*/
#include <linux/mm.h>
+#include <linux/scatterlist.h>
+#include <linux/sched.h>
+
+#include <asm/page.h>
#include "mthca_memfree.h"
#include "mthca_dev.h"
};
struct mthca_user_db_table {
- struct semaphore mutex;
+ struct mutex mutex;
struct {
u64 uvirt;
struct scatterlist mem;
} page[0];
};
-void mthca_free_icm(struct mthca_dev *dev, struct mthca_icm *icm)
+static void mthca_free_icm_pages(struct mthca_dev *dev, struct mthca_icm_chunk *chunk)
{
- struct mthca_icm_chunk *chunk, *tmp;
int i;
+ if (chunk->nsg > 0)
+ pci_unmap_sg(dev->pdev, chunk->mem, chunk->npages,
+ PCI_DMA_BIDIRECTIONAL);
+
+ for (i = 0; i < chunk->npages; ++i)
+ __free_pages(sg_page(&chunk->mem[i]),
+ get_order(chunk->mem[i].length));
+}
+
+static void mthca_free_icm_coherent(struct mthca_dev *dev, struct mthca_icm_chunk *chunk)
+{
+ int i;
+
+ for (i = 0; i < chunk->npages; ++i) {
+ dma_free_coherent(&dev->pdev->dev, chunk->mem[i].length,
+ lowmem_page_address(sg_page(&chunk->mem[i])),
+ sg_dma_address(&chunk->mem[i]));
+ }
+}
+
+void mthca_free_icm(struct mthca_dev *dev, struct mthca_icm *icm, int coherent)
+{
+ struct mthca_icm_chunk *chunk, *tmp;
+
if (!icm)
return;
list_for_each_entry_safe(chunk, tmp, &icm->chunk_list, list) {
- if (chunk->nsg > 0)
- pci_unmap_sg(dev->pdev, chunk->mem, chunk->npages,
- PCI_DMA_BIDIRECTIONAL);
-
- for (i = 0; i < chunk->npages; ++i)
- __free_pages(chunk->mem[i].page,
- get_order(chunk->mem[i].length));
+ if (coherent)
+ mthca_free_icm_coherent(dev, chunk);
+ else
+ mthca_free_icm_pages(dev, chunk);
kfree(chunk);
}
kfree(icm);
}
+static int mthca_alloc_icm_pages(struct scatterlist *mem, int order, gfp_t gfp_mask)
+{
+ struct page *page;
+
+ page = alloc_pages(gfp_mask, order);
+ if (!page)
+ return -ENOMEM;
+
+ sg_set_page(mem, page, PAGE_SIZE << order, 0);
+ return 0;
+}
+
+static int mthca_alloc_icm_coherent(struct device *dev, struct scatterlist *mem,
+ int order, gfp_t gfp_mask)
+{
+ void *buf = dma_alloc_coherent(dev, PAGE_SIZE << order, &sg_dma_address(mem),
+ gfp_mask);
+ if (!buf)
+ return -ENOMEM;
+
+ sg_set_buf(mem, buf, PAGE_SIZE << order);
+ BUG_ON(mem->offset);
+ sg_dma_len(mem) = PAGE_SIZE << order;
+ return 0;
+}
+
struct mthca_icm *mthca_alloc_icm(struct mthca_dev *dev, int npages,
- unsigned int gfp_mask)
+ gfp_t gfp_mask, int coherent)
{
struct mthca_icm *icm;
struct mthca_icm_chunk *chunk = NULL;
int cur_order;
+ int ret;
+
+ /* We use sg_set_buf for coherent allocs, which assumes low memory */
+ BUG_ON(coherent && (gfp_mask & __GFP_HIGHMEM));
icm = kmalloc(sizeof *icm, gfp_mask & ~(__GFP_HIGHMEM | __GFP_NOWARN));
