diff options
author | Christoph Lameter <clameter@sgi.com> | 2007-07-17 07:03:18 -0400 |
---|---|---|
committer | Linus Torvalds <torvalds@woody.linux-foundation.org> | 2007-07-17 13:23:01 -0400 |
commit | 2492268472e7d326a6fe10f92f9211c4578f2482 (patch) | |
tree | 5f668469190b96bc0db13f836d774ae73cf385ca /Documentation | |
parent | 8e1f936b73150f5095448a0fee6d4f30a1f9001d (diff) |
SLUB: change error reporting format to follow lockdep loosely
Changes the error reporting format to loosely follow lockdep.
If data corruption is detected then we generate the following lines:
============================================
BUG <slab-cache>: <problem>
--------------------------------------------
INFO: <more information> [possibly multiple times]
<object dump>
FIX <slab-cache>: <remedial action>
This also adds some more intelligence to the data corruption detection. Its
now capable of figuring out the start and end.
Add a comment on how to configure SLUB so that a production system may
continue to operate even though occasional slab corruption occur through
a misbehaving kernel component. See "Emergency operations" in
Documentation/vm/slub.txt.
[akpm@linux-foundation.org: build fix]
Signed-off-by: Christoph Lameter <clameter@sgi.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Diffstat (limited to 'Documentation')
-rw-r--r-- | Documentation/vm/slub.txt | 137 |
1 files changed, 89 insertions, 48 deletions
diff --git a/Documentation/vm/slub.txt b/Documentation/vm/slub.txt index df812b03b65d..d17f324db9f5 100644 --- a/Documentation/vm/slub.txt +++ b/Documentation/vm/slub.txt | |||
@@ -127,13 +127,20 @@ SLUB Debug output | |||
127 | 127 | ||
128 | Here is a sample of slub debug output: | 128 | Here is a sample of slub debug output: |
129 | 129 | ||
130 | *** SLUB kmalloc-8: Redzone Active@0xc90f6d20 slab 0xc528c530 offset=3360 flags=0x400000c3 inuse=61 freelist=0xc90f6d58 | 130 | ==================================================================== |
131 | Bytes b4 0xc90f6d10: 00 00 00 00 00 00 00 00 5a 5a 5a 5a 5a 5a 5a 5a ........ZZZZZZZZ | 131 | BUG kmalloc-8: Redzone overwritten |
132 | Object 0xc90f6d20: 31 30 31 39 2e 30 30 35 1019.005 | 132 | -------------------------------------------------------------------- |
133 | Redzone 0xc90f6d28: 00 cc cc cc . | 133 | |
134 | FreePointer 0xc90f6d2c -> 0xc90f6d58 | 134 | INFO: 0xc90f6d28-0xc90f6d2b. First byte 0x00 instead of 0xcc |
135 | Last alloc: get_modalias+0x61/0xf5 jiffies_ago=53 cpu=1 pid=554 | 135 | INFO: Slab 0xc528c530 flags=0x400000c3 inuse=61 fp=0xc90f6d58 |
136 | Filler 0xc90f6d50: 5a 5a 5a 5a 5a 5a 5a 5a ZZZZZZZZ | 136 | INFO: Object 0xc90f6d20 @offset=3360 fp=0xc90f6d58 |
137 | INFO: Allocated in get_modalias+0x61/0xf5 age=53 cpu=1 pid=554 | ||
138 | |||
139 | Bytes b4 0xc90f6d10: 00 00 00 00 00 00 00 00 5a 5a 5a 5a 5a 5a 5a 5a ........ZZZZZZZZ | ||
140 | Object 0xc90f6d20: 31 30 31 39 2e 30 30 35 1019.005 | ||
141 | Redzone 0xc90f6d28: 00 cc cc cc . | ||
142 | Padding 0xc90f6d50: 5a 5a 5a 5a 5a 5a 5a 5a ZZZZZZZZ | ||
143 | |||
137 | [<c010523d>] dump_trace+0x63/0x1eb | 144 | [<c010523d>] dump_trace+0x63/0x1eb |
138 | [<c01053df>] show_trace_log_lvl+0x1a/0x2f | 145 | [<c01053df>] show_trace_log_lvl+0x1a/0x2f |
139 | [<c010601d>] show_trace+0x12/0x14 | 146 | [<c010601d>] show_trace+0x12/0x14 |
@@ -155,74 +162,108 @@ Filler 0xc90f6d50: 5a 5a 5a 5a 5a 5a 5a 5a ZZZZZZZZ | |||
155 | [<c0104112>] sysenter_past_esp+0x5f/0x99 | 162 | [<c0104112>] sysenter_past_esp+0x5f/0x99 |
156 | [<b7f7b410>] 0xb7f7b410 | 163 | [<b7f7b410>] 0xb7f7b410 |
157 | ======================= | 164 | ======================= |
158 | @@@ SLUB kmalloc-8: Restoring redzone (0xcc) from 0xc90f6d28-0xc90f6d2b | ||
159 | 165 | ||
166 | FIX kmalloc-8: Restoring Redzone 0xc90f6d28-0xc90f6d2b=0xcc | ||
160 | 167 | ||
168 | If SLUB encounters a corrupted object (full detection requires the kernel | ||
