summaryrefslogtreecommitdiffstats
path: root/net
Commit message (Collapse)AuthorAgeFilesLines
* net: skbuff: don't zero tc members when freeing skbFlorian Westphal2015-01-021-7/+0
| | | | | | | | | | | | | | | Not needed, only four cases: - kfree_skb (or one of its aliases). Don't need to zero, memory will be freed. - kfree_skb_partial and head was stolen: memory will be freed. - skb_morph: The skb header fields (including tc ones) will be copied over from the 'to-be-morphed' skb right after skb_release_head_state returns. - skb_segment: Same as before, all the skb header fields are copied over from the original skb right away. Signed-off-by: Florian Westphal <fw@strlen.de> Signed-off-by: David S. Miller <davem@davemloft.net>
* Merge branch 'for-upstream' of ↵David S. Miller2015-01-0224-1687/+2869
|\ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | git://git.kernel.org/pub/scm/linux/kernel/git/bluetooth/bluetooth-next Johan Hedberg say: ==================== pull request: bluetooth-next 2014-12-31 Here's the first batch of bluetooth patches for 3.20. - Cleanups & fixes to ieee802154 drivers - Fix synchronization of mgmt commands with respective HCI commands - Add self-tests for LE pairing crypto functionality - Remove 'BlueFritz!' specific handling from core using a new quirk flag - Public address configuration support for ath3012 - Refactor debugfs support into a dedicated file - Initial support for LE Data Length Extension feature from Bluetooth 4.2 Please let me know if there are any issues pulling. Thanks. ==================== Signed-off-by: David S. Miller <davem@davemloft.net>
| * Bluetooth: Add timing information to ECDH test case runsMarcel Holtmann2014-12-301-1/+9
| | | | | | | | | | | | | | | | After successful completion of the ECDH test cases, print the time it took to run them. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
| * Bluetooth: Add timing information to SMP test case runsMarcel Holtmann2014-12-301-1/+9
| | | | | | | | | | | | | | | | After successful completion of the SMP test cases, print the time it took to run them. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
| * Bluetooth: Add LE Secure Connections tests for SMPJohan Hedberg2014-12-301-0/+186
| | | | | | | | | | | | | | | | This patch adds SMP self-tests for the Secure Connections crypto functions. The sample data has been taken from the core specification. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
| * Bluetooth: Add legacy SMP testsJohan Hedberg2014-12-301-0/+95
| | | | | | | | | | | | | | | | This patch adds self-tests for legacy SMP crypto functions. The sample data has been taken from the core specification. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
| * Bluetooth: Add skeleton for SMP self-testsJohan Hedberg2014-12-304-0/+66
| | | | | | | | | | | | | | | | This patch adds the initial skeleton and kernel config option for SMP self-tests. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
| * Bluetooth: Add support for ECDH test casesJohan Hedberg2014-12-302-1/+175
| | | | | | | | | | | | | | | | This patch adds the test cases for ECDH cryptographic functionality used by Bluetooth Low Energy Secure Connections feature. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
| * Bluetooth: Add support for self testing frameworkMarcel Holtmann2014-12-305-0/+128
| | | | | | | | | | | | | | | | | | | | | | | | | | This add support for the Bluetooth self testing framework that allows running certain test cases of sample data to ensure correctness of its basic functionality. With this patch only the basic framework will be added. It contains the build magic that allows running this at module loading time or at late_initcall stage when built into the kernel image. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
| * Bluetooth: Fix const declarations for smp_f5 and smp_f6Johan Hedberg2014-12-301-3/+4
| | | | | | | | | | | | | | | | | | These SMP crypto functions should have all their input parameters declared as const. This patch fixes the parameters that were missing the const declaration. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
| * Bluetooth: Remove BlueFritz! specific check from initializationMarcel Holtmann2014-12-261-3/+9
| | | | | | | | | | | | | | | | | | | | | | The AVM BlueFritz! USB controllers had a special handling in the Bluetooth core when it comes to reading the supported commands. Both drivers now set the HCI_QUIRK_BROKEN_LOCAL_COMMANDS and with that it is no longer needed to look for vendor specific details. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
| * Bluetooth: Remove duplicate constant for RFCOMM PSMMarcel Holtmann2014-12-201-2/+2
| | | | | | | | | | | | | | | | The RFCOMM_PSM constant is actually a duplicate. So remove it and use the L2CAP_PSM_RFCOMM constant instead. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
| * Bluetooth: Create debugfs directory for each connection handleMarcel Holtmann2014-12-204-0/+21
| | | | | | | | | | | | | | | | | | For every internal representation of a Bluetooth connection which is identified by hci_conn, create a debugfs directory with the handle number as directory name. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