if (!icm)
if (!chunk)
goto fail;
+ sg_init_table(chunk->mem, MTHCA_ICM_CHUNK_LEN);
chunk->npages = 0;
chunk->nsg = 0;
list_add_tail(&chunk->list, &icm->chunk_list);
while (1 << cur_order > npages)
--cur_order;
- chunk->mem[chunk->npages].page = alloc_pages(gfp_mask, cur_order);
- if (chunk->mem[chunk->npages].page) {
- chunk->mem[chunk->npages].length = PAGE_SIZE << cur_order;
- chunk->mem[chunk->npages].offset = 0;
+ if (coherent)
+ ret = mthca_alloc_icm_coherent(&dev->pdev->dev,
+ &chunk->mem[chunk->npages],
+ cur_order, gfp_mask);
+ else
+ ret = mthca_alloc_icm_pages(&chunk->mem[chunk->npages],
+ cur_order, gfp_mask);
+
+ if (!ret) {
+ ++chunk->npages;
- if (++chunk->npages == MTHCA_ICM_CHUNK_LEN) {
+ if (coherent)
+ ++chunk->nsg;
+ else if (chunk->npages == MTHCA_ICM_CHUNK_LEN) {
chunk->nsg = pci_map_sg(dev->pdev, chunk->mem,
chunk->npages,
PCI_DMA_BIDIRECTIONAL);
if (chunk->nsg <= 0)
goto fail;
+ }
+ if (chunk->npages == MTHCA_ICM_CHUNK_LEN)
chunk = NULL;
- }
npages -= 1 << cur_order;
} else {
}
}
- if (chunk) {
+ if (!coherent && chunk) {
chunk->nsg = pci_map_sg(dev->pdev, chunk->mem,
chunk->npages,
PCI_DMA_BIDIRECTIONAL);
return icm;
fail:
- mthca_free_icm(dev, icm);
+ mthca_free_icm(dev, icm, coherent);
return NULL;
}
int ret = 0;
u8 status;
- down(&table->mutex);
+ mutex_lock(&table->mutex);
if (table->icm[i]) {
++table->icm[i]->refcount;
table->icm[i] = mthca_alloc_icm(dev, MTHCA_TABLE_CHUNK_SIZE >> PAGE_SHIFT,
(table->lowmem ? GFP_KERNEL : GFP_HIGHUSER) |
- __GFP_NOWARN);
+ __GFP_NOWARN, table->coherent);
if (!table->icm[i]) {
ret = -ENOMEM;
goto out;
if (mthca_MAP_ICM(dev, table->icm[i], table->virt + i * MTHCA_TABLE_CHUNK_SIZE,
&status) || status) {
- mthca_free_icm(dev, table->icm[i]);
+ mthca_free_icm(dev, table->icm[i], table->coherent);
table->icm[i] = NULL;
ret = -ENOMEM;
goto out;
++table->icm[i]->refcount;
out:
- up(&table->mutex);
+ mutex_unlock(&table->mutex);
return ret;
}
i = (obj & (table->num_obj - 1)) * table->obj_size / MTHCA_TABLE_CHUNK_SIZE;
- down(&table->mutex);
+ mutex_lock(&table->mutex);
if (--table->icm[i]->refcount == 0) {
mthca_UNMAP_ICM(dev, table->virt + i * MTHCA_TABLE_CHUNK_SIZE,
- MTHCA_TABLE_CHUNK_SIZE >> 12, &status);
- mthca_free_icm(dev, table->icm[i]);
+ MTHCA_TABLE_CHUNK_SIZE / MTHCA_ICM_PAGE_SIZE,
+ &status);
+ mthca_free_icm(dev, table->icm[i], table->coherent);
table->icm[i] = NULL;
}
- up(&table->mutex);
+ mutex_unlock(&table->mutex);
}
-void *mthca_table_find(struct mthca_icm_table *table, int obj)
