An eighth sweep, over the part no earlier round named: the 6,300 lines of
tests, and once more over the seventh sweep's own commit, which was itself
unreviewed.
Four of the shrunk fuzz fixtures name one widget two or three times. `width`,
`sized` and `align` set a rule on the widget they are given and return its own
id -- only `pad` and `wrapper` make a new one -- so `let sized =
wrapped.width(76).add(..)` and the `let aligned = sized` beside it are three
names for one text. Each name then went into the list of ids the case compares
warm against cold, so a case that says it checks six boxes checks four, and
three doc comments quote that inflated count as the size of the tree the
shrinker reduced to. Measured: `plant` and `plant_fixed` list 6 and hold 4,
`plant_pair` lists 4 and holds 3, `plant_scrolled` lists 8 and holds 7. The
aliases are gone and the counts say what the fixtures build; each rebuilt
fixture was diffed against the old one, and both the widget slots and every
region are identical, for both settings of `swapped`.
`assert_same_regions` sits at the top of `unsettled.rs` and six tests call it.
Seven more spell its body out instead, byte for byte. They call it now, and it
is `#[track_caller]` so the panic names the case.
`tests/gpu/mod.rs` holds the adapter probe and the surface configuration that
`draw_cost` and `chain_cost` had a copy of each -- `config` identical, and the
probe identical but for the feature it asks for. The leak's justification lived
in one file with the other referring to it; it now sits on the thing it is
about. Shared through `#[path]`, the way `scenario/mod.rs` already is.
The mask a widget is clipped by was resolved in three places, two of them a
byte-identical closure. `mask_bounds` takes the slot rather than the widget,
because the third site deliberately reads the slot it saved before the frame:
that a redraw keeps the slot is what it is checking.
`Layered::_revision` was a field nothing reads, incremented to mark the widget
dirty. Two tests in the same file already do that with
`get_dyn_mut`, which is what the underscore was hiding.
`plan.rs` claimed every simplification is strictly smaller, and asserted `<=`.
Measured: 53 of one tree's 101 simplifications keep the widget count, since a
dropped alignment and a simpler leaf both do. The assertion is right and the
claim was not; the comment now gives the argument that does hold.
`generated.rs` said "Seven that have never failed" and "the nine the others
check" of a ten-seed array. The `should_panic` scroll test ended in an
`h.frame()` that cannot run, since `set_root` lays out and is where the panic
comes from. Two `drop(tree)` at the end of their own scope did nothing.
Format, clippy with and without layout-diagnostics, and the suite (131 + 19 +
13 + 4) are clean. The cold dump over 400 depth-5 trees is byte-identical to
f8aa0c5 across all 34,490 boxes. No library code changed, so the seed scans
have nothing to find. Both GPU rigs were rebuilt and run: chain cost +470% at
depth 64, draw cost ~4.4 us per layer.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Two findings from a sweep over the WGSL prelude and the position widgets,
scoped against upstream/main at ca2b4b2.
`module_source` already builds each shader's preamble from iris_core's own
constants, so the move-chain work's second copy of `MOVE_NONE` and
`CHAIN_LIMIT` -- under "keep in step with iris_core::CHAIN_LIMIT" -- asked a
reader by hand for what the mechanism beside it exists to do. Both are
injected now, with `MASK_NONE` beside them replacing a bare literal, and the
shader declares none of them.
`Scroll`'s `content_len` is never less than its box, so `slack` and the
`anchor` computed from it were always zero whatever the alignment: the
framework centres short content by placing the answer in the whole box, and
the comment credited arithmetic that could not have done it. The same belief
guarded the fits-in-the-box contract with `align == NEG`, so at the default
alignment -- the middle -- every box change redrew the scroll, measured as 1
widget against 0 at TOP_LEFT. `align` now has no reader at all.
`UiSpan::translated` and `UiRegion::translated` are reachable only from each
other and from nothing else.
Format, clippy with and without layout-diagnostics, and the 131-test suite
are clean. The cold dump over 400 depth-5 trees is byte-identical to
1096c31, and all three seed scans pass: 400 at depth 5 in 69.07s, 1,000 at
depth 6 in 169.29s, 2,000 at depth 4 in 300.75s.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
`Branch` picks which of two subtrees to draw by comparing a measured pixel
length with a threshold. That is exactly what the oracle wants -- it is how a
widget believing a measurement a cold start would not have given it becomes a
different tree -- and exactly what a rig measuring cost must not have: the
fixture's shape moves with the thing being measured.
It has been moving. Seed 1 at depth 8 draws 88 widgets and writes 2,298
primitives a frame at `5ed9e87`, and 115 and 8,209 at `bd6de71` -- three and
a half times the work -- so the handoff's "fixed point cost 3x" compared two
different workloads and is withdrawn. Measured on one tree instead, with
`Edits::fixed_branches`, `5ed9e87` is 1,761M instructions and ~699M cycles
against this head's 2,093M and ~819M, while drawing 100 widgets against 97
and writing 4,272 primitives against 3,951. Fixed point costs something like
a fifth to a quarter, not three times.
The oracle keeps measured branches: `fixed_branches` is false by default and
only the rig sets it. A branch consumes its randomness either way, so both
grow the same ids.
**Check the work counters before comparing two commits' times.** The rig
prints drawn widgets, widget draws and primitive writes for this reason;
an undrawn `leftover` child still moves them, which no flag can remove.
