X-Git-Url: http://git.droids-corp.org/?a=blobdiff_plain;f=lib%2Flibrte_mempool%2Frte_mempool.c;h=9f1a4253b75f5568c9cb0c73d258aac71d142894;hb=5751ff40fed0f7dc359e983d80c768126328561f;hp=d78d02b7b8c514cb631d3eee867501c7779ab382;hpb=a0fd91cefcc054b770dec6d8fb60db0d3145c45a;p=dpdk.git diff --git a/lib/librte_mempool/rte_mempool.c b/lib/librte_mempool/rte_mempool.c index d78d02b7b8..9f1a4253b7 100644 --- a/lib/librte_mempool/rte_mempool.c +++ b/lib/librte_mempool/rte_mempool.c @@ -1,37 +1,9 @@ -/*- - * BSD LICENSE - * - * Copyright(c) 2010-2014 Intel Corporation. All rights reserved. - * Copyright(c) 2016 6WIND S.A. - * All rights reserved. - * - * Redistribution and use in source and binary forms, with or without - * modification, are permitted provided that the following conditions - * are met: - * - * * Redistributions of source code must retain the above copyright - * notice, this list of conditions and the following disclaimer. - * * Redistributions in binary form must reproduce the above copyright - * notice, this list of conditions and the following disclaimer in - * the documentation and/or other materials provided with the - * distribution. - * * Neither the name of Intel Corporation nor the names of its - * contributors may be used to endorse or promote products derived - * from this software without specific prior written permission. - * - * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS - * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT - * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR - * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT - * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, - * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT - * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, - * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY - * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT - * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE - * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. +/* SPDX-License-Identifier: BSD-3-Clause + * Copyright(c) 2010-2014 Intel Corporation. + * Copyright(c) 2016 6WIND S.A. */ +#include #include #include #include @@ -55,7 +27,6 @@ #include #include #include -#include #include #include #include @@ -128,8 +99,31 @@ static unsigned optimize_object_size(unsigned obj_size) return new_obj_size * RTE_MEMPOOL_ALIGN; } +static int +find_min_pagesz(const struct rte_memseg_list *msl, void *arg) +{ + size_t *min = arg; + + if (msl->page_sz < *min) + *min = msl->page_sz; + + return 0; +} + +static size_t +get_min_page_size(void) +{ + size_t min_pagesz = SIZE_MAX; + + rte_memseg_list_walk(find_min_pagesz, &min_pagesz); + + return min_pagesz == SIZE_MAX ? (size_t) getpagesize() : min_pagesz; +} + + static void -mempool_add_elem(struct rte_mempool *mp, void *obj, phys_addr_t physaddr) +mempool_add_elem(struct rte_mempool *mp, __rte_unused void *opaque, + void *obj, rte_iova_t iova) { struct rte_mempool_objhdr *hdr; struct rte_mempool_objtlr *tlr __rte_unused; @@ -137,7 +131,7 @@ mempool_add_elem(struct rte_mempool *mp, void *obj, phys_addr_t physaddr) /* set mempool ptr in header */ hdr = RTE_PTR_SUB(obj, sizeof(*hdr)); hdr->mp = mp; - hdr->physaddr = physaddr; + hdr->iova = iova; STAILQ_INSERT_TAIL(&mp->elt_list, hdr, next); mp->populated_size++; @@ -146,9 +140,6 @@ mempool_add_elem(struct rte_mempool *mp, void *obj, phys_addr_t