diff --git a/tools/include/nolibc/Makefile b/tools/include/nolibc/Makefile
index 1958dda..7455097 100644
--- a/tools/include/nolibc/Makefile
+++ b/tools/include/nolibc/Makefile
@@ -17,19 +17,17 @@ endif
 # it defaults to this nolibc directory.
 OUTPUT ?= $(CURDIR)/
 
-ifeq ($(V),1)
-Q=
-else
-Q=@
-endif
-
-arch_files := arch.h $(wildcard arch-*.h)
+architectures := arm arm64 loongarch m68k mips powerpc riscv s390 sh sparc x86
+arch_files := arch.h $(addsuffix .h, $(addprefix arch-, $(architectures)))
 all_files := \
+		byteswap.h \
 		compiler.h \
 		crt.h \
 		ctype.h \
 		dirent.h \
 		elf.h \
+		endian.h \
+		err.h \
 		errno.h \
 		fcntl.h \
 		getopt.h \
@@ -96,12 +94,10 @@ help:
 
 # installs headers for all archs at once.
 headers:
-	$(Q)mkdir -p "$(OUTPUT)sysroot"
 	$(Q)mkdir -p "$(OUTPUT)sysroot/include"
 	$(Q)cp --parents $(arch_files) $(all_files) "$(OUTPUT)sysroot/include/"
 
 headers_standalone: headers
-	$(Q)$(MAKE) -C $(srctree) headers
 	$(Q)$(MAKE) -C $(srctree) headers_install INSTALL_HDR_PATH=$(OUTPUT)sysroot
 
 CFLAGS_s390 := -m64
diff --git a/tools/include/nolibc/arch-arm.h b/tools/include/nolibc/arch-arm.h
index 251c425..a4d3a77 100644
--- a/tools/include/nolibc/arch-arm.h
+++ b/tools/include/nolibc/arch-arm.h
@@ -50,7 +50,7 @@
 
 #endif /* end THUMB */
 
-#define my_syscall0(num)                                                      \
+#define __nolibc_syscall0(num)                                                \
 ({                                                                            \
 	register long _num  __asm__(_NOLIBC_SYSCALL_REG) = (num);             \
 	register long _arg1 __asm__ ("r0");                                   \
@@ -67,7 +67,7 @@
 	_arg1;                                                                \
 })
 
-#define my_syscall1(num, arg1)                                                \
+#define __nolibc_syscall1(num, arg1)                                          \
 ({                                                                            \
 	register long _num  __asm__(_NOLIBC_SYSCALL_REG) = (num);             \
 	register long _arg1 __asm__ ("r0") = (long)(arg1);                    \
@@ -84,7 +84,7 @@
 	_arg1;                                                                \
 })
 
-#define my_syscall2(num, arg1, arg2)                                          \
+#define __nolibc_syscall2(num, arg1, arg2)                                    \
 ({                                                                            \
 	register long _num  __asm__(_NOLIBC_SYSCALL_REG) = (num);             \
 	register long _arg1 __asm__ ("r0") = (long)(arg1);                    \
@@ -102,7 +102,7 @@
 	_arg1;                                                                \
 })
 
-#define my_syscall3(num, arg1, arg2, arg3)                                    \
+#define __nolibc_syscall3(num, arg1, arg2, arg3)                              \
 ({                                                                            \
 	register long _num  __asm__(_NOLIBC_SYSCALL_REG) = (num);             \
 	register long _arg1 __asm__ ("r0") = (long)(arg1);                    \
@@ -121,7 +121,7 @@
 	_arg1;                                                                \
 })
 
-#define my_syscall4(num, arg1, arg2, arg3, arg4)                              \
+#define __nolibc_syscall4(num, arg1, arg2, arg3, arg4)                        \
 ({                                                                            \
 	register long _num  __asm__(_NOLIBC_SYSCALL_REG) = (num);             \
 	register long _arg1 __asm__ ("r0") = (long)(arg1);                    \
@@ -141,7 +141,7 @@
 	_arg1;                                                                \
 })
 
-#define my_syscall5(num, arg1, arg2, arg3, arg4, arg5)                        \
+#define __nolibc_syscall5(num, arg1, arg2, arg3, arg4, arg5)                  \
 ({                                                                            \
 	register long _num  __asm__(_NOLIBC_SYSCALL_REG) = (num);             \
 	register long _arg1 __asm__ ("r0") = (long)(arg1);                    \
@@ -162,7 +162,7 @@
 	_arg1;                                                                \
 })
 
-#define my_syscall6(num, arg1, arg2, arg3, arg4, arg5, arg6)                  \
+#define __nolibc_syscall6(num, arg1, arg2, arg3, arg4, arg5, arg6)            \
 ({                                                                            \
 	register long _num  __asm__(_NOLIBC_SYSCALL_REG) = (num);             \
 	register long _arg1 __asm__ ("r0") = (long)(arg1);                    \
diff --git a/tools/include/nolibc/arch-arm64.h b/tools/include/nolibc/arch-arm64.h
index 080a55a..28b3c75 100644
--- a/tools/include/nolibc/arch-arm64.h
+++ b/tools/include/nolibc/arch-arm64.h
@@ -22,7 +22,7 @@
  *     don't have to experience issues with register constraints.
  */
 
-#define my_syscall0(num)                                                      \
+#define __nolibc_syscall0(num)                                                \
 ({                                                                            \
 	register long _num  __asm__ ("x8") = (num);                           \
 	register long _arg1 __asm__ ("x0");                                   \
@@ -36,7 +36,7 @@
 	_arg1;                                                                \
 })
 
-#define my_syscall1(num, arg1)                                                \
+#define __nolibc_syscall1(num, arg1)                                          \
 ({                                                                            \
 	register long _num  __asm__ ("x8") = (num);                           \
 	register long _arg1 __asm__ ("x0") = (long)(arg1);                    \
@@ -51,7 +51,7 @@
 	_arg1;                                                                \
 })
 
-#define my_syscall2(num, arg1, arg2)                                          \
+#define __nolibc_syscall2(num, arg1, arg2)                                    \
 ({                                                                            \
 	register long _num  __asm__ ("x8") = (num);                           \
 	register long _arg1 __asm__ ("x0") = (long)(arg1);                    \
@@ -67,7 +67,7 @@
 	_arg1;                                                                \
 })
 
-#define my_syscall3(num, arg1, arg2, arg3)                                    \
+#define __nolibc_syscall3(num, arg1, arg2, arg3)                              \
 ({                                                                            \
 	register long _num  __asm__ ("x8") = (num);                           \
 	register long _arg1 __asm__ ("x0") = (long)(arg1);                    \
@@ -84,7 +84,7 @@
 	_arg1;                                                                \
 })
 
-#define my_syscall4(num, arg1, arg2, arg3, arg4)                              \
+#define __nolibc_syscall4(num, arg1, arg2, arg3, arg4)                        \
 ({                                                                            \
 	register long _num  __asm__ ("x8") = (num);                           \
 	register long _arg1 __asm__ ("x0") = (long)(arg1);                    \
@@ -102,7 +102,7 @@
 	_arg1;                                                                \
 })
 
-#define my_syscall5(num, arg1, arg2, arg3, arg4, arg5)                        \
+#define __nolibc_syscall5(num, arg1, arg2, arg3, arg4, arg5)                  \
 ({                                                                            \
 	register long _num  __asm__ ("x8") = (num);                           \
 	register long _arg1 __asm__ ("x0") = (long)(arg1);                    \
@@ -121,7 +121,7 @@
 	_arg1;                                                                \
 })
 
-#define my_syscall6(num, arg1, arg2, arg3, arg4, arg5, arg6)                  \
+#define __nolibc_syscall6(num, arg1, arg2, arg3, arg4, arg5, arg6)            \
 ({                                                                            \
 	register long _num  __asm__ ("x8") = (num);                           \
 	register long _arg1 __asm__ ("x0") = (long)(arg1);                    \
diff --git a/tools/include/nolibc/arch-loongarch.h b/tools/include/nolibc/arch-loongarch.h
index c894176..86fb34b 100644
--- a/tools/include/nolibc/arch-loongarch.h
+++ b/tools/include/nolibc/arch-loongarch.h
@@ -24,7 +24,7 @@
 #define _NOLIBC_SYSCALL_CLOBBERLIST \
 	"memory", "$t0", "$t1", "$t2", "$t3", "$t4", "$t5", "$t6", "$t7", "$t8"
 
-#define my_syscall0(num)                                                      \
+#define __nolibc_syscall0(num)                                                \
 ({                                                                            \
 	register long _num  __asm__ ("a7") = (num);                           \
 	register long _arg1 __asm__ ("a0");                                   \
@@ -38,7 +38,7 @@
 	_arg1;                                                                \
 })
 
-#define my_syscall1(num, arg1)                                                \
+#define __nolibc_syscall1(num, arg1)                                          \
 ({                                                                            \
 	register long _num  __asm__ ("a7") = (num);                           \
 	register long _arg1 __asm__ ("a0") = (long)(arg1);		      \
@@ -52,7 +52,7 @@
 	_arg1;                                                                \
 })
 
-#define my_syscall2(num, arg1, arg2)                                          \
+#define __nolibc_syscall2(num, arg1, arg2)                                    \
 ({                                                                            \
 	register long _num  __asm__ ("a7") = (num);                           \
 	register long _arg1 __asm__ ("a0") = (long)(arg1);                    \
@@ -68,7 +68,7 @@
 	_arg1;                                                                \
 })
 
-#define my_syscall3(num, arg1, arg2, arg3)                                    \
+#define __nolibc_syscall3(num, arg1, arg2, arg3)                              \
 ({                                                                            \
 	register long _num  __asm__ ("a7") = (num);                           \
 	register long _arg1 __asm__ ("a0") = (long)(arg1);                    \
@@ -85,7 +85,7 @@
 	_arg1;                                                                \
 })
 
-#define my_syscall4(num, arg1, arg2, arg3, arg4)                              \
+#define __nolibc_syscall4(num, arg1, arg2, arg3, arg4)                        \
 ({                                                                            \
 	register long _num  __asm__ ("a7") = (num);                           \
 	register long _arg1 __asm__ ("a0") = (long)(arg1);                    \
@@ -103,7 +103,7 @@
 	_arg1;                                                                \
 })
 
-#define my_syscall5(num, arg1, arg2, arg3, arg4, arg5)                        \
+#define __nolibc_syscall5(num, arg1, arg2, arg3, arg4, arg5)                  \
 ({                                                                            \
 	register long _num  __asm__ ("a7") = (num);                           \
 	register long _arg1 __asm__ ("a0") = (long)(arg1);                    \
@@ -122,7 +122,7 @@
 	_arg1;                                                                \
 })
 
-#define my_syscall6(num, arg1, arg2, arg3, arg4, arg5, arg6)                  \
+#define __nolibc_syscall6(num, arg1, arg2, arg3, arg4, arg5, arg6)            \
 ({                                                                            \
 	register long _num  __asm__ ("a7") = (num);                           \
 	register long _arg1 __asm__ ("a0") = (long)(arg1);                    \
diff --git a/tools/include/nolibc/arch-m68k.h b/tools/include/nolibc/arch-m68k.h
index 2a4fbad..81d34c2 100644
--- a/tools/include/nolibc/arch-m68k.h
+++ b/tools/include/nolibc/arch-m68k.h
@@ -15,7 +15,7 @@
 
 #define _NOLIBC_SYSCALL_CLOBBERLIST "memory"
 
-#define my_syscall0(num)                                                      \
+#define __nolibc_syscall0(num)                                                \
 ({                                                                            \
 	register long _num __asm__ ("d0") = (num);                            \
 									      \
@@ -28,7 +28,7 @@
 	_num;                                                                 \
 })
 
-#define my_syscall1(num, arg1)                                                \
+#define __nolibc_syscall1(num, arg1)                                          \
 ({                                                                            \
 	register long _num __asm__ ("d0") = (num);                            \
 	register long _arg1 __asm__ ("d1") = (long)(arg1);                    \
@@ -42,7 +42,7 @@
 	_num;                                                                 \
 })
 
-#define my_syscall2(num, arg1, arg2)                                          \
+#define __nolibc_syscall2(num, arg1, arg2)                                    \
 ({                                                                            \
 	register long _num __asm__ ("d0") = (num);                            \
 	register long _arg1 __asm__ ("d1") = (long)(arg1);                    \
@@ -57,7 +57,7 @@
 	_num;                                                                 \
 })
 
-#define my_syscall3(num, arg1, arg2, arg3)                                    \
+#define __nolibc_syscall3(num, arg1, arg2, arg3)                              \
 ({                                                                            \
 	register long _num __asm__ ("d0")  = (num);                           \
 	register long _arg1 __asm__ ("d1") = (long)(arg1);                    \
@@ -73,7 +73,7 @@
 	_num;                                                                 \
 })
 
-#define my_syscall4(num, arg1, arg2, arg3, arg4)                              \
+#define __nolibc_syscall4(num, arg1, arg2, arg3, arg4)                        \
 ({                                                                            \
 	register long _num __asm__ ("d0") = (num);                            \
 	register long _arg1 __asm__ ("d1") = (long)(arg1);                    \
@@ -90,7 +90,7 @@
 	_num;                                                                 \
 })
 
-#define my_syscall5(num, arg1, arg2, arg3, arg4, arg5)                        \
+#define __nolibc_syscall5(num, arg1, arg2, arg3, arg4, arg5)                  \
 ({                                                                            \
 	register long _num __asm__ ("d0") = (num);                            \
 	register long _arg1 __asm__ ("d1") = (long)(arg1);                    \
@@ -108,7 +108,7 @@
 	_num;                                                                 \
 })
 
-#define my_syscall6(num, arg1, arg2, arg3, arg4, arg5, arg6)                  \
+#define __nolibc_syscall6(num, arg1, arg2, arg3, arg4, arg5, arg6)            \
 ({                                                                            \
 	register long _num __asm__ ("d0")  = (num);                           \
 	register long _arg1 __asm__ ("d1") = (long)(arg1);                    \
diff --git a/tools/include/nolibc/arch-mips.h b/tools/include/nolibc/arch-mips.h
index a72506c..bb9d580 100644
--- a/tools/include/nolibc/arch-mips.h
+++ b/tools/include/nolibc/arch-mips.h
@@ -39,11 +39,19 @@
  *   - stack is 16-byte aligned
  */
 
+#if !defined(__mips_isa_rev) || __mips_isa_rev < 6
+#define _NOLIBC_SYSCALL_CLOBBER_HI_LO "hi", "lo"
+#else
+#define _NOLIBC_SYSCALL_CLOBBER_HI_LO "$0"
+#endif
+
 #if defined(_ABIO32)
 
 #define _NOLIBC_SYSCALL_CLOBBERLIST \
-	"memory", "cc", "at", "v1", "hi", "lo", \
-	"t0", "t1", "t2", "t3", "t4", "t5", "t6", "t7", "t8", "t9"
+	"memory", "cc", "at", "v1", \
+	"t0", "t1", "t2", "t3", "t4", "t5", "t6", "t7", "t8", "t9", \
+	_NOLIBC_SYSCALL_CLOBBER_HI_LO
+
 #define _NOLIBC_SYSCALL_STACK_RESERVE "addiu $sp, $sp, -32\n"
 #define _NOLIBC_SYSCALL_STACK_UNRESERVE "addiu $sp, $sp, 32\n"
 
@@ -52,14 +60,15 @@
 /* binutils, GCC and clang disagree about register aliases, use numbers instead. */
 #define _NOLIBC_SYSCALL_CLOBBERLIST \
 	"memory", "cc", "at", "v1", \
-	"10", "11", "12", "13", "14", "15", "24", "25"
+	"10", "11", "12", "13", "14", "15", "24", "25", \
+	_NOLIBC_SYSCALL_CLOBBER_HI_LO
 
 #define _NOLIBC_SYSCALL_STACK_RESERVE
 #define _NOLIBC_SYSCALL_STACK_UNRESERVE
 
 #endif /* _ABIO32 */
 
-#define my_syscall0(num)                                                      \
+#define __nolibc_syscall0(num)                                                \
 ({                                                                            \
 	register long _num __asm__ ("v0") = (num);                            \
 	register long _arg4 __asm__ ("a3");                                   \
@@ -75,7 +84,7 @@
 	_arg4 ? -_num : _num;                                                 \
 })
 
-#define my_syscall1(num, arg1)                                                \
+#define __nolibc_syscall1(num, arg1)                                          \
 ({                                                                            \
 	register long _num __asm__ ("v0") = (num);                            \
 	register long _arg1 __asm__ ("a0") = (long)(arg1);                    \
@@ -93,7 +102,7 @@
 	_arg4 ? -_num : _num;                                                 \
 })
 
-#define my_syscall2(num, arg1, arg2)                                          \
+#define __nolibc_syscall2(num, arg1, arg2)                                    \
 ({                                                                            \
 	register long _num __asm__ ("v0") = (num);                            \
 	register long _arg1 __asm__ ("a0") = (long)(arg1);                    \
@@ -112,7 +121,7 @@
 	_arg4 ? -_num : _num;                                                 \
 })
 
-#define my_syscall3(num, arg1, arg2, arg3)                                    \
+#define __nolibc_syscall3(num, arg1, arg2, arg3)                              \
 ({                                                                            \
 	register long _num __asm__ ("v0")  = (num);                           \
 	register long _arg1 __asm__ ("a0") = (long)(arg1);                    \
@@ -132,7 +141,7 @@
 	_arg4 ? -_num : _num;                                                 \
 })
 
-#define my_syscall4(num, arg1, arg2, arg3, arg4)                              \
+#define __nolibc_syscall4(num, arg1, arg2, arg3, arg4)                        \
 ({                                                                            \
 	register long _num __asm__ ("v0") = (num);                            \
 	register long _arg1 __asm__ ("a0") = (long)(arg1);                    \
@@ -154,7 +163,7 @@
 
 #if defined(_ABIO32)
 
-#define my_syscall5(num, arg1, arg2, arg3, arg4, arg5)                        \
+#define __nolibc_syscall5(num, arg1, arg2, arg3, arg4, arg5)                  \
 ({                                                                            \
 	register long _num __asm__ ("v0") = (num);                            \
 	register long _arg1 __asm__ ("a0") = (long)(arg1);                    \
@@ -176,7 +185,7 @@
 	_arg4 ? -_num : _num;                                                 \
 })
 
-#define my_syscall6(num, arg1, arg2, arg3, arg4, arg5, arg6)                  \
+#define __nolibc_syscall6(num, arg1, arg2, arg3, arg4, arg5, arg6)            \
 ({                                                                            \
 	register long _num __asm__ ("v0")  = (num);                           \
 	register long _arg1 __asm__ ("a0") = (long)(arg1);                    \
@@ -203,7 +212,7 @@
 
 #else /* _ABIN32 || _ABI64 */
 
-#define my_syscall5(num, arg1, arg2, arg3, arg4, arg5)                        \
+#define __nolibc_syscall5(num, arg1, arg2, arg3, arg4, arg5)                  \
 ({                                                                            \
 	register long _num __asm__ ("v0") = (num);                            \
 	register long _arg1 __asm__ ("$4") = (long)(arg1);                    \
@@ -222,7 +231,7 @@
 	_arg4 ? -_num : _num;                                                 \
 })
 
-#define my_syscall6(num, arg1, arg2, arg3, arg4, arg5, arg6)                  \
+#define __nolibc_syscall6(num, arg1, arg2, arg3, arg4, arg5, arg6)            \
 ({                                                                            \
 	register long _num __asm__ ("v0")  = (num);                           \
 	register long _arg1 __asm__ ("$4") = (long)(arg1);                    \
diff --git a/tools/include/nolibc/arch-powerpc.h b/tools/include/nolibc/arch-powerpc.h
index e0c7e0b..ef87886 100644
--- a/tools/include/nolibc/arch-powerpc.h
+++ b/tools/include/nolibc/arch-powerpc.h
@@ -25,7 +25,7 @@
 #define _NOLIBC_SYSCALL_CLOBBERLIST \
 	"memory", "cr0", "r12", "r11", "r10", "r9"
 
-#define my_syscall0(num)                                                     \
+#define __nolibc_syscall0(num)                                               \
 ({                                                                           \
 	register long _ret  __asm__ ("r3");                                  \
 	register long _num  __asm__ ("r0") = (num);                          \
@@ -42,7 +42,7 @@
 	_ret;                                                                \
 })
 
-#define my_syscall1(num, arg1)                                               \
+#define __nolibc_syscall1(num, arg1)                                         \
 ({                                                                           \
 	register long _ret  __asm__ ("r3");                                  \
 	register long _num  __asm__ ("r0") = (num);                          \
@@ -61,7 +61,7 @@
 })
 
 
-#define my_syscall2(num, arg1, arg2)                                         \
+#define __nolibc_syscall2(num, arg1, arg2)                                   \
 ({                                                                           \
 	register long _ret  __asm__ ("r3");                                  \
 	register long _num  __asm__ ("r0") = (num);                          \
@@ -81,7 +81,7 @@
 })
 
 
-#define my_syscall3(num, arg1, arg2, arg3)                                   \
+#define __nolibc_syscall3(num, arg1, arg2, arg3)                             \
 ({                                                                           \
 	register long _ret  __asm__ ("r3");                                  \
 	register long _num  __asm__ ("r0") = (num);                          \
@@ -102,7 +102,7 @@
 })
 
 
-#define my_syscall4(num, arg1, arg2, arg3, arg4)                             \
+#define __nolibc_syscall4(num, arg1, arg2, arg3, arg4)                       \
 ({                                                                           \
 	register long _ret  __asm__ ("r3");                                  \
 	register long _num  __asm__ ("r0") = (num);                          \
@@ -125,7 +125,7 @@
 })
 
 
-#define my_syscall5(num, arg1, arg2, arg3, arg4, arg5)                       \
+#define __nolibc_syscall5(num, arg1, arg2, arg3, arg4, arg5)                 \
 ({                                                                           \
 	register long _ret  __asm__ ("r3");                                  \
 	register long _num  __asm__ ("r0") = (num);                          \
@@ -148,7 +148,7 @@
 	_ret;                                                                \
 })
 
-#define my_syscall6(num, arg1, arg2, arg3, arg4, arg5, arg6)                 \
+#define __nolibc_syscall6(num, arg1, arg2, arg3, arg4, arg5, arg6)           \
 ({                                                                           \
 	register long _ret  __asm__ ("r3");                                  \
 	register long _num  __asm__ ("r0") = (num);                          \
diff --git a/tools/include/nolibc/arch-riscv.h b/tools/include/nolibc/arch-riscv.h
index 1c00cac..386ebb9 100644
--- a/tools/include/nolibc/arch-riscv.h
+++ b/tools/include/nolibc/arch-riscv.h
@@ -21,7 +21,7 @@
  *     so that we don't have to experience issues with register constraints.
  */
 
-#define my_syscall0(num)                                                      \
+#define __nolibc_syscall0(num)                                                \
 ({                                                                            \
 	register long _num  __asm__ ("a7") = (num);                           \
 	register long _arg1 __asm__ ("a0");                                   \
@@ -35,7 +35,7 @@
 	_arg1;                                                                \
 })
 
-#define my_syscall1(num, arg1)                                                \
+#define __nolibc_syscall1(num, arg1)                                          \
 ({                                                                            \
 	register long _num  __asm__ ("a7") = (num);                           \
 	register long _arg1 __asm__ ("a0") = (long)(arg1);		      \
@@ -49,7 +49,7 @@
 	_arg1;                                                                \
 })
 
-#define my_syscall2(num, arg1, arg2)                                          \
+#define __nolibc_syscall2(num, arg1, arg2)                                    \
 ({                                                                            \
 	register long _num  __asm__ ("a7") = (num);                           \
 	register long _arg1 __asm__ ("a0") = (long)(arg1);                    \
@@ -65,7 +65,7 @@
 	_arg1;                                                                \
 })
 
-#define my_syscall3(num, arg1, arg2, arg3)                                    \
+#define __nolibc_syscall3(num, arg1, arg2, arg3)                              \
 ({                                                                            \
 	register long _num  __asm__ ("a7") = (num);                           \
 	register long _arg1 __asm__ ("a0") = (long)(arg1);                    \
@@ -82,7 +82,7 @@
 	_arg1;                                                                \
 })
 
-#define my_syscall4(num, arg1, arg2, arg3, arg4)                              \
+#define __nolibc_syscall4(num, arg1, arg2, arg3, arg4)                        \
 ({                                                                            \
 	register long _num  __asm__ ("a7") = (num);                           \
 	register long _arg1 __asm__ ("a0") = (long)(arg1);                    \
@@ -100,7 +100,7 @@
 	_arg1;                                                                \
 })
 
-#define my_syscall5(num, arg1, arg2, arg3, arg4, arg5)                        \
+#define __nolibc_syscall5(num, arg1, arg2, arg3, arg4, arg5)                  \
 ({                                                                            \
 	register long _num  __asm__ ("a7") = (num);                           \
 	register long _arg1 __asm__ ("a0") = (long)(arg1);                    \
@@ -119,7 +119,7 @@
 	_arg1;                                                                \
 })
 
