large binaries
Abraham Egnor
hello.hs:
module Main(main) where
main = putStr "Hello world!\n"
hello.c:
#include <stdio.h>
int main(void)
{
printf("Hello world!\n");
return 0;
}
[abe@shiva:~/src/test] $ ghc hello.hs -o hello_hs
[abe@shiva:~/src/test] $ gcc hello.c -o hello_c
[abe@shiva:~/src/test] $ ls -l hello_*
-rwxr-xr-x 1 abe engy 13712 Jul 18 11:34 hello_c
-rwxr-xr-x 1 abe engy 299900 Jul 18 11:33 hello_hs
Why is the binary made by ghc 20 times bigger than the one made by gcc?
Abe
_______________________________________________
Haskell-Cafe mailing list
Haskel...@haskell.org
http://www.haskell.org/mailman/listinfo/haskell-cafe
Jon Cast
> This is something I just noticed...
> hello.hs:
> module Main(main) where
> main = putStr "Hello world!\n"
> hello.c:
> #include <stdio.h>
> int main(void)
> {
> printf("Hello world!\n");
> return 0;
> }
> [abe@shiva:~/src/test] $ ghc hello.hs -o hello_hs
> [abe@shiva:~/src/test] $ gcc hello.c -o hello_c
> [abe@shiva:~/src/test] $ ls -l hello_*
> -rwxr-xr-x 1 abe engy 13712 Jul 18 11:34 hello_c
> -rwxr-xr-x 1 abe engy 299900 Jul 18 11:33 hello_hs
>
> Why is the binary made by ghc 20 times bigger than the one made by gcc?
I don't know for certain, but I've got a couple of guesses:
1. hello_hs is probably statically linked. hello_c is probably
dynamically linked.
2. "Hello world!\n" in Haskell is boxed; in C it's un-boxed. Ditto
for putStr vs. printf. The compiled code for the Haskell main appears
to have a lot of code to deal with those complications; multiply that
by however many functions are called by putStr.
> Abe
Jon Cast
Andreas Wollschlaeger
> Abraham Egnor <aeg...@antioch-college.edu> wrote:
> > This is something I just noticed...
> >
> > hello.hs:
> > module Main(main) where
> >
> > main = putStr "Hello world!\n"
> >
> > hello.c:
> > #include <stdio.h>
> >
> > int main(void)
> > {
> > printf("Hello world!\n");
> > return 0;
> > }
> >
> > [abe@shiva:~/src/test] $ ghc hello.hs -o hello_hs
> > [abe@shiva:~/src/test] $ gcc hello.c -o hello_c
> > [abe@shiva:~/src/test] $ ls -l hello_*
> > -rwxr-xr-x 1 abe engy 13712 Jul 18 11:34 hello_c
> > -rwxr-xr-x 1 abe engy 299900 Jul 18 11:33 hello_hs
> >
> > Why is the binary made by ghc 20 times bigger than the one made by gcc?
>
> I don't know for certain, but I've got a couple of guesses:
>
> 1. hello_hs is probably statically linked. hello_c is probably
> dynamically linked.
>
Right. The ldd command gives the dependencies (on Linux):
awo@asterix:~> ldd hello
libc.so.6 => /lib/libc.so.6 (0x40022000)
/lib/ld-linux.so.2 => /lib/ld-linux.so.2 (0x40000000)
If you compile hello.c as static binary, like
gcc -static -o hello hello.c
then the resulting exe is rather big, too ;-)
awo@asterix:~> ls -l hello
-rwxr-xr-x 1 awo users 1486299 Jul 18 21:40 hello
So ghc isnt so bad, really :-)
Greetings
Andreas
Abraham Egnor
> > dynamically linked.
> >
>
> Right. The ldd command gives the dependencies (on Linux):
>
> awo@asterix:~> ldd hello
> libc.so.6 => /lib/libc.so.6 (0x40022000)
> /lib/ld-linux.so.2 => /lib/ld-linux.so.2 (0x40000000)
>
> If you compile hello.c as static binary, like
>
> gcc -static -o hello hello.c
>
> then the resulting exe is rather big, too ;-)
>
> awo@asterix:~> ls -l hello
> -rwxr-xr-x 1 awo users 1486299 Jul 18 21:40 hello
>
> So ghc isnt so bad, really :-)
Okay, that makes sense.
Is there some reason haskell binaries have to be statically linked? I
can't seem to find a way to make them otherwise, at least with ghc.
Abe
Alastair Reid
> Is there some reason haskell binaries have to be statically linked?
> I can't seem to find a way to make them otherwise, at least with
> ghc.