physaddr) tlr = __mempool_get_trailer(obj); tlr->cookie = RTE_MEMPOOL_TRAILER_COOKIE; #endif - - /* enqueue in ring */ - rte_mempool_ops_enqueue_bulk(mp, &obj, 1); } /* call obj_cb() for each mempool element */ @@ -199,7 +190,11 @@ rte_mempool_calc_obj_size(uint32_t elt_size, uint32_t flags, sz->header_size = RTE_ALIGN_CEIL(sz->header_size, RTE_MEMPOOL_ALIGN); +#ifdef RTE_LIBRTE_MEMPOOL_DEBUG sz->trailer_size = sizeof(struct rte_mempool_objtlr); +#else + sz->trailer_size = 0; +#endif /* element size is 8 bytes-aligned at least */ sz->elt_size = RTE_ALIGN_CEIL(elt_size, sizeof(uint64_t)); @@ -232,10 +227,13 @@ rte_mempool_calc_obj_size(uint32_t elt_size, uint32_t flags, /* - * Calculate maximum amount of memory required to store given number of objects. + * Internal function to calculate required memory chunk size shared + * by default implementation of the corresponding callback and + * deprecated external function. */ size_t -rte_mempool_xmem_size(uint32_t elt_num, size_t total_elt_sz, uint32_t pg_shift) +rte_mempool_calc_mem_size_helper(uint32_t elt_num, size_t total_elt_sz, + uint32_t pg_shift) { size_t obj_per_page, pg_num, pg_sz; @@ -254,29 +252,40 @@ rte_mempool_xmem_size(uint32_t elt_num, size_t total_elt_sz, uint32_t pg_shift) return pg_num << pg_shift; } +/* + * Calculate maximum amount of memory required to store given number of objects. + */ +size_t +rte_mempool_xmem_size(uint32_t elt_num, size_t total_elt_sz, uint32_t pg_shift, + __rte_unused unsigned int flags) +{ + return rte_mempool_calc_mem_size_helper(elt_num, total_elt_sz, + pg_shift); +} + /* * Calculate how much memory would be actually required with the * given memory footprint to store required number of elements. */ ssize_t rte_mempool_xmem_usage(__rte_unused void *vaddr, uint32_t elt_num, - size_t total_elt_sz, const phys_addr_t paddr[], uint32_t pg_num, - uint32_t pg_shift) + size_t total_elt_sz, const rte_iova_t iova[], uint32_t pg_num, + uint32_t pg_shift, __rte_unused unsigned int flags) { uint32_t elt_cnt = 0; - phys_addr_t start, end; - uint32_t paddr_idx; + rte_iova_t start, end; + uint32_t iova_idx; size_t pg_sz = (size_t)1 << pg_shift; - /* if paddr is NULL, assume contiguous memory */ - if (paddr == NULL) { + /* if iova is NULL, assume contiguous memory */ + if (iova == NULL) { start = 0; end = pg_sz * pg_num; - paddr_idx = pg_num; + iova_idx = pg_num; } else { - start = paddr[0]; - end = paddr[0] + pg_sz; - paddr_idx = 1; + start = iova[0]; + end = iova[0] + pg_sz; + iova_idx = 1; } while (elt_cnt < elt_num) { @@ -284,15 +293,15 @@ rte_mempool_xmem_usage(__rte_unused void *vaddr, uint32_t elt_num, /* enough contiguous memory, add an object */ start += total_elt_sz; elt_cnt++; - } else if (paddr_idx < pg_num) { + } else if (iova_idx < pg_num) { /* no room to store one obj, add a page */ - if (end == paddr[paddr_idx]) { + if (end == iova[iova_idx]) { end += pg_sz; } else { - start = paddr[paddr_idx]; - end = paddr[paddr_idx] + pg_sz; + start = iova[iova_idx]; + end = iova[iova_idx] + pg_sz; } - paddr_idx++; + iova_idx++; } else { /* no more page, return how many elements fit */ @@ -300,7 +309,7 @@ rte_mempool_xmem_usage(__rte_unused void *vaddr, uint32_t elt_num, } } - return (size_t)paddr_idx << pg_shift; + return (size_t)iova_idx << pg_shift; } /* free a memchunk allocated with rte_memzone_reserve() */ @@ -336,42 +345,50 @@ rte_mempool_free_memchunks(struct