Bug#982422: debootstrap: segmentation fault around ldconfig during debootstrapping bullseye/arm64 on a buster/amd64 host

[Josef Hahn <ale@ginger.ws>]

Package: debootstrap
Version: 1.0.114
Severity: normal

I'm struggling with a problem that I currently have with debootstrapping an
arm64 system on my amd64 machine. The attached script can reproduce the issue
here.

It crashes here on a debian buster amd64, in the 2nd debootstrap stage:

...
W: Failure trying to run:  /sbin/ldconfig
W: See //debootstrap/debootstrap.log for details

...
2021-02-10 01:05:52 URL:http://deb.debian.org/debian/pool/main/x/xz-
utils/liblzma5_5.2.5-1.0_arm64.deb [164436/164436] ->
"/...//var/cache/apt/archives/partial/liblzma5_5.2.5-1.0_arm64.deb" [1]
2021-02-10 01:05:52
URL:http://deb.debian.org/debian/pool/main/z/zlib/zlib1g_1.2.11.dfsg-2_arm64.deb
[87944/87944] ->
"/...//var/cache/apt/archives/partial/zlib1g_1%3a1.2.11.dfsg-2_arm64.deb" [1]
qemu: uncaught target signal 11 (Segmentation fault) - core dumped

It works if I debootstrap a buster system instead. Then trying to upgrade to
bullseye in the same chroot crashes with a segfault during the upgrade of libc-
bin. It also segfaults when I just open a chrooted shell after the 1st stage
and then run "ldconfig". That worked here around new year, but then broke.

This looks like a software bug somewhere. I'm not sure if I'm right here. But
maybe you can tell me.

#!/bin/bash

set -e

LOOPFILE=removemelater.loopfile
INNERROOT=removemelater.loopmnt
LOOPDEV=/dev/loop5

reproduce() {
    dd if=/dev/zero of=$LOOPFILE bs=1M count=2048
    losetup $LOOPDEV $LOOPFILE
    mkfs.ext4 $LOOPDEV
    mkdir $INNERROOT
    mount $LOOPDEV $INNERROOT
    debootstrap --arch=arm64 --foreign bullseye $INNERROOT
    mount -o bind /sys $INNERROOT/sys
    mount -o bind /dev $INNERROOT/dev
    mount -o bind /dev/pts $INNERROOT/dev/pts
    mount -o bind /proc $INNERROOT/proc
    chroot $INNERROOT /bin/bash -c "/debootstrap/debootstrap --second-stage"
}

reproduce || cat $INNERROOT/debootstrap/debootstrap.log