> Date: Tue, 13 Jul 2010 12:45:08 +0200
> From: Claudio Jeker <[email protected]>
>
> On Sun, Jun 27, 2010 at 09:48:03AM -0700, Matthew Dempsky wrote:
> > Anyone still using ioprbs(4)? If so, please test the diff below and
> > report back to me.
> >
>
> No bad effects seen after light testing on a ADAPTEC 3200S that I found in
> a dark corner of the basement. Dmesg below.
>
> iop0 at pci0 dev 11 function 1 "DPT SmartRAID (I2O)" rev 0x02: : <ADAPTEC
> 3200S>, irq 11
Noticed that iop(4) screws up the dmesg printing. Diff below fixes it.
ok?
Index: iop_pci.c
===================================================================
RCS file: /cvs/src/sys/dev/pci/iop_pci.c,v
retrieving revision 1.7
diff -u -p -r1.7 iop_pci.c
--- iop_pci.c 8 Apr 2010 00:23:53 -0000 1.7
+++ iop_pci.c 13 Jul 2010 11:34:42 -0000
@@ -98,7 +98,6 @@ iop_pci_attach(struct device *parent, st
sc = (struct iop_softc *)self;
pa = (struct pci_attach_args *)aux;
pc = pa->pa_pc;
- printf(": ");
/*
* The kernel always uses the first memory mapping to communicate
@@ -110,14 +109,14 @@ iop_pci_attach(struct device *parent, st
break;
}
if (i == PCI_MAPREG_END) {
- printf("can't find mapping\n");
+ printf(": can't find mapping\n");
return;
}
/* Map the register window. */
if (pci_mapreg_map(pa, i, PCI_MAPREG_TYPE_MEM, 0, &sc->sc_iot,
&sc->sc_ioh, NULL, NULL, 0x40000)) {
- printf("%s: can't map register window\n", sc->sc_dv.dv_xname);
+ printf(": can't map register window\n");
return;
}
@@ -127,14 +126,14 @@ iop_pci_attach(struct device *parent, st
/* Map and establish the interrupt. XXX IPL_BIO. */
if (pci_intr_map(pa, &ih)) {
- printf("can't map interrupt\n");
+ printf(": can't map interrupt\n");
return;
}
intrstr = pci_intr_string(pc, ih);
sc->sc_ih = pci_intr_establish(pc, ih, IPL_BIO, iop_intr, sc,
sc->sc_dv.dv_xname);
if (sc->sc_ih == NULL) {
- printf("can't establish interrupt");
+ printf(": can't establish interrupt");
if (intrstr != NULL)
printf(" at %s", intrstr);
printf("\n");
