Split test error codes out for socket broker test.

Try and determine what could be causing
SBOX_TEST_FAILED_TO_RUN_TEST by splitting this out to individual
failures.

This will allow further diagnosis of why a test is flaky on the bots

BUG=1264188

Change-Id: Ic5afc6f9588f0b7b3a23f1adf490b9e3ff3ff6f3
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/3251432
Commit-Queue: Will Harris <wfh@chromium.org>
Reviewed-by: Alex Gough <ajgo@chromium.org>
Cr-Commit-Position: refs/heads/main@{#936068}
NOKEYCHECK=True
GitOrigin-RevId: 368a4df09ff17103504c22f22cddf11de4af6bb3
1 file changed
tree: dfb7353bbfeef38f98a969d0b4cc10e285d17721
  1. linux/
  2. mac/
  3. policy/
  4. win/
  5. BUILD.gn
  6. COMMON_METADATA
  7. constants.h
  8. DEPS
  9. DIR_METADATA
  10. features.gni
  11. ipc.dict
  12. OWNERS
  13. README.md
  14. sandbox_export.h
README.md

Sandbox Library

This directory contains platform-specific sandboxing libraries. Sandboxing is a technique that can improve the security of an application by separating untrustworthy code (or code that handles untrustworthy data) and restricting its privileges and capabilities.

Each platform relies on the operating system's process primitive to isolate code into distinct security principals, and platform-specific technologies are used to implement the privilege reduction. At a high-level:

  • mac/ uses the Seatbelt sandbox. See the detailed design for more.
  • linux/ uses namespaces and Seccomp-BPF. See the detailed design for more.
  • win/ uses a combination of restricted tokens, distinct job objects, alternate desktops, and integrity levels. See the detailed design for more.

Built on top of the low-level sandboxing library is the //sandbox/policy component, which provides concrete policies and helper utilities for sandboxing specific Chromium processes and services. The core sandbox library cannot depend on the policy component.