aboutsummaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2019-07-06killall: better handling of long names.Elliott Hughes
Change names_to_pid() so that we can actually match shell scripts with long names (the code to get the shell script's name was correct, but there was an extra test preventing us from actually comparing it to the sought name). In kill.c itself, remove a dead test for -l and switch to the FLAG() macro. Also extend the tests to explicitly cover long and short names.
2019-07-06ifconfig.test: disable the pointopoint tests.Elliott Hughes
These tests don't work for me as root on either my Debian desktop or my Android devices. The original mail thread implies that they were as close as the original comitter could get to _something_ that seemed to work, even if the kernel doesn't seem to bother with this: http://lists.landley.net/pipermail/toybox-landley.net/2014-November/003795.html toybox ifconfig *is* missing support for `-pointtopoint` and `pointopoint` (without an addess), similar for `broadcast` according to the man page. But since we don't appear to have a way to test this (other than looking at strace output!) I'm leaning towards YAGNI anyway...
2019-07-06Add comment and test.Rob Landley
2019-07-06Have env exec, not recurse to builtin.Denys Nykula
My app runs npm i -S better-sqlite3, and node-gyp-bin from the npm upstream has #!/usr/bin/env sh, expecting the system shell.
2019-07-03nl: nl -vEric Molitor
Implement nl -v supporting negative and zero starting values.
2019-07-01netcat: Add UNIX domain socket support (-U)Tom Cherry
Other implementations of netcat support UNIX domain sockets via -U, so this change adds it to toybox too. Example usage: Create ./uds as a UNIX domain socket and listen on it, forwarding messages to bash: > nc -U -s ./uds -L /bin/bash Connect to ./uds as a UNIX domain socket > nc -U ./uds
2019-06-29More toysh flow control plumbing.Rob Landley
2019-06-29Bash Compatibility PatchEric Molitor
I know that you are working on toysh which I'm looking forward to. In the meantime below is a patch to improve compatibility with older/odd versions of bash. This fixed a minor build issue I was having on MacOS (which was using zsh in emulated bash mode) as well as an oddball embedded SDK using a non-gnu version of bash. I believe these changes are minimal and should be safe to apply, if not I wanted to at least get them on the mailing list in case others ran into these issues.
2019-06-28modinfo: various small fixes.Elliott Hughes
Improve Android support (see code comments). Remove unnecessary fixed-length limits. Show error if module not found (plus test). Expand help text. Switch to FLAG macro. Stop hard-coding module assumptions in the tests.
2019-06-28split.test: don't rely on bash process substitution.Elliott Hughes
mksh doesn't support process substitution, so use an explicit temporary file.
2019-06-28file, stat: various small improvements.Elliott Hughes
file now shows the target of a symbolic link and calls out broken symbolic links. file now shows the device type for block/character special files. file now shows specific reason when it can't open. stat now includes the device type, plus a little more space between the number of blocks and the human-readable file type. Adjusted tests accordingly, which actually makes more of them pass on the host as a convenient side-effect, but I actually made these changes because I've been finding the desktop file and stat output more convenient in these cases.
2019-06-27touch.test: add missing `TZ=utc`s.Elliott Hughes
I don't know why we fixed the other tests but not this one. Before this patch it's failing with/without TEST_HOST. After this patch, it passes both.
2019-06-26testcmd shouldn't use shell builtin for TEST_HOST unless there isn'tRob Landley
one in $PATH.
2019-06-26losetup: fix Android.Elliott Hughes
Use /dev/block/loop* more uniformly, and teach the tests which to expect.
2019-06-26Fix cp.test to pass even if you're root or have a restrictive umask.Elliott Hughes
2019-06-25killall.test: don't assume the shell is /bin/sh.Elliott Hughes
The killall test still fails with toybox. Seems like right now killall and pidof share name to pid logic, but playing about with the desktop killall (which does pass this test), it seems like killall actually behaves more like pgrep than pidof (which does seem to be this limited on the desktop too)...
2019-06-25Make chgrp and chown tests work on Android.Elliott Hughes
There's no /etc/passwd and /etc/group, but there are enough users and groups that we can test with. ("bin" and "daemon" were added for LTP; "shell" is what you get with `adb shell`.)
2019-06-25blkid: adapt tests for util-linux blkid.Elliott Hughes
util-linux's blkid doesn't support reading from stdin, so move that to being a special toyonly test rather than the default. toybox blkid differs from util-linux in that it doesn't seem to find the LABEL for the checked-in example f2fs file system (the offset is wrong, but also f2fs uses LE16 strings), nor does it support the SEC_TYPE that util-linux blkid shows for several of the test file systems.
2019-06-25Start over on toysh.Rob Landley
Basic line continuation logic (to prompt with $PS2). Doesn't use result yet.
2019-06-25losetup: minor fixes.Elliott Hughes
Fix `losetup -f` to not fail with an error. Add the missing \n for `losetup -f --show FILE`. Use decimal for the device number, like the desktop losetup. Switch to the FLAG macro. Make the tests runnable as tests, and expand coverage a bit. With this patch, the tests pass both with and without TEST_HOST on the desktop. Note though that this patch is part of fixing some real-life losetup issues, not part of the "test cleanup" I'm also looking at. losetup is low down that list!
2019-06-25lib/args.c: add quotes.Elliott Hughes
This failure was a little hard to parse: tar: Unknown option mode (see "tar --help") This would have been clearer: tar: Unknown option 'mode' (see "tar --help")
2019-06-25tar: implement --mode.Elliott Hughes
Used by build/make/tools/mktarball.sh in AOSP. (Which is why today's switch to toybox tar got reverted.)
2019-06-24diff: fix diff of stdin for systems without /tmp.Elliott Hughes
xtempfile() alreay does the right thing, so switch to that. Also use xsendfile() for the actual copying. Fixes the cat tests on Android.