-#define my_syscall6(num, arg1, arg2, arg3, arg4, arg5, arg6)                  \
+#define __nolibc_syscall6(num, arg1, arg2, arg3, arg4, arg5, arg6)            \
 ({                                                                            \
 	register long _num  __asm__ ("a7") = (num);                           \
 	register long _arg1 __asm__ ("a0") = (long)(arg1);                    \
diff --git a/tools/include/nolibc/arch-s390.h b/tools/include/nolibc/arch-s390.h
index 904281e..4e69123 100644
--- a/tools/include/nolibc/arch-s390.h
+++ b/tools/include/nolibc/arch-s390.h
@@ -28,7 +28,7 @@
  *
  */
 
-#define my_syscall0(num)						\
+#define __nolibc_syscall0(num)						\
 ({									\
 	register long _num __asm__ ("1") = (num);			\
 	register long _rc __asm__ ("2");				\
@@ -42,7 +42,7 @@
 	_rc;								\
 })
 
-#define my_syscall1(num, arg1)						\
+#define __nolibc_syscall1(num, arg1)					\
 ({									\
 	register long _num __asm__ ("1") = (num);			\
 	register long _arg1 __asm__ ("2") = (long)(arg1);		\
@@ -56,7 +56,7 @@
 	_arg1;								\
 })
 
-#define my_syscall2(num, arg1, arg2)					\
+#define __nolibc_syscall2(num, arg1, arg2)				\
 ({									\
 	register long _num __asm__ ("1") = (num);			\
 	register long _arg1 __asm__ ("2") = (long)(arg1);		\
@@ -71,7 +71,7 @@
 	_arg1;								\
 })
 
-#define my_syscall3(num, arg1, arg2, arg3)				\
+#define __nolibc_syscall3(num, arg1, arg2, arg3)			\
 ({									\
 	register long _num __asm__ ("1") = (num);			\
 	register long _arg1 __asm__ ("2") = (long)(arg1);		\
@@ -87,7 +87,7 @@
 	_arg1;								\
 })
 
-#define my_syscall4(num, arg1, arg2, arg3, arg4)			\
+#define __nolibc_syscall4(num, arg1, arg2, arg3, arg4)			\
 ({									\
 	register long _num __asm__ ("1") = (num);			\
 	register long _arg1 __asm__ ("2") = (long)(arg1);		\
@@ -104,7 +104,7 @@
 	_arg1;								\
 })
 
-#define my_syscall5(num, arg1, arg2, arg3, arg4, arg5)			\
+#define __nolibc_syscall5(num, arg1, arg2, arg3, arg4, arg5)		\
 ({									\
 	register long _num __asm__ ("1") = (num);			\
 	register long _arg1 __asm__ ("2") = (long)(arg1);		\
@@ -123,7 +123,7 @@
 	_arg1;								\
 })
 
-#define my_syscall6(num, arg1, arg2, arg3, arg4, arg5, arg6)		\
+#define __nolibc_syscall6(num, arg1, arg2, arg3, arg4, arg5, arg6)	\
 ({									\
 	register long _num __asm__ ("1") = (num);			\
 	register long _arg1 __asm__ ("2") = (long)(arg1);		\
@@ -167,8 +167,8 @@ struct s390_mmap_arg_struct {
 };
 
 static __attribute__((unused))
-void *sys_mmap(void *addr, size_t length, int prot, int flags, int fd,
-	       off_t offset)
+void *_sys_mmap(void *addr, size_t length, int prot, int flags, int fd,
+		off_t offset)
 {
 	struct s390_mmap_arg_struct args = {
 		.addr = (unsigned long)addr,
@@ -179,22 +179,22 @@ void *sys_mmap(void *addr, size_t length, int prot, int flags, int fd,
 		.offset = (unsigned long)offset
 	};
 
-	return (void *)my_syscall1(__NR_mmap, &args);
+	return (void *)__nolibc_syscall1(__NR_mmap, &args);
 }
-#define sys_mmap sys_mmap
+#define _sys_mmap _sys_mmap
 
 static __attribute__((unused))
-pid_t sys_fork(void)
+pid_t _sys_fork(void)
 {
-	return my_syscall5(__NR_clone, 0, SIGCHLD, 0, 0, 0);
+	return __nolibc_syscall5(__NR_clone, 0, SIGCHLD, 0, 0, 0);
 }
-#define sys_fork sys_fork
+#define _sys_fork _sys_fork
 
 static __attribute__((unused))
-pid_t sys_vfork(void)
+pid_t _sys_vfork(void)
 {
-	return my_syscall5(__NR_clone, 0, CLONE_VM | CLONE_VFORK | SIGCHLD, 0, 0, 0);
+	return __nolibc_syscall5(__NR_clone, 0, CLONE_VM | CLONE_VFORK | SIGCHLD, 0, 0, 0);
 }
-#define sys_vfork sys_vfork
+#define _sys_vfork _sys_vfork
 
 #endif /* _NOLIBC_ARCH_S390_H */
diff --git a/tools/include/nolibc/arch-sh.h b/tools/include/nolibc/arch-sh.h
index 7a42119..b5a64ce 100644
--- a/tools/include/nolibc/arch-sh.h
+++ b/tools/include/nolibc/arch-sh.h
@@ -19,7 +19,7 @@
  *   - syscall return value is in r0
  */
 
-#define my_syscall0(num)                                                      \
+#define __nolibc_syscall0(num)                                                \
 ({                                                                            \
 	register long _num __asm__ ("r3") = (num);                            \
 	register long _ret __asm__ ("r0");                                    \
@@ -33,7 +33,7 @@
 	_ret;                                                                 \
 })
 
-#define my_syscall1(num, arg1)                                                \
+#define __nolibc_syscall1(num, arg1)                                          \
 ({                                                                            \
 	register long _num __asm__ ("r3") = (num);                            \
 	register long _ret __asm__ ("r0");                                    \
@@ -48,7 +48,7 @@
 	_ret;                                                                 \
 })
 
-#define my_syscall2(num, arg1, arg2)                                          \
+#define __nolibc_syscall2(num, arg1, arg2)                                    \
 ({                                                                            \
 	register long _num __asm__ ("r3") = (num);                            \
 	register long _ret __asm__ ("r0");                                    \
@@ -64,7 +64,7 @@
 	_ret;                                                                 \
 })
 
-#define my_syscall3(num, arg1, arg2, arg3)                                    \
+#define __nolibc_syscall3(num, arg1, arg2, arg3)                              \
 ({                                                                            \
 	register long _num __asm__ ("r3") = (num);                            \
 	register long _ret __asm__ ("r0");                                    \
@@ -81,7 +81,7 @@
 	_ret;                                                                 \
 })
 
-#define my_syscall4(num, arg1, arg2, arg3, arg4)                              \
+#define __nolibc_syscall4(num, arg1, arg2, arg3, arg4)                        \
 ({                                                                            \
 	register long _num __asm__ ("r3") = (num);                            \
 	register long _ret __asm__ ("r0");                                    \
@@ -99,7 +99,7 @@
 	_ret;                                                                 \
 })
 
-#define my_syscall5(num, arg1, arg2, arg3, arg4, arg5)                        \
+#define __nolibc_syscall5(num, arg1, arg2, arg3, arg4, arg5)                  \
 ({                                                                            \
 	register long _num __asm__ ("r3") = (num);                            \
 	register long _ret __asm__ ("r0");                                    \
@@ -119,7 +119,7 @@
 	_ret;                                                                 \
 })
 
-#define my_syscall6(num, arg1, arg2, arg3, arg4, arg5, arg6)                  \
+#define __nolibc_syscall6(num, arg1, arg2, arg3, arg4, arg5, arg6)            \
 ({                                                                            \
 	register long _num __asm__ ("r3") = (num);                            \
 	register long _ret __asm__ ("r0");                                    \
diff --git a/tools/include/nolibc/arch-sparc.h b/tools/include/nolibc/arch-sparc.h
index 2ebb568..240539d 100644
--- a/tools/include/nolibc/arch-sparc.h
+++ b/tools/include/nolibc/arch-sparc.h
@@ -38,7 +38,7 @@
 
 #endif /* __arch64__ */
 
-#define my_syscall0(num)                                                      \
+#define __nolibc_syscall0(num)                                                \
 ({                                                                            \
 	register long _num  __asm__ ("g1") = (num);                           \
 	register long _arg1 __asm__ ("o0");                                   \
@@ -52,7 +52,7 @@
 	_arg1;                                                                \
 })
 
-#define my_syscall1(num, arg1)                                                \
+#define __nolibc_syscall1(num, arg1)                                          \
 ({                                                                            \
 	register long _num  __asm__ ("g1") = (num);                           \
 	register long _arg1 __asm__ ("o0") = (long)(arg1);                    \
@@ -66,7 +66,7 @@
 	_arg1;                                                                \
 })
 
-#define my_syscall2(num, arg1, arg2)                                          \
+#define __nolibc_syscall2(num, arg1, arg2)                                    \
 ({                                                                            \
 	register long _num  __asm__ ("g1") = (num);                           \
 	register long _arg1 __asm__ ("o0") = (long)(arg1);                    \
@@ -81,7 +81,7 @@
 	_arg1;                                                                \
 })
 
-#define my_syscall3(num, arg1, arg2, arg3)                                    \
+#define __nolibc_syscall3(num, arg1, arg2, arg3)                              \
 ({                                                                            \
 	register long _num  __asm__ ("g1") = (num);                           \
 	register long _arg1 __asm__ ("o0") = (long)(arg1);                    \
@@ -97,7 +97,7 @@
 	_arg1;                                                                \
 })
 
-#define my_syscall4(num, arg1, arg2, arg3, arg4)                              \
+#define __nolibc_syscall4(num, arg1, arg2, arg3, arg4)                        \
 ({                                                                            \
 	register long _num  __asm__ ("g1") = (num);                           \
 	register long _arg1 __asm__ ("o0") = (long)(arg1);                    \
@@ -114,7 +114,7 @@
 	_arg1;                                                                \
 })
 
-#define my_syscall5(num, arg1, arg2, arg3, arg4, arg5)                        \
+#define __nolibc_syscall5(num, arg1, arg2, arg3, arg4, arg5)                  \
 ({                                                                            \
 	register long _num  __asm__ ("g1") = (num);                           \
 	register long _arg1 __asm__ ("o0") = (long)(arg1);                    \
@@ -132,7 +132,7 @@
 	_arg1;                                                                \
 })
 
-#define my_syscall6(num, arg1, arg2, arg3, arg4, arg5, arg6)                  \
+#define __nolibc_syscall6(num, arg1, arg2, arg3, arg4, arg5, arg6)            \
 ({                                                                            \
 	register long _num  __asm__ ("g1") = (num);                           \
 	register long _arg1 __asm__ ("o0") = (long)(arg1);                    \
@@ -175,12 +175,12 @@ void __attribute__((weak, noreturn)) __nolibc_entrypoint __no_stack_protector _s
 static pid_t getpid(void);
 
 static __attribute__((unused))
-pid_t sys_fork(void)
+pid_t _sys_fork(void)
 {
 	pid_t parent, ret;
 
 	parent = getpid();
-	ret = my_syscall0(__NR_fork);
+	ret = __nolibc_syscall0(__NR_fork);
 
 	/* The syscall returns the parent pid in the child instead of 0 */
 	if (ret == parent)
@@ -188,15 +188,15 @@ pid_t sys_fork(void)
 	else
 		return ret;
 }
-#define sys_fork sys_fork
+#define _sys_fork _sys_fork
 
 static __attribute__((unused))
-pid_t sys_vfork(void)
+pid_t _sys_vfork(void)
 {
 	pid_t parent, ret;
 
 	parent = getpid();
-	ret = my_syscall0(__NR_vfork);
+	ret = __nolibc_syscall0(__NR_vfork);
 
 	/* The syscall returns the parent pid in the child instead of 0 */
 	if (ret == parent)
@@ -204,6 +204,6 @@ pid_t sys_vfork(void)
 	else
 		return ret;
 }
-#define sys_vfork sys_vfork
+#define _sys_vfork _sys_vfork
 
 #endif /* _NOLIBC_ARCH_SPARC_H */
diff --git a/tools/include/nolibc/arch-x86.h b/tools/include/nolibc/arch-x86.h
index f6c43ac..769ba01 100644
--- a/tools/include/nolibc/arch-x86.h
+++ b/tools/include/nolibc/arch-x86.h
@@ -30,7 +30,7 @@
  */
 #define __ARCH_WANT_SYS_OLD_SELECT
 
-#define my_syscall0(num)                                                      \
+#define __nolibc_syscall0(num)                                                \
 ({                                                                            \
 	long _ret;                                                            \
 	register long _num __asm__ ("eax") = (num);                           \
@@ -44,7 +44,7 @@
 	_ret;                                                                 \
 })
 
-#define my_syscall1(num, arg1)                                                \
+#define __nolibc_syscall1(num, arg1)                                          \
 ({                                                                            \
 	long _ret;                                                            \
 	register long _num __asm__ ("eax") = (num);                           \
@@ -60,7 +60,7 @@
 	_ret;                                                                 \
 })
 
-#define my_syscall2(num, arg1, arg2)                                          \
+#define __nolibc_syscall2(num, arg1, arg2)                                    \
 ({                                                                            \
 	long _ret;                                                            \
 	register long _num __asm__ ("eax") = (num);                           \
@@ -77,7 +77,7 @@
 	_ret;                                                                 \
 })
 
-#define my_syscall3(num, arg1, arg2, arg3)                                    \
+#define __nolibc_syscall3(num, arg1, arg2, arg3)                              \
 ({                                                                            \
 	long _ret;                                                            \
 	register long _num __asm__ ("eax") = (num);                           \
@@ -95,7 +95,7 @@
 	_ret;                                                                 \
 })
 
-#define my_syscall4(num, arg1, arg2, arg3, arg4)                              \
+#define __nolibc_syscall4(num, arg1, arg2, arg3, arg4)                        \
 ({                                                                            \
 	long _ret;                                                            \
 	register long _num __asm__ ("eax") = (num);                           \
@@ -114,7 +114,7 @@
 	_ret;                                                                 \
 })
 
-#define my_syscall5(num, arg1, arg2, arg3, arg4, arg5)                        \
+#define __nolibc_syscall5(num, arg1, arg2, arg3, arg4, arg5)                  \
 ({                                                                            \
 	long _ret;                                                            \
 	register long _num __asm__ ("eax") = (num);                           \
@@ -134,27 +134,27 @@
 	_ret;                                                                 \
 })
 
-#define my_syscall6(num, arg1, arg2, arg3, arg4, arg5, arg6)	\
-({								\
-	long _eax  = (long)(num);				\
-	long _arg6 = (long)(arg6); /* Always in memory */	\
-	__asm__ volatile (					\
-		"pushl	%[_arg6]\n\t"				\
-		"pushl	%%ebp\n\t"				\
-		"movl	4(%%esp),%%ebp\n\t"			\
-		"int	$0x80\n\t"				\
-		"popl	%%ebp\n\t"				\
-		"addl	$4,%%esp\n\t"				\
-		: "+a"(_eax)		/* %eax */		\
-		: "b"(arg1),		/* %ebx */		\
-		  "c"(arg2),		/* %ecx */		\
-		  "d"(arg3),		/* %edx */		\
-		  "S"(arg4),		/* %esi */		\
-		  "D"(arg5),		/* %edi */		\
-		  [_arg6]"m"(_arg6)	/* memory */		\
-		: "memory", "cc"				\
-	);							\
-	_eax;							\
+#define __nolibc_syscall6(num, arg1, arg2, arg3, arg4, arg5, arg6)	      \
+({								              \
+	long _eax  = (long)(num);				              \
+	long _arg6 = (long)(arg6); /* Always in memory */	              \
+	__asm__ volatile (					              \
+		"pushl	%[_arg6]\n\t"				              \
+		"pushl	%%ebp\n\t"				              \
+		"movl	4(%%esp),%%ebp\n\t"			              \
+		"int	$0x80\n\t"				              \
+		"popl	%%ebp\n\t"				              \
+		"addl	$4,%%esp\n\t"				              \
+		: "+a"(_eax)		/* %eax */		              \
+		: "b"(arg1),		/* %ebx */		              \
+		  "c"(arg2),		/* %ecx */		              \
+		  "d"(arg3),		/* %edx */		              \
+		  "S"(arg4),		/* %esi */		              \
+		  "D"(arg5),		/* %edi */		              \
+		  [_arg6]"m"(_arg6)	/* memory */		              \
+		: "memory", "cc"				              \
+	);							              \
+	_eax;							              \
 })
 
 #ifndef NOLIBC_NO_RUNTIME
@@ -200,7 +200,7 @@ void __attribute__((weak, noreturn)) __nolibc_entrypoint __no_stack_protector _s
  *
  */
 
-#define my_syscall0(num)                                                      \
+#define __nolibc_syscall0(num)                                                \
 ({                                                                            \
 	long _ret;                                                            \
 	register long _num  __asm__ ("rax") = (num);                          \
@@ -214,7 +214,7 @@ void __attribute__((weak, noreturn)) __nolibc_entrypoint __no_stack_protector _s
 	_ret;                                                                 \
 })
 
-#define my_syscall1(num, arg1)                                                \
+#define __nolibc_syscall1(num, arg1)                                          \
 ({                                                                            \
 	long _ret;                                                            \
 	register long _num  __asm__ ("rax") = (num);                          \
@@ -230,7 +230,7 @@ void __attribute__((weak, noreturn)) __nolibc_entrypoint __no_stack_protector _s
 	_ret;                                                                 \
 })
 
-#define my_syscall2(num, arg1, arg2)                                          \
+#define __nolibc_syscall2(num, arg1, arg2)                                    \
 ({                                                                            \
 	long _ret;                                                            \
 	register long _num  __asm__ ("rax") = (num);                          \
@@ -247,7 +247,7 @@ void __attribute__((weak, noreturn)) __nolibc_entrypoint __no_stack_protector _s
 	_ret;                                                                 \
 })
 
-#define my_syscall3(num, arg1, arg2, arg3)                                    \
+#define __nolibc_syscall3(num, arg1, arg2, arg3)                              \
 ({                                                                            \
 	long _ret;                                                            \
 	register long _num  __asm__ ("rax") = (num);                          \
@@ -265,7 +265,7 @@ void __attribute__((weak, noreturn)) __nolibc_entrypoint __no_stack_protector _s
 	_ret;                                                                 \
 })
 
-#define my_syscall4(num, arg1, arg2, arg3, arg4)                              \
+#define __nolibc_syscall4(num, arg1, arg2, arg3, arg4)                        \
 ({                                                                            \
 	long _ret;                                                            \
 	register long _num  __asm__ ("rax") = (num);                          \
@@ -284,7 +284,7 @@ void __attribute__((weak, noreturn)) __nolibc_entrypoint __no_stack_protector _s
 	_ret;                                                                 \
 })
 
-#define my_syscall5(num, arg1, arg2, arg3, arg4, arg5)                        \
+#define __nolibc_syscall5(num, arg1, arg2, arg3, arg4, arg5)                  \
 ({                                                                            \
 	long _ret;                                                            \
 	register long _num  __asm__ ("rax") = (num);                          \
@@ -304,7 +304,7 @@ void __attribute__((weak, noreturn)) __nolibc_entrypoint __no_stack_protector _s
 	_ret;                                                                 \
 })
 
-#define my_syscall6(num, arg1, arg2, arg3, arg4, arg5, arg6)                  \
+#define __nolibc_syscall6(num, arg1, arg2, arg3, arg4, arg5, arg6)            \
 ({                                                                            \
 	long _ret;                                                            \
 	register long _num  __asm__ ("rax") = (num);                          \
diff --git a/tools/include/nolibc/byteswap.h b/tools/include/nolibc/byteswap.h
new file mode 100644
index 0000000..45bbf96
--- /dev/null
+++ b/tools/include/nolibc/byteswap.h
@@ -0,0 +1,21 @@
+/* SPDX-License-Identifier: LGPL-2.1 OR MIT */
+/*
+ * Byte swapping for NOLIBC
+ * Copyright (C) 2026 Thomas Weißschuh <linux@weissschuh.net>
+ */
+
+/* make sure to include all global symbols */
+#include "nolibc.h"
+
+#ifndef _NOLIBC_BYTESWAP_H
+#define _NOLIBC_BYTESWAP_H
+
+#include "stdint.h"
+
+#include <linux/swab.h>
+
+#define bswap_16(_x) __swab16(_x)
+#define bswap_32(_x) __swab32(_x)
+#define bswap_64(_x) __swab64(_x)
+
+#endif /* _NOLIBC_BYTESWAP_H */
diff --git a/tools/include/nolibc/compiler.h b/tools/include/nolibc/compiler.h
index a8c7619..b56570b 100644
--- a/tools/include/nolibc/compiler.h
+++ b/tools/include/nolibc/compiler.h
@@ -47,6 +47,12 @@
 #  define __nolibc_fallthrough do { } while (0)
 #endif /* __nolibc_has_attribute(fallthrough) */
 
+#if defined(__STDC_VERSION__)
+#  define __nolibc_stdc_version __STDC_VERSION__
+#else
+#  define __nolibc_stdc_version 0
+#endif
+
 #define __nolibc_version(_major, _minor, _patch) ((_major) * 10000 + (_minor) * 100 + (_patch))
 
 #ifdef __GNUC__
@@ -63,7 +69,7 @@
 #  define __nolibc_clang_version 0
 #endif /* __clang__ */
 
-#if __STDC_VERSION__ >= 201112L || \
+#if __nolibc_stdc_version >= 201112L || \
 	__nolibc_gnuc_version >= __nolibc_version(4, 6, 0) || \
 	__nolibc_clang_version >= __nolibc_version(3, 0, 0)
 #  define __nolibc_static_assert(_t) _Static_assert(_t, "")
@@ -71,4 +77,17 @@
 #  define __nolibc_static_assert(_t)
 #endif
 
+/* Make the optimizer believe the variable can be manipulated arbitrarily. */
+#define _NOLIBC_OPTIMIZER_HIDE_VAR(var)	__asm__ ("" : "+r" (var))
+
+#if __nolibc_has_feature(undefined_behavior_sanitizer)
+#  if defined(__clang__)
+#    define __nolibc_no_sanitize_undefined __attribute__((no_sanitize("function")))
+#  else
+#    define __nolibc_no_sanitize_undefined __attribute__((no_sanitize_undefined))
+#  endif
+#else
+#  define __nolibc_no_sanitize_undefined
+#endif
+
 #endif /* _NOLIBC_COMPILER_H */
diff --git a/tools/include/nolibc/crt.h b/tools/include/nolibc/crt.h
index d926299..d8ce91f 100644
--- a/tools/include/nolibc/crt.h
+++ b/tools/include/nolibc/crt.h
@@ -17,6 +17,7 @@ const unsigned long *_auxv __attribute__((weak));
 void _start(void);
 static void __stack_chk_init(void);
 static void exit(int);
+static char *strrchr(const char *s, int c);
 
 extern void (*const __preinit_array_start[])(int, char **, char**) __attribute__((weak));
 extern void (*const __preinit_array_end[])(int, char **, char**) __attribute__((weak));
@@ -27,11 +28,26 @@ extern void (*const __init_array_end[])(int, char **, char**) __attribute__((wea
 extern void (*const __fini_array_start[])(void) __attribute__((weak));
 extern void (*const __fini_array_end[])(void) __attribute__((weak));
 
+#ifndef NOLIBC_IGNORE_ERRNO
+extern char *program_invocation_name __attribute__((weak));
+extern char *program_invocation_short_name __attribute__((weak));
+
+static __inline__
+char *__nolibc_program_invocation_short_name(char *long_name)
+{
+
+	char *short_name;
+
+	short_name = strrchr(long_name, '/');
+	if (!short_name || !short_name[0])
+		return long_name;
+
+	return short_name + 1;
+}
+#endif /* NOLIBC_IGNORE_ERRNO */
+
 void _start_c(long *sp);
-__attribute__((weak,used))
-#if __nolibc_has_feature(undefined_behavior_sanitizer)
-	__attribute__((no_sanitize("function")))
-#endif
+__attribute__((weak,used)) __nolibc_no_sanitize_undefined
 void _start_c(long *sp)
 {
 	long argc;
@@ -76,6 +92,13 @@ void _start_c(long *sp)
 		;
 	_auxv = auxv;
 
+#ifndef NOLIBC_IGNORE_ERRNO
+	if (argc > 0 && argv[0]) {
+		program_invocation_name = argv[0];
+		program_invocation_short_name = __nolibc_program_invocation_short_name(argv[0]);
+	}
+#endif /* NOLIBC_IGNORE_ERRNO */
+
 	for (ctor_func = __preinit_array_start; ctor_func < __preinit_array_end; ctor_func++)
 		(*ctor_func)(argc, argv, envp);
 	for (ctor_func = __init_array_start; ctor_func < __init_array_end; ctor_func++)
diff --git a/tools/include/nolibc/dirent.h b/tools/include/nolibc/dirent.h
index 61a122a..4e02ef2 100644
--- a/tools/include/nolibc/dirent.h
+++ b/tools/include/nolibc/dirent.h
@@ -73,7 +73,7 @@ int readdir_r(DIR *dirp, struct dirent *entry, struct dirent **result)
 
 	fd = ~i;
 
