From 9a0bd2ce954a8033987b9b3118ee8bccabb0b8f2 Mon Sep 17 00:00:00 2001 From: Sebastian Andrzej Siewior Date: Thu, 15 Sep 2016 17:16:44 +0200 Subject: [PATCH 038/365] iommu/vt-d: don't disable preemption while accessing deferred_flush() get_cpu() disables preemption and returns the current CPU number. The CPU number is later only used once while retrieving the address of the local's CPU deferred_flush pointer. We can instead use raw_cpu_ptr() while we remain preemptible. The worst thing that can happen is that flush_unmaps_timeout() is invoked multiple times: once by taskA after seeing HIGH_WATER_MARK and then preempted to another CPU and then by taskB which saw HIGH_WATER_MARK on the same CPU as taskA. It is also likely that ->size got from HIGH_WATER_MARK to 0 right after its read because another CPU invoked flush_unmaps_timeout() for this CPU. The access to flush_data is protected by a spinlock so even if we get migrated to another CPU or preempted - the data structure is protected. While at it, I marked deferred_flush static since I can't find a reference to it outside of this file. Signed-off-by: Sebastian Andrzej Siewior --- drivers/iommu/intel-iommu.c | 8 ++------ 1 file changed, 2 insertions(+), 6 deletions(-) diff --git a/drivers/iommu/intel-iommu.c b/drivers/iommu/intel-iommu.c index 593a4bfcba42..fa7d78442ee7 100644 --- a/drivers/iommu/intel-iommu.c +++ b/drivers/iommu/intel-iommu.c @@ -480,7 +480,7 @@ struct deferred_flush_data { struct deferred_flush_table *tables; }; -DEFINE_PER_CPU(struct deferred_flush_data, deferred_flush); +static DEFINE_PER_CPU(struct deferred_flush_data, deferred_flush); /* bitmap for indexing intel_iommus */ static int g_num_of_iommus; @@ -3742,10 +3742,8 @@ static void add_unmap(struct dmar_domain *dom, unsigned long iova_pfn, struct intel_iommu *iommu; struct deferred_flush_entry *entry; struct deferred_flush_data *flush_data; - unsigned int cpuid; - cpuid = get_cpu(); - flush_data = per_cpu_ptr(&deferred_flush, cpuid); + flush_data = raw_cpu_ptr(&deferred_flush); /* Flush all CPUs' entries to avoid deferring too much. If * this becomes a bottleneck, can just flush us, and rely on @@ -3778,8 +3776,6 @@ static void add_unmap(struct dmar_domain *dom, unsigned long iova_pfn, } flush_data->size++; spin_unlock_irqrestore(&flush_data->lock, flags); - - put_cpu(); } static void intel_unmap(struct device *dev, dma_addr_t dev_addr, size_t size) -- 2.28.0