Based on kernel version 4.16.1. Page generated on 2018-04-09 11:53 EST.
1 Short users guide for SLUB 2 -------------------------- 3 4 The basic philosophy of SLUB is very different from SLAB. SLAB 5 requires rebuilding the kernel to activate debug options for all 6 slab caches. SLUB always includes full debugging but it is off by default. 7 SLUB can enable debugging only for selected slabs in order to avoid 8 an impact on overall system performance which may make a bug more 9 difficult to find. 10 11 In order to switch debugging on one can add an option "slub_debug" 12 to the kernel command line. That will enable full debugging for 13 all slabs. 14 15 Typically one would then use the "slabinfo" command to get statistical 16 data and perform operation on the slabs. By default slabinfo only lists 17 slabs that have data in them. See "slabinfo -h" for more options when 18 running the command. slabinfo can be compiled with 19 20 gcc -o slabinfo tools/vm/slabinfo.c 21 22 Some of the modes of operation of slabinfo require that slub debugging 23 be enabled on the command line. F.e. no tracking information will be 24 available without debugging on and validation can only partially 25 be performed if debugging was not switched on. 26 27 Some more sophisticated uses of slub_debug: 28 ------------------------------------------- 29 30 Parameters may be given to slub_debug. If none is specified then full 31 debugging is enabled. Format: 32 33 slub_debug=<Debug-Options> Enable options for all slabs 34 slub_debug=<Debug-Options>,<slab name> 35 Enable options only for select slabs 36 37 Possible debug options are 38 F Sanity checks on (enables SLAB_DEBUG_CONSISTENCY_CHECKS 39 Sorry SLAB legacy issues) 40 Z Red zoning 41 P Poisoning (object and padding) 42 U User tracking (free and alloc) 43 T Trace (please only use on single slabs) 44 A Toggle failslab filter mark for the cache 45 O Switch debugging off for caches that would have 46 caused higher minimum slab orders 47 - Switch all debugging off (useful if the kernel is 48 configured with CONFIG_SLUB_DEBUG_ON) 49 50 F.e. in order to boot just with sanity checks and red zoning one would specify: 51 52 slub_debug=FZ 53 54 Trying to find an issue in the dentry cache? Try 55 56 slub_debug=,dentry 57 58 to only enable debugging on the dentry cache. 59 60 Red zoning and tracking may realign the slab. We can just apply sanity checks 61 to the dentry cache with 62 63 slub_debug=F,dentry 64 65 Debugging options may require the minimum possible slab order to increase as 66 a result of storing the metadata (for example, caches with PAGE_SIZE object 67 sizes). This has a higher liklihood of resulting in slab allocation errors 68 in low memory situations or if there's high fragmentation of memory. To 69 switch off debugging for such caches by default, use 70 71 slub_debug=O 72 73 In case you forgot to enable debugging on the kernel command line: It is 74 possible to enable debugging manually when the kernel is up. Look at the 75 contents of: 76 77 /sys/kernel/slab/<slab name>/ 78 79 Look at the writable files. Writing 1 to them will enable the 80 corresponding debug option. All options can be set on a slab that does 81 not contain objects. If the slab already contains objects then sanity checks 82 and tracing may only be enabled. The other options may cause the realignment 83 of objects. 84 85 Careful with tracing: It may spew out lots of information and never stop if 86 used on the wrong slab. 87 88 Slab merging 89 ------------ 90 91 If no debug options are specified then SLUB may merge similar slabs together 92 in order to reduce overhead and increase cache hotness of objects. 93 slabinfo -a displays which slabs were merged together. 94 95 Slab validation 96 --------------- 97 98 SLUB can validate all object if the kernel was booted with slub_debug. In 99 order to do so you must have the slabinfo tool. Then you can do 100 101 slabinfo -v 102 103 which will test all objects. Output will be generated to the syslog. 104 105 This also works in a more limited way if boot was without slab debug. 106 In that case slabinfo -v simply tests all reachable objects. Usually 107 these are in the cpu slabs and the partial slabs. Full slabs are not 108 tracked by SLUB in a non debug situation. 109 110 Getting more performance 111 ------------------------ 112 113 To some degree SLUB's performance is limited by the need to take the 114 list_lock once in a while to deal with partial slabs. That overhead is 115 governed by the order of the allocation for each slab. The allocations 116 can be influenced by kernel parameters: 117 118 slub_min_objects=x (default 4) 119 slub_min_order=x (default 0) 120 slub_max_order=x (default 3 (PAGE_ALLOC_COSTLY_ORDER)) 121 122 slub_min_objects allows to specify how many objects must at least fit 123 into one slab in order for the allocation order to be acceptable. 