Tape drives are returning an "Input/output error" when a drive, the host, or the bus is reset. This behavior is expected when a tape is present to prevent data loss. However, the driver currently drops these errors even for empty drives due to the following change: Commit: 9604eea5bd3ae1fa3c098294f4fc29ad687141ea Subject: scsi: st: Add third-party power-on reset handling Link: https://github.com/torvalds/linux/commit/9604eea5bd3ae1fa3c098294f4fc29ad687141ea This issue is causing several tape software applications to crash on startup or when performing drive health checks, as noted in the following CERN CTA Tape software discussion: https://cta-community.web.cern.ch/t/input-output-error-from-tape-drive-device-dev-nst0/302 To correct this behavior, it is necessary to either check for the presence of a tape before blocking the device or revise the drive's readiness verification at the beginning of the flush function. Signed-off-by: Rafael Rocha <rrochavi@xxxxxxxx> --- drivers/scsi/st.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/drivers/scsi/st.c b/drivers/scsi/st.c index 0d8ce1a92168..10bda3543e93 100644 --- a/drivers/scsi/st.c +++ b/drivers/scsi/st.c @@ -834,6 +834,9 @@ static int flush_buffer(struct scsi_tape *STp, int seek_next) int backspace, result; struct st_partstat *STps; + if (STp->ready != ST_READY) + return 0; + /* * If there was a bus reset, block further access * to this device. @@ -841,8 +844,6 @@ static int flush_buffer(struct scsi_tape *STp, int seek_next) if (STp->pos_unknown) return (-EIO); - if (STp->ready != ST_READY) - return 0; STps = &(STp->ps[STp->partition]); if (STps->rw == ST_WRITING) /* Writing */ return st_flush_write_buffer(STp); -- 2.39.3 (Apple Git-146)
