If a system administrator prefixes /opt/ast/bin to the path and
then invokes the shell in restricted mode, they clearly intend for
the user to run those AST utilities.
Similarly, if a system administrator sets a PATH for a restricted
shell that includes libraries listed in the .paths file, they must
have intended for the user to use those loadable built-ins, as they
will be associated with the pathnames of their respective
libraries. Since the user cannot change PATH or use the builtin
command, they still cannot load just any built-in they choose.
src/cmd/ksh93/sh/path.c:
- Remove SH_RESTRICTED check when handling path-bound builtins
or dynamic libaries containining builtins in $PATH.
src/cmd/ksh93/tests/builtins.sh:
- Add test verifying a restricted user can use /opt/ast/bin/cat
via a PATH search.
Progresses: https://github.com/ksh93/ksh/issues/138
This commit fixes BUG_CSUBSTDO, which could break stdout inside of
non-forking command substitutions. The breakage only occurred when
stdout was closed outside of the command substitution and a file
descriptor other than stdout was redirected in the command substitution
(such as stderr). Thanks to the ast-open-history repo, I was able to
identify and backport the bugfix from ksh93v- 2012-08-24.
This backport may fix other bugs as well. On 93v- 2012-08-24 it
fixed the regression below, though it was not triggered on 93u+(m).
src/cmd/ksh93/tests/heredoc.sh
487 print foo > $tmp/foofile
488 x=$( $SHELL 2> /dev/null 'read <<< $(<'"$tmp"'/foofile) 2> /dev/null;print -r "$REPLY"')
489 [[ $x == foo ]] || err_exit '<<< $(<file) not working'
src/cmd/ksh93/sh/io.c: sh_open():
- If the just-opened file descriptor exists in sftable and is
flagged with SF_STRING (as in non-forking command substitutions,
among other situations), then move the file descriptor to a
number >= 10.
src/cmd/ksh93/tests/io.sh:
- Add a regression test for BUG_CSUBSTDO, adapted from the one in
modernish.
The current version of 93u+m does not have proper support for the
LC_TIME variable. Setting LC_TIME has no effect on printf %T, and
if the locale is invalid no error message is shown:
$ LC_TIME=ja_JP.UTF-8
$ printf '%T\n' now
Wed Apr 7 15:18:13 PDT 2021
$ LC_TIME=invalid.locale
$ # No error message
src/cmd/ksh93/data/variables.c,
src/cmd/ksh93/include/variables.h,
src/cmd/ksh93/sh/init.c:
- Add support for the $LC_TIME variable. ksh93v- attempted to add
support for LC_TIME, but the patch from that version was extended
because the variable still didn't function correctly.
src/cmd/ksh93/tests/variables.sh:
- Add LC_TIME to the regression tests for LC_* variables.
$ /usr/local/bin/ksh -c 'readonly v=1; export v'
/usr/local/bin/ksh: export: v: is read only
Every POSIX shell (even zsh, as of 5.8) allows this. So did ksh,
until the referenced commit.
src/cmd/ksh93/bltins/typeset.c: setall():
- Allow setting attributes on a readonly variable if any of
NV_ASSIGN (== NV_NOFREE), NV_EXPORT or NV_RDONLY are the only
flag bits that are set. This allows readonly, export, typeset -r,
typeset -x, and typeset -rx on variable arguments without an
assignment. Note that NV_ASSIGN is set for the first variable
argument even though it is not an assignment, so we must allow
it. The logic (or lack thereof) of that is yet to be worked out.
src/cmd/ksh93/tests/readonly.sh:
- Tests.
Resolves: https://github.com/ksh93/ksh/issues/258
The typecast fix was insufficient, avoiding the crash only when
compiling with optimisation disabled. The real problem is that
put_lineno() was passed a misaligned pointer, and that the value
didn't actually contain a double but a string. The bug occurred
when restoring the LINENO value upon exiting a virtual subshell.
Thanks to Harald van Dijk for figuring out the fix.
src/cmd/ksh93/sh/subshell.c: nv_restore():
- When restoring a special variable as defined by nv_cover(),
do not pass either the np->nvflag bits or NV_NOFREE. Why?
* The np->nvflag bits are not needed. They are also harmful
because they may include the NV_INTEGER bit. This is set
when the value is numeric. However, nv_getval() always
returns the value in string form, converting it if it is
numeric. So the NV_INTEGER flag should never be passed
to nv_putval() when it uses the result of nv_getval().
* According to nval.3, the NV_NOFREE flag stops nv_putval() from
creating a copy of the value. But this should be unnecessary
because the earlier _nv_unset(mp,NV_RDONLY|NV_CLONE) should
ensure there is no previous value. In addition, the NV_NOFREE
flag triggered another bug that caused the value of SECONDS to
be corrupted upon restoring it when exiting a virtual subshell.
- When restoring a regular variable, copy the entire nvalue union
and not just the 'cp' member. In practice this worked because
no current member of the nvalue union is larger than a pointer.
However, there is no guarantee it will stay that way.
src/cmd/ksh93/tests/leaks.sh:
- Add disabled test for a memory leak that was discovered in the
course of dealing with this bug. The fix doesn't introduce or
influence it. It will have to be dealt with later.
src/cmd/ksh93/tests/locale.sh:
- Add test for restoring locale on leaving virtual subshell.
https://github.com/ksh93/ksh/issues/253#issuecomment-815290154
src/cmd/ksh93/tests/variables.sh:
- Test against corruption of SECONDS on leaving virtual subshell.
https://github.com/ksh93/ksh/issues/253#issuecomment-815191052
Co-authored-by: Harald van Dijk <harald@gigawatt.nl>
Progresses: https://github.com/ksh93/ksh/issues/253
These fixes are applied rather blindly as no one has yet managed to
understand the almost entirely uncommented arrays and variables
handling code (arrays.c, name.c, nvdisc.c, nvtree.c, nvtype.c).
Hopefully we'll figure all that out at some point. In the meantime
these backported fixes appear to work fine, and these bugs impact
the usability of 'enum', so I'm just going to have to violate my
own policy and backport these fixes without understanding them.
Thanks to @JohnoKing for putting in a lot of work tracing these.
Further discussion at: https://github.com/ksh93/ksh/issues/87
src/cmd/ksh93/sh/array.c:
- nv_arraysettype():
* Further simplify the function. After my initial simplification
of it (re: 5491fe97), I don't believe there's actually a need
to save a duplicate copy of the value. Use the pointer returned
by nv_getval() directly to restore the value.
* Cope with a null value (nv_getval() returning a NULL pointer).
This is needed for compatibility with the backported fix in
nvtype.c (below).
