The first cut of the form-change refusal asked only about def and defonce
and asserted in its comment that defconst was another arm's business. It
was not: defconst to def at the same type fell past every arm, and defonce
to defconst fell past them into the consts republish, which stores the
declared value over live storage at the frame boundary. One refusal over
gconst and grerun together now covers all six directions.
And two coverage gaps closed by running rather than reasoning: reload-v6
carries a (def dial i64 5) the host was never built with, so the x86 image
path executes and its 5 shows in the transcript's arithmetic; dev-rerun's
echo reads counter in its initialiser and follows it 40, 41, 42, 43 across
re-runs, where a captured first answer would print 40 four times.
The trio the author decided on 2026-09-20 is now all built: def is CL's
defparameter — its initialiser runs on every daemon re-run, unguarded, so
an edited initialiser repaints the same storage on C-c C-c plus re-run —
defonce (Clojure's name for CL's defvar, per the author) initialises once
behind the .init~once. flag, and defconst stays the image.
One parse arm reads both forms; the difference is Ast.reinit, carried to
Tast.global's grerun. Emit.startup_plan gives a def no guard flag, and
Check.check_global lifts every def initialiser — zero and literal
included — into global/<n>, so the host's startup reaches it through the
function cell and a re-evaluated def swaps it (Session's def_inits;
Emit.redefinition declares the cell for a non-sibling target). The old
defvar spelling is refused with the rename and both compiling spellings,
and every program, test, doc and editor list is swept — except sand.flan,
the author's live WIP, whose seven defvar lines are flagged in FIX.org
and keep its three dependent tests red on this branch.
The wait recognised one kind of stop and waited five seconds for all the
others, then said the program was not reaching a frame boundary — on a reply
that named the condition two fields along.
Which stop is the thunk's is decided by the agent's stop generation rather
than by the condition's name, so a break entered on the same class as the one
it was evaluated inside is still told apart from it. A (pause) is answered as
a (pause) wherever it came from, the flag having never been what made one
deliberate.
- SA_NODEFER. sigaction without it blocks the handler's own signal for the
whole handler, and here the handler is the park — it never returns. A
hardware SIGSEGV delivered while SIGSEGV is blocked is not handled: the
kernel forces the default action. Fault, park, eval something at the
break loop that faults, daemon gone, exactly the author's session one
level in. Measured both ways; flan_crash_entered is cleared before the
hook so each break-loop fault still gets its line, and the case is pinned
(trap_park ~refault:true), confirmed to fail without the flag.
- Scope the handler to the thread it was armed on. A disposition is per
process and a merged dev session is one process, so this was shadowing
OCaml's SIGSEGV handler — and Stack_overflow — for the daemon's whole
life. Other threads chain to what was installed before. Arming per run
would leave the parked prompt's evaluations unprotected, since those are
program code too; the comment says so. Also makes the per-thread
sigaltstack honest.
- Sweep dyn-view.flan and string-eq.flan, which dev-loop added after the
first sweep. string-eq:46 wanted the aliasing outright: its comment is
about two slices sharing a base pointer.
- A StorageExhausted row for bytes, asserting the retry copies once and
whole rather than re-evaluating its argument.
- Gate the flan_dev_crash_enable declare to dev builds, so this lane adds
no dev-only text to a release module. flan_bytes_dup stays ungated: a
release build really calls it.
- Guard the section for wasm32, which compiles this file and has no
signals.
The INSERTIONSORT crash, all three rulings (FIX.org 2026-09-20):
- (bytes s) allocates a writable copy through the allocator surface —
context or (bytes s a), StorageExhausted with retry, a registry note in
dev builds (flan_bytes_dup, lowered like vec-new). (bytes-view s) is the
old zero-cost reinterpret, renamed, read-only by convention; every
in-repo reader swept over to it. (string b) unchanged.
- String constants were already read-only on both backends at -O0; now
pinned — bytes-copy.flan rows on LLVM/-O0/--x86, and dies_segv rows
asserting the write-through-view trap on both backends.
- A dev build installs a SIGSEGV/SIGBUS handler by the same dev-only
constructor slot that arms the registry: one line naming the address and
the innermost frame, then the trap-hook park — stopped, not dead, the
daemon serving. No agent: message and re-raise. Release builds untouched.
Pinned by trap_park over dev-segv.flan.
Five fixes off the independent review, plus the author's u8 ruling.
x86 parity: the bad-index block always ran x86 (it is flan dev's
default) and now says so with an explicit --x86; the condition render
gets an assertion under the x86 backend too, beside the LLVM one, and
a user error is pinned as carrying no site on both.
ArithError's layout is now pinned: {i32 op; i64 lhs, rhs} in C against
the prelude's defstruct, read field by field through the break loop's
render, driven from the editor through a divide under a restart-case.
That also covers condition and site on LLVM.
Three refusals that were wrong: trap_site tested the prefix "err"
and so ate any site whose path began with those letters; source_line
let Sys_error from input_line escape and take the whole break reply
with it, leaking the handle; and a condition with no fields was
reported as a name no struct has. The daemon now sends its own field
count and the buffer tells the two empties apart.
