| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <202205120007.cudgBgVI-lkp@intel.com>
Date: Thu, 12 May 2022 00:16:41 +0800
From: kernel test robot <lkp@...el.com>
To: "Kirill A. Shutemov" <kirill.shutemov@...ux.intel.com>
Cc: kbuild-all@...ts.01.org, linux-kernel@...r.kernel.org
Subject: [intel-tdx:guest-unaccepted-memory 3/15] mm/page_alloc.c:1183:12:
warning: 'page_needs_acceptance' is used uninitialized
tree: https://github.com/intel/tdx.git guest-unaccepted-memory
head: 83a8442434ff3bbf432df7508f1fefd447ca2c86
commit: 0845556ccd9555f5245d3bdd6cb646c2ea3fa9c2 [3/15] mm: Add support for unaccepted memory
config: x86_64-randconfig-a016-20220509 (https://download.01.org/0day-ci/archive/20220512/202205120007.cudgBgVI-lkp@intel.com/config)
compiler: gcc-11 (Debian 11.2.0-20) 11.2.0
reproduce (this is a W=1 build):
# https://github.com/intel/tdx/commit/0845556ccd9555f5245d3bdd6cb646c2ea3fa9c2
git remote add intel-tdx https://github.com/intel/tdx.git
git fetch --no-tags intel-tdx guest-unaccepted-memory
git checkout 0845556ccd9555f5245d3bdd6cb646c2ea3fa9c2
# save the config file
mkdir build_dir && cp config build_dir/.config
make W=1 O=build_dir ARCH=x86_64 SHELL=/bin/bash
If you fix the issue, kindly add following tag as appropriate
Reported-by: kernel test robot <lkp@...el.com>
All warnings (new ones prefixed by >>):
mm/page_alloc.c: In function '__free_one_page':
>> mm/page_alloc.c:1183:12: warning: 'page_needs_acceptance' is used uninitialized [-Wuninitialized]
1183 | if (!page_needs_acceptance && (fpi_flags & FPI_UNACCEPTED_SLOWPATH))
| ^
vim +/page_needs_acceptance +1183 mm/page_alloc.c
1056
1057 /*
1058 * Freeing function for a buddy system allocator.
1059 *
1060 * The concept of a buddy system is to maintain direct-mapped table
1061 * (containing bit values) for memory blocks of various "orders".
1062 * The bottom level table contains the map for the smallest allocatable
1063 * units of memory (here, pages), and each level above it describes
1064 * pairs of units from the levels below, hence, "buddies".
1065 * At a high level, all that happens here is marking the table entry
1066 * at the bottom level available, and propagating the changes upward
1067 * as necessary, plus some accounting needed to play nicely with other
1068 * parts of the VM system.
1069 * At each level, we keep a list of pages, which are heads of continuous
1070 * free pages of length of (1 << order) and marked with PageBuddy.
1071 * Page's order is recorded in page_private(page) field.
1072 * So when we are allocating or freeing one, we can derive the state of the
1073 * other. That is, if we allocate a small block, and both were
1074 * free, the remainder of the region must be split into blocks.
1075 * If a block is freed, and its buddy is also free, then this
1076 * triggers coalescing into a block of larger size.
1077 *
1078 * -- nyc
1079 */
1080
1081 static inline void __free_one_page(struct page *page,
1082 unsigned long pfn,
1083 struct zone *zone, unsigned int order,
1084 int migratetype, fpi_t fpi_flags)
1085 {
1086 struct capture_control *capc = task_capc(zone);
1087 unsigned int max_order = pageblock_order;
1088 unsigned long buddy_pfn;
1089 unsigned long combined_pfn;
1090 struct page *buddy;
1091 bool to_tail;
1092 bool page_needs_acceptance;
1093
1094 VM_BUG_ON(!zone_is_initialized(zone));
1095 VM_BUG_ON_PAGE(page->flags & PAGE_FLAGS_CHECK_AT_PREP, page);
1096
1097 VM_BUG_ON(migratetype == -1);
1098 if (likely(!is_migrate_isolate(migratetype)))
1099 __mod_zone_freepage_state(zone, 1 << order, migratetype);
1100
1101 VM_BUG_ON_PAGE(pfn & ((1 << order) - 1), page);
1102 VM_BUG_ON_PAGE(bad_range(zone, page), page);
1103
1104 if (PageUnaccepted(page)) {
1105 page_needs_acceptance = true;
1106 __ClearPageUnaccepted(page);
1107 }
1108
1109 continue_merging:
1110 while (order < max_order) {
1111 if (compaction_capture(capc, page, order, migratetype)) {
1112 __mod_zone_freepage_state(zone, -(1 << order),
1113 migratetype);
1114 return;
1115 }
1116 buddy_pfn = __find_buddy_pfn(pfn, order);
1117 buddy = page + (buddy_pfn - pfn);
1118
1119 if (!page_is_buddy(page, buddy, order))
1120 goto done_merging;
1121 /*
1122 * Our buddy is free or it is CONFIG_DEBUG_PAGEALLOC guard page,
1123 * merge with it and move up one order.
