Hello Gospodin,
Gospodin Dimitrov wrote:
I tried version 1.21 rc2, but in this version
errors again are showing 0 and errsize too is 0 bytes, but non-trimmed
shows errorsize...
This is because a sector is now not considered bad until it has been
tried in a single-sector read. I think this is more accurate that
marking 128 non-trimmed sectors as bad just to mark 127 of them as good
later.
this make option --max-errors=[n] useless... because errors are 0 to the
end...
'errors' begin counting just after the copying phase, so they are not 0
to the end. But you are right that now they start counting later.
Please, tell us what is the effect you want to achieve. If
--max-error-rate or --exit-on-error don't do what you need, maybe
--max-errors can be extended, or a new option can be implemented.
in version 1.19 "errors" are showing number of bad blocks (and errsize
is showing badblocks*4096 in bytes (or at least I think so))
'errsize' is the sum of the sizes of all the bad-sector blocks. (In 1.19
also of non-trimmed and non-scraped blocks). In your case the mean size
of the bad areas seems to be about 4 KiB.
Best regards,
Antonio.
_______________________________________________
Bug-ddrescue mailing list
[email protected]
https://lists.gnu.org/mailman/listinfo/bug-ddrescue