-	ret = sys_getdents64(fd, ldir, sizeof(buf));
+	ret = _sys_getdents64(fd, ldir, sizeof(buf));
 	if (ret < 0)
 		return -ret;
 	if (ret == 0) {
@@ -86,7 +86,7 @@ int readdir_r(DIR *dirp, struct dirent *entry, struct dirent **result)
 	 * readdir() can only return one entry at a time.
 	 * Make sure the non-returned ones are not skipped.
 	 */
-	ret = sys_lseek(fd, ldir->d_off, SEEK_SET);
+	ret = _sys_lseek(fd, ldir->d_off, SEEK_SET);
 	if (ret < 0)
 		return -ret;
 
diff --git a/tools/include/nolibc/endian.h b/tools/include/nolibc/endian.h
new file mode 100644
index 0000000..ccc016e
--- /dev/null
+++ b/tools/include/nolibc/endian.h
@@ -0,0 +1,32 @@
+/* SPDX-License-Identifier: LGPL-2.1 OR MIT */
+/*
+ * Byte order conversion for NOLIBC
+ * Copyright (C) 2026 Thomas Weißschuh <linux@weissschuh.net>
+ */
+
+/* make sure to include all global symbols */
+#include "nolibc.h"
+
+#ifndef _NOLIBC_ENDIAN_H
+#define _NOLIBC_ENDIAN_H
+
+#include "stdint.h"
+
+#include <asm/byteorder.h>
+
+#define htobe16(_x) __cpu_to_be16(_x)
+#define htole16(_x) __cpu_to_le16(_x)
+#define be16toh(_x) __be16_to_cpu(_x)
+#define le16toh(_x) __le16_to_cpu(_x)
+
+#define htobe32(_x) __cpu_to_be32(_x)
+#define htole32(_x) __cpu_to_le32(_x)
+#define be32toh(_x) __be32_to_cpu(_x)
+#define le32toh(_x) __le32_to_cpu(_x)
+
+#define htobe64(_x) __cpu_to_be64(_x)
+#define htole64(_x) __cpu_to_le64(_x)
+#define be64toh(_x) __be64_to_cpu(_x)
+#define le64toh(_x) __le64_to_cpu(_x)
+
+#endif /* _NOLIBC_ENDIAN_H */
diff --git a/tools/include/nolibc/err.h b/tools/include/nolibc/err.h
new file mode 100644
index 0000000..e22ae87
--- /dev/null
+++ b/tools/include/nolibc/err.h
@@ -0,0 +1,87 @@
+/* SPDX-License-Identifier: LGPL-2.1 OR MIT */
+/*
+ * formatted error message for NOLIBC
+ * Copyright (C) 2026 Thomas Weißschuh <linux@weissschuh.net>
+ */
+
+/* make sure to include all global symbols */
+#include "nolibc.h"
+
+#ifndef _NOLIBC_ERR_H
+#define _NOLIBC_ERR_H
+
+#include "errno.h"
+#include "stdarg.h"
+#include "sys.h"
+
+static __attribute__((unused))
+void vwarn(const char *fmt, va_list args)
+{
+	fprintf(stderr, "%s: ", program_invocation_short_name);
+	vfprintf(stderr, fmt, args);
+	fprintf(stderr, ": %m\n");
+}
+
+static __attribute__((unused))
+void vwarnx(const char *fmt, va_list args)
+{
+	fprintf(stderr, "%s: ", program_invocation_short_name);
+	vfprintf(stderr, fmt, args);
+	fprintf(stderr, "\n");
+}
+
+static __attribute__((unused))
+void warn(const char *fmt, ...)
+{
+	va_list args;
+
+	va_start(args, fmt);
+	vwarn(fmt, args);
+	va_end(args);
+}
+
+static __attribute__((unused))
+void warnx(const char *fmt, ...)
+{
+	va_list args;
+
+	va_start(args, fmt);
+	vwarnx(fmt, args);
+	va_end(args);
+}
+
+static __attribute__((noreturn, unused))
+void verr(int eval, const char *fmt, va_list args)
+{
+	vwarn(fmt, args);
+	exit(eval);
+}
+
+static __attribute__((noreturn, unused))
+void verrx(int eval, const char *fmt, va_list args)
+{
+	warnx(fmt, args);
+	exit(eval);
+}
+
+static __attribute__((noreturn, unused))
+void err(int eval, const char *fmt, ...)
+{
+	va_list args;
+
+	va_start(args, fmt);
+	verr(eval, fmt, args);
+	va_end(args);
+}
+
+static __attribute__((noreturn, unused))
+void errx(int eval, const char *fmt, ...)
+{
+	va_list args;
+
+	va_start(args, fmt);
+	verrx(eval, fmt, args);
+	va_end(args);
+}
+
+#endif /* _NOLIBC_ERR_H */
diff --git a/tools/include/nolibc/errno.h b/tools/include/nolibc/errno.h
index 08a33c4..bab8369 100644
--- a/tools/include/nolibc/errno.h
+++ b/tools/include/nolibc/errno.h
@@ -15,8 +15,12 @@
 #ifndef NOLIBC_IGNORE_ERRNO
 #define SET_ERRNO(v) do { errno = (v); } while (0)
 int errno __attribute__((weak));
+char *program_invocation_name __attribute__((weak)) = "";
+char *program_invocation_short_name __attribute__((weak)) = "";
 #else
 #define SET_ERRNO(v) do { } while (0)
+#define program_invocation_name ""
+#define program_invocation_short_name ""
 #endif
 
 
diff --git a/tools/include/nolibc/fcntl.h b/tools/include/nolibc/fcntl.h
index bff2e54..ed2f555 100644
--- a/tools/include/nolibc/fcntl.h
+++ b/tools/include/nolibc/fcntl.h
@@ -19,9 +19,9 @@
  */
 
 static __attribute__((unused))
-int sys_openat(int dirfd, const char *path, int flags, mode_t mode)
+int _sys_openat(int dirfd, const char *path, int flags, mode_t mode)
 {
-	return my_syscall4(__NR_openat, dirfd, path, flags, mode);
+	return __nolibc_syscall4(__NR_openat, dirfd, path, flags, mode);
 }
 
 static __attribute__((unused))
@@ -37,7 +37,7 @@ int openat(int dirfd, const char *path, int flags, ...)
 		va_end(args);
 	}
 
-	return __sysret(sys_openat(dirfd, path, flags, mode));
+	return __sysret(_sys_openat(dirfd, path, flags, mode));
 }
 
 /*
@@ -45,9 +45,9 @@ int openat(int dirfd, const char *path, int flags, ...)
  */
 
 static __attribute__((unused))
-int sys_open(const char *path, int flags, mode_t mode)
+int _sys_open(const char *path, int flags, mode_t mode)
 {
-	return my_syscall4(__NR_openat, AT_FDCWD, path, flags, mode);
+	return __nolibc_syscall4(__NR_openat, AT_FDCWD, path, flags, mode);
 }
 
 static __attribute__((unused))
@@ -63,7 +63,7 @@ int open(const char *path, int flags, ...)
 		va_end(args);
 	}
 
-	return __sysret(sys_open(path, flags, mode));
+	return __sysret(_sys_open(path, flags, mode));
 }
 
 #endif /* _NOLIBC_FCNTL_H */
diff --git a/tools/include/nolibc/nolibc.h b/tools/include/nolibc/nolibc.h
index 9c7f43b..f4120f6 100644
--- a/tools/include/nolibc/nolibc.h
+++ b/tools/include/nolibc/nolibc.h
@@ -12,24 +12,24 @@
  *
  * Syscalls are split into 3 levels:
  *   - The lower level is the arch-specific syscall() definition, consisting in
- *     assembly code in compound expressions. These are called my_syscall0() to
- *     my_syscall6() depending on the number of arguments. All input arguments
+ *     assembly code in compound expressions. These are called __nolibc_syscall0() to
+ *     __nolibc_syscall6() depending on the number of arguments. All input arguments
  *     are castto a long stored in a register. These expressions always return
  *     the syscall's return value as a signed long value which is often either
  *     a pointer or the negated errno value.
  *
  *   - The second level is mostly architecture-independent. It is made of
- *     static functions called sys_<name>() which rely on my_syscallN()
+ *     static functions called _sys_<name>() which rely on __nolibc_syscallN()
  *     depending on the syscall definition. These functions are responsible
  *     for exposing the appropriate types for the syscall arguments (int,
  *     pointers, etc) and for setting the appropriate return type (often int).
  *     A few of them are architecture-specific because the syscalls are not all
  *     mapped exactly the same among architectures. For example, some archs do
- *     not implement select() and need pselect6() instead, so the sys_select()
+ *     not implement select() and need pselect6() instead, so the _sys_select()
  *     function will have to abstract this.
  *
  *   - The third level is the libc call definition. It exposes the lower raw
- *     sys_<name>() calls in a way that looks like what a libc usually does,
+ *     _sys_<name>() calls in a way that looks like what a libc usually does,
  *     takes care of specific input values, and of setting errno upon error.
  *     There can be minor variations compared to standard libc calls.
  *
@@ -130,6 +130,9 @@
 #include "getopt.h"
 #include "poll.h"
 #include "math.h"
+#include "err.h"
+#include "byteswap.h"
+#include "endian.h"
 
 /* Used by programs to avoid std includes */
 #define NOLIBC
diff --git a/tools/include/nolibc/poll.h b/tools/include/nolibc/poll.h
index e854c94..dbcf883 100644
--- a/tools/include/nolibc/poll.h
+++ b/tools/include/nolibc/poll.h
@@ -21,7 +21,7 @@
  */
 
 static __attribute__((unused))
-int sys_poll(struct pollfd *fds, int nfds, int timeout)
+int _sys_poll(struct pollfd *fds, int nfds, int timeout)
 {
 #if defined(__NR_ppoll_time64)
 	struct __kernel_timespec t;
@@ -30,7 +30,7 @@ int sys_poll(struct pollfd *fds, int nfds, int timeout)
 		t.tv_sec  = timeout / 1000;
 		t.tv_nsec = (timeout % 1000) * 1000000;
 	}
-	return my_syscall5(__NR_ppoll_time64, fds, nfds, (timeout >= 0) ? &t : NULL, NULL, 0);
+	return __nolibc_syscall5(__NR_ppoll_time64, fds, nfds, (timeout >= 0) ? &t : NULL, NULL, 0);
 #else
 	struct __kernel_old_timespec t;
 
@@ -38,14 +38,14 @@ int sys_poll(struct pollfd *fds, int nfds, int timeout)
 		t.tv_sec  = timeout / 1000;
 		t.tv_nsec = (timeout % 1000) * 1000000;
 	}
-	return my_syscall5(__NR_ppoll, fds, nfds, (timeout >= 0) ? &t : NULL, NULL, 0);
+	return __nolibc_syscall5(__NR_ppoll, fds, nfds, (timeout >= 0) ? &t : NULL, NULL, 0);
 #endif
 }
 
 static __attribute__((unused))
 int poll(struct pollfd *fds, int nfds, int timeout)
 {
-	return __sysret(sys_poll(fds, nfds, timeout));
+	return __sysret(_sys_poll(fds, nfds, timeout));
 }
 
 #endif /* _NOLIBC_POLL_H */
diff --git a/tools/include/nolibc/sched.h b/tools/include/nolibc/sched.h
index 3222156..7a5f6d9 100644
--- a/tools/include/nolibc/sched.h
+++ b/tools/include/nolibc/sched.h
@@ -19,15 +19,15 @@
  */
 
 static __attribute__((unused))
-int sys_setns(int fd, int nstype)
+int _sys_setns(int fd, int nstype)
 {
-	return my_syscall2(__NR_setns, fd, nstype);
+	return __nolibc_syscall2(__NR_setns, fd, nstype);
 }
 
 static __attribute__((unused))
 int setns(int fd, int nstype)
 {
-	return __sysret(sys_setns(fd, nstype));
+	return __sysret(_sys_setns(fd, nstype));
 }
 
 
@@ -36,15 +36,15 @@ int setns(int fd, int nstype)
  */
 
 static __attribute__((unused))
-int sys_unshare(int flags)
+int _sys_unshare(int flags)
 {
-	return my_syscall1(__NR_unshare, flags);
+	return __nolibc_syscall1(__NR_unshare, flags);
 }
 
 static __attribute__((unused))
 int unshare(int flags)
 {
-	return __sysret(sys_unshare(flags));
+	return __sysret(_sys_unshare(flags));
 }
 
 #endif /* _NOLIBC_SCHED_H */
diff --git a/tools/include/nolibc/signal.h b/tools/include/nolibc/signal.h
index ac13e53..99a0489 100644
--- a/tools/include/nolibc/signal.h
+++ b/tools/include/nolibc/signal.h
@@ -20,7 +20,7 @@ int raise(int signal);
 __attribute__((weak,unused,section(".text.nolibc_raise")))
 int raise(int signal)
 {
-	return sys_kill(sys_getpid(), signal);
+	return _sys_kill(_sys_getpid(), signal);
 }
 
 #endif /* _NOLIBC_SIGNAL_H */
diff --git a/tools/include/nolibc/stackprotector.h b/tools/include/nolibc/stackprotector.h
index 7123aa0..ae8b1d3 100644
--- a/tools/include/nolibc/stackprotector.h
+++ b/tools/include/nolibc/stackprotector.h
@@ -24,9 +24,9 @@ __attribute__((weak,used,noreturn,section(".text.nolibc_stack_chk")))
 void __stack_chk_fail(void)
 {
 	pid_t pid;
-	my_syscall3(__NR_write, STDERR_FILENO, "!!Stack smashing detected!!\n", 28);
-	pid = my_syscall0(__NR_getpid);
-	my_syscall2(__NR_kill, pid, SIGABRT);
+	__nolibc_syscall3(__NR_write, STDERR_FILENO, "!!Stack smashing detected!!\n", 28);
+	pid = __nolibc_syscall0(__NR_getpid);
+	__nolibc_syscall2(__NR_kill, pid, SIGABRT);
 	for (;;);
 }
 
@@ -42,7 +42,7 @@ uintptr_t __stack_chk_guard;
 
 static __no_stack_protector void __stack_chk_init(void)
 {
-	my_syscall3(__NR_getrandom, &__stack_chk_guard, sizeof(__stack_chk_guard), 0);
+	__nolibc_syscall3(__NR_getrandom, &__stack_chk_guard, sizeof(__stack_chk_guard), 0);
 	/* a bit more randomness in case getrandom() fails, ensure the guard is never 0 */
 	if (__stack_chk_guard != (uintptr_t) &__stack_chk_guard)
 		__stack_chk_guard ^= (uintptr_t) &__stack_chk_guard;
diff --git a/tools/include/nolibc/std.h b/tools/include/nolibc/std.h
index b9a1161..dc2dc46 100644
--- a/tools/include/nolibc/std.h
+++ b/tools/include/nolibc/std.h
@@ -19,7 +19,7 @@
 #include <linux/types.h>
 
 /* those are commonly provided by sys/types.h */
-typedef unsigned int          dev_t;
+typedef uint64_t              dev_t;
 typedef uint64_t              ino_t;
 typedef unsigned int         mode_t;
 typedef   signed int          pid_t;
diff --git a/tools/include/nolibc/stddef.h b/tools/include/nolibc/stddef.h
index ecbd13e..a397634 100644
--- a/tools/include/nolibc/stddef.h
+++ b/tools/include/nolibc/stddef.h
@@ -18,7 +18,7 @@
 #endif
 
 #ifndef offsetof
-#define offsetof(TYPE, FIELD) ((size_t) &((TYPE *)0)->FIELD)
+#define offsetof(TYPE, FIELD) __builtin_offsetof(TYPE, FIELD)
 #endif
 
 #endif /* _NOLIBC_STDDEF_H */
diff --git a/tools/include/nolibc/stdio.h b/tools/include/nolibc/stdio.h
index 233318b..ebdd413 100644
--- a/tools/include/nolibc/stdio.h
+++ b/tools/include/nolibc/stdio.h
@@ -291,156 +291,371 @@ int fseek(FILE *stream, long offset, int whence)
 }
 
 
-/* minimal printf(). It supports the following formats:
- *  - %[l*]{d,u,c,x,p}
- *  - %s
- *  - unknown modifiers are ignored.
+/* printf(). Supports most of the normal integer and string formats.
+ *  - %[#0-+ ][width|*[.precision|*}][{l,t,z,ll,L,j,q}]{c,d,i,u,o,x,X,p,s,m,%}
+ *  - %% generates a single %
+ *  - %m outputs strerror(errno).
+ *  - %X outputs a..f the same as %x.
+ *  - No support for floating point or wide characters.
+ *  - Invalid formats are copied to the output buffer.
+ *
+ * Called by vfprintf() and snprintf() to do the actual formatting.
+ * The callers provide a callback function to save the formatted data.
+ * The callback function is called multiple times:
+ *  - for each group of literal characters in the format string.
+ *  - for field padding.
+ *  - for each conversion specifier.
+ *  - with (NULL, 0) at the end of the __nolibc_printf.
+ * If the callback returns non-zero __nolibc_printf() immediately returns -1.
  */
-typedef int (*__nolibc_printf_cb)(intptr_t state, const char *buf, size_t size);
 
-static __attribute__((unused, format(printf, 4, 0)))
-int __nolibc_printf(__nolibc_printf_cb cb, intptr_t state, size_t n, const char *fmt, va_list args)
+typedef int (*__nolibc_printf_cb)(void *state, const char *buf, size_t size);
+
+/* This code uses 'flag' variables that are indexed by the low 6 bits
+ * of characters to optimise checks for multiple characters.
+ *
+ * _NOLIBC_PF_FLAGS_CONTAIN(flags, 'a', 'b'. ...)
+ * returns non-zero if the bit for any of the specified characters is set.
+ *
+ * _NOLIBC_PF_CHAR_IS_ONE_OF(ch, 'a', 'b'. ...)
+ * returns the flag bit for ch if it is one of the specified characters.
+ * All the characters must be in the same 32 character block (non-alphabetic,
+ * upper case, or lower case) of the ASCII character set.
+ */
+#define _NOLIBC_PF_FLAG(ch) (1u << ((ch) & 0x1f))
+#define _NOLIBC_PF_FLAG_NZ(ch) ((ch) ? _NOLIBC_PF_FLAG(ch) : 0)
+#define _NOLIBC_PF_FLAG8(cmp_1, cmp_2, cmp_3, cmp_4, cmp_5, cmp_6, cmp_7, cmp_8, ...) \
+	(_NOLIBC_PF_FLAG_NZ(cmp_1) | _NOLIBC_PF_FLAG_NZ(cmp_2) | \
+	 _NOLIBC_PF_FLAG_NZ(cmp_3) | _NOLIBC_PF_FLAG_NZ(cmp_4) | \
+	 _NOLIBC_PF_FLAG_NZ(cmp_5) | _NOLIBC_PF_FLAG_NZ(cmp_6) | \
+	 _NOLIBC_PF_FLAG_NZ(cmp_7) | _NOLIBC_PF_FLAG_NZ(cmp_8))
+#define _NOLIBC_PF_FLAGS_CONTAIN(flags, ...) \
+	((flags) & _NOLIBC_PF_FLAG8(__VA_ARGS__, 0, 0, 0, 0, 0, 0, 0))
+#define _NOLIBC_PF_CHAR_IS_ONE_OF(ch, cmp_1, ...) \
+	((unsigned int)(ch) - (cmp_1 & 0xe0) > 0x1f ? 0 : \
+		_NOLIBC_PF_FLAGS_CONTAIN(_NOLIBC_PF_FLAG(ch), cmp_1, __VA_ARGS__))
+
+static __attribute__((unused, format(printf, 3, 0)))
+int __nolibc_printf(__nolibc_printf_cb cb, void *state, const char *fmt, va_list args)
 {
-	char escape, lpref, c;
+	char ch;
 	unsigned long long v;
-	unsigned int written, width;
-	size_t len, ofs, w;
-	char tmpbuf[21];
+	long long signed_v;
+	int written, width, precision, len;
+	unsigned int flags, ch_flag;
+	char outbuf[2 + 31 + 22 + 1];
+	char *out;
 	const char *outstr;
+	unsigned int sign_prefix;
+	int got_width;
 
-	written = ofs = escape = lpref = 0;
+	written = 0;
 	while (1) {
-		c = fmt[ofs++];
+		outstr = fmt;
+		ch = *fmt++;
+		if (!ch)
+			break;
+
 		width = 0;
+		flags = 0;
+		if (ch != '%') {
+			while (*fmt && *fmt != '%')
+				fmt++;
+			/* Output characters from the format string. */
+			len = fmt - outstr;
+			goto do_output;
+		}
 
-		if (escape) {
-			/* we're in an escape sequence, ofs == 1 */
-			escape = 0;
+		/* we're in a format sequence */
 
-			/* width */
-			while (c >= '0' && c <= '9') {
-				width *= 10;
-				width += c - '0';
+		/* Conversion flag characters */
+		while (1) {
+			ch = *fmt++;
+			ch_flag = _NOLIBC_PF_CHAR_IS_ONE_OF(ch, ' ', '#', '+', '-', '0');
+			if (!ch_flag)
+				break;
+			flags |= ch_flag;
+		}
 
-				c = fmt[ofs++];
+		/* Width and precision */
+		for (got_width = 0;; ch = *fmt++) {
+			if (ch == '*') {
+				precision = va_arg(args, int);
+				ch = *fmt++;
+			} else {
+				for (precision = 0; ch >= '0' && ch <= '9'; ch = *fmt++)
+					precision = precision * 10 + (ch - '0');
+			}
+			if (got_width)
+				break;
+			width = precision;
+			if (ch != '.') {
+				/* Default precision for strings */
+				precision = -1;
+				break;
+			}
+			got_width = 1;
+		}
+		/* A negative width (e.g. from "%*s") requests left justify. */
+		if (width < 0) {
+			width = -width;
+			flags |= _NOLIBC_PF_FLAG('-');
+		}
+
+		/* Length modifier.
+		 * They miss the conversion flags characters " #+-0" so can go into flags.
+		 * Change both L and ll to j (all always 64bit).
+		 */
+		if (ch == 'L')
+			ch = 'j';
+		ch_flag = _NOLIBC_PF_CHAR_IS_ONE_OF(ch, 'l', 't', 'z', 'j', 'q');
+		if (ch_flag != 0) {
+			if (ch == 'l' && fmt[0] == 'l') {
+				fmt++;
+				ch_flag = _NOLIBC_PF_FLAG('j');
+			}
+			flags |= ch_flag;
+			ch = *fmt++;
+		}
+
+		/* Conversion specifiers. */
+
+		/* Numeric and pointer conversion specifiers.
+		 *
+		 * Use an explicit bound check (rather than _NOLIBC_PF_CHAR_IS_ONE_OF())
+		 * so that 'X' can be allowed through.
+		 * 'X' gets treated and 'x' because _NOLIBC_PF_FLAG() returns the same
+		 * value for both.
+		 *
+		 * We need to check for "%p" or "%#x" later, merging here gives better code.
+		 * But '#' collides with 'c' so shift right.
+		 */
+		ch_flag = _NOLIBC_PF_FLAG(ch) | (flags & _NOLIBC_PF_FLAG('#')) >> 1;
+		if (((ch >= 'a' && ch <= 'z') || ch == 'X') &&
+		    _NOLIBC_PF_FLAGS_CONTAIN(ch_flag, 'c', 'd', 'i', 'u', 'o', 'x', 'p', 's')) {
+			/* 'long' is needed for pointer/string conversions and ltz lengths.
+			 * A single test can be used provided 'p' (the same bit as '0')
+			 * is masked from flags.
+			 */
+			if (_NOLIBC_PF_FLAGS_CONTAIN(ch_flag | (flags & ~_NOLIBC_PF_FLAG('p')),
+						     'p', 's', 'l', 't', 'z')) {
+				v = va_arg(args, unsigned long);
+				signed_v = (long)v;
+			} else if (_NOLIBC_PF_FLAGS_CONTAIN(flags, 'j', 'q')) {
+				v = va_arg(args, unsigned long long);
+				signed_v = v;
+			} else {
+				v = va_arg(args, unsigned int);
+				signed_v = (int)v;
+			}
+
+			if (ch == 'c') {
+				/* "%c" - single character. */
+				outbuf[0] = v;
+				len = 1;
+				outstr = outbuf;
+				goto do_output;
 			}
 
-			if (c == 'c' || c == 'd' || c == 'u' || c == 'x' || c == 'p') {
-				char *out = tmpbuf;
-
-				if (c == 'p')
-					v = va_arg(args, unsigned long);
-				else if (lpref) {
-					if (lpref > 1)
-						v = va_arg(args, unsigned long long);
-					else
-						v = va_arg(args, unsigned long);
-				} else
-					v = va_arg(args, unsigned int);
-
-				if (c == 'd') {
-					/* sign-extend the value */
-					if (lpref == 0)
-						v = (long long)(int)v;
-					else if (lpref == 1)
-						v = (long long)(long)v;
+			if (ch == 's') {
+				/* "%s" - character string. */
+				outstr = (const char  *)(uintptr_t)v;
+				if (!outstr) {
+					outstr = "(null)";
+					/* Match glibc, nothing output if precision too small */
+					len = precision < 0 || precision >= 6 ? 6 : 0;
+					goto do_output;
 				}
+				goto do_strlen_output;
+			}
 
