Skip to content

Commit 2371077

Browse files
workingjubileemichaelwoerister
workingjubilee
authored and
michaelwoerister
committed
Allow copying example shell scripts
1 parent 306a7d2 commit 2371077

File tree

1 file changed

+78
-78
lines changed

1 file changed

+78
-78
lines changed

‎summarize/README.md

Lines changed: 78 additions & 78 deletions
Original file line numberDiff line numberDiff line change
@@ -7,7 +7,7 @@ Summarize is a tool to produce a human readable summary of `measureme` profiling
77
To use this tool you will first want to install it:
88

99
```bash
10-
$ cargo install --git https://github.com/rust-lang/measureme --branch stable summarize
10+
cargo install --git https://github.com/rust-lang/measureme --branch stable summarize
1111
```
1212

1313
## Profiling the nightly compiler
@@ -23,9 +23,9 @@ profile the [regex][regex-crate] crate.
2323
[regex-crate]: https://github.com/rust-lang/regex
2424

2525
```bash
26-
$ git clone https://github.com/rust-lang/regex.git
27-
$ cd regex
28-
$ cargo +nightly rustc -- -Z self-profile
26+
git clone https://github.com/rust-lang/regex.git
27+
cd regex
28+
cargo +nightly rustc -- -Z self-profile
2929
```
3030

3131
The commands above will run `rustc` with the flag that enables profiling. You should now
@@ -38,36 +38,36 @@ You can now use the `summarize` tool we installed in the previous section to vie
3838
contents of these files:
3939

4040
```bash
41-
$ summarize summarize regex-{pid}.mm_profdata
42-
+------------------------+-----------+-----------------+------------+------------+--------------+-----------------------+
43-
| Item | Self time | % of total time | Item count | Cache hits | Blocked time | Incremental load time |
44-
+------------------------+-----------+-----------------+------------+------------+--------------+-----------------------+
45-
| LLVM_emit_obj | 4.51s | 41.432 | 141 | 0 | 0.00ns | 0.00ns |
46-
+------------------------+-----------+-----------------+------------+------------+--------------+-----------------------+
47-
| LLVM_module_passes | 1.05s | 9.626 | 140 | 0 | 0.00ns | 0.00ns |
48-
+------------------------+-----------+-----------------+------------+------------+--------------+-----------------------+
49-
| LLVM_make_bitcode | 712.94ms | 6.543 | 140 | 0 | 0.00ns | 0.00ns |
50-
+------------------------+-----------+-----------------+------------+------------+--------------+-----------------------+
51-
| typeck_tables_of | 542.23ms | 4.976 | 17470 | 16520 | 0.00ns | 0.00ns |
52-
+------------------------+-----------+-----------------+------------+------------+--------------+-----------------------+
53-
| codegen | 366.82ms | 3.366 | 141 | 0 | 0.00ns | 0.00ns |
54-
+------------------------+-----------+-----------------+------------+------------+--------------+-----------------------+
55-
| optimized_mir | 188.22ms | 1.727 | 11668 | 9114 | 0.00ns | 0.00ns |
56-
+------------------------+-----------+-----------------+------------+------------+--------------+-----------------------+
57-
| mir_built | 156.30ms | 1.434 | 2040 | 1020 | 0.00ns | 0.00ns |
58-
+------------------------+-----------+-----------------+------------+------------+--------------+-----------------------+
59-
| evaluate_obligation | 151.95ms | 1.394 | 33134 | 23817 | 0.00ns | 0.00ns |
60-
+------------------------+-----------+-----------------+------------+------------+--------------+-----------------------+
61-
| LLVM_compress_bitcode | 126.55ms | 1.161 | 140 | 0 | 0.00ns | 0.00ns |
62-
+------------------------+-----------+-----------------+------------+------------+--------------+-----------------------+
63-
| codegen crate | 119.08ms | 1.093 | 1 | 0 | 0.00ns | 0.00ns |
64-
+------------------------+-----------+-----------------+------------+------------+--------------+-----------------------+
65-
| mir_const | 117.82ms | 1.081 | 1050 | 30 | 0.00ns | 0.00ns |
66-
+------------------------+-----------+-----------------+------------+------------+--------------+-----------------------+
67-
68-
(rows elided)
69-
70-
Total cpu time: 10.896488447s
41+
summarize summarize regex-{pid}.mm_profdata
42+
# +------------------------+-----------+-----------------+------------+------------+--------------+-----------------------+
43+
# | Item | Self time | % of total time | Item count | Cache hits | Blocked time | Incremental load time |
44+
# +------------------------+-----------+-----------------+------------+------------+--------------+-----------------------+
45+
# | LLVM_emit_obj | 4.51s | 41.432 | 141 | 0 | 0.00ns | 0.00ns |
46+
# +------------------------+-----------+-----------------+------------+------------+--------------+-----------------------+
47+
# | LLVM_module_passes | 1.05s | 9.626 | 140 | 0 | 0.00ns | 0.00ns |
48+
# +------------------------+-----------+-----------------+------------+------------+--------------+-----------------------+
49+
# | LLVM_make_bitcode | 712.94ms | 6.543 | 140 | 0 | 0.00ns | 0.00ns |
50+
# +------------------------+-----------+-----------------+------------+------------+--------------+-----------------------+
51+
# | typeck_tables_of | 542.23ms | 4.976 | 17470 | 16520 | 0.00ns | 0.00ns |
52+
# +------------------------+-----------+-----------------+------------+------------+--------------+-----------------------+
53+
# | codegen | 366.82ms | 3.366 | 141 | 0 | 0.00ns | 0.00ns |
54+
# +------------------------+-----------+-----------------+------------+------------+--------------+-----------------------+
55+
# | optimized_mir | 188.22ms | 1.727 | 11668 | 9114 | 0.00ns | 0.00ns |
56+
# +------------------------+-----------+-----------------+------------+------------+--------------+-----------------------+
57+
# | mir_built | 156.30ms | 1.434 | 2040 | 1020 | 0.00ns | 0.00ns |
58+
# +------------------------+-----------+-----------------+------------+------------+--------------+-----------------------+
59+
# | evaluate_obligation | 151.95ms | 1.394 | 33134 | 23817 | 0.00ns | 0.00ns |
60+
# +------------------------+-----------+-----------------+------------+------------+--------------+-----------------------+
61+
# | LLVM_compress_bitcode | 126.55ms | 1.161 | 140 | 0 | 0.00ns | 0.00ns |
62+
# +------------------------+-----------+-----------------+------------+------------+--------------+-----------------------+
63+
# | codegen crate | 119.08ms | 1.093 | 1 | 0 | 0.00ns | 0.00ns |
64+
# +------------------------+-----------+-----------------+------------+------------+--------------+-----------------------+
65+
# | mir_const | 117.82ms | 1.081 | 1050 | 30 | 0.00ns | 0.00ns |
66+
# +------------------------+-----------+-----------------+------------+------------+--------------+-----------------------+
67+
#
68+
# (rows elided)
69+
#
70+
# Total cpu time: 10.896488447s
7171
```
7272

7373
## Profiling your own build of rustc
@@ -80,41 +80,41 @@ You can also profile your own custom build of rustc. First you'll have to clone
8080
[compiling-rust]: https://rustc-dev-guide.rust-lang.org/building/how-to-build-and-run.html
8181

8282
```bash
83-
$ git clone https://github.com/rust-lang/rust.git
84-
$ ./x.py build
83+
git clone https://github.com/rust-lang/rust.git
84+
./x.py build
8585
# This will take a while...
86-
$ rustup toolchain link mytoolchain build/x86_64-unknown-linux-gnu/stage1
86+
rustup toolchain link mytoolchain build/x86_64-unknown-linux-gnu/stage1
8787
```
8888

8989
Where `mytoolchain` is the name of your custom toolchain. Now we do more or less the same
9090
as before: (with regex as example)
9191

9292
```bash
93-
$ git clone https://github.com/rust-lang/regex.git
94-
$ cd regex
95-
$ cargo +mytoolchain rustc -- -Z self-profile
96-
$ summarize summarize regex-{pid}.mm_profdata
97-
+------------------------+-----------+-----------------+------------+------------+--------------+-----------------------+
98-
| Item | Self time | % of total time | Item count | Cache hits | Blocked time | Incremental load time |
99-
+------------------------+-----------+-----------------+------------+------------+--------------+-----------------------+
100-
| LLVM_emit_obj | 4.51s | 41.432 | 141 | 0 | 0.00ns | 0.00ns |
101-
+------------------------+-----------+-----------------+------------+------------+--------------+-----------------------+
102-
| LLVM_module_passes | 1.05s | 9.626 | 140 | 0 | 0.00ns | 0.00ns |
103-
+------------------------+-----------+-----------------+------------+------------+--------------+-----------------------+
104-
| LLVM_make_bitcode | 712.94ms | 6.543 | 140 | 0 | 0.00ns | 0.00ns |
105-
+------------------------+-----------+-----------------+------------+------------+--------------+-----------------------+
106-
| typeck_tables_of | 542.23ms | 4.976 | 17470 | 16520 | 0.00ns | 0.00ns |
107-
+------------------------+-----------+-----------------+------------+------------+--------------+-----------------------+
108-
| codegen | 366.82ms | 3.366 | 141 | 0 | 0.00ns | 0.00ns |
109-
+------------------------+-----------+-----------------+------------+------------+--------------+-----------------------+
110-
| optimized_mir | 188.22ms | 1.727 | 11668 | 9114 | 0.00ns | 0.00ns |
111-
+------------------------+-----------+-----------------+------------+------------+--------------+-----------------------+
112-
| mir_built | 156.30ms | 1.434 | 2040 | 1020 | 0.00ns | 0.00ns |
113-
+------------------------+-----------+-----------------+------------+------------+--------------+-----------------------+
114-
115-
(rows elided)
116-
117-
Total cpu time: 10.896488447s
93+
git clone https://github.com/rust-lang/regex.git
94+
cd regex
95+
cargo +mytoolchain rustc -- -Z self-profile
96+
summarize summarize regex-{pid}.mm_profdata
97+
# +------------------------+-----------+-----------------+------------+------------+--------------+-----------------------+
98+
# | Item | Self time | % of total time | Item count | Cache hits | Blocked time | Incremental load time |
99+
# +------------------------+-----------+-----------------+------------+------------+--------------+-----------------------+
100+
# | LLVM_emit_obj | 4.51s | 41.432 | 141 | 0 | 0.00ns | 0.00ns |
101+
# +------------------------+-----------+-----------------+------------+------------+--------------+-----------------------+
102+
# | LLVM_module_passes | 1.05s | 9.626 | 140 | 0 | 0.00ns | 0.00ns |
103+
# +------------------------+-----------+-----------------+------------+------------+--------------+-----------------------+
104+
# | LLVM_make_bitcode | 712.94ms | 6.543 | 140 | 0 | 0.00ns | 0.00ns |
105+
# +------------------------+-----------+-----------------+------------+------------+--------------+-----------------------+
106+
# | typeck_tables_of | 542.23ms | 4.976 | 17470 | 16520 | 0.00ns | 0.00ns |
107+
# +------------------------+-----------+-----------------+------------+------------+--------------+-----------------------+
108+
# | codegen | 366.82ms | 3.366 | 141 | 0 | 0.00ns | 0.00ns |
109+
# +------------------------+-----------+-----------------+------------+------------+--------------+-----------------------+
110+
# | optimized_mir | 188.22ms | 1.727 | 11668 | 9114 | 0.00ns | 0.00ns |
111+
# +------------------------+-----------+-----------------+------------+------------+--------------+-----------------------+
112+
# | mir_built | 156.30ms | 1.434 | 2040 | 1020 | 0.00ns | 0.00ns |
113+
# +------------------------+-----------+-----------------+------------+------------+--------------+-----------------------+
114+
#
115+
# (rows elided)
116+
#
117+
# Total cpu time: 10.896488447s
118118
```
119119

