From nobody Tue Apr 7 22:18:01 2026 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 449B4C4332F for ; Wed, 19 Oct 2022 18:35:22 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S229882AbiJSSfU (ORCPT ); Wed, 19 Oct 2022 14:35:20 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:51312 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S229905AbiJSSfL (ORCPT ); Wed, 19 Oct 2022 14:35:11 -0400 Received: from mail-qt1-x843.google.com (mail-qt1-x843.google.com [IPv6:2607:f8b0:4864:20::843]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 30F1A1552DF for ; Wed, 19 Oct 2022 11:35:08 -0700 (PDT) Received: by mail-qt1-x843.google.com with SMTP id l28so12267773qtv.4 for ; Wed, 19 Oct 2022 11:35:08 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20210112; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:from:to:cc:subject:date :message-id:reply-to; bh=jGXe7jLtYGY1YP2N1w0hr8ZmPtDOoXLhtl5xhD+f2+E=; b=Da6KovRipl4p7VJ+YDz1DhvDQ5tXjiDXp95hz6yoE8+dUvdVeC7zs8bqlvWiKHxLE2 KZd6lpMNIxrI4VMvmYySzEw+jYAUYd3gLeywm2jwToOjpKepW07Z2+nO/FnTS9JqQgzI LbLh22zZHCAe9lUUjnuSnr7dHbhRqU/Jfp2STOW4UGGZzq3AsPtkCLjjegB2kD4t6AQ/ SFAEsMk8JlbdKdfqdY3WaBsMwnQ8A3vRwk9KRrIGyA7rByeuijBoDXJLtVg8bL4nbGiL jOa/1crNsFXjA04Cx5K5sKnmauRKCoSb+khv4lze5QRbXFpUyL1akTCsj5esGcazF8Dk PfCw== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:x-gm-message-state:from:to:cc :subject:date:message-id:reply-to; bh=jGXe7jLtYGY1YP2N1w0hr8ZmPtDOoXLhtl5xhD+f2+E=; b=IrryWsNkR9H64Wccu6Sp6d31PeYKjpDhF98FDOUMrsxJjD63LbbPiY3Qy6y/lkglNc QDsvsNsBcw6HZf3/WApaomqjGUTj3fBUkVYqSpWNOOaLETlegDojOudMggfpbkj9YTaC TMPJRPWF3iu4Y7LUQQ+wO6SOsaG/1zABH5OKF5rCKe+Pc6MEARrSnzqPa7n5b/oslqFf ssiKazXaKMbs3m/CDsUfm2w0sD/yIAYmKv1ERY2NblxEMdh9Fm48nz1FX1oLv2CGR/Oq 0tkYaElPA8OveaVkpdTLRyNxhok590NatzmANzUrfBy5l5hbYapVrpaz+aXK4sYgZqAw yHFg== X-Gm-Message-State: ACrzQf3gr2WRWPCZ7y9MbAYmZh9tSuJnGf0WnTb9H2fM4Sb5lRuKwk4Q wHJRrgnLYWH88b6tJ2ccnBg= X-Google-Smtp-Source: AMsMyM7zG0wd+P6EZxgab7HjpBCmzgcwheAjoVWWXcPKbfU4Lk0JMfcz9TWxjDwG6xoPEzOg6Wr7ng== X-Received: by 2002:ac8:7d4c:0:b0:399:b8eb:749e with SMTP id h12-20020ac87d4c000000b00399b8eb749emr7556022qtb.446.1666204507249; Wed, 19 Oct 2022 11:35:07 -0700 (PDT) Received: from sophie ([143.244.47.100]) by smtp.gmail.com with ESMTPSA id g16-20020a05620a40d000b006eeb3165554sm5537313qko.19.2022.10.19.11.35.06 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Wed, 19 Oct 2022 11:35:06 -0700 (PDT) From: Rebecca Mckeever To: Mike Rapoport , linux-mm@kvack.org, linux-kernel@vger.kernel.org Cc: David Hildenbrand , Rebecca Mckeever Subject: [PATCH v3 2/5] memblock tests: add top-down NUMA tests for memblock_alloc_exact_nid_raw Date: Wed, 19 Oct 2022 13:34:09 -0500 Message-Id: X-Mailer: git-send-email 2.25.1 In-Reply-To: References: MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" Add tests for memblock_alloc_exact_nid_raw() where the simulated physical memory is set up with multiple NUMA nodes. Additionally, all of these tests set nid !=3D NUMA_NO_NODE. These tests are run with a top-down allocation direction. The tested scenarios are: Range unrestricted: - region can be allocated in the specific node requested: + there are no previously reserved regions + the requested node is partially reserved but has enough space Range restricted: - region can be allocated in the specific node requested after dropping min_addr: + range partially overlaps with two different nodes, where the first node is the requested node + range partially overlaps with two different nodes, where the requested node ends before min_addr + range overlaps with multiple nodes along node boundaries, and the requested node ends before min_addr Signed-off-by: Rebecca Mckeever Acked-by: David Hildenbrand --- .../memblock/tests/alloc_exact_nid_api.c | 344 ++++++++++++++++++ .../memblock/tests/alloc_exact_nid_api.h | 16 + 2 files changed, 360 insertions(+) diff --git a/tools/testing/memblock/tests/alloc_exact_nid_api.c b/tools/tes= ting/memblock/tests/alloc_exact_nid_api.c index 6406496623a0..79150784b373 100644 --- a/tools/testing/memblock/tests/alloc_exact_nid_api.c +++ b/tools/testing/memblock/tests/alloc_exact_nid_api.c @@ -4,6 +4,349 @@ =20 #define FUNC_NAME "memblock_alloc_exact_nid_raw" =20 +/* + * contains the fraction of MEM_SIZE contained in each node in basis point + * units (one hundredth of 1% or 1/10000) + */ +static const unsigned int node_fractions[] =3D { + 2500, /* 1/4 */ + 625, /* 1/16 */ + 1250, /* 1/8 */ + 1250, /* 1/8 */ + 625, /* 1/16 */ + 625, /* 1/16 */ + 2500, /* 1/4 */ + 625, /* 1/16 */ +}; + +/* + * A test that tries to allocate a memory region in a specific NUMA node t= hat + * has enough memory to allocate a region of the requested size. + * Expect to allocate an aligned region at the end of the requested node. + */ +static int alloc_exact_nid_top_down_numa_simple_check(void) +{ + int nid_req =3D 3; + struct memblock_region *new_rgn =3D &memblock.reserved.regions[0]; + struct memblock_region *req_node =3D &memblock.memory.regions[nid_req]; + void *allocated_ptr =3D NULL; + phys_addr_t size; + phys_addr_t min_addr; + phys_addr_t max_addr; + + PREFIX_PUSH(); + setup_numa_memblock(node_fractions); + + ASSERT_LE(SZ_4, req_node->size); + size =3D req_node->size / SZ_4; + min_addr =3D memblock_start_of_DRAM(); + max_addr =3D memblock_end_of_DRAM(); + + allocated_ptr =3D memblock_alloc_exact_nid_raw(size, SMP_CACHE_BYTES, + min_addr, max_addr, + nid_req); + + ASSERT_NE(allocated_ptr, NULL); + ASSERT_MEM_NE(allocated_ptr, 0, size); + + ASSERT_EQ(new_rgn->size, size); + ASSERT_EQ(new_rgn->base, region_end(req_node) - size); + ASSERT_LE(req_node->base, new_rgn->base); + + ASSERT_EQ(memblock.reserved.cnt, 1); + ASSERT_EQ(memblock.reserved.total_size, size); + + test_pass_pop(); + + return 0; +} + +/* + * A test that tries to allocate a memory region in a specific NUMA node t= hat + * is partially reserved but has enough memory for the allocated region: + * + * | +---------------------------------------+ | + * | | requested | | + * +-----------+---------------------------------------+----------+ + * + * | +------------------+ +-----+ | + * | | reserved | | new | | + * +-----------+------------------+--------------+-----+----------+ + * + * Expect to allocate an aligned region at the end of the requested node. = The + * region count and total size get updated. + */ +static int alloc_exact_nid_top_down_numa_part_reserved_check(void) +{ + int nid_req =3D 4; + struct memblock_region *new_rgn =3D &memblock.reserved.regions[1]; + struct memblock_region *req_node =3D &memblock.memory.regions[nid_req]; + void *allocated_ptr =3D NULL; + struct region r1; + phys_addr_t size; + phys_addr_t min_addr; + phys_addr_t max_addr; + + PREFIX_PUSH(); + setup_numa_memblock(node_fractions); + + ASSERT_LE(SZ_8, req_node->size); + r1.base =3D req_node->base; + r1.size =3D req_node->size / SZ_2; + size =3D r1.size / SZ_4; + min_addr =3D memblock_start_of_DRAM(); + max_addr =3D memblock_end_of_DRAM(); + + memblock_reserve(r1.base, r1.size); + allocated_ptr =3D memblock_alloc_exact_nid_raw(size, SMP_CACHE_BYTES, + min_addr, max_addr, + nid_req); + + ASSERT_NE(allocated_ptr, NULL); + ASSERT_MEM_NE(allocated_ptr, 0, size); + + ASSERT_EQ(new_rgn->size, size); + ASSERT_EQ(new_rgn->base, region_end(req_node) - size); + ASSERT_LE(req_node->base, new_rgn->base); + + ASSERT_EQ(memblock.reserved.cnt, 2); + ASSERT_EQ(memblock.reserved.total_size, size + r1.size); + + test_pass_pop(); + + return 