-				switch (c) {
-				case 'c':
-					out[0] = v;
-					out[1] = 0;
-					break;
-				case 'd':
-					i64toa_r(v, out);
-					break;
-				case 'u':
-					u64toa_r(v, out);
-					break;
-				case 'p':
-					*(out++) = '0';
-					*(out++) = 'x';
-					__nolibc_fallthrough;
-				default: /* 'x' and 'p' above */
-					u64toh_r(v, out);
-					break;
+			/* The 'sign_prefix' can be zero, one or two ("0x") characters.
+			 * Prepended least significant byte first stopping on a zero byte.
+			 */
+			sign_prefix = 0;
+
+			if (_NOLIBC_PF_FLAGS_CONTAIN(ch_flag, 'd', 'i')) {
+				/* "%d" and "%i" - signed decimal numbers. */
+				if (signed_v < 0) {
+					sign_prefix = '-';
+					v = -(signed_v + 1);
+					v++;
+				} else if (_NOLIBC_PF_FLAGS_CONTAIN(flags, '+')) {
+					sign_prefix = '+';
+				} else if (_NOLIBC_PF_FLAGS_CONTAIN(flags, ' ')) {
+					sign_prefix = ' ';
 				}
-				outstr = tmpbuf;
+			} else {
+				/* "#o" requires that the output always starts with a '0'.
+				 * This needs another check after any zero padding to avoid
+				 * adding an extra leading '0'.
+				 */
+				if (_NOLIBC_PF_FLAGS_CONTAIN(ch_flag, 'o') &&
+						_NOLIBC_PF_FLAGS_CONTAIN(ch_flag, '#' - 1))
+					sign_prefix = '0';
 			}
-			else if (c == 's') {
-				outstr = va_arg(args, char *);
-				if (!outstr)
-					outstr="(null)";
+
+			/* The value is converted offset into the buffer so that
+			 * 31 zero pad characters and the sign/prefix can be added in front.
+			 * The longest digit string is 22 + 1 for octal conversions.
+			 */
+			out = outbuf + 2 + 31;
+
+			if (v == 0) {
+				/* There are special rules for zero. */
+				if (_NOLIBC_PF_FLAGS_CONTAIN(ch_flag, 'p')) {
+					/* "%p" match glibc, precision is ignored */
+					outstr = "(nil)";
+					len = 5;
+					goto do_output;
+				}
+				if (!precision) {
+					/* Explicit %nn.0d, no digits output (except for %#.0o) */
+					len = 0;
+					goto prepend_sign;
+				}
+				/* All other formats (including "%#x") just output "0". */
+				out[0] = '0';
+				len = 1;
+			} else {
+				/* Convert the number to ascii in the required base. */
+				unsigned long long recip;
+				unsigned int base;
+				if (_NOLIBC_PF_FLAGS_CONTAIN(ch_flag, 'd', 'i', 'u')) {
+					base = 10;
+					recip = _NOLIBC_U64TOA_RECIP(10);
+				} else if (_NOLIBC_PF_FLAGS_CONTAIN(ch_flag, 'o')) {
+					base = 8;
+					recip = _NOLIBC_U64TOA_RECIP(8);
+				} else {
+					base = 16;
+					recip = _NOLIBC_U64TOA_RECIP(16);
+					if (_NOLIBC_PF_FLAGS_CONTAIN(ch_flag, 'p', '#' - 1)) {
+						/* "%p" and "%#x" need "0x" prepending. */
+						sign_prefix = '0' << 8 | 'x';
+					}
+				}
+				len = _nolibc_u64toa_base(v, out, base, recip);
 			}
-			else if (c == 'm') {
-#ifdef NOLIBC_IGNORE_ERRNO
-				outstr = "unknown error";
-#else
-				outstr = strerror(errno);
-#endif /* NOLIBC_IGNORE_ERRNO */
+
+			/* Add zero padding */
+			if (precision < 0) {
+				/* No explicit precision (or negative from "%.*s"). */
+				if (!_NOLIBC_PF_FLAGS_CONTAIN(flags, '0'))
+					goto no_zero_padding;
+				if (_NOLIBC_PF_FLAGS_CONTAIN(flags, '-'))
+					/* Left justify overrides zero pad */
+					goto no_zero_padding;
+				/* eg "%05d", Zero pad to field width less sign.
+				 * Note that precision can end up negative so all
+				 * the variables have to be 'signed int'.
+				 */
+				precision = width;
+				if (sign_prefix) {
+					precision--;
+					if (sign_prefix >= 256)
+						precision--;
+				}
 			}
-			else if (c == '%') {
-				/* queue it verbatim */
-				continue;
+			if (precision > 31)
+				/* Don't run off the start of outbuf[], arbitrary limit
+				 * longer than the longest number field. */
+				precision = 31;
+			for (; len < precision; len++) {
+				/* Stop gcc generating horrid code and memset(). */
+				_NOLIBC_OPTIMIZER_HIDE_VAR(len);
+				*--out = '0';
 			}
-			else {
-				/* modifiers or final 0 */
-				if (c == 'l') {
-					/* long format prefix, maintain the escape */
-					lpref++;
-				} else if (c == 'j') {
-					lpref = 2;
+no_zero_padding:
+
+			/* %#o has set sign_prefix to '0', but we don't want so add an extra
+			 * leading zero here.
+			 * Since the only other byte values of sign_prefix are ' ', '+' and '-'
+			 * it is enough to check that out[] doesn't already start with sign_prefix.
+			 */
+			if (sign_prefix - *out) {
+prepend_sign:
+				/* Add the 0, 1 or 2 ("0x") sign/prefix characters at the front. */
+				for (; sign_prefix; sign_prefix >>= 8) {
+					/* Force gcc to increment len inside the loop. */
+					_NOLIBC_OPTIMIZER_HIDE_VAR(len);
+					len++;
+					*--out = sign_prefix;
 				}
-				escape = 1;
-				goto do_escape;
 			}
-			len = strlen(outstr);
-			goto flush_str;
+			outstr = out;
+			goto do_output;
 		}
 
-		/* not an escape sequence */
-		if (c == 0 || c == '%') {
-			/* flush pending data on escape or end */
-			escape = 1;
-			lpref = 0;
-			outstr = fmt;
-			len = ofs - 1;
-		flush_str:
-			if (n) {
-				w = len < n ? len : n;
-				n -= w;
-				while (width-- > w) {
-					if (cb(state, " ", 1) != 0)
-						return -1;
-					written += 1;
-				}
-				if (cb(state, outstr, w) != 0)
-					return -1;
-			}
+		if (ch == 'm') {
+#ifdef NOLIBC_IGNORE_ERRNO
+			outstr = "unknown error";
+#else
+			outstr = strerror(errno);
+#endif /* NOLIBC_IGNORE_ERRNO */
+			goto do_strlen_output;
+		}
 
-			written += len;
-		do_escape:
-			if (c == 0)
-				break;
-			fmt += ofs;
-			ofs = 0;
-			continue;
+		if (ch != '%') {
+			/* Invalid format: back up to output the format characters */
+			fmt = outstr + 1;
+			/* and output a '%' now. */
 		}
+		/* %% is documented as a 'conversion specifier'.
+		 * Any flags, precision or length modifier are ignored.
+		 */
+		len = 1;
+		width = 0;
+		outstr = fmt - 1;
+		goto do_output;
+
+do_strlen_output:
+		/* Open coded strnlen() (slightly smaller). */
+		for (len = 0; precision < 0 || len < precision; len++)
+			if (!outstr[len])
+				break;
 
-		/* literal char, just queue it */
+do_output:
+		written += len;
+
+		/* Stop gcc back-merging this code into one of the conditionals above. */
+		_NOLIBC_OPTIMIZER_HIDE_VAR(len);
+
+		/* Output the characters on the required side of any padding. */
+		width -= len;
+		flags = _NOLIBC_PF_FLAGS_CONTAIN(flags, '-');
+		if (flags && cb(state, outstr, len) != 0)
+			return -1;
+		while (width > 0) {
+			/* Output pad in 16 byte blocks with the small block first. */
+			int pad_len = ((width - 1) & 15) + 1;
+			width -= pad_len;
+			written += pad_len;
+			if (cb(state, "                ", pad_len) != 0)
+				return -1;
+		}
+		if (!flags && cb(state, outstr, len) != 0)
+			return -1;
 	}
+
+	/* Request a final '\0' be added to the snprintf() output.
+	 * This may be the only call of the cb() function.
+	 */
+	if (cb(state, NULL, 0) != 0)
+		return -1;
+
 	return written;
 }
 
-static int __nolibc_fprintf_cb(intptr_t state, const char *buf, size_t size)
+static int __nolibc_fprintf_cb(void *stream, const char *buf, size_t size)
 {
-	return _fwrite(buf, size, (FILE *)state);
+	return _fwrite(buf, size, stream);
 }
 
 static __attribute__((unused, format(printf, 2, 0)))
 int vfprintf(FILE *stream, const char *fmt, va_list args)
 {
-	return __nolibc_printf(__nolibc_fprintf_cb, (intptr_t)stream, SIZE_MAX, fmt, args);
+	return __nolibc_printf(__nolibc_fprintf_cb, stream, fmt, args);
 }
 
 static __attribute__((unused, format(printf, 1, 0)))
@@ -498,26 +713,54 @@ int dprintf(int fd, const char *fmt, ...)
 	return ret;
 }
 
-static int __nolibc_sprintf_cb(intptr_t _state, const char *buf, size_t size)
+struct __nolibc_sprintf_cb_state {
+	char *buf;
+	size_t space;
+};
+
+static int __nolibc_sprintf_cb(void *v_state, const char *buf, size_t size)
 {
-	char **state = (char **)_state;
+	struct __nolibc_sprintf_cb_state *state = v_state;
+	size_t space = state->space;
+	char *tgt;
+
+	/* Truncate the request to fit in the output buffer space.
+	 * The last byte is reserved for the terminating '\0'.
+	 * state->space can only be zero for snprintf(NULL, 0, fmt, args)
+	 * so this normally lets through calls with 'size == 0'.
+	 */
+	if (size >= space) {
+		if (space <= 1)
+			return 0;
+		size = space - 1;
+	}
+	tgt = state->buf;
+
+	/* __nolibc_printf() ends with cb(state, NULL, 0) to request the output
+	 * buffer be '\0' terminated.
+	 * That will be the only cb() call for, eg, snprintf(buf, sz, "").
+	 * Zero lengths can occur at other times (eg "%s" for an empty string).
+	 * Unconditionally write the '\0' byte to reduce code size, it is
+	 * normally overwritten by the data being output.
+	 * There is no point adding a '\0' after copied data - there is always
+	 * another call.
+	 */
+	*tgt = '\0';
+	if (size) {
+		state->space = space - size;
+		state->buf = tgt + size;
+		memcpy(tgt, buf, size);
+	}
 
-	memcpy(*state, buf, size);
-	*state += size;
 	return 0;
 }
 
 static __attribute__((unused, format(printf, 3, 0)))
 int vsnprintf(char *buf, size_t size, const char *fmt, va_list args)
 {
-	char *state = buf;
-	int ret;
+	struct __nolibc_sprintf_cb_state state = { .buf = buf, .space = size };
 
-	ret = __nolibc_printf(__nolibc_sprintf_cb, (intptr_t)&state, size, fmt, args);
-	if (ret < 0)
-		return ret;
-	buf[(size_t)ret < size ? (size_t)ret : size - 1] = '\0';
-	return ret;
+	return __nolibc_printf(__nolibc_sprintf_cb, &state, fmt, args);
 }
 
 static __attribute__((unused, format(printf, 3, 4)))
@@ -552,6 +795,56 @@ int sprintf(char *buf, const char *fmt, ...)
 	return ret;
 }
 
+static __attribute__((unused, format(printf, 2, 0)))
+int __nolibc_vasprintf(char **strp, const char *fmt, va_list args1, va_list args2)
+{
+	int len1, len2;
+	char *buf;
+
+	len1 = vsnprintf(NULL, 0, fmt, args1);
+	if (len1 < 0)
+		return -1;
+
+	buf = malloc(len1 + 1);
+	if (!buf)
+		return -1;
+
+	len2 = vsnprintf(buf, len1 + 1, fmt, args2);
+	if (len2 < 0) {
+		free(buf);
+		return -1;
+	}
+
+	*strp = buf;
+	return len1;
+}
+
+static __attribute__((unused, format(printf, 2, 0)))
+int vasprintf(char **strp, const char *fmt, va_list args)
+{
+	va_list args2;
+	int ret;
+
+	va_copy(args2, args);
+	ret = __nolibc_vasprintf(strp, fmt, args, args2);
+	va_end(args2);
+
+	return ret;
+}
+
+static __attribute__((unused, format(printf, 2, 3)))
+int asprintf(char **strp, const char *fmt, ...)
+{
+	va_list args;
+	int ret;
+
+	va_start(args, fmt);
+	ret = vasprintf(strp, fmt, args);
+	va_end(args);
+
+	return ret;
+}
+
 static __attribute__((unused))
 int vsscanf(const char *str, const char *format, va_list args)
 {
@@ -683,13 +976,29 @@ int setvbuf(FILE *stream __attribute__((unused)),
 }
 
 static __attribute__((unused))
-const char *strerror(int errno)
+int strerror_r(int errnum, char *buf, size_t buflen)
+{
+	if (buflen < 18)
+		return ERANGE;
+
+	__builtin_memcpy(buf, "errno=", 6);
+	i64toa_r(errnum, buf + 6);
+	return 0;
+}
+
+static __attribute__((unused))
+const char *strerror(int errnum)
 {
-	static char buf[18] = "errno=";
+	static char buf[18];
+	char *b = buf;
+
+	/* Force gcc to use 'register offset' to access buf[]. */
+	_NOLIBC_OPTIMIZER_HIDE_VAR(b);
 
-	i64toa_r(errno, &buf[6]);
+	/* Use strerror_r() to avoid having the only .data in small programs. */
+	strerror_r(errnum, b, sizeof(buf));
 
-	return buf;
+	return b;
 }
 
 #endif /* _NOLIBC_STDIO_H */
diff --git a/tools/include/nolibc/stdlib.h b/tools/include/nolibc/stdlib.h
index f184e10..1816c23 100644
--- a/tools/include/nolibc/stdlib.h
+++ b/tools/include/nolibc/stdlib.h
@@ -55,7 +55,7 @@ void abort(void);
 __attribute__((weak,unused,noreturn,section(".text.nolibc_abort")))
 void abort(void)
 {
-	sys_kill(sys_getpid(), SIGABRT);
+	_sys_kill(_sys_getpid(), SIGABRT);
 	for (;;);
 }
 
@@ -145,9 +145,9 @@ void *malloc(size_t len)
 static __attribute__((unused))
 void *calloc(size_t size, size_t nmemb)
 {
-	size_t x = size * nmemb;
+	size_t x;
 
-	if (__builtin_expect(size && ((x / size) != nmemb), 0)) {
+	if (__builtin_expect(__builtin_mul_overflow(size, nmemb, &x), 0)) {
 		SET_ERRNO(ENOMEM);
 		return NULL;
 	}
@@ -188,36 +188,91 @@ void *realloc(void *old_ptr, size_t new_size)
 	return ret;
 }
 
-/* Converts the unsigned long integer <in> to its hex representation into
+/* Converts the unsigned 64bit integer <in> to base <base> ascii into
  * buffer <buffer>, which must be long enough to store the number and the
- * trailing zero (17 bytes for "ffffffffffffffff" or 9 for "ffffffff"). The
- * buffer is filled from the first byte, and the number of characters emitted
- * (not counting the trailing zero) is returned. The function is constructed
- * in a way to optimize the code size and avoid any divide that could add a
- * dependency on large external functions.
+ * trailing zero. The buffer is filled from the first byte, and the number
+ * of characters emitted (not counting the trailing zero) is returned.
+ * The function uses 'multiply by reciprocal' for the divisions and
+ * requires the caller pass the correct reciprocal.
+ *
+ * Note that unlike __div64_const32() in asm-generic/div64.h there isn't
+ * an extra shift done (by ___p), the reciprocal has to be lower resulting
+ * in a slightly low quotient.
+ * Keep things simple by correcting for the error.
+ * This also saves calculating the 'low * low' product (e2 below) which is
+ * very unlikely to be significant.
+ *
+ * Some maths:
+ *	recip = p2 / base - e1;		// With e1 < base.
+ *	q = (recip * in - e2) / p2;	// With e2 < p2.
+ *        = base / in - (e1 * in + e2) / p2;
+ *        > base / in - (e1 * p2 + p2) / p2;
+ *        = base / in - ((e1 + 1) * p2) / p2;
+ *        > base / in - base;
+ * So the maximum error is less than 'base'.
+ * Hence the largest possible digit is '2 * base - 1'.
+ * For base 10 e1 is 6 and you can get digits of 15 (eg from 2**64-1).
+ * Error e1 is largest for a base that is a factor of 2**64+1, the smallest is 274177
+ * and converting 2**42-1 in base 274177 does generate a digit of 274177+274175.
+ * This all means only a single correction is needed rather than a loop.
+ *
+ * __int128 isn't used for mips because gcc prior to 10.0 will call
+ * __multi3 for MIPS64r6. The same also happens for SPARC and clang.
  */
-static __attribute__((unused))
-int utoh_r(unsigned long in, char *buffer)
+#define _NOLIBC_U64TOA_RECIP(base) ((base) & 1 ? ~0ull / (base) : (1ull << 63) / ((base) / 2))
+static __attribute__((unused, noinline))
+int _nolibc_u64toa_base(uint64_t in, char *buffer, unsigned int base, uint64_t recip)
 {
-	signed char pos = (~0UL > 0xfffffffful) ? 60 : 28;
-	int digits = 0;
-	int dig;
+	unsigned int digits = 0;
+	unsigned int dig;
+	uint64_t q;
+	char *p;
 
+	/* Generate least significant digit first */
 	do {
-		dig = in >> pos;
-		in -= (uint64_t)dig << pos;
-		pos -= 4;
-		if (dig || digits || pos < 0) {
-			if (dig > 9)
-				dig += 'a' - '0' - 10;
-			buffer[digits++] = '0' + dig;
+#if defined(__SIZEOF_INT128__) && !defined(__mips__) && !defined(__sparc__)
+		q = ((unsigned __int128)in * recip) >> 64;
+#else
+		uint64_t p = (uint32_t)in * (recip >> 32);
+		q = (in >> 32) * (recip >> 32) + (p >> 32);
+		p = (uint32_t)p + (in >> 32) * (uint32_t)recip;
+		q += p >> 32;
+#endif
+		dig = in - q * base;
+		/* Correct for any rounding errors */
+		if (dig >= base) {
+			dig -= base;
+			q++;
 		}
-	} while (pos >= 0);
+		if (dig > 9)
+			dig += 'a' - '0' - 10;
+		buffer[digits++] = '0' + dig;
+	} while ((in = q));
 
 	buffer[digits] = 0;
+
+	/* Order reverse to result */
+	for (p = buffer + digits - 1; p > buffer; buffer++, p--) {
+		dig = *buffer;
+		*buffer = *p;
+		*p = dig;
+	}
+
 	return digits;
 }
 
+/* Converts the unsigned long integer <in> to its hex representation into
+ * buffer <buffer>, which must be long enough to store the number and the
+ * trailing zero (17 bytes for "ffffffffffffffff" or 9 for "ffffffff"). The
+ * buffer is filled from the first byte, and the number of characters emitted
+ * (not counting the trailing zero) is returned.
+ */
+static __inline__ __attribute__((unused))
+int utoh_r(unsigned long in, char *buffer)
+{
+	return _nolibc_u64toa_base(in, buffer, 16, _NOLIBC_U64TOA_RECIP(16));
+}
+
 /* converts unsigned long <in> to an hex string using the static itoa_buffer
  * and returns the pointer to that string.
  */
@@ -233,30 +288,11 @@ char *utoh(unsigned long in)
  * trailing zero (21 bytes for 18446744073709551615 in 64-bit, 11 for
  * 4294967295 in 32-bit). The buffer is filled from the first byte, and the
  * number of characters emitted (not counting the trailing zero) is returned.
- * The function is constructed in a way to optimize the code size and avoid
- * any divide that could add a dependency on large external functions.
  */
-static __attribute__((unused))
+static __inline__ __attribute__((unused))
 int utoa_r(unsigned long in, char *buffer)
 {
-	unsigned long lim;
-	int digits = 0;
-	int pos = (~0UL > 0xfffffffful) ? 19 : 9;
-	int dig;
-
-	do {
-		for (dig = 0, lim = 1; dig < pos; dig++)
-			lim *= 10;
-
-		if (digits || in >= lim || !pos) {
-			for (dig = 0; in >= lim; dig++)
-				in -= lim;
-			buffer[digits++] = '0' + dig;
-		}
-	} while (pos--);
-
-	buffer[digits] = 0;
-	return digits;
+	return _nolibc_u64toa_base(in, buffer, 10, _NOLIBC_U64TOA_RECIP(10));
 }
 
 /* Converts the signed long integer <in> to its string representation into
@@ -324,34 +360,12 @@ char *utoa(unsigned long in)
  * buffer <buffer>, which must be long enough to store the number and the
  * trailing zero (17 bytes for "ffffffffffffffff"). The buffer is filled from
  * the first byte, and the number of characters emitted (not counting the
- * trailing zero) is returned. The function is constructed in a way to optimize
- * the code size and avoid any divide that could add a dependency on large
- * external functions.
+ * trailing zero) is returned.
  */
-static __attribute__((unused))
+static __inline__ __attribute__((unused))
 int u64toh_r(uint64_t in, char *buffer)
 {
-	signed char pos = 60;
-	int digits = 0;
-	int dig;
-
-	do {
-		if (sizeof(long) >= 8) {
-			dig = (in >> pos) & 0xF;
-		} else {
-			/* 32-bit platforms: avoid a 64-bit shift */
-			uint32_t d = (pos >= 32) ? (in >> 32) : in;
-			dig = (d >> (pos & 31)) & 0xF;
-		}
-		if (dig > 9)
-			dig += 'a' - '0' - 10;
-		pos -= 4;
-		if (dig || digits || pos < 0)
-			buffer[digits++] = '0' + dig;
-	} while (pos >= 0);
-
-	buffer[digits] = 0;
-	return digits;
+	return _nolibc_u64toa_base(in, buffer, 16, _NOLIBC_U64TOA_RECIP(16));
 }
 
 /* converts uint64_t <in> to an hex string using the static itoa_buffer and
@@ -368,31 +382,12 @@ char *u64toh(uint64_t in)
  * buffer <buffer>, which must be long enough to store the number and the
  * trailing zero (21 bytes for 18446744073709551615). The buffer is filled from
  * the first byte, and the number of characters emitted (not counting the
- * trailing zero) is returned. The function is constructed in a way to optimize
- * the code size and avoid any divide that could add a dependency on large
- * external functions.
+ * trailing zero) is returned.
  */
-static __attribute__((unused))
+static __inline__ __attribute__((unused))
 int u64toa_r(uint64_t in, char *buffer)
 {
-	unsigned long long lim;
-	int digits = 0;
-	int pos = 19; /* start with the highest possible digit */
-	int dig;
-
-	do {
-		for (dig = 0, lim = 1; dig < pos; dig++)
-			lim *= 10;
-
-		if (digits || in >= lim || !pos) {
-			for (dig = 0; in >= lim; dig++)
-				in -= lim;
-			buffer[digits++] = '0' + dig;
-		}
-	} while (pos--);
-
-	buffer[digits] = 0;
-	return digits;
+	return _nolibc_u64toa_base(in, buffer, 10, _NOLIBC_U64TOA_RECIP(10));
 }
 
 /* Converts the signed 64-bit integer <in> to its string representation into
diff --git a/tools/include/nolibc/sys.h b/tools/include/nolibc/sys.h
index 403ee9c..6335fd5 100644
--- a/tools/include/nolibc/sys.h
+++ b/tools/include/nolibc/sys.h
@@ -63,7 +63,7 @@ static __inline__ int __nolibc_enosys(const char *syscall, ...)
  *   - the "internal" ones, which matches the raw syscall interface at the
  *     kernel level, which may sometimes slightly differ from the documented
  *     libc-level ones. For example most of them return either a valid value
- *     or -errno. All of these are prefixed with "sys_". They may be called
+ *     or -errno. All of these are prefixed with "_sys_". They may be called
  *     by non-portable applications if desired.
  *
  *   - the "exported" ones, whose interface must closely match the one
@@ -85,15 +85,15 @@ static __inline__ int __nolibc_enosys(const char *syscall, ...)
  */
 
 static __attribute__((unused))
-void *sys_brk(void *addr)
+void *_sys_brk(void *addr)
 {
-	return (void *)my_syscall1(__NR_brk, addr);
+	return (void *)__nolibc_syscall1(__NR_brk, addr);
 }
 
 static __attribute__((unused))
 int brk(void *addr)
 {
-	void *ret = sys_brk(addr);
+	void *ret = _sys_brk(addr);
 
 	if (!ret) {
 		SET_ERRNO(ENOMEM);
@@ -106,9 +106,9 @@ static __attribute__((unused))
 void *sbrk(intptr_t inc)
 {
 	/* first call to find current end */
-	void *ret = sys_brk(NULL);
+	void *ret = _sys_brk(NULL);
 
