)]}' { "log": [ { "commit": "c0560b9c523341516eabf0f3b51832256caa7bbb", "tree": "4ffbee34cbdf1988cfcf72cd48cbb147d23a48be", "parents": [ "36cbb2452cbafca64dcdd3578047433787900cf0" ], "author": { "name": "Joe Perches", "email": "joe@perches.com", "time": "Thu Nov 06 12:53:41 2014 -0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sat Nov 08 21:22:54 2014 -0500" }, "message": "dccp: Convert DCCP_WARN to net_warn_ratelimited\n\nRemove the dependency on the \"warning\" sysctl (net_msg_warn)\nwhich is only used by the LIMIT_NETDEBUG macro.\n\nConvert the LIMIT_NETDEBUG use in DCCP_WARN to the more\ncommon net_warn_ratelimited mechanism.\n\nThis still ratelimits based on the net_ratelimit()\nfunction, but removes the check for the sysctl.\n\nSigned-off-by: Joe Perches \u003cjoe@perches.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "36cbb2452cbafca64dcdd3578047433787900cf0", "tree": "3e50381574c3850fc2a6c76a14777daa821f85a0", "parents": [ "f46ad73ac6965f71d1a2ba217c4eb5b644e62949" ], "author": { "name": "Rick Jones", "email": "rick.jones2@hp.com", "time": "Thu Nov 06 10:37:54 2014 -0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Nov 07 15:45:50 2014 -0500" }, "message": "udp: Increment UDP_MIB_IGNOREDMULTI for arriving unmatched multicasts\n\nAs NIC multicast filtering isn\u0027t perfect, and some platforms are\nquite content to spew broadcasts, we should not trigger an event\nfor skb:kfree_skb when we do not have a match for such an incoming\ndatagram. We do though want to avoid sweeping the matter under the\nrug entirely, so increment a suitable statistic.\n\nThis incorporates feedback from David L. Stevens, Karl Neiss and Eric\nDumazet.\n\nV3 - use bool per David Miller\n\nSigned-off-by: Rick Jones \u003crick.jones2@hp.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "f46ad73ac6965f71d1a2ba217c4eb5b644e62949", "tree": "ebd282b43aa8427a79284448748df195dac08993", "parents": [ "df32dd2054b6edcbdfd3a31aec24e7fd0edba2ac" ], "author": { "name": "Oliver Neukum", "email": "oneukum@suse.de", "time": "Thu Nov 06 15:19:14 2014 +0100" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Nov 07 15:17:45 2014 -0500" }, "message": "cdc-ether: implement MULTICAST flag on the device\n\nOlivier having laid the groundwork this patch transmits the\nmulticast flag to the device to save some bus traffic.\n\nSigned-off-by: Oliver Neukum \u003coneukum@suse.de\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "df32dd2054b6edcbdfd3a31aec24e7fd0edba2ac", "tree": "3333516181f00f0be12f1e05fe878782f80b7f7f", "parents": [ "f10f9fb2167bfb5f02a63eb320813ff3c71eec98" ], "author": { "name": "stephen hemminger", "email": "stephen@networkplumber.org", "time": "Mon Nov 03 12:42:34 2014 -0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Nov 07 14:49:37 2014 -0500" }, "message": "uapi: resort Kbuild entries\n\nThe entries in the Kbuild files are incorrectly sorted.\nMatters for aesthetics only.\n\nSigned-off-by: Stephen Hemminger \u003cstephen@networkplumber.org\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "f10f9fb2167bfb5f02a63eb320813ff3c71eec98", "tree": "4decc8a43a7786c3178beadb2bba69896d146c64", "parents": [ "424c4f7875ee21e803f6ebeb8799b6185eda7864" ], "author": { "name": "Andy Shevchenko", "email": "andriy.shevchenko@linux.intel.com", "time": "Fri Nov 07 16:46:42 2014 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Nov 07 12:25:23 2014 -0500" }, "message": "stmmac: platform: fix sparse warnings\n\nThis patch fixes the following sparse warnings. One is fixed by casting return\nvalue to a return type of the function. The others by creating a specific\nstmmac_platform.h which provides the bits related to the platform driver.\n\ndrivers/net/ethernet/stmicro/stmmac/dwmac-meson.c:59:29: warning: incorrect type in return expression (different address spaces)\ndrivers/net/ethernet/stmicro/stmmac/dwmac-meson.c:59:29: expected void *\ndrivers/net/ethernet/stmicro/stmmac/dwmac-meson.c:59:29: got void [noderef] \u003casn:2\u003e*reg\n\ndrivers/net/ethernet/stmicro/stmmac/dwmac-meson.c:64:29: warning: symbol \u0027meson6_dwmac_data\u0027 was not declared. Should it be static?\ndrivers/net/ethernet/stmicro/stmmac/dwmac-sti.c:354:29: warning: symbol \u0027stih4xx_dwmac_data\u0027 was not declared. Should it be static?\ndrivers/net/ethernet/stmicro/stmmac/dwmac-sti.c:361:29: warning: symbol \u0027stid127_dwmac_data\u0027 was not declared. Should it be static?\ndrivers/net/ethernet/stmicro/stmmac/dwmac-sunxi.c:133:29: warning: symbol \u0027sun7i_gmac_data\u0027 was not declared. Should it be static?\n\nSigned-off-by: Andy Shevchenko \u003candriy.shevchenko@linux.intel.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "424c4f7875ee21e803f6ebeb8799b6185eda7864", "tree": "57053dedac0c9108b2c16a34ae873b74da75017a", "parents": [ "7f6d46704fe0895f33b81552244344422c1c647c" ], "author": { "name": "Andy Shevchenko", "email": "andriy.shevchenko@linux.intel.com", "time": "Fri Nov 07 16:53:12 2014 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Nov 07 12:20:43 2014 -0500" }, "message": "stmmac: remove custom implementation of print_hex_dump()\n\nThere is a kernel helper to dump buffers in a hexdecimal format. This patch\nsubstitutes the open coded function by calling that helper.\n\nThe output is slightly changed:\n - no lead space\n - ASCII part will be printed along with the dump\n - offset is longer than 3 characters (now 8)\n\nSigned-off-by: Andy Shevchenko \u003candriy.shevchenko@linux.intel.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "7f6d46704fe0895f33b81552244344422c1c647c", "tree": "b39b0d73e155780b340a337637c11c557d2181ca", "parents": [ "4e84b496fd2a226883920e0e0de4ed3f94898adf", "bfe1be38fcee0e13ad53175d0b530707c20f93ec" ], "author": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Nov 07 12:13:40 2014 -0500" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Nov 07 12:13:40 2014 -0500" }, "message": "Merge branch \u0027iov_iter\u0027\n\nHerbert Xu says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\nReplace skb_copy_datagram_const_iovec with iterator version\n\nThis patch series adds the helper skb_copy_datagram_iter, which\nis meant to replace both skb_copy_datagram_iovec and its evil\ntwin skb_copy_datagram_const_iovec.\n\nIt then converts tun and macvtap over to the new helper and finally\nremoves skb_copy_datagram_const_iovec which is only used by tun\nand macvtap.\n\nThe copy_to_iter return value issue pointed out by Al has now been\nfixed.\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "bfe1be38fcee0e13ad53175d0b530707c20f93ec", "tree": "b39b0d73e155780b340a337637c11c557d2181ca", "parents": [ "6c36d2e26cda1ad3e2c4b90dd843825fc62fe5b4" ], "author": { "name": "Herbert Xu", "email": "herbert@gondor.apana.org.au", "time": "Fri Nov 07 21:22:26 2014 +0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Nov 07 12:13:34 2014 -0500" }, "message": "net: Kill skb_copy_datagram_const_iovec\n\nNow that both macvtap and tun are using skb_copy_datagram_iter, we\ncan kill the abomination that is skb_copy_datagram_const_iovec.\n\nSigned-off-by: Herbert Xu \u003cherbert@gondor.apana.org.au\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "6c36d2e26cda1ad3e2c4b90dd843825fc62fe5b4", "tree": "1d98b6d2753df810746d49d0a62dfbb547087cb6", "parents": [ "e0b46d0ee9c240c7430a47e9b0365674d4a04522" ], "author": { "name": "Herbert Xu", "email": "herbert@gondor.apana.org.au", "time": "Fri Nov 07 21:22:25 2014 +0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Nov 07 12:13:34 2014 -0500" }, "message": "macvtap: Use iovec iterators\n\nThis patch removes the use of skb_copy_datagram_const_iovec in\nfavour of the iovec iterator-based skb_copy_datagram_iter.\n\nSigned-off-by: Herbert Xu \u003cherbert@gondor.apana.org.au\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "e0b46d0ee9c240c7430a47e9b0365674d4a04522", "tree": "383f614a420a3b5b843ee6147e39d3449e6d06e1", "parents": [ "a8f820aa4066d2c97e75ecd1bbca8a7920b66f2c" ], "author": { "name": "Herbert Xu", "email": "herbert@gondor.apana.org.au", "time": "Fri Nov 07 21:22:23 2014 +0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Nov 07 12:13:34 2014 -0500" }, "message": "tun: Use iovec iterators\n\nThis patch removes the use of skb_copy_datagram_const_iovec in\nfavour of the iovec iterator-based skb_copy_datagram_iter.\n\nSigned-off-by: Herbert Xu \u003cherbert@gondor.apana.org.au\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "a8f820aa4066d2c97e75ecd1bbca8a7920b66f2c", "tree": "69854ea942633a137e153ddd360aaa39aa2bdae6", "parents": [ "4e84b496fd2a226883920e0e0de4ed3f94898adf" ], "author": { "name": "Herbert Xu", "email": "herbert@gondor.apana.org.au", "time": "Fri Nov 07 21:22:22 2014 +0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Nov 07 12:13:34 2014 -0500" }, "message": "inet: Add skb_copy_datagram_iter\n\nThis patch adds skb_copy_datagram_iter, which is identical to\nskb_copy_datagram_iovec except that it operates on iov_iter\ninstead of iovec.\n\nEventually all users of skb_copy_datagram_iovec should switch\nover to iov_iter and then we can remove skb_copy_datagram_iovec.\n\nSigned-off-by: Herbert Xu \u003cherbert@gondor.apana.org.au\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "4e84b496fd2a226883920e0e0de4ed3f94898adf", "tree": "662e314870fabf5d6e3bf25b955bd537eb72de54", "parents": [ "5c91ae08e4500ebc8efd584c8b87c4040d5393ea", "9d01412ae76fec5274a3d94a28a3552a742a60dc" ], "author": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 22:01:18 2014 -0500" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 22:01:18 2014 -0500" }, "message": "Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net\n" }, { "commit": "5c91ae08e4500ebc8efd584c8b87c4040d5393ea", "tree": "9410a4b74a22553c71df13e81a52425da31a1a7e", "parents": [ "a1f5313cc4325a1e0feb3dd1a20fca1844dce72b" ], "author": { "name": "Tom Herbert", "email": "therbert@google.com", "time": "Thu Nov 06 18:06:01 2014 -0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 21:59:55 2014 -0500" }, "message": "vxlan: Fix to enable UDP checksums on interface\n\nAdd definition to vxlan nla_policy for UDP checksum. This is necessary\nto enable UDP checksums on VXLAN.\n\nIn some instances, enabling UDP checksums can improve performance on\nreceive for devices that return legacy checksum-unnecessary for UDP/IP.\nAlso, UDP checksum provides some protection against VNI corruption.\n\nTesting:\n\nRan 200 instances of TCP_STREAM and TCP_RR on bnx2x.\n\nTCP_STREAM\n IPv4, without UDP checksums\n 14.41% TX CPU utilization\n 25.71% RX CPU utilization\n 9083.4 Mbps\n IPv4, with UDP checksums\n 13.99% TX CPU utilization\n 13.40% RX CPU utilization\n 9095.65 Mbps\n\nTCP_RR\n IPv4, without UDP checksums\n 94.08% TX CPU utilization\n 156/248/462 90/95/99% latencies\n 1.12743e+06 tps\n IPv4, with UDP checksums\n 94.43% TX CPU utilization\n 158/250/462 90/95/99% latencies\n 1.13345e+06 tps\n\nSigned-off-by: Tom Herbert \u003ctherbert@google.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "a1f5313cc4325a1e0feb3dd1a20fca1844dce72b", "tree": "bfda906e32b1522b4d1c051cf207adb6ca7138f8", "parents": [ "6b798d70d07a38e9ba0a32316f76495f9dcfc343", "f5eecbbef0b7d28b171b3b873e1e4a69aabd7b58" ], "author": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 19:00:26 2014 -0500" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 19:00:26 2014 -0500" }, "message": "Merge branch \u0027amd-xgbe-next\u0027\n\nTom Lendacky says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\namd-xgbe: AMD XGBE driver updates 2014-11-06\n\nThe following series of patches fixes a couple of bugs that slipped\nthrough my last series.\n\n- Free channel structure after freeing the per channel interrupts\n- If an skb error allocation occurs during receive processing check\n whether more descriptors are associated with the packet or whether\n to start on a new packet\n\nThis patch series is based on net-next.\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "f5eecbbef0b7d28b171b3b873e1e4a69aabd7b58", "tree": "bfda906e32b1522b4d1c051cf207adb6ca7138f8", "parents": [ "e98c72c94205c59745f9ed0ac2837d5b83084a46" ], "author": { "name": "Lendacky, Thomas", "email": "Thomas.Lendacky@amd.com", "time": "Thu Nov 06 17:02:19 2014 -0600" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 19:00:15 2014 -0500" }, "message": "amd-xgbe: Check for complete packet on skb allocation error\n\nIf the skb allocation fails during receive processing, the driver would\ncontinue reading descriptors without first determining if there were\nany more descriptors for the current packet. Update the code to check\nwhether more descriptors are associated with the current packet or\nwhether to move on to the next descriptor as a new packet.