Nits taken: an over-long site is dropped rather than silently
truncated into a plausible one; the caret pads with the source line's
own tabs; the headline says when it has cut the field list;
flan-cnr-layout is live again as the single spelling of that request
rather than dead beside an inlined copy.
And the ruling: a u8 renders as 97 (\a) where a person is inspecting
and stays 97 where the program is printing.
A break nested inside a trap's break — a fix candidate evaluated at a
bounds stop raising its own error — copied the outer trap's site into
its snapshot, which is a caret pointing at an unrelated line under the
inner condition's name. snap_push now consumes the global: each
snapshot owns its copy, a nested entry that set no fresh site gets
none, and the outer break keeps its own. Pinned end to end.
The fields section says why the values are missing once, above the
rows, instead of repeating the sentence per field; a row keeps its own
reason only when it has one (no printer for its type).
strip_rebind and shown_names get direct coverage, including the clean
strip a real frame almost never reaches.
The break loop used to discard the pointer it was handed, so the buffer
could name a BoundsError's fields and never show 648. Now the snapshot
stashes it, flan_agent_condition hands it back on the stopped thread,
and a daemon-built thunk — locals pointed at the condition — renders
each field. Delivered at-stop, so a resume-and-restop cannot get the
old type read over the new pointer.
The trap sites publish their loc around the hook call, the snapshot
copies it, and break answers :site with the line's text as :source —
the frame lines say where each call was; this is the only record of
the indexing itself.
Compiler temps are hidden from the locals listing rather than refused
as s4; a shadowing rebind strips its ~N except where the outer binding
is on the same list, where both keep their raw spelling.
The bug review found: [start_on] claimed [started] at the top and every
failure exit left it claimed. Under [flan dev] the constructor is the first
caller and reports to nobody, so a path nothing could bind disarmed the
program's own (agent/start ...) as well — it answered 0 with no socket, no
listener and no hooks, where before this lane the explicit form answered -1.
Success reported for nothing at all is worse than the error it replaced.
So every way out that is not a listening socket unwinds: the fd is closed, a
file the bind managed to make is unlinked, and [started] goes back to 0 so a
later start is a real attempt. Pinned by running the zero-argument fixture
with FLAN_AGENT_SOCKET pointing nowhere — constructor fails silently, main's
own call then fails loudly, "cannot listen" and exit 1.
Two arguments to (agent/start) are refused, which nothing held: the macro's
[& args] cannot say "one at most", so what says it is the expansion splicing
every argument into a function that declares one. The message names
agent/start-at and carries the expanded-from note, and that is what the
acceptance row asserts.
And the reply a delivery gets when there is no agent in the process, which
nothing held either. dev-noagent.flan parks, so it was never this case;
dev-noagent-running.flan keeps running, and the answer is a refusal naming the
socket that could not be reached — not install_note's "queued", which would
promise a poll with nothing to drain. Which leaves that note unreachable in
all three shapes rather than merely unpinned, worked through in FIX.org.
FIX.org also now says what an exported FLAN_AGENT_SOCKET would do: start_on
unlinks before it binds, so an agent-linked program started in that
environment takes the path away from whoever bound it first.
A constructor in the agent package binds FLAN_AGENT_SOCKET when it is set,
which is the daemon and nothing else — both shapes set it, before the fork in
--two-process and before the exec in the merged build. So a program under
[flan dev] that calls (agent/poll) and has no (agent/start) in it takes
redefinitions anyway, and one that does call start meets an agent that is
already listening and gets a no-op.
The window this closes was the complaint in DISCUSS.org: a program that opens
a window before starting its agent leaves the daemon waiting on a socket that
does not exist yet. Bound here, the socket exists before main whatever the
program does afterwards — so test_dev.ml's late-agent row asserts the negation
of what it used to. The delivery sent during the sleep no longer carries "the
program has not called (agent/start ...) yet", because that is no longer true
of it; what is still late, and still asserted, is the poll that installs it.
It reaches exactly as far as the linker does. Reach prunes a package nothing
calls into, so a program that mentions the agent nowhere does not link this
file and has no constructor to run: auto-start is for a program that polls and
has dropped its start call, not for one that says nothing about the agent at
all. That limit and the release-build residual are in FIX.org, along with the
daemon branch that can no longer be reached.
agent-nostart.flan is the pin, and its two numbers are the honest ones: 1
before anything could arrive, 1000 after the wait, because a listener bound
before main is still not an install.
The migration transcript runs on x86, because that is what flan dev takes
unasked. What is backend-specific about any of this is one thing — whether
the registration thunk reaches the runtime at all — and the evidence for
LLVM was that the IR contained the call, which is emission and not
execution. x86.ml's own header claimed for some time that it did not emit
flan_reload_call, which is exactly the kind of sentence not to trust twice.
So: the same program under flan dev --llvm, one instance, one slot added,
and the four answers that say the migration happened. Short on purpose —
everything past the thunk is flan_dyn.c's, and flan_dyn.c does not know
who called it.
CLHS 4.3.6's update protocol, minus the user hook, on the dyn side's
defclass. Redefining a class used to be silent: a class is sugar for a
constructor defn, so the edit replaced a body and the instances already in
the program kept their old keys for ever.
