Re: [PATCH][2.6-mm] Fix CONFIG_SMT oops on UP

[Zwane Mwaikambo]

On Sat, 24 Jan 2004, Andrew Morton wrote:

> Zwane Mwaikambo <zwane@xxxxxxxxxxxxxxxx> wrote:
> >
> > +	cpu_sibling_map[0] = CPU_MASK_NONE;
>
> alas, this will not compile with NR_CPUS > 4*BITS_PER_LONG because this:
>
> 	#define CPU_MASK_NONE   { {[0 ... CPU_ARRAY_SIZE-1] =  0UL} }
>
> is not a suitable rhs - it can only be used for initalisers.  Fixing this
> would be appreciated.
>
> Meanwhile, I'll use cpus_clear() in there.

There was actually other breakage too, so how about;

Index: linux-2.6.2-rc1-mm2/arch/i386/kernel/smpboot.c
===================================================================
RCS file: /home/cvsroot/linux-2.6.2-rc1-mm2/arch/i386/kernel/smpboot.c,v
retrieving revision 1.1.1.1
diff -u -p -B -r1.1.1.1 smpboot.c
--- linux-2.6.2-rc1-mm2/arch/i386/kernel/smpboot.c	23 Jan 2004 22:22:46 -0000	1.1.1.1
+++ linux-2.6.2-rc1-mm2/arch/i386/kernel/smpboot.c	24 Jan 2004 18:46:41 -0000
@@ -951,6 +951,8 @@ static void __init smp_boot_cpus(unsigne

 	current_thread_info()->cpu = 0;
 	smp_tune_scheduling();
+	cpus_clear(cpu_sibling_map[0]);
+	cpu_set(0, cpu_sibling_map[0]);

 	/*
 	 * If we couldn't find an SMP configuration at boot time,
@@ -1083,7 +1085,7 @@ static void __init smp_boot_cpus(unsigne
 	 * efficiently.
 	 */
 	for (cpu = 0; cpu < NR_CPUS; cpu++)
-		cpu_sibling_map[cpu] = CPU_MASK_NONE;
+		cpus_clear(cpu_sibling_map[cpu]);

 	for (cpu = 0; cpu < NR_CPUS; cpu++) {
 		int siblings = 0;
@@ -1273,7 +1275,7 @@ __init void arch_init_sched_domains(void
 		for_each_cpu_mask(j, cpu_domain->span) {
 			struct sched_group *cpu = &sched_group_cpus[j];

-			cpu->cpumask = CPU_MASK_NONE;
+			cpus_clear(cpu->cpumask);
 			cpu_set(j, cpu->cpumask);

 			if (!first_cpu)
-
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/