Re: [patch] x86: test case for the RODATA config option

[Nick Piggin <nickpiggin@yahoo.com.au>]

On Wednesday 23 January 2008 09:44, Arjan van de Ven wrote:
> From: Arjan van de Ven <arjan@linux.intel.com>
> Subject: x86: test case for the RODATA config option
>
> This patch adds a test module for the DEBUG_RODATA config
> option to make sure change_page_attr() did indeed make
> "const" data read only.
>
> This testcase both tests the DEBUG_RODATA code as well as
> the change_page_attr() code for correct operation.
>
> When the tests/ patch gets merged, this module should move
> to the tests/ directory.
>
> Signed-off-by: Arjan van de Ven <arjan@linux.intel.com>
> ---
>  arch/x86/Kconfig.debug        |    8 +++++
>  arch/x86/kernel/Makefile_32   |    1
>  arch/x86/kernel/Makefile_64   |    2 +
>  arch/x86/kernel/test_rodata.c |   65
> ++++++++++++++++++++++++++++++++++++++++++ arch/x86/mm/init_32.c         | 
>   3 +
>  arch/x86/mm/init_64.c         |    3 +
>  6 files changed, 82 insertions(+)
>
> Index: linux-2.6.24-rc8/arch/x86/Kconfig.debug
> ===================================================================
> --- linux-2.6.24-rc8.orig/arch/x86/Kconfig.debug
> +++ linux-2.6.24-rc8/arch/x86/Kconfig.debug
> @@ -57,6 +57,14 @@ config DEBUG_RODATA
>  	  portion of the kernel code won't be covered by a 2MB TLB anymore.
>  	  If in doubt, say "N".
>
> +config DEBUG_RODATA_TEST
> +	tristate "Testcase for the DEBUG_RODATA feature"
> +	depends on DEBUG_RODATA && m
> +	help
> +	  This option enables a testcase for the DEBUG_RODATA
> +	  feature as well as for the change_page_attr() infrastructure.
> +	  If in doubt, say "N"
> +
>  config 4KSTACKS
>  	bool "Use 4Kb for kernel stacks instead of 8Kb"
>  	depends on DEBUG_KERNEL
> Index: linux-2.6.24-rc8/arch/x86/mm/init_32.c
> ===================================================================
> --- linux-2.6.24-rc8.orig/arch/x86/mm/init_32.c
> +++ linux-2.6.24-rc8/arch/x86/mm/init_32.c
> @@ -790,6 +790,9 @@ static int noinline do_test_wp_bit(void)
>
>  #ifdef CONFIG_DEBUG_RODATA
>
> +const int rodata_test_data;
> +EXPORT_SYMBOL_GPL(rodata_test_data);
> +
>  void mark_rodata_ro(void)
>  {
>  	unsigned long start = PFN_ALIGN(_text);
> Index: linux-2.6.24-rc8/arch/x86/mm/init_64.c
> ===================================================================
> --- linux-2.6.24-rc8.orig/arch/x86/mm/init_64.c
> +++ linux-2.6.24-rc8/arch/x86/mm/init_64.c
> @@ -590,6 +590,9 @@ void free_initmem(void)
>
>  #ifdef CONFIG_DEBUG_RODATA
>
> +const int rodata_test_data = 5;

I guess this should match the 32-bit case, and be zero instead of
5?

Can you disallow building as a module, and put this in the test
code? It could be run from the end of mark_rodata_ro()...