Rob Landley [Tue, 10 Mar 2015 16:07:28 +0000 (11:07 -0500)]
Use xsignal() instead of signal().
Rob Landley [Mon, 9 Mar 2015 20:06:10 +0000 (15:06 -0500)]
Fix thinko (don't &toybuf to get scratch space) and add -v option.
Rob Landley [Mon, 9 Mar 2015 19:52:32 +0000 (14:52 -0500)]
Upgrade oneit with -r (restart), -3 (send exiting PID values to child), and signal handling.
Rob Landley [Fri, 6 Mar 2015 22:50:54 +0000 (16:50 -0600)]
sed depends on -r to be rightmost flag to simplify the REG_EXTENDED test.
(If a flag is 1, you can multiply it by the value you want and get that value
else zero without a branch or conditional assignment.) So move -E one to the
left so FLAG_r is 1 again, and add a [+Er] suffix instead of testing
FLAG_r | FLAG_E in the users.
Rob Landley [Fri, 6 Mar 2015 21:12:52 +0000 (15:12 -0600)]
The "no }" error with find | xargs sed is because toy_init() wasn't blanking the GLOBALS() block on a recursive call. (How that ever worked...)
Rob Landley [Fri, 6 Mar 2015 21:11:38 +0000 (15:11 -0600)]
Yank $STRIP from config (and STRIP=no weirdness) and just allow strip to fail.
Hyejin Kim [Wed, 4 Mar 2015 13:29:57 +0000 (07:29 -0600)]
scripts/mkflags.c : main function need return 0 if normal exit
Rob Landley [Wed, 4 Mar 2015 04:06:55 +0000 (22:06 -0600)]
Quick cleanup pass on tar.
Rob Landley [Tue, 3 Mar 2015 02:27:50 +0000 (20:27 -0600)]
On 64 bit, subtracting two pointers produces a long result. On 32 bit, it's an int. Even though long _is_ 32 bits on a 32 bit systems, gcc warns about it because reasons.
Also, the warning being that "expects int, but type is wchar_t"... no, type
is not wchar_t. Type is probably long. Specify the ACTUAL TYPE, not the random
typedef alias for it. If the translated type _did_ match, there wouldn't
be a warning! (This is why c89 promoted all arguments to int, precisely
so this wasn't a problem.)
Rob Landley [Sun, 1 Mar 2015 22:43:01 +0000 (16:43 -0600)]
Fix several printf_format warnings.
Rob Landley [Sun, 1 Mar 2015 22:35:05 +0000 (16:35 -0600)]
Only switch on printf format warnings for error_exit() and friends when TOYBOX_DEBUG enabled.
One again gcc manages to be incompetent about producing warnings. For example,
in bzcat we error_exit() with a message that's an error code fed into an
array of string constants, and apparently dereferencing an array of string
literals doesn't give you a string literal according to -Wformat-security.
Not breaking the code to humor the compiler here, the compiler is wrong.
Elliott Hughes [Sun, 1 Mar 2015 22:16:50 +0000 (16:16 -0600)]
let the compiler check format strings
i'll be AFK for a week, so here's the patch i've been using this
evening to find other format string mistakes. BSD uses __printflike
and takes two arguments instead of hard-coding (1,2), but i figured
that as long as you don't need the generality you'd prefer not to have
it. and it's easy enough to retrofit if we ever do have a formatting
function that takes other arguments.
Elliott Hughes [Sun, 1 Mar 2015 22:11:50 +0000 (16:11 -0600)]
fix format problems in tar.c
%o is unsigned, but off_t is signed.
* takes an int.
an error_msg call was missing an argument.
only one of these is an actual error, but i'd like to fix the others
too so that we (toybox, but if not, then Android) can turn on format
string warnings to prevent future bugs like the stat.c LP32 ones.
Rob Landley [Sun, 1 Mar 2015 22:08:51 +0000 (16:08 -0600)]
Patches from Elliott Hughes to fix various printf() format strings.
(Don't specify the space flag when it's the default behavior. The ifconfig
stuff was actual bugs.)
Rob Landley [Sun, 1 Mar 2015 21:58:40 +0000 (15:58 -0600)]
Patches from Elliott Hughes to add missing arguments to error_exit() calls.
Elliott Hughes [Sun, 1 Mar 2015 21:48:49 +0000 (15:48 -0600)]
remove accidental space in format string
the ' ' flag makes no sense for %c.