-	if (ret && sys_brk(ret + inc) == ret + inc)
+	if (ret && _sys_brk(ret + inc) == ret + inc)
 		return ret + inc;
 
 	SET_ERRNO(ENOMEM);
@@ -122,27 +122,27 @@ void *sbrk(intptr_t inc)
  */
 
 static __attribute__((unused))
-int sys_chdir(const char *path)
+int _sys_chdir(const char *path)
 {
-	return my_syscall1(__NR_chdir, path);
+	return __nolibc_syscall1(__NR_chdir, path);
 }
 
 static __attribute__((unused))
 int chdir(const char *path)
 {
-	return __sysret(sys_chdir(path));
+	return __sysret(_sys_chdir(path));
 }
 
 static __attribute__((unused))
-int sys_fchdir(int fildes)
+int _sys_fchdir(int fildes)
 {
-	return my_syscall1(__NR_fchdir, fildes);
+	return __nolibc_syscall1(__NR_fchdir, fildes);
 }
 
 static __attribute__((unused))
 int fchdir(int fildes)
 {
-	return __sysret(sys_fchdir(fildes));
+	return __sysret(_sys_fchdir(fildes));
 }
 
 
@@ -151,19 +151,19 @@ int fchdir(int fildes)
  */
 
 static __attribute__((unused))
-int sys_chmod(const char *path, mode_t mode)
+int _sys_chmod(const char *path, mode_t mode)
 {
 #if defined(__NR_fchmodat)
-	return my_syscall4(__NR_fchmodat, AT_FDCWD, path, mode, 0);
+	return __nolibc_syscall4(__NR_fchmodat, AT_FDCWD, path, mode, 0);
 #else
-	return my_syscall2(__NR_chmod, path, mode);
+	return __nolibc_syscall2(__NR_chmod, path, mode);
 #endif
 }
 
 static __attribute__((unused))
 int chmod(const char *path, mode_t mode)
 {
-	return __sysret(sys_chmod(path, mode));
+	return __sysret(_sys_chmod(path, mode));
 }
 
 
@@ -172,19 +172,19 @@ int chmod(const char *path, mode_t mode)
  */
 
 static __attribute__((unused))
-int sys_chown(const char *path, uid_t owner, gid_t group)
+int _sys_chown(const char *path, uid_t owner, gid_t group)
 {
 #if defined(__NR_fchownat)
-	return my_syscall5(__NR_fchownat, AT_FDCWD, path, owner, group, 0);
+	return __nolibc_syscall5(__NR_fchownat, AT_FDCWD, path, owner, group, 0);
 #else
-	return my_syscall3(__NR_chown, path, owner, group);
+	return __nolibc_syscall3(__NR_chown, path, owner, group);
 #endif
 }
 
 static __attribute__((unused))
 int chown(const char *path, uid_t owner, gid_t group)
 {
-	return __sysret(sys_chown(path, owner, group));
+	return __sysret(_sys_chown(path, owner, group));
 }
 
 
@@ -193,15 +193,15 @@ int chown(const char *path, uid_t owner, gid_t group)
  */
 
 static __attribute__((unused))
-int sys_chroot(const char *path)
+int _sys_chroot(const char *path)
 {
-	return my_syscall1(__NR_chroot, path);
+	return __nolibc_syscall1(__NR_chroot, path);
 }
 
 static __attribute__((unused))
 int chroot(const char *path)
 {
-	return __sysret(sys_chroot(path));
+	return __sysret(_sys_chroot(path));
 }
 
 
@@ -210,15 +210,15 @@ int chroot(const char *path)
  */
 
 static __attribute__((unused))
-int sys_close(int fd)
+int _sys_close(int fd)
 {
-	return my_syscall1(__NR_close, fd);
+	return __nolibc_syscall1(__NR_close, fd);
 }
 
 static __attribute__((unused))
 int close(int fd)
 {
-	return __sysret(sys_close(fd));
+	return __sysret(_sys_close(fd));
 }
 
 
@@ -227,15 +227,15 @@ int close(int fd)
  */
 
 static __attribute__((unused))
-int sys_dup(int fd)
+int _sys_dup(int fd)
 {
-	return my_syscall1(__NR_dup, fd);
+	return __nolibc_syscall1(__NR_dup, fd);
 }
 
 static __attribute__((unused))
 int dup(int fd)
 {
-	return __sysret(sys_dup(fd));
+	return __sysret(_sys_dup(fd));
 }
 
 
@@ -244,7 +244,7 @@ int dup(int fd)
  */
 
 static __attribute__((unused))
-int sys_dup2(int old, int new)
+int _sys_dup2(int old, int new)
 {
 #if defined(__NR_dup3)
 	int ret, nr_fcntl;
@@ -256,20 +256,20 @@ int sys_dup2(int old, int new)
 #endif
 
 	if (old == new) {
-		ret = my_syscall2(nr_fcntl, old, F_GETFD);
+		ret = __nolibc_syscall2(nr_fcntl, old, F_GETFD);
 		return ret < 0 ? ret : old;
 	}
 
-	return my_syscall3(__NR_dup3, old, new, 0);
+	return __nolibc_syscall3(__NR_dup3, old, new, 0);
 #else
-	return my_syscall2(__NR_dup2, old, new);
+	return __nolibc_syscall2(__NR_dup2, old, new);
 #endif
 }
 
 static __attribute__((unused))
 int dup2(int old, int new)
 {
-	return __sysret(sys_dup2(old, new));
+	return __sysret(_sys_dup2(old, new));
 }
 
 
@@ -279,15 +279,15 @@ int dup2(int old, int new)
 
 #if defined(__NR_dup3)
 static __attribute__((unused))
-int sys_dup3(int old, int new, int flags)
+int _sys_dup3(int old, int new, int flags)
 {
-	return my_syscall3(__NR_dup3, old, new, flags);
+	return __nolibc_syscall3(__NR_dup3, old, new, flags);
 }
 
 static __attribute__((unused))
 int dup3(int old, int new, int flags)
 {
-	return __sysret(sys_dup3(old, new, flags));
+	return __sysret(_sys_dup3(old, new, flags));
 }
 #endif
 
@@ -297,15 +297,15 @@ int dup3(int old, int new, int flags)
  */
 
 static __attribute__((unused))
-int sys_execve(const char *filename, char *const argv[], char *const envp[])
+int _sys_execve(const char *filename, char *const argv[], char *const envp[])
 {
-	return my_syscall3(__NR_execve, filename, argv, envp);
+	return __nolibc_syscall3(__NR_execve, filename, argv, envp);
 }
 
 static __attribute__((unused))
 int execve(const char *filename, char *const argv[], char *const envp[])
 {
-	return __sysret(sys_execve(filename, argv, envp));
+	return __sysret(_sys_execve(filename, argv, envp));
 }
 
 
@@ -314,16 +314,16 @@ int execve(const char *filename, char *const argv[], char *const envp[])
  */
 
 static __attribute__((noreturn,unused))
-void sys_exit(int status)
+void _sys_exit(int status)
 {
-	my_syscall1(__NR_exit, status & 255);
+	__nolibc_syscall1(__NR_exit, status & 255);
 	while(1); /* shut the "noreturn" warnings. */
 }
 
 static __attribute__((noreturn,unused))
 void _exit(int status)
 {
-	sys_exit(status);
+	_sys_exit(status);
 }
 
 static __attribute__((noreturn,unused))
@@ -337,18 +337,18 @@ void exit(int status)
  * pid_t fork(void);
  */
 
-#ifndef sys_fork
+#ifndef _sys_fork
 static __attribute__((unused))
-pid_t sys_fork(void)
+pid_t _sys_fork(void)
 {
 #if defined(__NR_clone)
 	/* note: some archs only have clone() and not fork(). Different archs
 	 * have a different API, but most archs have the flags on first arg and
 	 * will not use the rest with no other flag.
 	 */
-	return my_syscall5(__NR_clone, SIGCHLD, 0, 0, 0, 0);
+	return __nolibc_syscall5(__NR_clone, SIGCHLD, 0, 0, 0, 0);
 #else
-	return my_syscall0(__NR_fork);
+	return __nolibc_syscall0(__NR_fork);
 #endif
 }
 #endif
@@ -356,18 +356,18 @@ pid_t sys_fork(void)
 static __attribute__((unused))
 pid_t fork(void)
 {
-	return __sysret(sys_fork());
+	return __sysret(_sys_fork());
 }
 
-#ifndef sys_vfork
+#ifndef _sys_vfork
 static __attribute__((unused))
-pid_t sys_vfork(void)
+pid_t _sys_vfork(void)
 {
 #if defined(__NR_clone)
-	/* See the note in sys_fork(). */
-	return my_syscall5(__NR_clone, CLONE_VM | CLONE_VFORK | SIGCHLD, 0, 0, 0, 0);
+	/* See the note in _sys_fork(). */
+	return __nolibc_syscall5(__NR_clone, CLONE_VM | CLONE_VFORK | SIGCHLD, 0, 0, 0, 0);
 #elif defined(__NR_vfork)
-	return my_syscall0(__NR_vfork);
+	return __nolibc_syscall0(__NR_vfork);
 #endif
 }
 #endif
@@ -375,7 +375,7 @@ pid_t sys_vfork(void)
 static __attribute__((unused))
 pid_t vfork(void)
 {
-	return __sysret(sys_vfork());
+	return __sysret(_sys_vfork());
 }
 
 /*
@@ -383,15 +383,15 @@ pid_t vfork(void)
  */
 
 static __attribute__((unused))
-int sys_fsync(int fd)
+int _sys_fsync(int fd)
 {
-	return my_syscall1(__NR_fsync, fd);
+	return __nolibc_syscall1(__NR_fsync, fd);
 }
 
 static __attribute__((unused))
 int fsync(int fd)
 {
-	return __sysret(sys_fsync(fd));
+	return __sysret(_sys_fsync(fd));
 }
 
 
@@ -400,15 +400,15 @@ int fsync(int fd)
  */
 
 static __attribute__((unused))
-int sys_getdents64(int fd, struct linux_dirent64 *dirp, int count)
+int _sys_getdents64(int fd, struct linux_dirent64 *dirp, int count)
 {
-	return my_syscall3(__NR_getdents64, fd, dirp, count);
+	return __nolibc_syscall3(__NR_getdents64, fd, dirp, count);
 }
 
 static __attribute__((unused))
 int getdents64(int fd, struct linux_dirent64 *dirp, int count)
 {
-	return __sysret(sys_getdents64(fd, dirp, count));
+	return __sysret(_sys_getdents64(fd, dirp, count));
 }
 
 
@@ -417,19 +417,19 @@ int getdents64(int fd, struct linux_dirent64 *dirp, int count)
  */
 
 static __attribute__((unused))
-uid_t sys_geteuid(void)
+uid_t _sys_geteuid(void)
 {
 #if defined(__NR_geteuid32)
-	return my_syscall0(__NR_geteuid32);
+	return __nolibc_syscall0(__NR_geteuid32);
 #else
-	return my_syscall0(__NR_geteuid);
+	return __nolibc_syscall0(__NR_geteuid);
 #endif
 }
 
 static __attribute__((unused))
 uid_t geteuid(void)
 {
-	return sys_geteuid();
+	return _sys_geteuid();
 }
 
 
@@ -438,15 +438,15 @@ uid_t geteuid(void)
  */
 
 static __attribute__((unused))
-pid_t sys_getpgid(pid_t pid)
+pid_t _sys_getpgid(pid_t pid)
 {
-	return my_syscall1(__NR_getpgid, pid);
+	return __nolibc_syscall1(__NR_getpgid, pid);
 }
 
 static __attribute__((unused))
 pid_t getpgid(pid_t pid)
 {
-	return __sysret(sys_getpgid(pid));
+	return __sysret(_sys_getpgid(pid));
 }
 
 
@@ -455,15 +455,15 @@ pid_t getpgid(pid_t pid)
  */
 
 static __attribute__((unused))
-pid_t sys_getpgrp(void)
+pid_t _sys_getpgrp(void)
 {
-	return sys_getpgid(0);
+	return _sys_getpgid(0);
 }
 
 static __attribute__((unused))
 pid_t getpgrp(void)
 {
-	return sys_getpgrp();
+	return _sys_getpgrp();
 }
 
 
@@ -472,15 +472,15 @@ pid_t getpgrp(void)
  */
 
 static __attribute__((unused))
-pid_t sys_getpid(void)
+pid_t _sys_getpid(void)
 {
-	return my_syscall0(__NR_getpid);
+	return __nolibc_syscall0(__NR_getpid);
 }
 
 static __attribute__((unused))
 pid_t getpid(void)
 {
-	return sys_getpid();
+	return _sys_getpid();
 }
 
 
@@ -489,15 +489,15 @@ pid_t getpid(void)
  */
 
 static __attribute__((unused))
-pid_t sys_getppid(void)
+pid_t _sys_getppid(void)
 {
-	return my_syscall0(__NR_getppid);
+	return __nolibc_syscall0(__NR_getppid);
 }
 
 static __attribute__((unused))
 pid_t getppid(void)
 {
-	return sys_getppid();
+	return _sys_getppid();
 }
 
 
@@ -506,15 +506,15 @@ pid_t getppid(void)
  */
 
 static __attribute__((unused))
-pid_t sys_gettid(void)
+pid_t _sys_gettid(void)
 {
-	return my_syscall0(__NR_gettid);
+	return __nolibc_syscall0(__NR_gettid);
 }
 
 static __attribute__((unused))
 pid_t gettid(void)
 {
-	return sys_gettid();
+	return _sys_gettid();
 }
 
 #ifndef NOLIBC_NO_RUNTIME
@@ -536,19 +536,19 @@ int getpagesize(void)
  */
 
 static __attribute__((unused))
-uid_t sys_getuid(void)
+uid_t _sys_getuid(void)
 {
 #if defined(__NR_getuid32)
-	return my_syscall0(__NR_getuid32);
+	return __nolibc_syscall0(__NR_getuid32);
 #else
-	return my_syscall0(__NR_getuid);
+	return __nolibc_syscall0(__NR_getuid);
 #endif
 }
 
 static __attribute__((unused))
 uid_t getuid(void)
 {
-	return sys_getuid();
+	return _sys_getuid();
 }
 
 
@@ -557,15 +557,15 @@ uid_t getuid(void)
  */
 
 static __attribute__((unused))
-int sys_kill(pid_t pid, int signal)
+int _sys_kill(pid_t pid, int signal)
 {
-	return my_syscall2(__NR_kill, pid, signal);
+	return __nolibc_syscall2(__NR_kill, pid, signal);
 }
 
 static __attribute__((unused))
 int kill(pid_t pid, int signal)
 {
-	return __sysret(sys_kill(pid, signal));
+	return __sysret(_sys_kill(pid, signal));
 }
 
 
@@ -574,19 +574,19 @@ int kill(pid_t pid, int signal)
  */
 
 static __attribute__((unused))
-int sys_link(const char *old, const char *new)
+int _sys_link(const char *old, const char *new)
 {
 #if defined(__NR_linkat)
-	return my_syscall5(__NR_linkat, AT_FDCWD, old, AT_FDCWD, new, 0);
+	return __nolibc_syscall5(__NR_linkat, AT_FDCWD, old, AT_FDCWD, new, 0);
 #else
-	return my_syscall2(__NR_link, old, new);
+	return __nolibc_syscall2(__NR_link, old, new);
 #endif
 }
 
 static __attribute__((unused))
 int link(const char *old, const char *new)
 {
-	return __sysret(sys_link(old, new));
+	return __sysret(_sys_link(old, new));
 }
 
 
@@ -595,14 +595,14 @@ int link(const char *old, const char *new)
  */
 
 static __attribute__((unused))
-off_t sys_lseek(int fd, off_t offset, int whence)
+off_t _sys_lseek(int fd, off_t offset, int whence)
 {
 #if defined(__NR_llseek)
 	__kernel_loff_t loff = 0;
 	off_t result;
 	int ret;
 
-	ret = my_syscall5(__NR_llseek, fd, offset >> 32, (uint32_t)offset, &loff, whence);
+	ret = __nolibc_syscall5(__NR_llseek, fd, offset >> 32, (uint32_t)offset, &loff, whence);
 	if (ret < 0)
 		result = ret;
 	else
@@ -610,14 +610,14 @@ off_t sys_lseek(int fd, off_t offset, int whence)
 
 	return result;
 #else
-	return my_syscall3(__NR_lseek, fd, offset, whence);
+	return __nolibc_syscall3(__NR_lseek, fd, offset, whence);
 #endif
 }
 
 static __attribute__((unused))
 off_t lseek(int fd, off_t offset, int whence)
 {
-	return __sysret(sys_lseek(fd, offset, whence));
+	return __sysret(_sys_lseek(fd, offset, whence));
 }
 
 
@@ -626,19 +626,19 @@ off_t lseek(int fd, off_t offset, int whence)
  */
 
 static __attribute__((unused))
-int sys_mkdir(const char *path, mode_t mode)
+int _sys_mkdir(const char *path, mode_t mode)
 {
 #if defined(__NR_mkdirat)
-	return my_syscall3(__NR_mkdirat, AT_FDCWD, path, mode);
+	return __nolibc_syscall3(__NR_mkdirat, AT_FDCWD, path, mode);
 #else
-	return my_syscall2(__NR_mkdir, path, mode);
+	return __nolibc_syscall2(__NR_mkdir, path, mode);
 #endif
 }
 
 static __attribute__((unused))
 int mkdir(const char *path, mode_t mode)
 {
-	return __sysret(sys_mkdir(path, mode));
+	return __sysret(_sys_mkdir(path, mode));
 }
 
 /*
@@ -646,19 +646,19 @@ int mkdir(const char *path, mode_t mode)
  */
 
 static __attribute__((unused))
-int sys_rmdir(const char *path)
+int _sys_rmdir(const char *path)
 {
 #if defined(__NR_rmdir)
-	return my_syscall1(__NR_rmdir, path);
+	return __nolibc_syscall1(__NR_rmdir, path);
 #else
-	return my_syscall3(__NR_unlinkat, AT_FDCWD, path, AT_REMOVEDIR);
+	return __nolibc_syscall3(__NR_unlinkat, AT_FDCWD, path, AT_REMOVEDIR);
 #endif
 }
 
 static __attribute__((unused))
 int rmdir(const char *path)
 {
-	return __sysret(sys_rmdir(path));
+	return __sysret(_sys_rmdir(path));
 }
 
 
@@ -667,19 +667,19 @@ int rmdir(const char *path)
  */
 
 static __attribute__((unused))
-long sys_mknod(const char *path, mode_t mode, dev_t dev)
+long _sys_mknod(const char *path, mode_t mode, dev_t dev)
 {
 #if defined(__NR_mknodat)
-	return my_syscall4(__NR_mknodat, AT_FDCWD, path, mode, dev);
+	return __nolibc_syscall4(__NR_mknodat, AT_FDCWD, path, mode, dev);
 #else
-	return my_syscall3(__NR_mknod, path, mode, dev);
+	return __nolibc_syscall3(__NR_mknod, path, mode, dev);
 #endif
 }
 
 static __attribute__((unused))
 int mknod(const char *path, mode_t mode, dev_t dev)
 {
-	return __sysret(sys_mknod(path, mode, dev));
+	return __sysret(_sys_mknod(path, mode, dev));
 }
 
 
@@ -689,15 +689,15 @@ int mknod(const char *path, mode_t mode, dev_t dev)
  */
 
 static __attribute__((unused))
-int sys_pipe2(int pipefd[2], int flags)
+int _sys_pipe2(int pipefd[2], int flags)
 {
-	return my_syscall2(__NR_pipe2, pipefd, flags);
+	return __nolibc_syscall2(__NR_pipe2, pipefd, flags);
 }
 
 static __attribute__((unused))
 int pipe2(int pipefd[2], int flags)
 {
-	return __sysret(sys_pipe2(pipefd, flags));
+	return __sysret(_sys_pipe2(pipefd, flags));
 }
 
 static __attribute__((unused))
@@ -712,15 +712,15 @@ int pipe(int pipefd[2])
  */
 
 static __attribute__((unused))
-int sys_pivot_root(const char *new, const char *old)
+int _sys_pivot_root(const char *new, const char *old)
 {
-	return my_syscall2(__NR_pivot_root, new, old);
+	return __nolibc_syscall2(__NR_pivot_root, new, old);
 }
 
 static __attribute__((unused))
 int pivot_root(const char *new, const char *old)
 {
-	return __sysret(sys_pivot_root(new, old));
+	return __sysret(_sys_pivot_root(new, old));
 }
 
 
@@ -729,15 +729,15 @@ int pivot_root(const char *new, const char *old)
  */
 
 static __attribute__((unused))
-ssize_t sys_read(int fd, void *buf, size_t count)
+ssize_t _sys_read(int fd, void *buf, size_t count)
 {
-	return my_syscall3(__NR_read, fd, buf, count);
+	return __nolibc_syscall3(__NR_read, fd, buf, count);
 }
 
 static __attribute__((unused))
 ssize_t read(int fd, void *buf, size_t count)
 {
-	return __sysret(sys_read(fd, buf, count));
+	return __sysret(_sys_read(fd, buf, count));
 }
 
 
@@ -746,15 +746,15 @@ ssize_t read(int fd, void *buf, size_t count)
  */
 
 static __attribute__((unused))
-int sys_sched_yield(void)
+int _sys_sched_yield(void)
 {
-	return my_syscall0(__NR_sched_yield);
+	return __nolibc_syscall0(__NR_sched_yield);
 }
 
 static __attribute__((unused))
 int sched_yield(void)
 {
-	return __sysret(sys_sched_yield());
+	return __sysret(_sys_sched_yield());
 }
 
 
@@ -763,15 +763,15 @@ int sched_yield(void)
  */
 
 static __attribute__((unused))
-int sys_setpgid(pid_t pid, pid_t pgid)
+int _sys_setpgid(pid_t pid, pid_t pgid)
 {
-	return my_syscall2(__NR_setpgid, pid, pgid);
+	return __nolibc_syscall2(__NR_setpgid, pid, pgid);
 }
 
 static __attribute__((unused))
 int setpgid(pid_t pid, pid_t pgid)
 {
-	return __sysret(sys_setpgid(pid, pgid));
+	return __sysret(_sys_setpgid(pid, pgid));
 }
 
 /*
@@ -790,15 +790,15 @@ pid_t setpgrp(void)
  */
 
 static __attribute__((unused))
-pid_t sys_setsid(void)
+pid_t _sys_setsid(void)
 {
-	return my_syscall0(__NR_setsid);
+	return __nolibc_syscall0(__NR_setsid);
 }
 
 static __attribute__((unused))
 pid_t setsid(void)
 {
-	return __sysret(sys_setsid());
+	return __sysret(_sys_setsid());
 }
 
 
@@ -807,19 +807,19 @@ pid_t setsid(void)
  */
 
 static __attribute__((unused))
-int sys_symlink(const char *old, const char *new)
+int _sys_symlink(const char *old, const char *new)
 {
 #if defined(__NR_symlinkat)
-	return my_syscall3(__NR_symlinkat, old, AT_FDCWD, new);
+	return __nolibc_syscall3(__NR_symlinkat, old, AT_FDCWD, new);
 #else
-	return my_syscall2(__NR_symlink, old, new);
+	return __nolibc_syscall2(__NR_symlink, old, new);
 #endif
 }
 
 static __attribute__((unused))
 int symlink(const char *old, const char *new)
 {
-	return __sysret(sys_symlink(old, new));
+	return __sysret(_sys_symlink(old, new));
 }
 
 
@@ -828,15 +828,15 @@ int symlink(const char *old, const char *new)
  */
 
 static __attribute__((unused))
-mode_t sys_umask(mode_t mode)
+mode_t _sys_umask(mode_t mode)
 {
-	return my_syscall1(__NR_umask, mode);
+	return __nolibc_syscall1(__NR_umask, mode);
 }
 
 static __attribute__((unused))
 mode_t umask(mode_t mode)
 {
-	return sys_umask(mode);
+	return _sys_umask(mode);
 }
 
 
@@ -845,15 +845,15 @@ mode_t umask(mode_t mode)
  */
 
 static __attribute__((unused))
-int sys_umount2(const char *path, int flags)
+int _sys_umount2(const char *path, int flags)
 {
-	return my_syscall2(__NR_umount2, path, flags);
+	return __nolibc_syscall2(__NR_umount2, path, flags);
 }
 
 static __attribute__((unused))
 int umount2(const char *path, int flags)
 {
-	return __sysret(sys_umount2(path, flags));
+	return __sysret(_sys_umount2(path, flags));
 }
 
