Re: serious performance regression due to NX patch

[Ingo Molnar]

On Mon, 12 Jul 2004, Linus Torvalds wrote:

> > so ... this should be #ifndef ia64?
> 
> No. Make it a CONFIG_DEFAULT_NOEXEC and make the relevant architectures do
> a
> 
> 	define_bool DEFAULT_NOEXEC y
> 
> in their Kconfig files.
> 
> In general, we should _never_ use an architecture-define. They just
> always end up becoming more and more hairy, and less and less obvious
> what they are all about.
> 
> So instead, make a readable and explicit config define, and let each
> architecture just set it (or not) as they wish.

ok - patch below. ia64 can now define it - the default is that legacy
binaries have executability expectations.

	Ingo

--- linux/fs/binfmt_elf.c.orig
+++ linux/fs/binfmt_elf.c
@@ -627,8 +627,14 @@ static int load_elf_binary(struct linux_
 				executable_stack = EXSTACK_DISABLE_X;
 			break;
 		}
+#ifndef CONFIG_DEFAULT_NOEXEC
+	/*
+	 * Legacy binaries (unless the arch defaults to noexec) have an
+	 * expectation of executability - turn it on:
+	 */
 	if (i == elf_ex.e_phnum)
 		def_flags |= VM_EXEC | VM_MAYEXEC;
+#endif
 
 	/* Some simple consistency checks for the interpreter */
 	if (elf_interpreter) {
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/