rte_mempool *mp) } } +static int +mempool_ops_alloc_once(struct rte_mempool *mp) +{ + int ret; + + /* create the internal ring if not already done */ + if ((mp->flags & MEMPOOL_F_POOL_CREATED) == 0) { + ret = rte_mempool_ops_alloc(mp); + if (ret != 0) + return ret; + mp->flags |= MEMPOOL_F_POOL_CREATED; + } + return 0; +} + /* 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_phys(struct rte_mempool *mp, char *vaddr, - phys_addr_t paddr, size_t len, rte_mempool_memchunk_free_cb_t *free_cb, +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) { - unsigned total_elt_sz; unsigned i = 0; size_t off; struct rte_mempool_memhdr *memhdr; int ret; - /* create the internal ring if not already done */ - if ((mp->flags & MEMPOOL_F_POOL_CREATED) == 0) { - ret = rte_mempool_ops_alloc(mp); - if (ret != 0) - return ret; - mp->flags |= MEMPOOL_F_POOL_CREATED; - } + ret = mempool_ops_alloc_once(mp); + if (ret != 0) + return ret; /* mempool is already populated */ if (mp->populated_size >= mp->size) return -ENOSPC; - total_elt_sz = mp->header_size + mp->elt_size + mp->trailer_size; - memhdr = rte_zmalloc("MEMPOOL_MEMHDR", sizeof(*memhdr), 0); if (memhdr == NULL) return -ENOMEM; memhdr->mp = mp; memhdr->addr = vaddr; - memhdr->phys_addr = paddr; + memhdr->iova = iova; memhdr->len = len; memhdr->free_cb = free_cb; memhdr->opaque = opaque; @@ -381,32 +398,45 @@ rte_mempool_populate_phys(struct rte_mempool *mp, char *vaddr, else off = RTE_PTR_ALIGN_CEIL(vaddr, RTE_CACHE_LINE_SIZE) - vaddr; - while (off + total_elt_sz <= len && mp->populated_size < mp->size) { - off += mp->header_size; - if (paddr == RTE_BAD_PHYS_ADDR) - mempool_add_elem(mp, (char *)vaddr + off, - RTE_BAD_PHYS_ADDR); - else - mempool_add_elem(mp, (char *)vaddr + off, paddr + off); - off += mp->elt_size + mp->trailer_size; - i++; + if (off > len) { + ret = -EINVAL; + goto fail; } + i = rte_mempool_ops_populate(mp, mp->size - mp->populated_size, + (char *)vaddr + off, + (iova == RTE_BAD_IOVA) ? RTE_BAD_IOVA : (iova + off), + len - off, mempool_add_elem, NULL); + /* not enough room to store one object */ - if (i == 0) - return -EINVAL; + if (i == 0) { + ret = -EINVAL; + goto fail; + } STAILQ_INSERT_TAIL(&mp->mem_list, memhdr, next); mp->nb_mem_chunks++; return i; + +fail: + rte_free(memhdr); + return ret; +} + +int +rte_mempool_populate_phys(struct rte_mempool *mp, char *vaddr, + phys_addr_t paddr, size_t len, rte_mempool_memchunk_free_cb_t *free_cb, + void *opaque) +{ + return rte_mempool_populate_iova(mp, vaddr, paddr, len, free_cb, opaque); } /* Add objects in the pool, using a table of physical pages. Return the * number of objects added, or a negative value on error. */ int -rte_mempool_populate_phys_tab(struct rte_mempool *mp, char *vaddr, - const phys_addr_t paddr[], uint32_t pg_num, uint32_t pg_shift, +rte_mempool_populate_iova_tab(struct rte_mempool *mp, char *vaddr, + const rte_iova_t iova[], uint32_t pg_num, uint32_t pg_shift, rte_mempool_memchunk_free_cb_t *free_cb, void *opaque) { uint32_t i, n; @@ -417,19 +447,19 @@ rte_mempool_populate_phys_tab(struct rte_mempool *mp, char *vaddr, if (mp->nb_mem_chunks != 0) return -EEXIST; - if (mp->flags & MEMPOOL_F_NO_PHYS_CONTIG) - return rte_mempool_populate_phys(mp, vaddr, RTE_BAD_PHYS_ADDR, + if (mp->flags & MEMPOOL_F_NO_IOVA_CONTIG) + return rte_mempool_populate_iova(mp, vaddr, RTE_BAD_IOVA, pg_num * pg_sz, free_cb, opaque); for (i = 0; i < pg_num && mp->populated_size < mp->size; i += n) { /* populate with the largest group of contiguous pages */ for (n = 1; (i + n) < pg_num && - paddr[i] + pg_sz == paddr[i+n]; n++) + iova[i + n - 1] + pg_sz == iova[i + n]; n++) ; - ret = rte_mempool_populate_phys(mp, vaddr + i * pg_sz, - paddr[i], n * pg_sz, free_cb, opaque); + ret = rte_mempool_populate_iova(mp, vaddr + i * pg_sz, + iova[i], n * pg_sz, free_cb, opaque); if (ret < 0) { rte_mempool_free_memchunks(mp); return ret; @@ -441,6 +471,15 @@ rte_mempool_populate_phys_tab(struct rte_mempool *mp, char *vaddr, return cnt; } +int +rte_mempool_populate_phys_tab(struct rte_mempool *mp, char *vaddr, + const phys_addr_t paddr[], uint32_t pg_num, uint32_t pg_shift, + rte_mempool_memchunk_free_cb_t *free_cb, void *opaque) +{ + return rte_mempool_populate_iova_tab(mp, vaddr, paddr, pg_num, pg_shift, + free_cb, opaque); +} + /* Populate the mempool with a virtual area. Return the number of * objects added, or a negative value on error. */ @@ -449,7 +488,7 @@ 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) { - phys_addr_t paddr; + rte_iova_t iova; size_t off, phys_len; int ret, cnt = 0; @@ -462,34 +501,31 @@ rte_mempool_populate_virt(struct rte_mempool *mp, char *addr, if (RTE_ALIGN_CEIL(len, pg_sz) != len) return -EINVAL; - if (mp->flags & MEMPOOL_F_NO_PHYS_CONTIG) - return rte_mempool_populate_phys(mp, addr, RTE_BAD_PHYS_ADDR, + if (mp->flags & MEMPOOL_F_NO_IOVA_CONTIG) + return rte_mempool_populate_iova(mp, addr, RTE_BAD_IOVA, len, free_cb, opaque); for (off = 0; off + pg_sz <= len && mp->populated_size < mp->size; off += phys_len) { - paddr = rte_mem_virt2phy(addr + off); - /* required for xen_dom0 to get the machine address */ - paddr = rte_mem_phy2mch(-1, paddr); + iova = rte_mem_virt2iova(addr + off); - if (paddr == RTE_BAD_PHYS_ADDR) { + if (iova == RTE_BAD_IOVA && rte_eal_has_hugepages()) { ret = -EINVAL; goto fail; } /* populate with the largest group of contiguous pages */ for (phys_len = pg_sz; off + phys_len < len; phys_len += pg_sz) { - phys_addr_t paddr_tmp; + rte_iova_t iova_tmp; - paddr_tmp = rte_mem_virt2phy(addr + off + phys_len); - paddr_tmp = rte_mem_phy2mch(-1, paddr_tmp); + iova_tmp = rte_mem_virt2iova(addr + off + phys_len); - if (paddr_tmp != paddr + phys_len) + if (iova_tmp != iova + phys_len) break; } - ret = rte_mempool_populate_phys(mp, addr + off, paddr, + ret = rte_mempool_populate_iova(mp, addr + off, iova, phys_len, free_cb, opaque); if (ret < 0) goto fail; @@ -512,31 +548,97 @@ rte_mempool_populate_virt(struct rte_mempool *mp, char *addr, int rte_mempool_populate_default(struct rte_mempool *mp) { - int mz_flags = RTE_MEMZONE_1GB|RTE_MEMZONE_SIZE_HINT_ONLY; + unsigned int mz_flags = RTE_MEMZONE_1GB|RTE_MEMZONE_SIZE_HINT_ONLY; char mz_name[RTE_MEMZONE_NAMESIZE]; const struct rte_memzone *mz; - size_t size, total_elt_sz, align, pg_sz, pg_shift; - phys_addr_t paddr; + ssize_t mem_size; + size_t align, pg_sz, pg_shift; + rte_iova_t iova; unsigned mz_id, n; int ret; + bool no_contig, try_contig, no_pageshift; + + ret = mempool_ops_alloc_once(mp); + if (ret != 0) + return ret; /* mempool must not be populated */ if (mp->nb_mem_chunks != 0) return -EEXIST; - if (rte_eal_has_hugepages()) { - pg_shift = 0; /* not needed, zone is physically contiguous */ + no_contig = mp->flags & MEMPOOL_F_NO_IOVA_CONTIG; + + /* + * the following section calculates