gzip: use common bbunzip infrastructure - ~700 bytes code less

4 files changed, 185 insertions, 99 deletions

diff --git a/archival/Kbuild b/archival/Kbuild
index 011feee5f..07b442f15 100644
--- a/archival/Kbuild
+++ b/archival/Kbuild
@@ -14,7 +14,7 @@ lib-$(CONFIG_CPIO)		+= cpio.o
 lib-$(CONFIG_DPKG)              += dpkg.o
 lib-$(CONFIG_DPKG_DEB)          += dpkg_deb.o
 lib-$(CONFIG_GUNZIP)            += bbunzip.o
-lib-$(CONFIG_GZIP)              += gzip.o
+lib-$(CONFIG_GZIP)              += gzip.o bbunzip.o
 lib-$(CONFIG_RPM2CPIO)          += rpm2cpio.o
 lib-$(CONFIG_RPM)               += rpm.o
 lib-$(CONFIG_TAR)               += tar.o
diff --git a/archival/bbunzip.c b/archival/bbunzip.c
index f7c861256..e16e6b083 100644
--- a/archival/bbunzip.c
+++ b/archival/bbunzip.c
@@ -9,12 +9,12 @@
 #include "unarchive.h"
 
 enum {
-        OPT_STDOUT = 1,
-        OPT_FORCE = 2,
+        OPT_STDOUT = 0x1,
+        OPT_FORCE = 0x2,
 /* gunzip only: */
-        OPT_TEST = 4,
-        OPT_DECOMPRESS = 8,
-        OPT_VERBOSE = 0x10,
+        OPT_VERBOSE = 0x4,
+        OPT_DECOMPRESS = 0x8,
+        OPT_TEST = 0x10,
 };
 
 static
@@ -33,8 +33,7 @@ int open_to_or_warn(int to_fd, const char *filename, int flags, int mode)
         return 0;
 }
 
-static
-int unpack(char **argv,
+int bbunpack(char **argv,
         char* (*make_new_name)(char *filename),
         USE_DESKTOP(long long) int (*unpacker)(void)
 )
@@ -173,7 +172,7 @@ int bunzip2_main(int argc, char **argv)
         if (applet_name[2] == 'c')
                 option_mask32 |= OPT_STDOUT;
 
-        return unpack(argv, make_new_name_bunzip2, unpack_bunzip2);
+        return bbunpack(argv, make_new_name_bunzip2, unpack_bunzip2);
 }
 
 #endif
@@ -267,13 +266,13 @@ USE_DESKTOP(long long) int unpack_gunzip(void)
 int gunzip_main(int argc, char **argv);
 int gunzip_main(int argc, char **argv)
 {
-        getopt32(argc, argv, "cftdv");
+        getopt32(argc, argv, "cfvdt");
         argv += optind;
         /* if called as zcat */
         if (applet_name[1] == 'c')
                 option_mask32 |= OPT_STDOUT;
 
-        return unpack(argv, make_new_name_gunzip, unpack_gunzip);
+        return bbunpack(argv, make_new_name_gunzip, unpack_gunzip);
 }
 
 #endif
@@ -311,7 +310,7 @@ int unlzma_main(int argc, char **argv)
         if (applet_name[4] == 'c')
                 option_mask32 |= OPT_STDOUT;
 
-        return unpack(argv, make_new_name_unlzma, unpack_unlzma);
+        return bbunpack(argv, make_new_name_unlzma, unpack_unlzma);
 }
 
 #endif
@@ -350,7 +349,7 @@ int uncompress_main(int argc, char **argv)
         getopt32(argc, argv, "cf");
         argv += optind;
 
-        return unpack(argv, make_new_name_uncompress, unpack_uncompress);
+        return bbunpack(argv, make_new_name_uncompress, unpack_uncompress);
 }
 
 #endif
diff --git a/archival/gzip.c b/archival/gzip.c
index c8444ac25..90075272d 100644
--- a/archival/gzip.c
+++ b/archival/gzip.c
@@ -2074,6 +2074,7 @@ static void zip(int in, int out)
 
