| Age | Commit message (Collapse) | Author |
|---|
|
http://lists.landley.net/pipermail/toybox-landley.net/2020-December/012169.html
|
|
mksh doesn't support the =~ extension, so my previous change -- while
fixing the host -- broke Android.
|
|
The GNU tar manual, when talking about the `tar --transform` option that I
need to implement, describes the 'x' flag by saying "regexp is an extended
regular expression (see section 'Extended regular expressions' in GNU sed)".
Only it turns out that even the latest GNU sed doesn't actually have
that flag. It's unique to `tar --transform`. That link is just telling
you that the sed manual will explain extended regular expressions, not
that GNU sed also supports the 'x' flag.
So I don't know whether we want this in toybox sed after all. (It made
sense that sed would have such a flag, but no sed that I know of
actually does.)
|
|
Based loosely on the Plan9/Inferno utility, and a convenient way to go back
and forth between code points and utf8 sequences.
This patch also fixes a couple of bugs in wctoutf8 (and the tests for this
toy effectively serve as unit tests for wctoutf8/utf8towc).
|
|
The rest of `make tests` all pass on a Raspberry Pi 400.
|
|
|
|
|
|
Add new toy `pwgen`.
A usefule password generation utility.
|
|
The AOSP build doesn't use tr (or anything that's still in pending), but
the kernel folks have been more aggressive. They found that tr's
pathological flushing was adding minutes to their build times.
Just removing the fflush() made tr significantly faster for my trivial
test, but still slow, with all the time going into stdio. Rewriting the
loop to modify toybuf in place and then do one write per read made most
of the difference, but special-casing the "neither -d nor -s" case made
a measurable difference too on a Xeon.
Bug: http://b/174773617
|
|
sed lies to fool autoconf, but also confuses the toybox test suite.
Without this patch we're skipping all "toyonly" tests on both GNU and
toybox sed.
|
|
|
|
Change 5109da9b3e6a898c8e0ad647303a1b375e3d97d3 caused test.test to call
chmod with mode `u+s+s` which passes on the host (where you have a
toybox test but a GNU chmod) but fails on Android where chmod is toybox
too.
Add the missing loop to string_to_mode(), which means this will also
affect other toys, but that seems like a feature (and, for example, GNU
mkdir also accepts a mode like `a=r+w+x`).
|
|
|
|
|
|
|
|
tests to catch this. While there add -k to test sticky bit.
|
|
|
|
|
|
Can't use the same set_terminal() logic as ptys because it not displaying
data, it should just accurately copy it.
|
|
|
|
|
|
|
|
Use 0 optstr prefix, fix bug initializing toys.envc one too low.
|
|
|
|
I didn't know until implementing this that --iso is actually called
--iso-8601 and that GNU date will actually accept any prefix. --iso-8
works fine too. I've assumed that --iso (that I always used) and
--iso-8601 (as given in the documentation) are the only two that matter.
|
|
|
|
|
|
I also promised to fix readelf. Where in file(1) I made no attempt to
say what was bad (or even to change `goto bad` to explicitly say that
*anything* was bad), I believe that readelf is much more likely to be
shown invalid ELF files, and that it would be useful to have some clue
as to what's wrong. Relatedly, this patch removes all existing
error_exit() calls in case it's being used on multiple files.
Again, this survived ~24hrs of AFL++ trying to blow its house down.
Test: ~/AFLplusplus/afl-fuzz -i tests/files/elf -o fuzz-out -- ./readelf -a @@
|
|
I promised months ago I'd fix this, and there was a (not visible to the
public but filed by a member of the public) bug filed against Android in
the meantime, but judged No Security Impact because "toybox is not a
security boundary". Anyway, it seemed high time I learned about fuzzing
command-line tools with AFL++, so here we are.
With these patches (and starting from the ELF files in test/files/elf),
toybox file survived ~24hours against AFL++. Amusingly it corrupted the
ELF files hard enough that it also managed to find a bug in the code
for MS-DOS executables, which is the motivation for the final hunk in
this patch.
Bug: http://b/159065007
Test: ~/AFLplusplus/afl-fuzz -i tests/files/elf -o fuzz-out -- ./file @@
|
|
|
|
Android is introducing a new binary XML format that is a drop-in
replacement for many existing .xml files written by system_server.
Since engineers may be surprised when encountering this new format,
add it to the "file" tool to aid identification in the field.
|
|
Even though the gcc toolchain ./configure is told it's cross compiling,
it optimizes for the current machine ANYWAY. So force it to be generic.
|
|
output depending on whether address was hex or decimal.
|
|
This makes it more likely that you can actually debug something like
"sed: bad regex: empty (sub)expression" or
"sed: bad regex: parentheses not balanced" from a build failure log,
where you don't necessarily know where the failure came from.
This also seems like it might be useful generally, although GNU grep
doesn't include this detail in its error messages, and busybox doesn't
even seem to notice that regcomp() failed?
(Realistically if the Android build team wants to move forward with "one
true regex implementation", we're going to have to add some GNU-isms to
the Android regex implementation. But we'd need to find them first! Note
that the two examples given above are real examples from failed buildbot
builds, but they occur early in the respective builds so there are
likely many more to look at after these. Interestingly, the first of the
two appears to be the more general case of something disallowed by POSIX
that xregcomp() already has a workaround for.)
|
|
I added a #include above this, which caused subtle breakages on 32-bit
systems. Move it to the top of the file to fix it and avoid making a
similar mistake in future.
|
|
One reason to use toybox on the host is to get the same behavior across
Android/Linux/macOS. Unfortunately (as we've seen from a few bugs) one
area where that doesn't quite work is that toybox uses the libc regular
expression implementation. That's fine, and mostly what users want, but
those folks trying to get the exact same behavior everywhere might want
to switch in a known regex implementation (bionic's NetBSD regex
implementation, say) for increased consistency.
That actually works pretty well, but portability.h has an #ifndef test
for REG_STARTEND before including <regex.h> that gets in the way. To
make up for that, this patch removes the unnecessary #include <regex.h>
from grep.c itself.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
