slub.txt 12.9 KB
Newer Older
Christoph Lameter's avatar
Christoph Lameter committed
1 2 3 4 5
Short users guide for SLUB
--------------------------

The basic philosophy of SLUB is very different from SLAB. SLAB
requires rebuilding the kernel to activate debug options for all
Christoph Lameter's avatar
Christoph Lameter committed
6
slab caches. SLUB always includes full debugging but it is off by default.
Christoph Lameter's avatar
Christoph Lameter committed
7 8 9 10
SLUB can enable debugging only for selected slabs in order to avoid
an impact on overall system performance which may make a bug more
difficult to find.

11
In order to switch debugging on one can add an option "slub_debug"
Christoph Lameter's avatar
Christoph Lameter committed
12 13 14 15 16 17 18 19
to the kernel command line. That will enable full debugging for
all slabs.

Typically one would then use the "slabinfo" command to get statistical
data and perform operation on the slabs. By default slabinfo only lists
slabs that have data in them. See "slabinfo -h" for more options when
running the command. slabinfo can be compiled with

20
gcc -o slabinfo tools/vm/slabinfo.c
Christoph Lameter's avatar
Christoph Lameter committed
21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37

Some of the modes of operation of slabinfo require that slub debugging
be enabled on the command line. F.e. no tracking information will be
available without debugging on and validation can only partially
be performed if debugging was not switched on.

Some more sophisticated uses of slub_debug:
-------------------------------------------

Parameters may be given to slub_debug. If none is specified then full
debugging is enabled. Format:

slub_debug=<Debug-Options>       Enable options for all slabs
slub_debug=<Debug-Options>,<slab name>
				Enable options only for select slabs

Possible debug options are
38 39
	F		Sanity checks on (enables SLAB_DEBUG_CONSISTENCY_CHECKS
			Sorry SLAB legacy issues)
Christoph Lameter's avatar
Christoph Lameter committed
40 41 42 43
	Z		Red zoning
	P		Poisoning (object and padding)
	U		User tracking (free and alloc)
	T		Trace (please only use on single slabs)
44
	A		Toggle failslab filter mark for the cache
45 46
	O		Switch debugging off for caches that would have
			caused higher minimum slab orders
47 48
	-		Switch all debugging off (useful if the kernel is
			configured with CONFIG_SLUB_DEBUG_ON)
Christoph Lameter's avatar
Christoph Lameter committed
49 50 51 52 53 54 55

F.e. in order to boot just with sanity checks and red zoning one would specify:

	slub_debug=FZ

Trying to find an issue in the dentry cache? Try

56
	slub_debug=,dentry
Christoph Lameter's avatar
Christoph Lameter committed
57 58 59 60 61 62

to only enable debugging on the dentry cache.

Red zoning and tracking may realign the slab.  We can just apply sanity checks
to the dentry cache with

63
	slub_debug=F,dentry
Christoph Lameter's avatar
Christoph Lameter committed
64

65 66 67 68 69 70 71 72
Debugging options may require the minimum possible slab order to increase as
a result of storing the metadata (for example, caches with PAGE_SIZE object
sizes).  This has a higher liklihood of resulting in slab allocation errors
in low memory situations or if there's high fragmentation of memory.  To
switch off debugging for such caches by default, use

	slub_debug=O

Christoph Lameter's avatar
Christoph Lameter committed
73 74 75 76
In case you forgot to enable debugging on the kernel command line: It is
possible to enable debugging manually when the kernel is up. Look at the
contents of:

77
/sys/kernel/slab/<slab name>/
Christoph Lameter's avatar
Christoph Lameter committed
78 79 80 81 82 83 84 85 86 87

Look at the writable files. Writing 1 to them will enable the
corresponding debug option. All options can be set on a slab that does
not contain objects. If the slab already contains objects then sanity checks
and tracing may only be enabled. The other options may cause the realignment
of objects.

Careful with tracing: It may spew out lots of information and never stop if
used on the wrong slab.