The ELF dynamic linking format seems to be designed on the assumption
of poor code reuse (e.g., C code) where calls from one module to
another are rare and can therefore be expensive.
Haskell code has very high levels of reuse and calls from one module
to another (especially calls into Prelude, List, Monad, etc) are very
common so dynamic linking imposes a very high overhead.
This isn't a reason to not support dynamic linking but it's a reason
to make it a low priority. Incidentally, it avoids confusion between
the performance overhead of lazy evaluation and the performance
overhead of dynamic linking of modular code.
--
Alastair Reid alas...@reid-consulting-uk.ltd.uk
Reid Consulting (UK) Limited http://www.reid-consulting-uk.ltd.uk/alastair/
Simon Marlow
> > I can't seem to find a way to make them otherwise, at least with
> > ghc.
>
> The ELF dynamic linking format seems to be designed on the assumption
> of poor code reuse (e.g., C code) where calls from one module to
> another are rare and can therefore be expensive.
>
> Haskell code has very high levels of reuse and calls from one module
> to another (especially calls into Prelude, List, Monad, etc) are very
> common so dynamic linking imposes a very high overhead.
>
> This isn't a reason to not support dynamic linking but it's a reason
> to make it a low priority. Incidentally, it avoids confusion between
> the performance overhead of lazy evaluation and the performance
> overhead of dynamic linking of modular code.
Yes, and there are various other reasons too. There's a FAQ question in
GHC's (rather well hidden) FAQ:
http://www.haskell.org/ghc/docs/latest/html/users_guide/faq.html
and a search through the glasgow-haskell-users archives will turn up
previous discussions.
Also, we do (or did) have support for dynamic libraries on Windows.
Cheers,
Simon
Malcolm Wallace
It would not be entirely fair to lay all the blame for large Haskell
binaries entirely at the door of static vs. dynamic linking.
After all, the Haskell version is dynamically linked against exactly
the same shared libraries as the C version, at least on my machine:
ldd Hello (Hello.hs)
libm.so.6 => /lib/libm.so.6 (0x40022000)
libc.so.6 => /lib/libc.so.6 (0x40044000)
/lib/ld-linux.so.2 => /lib/ld-linux.so.2 (0x40000000)
Of course, it is static linking against the *Haskell* runtime system,
Prelude and Libraries that is the cause of binary bloat. Quite simply,
lots of extra stuff is dragged in that isn't visible in the apparently
simple source program. For instance, I can find all the following symbols
in the binary for "hello world" (compiled with nhc98):
putStr, shows, showChar, showParen, showString, fromCString,
toCString, hGetFileName, hPutChar, hPutStr, error, flip, id, init,
length, not, putChar, putStrLn, seq, show, subtract, exitWith,
instance Bounded Int (maxBound, minBound), instance Enum Ordering
(succ, pred, toEnum, fromEnum, enumFrom, enumFromThen, enumFromTo,
enumFromThenTo), instance Enum ErrNo (succ, pred, toEnum,
fromEnum, enumFrom, enumFromThen, enumFromTo, enumFromThenTo),
instance Monad IO (>>=, >>, return, fail), instance Eq ErrNo (==,
/=), instance Eq Int (==, /=), instance Eq Ordering (==, /=),
instance Num Int (+, -, *, negate, abs, signum, fromInteger),
instance Ord Int (compare, <, <=, >=, >, max, min), instance Show
ErrNo (show, showsPrec, showList), instance Show IOError (show,
showsPrec, showList), instance Show Int (show, showsPrec, showList)
This is not the fault of any particular implementation - the ghc-built
binary has a similar collection - rather it is dictated by the nature
of the language and its standard libraries. Because Prelude functions
are small and re-usable, they do get used all over the place in the
implementation of other parts of the Prelude, so you end up with a
huge dependency graph hiding underneath the simplest of calls.
In fact, most of the extra stuff in "Hello World" is there purely to
handle all possible error conditions in the I/O monad. Several years
ago, Colin Runciman and I did the experiment of removing all the nice
error-handling stuff from the prelude (and eliminating a few classes
too I think), to see just how small we could squash "Hello World".
The idea was to target embedded systems where memory is a scarce
resource, and fancy error-reporting is pointless (a single red LED
would do). IIRC, we managed to achieve a size of 25kb, compiled
with nhc98, which don't forget includes a bytecode interpreter in
the runtime system.