120120
Note that your custom build of the compiler must not use a newer version of the
@@ -147,22 +147,22 @@ The `diff` sub command allows you to compare the performance of two different pr
147147
The output is a table like that of the `summarize` sub command but it instead shows the differences in each metric.
148148

149149
```bash
150-
$ summarize diff base-profile.mm_profdata changed-profile.mm_profdata
151-
+---------------------------+--------------+------------+------------+--------------+-----------------------+
152-
| Item | Self Time | Item count | Cache hits | Blocked time | Incremental load time |
153-
+---------------------------+--------------+------------+------------+--------------+-----------------------+
154-
| LLVM_module_passes | -66.626471ms | +0 | +0 | +0ns | +0ns |
155-
+---------------------------+--------------+------------+------------+--------------+-----------------------+
156-
| LLVM_emit_obj | -38.700719ms | +0 | +0 | +0ns | +0ns |
157-
+---------------------------+--------------+------------+------------+--------------+-----------------------+
158-
| LLVM_make_bitcode | +32.006706ms | +0 | +0 | +0ns | +0ns |
159-
+---------------------------+--------------+------------+------------+--------------+-----------------------+
160-
| mir_borrowck | -12.808322ms | +0 | +0 | +0ns | +0ns |
161-
+---------------------------+--------------+------------+------------+--------------+-----------------------+
162-
| typeck_tables_of | -10.325247ms | +0 | +0 | +0ns | +0ns |
163-
+---------------------------+--------------+------------+------------+--------------+-----------------------+
164-
(rows elided)
165-
Total cpu time: -155.177548ms
150+
summarize diff base-profile.mm_profdata changed-profile.mm_profdata
151+
# +---------------------------+--------------+------------+------------+--------------+-----------------------+
152+
# | Item | Self Time | Item count | Cache hits | Blocked time | Incremental load time |
153+
# +---------------------------+--------------+------------+------------+--------------+-----------------------+
154+
# | LLVM_module_passes | -66.626471ms | +0 | +0 | +0ns | +0ns |
155+
# +---------------------------+--------------+------------+------------+--------------+-----------------------+
156+
# | LLVM_emit_obj | -38.700719ms | +0 | +0 | +0ns | +0ns |
157+
# +---------------------------+--------------+------------+------------+--------------+-----------------------+
158+
# | LLVM_make_bitcode | +32.006706ms | +0 | +0 | +0ns | +0ns |
159+
# +---------------------------+--------------+------------+------------+--------------+-----------------------+
160+
# | mir_borrowck | -12.808322ms | +0 | +0 | +0ns | +0ns |
161+
# +---------------------------+--------------+------------+------------+--------------+-----------------------+
162+
# | typeck_tables_of | -10.325247ms | +0 | +0 | +0ns | +0ns |
163+
# +---------------------------+--------------+------------+------------+--------------+-----------------------+
164+
# (rows elided)
165+
# Total cpu time: -155.177548ms
166166
```
167167

168168
The table is sorted by the absolute value of `Self time` descending.

0 commit comments

Comments
 (0)