2019-06-24Fix FAILCOUNT arithmetic for mksh.Elliott Hughes
mksh doesn't support $[], only $(()).
2019-06-23Switch cp.c to FLAG() macros.Rob Landley
2019-06-22Don't test shell builtin for kill.Rob Landley
2019-06-22Make testcmd work if the path to $PWD has a space in it.Rob Landley
2019-06-22Use FORCE_FLAGS in kill.c.Elliott Hughes
Otherwise kill flags don't work unless killall5 is also configured. Also switch to the FLAG() macro.
2019-06-22Fix mv with trailing slash in source.Denys Nykula
Press tab, have bash complete dir name with a slash, notice musl rename() dislikes that. Replace trailing slash in the cp loop with a null character, if the command name is mv. Add the slash back if an error occurs.
2019-06-21bc: fix a few bugs found in upstreamGavin Howard
2019-06-21bc: remove a power test that is now an errorGavin Howard
2019-06-21Fix cp -r dir/. symlink child.makepost
Remove the existing link before trying to re-create, passing the test. Add -p to the -r test as a regression guard, portage calls cp with both.
2019-06-20The O_PATH breaks stuff, have to work around Android's selinux rules for /Rob Landley
some other way.
2019-06-20Teach dlist_pop() to work on a dlist_terminate()d listRob Landley
2019-06-19Avoid a double-definition (iphdr) from netinet/in.h and linux/ip_tunnel.h by ↵Mike Bennett
locally defining the single constant definition ip.c uses (IP_DF) if not already defined.
2019-06-12bc: fix previous patchGavin Howard
I put the check into the wrong if. That was my bad. Again.
2019-06-11Add DIRTREE_STATLESS to return entries we couldn't stat().Rob Landley
Sets ->again |= 2 when that happens and ->st is zeroed. While we're there, don't memset(st) and then memcpy(st) over it, and use O_PATH in the open().
2019-06-11Move sig_to_num and num_to_sig into portability.c.Elliott Hughes
macOS doesn't have real-time signals, and it has a slightly different set of non-portable signals from Linux. With this, the toybox `kill -l` output matches the default macOS kill(1).
2019-06-11bc: fix a bug in powerGavin Howard
This bug is that an error should be returned when the user tries to take 0 to a negative power, since that is undefined, but bc would return 0.
2019-06-11find: add -true/-false.Elliott Hughes
Used near the end of the AOSP build. Almost there! (This patch also fiddles with the help text to be able to slip the new options in without requiring so much extra space.)
2019-06-10kill: add trivial -l tests.Elliott Hughes
(`kill -l HUP` and `kill -l 1` both said "HUP" instead of giving you back the one you didn't provide, before my real-time signal patch.)
2019-06-10Add failing test for cp -r dir/. symlink child.makepost
Gentoo packages that build multiple variants, like once for every version of a slotted dependency, currently fail to install if their source dir includes a relative symlink to own child. Affects lots of Python infrastructure, for example meson and setuptools. You've already run into this issue, since cp.test has a todo. It's from 2008 though, so I guess I'll bump with an expanded test case. Libc is musl-1.1.22.
2019-06-10Improve signal name<->number conversions.Elliott Hughes
Include all the signals, not just the POSIX ones. In particular, improve support for real-time signals. My attempt to switch AOSP builds over to toybox timeout got reverted when I broke the ART build bots which use SIGRTMIN+2. Also fix `kill -l 3` to show "QUIT" and `kill -l QUIT` to show "3". Also make the `kill -l` output include numbers and names, and format it to look better on a 80x24 tty, because it's always August 1978 somewhere.
2019-06-09Inverted test.Rob Landley
2019-06-08macOS: numerous fixes.Elliott Hughes
This patch adds a BSD version of xgetmountlist (for the path == NULL case only), tested on macOS. It also papers over the differences between macOS' and Linux's xattr APIs. For once I think the macOS one is better. The imitation of mknodat I've had to write swings things back in Linux's favor though. BSD calls f_frsize by the name f_iosize instead. (FWIW, it looks like this is meaningless on Linux and actually meaningful on macOS.) I've added one #if to toys/ --- I'm calling pathconf in stat.c to work around the absence of f_namelen, and have left a TODO with an explanation. I'm not sure what the best fix is here, so punting. No-one can agree what f_fsid is, even if they're all basically the same, so work around the `val` versus `__val` issue between macOS and Linux. With this patch, it's now possible to build cp/mv/install and stat for macOS too. (Which completes the set of "toybox commands currently used on Linux as part of the AOSP build" if you ignore stuff that deals with processes, which I doubt we'll ever be able to support for lack of any API.)
2019-06-07Remove getprop, setprop, start, and stop from toyboxTom Cherry
These are Android specific, so not really helping the outside community, and are getting more and more Android dependencies to work correctly, so let's drop these from toybox and build them within Android. Change-Id: Ic6022f1f506e10868c61f55d64fa4e7c1b14eba2
2019-06-07Implement stat %C and find %Z.Elliott Hughes
We'd documented find %Z but not implemented it. We'd neither documented nor implemented stat's corresponding %C (they'd already taken %Z for ctime, which is ironic because %c/%C sounds more obvious than %z/%Z for that to me).
2019-06-05macOS build parallelism.Elliott Hughes
Macs are slow enough without crippling them further!
2019-06-03Clean up remaining TAIL_SEEK references.Elliott Hughes
2019-06-03Remove TAIL_SEEK config option, just always include it.Rob Landley
Yeah, it's twice the size and two codepaths, but seekable is the common case, it won't work in pipelines without a non-seek codepath, and the performance penalty not using seek on large files is a enormous.