+void *mthca_table_find(struct mthca_icm_table *table, int obj, dma_addr_t *dma_handle)
{
- int idx, offset, i;
+ int idx, offset, dma_offset, i;
struct mthca_icm_chunk *chunk;
struct mthca_icm *icm;
struct page *page = NULL;
if (!table->lowmem)
return NULL;
- down(&table->mutex);
+ mutex_lock(&table->mutex);
idx = (obj & (table->num_obj - 1)) * table->obj_size;
icm = table->icm[idx / MTHCA_TABLE_CHUNK_SIZE];
- offset = idx % MTHCA_TABLE_CHUNK_SIZE;
+ dma_offset = offset = idx % MTHCA_TABLE_CHUNK_SIZE;
if (!icm)
goto out;
list_for_each_entry(chunk, &icm->chunk_list, list) {
for (i = 0; i < chunk->npages; ++i) {
- if (chunk->mem[i].length >= offset) {
- page = chunk->mem[i].page;
- break;
+ if (dma_handle && dma_offset >= 0) {
+ if (sg_dma_len(&chunk->mem[i]) > dma_offset)
+ *dma_handle = sg_dma_address(&chunk->mem[i]) +
+ dma_offset;
+ dma_offset -= sg_dma_len(&chunk->mem[i]);
+ }
+ /* DMA mapping can merge pages but not split them,
+ * so if we found the page, dma_handle has already
+ * been assigned to. */
+ if (chunk->mem[i].length > offset) {
+ page = sg_page(&chunk->mem[i]);
+ goto out;
}
offset -= chunk->mem[i].length;
}
}
out:
- up(&table->mutex);
+ mutex_unlock(&table->mutex);
return page ? lowmem_page_address(page) + offset : NULL;
}
struct mthca_icm_table *mthca_alloc_icm_table(struct mthca_dev *dev,
u64 virt, int obj_size,
int nobj, int reserved,
- int use_lowmem)
+ int use_lowmem, int use_coherent)
{
struct mthca_icm_table *table;
+ int obj_per_chunk;
int num_icm;
unsigned chunk_size;
int i;
u8 status;
- num_icm = obj_size * nobj / MTHCA_TABLE_CHUNK_SIZE;
+ obj_per_chunk = MTHCA_TABLE_CHUNK_SIZE / obj_size;
+ num_icm = DIV_ROUND_UP(nobj, obj_per_chunk);
table = kmalloc(sizeof *table + num_icm * sizeof *table->icm, GFP_KERNEL);
if (!table)
table->num_obj = nobj;
table->obj_size = obj_size;
table->lowmem = use_lowmem;
- init_MUTEX(&table->mutex);
+ table->coherent = use_coherent;
+ mutex_init(&table->mutex);
for (i = 0; i < num_icm; ++i)
table->icm[i] = NULL;
table->icm[i] = mthca_alloc_icm(dev, chunk_size >> PAGE_SHIFT,
(use_lowmem ? GFP_KERNEL : GFP_HIGHUSER) |
- __GFP_NOWARN);
+ __GFP_NOWARN, use_coherent);
if (!table->icm[i])
goto err;
if (mthca_MAP_ICM(dev, table->icm[i], virt + i * MTHCA_TABLE_CHUNK_SIZE,
&status) || status) {
- mthca_free_icm(dev, table->icm[i]);
+ mthca_free_icm(dev, table->icm[i], table->coherent);
table->icm[i] = NULL;
goto err;
}
for (i = 0; i < num_icm; ++i)
if (table->icm[i]) {
mthca_UNMAP_ICM(dev, virt + i * MTHCA_TABLE_CHUNK_SIZE,
- MTHCA_TABLE_CHUNK_SIZE >> 12, &status);
- mthca_free_icm(dev, table->icm[i]);
