Date: Thu, 11 Jun 2026 11:46:14 +0000 From: bugzilla-noreply@freebsd.org To: doc@FreeBSD.org Subject: [Bug 294338] FreeBSD 14.0-RELEASE Release Notes: Upgrading from Previous Releases of FreeBSD misses several boot loader usecases Message-ID: <bug-294338-9-rAIEbhRJva@https.bugs.freebsd.org/bugzilla/> In-Reply-To: <bug-294338-9@https.bugs.freebsd.org/bugzilla/>
index | next in thread | previous in thread | raw e-mail
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=294338 --- Comment #9 from Michael Osipov <michaelo@FreeBSD.org> --- Apologies, it took me way longer than I have expected. (In reply to Vladlen Popolitov from comment #4) In my case, the BIOS doesn't switch at all. The controller I have does not have a suitable UEFI OpROM, only legacy OpROM so it boots via CSM only. Moreover, I can only select the controller itself in EUFI setup and then need to enter WebBIOS (LSI) to set the one and only bootable disk, no failover possible. My point is consistency over all possible setup scenarios you cannot foresee. If I choose to update the root pool after system upgrade and forgot to upgrade the loader on the other non-boot drive in the mirror then I will not be able to boot. Additionally, during recovery I want to copy the geometry and the boot code of the working drive on the new replacement drive without thinking whether I have or have not updated the boot loader back then. Yet in another sever I have provisioned recently (DL320 Gen11 with MR480i-o) I do see all enclosure slots in the EUFI setup and configured for sequential boot retry, but also here I want to be sure that the failover drive is exactly configured as the failed boot drive. (In reply to Vladlen Popolitov from comment #6) I'd even generalize your proposal, this isn't just BIOS it is UEFI: Update all disks which are part of the boot configuration. (In reply to Toomas Soome from comment #7) Exactly, uniform, swap slots, boot again, get a spare drive and fix the issue. Also a general note that Vladlen you should take a look at the broken FAT12 copy onto a valid ESP which was fixed in db8b56134506840832bec2d1ce07b9e00d4d6d71. This should be mentioned in the upgrade notes. I have here four different setups: HPE ProLiant DL360 Gen9: UEFI boot + UFS HPE ProLiant DL380 Gen7: BIOS boot (from UEFI) + ZFS Fujitsu PRIMERGY RX300 S7: BIOS boot (from UEFI) + ZFS HPE ProLiant DL320 Gen11: UEFI boot + ZFS They all have different storage controllers and different UEFI options to configure boot. At least from FreeBSD side I need to be sure that I have everything on my drives to have them bootable anytime. -- You are receiving this mail because: You are the assignee for the bug.home | help
Want to link to this message? Use this
URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?bug-294338-9-rAIEbhRJva>