124 In general slub will be able to perform this number of allocations 125 on a slab without consulting centralized resources (list_lock) where 126 contention may occur. 127 128 slub_min_order specifies a minim order of slabs. A similar effect like 129 slub_min_objects. 130 131 slub_max_order specified the order at which slub_min_objects should no 132 longer be checked. This is useful to avoid SLUB trying to generate 133 super large order pages to fit slub_min_objects of a slab cache with 134 large object sizes into one high order page. Setting command line 135 parameter debug_guardpage_minorder=N (N > 0), forces setting 136 slub_max_order to 0, what cause minimum possible order of slabs 137 allocation. 138 139 SLUB Debug output 140 ----------------- 141 142 Here is a sample of slub debug output: 143 144 ==================================================================== 145 BUG kmalloc-8: Redzone overwritten 146 -------------------------------------------------------------------- 147 148 INFO: 0xc90f6d28-0xc90f6d2b. First byte 0x00 instead of 0xcc 149 INFO: Slab 0xc528c530 flags=0x400000c3 inuse=61 fp=0xc90f6d58 150 INFO: Object 0xc90f6d20 @offset=3360 fp=0xc90f6d58 151 INFO: Allocated in get_modalias+0x61/0xf5 age=53 cpu=1 pid=554 152 153 Bytes b4 0xc90f6d10: 00 00 00 00 00 00 00 00 5a 5a 5a 5a 5a 5a 5a 5a ........ZZZZZZZZ 154 Object 0xc90f6d20: 31 30 31 39 2e 30 30 35 1019.005 155 Redzone 0xc90f6d28: 00 cc cc cc . 156 Padding 0xc90f6d50: 5a 5a 5a 5a 5a 5a 5a 5a ZZZZZZZZ 157 158 [<c010523d>] dump_trace+0x63/0x1eb 159 [<c01053df>] show_trace_log_lvl+0x1a/0x2f 160 [<c010601d>] show_trace+0x12/0x14 161 [<c0106035>] dump_stack+0x16/0x18 162 [<c017e0fa>] object_err+0x143/0x14b 163 [<c017e2cc>] check_object+0x66/0x234 164 [<c017eb43>] __slab_free+0x239/0x384 165 [<c017f446>] kfree+0xa6/0xc6 166 [<c02e2335>] get_modalias+0xb9/0xf5 167 [<c02e23b7>] dmi_dev_uevent+0x27/0x3c 168 [<c027866a>] dev_uevent+0x1ad/0x1da 169 [<c0205024>] kobject_uevent_env+0x20a/0x45b 170 [<c020527f>] kobject_uevent+0xa/0xf 171 [<c02779f1>] store_uevent+0x4f/0x58 172 [<c027758e>] dev_attr_store+0x29/0x2f 173 [<c01bec4f>] sysfs_write_file+0x16e/0x19c 174 [<c0183ba7>] vfs_write+0xd1/0x15a 175 [<c01841d7>] sys_write+0x3d/0x72 176 [<c0104112>] sysenter_past_esp+0x5f/0x99 177 [<b7f7b410>] 0xb7f7b410 178 ======================= 179 180 FIX kmalloc-8: Restoring Redzone 0xc90f6d28-0xc90f6d2b=0xcc 181 182 If SLUB encounters a corrupted object (full detection requires the kernel 183 to be booted with slub_debug) then the following output will be dumped 184 into the syslog: 185 186 1. Description of the problem encountered 187 188 This will be a message in the system log starting with 189 190 =============================================== 191 BUG <slab cache affected>: <What went wrong> 192 ----------------------------------------------- 193 194 INFO: <corruption start>-<corruption_end> <more info> 195 INFO: Slab <address> <slab information> 196 INFO: Object <address> <object information> 197 INFO: Allocated in <kernel function> age=<jiffies since alloc> cpu=<allocated by 198 cpu> pid=<pid of the process> 199 INFO: Freed in <kernel function> age=<jiffies since free> cpu=<freed by cpu> 200 pid=<pid of the process> 201 202 (Object allocation / free information is only available if SLAB_STORE_USER is 203 set for the slab. slub_debug sets that option) 204 205 2. The object contents if an object was involved. 206 207 Various types of lines can follow the BUG SLUB line: 208 209 Bytes b4 <address> : <bytes> 210 Shows a few bytes before the object where the problem was detected. 211 Can be useful if the corruption does not stop with the start of the 212 object. 213 214 Object <address> : <bytes> 215 The bytes of the object. If the object is inactive then the bytes 216 typically contain poison values. Any non-poison value shows a 217 corruption by a write after free. 218 219 Redzone <address> : <bytes> 220 The Redzone following the object. The Redzone is used to detect 221 writes after the object. All bytes should always have the same 222 value. If there is any deviation then it is due to a write after 223 the object boundary. 224 225 (Redzone information is only available if SLAB_RED_ZONE is set. 226 slub_debug sets that option) 227 228 Padding <address> : <bytes> 229 Unused data to fill up the space in order to get the next object 230 properly aligned. In the debug case we make sure that there are 231 at least 4 bytes of padding. This allows the detection of writes 232 before the object. 