Opening and parsing on the browser main thread froze the tab for the
whole load; the file selector path already loads in a worker thread,
so boot now uses the same thread, progress modal, and error dialogs.
The preloaded data package made every ?t= load download embed.tracy
as well. The startup glue now fetches embed.tracy only when no URL is
given, into the same /url.tracy slot, and the boot code has a single
open path.
Version and integrity checks throw from the View constructor, which
was outside the try/catch guarding the file open, so a bad argv trace
terminated the process. Catch and report it there as well.
The compare load thread caught only UnsupportedVersion, so a
corrupted or otherwise unloadable second trace escaped the thread as
std::terminate. Catch the same set as the main trace load and route
it to the existing bad-version dialog.
Exception catching was off by default in the web build, so the
try/catch handlers around trace loading were dead code and any
failure unwound through the page.
Loading on the browser main thread cannot start more than the 8
pre-created workers: beyond that, thread creation waits for the main
thread to return to its event loop, which the load never does.
Decompression threads were only an optimization, so on emscripten
spawn workers for at most 4 streams and decompress the rest on the
loading thread.
glfw 3.5.1 enables OBJC build-wide under CMake 4.x, remapping nfd's
nfd_cocoa.m to a language its scope never loaded; generate then fails on
CMAKE_OBJC_COMPILE_OBJECT. Initialize OBJC at the top level so sibling
scopes get the rules.
The raw sumSq paired with a self/running mean yielded the raw variance
inflated by the squared mean gap. Accumulate the vector's own sum of
squares alongside its sum, like compare and frame statistics do.
Per-thread lists were dropped for threads without zones or samples, a rule
from when only profiled threads consumed them; the CPU data graph uses them
for any visible thread, so saved traces silently lost wake and wait data.
Block sizes weren't checked against the file length, so truncated files
made the decompression threads read off the mapping. Validate framing
(u32 size + payload) before the mapping and threads exist; also reject
zero-stream or empty files.
The client decoded sched_switch, sched_waking and drm_vblank_event at
fixed positions behind an 8-byte common header. The kernel can extend the
header, which shifts the field positions.
The format files are generated from the live trace_entry (offsetof/ALIGN),
making them the authoritative layout.
Sections are packed into timeline rows purely by length, so sections
from different categories share rows and the same category can move
between rows along the timeline. When categories mark different kinds of
activity (game levels, loading screens, cutscenes) this makes the row
layout hard to read.
The new "Group by category" option in the section options lays out each
category in its own set of rows, in category identifier order, with the
existing length-first packing applied within a category. Off by default,
so the current layout is unchanged unless the option is enabled. Only
shown when the trace has more than one section category.
Pandoc 3.11 appends the LaTeX caption label to Markdown table
captions as a {#id} identifier; strip it in tablecaption() since the
manual renderer has no identifiers.
