Re: [PATCH] i386/gcc bug with do_test_wp_bit

[Zachary Amsden]

I agree the code obviously looks ok, I will verify against 2.6.9-rc3 
anyways and send an updated patch that gets a non-tweaked gcc 3.3.3 
build to boot.  Additional fixes are needed for regparm compliance 
between prototypes and functions; I've done these for i386, but may have 
missed some in other architectures.  I'd also like to get rid of the 
apparently harmless lvalue warnings I get from gcc 3.3.3.

I checked the gcc docs but couldn't find any versions where noinline was 
unsupported, although I should have assumed otherwise.

Zach

Linus Torvalds wrote:

> On Tue, 5 Oct 2004, Zachary Amsden wrote:
>  
>
> > I've included a small trivial fix that is harmless for users not using 
> > gcc 3.3.3.  Testing: my 2.6 kernel now boots when compiled with gcc 
> > 3.3.3 compiler.
> >    
>
> Thanks. However, it really should use the "noinline" define that we have 
> for this, and that doesn't upset older versions of gcc that don't 
> understand the "noinline" attribute. 
>
> Also, declaration and definition should match, even if gcc doesn't seem to 
> catch that.
>
> So here's the version I committed. It should be obviously ok, but it's 
> still a good idea to verify..
>
> 		Linus
>
> ---
> # This is a BitKeeper generated diff -Nru style patch.
> #
> # ChangeSet
> #   2004/10/05 18:51:53-07:00 torvalds@xxxxxxxxxxxx 
> #   i386: mark do_test_wp_bit() noinline
> #   
> #   As reported by Zachary Amsden <zach@xxxxxxxxxx>,
> #   some gcc versions will inline the function even when
> #   it is declared after the call-site. This particular
> #   function must not be inlined, since the exception
> #   recovery doesn't like __init sections (which the caller
> #   is in).
> # 
> # arch/i386/mm/init.c
> #   2004/10/05 18:51:43-07:00 torvalds@xxxxxxxxxxxx +2 -2
> #   i386: mark do_test_wp_bit() noinline
> #   
> #   As reported by Zachary Amsden <zach@xxxxxxxxxx>,
> #   some gcc versions will inline the function even when
> #   it is declared after the call-site. This particular
> #   function must not be inlined, since the exception
> #   recovery doesn't like __init sections (which the caller
> #   is in).
> # 
> diff -Nru a/arch/i386/mm/init.c b/arch/i386/mm/init.c
> --- a/arch/i386/mm/init.c	2004-10-05 19:04:53 -07:00
> +++ b/arch/i386/mm/init.c	2004-10-05 19:04:53 -07:00
> @@ -45,7 +45,7 @@
> DEFINE_PER_CPU(struct mmu_gather, mmu_gathers);
> unsigned long highstart_pfn, highend_pfn;
>
> -static int do_test_wp_bit(void);
> +static int noinline do_test_wp_bit(void);
>
> /*
>  * Creates a middle page table and puts a pointer to it in the
> @@ -673,7 +673,7 @@
>  * This function cannot be __init, since exceptions don't work in that
>  * section.  Put this after the callers, so that it cannot be inlined.
>  */
> -static int do_test_wp_bit(void)
> +static int noinline do_test_wp_bit(void)
> {
> 	char tmp_reg;
> 	int flag;
>  

-
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/