- array_putval(): If the array's value pointer (up->cp) is a
pointer to the empty string, it is set to NULL before calling
nv_putv() to prevent an empty string from being deleted. Backport
a fix from 93v- that restores the pointer to the empty string if
the NV_NOFREE attribute is set. Removing it somehow causes these
regressions:
enum.sh[86]: ${array[@]} doesn't yield all values for
associative enum arrays (expected 'green blue blue red
yellow green red orange'; got 'green blue blue yellow
green orange')
enum.sh[94]: unsetting associative enum array does not work
(got 'Color_t -A Colors=([foo]=red [rood]=red)')
enum.sh[116]: assigning first enum element to indexed array
failed (expected 'red red'; got 'BUG BUG')
- nv_associative(): Do not increase the 'nelem' (number of
elements) value of the array's 'header' struct if the array is
associative and of an enum type. The original 93v- fix only
checked for the NV_INTEGER attribute, but backporting that caused
several regressions. Using a debug output command I've determined
that the exact value of 'type' is somehow consistently set to
0x26 if the array is associative and of an enum type, which is
NV_INTEGER | NV_LTOU | NV_RJUST as defined in include/nval.h. I
cannot find where/how that value is determined. In any case this
fix, based on but more specific than the 93v- one, appears to
work fine. Removing it somehow causes this regression:
enum.sh[94]: unsetting associative enum array does not work
(got 'Color_t -A Colors=()')
src/cmd/ksh93/sh/nvtype.c: nv_settype():
- Another fix backported from 93v-. If the variable is an array,
also set the type of element 0 of that array using a call to
nv_arraysettype(). The value may be null. Removing this somehow
causes this regression:
enum.sh[94]: unsetting associative enum array does not work
(got 'Color_t -A Colors=()')
src/cmd/ksh93/tests/enum.sh:
- Add tests for all the bugs fixed here, plus some hypothetical
bugs (e.g., do the same tests for indexed enum type arrays as for
associative enum type arrays, even though indexed enum type
arrays didn't have all the same problems).
Co-authored-by: Johnothan King <johnothanking@protonmail.com>
Resolves: https://github.com/ksh93/ksh/issues/87
Simple reproducer:
set -A arr a b c d; : ${arr[1..2]}; unset arr[1]; echo ${arr[@]}
Output:
a
Expected output:
a c d
The ${arr[1..2]} expansion broke the subsequent 'unset' command
so that it unsets element 1 and on, instead of only 1.
This regression was introduced in nv_endsubscript() on 2009-07-31:
https://github.com/ksh93/ast-open-history/commit/c47896b4/src/cmd/ksh93/sh/array.c
That change checks for the ARRAY_SCAN attribute which enables
processing ranges of array elements instead of single array
elements, and restores it after. That restore is evidently not
correct as it causes the subsequent unset command to malfunction.
If we revert that change, the bug disappears and the regression
tests show no failures. However, I don't know what this was meant
to accomplish and what other bug we might introduce by reverting
this. However, no corresponding regression test was added along
with the 2009-07-31 change, nor is there any corresponding message
in the changelog. So this looks to be one of those mystery changes
that we'll never know the reason for.
Since we currently have proof that this change causes breakage and
no evidence that it fixes anything, I'll go ahead and revert it
(and add a regression test, of course). If that causes another
regression, hopefully someone will find it at some point.
src/cmd/ksh93/sh/array.c: nv_endsubscript():
- Revert the 2009-07-31 change that saves/restores the ARRAY_SCAN
attribute.
- Keep the 'ap' pointer as it is now used by newer code. Move the
declaration up to the beginning of the block, as is customary.
src/cmd/ksh93/sh/init.c:
- Cosmetic change: remove an unused array_scan() macro that I found
when grepping the code for ARRAY_SCAN. The macro was introduced
in version 2001-06-01 but the code that used it was replaced in
version 2001-07-04, without removing the macro itself.
Resolves: https://github.com/ksh93/ksh/issues/254
Ksh currently restricts readonly scalar variables from having their
values directly changed via a value assignment. However, since ksh
allows variable attributes to be altered, the variable's value can
be indirectly altered. For instance, if TMOUT=900 (for a 15 minute
idle timeout) was set to readonly, all that is needed to alter the
value of TMOUT from 900 to 0 is to issue 'typeset -R1 TMOUT',
perhaps followed by a 'typeset -i TMOUT' to turn off the shell's
timeout value.
In addition, there are problems with arrays. The following is
incorrectly allowed:
typeset -a arr=((a b c) 1)
readonly arr
arr[0][1]=d
arr=(alphas=(a b c);name=x)
readonly arr.alphas
arr.alphas[1]=([b]=5)
arr=(alphas=(a b c);name=x)
readonly arr.alphas
arr.alphas[1]=(b)
typeset -C arr=(typeset -r -a alphas=(a b c);name=x)
arr.alphas[1]=()
src/cmd/ksh93/bltins/typeset.c: setall():
- Relocate readonly attribute check higher up the code and widen
its application to issue an error message if the pre-existing
name-pair has the readonly bit flag set.
- To avoid compatibility problems, don't check for readonly if
NV_RDONLY is the only attribute set (ignoring NV_NOFREE). This
allows 'readonly foo; readonly foo' to keep working.
src/cmd/ksh93/sh/array.c: nv_endsubscript():
- Apply a readonly flag check when an array subscript or append
assignment occurs, but allow type variables (typeset -T) as they
utilize '-r' for 'required' sub-variables.
src/cmd/ksh93/tests/readonly.sh:
- New file. Create readonly tests that validate the warning message
and validate that the readonly variable did not change.
src/cmd/ksh93/sh/streval.c:
- Bump MAXLEVEL from 9 to 1024 as a workaround for arithmetic
expansion, avoiding a spurious error about too much recursion
when the readonly.sh tests are run. This change is backported
from ksh 93v-.
TODO: debug a spurious increase in arithmetic recursion level
variable when readonly.sh tests with 'typeset -i' are run.
That is a different bug for a different commit.
Co-authored-by: Martijn Dekker <martijn@inlv.org>
This commit adds an UNREACHABLE() macro that expands to either the
__builtin_unreachable() compiler builtin (for release builds) or
abort(3) (for development builds). This is used to mark code paths
that are never to be reached.
It also adds the 'noreturn' attribute to functions that never
return: path_exec(), sh_done() and sh_syntax(). The UNREACHABLE()
macro is not added after calling these.
The purpose of these is:
* to slightly improve GCC/Clang compiler optimizations;
* to fix a few compiler warnings;
* to add code clarity.
