reduce static linking overhead from TLS support by inlining mmap syscall

the external mmap function is heavy because it has to handle error
reporting that the kernel cannot do, and has to do some locking for
arcane race-condition-avoidance purposes. for allocating initial TLS,
we do not need any of that; the raw syscall suffices.

on i386, this change shaves off 13% of the size of .text for the empty
program.

diff --git a/src/env/__init_tls.c b/src/env/__init_tls.c
index 5c17681..dbfe62e 100644 (file)
--- a/src/env/__init_tls.c
+++ b/src/env/__init_tls.c
@@ -5,6 +5,7 @@
 #include "pthread_impl.h"
 #include "libc.h"
 #include "atomic.h"
+#include "syscall.h"
 
 #ifndef SHARED
 
@@ -87,8 +88,15 @@ void __init_tls(size_t *aux)
 
        libc.tls_size = 2*sizeof(void *)+T.size+T.align+sizeof(struct pthread);
 
-       mem = __mmap(0, libc.tls_size, PROT_READ|PROT_WRITE,
+       mem = (void *)__syscall(
+#ifdef SYS_mmap2
+               SYS_mmap2,
+#else
+               SYS_mmap,
+#endif
+               0, libc.tls_size, PROT_READ|PROT_WRITE,
                MAP_ANONYMOUS|MAP_PRIVATE, -1, 0);
+
        if (!__install_initial_tls(__copy_tls(mem))) a_crash();
 }
 #else