Three pieces. A registry in flan_dyn.c holding each class's current slot
list and a generation, made only of interned kw_entry pointers so the
collector has nothing to trace in it and no root to push for it. A uint32
generation on the instance, fitted into the padding kind and mark leave in
front of len's alignment — sizeof(flan_obj) is 48 with it and was 48
without, and flan_dyn_obj_size is there so a later field that moves it
fails a test. And a registration thunk per reload, run by the agent
through flan_reload_call after the module's bodies are published: it has
to be a thunk, because the case this exists for is a class redefined and
not constructed.
Migration is lazy, at want_map, len's map arm and dyn_equal's. Slots kept
by name, gained slots nil, dropped slots gone, identity preserved, entries
rebuilt in the class's order so a migrated instance is indistinguishable
from a fresh one. Equality migrates both operands first, so it is over the
class as it is now.
The session had to stop refusing the constructor's signature change, and
does so only for a defclass and only when no compiled caller is left
behind. The checker gets there first in practice; the walk in eval holds
the reason locally rather than inheriting it.
The registry is advisory: a class instance is an open map, so a key a raw
put wrote that the class never declared is dropped by the next migration.
FIX.org says that plainly rather than pretending enforcement.
DISCUSS.org's "need a value-producing array constructor": the author
wanted grid filled with 255 as part of its declaration and could not
write it. (array n T) produces the zeroed array only, and dotimes is
Unit, so it can mutate a place that already exists but cannot be the
initialiser expression -- which has to produce the whole value in one
go. The grid was declared zeroed and filled in main instead.
Two forms, both expressions, both any rank:
(array-fill [rows cols] 255) every element that value
(array-gen [rows cols] cell) every element (cell i j)
Spelled apart rather than one form dispatching on the third element's
type, because an array *of* function values is a thing to want and one
form would have to decide whether (array-fill [4] f) meant four copies
of f or four calls of it.
The dimensions are read in Parse, and that is the whole reason they are
recognised there: handed through as an ordinary call, [rows cols] is an
array literal of two names, and where those names are defconsts it is a
perfectly good two-element array of integers -- the wrong reading, and a
silent one. Read in Parse they are the same len the [n T] type spelling
takes, resolved by the same array_len, with one extra condition of their
own: the fill counts in i32 like every index in the language, so a
dimension no i32 can reach has no loop that could end.
The lowering is a loop over a slot, not an aggregate. Tast.Arr is the
node the backends have and both build it element by element from a list
as long as the array; a fill of [600 [800 u8]] is half a million
elements and there is no list to be had. So these bind the array to a
slot, zero it, run one While per dimension writing through Set of a
Pindex, and answer with the slot -- While, Set and Pindex, which is the
argument check_loop already makes for recur. Nothing new reaches a
backend and all three get the form with no edit. The value stays
value-like: the slot is the form's own, and the Local at the end copies
out the way any array-typed expression does.
Row-major is pinned, not incidental: the first dimension is the
outermost loop, and a generator that counts observes it. The fill value
and the generator value are each bound once before any loop starts, so
(array-fill [n] (next-id)) is one call and n copies of its answer.
What falls out for the defvar the note was written about, and neither
half is a carve-out:
(defvar grid [rows [cols u8]] (array-fill [rows cols] 255))
is the spelling that works -- a typed global with a computed
initialiser, which is the startup-lifted path with the init-once guard
that defvar already had, so the fill runs once and the value survives a
re-run like any other computed one. The three-element spelling means
what the 2026-09-20 rule says it means: not a type, so a dyn global, and
a typed fixed array crosses into dyn only as a view of storage that
outlives the view. A freshly built array is a temporary, so it is
refused -- by the element rule where the elements are themselves an
array, by the lifetime rule where they are one of the three scalars a
view carries. Both refusals are the ones any other temporary gets.
The type an array-fill builds never goes through resolve, so resolve's
own guard is asked again where it is built: a fixed array of function
values would be zeroed, and a zeroed function value is a null pointer.
[flan_merged_park] drained the agent's ring on one of the two flags that
wake it. [program_poll] — which an expression sets, by way of [Program.wake]
— polled and went back to sleep; [program_asked] broke out of the loop and
re-entered [flan_program_main] with the queue untouched. A plain
redefinition sets neither, so a body delivered to a parked program was still
in the ring when the run it was delivered for started, and installed at that
run's first frame boundary instead: everything main did before its first
(agent/poll) ran the body the person had already replaced, and the change
showed up one run late. A redefined main is the whole of a run, so it would
have had to be asked for twice.
Both flags drain now, and the exit drains before it leaves. The re-run is
still tested first and cannot be starved: the flag is latched at the top of
the round and nothing in the round can clear it.
The transcript row in test_dev.ml asserted the old ordering by name — two
lines out of the second run, the first of them the stale body — so it is a
line shorter now, and the absence of that line is the claim. The park-note
fixture grew a print of the redefinable body before its first poll, which is
what makes the new row able to see which body the re-run started with.
This is what the note the delivery is answered with has been promising: a
module queued against a park installs no later than the program's next run.