169 | to be booted with slub_debug) then the following output will be dumped | ||
170 | into the syslog: | ||
161 | 171 | ||
162 | If SLUB encounters a corrupted object then it will perform the following | 172 | 1. Description of the problem encountered |
163 | actions: | ||
164 | |||
165 | 1. Isolation and report of the issue | ||
166 | 173 | ||
167 | This will be a message in the system log starting with | 174 | This will be a message in the system log starting with |
168 | 175 | ||
169 | *** SLUB <slab cache affected>: <What went wrong>@<object address> | 176 | =============================================== |
170 | offset=<offset of object into slab> flags=<slabflags> | 177 | BUG <slab cache affected>: <What went wrong> |
171 | inuse=<objects in use in this slab> freelist=<first free object in slab> | 178 | ----------------------------------------------- |
172 | 179 | ||
173 | 2. Report on how the problem was dealt with in order to ensure the continued | 180 | INFO: <corruption start>-<corruption_end> <more info> |
174 | operation of the system. | 181 | INFO: Slab <address> <slab information> |
182 | INFO: Object <address> <object information> | ||
183 | INFO: Allocated in <kernel function> age=<jiffies since alloc> cpu=<allocated by | ||
184 | cpu> pid=<pid of the process> | ||
185 | INFO: Freed in <kernel function> age=<jiffies since free> cpu=<freed by cpu> | ||
186 | pid=<pid of the process> | ||
175 | 187 | ||
176 | These are messages in the system log beginning with | 188 | (Object allocation / free information is only available if SLAB_STORE_USER is |
177 | 189 | set for the slab. slub_debug sets that option) | |
178 | @@@ SLUB <slab cache affected>: <corrective action taken> | ||
179 | 190 | ||
191 | 2. The object contents if an object was involved. | ||
180 | 192 | ||
181 | In the above sample SLUB found that the Redzone of an active object has | 193 | Various types of lines can follow the BUG SLUB line: |
182 | been overwritten. Here a string of 8 characters was written into a slab that | ||
183 | has the length of 8 characters. However, a 8 character string needs a | ||
184 | terminating 0. That zero has overwritten the first byte of the Redzone field. | ||
185 | After reporting the details of the issue encountered the @@@ SLUB message | ||
186 | tell us that SLUB has restored the redzone to its proper value and then | ||
187 | system operations continue. | ||
188 | |||
189 | Various types of lines can follow the @@@ SLUB line: | ||
190 | 194 | ||
191 | Bytes b4 <address> : <bytes> | 195 | Bytes b4 <address> : <bytes> |
192 | Show a few bytes before the object where the problem was detected. | 196 | Shows a few bytes before the object where the problem was detected. |
193 | Can be useful if the corruption does not stop with the start of the | 197 | Can be useful if the corruption does not stop with the start of the |
194 | object. | 198 | object. |
195 | 199 | ||
196 | Object <address> : <bytes> | 200 | Object <address> : <bytes> |
197 | The bytes of the object. If the object is inactive then the bytes | 201 | The bytes of the object. If the object is inactive then the bytes |
198 | typically contain poisoning values. Any non-poison value shows a | 202 | typically contain poison values. Any non-poison value shows a |
199 | corruption by a write after free. | 203 | corruption by a write after free. |
200 | 204 | ||
201 | Redzone <address> : <bytes> | 205 | Redzone <address> : <bytes> |
202 | The redzone following the object. The redzone is used to detect | 206 | The Redzone following the object. The Redzone is used to detect |
203 | writes after the object. All bytes should always have the same | 207 | writes after the object. All bytes should always have the same |
204 | value. If there is any deviation then it is due to a write after | 208 | value. If there is any deviation then it is due to a write after |
205 | the object boundary. | 209 | the object boundary. |
