- 24 Nov, 2015 6 commits
-
-
Saurabh Sengar authored
in case of error no need to set num_tx and num_rx = 1, because in case of error these variables will remain unchanged by of_property_read_u32 ie 1 only Signed-off-by:
Saurabh Sengar <saurabh.truth@gmail.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Saurabh Sengar authored
this patch does the following: 1 . remove unnecessary if, else condition 2 . reduce one variable 3 . change the return type of 2 functions to void as there return values turn out to be 0 always after above changes Signed-off-by:
-
David S. Miller authored
Merge tag 'linux-can-next-for-4.5-20151123' of git://git.kernel.org/pub/scm/linux/kernel/git/mkl/linux-can-next Marc Kleine-Budde says: ==================== pull-request: can-next 2015-11-23 this is a pull request of a single patch for net-next/master. The patch by Kedareswara rao Appana converts the xilinx CAN driver to runtime_pm. ==================== Signed-off-by:
-
Geert Uytterhoeven authored
Commit 09605cc1 ("net ipv4: use preferred log methods") replaced a few calls of pr_cont() after a console print without a trailing newline by pr_info(), causing lines to be split during IP autoconfiguration, like: . , OK IP-Config: Got DHCP answer from 192.168.97.254, my address is 192.168.97.44 Convert these back to using pr_cont(), so it prints again: ., OK IP-Config: Got DHCP answer from 192.168.97.254, my address is 192.168.97.44 Absorb the printing of "my address ..." into the previous call to pr_info(), as there's no reason to use a continuation there. Convert one more pr_info() to print nameservers while we're at it. Fixes: 09605cc1 ("net ipv4: use preferred log methods") Signed-off-by:
Geert Uytterhoeven <geert+renesas@glider.be> Signed-off-by:
-
Arnd Bergmann authored
After the introduction of the switch gpio reset API, I'm getting build errors in configurations that disable CONFIG_GPIOLIB: net/dsa/dsa.c:783:16: error: implicit declaration of function 'gpio_to_desc' [-Werror=implicit-function-declaration] The reason is that linux/gpio/consumer.h is not automatically included without gpiolib support. This adds an explicit #include statement to make it compile in all configurations. The reset functionality will not work without gpiolib, which is what you get when disabling the feature. As far as I can tell, gpiolib is supported on all architectures on which you can have DSA at the moment. Signed-off-by:
Arnd Bergmann <arnd@arndb.de> Fixes: cc30c163 ("net: dsa: Add support for a switch reset gpio") Acked-by:
Andrew Lunn <andrew@lunn.ch> Signed-off-by:
-
git://git.kernel.org/pub/scm/linux/kernel/git/jkirsher/next-queueDavid S. Miller authored
Jeff Kirsher says: ==================== Intel Wired LAN Driver Updates 2015-11-23 This series contains updates to ixgbe, ixgbevf, fm10k, i40e and i40evf. Jacob fixes an issue where VF could attempt to read queues it does not own, so prevent this we check queue 0 before we continue. Matthew fixes the MTU for jumbo frames for fm10k. Julia Lawall cleans up a unneeded NULL test in ixgbe. Mark cleans up a redundant header inclusion. Adds KR mode support for CS4227 chip. Cleaned up diagnostic code, which is no longer needed, for the CS4227 chip. Jean Sacren fixes kernel documentation for ixgbe. Alex Duyck fixes an fm10k and ixgbe issue in which the polling routine would increase the budget for receive to at least 1 per queue if multiple queues were present. This would result in receive packets being processed when the budget was 0 which is meant to indicate that no receive can be handled. Also fixes an ixgbevf performance issue where netperf test will starve for memory in the time form one transmit interrupt to the next, so limit lowest interrupt rate for adaptive interrupt moderation to 12K. Fixed up ixgbe and ixgbevf to use napi_schedule_irqoff() where the drivers were run from hard interrupt context or with interrupts already disabled in netpoll. Jesse fixes a compiler warning about an unused variable for i40evf. John Greene fixes an issue with ixgbevf, where if the VF driver is loaded while the corresponding PH interface is down, the driver assigns a random MAC address, can be overwritten with the value of hw->mac.perm_addr which is 0 at that point. So avoid this case by initializing hw->mac.perm_addr to the randomly generated address and do not set it unless we receive an ACK from ixgbe. Rasmus Villemoes cleans up some confusing code in i40e debugfs code. ==================== Signed-off-by:
