From: James Bottomley Date: Sat, 10 Jun 2006 15:51:23 +0000 (-0500) Subject: [SCSI] spi transport: don't allow dt to be set on SE or HVD buses X-Git-Url: http://git.lede-project.org./?a=commitdiff_plain;h=60eef25701d25e99c991dd0f4a9f3832a0c3ad3e;p=openwrt%2Fstaging%2Fblogic.git [SCSI] spi transport: don't allow dt to be set on SE or HVD buses This is really just a belt and braces test. The standards require disks to respond DT not capable on a non-LVD bus ... however, not all disks follow the standards ... Signed-off-by: James Bottomley --- diff --git a/drivers/scsi/scsi_transport_spi.c b/drivers/scsi/scsi_transport_spi.c index 5829f97a2068..ace49d5bd9c4 100644 --- a/drivers/scsi/scsi_transport_spi.c +++ b/drivers/scsi/scsi_transport_spi.c @@ -785,6 +785,7 @@ spi_dv_device_internal(struct scsi_device *sdev, u8 *buffer) { struct spi_internal *i = to_spi_internal(sdev->host->transportt); struct scsi_target *starget = sdev->sdev_target; + struct Scsi_Host *shost = sdev->host; int len = sdev->inquiry_len; /* first set us up for narrow async */ DV_SET(offset, 0); @@ -844,6 +845,14 @@ spi_dv_device_internal(struct scsi_device *sdev, u8 *buffer) if (spi_min_period(starget) == 8) DV_SET(pcomp_en, 1); } + /* now that we've done all this, actually check the bus + * signal type (if known). Some devices are stupid on + * a SE bus and still claim they can try LVD only settings */ + if (i->f->get_signalling) + i->f->get_signalling(shost); + if (spi_signalling(shost) == SPI_SIGNAL_SE || + spi_signalling(shost) == SPI_SIGNAL_HVD) + DV_SET(dt, 0); /* Do the read only INQUIRY tests */ spi_dv_retrain(sdev, buffer, buffer + sdev->inquiry_len, spi_dv_device_compare_inquiry);