz" extension. It also removes
ever-growing magic checking code in rpm2cpio and get_header_tar -
we reuse one which lives in setup_unzip_on_fd.

function                                             old     new   delta
unpack_gz_stream                                       7     566    +559
check_signature16                                      -      70     +70
setup_unzip_on_fd                                     99     142     +43
handle_SIGCHLD                                         -      41     +41
unpack_bz2_stream                                    342     376     +34
unzip_main                                          2352    2385     +33
bbunpack                                             503     533     +30
open_transformer                                      74     102     +28
unpack_Z_stream                                     1278    1304     +26
unpack_gunzip                                        101     123     +22
init_transformer_aux_data                              -      18     +18
unpack_xz_stream                                    2388    2402     +14
open_zipped                                          131     141     +10
rpm_main                                            1358    1363      +5
get_header_tar_lzma                                   52      57      +5
get_header_tar_bz2                                    52      57      +5
unpack_lzma_stream                                  2698    2702      +4
hash_find                                            234     233      -1
get_header_tar                                      1759    1733     -26
get_header_tar_gz                                     92      57     -35
unpack_uncompress                                     51      12     -39
rpm2cpio_main                                        201     147     -54
unpack_unxz                                           67      12     -55
unpack_bz2_stream_prime                               55       -     -55
get_header_tar_Z                                      86       -     -86
unpack_gz_stream_with_info                           539       -    -539
------------------------------------------------------------------------------
(add/remove: 3/3 grow/shrink: 14/6 up/down: 947/-890)          Total: 57 bytes

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

10 files changed, 95 insertions, 138 deletions

diff --git a/archival/libarchive/decompress_bunzip2.c b/archival/libarchive/decompress_bunzip2.c
index c4640d489..dc252bb82 100644
--- a/archival/libarchive/decompress_bunzip2.c
+++ b/archival/libarchive/decompress_bunzip2.c
@@ -721,7 +721,7 @@ void FAST_FUNC dealloc_bunzip(bunzip_data *bd)
 
 /* Decompress src_fd to dst_fd.  Stops at end of bzip data, not end of file. */
 IF_DESKTOP(long long) int FAST_FUNC
-unpack_bz2_stream(int src_fd, int dst_fd)
+unpack_bz2_stream(transformer_aux_data_t *aux, int src_fd, int dst_fd)
 {
         IF_DESKTOP(long long total_written = 0;)
         bunzip_data *bd;
@@ -729,6 +729,9 @@ unpack_bz2_stream(int src_fd, int dst_fd)
         int i;
         unsigned len;
 
+        if (check_signature16(aux, src_fd, BZIP2_MAGIC))
+                return -1;
+
         outbuf = xmalloc(IOBUF_SIZE);
         len = 0;
         while (1) { /* "Process one BZ... stream" loop */
@@ -794,17 +797,6 @@ unpack_bz2_stream(int src_fd, int dst_fd)
         return i ? i : IF_DESKTOP(total_written) + 0;
 }
 
-IF_DESKTOP(long long) int FAST_FUNC
-unpack_bz2_stream_prime(int src_fd, int dst_fd)
-{
-        uint16_t magic2;
-        xread(src_fd, &magic2, 2);
-        if (magic2 != BZIP2_MAGIC) {
-                bb_error_msg_and_die("invalid magic");
-        }
-        return unpack_bz2_stream(src_fd, dst_fd);
-}
-
 #ifdef TESTING
 
 static char *const bunzip_errors[] = {
@@ -819,7 +811,7 @@ int main(int argc, char **argv)
         int i;
         char c;
 
-        int i = unpack_bz2_stream_prime(0, 1);
+        int i = unpack_bz2_stream(0, 1);
         if (i < 0)
                 fprintf(stderr, "%s\n", bunzip_errors[-i]);
         else if (read(STDIN_FILENO, &c, 1))
diff --git a/archival/libarchive/decompress_gunzip.c b/archival/libarchive/decompress_gunzip.c
index 50873e3f6..f1c9a79e5 100644
--- a/archival/libarchive/decompress_gunzip.c
+++ b/archival/libarchive/decompress_gunzip.c
@@ -1034,22 +1034,22 @@ inflate_unzip_internal(STATE_PARAM int in, int out)
 /* For unzip */
 
