3.2.93-rc1 review patch. If anyone has any objections, please let me know. ------------------ From: Nicholas Piggin <npiggin@xxxxxxxxx> commit 34f19ff1b5a0d11e46df479623d6936460105c9f upstream. Emergency stacks have their thread_info mostly uninitialised, which in particular means garbage preempt_count values. Emergency stack code runs with interrupts disabled entirely, and is used very rarely, so this has been unnoticed so far. It was found by a proposed new powerpc watchdog that takes a soft-NMI directly from the masked_interrupt handler and using the emergency stack. That crashed at BUG_ON(in_nmi()) in nmi_enter(). preempt_count()s were found to be garbage. To fix this, zero the entire THREAD_SIZE allocation, and initialize the thread_info. Reported-by: Abdul Haleem <abdhalee@xxxxxxxxxxxxxxxxxx> Signed-off-by: Nicholas Piggin <npiggin@xxxxxxxxx> [mpe: Move it all into setup_64.c, use a function not a macro. Fix crashes on Cell by setting preempt_count to 0 not HARDIRQ_OFFSET] Signed-off-by: Michael Ellerman <mpe@xxxxxxxxxxxxxx> [bwh: Backported to 3.2: - There's only one emergency stack - No need to call klp_init_thread_info() - Add the ti variable in emergency_stack_init()] Signed-off-by: Ben Hutchings <ben@xxxxxxxxxxxxxxx> --- arch/powerpc/kernel/setup_64.c | 31 ++++++++++++++++++++++++++++--- 1 file changed, 28 insertions(+), 3 deletions(-) --- a/arch/powerpc/kernel/setup_64.c +++ b/arch/powerpc/kernel/setup_64.c @@ -509,6 +509,23 @@ static void __init exc_lvl_early_init(vo #endif /* + * Emergency stacks are used for a range of things, from asynchronous + * NMIs (system reset, machine check) to synchronous, process context. + * We set preempt_count to zero, even though that isn't necessarily correct. To + * get the right value we'd need to copy it from the previous thread_info, but + * doing that might fault causing more problems. + * TODO: what to do with accounting? + */ +static void emerg_stack_init_thread_info(struct thread_info *ti, int cpu) +{ + ti->task = NULL; + ti->cpu = cpu; + ti->preempt_count = 0; + ti->local_flags = 0; + ti->flags = 0; +} + +/* * Stack space used when we detect a bad kernel stack pointer, and * early in SMP boots before relocation is enabled. */ @@ -525,12 +542,20 @@ static void __init emergency_stack_init( * Since we use these as temporary stacks during secondary CPU * bringup, we need to get at them in real mode. This means they * must also be within the RMO region. + * + * The IRQ stacks allocated elsewhere in this file are zeroed and + * initialized in kernel/irq.c. These are initialized here in order + * to have emergency stacks available as early as possible. */ limit = min(safe_stack_limit(), ppc64_rma_size); for_each_possible_cpu(i) { unsigned long sp; + struct thread_info *ti; sp = memblock_alloc_base(THREAD_SIZE, THREAD_SIZE, limit); + ti = __va(sp); + memset(ti, 0, THREAD_SIZE); + emerg_stack_init_thread_info(ti, i); sp += THREAD_SIZE; paca[i].emergency_sp = __va(sp); }