 
 /* ======================================================================== */
+#if 0
 static void abort_gzip(int ATTRIBUTE_UNUSED ignored)
 {
         exit(1);
@@ -2082,92 +2083,6 @@ static void abort_gzip(int ATTRIBUTE_UNUSED ignored)
 int gzip_main(int argc, char **argv);
 int gzip_main(int argc, char **argv)
 {
-        enum {
-                OPT_tostdout = 0x1,
-                OPT_force = 0x2,
-        };
-
-        unsigned opt;
-        int inFileNum;
-        int outFileNum;
-        int i;
-        struct stat statBuf;
-
-        opt = getopt32(argc, argv, "cf123456789qv" USE_GUNZIP("d"));
-        //if (opt & 0x1) // -c
-        //if (opt & 0x2) // -f
-        /* Ignore 1-9 (compression level) options */
-        //if (opt & 0x4) // -1
-        //if (opt & 0x8) // -2
-        //if (opt & 0x10) // -3
-        //if (opt & 0x20) // -4
-        //if (opt & 0x40) // -5
-        //if (opt & 0x80) // -6
-        //if (opt & 0x100) // -7
-        //if (opt & 0x200) // -8
-        //if (opt & 0x400) // -9
-        //if (opt & 0x800) // -q
-        //if (opt & 0x1000) // -v
-#if ENABLE_GUNZIP /* gunzip_main may not be visible... */
-        if (opt & 0x2000) { // -d
-                /* FIXME: getopt32 should not depend on optind */
-                optind = 1;
-                return gunzip_main(argc, argv);
-        }
-#endif
-
-        /* Comment?? */
-        if (signal(SIGINT, SIG_IGN) != SIG_IGN) {
-                signal(SIGINT, abort_gzip);
-        }
-#ifdef SIGTERM
-        if (signal(SIGTERM, SIG_IGN) != SIG_IGN) {
-                signal(SIGTERM, abort_gzip);
-        }
-#endif
-#ifdef SIGHUP
-        if (signal(SIGHUP, SIG_IGN) != SIG_IGN) {
-                signal(SIGHUP, abort_gzip);
-        }
-#endif
-
-        ptr_to_globals = xzalloc(sizeof(struct global1) + sizeof(struct global2));
-        ptr_to_globals++;
-        G2.l_desc.dyn_tree    = G2.dyn_ltree;
-        G2.l_desc.static_tree = G2.static_ltree;
-        G2.l_desc.extra_bits  = extra_lbits;
-        G2.l_desc.extra_base  = LITERALS + 1;
-        G2.l_desc.elems       = L_CODES;
-        G2.l_desc.max_length  = MAX_BITS;
-        //G2.l_desc.max_code    = 0;
-
-        G2.d_desc.dyn_tree    = G2.dyn_dtree;
-        G2.d_desc.static_tree = G2.static_dtree;
-        G2.d_desc.extra_bits  = extra_dbits;
-        //G2.d_desc.extra_base  = 0;
-        G2.d_desc.elems       = D_CODES;
-        G2.d_desc.max_length  = MAX_BITS;
-        //G2.d_desc.max_code    = 0;
-
-        G2.bl_desc.dyn_tree    = G2.bl_tree;
-        //G2.bl_desc.static_tree = NULL;
-        G2.bl_desc.extra_bits  = extra_blbits,
-        //G2.bl_desc.extra_base  = 0;
-        G2.bl_desc.elems       = BL_CODES;
-        G2.bl_desc.max_length  = MAX_BL_BITS;
-        //G2.bl_desc.max_code    = 0;
-
-        /* Allocate all global buffers (for DYN_ALLOC option) */
-        ALLOC(uch, G1.l_buf, INBUFSIZ);
-        ALLOC(uch, G1.outbuf, OUTBUFSIZ);
-        ALLOC(ush, G1.d_buf, DIST_BUFSIZE);
-        ALLOC(uch, G1.window, 2L * WSIZE);
-        ALLOC(ush, G1.prev, 1L << BITS);
-
-        /* Initialise the CRC32 table */
-        G1.crc_32_tab = crc32_filltable(0);
-
-        clear_bufs();
 
