NCR5380: bit MR_DMA_MODE set twice in NCR5380_transfer_dma()
Subsystems (2)
PATCH 1/1 NCR5380: bit MR_DMA_MODE set twice in NCR5380_transfer_dma()
Besides keeping the line short, the second setting of the MR_DMA_MODE bit was removed.
Signed-off-by Roel Kluin <[email protected]>
---
Please review, maybe some other bit should be set?
drivers/scsi/NCR5380.c +3 -1
@@ -1857,7 +1857,9 @@ static int NCR5380_transfer_dma(struct Scsi_Host *instance, unsigned char *phase#endif /* KLL May need eop and parity in 53c400 */ if (hostdata->flags &FLAG_NCR53C400 ) -NCR5380_write (MODE_REG ,MR_BASE |MR_DMA_MODE |MR_ENABLE_PAR_CHECK |MR_ENABLE_PAR_INTR |MR_ENABLE_EOP_INTR |MR_DMA_MODE |MR_MONITOR_BSY ); +NCR5380_write (MODE_REG ,MR_BASE |MR_DMA_MODE | +MR_ENABLE_PAR_CHECK |MR_ENABLE_PAR_INTR | +MR_ENABLE_EOP_INTR |MR_MONITOR_BSY );else NCR5380_write (MODE_REG ,MR_BASE |MR_DMA_MODE );#endif /* def REAL_DMA */
Quoted 6 lines · On Tue, 06 Oct 2009 22:17:43 +0200 Roel Kluin <[email protected]> wrote:
Besides keeping the line short, the second setting of the
MR_DMA_MODEbit was removed.Signed-off-by Roel Kluin <[email protected]>---
Please review, maybe some other bit should be set?
The only ones left are MR_BLOCK_DMA_MODE and MR_ARBITRATE.
No, I don't know if they hould have been set either. Let that sleeping dog lie, I suspect.
Quoted 15 lines
drivers/scsi/NCR5380.c +3
-1@@ -1857,7 +1857,9 @@ static int NCR5380_transfer_dma(struct Scsi_Host *instance, unsigned char *phase#endif /* KLL May need eop and parity in 53c400 */ if (hostdata->flags &FLAG_NCR53C400 ) -NCR5380_write (MODE_REG ,MR_BASE |MR_DMA_MODE |MR_ENABLE_PAR_CHECK |MR_ENABLE_PAR_INTR |MR_ENABLE_EOP_INTR |MR_DMA_MODE |MR_MONITOR_BSY ); +NCR5380_write (MODE_REG ,MR_BASE |MR_DMA_MODE | +MR_ENABLE_PAR_CHECK |MR_ENABLE_PAR_INTR | +MR_ENABLE_EOP_INTR |MR_MONITOR_BSY );else NCR5380_write (MODE_REG ,MR_BASE |MR_DMA_MODE );#endif /* def REAL_DMA */