summaryrefslogtreecommitdiff
path: root/arch/sparc/kernel/pci_sun4v.c
diff options
context:
space:
mode:
authorSowmini Varadhan <sowmini.varadhan@oracle.com>2015-04-09 15:33:31 -0400
committerDavid S. Miller <davem@davemloft.net>2015-04-18 12:32:59 -0700
commitbb620c3d3925aec0ed4f21010c86df08ec18a8c7 (patch)
treef3adfa12c95f9aa6dd844fd47fd486332bbd0a2e /arch/sparc/kernel/pci_sun4v.c
parentff7d37a502022149655c18035b99a53391be0383 (diff)
sparc: Make sparc64 use scalable lib/iommu-common.c functions
In iperf experiments running linux as the Tx side (TCP client) with 10 threads results in a severe performance drop when TSO is disabled, indicating a weakness in the software that can be avoided by using the scalable IOMMU arena DMA allocation. Baseline numbers before this patch: with default settings (TSO enabled) : 9-9.5 Gbps Disable TSO using ethtool- drops badly: 2-3 Gbps. After this patch, iperf client with 10 threads, can give a throughput of at least 8.5 Gbps, even when TSO is disabled. Signed-off-by: Sowmini Varadhan <sowmini.varadhan@oracle.com> Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: David S. Miller <davem@davemloft.net>
Diffstat (limited to 'arch/sparc/kernel/pci_sun4v.c')
-rw-r--r--arch/sparc/kernel/pci_sun4v.c183
1 files changed, 82 insertions, 101 deletions
diff --git a/arch/sparc/kernel/pci_sun4v.c b/arch/sparc/kernel/pci_sun4v.c
index 47ddbd496a1e..d2fe57dad433 100644
--- a/arch/sparc/kernel/pci_sun4v.c
+++ b/arch/sparc/kernel/pci_sun4v.c
@@ -15,6 +15,7 @@
#include <linux/export.h>
#include <linux/log2.h>
#include <linux/of_device.h>
+#include <linux/iommu-common.h>
#include <asm/iommu.h>
#include <asm/irq.h>
@@ -155,15 +156,13 @@ static void *dma_4v_alloc_coherent(struct device *dev, size_t size,
iommu = dev->archdata.iommu;
- spin_lock_irqsave(&iommu->lock, flags);
- entry = iommu_range_alloc(dev, iommu, npages, NULL);
- spin_unlock_irqrestore(&iommu->lock, flags);
+ entry = iommu_tbl_range_alloc(dev, &iommu->tbl, npages, NULL,
+ (unsigned long)(-1), 0);
if (unlikely(entry == DMA_ERROR_CODE))
goto range_alloc_fail;
- *dma_addrp = (iommu->page_table_map_base +
- (entry << IO_PAGE_SHIFT));
+ *dma_addrp = (iommu->tbl.table_map_base + (entry << IO_PAGE_SHIFT));
ret = (void *) first_page;
first_page = __pa(first_page);
@@ -188,45 +187,46 @@ static void *dma_4v_alloc_coherent(struct device *dev, size_t size,
return ret;
iommu_map_fail:
- /* Interrupts are disabled. */
- spin_lock(&iommu->lock);
- iommu_range_free(iommu, *dma_addrp, npages);
- spin_unlock_irqrestore(&iommu->lock, flags);
+ iommu_tbl_range_free(&iommu->tbl, *dma_addrp, npages, DMA_ERROR_CODE);
range_alloc_fail:
free_pages(first_page, order);
return NULL;
}
+static void dma_4v_iommu_demap(void *demap_arg, unsigned long entry,
+ unsigned long npages)
+{
+ u32 devhandle = *(u32 *)demap_arg;
+ unsigned long num, flags;
+
+ local_irq_save(flags);
+ do {
+ num = pci_sun4v_iommu_demap(devhandle,
+ HV_PCI_TSBID(0, entry),
+ npages);
+
+ entry += num;
+ npages -= num;
