This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion

Understanding defect fixes listed in Compiler Release Notes

I see that the lists of corrections / known defects in the release notes of each new compiler revision are tagged with a unique identifier (ex. SDCOMP-15294). Is it possible to find further information about these items such as steps to reproduce? Or are the descriptions as they appear in the release notes the only publicly accessible information?

I'm using a legacy compiler, and attempting to produce an errata-style document based on the defect fixes identified in subsequent compiler release notes, which is obviously far from ideal, but it's the best I've got. I've identified 30 'corrections' listed in newer compiler versions which I would like to have information on how to reproduce in order to be able to either discount the issue as not affecting our particular tool chain, or be able to explain how it is mitigated in our code base (ex. the operation which causes the issue is not performed anywhere). Is it possible to access this kind of additional information?

Thanks,

  • Hi Michael,

    I'm Product Manager for the Arm Compilers, thanks for getting in touch! Always good to hear from a long-term Arm Compiler user.

    I'm assuming that you've already considered, and discounted, the possibility of migrating your code to a later compiler, as being excessive in either work or risk. These conversations are easier on later compiler: we now have safety-qualified variants of the Arm Compilers which are designed to solve exactly these kinds of challenges. Each update release of the safety-qualified compilers includes a defect report giving defects of every errata fix made in that release, designed specifically to assess defect impact or mitigation.

    Looking at SDCOMP-15294 in particular, that was an errata fix that appeared in compiler 5.0u1 around the end of 2011. For that particular errata fix it does look like we have a reproducer. I'll send you a private message to continue the conversation, if you could get me the list of 30 or so fixes that interest you we can take a look at what might be possible.

    Thanks,

    Paul.