This is a MFC of all the commits listed below.
My original goal of this change was to only merge the move of the tests
from tools/regression/bin/ into the new layout (which include tests for
sh(1) and other tools as well). However, doing so is tricky due to the
ongoing work in sh(1) and, especially, the many changes to its tests
since stable/10 was first branched.
Merging everything is the simplest way to achieve this goal and, as a
bonus point, we get various fixes and miscellaneous improvements into
the branch.
Per jilles' suggestion, I'm avoiding the merge of a couple of changes
(r256850 and r257506) that required depending kernel changes. I'm also
avoiding very recent changes that have not had a long enough time to be
validated in current.
This is "make tinderbox" clean.
r256735 sh: Remove one syscall when waiting for a foreground job.
r257399 sh: Allow trapping SIGINT/SIGQUIT after ignore because of '&'.
r257504 sh: Reorder union node to reduce its size on 64-bit platforms.
r257920 sh: Add a test case for would-be assignments that are not due to quoting.
r257929 sh: Properly quote alias output from command -v.
r258489 sh: Add tests for the </dev/null implicit in a background command.
r258533 sh: Add more tests for the </dev/null implicit in a background command.
r258535 sh: Make <&0 disable the </dev/null implicit in a background command.
r258776 sh: Prefer memcpy() to strcpy() in most cases. Remove the scopy macro.
r259047 sh: Split set -x output into a separate function.
r259210 Migrate tools/regression/bin/ tests to the new layout.
r259844 sh: Remove an unused variable.
r259846 sh: Initialize OPTIND=1 even if it came from the environment.
r259874 sh: Simplify code related to PPID variable.
r259946 sh: Don't check input for non-whitespace if history is disabled.
r260246 sh(1): Discourage use of -e.
r260506 Run the sh(1) and test(1) tests as unprivileged.
r260586 Mark the bin/pax tests as requiring perl.
r260634 Use TAP_TESTS_PERL to register the legacy_test in bin/pax.
r260635 Replace hand-crafted Kyuafiles with automatic generation.
r260654 sh: Remove SIGWINCH handler and just check for resize before every read.
r261121 sh: Add test for nested alias.
r261125 sh: Solve the alias recursion problem in a less hackish way.
r261141 sh: Do not depend on parse/execute split in new alias test.
r261160 sh: Add tests for alias names after another alias.
r261192 sh: Allow aliases to force alias substitution on the following word.
r262533 sh: Make expari() static.
r262565 sh: Do not corrupt internal representation if LINENO inner expansion fails.
r262697 sh: Simplify expari().
Reviewed by: jilles
Formerly, return always returned from a function if it was called from a
function, even if there was a closer dot script. This was for compatibility
with the Bourne shell which only allowed returning from functions.
Other modern shells and POSIX return from the function or the dot script,
whichever is closest.
Git 1.8.4's rebase --continue depends on the POSIX behaviour.
Reported by: Christoph Mallon, avg
Replace the RESET blocks with regular functions and a reset() function that
calls them all.
This code generation tool is unusual and does not appear to provide much
benefit. I do not think isolating the knowledge about which modules need to
be reset is worth an almost 500-line build tool and wider scope for
variables used by the reset functions.
Also, relying on reset functions is often wrong: the cleanup should be done
in exception handlers so that no stale state remains after 'command eval'
and the like.
If a stack mark is set while the current stack block is empty, the stack
block may move later on (because of realloc()) and the stack mark needs to
be updated. This updating does not happen after popstackmark() has been
called; therefore, call setstackmark() again if the stack mark is still
being used.
For some reason, this only affects a few users. I cannot reproduce it. The
situation seems quite rare as well because an empty stack block would
usually be freed (by popstackmark()) before execution reaches a
setstackmark() call.
PR: 175922
Tested by: KT Sin
Expand here documents at the same point other redirections are expanded but
use a non-fork subshell environment (like simple command substitutions) for
compatibility. Substitition errors result in an empty here document like
before.
