[Octave-bug-tracker] [bug #54572] int64 does not saturate correctly in negative direction

[Dan Sebald]

Follow-up Comment #45, bug #54572 (project octave):

One last comment.  I wondered if "goto" is something an optimizing compiler
will not remove and whether it retains an extra jump, e.g.,

https://stackoverflow.com/a/41609491

Given the current goto is rather short, it doesn't add too much extra code
space to apply the saturation code in all cases.  The numbers are below, and
there seems to be no difference one way or another:


50 TRIALS

64-BIT SIGNED MULT
       CURRENT_MULT   CURRENT_NOGOTO_MULT   D_FLOAT_MULT

tgood     14.580           14.792             14.340
tover     12.584           12.188             11.520

32-BIT SIGNED MULT
       CURRENT_MULT   CURRENT_NOGOTO_MULT   D_FLOAT_MULT

tgood     7.4480           7.4520             7.4160
tover     5.8120           5.8160             5.9000

16-BIT SIGNED MULT
       CURRENT_MULT   CURRENT_NOGOTO_MULT   D_FLOAT_MULT

tgood     4.3440           4.3440             4.4000
tover     3.7720           3.7760             5.2840

8-BIT SIGNED MULT
       CURRENT_MULT   CURRENT_NOGOTO_MULT   D_FLOAT_MULT

tgood     2.7200           2.6200             1.9640
tover     2.5840           2.7320             3.7680


I'm attaching the diff FWIW.

... Close this bug report.

(file #44921)
    _______________________________________________________

Additional Item Attachment:

File name: octave-int_remove_goto-djs2018aug31.patch Size:3 KB


    _______________________________________________________

Reply to this item at:

  <http://savannah.gnu.org/bugs/?54572>

_______________________________________________
  Message sent via Savannah
  https://savannah.gnu.org/