\n\nSigned-off-by: Tom Lendacky \u003cthomas.lendacky@amd.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "e98c72c94205c59745f9ed0ac2837d5b83084a46", "tree": "b441cd6b39be96e098cdb7f6fc988de838de70e0", "parents": [ "6b798d70d07a38e9ba0a32316f76495f9dcfc343" ], "author": { "name": "Lendacky, Thomas", "email": "Thomas.Lendacky@amd.com", "time": "Thu Nov 06 17:02:13 2014 -0600" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 19:00:15 2014 -0500" }, "message": "amd-xgbe: Free channel/ring structures later\n\nThe channel structure is freed before freeing the per channel\ninterrupts resulting in a kernel oops. Move the call to free\nthe channel structure to after the freeing of the per channel\ninterrupts.\n\nSigned-off-by: Tom Lendacky \u003cthomas.lendacky@amd.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "9d01412ae76fec5274a3d94a28a3552a742a60dc", "tree": "7c18687df1974b08c44c313d62b9a0fd9a2e79cd", "parents": [ "f6b7734ba7b9c8c85c897c7b4b4e526bd5d7c978" ], "author": { "name": "Manish Chopra", "email": "manish.chopra@qlogic.com", "time": "Thu Nov 06 07:58:51 2014 -0500" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 16:43:29 2014 -0500" }, "message": "netxen: Fix link event handling.\n\no Poll for the link events only if firmware doesn\u0027t have capability\n to notify the driver for the link events.\n\nSigned-off-by: Manish Chopra \u003cmanish.chopra@qlogic.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "f6b7734ba7b9c8c85c897c7b4b4e526bd5d7c978", "tree": "0ff643e695c8cc48730b636631d4256730e44228", "parents": [ "44aa91ab2bb862540daa81403a1bc507496260fe" ], "author": { "name": "Govindarajulu Varadarajan", "email": "_govind@gmx.com", "time": "Thu Nov 06 15:21:39 2014 +0530" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 16:42:04 2014 -0500" }, "message": "enic: update desc properly in rx_copybreak\n\nWhen we reuse the rx buffer, we need to update the desc. If not hardware sees\nstale value.\n\nIn the following crash, when mtu is changed, hardware sees old rx buffer value\nand crashes on skb_put.\n\nFix this by using enic_queue_rq_desc helper function which updates the necessary\ndesc.\n\n[ 64.657376] skbuff: skb_over_panic: text:ffffffffa041f55d len:9010 put:9010 head:ffff8800d3ca9fc0 data:ffff8800d3caa000 tail:0x2372 end:0x640 dev:enp0s3\n[ 64.659965] ------------[ cut here ]------------\n[ 64.661322] kernel BUG at net/core/skbuff.c:100!\n[ 64.662644] invalid opcode: 0000 [#1] PREEMPT SMP\n[ 64.664001] Modules linked in: rpcsec_gss_krb5 auth_rpcgss oid_registry nfsv4 cirrus ttm drm_kms_helper drm enic psmouse microcode evdev serio_raw syscopyarea sysfillrect sysimgblt i2c_piix4 i2c_core pcspkr nfs lockd grace sunrpc fscache ext4 crc16 mbcache jbd2 sd_mod ata_generic virtio_balloon ata_piix libata uhci_hcd virtio_pci virtio_ring usbcore usb_common virtio scsi_mod\n[ 64.664834] CPU: 0 PID: 0 Comm: swapper/0 Tainted: G W 3.17.0-netnext-10335-g942396b-dirty #273\n[ 64.664834] Hardware name: Bochs Bochs, BIOS Bochs 01/01/2011\n[ 64.664834] task: ffffffff81a1d580 ti: ffffffff81a00000 task.ti: ffffffff81a00000\n[ 64.664834] RIP: 0010:[\u003cffffffff81392cf1\u003e] [\u003cffffffff81392cf1\u003e] skb_panic+0x61/0x70\n[ 64.664834] RSP: 0018:ffff880210603d48 EFLAGS: 00010292\n[ 64.664834] RAX: 000000000000008c RBX: ffff88020b0f6930 RCX: 0000000000000000\n[ 64.664834] RDX: 000000000000008c RSI: ffffffff8178b288 RDI: 00000000ffffffff\n[ 64.664834] RBP: ffff880210603d68 R08: 0000000000000001 R09: 0000000000000001\n[ 64.664834] R10: 00000000000005ce R11: 0000000000000001 R12: ffff88020b1f0b40\n[ 64.664834] R13: 000000000000a332 R14: ffff880209a1a000 R15: 0000000000000001\n[ 64.664834] FS: 0000000000000000(0000) GS:ffff880210600000(0000) knlGS:0000000000000000\n[ 64.664834] CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b\n[ 64.664834] CR2: 00007f6752935e48 CR3: 0000000035743000 CR4: 00000000000006f0\n[ 64.664834] Stack:\n[ 64.664834] ffff8800d3caa000 0000000000002372 0000000000000640 ffff88020b1f0000\n[ 64.664834] ffff880210603d78 ffffffff81392d54 ffff880210603e08 ffffffffa041f55d\n[ 64.664834] 0000000000000296 ffffffff00000000 00008e7e00008e7e ffff880200002332\n[ 64.664834] Call Trace:\n[ 64.664834] \u003cIRQ\u003e\n[ 64.664834]\n[ 64.664834] [\u003cffffffff81392d54\u003e] skb_put+0x54/0x60\n[ 64.664834] [\u003cffffffffa041f55d\u003e] enic_rq_service.constprop.47+0x3ad/0x730 [enic]\n[ 64.664834] [\u003cffffffffa041fa79\u003e] enic_poll_msix_rq+0x199/0x370 [enic]\n[ 64.664834] [\u003cffffffff813a5499\u003e] net_rx_action+0x139/0x210\n[ 64.664834] [\u003cffffffff81290db3\u003e] ? __this_cpu_preempt_check+0x13/0x20\n[ 64.664834] [\u003cffffffff8106110e\u003e] __do_softirq+0x14e/0x280\n[ 64.664834] [\u003cffffffff8106152e\u003e] irq_exit+0x8e/0xb0\n[ 64.664834] [\u003cffffffff8100fd21\u003e] do_IRQ+0x61/0x100\n[ 64.664834] [\u003cffffffff814a2bf2\u003e] common_interrupt+0x72/0x72\n\nfixes: a03bb56e67c357980dae886683733dab5583dc14 (\"enic: implement rx_copybreak\")\nSigned-off-by: Govindarajulu Varadarajan \u003c_govind@gmx.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "44aa91ab2bb862540daa81403a1bc507496260fe", "tree": "904965c075354ec137a1d46e08e2c4d7534ebe22", "parents": [ "c811906726ee28b638284b1965d728a9d3fe8d7d" ], "author": { "name": "Govindarajulu Varadarajan", "email": "_govind@gmx.com", "time": "Thu Nov 06 15:21:38 2014 +0530" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 16:42:04 2014 -0500" }, "message": "enic: handle error condition properly in enic_rq_indicate_buf\n\nIn case of error in rx path, we free the buf-\u003eos_buf but we do not make it NULL.\nIn next iteration we use the skb which is already freed. This causes the\nfollowing crash.\n\n[ 886.154772] general protection fault: 0000 [#1] PREEMPT SMP\n[ 886.154851] Modules linked in: rpcsec_gss_krb5 auth_rpcgss oid_registry nfsv4 microcode evdev cirrus ttm drm_kms_helper drm enic syscopyarea sysfillrect sysimgblt psmouse i2c_piix4 serio_raw pcspkr i2c_core nfs lockd grace sunrpc fscache ext4 crc16 mbcache jbd2 sd_mod crc_t10dif crct10dif_common ata_generic ata_piix virtio_balloon libata scsi_mod uhci_hcd usbcore virtio_pci virtio_ring virtio usb_common\n[ 886.155199] CPU: 0 PID: 0 Comm: swapper/0 Tainted: G W 3.17.0-netnext-05668-g876bc7f #272\n[ 886.155263] Hardware name: Bochs Bochs, BIOS Bochs 01/01/2011\n[ 886.155304] task: ffffffff81a1d580 ti: ffffffff81a00000 task.ti: ffffffff81a00000\n[ 886.155356] RIP: 0010:[\u003cffffffff81384030\u003e] [\u003cffffffff81384030\u003e] kfree_skb_list+0x10/0x30\n[ 886.155418] RSP: 0018:ffff880210603d48 EFLAGS: 00010206\n[ 886.155456] RAX: 0000000000000020 RBX: 0000000000000000 RCX: 0000000000000000\n[ 886.155504] RDX: 0000000000000000 RSI: 0000000000000001 RDI: 004500084e000017\n[ 886.155553] RBP: ffff880210603d50 R08: 00000000fe13d1b6 R09: 0000000000000001\n[ 886.155601] R10: 0000000000000000 R11: 0000000000000000 R12: ffff880209ff2f00\n[ 886.155650] R13: ffff88020ac0fe40 R14: ffff880209ff2f00 R15: ffff8800da8e3a80\n[ 886.155699] FS: 0000000000000000(0000) GS:ffff880210600000(0000) knlGS:0000000000000000\n[ 886.155774] CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b\n[ 886.155814] CR2: 00007f0e0c925000 CR3: 0000000035e8b000 CR4: 00000000000006f0\n[ 886.155865] Stack:\n[ 886.155882] 0000000000000000 ffff880210603d78 ffffffff81383f79 ffff880209ff2f00\n[ 886.155942] ffff88020b0c0b40 000000000000c000 ffff880210603d90 ffffffff81383faf\n[ 886.156001] ffff880209ff2f00 ffff880210603da8 ffffffff8138406d ffff88020b1b08c0\n[ 886.156061] Call Trace:\n[ 886.156080] \u003cIRQ\u003e\n[ 886.156095]\n[ 886.156112] [\u003cffffffff81383f79\u003e] skb_release_data+0xa9/0xc0\n[ 886.157656] [\u003cffffffff81383faf\u003e] skb_release_all+0x1f/0x30\n[ 886.159195] [\u003cffffffff8138406d\u003e] consume_skb+0x1d/0x40\n[ 886.160719] [\u003cffffffff813942e5\u003e] __dev_kfree_skb_any+0x35/0x40\n[ 886.162224] [\u003cffffffffa02dc1d5\u003e] enic_rq_service.constprop.47+0xe5/0x5a0 [enic]\n[ 886.163756] [\u003cffffffffa02dc829\u003e] enic_poll_msix_rq+0x199/0x370 [enic]\n[ 886.164730] [\u003cffffffff81397e29\u003e] net_rx_action+0x139/0x210\n[ 886.164730] [\u003cffffffff8105fb2e\u003e] __do_softirq+0x14e/0x280\n[ 886.164730] [\u003cffffffff8105ff2e\u003e] irq_exit+0x8e/0xb0\n[ 886.164730] [\u003cffffffff8100fc1d\u003e] do_IRQ+0x5d/0x100\n[ 886.164730] [\u003cffffffff81496832\u003e] common_interrupt+0x72/0x72\n\nfixes: a03bb56e67c357980dae886683733dab5583dc14 (\"enic: implement rx_copybreak\")\nSigned-off-by: Govindarajulu Varadarajan \u003c_govind@gmx.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "c811906726ee28b638284b1965d728a9d3fe8d7d", "tree": "7662a703252362ffbca459f891a34f73e6140c8c", "parents": [ "b31f65fb4383a49bdcfa465176754b37e44e1e17", "364d1798efdf13c7f2b9d902228adf8e84f1d963" ], "author": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 16:40:47 2014 -0500" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 16:40:47 2014 -0500" }, "message": "Merge branch \u0027mlx5-net\u0027\n\nEli Cohen says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\nmlx5_core fixes for 3.18\n\nthe following two patches fix races to could lead to kernel panic in some cases.\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "364d1798efdf13c7f2b9d902228adf8e84f1d963", "tree": "7662a703252362ffbca459f891a34f73e6140c8c", "parents": [ "a158906dd7d4379e85ec371a14edfe1ce5f2318d" ], "author": { "name": "Eli Cohen", "email": "eli@dev.mellanox.co.il", "time": "Thu Nov 06 12:51:22 2014 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 16:40:36 2014 -0500" }, "message": "net/mlx5_core: Fix race on driver load\n\nWhen events arrive at driver load, the event handler gets called even before\nthe spinlock and list are initialized. Fix this by moving the initialization\nbefore EQs creation.\n\nSigned-off-by: Eli Cohen \u003celi@mellanox.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "a158906dd7d4379e85ec371a14edfe1ce5f2318d", "tree": "435962b39e77df458649a115deabe4af45b85ece", "parents": [ "b31f65fb4383a49bdcfa465176754b37e44e1e17" ], "author": { "name": "Eli Cohen", "email": "eli@dev.mellanox.co.il", "time": "Thu Nov 06 12:51:21 2014 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 16:40:35 2014 -0500" }, "message": "net/mlx5_core: Fix race in create EQ\n\nAfter the EQ is created, it can possibly generate interrupts and the interrupt\nhandler is referencing eq-\u003edev. It is therefore required to set eq-\u003edev before\ncalling request_irq() so if an event is generated before request_irq() returns,\nwe will have a valid eq-\u003edev field.\n\nSigned-off-by: Eli Cohen \u003celi@mellanox.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "6b798d70d07a38e9ba0a32316f76495f9dcfc343", "tree": "f1c6e7acc61511454ab27e8ecdb00dfdcc8766e8", "parents": [ "271d70f4b7f2990c84138be267f8e42c33506cc7", "a85311bf1f9f8185682990cafdd4e0572c0ed373" ], "author": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 16:33:13 2014 -0500" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 16:33:13 2014 -0500" }, "message": "Merge branch \u0027net_next_ovs\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/pshelar/openvswitch\n\nPravin B Shelar says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\nOpen vSwitch\n\nFirst two patches are related to OVS MPLS support. Rest of patches\nare mostly refactoring and minor improvements to openvswitch.\n\nv1-v2:\n - Fix conflicts due to \"gue: Remote checksum offload\"\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "271d70f4b7f2990c84138be267f8e42c33506cc7", "tree": "e943ec597065f6d33713272d4e5805f3de7c052f", "parents": [ "4e865a5a997005a93dc0833683248dd185ea2c58", "8c4ee3e70646d6bfdda55afe84016b914c7fdd0e" ], "author": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 15:16:44 2014 -0500" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 15:16:44 2014 -0500" }, "message": "Merge branch \u0027sunvnet-next\u0027\n\nSowmini Varadhan says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\nsunvnet: bug fixes\n\nThis patch series has a coding-style fix and a bug fix.