1124 */
1125 if (page_is_guard(buddy))
1126 clear_page_guard(zone, buddy, order, migratetype);
1127 else
1128 del_page_from_free_list(buddy, zone, order);
1129
1130 /* Mark page unaccepted if any of merged pages were unaccepted */
1131 if (PageUnaccepted(buddy)) {
1132 page_needs_acceptance = true;
1133 __ClearPageUnaccepted(buddy);
1134 }
1135
1136 combined_pfn = buddy_pfn & pfn;
1137 page = page + (combined_pfn - pfn);
1138 pfn = combined_pfn;
1139 order++;
1140 }
1141 if (order < MAX_ORDER - 1) {
1142 /* If we are here, it means order is >= pageblock_order.
1143 * We want to prevent merge between freepages on pageblock
1144 * without fallbacks and normal pageblock. Without this,
1145 * pageblock isolation could cause incorrect freepage or CMA
1146 * accounting or HIGHATOMIC accounting.
1147 *
1148 * We don't want to hit this code for the more frequent
1149 * low-order merging.
1150 */
1151 int buddy_mt;
1152
1153 buddy_pfn = __find_buddy_pfn(pfn, order);
1154 buddy = page + (buddy_pfn - pfn);
1155
1156 if (!page_is_buddy(page, buddy, order))
1157 goto done_merging;
1158 buddy_mt = get_pageblock_migratetype(buddy);
1159
1160 if (migratetype != buddy_mt
1161 && (!migratetype_is_mergeable(migratetype) ||
1162 !migratetype_is_mergeable(buddy_mt)))
1163 goto done_merging;
1164 max_order = order + 1;
1165 goto continue_merging;
1166 }
1167
1168 done_merging:
1169 set_buddy_order(page, order);
1170
1171 /*
1172 * The page gets marked as PageUnaccepted() if any of merged-in pages
1173 * is PageUnaccepted().
1174 *
1175 * New pages, just being added to buddy allocator, do not have
1176 * PageUnaccepted() set. FPI_UNACCEPTED_SLOWPATH indicates that the
1177 * page is new and page_is_unaccepted() check is required to
1178 * determinate if acceptance is required.
1179 *
1180 * Avoid calling page_is_unaccepted() if it is known that the page
1181 * needs acceptance. It can be costly.
1182 */
> 1183 if (!page_needs_acceptance && (fpi_flags & FPI_UNACCEPTED_SLOWPATH))
1184 page_needs_acceptance = page_is_unaccepted(page, order);
1185 if (page_needs_acceptance)
1186 __SetPageUnaccepted(page);
1187
1188 if (fpi_flags & FPI_TO_TAIL)
1189 to_tail = true;
1190 else if (is_shuffle_order(order))
1191 to_tail = shuffle_pick_tail();
1192 else
1193 to_tail = buddy_merge_likely(pfn, buddy_pfn, page, order);
1194
1195 if (to_tail)
1196 add_to_free_list_tail(page, zone, order, migratetype);
1197 else
1198 add_to_free_list(page, zone, order, migratetype);
1199
1200 /* Notify page reporting subsystem of freed page */
1201 if (!(fpi_flags & FPI_SKIP_REPORT_NOTIFY))
1202 page_reporting_notify_free(order);
1203 }
1204
--
0-DAY CI Kernel Test Service
https://01.org/lkp
Powered by blists - more mailing lists