Changes of note:
src/cmd/ksh93/sh/io.c: outexcept():
- Avoid using __builtin_unreachable() here since errormsg can
return despite using ERROR_system(1), as shp->jmplist->mode is
temporarily set to 0. See: https://github.com/att/ast/issues/1336
src/cmd/ksh93/tests/io.sh:
- Add a regression test for the ksh2020 bug referenced above.
src/lib/libast/features/common:
- Detect the existence of either the C11 stdnoreturn.h header or
the GCC noreturn attribute, preferring the former when available.
- Test for the existence of __builtin_unreachable(). Use it for
release builds. On development builds, use abort() instead, which
crahses reliably for debugging when unreachable code is reached.
Co-authored-by: Martijn Dekker <martijn@inlv.org>
This commit fixes a bug in the ksh uname builtin's -d option that could
change the output of -o (I was only able to reproduce this on Linux):
$ builtin uname
$ uname -o
GNU/Linux
$ uname -d
(none)
$ uname -o
(none)
I identified this patch from ksh2020 as a fix for this bug:
<https://github.com/att/ast/pull/1187>
The linked patch was meant to fix a crash in 'uname -d', although I've
had no luck reproducing it: <https://github.com/att/ast/issues/1184>
src/lib/libcmd/uname.c:
- Pass correct buffer to getdomainname() while executing uname -d.
src/cmd/ksh93/tests/builtins.sh:
- Add a regression test for the reported 'uname -d' crash.
- Add a regression test for the output of 'uname -o' after 'uname -d'.
- To handle potential crashes when running the regression tests in older
versions of ksh, fork the command substitutions that run 'uname -d'.
This bug was first reported at <https://github.com/att/ast/issues/8>.
The 'cd' command currently takes the value of $OLDPWD from the
wrong scope. In the following example 'cd -' will change the
directory to /bin instead of /tmp:
$ OLDPWD=/bin ksh93 -c 'OLDPWD=/tmp cd -'
/bin
src/cmd/ksh93/bltins/cd_pwd.c:
- Use sh_scoped() to obtain the correct value of $OLDPWD.
- Fix a use-after-free bug. Make the 'oldpwd' variable a static
char that points to freeable memory. Each time cd is used, this
variable is freed if it points to a freeable memory address and
isn't also a pointer to shp->pwd.
src/cmd/ksh93/sh/path.c: path_pwd():
- Simplify and add comments.
- Scope $PWD properly.
src/cmd/ksh93/tests/builtins.sh,
src/cmd/ksh93/tests/leaks.sh:
- Backport the ksh2020 regression tests for 'cd -' when $OLDPWD is
set.
- Add test for $OLDPWD and $PWD after subshare.
- Add test for $PWD after 'cd'.
- Add test for possible memory leak.
- Add testing for 'unset' on OLDPWD and PWD.
src/cmd/ksh93/COMPATIBILITY:
- Add compatibility note about changes to $PWD and $OLDPWD.
Co-authored-by: Martijn Dekker <martijn@inlv.org>
src/cmd/ksh93/tests/variables.sh: LC_* error tests:
- Since operating systems validate locale strings differently,
try a few different bad locale strings to find one that makes
setlocale(2) fail, fixing test failures on OpenBSD and Debian.
- Restore warning removed in aed5c6d7, issuing it if none of the
bad locale strings produce a diagnostic.
- Reenable test for diagnostic message disabled in aed5c6d7.
Co-authored-by: Martijn Dekker <martijn@inlv.org>
This bug was originally reported at <https://github.com/att/ast/issues/1467>.
A crash can occur when using the 'b' or 'B' vi mode commands to go back
one word. I was able to reproduce these crashes with 100% consistency on
an OpenBSD virtual machine when ksh is compiled with -D_std_malloc.
Reproducer:
$ set -o vi
$ asdf <ESC> <b or B>
The fix is based on Matthew DeVore's analysis:
> I suspect this is caused by this line:
>> while (vi_isalph(tcur_virt) && tcur_virt >= first_virt) --tcur_virt;
> which is in the b codepath. It checks vi_isalph(tcur_virt) before checking
> if tcur_virt is in range. These two clauses should be reversed. Note that
> line 316 is a similar check for pressing B, and there the tcur_virt value
> is checked first.
src/cmd/ksh93/edit/vi.c:
- Check tcur_virt before using isalph() or isblank() to fix both crashes.
At the start of the backword() while loop this check was performed
twice, so the redundant check has been removed.
src/cmd/ksh93/tests/pty.sh:
- Add a regression test for the b, B, w and W editor commands.
src/cmd/ksh93/bltins/test.c:
- Fix the following compiler warnings from clang:
test.c:554:11: warning: assigning to 'char *' from 'const char []'
discards qualifiers
[-Wincompatible-pointer-types-discards-qualifiers]
e_msg = e_badop;
^ ~~~~~~~
test.c:556:11: warning: assigning to 'char *' from 'const char []'
discards qualifiers
[-Wincompatible-pointer-types-discards-qualifiers]
e_msg = e_unsupported_op;
^ ~~~~~~~~~~~~~~~~
test.c:560:1: warning: control may reach end of non-void function
[-Wreturn-type]
src/cmd/ksh93/tests/builtins.sh:
- Fix regression test by updating error message text.
When test is passed the '=~' operator, it will silently fail with
exit status 1:
$ test foo =~ foo; echo $?
1
This bug is caused by test_binop reaching the 'NOTREACHED' area of
code. The bugfix was adapted from ksh2020:
https://github.com/att/ast/issues/1152
src/cmd/ksh93/bltins/test.c: test_binop():
- Error out with a message suggesting usage of '[[ ... ]]' if '=~'
is passed to the test builtin.
- Special-case TEST_END (']]') as that is not really an operator.
Co-authored-by: Martijn Dekker <martijn@inlv.org>
src/lib/libast/tm/tminit.c:
- Commit 9f43f8d1, in addition to backporting fixes from ksh93v-, also
backported this bug:
$ printf '%(%Z)T' now
PPT # Should be PDT
Reapply the ksh2020 bugfix to fix the %Z time
format again.
src/cmd/ksh93/tests/builtins.sh:
- Add a regression test so this bug (hopefully) isn't backported from
ksh93v- again).
Every so often, a commit's GitHub CI run throws the following
regression test failure:
sigchld.sh[57]: expected '2 background' -- got '3' (DELAY=0.02)
When I re-run the job, the failure usually goes away.
In 712261c8 the DELAY variable was changed from 0.2 to 0.02 to
speed up the first SIGCHLD test. It's possible the GitHub CI
runners are just too slow or too heavily loaded for that.
src/cmd/ksh93/tests/sigchld.sh:
- Restore 0.2 value for 'float DELAY'.
Moving the 'err_exit' and 'warning' alias definitions in the
regression tests to one _common file introduced a bug: they are no
longer expanded at compile time when the tests are run with shcomp,
resulting in a 'command not found' (at best) on trying to execute
one. shcomp requires that the alias definitions need to be present
in the file itself. But that means maintaining 50-odd copies again.
I'd rather add a hack to shtests to avoid this.
src/cmd/ksh93/tests/shtests:
- Before running a test with shcomp, physically concatenate _common
and the test script together into a temporary file, minus the '.'
command that includes _common, and compile that with shcomp.
These are minor fixes I've accumulated over time. The following
changes are somewhat notable:
- Added a missing entry for 'typeset -s' to the man page.
- Add strftime(3) to the 'see also' section. This and the date(1)
addition are meant to add onto the documentation for 'printf %T'.
- Removed the man page the entry for ksh reading $PWD/.profile on
login. That feature was removed in commit aa7713c2.
- Added date(1) to the 'see also' section of the man page.
- Note that the 'hash' command can be used instead of 'alias -t' to
workaround one of the caveats listed in the man page.
- Use an 'out of memory' error message rather than 'out of space'
when memory allocation fails.
- Replaced backticks with quotes in some places for consistency.
- Added missing documentation for the %P date format.
- Added missing documentation for the printf %Q and %p formats
(backported from ksh2020: https://github.com/att/ast/pull/1032).
- The comments that show each builtin's options have been updated.
This was failing again on FreeBSD. Replicating the test in a real
session worked as expected.
Apparently, we just cannot rely on external 'vi' utilities playing
well with pty. This test has caused enough trouble. Removed.
Upon encountering two filenames with multibyte characters starting
with the same byte, a partial multibyte character was completed.
Reproducer (to run in UTF-8 locale):
$ touch XXXá XXXë
$ : XX <== pres tab
$ : XXX^? <== partial multibyte character appears
Note: á is $'\xc3\xa1' and ë is $'\xc3\xab' (same initial byte).
src/cmd/ksh93/edit/completion.c:
- Add multibyte support to the charcmp() and overlaid() functions.
Thanks to Harald van Dijk for useful code and suggestions.
- Add a few missing mbinit() calls. The state of multibyte
processing must be reset before starting a new loop in case a
previous processing run was interrupted mid-character.
src/cmd/ksh93/tests/pty.sh:
- Add test based on Harald's reproducer.
Resolves: https://github.com/ksh93/ksh/issues/223
Until now, when performing any tilde expansion like ~/foo or
~user/foo, ksh added a placeholder built-in command called
'.sh.tilde', ostensibly with the intention to allow users to
override it with a shell function or custom builtin. The multishell
ksh93 repo <https://github.com/multishell/ksh93/> shows this was
added sometime between 2002-06-28 and 2004-02-29. However, it has
never worked and crashed the shell.
This commit replaces that with something that works. Specific tilde
expansions can now be overridden using .set or .get discipline
functions associated with the .sh.tilde variable (see manual,
Discipline Functions).
For example, you can use either of:
.sh.tilde.set()
{
case ${.sh.value} in
'~tmp') .sh.value=${XDG_RUNTIME_DIR:-${TMPDIR:-/tmp}} ;;
'~doc') .sh.value=~/Documents ;;
'~ksh') .sh.value=/usr/local/src/ksh93/ksh ;;
esac
}
.sh.tilde.get()
{
case ${.sh.tilde} in
'~tmp') .sh.value=${XDG_RUNTIME_DIR:-${TMPDIR:-/tmp}} ;;
'~doc') .sh.value=~/Documents ;;
'~ksh') .sh.value=/usr/local/src/ksh93/ksh ;;
esac
}
src/cmd/ksh93/include/variables.h,
src/cmd/ksh93/data/variables.c:
- Add SH_TILDENOD for a new ${.sh.tilde} predefined variable.
It is initially unset.
src/cmd/ksh93/sh/macro.c:
- sh_btilde(): Removed.
- tilde_expand2(): Rewritten. I started out with the tiny version
of this function from the 2002-06-28 version of ksh. It uses the
stack instead of sfio, which is more efficient. A bugfix for
$HOME == '/' was retrofitted so that ~/foo does not become
//foo instead of /foo. The rest is entirely new code.
To implement the override functionality, it now checks if
${.sh.tilde} has any discipline function associated with it.
If it does, it assigns the tilde expression to ${.sh.tilde} using
nv_putval(), triggering the .set discipline, and then reads it
back using nv_getval(), triggering the .get discipline. The
resulting value is used if it is nonempty and does not still
start with a tilde.
src/cmd/ksh93/bltins/typeset.c,
src/cmd/ksh93/tests/builtins.sh:
- Since ksh no longer adds a dummy '.sh.tilde' builtin, remove the
ad-hoc hack that suppressed it from the output of 'builtin'.
src/cmd/ksh93/tests/tilde.sh:
- Add tests verifying everything I can think of, as well as tests
for bugs found and fixed during this rewrite.
src/cmd/ksh93/tests/pty.sh:
- Add test verifying that the .sh.tilde.set() discipline does not
modify the exit status value ($?) when performing tilde expansion
as part of tab completion.
src/cmd/ksh93/sh.1:
- Instead of "tilde substitution", call the basic mechanism "tilde
expansion", which is the term used everywhere else (including the
1995 Bolsky/Korn ksh book).
- Document the new override feature.
Resolves: https://github.com/ksh93/ksh/issues/217
In ksh93r a crash can occur after switching from emacs mode to vi
mode[*]:
$ ENV=/./dev/null ksh2006 -o emacs
$ echo ${.sh.version}
Version M 1993-12-28 r
$ set -o vi
$ <Esc> <r> <r> # This triggers the memory fault
Commit 129614b9 added the OpenSUSE patch for this crash. This commit
adds the regression test for it.
[*]: https://bugzilla.opensuse.org/show_bug.cgi?id=179917
A discipline function could incorrectly influence the value of $?
(exit status of last command) outside its context if it was
triggered without another command being run, e.g. when a prompt
variable is read, or COLUMNS or LINES is set.
Reproducers include:
PS1 prompt:
$ PS1.get() { true; }
$ false
$ echo $?
0
PS2 prompt:
$ PS2.get() { return 13; }
$ \
>
$ echo $?
13
The set discipline is affected too, e.g. COLUMNS and LINES:
$ COLUMNS.set() { return 13; }
$ true
$ (press return)
$ echo $?
13
There are probably other contexts where the shell reads or changes
variables without running commands, allowing their get or set
disciplines to influence $?. So this commit makes ksh save $? for
all .get, .set, .append, and .unset discipline calls.
src/cmd/ksh93/sh/nvdisc.c:
- assign(): Save/restore $? when running a .set/.append/.unset
discipline function.