As a result, a fork is avoided for short (<4K) expanded here documents.
Unexpanded here documents (with quoted end marker after <<) are not affected
by this change. They already only forked when >4K.
Side effects:
* Order of expansion is slightly different.
* Slow expansions are not executed in parallel with the redirected command.
* A non-fork subshell environment is subtly different from a forked process.
If a loop contained certain commands (such as redirected compound commands),
the temporary memory for the redirection was not freed between iterations of
the loop but only after the loop.
Put a stackmark in evaltree(), freeing memory whenever a node has been
evaluated. Some other stackmarks are then redundant; remove them.
Example:
while :; do { :; } </dev/null; done
If there is a write error on stdout, a message will be printed (to stderr)
and the exit status will be changed to 2 if it would have been 0 or 1.
PR: bin/158206
This reduces code duplication and code size.
/usr/bin/printf is not affected.
Side effect: different error messages when certain builtins are passed
invalid options.
Examples:
export x=~
now expands the tilde
local y=$1
is now safe, even if $1 contains IFS characters or metacharacters.
For a word to "look like an assignment", it must start with a name followed
by an equals sign, none of which may be quoted.
The special treatment applies when the first word (potentially after
"command") is "export", "readonly" or "local". There may be quoting
characters but no expansions. If "local" is overridden with a function there
is no special treatment ("export" and "readonly" cannot be overridden with a
function).
If things like
local arr=(1 2 3)
are ever allowed in the future, they cannot call a "local" function. This
would either be a run-time error or it would call the builtin.
This matches Austin Group bug #351, planned for the next issue of POSIX.1.
PR: bin/166771
This uses vfork() for simple commands and command substitutions containing a
single simple command, invoking an external program under certain conditions
(no redirections or variable assignments, non-interactive shell, no job
control). These restrictions limit the amount of code executed in a vforked
child.
There is a large speedup (for example 35%) in microbenchmarks. The
difference in buildkernel is smaller (for example 0.5%) but still
statistically significant. See
http://lists.freebsd.org/pipermail/freebsd-hackers/2012-January/037581.html
for some numbers.
The use of vfork() can be disabled by setting a variable named
SH_DISABLE_VFORK.
In the first command of a 'for', $? should be the exit status of the last
pipeline (command substitution in the word list or command before 'for'),
not always 0.
Before this fix, only the first statement of the trap was executed if
evalskip was set. This is for example the case when:
o "-e" is set for this shell
o a trap is set on EXIT
o a function returns 1 and causes the script to abort
Reviewed by: jilles
MFC after: 2 weeks
Also, rework evalcase() to not evaluate any tree. Instead, return the
NCLISTFALLTHRU node and handle it in evaltree().
Fixed bugs:
* If a ;& list with non-zero exit status is followed by an empty ;; or final
list, the exit status of the case command should be equal to the exit
status of the ;& list, not 0.
* An empty ;& case should not reset $?.
* If no pattern is matched, POSIX says the exit status shall be 0 (even if
there are command substitutions).
* If a pattern is matched and there are no command substitutions, the first
command should see the $? from before the case command, not always 0.
This is not necessary: errors are already caught in evalbackcmd() and
forcelocal handles changes to variables.
Note that this depends on r223024.
MFC after: 4 weeks
Free expanded case text before executing commands.
Remove impossible evalskip checks (expanding an argument cannot set
evalskip anymore since $(break) and the like are properly executed in a
subshell environment).
Replacing ;; with the new control operator ;& will cause the next list to be
executed as well without checking its pattern, continuing until a list ends
with ;; or until the end of the case statement. This is like omitting
"break" in a C "switch" statement.
The sequence ;& was formerly invalid.
This feature is proposed for the next POSIX issue in Austin Group issue
#449.
The eval special builtin now runs the code with EV_EXIT if it was run
with EV_EXIT itself.
In particular, this eliminates one fork when a command substitution contains
an eval command that ends with an external program or a subshell.
This is similar to what r220978 did for functions.
In optimized command substitution, save and restore any variables changed by
expansions (${var=value} and $((var=assigned))), instead of trying to
determine if an expansion may cause such changes.
If $! is referenced in optimized command substitution, do not cause jobs to
be remembered longer.
This fixes $(jobs $!) again, simplifies the man page and shortens the code.
The function name expandstr() and the general idea of doing this kind of
expansion by treating the text as a here document without end marker is from
dash.
All variants of parameter expansion and arithmetic expansion also work (the
latter is not required by POSIX but it does not take extra code and many
other shells also allow it).
Command substitution is prevented because I think it causes too much code to
be re-entered (for example creating an unbounded recursion of trace lines).
Unfortunately, our LINENO is somewhat crude, otherwise PS4='$LINENO+ ' would
be quite useful.
For backgrounded pipelines and subshells, the previous value of $? was being
preserved, which is incorrect.
For backgrounded simple commands containing a command substitution, the
status of the last command substitution was returned instead of 0.
If fork() fails, this is an error.
If EV_EXIT causes an exit, use the exception mechanism to unwind
redirections and local variables. This way, if the final command is a
redirected command, an EXIT trap now executes without the redirections.
Because of these changes, EV_EXIT can now be inherited by the body of a
function, so do so. This means that a function no longer prevents a fork
before an exec being skipped, such as in
f() { head -1 /etc/passwd; }; echo $(f)
Wrapping a single builtin in a function may still cause an otherwise
unnecessary fork with command substitution, however.
An exit command or -e failure still invokes the EXIT trap with the
original redirections and local variables in place.
Note: this depends on SHELLPROC being gone. A SHELLPROC depended on
keeping the redirections and local variables and only cleaning up the
state to restore them.
These are called "shell procedures" in the source.
If execve() failed with [ENOEXEC], the shell would reinitialize itself
and execute the program as a script. This requires a fair amount of code
which is not frequently used (most scripts have a #! magic number).
Therefore just execute a new instance of sh (_PATH_BSHELL) to run the
script.
An error message is written, the builtin is not executed, nonzero exit
status is returned but the shell does not abort.
This was already checked for special builtins and external commands, with
the same consequences except that the shell aborts for special builtins.
Obtained from: NetBSD
Change the criterion for builtins to be safe to execute in the same process
in optimized command substitution from a blacklist of only cd, . and eval to
a whitelist.
This avoids clobbering the main shell environment such as by $(exit 4) and
$(set -x).
The builtins jobid, jobs, times and trap can still show information not
available in a child process; this is deliberately permitted. (Changing
traps is not.)
For some builtins, whether they are safe depends on the arguments passed to
them. Some of these are always considered unsafe to keep things simple; this
only harms efficiency a little in the rare case they are used alone in a
command substitution.
Before considering to execute a command substitution in the same process,
check if any of the expansions may have a side effect; if so, execute it in
a new process just like happens if it is not a single simple command.
Although the check happens at run time, it is a static check that does not
depend on current state. It is triggered by:
- expanding $! (which may cause the job to be remembered)
- ${var=value} default value assignment
- assignment operators in arithmetic
- parameter substitutions in arithmetic except ${#param}, $$, $# and $?
- command substitutions in arithmetic
This means that $((v+1)) does not prevent optimized command substitution,
whereas $(($v+1)) does, because $v might expand to something containing
assignment operators.
Scripts should not depend on these exact details for correctness. It is also
imaginable to have the shell fork if and when a side effect is encountered
or to create a new temporary namespace for variables.
Due to the $! change, the construct $(jobs $!) no longer works. The value of
$! should be stored in a variable outside command substitution first.
Command substitutions consisting of a single simple command are executed in
the main shell process but this should be invisible apart from performance
and very few exceptions such as $(trap).
* Prefer one CHECKSTRSPACE with multiple USTPUTC to multiple STPUTC.
* Add STPUTS macro (based on function) and use it instead of loops that add
nul-terminated strings to the stack string.
No functional change is intended, but code size is about 1K less on i386.