 IF_DESKTOP(long long) int FAST_FUNC
-inflate_unzip(inflate_unzip_result *res, off_t compr_size, int in, int out)
+inflate_unzip(transformer_aux_data_t *aux, int in, int out)
 {
         IF_DESKTOP(long long) int n;
         DECLARE_STATE;
 
         ALLOC_STATE;
 
-        to_read = compr_size;
+        to_read = aux->bytes_in;
 //      bytebuffer_max = 0x8000;
         bytebuffer_offset = 4;
         bytebuffer = xmalloc(bytebuffer_max);
         n = inflate_unzip_internal(PASS_STATE in, out);
         free(bytebuffer);
 
-        res->crc = gunzip_crc;
-        res->bytes_out = gunzip_bytes_out;
+        aux->crc32 = gunzip_crc;
+        aux->bytes_out = gunzip_bytes_out;
         DEALLOC_STATE;
         return n;
 }
@@ -1107,7 +1107,7 @@ static uint32_t buffer_read_le_u32(STATE_PARAM_ONLY)
         return res;
 }
 
-static int check_header_gzip(STATE_PARAM unpack_info_t *info)
+static int check_header_gzip(STATE_PARAM transformer_aux_data_t *aux)
 {
         union {
                 unsigned char raw[8];
@@ -1169,8 +1169,8 @@ static int check_header_gzip(STATE_PARAM unpack_info_t *info)
                 }
         }
 
-        if (info)
-                info->mtime = SWAP_LE32(header.formatted.mtime);
+        if (aux)
+                aux->mtime = SWAP_LE32(header.formatted.mtime);
 
         /* Read the header checksum */
         if (header.formatted.flags & 0x02) {
@@ -1182,12 +1182,15 @@ static int check_header_gzip(STATE_PARAM unpack_info_t *info)
 }
 
 IF_DESKTOP(long long) int FAST_FUNC
-unpack_gz_stream_with_info(int src_fd, int dst_fd, unpack_info_t *info)
+unpack_gz_stream(transformer_aux_data_t *aux, int src_fd, int dst_fd)
 {
         uint32_t v32;
         IF_DESKTOP(long long) int total, n;
         DECLARE_STATE;
 
+        if (check_signature16(aux, src_fd, GZIP_MAGIC))
+                return -1;
+
         total = 0;
 
         ALLOC_STATE;
@@ -1197,7 +1200,7 @@ unpack_gz_stream_with_info(int src_fd, int dst_fd, unpack_info_t *info)
         gunzip_src_fd = src_fd;
 
  again:
-        if (!check_header_gzip(PASS_STATE info)) {
+        if (!check_header_gzip(PASS_STATE aux)) {
                 bb_error_msg("corrupted data");
                 total = -1;
                 goto ret;
@@ -1248,9 +1251,3 @@ unpack_gz_stream_with_info(int src_fd, int dst_fd, unpack_info_t *info)
         DEALLOC_STATE;
         return total;
 }
-
-IF_DESKTOP(long long) int FAST_FUNC
-unpack_gz_stream(int in, int out)
-{
-        return unpack_gz_stream_with_info(in, out, NULL);
-}
diff --git a/archival/libarchive/decompress_uncompress.c b/archival/libarchive/decompress_uncompress.c
index 289f9e233..e9bbfb9bd 100644
--- a/archival/libarchive/decompress_uncompress.c
+++ b/archival/libarchive/decompress_uncompress.c
@@ -73,7 +73,7 @@
  */
 
 IF_DESKTOP(long long) int FAST_FUNC
-unpack_Z_stream(int src_fd, int dst_fd)
+unpack_Z_stream(transformer_aux_data_t *aux, int src_fd, int dst_fd)
 {
         IF_DESKTOP(long long total_written = 0;)
         IF_DESKTOP(long long) int retval = -1;
@@ -103,6 +103,9 @@ unpack_Z_stream(int src_fd, int dst_fd)
         /* block compress mode -C compatible with 2.0 */
         int block_mode; /* = BLOCK_MODE; */
 
