perf(proot): enable syscall patching by default

This commit is contained in:
Kurt Himebauch committed 2026-10-27 14:29:58 -04:00
1 parent 4b230cbd0f
commit 785d1bf1da
2 files changed
+15 -9

No files matched your search

+14 -8
View File
@@ -134,10 +134,11 @@ netlink and SysV shm (`netif.c` and `sysv.c` already cover these); and the long-
## First implementation and Thor validation (2026-10-27)
`tools/proot/fastpath/svc-patch.h`, included by `fastpath.c`, implements an opt-in first
slice of item 1. Set `PROOT_FP_PATCH_SVC=1` in a fresh process using the replacement
`libblfastpath.so`. The ordinary fast-path configuration and tracer-key check must also
succeed. It is disabled by default, skips processes that already have multiple threads,
`tools/proot/fastpath/svc-patch.h`, included by `fastpath.c`, implements the first
slice of item 1. It runs automatically in fresh processes using the replacement
`libblfastpath.so` when the ordinary fast-path configuration and tracer-key check
succeed. Set `PROOT_FP_PATCH_SVC=0` to disable it for troubleshooting or comparisons;
`PROOT_FP_PATCH_SVC=1` also explicitly enables it. It skips processes that already have multiple threads,
and declines SVE/SME hardware because the relay preserves the base ARM64 register state.
The scanner visits executable segments of the loaded `libc.so.6` and
@@ -180,6 +181,11 @@ NZCV, FPCR/FPSR, four simultaneous threads, the multi-thread patching guard, and
Null arguments and unsupported flags fall back to PRoot. The instruction decoder
and exact ±128 MiB branch limits also have a host regression test.
Default-policy validation on the Thor confirmed that an unset `PROOT_FP_PATCH_SVC`
and explicit `1` both patched 14 sites; explicit `0` left all 415 libc and 56 linker
syscall sites unchanged. The register/thread/fork check also passed with the
environment flag unset.
Five alternating pairs of `bench/svc-workload.c` runs compared the same replacement
library with patching off/on, with fast-path stats enabled in both. The workload
gets `fstatat` and `__open64_nocancel` directly from libc, bypassing the preload's
@@ -234,7 +240,7 @@ the start and completed with `No Error`. Existing installed games were retained.
The repeat off/on pair used the exact same replacement library, SHA-256
`ab605c021aee271cd49fa1cb337d7ff0c762dbac198b711e9818639c0cd4b17d`.
The first off run used the earlier adjacent-only build, with patching disabled.
Only the opt-in environment flag changed for the final pair. Times below come
Only the patching environment flag changed for the final pair. Times below come
from Steam's one-second-resolution `content_log.txt`; total starts at Reconfiguring
and ends at scheduler completion, excluding session startup:
@@ -281,7 +287,7 @@ The installed PRoot pack emitted no profiling start/counter records despite the
queued profiling setting; the profiler report had `hasCounters: false`. That is a
measurement gap, not evidence of zero remaining traps.
Before enabling this by default: validate signal/cancellation/unwind behavior,
Further validation includes signal/cancellation/unwind behavior,
additional glibc builds and workloads, startup cost and Steam/CEF tree RSS; obtain
matching patch-0018 tracer measurements; and resolve the existing SD socket mismatch.
Additional syscall families, Android-blocked syscall emulation and build-id caching
@@ -305,7 +311,7 @@ Run `svc-probe` inside the app's guest environment, not from `adb shell`.
`svc-check` embeds the fast path and requires its normal enabled tracer environment.
For `svc-workload`, pass a prepared directory containing `file` and a `link` symlink
to it; add any second argument to enable timing. Use the documented library override
workflow in [proot-performance](proot-performance.md) for a fresh session, adding
`PROOT_FP_PATCH_SVC=1` through `droiddeckctl env set` before starting it.
workflow in [proot-performance](proot-performance.md) for a fresh session. No extra
environment setting is needed; use `PROOT_FP_PATCH_SVC=0` for the unpatched comparison.
`svc-storage` takes a new, unused directory name; it creates and removes its own test
files and directory. It writes 64 MiB plus 16 MiB of file extensions per run.