Comment 3 for bug 1999094

Confirmed that this commit

commit ec95f3b24425429ccce89ed6ee35bb318de029fa
Author: Al Viro <email address hidden>
Date: Mon Oct 3 20:26:08 2022 -0400

    fix coredump breakage

    Let me count the ways in which I'd screwed up:

    * when emitting a page, handling of gaps in coredump should happen
    before fetching the current file position.
    * fix for a problem that occurs on rather uncommon setups (and hadn't
    been observed in the wild) had been sent very late in the cycle.
    * ... with badly insufficient testing, introducing an easily
    reproducible breakage. Without giving it time to soak in -next.

    Fucked-up-by: Al Viro <email address hidden>
    Reported-by: "J. R. Okajima" <email address hidden>
    Tested-by: "J. R. Okajima" <email address hidden>
    Fixes: 06bbaa6dc53c "[coredump] don't use __kernel_write() on kmap_local_page()"
    Cc: <email address hidden> # v6.0-only
    Signed-off-by: Al Viro <email address hidden>

Fixes the issue.

This issue does not only affect vDSO vma's as the original test case was suggesting, but seems to completely breaks coredump debugging in gdb. Probably worth a respin as this is a very user visible issue.