- lookup(): Save/restore $? when running a .get discipline.
src/cmd/ksh93/tests/pty.sh:
- Add a regression test for $? after displaying a prompt
and when setting a LINES.set discipline function.
src/cmd/ksh93/tests/return.sh:
- The above test fails in script form on ksh93u+ and ksh2020, as
it exposes another form of #117 that occurs after running a
subshell. Add the above regression test here as well
(re: 092b90da).
Co-authored-by: Martijn Dekker <martijn@inlv.org>
For most numeric types the last provided one wins out. This commit
closes the gap for -F and -i numerics to not be covered up by other
preceding float types. Note: -u for requesting an unsigned float or
integer was considered and decided to be left alone as it stands,
so as to not allow the variable to become an uppercased string if
the requested options ended with a -u. As it stands for a case when
multiple numeric types are requested, a -u option may be applied
after the last numeric type is processed.
Examples:
-EF becomes -F
-Fi becomes -i
-Fu becomes -F
-uF becomes -F
-Fui becomes -i (because isfloat==1, unsigned is not applied)
-Fiu becomes -iu (isfloat is reset and allows unsigned to be set)
src/cmd/ksh93/bltins/typeset.c: b_typeset():
- Reset attribute bit flags for -E and -X when -F is requested by
adding in NV_EXPNOTE to be removed.
- For -i option if a float precedes it, reset isfloat and -E/-F
attribute bit flags.
- Take into account the impact of the shortint flag on floats.
src/cmd/ksh93/tests/attributes.sh:
- Add some validation tests to confirm that, when a -F follows
either -E or -X, -F is used.
- Add some validation tests to confirm that, when -F/E/X precede
a -i, the variable becomes an integer and not a float.
- Add in various tests when -s followed a float.
There was an issue with tilde expansion if the HOME var is unset.
$ unset HOME
$ echo ~
martijn
Only the username is returned. Users are more likely to expect the
current user's home directory as configured in the OS.
POSIXly, the expansion of ~ is based on the value of HOME. If HOME
is unset, the results are unspecified. After unsetting HOME, in
bash, ~ returns the user's home directory as specified by the OS,
whereas in all other shells, ~ expands to the empty string. Only
ksh93 returns the username. The behaviour of bash is more useful.
Discussion:
https://github.com/ksh93/ksh/pull/225#issuecomment-799074107
src/cmd/ksh93/sh/macro.c,
src/cmd/ksh93/tests/tilde.sh:
- sh_tilde(): Backport fix by Mike Gilbert from ksh2020.
See: https://github.com/att/ast/issues/1391https://github.com/att/ast/pull/1396https://github.com/att/ast/commit/070d365d
- Add test.
src/cmd/ksh93/COMPATIBILITY:
- Note this change.
src/cmd/ksh93/tests/_common:
- Commit aed5c6d7 renamed the err_exit function,
breaking a few tests in glob.sh that call the function
directly instead of using the alias. Restore the function.
src/cmd/ksh93/tests/builtins.sh:
- The dtksh builtins don't have optget option parsing, so
skip the unrecognized options test for those (this of
course only has relevance when running dtksh against the
regression tests).
src/cmd/ksh93/tests/pty.sh:
- If the vi editor couldn't be found on the $PATH, skip the
regression test that involves it.
src/cmd/ksh93/tests/_common:
- Added. This keeps one common version of 'err_exit', 'warning',
and other init code.
src/cmd/ksh93/tests/*.sh:
- Source _common as a dot script.
- Remove 50-odd, occasionally slightly different, versions of the
common code.
- Some minor tweaks.
This commit fixes a long-standing bug (present since at least
ksh93r) that caused a file descriptor leak when passing a process
substitution to a function, or (if compiled with SHOPT_SPAWN) to a
nonexistent command.
The leaks only occurred when ksh was compiled with SHOPT_DEVFD; the
FIFO method was unaffected.
src/cmd/ksh93/sh/xec.c: sh_exec():
- When a process substitution is passed to a built-in, the
remaining file descriptor is closed with sh_iorestore. Do the
same thing when passing a process substitution to a function.
This is done by delaying the sh_iorestore() call to 'setexit:'
where both built-ins and functions terminate and set the exit
status ($?).
This means that call now will not be executed if a longjmp is
done, e.g. due to an error in a special built-in. However, there
is already another sh_iorestore() call in main.c, exfile(), line
418, that handles that scenario.
- sh_ntfork() can fail, so rather than assume it will succeed,
handle a failure by closing extra file descriptors with
sh_iorestore(). This fixes the leak on command not found with
SHOPT_SPAWN.
src/cmd/ksh93/include/defs.h:
- Since the file descriptor leaks are now fixed, remove the
workaround that forced ksh to use the FIFO method.
src/cmd/ksh93/SHOPT.sh:
- Add SHOPT_DEVFD as a configurable option (default: probe).
src/cmd/ksh93/tests/io.sh:
- Add a regression test for the 'not found' file descriptor leak.
- Add a test to ensure it keeps working with 'command'.
Fixes: https://github.com/ksh93/ksh/issues/67
src/cmd/ksh93/tests/arrays.sh,
src/cmd/ksh93/tests/arrays2.sh:
- Backport some regression tests from ksh93v- for associative
arrays.
src/cmd/ksh93/tests/basic.sh:
- Add ksh93v- regression tests for background process output in
backtick and shared-state command substitutions as well as
functions used in command substitutions.
- Add regression tests for using EXIT traps in subshells. In
ksh93v- and ksh2020 EXIT traps don't work in forked subshells:
https://github.com/att/ast/issues/1452
- The trap builtin shouldn't segfault after receiving an invalid
signal name. ksh2020 regression:
https://github.com/att/ast/issues/1403
- Add a test to make sure invalid flags don't crash ksh.
ksh2020 regression: https://github.com/att/ast/issues/1284
- Test for an illegal seek error when using the 'join' command with
process substitutions. ksh93v- regression:
https://www.mail-archive.com/ast-users@lists.research.att.com/msg00816.html
src/cmd/ksh93/tests/bracket.sh:
- Add some regression tests from ksh93v- for the -eq test operator.
src/cmd/ksh93/tests/builtins.sh:
- Move the regression test for 'exit' in an interactive shell to
the exit.sh script.
- Test for assignments preceding the command builtin persisting
after an error. ksh2020 regression:
https://github.com/att/ast/issues/1402
- The chmod builtin should modify the permissions of all files
passed to it. ksh2020 regression:
https://github.com/att/ast/issues/949
- Add regression tests for the cd builtin. In ksh93v- 2013-10-10
alpha, using cd on a directory without an execute bit doesn't
cause an error. The test for using cd on a normal file was
backported from ksh93v-.
