Commit Graph

5608 Commits

Author SHA1 Message Date
Rob Landley
7c94bed234 Big dead code elimination pass from Garrett. 2006-05-03 21:58:45 +00:00
Rob Landley
db2ab89b7c Build break fix from Bernhard fischer, fixing fallout from svn 14974. (My bad.) 2006-05-03 21:39:28 +00:00
Rob Landley
f76cd964ec Whitespace and documentation cleanup from Dennis Vlasenko. 2006-05-03 21:23:15 +00:00
Rob Landley
425e7584a4 Patch from Rich Felker to make ls use libc's qsort. 2006-05-03 20:22:03 +00:00
Rob Landley
1801e9cde7 Rich Felker writes:
mistake in a macro caused statement before declaration.
2006-05-03 20:19:14 +00:00
Rob Landley
035ef5659e Patch from Garrett Kajmowicz to move a lot of #defines into the
header file.
2006-05-03 20:05:18 +00:00
Rob Landley
2584e9b39f Remove HDIO_GETGEO_BIG which is dead (removed from 2.6) and was pointless back
in 2.4 (it fetched the same hd_geometry struct only with uint cylinders instead
of ushort cylinders: nothing that big has ever accurately reported gemoetry,
it's all LBA.

Also stop declaring the hd_geometry struct static: it's two chars, a short, and
a long for a grand total of 8 bytes.  That can go on the stack.
2006-05-03 20:00:00 +00:00
Rob Landley
0f0b645f73 Use ATTRIBUTE_NORETURN as suggested by tito, plus a few in-passing cleanups. 2006-05-03 18:28:06 +00:00
Rob Landley
c5598177bc Patch from Jason Schoon to add optional SIGUSR1 support to dd. 2006-05-02 22:44:04 +00:00
Rob Landley
67d5b8b5b1 Minor tweaks: remove traces of old $COMMAND environment variable. 2006-05-02 21:39:04 +00:00
Rob Landley
2631486f1b Patch from Shaun Jackman:
ls has an ugly bug. ls uses an array of pointers, the elements of
which are all in a linked list. To free the elements, instead of
freeing all the elements in the array, array[0..nelements], it frees
by iterating the linked list starting at array[0], which it assumes is
the head of the list. Unfortunately, ls also sorts the array! So,
array[0] is no longer the head, but somewhere in the middle of the
linked list. This patch fixes this bug, and also adds an
ENABLE_FEATURE_CLEAN_UP stanza.
2006-05-02 19:46:52 +00:00
Rob Landley
6389ff1115 Patch from tito removing obsolete code. 2006-05-01 19:28:53 +00:00
Rob Landley
c488f87953 Notes on portability, and on when #include <linux/blah> is appropriate. 2006-05-01 05:26:01 +00:00
Rob Landley
73f54702bc Fixup a build break if ENABLE_SHOW_USAGE switched off. (Fallout from the
compressed usage messages patch).
2006-05-01 00:53:40 +00:00
Rob Landley
3ea66a805d Patches from Denis Vlasenko to rework "current_name" handling and remove
print_current_name().
2006-04-29 20:08:40 +00:00
Rob Landley
8a6254994c Patch from Denis Vlasenko:
* Do not initialize globals to 0, it is done automatically
* unsigned short -> uint16_t,  unsigned int -> uint32_t
  where appropriate (did it ever work on Alphas?)
* triple sync() is silly - removed
* check_zone_nr uses check_zone_nr2 now
* remove trailing periods from messages, uppercase first letter
2006-04-29 20:05:17 +00:00
Rob Landley
9950cab8e0 Patch from Denis Vlasenko:
ix bug (inode_map and zone_map are char* pointers,
sizeof() on them gives 4 or 8)
2006-04-29 20:03:54 +00:00
Rob Landley
5bc3f05a23 Remove some unused code pointed out by Tito, plus a slightly more graceful
way of handling endianness.
2006-04-29 19:11:24 +00:00
Rob Landley
a3e4f38f7b Style fixes from Denis Vlasenko. 2006-04-29 16:06:31 +00:00
Rob Landley
5f8b5ecda2 Another patch from Denis Vlasenko:
Make hdparm smaller:
* Make bb_ioctl return the status
* Replace ioctl with bb_ioctl in a few places
* Add bb_ioctl_alt, use where appropriate (four places)
* unsigned char args1[4+512] = {WIN_IDENTIFY,0,0,1,};
  ate 0.5k of rodata, fix that
2006-04-29 16:03:40 +00:00
Rob Landley
e5b281f89e From Dennis Vlasenko, Make hdparm smaller:
* Use combined no_scsi_no_xt()
 * Have common unsigned char args[4] = {WIN_SETFEATURES,0,0,0};
2006-04-29 15:49:18 +00:00
Rob Landley
2e2d752d49 Cleanup to hdparm by tito. 2006-04-29 15:23:33 +00:00
Mike Frysinger
2514042661 termios.h is already included so dont include sys/termios.h as well 2006-04-29 04:45:22 +00:00
Mike Frysinger
02fe81033f detect more obsolete headers 2006-04-29 04:44:08 +00:00
Mike Frysinger
59c6baedf7 just include fcntl.h not sys/fcntl.h 2006-04-29 04:43:50 +00:00
Mike Frysinger
706f8f6de5 just include fcntl.h not sys/fcntl.h 2006-04-29 04:43:31 +00:00
Mike Frysinger
f054be12de Rob Sullivan writes: rewrite the regex handling to improve stability 2006-04-29 04:21:10 +00:00
Mike Frysinger
12ef09b818 compile.h is incorrect; create a new do_link.h and use that for bb_mkdep and usage 2006-04-29 04:19:10 +00:00
Mike Frysinger
fe6bb3853c use pretty output for scripts/bb_mkdep and scripts/usage 2006-04-29 04:12:36 +00:00
Bernhard Reutner-Fischer
3737c03630 - rerun gcse pass after reload, enable whole-program iff we are not building
the lib (due to bug in gcc; see comment.)
I would have run the testsuite to check if everything is still behaving, but
unfortunately i cannot run the testsuite anymore: it just sits there and does
nothing even with a pristine checkout and an old gcc.
srcdir=/tmp/busybox/testsuite /tmp/busybox/testsuite/runtest 
PASS: Applet order
FAIL: Common typos
PASS: Obsolete function usage
PASS: Obsolete function usage
<spins, ^C>
make[1]: *** [check] Interrupt
make: *** [check] Interrupt

