- 01 Oct, 2018 3 commits
-
-
Gavi Teitz authored
Enabled checksum and TSO offloads for the representors, in order to increase their performance, which is required to increase the performance of flows that cannot be offloaded. Checksum offloads contribute to a general acceleration of all traffic (to around 150%), whereas the TSO offload contributes to a prominent acceleration of the representor's TX for traffic flows with larger than MTU sized packets (to around 200%). This is the usual case for TCP streams, as the PF, which serves as the uplink representor, and the VF representors employ GRO before forwarding the packets to the representor. GRO was enabled implicitly for the representors beforehand, and is explicitly enabled here to ensure that the representors preserve the performance boost it provides (of around 200%) when working in tandem with the TSO offload by the forwardee, which is the standard case as both the PF and the VF representors employ HW TSO. The impact of these changes can be seen in the following measurements taken on a setup of a VM over a VF, connected to OVS via the VF representor, to an external host: Before current changes: TCP Throughput [Gb/s] External host to VM ~ 10.5 VM to external host ~ 23.5 With just checksum offloads enabled: TCP Throughput [Gb/s] External host to VM ~ 14.9 VM to external host ~ 28.5 With the TSO offload also enabled: TCP Throughput [Gb/s] External host to VM ~ 30.5 Signed-off-by:
Gavi Teitz <gavi@mellanox.com> Reviewed-by:
Or Gerlitz <ogerlitz@mellanox.com> Signed-off-by:
Saeed Mahameed <saeedm@mellanox.com>
-
Gavi Teitz authored
The representors' RQ size was not large enough for them to achieve high enough performance, and therefore needed to be enlarged, while suffering a minimum hit to its memory usage. To achieve this the representors RQ size was increased, and its type was changed to be a striding RQ if it is supported. Towards that goal the following changes were made: * Extracted the sequence for setting the standard netdev's RQ parmas into a function * Replaced the sequence for setting the representor's RQ params with the standard sequence The impact of this change can be seen in the following measurements taken on a setup of a VM over a VF, connected to OVS via the VF representor, to an external host: Before current change: TCP Throughput [Gb/s] VM to external host ~ 7.2 With the current change (measured with a striding RQ): TCP Throughput [Gb/s] VM to external host ~ 23.5 Each representor now consumes 2 [MB] of memory for its packet buffers. Signed-off-by:
Gavi Teitz <gavi@mellanox.com> Reviewed-by:
Or Gerlitz <ogerlitz@mellanox.com> Signed-off-by:
Saeed Mahameed <saeedm@mellanox.com>
-
Or Gerlitz authored
Allow using partial masks for L3 addresses and L4 ports across the place. Signed-off-by:
Or Gerlitz <ogerlitz@mellanox.com> Signed-off-by:
Saeed Mahameed <saeedm@mellanox.com>
-
- 29 Sep, 2018 11 commits
-
-
Yifeng Sun authored
This patch fixes the bug that all datapath and vport ops are returning wrong values (OVS_FLOW_CMD_NEW or OVS_DP_CMD_NEW) in their replies. Signed-off-by:
Yifeng Sun <pkusunyifeng@gmail.com> Acked-by:
Pravin B Shelar <pshelar@ovn.org> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Vakul Garg authored
Structure 'tls_rec' contains sg_aead_in and sg_aead_out which point to a aad_space and then chain scatterlists sg_plaintext_data, sg_encrypted_data respectively. Rather than using chained scatterlists for plaintext and encrypted data in aead_req, it is efficient to store aad_space in sg_encrypted_data and sg_plaintext_data itself in the first index and get rid of sg_aead_in, sg_aead_in and further chaining. This requires increasing size of sg_encrypted_data & sg_plaintext_data arrarys by 1 to accommodate entry for aad_space. The code which uses sg_encrypted_data and sg_plaintext_data has been modified to skip first index as it points to aad_space. Signed-off-by:
Vakul Garg <vakul.garg@nxp.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
David S. Miller authored
Jon Maloy says: ==================== tipc: make connection setup more robust In this series we make a few improvements to the connection setup and probing mechanism, culminating in the last commit where we make it possible for a client socket to make multiple setup attempts in case it encounters receive buffer overflow at the listener socket. ==================== Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Tung Nguyen authored