- Backport a ksh93v- regression test for the exit status
from 'kill %'.
src/cmd/ksh93/tests/case.sh:
- Test for a segfault when ksh handles an invalid character class
in a pattern. ksh2020 regression:
https://github.com/att/ast/issues/1409
src/cmd/ksh93/tests/exit.sh:
- Add regression tests from ksh2020 for the 'exit' builtin:
https://github.com/att/ast/commit/d9491d46
src/cmd/ksh93/tests/io.sh:
- Add a regression test from ksh93v- for a process substitution
hang. This test fails in the 93v- 2013 alpha but succeeds in
the 2014 beta.
src/cmd/ksh93/tests/math.sh:
- 'typeset -s foo=30000' adds garbage to $foo in ksh93u+, ksh93v-
and ksh2020:
$ typeset -s foo=30000
$ echo $foo
5#1430000
This bug was fixed in commit 88a6baa1, but that commit didn't
add a regression test for it.
src/cmd/ksh93/tests/variables.sh:
- Add a regression test for $PS4 incorrectly unsetting
${.sh.subshell}: https://github.com/att/ast/issues/1092
This is an attempt to avoid fairly rare intermittent failures
on the GitHub CI runners. Apparently, they are sometimes so
slow that typeahead can still interfere with a test.
On systems where ksh needs to use the older and less secure FIFO
method for process substitutions (which is currently all of them as
the more modern and solid /dev/fd method is still broken, see #67),
process substitutions could leave background processes hanging in
these two scenarios:
1. If the parent process exits without opening a pipe to the child
process forked by the process substitution. The fifo_check()
function in xec.c, which is periodically called to check if the
parent process still exists while waiting for it to open the
FIFO, verified the parent process's existence by checking if the
PPID had reverted to 1, the traditional PID of init. However,
POSIX specifies that the PPID can revert to any implementation-
defined system process in that case. So this breaks on certain
systems, causing unused process substitutions to hang around
forever as they never detect that the parent disappeared.
The fix is to save the current PID before forking and having the
child check if the PPID has changed from that saved PID.
2. If command invoked from the main shell is passed a process
substitution, but terminates without opening the pipe to the
process substitution. In that case, the parent process never
disappears in the first place, because the parent process is the
main shell. So the same infinite wait occurs in unused process
substitutions, even after correcting problem 1.
The fix is to remember all FIFOs created for any number of
process substitutions passed to a single command, and unlink any
remaining FIFOs as they represent unused command substitutions.
Unlinking them FIFOs causes sh_open() in the child to fail with
ENOENT on the next periodic check, which can easily be handled.
Fixing these problems causes the FIFO method to act identically to
the /dev/fd method, which is good for compatibility. Even when #67
is fixed this will still be important, as ksh also runs on systems
that do not have /dev/fd (such as AIX, HP-UX, and QNX), so will
fall back to using FIFOs.
--- Fix problem 1 ---
src/cmd/ksh93/sh/xec.c:
- Add new static fifo_save_ppid variable.
- sh_exec(): If a FIFO is defined, save the current PID in
fifo_save_ppid for the forked child to use.
- fifo_check(): Compare PPID against the saved value instead of 1.
--- Fix problem 2 ---
To keep things simple I'm abusing the name-value pair routines used
for variables for this purpose. The overhead is negligible. A more
elegant solution is possible but would involve adding more code.
src/cmd/ksh93/include/defs.h: _SH_PRIVATE:
- Define new sh.fifo_tree pointer to a new FIFO cleanup tree.
src/cmd/ksh93/sh/args.c: sh_argprocsubs():
- After launching a process substitution in the background,
add the FIFO to the cleanup list before freeing it.
src/cmd/ksh93/sh/xec.c:
- Add fifo_cleanup() that unlinks all FIFOs in the cleanup list and
clears/closes the list. They should only still exist if the
command never used them, however, just run 'unlink' and don't
check for existence first as that would only add overhead.
- sh_exec():
* Call fifo_cleanup() on finishing all simple commands (when
setting $?) or when a special builtin fails.
* When forking, clear/close the cleanup list; we do not want
children doing duplicate cleanup, particularly as this can
interfere when using multiple process substitutions in one
command.
* Process substitution handling:
> Change FIFO check frequency from 500ms to 50ms.
Note that each check sends a signal that interrupts open(2),
causing sh_open() to reinvoke it. This causes sh_open() to
fail with ENOENT on the next check when the FIFO no longer
exists, so we do not need to add an additional check for
existence to fifo_check(). Unused process substitutions now
linger for a maximum of 50ms.
> Do not issue an error message if errno == ENOENT.
- sh_funct(): Process substitutions can be passed to functions as
well, and we do not want commands within the function to clean up
the FIFOs for the process substitutions passed to it from the
outside. The problem is solved by simply saving fifo_tree in a
local variable, setting it to null before running the function,
and cleaning it up before restoring the parent one at the end.
Since sh_funct() is called recursively for multiple-level
function calls, this correctly gives each function a locally
scoped fifo_tree.
--- Tests ---
src/cmd/ksh93/tests/io.sh:
- Add tests covering the failing scenarios.
Co-authored-by: Martijn Dekker <martijn@inlv.org>
This commit fixes two interrelated problems.
1. The -v unary test/[/[[ operator is documented to test if a
variable is set. However, it always returns true for variable
names with a numeric attribute, even if the variable has not
been given a value. Reproducer:
$ ksh -o nounset -c 'typeset -i n; [[ -v n ]] && echo $n'
ksh: n: parameter not set
That is clearly wrong; 'echo $n' should never be reached and the
error should not occur, and does not occur on mksh or bash.
2. Fixing the previous problem revealed serious breakage in short
integer type variables that was being masked. After applying
that fix and then executing 'typeset -si var=0':
- The conditional assignment expansions ${var=123} and
${var:=123} assigned 123 to var, even though it was set to 0.
- The expansions ${var+s} and ${var:+n} incorrectly acted as if
the variable was unset and empty, respectively.
- '[[ -v var ]]' and 'test -v var' incorrectly returned false.
The problems were caused by a different storage method for short
ints. Their values were stored directly in the 'union Value'
member of the Namval_t struct, instead of allocated on the stack
and referred to by a pointer, as regular integers and all other
types do. This inherently broke nv_isnull() as this leaves no
way to distinguish between a zero value and no value at all.
(I'm also pretty sure it's undefined behaviour in C to check for
a null pointer at the address where a short int is stored.)
