Blame view
Documentation/trace/tracepoint-analysis.txt
12 KB
bb7222208 tracing, document... |
1 2 3 4 5 6 7 8 9 10 11 |
Notes on Analysing Behaviour Using Events and Tracepoints Documentation written by Mel Gorman PCL information heavily based on email from Ingo Molnar 1. Introduction =============== Tracepoints (see Documentation/trace/tracepoints.txt) can be used without creating custom kernel modules to register probe functions using the event tracing infrastructure. |
b41df645c Documentation: Up... |
12 13 |
Simplistically, tracepoints represent important events that can be taken in conjunction with other tracepoints to build a "Big Picture" of |
bb7222208 tracing, document... |
14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 |
what is going on within the system. There are a large number of methods for gathering and interpreting these events. Lacking any current Best Practises, this document describes some of the methods that can be used. This document assumes that debugfs is mounted on /sys/kernel/debug and that the appropriate tracing options have been configured into the kernel. It is assumed that the PCL tool tools/perf has been installed and is in your path. 2. Listing Available Events =========================== 2.1 Standard Utilities ---------------------- All possible events are visible from /sys/kernel/debug/tracing/events. Simply calling $ find /sys/kernel/debug/tracing/events -type d will give a fair indication of the number of events available. |
b41df645c Documentation: Up... |
34 |
2.2 PCL (Performance Counters for Linux) |
bb7222208 tracing, document... |
35 |
------- |
b41df645c Documentation: Up... |
36 |
Discovery and enumeration of all counters and events, including tracepoints, |
bb7222208 tracing, document... |
37 |
are available with the perf tool. Getting a list of available events is a |
b41df645c Documentation: Up... |
38 |
simple case of: |
bb7222208 tracing, document... |
39 40 41 42 43 44 45 46 |
$ perf list 2>&1 | grep Tracepoint ext4:ext4_free_inode [Tracepoint event] ext4:ext4_request_inode [Tracepoint event] ext4:ext4_allocate_inode [Tracepoint event] ext4:ext4_write_begin [Tracepoint event] ext4:ext4_ordered_write_end [Tracepoint event] [ .... remaining output snipped .... ] |
b41df645c Documentation: Up... |
47 |
3. Enabling Events |
bb7222208 tracing, document... |
48 |
================== |
b41df645c Documentation: Up... |
49 |
3.1 System-Wide Event Enabling |
bb7222208 tracing, document... |
50 51 52 53 |
------------------------------ See Documentation/trace/events.txt for a proper description on how events can be enabled system-wide. A short example of enabling all events related |
b41df645c Documentation: Up... |
54 |
to page allocation would look something like: |
bb7222208 tracing, document... |
55 56 |
$ for i in `find /sys/kernel/debug/tracing/events -name "enable" | grep mm_`; do echo 1 > $i; done |
b41df645c Documentation: Up... |
57 |
3.2 System-Wide Event Enabling with SystemTap |
bb7222208 tracing, document... |
58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 |
--------------------------------------------- In SystemTap, tracepoints are accessible using the kernel.trace() function call. The following is an example that reports every 5 seconds what processes were allocating the pages. global page_allocs probe kernel.trace("mm_page_alloc") { page_allocs[execname()]++ } function print_count() { printf ("%-25s %-s ", "#Pages Allocated", "Process Name") foreach (proc in page_allocs-) printf("%-25d %s ", page_allocs[proc], proc) printf (" ") delete page_allocs } probe timer.s(5) { print_count() } |
b41df645c Documentation: Up... |
84 |
3.3 System-Wide Event Enabling with PCL |
bb7222208 tracing, document... |
85 86 87 88 89 90 |
--------------------------------------- By specifying the -a switch and analysing sleep, the system-wide events for a duration of time can be examined. $ perf stat -a \ |
90a5d5af7 mm-tracepoint: fi... |
91 92 |
-e kmem:mm_page_alloc -e kmem:mm_page_free \ -e kmem:mm_page_free_batched \ |
bb7222208 tracing, document... |
93 94 95 96 |
sleep 10 Performance counter stats for 'sleep 10': 9630 kmem:mm_page_alloc |
90a5d5af7 mm-tracepoint: fi... |
97 98 |
2143 kmem:mm_page_free 7424 kmem:mm_page_free_batched |
bb7222208 tracing, document... |
99 100 101 102 103 |
10.002577764 seconds time elapsed Similarly, one could execute a shell and exit it as desired to get a report at that point. |
b41df645c Documentation: Up... |
104 |
3.4 Local Event Enabling |
bb7222208 tracing, document... |
105 106 107 108 |
------------------------ Documentation/trace/ftrace.txt describes how to enable events on a per-thread basis using set_ftrace_pid. |
b41df645c Documentation: Up... |
109 |
3.5 Local Event Enablement with PCL |
bb7222208 tracing, document... |
110 |
----------------------------------- |
b41df645c Documentation: Up... |
111 |
Events can be activated and tracked for the duration of a process on a local |
bb7222208 tracing, document... |
112 |
basis using PCL such as follows. |
90a5d5af7 mm-tracepoint: fi... |
113 114 |
$ perf stat -e kmem:mm_page_alloc -e kmem:mm_page_free \ -e kmem:mm_page_free_batched ./hackbench 10 |
bb7222208 tracing, document... |
115 116 117 118 119 |
Time: 0.909 Performance counter stats for './hackbench 10': 17803 kmem:mm_page_alloc |
90a5d5af7 mm-tracepoint: fi... |
120 121 |
12398 kmem:mm_page_free 4827 kmem:mm_page_free_batched |
bb7222208 tracing, document... |
122 123 |
0.973913387 seconds time elapsed |
b41df645c Documentation: Up... |
124 |
4. Event Filtering |
bb7222208 tracing, document... |
125 126 127 128 129 |
================== Documentation/trace/ftrace.txt covers in-depth how to filter events in ftrace. Obviously using grep and awk of trace_pipe is an option as well as any script reading trace_pipe. |
b41df645c Documentation: Up... |
130 |
5. Analysing Event Variances with PCL |
bb7222208 tracing, document... |
131 132 133 |
===================================== Any workload can exhibit variances between runs and it can be important |
b41df645c Documentation: Up... |
134 |
to know what the standard deviation is. By and large, this is left to the |
bb7222208 tracing, document... |
135 136 |
performance analyst to do it by hand. In the event that the discrete event occurrences are useful to the performance analyst, then perf can be used. |
90a5d5af7 mm-tracepoint: fi... |
137 138 |
$ perf stat --repeat 5 -e kmem:mm_page_alloc -e kmem:mm_page_free -e kmem:mm_page_free_batched ./hackbench 10 |
bb7222208 tracing, document... |
139 140 141 142 143 144 145 146 147 |
Time: 0.890 Time: 0.895 Time: 0.915 Time: 1.001 Time: 0.899 Performance counter stats for './hackbench 10' (5 runs): 16630 kmem:mm_page_alloc ( +- 3.542% ) |
90a5d5af7 mm-tracepoint: fi... |
148 149 |
11486 kmem:mm_page_free ( +- 4.771% ) 4730 kmem:mm_page_free_batched ( +- 2.325% ) |
bb7222208 tracing, document... |
150 151 152 153 154 155 156 |
0.982653002 seconds time elapsed ( +- 1.448% ) In the event that some higher-level event is required that depends on some aggregation of discrete events, then a script would need to be developed. Using --repeat, it is also possible to view how events are fluctuating over |
b41df645c Documentation: Up... |
157 |
time on a system-wide basis using -a and sleep. |
bb7222208 tracing, document... |
158 |
|
90a5d5af7 mm-tracepoint: fi... |
159 160 |
$ perf stat -e kmem:mm_page_alloc -e kmem:mm_page_free \ -e kmem:mm_page_free_batched \ |
bb7222208 tracing, document... |
161 162 163 164 165 |
-a --repeat 10 \ sleep 1 Performance counter stats for 'sleep 1' (10 runs): 1066 kmem:mm_page_alloc ( +- 26.148% ) |
90a5d5af7 mm-tracepoint: fi... |
166 167 |
182 kmem:mm_page_free ( +- 5.464% ) 890 kmem:mm_page_free_batched ( +- 30.079% ) |
bb7222208 tracing, document... |
168 169 |
1.002251757 seconds time elapsed ( +- 0.005% ) |
b41df645c Documentation: Up... |
170 |
6. Higher-Level Analysis with Helper Scripts |
bb7222208 tracing, document... |
171 172 173 174 175 176 177 178 179 |
============================================ When events are enabled the events that are triggering can be read from /sys/kernel/debug/tracing/trace_pipe in human-readable format although binary options exist as well. By post-processing the output, further information can be gathered on-line as appropriate. Examples of post-processing might include o Reading information from /proc for the PID that triggered the event o Deriving a higher-level event from a series of lower-level events. |
b41df645c Documentation: Up... |
180 |
o Calculating latencies between two events |
bb7222208 tracing, document... |
181 182 183 |
Documentation/trace/postprocess/trace-pagealloc-postprocess.pl is an example script that can read trace_pipe from STDIN or a copy of a trace. When used |
b41df645c Documentation: Up... |
184 |
on-line, it can be interrupted once to generate a report without exiting |
bb7222208 tracing, document... |
185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 |
and twice to exit. Simplistically, the script just reads STDIN and counts up events but it also can do more such as o Derive high-level events from many low-level events. If a number of pages are freed to the main allocator from the per-CPU lists, it recognises that as one per-CPU drain even though there is no specific tracepoint for that event o It can aggregate based on PID or individual process number o In the event memory is getting externally fragmented, it reports on whether the fragmentation event was severe or moderate. o When receiving an event about a PID, it can record who the parent was so that if large numbers of events are coming from very short-lived processes, the parent process responsible for creating all the helpers can be identified |
b41df645c Documentation: Up... |
201 |
7. Lower-Level Analysis with PCL |
bb7222208 tracing, document... |
202 |
================================ |
b41df645c Documentation: Up... |
203 |
There may also be a requirement to identify what functions within a program |
bb7222208 tracing, document... |
204 |
were generating events within the kernel. To begin this sort of analysis, the |
b41df645c Documentation: Up... |
205 |
data must be recorded. At the time of writing, this required root: |
bb7222208 tracing, document... |
206 207 |
$ perf record -c 1 \ |
90a5d5af7 mm-tracepoint: fi... |
208 209 |
-e kmem:mm_page_alloc -e kmem:mm_page_free \ -e kmem:mm_page_free_batched \ |
bb7222208 tracing, document... |
210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 |
