https://bugs.kde.org/show_bug.cgi?id=526441
--- Comment #1 from Don <[email protected]> --- Astra additional review, jic: Confirmed with Qt 6.11.2 and a regular file: moving the underlying offset to EOF through a duplicated descriptor leaves Qt’s position unchanged. Repeated calls then return `atEnd() == false` and `readLineInto() == false`. Duplicated descriptors share their file offset, as documented in https://man7.org/linux/man-pages/man2/dup.2.html, so this reproducer requires no kernel bug. Failed reads empty the buffer, causing KConfig to `continue` without incrementing `errorCount`. A persistent mismatch therefore produces an unbounded loop. This confirms the failure mode, but does not establish what caused the mismatch in the original suspend/resume incident. -- You are receiving this mail because: You are watching all bug changes.