0; +} + +/* + * A test that tries to allocate a memory region that spans over the min_a= ddr + * and max_addr range and overlaps with two different nodes, where the fir= st + * node is the requested node: + * + * min_addr + * | max_addr + * | | + * v v + * | +-----------------------+-----------+ | + * | | requested | node3 | | + * +-----------+-----------------------+-----------+--------------+ + * + + + * | +-----------+ | + * | | rgn | | + * +-----------------------+-----------+--------------------------+ + * + * Expect to drop the lower limit and allocate a memory region that ends at + * the end of the requested node. + */ +static int alloc_exact_nid_top_down_numa_split_range_low_check(void) +{ + int nid_req =3D 2; + struct memblock_region *new_rgn =3D &memblock.reserved.regions[0]; + struct memblock_region *req_node =3D &memblock.memory.regions[nid_req]; + void *allocated_ptr =3D NULL; + phys_addr_t size =3D SZ_512; + phys_addr_t min_addr; + phys_addr_t max_addr; + phys_addr_t req_node_end; + + PREFIX_PUSH(); + setup_numa_memblock(node_fractions); + + req_node_end =3D region_end(req_node); + min_addr =3D req_node_end - SZ_256; + max_addr =3D min_addr + size; + + allocated_ptr =3D memblock_alloc_exact_nid_raw(size, SMP_CACHE_BYTES, + min_addr, max_addr, + nid_req); + + ASSERT_NE(allocated_ptr, NULL); + ASSERT_MEM_NE(allocated_ptr, 0, size); + + ASSERT_EQ(new_rgn->size, size); + ASSERT_EQ(new_rgn->base, req_node_end - size); + ASSERT_LE(req_node->base, new_rgn->base); + + ASSERT_EQ(memblock.reserved.cnt, 1); + ASSERT_EQ(memblock.reserved.total_size, size); + + test_pass_pop(); + + return 0; +} + +/* + * A test that tries to allocate a memory region that spans over the min_a= ddr + * and max_addr range and overlaps with two different nodes, where the req= uested + * node ends before min_addr: + * + * min_addr + * | max_addr + * | | + * v v + * | +---------------+ +-------------+---------+ | + * | | requested | | node1 | node2 | | + * +----+---------------+--------+-------------+---------+----------+ + * + + + * | +---------+ | + * | | rgn | | + * +----------+---------+-------------------------------------------+ + * + * Expect to drop the lower limit and allocate a memory region that ends at + * the end of the requested node. + */ +static int alloc_exact_nid_top_down_numa_no_overlap_split_check(void) +{ + int nid_req =3D 2; + struct memblock_region *new_rgn =3D &memblock.reserved.regions[0]; + struct memblock_region *req_node =3D &memblock.memory.regions[nid_req]; + struct memblock_region *node2 =3D &memblock.memory.regions[6]; + void *allocated_ptr =3D NULL; + phys_addr_t size; + phys_addr_t min_addr; + phys_addr_t max_addr; + + PREFIX_PUSH(); + setup_numa_memblock(node_fractions); + + size =3D SZ_512; + min_addr =3D node2->base - SZ_256; + max_addr =3D min_addr + size; + + allocated_ptr =3D memblock_alloc_exact_nid_raw(size, SMP_CACHE_BYTES, + min_addr, max_addr, + nid_req); + + ASSERT_NE(allocated_ptr, NULL); + ASSERT_MEM_NE(allocated_ptr, 0, size); + + ASSERT_EQ(new_rgn->size, size); + ASSERT_EQ(new_rgn->base, region_end(req_node) - size); + ASSERT_LE(req_node->base, new_rgn->base); + + ASSERT_EQ(memblock.reserved.cnt, 1); + ASSERT_EQ(memblock.reserved.total_size, size); + + test_pass_pop(); + + return 0; +} + +/* + * A test that tries to allocate memory within min_addr and max_add range = when + * the requested node and the range do not overlap, and requested node ends + * before min_addr. The range overlaps with multiple nodes along node + * boundaries: + * + * min_addr + * | max_addr + * | | + * v v + * |-----------+ +----------+----...----+----------+ | + * | requested | | min node | ... | max node | | + * +-----------+-----------+----------+----...