diff mbox series

[3/4] UBUNTU: SAUCE: swiotlb: allocate memory in a cache-friendly way

Message ID 20220505123949.12030-4-tim.gardner@canonical.com
State New
Headers show
Series swiotlb patch needed for CVM | expand

Commit Message

Tim Gardner May 5, 2022, 12:39 p.m. UTC 
From: Chao Gao <chao.gao@intel.com>

BugLink: https://bugs.launchpad.net/bugs/1971701

Currently, swiotlb uses a global index to indicate the starting point
of next search. The index increases from 0 to the number of slots - 1
and then wraps around. It is straightforward but not cache-friendly
because the "oldest" slot in swiotlb tends to be used first.

Freed slots are probably accessed right before being freed, especially
in VM's case (device backends access them in DMA_TO_DEVICE mode; guest
accesses them in other DMA modes). Thus those just freed slots may
reside in cache. Then reusing those just freed slots can reduce cache
misses.

To that end, maintain a free list for free slots and insert freed slots
from the head and searching for free slots always starts from the head.

With this optimization, network throughput of sending data from host to
guest, measured by iperf3, increases by 7%.

A bad side effect of this patch is we cannot use a large stride to skip
unaligned slots when there is an alignment requirement. Currently, a
large stride is used when a) device has an alignment requirement, stride
is calculated according to the requirement; b) the requested size is
larger than PAGE_SIZE. For x86 with 4KB page size, stride is set to 2.

For case a), few devices have an alignment requirement; the impact is
limited. For case b) this patch probably leads to one (or more if page size
is larger than 4K) additional lookup; but as the "io_tlb_slot" struct of
free slots are also accessed when freeing slots, they probably resides in
CPU cache as well and then the overhead is almost negligible.

Suggested-by: Andi Kleen <ak@linux.intel.com>
Signed-off-by: Chao Gao <chao.gao@intel.com>
(cherry picked from commit 3a91fd70a99451e91c0a0bde5637cbcf7748252c https://github.com/intel/tdx.git)
Signed-off-by: Tim Gardner <tim.gardner@canonical.com>
---
 include/linux/swiotlb.h | 13 ++++++----
 kernel/dma/swiotlb.c    | 53 ++++++++++++++---------------------------
 2 files changed, 26 insertions(+), 40 deletions(-)
diff mbox series

Patch

diff --git a/include/linux/swiotlb.h b/include/linux/swiotlb.h
index 47758796d2c0..aabca568bd15 100644
--- a/include/linux/swiotlb.h
+++ b/include/linux/swiotlb.h
@@ -64,6 +64,12 @@  dma_addr_t swiotlb_map(struct device *dev, phys_addr_t phys,
 #ifdef CONFIG_SWIOTLB
 extern enum swiotlb_force swiotlb_force;
 
+struct io_tlb_slot {
+	phys_addr_t orig_addr;
+	size_t alloc_size;
+	struct list_head node;
+};
+
 /**
  * struct io_tlb_mem - IO TLB Memory Pool Descriptor
  *
@@ -98,16 +104,13 @@  struct io_tlb_mem {
 	void *vaddr;
 	unsigned long nslabs;
 	unsigned long used;
-	unsigned int index;
+	struct list_head free_slots;
 	spinlock_t lock;
 	struct dentry *debugfs;
 	bool late_alloc;
 	bool force_bounce;
 	bool for_alloc;
-	struct io_tlb_slot {
-		phys_addr_t orig_addr;
-		size_t alloc_size;
-	} *slots;
+	struct io_tlb_slot *slots;
 	unsigned long *bitmap;
 };
 extern struct io_tlb_mem io_tlb_default_mem;
diff --git a/kernel/dma/swiotlb.c b/kernel/dma/swiotlb.c
index 55cd20f068fd..180e02f5fd26 100644
--- a/kernel/dma/swiotlb.c
+++ b/kernel/dma/swiotlb.c
@@ -227,7 +227,7 @@  static void swiotlb_init_io_tlb_mem(struct io_tlb_mem *mem, phys_addr_t start,
 	mem->nslabs = nslabs;
 	mem->start = start;
 	mem->end = mem->start + bytes;
-	mem->index = 0;
+	INIT_LIST_HEAD(&mem->free_slots);
 	mem->late_alloc = late_alloc;
 
 	if (swiotlb_force == SWIOTLB_FORCE)
@@ -238,6 +238,7 @@  static void swiotlb_init_io_tlb_mem(struct io_tlb_mem *mem, phys_addr_t start,
 		__set_bit(i, mem->bitmap);
 		mem->slots[i].orig_addr = INVALID_PHYS_ADDR;
 		mem->slots[i].alloc_size = 0;
+		list_add_tail(&mem->slots[i].node, &mem->free_slots);
 	}
 