206 | 210 | ||
207 | Freepointer | 211 | (Redzone information is only available if SLAB_RED_ZONE is set. |
208 | The pointer to the next free object in the slab. May become | 212 | slub_debug sets that option) |
209 | corrupted if overwriting continues after the red zone. | ||
210 | |||
211 | Last alloc: | ||
212 | Last free: | ||
213 | Shows the address from which the object was allocated/freed last. | ||
214 | We note the pid, the time and the CPU that did so. This is usually | ||
215 | the most useful information to figure out where things went wrong. | ||
216 | Here get_modalias() did an kmalloc(8) instead of a kmalloc(9). | ||
217 | 213 | ||
218 | Filler <address> : <bytes> | 214 | Padding <address> : <bytes> |
219 | Unused data to fill up the space in order to get the next object | 215 | Unused data to fill up the space in order to get the next object |
220 | properly aligned. In the debug case we make sure that there are | 216 | properly aligned. In the debug case we make sure that there are |
221 | at least 4 bytes of filler. This allow for the detection of writes | 217 | at least 4 bytes of padding. This allows the detection of writes |
222 | before the object. | 218 | before the object. |
223 | 219 | ||
224 | Following the filler will be a stackdump. That stackdump describes the | 220 | 3. A stackdump |
225 | location where the error was detected. The cause of the corruption is more | 221 | |
226 | likely to be found by looking at the information about the last alloc / free. | 222 | The stackdump describes the location where the error was detected. The cause |
223 | of the corruption is may be more likely found by looking at the function that | ||
224 | allocated or freed the object. | ||
225 | |||
226 | 4. Report on how the problem was dealt with in order to ensure the continued | ||
227 | operation of the system. | ||
228 | |||
229 | These are messages in the system log beginning with | ||
230 | |||
231 | FIX <slab cache affected>: <corrective action taken> | ||
232 | |||
233 | In the above sample SLUB found that the Redzone of an active object has | ||
234 | been overwritten. Here a string of 8 characters was written into a slab that | ||
235 | has the length of 8 characters. However, a 8 character string needs a | ||
236 | terminating 0. That zero has overwritten the first byte of the Redzone field. | ||
237 | After reporting the details of the issue encountered the FIX SLUB message | ||
238 | tell us that SLUB has restored the Redzone to its proper value and then | ||
239 | system operations continue. | ||
240 | |||
241 | Emergency operations: | ||
242 | --------------------- | ||
243 | |||
244 | Minimal debugging (sanity checks alone) can be enabled by booting with | ||
245 | |||
246 | slub_debug=F | ||
247 | |||
248 | This will be generally be enough to enable the resiliency features of slub | ||
249 | which will keep the system running even if a bad kernel component will | ||
250 | keep corrupting objects. This may be important for production systems. | ||
251 | Performance will be impacted by the sanity checks and there will be a | ||
252 | continual stream of error messages to the syslog but no additional memory | ||
253 | will be used (unlike full debugging). | ||
254 | |||
255 | No guarantees. The kernel component still needs to be fixed. Performance | ||
256 | may be optimized further by locating the slab that experiences corruption | ||
257 | and enabling debugging only for that cache | ||
258 | |||
259 | I.e. | ||
260 | |||
261 | slub_debug=F,dentry | ||
262 | |||
263 | If the corruption occurs by writing after the end of the object then it | ||
264 | may be advisable to enable a Redzone to avoid corrupting the beginning | ||
265 | of other objects. | ||
266 | |||
267 | slub_debug=FZ,dentry | ||
227 | 268 | ||
228 | Christoph Lameter, <clameter@sgi.com>, May 23, 2007 | 269 | Christoph Lameter, <clameter@sgi.com>, May 30, 2007 |