This commit recognizes an existing pattern: many operations need some
combination of a set of variables, a way to detect cancellation, and
sometimes a parser. For example, tab completion needs a parser to execute
custom completions, the variable set, should cancel on SIGINT. Background
autosuggestions don't need a parser, but they do need the variables and
should cancel if the user types something new. Etc.
This introduces a new triple operation_context_t that wraps these concepts
up. This simplifies many method signatures and argument passing.
When executing a buffered block or builtin, the usual approach is to
execute, collect output in a string, and then output that string to
stdout or whatever the redirections say. Similarly for stderr.
If we get no output, then we can elide the outputting which means
skipping the background thread. In this case we just mark the process as
finished immediately.
We do this in multiple locations which is confusing. Factor them all
together into a new function run_internal_process_or_short_circuit.
for-loops that were not inside a function could overwrite global
and universal variables with the loop variable. Avoid this by making
for-loop-variables local variables in their enclosing scope.
This means that if someone does:
set a global
for a in local; end
echo $a
The local $a will shadow the global one (but not be visible in child
scopes). Which is surprising, but less dangerous than the previous
behavior.
The detection whether the loop is running inside a function was failing
inside command substitutions. Remove this special handling of functions
alltogether, it's not needed anymore.
Fixes#6480
'fish_test_helper print_pid_then_sleep' tried to sleep for .5 seconds,
but instead it divided by .5 so it actually slept for 2 seconds.
This exceeds the maximum value on NetBSD so it wasn't sleeping at all
there.
Fixes#6476
Empty items are used as sentinels to indicate that we've reached the end of
history, so they should not be added as actual items. Enforce this.
Fixes#6032
This reduces the syscall count for `fish -c exit` from 651 to 566.
We don't attempt to *cache* the pgrp or anything, we just call it once
when we're about to execute the job to see if we are in foreground and
to assign it to the job, instead of once for checking foreground and
once to give it to the job.
Caching it with a simple `static` would get the count down to 480, but
it's possible for fish to have its pgroup changed.
Store the entire function declaration, not just its job list.
This allows us to extract the body of the function complete with any
leading comments and indents.
Fixes#5285
In particular, this allows `true && time true`, or `true; and time true`,
and both `time not true` as well as `not time true` (like bash).
time is valid only as job _prefix_, so `true | time true` could call
`/bin/time` (same in bash)
See discussion in #6442
Extend the commit 8e17d29e04 to block processes, for example:
begin ; stuff ; end
or if/while blocks as well.
Note there's an existing optimization where we do not create a job for a
block if it has no redirections.
job_promote attempts to bring the most recently "touched" job to the front
of the job list. It did this via:
std::rotate(begin, job, end)
However this has the effect of pushing job-1 to the end. That is,
promoting '2' in [1, 2, 3] would result in [2, 3, 1].
Correct this by replacing it with:
std::rotate(begin, job, job+1);
now we get the desired [2, 1, 3].
Also add a test.
This PR is aimed at improving how job ids are assigned. In particular,
previous to this commit, a job id would be consumed by functions (and
thus aliases). Since it's usual to use functions as command wrappers
this results in awkward job id assignments.
For example if the user is like me and just made the jump from vim -> neovim
then the user might create the following alias:
```
alias vim=nvim
```
Previous to this commit if the user ran `vim` after setting up this
alias, backgrounded (^Z) and ran `jobs` then the output might be:
```
Job Group State Command
2 60267 stopped nvim $argv
```
If the user subsequently opened another vim (nvim) session, backgrounded
and ran jobs then they might see what follows:
```
Job Group State Command
4 70542 stopped nvim $argv
2 60267 stopped nvim $argv
```
These job ids feel unnatural, especially when transitioning away from
e.g. bash where job ids are sequentially incremented (and aliases/functions
don't consume a job id).
See #6053 for more details.
As @ridiculousfish pointed out in
https://github.com/fish-shell/fish-shell/issues/6053#issuecomment-559899400,
we want to elide a job's job id if it corresponds to a single function in the
foreground. This translates to the following prerequisites:
- A job must correspond to a single process (i.e. the job continuation
must be empty)
- A job must be in the foreground (i.e. `&` wasn't appended)
- The job's single process must resolve to a function invocation
If all of these conditions are true then we should mark a job as
"internal" and somehow remove it from consideration when any
infrastructure tries to interact with jobs / job ids.
