On 07/09/12 17:21, Marc Zyngier wrote: > On 07/09/12 17:18, Jessica Allison wrote: >> On 07/09/12 16:08, Marc Zyngier wrote: >>> On 07/09/12 14:04, Jessica Allison wrote: >>>> On 07/09/12 12:22, Marc Zyngier wrote: >>>>> On 07/09/12 10:15, Jessica Allison wrote: >>>>>> Hi, I'm trying to get guest networking to function on the Cortex-A15x2 >>>>>> FastModel. >>>>> [...] >>>>> >>>>>> My only idea would be that the guest kernel device tree does not include >>>>>> the SMC device? Or does anyone have another idea why the lan9118 network >>>>>> card does not seem to show up in the guest? Has anyone ever managed to >>>>>> get guest networking to function? >>>>> Works well enough for me: >>>>> >>>>> <demo> >>>>> root@linaro:~# dmesg|grep smsc >>>>> [ 0.827410] smsc911x: Driver version 2008-10-21 >>>>> [ 0.835281] libphy: smsc911x-mdio: probed >>>>> [ 0.838724] smsc911x 1a000000.ethernet: eth0: attached PHY driver [Generic PHY] (mii_bus:phy_addr=1a000000.etherne:01, irq=-1) >>>>> [ 0.850633] smsc911x 1a000000.ethernet: eth0: MAC Address: 52:54:00:12:34:56 >>>>> [ 1.065648] smsc911x 1a000000.ethernet: eth0: SMSC911x/921x identified at 0xf0060000, IRQ: 47 >>>>> root@linaro:~# ifconfig eth0 >>>>> eth0 Link encap:Ethernet HWaddr 52:54:00:12:34:56 >>>>> inet addr:10.0.2.15 Bcast:10.0.2.255 Mask:255.255.255.0 >>>>> UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 >>>>> RX packets:16 errors:0 dropped:0 overruns:0 frame:0 >>>>> TX packets:19 errors:0 dropped:0 overruns:0 carrier:0 >>>>> collisions:0 txqueuelen:1000 >>>>> RX bytes:3394 (3.3 KB) TX bytes:2942 (2.9 KB) >>>>> Interrupt:47 >>>>> >>>>> root@linaro:~# ping -c3 10.0.2.2 >>>>> PING 10.0.2.2 (10.0.2.2) 56(84) bytes of data. >>>>> 64 bytes from 10.0.2.2: icmp_req=1 ttl=255 time=1.47 ms >>>>> 64 bytes from 10.0.2.2: icmp_req=2 ttl=255 time=1.34 ms >>>>> 64 bytes from 10.0.2.2: icmp_req=3 ttl=255 time=1.38 ms >>>>> >>>>> --- 10.0.2.2 ping statistics --- >>>>> 3 packets transmitted, 3 received, 0% packet loss, time 2002ms >>>>> rtt min/avg/max/mdev = 1.343/1.403/1.478/0.063 ms >>>>> root@linaro:~# wget http://www.google.com >>>>> --2012-09-07 11:19:20-- http://www.google.com/ >>>>> Resolving www.google.com... 173.194.78.106, 173.194.78.104, 173.194.78.105, ... >>>>> Connecting to www.google.com|173.194.78.106|:80... connected. >>>>> HTTP request sent, awaiting response... 302 Found >>>>> Location: http://www.google.co.uk/ [following] >>>>> --2012-09-07 11:19:22-- http://www.google.co.uk/ >>>>> Resolving www.google.co.uk... 173.194.78.94, 2a00:1450:400c:c05::5e >>>>> Connecting to www.google.co.uk|173.194.78.94|:80... connected. >>>>> HTTP request sent, awaiting response... 200 OK >>>>> Length: unspecified [text/html] >>>>> Saving to: `index.html' >>>>> >>>>> [ <=> ] 11,010 --.