\n\nThe coding style fix (patch 1) is the extra indentation flagged by\nBen Hutchings:\n http://marc.info/?l\u003dlinux-netdev\u0026m\u003d141529243409594\u0026w\u003d2\n\nThe bugfix (patch 2) is the following:\nwhen vnet_event_napi() is called as part of napi_resume\n(i.e., continuation of a previous NAPI read that was truncated\ndue to budget constraints), and then finds no more packets to read,\nthe code was trying to avoid an additional trip through ldc_rx\nas an optimization. However, when this corner case happens, we would\nneed to reset a number of dring state bits such as rcv_nxt carefully,\nwhich quickly becomes complex and hacky. The cleaner solution\nis to just roll back to vnet_poll, re-enable interrupts and set up\ndring state as was done in the pre-NAPI version of the driver.\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "8c4ee3e70646d6bfdda55afe84016b914c7fdd0e", "tree": "e943ec597065f6d33713272d4e5805f3de7c052f", "parents": [ "6c3ce8a30c1bafa8952461fb76e66e2da338b7ed" ], "author": { "name": "Sowmini Varadhan", "email": "sowmini.varadhan@oracle.com", "time": "Thu Nov 06 14:51:08 2014 -0500" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 15:16:30 2014 -0500" }, "message": "sunvnet: Return from vnet_napi_event() if no packets to read\n\nvnet_event_napi() may be called as part of the NAPI -\u003epoll,\nto resume reading descriptor rings. When no data is available,\ndescriptor ring state (e.g., rcv_nxt) needs to be reset\ncarefully to stay in lock-step with ldc_read(). In the interest\nof simplicity, the best way to do this is to return from\nvnet_event_napi() when there are no more packets to read.\nThe next trip through ldc_rx will correctly set up the dring state.\n\nSigned-off-by: Sowmini Varadhan \u003csowmini.varadhan@oracle.com\u003e\nTested-by: David Stevens \u003cdavid.stevens@oracle.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "6c3ce8a30c1bafa8952461fb76e66e2da338b7ed", "tree": "68b0c2434841ed0d232dcd2ce0ca8d29820a7652", "parents": [ "4e865a5a997005a93dc0833683248dd185ea2c58" ], "author": { "name": "Sowmini Varadhan", "email": "sowmini.varadhan@oracle.com", "time": "Thu Nov 06 14:51:02 2014 -0500" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 15:16:30 2014 -0500" }, "message": "sunvnet: Fix indentation in maybe_tx_wakeup()\n\nremove redundant tab.\n\nSigned-off-by: Sowmini Varadhan \u003csowmini.varadhan@oracle.com\u003e\nReported-by: Ben Hutchings \u003cben@decadent.org.uk\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "4e865a5a997005a93dc0833683248dd185ea2c58", "tree": "94252a1941445ef9c74800a898d2cf947bbc2ca1", "parents": [ "926c512685ddd8f26f1c789218391530ccd54a35", "662412d14bfa6a672626e4470cab73b75c8b42f0" ], "author": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 15:14:36 2014 -0500" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 15:14:36 2014 -0500" }, "message": "Merge branch \u0027r8152-next\u0027\n\nHayes Wang says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\nr8152: rtl_ops_init modify\n\nInitialize the ops through tp-\u003eversion. This could skip checking\neach VID/PID.\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "662412d14bfa6a672626e4470cab73b75c8b42f0", "tree": "94252a1941445ef9c74800a898d2cf947bbc2ca1", "parents": [ "55b6547547f9a6719b62fb07d656863142e028a8" ], "author": { "name": "hayeswang", "email": "hayeswang@realtek.com", "time": "Thu Nov 06 12:47:40 2014 +0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 15:14:31 2014 -0500" }, "message": "r8152: remove the definitions of the PID\n\nThe PIDs are only used in the id table, so the definitions are\nunnacessary. Remove them wouldn\u0027t have confusion.\n\nSigned-off-by: Hayes Wang \u003chayeswang@realtek.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "55b6547547f9a6719b62fb07d656863142e028a8", "tree": "537abd3301e6da684eb44d869c9f78fb0a738fd9", "parents": [ "82cf94cbcf95356ba9ba8c2533291a7792bbe5e7" ], "author": { "name": "hayeswang", "email": "hayeswang@realtek.com", "time": "Thu Nov 06 12:47:39 2014 +0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 15:14:30 2014 -0500" }, "message": "r8152: modify rtl_ops_init\n\nReplace using VID/PID with using tp-\u003eversion to initialize the ops.\n\nSigned-off-by: Hayes Wang \u003chayeswang@realtek.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "82cf94cbcf95356ba9ba8c2533291a7792bbe5e7", "tree": "54ae2d254a805232fbb57bd491737fb7d6d68491", "parents": [ "926c512685ddd8f26f1c789218391530ccd54a35" ], "author": { "name": "hayeswang", "email": "hayeswang@realtek.com", "time": "Thu Nov 06 12:47:38 2014 +0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 15:14:30 2014 -0500" }, "message": "r8152: move r8152b_get_version\n\nMove r8152b_get_version() to the location before rtl_ops_init().\nThen, the rtl_ops_init() could use tp-\u003eversion.\n\nSigned-off-by: Hayes Wang \u003chayeswang@realtek.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "926c512685ddd8f26f1c789218391530ccd54a35", "tree": "01d9d6821c95e47080a8126dfb3316996793365c", "parents": [ "450834977796cc74d1265d7dfe69cf6767537dfc" ], "author": { "name": "Joe Perches", "email": "joe@perches.com", "time": "Wed Nov 05 15:42:09 2014 -0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 15:11:11 2014 -0500" }, "message": "sock.h: Remove unused NETDEBUG macro\n\nIt\u0027s unused now, just delete it.\n\nSigned-off-by: Joe Perches \u003cjoe@perches.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "450834977796cc74d1265d7dfe69cf6767537dfc", "tree": "e0c644e9978c7030c4109759d4b33e38fb002fb8", "parents": [ "cbffccc970394f82e397fccbeb136eeaffb3c552" ], "author": { "name": "Joe Perches", "email": "joe@perches.com", "time": "Wed Nov 05 15:36:08 2014 -0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 15:11:10 2014 -0500" }, "message": "net: esp: Convert NETDEBUG to pr_info\n\nCommit 64ce207306de (\"[NET]: Make NETDEBUG pure printk wrappers\")\noriginally had these NETDEBUG printks as always emitting.\n\nCommit a2a316fd068c (\"[NET]: Replace CONFIG_NET_DEBUG with sysctl\")\nadded a net_msg_warn sysctl to these NETDEBUG uses.\n\nConvert these NETDEBUG uses to normal pr_info calls.\n\nThis changes the output prefix from \"ESP: \" to include\n\"IPSec: \" for the ipv4 case and \"IPv6: \" for the ipv6 case.\n\nThese output lines are now like the other messages in the files.\n\nOther miscellanea:\n\nNeaten the arithmetic spacing to be consistent with other\narithmetic spacing in the files.\n\nSigned-off-by: Joe Perches \u003cjoe@perches.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "cbffccc970394f82e397fccbeb136eeaffb3c552", "tree": "de1fdbcb75ec938df78a49921531c694839ff080", "parents": [ "f03ae5f9c8560f0a8dcad8ae6ea18efc981d8e35" ], "author": { "name": "Joe Perches", "email": "joe@perches.com", "time": "Wed Nov 05 14:39:21 2014 -0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 15:11:10 2014 -0500" }, "message": "net; ipv[46] - Remove 2 unnecessary NETDEBUG OOM messages\n\nThese messages aren\u0027t useful as there\u0027s a generic dump_stack()\non OOM.\n\nNeaten the comment and if test above the OOM by separating the\nassign in if into an allocation then if test.\n\nSigned-off-by: Joe Perches \u003cjoe@perches.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "f03ae5f9c8560f0a8dcad8ae6ea18efc981d8e35", "tree": "820a40a7e80d70fc976f3f10039e35f417bff64e", "parents": [ "0c6965dd317185cef3d706e5785bb032f819faa1" ], "author": { "name": "Andrew Lunn", "email": "andrew@lunn.ch", "time": "Wed Nov 05 20:01:59 2014 +0100" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 15:08:05 2014 -0500" }, "message": "dsa: mv88e6171: Add support for mv88e6172\n\nThe mv88e6172 is very similar to the mv88e6171. So extend the\nmv88e6171 driver to support it.\n\nSigned-off-by: Andrew Lunn \u003candrew@lunn.ch\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "b31f65fb4383a49bdcfa465176754b37e44e1e17", "tree": "a7885358f08a9e2b714aca0386f91ca825ce266d", "parents": [ "5816c3dafb6c63fd5c7b9f3f707c8565811d9916" ], "author": { "name": "Andrew Lunn", "email": "andrew@lunn.ch", "time": "Wed Nov 05 19:47:28 2014 +0100" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 15:06:28 2014 -0500" }, "message": "net: dsa: slave: Fix autoneg for phys on switch MDIO bus\n\nWhen the ports phys are connected to the switches internal MDIO bus,\nwe need to connect the phy to the slave netdev, otherwise\nauto-negotiation etc, does not work.\n\nSigned-off-by: Andrew Lunn \u003candrew@lunn.ch\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "0c6965dd317185cef3d706e5785bb032f819faa1", "tree": "fb4750a2b62514f3aeaf400e9c19697f6799c888", "parents": [ "f81c6906dcfb5c16c50a7ae5ecff60294c443f65" ], "author": { "name": "Jiri Pirko", "email": "jiri@resnulli.us", "time": "Wed Nov 05 20:51:51 2014 +0100" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 15:04:41 2014 -0500" }, "message": "sched: fix act file names in header comment\n\nFixes: 4bba3925 (\"[PKT_SCHED]: Prefix tc actions with act_\")\nSigned-off-by: Jiri Pirko \u003cjiri@resnulli.us\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "5816c3dafb6c63fd5c7b9f3f707c8565811d9916", "tree": "5c505476604a3bead127e1dc18d76d24210a8b1a", "parents": [ "4484d0524e86dcb2fb5ad0805b09d3e74559ebd5" ], "author": { "name": "Ryo Munakata", "email": "ryomnktml@gmail.com", "time": "Wed Nov 05 23:45:58 2014 +0900" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 14:59:19 2014 -0500" }, "message": "net/9p: remove a comment about pref member which doesn\u0027t exist\n\nSigned-off-by: Ryo Munakata \u003cryomnktml@gmail.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "4484d0524e86dcb2fb5ad0805b09d3e74559ebd5", "tree": "f492c0bd8d4c53b8886c68304833caea981d0255", "parents": [ "2c2a9cbd64387d6b70ac5db013e9bfe9412c7354" ], "author": { "name": "Mugunthan V N", "email": "mugunthanvnm@ti.com", "time": "Wed Nov 05 18:33:31 2014 +0530" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 14:58:04 2014 -0500" }, "message": "drivers: net: cpsw: remove cpsw_ale_stop from cpsw_ale_destroy\n\nwhen cpsw is build as modulea and simple insert and removal of module\ncreates a deadlock, due to delete timer. the timer is created and destroyed\nin cpsw_ale_start and cpsw_ale_stop which are from device open and close.\n\nroot@am437x-evm:~# modprobe -r ti_cpsw\n[ 158.505333] INFO: trying to register non-static key.\n[ 158.510623] the code is fine but needs lockdep annotation.\n[ 158.516448] turning off the locking correctness validator.\n[ 158.522282] CPU: 0 PID: 1339 Comm: modprobe Not tainted 3.14.23-00445-gd41c88f #44\n[ 158.530359] [\u003cc0015380\u003e] (unwind_backtrace) from [\u003cc0012088\u003e] (show_stack+0x10/0x14)\n[ 158.538603] [\u003cc0012088\u003e] (show_stack) from [\u003cc054ad70\u003e] (dump_stack+0x78/0x94)\n[ 158.546295] [\u003cc054ad70\u003e] (dump_stack) from [\u003cc0088008\u003e] (__lock_acquire+0x176c/0x1b74)\n[ 158.554711] [\u003cc0088008\u003e] (__lock_acquire) from [\u003cc0088944\u003e] (lock_acquire+0x9c/0x104)\n[ 158.563043] [\u003cc0088944\u003e] (lock_acquire) from [\u003cc004e520\u003e] (del_timer_sync+0x44/0xd8)\n[ 158.571289] [\u003cc004e520\u003e] (del_timer_sync) from [\u003cbf2eac1c\u003e] (cpsw_ale_destroy+0x10/0x3c [ti_cpsw])\n[ 158.580821] [\u003cbf2eac1c\u003e] (cpsw_ale_destroy [ti_cpsw]) from [\u003cbf2eb268\u003e] (cpsw_remove+0x30/0xa0 [ti_cpsw])\n[ 158.591000] [\u003cbf2eb268\u003e] (cpsw_remove [ti_cpsw]) from [\u003cc035ef44\u003e] (platform_drv_remove+0x18/0x1c)\n[ 158.600527] [\u003cc035ef44\u003e] (platform_drv_remove) from [\u003cc035d8bc\u003e] (__device_release_driver+0x70/0xc8)\n[ 158.610236] [\u003cc035d8bc\u003e] (__device_release_driver) from [\u003cc035e0d4\u003e] (driver_detach+0xb4/0xb8)\n[ 158.619386] [\u003cc035e0d4\u003e] (driver_detach) from [\u003cc035d6e4\u003e] (bus_remove_driver+0x4c/0x90)\n[ 158.627988] [\u003cc035d6e4\u003e] (bus_remove_driver) from [\u003cc00af2a8\u003e] (SyS_delete_module+0x10c/0x198)\n[ 158.637144] [\u003cc00af2a8\u003e] (SyS_delete_module) from [\u003cc000e580\u003e] (ret_fast_syscall+0x0/0x48)\n[ 179.524727] INFO: rcu_sched detected stalls on CPUs/tasks: {} (detected by 0, t\u003d2102 jiffies, g\u003d1487, c\u003d1486, q\u003d6)\n[ 179.535741] INFO: Stall ended before state dump start\n\nSigned-off-by: Mugunthan V N \u003cmugunthanvnm@ti.