It now installs before that run's first frame rather than during it.
(defmacro do-grid [[r rows c cols] & body] ...) — positional names, a [ ]
pattern wherever an argument is a vector, and & for the tail. The reading of
the list lives in Expand, below both sides that need it: Parse turns it into
the bindings a macro body opens with, and Macro checks a call against the same
reading before expanding it, so arity and shape are refused with the call's own
location rather than with the Loc.from_macro stamp every node of an expansion
carries.
The breaking half: [args] used to bind the whole argument list and now binds
the first argument. The whole list is [& args], and every defmacro in the tree
— prelude, vendor, tests, the elisp fixtures — was migrated to it. One grammar,
not a legacy mode.
[merged_serve] waited up to ten seconds for the program to bind agent.sock
before it started [accept_loop]. The listening socket was already up, so an
editor connected fine and then heard nothing: every first request of every
session cost the whole wait when the program calls (agent/start ...) late —
sand.flan starts it after rl/init-window returns — or never.
Nothing the editor asks needs that socket. In one process a delivery is a
call into flan_agent.c, not a connect, and the two-process daemon has already
waited for the bind in [two_process] and fails if it never comes. What the
wait was for is the sentence a program with no agent deserves, and a sentence
does not have to be in front of the loop to be said. So it is a deadline the
session passes ([agent_check]) rather than a wait it does: read from the
accept loop between connections and from [serve] before each request, because
an editor holds one connection for a whole session and the loop is not
cycling while it is attached. No thread, for the reason lib/dune gives about
what the merged link does to this library's dependencies.
Delivery stays honest either way. A program that links no agent at all
refuses through [over_socket]'s ENOENT, as before. One that has the agent but
has not started it takes the module into the ring and is answered with a note
that promises the poll and not a frame: a program with no (agent/poll) in it
never installs this, and "at its next frame boundary" would be the reply that
makes a redefinition look applied when it is not. C-x C-e's five-second
timeout gets the same distinction instead of asking whether a program that
has not got to its loop yet is calling the poll in it.
And the note a parked program's delivery carries is now said once per park.
A finished program is parked, so re-evaluating while a run's output is on the
screen repeated a paragraph on every C-c C-c. The first delivery of each park
explains itself; the rest say the one line that is the claim. [rerun] clears
the flag as well as [eval] does, so a new park is a new reader.
test_dev.ml's abort asks were a bare Wire.send/Wire.recv pair against a
daemon that the abort itself is killing. In a merged flan dev the daemon
is the program: flan_agent.c's listener writes ok and sets aborting, and
the break loop's next pass _exit(134)s from the program thread while the
editor's reply is still being composed on the serve thread. Nothing
orders the two, so the reply arrives or the socket closes, at random.
When it closed, Wire.recv raised Closed, nothing caught it, and the test
binary died with no FAIL line and every case after it unrun. Four runs
out of four at the null-allocator trap.
Both endings mean the same thing and neither is the assertion: the
waitpid wait underneath each row is what says the program went. aborted
answers None for the end that arrived as an exit; the three sites that
abort a stopped or trapped program go through it. The fourth abort is
refused by a running program and ends nothing, so it is left alone.
trap_park's describe poll is guarded with the other answer: these traps
park because flan_trap_hook is installed, so a socket closing there is
the trap having ended the program instead of stopping it, which is the
failure that row already names. And SIGPIPE is ignored for the
watchdog's reason, so a send into the socket a dead daemon left behind
cannot kill the binary silently from the write side.
The claim classes were built for, made against a real daemon rather than
at the session's report: dev-class.flan is compiled with one method for
one class, a second method is delivered into the live process, and the
call through the generic's cell answers with the new method's body while
the original one goes on answering. That is what a generic being exactly
one top-level name buys, and it is now pinned rather than argued.
Two things came out of writing it, neither of them about classes.
The x86 backend's redefinition module never emitted the per-type dyn
descriptors. Emit.redefinition has always emitted them, by going through
finish; the x86 twin ended at the rodata section and stopped. Nothing
had reached it, because a redefined body had to construct a struct
holding a dyn to need one, and until NoMethod there was no such struct a
compiler-written body could build. What it looks like is not a bad read
at run time but a link failure — desc_of mints a local label, the body
references it, and ld refuses the module with an undefined symbol. One
line, beside the same call in the executable path.
And the thing the test had to be written around: a dyn value answered by
eval-expr does not come back in the reply's :value at all. It renders to
the program's own stdout, which reaches a *later* reply's :output — the
dyn-global rows already read one that way and say so. So every answer
here is compared inside the expression, and what crosses the wire is a
typed 1 or 0. Left as it is; where a dyn expression's value should
surface is a question about the editor protocol, not about this lane.
The session-level test stays: it pins which names an added method
reports for installation, which is the half a daemon test cannot see.
[Check.memory_sites] is the pass [Check.no_gc]'s shape: it runs over the
finished program, answers a diagnostic list, and tells nothing downstream
that it ran. Two classes on the diagnostic's kind — the collector's heap
and an allocator the program named — so the CLI, the daemon and the editor
dispatch on one field and none of them reads a message to find the class.