./hackbench 10 Time: 0.894 [ perf record: Captured and wrote 0.733 MB perf.data (~32010 samples) ] Note the use of '-c 1' to set the event period to sample. The default sample period is quite high to minimise overhead but the information collected can be very coarse as a result. This record outputted a file called perf.data which can be analysed using perf report. $ perf report # Samples: 30922 # # Overhead Command Shared Object # ........ ......... ................................ # 87.27% hackbench [vdso] 6.85% hackbench /lib/i686/cmov/libc-2.9.so 2.62% hackbench /lib/ld-2.9.so 1.52% perf [vdso] 1.22% hackbench ./hackbench 0.48% hackbench [kernel] 0.02% perf /lib/i686/cmov/libc-2.9.so 0.01% perf /usr/bin/perf 0.01% perf /lib/ld-2.9.so 0.00% hackbench /lib/i686/cmov/libpthread-2.9.so # # (For more details, try: perf report --sort comm,dso,symbol) # |
b41df645c Documentation: Up... |
240 241 |
According to this, the vast majority of events triggered on events within the VDSO. With simple binaries, this will often be the case so let's |
bb7222208 tracing, document... |
242 |
take a slightly different example. In the course of writing this, it was |
b41df645c Documentation: Up... |
243 244 |
noticed that X was generating an insane amount of page allocations so let's look at it: |
bb7222208 tracing, document... |
245 246 |
$ perf record -c 1 -f \ |
90a5d5af7 mm-tracepoint: fi... |
247 248 |
-e kmem:mm_page_alloc -e kmem:mm_page_free \ -e kmem:mm_page_free_batched \ |
bb7222208 tracing, document... |
249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 |
-p `pidof X` This was interrupted after a few seconds and $ perf report # Samples: 27666 # # Overhead Command Shared Object # ........ ....... ....................................... # 51.95% Xorg [vdso] 47.95% Xorg /opt/gfx-test/lib/libpixman-1.so.0.13.1 0.09% Xorg /lib/i686/cmov/libc-2.9.so 0.01% Xorg [kernel] # # (For more details, try: perf report --sort comm,dso,symbol) # |
b41df645c Documentation: Up... |
266 267 |
So, almost half of the events are occurring in a library. To get an idea which symbol: |
bb7222208 tracing, document... |
268 269 270 271 272 273 274 275 276 277 278 279 280 281 |
$ perf report --sort comm,dso,symbol # Samples: 27666 # # Overhead Command Shared Object Symbol # ........ ....... ....................................... ...... # 51.95% Xorg [vdso] [.] 0x000000ffffe424 47.93% Xorg /opt/gfx-test/lib/libpixman-1.so.0.13.1 [.] pixmanFillsse2 0.09% Xorg /lib/i686/cmov/libc-2.9.so [.] _int_malloc 0.01% Xorg /opt/gfx-test/lib/libpixman-1.so.0.13.1 [.] pixman_region32_copy_f 0.01% Xorg [kernel] [k] read_hpet 0.01% Xorg /opt/gfx-test/lib/libpixman-1.so.0.13.1 [.] get_fast_path 0.00% Xorg [kernel] [k] ftrace_trace_userstack |
b41df645c Documentation: Up... |
282 |
To see where within the function pixmanFillsse2 things are going wrong: |
bb7222208 tracing, document... |
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 |
$ perf annotate pixmanFillsse2 [ ... ] 0.00 : 34eeb: 0f 18 08 prefetcht0 (%eax) : } : : extern __inline void __attribute__((__gnu_inline__, __always_inline__, _ : _mm_store_si128 (__m128i *__P, __m128i __B) : { : *__P = __B; 12.40 : 34eee: 66 0f 7f 80 40 ff ff movdqa %xmm0,-0xc0(%eax) 0.00 : 34ef5: ff 12.40 : 34ef6: 66 0f 7f 80 50 ff ff movdqa %xmm0,-0xb0(%eax) 0.00 : 34efd: ff 12.39 : 34efe: 66 0f 7f 80 60 ff ff movdqa %xmm0,-0xa0(%eax) 0.00 : 34f05: ff 12.67 : 34f06: 66 0f 7f 80 70 ff ff movdqa %xmm0,-0x90(%eax) 0.00 : 34f0d: ff 12.58 : 34f0e: 66 0f 7f 40 80 movdqa %xmm0,-0x80(%eax) 12.31 : 34f13: 66 0f 7f 40 90 movdqa %xmm0,-0x70(%eax) 12.40 : 34f18: 66 0f 7f 40 a0 movdqa %xmm0,-0x60(%eax) 12.31 : 34f1d: 66 0f 7f 40 b0 movdqa %xmm0,-0x50(%eax) At a glance, it looks like the time is being spent copying pixmaps to the card. Further investigation would be needed to determine why pixmaps are being copied around so much but a starting point would be to take an ancient build of libpixmap out of the library path where it was totally forgotten about from months ago! |