233 234 3. A stackdump 235 236 The stackdump describes the location where the error was detected. The cause 237 of the corruption is may be more likely found by looking at the function that 238 allocated or freed the object. 239 240 4. Report on how the problem was dealt with in order to ensure the continued 241 operation of the system. 242 243 These are messages in the system log beginning with 244 245 FIX <slab cache affected>: <corrective action taken> 246 247 In the above sample SLUB found that the Redzone of an active object has 248 been overwritten. Here a string of 8 characters was written into a slab that 249 has the length of 8 characters. However, a 8 character string needs a 250 terminating 0. That zero has overwritten the first byte of the Redzone field. 251 After reporting the details of the issue encountered the FIX SLUB message 252 tells us that SLUB has restored the Redzone to its proper value and then 253 system operations continue. 254 255 Emergency operations: 256 --------------------- 257 258 Minimal debugging (sanity checks alone) can be enabled by booting with 259 260 slub_debug=F 261 262 This will be generally be enough to enable the resiliency features of slub 263 which will keep the system running even if a bad kernel component will 264 keep corrupting objects. This may be important for production systems. 265 Performance will be impacted by the sanity checks and there will be a 266 continual stream of error messages to the syslog but no additional memory 267 will be used (unlike full debugging). 268 269 No guarantees. The kernel component still needs to be fixed. Performance 270 may be optimized further by locating the slab that experiences corruption 271 and enabling debugging only for that cache 272 273 I.e. 274 275 slub_debug=F,dentry 276 277 If the corruption occurs by writing after the end of the object then it 278 may be advisable to enable a Redzone to avoid corrupting the beginning 279 of other objects. 280 281 slub_debug=FZ,dentry 282 283 Extended slabinfo mode and plotting 284 ----------------------------------- 285 286 The slabinfo tool has a special 'extended' ('-X') mode that includes: 287 - Slabcache Totals 288 - Slabs sorted by size (up to -N <num> slabs, default 1) 289 - Slabs sorted by loss (up to -N <num> slabs, default 1) 290 291 Additionally, in this mode slabinfo does not dynamically scale sizes (G/M/K) 292 and reports everything in bytes (this functionality is also available to 293 other slabinfo modes via '-B' option) which makes reporting more precise and 294 accurate. Moreover, in some sense the `-X' mode also simplifies the analysis 295 of slabs' behaviour, because its output can be plotted using the 296 slabinfo-gnuplot.sh script. So it pushes the analysis from looking through 297 the numbers (tons of numbers) to something easier -- visual analysis. 298 299 To generate plots: 300 a) collect slabinfo extended records, for example: 301 302 while [ 1 ]; do slabinfo -X >> FOO_STATS; sleep 1; done 303 304 b) pass stats file(-s) to slabinfo-gnuplot.sh script: 305 slabinfo-gnuplot.sh FOO_STATS [FOO_STATS2 .. FOO_STATSN] 306 307 The slabinfo-gnuplot.sh script will pre-processes the collected records 308 and generates 3 png files (and 3 pre-processing cache files) per STATS 309 file: 310 - Slabcache Totals: FOO_STATS-totals.png 311 - Slabs sorted by size: FOO_STATS-slabs-by-size.png 312 - Slabs sorted by loss: FOO_STATS-slabs-by-loss.png 313 314 Another use case, when slabinfo-gnuplot can be useful, is when you need 315 to compare slabs' behaviour "prior to" and "after" some code modification. 316 To help you out there, slabinfo-gnuplot.sh script can 'merge' the 317 `Slabcache Totals` sections from different measurements. To visually 318 compare N plots: 319 320 a) Collect as many STATS1, STATS2, .. STATSN files as you need 321 while [ 1 ]; do slabinfo -X >> STATS<X>; sleep 1; done 322 323 b) Pre-process those STATS files 324 slabinfo-gnuplot.sh STATS1 STATS2 .. STATSN 325 326 c) Execute slabinfo-gnuplot.sh in '-t' mode, passing all of the 327 generated pre-processed *-totals 328 slabinfo-gnuplot.sh -t STATS1-totals STATS2-totals .. STATSN-totals 329 330 This will produce a single plot (png file). 331 332 Plots, expectedly, can be large so some fluctuations or small spikes 333 can go unnoticed. To deal with that, `slabinfo-gnuplot.sh' has two 334 options to 'zoom-in'/'zoom-out': 335 a) -s %d,%d overwrites the default image width and heigh 336 b) -r %d,%d specifies a range of samples to use (for example, 337 in `slabinfo -X >> FOO_STATS; sleep 1;' case, using 338 a "-r 40,60" range will plot only samples collected 339 between 40th and 60th seconds). 340 341 Christoph Lameter, May 30, 2007 342 Sergey Senozhatsky, October 23, 2015