Description of problem: grub2 works fine on IBM PPC systems, but doesn't work out of the box on PowerMac, grub2-install aborts with '/dev/sda2 is not a PReP partition' These steps were necessary to get it to work on my G5: mount -t hfs /dev/sda2 /mnt # HFS partition where the grub image on Apple HW # needs to be installed/copied onto # rm -f /mnt/* # remove yaboot, not enough space cd /usr/lib/grub/powerpc-ieee1275 cat > /tmp/grub.cfg <<EOT insmod part_apple insmod ext2 insmod lvm insmod boot insmod hfs insmod gzio search --no-floppy --fs-uuid --set=root a99ca62e-2649-465c-a06f-ff9203061198 set prefix=($root)/grub2 configfile /grub2/grub.cfg EOT # a99ca62e-2649-465c-a06f-ff9203061198 is the UUID of my /boot partition: # df /boot| cut -d ' ' -f 1 | xargs blkid grub2-mkimage -O powerpc-ieee1275 -o /mnt/grub.img -c /tmp/grub.cfg --prefix "(ieee1275/hd,apple2)" {blocklist,boot,bufio,cat,cmp,configfile,cpio_be,cpio,crc64,datehook,date,datetime,dm_nv,echo,elf,ext2,fat,font,fshelp,functional_test,gettext,gptsync,gzio,halt,hello,help,hfs,hfsplus,ieee1275_fb,keylayouts,keystatus,linux,loadenv,loopback,ls,lvm,lzopio,mdraid09,mdraid1x,memdisk,memrw,minicmd,minix2,minix3,minix,msdospart,net,newc,normal,ofnet,part_apple,part_bsd,part_dvh,part_gpt,part_msdos,part_plan,parttool,pbkdf2,priority_queue,probe,raid5rec,raid6rec,read,reboot,regexp,romfs,scsi,search_fs_file,search_fs_uuid,search_label,search,serial,setjmp,sleep,squash4,suspend,terminal,time,trig,true,videotest,xzio}.mod # We can't install all modules as the HFS partition is only 1048 kb # We also can't use the *gfx modules as those mess up the display pretty bad umount /mnt/ # This installs a minimal grub.cfg file on the HFS partition which then searches and loads the real grub config file on the /boot partition. This avoids having to update the files on the HFS partition during each kernel update. Version-Release number of selected component (if applicable): grub2-2.0-0.22.fc17.ppc64
It seems like that wasn't the exact error message - was it? pjones is working on a upgrade to beta4 - please try with a 0.23 build from git which is very close to upstream. Please try to run grub2-install with bash -x and see why it fails. I guess it fails because grub2-probe returns something else than 41 (PPC PReP Boot)? Isn't that the right thing to do? grub2-install should automatically only include the necessary modules in the image and it should thus fit on the partition once you get that far. (Note that 20_ppc_terminfo haven't been upstreamed and might conflict with recent upstream development. It seems strange that it should be priority 20 - It looks more like a 01 thing.)
correct, the last few lines of the bash -x output are: + '[' -z /dev/sda2 ']' ++ /usr/sbin/grub2-probe -m '' -d /dev/sda2 -t msdos_parttype + '[' '' '!=' 41 ']' + gettext 'The chosen partition is not a PReP partition.' > /usr/sbin/grub2-probe -m '' -d /dev/sda2 -t msdos_parttype > echo $? 0 > As I said, Apple PowerMacs don't have PReP partitions, they use HFS partitions for their boot files.
So this is a Open Firmware system and it should have ended up in the True branch of the if statement? What is the previous line of -x output ... or what is "${grub_modinfo_target_cpu}-${grub_modinfo_platform}"? Apparently not "powerpc-ieee1275"?
Created attachment 579631 [details] output of sh -x grub2-install /dev/sda2
(In reply to comment #3) > "${grub_modinfo_target_cpu}-${grub_modinfo_platform}"? Apparently not > "powerpc-ieee1275"? Baaah. Obviously it _is_ "powerpc-ieee1275" and IEEE1275 _is_ Open Firmware. Anyway: Do the non-PReP branch of the if statement work for your case? It seems like http://bzr.savannah.gnu.org/lh/grub/trunk/grub/revision/3764 got something wrong. Perhaps "powerpc-ieee1275" in some cases still should reach install_device= . (http://bzr.savannah.gnu.org/lh/grub/trunk/grub/revision/4146 refactored it but didn't change this aspect.) Anyway, I guess you as a PPC expert is most qualified to propose and discuss a fix upstream ... unless you can bribe one of the other RH guys to do it. (The trace shows that it generated a small /boot/grub2/powerpc-ieee1275/core.elf that will look for grub.cfg and modules in (I presume) /boot/grub2. Do that work for you? And the reason the probe fails seems to be that it isn't msdos partitioned at all ... which is one reason it shouldn't end up in that branch.)
Created attachment 580140 [details] patch to add support for grub2 on PowerMac HFS partitions patch that sets up grub2 on a PowerMac G5 with an HFS partition and hopefully doesn't break anything on the other archs
What grub version do that patch apply to? (It fails everywhere here.) It also seems like the patch try to fix the strange tab indent at the same time - that makes it hard to review.
this is for 2.00~beta4 or better grub2-2.0-0.23.fc17
Right - it apply on top of grub-1.99-Fix-tests-of-zeroed-partition.patch ... which haven't been upstreamed. Note to self: The patch moves some code around, but the only real change is in cases which otherwise would have ended up in "The chosen partition is not a PReP partition."
grub2-2.0-0.24.beta4.fc17 has been pushed to f17 stable - closing. - but please send the patch upstream too!
Feedback from phcoder upstream: grub-2.00~beta4-add-support-for-PowerMac-HFS-partitions.patch: Unsafe if partition is mounted ... this section writes to the potentially mounted filesystem directly ... hmount has nothing to do with standard mount. I'm speaking about simultaneous access through htools and standard way ... htools is like mtools and will not do proper locking
Created attachment 583531 [details] patch to use mkfs
Actually this whole thread results from a simple misconception as to how to install GRUB on HFS(+) in the first place. Usual way is to mount it on /boot/grub and use grub-install *without* disk argument
This patch no longer apply on top of beta6.
This message is a reminder that Fedora 17 is nearing its end of life. Approximately 4 (four) weeks from now Fedora will stop maintaining and issuing updates for Fedora 17. It is Fedora's policy to close all bug reports from releases that are no longer maintained. At that time this bug will be closed as WONTFIX if it remains open with a Fedora 'version' of '17'. Package Maintainer: If you wish for this bug to remain open because you plan to fix it in a currently maintained version, simply change the 'version' to a later Fedora version prior to Fedora 17's end of life. Bug Reporter: Thank you for reporting this issue and we are sorry that we may not be able to fix it before Fedora 17 is end of life. If you would still like to see this bug fixed and are able to reproduce it against a later version of Fedora, you are encouraged change the 'version' to a later Fedora version prior to Fedora 17's end of life. Although we aim to fix as many bugs as possible during every release's lifetime, sometimes those efforts are overtaken by events. Often a more recent Fedora release includes newer upstream software that fixes bugs or makes them obsolete.
Fedora 17 changed to end-of-life (EOL) status on 2013-07-30. Fedora 17 is no longer maintained, which means that it will not receive any further security or bug fix updates. As a result we are closing this bug. If you can reproduce this bug against a currently maintained version of Fedora please feel free to reopen this bug against that version. Thank you for reporting this bug and we are sorry it could not be fixed.