• State TI Thinks Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 1 answer
  • Subscribers 53 subscribers
  • Views 484 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Original question:

Compiler/ARM-CGT: Information about known issues with ARM compiler

Compiler/ARM-CGT: Follow-up on CODEGEN-4304 "runtime failure on unaligned access (on some processor variants, not all"

1138
Expert 1226 points
Part Number: ARM-CGT

Tool/software: TI C/C++ Compiler

Earlier, George Mock had the following to say about CODEGEN-4304:

https://e2e.ti.com/support/tools/ccs/f/81/p/814000/3017692#3017692

1138

CODEGEN-4304  "runtime failure on unaligned access (on some processor variants, not all)":

I see that this has a test case attached to it, but no further analysis.  This does not seem to be a case where the user has done something wrong, leading to an unaligned access, but rather some kind of issue with the compiler or library.  What are the scenarios that can trigger this problem?

This one has only been partially analyzed.  No firm conclusions about the cause are available.

Can you please update us on this issue?

--thx

  • 0
    TI__Guru**** 243670 points

    Unfortunately, there is no progress to report.

    Thanks and regards,

    -George

  • 0 in reply to George Mock
    Expert 1226 points

    The "headline" of the incident report for CODEGEN-4304 states makes a reference to "on some processor variants, not all".

    Which processor variants are at risk?

    --thx

  • 0 in reply to 1138
    TI__Guru**** 243670 points

    Analysis is ongoing, and still not complete.  Here is what know so far.  The issue is with either the compiler support for certain floating-point units, or it could be a simulator issue when modeling the floating-point unit.  We know the problem is not present on Cortex-M0 (6M0), Cortex-M3, or Cortex-M4, or if you don't use the floating-point unit.

    Thanks and regards,

    -George

  • 0 in reply to George Mock
    Expert 1226 points

    The example in the incident report happened to involve using C++, not C, including usage of std::map and std::pair.

    However, it sounds like the underlying problem is likely to be present in some way with C compilation as well (we specifically use the --c99 flag with the v18.1.X compilers, although we're considering updating to the --c11 flag with the v18.12.X compilers).  Does this seem likely?

    --thx

  • 0 in reply to 1138
    TI__Guru**** 243670 points

    1138 said:
    it sounds like the underlying problem is likely to be present in some way with C compilation as well

    That is correct.

    Thanks and regards,

    -George