         if (optind == argc) {
                 G1.time_stamp = 0;
@@ -2240,3 +2155,87 @@ int gzip_main(int argc, char **argv)
 
         return 0; //##G1.exit_code;
 }
+#endif
+
+int bbunpack(char **argv,
+        char* (*make_new_name)(char *filename),
+        USE_DESKTOP(long long) int (*unpacker)(void)
+);
+
+static
+char* make_new_name_gzip(char *filename)
+{
+        return xasprintf("%s.gz", filename);
+}
+
+static
+USE_DESKTOP(long long) int pack_gzip(void)
+{
+        struct stat s;
+        
+        G1.time_stamp = 0;
+        if (!fstat(STDIN_FILENO, &s))
+                G1.time_stamp = s.st_ctime;
+        zip(STDIN_FILENO, STDOUT_FILENO);
+        return 0;
+}
+
+int gzip_main(int argc, char **argv);
+int gzip_main(int argc, char **argv)
+{
+        unsigned opt;
+
+        /* Must match bbunzip's constants OPT_STDOUT, OPT_FORCE! */
+        opt = getopt32(argc, argv, "cfv" USE_GUNZIP("d") "q123456789" );
+        option_mask32 &= 0x7; /* Clear -d, ignore -q, -0..9 */
+        //if (opt & 0x1) // -c
+        //if (opt & 0x2) // -f
+        //if (opt & 0x4) // -v
+#if ENABLE_GUNZIP /* gunzip_main may not be visible... */
+        if (opt & 0x8) { // -d
+                /* FIXME: getopt32 should not depend on optind */
+                optind = 1;
+                return gunzip_main(argc, argv);
+        }
+#endif
+
+        ptr_to_globals = xzalloc(sizeof(struct global1) + sizeof(struct global2));
+        ptr_to_globals++;
+        G2.l_desc.dyn_tree    = G2.dyn_ltree;
+        G2.l_desc.static_tree = G2.static_ltree;
+        G2.l_desc.extra_bits  = extra_lbits;
+        G2.l_desc.extra_base  = LITERALS + 1;
+        G2.l_desc.elems       = L_CODES;
+        G2.l_desc.max_length  = MAX_BITS;
+        //G2.l_desc.max_code    = 0;
+
+        G2.d_desc.dyn_tree    = G2.dyn_dtree;
+        G2.d_desc.static_tree = G2.static_dtree;
+        G2.d_desc.extra_bits  = extra_dbits;
+        //G2.d_desc.extra_base  = 0;
+        G2.d_desc.elems       = D_CODES;
+        G2.d_desc.max_length  = MAX_BITS;
+        //G2.d_desc.max_code    = 0;
+
+        G2.bl_desc.dyn_tree    = G2.bl_tree;
+        //G2.bl_desc.static_tree = NULL;
+        G2.bl_desc.extra_bits  = extra_blbits,
+        //G2.bl_desc.extra_base  = 0;
+        G2.bl_desc.elems       = BL_CODES;
+        G2.bl_desc.max_length  = MAX_BL_BITS;
+        //G2.bl_desc.max_code    = 0;
+
+        /* Allocate all global buffers (for DYN_ALLOC option) */
+        ALLOC(uch, G1.l_buf, INBUFSIZ);
+        ALLOC(uch, G1.outbuf, OUTBUFSIZ);
+        ALLOC(ush, G1.d_buf, DIST_BUFSIZE);
+        ALLOC(uch, G1.window, 2L * WSIZE);
+        ALLOC(ush, G1.prev, 1L << BITS);
+
+        /* Initialise the CRC32 table */
+        G1.crc_32_tab = crc32_filltable(0);
+
+        clear_bufs();
+
+        return bbunpack(argv, make_new_name_gzip, pack_gzip);
+}
diff --git a/docs/keep_data_small.txt b/docs/keep_data_small.txt
new file mode 100644
index 000000000..88cc2bc66
--- /dev/null
+++ b/docs/keep_data_small.txt
@@ -0,0 +1,88 @@
+        Keeping data small
+
+When many applets are compiled into busybox, all rw data and
+bss for each applet are concatenated. Including those from libc,