The fix is to store short ints like other variables and refer
to them by pointers. The NV_INT16P combined bit mask already
existed for this, but nv_putval() did not yet support it.
src/cmd/ksh93/bltins/test.c: test_unop():
- Fix problem 1. For -v, only check nv_isnull() and do not check
for the NV_INTEGER attribute (which, by the way, is also used
for float variables by combining it with other bits).
See also 5aba0c72 where we recently fixed nv_isnull() to
work properly for all variable types including short ints.
src/cmd/ksh93/sh/name.c: nv_putval():
- Fix problem 2, part 1. Add support for NV_INT16P. The code is
simply copied and adapted from the code for regular integers, a
few lines further on. The regular NV_SHORT code is kept as this
is still used for some special variables like ${.sh.level}.
src/cmd/ksh93/bltins/typeset.c: b_typeset():
- Fix problem 2, part 2. Use NV_INT16P instead of NV_SHORT.
src/cmd/ksh93/tests/attributes.sh:
- Add set/unset/empty/nonempty tests for all numeric types.
src/cmd/ksh93/tests/bracket.sh,
src/cmd/ksh93/tests/comvar.sh:
- Update a couple of existing tests.
- Add test for [[ -v var ]] and [[ -n ${var+s} ]] on unset
and empty variables with many attributes.
src/cmd/ksh93/COMPATIBILITY:
- Add a note detailing the change to test -v.
src/cmd/ksh93/data/builtins.c,
src/cmd/ksh93/sh.1:
- Correct 'typeset -C' documentation. Variables declared as
compound are *not* initially unset, but initially have the empty
compound value. 'typeset' outputs them as:
typeset -C foo=()
and not:
typeset -C foo
and nv_isnull() is never true for them. This may or may not
technically be a bug. I don't think it's worth changing, but
it should at least be documented correctly.
These expansions are supposed to yield all variable names beginning
with the indicated prefix. This should include the variable name
that is identical to the prefix (as 'prefix' begins with 'prefix').
This bugfix is backported from the abandoned ksh 93v- beta, so AT&T
intended this change. It also makes ksh work like bash in this.
src/cmd/ksh93/sh/macro.c: varsub(): M_NAMESCAN:
- Check if the prefix itself exists. If so, start with that.
src/cmd/ksh93/tests/variables.sh:
- Add tests for these expansions.
src/cmd/ksh93/sh.1:
- Fix the incomplete documentation of these expansions.
src/cmd/ksh93/COMPATIBILITY:
- Note the change as it's potentially incompatible in corner cases.
Resolves: https://github.com/ksh93/ksh/issues/183
This commit fixes at least three bugs:
1. When issuing 'typeset -p' for unset variables typeset as short
integer, a value of 0 was incorrectly diplayed.
2. ${x=y} and ${x:=y} were still broken for short integer types
(re: 9f2389ed). ${x+set} and ${x:+nonempty} were also broken.
3. A memory fault could occur if typeset -l followed a -s option
with integers. Additonally, now the last -s/-l wins out as the
option to utilize instead of it always being short.
src/cmd/ksh93/include/name.h:
- Fix the nv_isnull() macro by removing the direct exclusion of
short integers from this set/unset test. This breaks few things
(only ${.sh.subshell} and ${.sh.level}, as far as we can tell)
while potentially correcting many aspects of short integer use
(at least bugs 1 and 2 above), as this macro is widely used.
- union Value: add new pid_t *pidp pointer member for PID values
(see further below).
src/cmd/ksh93/bltins/typeset.c: b_typeset():
- To fix bug 3 above, unset the 'shortint' flag and NV_SHORT
attribute bit upon encountering the -l optiobn.
*** To fix ${.sh.subshell} to work with the new nv_isnull():
src/cmd/ksh93/sh/defs.h:
- Add new 'realsubshell' member to the shgd (aka shp->gd) struct
which will be the integer value for ${.sh.subshell}.
src/cmd/ksh93/sh/init.c,
src/cmd/ksh93/data/variables.c:
- Initialize SH_SUBSHELLNOD as a pointer to shgd->realsubshell
instead of using a short value (.s) directly. Using a pointer
allows nv_isnull() to return a positive for ${.sh.subshell} as
a non-null pointer is what it checks for.
- While we're at it, initialize PPIDNOD ($PPID) and SH_PIDNOD
(${.sh.pid}) using the new pdip union member, which is more
correct as they are values of type pid_t.
src/cmd/ksh93/sh/subshell.c,
src/cmd/ksh93/sh/xec.c:
- Update the ${.sh.subshell} increases/decreases to refer to
shgd->realsubshell (a.k.a. shp->gd->realsubshell).
*** To fix ${.sh.level} after changing nv_isnull():
src/cmd/ksh93/sh/macro.c: varsub():
- Add a specific exception for SH_LEVLNOD to the nv_isnull() test,
so that ${.sh.level} is always considered to be set. Its handling
throughout the code is too complex/special for a simple fix, so
we have to special-case it, at least for now.
*** Regression test additions:
src/cmd/ksh93/tests/attributes.sh:
- Add in missing short integer tests and correct the one that
existed. The -si test now yields 'typeset -x -r -s -i foo'
instead of 'typeset -x -r -s -i foo=0' which brings it in line
with all the others.
- Add in some other -l attribute tests for floats. Note, -lX test
was not added as the size of long double is platform dependent.
src/cmd/ksh93/tests/variables.sh:
- Add tests for ${x=y} and ${x:=y} used on short int variables.
Co-authored-by: Martijn Dekker <martijn@inlv.org>
The fix for '.' and '..' in regular globbing broke '.' and '..' in
globstar. No globstar pattern that contains '.' or '..' as any
pathname component still matched. This commit fixes that.
This commit also makes symlink/** mostly work, which it never has
done in any ksh93 version. It is correct and expected that symlinks
found by patterns are not resolved, but symlinks were not resolved
even when specified as explicit non-pattern pathname components.
For example, /tmp/** breaks if /tmp is a symlink (e.g. on macOS),
which looks like a bug.
src/lib/libast/include/glob.h,
src/lib/libast/misc/glob.c: glob_dir():
- Make symlink/** work. we can check if the string pointed to by
pat is exactly equal to *. If so, we are doing regular globbing
for that particular pathname element, and it's okay to resolve
symlinks. If not (if it's **), we're doing globstar and we should
not be matching symlinks.
- Let's also introduce proper identification of symlinks (GLOB_SYM)
and not lump them in with other special files (GLOB_DEV).
