bjorn3 opened issue #4971:
This may allow for reproducing some miscompilations more reliably with less risk that changes in functions called before the miscompiled function affect if the miscompilation results in a crash or not. In addition it could be used as defence-in-depth.
cfallin commented on issue #4971:
This is a great idea; it's part of the "chaos mode" general approach that we should build at some point (#4134).
Last updated: Nov 22 2024 at 16:03 UTC