[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Qemu-devel] [PATCH v9 01/22] virtio-net: byteswap virtio-net header
From: |
Greg Kurz |
Subject: |
[Qemu-devel] [PATCH v9 01/22] virtio-net: byteswap virtio-net header |
Date: |
Tue, 24 Jun 2014 19:11:32 +0200 |
User-agent: |
StGit/0.17-dirty |
From: Cédric Le Goater <address@hidden>
TCP connectivity fails when the guest has a different endianness.
The packets are silently dropped on the host by the tap backend
when they are read from user space because the endianness of the
virtio-net header is in the wrong order. These lines may appear
in the guest console:
[ 454.709327] skbuff: bad partial csum: csum=8704/4096 len=74
[ 455.702554] skbuff: bad partial csum: csum=8704/4096 len=74
The issue that got first spotted with a ppc64le PowerKVM guest,
but it also exists for the less common case of a x86_64 guest run
by a big-endian ppc64 TCG hypervisor.
Signed-off-by: Cédric Le Goater <address@hidden>
[ Ported from PowerKVM,
Greg Kurz <address@hidden> ]
Signed-off-by: Greg Kurz <address@hidden>
---
This patch had been mistakenly folded into patch 14/20 in v8
hw/net/virtio-net.c | 17 +++++++++++++++++
1 file changed, 17 insertions(+)
diff --git a/hw/net/virtio-net.c b/hw/net/virtio-net.c
index 00b5e07..c39db7c 100644
--- a/hw/net/virtio-net.c
+++ b/hw/net/virtio-net.c
@@ -875,6 +875,14 @@ static int virtio_net_has_buffers(VirtIONetQueue *q, int
bufsize)
return 1;
}
+static void virtio_net_hdr_swap(struct virtio_net_hdr *hdr)
+{
+ tswap16s(&hdr->hdr_len);
+ tswap16s(&hdr->gso_size);
+ tswap16s(&hdr->csum_start);
+ tswap16s(&hdr->csum_offset);
+}
+
/* dhclient uses AF_PACKET but doesn't pass auxdata to the kernel so
* it never finds out that the packets don't have valid checksums. This
* causes dhclient to get upset. Fedora's carried a patch for ages to
@@ -910,6 +918,7 @@ static void receive_header(VirtIONet *n, const struct iovec
*iov, int iov_cnt,
void *wbuf = (void *)buf;
work_around_broken_dhclient(wbuf, wbuf + n->host_hdr_len,
size - n->host_hdr_len);
+ virtio_net_hdr_swap(wbuf);
iov_from_buf(iov, iov_cnt, 0, buf, sizeof(struct virtio_net_hdr));
} else {
struct virtio_net_hdr hdr = {
@@ -1118,6 +1127,14 @@ static int32_t virtio_net_flush_tx(VirtIONetQueue *q)
exit(1);
}
+ if (n->has_vnet_hdr) {
+ if (out_sg[0].iov_len < n->guest_hdr_len) {
+ error_report("virtio-net header incorrect");
+ exit(1);
+ }
+ virtio_net_hdr_swap((void *) out_sg[0].iov_base);
+ }
+
/*
* If host wants to see the guest header as is, we can
* pass it on unchanged. Otherwise, copy just the parts
- [Qemu-devel] [PATCH v9 00/22] legacy virtio support for cross-endian targets, Greg Kurz, 2014/06/24
- [Qemu-devel] [PATCH v9 01/22] virtio-net: byteswap virtio-net header,
Greg Kurz <=
- [Qemu-devel] [PATCH v9 02/22] virtio-serial: don't migrate the config space, Greg Kurz, 2014/06/24
- [Qemu-devel] [PATCH v9 03/22] virtio: introduce device specific migration calls, Greg Kurz, 2014/06/24
- [Qemu-devel] [PATCH v9 04/22] virtio-net: implement per-device migration calls, Greg Kurz, 2014/06/24
- [Qemu-devel] [PATCH v9 05/22] virtio-blk: implement per-device migration calls, Greg Kurz, 2014/06/24
- [Qemu-devel] [PATCH v9 06/22] virtio-serial: implement per-device migration calls, Greg Kurz, 2014/06/24
- [Qemu-devel] [PATCH v9 07/22] virtio-balloon: implement per-device migration calls, Greg Kurz, 2014/06/24
- [Qemu-devel] [PATCH v9 08/22] virtio-rng: implement per-device migration calls, Greg Kurz, 2014/06/24
- [Qemu-devel] [PATCH v9 09/22] virtio: add subsections to the migration stream, Greg Kurz, 2014/06/24
- Re: [Qemu-devel] [PATCH v9 00/22] legacy virtio support for cross-endian targets, Michael S. Tsirkin, 2014/06/24