+ MTHCA_TABLE_CHUNK_SIZE / MTHCA_ICM_PAGE_SIZE,
+ &status);
+ mthca_free_icm(dev, table->icm[i], table->coherent);
}
kfree(table);
for (i = 0; i < table->num_icm; ++i)
if (table->icm[i]) {
mthca_UNMAP_ICM(dev, table->virt + i * MTHCA_TABLE_CHUNK_SIZE,
- MTHCA_TABLE_CHUNK_SIZE >> 12, &status);
- mthca_free_icm(dev, table->icm[i]);
+ MTHCA_TABLE_CHUNK_SIZE / MTHCA_ICM_PAGE_SIZE,
+ &status);
+ mthca_free_icm(dev, table->icm[i], table->coherent);
}
kfree(table);
{
return dev->uar_table.uarc_base +
uar->index * dev->uar_table.uarc_size +
- page * 4096;
+ page * MTHCA_ICM_PAGE_SIZE;
}
int mthca_map_user_db(struct mthca_dev *dev, struct mthca_uar *uar,
struct mthca_user_db_table *db_tab, int index, u64 uaddr)
{
+ struct page *pages[1];
int ret = 0;
u8 status;
int i;
if (index < 0 || index > dev->uar_table.uarc_size / 8)
return -EINVAL;
- down(&db_tab->mutex);
+ mutex_lock(&db_tab->mutex);
i = index / MTHCA_DB_REC_PER_PAGE;
}
ret = get_user_pages(current, current->mm, uaddr & PAGE_MASK, 1, 1, 0,
- &db_tab->page[i].mem.page, NULL);
+ pages, NULL);
if (ret < 0)
goto out;
- db_tab->page[i].mem.length = 4096;
- db_tab->page[i].mem.offset = uaddr & ~PAGE_MASK;
+ sg_set_page(&db_tab->page[i].mem, pages[0], MTHCA_ICM_PAGE_SIZE,
+ uaddr & ~PAGE_MASK);
ret = pci_map_sg(dev->pdev, &db_tab->page[i].mem, 1, PCI_DMA_TODEVICE);
if (ret < 0) {
- put_page(db_tab->page[i].mem.page);
+ put_page(pages[0]);
goto out;
}
ret = -EINVAL;
if (ret) {
pci_unmap_sg(dev->pdev, &db_tab->page[i].mem, 1, PCI_DMA_TODEVICE);
- put_page(db_tab->page[i].mem.page);
+ put_page(sg_page(&db_tab->page[i].mem));
goto out;
}
db_tab->page[i].refcount = 1;
out:
- up(&db_tab->mutex);
+ mutex_unlock(&db_tab->mutex);
return ret;
}
* pages until we clean up the whole db table.
*/
- down(&db_tab->mutex);
+ mutex_lock(&db_tab->mutex);
--db_tab->page[index / MTHCA_DB_REC_PER_PAGE].refcount;
- up(&db_tab->mutex);
+ mutex_unlock(&db_tab->mutex);
}
struct mthca_user_db_table *mthca_init_user_db_tab(struct mthca_dev *dev)
if (!mthca_is_memfree(dev))
return NULL;
- npages = dev->uar_table.uarc_size / 4096;
+ npages = dev->uar_table.uarc_size / MTHCA_ICM_PAGE_SIZE;
db_tab = kmalloc(sizeof *db_tab + npages * sizeof *db_tab->page, GFP_KERNEL);
if (!db_tab)
return ERR_PTR(-ENOMEM);
- init_MUTEX(&db_tab->mutex);
+ mutex_init(&db_tab->mutex);
for (i = 0; i < npages; ++i) {
db_tab->page[i].refcount = 0;
db_tab->page[i].uvirt = 0;
+ sg_init_table(&db_tab->page[i].mem, 1);
}
return db_tab;
if (!mthca_is_memfree(dev))
return;
- for (i = 0; i < dev->uar_table.uarc_size / 4096; ++i) {
