So I am making progress but saw a strange bit.
#parted /dev/sdb mkpart uboot ext3 4 516
# parted /dev/sdb print Model: Generic- Multi-Card (scsi) Disk /dev/sdb: 7969MB Sector size (logical/physical): 512B/512B Partition Table: gpt Disk Flags:
Number Start End Size File system Name Flags 1 4194kB 516MB 512MB fat32 uboot
I formated it with:
#mkfs -t ext3 /dev/sdb1
# parted /dev/sdb print Model: Generic- Multi-Card (scsi) Disk /dev/sdb: 7969MB Sector size (logical/physical): 512B/512B Partition Table: gpt Disk Flags:
Number Start End Size File system Name Flags 1 4194kB 516MB 512MB ext3 uboot
So why did parted mislabel the partition?
On 08/08/2014 06:21 AM, Robert Moskowitz issued this missive:
So I am making progress but saw a strange bit.
#parted /dev/sdb mkpart uboot ext3 4 516
# parted /dev/sdb print Model: Generic- Multi-Card (scsi) Disk /dev/sdb: 7969MB Sector size (logical/physical): 512B/512B Partition Table: gpt Disk Flags:
Number Start End Size File system Name Flags 1 4194kB 516MB 512MB fat32 uboot
I formated it with:
#mkfs -t ext3 /dev/sdb1
# parted /dev/sdb print Model: Generic- Multi-Card (scsi) Disk /dev/sdb: 7969MB Sector size (logical/physical): 512B/512B Partition Table: gpt Disk Flags:
Number Start End Size File system Name Flags 1 4194kB 516MB 512MB ext3 uboot
So why did parted mislabel the partition?
I don't think "uboot" is a valid partition type--it should be "primary", "logical" or "extended". Thus the format of the command should have been:
# parted /dev/sdb mkpart primary ext3 4 516 # parted /dev/sdb name 1 uboot
I've noticed that parted sometimes makes some weird decisions if parameters it expects are missing or mis-specified. ---------------------------------------------------------------------- - Rick Stevens, Systems Engineer, AllDigital ricks@alldigital.com - - AIM/Skype: therps2 ICQ: 22643734 Yahoo: origrps2 - - - - A friend said he climbed to the top of Mount Rainier. My view is - - that if there's no elevator, it must not be that interesting. - ----------------------------------------------------------------------
On 08/08/2014 12:38 PM, Rick Stevens wrote:
On 08/08/2014 06:21 AM, Robert Moskowitz issued this missive:
So I am making progress but saw a strange bit.
#parted /dev/sdb mkpart uboot ext3 4 516
# parted /dev/sdb print Model: Generic- Multi-Card (scsi) Disk /dev/sdb: 7969MB Sector size (logical/physical): 512B/512B Partition Table: gpt Disk Flags:
Number Start End Size File system Name Flags 1 4194kB 516MB 512MB fat32 uboot
I formated it with:
#mkfs -t ext3 /dev/sdb1
# parted /dev/sdb print Model: Generic- Multi-Card (scsi) Disk /dev/sdb: 7969MB Sector size (logical/physical): 512B/512B Partition Table: gpt Disk Flags:
Number Start End Size File system Name Flags 1 4194kB 516MB 512MB ext3 uboot
So why did parted mislabel the partition?
I don't think "uboot" is a valid partition type--it should be "primary", "logical" or "extended". Thus the format of the command should have been:
# parted /dev/sdb mkpart primary ext3 4 516 # parted /dev/sdb name 1 ubootI've noticed that parted sometimes makes some weird decisions if parameters it expects are missing or mis-specified.
Somewhere in my reading and testing I got that mkpart format was
mkpart part-type label fs-type start end
and then part-type became dropped if the partition type is gpt.
Seems on further reading you are probably right. There is a separate labels command.
- Rick Stevens, Systems Engineer, AllDigital ricks@alldigital.com -
- AIM/Skype: therps2 ICQ: 22643734 Yahoo: origrps2 -
- A friend said he climbed to the top of Mount Rainier. My view is -
- that if there's no elevator, it must not be that interesting. -
On Aug 8, 2014, at 10:38 AM, Rick Stevens ricks@alldigital.com wrote:
I don't think "uboot" is a valid partition type--it should be "primary", "logical" or "extended". Thus the format of the command should have been:
Since the disk is GPT there's no such distinction among partitions. Behavior wise they're primary partitions, and on-disk they're most like primary partitions, but it's better to just say they're partitions with no distinction.
# parted /dev/sdb mkpart primary ext3 4 516 # parted /dev/sdb name 1 uboot
I've noticed that parted sometimes makes some weird decisions if parameters it expects are missing or mis-specified.
