#include <rte_string_fns.h>
#include <rte_spinlock.h>
#include <rte_tailq.h>
+#include <rte_function_versioning.h>
#include "rte_mempool.h"
+#include "rte_mempool_trace.h"
TAILQ_HEAD(rte_mempool_list, rte_tailq_entry);
#define CALC_CACHE_FLUSHTHRESH(c) \
((typeof(c))((c) * CACHE_FLUSHTHRESH_MULTIPLIER))
+#if defined(RTE_ARCH_X86)
/*
* return the greatest common divisor between a and b (fast algorithm)
*
}
/*
- * Depending on memory configuration, objects addresses are spread
+ * Depending on memory configuration on x86 arch, objects addresses are spread
* between channels and ranks in RAM: the pool allocator will add
* padding between objects. This function return the new size of the
* object.
*/
-static unsigned optimize_object_size(unsigned obj_size)
+static unsigned int
+arch_mem_object_align(unsigned int obj_size)
{
unsigned nrank, nchan;
unsigned new_obj_size;
new_obj_size++;
return new_obj_size * RTE_MEMPOOL_ALIGN;
}
+#else
+static unsigned int
+arch_mem_object_align(unsigned int obj_size)
+{
+ return obj_size;
+}
+#endif
struct pagesz_walk_arg {
int socket_id;
*/
if ((flags & MEMPOOL_F_NO_SPREAD) == 0) {
unsigned new_size;
- new_size = optimize_object_size(sz->header_size + sz->elt_size +
- sz->trailer_size);
+ new_size = arch_mem_object_align
+ (sz->header_size + sz->elt_size + sz->trailer_size);
sz->trailer_size = new_size - sz->header_size - sz->elt_size;
}
return 0;
}
+__vsym int
+rte_mempool_populate_iova_v21(struct rte_mempool *mp, char *vaddr,
+ rte_iova_t iova, size_t len, rte_mempool_memchunk_free_cb_t *free_cb,
+ void *opaque);
+
/* Add objects in the pool, using a physically contiguous memory
* zone. Return the number of objects added, or a negative value
* on error.
*/
-int
-rte_mempool_populate_iova(struct rte_mempool *mp, char *vaddr,
+__vsym int
+rte_mempool_populate_iova_v21(struct rte_mempool *mp, char *vaddr,
rte_iova_t iova, size_t len, rte_mempool_memchunk_free_cb_t *free_cb,
void *opaque)
{
if (mp->flags & MEMPOOL_F_NO_CACHE_ALIGN)
off = RTE_PTR_ALIGN_CEIL(vaddr, 8) - vaddr;
else
- off = RTE_PTR_ALIGN_CEIL(vaddr, RTE_CACHE_LINE_SIZE) - vaddr;
+ off = RTE_PTR_ALIGN_CEIL(vaddr, RTE_MEMPOOL_ALIGN) - vaddr;
if (off > len) {
- ret = -EINVAL;
+ ret = 0;
goto fail;
}
/* not enough room to store one object */
if (i == 0) {
- ret = -EINVAL;
+ ret = 0;
goto fail;
}
STAILQ_INSERT_TAIL(&mp->mem_list, memhdr, next);
mp->nb_mem_chunks++;
+
+ rte_mempool_trace_populate_iova(mp, vaddr, iova, len, free_cb, opaque);
return i;
fail:
return ret;
}
+BIND_DEFAULT_SYMBOL(rte_mempool_populate_iova, _v21, 21);
+MAP_STATIC_SYMBOL(
+ int rte_mempool_populate_iova(struct rte_mempool *mp, char *vaddr,
+ rte_iova_t iova, size_t len,
+ rte_mempool_memchunk_free_cb_t *free_cb,
+ void *opaque),
+ rte_mempool_populate_iova_v21);
+
+__vsym int
+rte_mempool_populate_iova_v20(struct rte_mempool *mp, char *vaddr,
+ rte_iova_t iova, size_t len, rte_mempool_memchunk_free_cb_t *free_cb,
