XPost: linux.debian.bugs.dist 
 From: steve@einval.com 
  
 Package: partman-efi 
 Severity: normal 
 Tags: d-i 
  
 Just found in d-i testing on a VM which has been swapped backwards and 
 forwards between testing BIOS and EFI installations. 
  
 On a machine with an existing partition in the right place that 
 matches our default location for the ESP (1 MiB from the beginning of 
 the disk), we will end up trying to mount that partition as our new 
 ESP. If it's *not* a VFAT-formatteD ESP, then things will go badly. 
  
 In my particular case, the test failed as it tried to mount an 
 existing ext4-formatted fs (the rootfs from a previous BIOS-boot 
 test), and this failed with EINVAL. 
  
 We should actually validate that the ESP we're proposing to use is 
 *actually usable*. It's better to fail up front here with a "this is 
 not as expected" than later on with a generic "failed" message. 
  
 -- System Information: 
 Debian Release: 12.11 
   APT prefers oldstable-updates 
   APT policy: (500, 'oldstable-updates'), (500, 'oldstable-security'), (500, 
 'oldstable') 
 Architecture: amd64 (x86_64) 
 Foreign Architectures: i386 
  
 Kernel: Linux 6.1.0-38-amd64 (SMP w/12 CPU threads; PREEMPT) 
 Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8 (charmap=UTF-8), 
 ANGUAGE=en_GB:en 
 Shell: /bin/sh linked to /usr/bin/dash 
 Init: systemd (via /run/systemd/system) 
 LSM: AppArmor: enabled 
  
 --- SoupGate-Win32 v1.05 
  * Origin: you cannot sedate... all the things you hate (1:229/2)