On Tue, Dec 23, 2003 at 10:12:15AM +0000, Mark Powell wrote:
Davey wrote:
quoted
Some time ago Stefan and I had reported that both ethernet interfaces
cannot work well together. Sorry I have lost all old mails when I
reinstall my computer. We have find that this problem is caused by
that the base address of eht1 is wrong set. Follow by my patch for
linux-2.4.22 and linux-2.4.23. I have test the patch and it work well.
shouldn't that just be:
This is the right way to do it, we can't put that other hack in the
kernel. If somebody verifies this patch hen we can get it in.
-Matt
** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/
I have test the way that Mark Powell provide and it doesn't work well. When I run "ifconfig eth1 up" I get output "SIOCSIFFLAGS: Device or resource busy".
I think we should fix more code as follow:
file: driver/net/ibm_emac/ibm_ocp_enet.c
------------------------------------------------------------------------------------------------------
--line 1150
/* Request our interrupt lines */
rc = request_irq(dev->irq, emac_mac_irq, 0, "OCP EMAC MAC", dev);
if (rc != 0)
goto bail;
--- rc = request_irq(fep->wol_irq, emac_wakeup_irq, 0, "OCP EMAC Wakeup", dev);
+++ rc = request_irq(fep->wol_irq, emac_wakeup_irq, SHIRQ, "OCP EMAC Wakeup", dev);
if (rc != 0) {
free_irq(dev->irq, dev);
goto bail;
}
-------------------------------------------------------------------------------------------------------
Davey Wu
----- Original Message -----
From: "Matt Porter" <mporter@kernel.crashing.org>
To: "Mark Powell" <redacted>
Cc: <redacted>; "Davey" <redacted>
Sent: Wednesday, December 24, 2003 1:10 AM
Subject: Re: patch for ibm 405ep emac driver in linux-2.4.22 and 2.4.23
On Tue, Dec 23, 2003 at 10:12:15AM +0000, Mark Powell wrote:
quoted
Davey wrote:
quoted
Some time ago Stefan and I had reported that both ethernet interfaces
cannot work well together. Sorry I have lost all old mails when I
reinstall my computer. We have find that this problem is caused by
that the base address of eht1 is wrong set. Follow by my patch for
linux-2.4.22 and linux-2.4.23. I have test the patch and it work well.
shouldn't that just be:
This is the right way to do it, we can't put that other hack in the
kernel. If somebody verifies this patch hen we can get it in.
-Matt
** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/
On Tue, Dec 23, 2003 at 10:10:22AM -0700, Matt Porter wrote:
On Tue, Dec 23, 2003 at 10:12:15AM +0000, Mark Powell wrote:
quoted
Davey wrote:
quoted
Some time ago Stefan and I had reported that both ethernet interfaces
cannot work well together. Sorry I have lost all old mails when I
reinstall my computer. We have find that this problem is caused by
that the base address of eht1 is wrong set. Follow by my patch for
linux-2.4.22 and linux-2.4.23. I have test the patch and it work well.
shouldn't that just be:
This is the right way to do it, we can't put that other hack in the
kernel. If somebody verifies this patch hen we can get it in.
Matt, I checked old-OCP version of ibm405ep.c and indeed it had
EMAC1_BASE there. Obviously this is a typo introduced during
conversion to new OCP.
I checked these definitions when Stefan first reported the problem
but failed to notice this :-(.
Eugene.
** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/