lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  linux-hardening  linux-cve-announce  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date:	Fri, 22 May 2009 22:33:32 +0200 (CEST)
From:	Thomas Gleixner <tglx@...utronix.de>
To:	Rakib Mullick <rakib.mullick@...il.com>
cc:	Ingo Molnar <mingo@...e.hu>, Cyrill Gorcunov <gorcunov@...nvz.org>,
	akpm@...ux-foundation.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] x86,APIC: Detect lapic_is_integrated() once - use on
 and on.

On Fri, 22 May 2009, Rakib Mullick wrote:
>   Impact: Reduce text space through efficient coding
> 
>  Determine apic is integrated or not - by calling
> lapic_is_integrated() once from APIC_init_uniprocessor() and keep it
> in a variable integrated_lapic. Thus we can determine apic is
> integrated or not, by checking the variable instead of calling
> lapic_is_integrated() on and on. Marking lapic_is_integrated() as
> __init, which reduces some text space. Also allows us to get away from
> messy #ifdef and inline.
> 
> ---
> Signed-off-by: Rakib Mullick <rakib.mullick@...il.com>
> 
> --- linus/arch/x86/kernel/apic/apic.c	2009-05-21 22:22:15.000000000 +0600
> +++ rakib/arch/x86/kernel/apic/apic.c	2009-05-21 23:27:26.000000000 +0600
> @@ -127,6 +127,9 @@ early_param("nox2apic", setup_nox2apic);
> 
>  unsigned long mp_lapic_addr;
>  int disable_apic;
> +
> +static int integrated_lapic;
> +
>  /* Disable local APIC timer from the kernel commandline or via dmi quirk */
>  static int disable_apic_timer __cpuinitdata;
>  /* Local APIC timer works in C2 */
> @@ -188,13 +191,12 @@ static inline int lapic_get_version(void
>  /*
>   * Check, if the APIC is integrated or a separate chip
>   */
> -static inline int lapic_is_integrated(void)
> +void __init lapic_is_integrated(void)
>  {
> -#ifdef CONFIG_X86_64
> -	return 1;
> -#else
> -	return APIC_INTEGRATED(lapic_get_version());
> -#endif
> +	if (APIC_INTEGRATED(lapic_get_version()))
> +		integrated_lapic = 1;
> +	else
> +		integrated_lapic = 0;
>  }

I agree in general, but the patch is flawed as it prevents the
compiler to optimize the X86_64 case out.

What you really want is:

#ifdef CONFIG_X86_64
# define integrated_lapic (1)
#else
static int integrated_lapic __read_mostly;
#endif

static inline void lapic_is_integrated(void)
{
#ifdef CONFIG_X86_32
	integrate_lapic = APIC_INTEGRATED(lapic_get_version());
#endif
}

So that way the check is only done for 32bit systems and the compiler
will optimize out the code which depends on !integrated_lapic.

Thanks,

	tglx
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