[--warn-memory] on check and build prints them where errors go, in the
shape flycheck parses, without moving the exit status. [(:op "memory")]
answers the same list over a session's last checked program, needing no
program on the far end. [M-x flan-check-memory] paints it two faces fainter
than an error's, cleared by an edit or by asking again.
Two of the spec's own examples turned out not to allocate, and the
precision rule outranks the enumeration: (vec-new T) passes a capacity of
literal zero to flan_vec_init, which returns before the grow, and
flan_map_init takes no block at all and says so in its own comment. The
block arrives at the first push, which is the line that is marked. The
classifier reads the capacity argument rather than the symbol, which is
what lets slurp be marked through the same entry point vec-new is silent
through. FIX.org has the rest of the evidence.
Ten test binaries share a directory and had shared nothing in it but
watchdog.ml. Everything else each one needed it wrote out again: the
failure counter and its FAIL line, the three-line report tail, a poll,
a socket connect, the wait for a [flan dev] daemon to bind, a substring
search, and the Load -> Check -> Reach.link front half of a compile.
[listening] was the clearest case. Three copies, byte for byte apart
from one comment, and two of them said in that comment that they were
kept separate because "these three files have no module between them".
That was not true when it was written: watchdog.ml was already named in
the same (modules ...) stanzas. test_support.ml is the second such
module, wired the same way, and those two sentences go with the copies
they were explaining.
test_repl.ml's [quote] was Wire.quote character for character, in a file
that already links Wire and already names Wire.quote in a comment about
what the case below it is checking. It is Wire.quote now.
One real behaviour change, and it is a fix. [connect] existed twice over
with different retries: the agent's narrowed to ECONNREFUSED with a
comment saying why -- the socket file appears at bind, a moment before
listen -- while dev's and repl's retried any Unix_error, which meant an
ENOENT or an EACCES was retried to the full timeout before raising
something the reader still had to interpret. The shared one takes the
narrow version. Every caller connects to a socket [listening] has
already seen on disk, so the race it does catch is the only one left.
The rest is left where it is, on purpose. The three output-capturing
[run]s differ in what they wrap -- a pid suffix, a sanitizer environment,
a valgrind invocation -- and are not the same function. The report tails
in test_repl, test_web and the two sweep binaries print different things
for different reasons. The per-file scratch prefixes are the feature that
keeps two suites running at once from unlinking each other's sockets, so
the shared helper takes the prefix rather than choosing one. And the
[match Sys.command "command -v clang ..."] probes stay as they are:
their skip lines are output this suite pins.
bin/main.ml has the compile pipeline written out twice more. Left alone
-- this was a test/-scoped change and bin/ should not be reaching into a
test module -- and noted in FIX.org as what it actually needs, which is
the pipeline moving into lib/.
dune test: exit 0, and its output is the same line for line once the
temp-directory hash and the millisecond counts are normalised.
flan_merged_park called flan_dyn_root_reset, which emptied the collector's
root stack. The frames' roots had to go — main is left by longjmp, so they
name stack the next run overwrites — but the dyn globals' roots are on that
same stack, pushed once by the emitted main and never popped, and the park
took them with the frames.
The park is not a quiet state. It services evaluated thunks, a thunk
allocates, and an allocation collects. So a program with (defvar config dyn)
answered (get config :s) with its string before any thunk ran and with nil
after one that allocated past the heap's floor — a read of memory the sweep
had freed, answering nil by luck of what the freed words decoded as.
The emitted main now brackets its global pushes: flan_dyn_root_globals_begin
empties the stack, the pushes go on, flan_dyn_root_globals_end records how
many of them there are, and the park resets to that line instead of to zero.
Nothing between the two allocates, which is what keeps the globals from being
swept in the window where they are unrooted — and [begin] emptying the stack
rather than adding to it is what makes a re-entered main re-root the same
globals rather than push a second copy of each, which also closes the other
half: a re-run used to re-push roots over slots left dangling by the park.
Both emitters, because the dev loop's default backend is x86 and a fix in one
lowering is not a fix. A program with no dyn globals emits neither call and
its root stack still resets to empty, which is what an empty push list should
leave behind.
flan_dyn_root_pop now clamps at the globals rather than at zero. An
over-popping frame eating the globals is the one way that clamp could turn a
miscount into this same use-after-free.
Covered twice. test/dyn_ops.c's park mode is the runtime's half — a run, a
park with a collecting thunk in it, and another run, three times over,
asserting both that the global survives and that the frame's five hundred
objects do not. Under ASan the old reset reports heap-use-after-free in
flan_dyn_tag with the free in gc_sweep; under memcheck it reports 24 errors
and still prints the right answer, which is the shape of the bug. test_dev.ml
drives the whole daemon over its socket on both backends against
programs/dev-dyn-global.flan.
Not touched, and it wants a decision rather than a patch: a re-run re-enters
flan_program_main, which re-runs the lifted startup function, so every global
with a computed initialiser is reset by a re-run. That contradicts dev.ml's
own note and FIX.org item 1. It is independent of this — the roots are right
whether or not the values are re-initialised.
