Re: [PATCH net] virtio_net: Do not pull payload in skb->head

On 4/11/21 2:43 PM, Eric Dumazet wrote:

On Sun, Apr 11, 2021 at 11:32 PM Guenter Roeck [off-list ref] wrote:

quoted

On 4/11/21 2:23 PM, Eric Dumazet wrote:

quoted

On Sun, Apr 11, 2021 at 10:37 PM Guenter Roeck [off-list ref] wrote:

quoted

On 4/11/21 8:06 AM, Eric Dumazet wrote:

quoted

On Sun, Apr 11, 2021 at 3:43 PM Guenter Roeck [off-list ref] wrote:

quoted

This patch causes a virtio-net interface failure when booting sh4 images
in qemu. The test case is nothing special: Just try to get an IP address
using udhcpc. If it fails, udhcpc reports:

udhcpc: started, v1.33.0
udhcpc: sending discover
FAIL

Can you investigate where the incoming packet is dropped ?

Unless I am missing something, packets are not dropped. It looks more
like udhcpc gets bad indigestion in the receive path and exits immediately.
Plus, it doesn't happen all the time; sometimes it receives the discover
response and is able to obtain an IP address.

Overall this is quite puzzling since udhcpc exits immediately when the problem
is seen, no matter which option I give it on the command line; it should not
really do that.

Could you strace both cases and report differences you can spot ?

strace -o STRACE -f -s 1000 udhcpc

I'll give it a try. It will take a while; I'll need to add strace to my root
file systems first.

As a quick hack, I added some debugging into the kernel; it looks like
the data part of the dhcp discover response may get lost with your patch
in place.

Data is not lost, the payload is whole contained in skb frags, which
was expected from my patch.

Maybe this sh arch does something wrong in this case.

This could be checksuming...

Please check

nstat -n
<run udhcpc>
nstat

Does that tell you anything ?

/ # nstat -n
/ # udhcpc -n -q
udhcpc: started, v1.33.0
udhcpc: sending discover
/ # nstat
#kernel
IpInReceives                    1                  0.0
IpInDelivers                    1                  0.0
UdpIgnoredMulti                 1                  0.0
IpExtInBcastPkts                1                  0.0
IpExtInOctets                   576                0.0
IpExtInBcastOctets              576                0.0
IpExtInNoECTPkts                1                  0.0

Also, one interesting detail is that the problem is not seen all the time,
even with your patch in place. Not sure if I mentioned that before. strace
output in the success case (same image, with patch in place) looks as follows.

130   write(2, "udhcpc: sending discover\n", 25) = 25
130   socket(AF_PACKET, SOCK_DGRAM, htons(ETH_P_IP)) = 6
130   bind(6, {sa_family=AF_PACKET, sll_protocol=htons(ETH_P_IP), sll_ifindex=if_nametoindex("eth0"), sll_hatype=ARPHRD_NETROM, sll_pkttype=PACKET_HOST, sll_halen=6, sll_addr=[0xff, 0xff, 0xff, 0xf
130   sendto(6, "E\0\1H\0\0\0\0@\21y\246\0\0\0\0\377\377\377\377\0D\0C\0014\227r\1\1\6\0\5\16\36P\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0RT\0\0224V\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0
130   close(6)                          = 0
130   fcntl64(5, F_SETFD, FD_CLOEXEC)   = 0
130   clock_gettime64(CLOCK_MONOTONIC, {tv_sec=25, tv_nsec=202168729}) = 0
130   poll([{fd=3, events=POLLIN}, {fd=5, events=POLLIN}], 2, 3000) = 1 ([{fd=5, revents=POLLIN}])
130   read(3, 0x7bf47a73, 1)            = -1 EAGAIN (Resource temporarily unavailable)
130   recvmsg(5, {msg_name=NULL, msg_namelen=0, msg_iov=[{iov_base="E\20\2@\0\10\0\0@\21l\224\n\0\2\2\377\377\377\377\0C\0D\2,\230\23\2\1\6\0\5\16\36P\0\0\0\0\0\0\0\0\n\0\2\17\n\0\2\2\0\0\0\0RT\0\0
130   clock_gettime64(CLOCK_MONOTONIC, {tv_sec=25, tv_nsec=205504062}) = 0
130   fcntl64(5, F_SETFD, FD_CLOEXEC)   = 0
130   socket(AF_INET, SOCK_RAW, IPPROTO_RAW) = 6
130   ioctl(6, SIOCGIFINDEX, {ifr_name="eth0", }) = 0
130   ioctl(6, SIOCGIFHWADDR, {ifr_name="eth0", ifr_hwaddr={sa_family=ARPHRD_ETHER, sa_data=52:54:00:12:34:56}}) = 0
130   close(6)                          = 0
130   clock_gettime64(CLOCK_MONOTONIC, {tv_sec=25, tv_nsec=208676862}) = 0
130   write(2, "udhcpc: sending select for 10.0.2.15\n", 37) = 37
130   socket(AF_PACKET, SOCK_DGRAM, htons(ETH_P_IP)) = 6
130   bind(6, {sa_family=AF_PACKET, sll_protocol=htons(ETH_P_IP), sll_ifindex=if_nametoindex("eth0"), sll_hatype=ARPHRD_NETROM, sll_pkttype=PACKET_HOST, sll_halen=6, sll_addr=[0xff, 0xff, 0xff, 0xf
130   sendto(6, "E\0\1H\0\0\0\0@\21y\246\0\0\0\0\377\377\377\377\0D\0C\0014\25Y\1\1\6\0\5\16\36P\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0RT\0\0224V\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\
130   close(6)                          = 0
130   fcntl64(5, F_SETFD, FD_CLOEXEC)   = 0
130   clock_gettime64(CLOCK_MONOTONIC, {tv_sec=25, tv_nsec=213060729}) = 0
130   poll([{fd=3, events=POLLIN}, {fd=5, events=POLLIN}], 2, 3000) = 1 ([{fd=5, revents=POLLIN}])
130   read(3, 0x7bf47a73, 1)            = -1 EAGAIN (Resource temporarily unavailable)
130   recvmsg(5, {msg_name=NULL, msg_namelen=0, msg_iov=[{iov_base="E\20\2@\0\t\0\0@\21l\223\n\0\2\2\377\377\377\377\0C\0D\2,\225\23\2\1\6\0\5\16\36P\0\0\0\0\0\0\0\0\n\0\2\17\n\0\2\2\0\0\0\0RT\0\02
130   clock_gettime64(CLOCK_MONOTONIC, {tv_sec=25, tv_nsec=215453662}) = 0
130   write(2, "udhcpc: lease of 10.0.2.15 obtained, lease time 86400\n", 54) = 54

In that case, the output of nstat is as follows.

/ # nstat
#kernel
IpInReceives                    2                  0.0
IpInDelivers                    2                  0.0
UdpIgnoredMulti                 2                  0.0
IpExtInBcastPkts                2                  0.0
IpExtInOctets                   1152               0.0
IpExtInBcastOctets              1152               0.0
IpExtInNoECTPkts                2                  0.0

Guenter