+ } while (npages != 0);
+ local_irq_restore(flags);
+}
+
static void dma_4v_free_coherent(struct device *dev, size_t size, void *cpu,
dma_addr_t dvma, struct dma_attrs *attrs)
{
struct pci_pbm_info *pbm;
struct iommu *iommu;
- unsigned long flags, order, npages, entry;
+ unsigned long order, npages, entry;
u32 devhandle;
npages = IO_PAGE_ALIGN(size) >> IO_PAGE_SHIFT;
iommu = dev->archdata.iommu;
pbm = dev->archdata.host_controller;
devhandle = pbm->devhandle;
- entry = ((dvma - iommu->page_table_map_base) >> IO_PAGE_SHIFT);
-
- spin_lock_irqsave(&iommu->lock, flags);
-
- iommu_range_free(iommu, dvma, npages);
-
- do {
- unsigned long num;
-
- num = pci_sun4v_iommu_demap(devhandle, HV_PCI_TSBID(0, entry),
- npages);
- entry += num;
- npages -= num;
- } while (npages != 0);
-
- spin_unlock_irqrestore(&iommu->lock, flags);
-
+ entry = ((dvma - iommu->tbl.table_map_base) >> IO_PAGE_SHIFT);
+ dma_4v_iommu_demap(&devhandle, entry, npages);
+ iommu_tbl_range_free(&iommu->tbl, dvma, npages, DMA_ERROR_CODE);
order = get_order(size);
if (order < 10)
free_pages((unsigned long)cpu, order);
@@ -253,15 +253,13 @@ static dma_addr_t dma_4v_map_page(struct device *dev, struct page *page,
npages = IO_PAGE_ALIGN(oaddr + sz) - (oaddr & IO_PAGE_MASK);
npages >>= IO_PAGE_SHIFT;
- spin_lock_irqsave(&iommu->lock, flags);
- entry = iommu_range_alloc(dev, iommu, npages, NULL);
- spin_unlock_irqrestore(&iommu->lock, flags);
+ entry = iommu_tbl_range_alloc(dev, &iommu->tbl, npages, NULL,
+ (unsigned long)(-1), 0);
if (unlikely(entry == DMA_ERROR_CODE))
goto bad;
- bus_addr = (iommu->page_table_map_base +
- (entry << IO_PAGE_SHIFT));
+ bus_addr = (iommu->tbl.table_map_base + (entry << IO_PAGE_SHIFT));
ret = bus_addr | (oaddr & ~IO_PAGE_MASK);
base_paddr = __pa(oaddr & IO_PAGE_MASK);
prot = HV_PCI_MAP_ATTR_READ;
@@ -290,11 +288,7 @@ bad:
return DMA_ERROR_CODE;
iommu_map_fail:
- /* Interrupts are disabled. */
- spin_lock(&iommu->lock);
- iommu_range_free(iommu, bus_addr, npages);
- spin_unlock_irqrestore(&iommu->lock, flags);
-
+ iommu_tbl_range_free(&iommu->tbl, bus_addr, npages, DMA_ERROR_CODE);
return DMA_ERROR_CODE;
}
@@ -304,7 +298,7 @@ static void dma_4v_unmap_page(struct device *dev, dma_addr_t bus_addr,
{
struct pci_pbm_info *pbm;
struct iommu *iommu;
- unsigned long flags, npages;
+ unsigned long npages;
long entry;
u32 devhandle;
@@ -321,22 +315,9 @@ static void dma_4v_unmap_page(struct device *dev, dma_addr_t bus_addr,
npages = IO_PAGE_ALIGN(bus_addr + sz) - (bus_addr & IO_PAGE_MASK);
npages >>= IO_PAGE_SHIFT;
bus_addr &= IO_PAGE_MASK;
-
- spin_lock_irqsave(&iommu->lock, flags);
-
- iommu_range_free(iommu, bus_addr, npages);
-
- entry = (bus_addr - iommu->page_table_map_base) >> IO_PAGE_SHIFT;
- do {
- unsigned long num;
-
- num = pci_sun4v_iommu_demap(devhandle, HV_PCI_TSBID(0, entry),
- npages);
- entry += num;
- npages -= num;
- } while (npages != 0);
-
- spin_unlock_irqrestore(&iommu->lock, flags);
+ entry = (bus_addr - iommu->tbl.table_map_base) >> IO_PAGE_SHIFT;
