On Wed, Jan 29, 2025 at 09:39:42AM +0100, Christian Kühnke wrote: > Hi all, > > I have been sent here from the linux-ide mailing list. Over there, I have > reported an issue with the 6.6 stable series of the kernel, starting with > 6.6.51. The root cause is as follows: > > On 12.09.2024, commit 872f86e1757bbb0a334ee739b824e47c448f5ebc ("ata: > libata-scsi: Check ATA_QCFLAG_RTF_FILLED before using result_tf") was > applied to 6.6, adding checks of ATA_QCFLAG_RTF_FILLED to libata_scsi. The > patch seen in baseline commit 18676c6aab0863618eb35443e7b8615eea3535a9 > ("ata: libata-core: Set ATA_QCFLAG_RTF_FILLED in fill_result_tf()") should > have gone together with this. > > Without it, I receive errors retrieving SMART data from SATA disks via a > C602 SAS controller, apparently because in this situation > ATA_QCFLAG_RTF_FILLED is not set. > > I applied 18676c6aab0863618eb35443e7b8615eea3535a9 from baseline to 6.6.74 > and the problem went away. > > If you need any further information, do not hesitate to contact me (linux > user since 0.99.x, but only debugging it once every few years or so...). > Now queued up, thanks. greg k-h