Date: Sun, 8 May 2011 11:43:24 +0200 From: Bernhard Schmidt <bschmidt@freebsd.org> To: freebsd-wireless@freebsd.org Subject: Re: Ath issues on -CURRENT > Mar 31 Message-ID: <201105081143.24478.bschmidt@freebsd.org> In-Reply-To: <20110508082212.GA72727@solitaire.psych0tik.net> References: <20110508082212.GA72727@solitaire.psych0tik.net>
next in thread | previous in thread | raw e-mail | index | archive | help
--Boundary-00=_8WmxN9fgqRawn8x
Content-Type: Text/Plain;
charset="iso-8859-15"
Content-Transfer-Encoding: 7bit
On Sunday 08 May 2011 10:22:12 richo wrote:
> Hi,
>
> I use an atheros based pccard nic on my laptop with 9-CURRENT.
>
> The last time I was able to build a working kernel was 20 march (the current
> kernel I'm using is=20
> FreeBSD solitaire.psych0tik.net 9.0-CURRENT FreeBSD 9.0-CURRENT #1 r219786:=
> Sun Mar 20 15:26:26 EST 2011 richo@solitaire.psych0tik.net:/usr/obj/us=
> r/src/sys/GENERIC i386
>
> a few weeks later UPDATING included this message:
> 20110331:
> ath(4) has been split into bus- and device- modules. if_ath contains
> the HAL, the TX rate control and the network device code. if_ath_pci
> contains the PCI bus glue. For Atheros MIPS embedded systems, if_ath_a
> contains the AHB glue. Users need to load both if_ath_pci and if_ath
> in order to use ath on everything else.
>
> TO REPEAT: if_ath_ahb is not needed for normal users. Normal users onl
> need to load if_ath and if_ath_pci for ath(4) operation.
>
> I'm still using default config, but any kernels I build panic if I boot with
> the NIC plugged in. They boot fine with it unplugged, but if I plug in the
> card
> it immediately panics.
>
> Booting with verbose logging on, I see this (I typed this out by hand after
> writing it down, so it's possible that it's slightly incorrect. I did double
> check it, though):
>
> pcib2: allocated memory range (0xf6010000-0xf601ffff) for rid 10 of pci0:3:0:0
> unkown: Lazy allovation of 0x10000 bytes rid 0x10 type 3 at 0xf6010000
> cbb0 Opening memory:
> cbb0 Normal: (0xg6010000-0xf601ffff)
> cbb0: Opening memory:
> panic: duplcate map 10
>
> I have confirmed with pciconf -l that pci0:3:0:0 is my ath nic.
>
> I have the backtrace, but I only wrote down the function names. I can write
> down the whole lot if it will help though.
I was running into the same issue with a ral(4) cardbus adapter.
John Baldwin provided the attached patch which fixed the issue
for me.
--
Bernhard
--Boundary-00=_8WmxN9fgqRawn8x
Content-Type: text/x-patch;
charset="ISO-8859-1";
name="cardbus_cis.diff"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
filename="cardbus_cis.diff"
Index: sys/dev/cardbus/cardbus_cis.c
===================================================================
--- sys/dev/cardbus/cardbus_cis.c (revision 221619)
+++ sys/dev/cardbus/cardbus_cis.c (working copy)
@@ -430,7 +430,10 @@ cardbus_read_tuple_finish(device_t cbdev, device_t
{
if (res != CIS_CONFIG_SPACE) {
bus_release_resource(child, SYS_RES_MEMORY, rid, res);
+#if 0
+ /* XXX: Not anymore. */
bus_delete_resource(child, SYS_RES_MEMORY, rid);
+#endif
}
}
Index: sys/dev/pci/pci.c
===================================================================
--- sys/dev/pci/pci.c (revision 221619)
+++ sys/dev/pci/pci.c (working copy)
@@ -2576,6 +2576,17 @@ pci_add_map(device_t bus, device_t dev, int reg, s
uint16_t cmd;
struct resource *res;
+ /*
+ * The BAR may already exist if the device is a CardBus card
+ * whose CIS is stored in this BAR.
+ */
+ pm = pci_find_bar(dev, reg);
+ if (pm != NULL) {
+ maprange = pci_maprange(pm->pm_value);
+ barlen = maprange == 64 ? 2 : 1;
+ return (barlen);
+ }
+
pci_read_bar(dev, reg, &map, &testval);
if (PCI_BAR_MEM(map)) {
type = SYS_RES_MEMORY;
--Boundary-00=_8WmxN9fgqRawn8x--
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?201105081143.24478.bschmidt>