+        if (check_signature16(aux, src_fd, COMPRESS_MAGIC))
+                return -1;
+
         inbuf = xzalloc(IBUFSIZ + 64);
         outbuf = xzalloc(OBUFSIZ + 2048);
         htab = xzalloc(HSIZE);  /* wasn't zeroed out before, maybe can xmalloc? */
diff --git a/archival/libarchive/decompress_unlzma.c b/archival/libarchive/decompress_unlzma.c
index 3631b50cc..cfde8ea56 100644
--- a/archival/libarchive/decompress_unlzma.c
+++ b/archival/libarchive/decompress_unlzma.c
@@ -213,7 +213,7 @@ enum {
 
 
 IF_DESKTOP(long long) int FAST_FUNC
-unpack_lzma_stream(int src_fd, int dst_fd)
+unpack_lzma_stream(transformer_aux_data_t *aux UNUSED_PARAM, int src_fd, int dst_fd)
 {
         IF_DESKTOP(long long total_written = 0;)
         lzma_header_t header;
diff --git a/archival/libarchive/decompress_unxz.c b/archival/libarchive/decompress_unxz.c
index 3e5d4edca..79b48a152 100644
--- a/archival/libarchive/decompress_unxz.c
+++ b/archival/libarchive/decompress_unxz.c
@@ -38,7 +38,7 @@ static uint32_t xz_crc32(const uint8_t *buf, size_t size, uint32_t crc)
 #include "unxz/xz_dec_stream.c"
 
 IF_DESKTOP(long long) int FAST_FUNC
-unpack_xz_stream(int src_fd, int dst_fd)
+unpack_xz_stream(transformer_aux_data_t *aux, int src_fd, int dst_fd)
 {
         struct xz_buf iobuf;
         struct xz_dec *state;
@@ -49,13 +49,17 @@ unpack_xz_stream(int src_fd, int dst_fd)
                 global_crc32_table = crc32_filltable(NULL, /*endian:*/ 0);
 
         memset(&iobuf, 0, sizeof(iobuf));
-        /* Preload XZ file signature */
-        membuf = (void*) strcpy(xmalloc(2 * BUFSIZ), HEADER_MAGIC);
+        membuf = xmalloc(2 * BUFSIZ);
         iobuf.in = membuf;
-        iobuf.in_size = HEADER_MAGIC_SIZE;
         iobuf.out = membuf + BUFSIZ;
         iobuf.out_size = BUFSIZ;
 
+        if (!aux || aux->check_signature == 0) {
+                /* Preload XZ file signature */
+                strcpy((char*)membuf, HEADER_MAGIC);
+                iobuf.in_size = HEADER_MAGIC_SIZE;
+        } /* else: let xz code read & check it */
+
         /* Limit memory usage to about 64 MiB. */
         state = xz_dec_init(XZ_DYNALLOC, 64*1024*1024);
 
diff --git a/archival/libarchive/get_header_tar.c b/archival/libarchive/get_header_tar.c
index 8c699754b..80a709144 100644
--- a/archival/libarchive/get_header_tar.c
+++ b/archival/libarchive/get_header_tar.c
@@ -235,43 +235,18 @@ char FAST_FUNC get_header_tar(archive_handle_t *archive_handle)
              || memcmp(tar.magic, "\0\0\0\0", 5) != 0)
         ) {
 #if ENABLE_FEATURE_TAR_AUTODETECT
-                char FAST_FUNC (*get_header_ptr)(archive_handle_t *);
-                uint16_t magic2;
-
  autodetect:
-                magic2 = *(bb__aliased_uint16_t*)tar.name;
-                /* tar gz/bz autodetect: check for gz/bz2 magic.
-                 * If we see the magic, and it is the very first block,
-                 * we can switch to get_header_tar_gz/bz2/lzma().
-                 * Needs seekable fd. I wish recv(MSG_PEEK) works
-                 * on any fd... */
-# if ENABLE_FEATURE_SEAMLESS_GZ
-                if (magic2 == GZIP_MAGIC) {
-                        get_header_ptr = get_header_tar_gz;
-                } else
-# endif
-# if ENABLE_FEATURE_SEAMLESS_BZ2
-                if (magic2 == BZIP2_MAGIC
-                 && tar.name[2] == 'h' && isdigit(tar.name[3])
-                ) { /* bzip2 */
-                        get_header_ptr = get_header_tar_bz2;
-                } else
-# endif
-# if ENABLE_FEATURE_SEAMLESS_XZ
-                //TODO: if (magic2 == XZ_MAGIC1)...
-                //else
-# endif
-                        goto err;
                 /* Two different causes for lseek() != 0:
                  * unseekable fd (would like to support that too, but...),
                  * or not first block (false positive, it's not .gz/.bz2!) */
                 if (lseek(archive_handle->src_fd, -i, SEEK_CUR) != 0)
                         goto err;
-                while (get_header_ptr(archive_handle) == EXIT_SUCCESS)
-                        continue;
-                return EXIT_FAILURE;
+                if (setup_unzip_on_fd(archive_handle->src_fd, /*fail_if_not_detected:*/ 0) != 0)
  err:
-#endif /* FEATURE_TAR_AUTODETECT */
+                        bb_error_msg_and_die("invalid tar magic");
+                archive_handle->offset = 0;
+                goto again_after_align;
+#endif
                 bb_error_msg_and_die("invalid tar magic");
         }
 