+ void *opaque);
+
+__vsym int
+rte_mempool_populate_iova_v20(struct rte_mempool *mp, char *vaddr,
+ rte_iova_t iova, size_t len, rte_mempool_memchunk_free_cb_t *free_cb,
+ void *opaque)
+{
+ int ret;
+
+ ret = rte_mempool_populate_iova_v21(mp, vaddr, iova, len, free_cb,
+ opaque);
+ if (ret == 0)
+ ret = -EINVAL;
+
+ return ret;
+}
+VERSION_SYMBOL(rte_mempool_populate_iova, _v20, 20.0);
+
+static rte_iova_t
+get_iova(void *addr)
+{
+ struct rte_memseg *ms;
+
+ /* try registered memory first */
+ ms = rte_mem_virt2memseg(addr, NULL);
+ if (ms == NULL || ms->iova == RTE_BAD_IOVA)
+ /* fall back to actual physical address */
+ return rte_mem_virt2iova(addr);
+ return ms->iova + RTE_PTR_DIFF(addr, ms->addr);
+}
+
+__vsym int
+rte_mempool_populate_virt_v21(struct rte_mempool *mp, char *addr,
+ size_t len, size_t pg_sz, rte_mempool_memchunk_free_cb_t *free_cb,
+ void *opaque);
+
/* Populate the mempool with a virtual area. Return the number of
* objects added, or a negative value on error.
*/
-int
-rte_mempool_populate_virt(struct rte_mempool *mp, char *addr,
+__vsym int
+rte_mempool_populate_virt_v21(struct rte_mempool *mp, char *addr,
size_t len, size_t pg_sz, rte_mempool_memchunk_free_cb_t *free_cb,
void *opaque)
{
for (off = 0; off < len &&
mp->populated_size < mp->size; off += phys_len) {
- iova = rte_mem_virt2iova(addr + off);
-
- if (iova == RTE_BAD_IOVA && rte_eal_has_hugepages()) {
- ret = -EINVAL;
- goto fail;
- }
+ iova = get_iova(addr + off);
/* populate with the largest group of contiguous pages */
for (phys_len = RTE_MIN(
phys_len = RTE_MIN(phys_len + pg_sz, len - off)) {
rte_iova_t iova_tmp;
- iova_tmp = rte_mem_virt2iova(addr + off + phys_len);
+ iova_tmp = get_iova(addr + off + phys_len);
if (iova_tmp == RTE_BAD_IOVA ||
iova_tmp != iova + phys_len)
break;
}
- ret = rte_mempool_populate_iova(mp, addr + off, iova,
+ ret = rte_mempool_populate_iova_v21(mp, addr + off, iova,
phys_len, free_cb, opaque);
+ if (ret == 0)
+ continue;
if (ret < 0)
goto fail;
/* no need to call the free callback for next chunks */
cnt += ret;
}
+ rte_mempool_trace_populate_virt(mp, addr, len, pg_sz, free_cb, opaque);
return cnt;
fail:
rte_mempool_free_memchunks(mp);
return ret;
}
+BIND_DEFAULT_SYMBOL(rte_mempool_populate_virt, _v21, 21);
+MAP_STATIC_SYMBOL(
+ int rte_mempool_populate_virt(struct rte_mempool *mp,
+ char *addr, size_t len, size_t pg_sz,
+ rte_mempool_memchunk_free_cb_t *free_cb,
+ void *opaque),
+ rte_mempool_populate_virt_v21);
+
+__vsym int
+rte_mempool_populate_virt_v20(struct rte_mempool *mp, char *addr,
+ size_t len, size_t pg_sz, rte_mempool_memchunk_free_cb_t *free_cb,
+ void *opaque);
+
+__vsym int
+rte_mempool_populate_virt_v20(struct rte_mempool *mp, char *addr,
+ size_t len, size_t pg_sz, rte_mempool_memchunk_free_cb_t *free_cb,
+ void *opaque)
+{
+ int ret;
+
+ ret = rte_mempool_populate_virt_v21(mp, addr, len, pg_sz,
+ free_cb, opaque);
+
