Fixing a flaky test

[linzhp]

What type of PR is this?
Bug fix

What does this PR do? Why is it needed?
Adder_test.go generates two random int32 and adds them. There is a high chance that the sum overflow, which is an undefined behavior in C, and then caught by zig cc's ubsan (undefined behavior sanitizer), leading to test failure occasionally.

Generating smaller integers to avoid overflow.

Test plan

Run bazelisk test --runs_per_test 10 --extra_toolchains=@zig_sdk//toolchain:linux_amd64_gnu.2.31 //tests/core/cgo:opts_test

Before:

...
==================== Test output for //tests/core/cgo:opts_test (run 4 of 10):
SIGILL: illegal instruction
PC=0x4f19a7 m=4 sigcode=2
signal arrived during cgo execution
instruction bytes: 0x67 0xf 0xb9 0x0 0x8b 0x45 0xec 0x89 0x45 0xf0 0x8b 0x45 0xf4 0x3 0x45 0xf0
...
================================================================================
INFO: Found 1 test target...
Target //tests/core/cgo:opts_test up-to-date:
  bazel-bin/tests/core/cgo/opts_test_/opts_test
INFO: Elapsed time: 5.981s, Critical Path: 5.52s
INFO: 17 processes: 1 internal, 16 processwrapper-sandbox.
INFO: Build completed, 1 test FAILED, 17 total actions
//tests/core/cgo:opts_test                                               FAILED in 7 out of 10 in 0.1s
  Stats over 10 runs: max = 0.1s, min = 0.1s, avg = 0.1s, dev = 0.0s

After:

//tests/core/cgo:opts_test                                               PASSED in 0.1s
  Stats over 10 runs: max = 0.1s, min = 0.0s, avg = 0.1s, dev = 0.0s