Hannes Reinecke <hare@xxxxxxx> wrote: > Wouldn't the second truncate end up with a 4k file, and not an 8k? > IE the resulting file will be: > After step 1: 8k > After step 2: 4 > After step 3: 4k Yes, but the folio should still be an 8K folio, and it is: > pankaj-5833: netfs_folio: pfn=116fec i=0009e ix=00000-00001 inval-part as indicated by the inclusive folio index range ix=00000-00001. The problem is that the bottom four bytes of the file are getting cleared somewhere. They *should* be "XXXX", but they're all zeros. David