Nor is this the reload path. A defvar added by an evaluation gets its storage
from flan_dev_global (emit.ml's new_globals, x86.ml's counterpart) and there
is no flan_dyn_root_push anywhere on that path in either backend, so a dyn
global added to a live session is unrooted. That is a separate defect with a
separate fix, and nothing here makes it better or worse.
FLAN_DEV_CLIENT_GRACE dropped from 0.2s to 0.1s, so the daemon's Live
threshold (six graces) is 0.6s instead of 1.2s -- still three ticks
past accept_loop's own 0.2s select granularity, which is where the
margin has to live rather than in the test's own wait. The structural
await scales with it, from three seconds to one: the assertion was
never about the clock, only about serve() sitting in Wire.recv while a
client holds the socket, so the number just had to clear the threshold
with room, not any particular multiple of it.
The other awaits in this test are already ceilings that return the
moment their condition is true, so they cost what the daemon actually
takes and nothing was touched there. A trace of the eval-expr path
elsewhere in this file turned up a second five-second wait, twice over
-- but that one is a documented hang-detection timeout on a program
that structurally cannot reach a frame boundary, the exact shape the
task said to leave alone, so it stayed.
The x86 block asks the break loop's questions of an --x86 host so that the two
sets of answers can be read against each other rather than merely found
plausible. A store is where they could most easily differ: the place forms the
walk ends at are lowered by each backend's own `place', and the two disagree
about an Option — which is why what may be written is settled in session.ml
above both of them and not in either. So the same slot is written, read back,
and refused for a type that does not fit, on this backend too.
And `Check.expressions' carries the comment that was written for the function
it replaced; the one-expression entry beside it has its own line.
`(:op "set")` is `inspect`'s addressing with the arrow turned round: same
frame, same slot index, same path steps, same fingerprint check, and a list of
(path, expression) edits relative to what the buffer is showing. The thunk
stores and then renders the same place, so what comes back is what the program
holds afterwards rather than an echo of what was asked for.
`stopped-only` was not enough for a write. It asks whether the program is
stopped, and for a read that is the whole question — the worst a render can do
against the wrong stop is print something true of a different frame. A module
that *stores* through `flan/dev-slot` reaches its target through whatever
snapshot is on top when the store runs, so a resume and a second stop inside
the ~300ms build window lands it in the same slot index of a different stack.
Not a fault: a plausible shape, in the wrong place, silently. So the agent
grew `at-stop N` beside `stopped-only`, the generation `snap_push` already
mints for the restart machinery, checked on the game thread at the moment the
job is claimed. `stop` answers it, `inspect` carries it out on every reply, and
a write that names a stop the program has left is refused before anything is
built.
Three refusals about where rather than what, and they are in session.ml above
both backends because emit and x86 disagree about two of them. A data type's
case field has no address that does not also settle the tag. An option's
payload has none that does not settle whether there is one. A pointer would be
an address this end made up, which is the blessing the registry exists to
insist on.
`Check.expression` grew a `want` and `Check.expressions` a shared frame: the
first is why `3` into an `f32` field is an f32 three rather than "expected f32,
found i32", and the second is why two edits in one commit are not two `let`s
reading each other's storage.
The backend had a full dev story for everything except the one thing the
break loop reads: emit.ml pushes a shadow-stack frame on entry to every body
in a dev build and x86.ml pushed none, so an --x86 session answered backtrace,
locals, inspect and globals with "this program was not built with --dev",
which was false of it.
emit_fn now builds the same four words runtime/flan_dev.c describes -- the
previous head, a static descriptor, and a table of one address per slot, null
until the binding that fills it has run. The push is emitted at the head of
the body buffer rather than into the prologue's, because the prologue is where
the incoming registers still hold the arguments and every instruction of it
clobbers rax. The pop is one store at the epilogue label: this backend has one
epilogue and a return, the body's tail and the transfer exit all arrive there,
where emit.ml needs the same restore at five separate rets.
Two decisions are taken from emit.ml rather than reasoned out again, so that
the two dev builds answer a stopped frame identically. A function with no
named slot gets no table and reports no slots -- there the gate buys mem2reg
an alloca that does not escape, here it buys nothing but agreement, and
agreement is the only thing the break loop can check. And the descriptor's
bytes do not go through string_const: that counter is what decides whether an
expression thunk's module may be unloaded, and counting a descriptor in it
would pin every C-x C-e module's mapping for ever.
The descriptors are the one constant this backend emits that holds an address,
so they go in .data.rel.ro and not .rodata -- a relocation the loader applies
needs a section it may write, and a redefinition module is such an object.
Verified on an assembled module: R_X86_64_RELATIVE for the two pointers,
flan_frame_head through the GOT, no TEXTREL.
layout_ctx fills its globals table, which it never had to before. Reach's
ref_fingerprint asks whether a name is a global, and with an empty table no
name is one -- every frame would carry the hash of the empty set, the daemon
would recompute the real one, and the globals section would refuse a body
nobody had touched while its locals still read.
Release builds gain nothing: all of it is behind md.dev, and three corpus
programs emit byte-identical assembly before and after.
