2 ms·

So my concern with following this logic is that such tools that are writing to specific block locations would then be wiping unrelated inodes depending on the l

So my concern with following this logic is that such tools that are writing to specific block locations would then be wiping unrelated inodes depending on the low level logic. I tried emailing Colin but his last known email is invalid now. He may be working for the NSA so I probably won't be able to reach him. Some people there are responsive, some are not and a few are on here. So even if wear leveling prevents overwriting a file then such tools should in theory be a risk of data corruption. If this is the case then the tool should be updated to detect if the target is an SSD and abort with a scary message. Perhaps another route is to reach out to the coreutils team.

10987654321 4y ago
> writing to specific block locations would then be wiping unrelated inodes depending on the low level logic It won't wipe unrelated files as far as the file system sees it, but it might overwrite some previously discarded internal blocks. "inodes" is a too high-level concept in this context. You can read more about it on wikipedia which might be an authorative enough source for you? https://en.wikipedia.org/wiki/Wear_leveling https://en.wikipedia.org/wiki/Wear_leveling Also note that most ssds actually have more internal blocks available than what they present to the host device so that they can have a "cache" to be able to move things around internally, and also so they can mark certain positions as "bad" when writes to one internal block start failing and still operate properly.