From mboxrd@z Thu Jan 1 00:00:00 1970 From: jungseoklee85@gmail.com (Jungseok Lee) Date: Sun, 20 Dec 2015 20:07:46 +0900 Subject: [PATCH v8 9/4] arm64: remove irq_count and do_softirq_own_stack() In-Reply-To: <1450454507-8907-1-git-send-email-james.morse@arm.com> References: <1449226948-14251-1-git-send-email-james.morse@arm.com> <1450454507-8907-1-git-send-email-james.morse@arm.com> Message-ID: <68E0E2C7-A4C2-45C2-95C5-842FA807F918@gmail.com> To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org On Dec 19, 2015, at 1:01 AM, James Morse wrote: Hi James, > sysrq_handle_reboot() re-enables interrupts while on the irq stack. The > irq_stack implementation wrongly assumed this would only ever happen > via the softirq path, allowing it to update irq_count late, in > do_softirq_own_stack(). > > This means if an irq occurs in sysrq_handle_reboot(), during > emergency_restart() the stack will be corrupted, as irq_count wasn't > updated. > > Lose the optimisation, and instead of moving the adding/subtracting of > irq_count into irq_stack_entry/irq_stack_exit, remove it, and compare > sp_el0 (struct task_info) with sp & ~(THREAD_SIZE - 1). This tells us if Nit: s/task_info/thread_info > we are on a task stack, if so, we can safely switch to the irq stack. > Finally, remove do_softirq_own_stack(), we don't need it anymore. > > Reported-by: Will Deacon > Signed-off-by: James Morse > --- > arch/arm64/include/asm/irq.h | 2 -- > arch/arm64/kernel/entry.S | 17 +++++++++-------- > arch/arm64/kernel/irq.c | 38 +------------------------------------- > 3 files changed, 10 insertions(+), 47 deletions(-) > > diff --git a/arch/arm64/include/asm/irq.h b/arch/arm64/include/asm/irq.h > index 3bece4379bd9..b77197d941fc 100644 > --- a/arch/arm64/include/asm/irq.h > +++ b/arch/arm64/include/asm/irq.h > @@ -11,8 +11,6 @@ > #include > #include > > -#define __ARCH_HAS_DO_SOFTIRQ > - > struct pt_regs; > > DECLARE_PER_CPU(unsigned long [IRQ_STACK_SIZE/sizeof(long)], irq_stack); > diff --git a/arch/arm64/kernel/entry.S b/arch/arm64/kernel/entry.S > index 0667fb7d8bb1..6745a9041f99 100644 > --- a/arch/arm64/kernel/entry.S > +++ b/arch/arm64/kernel/entry.S > @@ -181,19 +181,20 @@ alternative_endif > .macro irq_stack_entry > mov x19, sp // preserve the original sp > > - this_cpu_ptr irq_stack, x25, x26 > - > /* > - * Check the lowest address on irq_stack for the irq_count value, > - * incremented by do_softirq_own_stack if we have re-enabled irqs > - * while on the irq_stack. > + * Compare sp and sp_el0, if the top ~(THREAD_SIZE - 1) bits match, > + * we are on a task stack, and should switch to the irq stack. > */ > - ldr x26, [x25] > - cbnz x26, 9998f // recursive use? > + mrs x26, sp_el0 // already masked Nit: How about using 'get_thread_info x26'? > + and x25, x19, #~(THREAD_SIZE - 1) > + cmp x25, x26 > + b.ne 9998f > > - /* switch to the irq stack */ > + this_cpu_ptr irq_stack, x25, x26 > mov x26, #IRQ_STACK_START_SP > add x26, x25, x26 > + > + /* switch to the irq stack */ > mov sp, x26 > > /* > diff --git a/arch/arm64/kernel/irq.c b/arch/arm64/kernel/irq.c > index ff7ebb710e51..2386b26c0712 100644 > --- a/arch/arm64/kernel/irq.c > +++ b/arch/arm64/kernel/irq.c > @@ -25,24 +25,14 @@ > #include > #include > #include > -#include > #include > #include > > unsigned long irq_err_count; > > -/* > - * irq stack only needs to be 16 byte aligned - not IRQ_STACK_SIZE aligned. > - * irq_stack[0] is used as irq_count, a non-zero value indicates the stack > - * is in use, and el?_irq() shouldn't switch to it. This is used to detect > - * recursive use of the irq_stack, it is lazily updated by > - * do_softirq_own_stack(), which is called on the irq_stack, before > - * re-enabling interrupts to process softirqs. > - */ > +/* irq stack only needs to be 16 byte aligned - not IRQ_STACK_SIZE aligned. */ > DEFINE_PER_CPU(unsigned long [IRQ_STACK_SIZE/sizeof(long)], irq_stack) __aligned(16); > > -#define IRQ_COUNT() (*per_cpu(irq_stack, smp_processor_id())) > - > int arch_show_interrupts(struct seq_file *p, int prec) > { > show_ipi_list(p, prec); > @@ -66,29 +56,3 @@ void __init init_IRQ(void) > if (!handle_arch_irq) > panic("No interrupt controller found."); > } > - > -/* > - * do_softirq_own_stack() is called from irq_exit() before __do_softirq() > - * re-enables interrupts, at which point we may re-enter el?_irq(). We > - * increase irq_count here so that el1_irq() knows that it is already on the > - * irq stack. > - * > - * Called with interrupts disabled, so we don't worry about moving cpu, or > - * being interrupted while modifying irq_count. > - * > - * This function doesn't actually switch stack. > - */ > -void do_softirq_own_stack(void) > -{ > - int cpu = smp_processor_id(); > - > - WARN_ON_ONCE(!irqs_disabled()); > - > - if (on_irq_stack(current_stack_pointer, cpu)) { > - IRQ_COUNT()++; > - __do_softirq(); > - IRQ_COUNT()--; > - } else { > - __do_softirq(); > - } > -} The approach, removal of the counter, looks good to me from a high level point of view although I have not played with this change seriously yet. If I spot something strange, I will report it. Best Regards Jungseok Lee