The new coverage stands up an --x86 daemon on dev-locals.flan and asks the
four questions, against the values the LLVM block above asserts of that same
program -- the claim is that the answers are the same and not merely
plausible. The existing --x86 two-process daemon gains one more: a backtrace
through a redefined body, which reports the installed body's own file, since
that descriptor travels in the module and a frame pointing at the host's would
name the body it replaced.
flan dev takes the hand-written backend unasked now: it is what that
backend was written for, it halves the C-c C-c round trip, and nothing it
builds outlives the session. Every other command is LLVM by default and
unmoved, which is what keeps lib/x86.ml's calling convention licensed.
--debug picks LLVM on its own rather than landing on the refusal: a
redefinition module from this backend carries no line table, so the one
flag someone types to get a debugger must not choose the backend without
one. --x86 --debug together is still refused, and now says which to drop.
What the flip costs is more than the forms this backend refuses. It
pushes no shadow stack, so a session built by it stops on an error and
cannot say where -- and the agent answered that with "this program was
not built with --dev", which is false of an x86 dev host. Dev.ask
rewrites it; the IR view refuses by name instead of failing to find a
define in a .s file; and a refusal at the host build or at C-c C-c names
--llvm rather than stopping at "unsupported".
flan-daemon-args carries the flag from Emacs, spliced into the one
argument list the daemon buffer's first line is now printed from.
`arena-new` is a builtin, so it is in no program's symbol table and `defs`
never mentioned it — which made the editor answer "the running program
defines no arena-new" about a name that works. The fix is not a better
refusal: it is that the seventy-eight names the checker answers without
being told are now in the reply, under a kind of their own.
check.ml carries the table, beside the arms it describes, because a table
in another file drifts from them with nothing said. test_flan reads both
the arms and the table and fails on either having a name the other does
not, in both directions — there is no reflecting over a match, so it reads
the source.
Each entry is a signature in `signature_of_fn`'s shape and one line. The
arms that do not have one shape say what is true instead of pretending:
`?` for an argument that may be left out, `|` for the types an arm really
takes, and the checker's own predicate names for the type-directed ones.
The three user-allocator names carry a bare name and no bracket list,
because they are refused wherever they are written.
`defs` grows a fifth string for the prose, and builtins are appended last
so a completion table does not bury the names being worked on. A defn has
no docstring to put there and does not get one here: the Tast keeps none,
and that is a different piece of work.
The editor reads the kind, not a special case. `flan-doc--where` and the
xref backend both answer before their empty-location branch, because a
global's missing location and a builtin's absent one are different facts
and only one of them is about the daemon.
Three gaps in the first pass. The park flushes after the poll, which is the
only thing that gets a printing expression's line out of a fully buffered pipe
before the next run — now pinned by its position in the transcript. The globals
section under a parked break answers with everything skipped, which is a
different reply from the timeout a job nobody polled would give, so it is asked
for rather than assumed. And the render thunk's five-second sentence no longer
asks a parked program whether it is calling agent/poll.
The park waited on one flag and could do one thing, so C-x C-e on (+ 1 1) was
refused for want of a frame boundary — an expression that needs nothing from
the program, in a process holding every global the run left.
It waits on two now. A re-run leaves the park; a wake drains the agent's ring
and waits again, with the state still PARKED, which is what makes running the
thunk there exactly as safe as running it at a frame boundary: while parked
there is no concurrency to be unsafe against. The break loop is the precedent
and CL's spawned worker is deliberately not copied.
A thunk that stops now stops on a parked thread, so the restart ops refuse on
whether a break is engaged rather than on the state, and a paused expression
against the park is resumable.
[close] was the only way out of the accept loop that anybody ever took, and
an editor that is killed rather than quit never sends it. Four of the eight
orphans found on the author's machine were that.
The signal is the socket, not a new op: Emacs opens one connection for the
whole session and drops it only on paths that send [close] first, so an
editor left open overnight is an editor still attached and the grace cannot
accumulate under it. Armed only once a client has been there, so a daemon
still waiting for its first one is untouched. Parked sessions get the short
clock and live ones six times it, because a parked program is invisible and
a running one is a window somebody may be looking at.
Under --two-process the compiled program is a child of the daemon, and the
daemon is the only thing that ever ends it: the program is a loop with no
reason of its own to stop, and lib/dev.ml kills it from a Fun.protect finally
block on the way out. That block does not run when the daemon is SIGKILLed —
by a harness tearing down a daemon it gave up on, by a watchdog, by anybody —
and what was left behind was a program with ppid 1, sleeping, holding a socket
nothing will ever connect to. Eight of those were on the machine this was
written on, the oldest six days old, and one had been minted by this suite.
The child now arms PR_SET_PDEATHSIG when it starts its agent, gated on a
FLAN_DEV_PARENT the daemon sets and nothing else does, and closes the race
where the daemon died before the arming by comparing getppid against that pid
rather than against 1 — under a subreaper an orphan is not reparented to init.
The signal is SIGPWR and not SIGTERM, because the daemon's ordinary teardown
already sends the child SIGTERM and the child's stderr is the daemon's own; a
handler there would print "the daemon is gone" into the daemon's output on
every clean close. The handler writes one sentence, unlinks the agent socket
it bound, and exits 0, and it does nothing else because it can run on the game
thread inside stdio's lock.