+ dma_4v_iommu_demap(&devhandle, entry, npages);
+ iommu_tbl_range_free(&iommu->tbl, bus_addr, npages, DMA_ERROR_CODE);
}
static int dma_4v_map_sg(struct device *dev, struct scatterlist *sglist,
@@ -371,14 +352,14 @@ static int dma_4v_map_sg(struct device *dev, struct scatterlist *sglist,
/* Init first segment length for backout at failure */
outs->dma_length = 0;
- spin_lock_irqsave(&iommu->lock, flags);
+ local_irq_save(flags);
iommu_batch_start(dev, prot, ~0UL);
max_seg_size = dma_get_max_seg_size(dev);
seg_boundary_size = ALIGN(dma_get_seg_boundary(dev) + 1,
IO_PAGE_SIZE) >> IO_PAGE_SHIFT;
- base_shift = iommu->page_table_map_base >> IO_PAGE_SHIFT;
+ base_shift = iommu->tbl.table_map_base >> IO_PAGE_SHIFT;
for_each_sg(sglist, s, nelems, i) {
unsigned long paddr, npages, entry, out_entry = 0, slen;
@@ -391,7 +372,8 @@ static int dma_4v_map_sg(struct device *dev, struct scatterlist *sglist,
/* Allocate iommu entries for that segment */
paddr = (unsigned long) SG_ENT_PHYS_ADDRESS(s);
npages = iommu_num_pages(paddr, slen, IO_PAGE_SIZE);
- entry = iommu_range_alloc(dev, iommu, npages, &handle);
+ entry = iommu_tbl_range_alloc(dev, &iommu->tbl, npages,
+ &handle, (unsigned long)(-1), 0);
/* Handle failure */
if (unlikely(entry == DMA_ERROR_CODE)) {
@@ -404,8 +386,7 @@ static int dma_4v_map_sg(struct device *dev, struct scatterlist *sglist,
iommu_batch_new_entry(entry);
/* Convert entry to a dma_addr_t */
- dma_addr = iommu->page_table_map_base +
- (entry << IO_PAGE_SHIFT);
+ dma_addr = iommu->tbl.table_map_base + (entry << IO_PAGE_SHIFT);
dma_addr |= (s->offset & ~IO_PAGE_MASK);
/* Insert into HW table */
@@ -451,7 +432,7 @@ static int dma_4v_map_sg(struct device *dev, struct scatterlist *sglist,
if (unlikely(err < 0L))
goto iommu_map_failed;
- spin_unlock_irqrestore(&iommu->lock, flags);
+ local_irq_restore(flags);
if (outcount < incount) {
outs = sg_next(outs);
@@ -469,7 +450,8 @@ iommu_map_failed:
vaddr = s->dma_address & IO_PAGE_MASK;
npages = iommu_num_pages(s->dma_address, s->dma_length,
IO_PAGE_SIZE);
- iommu_range_free(iommu, vaddr, npages);
+ iommu_tbl_range_free(&iommu->tbl, vaddr, npages,
+ DMA_ERROR_CODE);
/* XXX demap? XXX */
s->dma_address = DMA_ERROR_CODE;
s->dma_length = 0;
@@ -477,7 +459,7 @@ iommu_map_failed:
if (s == outs)
break;
}
- spin_unlock_irqrestore(&iommu->lock, flags);
+ local_irq_restore(flags);
return 0;
}
@@ -489,7 +471,7 @@ static void dma_4v_unmap_sg(struct device *dev, struct scatterlist *sglist,
struct pci_pbm_info *pbm;
struct scatterlist *sg;
struct iommu *iommu;
- unsigned long flags;
+ unsigned long flags, entry;
u32 devhandle;
BUG_ON(direction == DMA_NONE);
@@ -498,33 +480,27 @@ static void dma_4v_unmap_sg(struct device *dev, struct scatterlist *sglist,
pbm = dev->archdata.host_controller;
devhandle = pbm->devhandle;
- spin_lock_irqsave(&iommu->lock, flags);
+ local_irq_save(flags);
sg = sglist;
while (nelems--) {
dma_addr_t dma_handle = sg->dma_address;
unsigned int len = sg->dma_length;