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "2c2a9cbd64387d6b70ac5db013e9bfe9412c7354", "tree": "24457fbbaab2016171b13aae6c2e105c082b8f2a", "parents": [ "66f1c44887ba4f47d617f8ae21cf8e04e1892bd7" ], "author": { "name": "Karl Beldan", "email": "karl.beldan@rivierawaves.com", "time": "Wed Nov 05 15:32:59 2014 +0100" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 14:54:39 2014 -0500" }, "message": "net: mv643xx_eth: reclaim TX skbs only when released by the HW\n\nATM, txq_reclaim will dequeue and free an skb for each tx desc released\nby the hw that has TX_LAST_DESC set. However, in case of TSO, each\nhw desc embedding the last part of a segment has TX_LAST_DESC set,\nlosing the one-to-one \u0027last skb frag\u0027/\u0027TX_LAST_DESC set\u0027 correspondance,\nwhich causes data corruption.\n\nFix this by checking TX_ENABLE_INTERRUPT instead of TX_LAST_DESC, and\nwarn when trying to dequeue from an empty txq (which can be symptomatic\nof releasing skbs prematurely).\n\nFixes: 3ae8f4e0b98 (\u0027net: mv643xx_eth: Implement software TSO\u0027)\nReported-by: Slawomir Gajzner \u003cslawomir.gajzner@gmail.com\u003e\nReported-by: Julien D\u0027Ascenzio \u003cjdascenzio@yahoo.fr\u003e\nSigned-off-by: Karl Beldan \u003ckarl.beldan@rivierawaves.com\u003e\nCc: Ian Campbell \u003cijc@hellion.org.uk\u003e\nCc: Eric Dumazet \u003ceric.dumazet@gmail.com\u003e\nCc: Ezequiel Garcia \u003cezequiel.garcia@free-electrons.com\u003e\nCc: Sebastian Hesselbarth \u003csebastian.hesselbarth@gmail.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "f81c6906dcfb5c16c50a7ae5ecff60294c443f65", "tree": "8759bd88647af178390ac8c16e03b9a6f794c762", "parents": [ "7e5d7753956b374516530e156c5e8aa19652398d", "d98a4ffe0d7a8946c67d115f8b3bd8e5bb1df86d" ], "author": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 14:43:13 2014 -0500" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 14:43:13 2014 -0500" }, "message": "Merge branch \u0027sfc-next\u0027\n\nShradha Shah says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\nsfc: Clean up Siena SR-IOV support in preparation for EF10 SR-IOV support\n\nThis patch series provides a base and clean up for the upcoming\nEF10 SRIOV patches.\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "d98a4ffe0d7a8946c67d115f8b3bd8e5bb1df86d", "tree": "8759bd88647af178390ac8c16e03b9a6f794c762", "parents": [ "327c685eb745e3c56ed52734d3671dcf6f4f6940" ], "author": { "name": "Shradha Shah", "email": "sshah@solarflare.com", "time": "Wed Nov 05 12:16:46 2014 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 14:43:09 2014 -0500" }, "message": "sfc: Add NIC type operations to replace direct calls from efx.c into siena_sriov.c\n\nAlso add dummy functions where required to avoid NULL pointer dereference.\n\nSigned-off-by: Shradha Shah \u003csshah@solarflare.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "327c685eb745e3c56ed52734d3671dcf6f4f6940", "tree": "0a0234e53107df72e6e2048d3d026879ca693988", "parents": [ "2dc313eca30043978d749c1d37e3c62f3307575d" ], "author": { "name": "Shradha Shah", "email": "sshah@solarflare.com", "time": "Wed Nov 05 12:16:32 2014 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 14:43:08 2014 -0500" }, "message": "sfc: Rename implementations in siena_sriov.c to have a \u0027siena\u0027 prefix\n\nPatch in preparation for the upcoming EF10 sriov support.\n\nSigned-off-by: Shradha Shah \u003csshah@solarflare.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "2dc313eca30043978d749c1d37e3c62f3307575d", "tree": "a64d1082fc45a394d5e2d670a19d01456772e9cd", "parents": [ "7e5d7753956b374516530e156c5e8aa19652398d" ], "author": { "name": "Shradha Shah", "email": "sshah@solarflare.com", "time": "Wed Nov 05 12:16:18 2014 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 14:43:08 2014 -0500" }, "message": "sfc: Move the current VF state from efx_nic into siena_nic_data\n\nThis patch series provides a base and cleanup for the\nupcoming EF10 SRIOV support.\n\nThis patch moves the VF state into siena_nic_data as a basis to\nsave the VF state based on nic type.\n\nSigned-off-by: Shradha Shah \u003csshah@solarflare.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "7e5d7753956b374516530e156c5e8aa19652398d", "tree": "536746bc7043795210b9dd9fd4588dfcf2c37f5f", "parents": [ "17007aa53a1d870d57fdc6a316a3315eabb10f2a" ], "author": { "name": "Malcolm Crossley", "email": "malcolm.crossley@citrix.com", "time": "Wed Nov 05 10:50:22 2014 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 14:40:18 2014 -0500" }, "message": "xen-netback: remove unconditional __pskb_pull_tail() in guest Tx path\n\nUnconditionally pulling 128 bytes into the linear area is not required\nfor:\n\n- security: Every protocol demux starts with pskb_may_pull() to pull\n frag data into the linear area, if necessary, before looking at\n headers.\n\n- performance: Netback has already grant copied up-to 128 bytes from\n the first slot of a packet into the linear area. The first slot\n normally contain all the IPv4/IPv6 and TCP/UDP headers.\n\nThe unconditional pull would often copy frag data unnecessarily. This\nis a performance problem when running on a version of Xen where grant\nunmap avoids TLB flushes for pages which are not accessed. TLB\nflushes can now be avoided for \u003e 99% of unmaps (it was 0% before).\n\nGrant unmap TLB flush avoidance will be available in a future version\nof Xen (probably 4.6).\n\nSigned-off-by: Malcolm Crossley \u003cmalcolm.crossley@citrix.com\u003e\nSigned-off-by: David Vrabel \u003cdavid.vrabel@citrix.com\u003e\nAcked-by: Ian Campbell \u003cian.campbell@citrix.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "17007aa53a1d870d57fdc6a316a3315eabb10f2a", "tree": "727fa19f7954de4a4fcd4b44278424166004144b", "parents": [ "915af65619244eaf0644aad22e9299c7da303a8e", "7627fc074b9f7d2b0833f4f83c6b8e432d9f303e" ], "author": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 14:39:01 2014 -0500" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 14:39:01 2014 -0500" }, "message": "Merge branch \u0027stmmac-next\u0027\n\nAndy Shevchenko says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\nstmmac: pci: various cleanups and fixes\n\nThere are few cleanups and fixes regarding to stmmac PCI driver.\nThis has been tested on Intel Galileo board with recent net-next tree.\n\nSince v2:\n- drop patch 5/5 since it will be part of a big change across entire subsystem\n\nSince v1:\n- remove already applied patch\n- append patch 1/5\n- rework patch 3/5 to be functional compatible with original code\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "7627fc074b9f7d2b0833f4f83c6b8e432d9f303e", "tree": "727fa19f7954de4a4fcd4b44278424166004144b", "parents": [ "2a3e8e93bd0d63df279c63daf8d29fe661c0bea9" ], "author": { "name": "Andy Shevchenko", "email": "andriy.shevchenko@linux.intel.com", "time": "Wed Nov 05 12:27:29 2014 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 14:38:58 2014 -0500" }, "message": "stmmac: pci: convert to use dev_* macros\n\nInstead of pr_* macros let\u0027s use dev_* macros which provide device name.\n\nSigned-off-by: Andy Shevchenko \u003candriy.shevchenko@linux.intel.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "2a3e8e93bd0d63df279c63daf8d29fe661c0bea9", "tree": "6e61b2c73e8c551b40717d6439e790705ea5beb9", "parents": [ "3be3d81b629d5345664dbb07efeb5160e8f0b19f" ], "author": { "name": "Andy Shevchenko", "email": "andriy.shevchenko@linux.intel.com", "time": "Wed Nov 05 12:27:28 2014 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 14:38:57 2014 -0500" }, "message": "stmmac: pci: use managed resources\n\nMigrate pci driver to managed resources to reduce boilerplate error handling\ncode.\n\nSigned-off-by: Andy Shevchenko \u003candriy.shevchenko@linux.intel.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "3be3d81b629d5345664dbb07efeb5160e8f0b19f", "tree": "5c8898367018e4f386c004caee441d47e47057b2", "parents": [ "295f9d0bc3975e9885c73bfd82cdff9b739d2001" ], "author": { "name": "Andy Shevchenko", "email": "andriy.shevchenko@linux.intel.com", "time": "Wed Nov 05 12:27:27 2014 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 14:38:57 2014 -0500" }, "message": "stmmac: pci: convert to use dev_pm_ops\n\nConvert system PM callbacks to use dev_pm_ops. In addition remove the PCI calls\nrelated to a power state since the bus code cares about this already.\n\nSigned-off-by: Andy Shevchenko \u003candriy.shevchenko@linux.intel.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "295f9d0bc3975e9885c73bfd82cdff9b739d2001", "tree": "75daf996e5bc0fe936844ed68d10333b1423f1ad", "parents": [ "915af65619244eaf0644aad22e9299c7da303a8e" ], "author": { "name": "Andy Shevchenko", "email": "andriy.shevchenko@linux.intel.com", "time": "Wed Nov 05 12:27:26 2014 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 14:38:57 2014 -0500" }, "message": "stmmac: pci: use defined constant instead of magic number\n\nThe last standard PCI resource is defined as PCI_STD_RESOURCE_END. Thus, we\ncould use it instead of plain integer.\n\nSigned-off-by: Andy Shevchenko \u003candriy.shevchenko@linux.intel.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "915af65619244eaf0644aad22e9299c7da303a8e", "tree": "f43dee27a1ee6ec4721325a99263085385f4b193", "parents": [ "ea3dc9601bda69d8d695b57c4f7a997cd7039781" ], "author": { "name": "Andy Shevchenko", "email": "andriy.shevchenko@linux.intel.com", "time": "Wed Nov 05 11:45:32 2014 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 14:35:11 2014 -0500" }, "message": "stmmac: fix sparse warnings\n\nThis patch fixes the following sparse warnings.\n\ndrivers/net/ethernet/stmicro/stmmac/enh_desc.c:381:30: warning: symbol \u0027enh_desc_ops\u0027 was not declared. Should it be static?\ndrivers/net/ethernet/stmicro/stmmac/norm_desc.c:253:30: warning: symbol \u0027ndesc_ops\u0027 was not declared. Should it be static?\ndrivers/net/ethernet/stmicro/stmmac/stmmac_hwtstamp.c:141:33: warning: symbol \u0027stmmac_ptp\u0027 was not declared. Should it be static?\n\nThere is no functional change.\n\nSigned-off-by: Andy Shevchenko \u003candriy.shevchenko@linux.intel.com\u003e\nAcked-by: Giuseppe CAVALLARO \u003cpeppe.cavallaro@st.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "ea3dc9601bda69d8d695b57c4f7a997cd7039781", "tree": "cebdab7c6b281e52d54a96e30738539825448f92", "parents": [ "d50051407f136028108cfda068d55ef053a54fe1" ], "author": { "name": "Steffen Klassert", "email": "steffen.klassert@secunet.com", "time": "Wed Nov 05 08:03:50 2014 +0100" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 14:19:20 2014 -0500" }, "message": "ip6_tunnel: Add support for wildcard tunnel endpoints.\n\nThis patch adds support for tunnels with local or\nremote wildcard endpoints. With this we get a\nNBMA tunnel mode like we have it for ipv4 and\nsit tunnels.\n\nSigned-off-by: Steffen Klassert \u003csteffen.klassert@secunet.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "d50051407f136028108cfda068d55ef053a54fe1", "tree": "012d164d745342f453d5fed4619884839f51fbc0", "parents": [ "e1b2cb655060e081e73b384b1fc8b2e978f73467" ], "author": { "name": "Steffen Klassert", "email": "steffen.klassert@secunet.com", "time": "Wed Nov 05 08:02:48 2014 +0100" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 06 14:19:19 2014 -0500" }, "message": "ipv6: Allow sending packets through tunnels with wildcard endpoints\n\nCurrently we need the IP6_TNL_F_CAP_XMIT capabiltiy to transmit\npackets through an ipv6 tunnel. This capability is set when the\ntunnel gets configured, based on the tunnel endpoint addresses.\n\nOn tunnels with wildcard tunnel endpoints, we need to do the\ncapabiltiy checking on a per packet basis like it is done in\nthe receive path.\n\nThis patch extends ip6_tnl_xmit_ctl() to take local and remote\naddresses as parameters to allow for per packet capabiltiy\nchecking.\n\nSigned-off-by: Steffen Klassert \u003csteffen.klassert@secunet.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "a85311bf1f9f8185682990cafdd4e0572c0ed373", "tree": "ff8b3d0e0d3a7da142d371ade5a5bb9c065cd07b", "parents": [ "2fdb957d634a906ae8939bff23d45968307acbf7" ], "author": { "name": "Pravin B Shelar", "email": "pshelar@nicira.com", "time": "Sun Oct 19 12:03:40 2014 -0700" }, "committer": { "name": "Pravin B Shelar", "email": "pshelar@nicira.com", "time": "Wed Nov 05 23:52:35 2014 -0800" }, "message": "openvswitch: Avoid NULL mask check while building mask\n\nOVS does mask validation even if it does not need to convert\nnetlink mask attributes to mask structure. ovs_nla_get_match()\ncaller can pass NULL mask structure pointer if the caller does\nnot need mask. Therefore NULL check is required in SW_FLOW_KEY*\nmacros. Following patch does not convert mask netlink attributes\nif mask pointer is NULL, so we do not need these checks in\nSW_FLOW_KEY* macro.