LLVM-OPT-REPORT(1) LLVM LLVM-OPT-REPORT(1)

llvm-opt-report - generate optimization report from YAML

llvm-opt-report [options] [input]

llvm-opt-report is a tool to generate an optimization report from YAML optimization record files.

You need to create an input YAML optimization record file before running llvm-opt-report.

It provides information on the execution time, memory usage, and other details of each optimization pass.

$ clang -c foo.c -o foo.o -O3 -fsave-optimization-record

Then, you create a report using the llvm-opt-report command with the YAML optimization record file foo.opt.yaml as input.

$ llvm-opt-report foo.opt.yaml -o foo.lst

foo.lst is the generated optimization report.

< foo.c
 1          | void bar();
 2          | void foo() { bar(); }
 3          |
 4          | void Test(int *res, int *c, int *d, int *p, int n) {
 5          |   int i;
 6          |
 7          | #pragma clang loop vectorize(assume_safety)
 8     V4,1 |   for (i = 0; i < 1600; i++) {
 9          |     res[i] = (p[i] == 0) ? res[i] : res[i] + d[i];
10          |   }
11          |
12  U16     |   for (i = 0; i < 16; i++) {
13          |     res[i] = (p[i] == 0) ? res[i] : res[i] + d[i];
14          |   }
15          |
16 I        |   foo();
17          |
18          |   foo(); bar(); foo();
   I        |   ^
   I        |                 ^
19          | }
20          |

Symbols printed on the left side of the program indicate what kind of optimization was performed. The meanings of the symbols are as follows:

  • I: The function is inlined.
  • U: The loop is unrolled. The following number indicates the unroll factor.
  • V: The loop is vectorized. The following numbers indicate the vector length and the interleave factor.

NOTE:

If a specific line of code is output twice, it means that the same optimization pass was applied to that line of code twice, and the pass was able to further optimize the code on the second iteration.

If input is "-" or omitted, llvm-opt-report reads from standard input. Otherwise, it will read from the specified filename.

If the -o option is omitted, then llvm-opt-report will send its output to standard output. If the -o option specifies "-", then the output will also be sent to standard output.

Display available options.
Display the version of this program.
The format of the optimization record file. The Argument is one of the following:
  • yaml
  • yaml-strtab
  • bitstream
Do not demangle function names.
Output file.
Root for relative input paths.
Do not include vectorization factors, etc.

llvm-opt-report returns 0 on success. Otherwise, an error message is printed to standard error, and the tool returns 1.

Maintained by the LLVM Team (https://llvm.org/).

2003-2024, LLVM Project

2024-05-21 17