# Not really sure if this is a real leak or just ASAN not correctly handling the autorelease pool.
# The leak is reported both from the `crashpad_handler` as well as sentry itself.
# Adding a manual `[paths release]` "fixes" the `crashpad_handler` leak, but leads to a use-after-free in sentry.
# https://github.com/getsentry/crashpad/blob/9cd1a4dadb51b31665f5e50c5ffc25bb9d10571a/client/crash_report_database_mac.mm#L705
leak:contentsOfDirectoryAtPath

# This is a known issue introduced with libcurl 7.77.0 on macOS: https://github.com/getsentry/sentry-native/pull/827#issuecomment-1521956265
# While it should have been fixed with 7.78.0, tests with SDK-packaged version 7.85.0 still detect the same leak. 
leak:SCDynamicStoreCopyProxiesWithOptions

# This is a known issue in ASAN packaged with llvm15/16 (and below): https://github.com/google/sanitizers/issues/1501
# I cannot reproduce it with the current brew llvm package (18.1.7). TODO: remove when GHA macOS runner image updates the llvm package.
leak:realizeClassWithoutSwift

leak:Curl_getaddrinfo_ex

# macOS notify framework false positives from fork() during tests.
# The notify framework re-allocates internal data structures in the child
# process after fork(), which LSan reports as leaks since the parent's
# pointers are gone. Upstream: https://github.com/google/sanitizers/issues/764
leak:os_set_*_ptr_init
leak:os_map_*_init

# macOS Objective-C runtime false positive from _fetchInitializingClassList.
# Triggered during lazy class initialization in CoreFoundation/CFNumber.
# Upstream: https://github.com/google/sanitizers/issues/1781
leak:_fetchInitializingClassList