\n\nSigned-off-by: Pravin B Shelar \u003cpshelar@nicira.com\u003e\nAcked-by: Daniele Di Proietto \u003cddiproietto@vmware.com\u003e\nAcked-by: Andy Zhou \u003cazhou@nicira.com\u003e\n" }, { "commit": "2fdb957d634a906ae8939bff23d45968307acbf7", "tree": "98eee0e868f2b6ff95d715d7a0cdac169d3357d4", "parents": [ "41af73e9c17d5fb549fced2be97faeb4b8606fb8" ], "author": { "name": "Pravin B Shelar", "email": "pshelar@nicira.com", "time": "Sun Oct 19 11:19:51 2014 -0700" }, "committer": { "name": "Pravin B Shelar", "email": "pshelar@nicira.com", "time": "Wed Nov 05 23:52:35 2014 -0800" }, "message": "openvswitch: Refactor action alloc and copy api.\n\nThere are two separate API to allocate and copy actions list. Anytime\nOVS needs to copy action list, it needs to call both functions.\nFollowing patch moves action allocation to copy function to avoid\ncode duplication.\n\nSigned-off-by: Pravin B Shelar \u003cpshelar@nicira.com\u003e\nAcked-by: Jarno Rajahalme \u003cjrajahalme@nicira.com\u003e\n" }, { "commit": "41af73e9c17d5fb549fced2be97faeb4b8606fb8", "tree": "f2e52b72eee266b9d947e58fb778330ddd023c79", "parents": [ "d98612b8c1150cb73ecd45e94c62de053f89441c" ], "author": { "name": "Joe Stringer", "email": "joestringer@nicira.com", "time": "Sat Oct 18 16:14:14 2014 -0700" }, "committer": { "name": "Pravin B Shelar", "email": "pshelar@nicira.com", "time": "Wed Nov 05 23:52:35 2014 -0800" }, "message": "openvswitch: Move key_attr_size() to flow_netlink.h.\n\nflow-netlink has netlink related code.\n\nSigned-off-by: Joe Stringer \u003cjoestringer@nicira.com\u003e\nSigned-off-by: Pravin B Shelar \u003cpshelar@nicira.com\u003e\n" }, { "commit": "d98612b8c1150cb73ecd45e94c62de053f89441c", "tree": "2ed1870ee81bb17f5d6bbfaea04ea207f75836fc", "parents": [ "1a4e96a0e989200f7180264e27b22e9a85f3fcc8" ], "author": { "name": "Lorand Jakab", "email": "lojakab@cisco.com", "time": "Mon Oct 06 05:45:32 2014 -0700" }, "committer": { "name": "Pravin B Shelar", "email": "pshelar@nicira.com", "time": "Wed Nov 05 23:52:35 2014 -0800" }, "message": "openvswitch: Remove flow member from struct ovs_skb_cb\n\nThe \u0027flow\u0027 memeber was chosen for removal because it\u0027s only used\nin ovs_execute_actions() we can pass it as argument to this\nfunction.\n\nSigned-off-by: Lorand Jakab \u003clojakab@cisco.com\u003e\nSigned-off-by: Pravin B Shelar \u003cpshelar@nicira.com\u003e\n" }, { "commit": "1a4e96a0e989200f7180264e27b22e9a85f3fcc8", "tree": "d26b78452e2919e347486bbd25452a9248cfc936", "parents": [ "e1f9c356d2946ec717d035ccac5532b1c5709c52" ], "author": { "name": "Jarno Rajahalme", "email": "jrajahalme@nicira.com", "time": "Tue Sep 30 10:52:32 2014 -0700" }, "committer": { "name": "Pravin B Shelar", "email": "pshelar@nicira.com", "time": "Wed Nov 05 23:52:35 2014 -0800" }, "message": "openvswitch: Fix the type of struct ovs_key_nd nd_target field.\n\nShould be the same as other IPv6 address fields.\n\nCurrent master produces sparse warnings without this change.\n\nSigned-off-by: Jarno Rajahalme \u003cjrajahalme@nicira.com\u003e\nSigned-off-by: Pravin B Shelar \u003cpshelar@nicira.com\u003e\n" }, { "commit": "e1f9c356d2946ec717d035ccac5532b1c5709c52", "tree": "a006b187186abc7a7d0ebf3c56ff6b1e6d64febd", "parents": [ "cc3a5ae6f23336ddc1d136a66c0a6a60276e99a3" ], "author": { "name": "Chunhe Li", "email": "lichunhe@huawei.com", "time": "Mon Sep 08 13:17:21 2014 -0700" }, "committer": { "name": "Pravin B Shelar", "email": "pshelar@nicira.com", "time": "Wed Nov 05 23:52:35 2014 -0800" }, "message": "openvswitch: Drop packets when interdev is not up\n\nIf the internal device is not up, it should drop received\npackets. Sometimes it receive the broadcast or multicast\npackets, and the ip protocol stack will casue more cpu\nusage wasted.\n\nSigned-off-by: Chunhe Li \u003clichunhe@huawei.com\u003e\nSigned-off-by: Pravin B Shelar \u003cpshelar@nicira.com\u003e\n" }, { "commit": "cc3a5ae6f23336ddc1d136a66c0a6a60276e99a3", "tree": "6c3af8e0e35ca25c331659a7f6c5c057158ba1d4", "parents": [ "ca7105f278b3f7bd2c6f2b336c928f679054de4d" ], "author": { "name": "Andy Zhou", "email": "azhou@nicira.com", "time": "Mon Sep 08 13:14:22 2014 -0700" }, "committer": { "name": "Pravin B Shelar", "email": "pshelar@nicira.com", "time": "Wed Nov 05 23:52:34 2014 -0800" }, "message": "openvswitch: Refactor get_dp() function into multiple access APIs.\n\nAvoid recursive read_rcu_lock() by using the lighter weight\nget_dp_rcu() API. Add proper locking assertions to get_dp().\n\nSigned-off-by: Andy Zhou \u003cazhou@nicira.com\u003e\nSigned-off-by: Pravin B Shelar \u003cpshelar@nicira.com\u003e\n" }, { "commit": "ca7105f278b3f7bd2c6f2b336c928f679054de4d", "tree": "2ebb11ba2d8890820e4e66d1ae6d25cf5a5bf3f3", "parents": [ "738967b8bf57e582db1a23ce773c36fefd4b7d37" ], "author": { "name": "Joe Stringer", "email": "joestringer@nicira.com", "time": "Mon Sep 08 13:09:37 2014 -0700" }, "committer": { "name": "Pravin B Shelar", "email": "pshelar@nicira.com", "time": "Wed Nov 05 23:52:34 2014 -0800" }, "message": "openvswitch: Refactor ovs_flow_cmd_fill_info().\n\nSplit up ovs_flow_cmd_fill_info() to make it easier to cache parts of a\ndump reply. This will be used to streamline flow_dump in a future patch.\n\nSigned-off-by: Joe Stringer \u003cjoestringer@nicira.com\u003e\nAcked-by: Thomas Graf \u003ctgraf@noironetworks.com\u003e\nSigned-off-by: Pravin B Shelar \u003cpshelar@nicira.com\u003e\n" }, { "commit": "738967b8bf57e582db1a23ce773c36fefd4b7d37", "tree": "b8a6a56e9eaa280fc10371002bb6d8fef790ab04", "parents": [ "426cda5cc177301f9c196f3a9b6a1287051ba599" ], "author": { "name": "Andy Zhou", "email": "azhou@nicira.com", "time": "Mon Sep 08 00:35:02 2014 -0700" }, "committer": { "name": "Pravin B Shelar", "email": "pshelar@nicira.com", "time": "Wed Nov 05 23:52:34 2014 -0800" }, "message": "openvswitch: refactor do_output() to move NULL check out of fast path\n\nskb_clone() NULL check is implemented in do_output(), as past of the\ncommon (fast) path. Refactoring so that NULL check is done in the\nslow path, immediately after skb_clone() is called.\n\nBesides optimization, this change also improves code readability by\nmaking the skb_clone() NULL check consistent within OVS datapath\nmodule.\n\nSigned-off-by: Andy Zhou \u003cazhou@nicira.com\u003e\nSigned-off-by: Pravin B Shelar \u003cpshelar@nicira.com\u003e\n" }, { "commit": "426cda5cc177301f9c196f3a9b6a1287051ba599", "tree": "f2048a5349f4aedb4e9a65c692ef448d83e8598c", "parents": [ "1b760fb9a8a8c0babc15f886ee5740cb33744168" ], "author": { "name": "Jesse Gross", "email": "jesse@nicira.com", "time": "Mon Oct 06 05:08:38 2014 -0700" }, "committer": { "name": "Pravin B Shelar", "email": "pshelar@nicira.com", "time": "Wed Nov 05 23:52:34 2014 -0800" }, "message": "openvswitch: Additional logging for -EINVAL on flow setups.\n\nThere are many possible ways that a flow can be invalid so we\u0027ve\nadded logging for most of them. This adds logs for the remaining\npossible cases so there isn\u0027t any ambiguity while debugging.\n\nCC: Federico Iezzi \u003cfiezzi@enter.it\u003e\nSigned-off-by: Jesse Gross \u003cjesse@nicira.com\u003e\nAcked-by: Thomas Graf \u003ctgraf@noironetworks.com\u003e\nSigned-off-by: Pravin B Shelar \u003cpshelar@nicira.com\u003e\n" }, { "commit": "1b760fb9a8a8c0babc15f886ee5740cb33744168", "tree": "69662104f13eacf4c5fdef317c9b2db9ad36b983", "parents": [ "9b996e544a6bc7d201060fdcbdb5d4a9b734aa1b" ], "author": { "name": "Joe Stringer", "email": "joestringer@nicira.com", "time": "Sun Sep 07 22:11:08 2014 -0700" }, "committer": { "name": "Pravin B Shelar", "email": "pshelar@nicira.com", "time": "Wed Nov 05 23:52:34 2014 -0800" }, "message": "openvswitch: Remove redundant tcp_flags code.\n\nThese two cases used to be treated differently for IPv4/IPv6,\nbut they are now identical.\n\nSigned-off-by: Joe Stringer \u003cjoestringer@nicira.com\u003e\nAcked-by: Jarno Rajahalme \u003cjrajahalme@nicira.com\u003e\nSigned-off-by: Pravin B Shelar \u003cpshelar@nicira.com\u003e\n" }, { "commit": "9b996e544a6bc7d201060fdcbdb5d4a9b734aa1b", "tree": "460ba6f2685e4030871b43db2d3183bc1f781747", "parents": [ "25cd9ba0abc0749e5cb78e6493c6f6b3311ec6c5" ], "author": { "name": "Pravin B Shelar", "email": "pshelar@nicira.com", "time": "Tue May 06 18:41:20 2014 -0700" }, "committer": { "name": "Pravin B Shelar", "email": "pshelar@nicira.com", "time": "Wed Nov 05 23:52:34 2014 -0800" }, "message": "openvswitch: Move table destroy to dp-rcu callback.\n\nThs simplifies flow-table-destroy API. No need to pass explicit\nparameter about context.\n\nSigned-off-by: Pravin B Shelar \u003cpshelar@nicira.com\u003e\nAcked-by: Thomas Graf \u003ctgraf@redhat.com\u003e\n" }, { "commit": "25cd9ba0abc0749e5cb78e6493c6f6b3311ec6c5", "tree": "fbfb953b27fbdcc27d0a50a6e3444532f51f5ffa", "parents": [ "59b93b41e7fa71138734a911b11b044340dd16bd" ], "author": { "name": "Simon Horman", "email": "horms@verge.net.au", "time": "Mon Oct 06 05:05:13 2014 -0700" }, "committer": { "name": "Pravin B Shelar", "email": "pshelar@nicira.com", "time": "Wed Nov 05 23:52:33 2014 -0800" }, "message": "openvswitch: Add basic MPLS support to kernel\n\nAllow datapath to recognize and extract MPLS labels into flow keys\nand execute actions which push, pop, and set labels on packets.\n\nBased heavily on work by Leo Alterman, Ravi K, Isaku Yamahata and Joe Stringer.\n\nCc: Ravi K \u003crkerur@gmail.com\u003e\nCc: Leo Alterman \u003clalterman@nicira.com\u003e\nCc: Isaku Yamahata \u003cyamahata@valinux.co.jp\u003e\nCc: Joe Stringer \u003cjoe@wand.net.nz\u003e\nSigned-off-by: Simon Horman \u003chorms@verge.net.au\u003e\nSigned-off-by: Jesse Gross \u003cjesse@nicira.com\u003e\nSigned-off-by: Pravin B Shelar \u003cpshelar@nicira.com\u003e\n" }, { "commit": "59b93b41e7fa71138734a911b11b044340dd16bd", "tree": "7d5680f8c3eba5ce40a04b6115b41550c9d140f5", "parents": [ "e1b2cb655060e081e73b384b1fc8b2e978f73467" ], "author": { "name": "Pravin B Shelar", "email": "pshelar@nicira.com", "time": "Wed Nov 05 15:27:48 2014 -0800" }, "committer": { "name": "Pravin B Shelar", "email": "pshelar@nicira.com", "time": "Wed Nov 05 23:52:33 2014 -0800" }, "message": "net: Remove MPLS GSO feature.\n\nDevice can export MPLS GSO support in dev-\u003empls_features same way\nit export vlan features in dev-\u003evlan_features. So it is safe to\nremove NETIF_F_GSO_MPLS redundant flag.\n\nSigned-off-by: Pravin B Shelar \u003cpshelar@nicira.com\u003e\n" }, { "commit": "e1b2cb655060e081e73b384b1fc8b2e978f73467", "tree": "46b69489a8af03acbccbdf5038ec0cf5854f8191", "parents": [ "93ffbeab77ca6a49e075235faa04c8b032c9fdc1" ], "author": { "name": "Tom Herbert", "email": "therbert@google.com", "time": "Wed Nov 05 16:49:38 2014 -0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Nov 05 22:18:20 2014 -0500" }, "message": "fou: Fix typo in returning flags in netlink\n\nWhen filling netlink info, dport is being returned as flags. Fix\ninstances to return correct value.\n\nSigned-off-by: Tom Herbert \u003ctherbert@google.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "93ffbeab77ca6a49e075235faa04c8b032c9fdc1", "tree": "98d0e0dc1a8954e5418ccd1a9a72a3ca6b5fc2ac", "parents": [ "4c672e4b42bc8046d63a6eb0a2c6a450a501af32" ], "author": { "name": "hayeswang", "email": "hayeswang@realtek.com", "time": "Wed Nov 05 10:17:02 2014 +0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Nov 05 22:17:10 2014 -0500" }, "message": "r8152: disable the tasklet by default\n\nLet the tasklet only be enabled after open(), and be disabled for\nthe other situation. The tasklet is only necessary after open() for\ntx/rx, so it could be disabled by default.\n\nSigned-off-by: Hayes Wang \u003chayeswang@realtek.