 
@@ -862,19 +862,19 @@ int umount2(const char *path, int flags)
  */
 
 static __attribute__((unused))
-int sys_unlink(const char *path)
+int _sys_unlink(const char *path)
 {
 #if defined(__NR_unlinkat)
-	return my_syscall3(__NR_unlinkat, AT_FDCWD, path, 0);
+	return __nolibc_syscall3(__NR_unlinkat, AT_FDCWD, path, 0);
 #else
-	return my_syscall1(__NR_unlink, path);
+	return __nolibc_syscall1(__NR_unlink, path);
 #endif
 }
 
 static __attribute__((unused))
 int unlink(const char *path)
 {
-	return __sysret(sys_unlink(path));
+	return __sysret(_sys_unlink(path));
 }
 
 
@@ -883,15 +883,15 @@ int unlink(const char *path)
  */
 
 static __attribute__((unused))
-ssize_t sys_write(int fd, const void *buf, size_t count)
+ssize_t _sys_write(int fd, const void *buf, size_t count)
 {
-	return my_syscall3(__NR_write, fd, buf, count);
+	return __nolibc_syscall3(__NR_write, fd, buf, count);
 }
 
 static __attribute__((unused))
 ssize_t write(int fd, const void *buf, size_t count)
 {
-	return __sysret(sys_write(fd, buf, count));
+	return __sysret(_sys_write(fd, buf, count));
 }
 
 
@@ -900,15 +900,15 @@ ssize_t write(int fd, const void *buf, size_t count)
  */
 
 static __attribute__((unused))
-int sys_memfd_create(const char *name, unsigned int flags)
+int _sys_memfd_create(const char *name, unsigned int flags)
 {
-	return my_syscall2(__NR_memfd_create, name, flags);
+	return __nolibc_syscall2(__NR_memfd_create, name, flags);
 }
 
 static __attribute__((unused))
 int memfd_create(const char *name, unsigned int flags)
 {
-	return __sysret(sys_memfd_create(name, flags));
+	return __sysret(_sys_memfd_create(name, flags));
 }
 
 #endif /* _NOLIBC_SYS_H */
diff --git a/tools/include/nolibc/sys/ioctl.h b/tools/include/nolibc/sys/ioctl.h
index fc88068..f062a7b 100644
--- a/tools/include/nolibc/sys/ioctl.h
+++ b/tools/include/nolibc/sys/ioctl.h
@@ -19,11 +19,11 @@
  */
 
 static __attribute__((unused))
-long sys_ioctl(unsigned int fd, unsigned int cmd, unsigned long arg)
+long _sys_ioctl(unsigned int fd, unsigned int cmd, unsigned long arg)
 {
-	return my_syscall3(__NR_ioctl, fd, cmd, arg);
+	return __nolibc_syscall3(__NR_ioctl, fd, cmd, arg);
 }
 
-#define ioctl(fd, cmd, arg) __sysret(sys_ioctl(fd, cmd, (unsigned long)(arg)))
+#define ioctl(fd, cmd, arg) __sysret(_sys_ioctl(fd, cmd, (unsigned long)(arg)))
 
 #endif /* _NOLIBC_SYS_IOCTL_H */
diff --git a/tools/include/nolibc/sys/mman.h b/tools/include/nolibc/sys/mman.h
index 77084ac..91d77a5 100644
--- a/tools/include/nolibc/sys/mman.h
+++ b/tools/include/nolibc/sys/mman.h
@@ -13,10 +13,10 @@
 #include "../arch.h"
 #include "../sys.h"
 
-#ifndef sys_mmap
+#ifndef _sys_mmap
 static __attribute__((unused))
-void *sys_mmap(void *addr, size_t length, int prot, int flags, int fd,
-	       off_t offset)
+void *_sys_mmap(void *addr, size_t length, int prot, int flags, int fd,
+		off_t offset)
 {
 	int n;
 
@@ -27,14 +27,14 @@ void *sys_mmap(void *addr, size_t length, int prot, int flags, int fd,
 	n = __NR_mmap;
 #endif
 
-	return (void *)my_syscall6(n, addr, length, prot, flags, fd, offset);
+	return (void *)__nolibc_syscall6(n, addr, length, prot, flags, fd, offset);
 }
 #endif
 
 static __attribute__((unused))
 void *mmap(void *addr, size_t length, int prot, int flags, int fd, off_t offset)
 {
-	void *ret = sys_mmap(addr, length, prot, flags, fd, offset);
+	void *ret = _sys_mmap(addr, length, prot, flags, fd, offset);
 
 	if ((unsigned long)ret >= -4095UL) {
 		SET_ERRNO(-(long)ret);
@@ -44,16 +44,16 @@ void *mmap(void *addr, size_t length, int prot, int flags, int fd, off_t offset)
 }
 
 static __attribute__((unused))
-void *sys_mremap(void *old_address, size_t old_size, size_t new_size, int flags, void *new_address)
+void *_sys_mremap(void *old_address, size_t old_size, size_t new_size, int flags, void *new_address)
 {
-	return (void *)my_syscall5(__NR_mremap, old_address, old_size,
-				   new_size, flags, new_address);
+	return (void *)__nolibc_syscall5(__NR_mremap, old_address, old_size,
+					 new_size, flags, new_address);
 }
 
 static __attribute__((unused))
 void *mremap(void *old_address, size_t old_size, size_t new_size, int flags, void *new_address)
 {
-	void *ret = sys_mremap(old_address, old_size, new_size, flags, new_address);
+	void *ret = _sys_mremap(old_address, old_size, new_size, flags, new_address);
 
 	if ((unsigned long)ret >= -4095UL) {
 		SET_ERRNO(-(long)ret);
@@ -63,15 +63,15 @@ void *mremap(void *old_address, size_t old_size, size_t new_size, int flags, voi
 }
 
 static __attribute__((unused))
-int sys_munmap(void *addr, size_t length)
+int _sys_munmap(void *addr, size_t length)
 {
-	return my_syscall2(__NR_munmap, addr, length);
+	return __nolibc_syscall2(__NR_munmap, addr, length);
 }
 
 static __attribute__((unused))
 int munmap(void *addr, size_t length)
 {
-	return __sysret(sys_munmap(addr, length));
+	return __sysret(_sys_munmap(addr, length));
 }
 
 #endif /* _NOLIBC_SYS_MMAN_H */
diff --git a/tools/include/nolibc/sys/mount.h b/tools/include/nolibc/sys/mount.h
index e39ec02..8d3128e 100644
--- a/tools/include/nolibc/sys/mount.h
+++ b/tools/include/nolibc/sys/mount.h
@@ -20,10 +20,10 @@
  *           const void *data);
  */
 static __attribute__((unused))
-int sys_mount(const char *src, const char *tgt, const char *fst,
-	      unsigned long flags, const void *data)
+int _sys_mount(const char *src, const char *tgt, const char *fst,
+	       unsigned long flags, const void *data)
 {
-	return my_syscall5(__NR_mount, src, tgt, fst, flags, data);
+	return __nolibc_syscall5(__NR_mount, src, tgt, fst, flags, data);
 }
 
 static __attribute__((unused))
@@ -31,7 +31,7 @@ int mount(const char *src, const char *tgt,
 	  const char *fst, unsigned long flags,
 	  const void *data)
 {
-	return __sysret(sys_mount(src, tgt, fst, flags, data));
+	return __sysret(_sys_mount(src, tgt, fst, flags, data));
 }
 
 #endif /* _NOLIBC_SYS_MOUNT_H */
diff --git a/tools/include/nolibc/sys/prctl.h b/tools/include/nolibc/sys/prctl.h
index 0205907..3e11d3e 100644
--- a/tools/include/nolibc/sys/prctl.h
+++ b/tools/include/nolibc/sys/prctl.h
@@ -20,17 +20,17 @@
  */
 
 static __attribute__((unused))
-int sys_prctl(int option, unsigned long arg2, unsigned long arg3,
-			  unsigned long arg4, unsigned long arg5)
+int _sys_prctl(int option, unsigned long arg2, unsigned long arg3,
+			   unsigned long arg4, unsigned long arg5)
 {
-	return my_syscall5(__NR_prctl, option, arg2, arg3, arg4, arg5);
+	return __nolibc_syscall5(__NR_prctl, option, arg2, arg3, arg4, arg5);
 }
 
 static __attribute__((unused))
 int prctl(int option, unsigned long arg2, unsigned long arg3,
 		      unsigned long arg4, unsigned long arg5)
 {
-	return __sysret(sys_prctl(option, arg2, arg3, arg4, arg5));
+	return __sysret(_sys_prctl(option, arg2, arg3, arg4, arg5));
 }
 
 #endif /* _NOLIBC_SYS_PRCTL_H */
diff --git a/tools/include/nolibc/sys/ptrace.h b/tools/include/nolibc/sys/ptrace.h
index 72ca285..4c4820f 100644
--- a/tools/include/nolibc/sys/ptrace.h
+++ b/tools/include/nolibc/sys/ptrace.h
@@ -19,15 +19,15 @@
  * long ptrace(int op, pid_t pid, void *addr, void *data);
  */
 static __attribute__((unused))
-long sys_ptrace(int op, pid_t pid, void *addr, void *data)
+long _sys_ptrace(int op, pid_t pid, void *addr, void *data)
 {
-	return my_syscall4(__NR_ptrace, op, pid, addr, data);
+	return __nolibc_syscall4(__NR_ptrace, op, pid, addr, data);
 }
 
 static __attribute__((unused))
 ssize_t ptrace(int op, pid_t pid, void *addr, void *data)
 {
-	return __sysret(sys_ptrace(op, pid, addr, data));
+	return __sysret(_sys_ptrace(op, pid, addr, data));
 }
 
 #endif /* _NOLIBC_SYS_PTRACE_H */
diff --git a/tools/include/nolibc/sys/random.h b/tools/include/nolibc/sys/random.h
index cd5d25c..6d055b0 100644
--- a/tools/include/nolibc/sys/random.h
+++ b/tools/include/nolibc/sys/random.h
@@ -20,15 +20,15 @@
  */
 
 static __attribute__((unused))
-ssize_t sys_getrandom(void *buf, size_t buflen, unsigned int flags)
+ssize_t _sys_getrandom(void *buf, size_t buflen, unsigned int flags)
 {
-	return my_syscall3(__NR_getrandom, buf, buflen, flags);
+	return __nolibc_syscall3(__NR_getrandom, buf, buflen, flags);
 }
 
 static __attribute__((unused))
 ssize_t getrandom(void *buf, size_t buflen, unsigned int flags)
 {
-	return __sysret(sys_getrandom(buf, buflen, flags));
+	return __sysret(_sys_getrandom(buf, buflen, flags));
 }
 
 #endif /* _NOLIBC_SYS_RANDOM_H */
diff --git a/tools/include/nolibc/sys/reboot.h b/tools/include/nolibc/sys/reboot.h
index 38274c6..73cbca3 100644
--- a/tools/include/nolibc/sys/reboot.h
+++ b/tools/include/nolibc/sys/reboot.h
@@ -20,15 +20,15 @@
  */
 
 static __attribute__((unused))
-ssize_t sys_reboot(int magic1, int magic2, int cmd, void *arg)
+ssize_t _sys_reboot(int magic1, int magic2, int cmd, void *arg)
 {
-	return my_syscall4(__NR_reboot, magic1, magic2, cmd, arg);
+	return __nolibc_syscall4(__NR_reboot, magic1, magic2, cmd, arg);
 }
 
 static __attribute__((unused))
 int reboot(int cmd)
 {
-	return __sysret(sys_reboot(LINUX_REBOOT_MAGIC1, LINUX_REBOOT_MAGIC2, cmd, NULL));
+	return __sysret(_sys_reboot(LINUX_REBOOT_MAGIC1, LINUX_REBOOT_MAGIC2, cmd, NULL));
 }
 
 #endif /* _NOLIBC_SYS_REBOOT_H */
diff --git a/tools/include/nolibc/sys/resource.h b/tools/include/nolibc/sys/resource.h
index b990f91..f35de1c 100644
--- a/tools/include/nolibc/sys/resource.h
+++ b/tools/include/nolibc/sys/resource.h
@@ -20,10 +20,10 @@
  */
 
 static __attribute__((unused))
-int sys_prlimit64(pid_t pid, int resource,
-		  const struct rlimit64 *new_limit, struct rlimit64 *old_limit)
+int _sys_prlimit64(pid_t pid, int resource,
+		   const struct rlimit64 *new_limit, struct rlimit64 *old_limit)
 {
-	return my_syscall4(__NR_prlimit64, pid, resource, new_limit, old_limit);
+	return __nolibc_syscall4(__NR_prlimit64, pid, resource, new_limit, old_limit);
 }
 
 static __attribute__((unused))
@@ -32,7 +32,7 @@ int getrlimit(int resource, struct rlimit *rlim)
 	struct rlimit64 rlim64;
 	int ret;
 
-	ret = __sysret(sys_prlimit64(0, resource, NULL, &rlim64));
+	ret = __sysret(_sys_prlimit64(0, resource, NULL, &rlim64));
 	rlim->rlim_cur = rlim64.rlim_cur;
 	rlim->rlim_max = rlim64.rlim_max;
 
@@ -47,7 +47,7 @@ int setrlimit(int resource, const struct rlimit *rlim)
 		.rlim_max = rlim->rlim_max,
 	};
 
-	return __sysret(sys_prlimit64(0, resource, &rlim64, NULL));
+	return __sysret(_sys_prlimit64(0, resource, &rlim64, NULL));
 }
 
 #endif /* _NOLIBC_SYS_RESOURCE_H */
diff --git a/tools/include/nolibc/sys/select.h b/tools/include/nolibc/sys/select.h
index 80cb375..6d65d9e 100644
--- a/tools/include/nolibc/sys/select.h
+++ b/tools/include/nolibc/sys/select.h
@@ -61,7 +61,7 @@ typedef struct {
  */
 
 static __attribute__((unused))
-int sys_select(int nfds, fd_set *rfds, fd_set *wfds, fd_set *efds, struct timeval *timeout)
+int _sys_select(int nfds, fd_set *rfds, fd_set *wfds, fd_set *efds, struct timeval *timeout)
 {
 #if defined(__NR_pselect6_time64)
 	struct __kernel_timespec t;
@@ -70,7 +70,8 @@ int sys_select(int nfds, fd_set *rfds, fd_set *wfds, fd_set *efds, struct timeva
 		t.tv_sec  = timeout->tv_sec;
 		t.tv_nsec = (uint32_t)timeout->tv_usec * 1000;
 	}
-	return my_syscall6(__NR_pselect6_time64, nfds, rfds, wfds, efds, timeout ? &t : NULL, NULL);
+	return __nolibc_syscall6(__NR_pselect6_time64, nfds, rfds, wfds, efds,
+				 timeout ? &t : NULL, NULL);
 #else
 	struct __kernel_old_timespec t;
 
@@ -78,14 +79,15 @@ int sys_select(int nfds, fd_set *rfds, fd_set *wfds, fd_set *efds, struct timeva
 		t.tv_sec  = timeout->tv_sec;
 		t.tv_nsec = (uint32_t)timeout->tv_usec * 1000;
 	}
-	return my_syscall6(__NR_pselect6, nfds, rfds, wfds, efds, timeout ? &t : NULL, NULL);
+	return __nolibc_syscall6(__NR_pselect6, nfds, rfds, wfds, efds,
+				 timeout ? &t : NULL, NULL);
 #endif
 }
 
 static __attribute__((unused))
 int select(int nfds, fd_set *rfds, fd_set *wfds, fd_set *efds, struct timeval *timeout)
 {
-	return __sysret(sys_select(nfds, rfds, wfds, efds, timeout));
+	return __sysret(_sys_select(nfds, rfds, wfds, efds, timeout));
 }
 
 
diff --git a/tools/include/nolibc/sys/stat.h b/tools/include/nolibc/sys/stat.h
index 8b4d80e..07ae715 100644
--- a/tools/include/nolibc/sys/stat.h
+++ b/tools/include/nolibc/sys/stat.h
@@ -13,6 +13,7 @@
 #include "../arch.h"
 #include "../types.h"
 #include "../sys.h"
+#include "../sys/sysmacros.h"
 
 /*
  * int statx(int fd, const char *path, int flags, unsigned int mask, struct statx *buf);
@@ -23,10 +24,10 @@
  */
 
 static __attribute__((unused))
-int sys_statx(int fd, const char *path, int flags, unsigned int mask, struct statx *buf)
+int _sys_statx(int fd, const char *path, int flags, unsigned int mask, struct statx *buf)
 {
 #ifdef __NR_statx
-	return my_syscall5(__NR_statx, fd, path, flags, mask, buf);
+	return __nolibc_syscall5(__NR_statx, fd, path, flags, mask, buf);
 #else
 	return __nolibc_enosys(__func__, fd, path, flags, mask, buf);
 #endif
@@ -35,7 +36,7 @@ int sys_statx(int fd, const char *path, int flags, unsigned int mask, struct sta
 static __attribute__((unused))
 int statx(int fd, const char *path, int flags, unsigned int mask, struct statx *buf)
 {
-	return __sysret(sys_statx(fd, path, flags, mask, buf));
+	return __sysret(_sys_statx(fd, path, flags, mask, buf));
 }
 
 
@@ -45,21 +46,17 @@ int fstatat(int fd, const char *path, struct stat *buf, int flag)
 	struct statx statx;
 	long ret;
 
-	ret = __sysret(sys_statx(fd, path, flag | AT_NO_AUTOMOUNT, STATX_BASIC_STATS, &statx));
+	ret = __sysret(_sys_statx(fd, path, flag | AT_NO_AUTOMOUNT, STATX_BASIC_STATS, &statx));
 	if (ret == -1)
 		return ret;
 
-	buf->st_dev          = ((statx.stx_dev_minor & 0xff)
-			       | (statx.stx_dev_major << 8)
-			       | ((statx.stx_dev_minor & ~0xff) << 12));
+	buf->st_dev          = makedev(statx.stx_dev_major, statx.stx_dev_minor);
 	buf->st_ino          = statx.stx_ino;
 	buf->st_mode         = statx.stx_mode;
 	buf->st_nlink        = statx.stx_nlink;
 	buf->st_uid          = statx.stx_uid;
 	buf->st_gid          = statx.stx_gid;
-	buf->st_rdev         = ((statx.stx_rdev_minor & 0xff)
-			       | (statx.stx_rdev_major << 8)
-			       | ((statx.stx_rdev_minor & ~0xff) << 12));
+	buf->st_rdev         = makedev(statx.stx_rdev_major, statx.stx_rdev_minor);
 	buf->st_size         = statx.stx_size;
 	buf->st_blksize      = statx.stx_blksize;
 	buf->st_blocks       = statx.stx_blocks;
diff --git a/tools/include/nolibc/sys/syscall.h b/tools/include/nolibc/sys/syscall.h
index 4bf97f1..7f06314 100644
--- a/tools/include/nolibc/sys/syscall.h
+++ b/tools/include/nolibc/sys/syscall.h
@@ -10,10 +10,11 @@
 #ifndef _NOLIBC_SYS_SYSCALL_H
 #define _NOLIBC_SYS_SYSCALL_H
 
-#define __syscall_narg(_0, _1, _2, _3, _4, _5, _6, N, ...) N
-#define _syscall_narg(...) __syscall_narg(__VA_ARGS__, 6, 5, 4, 3, 2, 1, 0)
-#define _syscall(N, ...) __sysret(my_syscall##N(__VA_ARGS__))
-#define _syscall_n(N, ...) _syscall(N, __VA_ARGS__)
-#define syscall(...) _syscall_n(_syscall_narg(__VA_ARGS__), ##__VA_ARGS__)
+#define ___nolibc_syscall_narg(_0, _1, _2, _3, _4, _5, _6, N, ...) N
+#define __nolibc_syscall_narg(...) ___nolibc_syscall_narg(__VA_ARGS__, 6, 5, 4, 3, 2, 1, 0)
+#define __nolibc_syscall(N, ...) __nolibc_syscall##N(__VA_ARGS__)
+#define __nolibc_syscall_n(N, ...) __nolibc_syscall(N, __VA_ARGS__)
+#define _syscall(...) __nolibc_syscall_n(__nolibc_syscall_narg(__VA_ARGS__), ##__VA_ARGS__)
+#define syscall(...) __sysret(_syscall(__VA_ARGS__))
 
 #endif /* _NOLIBC_SYS_SYSCALL_H */
diff --git a/tools/include/nolibc/sys/sysmacros.h b/tools/include/nolibc/sys/sysmacros.h
index 37c33f0..ff9dd85 100644
--- a/tools/include/nolibc/sys/sysmacros.h
+++ b/tools/include/nolibc/sys/sysmacros.h
@@ -12,9 +12,26 @@
 
 #include "../std.h"
 
-/* WARNING, it only deals with the 4096 first majors and 256 first minors */
-#define makedev(major, minor) ((dev_t)((((major) & 0xfff) << 8) | ((minor) & 0xff)))
-#define major(dev) ((unsigned int)(((dev) >> 8) & 0xfff))
-#define minor(dev) ((unsigned int)((dev) & 0xff))
+static __inline__ dev_t __nolibc_makedev(unsigned int maj, unsigned int min)
+{
+	return (((dev_t)maj & ~0xfff) << 32) | ((maj & 0xfff) << 8) |
+	       (((dev_t)min & ~0xff) << 12) | (min & 0xff);
+}
+
+#define makedev(maj, min) __nolibc_makedev(maj, min)
+
+static __inline__ unsigned int __nolibc_major(dev_t dev)
+{
+	return ((dev >> 32) & ~0xfff) | ((dev >> 8) & 0xfff);
+}
+
+#define major(dev) __nolibc_major(dev)
+
+static __inline__ unsigned int __nolibc_minor(dev_t dev)
+{
+	return ((dev >> 12) & ~0xff) | (dev & 0xff);
+}
+
+#define minor(dev) __nolibc_minor(dev)
 
 #endif /* _NOLIBC_SYS_SYSMACROS_H */
diff --git a/tools/include/nolibc/sys/time.h b/tools/include/nolibc/sys/time.h
index afdb7e3..c3bb47f 100644
--- a/tools/include/nolibc/sys/time.h
+++ b/tools/include/nolibc/sys/time.h
@@ -13,21 +13,21 @@
 #include "../arch.h"
 #include "../sys.h"
 
-static int sys_clock_gettime(clockid_t clockid, struct timespec *tp);
+static int _sys_clock_gettime(clockid_t clockid, struct timespec *tp);
 
 /*
  * int gettimeofday(struct timeval *tv, struct timezone *tz);
  */
 
 static __attribute__((unused))
-int sys_gettimeofday(struct timeval *tv, struct timezone *tz)
+int _sys_gettimeofday(struct timeval *tv, struct timezone *tz)
 {
 	(void) tz; /* Non-NULL tz is undefined behaviour */
 
 	struct timespec tp;
 	int ret;
 
-	ret = sys_clock_gettime(CLOCK_REALTIME, &tp);
+	ret = _sys_clock_gettime(CLOCK_REALTIME, &tp);
 	if (!ret && tv) {
 		tv->tv_sec = tp.tv_sec;
 		tv->tv_usec = (uint32_t)tp.tv_nsec / 1000;
@@ -39,7 +39,7 @@ int sys_gettimeofday(struct timeval *tv, struct timezone *tz)
 static __attribute__((unused))
 int gettimeofday(struct timeval *tv, struct timezone *tz)
 {
-	return __sysret(sys_gettimeofday(tv, tz));
+	return __sysret(_sys_gettimeofday(tv, tz));
 }
 
 #endif /* _NOLIBC_SYS_TIME_H */
diff --git a/tools/include/nolibc/sys/timerfd.h b/tools/include/nolibc/sys/timerfd.h
index 29fd92b..cb7fef3 100644
--- a/tools/include/nolibc/sys/timerfd.h
+++ b/tools/include/nolibc/sys/timerfd.h
@@ -17,47 +17,47 @@
 
 
 static __attribute__((unused))
-int sys_timerfd_create(int clockid, int flags)
+int _sys_timerfd_create(int clockid, int flags)
 {
-	return my_syscall2(__NR_timerfd_create, clockid, flags);
+	return __nolibc_syscall2(__NR_timerfd_create, clockid, flags);
 }
 
 static __attribute__((unused))
 int timerfd_create(int clockid, int flags)
 {
-	return __sysret(sys_timerfd_create(clockid, flags));
+	return __sysret(_sys_timerfd_create(clockid, flags));
 }
 
 
 static __attribute__((unused))
-int sys_timerfd_gettime(int fd, struct itimerspec *curr_value)
+int _sys_timerfd_gettime(int fd, struct itimerspec *curr_value)
 {
 #if defined(__NR_timerfd_gettime64)
 	__nolibc_assert_time64_type(curr_value->it_value.tv_sec);
-	return my_syscall2(__NR_timerfd_gettime64, fd, curr_value);
+	return __nolibc_syscall2(__NR_timerfd_gettime64, fd, curr_value);
 #else
 	__nolibc_assert_native_time64();
-	return my_syscall2(__NR_timerfd_gettime, fd, curr_value);
+	return __nolibc_syscall2(__NR_timerfd_gettime, fd, curr_value);
 #endif
 }
 
 static __attribute__((unused))
 int timerfd_gettime(int fd, struct itimerspec *curr_value)
 {
-	return __sysret(sys_timerfd_gettime(fd, curr_value));
+	return __sysret(_sys_timerfd_gettime(fd, curr_value));
 }
 
 
 static __attribute__((unused))
-int sys_timerfd_settime(int fd, int flags,
-			const struct itimerspec *new_value, struct itimerspec *old_value)
+int _sys_timerfd_settime(int fd, int flags,
+			 const struct itimerspec *new_value, struct itimerspec *old_value)
 {
 #if defined(__NR_timerfd_settime64)
 	__nolibc_assert_time64_type(new_value->it_value.tv_sec);
-	return my_syscall4(__NR_timerfd_settime64, fd, flags, new_value, old_value);
+	return __nolibc_syscall4(__NR_timerfd_settime64, fd, flags, new_value, old_value);
 #else
 	__nolibc_assert_native_time64();
-	return my_syscall4(__NR_timerfd_settime, fd, flags, new_value, old_value);
+	return __nolibc_syscall4(__NR_timerfd_settime, fd, flags, new_value, old_value);
 #endif
 }
 
