Tutorial

  • 04/11/2022
  • Public Content

Generating a Vectorization Report

A vectorization report shows what loops in your code were vectorized and explains why other loops were not vectorized. To generate a vectorization report, use the
qopt-report-phase=vec
compiler options together with
qopt-report=1
or
qopt-report=2
.
Together with
qopt-report-phase=vec
,
qopt-report=1
generates a report with the loops in your code that were vectorized while
qopt-report-phase=vec
with
qopt-report=2
generates a report with both the loops in your code that were vectorized and the reason that other loops were not vectorized.
Because vectorization is turned off with the
O1
option, the compiler does not generate a vectorization report. To generate a vectorization report, compile your project with the
O2
,
qopt-report-phase=vec
,
qopt-report=1
options:
icc -std=c99 -O2 -D NOFUNCCALL -qopt-report=1 -qopt-report-phase=vec Multiply.c Driver.c -o MatVector
We replace the call to the Matvec function with an inline equivalent with
-D NOFUNCCALL
.
Recompile the program and then execute
MatVector
. Record the new execution time. The reduction in time is mostly due to auto-vectorization of the inner loop at line
145
noted in the vectorization report
matvec.optrpt
:
LOOP BEGIN at Driver.c(140,5) remark #25460: No loop optimizations reported LOOP BEGIN at Driver.c(143,9) remark #25460: No loop optimizations reported LOOP BEGIN at Driver.c(145,13) remark #15300: LOOP WAS VECTORIZED LOOP END LOOP BEGIN at Driver.c(145,13) <Remainder loop for vectorization> LOOP END LOOP END LOOP END
Your line and column numbers may be different.
qopt-report=2
with
qopt-report-phase=vec,loop
returns a list that also includes loops that were not vectorized or multi-versioned, along with the reason that the compiler did not vectorize them or multi-version the loop.
Recompile your project with the
qopt-report=2
and
qopt-report-phase=vec,loop
options.
icc -std=c99 -O2 -D NOFUNCCALL -qopt-report-phase=vec,loop -qopt-report=2 Multiply.c Driver.c -o MatVector
The vectorization report
Multiply.optrpt
indicates that the loop at line
37 in
Multiply.c
did not vectorize because it is not the innermost loop of the loop nest. Two versions of the innermost loop at line 49 were generated, and one version was vectorized.
LOOP BEGIN at Multiply.c(37,5) remark #15542: loop was not vectorized: inner loop was already vectorized LOOP BEGIN at Multiply.c(49,9) <Peeled loop for vectorization, Multiversioned v1> LOOP END LOOP BEGIN at Multiply.c(49,9) <Multiversioned v1> remark #25228: Loop multiversioned for Data Dependence remark #15300: LOOP WAS VECTORIZED LOOP END LOOP BEGIN at Multiply.c(49,9) <Alternate Alignment Vectorized Loop, Multiversioned v1> LOOP END LOOP BEGIN at Multiply.c(49,9) <Remainder loop for vectorization, Multiversioned v1> LOOP END LOOP BEGIN at Multiply.c(49,9) <Multiversioned v2> remark #15304: loop was not vectorized: non-vectorizable loop instance from multiversioning remark #25439: unrolled with remainder by 2 LOOP END LOOP BEGIN at Multiply.c(49,9) <Remainder, Multiversioned v2> LOOP END
  • Your line and column numbers may be different.
  • For more information on the
    qopt-report
    and
    qopt-report-phase
    compiler options, see the
    Compiler Options
    section in the
    Intel®
    C++
    Compiler Developer Guide and Reference
    .

Product and Performance Information

1

Performance varies by use, configuration and other factors. Learn more at www.Intel.com/PerformanceIndex.