On Wed, 12 Feb 2025, Corinna Vinschen wrote: > Hi Jeremy, > > your patch is basically fine, and I was about to push it, when I > realized that I don't quite understand this: > > What exactly is de-duplicated here? > > I have a drive mounted under C:\drvmount. > I create an additional mount entry: > > $ mount C:/drvmount /home/corinna/drv > > If I call mount, I see two mount entries, both pointing to the > /home/corinna/drv dir: > > $ mount | grep drvmount > C:/drvmount on /home/corinna/drv type ntfs (binary,user) > C:/drvmount on /home/corinna/drv type ntfs (binary,posix=0,noumount,auto) > > The first is the explicit mount, the second is the cygdrive entry. > If I disable the above de-dup code, the result is the same.
It was *supposed* to not return the second one. Maybe I broke it when trying to break out of the loop early... I will test this scenario and see why it doesn't work as expected. > However, either way, both point to the explicit mount point. > Wouldn't it be helpful to see /cygdrive/c/drvmount? > > C:/drvmount on /home/corinna/drv type ntfs (binary,user) > C:/drvmount on /cygdrive/c/drvmount type ntfs (binary,posix=0,noumount,auto) The existing code masked off of available_drives if it saw an explicit mount entry to the root of that drive letter. I assumed it was desired not to show a cygdrive mount entry in that case. If it is, maybe it could use cygdrive_posix_path instead of conv_to_posix_path? I'm kind of concerned that it could get confusing though... consider: drive mounted under C:\drvmount mount C: /something mount would show C: on /something C: on /cygdrive/c C:/drvmount on /cygdrive/c/drvmount and say nothing about /something/drvmount. Of course, the existing behavior I was trying to replicate is to say nothing about /cygdrive/c...