I saw two paths to implement these requirements:
- At the time of job creation calculate whether or not a job is
"internal" and use a separate list of job ids to track their ids.
Additionally introduce a new flag denoting that a job is internal so
that e.g. `jobs` doesn't list internal jobs
- I started implementing this route but quickly realized I was
computing the same information that would be computed later on (e.g.
"is this job a single process" and "is this jobs statement a
function"). Specifically I was computing data that populate_job_process
would end up computing later anyway. Additionally this added some
weird complexities to the job system (after the change there were two
job id lists AND an additional flag that had to be taken into
consideration)
- Once a function is about to be executed we release the current jobs
job id if the prerequisites are satisfied (which at this point have
been fully computed).
- I opted for this solution since it seems cleaner. In this
implementation "releasing a job id" is done by both calling
`release_job_id` and by marking the internal job_id member variable to
-1. The former operation allows subsequent child jobs to reuse that
same job id (so e.g. the situation described in Motivation doesn't
occur), and the latter ensures that no other job / job id
infrastructure will interact with these jobs because valid jobs have
positive job ids. The second operation causes job_id to become
non-const which leads to the list of code changes outside of `exec.c`
(i.e. a codemod from `job_t::job_id` -> `job_t::job_id()` and moving the
old member variable to a non-const private `job_t::job_id_`)
Note: Its very possible I missed something and setting the job id to -1
will break some other infrastructure, please let me know if so!
I tried to run `make/ninja lint`, but a bunch of non-relevant issues
appeared (e.g. `fatal error: 'config.h' file not found`). I did
successfully clang-format (`git clang-format -f`) and run tests, though.
This PR closes#6053.
It looks like the last status already contains the signal that cancelled
execution.
Also make `fish -c something` always return the last exit status of
"something", instead of hardcoded 127 if exited or signalled.
Fixes#6444
Previously, the block stack was a true stack. However in most cases, you
want to traverse the stack from the topmost frame down. This is awkward
to do with range-based for loops.
Switch it to pushing new blocks to the front of the block list.
This simplifies some traversals.
This was previously required so that, if there was a redirection to a
file, we would fork a process to create the file even if there was no
output. For example `echo -n >/tmp/file.txt` would have to create
file.txt even though it would be empty.
However now we open the file before fork, so we no longer need special
logic around this.
Do this only when splitting on IFS characters which usually contains
whitespace characters --- read --delimiter is unchanged; it still
consumes no more than one delimiter per variable. This seems better,
because it allows arbitrary delimiters in the last field.
Fixes#6406
user_supplied was used to distinguish IO redirections which were
explicit, vs those that came about through "transmogrphication." But
transmogrification is no more. Remove the flag.
As of GCC 7.4 (at least under macOS 10.10), the previous workaround of
casting a must-use result to `(void)` to avoid warnings about unused
code no longer works.
This workaround is uglier but it quiets these warnings.
The C++ spec (as of C++17/n4713) does not specify the sign of `wchar_t`,
saying only (in section 6.7.1: Fundamental Types)
> Type wchar_t shall have the same size, signedness, and alignment
> requirements (6.6.5) as one of the other integral types, called its
> underlying type.
On most *nix platforms on AMD64 architecture, `wchar_t` is a signed type
and can be compared with `int32_t` without incident, but on at least
some platforms (tested: clang under FreeBSD 12.1 on AARCH64), `wchar_t`
appears to be unsigned leading to sign comparison warnings:
```
../src/widecharwidth/widechar_width.h:512:48: warning: comparison of
integers of different signs: 'const wchar_t' and 'int32_t' (aka 'int')
[-Wsign-compare]
return where != std::end(arr) && where->lo <= c;
```
This patch forces the use of wchar_t for the range start/end values in
`widechar_range` and the associated comparison values.
Previously, if the user control-C'd out of a process, we would set a
bogus exit status in the process, but it was difficult to observe this
because we would be cancelling anyways. But set it properly.
If a Control-C is received during expanding a command substitution, we
may execute the job anyways, because we do not check for cancellation
after the expansion. Ensure that does not happen.
This should fix sporadic test failures in the cancellation unit test.
parser_t::eval indicates whether there was a parse error. It can be
easily confused with the status of the execution. Use a real type to
make it more clear.
