Re: [PATCH] i386: per-CPU double fault TSS and stack

2007-09-03 Thread Andi Kleen
> It costs 4.xx k space per CPU - perhaps a constraint for embedded? Not a major one. > >In fact I would prefer to just eliminate CONFIG_DOUBLEFAULT (imho > >it always a bad idea because the amount of code it saves is miniscule) > > instead of adding such a ifdef maze. > > It's configurable for

Re: [PATCH] i386: per-CPU double fault TSS and stack

2007-09-03 Thread Jan Beulich
>>> Andi Kleen <[EMAIL PROTECTED]> 01.09.07 12:33 >>> > >Can you cc the next version to Linus please? He's probably best qualified >to review the i386 double fault handler because he wrote it originally. >I must admit the code always scared me a bit. Will do. >> +#ifdef CONFIG_HOTPLUG_CPU >>

Re: [PATCH] i386: per-CPU double fault TSS and stack

2007-09-03 Thread Jan Beulich
Andi Kleen [EMAIL PROTECTED] 01.09.07 12:33 Can you cc the next version to Linus please? He's probably best qualified to review the i386 double fault handler because he wrote it originally. I must admit the code always scared me a bit. Will do. +#ifdef CONFIG_HOTPLUG_CPU +static void

Re: [PATCH] i386: per-CPU double fault TSS and stack

2007-09-01 Thread Andi Kleen
Can you cc the next version to Linus please? He's probably best qualified to review the i386 double fault handler because he wrote it originally. I must admit the code always scared me a bit. > +#ifdef CONFIG_HOTPLUG_CPU > +static void *noinline __init_refok > +#else > +static inline void

Re: [PATCH] i386: per-CPU double fault TSS and stack

2007-09-01 Thread Andi Kleen
Can you cc the next version to Linus please? He's probably best qualified to review the i386 double fault handler because he wrote it originally. I must admit the code always scared me a bit. +#ifdef CONFIG_HOTPLUG_CPU +static void *noinline __init_refok +#else +static inline void *__init