diff options
author | Eric Snow <ericsnowcurrently@gmail.com> | 2023-04-24 17:23:57 -0600 |
---|---|---|
committer | GitHub <noreply@github.com> | 2023-04-24 17:23:57 -0600 |
commit | df3173d28ef25a0f97d2cca8cf4e64e062a08d06 (patch) | |
tree | f2b6f378f81ceee48a9e710154b9d6c4b0f959a2 /Include/internal/pycore_obmalloc_init.h | |
parent | 01be52e42eac468b6511b56ee60cd1b99baf3848 (diff) | |
download | cpython-git-df3173d28ef25a0f97d2cca8cf4e64e062a08d06.tar.gz |
gh-101659: Isolate "obmalloc" State to Each Interpreter (gh-101660)
This is strictly about moving the "obmalloc" runtime state from
`_PyRuntimeState` to `PyInterpreterState`. Doing so improves isolation
between interpreters, specifically most of the memory (incl. objects)
allocated for each interpreter's use. This is important for a
per-interpreter GIL, but such isolation is valuable even without it.
FWIW, a per-interpreter obmalloc is the proverbial
canary-in-the-coalmine when it comes to the isolation of objects between
interpreters. Any object that leaks (unintentionally) to another
interpreter is highly likely to cause a crash (on debug builds at
least). That's a useful thing to know, relative to interpreter
isolation.
Diffstat (limited to 'Include/internal/pycore_obmalloc_init.h')
-rw-r--r-- | Include/internal/pycore_obmalloc_init.h | 6 |
1 files changed, 5 insertions, 1 deletions
diff --git a/Include/internal/pycore_obmalloc_init.h b/Include/internal/pycore_obmalloc_init.h index c9f197e72d..8ee72ff2d4 100644 --- a/Include/internal/pycore_obmalloc_init.h +++ b/Include/internal/pycore_obmalloc_init.h @@ -54,9 +54,13 @@ extern "C" { # error "NB_SMALL_SIZE_CLASSES should be less than 64" #endif -#define _obmalloc_state_INIT(obmalloc) \ +#define _obmalloc_global_state_INIT \ { \ .dump_debug_stats = -1, \ + } + +#define _obmalloc_state_INIT(obmalloc) \ + { \ .pools = { \ .used = _obmalloc_pools_INIT(obmalloc.pools), \ }, \ |