Simplify TLS and reduce startup latency

This change simplifies the thread-local storage support code. On Windows
and Mac OS X the startup latency of __enable_tls() has been reduced from
30ms to 1ms. On Windows, TLS memory accesses will now go much faster due
to better self-modifying code that prevents a function call and acquires
our thread information block pointer in a single instruction.
This commit is contained in:
Justine Tunney 2022-07-18 03:33:32 -07:00
parent 38c3fa63fe
commit b1d9d11be1
15 changed files with 136 additions and 312 deletions

View file

@ -72,7 +72,7 @@ privileged void ftracer(void) {
long stackuse;
struct FtraceTls *ft;
struct StackFrame *sf;
ft = (struct FtraceTls *)(__get_tls_inline() + 0x08);
ft = (struct FtraceTls *)(__get_tls_privileged() + 0x08);
if (_cmpxchg(&ft->once, false, true)) {
ft->lastaddr = -1;
ft->skew = GetNestingLevelImpl(__builtin_frame_address(0));