Christoph Lameter's avatar
Christoph Lameter committed
88
Slab merging
Christoph Lameter's avatar
Christoph Lameter committed
89 90
------------

Christoph Lameter's avatar
Christoph Lameter committed
91
If no debug options are specified then SLUB may merge similar slabs together
Christoph Lameter's avatar
Christoph Lameter committed
92 93 94
in order to reduce overhead and increase cache hotness of objects.
slabinfo -a displays which slabs were merged together.

Christoph Lameter's avatar
Christoph Lameter committed
95 96 97 98 99 100 101 102 103 104 105 106 107 108 109
Slab validation
---------------

SLUB can validate all object if the kernel was booted with slub_debug. In
order to do so you must have the slabinfo tool. Then you can do

slabinfo -v

which will test all objects. Output will be generated to the syslog.

This also works in a more limited way if boot was without slab debug.
In that case slabinfo -v simply tests all reachable objects. Usually
these are in the cpu slabs and the partial slabs. Full slabs are not
tracked by SLUB in a non debug situation.

Christoph Lameter's avatar
Christoph Lameter committed
110 111 112 113 114 115 116 117
Getting more performance
------------------------

To some degree SLUB's performance is limited by the need to take the
list_lock once in a while to deal with partial slabs. That overhead is
governed by the order of the allocation for each slab. The allocations
can be influenced by kernel parameters:

Christoph Lameter's avatar
Christoph Lameter committed
118
slub_min_objects=x		(default 4)
Christoph Lameter's avatar
Christoph Lameter committed
119
slub_min_order=x		(default 0)
120
slub_max_order=x		(default 3 (PAGE_ALLOC_COSTLY_ORDER))
Christoph Lameter's avatar
Christoph Lameter committed
121 122 123 124 125 126 127 128 129 130 131 132 133

slub_min_objects allows to specify how many objects must at least fit
into one slab in order for the allocation order to be acceptable.
In general slub will be able to perform this number of allocations
on a slab without consulting centralized resources (list_lock) where
contention may occur.

slub_min_order specifies a minim order of slabs. A similar effect like
slub_min_objects.

slub_max_order specified the order at which slub_min_objects should no
longer be checked. This is useful to avoid SLUB trying to generate
super large order pages to fit slub_min_objects of a slab cache with
134 135 136 137
large object sizes into one high order page. Setting command line
parameter debug_guardpage_minorder=N (N > 0), forces setting
slub_max_order to 0, what cause minimum possible order of slabs
allocation.
Christoph Lameter's avatar
Christoph Lameter committed
138

Christoph Lameter's avatar
Christoph Lameter committed
139 140 141 142 143
SLUB Debug output
-----------------

Here is a sample of slub debug output:

144 145 146 147 148 149 150 151 152 153 154 155 156 157
====================================================================
BUG kmalloc-8: Redzone overwritten
--------------------------------------------------------------------

INFO: 0xc90f6d28-0xc90f6d2b. First byte 0x00 instead of 0xcc
INFO: Slab 0xc528c530 flags=0x400000c3 inuse=61 fp=0xc90f6d58
INFO: Object 0xc90f6d20 @offset=3360 fp=0xc90f6d58
INFO: Allocated in get_modalias+0x61/0xf5 age=53 cpu=1 pid=554

Bytes b4 0xc90f6d10:  00 00 00 00 00 00 00 00 5a 5a 5a 5a 5a 5a 5a 5a ........ZZZZZZZZ
  Object 0xc90f6d20:  31 30 31 39 2e 30 30 35                         1019.005
 Redzone 0xc90f6d28:  00 cc cc cc                                     .
 Padding 0xc90f6d50:  5a 5a 5a 5a 5a 5a 5a 5a                         ZZZZZZZZ