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "4c672e4b42bc8046d63a6eb0a2c6a450a501af32", "tree": "a6f7f031f6c1995c8854775599919eb18fdfecc6", "parents": [ "1744bea1fa382f67263fdd9fee51d603fddb3da6" ], "author": { "name": "Daniel Borkmann", "email": "dborkman@redhat.com", "time": "Wed Nov 05 20:27:38 2014 +0100" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Nov 05 22:12:30 2014 -0500" }, "message": "ipv6: mld: fix add_grhead skb_over_panic for devs with large MTUs\n\nIt has been reported that generating an MLD listener report on\ndevices with large MTUs (e.g. 9000) and a high number of IPv6\naddresses can trigger a skb_over_panic():\n\nskbuff: skb_over_panic: text:ffffffff80612a5d len:3776 put:20\nhead:ffff88046d751000 data:ffff88046d751010 tail:0xed0 end:0xec0\ndev:port1\n ------------[ cut here ]------------\nkernel BUG at net/core/skbuff.c:100!\ninvalid opcode: 0000 [#1] SMP\nModules linked in: ixgbe(O)\nCPU: 3 PID: 0 Comm: swapper/3 Tainted: G O 3.14.23+ #4\n[...]\nCall Trace:\n \u003cIRQ\u003e\n [\u003cffffffff80578226\u003e] ? skb_put+0x3a/0x3b\n [\u003cffffffff80612a5d\u003e] ? add_grhead+0x45/0x8e\n [\u003cffffffff80612e3a\u003e] ? add_grec+0x394/0x3d4\n [\u003cffffffff80613222\u003e] ? mld_ifc_timer_expire+0x195/0x20d\n [\u003cffffffff8061308d\u003e] ? mld_dad_timer_expire+0x45/0x45\n [\u003cffffffff80255b5d\u003e] ? call_timer_fn.isra.29+0x12/0x68\n [\u003cffffffff80255d16\u003e] ? run_timer_softirq+0x163/0x182\n [\u003cffffffff80250e6f\u003e] ? __do_softirq+0xe0/0x21d\n [\u003cffffffff8025112b\u003e] ? irq_exit+0x4e/0xd3\n [\u003cffffffff802214bb\u003e] ? smp_apic_timer_interrupt+0x3b/0x46\n [\u003cffffffff8063f10a\u003e] ? apic_timer_interrupt+0x6a/0x70\n\nmld_newpack() skb allocations are usually requested with dev-\u003emtu\nin size, since commit 72e09ad107e7 (\"ipv6: avoid high order allocations\")\nwe have changed the limit in order to be less likely to fail.\n\nHowever, in MLD/IGMP code, we have some rather ugly AVAILABLE(skb)\nmacros, which determine if we may end up doing an skb_put() for\nadding another record. To avoid possible fragmentation, we check\nthe skb\u0027s tailroom as skb-\u003edev-\u003emtu - skb-\u003elen, which is a wrong\nassumption as the actual max allocation size can be much smaller.\n\nThe IGMP case doesn\u0027t have this issue as commit 57e1ab6eaddc\n(\"igmp: refine skb allocations\") stores the allocation size in\nthe cb[].\n\nSet a reserved_tailroom to make it fit into the MTU and use\nskb_availroom() helper instead. This also allows to get rid of\nigmp_skb_size().\n\nReported-by: Wei Liu \u003clw1a2.jing@gmail.com\u003e\nFixes: 72e09ad107e7 (\"ipv6: avoid high order allocations\")\nSigned-off-by: Daniel Borkmann \u003cdborkman@redhat.com\u003e\nCc: Eric Dumazet \u003cedumazet@google.com\u003e\nCc: Hannes Frederic Sowa \u003channes@stressinduktion.org\u003e\nCc: David L Stevens \u003cdavid.stevens@oracle.com\u003e\nAcked-by: Eric Dumazet \u003cedumazet@google.com\u003e\nAcked-by: Hannes Frederic Sowa \u003channes@stressinduktion.org\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "1744bea1fa382f67263fdd9fee51d603fddb3da6", "tree": "fa9d428d35a2c18571ccfecdf4bb8c3349cddef8", "parents": [ "e5a2c899957659cd1a9f789bc462f9c0b35f5150" ], "author": { "name": "Joe Perches", "email": "joe@perches.com", "time": "Tue Nov 04 15:37:03 2014 -0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Nov 05 22:04:55 2014 -0500" }, "message": "net: Convert SEQ_START_TOKEN/seq_printf to seq_puts\n\nUsing a single fixed string is smaller code size than using\na format and many string arguments.\n\nReduces overall code size a little.\n\n$ size net/ipv4/igmp.o* net/ipv6/mcast.o* net/ipv6/ip6_flowlabel.o*\n text\t data\t bss\t dec\t hex\tfilename\n 34269\t 7012\t 14824\t 56105\t db29\tnet/ipv4/igmp.o.new\n 34315\t 7012\t 14824\t 56151\t db57\tnet/ipv4/igmp.o.old\n 30078\t 7869\t 13200\t 51147\t c7cb\tnet/ipv6/mcast.o.new\n 30105\t 7869\t 13200\t 51174\t c7e6\tnet/ipv6/mcast.o.old\n 11434\t 3748\t 8580\t 23762\t 5cd2\tnet/ipv6/ip6_flowlabel.o.new\n 11491\t 3748\t 8580\t 23819\t 5d0b\tnet/ipv6/ip6_flowlabel.o.old\n\nSigned-off-by: Joe Perches \u003cjoe@perches.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "e5a2c899957659cd1a9f789bc462f9c0b35f5150", "tree": "4c9b8a6f89d961daf9ada9f5ee95f8b371ce3a04", "parents": [ "2c99cd914d4fed9160d98849c9dd38034616768e" ], "author": { "name": "Hannes Frederic Sowa", "email": "hannes@stressinduktion.org", "time": "Wed Nov 05 00:23:04 2014 +0100" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Nov 05 22:01:21 2014 -0500" }, "message": "fast_hash: avoid indirect function calls\n\nBy default the arch_fast_hash hashing function pointers are initialized\nto jhash(2). If during boot-up a CPU with SSE4.2 is detected they get\nupdated to the CRC32 ones. This dispatching scheme incurs a function\npointer lookup and indirect call for every hashing operation.\n\nrhashtable as a user of arch_fast_hash e.g. stores pointers to hashing\nfunctions in its structure, too, causing two indirect branches per\nhashing operation.\n\nUsing alternative_call we can get away with one of those indirect branches.\n\nAcked-by: Daniel Borkmann \u003cdborkman@redhat.com\u003e\nCc: Thomas Graf \u003ctgraf@suug.ch\u003e\nSigned-off-by: Hannes Frederic Sowa \u003channes@stressinduktion.org\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "2c99cd914d4fed9160d98849c9dd38034616768e", "tree": "fd6c145593ff1c901a9980765952700586d6fe20", "parents": [ "25de4668d094f00e44a8f2428dd3c1a4ecfa0053", "5cdec679671741bcca1f8280699a64b42c9fa2b4" ], "author": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Nov 05 21:50:43 2014 -0500" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Nov 05 21:50:43 2014 -0500" }, "message": "Merge branch \u0027amd-xgbe-next\u0027\n\nTom Lendacky says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\namd-xgbe: AMD XGBE driver updates 2014-11-04\n\nThe following series of patches includes functional updates to the\ndriver as well as some trivial changes for function renaming and\nspelling fixes.\n\n- Move channel and ring structure allocation into the device open path\n- Rename the pre_xmit function to dev_xmit\n- Explicitly use the u32 data type for the device descriptors\n- Use page allocation for the receive buffers\n- Add support for split header/payload receive\n- Add support for per DMA channel interrupts\n- Add support for receive side scaling (RSS)\n- Add support for ethtool receive side scaling commands\n- Fix the spelling of descriptors\n- After a PCS reset, sync the PCS and PHY modes\n- Add dependency on HAS_IOMEM to both the amd-xgbe and amd-xgbe-phy\n drivers\n\nThis patch series is based on net-next.\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "5cdec679671741bcca1f8280699a64b42c9fa2b4", "tree": "fd6c145593ff1c901a9980765952700586d6fe20", "parents": [ "474809b9e1daf429047195d2394c51a6b4b09f36" ], "author": { "name": "Lendacky, Thomas", "email": "Thomas.Lendacky@amd.com", "time": "Tue Nov 04 16:07:46 2014 -0600" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Nov 05 21:50:13 2014 -0500" }, "message": "amd-xgbe-phy: Let AMD_XGBE_PHY depend on HAS_IOMEM\n\nThe amd-xgbe-phy driver needs to perform ioremap calls, so add HAS_IOMEM\nto its build dependency.\n\nSigned-off-by: Tom Lendacky \u003cthomas.lendacky@amd.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "474809b9e1daf429047195d2394c51a6b4b09f36", "tree": "2be146a07eeffaa8f93214758b8fe37c55572e9c", "parents": [ "0c95a1faaa155bb49fc4acfec61242e80b8a8502" ], "author": { "name": "Lendacky, Thomas", "email": "Thomas.Lendacky@amd.com", "time": "Tue Nov 04 16:07:41 2014 -0600" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Nov 05 21:50:13 2014 -0500" }, "message": "amd-xgbe: Let AMD_XGBE depend on HAS_IOMEM\n\nThe amd-xgbe driver needs to perform ioremap calls, so add HAS_IOMEM\nto its build dependency.\n\nSigned-off-by: Tom Lendacky \u003cthomas.lendacky@amd.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "0c95a1faaa155bb49fc4acfec61242e80b8a8502", "tree": "c286514d1e855efce6257114d2a6363a1dc4bc6a", "parents": [ "a7beaf23007ac701ebffee1cb48d5d1aea5e7803" ], "author": { "name": "Lendacky, Thomas", "email": "Thomas.Lendacky@amd.com", "time": "Tue Nov 04 16:07:35 2014 -0600" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Nov 05 21:50:12 2014 -0500" }, "message": "amd-xgbe-phy: Sync PCS and PHY modes after reset\n\nThis patch adds support to sync the states of the PCS and the PHY\nafter a reset is performed. If the PCS and the PHY are not in the\nsame state after reset an extra mode change would be performed. This\nextra mode change might not be needed if the PCS and the PHY are\nsynced up after reset.\n\nSigned-off-by: Tom Lendacky \u003cthomas.lendacky@amd.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "a7beaf23007ac701ebffee1cb48d5d1aea5e7803", "tree": "b50a0bf679bdb3d3ca5ffba66c992e205366209e", "parents": [ "f6ac862845bbe9daac9a02c7690efa01bfb62fa5" ], "author": { "name": "Lendacky, Thomas", "email": "Thomas.Lendacky@amd.com", "time": "Tue Nov 04 16:07:29 2014 -0600" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Nov 05 21:50:12 2014 -0500" }, "message": "amd-xgbe: Fix a spelling error\n\nThis patch fixes the spelling of the word \"descriptor\" in a couple\nof locations.\n\nSigned-off-by: Tom Lendacky \u003cthomas.lendacky@amd.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "f6ac862845bbe9daac9a02c7690efa01bfb62fa5", "tree": "1d3394684f9b60aa7d47a7928e59c9689920d512", "parents": [ "5b9dfe299e55604af47cdca9d03d2e9d4fe2ad53" ], "author": { "name": "Lendacky, Thomas", "email": "Thomas.Lendacky@amd.com", "time": "Tue Nov 04 16:07:23 2014 -0600" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Nov 05 21:50:12 2014 -0500" }, "message": "amd-xgbe: Add receive side scaling ethtool support\n\nThis patch adds support for ethtool receive side scaling (RSS) commands.\nSupport is added to get/set the RSS hash key and the RSS lookup table.\n\nSigned-off-by: Tom Lendacky \u003cthomas.lendacky@amd.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "5b9dfe299e55604af47cdca9d03d2e9d4fe2ad53", "tree": "c011bc202e146d22f51b526c39344bf3887e7a53", "parents": [ "9227dc5e579b6b2ef58ad0d3d0d23ddac77846ef" ], "author": { "name": "Lendacky, Thomas", "email": "Thomas.Lendacky@amd.com", "time": "Tue Nov 04 16:07:02 2014 -0600" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Nov 05 21:50:12 2014 -0500" }, "message": "amd-xgbe: Provide support for receive side scaling\n\nThis patch provides support for receive side scaling (RSS). RSS allows\nfor spreading incoming network packets across the Rx queues. When used\nin conjunction with the per DMA channel interrupt support, this allows\nthe receive processing to be spread across multiple processors.\n\nSigned-off-by: Tom Lendacky \u003cthomas.lendacky@amd.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "9227dc5e579b6b2ef58ad0d3d0d23ddac77846ef", "tree": "cef535601eda5a1c79f6d6d96bf2b752993d72bb", "parents": [ "174fd2597b0bd8c19fce6a97e8b0f753ef4ce7cb" ], "author": { "name": "Lendacky, Thomas", "email": "Thomas.Lendacky@amd.com", "time": "Tue Nov 04 16:06:56 2014 -0600" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Nov 05 21:50:12 2014 -0500" }, "message": "amd-xgbe: Add support for per DMA channel interrupts\n\nThis patch provides support for interrupts that are generated by the\nTx/Rx DMA channel pairs of the device. This allows for Tx and Rx\nprocessing to run across multiple processsors.\n\nSigned-off-by: Tom Lendacky \u003cthomas.lendacky@amd.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "174fd2597b0bd8c19fce6a97e8b0f753ef4ce7cb", "tree": "0770a468c71da5dcaa7353537f822584df34907e", "parents": [ "08dcc47c06c79de31b9b2c0b4637f6119e5701fa" ], "author": { "name": "Lendacky, Thomas", "email": "Thomas.Lendacky@amd.com", "time": "Tue Nov 04 16:06:50 2014 -0600" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Nov 05 21:50:12 2014 -0500" }, "message": "amd-xgbe: Implement split header receive support\n\nProvide support for splitting IP packets so that the header and\npayload can be sent to different DMA addresses. This will allow\nthe IP header to be put into the linear part of the skb while the\npayload can be added as frags.\n\nSigned-off-by: Tom Lendacky \u003cthomas.lendacky@amd.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "08dcc47c06c79de31b9b2c0b4637f6119e5701fa", "tree": "55a56cf66275b02c33afb6e4a2d2296b3d1450d8", "parents": [ "aa96bd3c9fda52b9f57128798b49d662e5d4659c" ], "author": { "name": "Lendacky, Thomas", "email": "Thomas.Lendacky@amd.com", "time": "Tue Nov 04 16:06:44 2014 -0600" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Nov 05 21:50:12 2014 -0500" }, "message": "amd-xgbe: Use page allocations for Rx buffers\n\nUse page allocations for Rx buffers instead of pre-allocating skbs\nof a set size.