----+----------+------+ + * + + + * | +-----+ | + * | | rgn | | + * +-----+-----+----------------------------------------------------+ + * + * Expect to drop the lower limit and allocate a memory region that ends at + * the end of the requested node. + */ +static int alloc_exact_nid_top_down_numa_no_overlap_low_check(void) +{ + int nid_req =3D 0; + struct memblock_region *new_rgn =3D &memblock.reserved.regions[0]; + struct memblock_region *req_node =3D &memblock.memory.regions[nid_req]; + struct memblock_region *min_node =3D &memblock.memory.regions[2]; + struct memblock_region *max_node =3D &memblock.memory.regions[5]; + void *allocated_ptr =3D NULL; + phys_addr_t size =3D SZ_64; + phys_addr_t max_addr; + phys_addr_t min_addr; + + PREFIX_PUSH(); + setup_numa_memblock(node_fractions); + + min_addr =3D min_node->base; + max_addr =3D region_end(max_node); + + allocated_ptr =3D memblock_alloc_exact_nid_raw(size, SMP_CACHE_BYTES, + min_addr, max_addr, + nid_req); + + ASSERT_NE(allocated_ptr, NULL); + ASSERT_MEM_NE(allocated_ptr, 0, size); + + ASSERT_EQ(new_rgn->size, size); + ASSERT_EQ(new_rgn->base, region_end(req_node) - size); + + ASSERT_EQ(memblock.reserved.cnt, 1); + ASSERT_EQ(memblock.reserved.total_size, size); + + test_pass_pop(); + + return 0; +} + +/* Test case wrappers for NUMA tests */ +static int alloc_exact_nid_numa_simple_check(void) +{ + test_print("\tRunning %s...\n", __func__); + memblock_set_bottom_up(false); + alloc_exact_nid_top_down_numa_simple_check(); + + return 0; +} + +static int alloc_exact_nid_numa_part_reserved_check(void) +{ + test_print("\tRunning %s...\n", __func__); + memblock_set_bottom_up(false); + alloc_exact_nid_top_down_numa_part_reserved_check(); + + return 0; +} + +static int alloc_exact_nid_numa_split_range_low_check(void) +{ + test_print("\tRunning %s...\n", __func__); + memblock_set_bottom_up(false); + alloc_exact_nid_top_down_numa_split_range_low_check(); + + return 0; +} + +static int alloc_exact_nid_numa_no_overlap_split_check(void) +{ + test_print("\tRunning %s...\n", __func__); + memblock_set_bottom_up(false); + alloc_exact_nid_top_down_numa_no_overlap_split_check(); + + return 0; +} + +static int alloc_exact_nid_numa_no_overlap_low_check(void) +{ + test_print("\tRunning %s...\n", __func__); + memblock_set_bottom_up(false); + alloc_exact_nid_top_down_numa_no_overlap_low_check(); + + return 0; +} + +int __memblock_alloc_exact_nid_numa_checks(void) +{ + test_print("Running %s NUMA tests...\n", FUNC_NAME); + + alloc_exact_nid_numa_simple_check(); + alloc_exact_nid_numa_part_reserved_check(); + alloc_exact_nid_numa_split_range_low_check(); + alloc_exact_nid_numa_no_overlap_split_check(); + alloc_exact_nid_numa_no_overlap_low_check(); + + return 0; +} + int memblock_alloc_exact_nid_checks(void) { prefix_reset(); @@ -13,6 +356,7 @@ int memblock_alloc_exact_nid_checks(void) dummy_physical_memory_init(); =20 memblock_alloc_exact_nid_range_checks(); + memblock_alloc_exact_nid_numa_checks(); =20 dummy_physical_memory_cleanup(); =20 diff --git a/tools/testing/memblock/tests/alloc_exact_nid_api.h b/tools/tes= ting/memblock/tests/alloc_exact_nid_api.h index 4408719de3b9..cef419d55d2a 100644 --- a/tools/testing/memblock/tests/alloc_exact_nid_api.h +++ b/tools/testing/memblock/tests/alloc_exact_nid_api.h @@ -5,5 +5,21 @@ #include "common.h" =20 int memblock_alloc_exact_nid_checks(void); +int __memblock_alloc_exact_nid_numa_checks(void); + +#ifdef CONFIG_NUMA +static inline int memblock_alloc_exact_nid_numa_checks(void) +{ + __memblock_alloc_exact_nid_numa_checks(); + return 0; +} + +#else +static inline int memblock_alloc_exact_nid_numa_checks(void) +{ + return 0; +} + +#endif /* CONFIG_NUMA */ =20 #endif --=20 2.25.1