linux-kernel - Re: [PATCH] x86: fix wakeup

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]

Message-ID: <20081117165224.GH12081@elte.hu>
Date:	Mon, 17 Nov 2008 17:52:24 +0100
From:	Ingo Molnar <mingo@...e.hu>
To:	Yinghai Lu <yinghai@...nel.org>
Cc:	Thomas Gleixner <tglx@...utronix.de>,
	"H. Peter Anvin" <hpa@...or.com>,
	Andrew Morton <akpm@...ux-foundation.org>,
	"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] x86: fix wakeup_cpu with numaq/es7000 v2


* Yinghai Lu <yinghai@...nel.org> wrote:

> Impact: fix wakeup path with numaq and es7000

applied	to tip/x86/quirks, thanks Yinghai!

A couple of comments:

> +static inline void inquire_remote_apic(int apicid)
> +{
> +	if (apic_verbosity >= APIC_DEBUG)
> +		__inquire_remote_apic(apicid);
> +}

could you please make this a non-include-file function and only 
install the function vector if the platform necessiates it?

I.e.:

> +	void (*wait_for_init_deassert)(atomic_t *deassert);
> +	void (*smp_callin_clear_local_apic)(void);
> +	void (*store_NMI_vector)(unsigned short *high, unsigned short *low);
> +	void (*restore_NMI_vector)(unsigned short *high, unsigned short *low);
> +	void (*inquire_remote_apic)(int apicid);

should have NULL in the 'default' case - and only have some exception 
(quick) function installed in the non-default case.

That way the 3 separate wait_for_init_deassert() and 
inquire_remote_apic() inline functions can be removed and the 
exception be moved to es7000_32.c and numaq_32.c.

like you did it here properly to install the MIP-init sequence in the 
es7000 case:

> Index: linux-2.6/arch/x86/kernel/es7000_32.c
> ===================================================================
> --- linux-2.6.orig/arch/x86/kernel/es7000_32.c
> +++ linux-2.6/arch/x86/kernel/es7000_32.c
> @@ -40,6 +40,7 @@
>  #include <asm/smp.h>
>  #include <asm/apicdef.h>
>  #include <mach_mpparse.h>
> +#include <asm/setup.h>
>  
>  /*
>   * ES7000 chipsets
> @@ -161,6 +162,26 @@ es7000_rename_gsi(int ioapic, int gsi)
>  	return gsi;
>  }
>  
> +#ifdef CONFIG_ES7000_CLUSTERED_APIC
> +static int wakeup_secondary_cpu_via_mip(int cpu, unsigned long eip)
> +{
> +	unsigned long vect = 0, psaival = 0;
> +
> +	if (psai == NULL)
> +		return -1;
> +
> +	vect = ((unsigned long)__pa(eip)/0x1000) << 16;
> +	psaival = (0x1000000 | vect | cpu);
> +
> +	while (*psai & 0x1000000)
> +		;
> +
> +	*psai = psaival;
> +
> +	return 0;
> +}
> +#endif
> +
>  void __init
>  setup_unisys(void)
>  {
> @@ -176,6 +197,9 @@ setup_unisys(void)
>  	else
>  		es7000_plat = ES7000_CLASSIC;
>  	ioapic_renumber_irq = es7000_rename_gsi;
> +#ifdef CONFIG_ES7000_CLUSTERED_APIC
> +	x86_quirks->wakeup_secondary_cpu = wakeup_secondary_cpu_via_mip;
> +#endif
>  }

btw:

shouldnt CONFIG_ES7000_CLUSTERED_APIC be converted to a proper genapic 
variant (like we have in apic.c), instead of this #ifdef driven 
mechanism?

	Ingo
--
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/