Regards,
Malcolm
Alastair Reid
Malcolm Wallace <Malcolm...@cs.york.ac.uk>:
> For instance, I can find all the following symbols in the binary for
> "hello world" (compiled with nhc98):
> putStr, shows, showChar, showParen, showString, fromCString,
> toCString, hGetFileName, hPutChar, hPutStr, error, flip, id, init,
> length, not, putChar, putStrLn, seq, show, subtract, exitWith,
> instance Bounded Int (maxBound, minBound), instance Enum Ordering
> (succ, pred, toEnum, fromEnum, enumFrom, enumFromThen, enumFromTo,
> enumFromThenTo), instance Enum ErrNo (succ, pred, toEnum, fromEnum,
> enumFrom, enumFromThen, enumFromTo, enumFromThenTo), instance Monad
> IO (>>=, >>, return, fail), instance Eq ErrNo (==, /=), instance Eq
> Int (==, /=), instance Eq Ordering (==, /=), instance Num Int (+, -,
> *, negate, abs, signum, fromInteger), instance Ord Int (compare, <,
> <=, >=, >, max, min), instance Show ErrNo (show, showsPrec,
> showList), instance Show IOError (show, showsPrec, showList),
> instance Show Int (show, showsPrec, showList)
> [...]
> In fact, most of the extra stuff in "Hello World" is there purely to
> handle all possible error conditions in the I/O monad.
How many of those class methods could actually be executed and how
many are included just because the dictionary that contains them is
included? And how many of those dictionaries are included just
because they are reachable from the dictionary of a subclass?
Some compilers for C++ and Java are claimed to discard member
functions which cannot possibly be executed even if the corresponding
vtable is kept. The simplest (least effective) way of doing this is
to discard all enumFromThenTo methods if none of the live code refers
to enumFromThenTo. More effective is to discard the Int instance of
enumFromThenTo if all enumFromThenTo references can be shown to
involve other types.
--
Alastair Reid alas...@reid-consulting-uk.ltd.uk
Reid Consulting (UK) Limited http://www.reid-consulting-uk.ltd.uk/alastair/
Jon Cast
> > > > Is there some reason haskell binaries have to be statically
> > > > linked?
> It would not be entirely fair to lay all the blame for large Haskell
> binaries entirely at the door of static vs. dynamic linking.
Well, considering that compiling the C binary statically linked
produces an even bigger executable:
$ gcc -static hello.c -o hello_c
$ ls -l hello_c hello_hs
-rwxrwxr-x 1 jcast jcast 441624 Jul 23 20:56 hello_c
-rwxrwxr-x 1 jcast jcast 157028 Jul 18 14:08 hello_hs
I think dynamic linking is fair game :)
Well, look at the symbols I find in the statically linked C hello
world:
_Exit _GLOBAL_OFFSET_TABLE_ _IO_2_1_stderr_ _IO_2_1_stdin_
_IO_2_1_stdout_ _IO_adjust_column _IO_adjust_wcolumn _IO_cleanup
_IO_default_doallocate _IO_default_finish _IO_default_imbue
_IO_default_pbackfail _IO_default_read _IO_default_seek
_IO_default_seekoff _IO_default_seekpos _IO_default_setbuf
_IO_default_showmanyc _IO_default_stat _IO_default_sync
_IO_default_uflow _IO_default_underflow _IO_default_write
_IO_default_xsgetn _IO_default_xsputn _IO_do_write _IO_doallocbuf
_IO_fclose _IO_file_attach _IO_file_close _IO_file_close_it
_IO_file_doallocate _IO_file_finish _IO_file_fopen _IO_file_init
_IO_file_jumps _IO_file_open _IO_file_overflow _IO_file_read
_IO_file_seek _IO_file_seekoff _IO_file_setbuf _IO_file_stat
_IO_file_sync _IO_file_underflow _IO_file_write _IO_file_xsgetn
_IO_file_xsputn _IO_flockfile _IO_flush_all _IO_flush_all_linebuffered