Elliott Hughes [Sun, 1 Mar 2015 21:35:48 +0000 (15:35 -0600)]
fix copy/paste-o in useradd.test
Rob Landley [Fri, 27 Feb 2015 18:17:30 +0000 (12:17 -0600)]
Cleanup pass on ps.
Rob Landley [Fri, 27 Feb 2015 14:13:24 +0000 (08:13 -0600)]
0.5.2 release notes.
Rob Landley [Sun, 1 Mar 2015 01:54:03 +0000 (19:54 -0600)]
Minor toysh cleanup: remove another unnecessary option.
Isabella Bosia [Sat, 28 Feb 2015 21:21:13 +0000 (15:21 -0600)]
BSD sed uses -E instead of -r to enable ERE.
GNU sed silently supports -E as an alias for -r too.
Rob Landley [Sat, 28 Feb 2015 18:39:16 +0000 (12:39 -0600)]
Add LDFLAGS and STRIP=no support.
Rob Landley [Sat, 28 Feb 2015 18:35:12 +0000 (12:35 -0600)]
Add "make change" output to .hgignore
Rob Landley [Fri, 27 Feb 2015 03:07:33 +0000 (21:07 -0600)]
More 'splaining.
Rob Landley [Thu, 26 Feb 2015 02:42:24 +0000 (20:42 -0600)]
Added tag 0.5.2 for changeset
8ce8c5365f8b
Rob Landley [Wed, 25 Feb 2015 06:45:50 +0000 (00:45 -0600)]
stat -f didn't work on 32 bit systems because type sizes wrong (fix from Elliott Hughes).
Rob Landley [Wed, 18 Feb 2015 21:19:15 +0000 (15:19 -0600)]
Patch from Isaac Dunham to add -r, fixed up so it doesn't try to include two flag contexts simultaneously.
Yeongdeok Suh [Wed, 18 Feb 2015 19:06:26 +0000 (13:06 -0600)]
Separate 'userdel' from testing syntax.
When I test toybox with toybox/tests/*.test scripts,
I got many false FAILs from it. So, I tried to fix useradd.test file.
Rob Landley [Sun, 15 Feb 2015 21:44:26 +0000 (15:44 -0600)]
Add --remove-destination longopt for -F because the host version doesn't have a shortopt for it.
Rob Landley [Sun, 15 Feb 2015 21:27:43 +0000 (15:27 -0600)]
Only discard optargs in toy_exec() after finding the command to run (otherwise the fallback exevp() segfaults).
Rob Landley [Sat, 14 Feb 2015 20:59:30 +0000 (14:59 -0600)]
Make egrep and fgrep build standalone.
Rob Landley [Sat, 14 Feb 2015 20:54:36 +0000 (14:54 -0600)]
Fix --help for single.sh builds.
Rob Landley [Sat, 14 Feb 2015 20:05:00 +0000 (14:05 -0600)]
Split out unix2dos so scripts/single.sh and "make change" can build it standalone.
Rob Landley [Sat, 14 Feb 2015 07:08:15 +0000 (01:08 -0600)]
Move guts of "make change" to scripts/change.sh, don't try to build sh or help standalone.
sh builtin commands need multiplexer, help describes other enabled commands.
Rob Landley [Thu, 12 Feb 2015 22:41:59 +0000 (16:41 -0600)]
Fix bug introduced by last commit (print template instead of toybuf).
Also, xstrdup() the unmodified template because changing the environment string
could make the changed version show up in "ps".
Rob Landley [Wed, 11 Feb 2015 23:10:28 +0000 (17:10 -0600)]
Only apply mktemp directory to targets that aren't already paths.
Rob Landley [Wed, 11 Feb 2015 04:51:23 +0000 (22:51 -0600)]
Fix longopts (pointed out by Hyejin Kim).
Rob Landley [Mon, 9 Feb 2015 23:47:29 +0000 (17:47 -0600)]
One more cleanup pass on hwclock factoring out common code.
Rob Landley [Mon, 9 Feb 2015 23:33:59 +0000 (17:33 -0600)]
Inline rtc_open() and get_rtc_seconds(), perform resulting cleanups.
Rob Landley [Mon, 9 Feb 2015 22:34:24 +0000 (16:34 -0600)]
Cleanup/refactoring pass on hwclock.
Inline open_wall_clock_rtc() into rtc_open(), factor out xtzset(),
inline set_sysclock_from_hwclock(), set_hwclock_from_sysclock(),
and set_sysclock_timezone().