- Fix the bug with literal '.' and '..' components in globstar
patterns. In preceding code, the matchdir pointer gets set to the
complete glob pattern if we're doing globstar for the current
pathname element, null if not. The pat pointer gets set to the
elements of the pattern that are still left to be processed;
already-done elements are trimmed from it by increasing the
pointer. So, to do the right thing, we need to make sure that '.'
or '..' is skipped if, and only if, it is the final element in
the pattern (i.e., if pat does not contain a slash) and is not
specified literally as '.' or '..', i.e., only if '.' or '..' was
actually resolved from a glob pattern. After this change,
'**/.*', '**/../.*', etc. do the right thing, showing all your
hidden files and directories without undesirable '.' and '..'
results; '.' and '..' are skipped as final elements, unless you
literally specify '**/.', '**/..', '**/foo/bar/..', etc.
src/cmd/ksh93/COMPATIBILITY:
- Note the symlink/** globstar change.
src/cmd/ksh93/sh.1:
- Try to document the current globstar behaviour more exhausively.
src/cmd/ksh93/tests/glob.sh:
- Add tests. Try to cover all the corner cases.
src/cmd/ksh93/tests/shtests:
- Since tests in glob.sh do not use err_exit, they were not
counted. Special-case glob.sh for counting the tests: count the
lines starting with a test_* function call.
Resolves: https://github.com/ksh93/ksh/issues/146
Analysis: When a syntax error occurs, the shell performs a
longjmp(3) back to exfile() in main.c on line 417:
415| if(jmpval)
416| {
417| Sfio_t *top;
418| sh_iorestore((void*)shp,0,jmpval);
419| hist_flush(shp->gd->hist_ptr);
420| sfsync(shp->outpool);
The first thing it does is restore the file descriptor state
(sh_iorestore), then it flushes the history file (hist_flush), then
it synchronises sfio's logical stream state with the physical
stream state using (sfsync).
However, the fix applied in e999f6b1 caused sh_iorestore() to sync
all sfio streams unconditionally. So this was done before
hist_flush(), which caused unpredictable behaviour, including
temporary and/or permanent history corruption, as this also synched
shp->outpool before hist_flush() had a chance to do its thing.
The fix is to only call sfsync() in sh_iorestore() if we're
actually about to call ftruncate(2), and not otherwise.
Moral of the story: bug fixes should be as specific as possible to
minimise the risk of side effects.
src/cmd/ksh93/sh/io.c: sh_iorestore():
- Only call sfsync() if we're about to truncate a file.
src/cmd/ksh93/tests/pty.sh:
- Add test.
Thanks to Marc Wilson for reporting the bug and to Johnothan King
for finding the commit that introduced it.
Resolves: https://github.com/ksh93/ksh/issues/209
Relevant: https://github.com/att/ast/issues/61
These POSIX expansions first assign y to x if x is unset or empty,
respectively, and then they yield the value of x. This was not
working on any ksh93 version if x was typeset as numeric (integer
or float) but still unset, as in not assigned a value.
$ unset a; typeset -i a; printf '%q\n' "${a:=42}" "$a"
0
''
Expected output:
42
42
src/cmd/ksh93/sh/macro.c:
- Fix the test for set/unset variable. It was broken because it
only checked for the existence of the node, which exists after
'typeset', but did not check if a value had been assigned. This
additional check needs to be done with the nv_isnull() macro, but
only for expansions of the regular M_BRACE type. Special
expansions cannot have an unset state.
- As of commit 95294419, we know that an nv_optimize() call may be
needed before using nv_isnull() if the shell is compiled with
SHOPT_OPTIMIZE. Move the nv_optimize() call from that commit
forward to before the new check that calls nv_isnull(), and only
bother with it if the type is M_BRACE.
src/cmd/ksh93/tests/variables.sh:
- Add tests for this bug. Test float and integer, and also check
that ${a=b} and ${a:=b} correctly treat the value of 'b' as an
arithmetic expression of which the result is assigned to 'a' if
'a' was typeset as numeric.
src/cmd/ksh93/tests/attributes.sh,
src/cmd/ksh93/tests/comvar.sh,
src/cmd/ksh93/tests/nameref.sh,
src/cmd/ksh93/tests/types.sh:
- Fix a number of tests to report failures correctly.
Resolves: https://github.com/ksh93/ksh/issues/157
This commit fixes an arbitrary command execution vulnerability in
array subscripts used within the arithmetic subsystem.
One of the possible reproducers is:
var='1$(echo INJECTION >&2)' ksh -c \
'typeset -A a; ((a[$var]++)); typeset -p a'
Output before this commit:
INJECTION
typeset -A a=([1]=1)
The 'echo' command has been surreptitiously executed from an
external environment variable.
Output after this commit:
typeset -A a=(['1$(echo INJECTION >&2)']=1)
The value is correctly used as an array subscript and nothing in it
is parsed or executed. This is as it should be, as ksh93 supports
arbitrary subscripts for associative arrays.
If we think about it logically, the C-style arithmetic subsystem
simply has no business messing around with shell expansions or
quoting at all, because those don't belong to it. Shell expansions
and quotes are properly resolved by the main shell language before
the arithmetic subsystem is even invoked. It is particularly
important to maintain that separation because the shell expansion
mechanism also executes command substitutions.
Yet, the arithmetic subsystem subjected array subscripts that
contain `$` (and only array subscripts -- how oddly specific) to
an additional level of expansion and quote resolution. For some
unfathomable reason, there are two lines of code doing specifically
this. The vulnerability is fixed by simply removing those.
Incredibly, variants of this vulnerability are shared by bash, mksh
and zsh. Instead of fixing it, it got listed in Bash Pitfalls!
http://mywiki.wooledge.org/BashPitfalls#y.3D.24.28.28_array.5B.24x.5D_.29.29
src/cmd/ksh93/sh/arith.c:
- scope(): Remove these two lines that implement the vulnerability.
if(strchr(sub,'$'))
sub = sh_mactrim(shp,sub,0);
- scope(), arith(): Remove the NV_SUBQUOTE flag from two
nv_endsubscript() calls. That flag causes the array subscript to
retain the current level of shell quoting. The shell quotes
everything as in "double quotes" before invoking the arithmetic
subsystem, and the bad sh_mactrim() call removed one level of
quoting. Since we're no longer doing that, this flag should no
longer be passed, or subscripts may get extra backslash escapes.
src/cmd/ksh93/include/name.h,
src/cmd/ksh93/sh/array.c:
- nv_endsubscript(): The NV_SUBQUOTE flag was only passed from
arith.c. Since it is now unused, remove it.
src/cmd/ksh93/tests/arith.sh:
- Tweak some tests: fix typos, report wrong values.
- Add 21 tests. Most are based on reproducers contributed by
@stephane-chazelas and @hyenias. They verify that this
vulnerability is gone and that no quoting bugs were introduced.
Resolves: https://github.com/ksh93/ksh/issues/152
