If we get interrupted during a data phase and the DMA engine is still

pumping data despite our scsi data counters being at 0, something has
gone massively wrong.  The consequence of happily ignoring this is more
DMA phase errors and a disk full of spammed sectors.  Instead, panic on
the first occurance to hopefully limit the damage.

MFC After: 3 days

1 files changed, 6 insertions, 0 deletions

diff --git a/sys/dev/esp/ncr53c9x.c b/sys/dev/esp/ncr53c9x.c
index 8523a4b..f9f08b4 100644
--- a/sys/dev/esp/ncr53c9x.c
+++ b/sys/dev/esp/ncr53c9x.c
@@ -2228,6 +2228,11 @@ again:
 				 * a DATA transfer. Print a diagnostic
 				 * if the DMA counter and TC bit
 				 * appear to be out of sync.
+				 *
+				 * XXX This is fatal and usually means that
+				 *     the DMA engine is hopelessly out of
+				 *     sync with reality.  A disk is likely
+				 *     getting spammed at this point.
 				 */
 				device_printf(sc->sc_dev, "!TC on DATA XFER"
 				    " [intr %x, stat %x, step %d]"
@@ -2237,6 +2242,7 @@ again:
 				    sc->sc_espstep,
 				    sc->sc_prevphase,
 				    ecb ? ecb->dleft : -1);
+				panic("esp: unrecoverable DMA error");
 			}
 		}
 	}