page shift and page size values. + * + * these values impact the result of calc_mem_size operation, which + * returns the amount of memory that should be allocated to store the + * desired number of objects. when not zero, it allocates more memory + * for the padding between objects, to ensure that an object does not + * cross a page boundary. in other words, page size/shift are to be set + * to zero if mempool elements won't care about page boundaries. + * there are several considerations for page size and page shift here. + * + * if we don't need our mempools to have physically contiguous objects, + * 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 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. + * + * 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 + * we require depends on whether we want bigger or smaller pages. + * However, requesting each and every memory size is too much work, so + * what we'll do instead is walk through the page sizes available, pick + * the smallest one and set up page shift to match that one. We will be + * wasting some space this way, but it's much nicer than looping around + * trying to reserve each and every page size. + * + * However, since size calculation will produce page-aligned sizes, it + * makes sense to first try and see if we can reserve the entire memzone + * in one contiguous chunk as well (otherwise we might end up wasting a + * 1G page on a 10MB memzone). If we fail to get enough contiguous + * memory, then we'll go and reserve space page-by-page. + */ + no_pageshift = no_contig || rte_eal_iova_mode() == RTE_IOVA_VA; + try_contig = !no_contig && !no_pageshift && rte_eal_has_hugepages(); + + if (no_pageshift) { pg_sz = 0; - align = RTE_CACHE_LINE_SIZE; + pg_shift = 0; + } else if (try_contig) { + pg_sz = get_min_page_size(); + pg_shift = rte_bsf32(pg_sz); } else { pg_sz = getpagesize(); pg_shift = rte_bsf32(pg_sz); - align = pg_sz; } - total_elt_sz = mp->header_size + mp->elt_size + mp->trailer_size; for (mz_id = 0, n = mp->size; n > 0; mz_id++, n -= ret) { - size = rte_mempool_xmem_size(n, total_elt_sz, pg_shift); + size_t min_chunk_size; + unsigned int flags; + + if (try_contig || no_pageshift) + mem_size = rte_mempool_ops_calc_mem_size(mp, n, + 0, &min_chunk_size, &align); + else + mem_size = rte_mempool_ops_calc_mem_size(mp, n, + pg_shift, &min_chunk_size, &align); + + if (mem_size < 0) { + ret = mem_size; + goto fail; + } ret = snprintf(mz_name, sizeof(mz_name), RTE_MEMPOOL_MZ_FORMAT "_%d", mp->name, mz_id); @@ -545,34 +647,76 @@ rte_mempool_populate_default(struct rte_mempool *mp) goto fail; } - mz = rte_memzone_reserve_aligned(mz_name, size, - mp->socket_id, mz_flags, align); - /* not enough memory, retry with the biggest zone we have */ - if (mz == NULL) + flags = mz_flags; + + /* if we're trying to reserve contiguous memory, add appropriate + * memzone flag. + */ + if (try_contig) + flags |= RTE_MEMZONE_IOVA_CONTIG; + + mz = rte_memzone_reserve_aligned(mz_name, mem_size, + mp->socket_id, flags, align); + + /* if we were trying to allocate contiguous memory, failed and + * minimum required contiguous chunk fits minimum page, adjust + * memzone size to the page size, and try again. + */ + if (mz == NULL && try_contig && min_chunk_size <= pg_sz) { + try_contig = false; + flags &= ~RTE_MEMZONE_IOVA_CONTIG; + + mem_size = rte_mempool_ops_calc_mem_size(mp, n, + pg_shift, &min_chunk_size, &align); + if (mem_size < 0) { + ret = mem_size; + goto fail; + } + + mz = rte_memzone_reserve_aligned(mz_name, mem_size, + mp->socket_id, 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); + mp->socket_id, flags, + RTE_MAX(pg_sz, align)); + } if (mz == NULL) { ret = -rte_errno; goto fail; } - if (mp->flags & MEMPOOL_F_NO_PHYS_CONTIG) - paddr = RTE_BAD_PHYS_ADDR; + if (mz->len < min_chunk_size) { + rte_memzone_free(mz); + ret = -ENOMEM; + goto fail; + } + + if (no_contig) + iova = RTE_BAD_IOVA; else - paddr = mz->phys_addr; + iova = mz->iova; - if (rte_eal_has_hugepages() && !rte_xen_dom0_supported()) - ret = rte_mempool_populate_phys(mp, mz->addr, - paddr, mz->len, + if (no_pageshift || try_contig) + ret = rte_mempool_populate_iova(mp, mz->addr, + iova, mz->len, rte_mempool_memchunk_mz_free, (void *)(uintptr_t)mz); else ret = rte_mempool_populate_virt(mp, mz->addr, - mz->len, pg_sz, + RTE_ALIGN_FLOOR(mz->len, pg_sz), pg_sz, rte_mempool_memchunk_mz_free, (void *)(uintptr_t)mz); - if (ret < 0) + if (ret < 0) { + rte_memzone_free(mz); goto fail; + } } return mp->size; @@ -583,15 +727,18 @@ rte_mempool_populate_default(struct rte_mempool *mp) } /* return the memory size required for mempool objects in anonymous mem */ -static size_t +static ssize_t get_anon_size(const struct rte_mempool *mp) { - size_t size, total_elt_sz, pg_sz, pg_shift; + ssize_t size; + size_t pg_sz, pg_shift; + size_t min_chunk_size; + size_t align; pg_sz = getpagesize(); pg_shift = rte_bsf32(pg_sz); - total_elt_sz = mp->header_size + mp->elt_size + mp->trailer_size; - size = rte_mempool_xmem_size(mp->size, total_elt_sz, pg_shift); + size = rte_mempool_ops_calc_mem_size(mp, mp->size, pg_shift, + &min_chunk_size, &align); return size; } @@ -601,14 +748,25 @@ static void rte_mempool_memchunk_anon_free(struct rte_mempool_memhdr *memhdr, void *opaque) { - munmap(opaque, get_anon_size(memhdr->mp)); + ssize_t size; + + /* + * Calculate size since memhdr->len has contiguous chunk length + * which may be smaller if anon map is split into many contiguous + * chunks. Result must be the same as we calculated on populate. + */ + size = get_anon_size(memhdr->mp); + if (size < 0) + return; + + munmap(opaque, size); } /* populate the mempool with an anonymous mapping */ int rte_mempool_populate_anon(struct rte_mempool *mp) { - size_t size; + ssize_t size; int ret; char *addr; @@ -618,8 +776,17 @@ rte_mempool_populate_anon(struct rte_mempool *mp) return 0; } - /* get chunk of virtually continuous memory */ + ret = mempool_ops_alloc_once(mp); + if (ret != 0) + return ret; + size = get_anon_size(mp); + if (size < 0) { + rte_errno = -size; + return 0; + } + + /* get chunk of virtually continuous memory */ addr = mmap(NULL, size, PROT_READ | PROT_WRITE, MAP_SHARED | MAP_ANONYMOUS, -1, 0); if (addr == MAP_FAILED) { @@ -733,7 +900,7 @@ rte_mempool_create_empty(const char *name, unsigned n, unsigned elt_size, struct rte_tailq_entry *te = NULL; const struct rte_memzone *mz = NULL; size_t mempool_size; - int mz_flags = RTE_MEMZONE_1GB|RTE_MEMZONE_SIZE_HINT_ONLY; + unsigned int mz_flags = RTE_MEMZONE_1GB|RTE_MEMZONE_SIZE_HINT_ONLY; struct rte_mempool_objsz objsz; unsigned lcore_id; int ret; @@ -809,7 +976,6 @@ rte_mempool_create_empty(const char *name, unsigned n, unsigned elt_size, goto exit_unlock; } mp->mz = mz; - mp->socket_id = socket_id; mp->size = n; mp->flags = flags; mp->socket_id = socket_id; @@ -860,6 +1026,7 @@ rte_mempool_create(const char *name, unsigned n, unsigned elt_size, rte_mempool_obj_cb_t *obj_init, void *obj_init_arg, int socket_id, unsigned flags) { + int ret; struct rte_mempool *mp; mp = rte_mempool_create_empty(name, n, elt_size, cache_size, @@ -871,14 +1038,17 @@ rte_mempool_create(const char *name, unsigned n, unsigned elt_size, * Since we have 4 combinations of the SP/SC/MP/MC examine the flags to * set the correct index into the table of ops structs. */ - if (flags & (MEMPOOL_F_SP_PUT | MEMPOOL_F_SC_GET)) - rte_mempool_set_ops_byname(mp, "ring_sp_sc", NULL); + if ((flags & MEMPOOL_F_SP_PUT) && (flags & MEMPOOL_F_SC_GET)) + ret = rte_mempool_set_ops_byname(mp, "ring_sp_sc", NULL); else if (flags & MEMPOOL_F_SP_PUT) - rte_mempool_set_ops_byname(mp, "ring_sp_mc", NULL); + ret = rte_mempool_set_ops_byname(mp, "ring_sp_mc", NULL); else if (flags & MEMPOOL_F_SC_GET) - rte_mempool_set_ops_byname(mp, "ring_mp_sc", NULL); + ret = rte_mempool_set_ops_byname(mp, "ring_mp_sc", NULL); else - rte_mempool_set_ops_byname(mp, "ring_mp_mc", NULL); + ret = rte_mempool_set_ops_byname(mp, "ring_mp_mc", NULL); + + if (ret) + goto fail; /* call the mempool priv initializer */ if (mp_init) @@ -901,9 +1071,8 @@ rte_mempool_create(const char *name, unsigned n, unsigned elt_size, /* * Create the mempool over already allocated chunk of memory. * That external memory buffer can consists of physically disjoint pages. - * Setting vaddr to NULL, makes mempool to fallback to original behaviour - * and allocate space for mempool and it's elements as one big chunk of - * physically continuos memory. + * Setting vaddr to NULL, makes mempool to fallback to rte_mempool_create() + * behavior. */ struct rte_mempool * rte_mempool_xmem_create(const char *name, unsigned n, unsigned elt_size, @@ -911,7 +1080,7 @@ rte_mempool_xmem_create(const char *name, unsigned n, unsigned elt_size, rte_mempool_ctor_t *mp_init, void *mp_init_arg, rte_mempool_obj_cb_t *obj_init, void *obj_init_arg, int socket_id, unsigned flags, void *vaddr, - const phys_addr_t paddr[], uint32_t pg_num, uint32_t pg_shift) + const rte_iova_t iova[], uint32_t pg_num, uint32_t pg_shift) { struct rte_mempool *mp = NULL; int ret; @@ -923,7 +1092,7 @@ rte_mempool_xmem_create(const char *name, unsigned n, unsigned elt_size, obj_init, obj_init_arg, socket_id, flags); /* check that we have both VA and PA */ - if (paddr == NULL) { + if (iova == NULL) { rte_errno = EINVAL; return NULL; } @@ -943,7 +1112,7 @@ rte_mempool_xmem_create(const char *name, unsigned n, unsigned elt_size, if (mp_init) mp_init(mp, mp_init_arg); - ret = rte_mempool_populate_phys_tab(mp, vaddr, paddr, pg_num, pg_shift, + ret = rte_mempool_populate_iova_tab(mp, vaddr, iova, pg_num, pg_shift, NULL, NULL); if (ret < 0 || ret != (int)mp->size) goto fail; @@ -990,12 +1159,6 @@ rte_mempool_in_use_count(const struct rte_mempool *mp) return mp->size - rte_mempool_avail_count(mp); } -unsigned int -rte_mempool_count(const struct rte_mempool *mp) -{ - return rte_mempool_avail_count(mp); -} - /* dump the cache status */ static unsigned rte_mempool_dump_cache(FILE *f, const struct rte_mempool *mp) @@ -1039,7 +1202,7 @@ void rte_mempool_check_cookies(const struct rte_mempool *mp, /* Force to drop the "const" attribute. This is done only when * DEBUG is enabled */ tmp = (void *) obj_table_const; - obj_table = (void **) tmp; + obj_table = tmp; while (n--) { obj = obj_table[n]; @@ -1093,6 +1256,36 @@ void rte_mempool_check_cookies(const struct rte_mempool *mp, #endif } +void +rte_mempool_contig_blocks_check_cookies(const struct rte_mempool *mp, + void * const *first_obj_table_const, unsigned int n, int free) +{ +#ifdef RTE_LIBRTE_MEMPOOL_DEBUG + struct rte_mempool_info info; + const size_t total_elt_sz = + mp->header_size + mp->elt_size + mp->trailer_size; + unsigned int i, j; + + rte_mempool_ops_get_info(mp, &info); + + for (i = 0; i < n; ++i) { + void *first_obj = first_obj_table_const[i]; + + for (j = 0; j < info.contig_block_size; ++j) { + void *obj; + + obj = (void *)((uintptr_t)first_obj + j * total_elt_sz); + rte_mempool_check_cookies(mp, &obj, 1, free); + } + } +#else + RTE_SET_USED(mp); + RTE_SET_USED(first_obj_table_const); + RTE_SET_USED(n); + RTE_SET_USED(free); +#endif +} + #ifdef RTE_LIBRTE_MEMPOOL_DEBUG static void mempool_obj_audit(struct rte_mempool *mp, __rte_unused void *opaque, @@ -1158,6 +1351,7 @@ void rte_mempool_dump(FILE *f, struct rte_mempool *mp) { #ifdef RTE_LIBRTE_MEMPOOL_DEBUG + struct rte_mempool_info info; struct rte_mempool_debug_stats sum; unsigned lcore_id; #endif @@ -1172,7 +1366,7 @@ rte_mempool_dump(FILE *f, struct rte_mempool *mp) fprintf(f, "mempool <%s>@%p\n", mp->name, mp); fprintf(f, " flags=%x\n", mp->flags); fprintf(f, " pool=%p\n", mp->pool_data); - fprintf(f, " phys_addr=0x%" PRIx64 "\n", mp->mz->phys_addr); + fprintf(f, " iova=0x%" PRIx64 "\n", mp->mz->iova); fprintf(f, " nb_mem_chunks=%u\n", mp->nb_mem_chunks); fprintf(f, " size=%"PRIu32"\n", mp->size); fprintf(f, " populated_size=%"PRIu32"\n", mp->populated_size); @@ -1199,6 +1393,7 @@ rte_mempool_dump(FILE *f, struct rte_mempool *mp) /* sum and dump statistics */ #ifdef RTE_LIBRTE_MEMPOOL_DEBUG + rte_mempool_ops_get_info(mp, &info); memset(&sum, 0, sizeof(sum)); for (lcore_id = 0; lcore_id < RTE_MAX_LCORE; lcore_id++) { sum.put_bulk += mp->stats[lcore_id].put_bulk; @@ -1207,6 +1402,8 @@ rte_mempool_dump(FILE *f, struct rte_mempool *mp) sum.get_success_objs += mp->stats[lcore_id].get_success_objs; sum.get_fail_bulk += mp->stats[lcore_id].get_fail_bulk; sum.get_fail_objs += mp->stats[lcore_id].get_fail_objs; + sum.get_success_blks += mp->stats[lcore_id].get_success_blks; + sum.get_fail_blks += mp->stats[lcore_id].get_fail_blks; } fprintf(f, " stats:\n"); fprintf(f, " put_bulk=%"PRIu64"\n", sum.put_bulk); @@ -1215,6 +1412,11 @@ rte_mempool_dump(FILE *f, struct rte_mempool *mp) fprintf(f, " get_success_objs=%"PRIu64"\n", sum.get_success_objs); fprintf(f, " get_fail_bulk=%"PRIu64"\n", sum.get_fail_bulk); fprintf(f, " get_fail_objs=%"PRIu64"\n", sum.get_fail_objs); + if (info.contig_block_size > 0) { + fprintf(f, " get_success_blks=%"PRIu64"\n", + sum.get_success_blks); + fprintf(f, " get_fail_blks=%"PRIu64"\n", sum.get_fail_blks); + } #else fprintf(f, " no statistics available\n"); #endif @@ -1275,12 +1477,13 @@ void rte_mempool_walk(void (*func)(struct rte_mempool *, void *), { struct rte_tailq_entry *te = NULL; struct rte_mempool_list *mempool_list; + void *tmp_te; mempool_list = RTE_TAILQ_CAST(rte_mempool_tailq.head, rte_mempool_list); rte_rwlock_read_lock(RTE_EAL_MEMPOOL_RWLOCK); - TAILQ_FOREACH(te, mempool_list, next) { + TAILQ_FOREACH_SAFE(te, mempool_list, next, tmp_te) { (*func)((struct rte_mempool *) te->data, arg); }