-K/s in 0s >>>>> >>>>> 2012-09-07 11:19:26 (85.6 MB/s) - `index.html' saved [11010] >>>>> </demo> >>>>> >>>>> Slow, but it does work. >>>>> >>>>> So the question is about the DT you're using... My gut feeling is that >>>>> you're using the same DT as the model, which is the wrong thing to do. >>>>> >>>>> I use vexpress-v2f-2xv6-ca15x2.dts, which is a good match for what QEMU >>>>> emulates. >>>>> >>>>> M. >>>> With that DT I get a kernel bug though which I didn't get before: >>>> >>>> WARNING: at kernel/irq/irqdomain.c:135 irq_domain_legacy_revmap+0x20/0x48() >>>> [<c0013e4c>] (unwind_backtrace+0x0/0xf8) from [<c001d728>] >>>> (warn_slowpath_common+0x48/0x60) >>>> [<c001d728>] (warn_slowpath_common+0x48/0x60) from [<c001d808>] >>>> (warn_slowpath_null+0x1c/0x24) >>>> [<c001d808>] (warn_slowpath_null+0x1c/0x24) from [<c005f0b4>] >>>> (irq_domain_legacy_revmap+0x20/0x48) >>>> [<c005f0b4>] (irq_domain_legacy_revmap+0x20/0x48) from [<c005f80c>] >>>> (irq_create_mapping+0x20/0x120) >>>> [<c005f80c>] (irq_create_mapping+0x20/0x120) from [<c005f988>] >>>> (irq_create_of_mapping+0x7c/0xf0) >>>> [<c005f988>] (irq_create_of_mapping+0x7c/0xf0) from [<c02465e8>] >>>> (irq_of_parse_and_map+0x2c/0x34) >>>> [<c02465e8>] (irq_of_parse_and_map+0x2c/0x34) from [<c0246608>] >>>> (of_irq_to_resource+0x18/0x74) >>>> [<c0246608>] (of_irq_to_resource+0x18/0x74) from [<c0246688>] >>>> (of_irq_count+0x24/0x34) >>>> [<c0246688>] (of_irq_count+0x24/0x34) from [<c0246bc4>] >>>> (of_device_alloc+0x54/0x150) >>>> [<c0246bc4>] (of_device_alloc+0x54/0x150) from [<c0246cf4>] >>>> (of_platform_device_create_pdata+0x34/0x80) >>>> [<c0246cf4>] (of_platform_device_create_pdata+0x34/0x80) from >>>> [<c0246e18>] (of_platform_bus_create+0xd8/0x270) >>>> [<c0246e18>] (of_platform_bus_create+0xd8/0x270) from [<c02470ec>] >>>> (of_platform_populate+0x60/0x98) >>>> [<c02470ec>] (of_platform_populate+0x60/0x98) from [<c03bf4f8>] >>>> (v2m_dt_init+0x18/0x38) >>>> [<c03bf4f8>] (v2m_dt_init+0x18/0x38) from [<c03bb69c>] >>>> (customize_machine+0x1c/0x28) >>>> [<c03bb69c>] (customize_machine+0x1c/0x28) from [<c03b8874>] >>>> (do_one_initcall+0x90/0x160) >>>> [<c03b8874>] (do_one_initcall+0x90/0x160) from [<c03b8a40>] >>>> (kernel_init+0xfc/0x1bc) >>>> [<c03b8a40>] (kernel_init+0xfc/0x1bc) from [<c000eeac>] >>>> (kernel_thread_exit+0x0/0x8) >>>> ---[ end trace 1b75b31a2719ed1d ]--- >>>> >>>> Is that any serious? >>> Depends. Is that with qemu GIC or vgic? >>> >>> M. >> My host kernel is configured with >> >> CONFIG_HAVE_KVM_IRQCHIP=y >> CONFIG_KVM_MMIO=y >> CONFIG_KVM=y >> CONFIG_KVM_ARM_HOST=y >> CONFIG_KVM_ARM_VGIC=y >> >> which means I am using vgic? > No, that just mean you've compiled the code. Your qemu command line > needs something like: > "-machine type=vexpress-a15,accel=kvm,kernel_irqchip=on" > > M. OK. Running with kernel_irqchip=on gives me the same error. For now it doesn't seem to be a problem. The guest runs OK. Which is currently the best branch (most features and best performance) to work on from scm/linux/kernel/git/maz/arm-platforms.git? Jess _______________________________________________ kvmarm mailing list kvmarm@xxxxxxxxxxxxxxxxxxxxx https://lists.cs.columbia.edu/cucslists/listinfo/kvmarm