diff --git a/archival/libarchive/get_header_tar_bz2.c b/archival/libarchive/get_header_tar_bz2.c
index e012dec3b..0ee00df53 100644
--- a/archival/libarchive/get_header_tar_bz2.c
+++ b/archival/libarchive/get_header_tar_bz2.c
@@ -11,7 +11,7 @@ char FAST_FUNC get_header_tar_bz2(archive_handle_t *archive_handle)
         /* Can't lseek over pipes */
         archive_handle->seek = seek_by_read;
 
-        open_transformer(archive_handle->src_fd, unpack_bz2_stream_prime, "bunzip2");
+        open_transformer_with_sig(archive_handle->src_fd, unpack_bz2_stream, "bunzip2");
         archive_handle->offset = 0;
         while (get_header_tar(archive_handle) == EXIT_SUCCESS)
                 continue;
diff --git a/archival/libarchive/get_header_tar_gz.c b/archival/libarchive/get_header_tar_gz.c
index b9679b0bd..03284342b 100644
--- a/archival/libarchive/get_header_tar_gz.c
+++ b/archival/libarchive/get_header_tar_gz.c
@@ -8,25 +8,10 @@
 
 char FAST_FUNC get_header_tar_gz(archive_handle_t *archive_handle)
 {
-#if BB_MMU
-        uint16_t magic;
-#endif
-
         /* Can't lseek over pipes */
         archive_handle->seek = seek_by_read;
 
-        /* Check gzip magic only if open_transformer will invoke unpack_gz_stream (MMU case).
-         * Otherwise, it will invoke an external helper "gunzip -cf" (NOMMU case) which will
-         * need the header. */
-#if BB_MMU
-        xread(archive_handle->src_fd, &magic, 2);
-        /* Can skip this check, but error message will be less clear */
-        if (magic != GZIP_MAGIC) {
-                bb_error_msg_and_die("invalid gzip magic");
-        }
-#endif
-
-        open_transformer(archive_handle->src_fd, unpack_gz_stream, "gunzip");
+        open_transformer_with_sig(archive_handle->src_fd, unpack_gz_stream, "gunzip");
         archive_handle->offset = 0;
         while (get_header_tar(archive_handle) == EXIT_SUCCESS)
                 continue;
diff --git a/archival/libarchive/get_header_tar_lzma.c b/archival/libarchive/get_header_tar_lzma.c
index 666700729..d565a217d 100644
--- a/archival/libarchive/get_header_tar_lzma.c
+++ b/archival/libarchive/get_header_tar_lzma.c
@@ -14,7 +14,7 @@ char FAST_FUNC get_header_tar_lzma(archive_handle_t *archive_handle)
         /* Can't lseek over pipes */
         archive_handle->seek = seek_by_read;
 
-        open_transformer(archive_handle->src_fd, unpack_lzma_stream, "unlzma");
+        open_transformer_with_sig(archive_handle->src_fd, unpack_lzma_stream, "unlzma");
         archive_handle->offset = 0;
         while (get_header_tar(archive_handle) == EXIT_SUCCESS)
                 continue;
diff --git a/archival/libarchive/open_transformer.c b/archival/libarchive/open_transformer.c
index 743ffee02..693ae9995 100644
--- a/archival/libarchive/open_transformer.c
+++ b/archival/libarchive/open_transformer.c
@@ -6,24 +6,36 @@
 #include "libbb.h"
 #include "bb_archive.h"
 