-
- 23 Nov, 2015 34 commits
-
-
Julia Lawall authored
The vmci_transport_notify_ops structures are never modified, so declare them as const. Done with the help of Coccinelle. Signed-off-by:
Julia Lawall <Julia.Lawall@lip6.fr> Signed-off-by:
-
Julia Lawall authored
The in_cache_ops and eg_cache_ops structures are never modified, so declare them as const. Done with the help of Coccinelle. Signed-off-by:
-
David S. Miller authored
Nikolay Aleksandrov says: ==================== net: ipmr: cleanups and minor improvements Since I'll have to work with ipmr, I decided to clean it up and do some minor improvements. Functionally there're almost no changes except the SLAB_PANIC removal. Most of the patches just re-design some functions to be clearer and more concise and try to remove the ifdef web that was inside. There's more information in each commit. This is the first set, the end goal is to introduce complete netlink support and control over the mfc and vif devices. I've tried to test all of the setsockopt/getsockopt options, and also made builds with various ipmr kconfig options turned on and off. v2: change patch 7 to keep SLAB_PANIC and just drop the unnecessary null check ==================== Signed-off-by: -
Nikolay Aleksandrov authored
Factor out common vif init code used in both tunnel and pimreg initialization and create ipmr_init_vif_indev() function. Signed-off-by:
Nikolay Aleksandrov <nikolay@cumulusnetworks.com> Signed-off-by:
-
Nikolay Aleksandrov authored
Take rtnl in the beginning unconditionally as most options already need it (one exception - MRT_DONE, see the comment inside), make the lock/unlock places central and move out the switch() local variables. Signed-off-by:
-
Nikolay Aleksandrov authored
It's not necessary to check for null as SLAB_PANIC is used and we'll panic if the alloc fails, so just drop it. Signed-off-by:
-
Nikolay Aleksandrov authored
Trivial replace of ifdef with IS_BUILTIN(). Signed-off-by:
-
Nikolay Aleksandrov authored
Use a switch to determine if optname is correct and set val accordingly. This produces a much more straight-forward and readable code. Signed-off-by:
-
Nikolay Aleksandrov authored
Trivial code and comment style fixes, also removed some extra newlines, spaces and tabs. Signed-off-by:
-
Nikolay Aleksandrov authored
Add the helper pimsm_enabled() which replaces the old CONFIG_IP_PIMSM define and is used to check if any version of PIM-SM has been enabled. Use a single if defined(CONFIG_IP_PIMSM_V1) || defined(CONFIG_IP_PIMSM_V2) for the pim-sm shared code. This is okay w.r.t IGMPMSG_WHOLEPKT because only a VIFF_REGISTER device can send such packet, and it can't be created if pimsm_enabled() is false. Signed-off-by:
-
Nikolay Aleksandrov authored
Before mroute_reg_vif_num was defined only if any of the CONFIG_PIMSM_ options were set, but that's not really necessary as the size of the struct is the same in both cases (checked with pahole, both cases size is 3256 bytes) and we can remove some unnecessary ifdefs to simplify the code. Signed-off-by:
-
Nikolay Aleksandrov authored
Move the table id check in ipmr_new_table and make it return error pointer. We need this change for the upcoming netlink table manipulation support in order to avoid code duplication and a race condition. Signed-off-by:
-
Rasmus Villemoes authored
