fix(syscalls): review follow-ups for the bump allocator default

.github/workflows/benchmark-pr.yml

@@ -12,6 +12,13 @@ on:
       - 'executor/**'
       - 'bin/cli/**'
       - 'tooling/ethrex-fixtures/**'
+      # syscalls is linked into the guest ELF this job builds, so a change confined to
+      # it changes the bytes proven — a guest allocator swap moves cycles on every
+      # workload. Without it main's baseline would stay stale until some prover file
+      # happened to change, and the comparison guard would suppress the table until
+      # then. pr_main.yaml:99 already hashes 'syscalls/**' into the guest-ELF cache
+      # key; the two lists must agree on what rebuilds the guest.
+      - 'syscalls/**'
       # A baseline is only valid for the workload it measured, and the Makefile is
       # what defines that workload: it names the block and pins the URL and sha256
       # of the .bin this job fetches. Without it a repointed block would leave

syscalls/Cargo.toml

@@ -12,8 +12,11 @@ lazy_static = "1.5.0"
 rand = "0.9.2"
 # Doug Lea's malloc, behind `dlmalloc-alloc`: slower than the default bump allocator on
 # every workload measured, but it reclaims freed memory, so it is the allocator to pick
-# for continuations. `critical-section` gives the Sync a #[global_allocator] static
-# needs; its single-hart impl comes from `riscv` above. See `src/allocator.rs`.
+# for a guest whose cumulative allocation has no per-execution bound. Not a
+# continuations criterion: continuations are a prover-side split of a single guest
+# execution and change nothing about what the guest allocates. `critical-section` gives
+# the Sync a #[global_allocator] static needs; its single-hart impl comes from `riscv`
+# above. See `src/allocator.rs`.
 dlmalloc = { version = "0.2.14", default-features = false, optional = true }
 critical-section = { version = "1.2", optional = true }