/etc/adjtime is short enough we don't need to bother with a break.
The final "else" case in main() should always trigger because >0 in optstr
says "at most zero arguments", so the || at the end is always true, so
take the test off.
Rob Landley [Sun, 8 Feb 2015 22:38:38 +0000 (16:38 -0600)]
Promote chcon to "other".
Rob Landley [Sun, 8 Feb 2015 22:38:11 +0000 (16:38 -0600)]
Cleanup pass on chcon.
Rob Landley [Sun, 8 Feb 2015 22:33:24 +0000 (16:33 -0600)]
Rewrite of a patch from Elliott Hughes to implement touch -h and switch API to utimensat() which offers nanosecond granularity and UTIME_OMIT/UTIME_NOW.
Rob Landley [Sun, 8 Feb 2015 22:27:30 +0000 (16:27 -0600)]
For some reason, ifconfig was explicitly excluding "ifconfig lo:0 127.0.0.2" style alias address assignments. No idea why.
Elliott Hughes [Sun, 8 Feb 2015 01:51:27 +0000 (19:51 -0600)]
netcfg is being removed, but shows up in the toybox roadmap.
Rob Landley [Sun, 8 Feb 2015 01:45:23 +0000 (19:45 -0600)]
Shameless meddling.
Elliot Hughes [Sun, 8 Feb 2015 01:27:59 +0000 (19:27 -0600)]
Use $TMPDIR if set (necessary on Android, where there is no /tmp).
Include full template in error messages.
Don't report success on failure with -q.
Avoid unnecessary allocation.
Fix "xxxxxx" versus "XXXXXX" confusion.
Rob Landley [Sat, 7 Feb 2015 23:20:23 +0000 (17:20 -0600)]
Build standalone commands where "depends on" config entries need to be switched on.
Rob Landley [Sat, 7 Feb 2015 23:19:38 +0000 (17:19 -0600)]
Clean up the output of "make change".
Rob Landley [Sat, 7 Feb 2015 22:23:59 +0000 (16:23 -0600)]
Remove xexec_optargs().
Rob Landley [Sat, 7 Feb 2015 22:23:03 +0000 (16:23 -0600)]
Cleanup pass on env, removing exec_optargs().
Rob Landley [Sat, 7 Feb 2015 22:17:44 +0000 (16:17 -0600)]
Make toy_exec() check if argc is in optargs and deal with it there so we don't need a separate xexec_optargs().
Rob Landley [Sat, 7 Feb 2015 21:32:22 +0000 (15:32 -0600)]
Merge unshare and nsenter (promoting and cleaning up nsenter).
Needs more testing, don't have a test environment set up for this yet...
Rob Landley [Fri, 6 Feb 2015 23:31:52 +0000 (17:31 -0600)]
In rm, init using to AT_REMOVEDIR in the dir case earlier.
(If you had a chmod 000 directory and did rm -r on it without -f, after the prompt it would complain it was a directory.)
Rob Landley [Fri, 6 Feb 2015 22:42:44 +0000 (16:42 -0600)]
Alright, the Android guys agree with the musl guys: faccessat(AT_SYMLINK_NOFOLLOW) is not supported.
Rob Landley [Sun, 25 Jan 2015 20:04:51 +0000 (14:04 -0600)]
Commit 1061 said "this gets expand off my to-review list". Remove it from the list.
Rob Landley [Sun, 25 Jan 2015 19:11:58 +0000 (13:11 -0600)]
Cleanup mountpoint and remove it from the pending README.
Luis Felipe Strano Moraes [Thu, 22 Jan 2015 22:29:34 +0000 (16:29 -0600)]
bootchartd: Removing wrong free call to non-malloc'ed string.
Rob Landley [Thu, 22 Jan 2015 08:18:30 +0000 (02:18 -0600)]
Fix sed s//\[newline]/ line continuations.
The problem was that readline() was returning a newline at the end of each
string, which wasn't getting stripped in the parser and thus \ wasn't at the
end of a line for -f, it was escaping a literal newline, so the continuation
logic didn't trigger. Remove some redundant null checks while we're at it,
and don't bother terminating a string we don't return (yes we leak memory
in an error path, but it's about to error_exit() anyway).
Elliott Hughes [Tue, 20 Jan 2015 22:03:29 +0000 (16:03 -0600)]
fix hwclock's rtc selection
For systems using /dev/rtcN, /dev/rtc0 isn't necessarily the RTC
that's used to provide the system time at boot time. We need to search
for the RTC whose /sys/class/rtc/rtcN/hctosys contains "1".