Christoph Lameter's avatar
Christoph Lameter committed
158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179
  [<c010523d>] dump_trace+0x63/0x1eb
  [<c01053df>] show_trace_log_lvl+0x1a/0x2f
  [<c010601d>] show_trace+0x12/0x14
  [<c0106035>] dump_stack+0x16/0x18
  [<c017e0fa>] object_err+0x143/0x14b
  [<c017e2cc>] check_object+0x66/0x234
  [<c017eb43>] __slab_free+0x239/0x384
  [<c017f446>] kfree+0xa6/0xc6
  [<c02e2335>] get_modalias+0xb9/0xf5
  [<c02e23b7>] dmi_dev_uevent+0x27/0x3c
  [<c027866a>] dev_uevent+0x1ad/0x1da
  [<c0205024>] kobject_uevent_env+0x20a/0x45b
  [<c020527f>] kobject_uevent+0xa/0xf
  [<c02779f1>] store_uevent+0x4f/0x58
  [<c027758e>] dev_attr_store+0x29/0x2f
  [<c01bec4f>] sysfs_write_file+0x16e/0x19c
  [<c0183ba7>] vfs_write+0xd1/0x15a
  [<c01841d7>] sys_write+0x3d/0x72
  [<c0104112>] sysenter_past_esp+0x5f/0x99
  [<b7f7b410>] 0xb7f7b410
  =======================

180
FIX kmalloc-8: Restoring Redzone 0xc90f6d28-0xc90f6d2b=0xcc
Christoph Lameter's avatar
Christoph Lameter committed
181

182 183 184
If SLUB encounters a corrupted object (full detection requires the kernel
to be booted with slub_debug) then the following output will be dumped
into the syslog:
Christoph Lameter's avatar
Christoph Lameter committed
185

186
1. Description of the problem encountered
Christoph Lameter's avatar
Christoph Lameter committed
187 188 189

This will be a message in the system log starting with

190 191 192
===============================================
BUG <slab cache affected>: <What went wrong>
-----------------------------------------------
Christoph Lameter's avatar
Christoph Lameter committed
193

194 195 196 197 198 199 200
INFO: <corruption start>-<corruption_end> <more info>
INFO: Slab <address> <slab information>
INFO: Object <address> <object information>
INFO: Allocated in <kernel function> age=<jiffies since alloc> cpu=<allocated by
	cpu> pid=<pid of the process>
INFO: Freed in <kernel function> age=<jiffies since free> cpu=<freed by cpu>
	 pid=<pid of the process>
Christoph Lameter's avatar
Christoph Lameter committed
201

202 203
(Object allocation / free information is only available if SLAB_STORE_USER is
set for the slab. slub_debug sets that option)
Christoph Lameter's avatar
Christoph Lameter committed
204

205
2. The object contents if an object was involved.
Christoph Lameter's avatar
Christoph Lameter committed
206

207
Various types of lines can follow the BUG SLUB line:
Christoph Lameter's avatar
Christoph Lameter committed
208 209

Bytes b4 <address> : <bytes>
210
	Shows a few bytes before the object where the problem was detected.
Christoph Lameter's avatar
Christoph Lameter committed
211 212 213 214 215
	Can be useful if the corruption does not stop with the start of the
	object.

Object <address> : <bytes>
	The bytes of the object. If the object is inactive then the bytes
216
	typically contain poison values. Any non-poison value shows a
Christoph Lameter's avatar
Christoph Lameter committed
217 218 219
	corruption by a write after free.

Redzone <address> : <bytes>
220
	The Redzone following the object. The Redzone is used to detect
Christoph Lameter's avatar
Christoph Lameter committed
221 222 223 224
	writes after the object. All bytes should always have the same
	value. If there is any deviation then it is due to a write after
	the object boundary.

225 226
	(Redzone information is only available if SLAB_RED_ZONE is set.
	slub_debug sets that option)
Christoph Lameter's avatar
Christoph Lameter committed
227

228
Padding <address> : <bytes>
Christoph Lameter's avatar
Christoph Lameter committed
229 230
	Unused data to fill up the space in order to get the next object
	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
Christoph Lameter's avatar
Christoph Lameter committed
232 233
	before the object.

234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251
3. A stackdump

The stackdump describes the location where the error was detected. The cause
of the corruption is may be more likely found by looking at the function that
allocated or freed the object.