- unsigned long npages, entry;
+ unsigned long npages;
+ struct iommu_map_table *tbl = &iommu->tbl;
+ unsigned long shift = IO_PAGE_SHIFT;
if (!len)
break;
npages = iommu_num_pages(dma_handle, len, IO_PAGE_SIZE);
- iommu_range_free(iommu, dma_handle, npages);
-
- entry = ((dma_handle - iommu->page_table_map_base) >> IO_PAGE_SHIFT);
- while (npages) {
- unsigned long num;
-
- num = pci_sun4v_iommu_demap(devhandle, HV_PCI_TSBID(0, entry),
- npages);
- entry += num;
- npages -= num;
- }
-
+ entry = ((dma_handle - tbl->table_map_base) >> shift);
+ dma_4v_iommu_demap(&devhandle, entry, npages);
+ iommu_tbl_range_free(&iommu->tbl, dma_handle, npages,
+ DMA_ERROR_CODE);
sg = sg_next(sg);
}
- spin_unlock_irqrestore(&iommu->lock, flags);
+ local_irq_restore(flags);
}
static struct dma_map_ops sun4v_dma_ops = {
@@ -550,30 +526,33 @@ static void pci_sun4v_scan_bus(struct pci_pbm_info *pbm, struct device *parent)
}
static unsigned long probe_existing_entries(struct pci_pbm_info *pbm,
- struct iommu *iommu)
+ struct iommu_map_table *iommu)
{
- struct iommu_arena *arena = &iommu->arena;
- unsigned long i, cnt = 0;
+ struct iommu_pool *pool;
+ unsigned long i, pool_nr, cnt = 0;
u32 devhandle;
devhandle = pbm->devhandle;
- for (i = 0; i < arena->limit; i++) {
- unsigned long ret, io_attrs, ra;
-
- ret = pci_sun4v_iommu_getmap(devhandle,
- HV_PCI_TSBID(0, i),
- &io_attrs, &ra);
- if (ret == HV_EOK) {
- if (page_in_phys_avail(ra)) {
- pci_sun4v_iommu_demap(devhandle,
- HV_PCI_TSBID(0, i), 1);
- } else {
- cnt++;
- __set_bit(i, arena->map);
+ for (pool_nr = 0; pool_nr < iommu->nr_pools; pool_nr++) {
+ pool = &(iommu->pools[pool_nr]);
+ for (i = pool->start; i <= pool->end; i++) {
+ unsigned long ret, io_attrs, ra;
+
+ ret = pci_sun4v_iommu_getmap(devhandle,
+ HV_PCI_TSBID(0, i),
+ &io_attrs, &ra);
+ if (ret == HV_EOK) {
+ if (page_in_phys_avail(ra)) {
+ pci_sun4v_iommu_demap(devhandle,
+ HV_PCI_TSBID(0,
+ i), 1);
+ } else {
+ cnt++;
+ __set_bit(i, iommu->map);
+ }
}
}
}
-
return cnt;
}
@@ -603,20 +582,22 @@ static int pci_sun4v_iommu_init(struct pci_pbm_info *pbm)
/* Setup initial software IOMMU state. */
spin_lock_init(&iommu->lock);
iommu->ctx_lowest_free = 1;
- iommu->page_table_map_base = dma_offset;
+ iommu->tbl.table_map_base = dma_offset;
iommu->dma_addr_mask = dma_mask;
/* Allocate and initialize the free area map. */
sz = (num_tsb_entries + 7) / 8;
sz = (sz + 7UL) & ~7UL;
- iommu->arena.map = kzalloc(sz, GFP_KERNEL);
- if (!iommu->arena.map) {
+ iommu->tbl.map = kzalloc(sz, GFP_KERNEL);
+ if (!iommu->tbl.map) {
printk(KERN_ERR PFX "Error, kmalloc(arena.map) failed.\n");
return -ENOMEM;
}
- iommu->arena.limit = num_tsb_entries;
-
- sz = probe_existing_entries(pbm, iommu);
+ iommu_tbl_pool_init(&iommu->tbl, num_tsb_entries, IO_PAGE_SHIFT,
+ NULL, false /* no large_pool */,
+ 0 /* default npools */,
+ false /* want span boundary checking */);
+ sz = probe_existing_entries(pbm, &iommu->tbl);
if (sz)
printk("%s: Imported %lu TSB entries from OBP\n",
pbm->name, sz);