_IO_flush_all_lockp _IO_fopen _IO_fprintf _IO_free_backup_area
_IO_free_wbackup_area _IO_ftrylockfile _IO_funlockfile _IO_fwide
_IO_getdelim _IO_getline _IO_getline_info _IO_helper_jumps
_IO_helper_overflow _IO_init _IO_init_marker _IO_init_wmarker
_IO_iter_begin _IO_iter_end _IO_iter_file _IO_iter_next
_IO_least_marker _IO_least_wmarker _IO_link_in _IO_list_all
_IO_list_all_stamp _IO_list_lock _IO_list_resetlock _IO_list_unlock
_IO_marker_delta _IO_marker_difference _IO_new_do_write _IO_new_fclose
_IO_new_file_attach _IO_new_file_close_it _IO_new_file_finish
_IO_new_file_fopen _IO_new_file_init _IO_new_file_overflow
_IO_new_file_seekoff _IO_new_file_setbuf _IO_new_file_sync
_IO_new_file_underflow _IO_new_file_write _IO_new_file_xsputn
_IO_new_fopen _IO_no_init _IO_padn _IO_printf _IO_remove_marker
_IO_seekmark _IO_seekoff _IO_seekwmark _IO_setb _IO_sgetn
_IO_sputbackc _IO_sputbackwc _IO_sscanf _IO_stderr _IO_stdfile_0_lock
_IO_stdfile_1_lock _IO_stdfile_2_lock _IO_stdin _IO_stdin_used
_IO_stdout _IO_str_count _IO_str_finish _IO_str_init_readonly
_IO_str_init_static _IO_str_jumps _IO_str_overflow _IO_str_pbackfail
_IO_str_seekoff _IO_str_underflow _IO_sungetc _IO_sungetwc
_IO_switch_to_backup_area _IO_switch_to_get_mode
_IO_switch_to_main_get_area _IO_switch_to_main_wget_area
_IO_switch_to_wbackup_area _IO_switch_to_wget_mode _IO_un_link
_IO_unsave_markers _IO_unsave_wmarkers _IO_vfprintf _IO_vfscanf
_IO_vsscanf _IO_wdefault_doallocate _IO_wdefault_finish
_IO_wdefault_pbackfail _IO_wdefault_setbuf _IO_wdefault_uflow
_IO_wdefault_xsgetn _IO_wdefault_xsputn _IO_wdo_write _IO_wdoallocbuf
_IO_wfile_doallocate _IO_wfile_jumps _IO_wfile_overflow
_IO_wfile_seekoff _IO_wfile_setbuf _IO_wfile_sync _IO_wfile_underflow
_IO_wfile_xsputn _IO_wide_data_0 _IO_wide_data_1 _IO_wide_data_2
_IO_wmarker_delta _IO_wpadn _IO_wsetb __CTOR_END__ __CTOR_LIST__
__DTOR_END__ __DTOR_LIST__ __EH_FRAME_BEGIN__ __FRAME_BEGIN__
__FRAME_BEGIN__ __FRAME_BEGIN__ __FRAME_BEGIN__ __FRAME_BEGIN__
__FRAME_BEGIN__ __FRAME_BEGIN__ __FRAME_BEGIN__ __FRAME_END__
___brk_addr ___fxstat64 ___xstat64 __access __add_to_environ
__addmntent __after_morecore_hook __argz_add_sep __argz_count
__argz_create_sep __argz_stringify __atomic_writev_replacement __brk
__bss_start __calloc __cfree __clearenv __close __clz_tab __clz_tab
__ctype32_b __ctype32_tolower __ctype32_toupper __ctype32_wctrans
__ctype32_wctype __ctype32_width __ctype_b __ctype_tolower
__ctype_toupper __curbrk __cxa_atexit __data_start __daylight
__dcgettext __dcigettext __default_morecore __deregister_frame_info
__do_global_ctors_aux __do_global_dtors_aux __dso_handle
__elf_set___libc_atexit_element__cleanup__
__elf_set___libc_subfreeres_element_free_mem__
__elf_set___libc_subfreeres_element_free_mem__
__elf_set___libc_subfreeres_element_free_mem__
__elf_set___libc_subfreeres_element_free_mem__
__elf_set___libc_subfreeres_element_free_mem__
__elf_set___libc_subfreeres_element_free_mem__
__elf_set___libc_subfreeres_element_free_mem__
__elf_set___libc_subfreeres_element_free_mem__
__elf_set___libc_subfreeres_element_free_mem__
__elf_set___libc_subfreeres_element_free_mem__
__elf_set___libc_subfreeres_element_free_mem__
__elf_set___libc_subfreeres_element_free_mem__
__elf_set___libc_subfreeres_element_free_mem__
__elf_set___libc_subfreeres_element_free_mem__
__elf_set___libc_subfreeres_element_free_mem__
__elf_set___libc_subfreeres_element_freeres__ __endmntent __environ
__errno_location __evoke_link_warning_llseek __exit_funcs __fcloseall
__fcntl __flockfile __fpu_control __free __free_hook __fsetlocking
__ftrylockfile __funlockfile __fxstat64 __gconv __gconv_alias_compare
__gconv_alias_db __gconv_cache __gconv_close __gconv_close_transform
__gconv_compare_alias __gconv_compare_alias_cache __gconv_find_shlib
__gconv_find_transform __gconv_get_builtin_trans __gconv_get_path
__gconv_load_cache __gconv_lookup_cache __gconv_max_path_elem_len
__gconv_modules_db __gconv_open __gconv_path_elem __gconv_path_envvar
__gconv_read_conf __gconv_release_cache __gconv_release_shlib
__gconv_release_step __gconv_transform_ascii_internal
__gconv_transform_internal_ascii __gconv_transform_internal_ucs2
__gconv_transform_internal_ucs2reverse __gconv_transform_internal_ucs4
__gconv_transform_internal_ucs4le __gconv_transform_internal_utf8
__gconv_transform_ucs2_internal __gconv_transform_ucs2reverse_internal
__gconv_transform_ucs4_internal __gconv_transform_ucs4le_internal
__gconv_transform_utf8_internal __gconv_translit_find
__gconv_transliterate __get_avphys_pages __get_nprocs
__get_nprocs_conf __get_phys_pages __getclktck __getcwd __getdelim
__getdtablesize __getegid __geteuid __getgid __getmntent_r
__getpagesize __getpid __getrlimit __gettext_extract_plural
__gettext_free_exp __gettext_germanic_plural __gettexterror
__gettextlex __gettextparse __getuid __gmon_start__ __guess_grouping
__hasmntopt __have_no_fcntl64 __have_no_new_getrlimit __have_no_stat64
__init_misc __ioctl __isatty __isinf __isinfl __isnan __isnanl __kill
__libc_argc __libc_argv __libc_calloc __libc_check_standard_fds
__libc_close __libc_dlclose __libc_dlopen __libc_dlsym
__libc_enable_secure __libc_fatal __libc_fcntl __libc_free
__libc_init_first __libc_init_secure __libc_internal_tsd_get
__libc_internal_tsd_set __libc_longjmp __libc_lseek __libc_lseek64
__libc_mallinfo __libc_malloc __libc_malloc_initialized __libc_mallopt
__libc_memalign __libc_missing_32bit_uids __libc_multiple_libcs
__libc_open __libc_open64 __libc_pagesize __libc_pvalloc __libc_read
__libc_realloc __libc_setlocale_lock __libc_sigaction
__libc_siglongjmp __libc_stack_end __libc_start_main
__libc_tsd_DL_ERROR_data __libc_tsd_MALLOC_data __libc_valloc
__libc_write __libio_codecvt __libio_translit __llseek __localtime_r
__longjmp __lseek __lseek64 __mallinfo __malloc __malloc_check_init
__malloc_get_state __malloc_hook __malloc_initialize_hook
__malloc_set_state __malloc_stats __malloc_trim __malloc_usable_size
__mallopt __mbrlen __mbrtowc __mbsnrtowcs __memalign __memalign_hook
__memchr __mempcpy __mktime_internal __mmap __mon_yday __morecore
__mpn_add_n __mpn_addmul_1 __mpn_cmp __mpn_construct_double
__mpn_construct_float __mpn_construct_long_double __mpn_divrem
__mpn_extract_double __mpn_extract_long_double __mpn_impn_mul_n
__mpn_impn_mul_n_basecase __mpn_impn_sqr_n __mpn_impn_sqr_n_basecase
__mpn_lshift __mpn_mul __mpn_mul_1 __mpn_mul_n __mpn_rshift
__mpn_sub_n __mpn_submul_1 __mprotect __mremap __munmap __new_exitfn
__new_fclose __new_fopen __new_getrlimit __new_sys_errlist
__new_sys_nerr __offtime __open __open64 __overflow __posix_memalign
__printf_arginfo_table __printf_fp __printf_fphex
__printf_function_table __profil __profile_frequency __progname
__progname_full __pthread_atfork __pthread_initialize
__pthread_initialize_minimal __pthread_mutex_destroy
__pthread_mutex_init __pthread_mutex_lock __pthread_mutex_trylock