+ for (i = 0; i < dev->uar_table.uarc_size / MTHCA_ICM_PAGE_SIZE; ++i) {
if (db_tab->page[i].uvirt) {
mthca_UNMAP_ICM(dev, mthca_uarc_virt(dev, uar, i), 1, &status);
pci_unmap_sg(dev->pdev, &db_tab->page[i].mem, 1, PCI_DMA_TODEVICE);
- put_page(db_tab->page[i].mem.page);
+ put_page(sg_page(&db_tab->page[i].mem));
}
}
+
+ kfree(db_tab);
}
-int mthca_alloc_db(struct mthca_dev *dev, int type, u32 qn, __be32 **db)
+int mthca_alloc_db(struct mthca_dev *dev, enum mthca_db_type type,
+ u32 qn, __be32 **db)
{
int group;
int start, end, dir;
int ret = 0;
u8 status;
- down(&dev->db_tab->mutex);
+ mutex_lock(&dev->db_tab->mutex);
switch (type) {
case MTHCA_DB_TYPE_CQ_ARM:
page = dev->db_tab->page + end;
alloc:
- page->db_rec = dma_alloc_coherent(&dev->pdev->dev, 4096,
+ page->db_rec = dma_alloc_coherent(&dev->pdev->dev, MTHCA_ICM_PAGE_SIZE,
&page->mapping, GFP_KERNEL);
if (!page->db_rec) {
ret = -ENOMEM;
goto out;
}
- memset(page->db_rec, 0, 4096);
+ memset(page->db_rec, 0, MTHCA_ICM_PAGE_SIZE);
ret = mthca_MAP_ICM_page(dev, page->mapping,
mthca_uarc_virt(dev, &dev->driver_uar, i), &status);
if (!ret && status)
ret = -EINVAL;
if (ret) {
- dma_free_coherent(&dev->pdev->dev, 4096,
+ dma_free_coherent(&dev->pdev->dev, MTHCA_ICM_PAGE_SIZE,
page->db_rec, page->mapping);
goto out;
}
*db = (__be32 *) &page->db_rec[j];
out:
- up(&dev->db_tab->mutex);
+ mutex_unlock(&dev->db_tab->mutex);
return ret;
}
page = dev->db_tab->page + i;
- down(&dev->db_tab->mutex);
+ mutex_lock(&dev->db_tab->mutex);
page->db_rec[j] = 0;
if (i >= dev->db_tab->min_group2)
i >= dev->db_tab->max_group1 - 1) {
mthca_UNMAP_ICM(dev, mthca_uarc_virt(dev, &dev->driver_uar, i), 1, &status);
- dma_free_coherent(&dev->pdev->dev, 4096,
+ dma_free_coherent(&dev->pdev->dev, MTHCA_ICM_PAGE_SIZE,
page->db_rec, page->mapping);
page->db_rec = NULL;
++dev->db_tab->min_group2;
}
- up(&dev->db_tab->mutex);
+ mutex_unlock(&dev->db_tab->mutex);
}
int mthca_init_db_tab(struct mthca_dev *dev)
if (!dev->db_tab)
return -ENOMEM;
- init_MUTEX(&dev->db_tab->mutex);
+ mutex_init(&dev->db_tab->mutex);
- dev->db_tab->npages = dev->uar_table.uarc_size / 4096;
+ dev->db_tab->npages = dev->uar_table.uarc_size / MTHCA_ICM_PAGE_SIZE;
dev->db_tab->max_group1 = 0;
dev->db_tab->min_group2 = dev->db_tab->npages - 1;
mthca_UNMAP_ICM(dev, mthca_uarc_virt(dev, &dev->driver_uar, i), 1, &status);
- dma_free_coherent(&dev->pdev->dev, 4096,
+ dma_free_coherent(&dev->pdev->dev, MTHCA_ICM_PAGE_SIZE,
dev->db_tab->page[i].db_rec,
dev->db_tab->page[i].mapping);
}
git://ftp.safe.ca / safe / jmp / linux-2.6 / blobdiff