4. Report on how the problem was dealt with in order to ensure the continued
operation of the system.

These are messages in the system log beginning with

FIX <slab cache affected>: <corrective action taken>

In the above sample SLUB found that the Redzone of an active object has
been overwritten. Here a string of 8 characters was written into a slab that
has the length of 8 characters. However, a 8 character string needs a
terminating 0. That zero has overwritten the first byte of the Redzone field.
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
253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281
system operations continue.

Emergency operations:
---------------------

Minimal debugging (sanity checks alone) can be enabled by booting with

	slub_debug=F

This will be generally be enough to enable the resiliency features of slub
which will keep the system running even if a bad kernel component will
keep corrupting objects. This may be important for production systems.
Performance will be impacted by the sanity checks and there will be a
continual stream of error messages to the syslog but no additional memory
will be used (unlike full debugging).

No guarantees. The kernel component still needs to be fixed. Performance
may be optimized further by locating the slab that experiences corruption
and enabling debugging only for that cache

I.e.

	slub_debug=F,dentry

If the corruption occurs by writing after the end of the object then it
may be advisable to enable a Redzone to avoid corrupting the beginning
of other objects.

	slub_debug=FZ,dentry
Christoph Lameter's avatar
Christoph Lameter committed
282

283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340
Extended slabinfo mode and plotting
-----------------------------------

The slabinfo tool has a special 'extended' ('-X') mode that includes:
 - Slabcache Totals
 - Slabs sorted by size (up to -N <num> slabs, default 1)
 - Slabs sorted by loss (up to -N <num> slabs, default 1)

Additionally, in this mode slabinfo does not dynamically scale sizes (G/M/K)
and reports everything in bytes (this functionality is also available to
other slabinfo modes via '-B' option) which makes reporting more precise and
accurate. Moreover, in some sense the `-X' mode also simplifies the analysis
of slabs' behaviour, because its output can be plotted using the
slabinfo-gnuplot.sh script. So it pushes the analysis from looking through
the numbers (tons of numbers) to something easier -- visual analysis.

To generate plots:
a) collect slabinfo extended records, for example:

  while [ 1 ]; do slabinfo -X >> FOO_STATS; sleep 1; done

b) pass stats file(-s) to slabinfo-gnuplot.sh script:
  slabinfo-gnuplot.sh FOO_STATS [FOO_STATS2 .. FOO_STATSN]

The slabinfo-gnuplot.sh script will pre-processes the collected records
and generates 3 png files (and 3 pre-processing cache files) per STATS
file:
 - Slabcache Totals: FOO_STATS-totals.png
 - Slabs sorted by size: FOO_STATS-slabs-by-size.png
 - Slabs sorted by loss: FOO_STATS-slabs-by-loss.png

Another use case, when slabinfo-gnuplot can be useful, is when you need
to compare slabs' behaviour "prior to" and "after" some code modification.
To help you out there, slabinfo-gnuplot.sh script can 'merge' the
`Slabcache Totals` sections from different measurements. To visually
compare N plots:

a) Collect as many STATS1, STATS2, .. STATSN files as you need
  while [ 1 ]; do slabinfo -X >> STATS<X>; sleep 1; done

b) Pre-process those STATS files
  slabinfo-gnuplot.sh STATS1 STATS2 .. STATSN

c) Execute slabinfo-gnuplot.sh in '-t' mode, passing all of the
generated pre-processed *-totals
  slabinfo-gnuplot.sh -t STATS1-totals STATS2-totals .. STATSN-totals

This will produce a single plot (png file).

Plots, expectedly, can be large so some fluctuations or small spikes
can go unnoticed. To deal with that, `slabinfo-gnuplot.sh' has two
options to 'zoom-in'/'zoom-out':
 a) -s %d,%d  overwrites the default image width and heigh
 b) -r %d,%d  specifies a range of samples to use (for example,
              in `slabinfo -X >> FOO_STATS; sleep 1;' case, using
              a "-r 40,60" range will plot only samples collected
              between 40th and 60th seconds).

Christoph Lameter's avatar
Christoph Lameter committed
341
Christoph Lameter, May 30, 2007
342
Sergey Senozhatsky, October 23, 2015