__pthread_mutex_unlock __pthread_mutexattr_destroy
__pthread_mutexattr_init __pthread_mutexattr_settype __pthread_once
__pthread_rwlock_rdlock __pthread_rwlock_unlock __pvalloc __rawmemchr
__read __readlink __realloc __realloc_hook __register_frame_info
__register_printf_function __restore __restore_rt __sbrk __setenv
__setfpucw __setitimer __setmntent __sigaction __sigprocmask
__start___libc_atexit __start___libc_subfreeres __stop___libc_atexit
__stop___libc_subfreeres __stpcpy __strcasecmp __strcasecmp_l
__strchrnul __strdup __strerror_r __strncasecmp __strndup __strnlen
__strsep __strsep_g __strtod_internal __strtof_internal
__strtol_internal __strtold_internal __strtoll_internal
__strtoul_internal __strtoull_internal __syscall_error
__syscall_error_1 __sysconf __tcgetattr __tdelete __tdestroy __tens
__tfind __timezone __tsearch __twalk __tz_convert __tzfile_compute
__tzfile_default __tzfile_read __tzname __tzname_cur_max __tzname_max
__tzset __tzstring __ubp_memchr __udivdi3 __uflow __umoddi3 __uname
__underflow __unsetenv __use_tzfile __valloc __vfscanf __vsscanf
__wcrtomb __wcslen __wcsmbs_clone_conv __wcsmbs_gconv_fcts
__wcsmbs_last_locale __wcsmbs_load_conv __wcsmbs_named_conv __wcsnlen
__wcsrtombs __wmemcpy __wmemmove __wmempcpy __woverflow __write
__writev __wuflow __wunderflow __xstat64 _cleanup _dl_all_dirs
_dl_argv _dl_aux_init _dl_bind_not _dl_build_local_scope
_dl_cache_libcmp _dl_cache_libcmp _dl_catch_error
_dl_check_all_versions _dl_check_map_versions _dl_clktck _dl_close
_dl_correct_cache_id _dl_debug_bindings _dl_debug_fd
_dl_debug_initialize _dl_debug_mask _dl_debug_printf
_dl_debug_printf_c _dl_debug_state _dl_debug_vdprintf _dl_do_lookup
_dl_do_lookup_versioned _dl_dprintf _dl_dst_count _dl_dst_substitute
_dl_dynamic_weak _dl_get_origin _dl_global_scope
_dl_global_scope_alloc _dl_hwcap _dl_important_hwcaps
_dl_inhibit_rpath _dl_init _dl_init_all_dirs _dl_init_paths
_dl_initfirst _dl_initial_searchlist _dl_lazy _dl_load_cache_lookup
_dl_load_lock _dl_loaded _dl_lookup_symbol _dl_lookup_symbol_skip
_dl_lookup_versioned_symbol _dl_lookup_versioned_symbol_skip
_dl_main_searchlist _dl_map_object _dl_map_object_deps
_dl_map_object_from_fd _dl_mcount _dl_mcount_wrapper
_dl_mcount_wrapper_check _dl_new_object _dl_nloaded
_dl_non_dynamic_init _dl_open _dl_origin_path _dl_osversion
_dl_out_of_memory _dl_pagesize _dl_platform _dl_platformlen
_dl_profile _dl_profile_map _dl_receive_error _dl_reloc_bad_type
_dl_relocate_object _dl_rtld_map _dl_runtime_profile
_dl_runtime_resolve _dl_setup_hash _dl_signal_cerror _dl_signal_error
_dl_start _dl_start_profile _dl_starting_up _dl_sysdep_read_whole_file
_dl_unload_cache _dl_verbose _dl_x86_cap_flags _dl_x86_platforms
_edata _end _environ _errno _exit _fini _flushlbf _fp_hw
_fpioconst_pow10 _i18n_number_rewrite _init _itoa _itoa_base_table
_itoa_lower_digits _itoa_upper_digits _itowa _itowa_lower_digits
_itowa_upper_digits _libc_intl_domainname _longjmp _longjmp_unwind
_new_sys_errlist _new_sys_nerr _nl_C _nl_C_LC_ADDRESS _nl_C_LC_COLLATE
_nl_C_LC_CTYPE _nl_C_LC_CTYPE_class _nl_C_LC_CTYPE_class32
_nl_C_LC_CTYPE_class_alnum _nl_C_LC_CTYPE_class_alpha
_nl_C_LC_CTYPE_class_blank _nl_C_LC_CTYPE_class_cntrl
_nl_C_LC_CTYPE_class_digit _nl_C_LC_CTYPE_class_graph
_nl_C_LC_CTYPE_class_lower _nl_C_LC_CTYPE_class_print
_nl_C_LC_CTYPE_class_punct _nl_C_LC_CTYPE_class_space
_nl_C_LC_CTYPE_class_upper _nl_C_LC_CTYPE_class_xdigit
