TLS (GNU/C11 thread-local storage) support for static-linked programs

the design for TLS in dynamic-linked programs is mostly complete too,
but I have not yet implemented it. cost is nonzero but still low for
programs which do not use TLS and/or do not use threads (a few hundred
bytes of new code, plus dependency on memcpy). i believe it can be
made smaller at some point by merging __init_tls and __init_security
into __libc_start_main and avoiding duplicate auxv-parsing code.

at the same time, I've also slightly changed the logic pthread_create
uses to allocate guard pages to ensure that guard pages are not
counted towards commit charge.

2 files changed, 73 insertions, 0 deletions

diff --git a/src/env/__init_tls.c b/src/env/__init_tls.c
new file mode 100644
index 00000000..4a840cf2
--- /dev/null
+++ b/src/env/__init_tls.c
@@ -0,0 +1,71 @@
+#include <elf.h>
+#include <limits.h>
+#include "pthread_impl.h"
+#include "libc.h"
+#include "atomic.h"
+
+#ifndef SHARED
+
+static void *image;
+static size_t len, bss, align;
+
+void *__copy_tls(unsigned char *mem, size_t cnt)
+{
+	mem += ((uintptr_t)image - (uintptr_t)mem) & (align-1);
+	memcpy(mem, image, len);
+	return mem + len + bss;
+}
+
+static void *simple(void *p)
+{
+	*(void **)p = p;
+	return __set_thread_area(p) ? 0 : p;
+}
+
+weak_alias(simple, __install_initial_tls);
+
+void *__mmap(void *, size_t, int, int, int, off_t);
+
+#if ULONG_MAX == 0xffffffff
+typedef Elf32_Phdr Phdr;
+#else
+typedef Elf64_Phdr Phdr;
+#endif
+
+#define AUX_CNT 6
+
+void __init_tls(size_t *auxv)
+{
+	size_t i, aux[AUX_CNT] = { 0 };
+	unsigned char *p, *mem;
+	size_t n, d;
+	Phdr *phdr, *tls_phdr=0;
+	size_t base = 0;
+
+	for (; auxv[0]; auxv+=2) if (auxv[0]<AUX_CNT) aux[auxv[0]] = auxv[1];
+	p = (void *)aux[AT_PHDR];
+	for (p=(void *)aux[AT_PHDR]; aux[AT_PHNUM]--; p+=aux[AT_PHENT]) {
+		phdr = (void *)p;
+		if (phdr->p_type == PT_PHDR)
+			base = aux[AT_PHDR] - phdr->p_vaddr;
+		if (phdr->p_type == PT_TLS)
+			tls_phdr = phdr;
+	}
+	if (!tls_phdr) return;
+
+	libc.tls_size = len+bss+align+4*sizeof(size_t)+sizeof(struct pthread);
+
+	image = (void *)(base + tls_phdr->p_vaddr);
+	len = tls_phdr->p_filesz;
+	bss = tls_phdr->p_memsz - len;
+	align = tls_phdr->p_align;
+	if (align < 4*sizeof(size_t)) align = 4*sizeof(size_t);
+	mem = __mmap(0, libc.tls_size, PROT_READ|PROT_WRITE,
+		MAP_ANONYMOUS|MAP_PRIVATE, -1, 0);
+	if (mem == MAP_FAILED) a_crash();
+
+	if (!__install_initial_tls(__copy_tls(mem, 0))) a_crash();
+}
+#else
+void __init_tls(size_t *auxv) { }
+#endif
diff --git a/src/env/__libc_start_main.c b/src/env/__libc_start_main.c
index f31222b2..a38569d5 100644
--- a/src/env/__libc_start_main.c
+++ b/src/env/__libc_start_main.c
@@ -1,5 +1,6 @@
 #include "libc.h"
 
+void __init_tls(size_t *);
 void __init_security(size_t *);
 
 int __libc_start_main(
@@ -15,6 +16,7 @@ int __libc_start_main(
 	libc.ldso_fini = ldso_fini;
 	libc.fini = fini;
 
+	__init_tls((void *)auxv);
 	__init_security((void *)auxv);
 
 	/* Execute constructors (static) linked into the application */