Elliott Hughes [Tue, 20 Jan 2015 21:52:41 +0000 (15:52 -0600)]
fix typo in printf
Rob Landley [Sun, 18 Jan 2015 20:06:14 +0000 (14:06 -0600)]
Remove trailing whitespace.
Rob Landley [Sun, 18 Jan 2015 19:44:24 +0000 (13:44 -0600)]
Lift the basename/libgen.h shenanigans back out of portability.c and make it a static inline in portability.h, and prototype dirname() while we're at it.
Elliott Hughes [Sun, 18 Jan 2015 19:36:31 +0000 (13:36 -0600)]
fix non-glibc basename(3) usage
I considered a #else on the big #ifdef __GLIBC__ above, but that
seemed harder to follow.
Rob Landley [Fri, 16 Jan 2015 19:55:32 +0000 (13:55 -0600)]
Move pty.h back to toys.h (under LSB 4.1 headers).
Elliott Hughes [Fri, 16 Jan 2015 19:49:23 +0000 (13:49 -0600)]
AOSP master has <pty.h> and builds netcat/nc.
Rob Landley [Fri, 16 Jan 2015 19:43:09 +0000 (13:43 -0600)]
id.c #ifdefectomy. (ifdefs belong in headers, not in C code.)
Elliott Hughes [Fri, 16 Jan 2015 19:36:53 +0000 (13:36 -0600)]
This patch adds a TOYBOX_SELINUX configuration option to control both
the SELinux commands (such as chcon) and the SELinux-specific options
to regular commands (such as ls -Z).
This lets us #include <selinux/selinux.h> in portability.h.
I've also fixed chcon to insist on being given the a context argument.
This patch also adds -Z to id and fixes id's regular output (-G should
be separated by spaces, non-G output should be separated by commas,
and you don't want a double comma where the egid is omitted from the
list of groups).
Rob Landley [Fri, 16 Jan 2015 01:09:08 +0000 (19:09 -0600)]
Added tag 0.5.1 for changeset
62a7d617e1ce
Rob Landley [Wed, 14 Jan 2015 19:59:10 +0000 (13:59 -0600)]
Switch netcat to stop parsing command line at first nonoption argument instead of stopping after -l or -L.
Hyejin Kim tried "toyboxd nc nc -p 1234 -l rm mv ps" (which attempted to run
"nc rm mv ps") and "toybox nc -l -p 1234" (which attempted to run "-p 1234")
and found the behavior confusing. So make it work like xargs, where all
netcat options have to come before the start of any child command.
Rob Landley [Wed, 14 Jan 2015 19:54:34 +0000 (13:54 -0600)]
Make touch -d miliseconds work. (Still only 6 digits precision instead of 9, but it's a start.)
Elliott Hughes [Wed, 14 Jan 2015 09:43:49 +0000 (03:43 -0600)]
i found a few problems while manually smoke testing toybox chown versus toolbox (NetBSD) chown...
new test:
and here's the patch to fix "owner:" ":group" and the ":" special case:
Rob Landley [Wed, 14 Jan 2015 06:31:06 +0000 (00:31 -0600)]
Let chown build standalone.
Rob Landley [Tue, 13 Jan 2015 10:28:19 +0000 (04:28 -0600)]
sed bugfix: N or n at end of script would save the terminating NULL as the resume position, so the script would restart from beginning.
Rob Landley [Tue, 13 Jan 2015 09:35:37 +0000 (03:35 -0600)]
sed s/// can have line continuations in the replacement part, with or without a \ escaping the newline.
Rob Landley [Sun, 11 Jan 2015 16:17:58 +0000 (10:17 -0600)]
Promote printf.
Rob Landley [Sun, 11 Jan 2015 16:16:38 +0000 (10:16 -0600)]
One more bugfix for printf.c, with test suite entry. (Make %-3d etc work.)
Rob Landley [Sun, 11 Jan 2015 07:22:36 +0000 (01:22 -0600)]
Cleanup pass on printf.
Alas, passing a union as the last argument to printf does not appear to work
reliably, and there's no obvious way to manually assemble varargs in a portable
manner. So I have to repeat the printf once for each data type. Oh well.
Rob Landley [Sun, 11 Jan 2015 02:02:21 +0000 (20:02 -0600)]
More half-finished cleanup of printf.c, and more test suite entries.