Default socket receive buffer size for a listener socket is 2Mb. For each arriving empty SYN, the linux kernel allocates a 768 bytes buffer. This means that a listener socket can serve maximum 2700 simultaneous empty connection setup requests before it hits a receive buffer overflow, and much fewer if the SYN is carrying any significant amount of data. When this happens the setup request is rejected, and the client receives an ECONNREFUSED error. This commit mitigates this problem by letting the client socket try to retransmit the SYN message multiple times when it sees it rejected with the code TIPC_ERR_OVERLOAD. Retransmission is done at random intervals in the range of [100 ms, setup_timeout / 4], as many times as there is room for within the setup timeout limit. Signed-off-by:
Tung Nguyen <tung.q.nguyen@dektech.com.au> Acked-by:
Ying Xue <ying.xue@windriver.com> Signed-off-by:
Jon Maloy <jon.maloy@ericsson.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Jon Maloy authored
Messages intended for intitating a connection are currently indistinguishable from regular datagram messages. The TIPC protocol specification defines bit 17 in word 0 as a SYN bit to allow sanity check of such messages in the listening socket, but this has so far never been implemented. We do that in this commit. Acked-by:
Ying Xue <ying.xue@windriver.com> Signed-off-by:
Jon Maloy <jon.maloy@ericsson.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Jon Maloy authored
We refactor the function tipc_sk_filter_connect(), both to make it more readable and as a preparation for the next commit. Acked-by:
Ying Xue <ying.xue@windriver.com> Signed-off-by:
Jon Maloy <jon.maloy@ericsson.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Jon Maloy authored
We refactor this function as a preparation for the coming commits in the same series. Acked-by:
Ying Xue <ying.xue@windriver.com> Signed-off-by:
Jon Maloy <jon.maloy@ericsson.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Jon Maloy authored
The function tipc_msg_reverse() is reversing the header of a message while reusing the original buffer. We have seen at several occasions that this may have unfortunate side effects when the buffer to be reversed is a clone. In one of the following commits we will again need to reverse cloned buffers, so this is the right time to permanently eliminate this problem. In this commit we let the said function always consume the original buffer and replace it with a new one when applicable. Acked-by:
Ying Xue <ying.xue@windriver.com> Signed-off-by:
Jon Maloy <jon.maloy@ericsson.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Yuchung Cheng authored
Previously TCP initial receive buffer is ~87KB by default and the initial receive window is ~29KB (20 MSS). This patch changes the two numbers to 128KB and ~64KB (rounding down to the multiples of MSS) respectively. The patch also simplifies the calculations s.t. the two numbers are directly controlled by sysctl tcp_rmem[1]: 1) Initial receiver buffer budget (sk_rcvbuf): while this should be configured via sysctl tcp_rmem[1], previously tcp_fixup_rcvbuf() always override and set a larger size when a new connection establishes. 2) Initial receive window in SYN: previously it is set to 20 packets if MSS <= 1460. The number 20 was based on the initial congestion window of 10: the receiver needs twice amount to avoid being limited by the receive window upon out-of-order delivery in the first window burst. But since this only applies if the receiving MSS <= 1460, connection using large MTU (e.g. to utilize receiver zero-copy) may be limited by the receive window. With this patch TCP memory configuration is more straight-forward and more properly sized to modern high-speed networks by default. Several popular stacks have been announcing 64KB rwin in SYNs as well. Signed-off-by:
Yuchung Cheng <ycheng@google.com> Signed-off-by:
Wei Wang <weiwan@google.com> Signed-off-by:
Neal Cardwell <ncardwell@google.com> Signed-off-by:
Eric Dumazet <edumazet@google.com> Reviewed-by:
Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
David S. Miller authored
drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c: In function ‘hclge_get_sset_count’: drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c:496:31: error: ‘HNAE3_REVISION_ID_21’ undeclared (first use in this function); did you mean ‘FADT2_REVISION_ID’? if (hdev->pdev->revision >= HNAE3_REVISION_ID_21 || ^~~~~~~~~~~~~~~~~~~~ FADT2_REVISION_ID Signed-off-by:
David S. Miller <davem@davemloft.net>
-
David S. Miller authored
drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c: In function ‘hns3_self_test’: drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c:278:15: error: ‘HNS3_SELF_TEST_TYPE_NUM’ undeclared (first use in this function); did you mean ‘HNS3_SELF_TEST_TPYE_NUM’? int st_param[HNS3_SELF_TEST_TYPE_NUM][2]; ^~~~~~~~~~~~~~~~~~~~~~~ HNS3_SELF_TEST_TPYE_NUM Signed-off-by:
David S. Miller <davem@davemloft.net>
-
- 28 Sep, 2018 26 commits
-
-
Colin Ian King authored
Trivial fix to spelling mistake struct field name, rename it. Signed-off-by:
Colin Ian King <colin.king@canonical.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
git://git.kernel.org/pub/scm/linux/kernel/git/jkirsher/next-queueDavid S. Miller authored
Jeff Kirsher says: ==================== 100GbE Intel Wired LAN Driver Updates 2018-09-27 This series contains fixes to the ice driver only. Jake fixes a potential crash due to attempting to access the mutex which is already destroyed. Fix this by using rq.count and sq.count to determine if the queue was initialized. Fixed the current logic for checking the firmware version to properly handle situations when firmware major/minor versions differ and when the branch version differs. Bruce replaces a memcpy() with a direct assignment, which is preferred. Also updated the branding strings and device ids supported by the driver. Fixed the "ethtool -G" command in the driver, which was always returning EINVAL when changing the descriptor ring size. Brett update and clarified code comments. Anirudh updates the driver to ensure we query the firmware for the transmit scheduler node information before adding it to the driver database, to ensure we have the current information. Also update the "get capabilities" command to get device and function capabilities. ==================== Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Wei Yongjun authored
Fixes the following sparse warning: net/sched/sch_generic.c:944:6: warning: symbol 'qdisc_free_cb' was not declared. Should it be static? Fixes: 3a7d0d07 ("net: sched: extend Qdisc with rcu") Signed-off-by:
Wei Yongjun <weiyongjun1@huawei.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Nikolay Aleksandrov authored
This test adds an fdb entry with the sticky flag and sends traffic from a different port with the same mac as a source address expecting the entry to not change ports if the flag is operating correctly. Signed-off-by:
Nikolay Aleksandrov <nikolay@cumulusnetworks.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Nikolay Aleksandrov authored
We need to explicitly zero is_sticky when creating a new fdb, otherwise we might get a stale value for a new entry. Fixes: 435f2e7c ("net: bridge: add support for sticky fdb entries") Signed-off-by:
Nikolay Aleksandrov <nikolay@cumulusnetworks.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
David S. Miller authored
Salil Mehta says: ==================== Cleanups, minor additions & fixes for HNS3 driver This patch-set contains cleans-ups, minor changes and fixes to the HNS3 driver. ==================== Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Huazhong Tan authored
The user's coal configuration will be lost after reset, so the tx_coal and rx_coal fields are added to the struct hns_nic_priv to save the coal configuration and used to restore the user's configuration after the reset is complete. Fixes: bb6b94a8 ("net: hns3: Add reset interface implementation in client") Signed-off-by:
Huazhong Tan <tanhuazhong@huawei.com> Signed-off-by:
Yunsheng Lin <linyunsheng@huawei.com> Signed-off-by:
Salil Mehta <salil.mehta@huawei.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Huazhong Tan authored
The current hns3_get_max_available_channels returns the total number of queues for the device, which makes ethtool -L set the number of queues per channel queues incorrectly, so hns3_get_max_available_channels should return the maximum available number of queues per channel, depending on the total number of queues allocated and the hardware configurations. Signed-off-by:
Huazhong Tan <tanhuazhong@huawei.com> Signed-off-by:
Yunsheng Lin <linyunsheng@huawei.com> Signed-off-by:
Peng Li <lipeng321@huawei.com> Signed-off-by:
Salil Mehta <salil.mehta@huawei.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Huazhong Tan authored
hclge_tm_schd_info_update should return an error when num_tc is greater than alloc_tqps. This patch changes the return type of hnae3_register_ae_algo from void to int. Signed-off-by:
Huazhong Tan <tanhuazhong@huawei.com> Signed-off-by:
Yunsheng Lin <linyunsheng@huawei.com> Signed-off-by:
Peng Li <lipeng321@huawei.com> Signed-off-by:
Salil Mehta <salil.mehta@huawei.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Yunsheng Lin authored
Currently hns3_nic_change_mtu will try to down the netdev before setting mtu, and it does not up the netdev when the setting fails, which causes netdev not up problem. This patch fixes it by not returning when the setting fails. Fixes: a8e8b7ff ("net: hns3: Add support to change MTU in HNS3 hardware") Signed-off-by:
Yunsheng Lin <linyunsheng@huawei.com> Signed-off-by:
Peng Li <lipeng321@huawei.com> Signed-off-by:
Salil Mehta <salil.mehta@huawei.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Yunsheng Lin authored
The hardware expects a unit of 128 bytes when setting packet buffer. When calculating the packet buffer size, hclge_rx_buffer_calc does not round up the size as a unit of 128 byte, which may casue packet lost problem when stress testing. This patch fixes it by rounding up packet size when calculating. Fixes: 46a3df9f ("net: hns3: Add HNS3 Acceleration Engine & Compatibility Layer Support") Signed-off-by:
Yunsheng Lin <linyunsheng@huawei.com> Signed-off-by:
Peng Li <lipeng321@huawei.com> Signed-off-by:
Salil Mehta <salil.mehta@huawei.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Fuyun Liang authored
This patch adds serdes parallel inner loopback support for self test. Signed-off-by:
Fuyun Liang <liangfuyun1@huawei.com> Signed-off-by:
Peng Li <lipeng321@huawei.com> Signed-off-by:
Salil Mehta <salil.mehta@huawei.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Fuyun Liang authored
In fact, our implementation of mac loopback is the implementation of app loopback now. Current name is wrong. This patch renames mac loopback to app loopback. Signed-off-by:
Fuyun Liang <liangfuyun1@huawei.com> Signed-off-by:
Peng Li <lipeng321@huawei.com> Signed-off-by:
Salil Mehta <salil.mehta@huawei.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Fuyun Liang authored
Our loop mode includes mac loop, serdes loop and phy loop. Not all of them are related with mac. This patch corrects their names. Signed-off-by:
Fuyun Liang <liangfuyun1@huawei.com> Signed-off-by:
Peng Li <lipeng321@huawei.com> Signed-off-by:
Salil Mehta <salil.mehta@huawei.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Fuyun Liang authored
The extra mac address of pause param is used to do double check for pause frame. This patch set it to HW. If we do not do that, pfc pause frame will be transferred protocol stack when normal flow control mode is enabled. Signed-off-by:
Fuyun Liang <liangfuyun1@huawei.com> Signed-off-by:
Peng Li <lipeng321@huawei.com> Signed-off-by:
Salil Mehta <salil.mehta@huawei.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Peng Li authored
This patch adds support for sctp checksum offload. Signed-off-by:
Fuyun Liang <liangfuyun1@huawei.com> Signed-off-by:
Peng Li <lipeng321@huawei.com> Signed-off-by:
Salil Mehta <salil.mehta@huawei.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Arnd Bergmann authored
When B53_SERDES is a loadable module, a built-in srab driver still cannot reach it, so the previous fix is incomplete: b53_srab.c:(.text+0x3f4): undefined reference to `b53_serdes_init' drivers/net/dsa/b53/b53_srab.o:(.rodata+0xe64): undefined reference to `b53_serdes_link_state' drivers/net/dsa/b53/b53_srab.o:(.rodata+0xe74): undefined reference to `b53_serdes_link_set' drivers/net/dsa/b53/b53_srab.o:(.rodata+0xe88): undefined reference to `b53_serdes_an_restart' drivers/net/dsa/b53/b53_srab.o:(.rodata+0xea0): undefined reference to `b53_serdes_phylink_validate' drivers/net/dsa/b53/b53_srab.o:(.rodata+0xea4): undefined reference to `b53_serdes_config' Add a Kconfig dependency that forces srab to also be a module in this case, but allow it to be built-in when serdes is disabled or built-in. Fixes: 7a8c7f5c ("net: dsa: b53: Fix build with B53_SRAB enabled and not B53_SERDES") Signed-off-by:
Arnd Bergmann <arnd@arndb.de> Acked-by:
Florian Fainelli <f.fainelli@gmail.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Arnd Bergmann authored
The newly added runtime-pm support causes a harmless warning when CONFIG_PM is disabled: drivers/net/phy/mdio-bcm-unimac.c:330:12: error: 'unimac_mdio_resume' defined but not used [-Werror=unused-function] static int unimac_mdio_resume(struct device *d) drivers/net/phy/mdio-bcm-unimac.c:321:12: error: 'unimac_mdio_suspend' defined but not used [-Werror=unused-function] static int unimac_mdio_suspend(struct device *d) Marking the functions as __maybe_unused is the easiest workaround and avoids adding #ifdef checks. Fixes: b78ac6ec ("net: phy: mdio-bcm-unimac: Allow configuring MDIO clock divider") Signed-off-by:
Arnd Bergmann <arnd@arndb.de> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Arnd Bergmann authored
gcc points out that the length of the temporary buffer may not be sufficient for large numbers of leds: drivers/net/phy/mscc.c: In function 'vsc85xx_probe': drivers/net/phy/mscc.c:460:45: error: '-mode' directive writing 5 bytes into a region of size between 0 and 9 [-Werror=format-overflow=] ret = sprintf(led_dt_prop, "vsc8531,led-%d-mode", i); ^~~~~ drivers/net/phy/mscc.c:460:9: note: 'sprintf' output between 19 and 28 bytes into a destination of size 22 ret = sprintf(led_dt_prop, "vsc8531,led-%d-mode", i); ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ While we can make a reasonable assumption that the number of LEDs is small, the cost of making the buffer a little bigger is insignificant as well. Fixes: 11bfdabb ("net: phy: mscc: factorize code for LEDs mode") Signed-off-by:
Arnd Bergmann <arnd@arndb.de> Reviewed-by:
Alexandre Belloni <alexandre.belloni@bootlin.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Arnd Bergmann authored
The patch that removed the only users of the oldadv/newadv variables accidentally left the now-unused declarations behind: drivers/net/ethernet/freescale/dpaa/dpaa_ethtool.c: In function 'dpaa_set_pauseparam': drivers/net/ethernet/freescale/dpaa/dpaa_ethtool.c:185:14: error: unused variable 'oldadv' [-Werror=unused-variable] drivers/net/ethernet/freescale/dpaa/dpaa_ethtool.c:185:6: error: unused variable 'newadv' [-Werror=unused-variable] Fixes: 70814e81 ("net: ethernet: Add helper for set_pauseparam for Asym Pause") Signed-off-by:
Arnd Bergmann <arnd@arndb.de> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Wei Yongjun authored
Fixes the following sparse warning: drivers/net/ethernet/aquantia/atlantic/hw_atl/hw_atl_utils.c:873:5: warning: symbol 'aq_fw1x_set_power' was not declared. Should it be static? Fixes: a0da96c0 ("net: aquantia: implement WOL support") Signed-off-by:
Wei Yongjun <weiyongjun1@huawei.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Wei Yongjun authored
Fixes the following sparse warning: net/tls/tls_sw.c:655:16: warning: symbol 'get_rec' was not declared. Should it be static? Fixes: a42055e8 ("net/tls: Add support for async encryption of records for performance") Signed-off-by:
Wei Yongjun <weiyongjun1@huawei.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Wei Yongjun authored
Fixes the following sparse warning: net/core/gen_stats.c:166:1: warning: symbol '___gnet_stats_copy_basic' was not declared. Should it be static? Fixes: 5e111210 ("net/core: Add new basic hardware counter") Signed-off-by:
Wei Yongjun <weiyongjun1@huawei.com> Acked-by:
Eelco Chaudron <echaudro@redhat.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
YueHaibing authored
The method ndo_start_xmit() is defined as returning an 'netdev_tx_t', which is a typedef for an enum type, so make sure the implementation in this driver has returns 'netdev_tx_t' value, and change the function return type to netdev_tx_t. Found by coccinelle. Signed-off-by:
YueHaibing <yuehaibing@huawei.com> Acked-by:
Wei Liu <wei.liu2@citrix.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
YueHaibing authored
Fixes gcc '-Wunused-but-set-variable' warning: drivers/net/ethernet/qlogic/qed/qed_ooo.c: In function 'qed_ooo_delete_isles': drivers/net/ethernet/qlogic/qed/qed_ooo.c:354:30: warning: variable 'p_archipelago' set but not used [-Wunused-but-set-variable] drivers/net/ethernet/qlogic/qed/qed_ooo.c: In function 'qed_ooo_join_isles': drivers/net/ethernet/qlogic/qed/qed_ooo.c:463:30: warning: variable 'p_archipelago' set but not used [-Wunused-but-set-variable] Since commit 1eec2437 ("qed: Make OOO archipelagos into an array"), 'p_archipelago' is no longer in use. Signed-off-by:
YueHaibing <yuehaibing@huawei.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
YueHaibing authored
The method ndo_start_xmit() is defined as returning an 'netdev_tx_t', which is a typedef for an enum type, so make sure the implementation in this driver has returns 'netdev_tx_t' value, and change the function return type to netdev_tx_t. Found by coccinelle. Signed-off-by:
YueHaibing <yuehaibing@huawei.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-