@@ -65,7 +65,7 @@ static __attribute__((unused))
 int timerfd_settime(int fd, int flags,
 		    const struct itimerspec *new_value, struct itimerspec *old_value)
 {
-	return __sysret(sys_timerfd_settime(fd, flags, new_value, old_value));
+	return __sysret(_sys_timerfd_settime(fd, flags, new_value, old_value));
 }
 
 #endif /* _NOLIBC_SYS_TIMERFD_H */
diff --git a/tools/include/nolibc/sys/uio.h b/tools/include/nolibc/sys/uio.h
index 7ad42b9..06bf17d 100644
--- a/tools/include/nolibc/sys/uio.h
+++ b/tools/include/nolibc/sys/uio.h
@@ -19,30 +19,30 @@
  * ssize_t readv(int fd, const struct iovec *iovec, int count);
  */
 static __attribute__((unused))
-ssize_t sys_readv(int fd, const struct iovec *iovec, int count)
+ssize_t _sys_readv(int fd, const struct iovec *iovec, int count)
 {
-	return my_syscall3(__NR_readv, fd, iovec, count);
+	return __nolibc_syscall3(__NR_readv, fd, iovec, count);
 }
 
 static __attribute__((unused))
 ssize_t readv(int fd, const struct iovec *iovec, int count)
 {
-	return __sysret(sys_readv(fd, iovec, count));
+	return __sysret(_sys_readv(fd, iovec, count));
 }
 
 /*
  * ssize_t writev(int fd, const struct iovec *iovec, int count);
  */
 static __attribute__((unused))
-ssize_t sys_writev(int fd, const struct iovec *iovec, int count)
+ssize_t _sys_writev(int fd, const struct iovec *iovec, int count)
 {
-	return my_syscall3(__NR_writev, fd, iovec, count);
+	return __nolibc_syscall3(__NR_writev, fd, iovec, count);
 }
 
 static __attribute__((unused))
 ssize_t writev(int fd, const struct iovec *iovec, int count)
 {
-	return __sysret(sys_writev(fd, iovec, count));
+	return __sysret(_sys_writev(fd, iovec, count));
 }
 
 
diff --git a/tools/include/nolibc/sys/utsname.h b/tools/include/nolibc/sys/utsname.h
index 01023e1..2aaf873 100644
--- a/tools/include/nolibc/sys/utsname.h
+++ b/tools/include/nolibc/sys/utsname.h
@@ -28,15 +28,15 @@ struct utsname {
 };
 
 static __attribute__((unused))
-int sys_uname(struct utsname *buf)
+int _sys_uname(struct utsname *buf)
 {
-	return my_syscall1(__NR_uname, buf);
+	return __nolibc_syscall1(__NR_uname, buf);
 }
 
 static __attribute__((unused))
 int uname(struct utsname *buf)
 {
-	return __sysret(sys_uname(buf));
+	return __sysret(_sys_uname(buf));
 }
 
 #endif /* _NOLIBC_SYS_UTSNAME_H */
diff --git a/tools/include/nolibc/sys/wait.h b/tools/include/nolibc/sys/wait.h
index 9d9319b..7a1feb2 100644
--- a/tools/include/nolibc/sys/wait.h
+++ b/tools/include/nolibc/sys/wait.h
@@ -21,15 +21,15 @@
  */
 
 static __attribute__((unused))
-int sys_waitid(int which, pid_t pid, siginfo_t *infop, int options, struct rusage *rusage)
+int _sys_waitid(int which, pid_t pid, siginfo_t *infop, int options, struct rusage *rusage)
 {
-	return my_syscall5(__NR_waitid, which, pid, infop, options, rusage);
+	return __nolibc_syscall5(__NR_waitid, which, pid, infop, options, rusage);
 }
 
 static __attribute__((unused))
 int waitid(int which, pid_t pid, siginfo_t *infop, int options)
 {
-	return __sysret(sys_waitid(which, pid, infop, options, NULL));
+	return __sysret(_sys_waitid(which, pid, infop, options, NULL));
 }
 
 
diff --git a/tools/include/nolibc/time.h b/tools/include/nolibc/time.h
index f9257d6..9ba9307 100644
--- a/tools/include/nolibc/time.h
+++ b/tools/include/nolibc/time.h
@@ -33,69 +33,69 @@
  */
 
 static __attribute__((unused))
-int sys_clock_getres(clockid_t clockid, struct timespec *res)
+int _sys_clock_getres(clockid_t clockid, struct timespec *res)
 {
 #if defined(__NR_clock_getres_time64)
 	__nolibc_assert_time64_type(res->tv_sec);
-	return my_syscall2(__NR_clock_getres_time64, clockid, res);
+	return __nolibc_syscall2(__NR_clock_getres_time64, clockid, res);
 #else
 	__nolibc_assert_native_time64();
-	return my_syscall2(__NR_clock_getres, clockid, res);
+	return __nolibc_syscall2(__NR_clock_getres, clockid, res);
 #endif
 }
 
 static __attribute__((unused))
 int clock_getres(clockid_t clockid, struct timespec *res)
 {
-	return __sysret(sys_clock_getres(clockid, res));
+	return __sysret(_sys_clock_getres(clockid, res));
 }
 
 static __attribute__((unused))
-int sys_clock_gettime(clockid_t clockid, struct timespec *tp)
+int _sys_clock_gettime(clockid_t clockid, struct timespec *tp)
 {
 #if defined(__NR_clock_gettime64)
 	__nolibc_assert_time64_type(tp->tv_sec);
-	return my_syscall2(__NR_clock_gettime64, clockid, tp);
+	return __nolibc_syscall2(__NR_clock_gettime64, clockid, tp);
 #else
 	__nolibc_assert_native_time64();
-	return my_syscall2(__NR_clock_gettime, clockid, tp);
+	return __nolibc_syscall2(__NR_clock_gettime, clockid, tp);
 #endif
 }
 
 static __attribute__((unused))
 int clock_gettime(clockid_t clockid, struct timespec *tp)
 {
-	return __sysret(sys_clock_gettime(clockid, tp));
+	return __sysret(_sys_clock_gettime(clockid, tp));
 }
 
 static __attribute__((unused))
-int sys_clock_settime(clockid_t clockid, struct timespec *tp)
+int _sys_clock_settime(clockid_t clockid, struct timespec *tp)
 {
 #if defined(__NR_clock_settime64)
 	__nolibc_assert_time64_type(tp->tv_sec);
-	return my_syscall2(__NR_clock_settime64, clockid, tp);
+	return __nolibc_syscall2(__NR_clock_settime64, clockid, tp);
 #else
 	__nolibc_assert_native_time64();
-	return my_syscall2(__NR_clock_settime, clockid, tp);
+	return __nolibc_syscall2(__NR_clock_settime, clockid, tp);
 #endif
 }
 
 static __attribute__((unused))
 int clock_settime(clockid_t clockid, struct timespec *tp)
 {
-	return __sysret(sys_clock_settime(clockid, tp));
+	return __sysret(_sys_clock_settime(clockid, tp));
 }
 
 static __attribute__((unused))
-int sys_clock_nanosleep(clockid_t clockid, int flags, const struct timespec *rqtp,
-			struct timespec *rmtp)
+int _sys_clock_nanosleep(clockid_t clockid, int flags, const struct timespec *rqtp,
+			 struct timespec *rmtp)
 {
 #if defined(__NR_clock_nanosleep_time64)
 	__nolibc_assert_time64_type(rqtp->tv_sec);
-	return my_syscall4(__NR_clock_nanosleep_time64, clockid, flags, rqtp, rmtp);
+	return __nolibc_syscall4(__NR_clock_nanosleep_time64, clockid, flags, rqtp, rmtp);
 #else
 	__nolibc_assert_native_time64();
-	return my_syscall4(__NR_clock_nanosleep, clockid, flags, rqtp, rmtp);
+	return __nolibc_syscall4(__NR_clock_nanosleep, clockid, flags, rqtp, rmtp);
 #endif
 }
 
@@ -104,7 +104,7 @@ int clock_nanosleep(clockid_t clockid, int flags, const struct timespec *rqtp,
 		    struct timespec *rmtp)
 {
 	/* Directly return a positive error number */
-	return -sys_clock_nanosleep(clockid, flags, rqtp, rmtp);
+	return -_sys_clock_nanosleep(clockid, flags, rqtp, rmtp);
 }
 
 static __inline__
@@ -116,7 +116,7 @@ double difftime(time_t time1, time_t time2)
 static __inline__
 int nanosleep(const struct timespec *rqtp, struct timespec *rmtp)
 {
-	return __sysret(sys_clock_nanosleep(CLOCK_REALTIME, 0, rqtp, rmtp));
+	return __sysret(_sys_clock_nanosleep(CLOCK_REALTIME, 0, rqtp, rmtp));
 }
 
 
@@ -126,7 +126,7 @@ time_t time(time_t *tptr)
 	struct timeval tv;
 
 	/* note, cannot fail here */
-	sys_gettimeofday(&tv, NULL);
+	_sys_gettimeofday(&tv, NULL);
 
 	if (tptr)
 		*tptr = tv.tv_sec;
@@ -141,57 +141,57 @@ time_t time(time_t *tptr)
  */
 
 static __attribute__((unused))
-int sys_timer_create(clockid_t clockid, struct sigevent *evp, timer_t *timerid)
+int _sys_timer_create(clockid_t clockid, struct sigevent *evp, timer_t *timerid)
 {
-	return my_syscall3(__NR_timer_create, clockid, evp, timerid);
+	return __nolibc_syscall3(__NR_timer_create, clockid, evp, timerid);
 }
 
 static __attribute__((unused))
 int timer_create(clockid_t clockid, struct sigevent *evp, timer_t *timerid)
 {
-	return __sysret(sys_timer_create(clockid, evp, timerid));
+	return __sysret(_sys_timer_create(clockid, evp, timerid));
 }
 
 static __attribute__((unused))
-int sys_timer_delete(timer_t timerid)
+int _sys_timer_delete(timer_t timerid)
 {
-	return my_syscall1(__NR_timer_delete, timerid);
+	return __nolibc_syscall1(__NR_timer_delete, timerid);
 }
 
 static __attribute__((unused))
 int timer_delete(timer_t timerid)
 {
-	return __sysret(sys_timer_delete(timerid));
+	return __sysret(_sys_timer_delete(timerid));
 }
 
 static __attribute__((unused))
-int sys_timer_gettime(timer_t timerid, struct itimerspec *curr_value)
+int _sys_timer_gettime(timer_t timerid, struct itimerspec *curr_value)
 {
 #if defined(__NR_timer_gettime64)
 	__nolibc_assert_time64_type(curr_value->it_value.tv_sec);
-	return my_syscall2(__NR_timer_gettime64, timerid, curr_value);
+	return __nolibc_syscall2(__NR_timer_gettime64, timerid, curr_value);
 #else
 	__nolibc_assert_native_time64();
-	return my_syscall2(__NR_timer_gettime, timerid, curr_value);
+	return __nolibc_syscall2(__NR_timer_gettime, timerid, curr_value);
 #endif
 }
 
 static __attribute__((unused))
 int timer_gettime(timer_t timerid, struct itimerspec *curr_value)
 {
-	return __sysret(sys_timer_gettime(timerid, curr_value));
+	return __sysret(_sys_timer_gettime(timerid, curr_value));
 }
 
 static __attribute__((unused))
-int sys_timer_settime(timer_t timerid, int flags,
-		      const struct itimerspec *new_value, struct itimerspec *old_value)
+int _sys_timer_settime(timer_t timerid, int flags,
+		       const struct itimerspec *new_value, struct itimerspec *old_value)
 {
 #if defined(__NR_timer_settime64)
 	__nolibc_assert_time64_type(new_value->it_value.tv_sec);
-	return my_syscall4(__NR_timer_settime64, timerid, flags, new_value, old_value);
+	return __nolibc_syscall4(__NR_timer_settime64, timerid, flags, new_value, old_value);
 #else
 	__nolibc_assert_native_time64();
-	return my_syscall4(__NR_timer_settime, timerid, flags, new_value, old_value);
+	return __nolibc_syscall4(__NR_timer_settime, timerid, flags, new_value, old_value);
 #endif
 }
 
@@ -199,7 +199,7 @@ static __attribute__((unused))
 int timer_settime(timer_t timerid, int flags,
 		  const struct itimerspec *new_value, struct itimerspec *old_value)
 {
-	return __sysret(sys_timer_settime(timerid, flags, new_value, old_value));
+	return __sysret(_sys_timer_settime(timerid, flags, new_value, old_value));
 }
 
 #endif /* _NOLIBC_TIME_H */
diff --git a/tools/include/nolibc/unistd.h b/tools/include/nolibc/unistd.h
index bb5e80f..5882a68 100644
--- a/tools/include/nolibc/unistd.h
+++ b/tools/include/nolibc/unistd.h
@@ -31,15 +31,15 @@
  */
 
 static __attribute__((unused))
-int sys_faccessat(int fd, const char *path, int amode, int flag)
+int _sys_faccessat(int fd, const char *path, int amode, int flag)
 {
-	return my_syscall4(__NR_faccessat, fd, path, amode, flag);
+	return __nolibc_syscall4(__NR_faccessat, fd, path, amode, flag);
 }
 
 static __attribute__((unused))
 int faccessat(int fd, const char *path, int amode, int flag)
 {
-	return __sysret(sys_faccessat(fd, path, amode, flag));
+	return __sysret(_sys_faccessat(fd, path, amode, flag));
 }
 
 static __attribute__((unused))
@@ -54,7 +54,7 @@ int msleep(unsigned int msecs)
 {
 	struct timeval my_timeval = { msecs / 1000, (msecs % 1000) * 1000 };
 
-	if (sys_select(0, NULL, NULL, NULL, &my_timeval) < 0)
+	if (_sys_select(0, NULL, NULL, NULL, &my_timeval) < 0)
 		return (my_timeval.tv_sec * 1000) +
 			(my_timeval.tv_usec / 1000) +
 			!!(my_timeval.tv_usec % 1000);
@@ -67,7 +67,7 @@ unsigned int sleep(unsigned int seconds)
 {
 	struct timeval my_timeval = { seconds, 0 };
 
-	if (sys_select(0, NULL, NULL, NULL, &my_timeval) < 0)
+	if (_sys_select(0, NULL, NULL, NULL, &my_timeval) < 0)
 		return my_timeval.tv_sec + !!my_timeval.tv_usec;
 	else
 		return 0;
@@ -78,7 +78,7 @@ int usleep(unsigned int usecs)
 {
 	struct timeval my_timeval = { usecs / 1000000, usecs % 1000000 };
 
-	return sys_select(0, NULL, NULL, NULL, &my_timeval);
+	return _sys_select(0, NULL, NULL, NULL, &my_timeval);
 }
 
 static __attribute__((unused))
diff --git a/tools/testing/selftests/nolibc/Makefile b/tools/testing/selftests/nolibc/Makefile
index 0370489..6bace04 100644
--- a/tools/testing/selftests/nolibc/Makefile
+++ b/tools/testing/selftests/nolibc/Makefile
@@ -13,9 +13,9 @@ $(OUTPUT)/nolibc-test: CFLAGS = -nostdlib -nostdinc -static \
 	 -isystem $(top_srcdir)/tools/include/nolibc -isystem $(top_srcdir)/usr/include \
 	 $(CFLAGS_NOLIBC_TEST)
 $(OUTPUT)/nolibc-test: LDLIBS = $(if $(LLVM),,-lgcc)
-$(OUTPUT)/nolibc-test: nolibc-test.c nolibc-test-linkage.c | headers
+$(OUTPUT)/nolibc-test: $(NOLIBC_TEST_SOURCES) | headers
 
-$(OUTPUT)/libc-test: nolibc-test.c nolibc-test-linkage.c
+$(OUTPUT)/libc-test: $(NOLIBC_TEST_SOURCES)
 	$(call msg,CC,,$@)
 	$(Q)$(LINK.c) $^ -o $@
 
diff --git a/tools/testing/selftests/nolibc/Makefile.include b/tools/testing/selftests/nolibc/Makefile.include
index 66287fa..96fe2bc 100644
--- a/tools/testing/selftests/nolibc/Makefile.include
+++ b/tools/testing/selftests/nolibc/Makefile.include
@@ -5,6 +5,9 @@ _CFLAGS_STACKPROTECTOR ?= $(call try-run, \
 	echo 'void foo(void) {}' | $(CC) -x c - -o - -S $(CLANG_CROSS_FLAGS) $(__CFLAGS_STACKPROTECTOR) | grep -q __stack_chk_guard, \
 	$(__CFLAGS_STACKPROTECTOR))
 _CFLAGS_SANITIZER ?= $(call cc-option,-fsanitize=undefined -fsanitize-trap=all)
-CFLAGS_NOLIBC_TEST  ?= -Os -fno-ident -fno-asynchronous-unwind-tables -std=c89 -W -Wall -Wextra \
+CFLAGS_NOLIBC_TEST  ?= -Os -fno-ident -fno-asynchronous-unwind-tables -std=c89 \
+		-W -Wall -Wextra -Wundef \
 		$(call cc-option,-fno-stack-protector) $(call cc-option,-Wmissing-prototypes) \
 		$(_CFLAGS_STACKPROTECTOR) $(_CFLAGS_SANITIZER)
+
+NOLIBC_TEST_SOURCES := nolibc-test.c nolibc-test-linkage.c nolibc-test-ignore-errno.c
diff --git a/tools/testing/selftests/nolibc/Makefile.nolibc b/tools/testing/selftests/nolibc/Makefile.nolibc
index f570419..f30bc68 100644
--- a/tools/testing/selftests/nolibc/Makefile.nolibc
+++ b/tools/testing/selftests/nolibc/Makefile.nolibc
@@ -232,23 +232,14 @@ ifeq ($(origin XARCH),command line)
 CFLAGS_XARCH = $(CFLAGS_$(XARCH))
 endif
 
+LDLIBS_ppc = $(if $(LLVM),,-lgcc)
+LDLIBS = $(LDLIBS_$(XARCH))
+
 include Makefile.include
 
 CFLAGS  ?= $(CFLAGS_NOLIBC_TEST) $(CFLAGS_XARCH) $(CFLAGS_EXTRA)
 LDFLAGS :=
 
-LIBGCC := -lgcc
-
-ifeq ($(ARCH),x86)
-# Not needed on x86, probably not present for x32
-LIBGCC :=
-endif
-
-ifneq ($(LLVM),)
-# Not needed for clang
-LIBGCC :=
-endif
-
 # Modify CFLAGS based on LLVM=
 include $(srctree)/tools/scripts/Makefile.include
 
@@ -302,12 +293,12 @@ sysroot/$(ARCH)/include:
 	$(Q)$(MAKE) -C $(srctree)/tools/include/nolibc ARCH=$(ARCH) OUTPUT=$(CURDIR)/sysroot/ headers_standalone headers_check
 	$(Q)mv sysroot/sysroot sysroot/$(ARCH)
 
-nolibc-test: nolibc-test.c nolibc-test-linkage.c sysroot/$(ARCH)/include
+nolibc-test: $(NOLIBC_TEST_SOURCES) sysroot/$(ARCH)/include
 	$(QUIET_CC)$(CC) $(CFLAGS) $(LDFLAGS) -o $@ \
-	  -nostdlib -nostdinc -static -Isysroot/$(ARCH)/include nolibc-test.c nolibc-test-linkage.c $(LIBGCC)
+	  -nostdlib -nostdinc -static -Isysroot/$(ARCH)/include $(NOLIBC_TEST_SOURCES) $(LDLIBS)
 
-libc-test: nolibc-test.c nolibc-test-linkage.c
-	$(QUIET_CC)$(HOSTCC) -o $@ nolibc-test.c nolibc-test-linkage.c
+libc-test: $(NOLIBC_TEST_SOURCES)
+	$(QUIET_CC)$(HOSTCC) -o $@ $(NOLIBC_TEST_SOURCES)
 
 # local libc-test
 run-libc-test: libc-test
diff --git a/tools/testing/selftests/nolibc/nolibc-test-ignore-errno.c b/tools/testing/selftests/nolibc/nolibc-test-ignore-errno.c
new file mode 100644
index 0000000..aea816d
--- /dev/null
+++ b/tools/testing/selftests/nolibc/nolibc-test-ignore-errno.c
@@ -0,0 +1,6 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+
+#define NOLIBC_IGNORE_ERRNO
+
+/* Include all of nolibc and make sure everything compiles */
+#include <stdlib.h>
diff --git a/tools/testing/selftests/nolibc/nolibc-test.c b/tools/testing/selftests/nolibc/nolibc-test.c
index 1b9d3b2..d3c4fac 100644
--- a/tools/testing/selftests/nolibc/nolibc-test.c
+++ b/tools/testing/selftests/nolibc/nolibc-test.c
@@ -42,6 +42,9 @@
 #include <unistd.h>
 #include <limits.h>
 #include <ctype.h>
+#include <stdbool.h>
+#include <byteswap.h>
+#include <endian.h>
 
 #pragma GCC diagnostic ignored "-Wmissing-prototypes"
 
@@ -66,6 +69,8 @@ static const char *argv0;
 /* will be used by constructor tests */
 static int constructor_test_value;
 
+static const int is_le = __BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__;
+
 static const int is_nolibc =
 #ifdef NOLIBC
 	1
@@ -74,6 +79,30 @@ static const int is_nolibc =
 #endif
 ;
 
+static const int is_glibc =
+#ifdef __GLIBC__
+	1
+#else
+	0
+#endif
+;
+
+#if !defined(NOLIBC)
+/* Some disabled tests may not compile. */
+
+/* strlcat() and strlcpy() may not be in the system headers. */
+#undef strlcat
+#undef strlcpy
+#define strlcat(d, s, l) 0
+#define strlcpy(d, s, l) 0
+
+/* readdir_r() is likely to be marked deprecated */
+#undef readdir_r
+#define readdir_r(dir, dirent, result) ((errno = EINVAL), -1)
+
+#define _syscall(...) 0
+#endif
+
 /* definition of a series of tests */
 struct test {
 	const char *name;              /* test name */
@@ -142,21 +171,6 @@ static const char *errorname(int err)
 	}
 }
 
-static void align_result(size_t llen)
-{
-	const size_t align = 64;
-	char buf[align];
-	size_t n;
-
-	if (llen >= align)
-		return;
-
-	n = align - llen;
-	memset(buf, ' ', n);
-	buf[n] = '\0';
-	fputs(buf, stdout);
-}
-
 enum RESULT {
 	OK,
 	FAIL,
@@ -174,8 +188,10 @@ static void result(int llen, enum RESULT r)
 	else
 		msg = " [FAIL]";
 
-	align_result(llen);
-	puts(msg);
+	llen = 64 - llen;
+	if (llen < 0)
+		llen = 0;
+	printf("%*s%s\n", llen, "", msg);
 }
 
 /* The tests below are intended to be used by the macroes, which evaluate
@@ -304,10 +320,7 @@ int expect_syszr(int expr, int llen)
 {
 	int ret = 0;
 
-	if (errno == ENOSYS) {
-		llen += printf(" = ENOSYS");
-		result(llen, SKIPPED);
-	} else if (expr) {
+	if (expr) {
 		ret = 1;
 		llen += printf(" = %d %s ", expr, errorname(errno));
 		result(llen, FAIL);
@@ -347,10 +360,7 @@ int expect_sysne(int expr, int llen, int val)
 {
 	int ret = 0;
 
-	if (errno == ENOSYS) {
-		llen += printf(" = ENOSYS");
-		result(llen, SKIPPED);
-	} else if (expr == val) {
+	if (expr == val) {
 		ret = 1;
 		llen += printf(" = %d %s ", expr, errorname(errno));
 		result(llen, FAIL);
@@ -375,9 +385,7 @@ int expect_syserr2(int expr, int expret, int experr1, int experr2, int llen)
 	int _errno = errno;
 
 	llen += printf(" = %d %s ", expr, errorname(_errno));
-	if (errno == ENOSYS) {
-		result(llen, SKIPPED);
-	} else if (expr != expret || (_errno != experr1 && _errno != experr2)) {
+	if (expr != expret || (_errno != experr1 && _errno != experr2)) {
 		ret = 1;
 		if (experr2 == 0)
 			llen += printf(" != (%d %s) ", expret, errorname(experr1));
@@ -701,6 +709,37 @@ static void constructor2(int argc, char **argv, char **envp)
 		constructor_test_value |= 1 << 1;
 }
 