This code is pretty confused. The variable name 'bytes_not_copied' clearly indicates that the programmer knew the semantics of copy_{to,from}_user, but then the return value is checked for being negative and used as a -Exxx return value. I'm not sure this is the proper fix, but at least we get rid of the dead code which pretended to check for access faults. Signed-off-by:Rasmus Villemoes <linux@rasmusvillemoes.dk> Acked-by:
Shannon Nelson <shannon.nelson@intel.com> Signed-off-by:
Jeff Kirsher <jeffrey.t.kirsher@intel.com>
-
Emil Tantilov authored
If ixgbevf is loaded while the corresponding PF interface is down and the driver assigns a random MAC address, that address can be overwritten with the value of hw->mac.perm_addr, which would be 0 at that point. To avoid this case we init hw->mac.perm_addr to the randomly generated address and do not set it unless we receive ACK from ixgbe. Reported-by:
John Greene <jogreene@redhat.com> Signed-off-by:
Emil Tantilov <emil.s.tantilov@intel.com> Tested-by:
Krishneil Singh <Krishneil.k.singh@intel.com> Signed-off-by:
-
Emil Tantilov authored
replace some instances of memcpy for setting up the mac address with ether_addr_copy() Signed-off-by:
Darin Miller <darin.j.miller@intel.com> Signed-off-by:
-
Jesse Brandeburg authored
Compiler complained of an unused variable, which the driver was just using to store the result of a rd32 which is used to clear a register unconditionally. Just drop the unused variable and re-use one. Signed-off-by:
Jesse Brandeburg <jesse.brandeburg@intel.com> Tested-by:
Andrew Bowers <andrewx.bowers@intel.com> Signed-off-by:
-
Mark Rustad authored
Testing has now shown that the diagnostic code used with the CS4227 is no longer needed, so remove it. Signed-off-by:
Mark Rustad <mark.d.rustad@intel.com> Tested-by:
-
Alexander Duyck authored
The ixgbe_intr and ixgbe/ixgbevf_msix_clean_rings functions run from hard interrupt context or with interrupts already disabled in netpoll. They can use napi_schedule_irqoff() instead of napi_schedule() Signed-off-by:
Alexander Duyck <aduyck@mirantis.com> Tested-by:
-
Alexander Duyck authored
This patch is the ixgbevf version of commit 8ac34f10 "ixgbe: Limit lowest interrupt rate for adaptive interrupt moderation to 12K" The same logic applies here as well as the same results since a netperf test will starve for memory in the time from one Tx interrupt to the next. As a result the ixgbevf driver underperformed when compared to vhost_net. Signed-off-by:
-
Mark Rustad authored
KR auto-neg mode is what we will be using going forward. The SW interface for this mode is different that what was used for iXFI. Signed-off-by:
Phil Schmitt <phillip.j.schmitt@intel.com> Signed-off-by:
-
Alexander Duyck authored
This patch corrects an issue in which the polling routine would increase the budget for Rx to at least 1 per queue if multiple queues were present. This would result in Rx packets being processed when the budget was 0 which is meant to indicate that no Rx can be handled. Signed-off-by:
-
Alexander Duyck authored
This patch corrects an issue in which the polling routine would increase the budget for Rx to at least 1 per queue if multiple queues were present. This would result in Rx packets being processed when the budget was 0 which is meant to indicate that no Rx can be handled. Signed-off-by:
-
Jean Sacren authored
The commit dfaf891d ("ixgbe: Refactor the RSS configuration code") introduced a few kernel-doc errors: 1) The function name is missing; 2) The format is wrong; 3) The short description is redundant. Fix all the above for the correct execution of the kernel doc. Signed-off-by:
Jean Sacren <sakiwit@gmail.com> Tested-by:
-
Mark Rustad authored
Delete a redundant include of net/vxlan.h. Signed-off-by:
-
Julia Lawall authored
Remove unneeded NULL test. The semantic patch that makes this change is as follows: (http://coccinelle.lip6.fr/) // <smpl> @@ expression x; @@ -if (x != NULL) \(kmem_cache_destroy\|mempool_destroy\|dma_pool_destroy\)(x); // </smpl> Signed-off-by:
-
Jacob Keller authored
Based on hardware testing, the host interface supports up to 15368 bytes as the maximum frame size. To determine the correct MTU, we subtract 8 for the internal switch tag, 14 for the L2 header, and 4 for the appended FCS header, resulting in 15342 bytes of payload for our maximum MTU on jumbo frames. Signed-off-by:
Matthew Vick <matthew.vick@intel.com> Signed-off-by:
Jacob Keller <jacob.e.keller@intel.com> Acked-by:
Bruce Allan <bruce.w.allan@intel.com> Tested-by:
-
Jacob Keller authored
It is possible that the PF has not yet assigned resources to the VF. Although rare, this could result in the VF attempting to read queues it does not own and result in FUM or THI faults in the PF. To prevent this, check queue 0 before we continue in init_hw_vf. Signed-off-by:
-
David S. Miller authored
Phil Sutter says: ==================== improve fault-tolerance of rhashtable runtime-test The following series aims to improve lib/test_rhashtable in different situations: Patch 1 allows the kernel to reschedule so the test does not block too long on slow systems. Patch 2 fixes behaviour under pressure, retrying inserts in non-permanent error case (-EBUSY). Patch 3 auto-adjusts the upper table size limit according to the number of threads (in concurrency test). In fact, the current default is already too small. Patch 4 makes it possible to retry inserts even in supposedly permanent error case (-ENOMEM) to expose rhashtable's remaining problem of -ENOMEM being not as permanent as it is expected to be. Changes since v1: - Introduce insert_retry() which is then used in single-threaded test as well. - Do not retry inserts by default if -ENOMEM was returned. - Rename the retry counter to be a bit more verbose about what it contains. - Add patch 4 as a debugging aid. ==================== Signed-off-by: -
Phil Sutter authored
This is rather a hack to expose the current issue with rhashtable to under high pressure sometimes return -ENOMEM even though system memory is not exhausted and a consecutive insert may succeed. Signed-off-by:
Phil Sutter <phil@nwl.cc> Signed-off-by:
-
Phil Sutter authored
A maximum table size of 64k entries is insufficient for the multiple threads test even in default configuration (10 threads * 50000 objects = 500000 objects in total). Since we know how many objects will be inserted, calculate the max size unless overridden by parameter. Note that specifying the exact number of objects upon table init won't suffice as that value is being rounded down to the next power of two - anticipate this by rounding up to the next power of two in beforehand. Signed-off-by:
-
Phil Sutter authored
After adding cond_resched() calls to threadfunc(), a surprisingly high rate of insert failures occurred probably due to table resizes getting a better chance to run in background. To not soften up the remaining tests, retry inserts until they either succeed or fail permanently. Also change the non-threaded test to retry insert operations, too. Suggested-by:
Thomas Graf <tgraf@suug.ch> Signed-off-by:
-
Phil Sutter authored
This should fix for soft lockup bugs triggered on slow systems. Signed-off-by:
-
David S. Miller authored
Andrew Lunn says: ==================== DSA: GPIO to reset switches These two patches add support for using a GPIO to hard reset a switch during reset. v2: Thanks to a clue from Neil Armstrong, i figured out how to convert the gpio into a gpiod, while keeping the ACTIVE_LOW flag, so simplifiying the set/reset code. I have not included the Tested-by: from Phil Reid, since i made a lot of changes. ==================== Signed-off-by: -
Andrew Lunn authored
The device tree binding now allows a gpio to be specified which is attached to the switch chips reset line. If it is defined, perform a hardware reset on the switch during setup. Signed-off-by:
-
