Uploaded image for project: 'JDK'
  1. JDK
  2. JDK-8035283

Second phase of branch shortening doesn't account for loop alignment

    XMLWordPrintable

Details

    • b04
    • sparc

    Backports

      Description

        On SPARC with cbcond we can fail with:
        # Internal Error (output.cpp:1576), pid=3601, tid=15
        # guarantee((int)(blk_starts[i+1] - blk_starts[i]) >= (current_offset - blk_offset)) failed: shouldn't increase block size

        The customer hit the bug with the 32bit version on a virtual machine running on T-series. We didn't identify that machine as niagara so we used the default loop alignment (16 bytes instead of 4). I cannot reproduce the crash on 64bit, but the problem exists there as well.

        It very hard to write a unit test for the fix (at least I wasn't able to), since the code must specifically trigger block rotation in loop to make sure that the loop header starts with a compare.

        Attachments

          Issue Links

            backported by

            Backport - A issue that is required to port a Bug or Feature into another product release. This issue type is generally associated with the main Bug/Feature to represent each individual release of the port. JDK-8036023 Second phase of branch shortening doesn't account for loop alignment

            • P3 - Major loss of function.
            • Resolved

            Backport - A issue that is required to port a Bug or Feature into another product release. This issue type is generally associated with the main Bug/Feature to represent each individual release of the port. JDK-8036028 Second phase of branch shortening doesn't account for loop alignment

            • P3 - Major loss of function.
            • Resolved

            Backport - A issue that is required to port a Bug or Feature into another product release. This issue type is generally associated with the main Bug/Feature to represent each individual release of the port. JDK-8036839 Second phase of branch shortening doesn't account for loop alignment

            • P3 - Major loss of function.
            • Resolved

            Backport - A issue that is required to port a Bug or Feature into another product release. This issue type is generally associated with the main Bug/Feature to represent each individual release of the port. JDK-8037130 Second phase of branch shortening doesn't account for loop alignment

            • P3 - Major loss of function.
            • Resolved

            Backport - A issue that is required to port a Bug or Feature into another product release. This issue type is generally associated with the main Bug/Feature to represent each individual release of the port. JDK-8037237 Second phase of branch shortening doesn't account for loop alignment

            • P3 - Major loss of function.
            • Resolved

            Backport - A issue that is required to port a Bug or Feature into another product release. This issue type is generally associated with the main Bug/Feature to represent each individual release of the port. JDK-8039589 Second phase of branch shortening doesn't account for loop alignment

            • P3 - Major loss of function.
            • Resolved

            Backport - A issue that is required to port a Bug or Feature into another product release. This issue type is generally associated with the main Bug/Feature to represent each individual release of the port. JDK-8040350 Second phase of branch shortening doesn't account for loop alignment

            • P3 - Major loss of function.
            • Resolved

            Backport - A issue that is required to port a Bug or Feature into another product release. This issue type is generally associated with the main Bug/Feature to represent each individual release of the port. JDK-8045242 Second phase of branch shortening doesn't account for loop alignment

            • P3 - Major loss of function.
            • Resolved

            Backport - A issue that is required to port a Bug or Feature into another product release. This issue type is generally associated with the main Bug/Feature to represent each individual release of the port. JDK-8057282 Second phase of branch shortening doesn't account for loop alignment

            • P3 - Major loss of function.
            • Resolved

            Backport - A issue that is required to port a Bug or Feature into another product release. This issue type is generally associated with the main Bug/Feature to represent each individual release of the port. JDK-8060849 Second phase of branch shortening doesn't account for loop alignment

            • P3 - Major loss of function.
            • Resolved

            Backport - A issue that is required to port a Bug or Feature into another product release. This issue type is generally associated with the main Bug/Feature to represent each individual release of the port. JDK-8036029 Second phase of branch shortening doesn't account for loop alignment

            • P3 - Major loss of function.
            • Closed

            Backport - A issue that is required to port a Bug or Feature into another product release. This issue type is generally associated with the main Bug/Feature to represent each individual release of the port. JDK-8036838 Second phase of branch shortening doesn't account for loop alignment

            • P3 - Major loss of function.
            • Closed

            Backport - A issue that is required to port a Bug or Feature into another product release. This issue type is generally associated with the main Bug/Feature to represent each individual release of the port. JDK-8037668 Second phase of branch shortening doesn't account for loop alignment

            • P3 - Major loss of function.
            • Closed
            relates to

            Enhancement - null JDK-8056124 Hotspot should use PICL interface to get cacheline size on SPARC

            • P2 - Crashes, loss of data, severe memory leak.
            • Resolved

            Enhancement - null JDK-7063629 use cbcond in C2 generated code on T4

            • P4 - Minor loss of function, or other problem where easy workaround is present.
            • Closed

            Activity

              People

                iveresov Igor Veresov
                iveresov Igor Veresov
                Votes:
                0 Vote for this issue
                Watchers:
                6 Start watching this issue

                Dates

                  Created:
                  Updated:
                  Resolved: