I sent the first in a previous patch, but didn't notice the other issues below initially. (should I make a delta patch? and how do you do that?) --------------------------->8-------------8<------------------------------ With a postfix decrement timeouts will reach -1 rather than 0, so the errors do not appear.
Signed-off-by: Roel Kluin <[email protected]> --- diff --git a/drivers/scsi/pcmcia/nsp_cs.c b/drivers/scsi/pcmcia/nsp_cs.c index 11a61ea..70b60ad 100644 --- a/drivers/scsi/pcmcia/nsp_cs.c +++ b/drivers/scsi/pcmcia/nsp_cs.c @@ -530,7 +530,7 @@ static int nsp_negate_signal(struct scsi_cmnd *SCpnt, unsigned char mask, if (reg == 0xff) { break; } - } while ((time_out-- != 0) && (reg & mask) != 0); + } while ((--time_out != 0) && (reg & mask) != 0); if (time_out == 0) { nsp_msg(KERN_DEBUG, " %s signal off timeut", str); @@ -801,7 +801,7 @@ static void nsp_pio_read(struct scsi_cmnd *SCpnt) data->FifoCount = ocount; - if (time_out == 0) { + if (time_out < 0) { nsp_msg(KERN_DEBUG, "pio read timeout resid=%d this_residual=%d buffers_residual=%d", scsi_get_resid(SCpnt), SCpnt->SCp.this_residual, SCpnt->SCp.buffers_residual); @@ -897,7 +897,7 @@ static void nsp_pio_write(struct scsi_cmnd *SCpnt) data->FifoCount = ocount; - if (time_out == 0) { + if (time_out < 0) { nsp_msg(KERN_DEBUG, "pio write timeout resid=0x%x", scsi_get_resid(SCpnt)); } _______________________________________________ Linux PCMCIA reimplementation list http://lists.infradead.org/mailman/listinfo/linux-pcmcia
