perf: a file view re-reads the file only when a structural change really dropped it

Three tightenings, all measured with the reloads now reported on stderr.

`set_math_channels()` treats an identical list as no change at all. It rebuilt
unconditionally, and a rebuild replaces the buffer and drops its history -- which a file
view then has to re-read from disk. The comparison is the graphs', and it leaves the nodes'
`x`/`y` out deliberately: those are the canvas's own, no op reads them, and a graph that
merely *moved* should not cost the buffer.

`reload_file_view()` fires only when the store is actually empty. A change that really
happened leaves it so in a file view (the rebuild replaces it, and there is no producer to
refill it); one that found nothing to do leaves it full, and reading a 200 MB file again for
that is the waste. `ScopeModel::stored_slots()` is the one-line accessor it needed.

Every reload now says so, because a second read of a file is worth accounting for:

    [udpscope] reloaded <path> after a structural change: 402 samples back

Measured, `--open` a file:

    session with no graphs            0 reloads
    the math probe (3 real changes)   3 reloads, and all the readings stay equal
                                      (`math x2 402 samples vs 402 time slots`)

I had offered this as "one extra read at `--open` startup", and that was wrong: with no
session there is no `apply_math_channels()` call at all, so there was never an extra read
to remove. What the tightenings buy is that a no-op application is now free, and that a
reload cannot happen unless the buffer was really lost. Each *genuine* change still costs
one re-read -- a rebuild drops the history by design, and preserving it across a rebuild is
a model change rather than this one.

docs/claude/invariants-math-channels.md

@@ -42,7 +42,13 @@ Part of [CLAUDE.md](../../CLAUDE.md). These are rules that cost a defect to lear
   channel to an open waveform therefore emptied *every* channel -- `samples=0` on the sources too, which a user
   reads as "the plot is broken" -- and the same emptiness follows any structural change in a file view.
   `reload_file_view()` (in `ui.cpp`) re-reads the file after such a change, which is also what computes the
-  derived channels over it: the load's own pushes run the math once the set is in place. It is why `--open`
+  derived channels over it: the load's own pushes run the math once the set is in place. It fires only when the
+  buffer was *really* dropped -- `stored_slots() > 0` means the change found nothing to do -- and it says so
+  on stderr (`reloaded <path> after a structural change`), because a second read of a file is worth
+  accounting for. `set_math_channels()` in turn treats an identical list as no change at all, so a no-op
+  application (the session apply, a re-open) neither rebuilds nor drops nor re-reads; the comparison leaves the
+  nodes' `x`/`y` out on purpose, since those are the canvas's own and no op reads them -- a graph that only
+  *moved* is not a change, and rebuilding for it would drop the buffer to store two numbers. It is why `--open`
   works at all with a session that defines math channels, since `main()` loads the file *before* the UI (and
   so before the session's graphs) exists.
 - **Every path that fills a store must skip a derived channel, not just the packet path.**

src/math_channel.hpp

@@ -58,6 +58,17 @@ struct MathNode {
   double hi = 1.0;
   double x = 0.0;       ///< the editor's box position, logical pixels.
   double y = 0.0;
+
+  /// Whether two nodes are the same node for the *engine's* purposes.
+  ///
+  /// `x`/`y` are the canvas's own and no op reads them, so a graph that only moved is not a change:
+  /// rebuilding for it replaces the buffer and drops its history -- and in a file view that means
+  /// re-reading the file -- to store two numbers nothing but the editor draws with.
+  bool operator==(const MathNode& other) const {
+    return op == other.op && source == other.source && in_a == other.in_a && in_b == other.in_b &&
+           gain == other.gain && offset == other.offset && fc == other.fc && q == other.q &&
+           lo == other.lo && hi == other.hi;
+  }
 };
 
 /// A derived channel: a graph of nodes and which node is the answer.
@@ -67,6 +78,11 @@ struct MathChannel {
   bool enabled = false;
   std::vector<MathNode> nodes;
   int output = -1;      ///< the node whose value is the channel's.
+
+  bool operator==(const MathChannel& other) const {
+    return name == other.name && unit == other.unit && enabled == other.enabled &&
+           output == other.output && nodes == other.nodes;
+  }
 };
 
 /// The per-channel runtime: the filter memory and the per-node scratch. Kept apart from the

src/model.cpp

@@ -257,6 +257,11 @@ void ScopeModel::append_math_signals_locked() {
 
 void ScopeModel::set_math_channels(std::vector<MathChannel> channels) {
   std::lock_guard<std::mutex> lock(mutex_);
+  // An identical list is not a change. The rebuild below replaces the buffer and drops its history --
+  // which a file view then has to re-read from disk -- and `apply_math_channels()` runs this on every
+  // session start, where a session with no graphs hands over an empty list against a model that has
+  // none either. Comparing the graphs is what makes that call free.
+  if (channels == math_) return;
   math_ = std::move(channels);
   // Lift the previous synthetic entries out of the signal list; `append_math_signals_locked()`
   // puts the current ones back, so an edit is the same kind of structural rebuild as a CONFIG.
@@ -314,6 +319,11 @@ void ScopeModel::push_loaded_slot(const std::vector<f32>& values, f64 t, f64 abs
   push_math_sample(slot);
 }
 
+std::uint64_t ScopeModel::stored_slots() {
+  std::lock_guard<std::mutex> lock(mutex_);
+  return store_ ? store_->time().size() : 0;
+}
+
 void ScopeModel::push_math_sample(const std::vector<f32>& sample) {
   // The common case is no derived channels at all, and this runs once per *stored slot* -- so the
   // early-out is what keeps a stream with no math from paying for the feature.

src/model.hpp

@@ -281,6 +281,10 @@ class ScopeModel {
   /// evaluated from those, in order, so a math channel may use another one.
   void push_math_sample(const std::vector<f32>& sample);
 
+  /// How many slots the store currently holds. A file view reads it to tell a structural change that
+  /// dropped the buffer (0) from one that could not have (see `reload_file_view()` in ui.cpp).
+  std::uint64_t stored_slots();
+
   /// Pushes one stored slot of a **file** load -- the values, the shared clock and the absolute
   /// clock -- and then the derived channels, in the producer's own order and under the model's own
   /// lock. The loader cannot simply hold the lock and push into `live()` itself: every setter it

src/ui.cpp

@@ -4190,6 +4190,10 @@ gboolean on_app_key_pressed(GtkEventControllerKey*, guint keyval, guint, GdkModi
 /// math once the set is in place.
 bool reload_file_view(App& app) {
   if (!app.replaying && app.model->config().open_waveform.empty()) return false;
+  // Nothing to restore: a change that really happened leaves the store empty in a file view -- the
+  // rebuild replaces it and there is no producer to refill it -- while one that found nothing to do
+  // leaves it full. Reading the file again for the second kind is the waste this avoids.
+  if (app.model->stored_slots() > 0) return false;
   const std::string path =
       !app.waveform_name.empty() ? app.waveform_name : app.model->config().open_waveform;
   if (path.empty()) return false;
@@ -4199,6 +4203,10 @@ bool reload_file_view(App& app) {
     set_notice(app, "cannot reload " + path + " after the change: " + error, 6);
     return false;
   }
+  // Said out loud because it is the one place a file is read twice: a structural change really did
+  // drop the buffer, and the file is the only thing that can put it back.
+  std::fprintf(stderr, "[udpscope] reloaded %s after a structural change: %llu samples back\n",
+               path.c_str(), static_cast<unsigned long long>(report.samples));
   return true;
 }