 	/*
@@ -516,13 +517,6 @@  static inline unsigned long get_max_slots(unsigned long boundary_mask)
 	return nr_slots(boundary_mask + 1);
 }
 
-static unsigned int wrap_index(struct io_tlb_mem *mem, unsigned int index)
-{
-	if (index >= mem->nslabs)
-		return 0;
-	return index;
-}
-
 /*
  * Find a suitable number of IO TLB entries size that will fit this request and
  * allocate a buffer from that IO TLB pool.
@@ -531,46 +525,41 @@  static int swiotlb_find_slots(struct device *dev, phys_addr_t orig_addr,
 			      size_t alloc_size)
 {
 	struct io_tlb_mem *mem = dev->dma_io_tlb_mem;
+	struct io_tlb_slot *slot, *tmp;
 	unsigned long boundary_mask = dma_get_seg_boundary(dev);
 	dma_addr_t tbl_dma_addr =
 		phys_to_dma_unencrypted(dev, mem->start) & boundary_mask;
 	unsigned long max_slots = get_max_slots(boundary_mask);
 	unsigned int iotlb_align_mask =
 		dma_get_min_align_mask(dev) & ~(IO_TLB_SIZE - 1);
-	unsigned int nslots = nr_slots(alloc_size), stride;
-	unsigned int index, wrap, i;
+	unsigned int nslots = nr_slots(alloc_size);
+	unsigned int index, i;
 	unsigned int offset = swiotlb_align_offset(dev, orig_addr);
 	unsigned long flags;
 
 	BUG_ON(!nslots);
 
-	/*
-	 * For mappings with an alignment requirement don't bother looping to
-	 * unaligned slots once we found an aligned one.  For allocations of
-	 * PAGE_SIZE or larger only look for page aligned allocations.
-	 */
-	stride = (iotlb_align_mask >> IO_TLB_SHIFT) + 1;
-	if (alloc_size >= PAGE_SIZE)
-		stride = max(stride, stride << (PAGE_SHIFT - IO_TLB_SHIFT));
-
 	spin_lock_irqsave(&mem->lock, flags);
 	if (unlikely(nslots > mem->nslabs - mem->used))
 		goto not_found;
 
-	index = wrap = wrap_index(mem, ALIGN(mem->index, stride));
-	do {
+	list_for_each_entry_safe(slot, tmp, &mem->free_slots, node) {
+		index = slot - mem->slots;
 		if (orig_addr &&
 		    (slot_addr(tbl_dma_addr, index) & iotlb_align_mask) !=
 			    (orig_addr & iotlb_align_mask)) {
-			index = wrap_index(mem, index + 1);
 			continue;
 		}
 
-		/* Start from the next segment if no enough free entries */
-		if (io_tlb_offset(index) + nslots > IO_TLB_SEGSIZE) {
-			index = wrap_index(mem, round_up(index, IO_TLB_SEGSIZE));
+		if (io_tlb_offset(index) + nslots > IO_TLB_SEGSIZE)
+			continue;
+
+		/*
+		 * If requested size is larger than a page, ensure allocated
+		 * memory to be page aligned.
+		 */
+		if (alloc_size >= PAGE_SIZE && (slot_addr(tbl_dma_addr, index) & ~PAGE_MASK))
 			continue;
-		}
 
 		/*
 		 * If we find a slot that indicates we have 'nslots' number of
@@ -584,8 +573,7 @@  static int swiotlb_find_slots(struct device *dev, phys_addr_t orig_addr,
 					index + nslots)
 				goto found;
 		}
-		index = wrap_index(mem, index + stride);
-	} while (index != wrap);
+	}
 
 not_found:
 	spin_unlock_irqrestore(&mem->lock, flags);
@@ -596,15 +584,9 @@  static int swiotlb_find_slots(struct device *dev, phys_addr_t orig_addr,
 		__clear_bit(i, mem->bitmap);
 		mem->slots[i].alloc_size =
 			alloc_size - (offset + ((i - index) << IO_TLB_SHIFT));
+		list_del(&mem->slots[i].node);
 	}
 
-	/*
-	 * Update the indices to avoid searching in the next round.
-	 */
-	if (index + nslots < mem->nslabs)
-		mem->index = index + nslots;
-	else
-		mem->index = 0;
 	mem->used += nslots;
 
 	spin_unlock_irqrestore(&mem->lock, flags);
@@ -679,6 +661,7 @@  static void swiotlb_release_slots(struct device *dev, phys_addr_t tlb_addr)
 		__set_bit(i, mem->bitmap);
 		mem->slots[i].orig_addr = INVALID_PHYS_ADDR;
 		mem->slots[i].alloc_size = 0;
+		list_add(&mem->slots[i].node, &mem->free_slots);
 	}
 
 	mem->used -= nslots;