\n\nSigned-off-by: Tom Lendacky \u003cthomas.lendacky@amd.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "aa96bd3c9fda52b9f57128798b49d662e5d4659c", "tree": "ee69824197d8c5d7313b620177690b49f32dc310", "parents": [ "a9d41981e95651143125352f0233138efc17378a" ], "author": { "name": "Lendacky, Thomas", "email": "Thomas.Lendacky@amd.com", "time": "Tue Nov 04 16:06:37 2014 -0600" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Nov 05 21:50:12 2014 -0500" }, "message": "amd-xgbe: Use the u32 data type for descriptors\n\nThe Tx and Rx descriptors are unsigned 32 bit values. Use the u32\ntype, rather than unsigned int, to map these descriptors.\n\nSigned-off-by: Tom Lendacky \u003cthomas.lendacky@amd.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "a9d41981e95651143125352f0233138efc17378a", "tree": "a0e9acdcfc564cc67ac6c4cbb18341f8293e0366", "parents": [ "4780b7cae60cf10af4ae75bc5d6643f41d4c2969" ], "author": { "name": "Lendacky, Thomas", "email": "Thomas.Lendacky@amd.com", "time": "Tue Nov 04 16:06:32 2014 -0600" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Nov 05 21:50:12 2014 -0500" }, "message": "amd-xgbe: Rename pre_xmit function to dev_xmit\n\nThe pre_xmit function name implies that it performs operations prior\nto transmitting the packet when in fact it is responsible for setting\nup the descriptors and initiating the transmit. Rename this to\nfunction from pre_xmit to dev_xmit, which is consistent with the name\nused during receive processing - dev_read.\n\nSigned-off-by: Tom Lendacky \u003cthomas.lendacky@amd.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "4780b7cae60cf10af4ae75bc5d6643f41d4c2969", "tree": "1feb576159fdb1cbde6067336b83377f2ec8f93d", "parents": [ "25de4668d094f00e44a8f2428dd3c1a4ecfa0053" ], "author": { "name": "Lendacky, Thomas", "email": "Thomas.Lendacky@amd.com", "time": "Tue Nov 04 16:06:26 2014 -0600" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Nov 05 21:50:11 2014 -0500" }, "message": "amd-xgbe: Move ring allocation to device open\n\nMove the channel and ring tracking structures allocation to device\nopen. This will allow for future support to vary the number of Tx/Rx\nqueues without unloading the module.\n\nSigned-off-by: Tom Lendacky \u003cthomas.lendacky@amd.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "66f1c44887ba4f47d617f8ae21cf8e04e1892bd7", "tree": "669fbc755c6062740c6542ccdb855ca9ddfd9315", "parents": [ "1f37bf87aa7523d28e7e4c4f7bb5dba98faa3e00" ], "author": { "name": "Gregory Fong", "email": "gregory.0xf0@gmail.com", "time": "Tue Nov 04 11:21:21 2014 -0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Nov 05 17:13:34 2014 -0500" }, "message": "bridge: include in6.h in if_bridge.h for struct in6_addr\n\nif_bridge.h uses struct in6_addr ip6, but wasn\u0027t including the in6.h\nheader. Thomas Backlund originally sent a patch to do this, but this\nrevealed a redefinition issue: https://lkml.org/lkml/2013/1/13/116\n\nThe redefinition issue should have been fixed by the following Linux\ncommits:\nee262ad827f89e2dc7851ec2986953b5b125c6bc inet: defines IPPROTO_* needed for module alias generation\ncfd280c91253cc28e4919e349fa7a813b63e71e8 net: sync some IP headers with glibc\n\nand the following glibc commit:\n6c82a2f8d7c8e21e39237225c819f182ae438db3 Coordinate IPv6 definitions for Linux and glibc\n\nso actually include the header now.\n\nReported-by: Colin Guthrie \u003ccolin@mageia.org\u003e\nReported-by: Christiaan Welvaart \u003ccjw@daneel.dyndns.org\u003e\nReported-by: Thomas Backlund \u003ctmb@mageia.org\u003e\nCc: Florian Fainelli \u003cf.fainelli@gmail.com\u003e\nCc: Cong Wang \u003cxiyou.wangcong@gmail.com\u003e\nCc: David Miller \u003cdavem@davemloft.net\u003e\nSigned-off-by: Gregory Fong \u003cgregory.0xf0@gmail.com\u003e\nAcked-by: Cong Wang \u003cxiyou.wangcong@gmail.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "1f37bf87aa7523d28e7e4c4f7bb5dba98faa3e00", "tree": "587ea2b8eaff79f04b9f6cadce4089c1dd765b32", "parents": [ "46d3802627f60be1c17659a44c8d2d7a5e247023" ], "author": { "name": "Marcelo Leitner", "email": "mleitner@redhat.com", "time": "Tue Nov 04 17:15:08 2014 -0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Nov 05 16:59:49 2014 -0500" }, "message": "tcp: zero retrans_stamp if all retrans were acked\n\nUeki Kohei reported that when we are using NewReno with connections that\nhave a very low traffic, we may timeout the connection too early if a\nsecond loss occurs after the first one was successfully acked but no\ndata was transfered later. Below is his description of it:\n\nWhen SACK is disabled, and a socket suffers multiple separate TCP\nretransmissions, that socket\u0027s ETIMEDOUT value is calculated from the\ntime of the *first* retransmission instead of the *latest*\nretransmission.\n\nThis happens because the tcp_sock\u0027s retrans_stamp is set once then never\ncleared.\n\nTake the following connection:\n\n Linux remote-machine\n | |\n send#1----\u003e(*1)|--------\u003e data#1 ---------\u003e|\n | | |\n RTO : :\n | | |\n ---(*2)|----\u003e data#1(retrans) ----\u003e|\n | (*3)|\u003c---------- ACK \u003c----------|\n | | |\n | : :\n | : :\n | : :\n 16 minutes (or more) :\n | : :\n | : :\n | : :\n | | |\n send#2----\u003e(*4)|--------\u003e data#2 ---------\u003e|\n | | |\n RTO : :\n | | |\n ---(*5)|----\u003e data#2(retrans) ----\u003e|\n | | |\n | | |\n RTO*2 : :\n | | |\n | | |\n ETIMEDOUT\u003c----(*6)| |\n\n(*1) One data packet sent.\n(*2) Because no ACK packet is received, the packet is retransmitted.\n(*3) The ACK packet is received. The transmitted packet is acknowledged.\n\nAt this point the first \"retransmission event\" has passed and been\nrecovered from. Any future retransmission is a completely new \"event\".\n\n(*4) After 16 minutes (to correspond with retries2\u003d15), a new data\npacket is sent. Note: No data is transmitted between (*3) and (*4).\n\nThe socket\u0027s timeout SHOULD be calculated from this point in time, but\ninstead it\u0027s calculated from the prior \"event\" 16 minutes ago.\n\n(*5) Because no ACK packet is received, the packet is retransmitted.\n(*6) At the time of the 2nd retransmission, the socket returns\nETIMEDOUT.\n\nTherefore, now we clear retrans_stamp as soon as all data during the\nloss window is fully acked.\n\nReported-by: Ueki Kohei\nCc: Neal Cardwell \u003cncardwell@google.com\u003e\nCc: Yuchung Cheng \u003cycheng@google.com\u003e\nSigned-off-by: Marcelo Ricardo Leitner \u003cmleitner@redhat.com\u003e\nAcked-by: Neal Cardwell \u003cncardwell@google.com\u003e\nTested-by: Neal Cardwell \u003cncardwell@google.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "25de4668d094f00e44a8f2428dd3c1a4ecfa0053", "tree": "feb481e52578e7278538814191777d500cb2ccc1", "parents": [ "51f3d02b980a338cd291d2bc7629cdfb2568424b" ], "author": { "name": "WANG Cong", "email": "xiyou.wangcong@gmail.com", "time": "Tue Nov 04 10:59:47 2014 -0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Nov 05 16:59:04 2014 -0500" }, "message": "ipv6: move INET6_MATCH() to include/net/inet6_hashtables.h\n\nIt is only used in net/ipv6/inet6_hashtables.c.\n\nCc: David S. Miller \u003cdavem@davemloft.net\u003e\nSigned-off-by: Cong Wang \u003cxiyou.wangcong@gmail.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "51f3d02b980a338cd291d2bc7629cdfb2568424b", "tree": "f49898ea87db3d5e3f5c248ab21f6ed0a70785b8", "parents": [ "1d76c1d028975df8488d1ae18a76f268eb5efa93" ], "author": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Nov 05 16:46:40 2014 -0500" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Nov 05 16:46:40 2014 -0500" }, "message": "net: Add and use skb_copy_datagram_msg() helper.\n\nThis encapsulates all of the skb_copy_datagram_iovec() callers\nwith call argument signature \"skb, offset, msghdr-\u003emsg_iov, length\".\n\nWhen we move to iov_iters in the networking, the iov_iter object will\nsit in the msghdr.\n\nHaving a helper like this means there will be less places to touch\nduring that transformation.\n\nBased upon descriptions and patch from Al Viro.\n\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "1d76c1d028975df8488d1ae18a76f268eb5efa93", "tree": "01bfc4d3ef16fe7e5a4da0be1e7f3fd432e7495f", "parents": [ "890b7916d0965829ad1c457aa61f049a210c19f8", "a8d31c128bf574bed2fa29e0512b24d446018a50" ], "author": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Nov 05 16:34:47 2014 -0500" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Nov 05 16:34:47 2014 -0500" }, "message": "Merge branch \u0027gue-next\u0027\n\nTom Herbert says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\ngue: Remote checksum offload\n\nThis patch set implements remote checksum offload for\nGUE, which is a mechanism that provides checksum offload of\nencapsulated packets using rudimentary offload capabilities found in\nmost Network Interface Card (NIC) devices. The outer header checksum\nfor UDP is enabled in packets and, with some additional meta\ninformation in the GUE header, a receiver is able to deduce the\nchecksum to be set for an inner encapsulated packet. Effectively this\noffloads the computation of the inner checksum. Enabling the outer\nchecksum in encapsulation has the additional advantage that it covers\nmore of the packet than the inner checksum including the encapsulation\nheaders.\n\nRemote checksum offload is described in:\nhttp://tools.ietf.org/html/draft-herbert-remotecsumoffload-01\n\nThe GUE transmit and receive paths are modified to support the\nremote checksum offload option. The option contains a checksum\noffset and checksum start which are directly derived from values\nset in stack when doing CHECKSUM_PARTIAL. On receipt of the option, the\noperation is to calculate the packet checksum from \"start\" to end of\nthe packet (normally derived for checksum complete), and then set\nthe resultant value at checksum \"offset\" (the checksum field has\nalready been primed with the pseudo header). This emulates a NIC\nthat implements NETIF_F_HW_CSUM.\n\nThe primary purpose of this feature is to eliminate cost of performing\nchecksum calculation over a packet when encpasulating.\n\nIn this patch set:\n - Move fou_build_header into fou.c and split it into a couple of\n functions\n - Enable offloading of outer UDP checksum in encapsulation\n - Change udp_offload to support remote checksum offload, includes\n new GSO type and ensuring encapsulated layers (TCP) doesn\u0027t try to\n set a checksum covered by RCO\n - TX support for RCO with GUE. This is configured through ip_tunnel\n and set the option on transmit when packet being encapsulated is\n CHECKSUM_PARTIAL\n - RX support for RCO with GUE for normal and GRO paths. Includes\n resolving the offloaded checksum\n\nv2:\n Address comments from davem: Move accounting for private option\n field in gue_encap_hlen to patch in which we add the remote checksum\n offload option.\n\nTesting:\n\nI ran performance numbers using netperf TCP_STREAM and TCP_RR with 200\nstreams, comparing GUE with and without remote checksum offload (doing\nchecksum-unnecessary to complete conversion in both cases). These\nwere run on mlnx4 and bnx2x. Some mlnx4 results are below.