factor: factor2 variable is unused now, drop it

[oweals/busybox.git] / docs / keep_data_small.txt

diff --git a/docs/keep_data_small.txt b/docs/keep_data_small.txt
index 55f4fc95a6816a7c728be54042d004b6c4165027..218d4f2eeb6bbae9019713dfc667f8240bc42c3d 100644 (file)
--- a/docs/keep_data_small.txt
+++ b/docs/keep_data_small.txt
@@ -43,13 +43,23 @@ takes 55k of memory on 64-bit x86 kernel.
 
 On 32-bit kernel we need ~26k per applet.
 
 
 On 32-bit kernel we need ~26k per applet.
 

+Script:
+
+i=1000; while test $i != 0; do
+        echo -n .
+        busybox sleep 30 &
+        i=$((i - 1))
+done
+echo
+wait
+

 (Data from NOMMU arches are sought. Provide 'size busybox' output too)
 
 
                Example 1
 
 One example how to reduce global data usage is in
 (Data from NOMMU arches are sought. Provide 'size busybox' output too)
 
 
                Example 1
 
 One example how to reduce global data usage is in

-archival/libunarchive/decompress_unzip.c:
+archival/libarchive/decompress_gunzip.c:

 
 /* This is somewhat complex-looking arrangement, but it allows
  * to place decompressor state either in bss or in
 
 /* This is somewhat complex-looking arrangement, but it allows
  * to place decompressor state either in bss or in


@@ -65,7 +75,7 @@ archival/libunarchive/decompress_unzip.c:
 (see the rest of the file to get the idea)
 
 This example completely eliminates globals in that module.
 (see the rest of the file to get the idea)
 
 This example completely eliminates globals in that module.

-Required memory is allocated in inflate_gunzip() [its main module]
+Required memory is allocated in unpack_gz_stream() [its main module]

 and then passed down to all subroutines which need to access 'globals'
 as a parameter.
 
 and then passed down to all subroutines which need to access 'globals'
 as a parameter.
 


@@ -77,7 +87,7 @@ take a look at archival/gzip.c. Here all global data is replaced by
 single global pointer (ptr_to_globals) to allocated storage.
 
 In order to not duplicate ptr_to_globals in every applet, you can
 single 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
+reuse single common one. It is defined in libbb/ptr_to_globals.c

 as struct globals *const ptr_to_globals, but the struct globals is
 NOT defined in libbb.h. You first define your own struct:
 
 as struct globals *const ptr_to_globals, but the struct globals is
 NOT defined in libbb.h. You first define your own struct:
 


@@ -89,11 +99,19 @@ and then declare that ptr_to_globals is a pointer to it:
 
 ptr_to_globals is declared as constant pointer.
 This helps gcc understand that it won't change, resulting in noticeably
 
 ptr_to_globals is declared as constant pointer.
 This helps gcc understand that it won't change, resulting in noticeably

-smaller code. In order to assign it, use PTR_TO_GLOBALS macro:
+smaller code. In order to assign it, use SET_PTR_TO_GLOBALS macro:

 
 

-       PTR_TO_GLOBALS = xzalloc(sizeof(G));
+       SET_PTR_TO_GLOBALS(xzalloc(sizeof(G)));

 
 

-Typically it is done in <applet>_main().
+Typically it is done in <applet>_main(). Another variation is
+to use stack:
+
+int <applet>_main(...)
+{
+#undef G
+       struct globals G;
+       memset(&G, 0, sizeof(G));
+       SET_PTR_TO_GLOBALS(&G);

 
 Now you can reference "globals" by G.a, G.buf and so on, in any function.
 
 
 Now you can reference "globals" by G.a, G.buf and so on, in any function.
 


@@ -128,11 +146,9 @@ less readable, use #defines:
 #define sector (G.sector)
 
 
 #define sector (G.sector)
 
 

-               Word of caution
+               Finding non-shared duplicated strings

 
 

-If applet doesn't use much of global data, converting it to use
-one of above methods is not worth the resulting code obfuscation.
-If you have less than ~300 bytes of global data - don't bother.
+strings busybox | sort | uniq -c | sort -nr

 
 
                gcc's data alignment problem
 
 
                gcc's data alignment problem


@@ -204,3 +220,46 @@ Result (non-static busybox built against glibc):
    text    data     bss     dec     hex filename
  634416    2736   23856  661008   a1610 busybox
  632580    2672   22944  658196   a0b14 busybox_noalign
    text    data     bss     dec     hex filename
  634416    2736   23856  661008   a1610 busybox
  632580    2672   22944  658196   a0b14 busybox_noalign

+
+
+
+               Keeping code small
+
+Use scripts/bloat-o-meter to check whether introduced changes
+didn't generate unnecessary bloat. This script needs unstripped binaries
+to generate a detailed report. To automate this, just use
+"make bloatcheck". It requires busybox_old binary to be present,
+use "make baseline" to generate it from unmodified source, or
+copy busybox_unstripped to busybox_old before modifying sources
+and rebuilding.
+
+Set CONFIG_EXTRA_CFLAGS="-fno-inline-functions-called-once",
+produce "make bloatcheck", see the biggest auto-inlined functions.
+Now, set CONFIG_EXTRA_CFLAGS back to "", but add NOINLINE
+to some of these functions. In 1.16.x timeframe, the results were
+(annotated "make bloatcheck" output):
+
+function             old     new   delta
+expand_vars_to_list    -    1712   +1712 win
+lzo1x_optimize         -    1429   +1429 win
+arith_apply            -    1326   +1326 win
+read_interfaces        -    1163   +1163 loss, leave w/o NOINLINE
+logdir_open            -    1148   +1148 win
+check_deps             -    1148   +1148 loss
+rewrite                -    1039   +1039 win
+run_pipe             358    1396   +1038 win
+write_status_file      -    1029   +1029 almost the same, leave w/o NOINLINE
+dump_identity          -     987    +987 win
+mainQSort3             -     921    +921 win
+parse_one_line         -     916    +916 loss
+summarize              -     897    +897 almost the same
+do_shm                 -     884    +884 win
+cpio_o                 -     863    +863 win
+subCommand             -     841    +841 loss
+receive                -     834    +834 loss
+
+855 bytes saved in total.
+
+scripts/mkdiff_obj_bloat may be useful to automate this process: run
+"scripts/mkdiff_obj_bloat NORMALLY_BUILT_TREE FORCED_NOINLINE_TREE"
+and select modules which shrank.