Yes. It's designed to be stable: if you update parted from 2.0 to 3.1, chances are your app will still do the partitioning portion correctly. But if you're a mortal user looking for an interface that accurately communicates facts bidirectionally, parted is eyebrow raising.
Chris Murphy
On Aug 8, 2014, at 7:21 AM, Robert Moskowitz rgm@htt-consult.com wrote:
So I am making progress but saw a strange bit.
#parted /dev/sdb mkpart uboot ext3 4 516
# parted /dev/sdb print Model: Generic- Multi-Card (scsi) Disk /dev/sdb: 7969MB Sector size (logical/physical): 512B/512B Partition Table: gpt Disk Flags:
Number Start End Size File system Name Flags 1 4194kB 516MB 512MB fat32 uboot
I formated it with:
#mkfs -t ext3 /dev/sdb1
# parted /dev/sdb print Model: Generic- Multi-Card (scsi) Disk /dev/sdb: 7969MB Sector size (logical/physical): 512B/512B Partition Table: gpt Disk Flags:
Number Start End Size File system Name Flags 1 4194kB 516MB 512MB ext3 uboot
So why did parted mislabel the partition?
Yeah you're getting confused. It's really easy. It didn't format it ext3, it just sets it to a partition type GUID for that filesystem, which in the version of parted you're using is for Microsoft basic data (i.e. it would not have made any difference had you chosen fat32 or ext3, because on GPT disks parted uses the same partition type code until you get to the Fedora 21 version and then choosing ext3 would make it a partition type GUID for "Linux filesystem".
Parted is showing it as fat32 because it found a fat32 superblock. When you reformatted it ext3, parted found the ext3 superblock. So again, it's a throwback from when parted did formatting and fs resizing.
The behaviors change silently (including flags) between MBR and GPT partition schemes too. So you have to be aware of what partition method is being used.
For example "boot" flag for MBR disks sets the active bit for a partition; whereas on GPT the "boot" flag changes the partition type GUID to that of "EFI System partition". Yeah I know, you're welcome to start shooting gin at any time but I guarantee you it doesn't really help. It's sorta like taking Percocet for pain and realizing it still hurts, you just don't care.
Chris Murphy
On 08/08/2014 02:32 PM, Chris Murphy wrote:
On Aug 8, 2014, at 7:21 AM, Robert Moskowitz rgm@htt-consult.com wrote:
So I am making progress but saw a strange bit.
#parted /dev/sdb mkpart uboot ext3 4 516
# parted /dev/sdb print Model: Generic- Multi-Card (scsi) Disk /dev/sdb: 7969MB Sector size (logical/physical): 512B/512B Partition Table: gpt Disk Flags:
Number Start End Size File system Name Flags 1 4194kB 516MB 512MB fat32 uboot
I formated it with:
#mkfs -t ext3 /dev/sdb1
# parted /dev/sdb print Model: Generic- Multi-Card (scsi) Disk /dev/sdb: 7969MB Sector size (logical/physical): 512B/512B Partition Table: gpt Disk Flags:
Number Start End Size File system Name Flags 1 4194kB 516MB 512MB ext3 uboot
So why did parted mislabel the partition?
Yeah you're getting confused. It's really easy. It didn't format it ext3, it just sets it to a partition type GUID for that filesystem, which in the version of parted you're using is for Microsoft basic data (i.e. it would not have made any difference had you chosen fat32 or ext3, because on GPT disks parted uses the same partition type code until you get to the Fedora 21 version and then choosing ext3 would make it a partition type GUID for "Linux filesystem".
Parted is showing it as fat32 because it found a fat32 superblock. When you reformatted it ext3, parted found the ext3 superblock. So again, it's a throwback from when parted did formatting and fs resizing.
The behaviors change silently (including flags) between MBR and GPT partition schemes too. So you have to be aware of what partition method is being used.
And after I did all the setup for GPT, it (the Cubieboard) would not boot. So I looked at cards that I had that did boot, and THEY were all type msdos. So back to the beginning making the card msdos, and can't label the partitions, oh, but finally rebuilt and I did create a bootable card!
So now I have a good cookbook that I will eventually be able to post to the Redsleeve wiki (what I am build). Once I get the video to login, I will say success. I only get login so far to the serial console.
For example "boot" flag for MBR disks sets the active bit for a partition; whereas on GPT the "boot" flag changes the partition type GUID to that of "EFI System partition". Yeah I know, you're welcome to start shooting gin at any time but I guarantee you it doesn't really help. It's sorta like taking Percocet for pain and realizing it still hurts, you just don't care.
Chris Murphy
-
Chris Murphy -
Rick Stevens -
Robert Moskowitz