+if static bbox is built. When bbox is started, _all_ this data
+is allocated, not just that one part for selected applet.
+
+What "allocated" exactly means, depends on arch.
+On nommu it's probably bites the most, actually using real
+RAM for rwdata and bss. On i386, bss is lazily allocated
+by COWed zero pages. Not sure about rwdata - also COW?
+
+Small experiment measures "parasitic" bbox memory consumption.
+Here we start 1000 "busybox sleep 10" in parallel.
+bbox binary is practically allyesconfig static one,
+built against uclibc:
+
+bash-3.2# nmeter '%t %c %b %m %p %[pn]'
+23:17:28 ..........  0  0 168M  0 147
+23:17:29 ..........  0  0 168M  0 147
+23:17:30 U.........  0  0 168M  1 147
+23:17:31 SU........  0 188k 181M 244 391
+23:17:32 SSSSUUU...  0  0 223M 757 1147
+23:17:33 UUU.......  0  0 223M  0 1147
+23:17:34 U.........  0  0 223M  1 1147
+23:17:35 ..........  0  0 223M  0 1147
+23:17:36 ..........  0  0 223M  0 1147
+23:17:37 S.........  0  0 223M  0 1147
+23:17:38 ..........  0  0 223M  1 1147
+23:17:39 ..........  0  0 223M  0 1147
+23:17:40 ..........  0  0 223M  0 1147
+23:17:41 ..........  0  0 210M  0 906
+23:17:42 ..........  0  0 168M  1 147
+23:17:43 ..........  0  0 168M  0 147
+
+This requires 55M of memory. Thus 1 trivial busybox applet
+takes 55k of userspace memory (nmeter doesn't account for kernel-side
+allocations). Definitely can be improved.
+
+Thus we should avoid large global data in our applets,
+and should minimize usage of libc functions which implicitly use
+such structures in libc.
+
+        Example 1
+
+One example how to reduce global data usage is in
+archival/libunarchive/decompress_unzip.c:
+
+/* This is somewhat complex-looking arrangement, but it allows
+ * to place decompressor state either in bss or in
+ * malloc'ed space simply by changing #defines below.
+ * Sizes on i386:
+ * text    data     bss     dec     hex
+ * 5256       0     108    5364    14f4 - bss
+ * 4915       0       0    4915    1333 - malloc
+ */
+#define STATE_IN_BSS 0
+#define STATE_IN_MALLOC 1
+
+This example completely eliminates globals in that module.
+Required memory is allocated in inflate_gunzip() [its main module]
+and then passed down to all subroutines which need to access globals
+as a parameter.
+
+        Example 2
+
+In case you don't want to pass this additional parameter everywhere,
+take a look at archival/gzip.c. Here all global data is replaced by
+singe global pointer (ptr_to_globals) to allocated storage.
+
+In order to not duplicate ptr_to_globals in every applet, you can
+reuse single common one. It is defined in libbb/messages.c
+as void *ptr_to_globals, but is NOT declared in libbb.h.
+You first define a struct:
+
+struct my_globals { int a; char buf[1000]; };
+
+and then declare that ptr_to_globals is a pointer to it:
+
+extern struct my_globals *ptr_to_globals;
+#define G (*ptr_to_globals)
+
+Linker magic enures that these two merge into single pointer object.
+Now initialize it in <applet>_main():
+
+        ptr_to_globals = xzalloc(sizeof(G));
+
+and you can reference "globals" by G.a, G.buf and so on, in any function.