Re: [Bug-ddrescue] GNU ddrescue 1.18-pre2 released

[Scott Dwyer]

> Do you have the "reopen" patch that I can try on my 18pre2 copy
>         here so that I can see how the performance changes?
No, I do not have a patch for ddrescue. I wrote an independent program 
for myself to test and prove the theory. Antonio will have to make any 
changes to ddrescue as he wishes.

> For what ever reason, it still slows down for me with
>         --direct,  I've also noticed that using --direct causes
>         different "severe failure" behaviour when reading from USB
>         bridges, sometimes requiring a reboot to resolve.
The fact that you are using a USB bridge can answer why the --direct 
does not work. I have a USB adaptor, and it seems to work ok with IDE 
devices. But it will not show the errors with a SATA drive. It will slow 
down where the errors are, but will not report them as an error (so a 
bad drive will actually not seem to have any errors, but will just be 
really slow). It will instead return what I would call garbage data back 
in place of the errors. USB adaptors have their own code on them to 
access the drives, and you are at their mercy for what they can and can 
not do. They are great for the field in a pinch, but they are not good 
for hard core data recovery. If you really want to know what is going on 
with a drive, hook it direct to the computer.

So now I have to ask, are you using the USB bridge for these tests? If 
you have the drive hooked directly to the computer and you still notice 
the slowdown with the --direct option, then maybe ddrescue could use to 
be changed. But if you are basing this all on a USB adaptor, then 
ddrescue might not need to be changed at all.

Scott D.