I've just corrected that wrong double-printing.

Anyway. with gcc-4.1-HEAD from earlier today, we now have for
defconfig+ COMPILE_AT_ONCE:


880 -rwxr-xr-x  1 433 433 893476 Apr 28 21:41 busybox.oorig-4.0
860 -rwxr-xr-x  1 433 433 874560 Apr 28 21:36 busybox.oorig
844 -rwxr-xr-x  1 433 433 858752 Apr 28 21:49 busybox

   text    data     bss     dec     hex filename
 879920   11568 1039148 1930636  1d758c busybox.oorig-4.0
 862802   10192 1038796 1911790  1d2bee busybox.oorig
 848066    9100 1037536 1894702  1ce92e busybox

values for 4.0.2 just added for reference..
2006-04-28 20:02:29 +00:00
Bernhard Reutner-Fischer
15545b1db9 - print "Obsolet headers" instead of "Obsolete function usage" when checking
for obsolet includes.
2006-04-28 19:53:48 +00:00
Rob Landley
a2d9a1752c More usage.h cleanups, with collateral changes to httpd. Specifically,
most of our CONFIG entries switch stuff on instead of off when enabled, so fix
the gratuitously reversed INETD thing.
2006-04-28 19:38:04 +00:00
Rob Landley
0e6a3e14e4 Patch from tito to update output of hdparm. 2006-04-28 01:33:30 +00:00
Rob Landley
7e21d5f6b1 Patch from Dennis Vlasenko to add the option to compress help text. 2006-04-27 23:34:46 +00:00
Rob Landley
d5b9428bb6 Re-wordwrap the new tests, and fix seq.c to pass them. 2006-04-27 22:36:32 +00:00
Bernhard Reutner-Fischer
a345dc855f - add two regression tests for seq.
"seq two arguments equal, arbitrary negative step" fails but should pass..
2006-04-27 13:46:56 +00:00
Bernhard Reutner-Fischer
fc477f4bc6 - add note about crond 2006-04-26 19:40:15 +00:00
Bernhard Reutner-Fischer
06a81d368e - patch from Denis Vlasenko to make sizes PHONY and to add objsizes which prints
the sizes of the individual object files.
2006-04-26 10:10:25 +00:00
Rob Landley
5fe8d5b9c5 Convert a chunk of usage.h to USE_ and SKIP_ (more to do there), and fix a
DEFALT typo in du while I was there.
2006-04-25 22:40:53 +00:00
Rob Landley
7818a422bc Portability tweak from Shaun Jackman, don't include asm/page.h directly. 2006-04-25 18:42:23 +00:00
Rob Landley
997650b972 From Denis Vlasenko:
* CPU% = process[i].ticks/sum(process[i].ticks) * busy_cpu_ticks/total_cpu_ticks
* got rid of empty line under "Load average" line
  and used last line - +2 visible processes
* do not do float conversion of loadavg, just read it as string from /proc
* fix display on small screens
* dropped unused .stime, .utime fields
* a few variables were renamed
* style fixes
2006-04-24 23:13:46 +00:00
Bernhard Reutner-Fischer
56fbd21fa7 - return ret so we return something meaningful. 2006-04-24 17:41:29 +00:00
Mike Frysinger
70cbb6e026 Jason Schoon writes: Attached is a patch to display the usage message if fuser is called with no arguments, rather than doing no work and failing silently. 2006-04-21 22:04:05 +00:00
Mike Frysinger
10f55b1dae Jason Schoon writes: This makes the case where LFS is not defined, but fdisk is configured with large disk support use the pretty build rules. 2006-04-21 21:59:22 +00:00
Mike Frysinger
586ea61284 2006-04-20 Shaun Jackman <sjackman@gmail.com>
* Makefile: Build a bFLT binary.
        * Rules.mak: Ditto.
2006-04-21 01:18:07 +00:00
Ned Ludd
791976490b - passwd doesnt use salt with md5 passwords; bug #604 thanks taviso 2006-04-21 00:40:35 +00:00
Mike Frysinger
f162183527 kill signed warnings 2006-04-20 04:00:11 +00:00
Rob Landley
998dbee6d9 Patch from Jean Wolter:
it looks like the introduced support for character classes and
equivalence classes is not correct. The attached patch tries to fix
some symptoms and tries to make tr behave like gnu tr for the added
test cases. The patch

 - removes if clauses with side effects
 - fixes handling of buffer pointer (strcat added characters to the
   buffer without increasing the buffer pointer)
 - re-arranges character classes to match ASCII order

regards,
Jean
2006-04-19 22:22:06 +00:00
Rob Landley
5076eb4af9 svn 14824 broke -funsigned-char and possibly building out of tree, because
$(srcdir) can't bind early since it's set to a different value in each subdir.
If it binds early, it's blank, hence an empty -I which eats the next option
as a directory to look for #include files in.

So CFLAGS has to bind late, but the check_gcc stuff should only get run once.
I added a "make V=2" mode to show when check_gcc gets run, and turned WARNINGS
into a CFLAGS line since it was always getting added anyway.
2006-04-19 20:32:27 +00:00
Rob Landley
856489b63c New applet cksum, from Rob Sullivan. 2006-04-18 20:57:28 +00:00