On Mon, 2005-07-25 at 10:07 +0200, Hannes Reinecke wrote:
> But this is not the root problem.
>
> target5:0:10: FAST-160 WIDE SCSI 320.0 MB/s ST IU (6.25 ns, offset 127)
> scsi5: target 10 synchronous with period = 0x8, offset =
> 0x7f(RDSTRM|DT|IU|QAS)
>
> So, the ppr_options (which have generated the second line) are set ok,
> but they somehow haven't been transferred into the scsi_transport attribute.
Fortunately, HP gave me a U160 AIC card at OLS, so I can now boot it up
and see that the aic7xxx driver has the same problem. Attached is the
fix (basically it was a single bit variable rounding error).
James
diff --git a/drivers/scsi/aic7xxx/aic7xxx_osm.c b/drivers/scsi/aic7xxx/aic7xxx_osm.c
--- a/drivers/scsi/aic7xxx/aic7xxx_osm.c
+++ b/drivers/scsi/aic7xxx/aic7xxx_osm.c
@@ -1635,9 +1635,9 @@ ahc_send_async(struct ahc_softc *ahc, ch
spi_period(starget) = tinfo->curr.period;
spi_width(starget) = tinfo->curr.width;
spi_offset(starget) = tinfo->curr.offset;
- spi_dt(starget) = tinfo->curr.ppr_options & MSG_EXT_PPR_DT_REQ;
- spi_qas(starget) = tinfo->curr.ppr_options & MSG_EXT_PPR_QAS_REQ;
- spi_iu(starget) = tinfo->curr.ppr_options & MSG_EXT_PPR_IU_REQ;
+ spi_dt(starget) = tinfo->curr.ppr_options & MSG_EXT_PPR_DT_REQ ? 1 : 0;
+ spi_qas(starget) = tinfo->curr.ppr_options & MSG_EXT_PPR_QAS_REQ ? 1 : 0;
+ spi_iu(starget) = tinfo->curr.ppr_options & MSG_EXT_PPR_IU_REQ ? 1 : 0;
spi_display_xfer_agreement(starget);
break;
}
@@ -2435,8 +2435,10 @@ static void ahc_linux_set_dt(struct scsi
if (dt) {
period = 9; /* 12.5ns is the only period valid for DT */
ppr_options |= MSG_EXT_PPR_DT_REQ;
- } else if (period == 9)
+ } else if (period == 9) {
period = 10; /* if resetting DT, period must be >= 25ns */
+ ppr_options &= ~MSG_EXT_PPR_DT_REQ;
+ }
ahc_compile_devinfo(&devinfo, shost->this_id, starget->id, 0,
starget->channel + 'A', ROLE_INITIATOR);
-
: send the line "unsubscribe linux-scsi" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