_nl_C_LC_CTYPE_map_tolower _nl_C_LC_CTYPE_map_toupper
_nl_C_LC_CTYPE_tolower _nl_C_LC_CTYPE_toupper _nl_C_LC_CTYPE_width
_nl_C_LC_IDENTIFICATION _nl_C_LC_MEASUREMENT _nl_C_LC_MESSAGES
_nl_C_LC_MONETARY _nl_C_LC_NAME _nl_C_LC_NUMERIC _nl_C_LC_PAPER
_nl_C_LC_TELEPHONE _nl_C_LC_TIME _nl_C_codeset _nl_C_locobj _nl_C_name
_nl_POSIX_name _nl_category_name_sizes _nl_category_names
_nl_category_num_items _nl_category_postload _nl_current
_nl_current_LC_ADDRESS _nl_current_LC_COLLATE _nl_current_LC_CTYPE
_nl_current_LC_IDENTIFICATION _nl_current_LC_MEASUREMENT
_nl_current_LC_MESSAGES _nl_current_LC_MONETARY _nl_current_LC_NAME
_nl_current_LC_NUMERIC _nl_current_LC_PAPER _nl_current_LC_TELEPHONE
_nl_current_LC_TIME _nl_current_default_domain _nl_current_names
_nl_default_default_domain _nl_default_dirname _nl_domain_bindings
_nl_expand_alias _nl_explode_name _nl_find_domain _nl_find_language
_nl_find_locale _nl_find_msg _nl_free_domain_conv _nl_get_alt_digit
_nl_get_era_entry _nl_get_walt_digit _nl_init_domain_conv
_nl_init_era_entries _nl_load_domain _nl_load_locale
_nl_loaded_domains _nl_locale_file_list _nl_make_l10nflist
_nl_msg_cat_cntr _nl_normalize_codeset _nl_parse_alt_digit
_nl_postload_ctype _nl_postload_time _nl_remove_locale
_nl_select_era_entry _nl_state_lock _nl_unload_domain
_nl_unload_locale _nl_value_type_LC_ADDRESS _nl_value_type_LC_COLLATE
_nl_value_type_LC_CTYPE _nl_value_type_LC_IDENTIFICATION
_nl_value_type_LC_MEASUREMENT _nl_value_type_LC_MESSAGES
_nl_value_type_LC_MONETARY _nl_value_type_LC_NAME
_nl_value_type_LC_NUMERIC _nl_value_type_LC_PAPER
_nl_value_type_LC_TELEPHONE _nl_value_type_LC_TIME _nl_value_types
_pthread_cleanup_pop_restore _pthread_cleanup_push_defer _quicksort
_r_debug _setjmp _start _sys_errlist _sys_nerr _tens_in_limb
_tens_in_limb _tens_in_limb _tmbuf abort access add_dependency
add_derivation add_module add_name_to_object add_to_global addmntent
alias_compare aliasfile.1 alt_digits alt_digits_initialized arena_get2
arena_key arena_mem argz_add_sep argz_count argz_create_sep
argz_stringify blanks blanks brk bsearch buf.2 buffered_vfprintf
builtin_aliases builtin_modules cache cache_malloced cache_new
cache_size cachesize call_gmon_start calloc capstr category_to_name
cfree check_action chunk_align chunk_alloc chunk_free chunk_realloc
clearenv close codeset_idx.0 collseqmb collseqwc completed.1
compute_change compute_tzname_max curwd.0 data data_start daylight
dcgettext decompose_rpath default_gconv_path default_tzdir.0
derivation_compare detect_conflict disallow_malloc_check
dl_open_worker do_always_noconv do_dlclose do_dlopen do_dlsym
do_encoding do_in do_length do_max_length do_out do_release_all
do_release_shlib do_unshift dummy_bucket.3 empty_path_elem endmntent
env_path_list environ envlock era_initialized eras errno exit
expand_dynamic_string_token expected.1 expected_note.2
extend_alias_table fclose fcloseall fcntl fgets_unlocked
find_derivation find_module find_module_idx fini_dummy fixup flockfile
flush_cleanup fopen force_to_data force_to_data fprintf frame_dummy
fread_unlocked free free_atfork free_check free_derivation free_mem
free_mem free_mem free_mem free_mem free_mem free_mem free_mem
free_mem free_mem free_mem free_mem free_mem free_mem free_mem
free_modules_db free_starter freemem.1 freemem_size.2 freeres fromidx
fromlimit froms fseek ftrylockfile funlockfile gcc2_compiled.
gcc2_compiled. gcc2_compiled. gcc2_compiled. gcc2_compiled.
