[PATCH] fix 4k irqstacks on x86 (and add voyager support)

[PATCH] fix 4k irqstacks on x86 (and add voyager support)

[James Bottomley]

There's a bug in the x86 code in that it sets the boot CPU to zero. 
This isn't correct since some subarch's use physically indexed CPUs. 
However, subarchs have either set the boot cpu before irq_INIT() (or
just inherited the default zero from INIT_THREAD_INFO()), so it's safe
to believe current_thread_info()->cpu about the boot cpu.

James

===== arch/i386/kernel/i8259.c 1.28 vs edited =====
--- 1.28/arch/i386/kernel/i8259.c	Mon Apr 12 12:54:45 2004
+++ edited/arch/i386/kernel/i8259.c	Thu Apr 15 09:59:27 2004
@@ -445,6 +445,5 @@
 	if (boot_cpu_data.hard_math && !cpu_has_fpu)
 		setup_irq(FPU_IRQ, &fpu_irq);
 
-	current_thread_info()->cpu = 0;
-	irq_ctx_init(0);
+	irq_ctx_init(current_thread_info()->cpu);
 }
===== arch/i386/mach-voyager/voyager_smp.c 1.19 vs edited =====
--- 1.19/arch/i386/mach-voyager/voyager_smp.c	Sun Mar 14 05:23:02 2004
+++ edited/arch/i386/mach-voyager/voyager_smp.c	Thu Apr 15 09:52:49 2004
@@ -599,12 +599,10 @@
 	idle->thread.eip = (unsigned long) start_secondary;
 	unhash_process(idle);
 	/* init_tasks (in sched.c) is indexed logically */
-#if 0
-	// for AC kernels
-	stack_start.esp = (THREAD_SIZE + (__u8 *)TSK_TO_KSTACK(idle));
-#else
-	stack_start.esp = (void *) (1024 + PAGE_SIZE + (char *)idle->thread_info);
-#endif
+	stack_start.esp = (void *) idle->thread.esp;
+
+	irq_ctx_init(cpu);
+
 	/* Note: Don't modify initial ss override */
 	VDEBUG(("VOYAGER SMP: Booting CPU%d at 0x%lx[%x:%x], stack %p\n", cpu, 
 		(unsigned long)hijack_source.val, hijack_source.idt.Segment,
===== arch/i386/mach-voyager/voyager_thread.c 1.3 vs edited =====
--- 1.3/arch/i386/mach-voyager/voyager_thread.c	Wed Feb 12 21:35:38 2003
+++ edited/arch/i386/mach-voyager/voyager_thread.c	Thu Apr 15 09:11:35 2004
@@ -135,7 +135,7 @@
 	init_timer(&wakeup_timer);
 
 	sigfillset(&current->blocked);
-	current->tty = NULL;	/* get rid of controlling tty */
+	current->signal->tty = NULL;
 
 	printk(KERN_NOTICE "Voyager starting monitor thread\n");
 

Re: [PATCH] fix 4k irqstacks on x86 (and add voyager support)

[Zwane Mwaikambo]

On Thu, 15 Apr 2004, James Bottomley wrote:

> There's a bug in the x86 code in that it sets the boot CPU to zero.
> This isn't correct since some subarch's use physically indexed CPUs.
> However, subarchs have either set the boot cpu before irq_INIT() (or
> just inherited the default zero from INIT_THREAD_INFO()), so it's safe
> to believe current_thread_info()->cpu about the boot cpu.

There is also smp_boot_cpus() which sets it to zero yet again later on =)

Re: [PATCH] fix 4k irqstacks on x86 (and add voyager support)

[James Bottomley]

On Thu, 2004-04-15 at 10:29, Zwane Mwaikambo wrote:
> > There's a bug in the x86 code in that it sets the boot CPU to zero.
> > This isn't correct since some subarch's use physically indexed CPUs.
> > However, subarchs have either set the boot cpu before irq_INIT() (or
> > just inherited the default zero from INIT_THREAD_INFO()), so it's safe
> > to believe current_thread_info()->cpu about the boot cpu.
> 
> There is also smp_boot_cpus() which sets it to zero yet again later on =)

That's PC specific, not subarch generic, so it doesn't matter to me.

James

Re: [PATCH] fix 4k irqstacks on x86 (and add voyager support)

[Zwane Mwaikambo]

On Thu, 15 Apr 2004, James Bottomley wrote:

> On Thu, 2004-04-15 at 10:29, Zwane Mwaikambo wrote:
> > > There's a bug in the x86 code in that it sets the boot CPU to zero.
> > > This isn't correct since some subarch's use physically indexed CPUs.
> > > However, subarchs have either set the boot cpu before irq_INIT() (or
> > > just inherited the default zero from INIT_THREAD_INFO()), so it's safe
> > > to believe current_thread_info()->cpu about the boot cpu.
> >
> > There is also smp_boot_cpus() which sets it to zero yet again later on =)
>
> That's PC specific, not subarch generic, so it doesn't matter to me.

Sorry for being a bit slow here, doesn't this affect voyager at all?

init/main.c:init()
	smp_prepare_cpus()

arch/i386/kernel/smpboot.c:smp_prepare_cpus()
	smp_boot_cpus();

obj-$(CONFIG_X86_SMP)           += smp.o smpboot.o

Re: [PATCH] fix 4k irqstacks on x86 (and add voyager support)

[James Bottomley]

On Thu, 2004-04-15 at 11:16, Zwane Mwaikambo wrote:
> Sorry for being a bit slow here, doesn't this affect voyager at all?

No.  Voyager has a separate smp implementation.

James