Checked: fmt, clippy, 105 tests, the 100-seed generated oracle.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
`UiScalar` was `Len` without the `leftover` weight, which is the separation
canonical `main` already had as `Len` beside `LayoutLen` and this branch
collapsed. It is needed back for the queued clamp: a cap may not contain a
share, because a cap has to read the report a rule otherwise makes moot, and
a share puts the container's division into the same equation -- two
self-consistent assignments, which is the multiple-fixed-point failure
generated seed 13 punished for orthogonal sizing. `min(report, cap)` is not
a `LayoutLen` either: it is a sum of parts, and the smaller of two of them
is not one.
So `UiScalar` is `Len`, what was `Len` is `LayoutLen`, and the two say in
their docs which is which: a `Len` is pixels plus a fraction of a box -- a
position being the length from the box's start, which is why a span is two
of them -- and a `LayoutLen` is a `Len` plus a claim only a container
dividing its room can answer. `From<Len> for LayoutLen` is the one-way step
between them.
Names only; the shader's `UiScalar` is renamed with them. Checked: fmt,
clippy, 105 tests, and `tabs`, `minimal`, `view`, `text` and `random`
byte-identical at 1920x1200.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
`.width()` built a `SetSize` whose whole job was to answer `size_hint`, so
every declared length cost a widget, an `ActiveData` and a link of chain to
say one number. It is now a `SizeRule` per axis on `WidgetData`, beside
`region_node`, resolved by `Painter` where the widget is drawn. `SetSize`
and `MaxSize` are gone; `MaxSize` had no caller but its own builders.
That settles which of two answers is the size. A rule wins on the axis it
names and the `Size` returned by `draw` answers the rest, applied once in
`draw_inner` rather than by each widget that could carry one -- so the
widget under a rule never learns of it. `Painter::size_hint` reads the rule
first for the same reason: a rule that beats what a widget would draw has
to beat what it says about itself.
`declared_lens` still falls back to a non-leftover `size_hint`, which is
how an image or a gap gets its own pixel size rather than the whole offer.
That is the offer's business rather than a declaration's, and it falls away
when a widget occupies its reported size inside the box it was offered.
`known` and `declared` are separate because a share is a length to whoever
divides one and not to whoever composes a box: `.width(leftover(3))` is
known without drawing but cannot narrow anything.
Checked: fmt, clippy, 85 tests, and 100 generated seeds agreeing warm
against cold in 67.6 s. `minimal`, `text` and `view` render byte-identical
at 1920x1200; `tabs` differs only in the widget count it prints about
itself, which is two wrapper types smaller.
`dp` is coming, and then `abs` says which of the two it is not. The
component has always been a pixel count, so the name that admits it is
the one that leaves room for a second unit beside it.
Mechanical: the field on `Len` and `UiScalar`, their constructors,
`to_abs`/`get_abs`, the matching WGSL struct member and the locals
composing it. Field order and types are unchanged, so the `Pod` layout
the shader reads is the same bytes. `f32::abs` is untouched.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Every generated case changes one thing: four declared sizes, or one
span's children, or the output. A frame settling one dependency path
says nothing about a frame settling a set of them that overlap, which
is the case the settle order exists for.
So two more: every declared size in the tree changing at once, and a
spread of widgets marked for redraw together. The second changes
nothing, which is the point -- no box may move, and the order the
dirty set is taken in is all that can make one. The hundred-seed sweep
is 1,000 comparisons now rather than 800, and passes.
`IRIS_PHASE=many` is the same load for the diagnostics rig, with
`IRIS_DIRTY` widgets marked per frame. It says what one repainting leaf
cannot: at 130 of 260 widgets, choosing which dirty widget to settle
next is 24.5% of the frame, because the dirty set is scanned once per
widget settled and a hash set is walked by capacity rather than by
length. Memoizing the depth walk inside one scan does not pay -- it
trades parent lookups for memo lookups and costs 4% more instructions --
so the fix is to stop rescanning, which changes the order widgets
settle in and wants agreeing first.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
A container measures a child by drawing it in a box it may not keep, so
one layout asks a text for a dozen widths and comes back to widths it
has already had -- the hottest text in the depth-8 tree draws 32 times.
Each ask re-ran the shaper, because the two caches in front of it held
one entry each and a trial width alternating with a final width evicts
the answer about to be wanted again. `perf record` put 63% of a resize
frame in text and 0.9% in `draw_inner`.
So keep more than one: a bounded store of shapings on `TextData`, keyed
by the text, the attrs and the width, holding the parley layout and the
glyphs placed from it. Bounding the store rather than each buffer is
what keeps it a fixed cost -- +4 MB on a tree of 4,000 texts, which is
19 MB less than the code before #16 holds after the same resizes.
`TextBuffer` now holds the glyphs of the shaping it is drawn as, which
is where `TextView::tex` was. That leaves one place to invalidate rather
than two, so the `MutDetect` flags on a view's text and attrs have no
reader and go, along with the `buf.changed = true` after every edit.
On a 40-row tree of distinct random paragraphs, 500 resize frames:
124.2M instructions per frame before, 17.7M after, and 45.9M when the
width never repeats. The five reference renders and the resize render
are byte-identical, and the 100-seed sweep passes.
`tests/revision_cost.rs` is that tree, written in the API subset
`43ce8c7` shares so the same source measures the code this replaced.
Report the worst frame and p99 beside the median, since a stutter is
what somebody sees. Count glyph placements, and count a text render per
ask rather than per shaping, so the store cannot hide how many times a
layout drew the same text.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
iris-aiandClaude Opus 5