Re: Fun with SSD and GPT wedges

[John Nemeth <jnemeth%cue.bc.ca@localhost>]

On Feb 13,  9:02am, Robert Elz wrote:
}
}     Date:        Tue, 12 Feb 2019 19:57:42 -0500
}     From:        Greg Troxel <gdt%lexort.com@localhost>
}     Message-ID:  <smusgws7beh.fsf%linuxpal.mit.edu@localhost>
} 
}   | I can see how we got here, but the situation seems wrong from a logical
}   | consistency point of view.  If gpt(8) is going to create wedges on
}   | adding a new partition, it should delete the wedge corresponding to a
}   | partition that it removes.
} 
} Logically yes, but I suspect that we wouldn't like the result, just
} as we didn't really like it when all gpt did was tell us that we needed
} to make the wedge manually in order to access a newly added gpt partition.
} 
} Perhaps a better solution would be to have gpt refuse to delete a
} partition if there is an existing wedge for it in the kernel?

     As stated before, gpt(8) is a tool for manipulating on-disk
data structures.  This is outside the scope of it.  Also, disklabel(8)
lets you do anything you want to a disklabel regardless of whether
the partitons are mounted.  Why would you expect gpt(8) to behave
differently?

} That would be more in line with how I (at least) would expect things
} to work ... logically consistent or not.
} 
}-- End of excerpt from Robert Elz