-#define ZIPPED (ENABLE_FEATURE_SEAMLESS_LZMA \
-        || ENABLE_FEATURE_SEAMLESS_BZ2 \
-        || ENABLE_FEATURE_SEAMLESS_GZ \
-        /* || ENABLE_FEATURE_SEAMLESS_Z */ \
-)
+void FAST_FUNC init_transformer_aux_data(transformer_aux_data_t *aux)
+{
+        memset(aux, 0, sizeof(*aux));
+}
 
-#if ZIPPED
-# include "bb_archive.h"
+int FAST_FUNC check_signature16(transformer_aux_data_t *aux, int src_fd, unsigned magic16)
+{
+        if (aux && aux->check_signature) {
+                uint16_t magic2;
+                if (full_read(src_fd, &magic2, 2) != 2 || magic2 != magic16) {
+                        bb_error_msg("invalid magic");
+#if 0 /* possible future extension */
+                        if (aux->check_signature > 1)
+                                xfunc_die();
 #endif
+                        return -1;
+                }
+        }
+        return 0;
+}
 
 /* transformer(), more than meets the eye */
-/*
- * On MMU machine, the transform_prog is removed by macro magic
- * in include/archive.h. On NOMMU, transformer is removed.
- */
+#if BB_MMU
 void FAST_FUNC open_transformer(int fd,
-        IF_DESKTOP(long long) int FAST_FUNC (*transformer)(int src_fd, int dst_fd),
-        const char *transform_prog)
+        int check_signature,
+        IF_DESKTOP(long long) int FAST_FUNC (*transformer)(transformer_aux_data_t *aux, int src_fd, int dst_fd)
+)
+#else
+void FAST_FUNC open_transformer(int fd, const char *transform_prog)
+#endif
 {
         struct fd_pair fd_pipe;
         int pid;
@@ -35,13 +47,18 @@ void FAST_FUNC open_transformer(int fd,
                 close(fd_pipe.rd); /* we don't want to read from the parent */
                 // FIXME: error check?
 #if BB_MMU
-                transformer(fd, fd_pipe.wr);
-                if (ENABLE_FEATURE_CLEAN_UP) {
-                        close(fd_pipe.wr); /* send EOF */
-                        close(fd);
+                {
+                        transformer_aux_data_t aux;
+                        init_transformer_aux_data(&aux);
+                        aux.check_signature = check_signature;
+                        transformer(&aux, fd, fd_pipe.wr);
+                        if (ENABLE_FEATURE_CLEAN_UP) {
+                                close(fd_pipe.wr); /* send EOF */
+                                close(fd);
+                        }
+                        /* must be _exit! bug was actually seen here */
+                        _exit(EXIT_SUCCESS);
                 }
-                /* must be _exit! bug was actually seen here */
-                _exit(EXIT_SUCCESS);
 #else
                 {
                         char *argv[4];
@@ -64,26 +81,21 @@ void FAST_FUNC open_transformer(int fd,
 }
 
 
+#if SEAMLESS_COMPRESSION
+
 /* Used by e.g. rpm which gives us a fd without filename,
  * thus we can't guess the format from filename's extension.
  */
-#if ZIPPED
-void FAST_FUNC setup_unzip_on_fd(int fd /*, int fail_if_not_detected*/)
+int FAST_FUNC setup_unzip_on_fd(int fd, int fail_if_not_detected)
 {
-        const int fail_if_not_detected = 1;
         union {
                 uint8_t b[4];
                 uint16_t b16[2];
                 uint32_t b32[1];
         } magic;
         int offset = -2;
-# if BB_MMU
-        IF_DESKTOP(long long) int FAST_FUNC (*xformer)(int src_fd, int dst_fd);
-        enum { xformer_prog = 0 };
-# else
-        enum { xformer = 0 };
-        const char *xformer_prog;
-# endif
+        USE_FOR_MMU(IF_DESKTOP(long long) int FAST_FUNC (*xformer)(transformer_aux_data_t *aux, int src_fd, int dst_fd);)
+        USE_FOR_NOMMU(const char *xformer_prog;)
 