+int test_program_invocation_name(void)
+{
+	char buf[100];
+	char *dirsep;
+	ssize_t r;
+	int fd;
+
+	fd = open("/proc/self/cmdline", O_RDONLY);
+	if (fd == -1)
+		return 1;
+
+	r = read(fd, buf, sizeof(buf));
+	close(fd);
+	if (r < 1 || r == sizeof(buf))
+		return 1;
+
+	buf[r - 1] = '\0';
+
+	if (strcmp(program_invocation_name, buf) != 0)
+		return 1;
+
+	dirsep = strrchr(buf, '/');
+	if (!dirsep || dirsep[1] == '\0')
+		return 1;
+
+	if (strcmp(program_invocation_short_name, dirsep + 1) != 0)
+		return 1;
+
+	return 0;
+}
+
 int run_startup(int min, int max)
 {
 	int test;
@@ -715,6 +754,7 @@ int run_startup(int min, int max)
 #ifdef NOLIBC
 	test_auxv = _auxv;
 #endif
+	bool proc = access("/proc", R_OK) == 0;
 
 	for (test = min; test >= 0 && test <= max; test++) {
 		int llen = 0; /* line length */
@@ -740,6 +780,7 @@ int run_startup(int min, int max)
 		CASE_TEST(constructor);      EXPECT_EQ(is_nolibc, constructor_test_value, 0x3); break;
 		CASE_TEST(linkage_errno);    EXPECT_PTREQ(1, linkage_test_errno_addr(), &errno); break;
 		CASE_TEST(linkage_constr);   EXPECT_EQ(1, linkage_test_constructor_test_value, 0x3); break;
+		CASE_TEST(prog_name);        EXPECT_ZR(proc, test_program_invocation_name()); break;
 		case __LINE__:
 			return ret; /* must be last */
 		/* note: do not set any defaults so as to permit holes above */
@@ -866,7 +907,7 @@ int test_file_stream(void)
 
 	errno = 0;
 	r = fwrite("foo", 1, 3, f);
-	if (r != 0 || errno != EBADF) {
+	if (r != 0 || ((is_nolibc || is_glibc) && errno != EBADF)) {
 		fclose(f);
 		return -1;
 	}
@@ -1408,7 +1449,7 @@ int run_syscall(int min, int max)
 		CASE_TEST(fork);              EXPECT_SYSZR(1, test_fork(FORK_STANDARD)); break;
 		CASE_TEST(getdents64_root);   EXPECT_SYSNE(1, test_getdents64("/"), -1); break;
 		CASE_TEST(getdents64_null);   EXPECT_SYSER(1, test_getdents64("/dev/null"), -1, ENOTDIR); break;
-		CASE_TEST(directories);       EXPECT_SYSZR(proc, test_dirent()); break;
+		CASE_TEST(directories);       EXPECT_SYSZR(is_nolibc && proc, test_dirent()); break;
 		CASE_TEST(getrandom);         EXPECT_SYSZR(1, test_getrandom()); break;
 		CASE_TEST(gettimeofday_tv);   EXPECT_SYSZR(1, gettimeofday(&tv, NULL)); break;
 		CASE_TEST(gettimeofday_tv_tz);EXPECT_SYSZR(1, gettimeofday(&tv, &tz)); break;
@@ -1442,6 +1483,7 @@ int run_syscall(int min, int max)
 		CASE_TEST(select_fault);      EXPECT_SYSER(1, select(1, (void *)1, NULL, NULL, 0), -1, EFAULT); break;
 		CASE_TEST(stat_blah);         EXPECT_SYSER(1, stat("/proc/self/blah", &stat_buf), -1, ENOENT); break;
 		CASE_TEST(stat_fault);        EXPECT_SYSER(1, stat(NULL, &stat_buf), -1, EFAULT); break;
+		CASE_TEST(stat_rdev);         EXPECT_SYSZR(1, ({ int ret = stat("/dev/null", &stat_buf); ret ?: stat_buf.st_rdev != makedev(1, 3); })); break;
 		CASE_TEST(stat_timestamps);   EXPECT_SYSZR(1, test_stat_timestamps()); break;
 		CASE_TEST(symlink_root);      EXPECT_SYSER(1, symlink("/", "/"), -1, EEXIST); break;
 		CASE_TEST(timer);             EXPECT_SYSZR(1, test_timer()); break;
@@ -1460,9 +1502,11 @@ int run_syscall(int min, int max)
 		CASE_TEST(readv_zero);        EXPECT_SYSZR(1, readv(0, NULL, 0)); break;
 		CASE_TEST(writev_badf);       EXPECT_SYSER(1, writev(-1, &iov_one, 1), -1, EBADF); break;
 		CASE_TEST(writev_zero);       EXPECT_SYSZR(1, writev(1, NULL, 0)); break;
-		CASE_TEST(ptrace);            EXPECT_SYSER(1, ptrace(PTRACE_CONT, getpid(), NULL, NULL), -1, ESRCH); break;
+		CASE_TEST(ptrace);            tmp = ptrace(PTRACE_CONT, getpid(), NULL, NULL); EXPECT_SYSER(tmp != -1 && errno != ENOSYS, tmp, -1, EFAULT); break;
 		CASE_TEST(syscall_noargs);    EXPECT_SYSEQ(1, syscall(__NR_getpid), getpid()); break;
 		CASE_TEST(syscall_args);      EXPECT_SYSER(1, syscall(__NR_statx, 0, NULL, 0, 0, NULL), -1, EFAULT); break;
+		CASE_TEST(_syscall_noargs);   EXPECT_SYSEQ(is_nolibc, _syscall(__NR_getpid), getpid()); break;
+		CASE_TEST(_syscall_args);     EXPECT_SYSEQ(is_nolibc, _syscall(__NR_statx, 0, NULL, 0, 0, NULL), -EFAULT); break;
 		CASE_TEST(namespace);         EXPECT_SYSZR(euid0 && proc, test_namespace()); break;
 		case __LINE__:
 			return ret; /* must be last */
@@ -1511,6 +1555,60 @@ int test_time_types(void)
 	return 0;
 }
 
+int test_malloc(void)
+{
+	size_t sz_array1, sz_array2, sz_array3;
+	int *array1, *array2, *array3;
+	int pagesize = getpagesize();
+	size_t idx;
+
+	if (pagesize < 0)
+		return 1;
+
+	/* Dependent on the page size, as that is the granularity of our allocator. */
+	sz_array1 = pagesize / 2;
+	array1 = malloc(sz_array1 * sizeof(*array1));
+	if (!array1)
+		return 2;
+
+	for (idx = 0; idx < sz_array1; idx++)
+		array1[idx] = idx;
+
+	sz_array2 = pagesize * 2;
+	array2 = calloc(sz_array2, sizeof(*array2));
+	if (!array2) {
+		free(array1);
+		return 3;
+	}
+
+	for (idx = 0; idx < sz_array2; idx++) {
+		if (array2[idx] != 0) {
+			free(array2);
+			return 4;
+		}
+		array2[idx] = idx + sz_array1;
+	}
+
+	/* Resize array1 into array3 and append array2 at the end. */
+	sz_array3 = sz_array1 + sz_array2;
+	array3 = realloc(array1, sz_array3 * sizeof(*array3));
+	if (!array3) {
+		free(array2);
+		free(array1);
+		return 5;
+	}
+	memcpy(array3 + sz_array1, array2, sizeof(*array2) * sz_array2);
+	free(array2);
+
+	/* The contents must be contiguous now. */
+	for (idx = 0; idx < sz_array3; idx++)
+		if (array3[idx] != (int)idx)
+			return 6;
+
+	free(array3);
+	return 0;
+}
+
 int run_stdlib(int min, int max)
 {
 	int test;
@@ -1637,6 +1735,22 @@ int run_stdlib(int min, int max)
 		CASE_TEST(memchr_foobar6_o);        EXPECT_STREQ(1, memchr("foobar", 'o', 6), "oobar"); break;
 		CASE_TEST(memchr_foobar3_b);        EXPECT_STRZR(1, memchr("foobar", 'b', 3)); break;
 		CASE_TEST(time_types);              EXPECT_ZR(is_nolibc, test_time_types()); break;
+		CASE_TEST(makedev);                 EXPECT_EQ(1, makedev(0x12, 0x34), 0x1234); break;
+		CASE_TEST(major);                   EXPECT_EQ(1, major(0x1234), 0x12); break;
+		CASE_TEST(minor);                   EXPECT_EQ(1, minor(0x1234), 0x34); break;
+		CASE_TEST(makedev_big);             EXPECT_EQ(1, makedev(0x11223344, 0x55667788), 0x1122355667734488); break;
+		CASE_TEST(major_big);               EXPECT_EQ(1, major(0x1122355667734488), 0x11223344); break;
+		CASE_TEST(minor_big);               EXPECT_EQ(1, minor(0x1122355667734488), 0x55667788); break;
+		CASE_TEST(malloc);                  EXPECT_ZR(1, test_malloc()); break;
+		CASE_TEST(bswap_16);                EXPECT_EQ(1, bswap_16(0x0123), 0x2301); break;
+		CASE_TEST(bswap_32);                EXPECT_EQ(1, bswap_32(0x01234567), 0x67452301); break;
+		CASE_TEST(bswap_64);                EXPECT_EQ(1, bswap_64(0x0123456789abcdef), 0xefcdab8967452301); break;
+		CASE_TEST(htobe16);                 EXPECT_EQ(1, htobe16(is_le ? 0x0123 : 0x2301), 0x2301); break;
+		CASE_TEST(htole16);                 EXPECT_EQ(1, htole16(is_le ? 0x0123 : 0x2301), 0x0123); break;
+		CASE_TEST(htobe32);                 EXPECT_EQ(1, htobe32(is_le ? 0x01234567 : 0x67452301), 0x67452301); break;
+		CASE_TEST(htole32);                 EXPECT_EQ(1, htole32(is_le ? 0x01234567 : 0x67452301), 0x01234567); break;
+		CASE_TEST(htobe64);                 EXPECT_EQ(1, htobe64(is_le ? 0x0123456789000000 : 0x8967452301), 0x8967452301); break;
+		CASE_TEST(htole64);                 EXPECT_EQ(1, htole64(is_le ? 0x0123456789 : 0x8967452301000000), 0x0123456789); break;
 
 		case __LINE__:
 			return ret; /* must be last */
@@ -1646,33 +1760,60 @@ int run_stdlib(int min, int max)
 	return ret;
 }
 
-#define EXPECT_VFPRINTF(c, expected, fmt, ...)				\
-	ret += expect_vfprintf(llen, c, expected, fmt, ##__VA_ARGS__)
+#define EXPECT_VFPRINTF(cond, expected, fmt, ...)				\
+	do { if (!(cond)) result(llen, SKIPPED); else ret += expect_vfprintf(llen, expected, fmt, ##__VA_ARGS__); } while (0)
 
-static int expect_vfprintf(int llen, int c, const char *expected, const char *fmt, ...)
+#define VFPRINTF_LEN 25
+static int expect_vfprintf(int llen, const char *expected, const char *fmt, ...)
 {
-	char buf[100];
+	char buf[VFPRINTF_LEN + 80];
+	unsigned int cmp_len;
 	va_list args;
-	ssize_t w;
-	int ret;
+	ssize_t written, expected_len;
 
+	/* Fill and terminate buf[] to check for overlong/absent writes */
+	memset(buf, 0xa5, sizeof(buf) - 1);
+	buf[sizeof(buf) - 1] = 0;
 
 	va_start(args, fmt);
-	/* Only allow writing 21 bytes, to test truncation */
-	w = vsnprintf(buf, 21, fmt, args);
+	/* Limit buffer length to test truncation */
+	written = vsnprintf(buf, VFPRINTF_LEN + 1, fmt, args);
 	va_end(args);
 
-	if (w != c) {
-		llen += printf(" written(%d) != %d", (int)w, c);
+	llen += printf(" \"%s\"", buf);
+
+	expected_len = strlen(expected);
+	if (expected_len > VFPRINTF_LEN) {
+		/* Indicate truncated in test output */
+		llen += printf("+");
+		cmp_len = VFPRINTF_LEN;
+	} else {
+		cmp_len = expected_len;
+	}
+
+	if (memcmp(expected, buf, cmp_len) || buf[cmp_len]) {
+		llen += printf(" should be \"%.*s\"", VFPRINTF_LEN, expected);
 		result(llen, FAIL);
 		return 1;
 	}
 
-	llen += printf(" \"%s\" = \"%s\"", expected, buf);
-	ret = strncmp(expected, buf, c);
+	if (written != expected_len) {
+		llen += printf(" written(%d) != %d", (int)written, (int)expected_len);
+		result(llen, FAIL);
+		return 1;
+	}
 
-	result(llen, ret ? FAIL : OK);
-	return ret;
+	/* Check for any overwrites after the actual data. */
+	while (++cmp_len < sizeof(buf) - 1) {
+		if ((unsigned char)buf[cmp_len] != 0xa5) {
+			llen += printf(" overwrote buf[%d] with 0x%x", cmp_len, buf[cmp_len]);
+			result(llen, FAIL);
+			return 1;
+		}
+	}
+
+	result(llen, OK);
+	return 0;
 }
 
 static int test_scanf(void)
@@ -1742,23 +1883,6 @@ static int test_scanf(void)
 	return 0;
 }
 
-int test_strerror(void)
-{
-	char buf[100];
-	ssize_t ret;
-
-	memset(buf, 'A', sizeof(buf));
-
-	errno = EINVAL;
-	ret = snprintf(buf, sizeof(buf), "%m");
-	if (is_nolibc) {
-		if (ret < 6 || memcmp(buf, "errno=", 6))
-			return 1;
-	}
-
-	return 0;
-}
-
 static int test_printf_error(void)
 {
 	int fd, ret, saved_errno;
@@ -1781,6 +1905,29 @@ static int test_printf_error(void)
 	return 0;
 }
 
+int test_asprintf(void)
+{
+	char *str;
+	int ret;
+
+	ret = asprintf(&str, "foo%s", "bar");
+	if (ret == -1)
+		return 1;
+
+	if (ret != 6) {
+		free(str);
+		return 2;
+	}
+
+	if (memcmp(str, "foobar", 6) != 0) {
+		free(str);
+		return 3;
+	}
+
+	free(str);
+	return 0;
+}
+
 static int run_printf(int min, int max)
 {
 	int test;
@@ -1793,24 +1940,58 @@ static int run_printf(int min, int max)
 		 * test numbers.
 		 */
 		switch (test + __LINE__ + 1) {
-		CASE_TEST(empty);        EXPECT_VFPRINTF(0, "", ""); break;
-		CASE_TEST(simple);       EXPECT_VFPRINTF(3, "foo", "foo"); break;
-		CASE_TEST(string);       EXPECT_VFPRINTF(3, "foo", "%s", "foo"); break;
-		CASE_TEST(number);       EXPECT_VFPRINTF(4, "1234", "%d", 1234); break;
-		CASE_TEST(negnumber);    EXPECT_VFPRINTF(5, "-1234", "%d", -1234); break;
-		CASE_TEST(unsigned);     EXPECT_VFPRINTF(5, "12345", "%u", 12345); break;
-		CASE_TEST(char);         EXPECT_VFPRINTF(1, "c", "%c", 'c'); break;
-		CASE_TEST(hex);          EXPECT_VFPRINTF(1, "f", "%x", 0xf); break;
-		CASE_TEST(pointer);      EXPECT_VFPRINTF(3, "0x1", "%p", (void *) 0x1); break;
-		CASE_TEST(uintmax_t);    EXPECT_VFPRINTF(20, "18446744073709551615", "%ju", 0xffffffffffffffffULL); break;
-		CASE_TEST(intmax_t);     EXPECT_VFPRINTF(20, "-9223372036854775807", "%jd", 0x8000000000000001LL); break;
-		CASE_TEST(truncation);   EXPECT_VFPRINTF(25, "01234567890123456789", "%s", "0123456789012345678901234"); break;
-		CASE_TEST(string_width); EXPECT_VFPRINTF(10, "         1", "%10s", "1"); break;
-		CASE_TEST(number_width); EXPECT_VFPRINTF(10, "         1", "%10d", 1); break;
-		CASE_TEST(width_trunc);  EXPECT_VFPRINTF(25, "                    ", "%25d", 1); break;
+		CASE_TEST(empty);        EXPECT_VFPRINTF(1, "", ""); break;
+		CASE_TEST(simple);       EXPECT_VFPRINTF(1, "foo", "foo"); break;
+		CASE_TEST(string);       EXPECT_VFPRINTF(1, "foo", "%s", "foo"); break;
+		CASE_TEST(number);       EXPECT_VFPRINTF(1, "1234", "%d", 1234); break;
+		CASE_TEST(negnumber);    EXPECT_VFPRINTF(1, "-1234", "%d", -1234); break;
+		CASE_TEST(num_sign);     EXPECT_VFPRINTF(1, "| 1|+2|+3|+4|5|", "|% d|%+d|% +d|%+ d|%#d|", 1, 2, 3, 4, 5); break;
+		CASE_TEST(unsigned);     EXPECT_VFPRINTF(1, "12345", "%u", 12345); break;
+		CASE_TEST(signed_max);   EXPECT_VFPRINTF(1, "2147483647", "%i", ~0u >> 1); break;
+		CASE_TEST(signed_min);   EXPECT_VFPRINTF(1, "-2147483648", "%i", (~0u >> 1) + 1); break;
+		CASE_TEST(unsigned_max); EXPECT_VFPRINTF(1, "4294967295", "%u", ~0u); break;
+		CASE_TEST(char);         EXPECT_VFPRINTF(1, "|c|d|   e|", "|%c|%.0c|%4c|", 'c', 'd', 'e'); break;
+		CASE_TEST(octal);        EXPECT_VFPRINTF(1, "|17|  0033||", "|%o|%6.4o|%.0o|", 017, 033, 0); break;
+		CASE_TEST(octal_max);    EXPECT_VFPRINTF(1, "1777777777777777777777", "%llo", ~0ULL); break;
+		CASE_TEST(octal_alt);    EXPECT_VFPRINTF(1, "|0|01|02|034|0|0|", "|%#o|%#o|%#02o|%#02o|%#.0o|%0-#o|", 0, 1, 2, 034, 0, 0); break;
+		CASE_TEST(hex_nolibc);   EXPECT_VFPRINTF(is_nolibc, "|f|d|", "|%x|%X|", 0xf, 0xd); break;
+		CASE_TEST(hex_libc);     EXPECT_VFPRINTF(!is_nolibc, "|f|D|", "|%x|%X|", 0xf, 0xd); break;
+		CASE_TEST(hex_alt);      EXPECT_VFPRINTF(1, "|0x1|  0x2|    0|", "|%#x|%#5x|%#5x|", 1, 2, 0); break;
+		CASE_TEST(hex_alt_prec); EXPECT_VFPRINTF(1, "| 0x02|0x03| 0x123|", "|%#5.2x|%#04x|%#6.2x|", 2, 3, 0x123); break;
+		CASE_TEST(hex_0_alt);    EXPECT_VFPRINTF(1, "|0|0000|   00|", "|%#x|%#04x|%#5.2x|", 0, 0, 0); break;
+		CASE_TEST(pointer);      EXPECT_VFPRINTF(1, "0x1", "%p", (void *) 0x1); break;
+		CASE_TEST(pointer_NULL); EXPECT_VFPRINTF(is_nolibc || is_glibc, "|(nil)|(nil)|", "|%p|%.4p|", (void *)0, (void *)0); break;
+		CASE_TEST(string_NULL);  EXPECT_VFPRINTF(is_nolibc || is_glibc, "|(null)||(null)|", "|%s|%.5s|%.6s|", (void *)0, (void *)0, (void *)0); break;
+		CASE_TEST(percent);      EXPECT_VFPRINTF(1, "a%d42%69%", "a%%d%d%%%d%%", 42, 69); break;
+		CASE_TEST(perc_qual);    EXPECT_VFPRINTF(is_nolibc || is_glibc, "a%d2", "a%-14l%d%d", 2); break;
+		CASE_TEST(invalid);      EXPECT_VFPRINTF(is_nolibc || is_glibc, "a%12yx3%y42%P", "a%12yx%d%y%d%P", 3, 42); break;
+		CASE_TEST(intmax_max);   EXPECT_VFPRINTF(1, "9223372036854775807", "%lld", ~0ULL >> 1); break;
+		CASE_TEST(intmax_min);   EXPECT_VFPRINTF(is_nolibc || is_glibc, "-9223372036854775808", "%Li", (~0ULL >> 1) + 1); break;
+		CASE_TEST(uintmax_max);  EXPECT_VFPRINTF(1, "18446744073709551615", "%ju", ~0ULL); break;
+		CASE_TEST(truncation);   EXPECT_VFPRINTF(1, "012345678901234567890123456789", "%s", "012345678901234567890123456789"); break;
+		CASE_TEST(string_width); EXPECT_VFPRINTF(1, "         1", "%10s", "1"); break;
+		CASE_TEST(string_trunc); EXPECT_VFPRINTF(1, "     12345", "%10.5s", "1234567890"); break;
+		CASE_TEST(string_var);   EXPECT_VFPRINTF(1, "| ab|ef | ij|kl |", "|%*.*s|%*.*s|%*.*s|%*.*s|", 3, 2, "abcd", -3, 2, "efgh", 3, -1, "ij", -3, -1, "kl"); break;
+		CASE_TEST(number_width); EXPECT_VFPRINTF(1, "         1", "%10d", 1); break;
+		CASE_TEST(number_left);  EXPECT_VFPRINTF(1, "|-5      |", "|%-8d|", -5); break;
+		CASE_TEST(string_align); EXPECT_VFPRINTF(1, "|foo     |", "|%-8s|", "foo"); break;
+		CASE_TEST(width_trunc);  EXPECT_VFPRINTF(1, "                             1", "%30d", 1); break;
+		CASE_TEST(width_tr_lft); EXPECT_VFPRINTF(1, "1                             ", "%-30d", 1); break;
+		CASE_TEST(number_pad);   EXPECT_VFPRINTF(1, "0000000005", "%010d", 5); break;
+		CASE_TEST(number_pad);   EXPECT_VFPRINTF(1, "|0000000005|0x1234|", "|%010d|%#01x|", 5, 0x1234); break;
+		CASE_TEST(num_pad_neg);  EXPECT_VFPRINTF(1, "-000000005", "%010d", -5); break;
+		CASE_TEST(num_pad_hex);  EXPECT_VFPRINTF(1, "00fffffffb", "%010x", -5); break;
+		CASE_TEST(num_pad_trunc);EXPECT_VFPRINTF(is_nolibc, "    0000000000000000000000000000005", "%035d", 5); break;  /* max 31 '0' can be added */
+		CASE_TEST(num_p_tr_libc);EXPECT_VFPRINTF(!is_nolibc, "00000000000000000000000000000000005", "%035d", 5); break;
+		CASE_TEST(number_prec);  EXPECT_VFPRINTF(1, "     00005", "%10.5d", 5); break;
+		CASE_TEST(num_prec_neg); EXPECT_VFPRINTF(1, "    -00005", "%10.5d", -5); break;
+		CASE_TEST(number_var);   EXPECT_VFPRINTF(1, "|    -00005|5 |", "|%*.*d|%*.*d|", 10, 5, -5, -2, -10, 5); break;
+		CASE_TEST(num_0_prec_0); EXPECT_VFPRINTF(1, "|| |+||||", "|%.0d|% .0d|%+.0d|%.0u|%.0x|%#.0x|", 0, 0, 0, 0, 0, 0); break;
+		CASE_TEST(errno);        errno = 22; EXPECT_VFPRINTF(is_nolibc, "errno=22", "%m"); break;
+		CASE_TEST(errno-neg);    errno = -22; EXPECT_VFPRINTF(is_nolibc, "errno=-22   ", "%-12m"); break;
 		CASE_TEST(scanf);        EXPECT_ZR(1, test_scanf()); break;
-		CASE_TEST(strerror);     EXPECT_ZR(1, test_strerror()); break;
 		CASE_TEST(printf_error); EXPECT_ZR(1, test_printf_error()); break;
+		CASE_TEST(asprintf);     EXPECT_ZR(1, test_asprintf()); break;
 		case __LINE__:
 			return ret; /* must be last */
 		/* note: do not set any defaults so as to permit holes above */
diff --git a/tools/testing/selftests/nolibc/run-tests.sh b/tools/testing/selftests/nolibc/run-tests.sh
index 3917cfb..cd43909 100755
--- a/tools/testing/selftests/nolibc/run-tests.sh
+++ b/tools/testing/selftests/nolibc/run-tests.sh
@@ -7,7 +7,7 @@ set -e
 
 trap 'echo Aborting...' 'ERR'
 
-crosstool_version=13.2.0
+crosstool_version=15.2.0
 hostarch=x86_64
 nproc=$(( $(nproc) + 2))
 cache_dir="${XDG_CACHE_HOME:-"$HOME"/.cache}"
