| Age | Commit message (Collapse) | Author |
|---|
|
(Moved header a while ago, forgot to check in function move.)
|
|
|
|
|
|
|
|
on files opened read-only, so revert last commit (force writeable logic).
|
|
These three are currently available in Android's ps but not in toybox ps.
|
|
|
|
(leading underscore version is 1<<NAME_TAG version, with 1LL for >31 shift),
suck lots of magic constants out of ps and use tag macros instead,
redo command line display so there's now 6 variants (CMD COMM ARGS from posix,
NAME CMDLINE from android, and COMMAND for completeness). Document more
cases where posix is nuts or widely ignored.
|
|
and slot numers into a structure. (Keeping multiple arrays in sync may have
been efficient but it was ugly.) Fix duplicate command name copying that
corrupted the name of kernel threads. Tighten up slot[] docs.
|
|
Valgrind complains that we add uninitalized memory to totals[] fields we never
read from. (If we didn't set it in entrylen() we don't use it during display,
they're testing the same flags), but valgrind doesn't understand that.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Includes tests for the new feature, and a failure case for the minimal
perms test as well.
Also some typo fixing / massaging the help text so it fits in 80
columns.
|
|
move slot[] documentation into C code, add PR %VSZ VIRT RES SHR
|
|
|
|
|
|
|
|
instead of carving up toybuf by hand. This makes breaking out the field
conversion logic into its own function less expensive.
|
|
Change readfileat() to pass back length of read.
|
|
1) It read st_dev instead of st_mode.
2) It reversed the semantics of absolute vs minimal ('-' prefixed) tests.
Add tests for these, and move the "unterminated -exec" test into the "Still
fails" section because it's still dumping core for me.
|
|
|
|
Four-digit years were being mangled by the code for two-digit years.
Move all the two-digit year code into the "we only saw two digits" case.
Add some new tests and fix existing tests.
|
|
functions to read major()/minor(), fix printf format warnings in error msg.
|
|
|
|
|
|
Both Android and GNU interpret -n to mean "show numeric users and groups",
despite what POSIX says.
|
|
This is equivalent to Android's historical "NAME" column, showing the
first element of /proc/pid/cmdline.
|
|
|
|
(The help infrastructure can collate and alphebetize short options,
long options haven't got a standardized help text format. Not a hard
requirement to have a short opt, but it's nice.)
|
|
"comm" is the command name without arguments, and "cmd" the command
line including arguments. Confusingly, "command" is a synonym for
the *latter*, not the former.
This implementation of "comm" matches the GNU ps behavior where the
names are truncated, but Android historically used the full name
from /proc/pid/cmdline instead, so this patch isn't sufficient to
let us match the Android behavior.
|
|
I actually thought ps was segfaulting, and it wasn't until I was in gdb
that I even noticed there was an error message at all, and even then I
had to read the source to work out what it was trying to tell me.
Before:
$ ps -o user,pid,ppid,vsize,rss,nice,wchan,addr,unknown,stat,cmd
ps: -o 'user,pid,ppid,vsize,rss,nice,wchan,addr,unknown,stat,cmd'@41
USER PID PPID VSZ RSS NI WCHAN ADDR$
After:
$ ps -o user,pid,ppid,vsize,rss,nice,wchan,addr,unknown,stat,cmd
ps: bad -o field 'user,pid,ppid,vsize,rss,nice,wchan,addr,unknown,stat,cmd'
^
|
|
As with ls, it doesn't seem like -Z should be guarded behind LSM
availability. On a non-SELinux system, the label is always "unconfined".
|
|
Used by Android.
|
|
(Matches beginning and EOF at once.)
|
|
|
|
This lets '(x)\1' match, as reported by Isabella Parakiss.
|
|
an extra newline because the test for whether we have an existing string to
append a newline to was checking if struct step had data appended to it,
and the /x/ regex is data appended to it. Change test to check for null
terminator at ->arg1 offset.
|
|
|
|
inspired by:
https://android-review.googlesource.com/#/c/152663/4/libc/bionic/fgetxattr.cpp
|
|
Also, if you mkdir "$(echo -e "one\ntwo"); chmod -r one*; ls -q one*
it honors -q.
|
|
the global so "install -g 0" doesn't alias to "cp --preserve 0" and error out.
through to --preserve 0"
|
|
On Android, the filesystem column is pretty wide. Actually measure the widths.
|
|
|
|
Adds XVFORK() macro, teaches xpopen_both() to call /proc/self/exe with NULL
argv (and converts cpio -p to use that), adds TOYBOX_FORK guards to some
unconverted commands.
|
|
Note: vfork(), like fork(), can return -1 if too many processes, and
we should notice and fail loudly.
|
|
Elliott Hughes found a bug https://android-review.googlesource.com/#/c/170020/
and Daniel K. Levy worked out the problem: the user/group/newer arguments
to find weren't consuming their arguments when not checking the results of
their comparison (because an earlier test had already caused their
parenthetical group to fail). This confused the argument parsing logic
and could lead to segfaults.
I applied a different fix that reorganized the existing tests instead of
adding a new test. (Looks like a big commit but it's mostly whitespace
due to extra curly brackets changing indendentation levels.)
|
