search for: checksum_hw

Another checksum offload problem was reported on xen-users, when using a domU as a firewall: http://lists.xensource.com/archives/html/xen-users/2006-04/msg01150.html It also fails without VLANs. The path from dom0->domU with ip_summed==CHECKSUM_HW/proto_csum_blank==1 is broken. - skb_checksum_setup() assumes that a checksum will definitely be calculated in dev_queue_xmit(), but the destination device is netback, which is advertising NETIF_F_IP_CSUM ---> no checksum. So, added reset_proto_csum_blank() to skbuff.h, and call it in...

Looks like the GSO is involved? I got this while running Dom0 only (no guests), with a BOINC/Rosetta@home application running on all 4 cores. changeset: 10649:8e55c5c11475 Build: x86_32p (pae). ------------[ cut here ]------------ kernel BUG at net/core/dev.c:1133! invalid opcode: 0000 [#1] SMP CPU: 0 EIP: 0061:[<c04dceb0>] Not tainted VLI EFLAGS: 00210297 (2.6.16.13-xen

...t/core/skbuff.c @@ -1896,6 +1896,29 @@ return 0; } +/** + * skb_pull_rcsum - pull skb and update receive checksum + * @skb: buffer to update + * @start: start of data before pull + * @len: length of data pulled + * + * This function performs an skb_pull on the packet and updates + * update the CHECKSUM_HW checksum. It should be used on receive + * path processing instead of skb_pull unless you know that the + * checksum difference is zero (e.g., a valid IP header) or you + * are setting ip_summed to CHECKSUM_NONE. + */ +unsigned char *skb_pull_rcsum(struct sk_buff *skb, unsigned int len) +{ + BUG_O...

...>+ "TCP/UDP packet, dropping a protocol" >+ " %d packet", skb->nh.iph->protocol); >+ goto out; > } > if ((skb->h.raw + skb->csum + 2) > skb->tail) >- goto out_kfree_skb; >+ goto out; > skb->ip_summed = CHECKSUM_HW; > } >+ >+ return 0; >+out: >+ return -EPROTO; >+} >+#else >+static inline int xen_checksum_setup(struct sk_buff *skb) {} > #endif >+ >+#define HARD_TX_LOCK(dev, cpu) { \ >+ if ((dev->features & NETIF_F_LLTX) == 0) { \ >+ spin_lock(&dev->xmit...

...ppears to be a good first * approximiation. */ -#define HAVE_NO_CSUM_OFFLOAD 1 +#define HAVE_CSUM_OFFLOAD 0 #define gso_size tso_size #define gso_segs tso_segs @@ -138,8 +140,12 @@ static inline int netif_needs_gso(struct unlikely(skb->ip_summed != CHECKSUM_HW)); } #else +#define HAVE_GSO 0 +#define HAVE_TSO 0 +#define HAVE_CSUM_OFFLOAD 0 #define netif_needs_gso(dev, skb) 0 #define dev_disable_gso_features(dev) ((void)0) +#define ethtool_op_set_tso(dev, data) (-ENOSYS) #endif #define GRANT_INVALID_REF 0 @@ -412,13 +418,12 @@ again: goto a...

Hello, I''m trying to setup Xen on a machine with 2 physical network devices using linux bonding without success. I haven''t received any answer on my question on xen-users, the only discussion on this seems to be at http://lists.xensource.com/archives/html/xen-devel/2005-10/msg01174.html Could someone shed some light on state/ plans regarding Xen and bonding? Regards --

Below is a fix for the current problem of checksum offload not working in a NAT''ed network. The cause is the NAT/iptables code incorrectly modifying the TCP/UDP checksum (for the checksum offload case). The original code assumes a valid checksum, which is not the case for checksum offload packets (which has a complimented, partial checksum for the hardware to use). The fix is to

Hi, Is the http://xenbits.xensource.com/linux-2.6-xen.hg tree still being updated? if not, what''s the preferred Linux tree to track that has all of the Xen bits? Thanks, Muli _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel

Unlike full virtualization in which the virtual machine provides the same platform interface as running natively on the hardware, paravirtualization requires modification to the guest operating system to work with the platform interface provided by the hypervisor. Xen was designed with performance in mind. Calls to the hypervisor are minimized, batched if necessary, and non-critical codepaths

Unlike full virtualization in which the virtual machine provides the same platform interface as running natively on the hardware, paravirtualization requires modification to the guest operating system to work with the platform interface provided by the hypervisor. Xen was designed with performance in mind. Calls to the hypervisor are minimized, batched if necessary, and non-critical codepaths

Unlike full virtualization in which the virtual machine provides the same platform interface as running natively on the hardware, paravirtualization requires modification to the guest operating system to work with the platform interface provided by the hypervisor. Xen was designed with performance in mind. Calls to the hypervisor are minimized, batched if necessary, and non-critical codepaths