         /* .gz and .bz2 both have 2-byte signature, and their
          * unpack_XXX_stream wants this header skipped. */
@@ -91,21 +103,15 @@ void FAST_FUNC setup_unzip_on_fd(int fd /*, int fail_if_not_detected*/)
         if (ENABLE_FEATURE_SEAMLESS_GZ
          && magic.b16[0] == GZIP_MAGIC
         ) {
-# if BB_MMU
-                xformer = unpack_gz_stream;
-# else
-                xformer_prog = "gunzip";
-# endif
+                USE_FOR_MMU(xformer = unpack_gz_stream;)
+                USE_FOR_NOMMU(xformer_prog = "gunzip";)
                 goto found_magic;
         }
         if (ENABLE_FEATURE_SEAMLESS_BZ2
          && magic.b16[0] == BZIP2_MAGIC
         ) {
-# if BB_MMU
-                xformer = unpack_bz2_stream;
-# else
-                xformer_prog = "bunzip2";
-# endif
+                USE_FOR_MMU(xformer = unpack_bz2_stream;)
+                USE_FOR_NOMMU(xformer_prog = "bunzip2";)
                 goto found_magic;
         }
         if (ENABLE_FEATURE_SEAMLESS_XZ
@@ -114,13 +120,8 @@ void FAST_FUNC setup_unzip_on_fd(int fd /*, int fail_if_not_detected*/)
                 offset = -6;
                 xread(fd, magic.b32, sizeof(magic.b32[0]));
                 if (magic.b32[0] == XZ_MAGIC2) {
-# if BB_MMU
-                        xformer = unpack_xz_stream;
-                        /* unpack_xz_stream wants fd at position 6, no need to seek */
-                        //xlseek(fd, offset, SEEK_CUR);
-# else
-                        xformer_prog = "unxz";
-# endif
+                        USE_FOR_MMU(xformer = unpack_xz_stream;)
+                        USE_FOR_NOMMU(xformer_prog = "unxz";)
                         goto found_magic;
                 }
         }
@@ -132,24 +133,23 @@ void FAST_FUNC setup_unzip_on_fd(int fd /*, int fail_if_not_detected*/)
                         IF_FEATURE_SEAMLESS_XZ("/xz")
                         " magic");
         xlseek(fd, offset, SEEK_CUR);
-        return;
+        return 1;
 
  found_magic:
-# if !BB_MMU
+# if BB_MMU
+        open_transformer_with_no_sig(fd, xformer);
+# else
         /* NOMMU version of open_transformer execs
          * an external unzipper that wants
          * file position at the start of the file */
         xlseek(fd, offset, SEEK_CUR);
+        open_transformer_with_sig(fd, xformer, xformer_prog);
 # endif
-        open_transformer(fd, xformer, xformer_prog);
+        return 0;
 }
-#endif /* ZIPPED */
 
 int FAST_FUNC open_zipped(const char *fname)
 {
-#if !ZIPPED
-        return open(fname, O_RDONLY);
-#else
         char *sfx;
         int fd;
 
@@ -162,20 +162,21 @@ int FAST_FUNC open_zipped(const char *fname)
                 sfx++;
                 if (ENABLE_FEATURE_SEAMLESS_LZMA && strcmp(sfx, "lzma") == 0)
                         /* .lzma has no header/signature, just trust it */
-                        open_transformer(fd, unpack_lzma_stream, "unlzma");
+                        open_transformer_with_sig(fd, unpack_lzma_stream, "unlzma");
                 else
                 if ((ENABLE_FEATURE_SEAMLESS_GZ && strcmp(sfx, "gz") == 0)
                  || (ENABLE_FEATURE_SEAMLESS_BZ2 && strcmp(sfx, "bz2") == 0)
                  || (ENABLE_FEATURE_SEAMLESS_XZ && strcmp(sfx, "xz") == 0)
                 ) {
-                        setup_unzip_on_fd(fd /*, fail_if_not_detected: 1*/);
+                        setup_unzip_on_fd(fd, /*fail_if_not_detected:*/ 1);
                 }
         }
 
         return fd;
-#endif
 }
 
+#endif /* SEAMLESS_COMPRESSION */
+
 void* FAST_FUNC xmalloc_open_zipped_read_close(const char *fname, size_t *maxsz_p)
 {
         int fd;