+ if (ret == 0)
+ ret = -EINVAL;
+
+ return ret;
+}
+VERSION_SYMBOL(rte_mempool_populate_virt, _v20, 20.0);
+
+/* Get the minimal page size used in a mempool before populating it. */
+int
+rte_mempool_get_page_size(struct rte_mempool *mp, size_t *pg_sz)
+{
+ bool need_iova_contig_obj;
+ bool alloc_in_ext_mem;
+ int ret;
+
+ /* check if we can retrieve a valid socket ID */
+ ret = rte_malloc_heap_socket_is_external(mp->socket_id);
+ if (ret < 0)
+ return -EINVAL;
+ alloc_in_ext_mem = (ret == 1);
+ need_iova_contig_obj = !(mp->flags & MEMPOOL_F_NO_IOVA_CONTIG);
+
+ if (!need_iova_contig_obj)
+ *pg_sz = 0;
+ else if (rte_eal_has_hugepages() || alloc_in_ext_mem)
+ *pg_sz = get_min_page_size(mp->socket_id);
+ else
+ *pg_sz = getpagesize();
+
+ rte_mempool_trace_get_page_size(mp, *pg_sz);
+ return 0;
+}
/* Default function to populate the mempool: allocate memory in memzones,
* and populate them. Return the number of objects added, or a negative
char mz_name[RTE_MEMZONE_NAMESIZE];
const struct rte_memzone *mz;
ssize_t mem_size;
- size_t align, pg_sz, pg_shift;
+ size_t align, pg_sz, pg_shift = 0;
rte_iova_t iova;
unsigned mz_id, n;
int ret;
bool need_iova_contig_obj;
- bool alloc_in_ext_mem;
+ size_t max_alloc_size = SIZE_MAX;
ret = mempool_ops_alloc_once(mp);
if (ret != 0)
* then just set page shift and page size to 0, because the user has
* indicated that there's no need to care about anything.
*
- * if we do need contiguous objects, there is also an option to reserve
- * the entire mempool memory as one contiguous block of memory, in
- * which case the page shift and alignment wouldn't matter as well.
+ * if we do need contiguous objects (if a mempool driver has its
+ * own calc_size() method returning min_chunk_size = mem_size),
+ * there is also an option to reserve the entire mempool memory
+ * as one contiguous block of memory.
*
* if we require contiguous objects, but not necessarily the entire
- * mempool reserved space to be contiguous, then there are two options.
- *
- * if our IO addresses are virtual, not actual physical (IOVA as VA
- * case), then no page shift needed - our memory allocation will give us
- * contiguous IO memory as far as the hardware is concerned, so
- * act as if we're getting contiguous memory.
+ * mempool reserved space to be contiguous, pg_sz will be != 0,
+ * and the default ops->populate() will take care of not placing
+ * objects across pages.
*
* if our IO addresses are physical, we may get memory from bigger
* pages, or we might get memory from smaller pages, and how much of it
*
* If we fail to get enough contiguous memory, then we'll go and
* reserve space in smaller chunks.
- *
- * We also have to take into account the fact that memory that we're
- * going to allocate from can belong to an externally allocated memory
- * area, in which case the assumption of IOVA as VA mode being
- * synonymous with IOVA contiguousness will not hold.