\n\nGRE/GUE\n TCP_STREAM\n IPv4, with remote checksum offload\n 9.71% TX CPU utilization\n 7.42% RX CPU utilization\n 36380 Mbps\n IPv4, without remote checksum offload\n 12.40% TX CPU utilization\n 7.36% RX CPU utilization\n 36591 Mbps\n TCP_RR\n IPv4, with remote checksum offload\n 77.79% CPU utilization\n\t91/144/216 90/95/99% latencies\n 1.95127e+06 tps\n IPv4, without remote checksum offload\n 78.70% CPU utilization\n 89/152/297 90/95/99% latencies\n 1.95458e+06 tps\n\nIPIP/GUE\n TCP_STREAM\n With remote checksum offload\n 10.30% TX CPU utilization\n 7.43% RX CPU utilization\n 36486 Mbps\n Without remote checksum offload\n 12.47% TX CPU utilization\n 7.49% RX CPU utilization\n 36694 Mbps\n TCP_RR\n With remote checksum offload\n 77.80% CPU utilization\n 87/153/270 90/95/99% latencies\n 1.98735e+06 tps\n Without remote checksum offload\n 77.98% CPU utilization\n 87/150/287 90/95/99% latencies\n 1.98737e+06 tps\n\nSIT/GUE\n TCP_STREAM\n With remote checksum offload\n 9.68% TX CPU utilization\n 7.36% RX CPU utilization\n 35971 Mbps\n Without remote checksum offload\n 12.95% TX CPU utilization\n 8.04% RX CPU utilization\n 36177 Mbps\n TCP_RR\n With remote checksum offload\n 79.32% CPU utilization\n 94/158/295 90/95/99% latencies\n 1.88842e+06 tps\n Without remote checksum offload\n 80.23% CPU utilization\n 94/149/226 90/95/99% latencies\n 1.90338e+06 tps\n\nVXLAN\n TCP_STREAM\n 35.03% TX CPU utilization\n 20.85% RX CPU utilization\n 36230 Mbps\n TCP_RR\n 77.36% CPU utilization\n 84/146/270 90/95/99% latencies\n 2.08063e+06 tps\n\nWe can also look at CPU time in csum_partial using perf (with bnx2x\nsetup). For GRE with TCP_STREAM I see:\n\n With remote checksum offload\n 0.33% TX\n 1.81% RX\n Without remote checksum offload\n 6.00% TX\n 0.51% RX\n\nI suspect the fact that time in csum_partial noticably increases\nwith remote checksum offload for RX is due to taking the cache miss on\nthe encapsulated header in that function. By similar reasoning, if on\nthe TX side the packet were not in cache (say we did a splice from a\nfile whose data was never touched by the CPU) the CPU savings for TX\nwould probably be more pronounced.\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "a8d31c128bf574bed2fa29e0512b24d446018a50", "tree": "01bfc4d3ef16fe7e5a4da0be1e7f3fd432e7495f", "parents": [ "b17f709a24013fcbb257f6f89b4d81ac9fdf0d18" ], "author": { "name": "Tom Herbert", "email": "therbert@google.com", "time": "Tue Nov 04 09:06:57 2014 -0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Nov 05 16:30:04 2014 -0500" }, "message": "gue: Receive side of remote checksum offload\n\nAdd processing of the remote checksum offload option in both the normal\npath as well as the GRO path. The implements patching the affected\nchecksum to derive the offloaded checksum.\n\nSigned-off-by: Tom Herbert \u003ctherbert@google.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "b17f709a24013fcbb257f6f89b4d81ac9fdf0d18", "tree": "2d279bf83c7bdbe770a90f784b6c3e2b78e7237d", "parents": [ "c1aa8347e73e4092411fbd96cc59531fb7e76d04" ], "author": { "name": "Tom Herbert", "email": "therbert@google.com", "time": "Tue Nov 04 09:06:56 2014 -0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Nov 05 16:30:03 2014 -0500" }, "message": "gue: TX support for using remote checksum offload option\n\nAdd if_tunnel flag TUNNEL_ENCAP_FLAG_REMCSUM to configure\nremote checksum offload on an IP tunnel. Add logic in gue_build_header\nto insert remote checksum offload option.\n\nSigned-off-by: Tom Herbert \u003ctherbert@google.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "c1aa8347e73e4092411fbd96cc59531fb7e76d04", "tree": "f683f9f9888b42dbdd42f70f1b92a3fa36cc992c", "parents": [ "e585f23636370320bc2071ca5ba2744ae37c3e51" ], "author": { "name": "Tom Herbert", "email": "therbert@google.com", "time": "Tue Nov 04 09:06:55 2014 -0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Nov 05 16:30:03 2014 -0500" }, "message": "gue: Protocol constants for remote checksum offload\n\nDefine a private flag for remote checksun offload as well as a length\nfor the option.\n\nSigned-off-by: Tom Herbert \u003ctherbert@google.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "e585f23636370320bc2071ca5ba2744ae37c3e51", "tree": "c1d9d60bd084099f3e65155dd920ee72d41a808e", "parents": [ "5024c33ac354577635c5671498891eb197f3ec4d" ], "author": { "name": "Tom Herbert", "email": "therbert@google.com", "time": "Tue Nov 04 09:06:54 2014 -0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Nov 05 16:30:03 2014 -0500" }, "message": "udp: Changes to udp_offload to support remote checksum offload\n\nAdd a new GSO type, SKB_GSO_TUNNEL_REMCSUM, which indicates remote\nchecksum offload being done (in this case inner checksum must not\nbe offloaded to the NIC).\n\nAdded logic in __skb_udp_tunnel_segment to handle remote checksum\noffload case.\n\nSigned-off-by: Tom Herbert \u003ctherbert@google.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "5024c33ac354577635c5671498891eb197f3ec4d", "tree": "bb59341e98a708195c3ff6fe3f704a6f0f2dc002", "parents": [ "4bcb877d257c87298aedead1ffeaba0d5df1991d" ], "author": { "name": "Tom Herbert", "email": "therbert@google.com", "time": "Tue Nov 04 09:06:53 2014 -0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Nov 05 16:30:03 2014 -0500" }, "message": "gue: Add infrastructure for flags and options\n\nAdd functions and basic definitions for processing standard flags,\nprivate flags, and control messages. This includes definitions\nto compute length of optional fields corresponding to a set of flags.\nFlag validation is in validate_gue_flags function. This checks for\nunknown flags, and that length of optional fields is \u003c\u003d length\nin guehdr hlen.\n\nSigned-off-by: Tom Herbert \u003ctherbert@google.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "4bcb877d257c87298aedead1ffeaba0d5df1991d", "tree": "00b616716253b6bed2510505e38f345b1a165efa", "parents": [ "63487babf08d6d67483c67ed21d8cea6674a44ec" ], "author": { "name": "Tom Herbert", "email": "therbert@google.com", "time": "Tue Nov 04 09:06:52 2014 -0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Nov 05 16:30:03 2014 -0500" }, "message": "udp: Offload outer UDP tunnel csum if available\n\nIn __skb_udp_tunnel_segment if outer UDP checksums are enabled and\nip_summed is not already CHECKSUM_PARTIAL, set up checksum offload\nif device features allow it.\n\nSigned-off-by: Tom Herbert \u003ctherbert@google.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "63487babf08d6d67483c67ed21d8cea6674a44ec", "tree": "f0e11d54e33a689a25bed43f7bd3c720b49a6bcd", "parents": [ "890b7916d0965829ad1c457aa61f049a210c19f8" ], "author": { "name": "Tom Herbert", "email": "therbert@google.com", "time": "Tue Nov 04 09:06:51 2014 -0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Nov 05 16:30:02 2014 -0500" }, "message": "net: Move fou_build_header into fou.c and refactor\n\nMove fou_build_header out of ip_tunnel.c and into fou.c splitting\nit up into fou_build_header, gue_build_header, and fou_build_udp.\nThis allows for other users for TX of FOU or GUE. Change ip_tunnel_encap\nto call fou_build_header or gue_build_header based on the tunnel\nencapsulation type. Similarly, added fou_encap_hlen and gue_encap_hlen\nfunctions which are called by ip_encap_hlen. New net/fou.h has\nprototypes and defines for this.\n\nAdded NET_FOU_IP_TUNNELS configuration. When this is set, IP tunnels\ncan use FOU/GUE and fou module is also selected.\n\nSigned-off-by: Tom Herbert \u003ctherbert@google.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "46d3802627f60be1c17659a44c8d2d7a5e247023", "tree": "546378e4ce446ed1b9883bb16307a3fb3cc9ca7a", "parents": [ "b994ca6b67ade73c2d69c00a449ad7006cd4546e", "777da230c5b98a6a3793f0525dd99e1e61c8a072" ], "author": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Nov 05 16:23:09 2014 -0500" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Nov 05 16:23:09 2014 -0500" }, "message": "Merge branch \u0027stmmac-net\u0027\n\nGiuseppe Cavallaro says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\nstmmac: review and fix lock and atomicity\n\nRecently some issues have been reported for the driver for locking mechanism\nand atomicity.\n\nIn fact, enabling DEBUG support to prove lock and to verify if sleeping while\natomic context some warnings occur at runtime. I have reproduced all on STi\nplatforms.\n\nConcerning the tx path, I had provided a patch time ago but\nI discarded the idea to completely remove locks; in this patch-set we can have\nsome useful fixes instead of.\n\nThis patch-set is to fix the atomicity in the PM stuff where I tried to collect\nall the points and advice reported in the past weeks.\nAs final result, on my side no warnings and no problem when suspend/resume the\ndriver on STi boxes.\n\nI also added a patch that fixes the locks for the EEE.\nAs pointed in some thread there was a design problem behind the eee\ninitialization and I have tried to fix that before.\nAs final result no issues when proving locks too.\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "777da230c5b98a6a3793f0525dd99e1e61c8a072", "tree": "546378e4ce446ed1b9883bb16307a3fb3cc9ca7a", "parents": [ "4741cf9cecf8af57dc612dd96b0056fa8e2f301d" ], "author": { "name": "Giuseppe CAVALLARO", "email": "peppe.cavallaro@st.com", "time": "Tue Nov 04 17:08:09 2014 +0100" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Nov 05 16:22:57 2014 -0500" }, "message": "stmmac: fix atomicity in pm routines\n\nThis patch is to fix the atomicity when suspend and resume the\ndriver. The clk api have been changed (as reported by Hao Liang)\nand the skb allocation is done out of the hw setup function and\ntaking care about the GFP flags.\n\nReported-by: Hao Liang \u003chliang1025@gmail.com\u003e\nSigned-off-by: Giuseppe Cavallaro \u003cpeppe.cavallaro@st.com\u003e\nCc: Alexey Khoroshilov \u003ckhoroshilov@ispras.ru\u003e\nCc: Hao Liang \u003chliang1025@gmail.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "4741cf9cecf8af57dc612dd96b0056fa8e2f301d", "tree": "21296c83a3e7b9f42f1be40608adf30fc2fffcda", "parents": [ "b9d73704aab92602fcadff26f61462a6445bd0cf" ], "author": { "name": "Giuseppe CAVALLARO", "email": "peppe.cavallaro@st.com", "time": "Tue Nov 04 17:08:08 2014 +0100" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Nov 05 16:22:57 2014 -0500" }, "message": "stmmac: fix concurrency in eee initialization.\n\nThis patch aims to fix the concurrency in eee initialization\ninside the stmmac driver and related warnings when enable\nDEBUG_ATOMIC_SLEEP.\n\nPrior this patch, the stmmac_eee_init could be called in several places\nas shown below:\n\nstmmac_open stmmac_resume PHY Layer\n | | |\n stmmac_hw_setup stmmac_adjust_link\n | | stmmac ethtool\n |__________________________|______________|\n |\n stmmac_eee_init\n\nThe patch removes the stmmac_eee_init call inside the stmmac_hw_setup\nthat is unnecessary. It is sufficient to call it in the adjust_link to\nalways guarantee that EEE is always configured at mac level too.\n\nFixing the lock protection now it is covered another case (not\nconsidered before). The stmmac_eee_init could be called by the ethtool\nso critical sections must be protected inside this function too.\n\nSigned-off-by: Giuseppe Cavallaro \u003cpeppe.cavallaro@st.com\u003e\nCc: Sebastian Andrzej Siewior \u003cbigeasy@linutronix.de\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "b9d73704aab92602fcadff26f61462a6445bd0cf", "tree": "c4fdce909bf102ee486064d2e07199f17a419184", "parents": [ "758a0ab59b9bed75d8c8fcaed3cb41f10a586793" ], "author": { "name": "Giuseppe CAVALLARO", "email": "peppe.cavallaro@st.com", "time": "Tue Nov 04 17:08:07 2014 +0100" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Nov 05 16:22:56 2014 -0500" }, "message": "stmmac: fix lock in stmmac_set_rx_mode\n\nWhen compile with CONFIG_PROVE_LOCKING the following warnings happen:\n\n[snip]\n\n HARDIRQ-ON-W at:\n [\u003cc0480c1c\u003e] _raw_spin_lock+0x3c/0x4c\n [\u003cc02c2828\u003e] stmmac_set_rx_mode+0x18/0x3c\n [\u003cc038b2cc\u003e] dev_set_rx_mode+0x1c/0x28\n [\u003cc038b38c\u003e] __dev_open+0xb4/0xf8\n [\u003cc038b5a8\u003e] __dev_change_flags+0x94/0x128\n [\u003cc038b6a8\u003e] dev_change_flags+0x10/0x48\n [\u003cc062afe0\u003e] ip_auto_config+0x1d4/0x1084\n [\u003cc000873c\u003e] do_one_initcall+0x108/0x15c\n [\u003cc060ec50\u003e] kernel_init_freeable+0x1a8/0x248\n [\u003cc0472cc0\u003e] kernel_init+0x8/0x160\n [\u003cc000dfc8\u003e] ret_from_fork+0x14/0x2c\n INITIAL USE at:\n [\u003cc0480c1c\u003e] _raw_spin_lock+0x3c/0x4c\n [\u003cc02c2828\u003e] stmmac_set_rx_mode+0x18/0x3c\n [\u003cc038b2cc\u003e] dev_set_rx_mode+0x1c/0x28\n [\u003cc038b38c\u003e] __dev_open+0xb4/0xf8\n [\u003cc038b5a8\u003e] __dev_change_flags+0x94/0x128\n [\u003cc038b6a8\u003e] dev_change_flags+0x10/0x48\n [\u003cc062afe0\u003e] ip_auto_config+0x1d4/0x1084\n [\u003cc000873c\u003e] do_one_initcall+0x108/0x15c\n [\u003cc060ec50\u003e] kernel_init_freeable+0x1a8/0x248\n [\u003cc0472cc0\u003e] kernel_init+0x8/0x160\n [\u003cc000dfc8\u003e] ret_from_fork+0x14/0x2c\n\nso the patch just removes the lock protection in the stmmac_set_rx_mode\n\nSigned-off-by: Giuseppe Cavallaro \u003cpeppe.cavallaro@st.com\u003e\nCc: Emilio Lopez \u003cemilio@elopez.com.ar\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" } ], "next": "758a0ab59b9bed75d8c8fcaed3cb41f10a586793" }