gconv_conf_filename gconv_module_ext gen_steps get_avphys_pages
get_nprocs get_nprocs_conf get_phys_pages get_proc_path getcwd
getdelim getdtablesize getegid getenv geteuid getgid getmntent_r
getpagesize getpid getrlimit getuid group_number group_number gsignal
guess_category_value hack_digit.0 hashfraction hasmntopt heap_trim
increment_counter index init init_dummy init_dummy initial internal
internal_trans_names.0 ioctl is_initialized.0 isatty isinf isinfl
isnan isnanl jump_table.0 kcount kcountsize kill known_compare
known_derivations known_values last_environ leaps list_all_lock
list_lock llseek loaded locale_alias_path.0 localtime localtime_offset
localtime_r lock lock lock lock lock lock lock.0 lock.1
log_hashfraction longjmp lose lowpc lseek lseek64 main main_arena
main_trim mallinfo malloc malloc_atfork malloc_check malloc_get_state
malloc_hook_ini malloc_set_state malloc_starter malloc_stats
malloc_trim malloc_usable_size mallopt map map match_symbol
max_capstrlen max_dirnamelen max_mmapped_mem max_n_mmaps
max_sbrked_mem maxmap mbrlen mbrtowc mbsnrtowcs mem2chunk_check
memalign memalign_check memalign_hook_ini memchr memcpy memmove
mempcpy memset mktime mmap mmap_threshold mmapped_mem modcounter.0
mount_proc mprotect mremap msg.9 msort_with_tmp munmap n_mmaps
n_mmaps_max narcs narcsp nbits.0 nbits.0 nbits.0 ncapstr new_do_write
new_heap nmap not_available nsamples null num_eras num_leaps
num_transitions num_types oact.0 object.2 old_tz once open open64
open_path open_verify openaux otimer.1 p.0 pagesize.1 path_proc
pc_offset pc_scale phys_pages.0 phys_pages_info plone plural_eval
plural_lookup plvar posix_memalign print_search_path printf
printf_funcs printf_unknown profil profil_counter profile_fixup
program_invocation_name program_invocation_short_name ptmalloc_init
ptmalloc_init_all ptmalloc_lock_all ptmalloc_unlock_all pvalloc qsort
raise rawmemchr read read_alias_file read_conf_file readlink realloc
realloc_check realloc_hook_ini receiver register_printf_function
release_handle result.1 rindex root rtld_search_dirs rule_dstoff
rule_stdoff run_fp running samples save_arena save_for_backup
save_for_wbackup save_free_hook save_malloc_hook sbrk sbrk_base
search_tree setenv setitimer setlocale setmntent sigaction sigfillset
siglongjmp sigprocmask sscanf stage state state state state stderr
stdin stdout step0_jumps.1 step1_jumps.2 step2_jumps.3 step3a_jumps.4
step3b_jumps.5 step4_jumps.6 step4_jumps.7 stpcpy strcasecmp strchr
strchrnul strcmp strcpy strdup strerror_r string_space
string_space_act string_space_max strncasecmp strncmp strndup strnlen
strpbrk strrchr strsep strstr strtod strtof strtol strtold strtoll
strtoq strtoul strtoull strtouq sys_errlist sys_nerr sysconf
system_dirs system_dirs_len tcgetattr tdelete tdestroy
tdestroy_recurse textsize tfind timelocal timezone to_mb to_wc
top_check top_pad tos trans_compare transcmp transitions
translit_from_idx translit_from_tbl translit_to_idx translit_to_tbl
transmem_list trecurse trim_threshold tsearch twalk type_idxs types
tz_rules tzname tzset tzset_internal tzset_lock tzstring_list uname
undefined_msg unsecure_envvars.0 unsetenv using_malloc_checking valloc
vfprintf vfscanf vsscanf walt_digits walt_digits_initialized wcrtomb
wcschr wcslen wcsnlen wcsrtombs wmemcpy wmemmove wmempcpy write writev
yycheck yydefact yydefgoto yypact yypgoto yyr1 yyr2 yytable
yytranslate zeroes zeroes zone_names
The point is, /both/ libraries have a lot of code re-use. Dynamic
linking puts that code re-use in some random .so in $foo/lib; static
linking puts it in your executable. That's the major difference.
(The minor difference is that most Haskell implementations use libc
more than most C implementations use the Haskell standard libraries.)
> This is not the fault of any particular implementation - the
> ghc-built binary has a similar collection - rather it is dictated by
> the nature of the language and its standard libraries. Because
> Prelude functions are small and re-usable, they do get used all over
> the place in the implementation of other parts of the Prelude, so
> you end up with a huge dependency graph hiding underneath the
> simplest of calls.
You mean like printf("Hello, world!\n")?
> In fact, most of the extra stuff in "Hello World" is there purely to
> handle all possible error conditions in the I/O monad.
You mean as opposed to C, where most of the extra stuff is there
purely to support number formatting?
> Several years ago, Colin Runciman and I did the experiment of
> removing all the nice error-handling stuff from the prelude (and
> eliminating a few classes too I think), to see just how small we
> could squash "Hello World". The idea was to target embedded systems
> where memory is a scarce resource, and fancy error-reporting is
> pointless (a single red LED would do). IIRC, we managed to achieve
> a size of 25kb, compiled with nhc98, which don't forget includes a
> bytecode interpreter in the runtime system.
> Regards,
> Malcolm
To sum up: code re-use within a library certainly happens in more
languages than just Haskell :) So don't blame code re-use for
Haskell-specific (or non-specific) problems, because code re-use is
not a Haskell-specific cause.
Jon Cast