A merged build arms none of this and must not. There the daemon is the
process, so a dead daemon is a dead program; its parent is whoever typed
flan dev, which is a shell or an emacs and not the session's owner. Merged
daemons do leak on this machine — four of the eight — but for a different
reason, which is that accept_loop has no idea whether its editor still exists.
That wants a mechanism that tracks the client rather than the parent, and it
is not this change.
The daemon's read end of the program's stdout pipe is also cloexec now. The
child used to inherit it, which is both a leak and the reason waiting for EOF
on that pipe could never have been the mechanism: the pipe always had a
reader. A printing program now takes SIGPIPE once its daemon is gone, which is
a second net under chatty programs and not a substitute — the test uses
dev-watch, which is silent, so that only the thing under test can end it.
The test spawns the two-process shape, sends one request so that what it kills
is demonstrably serving, reads the program's pid out of the daemon's
/proc children before killing it — it is a grandchild, so waitpid is ECHILD —
and SIGKILLs the daemon. SIGKILL and not SIGTERM: SIGTERM lets the finally
block run and the assertion would be green with this change reverted. It fails
that way, and passes with it, in well under a second.
Six refusals in the runtime called _exit(134) where every other error had
learned to park: no restart by that name, a restart taken with the wrong
arguments or with none, a defer that invoked one, a null allocator, and
free-all on something with no region. Under a merged flan dev the compiler is
in that process, so a program that named a restart nobody established took the
session down with it, which is the one thing the break loop exists to prevent.
They park now. Not through flan_break_hook, which is what bounds and
arithmetic use: that hook may answer by aiming a transfer channel, and these
six are called by emitted code that falls off the end with no channel anywhere
in the call, so a restart chosen against one would be accepted and dropped.
flan_trap_hook says the other thing instead — stop here, let everything be
read, and refuse the resume with a reason.
All six park, for two reasons rather than one. Four are guards that fire
before the operation they guard, so nothing is half done and the frame reads
like any other. The other two fire mid-transfer, with the frame's defers
possibly half run, and they park only to be looked at: stopping on a torn
unwind is strictly more than exiting before anyone can ask what tore it.
The break loop grew a per-snapshot resumable flag for it. Restarts are still
listed and still numbered, the terminal marks them untakeable and the socket
reports the same positions as unreachable, and the listener refuses a choice
with the trap's own sentence rather than the thunk-boundary one.
Standalone builds die exactly as they did: nothing installs the hook in a
program that did not import the agent, and the acceptance case for free-all
still wants exit 134 and the same message.
The review entry that asked for this named flan_exit_hook, which is normal
termination and not this at all; it is struck out with the correction.
The three named arms are not every way a step after the check can fail.
write_file makes the daemon's copy of the module's text before llc is called at
all, and a Sys_error from it went out through serve's guard with the session
already holding the declaration — the same stranded declaration under a
different exception. Dev.eval now restores on the way past anything that
escapes, and carries an accepted flag so that a module the agent has already
taken is never rolled back out from under the process.
That is also the seam the crash can be reproduced through. test_dev takes the
daemon's working directory away from it — its own, not the macro cache the two
failures above use, which cannot fail a build with no macro in it — evaluates a
defn, and then calls it. Without the restore the daemon lists probe-two in
describe and the call closes the socket: the null cell, the game thread, and
address 0.
Two seconds was measured on one warm machine and has no margin against the
6.8s cold build listening already records. The status and the absence of the
frame-boundary sentence are what discriminate; this bound exists only so a
timeout cannot be read as an answer, so it moves to 4.5 and says so.
The merged build's stdout is a 64K pipe back into the daemon's own process,
and the accept loop is the only thing reading it -- which it is not doing
while serve is answering a request. The two five-second waits for a frame
boundary now drain the pipe on every tick, so a program stopped inside fwrite
is one the daemon lets go rather than one it waits out and then accuses of
not calling agent/poll.
drain and not take: the text stays in the buffer until with_output puts it on
the reply, which is where the output an evaluation caused belongs. And the
drain sits beside the sleep rather than inside the select, because a readable
pipe would make the tick free and count the timeout out in a fraction of it.
dev-chatty.flan prints 4K a frame, which is the only fixture here that fills
the pipe at all; without the drain it fails in 5.1s with the old sentence.
The agent refuses a module it has no room to queue with "the program is not
calling agent/poll". For a running program that is the cause. For a parked one
it is confidently wrong: there is no game thread left to poll with, nothing
drains the ring until somebody runs the program again, and the ring is full
precisely because Dev.eval accepts evaluations while parked and promises they
install at the next run. At install sixty-five that promise breaks, and the
reply sent the reader to inspect a loop that is not running.
The agent cannot know this — parking is the merged shim's state — so the
daemon, which has just asked liveness, rewrites that one reply the way
abi_mismatch rewrites dlerror's. Every other refusal is still quoted in the
agent's own words, and a running program's is unchanged.
Pinned on Dev.refusal rather than end to end: filling the ring is sixty-four
real clang runs, and what the daemon does with the agent's words is the whole
of the change.
