Good morning! Could I ask for a very short code review for the following bug:
7161544 Workaround for brain dead BIOSes that refuse to boot off USB media that have no active DOS partition http://monaco.us.oracle.com/detail.jsf?cr=7161544 The fix is to mark the EFI system partition on the boot media as active when the MBR image is being assembled during distro constructor's boot checkpoint execution. This satisfies the BIOS enough to execute the boot sector of the MBR on the affected system. I've tested and confirmed this on the system that this was reported against plus on a couple BIOS and UEFI systems that were doing the correct thing already. Wevrev: http://jurassic.us.oracle.com/~npower/webrevs/webrev-active-esp/ Diff: --- old/usr/src/lib/install_boot/boot.py Sun Apr 15 14:37:38 2012 +++ new/usr/src/lib/install_boot/boot.py Sun Apr 15 14:37:38 2012 @@ -1023,11 +1023,15 @@ # target modules because we want precise control over the # geometry, plus it's not well suited to lofi devices. # Layout: - # type ESP, at 1M offset, partition size = uefi_image size + # - type ESP, active, at 1M offset, + # - partition size = uefi_image size + # The ESP is marked active, because despite the fact that we + # write GRUB to the MBR, some BIOSes are brain dead and won't + # boot USB media without an active DOS partition. esptype = Partition.name_to_num("EFI System") startblock = MB / BLOCKSIZE uefi_image_nblocks = uefi_image_size / BLOCKSIZE - fdiskentry = "%d:0:0:0:0:0:0:0:%d:%d" \ + fdiskentry = "%d:128:0:0:0:0:0:0:%d:%d" \ % (esptype, startblock, uefi_image_nblocks) self.logger.debug("Creating fdisk table on %s:\n\t%s", lofi_rdev, fdiskentry) Thanks! Niall _______________________________________________ caiman-discuss mailing list [email protected] http://mail.opensolaris.org/mailman/listinfo/caiman-discuss