Looking up a variable by a string literal implicitly constructs a wcstring.
By avoiding that, we get a noticeable reduction of temporary allocations.
$ HOME=. heaptrack ./fish -c true
heaptrack stats: # baseline
allocations: 7635
leaked allocations: 3277
temporary allocations: 602
heaptrack stats: # new
allocations: 7565
leaked allocations: 3267
temporary allocations: 530
This adds a test for the obscure case where an fd is redirected to
itself. This is tricky because the dup2 will not clear the CLO_EXEC bit.
So do it manually; also posix_spawn can't be used in this case.
The IO cleanup left file redirections open in the child. For example,
/bin/cmd < file.txt would redirect stdin but also leave the file open.
Ensure these get closed properly.
Prior to this fix, a file redirection was turned into an io_file_t. This is
annoying because every place where we want to apply the redirection, we
might fail due to open() failing. Switch to opening the file at the point
we resolve the redirection spec. This will simplify a lot of code.
Prior to this change, a process after it has been constructed by
parse_execution, but before it is executed, was given a list of
io_data_t redirections. The problem is that redirections have a
sensitive ownership policy because they hold onto fds. This made it
rather hard to reason about fd lifetime.
Change these to redirection_spec_t. This is a textual description
of a redirection after expansion. It does not represent an open file and
so its lifetime is no longer important.
This enables files to be held only on the stack, and are no longer owned
by a process of indeterminate lifetime.
fish has to ensure that the pipes it creates do not conflict with any
explicit fds named in redirections. Switch this code to using
autoclose_fd_t to make the ownership logic more explicit, and also
introduce fd_set_t to reduce the dependence on io_chain_t.
Prior to this fix, a job would hold onto any IO redirections from its
parent. For example:
begin
echo a
end < file.txt
The "echo a" job would hold a reference to the I/O redirection.
The problem is that jobs then extend the life of pipes until the job is
cleaned up. This can prevent pipes from closing, leading to hangs.
Fix this by not storing the block IO; this ensures that jobs do not
prolong the life of pipes.
Fixes#6397
Currently a job needs to know three things about its "parents:"
1. Any IO redirections for the block or function containing this job
2. The pgid for the parent job
3. Whether the parent job has been fully constructed (to defer self-disown)
These are all tracked in somewhat separate awkward ways. Collapse them
into a single new type job_lineage_t.
In preparation for concurrent execution, invert the control of function and
block execution. Allow a process to return an std::function that performs the
the execution. This can be run on either the main or a background thread
(eventually).
This splits a string into variables according to the shell's
tokenization rules, considering quoting, escaping etc.
This runs an automatic `unescape` on the string so it's presented like
it would be passed to the command. E.g.
printf '%s\n' a\ b
returns the tokens
printf
%s\n
a b
It might be useful to add another mode "--tokenize-raw" that doesn't
do that, but this seems to be the more useful of the two.
Fixes#3823.
Background fillthreads are used when we want to populate a buffer from an
external command. The most common is command substitution.
Prior to this commit, fish would spin up a fillthread whenever required.
This ended up being quite expensive.
Switch to using the iothread pool instead. This enables reusing the same
thread(s), which prevents needing to spawn new threads. This shows a big
perf win on the alias benchmark (766 -> 378 ms).
This reintroduces commits 22230a1a0d
and 9d7d70c204, now with the bug fixed.
The problem was when there was one thread waiting in the pool. We enqueue
an item onto the pool and attempt to wake up the thread. But before the
thread runs, we enqueue another item - this second enqueue will see the
thread waiting and attempt to wake it up as well. If the two work items
were dependent (reader/writer) then we would have a deadlock.
The fix is to check if the number of waiting threads is at least as large
as the queue. If the number of enqueued items exceeds the number of waiting
threads, then spawn a new thread always.
This added the function offset *again*, but it's already included in
the line for the current file.
And yes, I have explicitly tested a function file with a function
defined at a later line.
Fixes#6350
Since #6287, bare variable assignments do not parse, which broke
the "Unsupported use of '='" error message.
This commit catches parse errors that occur on bare variable assignments.
When a statement node fails to parse, then we check if there is at least one
prefixing variable assignment. If so, we emit the old error message.
See also #6347