Rob Landley [Tue, 6 Jan 2015 21:06:51 +0000 (15:06 -0600)]
Since "printf" is a shell builtin, printf.test wasn't actually testing anything.
Rob Landley [Tue, 6 Jan 2015 18:07:20 +0000 (12:07 -0600)]
More printf cleanup, and test suite entries.
Fixes bug introduced last time where toys.optargs was both snapshotted and used
directly and the two fell out of sync.
Rob Landley [Sun, 4 Jan 2015 09:50:52 +0000 (03:50 -0600)]
Fix sed backslash parsing in square bracket pattern sections.
Rob Landley [Sun, 4 Jan 2015 02:31:41 +0000 (20:31 -0600)]
Another cleanup pass on printf.
Rob Landley [Sat, 3 Jan 2015 22:25:36 +0000 (16:25 -0600)]
Move fflush() checking to xexit() and have exit paths in main() call that.
Rob Landley [Sat, 3 Jan 2015 03:28:39 +0000 (21:28 -0600)]
Cleanup pass on printf.
Rob Landley [Fri, 2 Jan 2015 06:44:27 +0000 (00:44 -0600)]
Need to update install.c for the changed OLDTOY() argument list.
Rob Landley [Thu, 1 Jan 2015 23:37:57 +0000 (17:37 -0600)]
Tweak status.html
Rob Landley [Thu, 1 Jan 2015 22:59:35 +0000 (16:59 -0600)]
Switch a lot of strncpy() calls to xstrncpy().
Rob Landley [Thu, 1 Jan 2015 22:49:55 +0000 (16:49 -0600)]
strncpy(optptr, hname, strlen(hname)) is really just strcpy().
Rob Landley [Thu, 1 Jan 2015 22:28:51 +0000 (16:28 -0600)]
strtol() doesn't return error indicator for overflow, it just sets errno. So add estrtol() (which clears errno first), and xstrtol() (which error_exit()s on overflow).
Rob Landley [Thu, 1 Jan 2015 22:19:40 +0000 (16:19 -0600)]
Debris from flag handling rewrite: don't allow -^A to actually trigger.
Elliott Hughes [Thu, 1 Jan 2015 16:58:59 +0000 (10:58 -0600)]
typo fixes
Rob Landley [Thu, 1 Jan 2015 04:38:48 +0000 (22:38 -0600)]
Fix for mix from Isaac Dunham (who can actually test it). (I tweaked the curly brackets.)
He says:
In the channel selection loop, "if (TT.chan)" had been combined with
a strcmp; the else conditionals were premised on if (TT.chan) alone.
The help text had been only partially updated for a couple option changes.
Rob Landley [Thu, 1 Jan 2015 03:30:59 +0000 (21:30 -0600)]
Redo option parsing infrastructure so #define FORCE_FLAGS can unzero flag macros for a disabled command (needed when multiple commands share infrastructure with a common set of flags).
This means the flag space is no longer packed, but leaves gaps where the zeroes
go. (Actual flag bit positions are the same for all configs.) Since the
option parsing needs to know where the holes are, the OPTSTR values are
now generated as part of flags.h with ascii 1 values for the disabled values.
(So generated/oldflags.h went away.)
This also means that the option string argument for OLDTOY() went away, it now
uses the same arguments as the NEWTOY() it references.
Rob Landley [Thu, 1 Jan 2015 03:23:37 +0000 (21:23 -0600)]
When CP_MORE was disabled, the d flag was still in the [-exclusion] list at the end and option parsing fell off the end looking for it (segfault) if DEBUG wasn't enabled to check for that.
Rob Landley [Wed, 31 Dec 2014 22:22:31 +0000 (16:22 -0600)]
When you include the posix header libgen.h, glibc #defines basename to some random other symbol name (because gnu) and this screws up nontrivial macro expansions of NEWTOY(basename), so work around it in portability.h.
Rob Landley [Tue, 30 Dec 2014 20:49:02 +0000 (14:49 -0600)]
Link to gmane web archive.
Rob Landley [Sun, 28 Dec 2014 20:44:09 +0000 (14:44 -0600)]
Another random link for the design page.
Rob Landley [Sun, 28 Dec 2014 20:38:14 +0000 (14:38 -0600)]
Fluff out help text.
Rob Landley [Sat, 27 Dec 2014 19:52:40 +0000 (13:52 -0600)]
Promote mix