*/
- /* check if we can retrieve a valid socket ID */
- ret = rte_malloc_heap_socket_is_external(mp->socket_id);
- if (ret < 0)
- return -EINVAL;
- alloc_in_ext_mem = (ret == 1);
need_iova_contig_obj = !(mp->flags & MEMPOOL_F_NO_IOVA_CONTIG);
+ ret = rte_mempool_get_page_size(mp, &pg_sz);
+ if (ret < 0)
+ return ret;
- if (!need_iova_contig_obj) {
- pg_sz = 0;
- pg_shift = 0;
- } else if (!alloc_in_ext_mem && rte_eal_iova_mode() == RTE_IOVA_VA) {
- pg_sz = 0;
- pg_shift = 0;
- } else if (rte_eal_has_hugepages() || alloc_in_ext_mem) {
- pg_sz = get_min_page_size(mp->socket_id);
- pg_shift = rte_bsf32(pg_sz);
- } else {
- pg_sz = getpagesize();
+ if (pg_sz != 0)
pg_shift = rte_bsf32(pg_sz);
- }
for (mz_id = 0, n = mp->size; n > 0; mz_id++, n -= ret) {
size_t min_chunk_size;
if (min_chunk_size == (size_t)mem_size)
mz_flags |= RTE_MEMZONE_IOVA_CONTIG;
- mz = rte_memzone_reserve_aligned(mz_name, mem_size,
+ /* Allocate a memzone, retrying with a smaller area on ENOMEM */
+ do {
+ mz = rte_memzone_reserve_aligned(mz_name,
+ RTE_MIN((size_t)mem_size, max_alloc_size),
mp->socket_id, mz_flags, align);
- /* don't try reserving with 0 size if we were asked to reserve
- * IOVA-contiguous memory.
- */
- if (min_chunk_size < (size_t)mem_size && mz == NULL) {
- /* not enough memory, retry with the biggest zone we
- * have
- */
- mz = rte_memzone_reserve_aligned(mz_name, 0,
- mp->socket_id, mz_flags, align);
- }
+ if (mz == NULL && rte_errno != ENOMEM)
+ break;
+
+ max_alloc_size = RTE_MIN(max_alloc_size,
+ (size_t)mem_size) / 2;
+ } while (mz == NULL && max_alloc_size >= min_chunk_size);
+
if (mz == NULL) {
ret = -rte_errno;
goto fail;
}
- if (mz->len < min_chunk_size) {
- rte_memzone_free(mz);
- ret = -ENOMEM;
- goto fail;
- }
-
if (need_iova_contig_obj)
iova = mz->iova;
else
mz->len, pg_sz,
rte_mempool_memchunk_mz_free,
(void *)(uintptr_t)mz);
+ if (ret == 0) /* should not happen */
+ ret = -ENOBUFS;
if (ret < 0) {
rte_memzone_free(mz);
goto fail;
}
}
+ rte_mempool_trace_populate_default(mp);
return mp->size;
fail:
}
ret = mempool_ops_alloc_once(mp);
- if (ret != 0)
- return ret;
+ if (ret < 0) {
+ rte_errno = -ret;
+ return 0;
+ }
size = get_anon_size(mp);
if (size < 0) {
ret = rte_mempool_populate_virt(mp, addr, size, getpagesize(),
rte_mempool_memchunk_anon_free, addr);
- if (ret == 0)
+ if (ret == 0) /* should not happen */
+ ret = -ENOBUFS;
+ if (ret < 0) {
+ rte_errno = -ret;
goto fail;
+ }
+ rte_mempool_trace_populate_anon(mp);
return mp->populated_size;
fail:
}
rte_mcfg_tailq_write_unlock();
+ rte_mempool_trace_free(mp);
rte_mempool_free_memchunks(mp);
rte_mempool_ops_free(mp);
rte_memzone_free(mp->mz);
mempool_cache_init(cache, size);
+ rte_mempool_trace_cache_create(size, socket_id, cache);
return cache;
}
void
rte_mempool_cache_free(struct rte_mempool_cache *cache)
{
+ rte_mempool_trace_cache_free(cache);
rte_free(cache);
}
rte_mcfg_tailq_write_unlock();
rte_mcfg_mempool_write_unlock();
+ rte_mempool_trace_create_empty(name, n, elt_size, cache_size,
+ private_data_size, flags, mp);
return mp;
exit_unlock:
if (obj_init)
rte_mempool_obj_iter(mp, obj_init, obj_init_arg);
+ rte_mempool_trace_create(name, n, elt_size, cache_size,
+ private_data_size, mp_init, mp_init_arg, obj_init,
+ obj_init_arg, flags, mp);
return mp;
fail:
git.droids-corp.org / dpdk.git / blobdiff