| * Bluetooth: Store default and maximum LE data length settingsMarcel Holtmann2014-12-202-0/+67
| | | | | | | | | | | | | | | | | | | | | | When the controller supports the LE Data Length Extension feature, the default and maximum data length are read and now stored. For backwards compatibility all values are initialized to the data length values from Bluetooth 4.1 and earlier specifications. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
| * Bluetooth: Enable basics for LE Data Length Extension featureMarcel Holtmann2014-12-201-0/+14
| | | | | | | | | | | | | | | | | | When the controller supports the new LE Data Length Extension feature from Bluetooth 4.2 specification, enable the new events and read the values for default and maxmimum data length supported by the controller. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
| * Bluetooth: Move LE debugfs file creation into hci_debugfs.cMarcel Holtmann2014-12-202-478/+474
| | | | | | | | | | | | | | | | This patch moves the creation of the debugs files for LE controllers into hci_debugfs.c file. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
| * Bluetooth: Move BR/EDR debugfs file creation into hci_debugfs.cMarcel Holtmann2014-12-202-349/+350
| | | | | | | | | | | | | | | | This patch moves the creation of the debugs files for BR/EDR controllers into hci_debugfs.c file. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
| * Bluetooth: Move common debugfs file creation into hci_debugfs.cMarcel Holtmann2014-12-202-201/+200
| | | | | | | | | | | | | | | | This patch moves the creation of the debugs files common for all controllers into hci_debugfs.c file. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
| * Bluetooth: Add skeleton functions for debugfs creationMarcel Holtmann2014-12-204-1/+74
| | | | | | | | | | | | | | | | | | | | | | | | | | | | The debugfs file creation has been part of the core initialization handling of controllers. With the introduction of Bluetooth 4.2 core specification, the number of debugfs files is increasing even further. To avoid cluttering the core controller handling, create a separate file hci_debugfs.c to centralize all debugfs file creation. For now leave the current files in the core, but in the future all debugfs file creation will be moved. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
| * Bluetooth: Support static address when BR/EDR has been disabledMarcel Holtmann2014-12-202-2/+14
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Every BR/EDR/LE dual-mode controller requires to have a public address and so far that has become the identity address and own address. The only way to change that behavior was with a force_static_address debugfs option. However the host can actually disable the BR/EDR part of a dual-mode controller and turn into a single mode LE controller. In that case it makes perfect sense for a host to use a chosen static address instead of the public address. So if the host disables BR/EDR and configures a static address, then that static address is used as identity address and own address. If the host does not configure a static address, then the public address is used as before. Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
| * ieee802154: iface: move multiple node type checkAlexander Aring2014-12-201-14/+9
| | | | | | | | | | | | | | | | This patch moves the handling for checking on multiple node type interface to the corresponding concurrent iface check function. Signed-off-by: Alexander Aring <alex.aring@gmail.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
| * mac802154: iface: check concurrent ifacesAlexander Aring2014-12-201-0/+77
| | | | | | | | | | | | | | | | | | | | This patch adds a check for concurrent interfaces while calling interface up. This avoids to have different mac parameters on one phy. Otherwise it could be that a interface can overwrite current phy mac settings which is set by an another interface. Signed-off-by: Alexander Aring <alex.aring@gmail.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
| * Bluetooth: Move hci_update_page_scan to hci_request.cJohan Hedberg2014-12-192-56/+56
| | | | | | | | | | | | | | | | | | This is a left-over from the patch that created hci_request.c. The hci_update_page_scan functions should have been moved from hci_core.c there. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
| * Bluetooth: Add return parameter to cmd_complete callbacksJohan Hedberg2014-12-191-40/+45
| | | | | | | | | | | | | | | | | | | | | | | | | | The cmd_complete callbacks for pending mgmt commands may fail e.g. in the case of memory allocation. Previously this error would be caught and returned to user space in the form of a failed write on the mgmt socket (when the error happened in the mgmt command handler) but with the introduction of the generic cmd_complete callback this information was lost. This patch returns the feature by making cmd_complete callbacks return int instead of void. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
| * Bluetooth: Fix Add Device to wait for HCI before sending cmd_completeJohan Hedberg2014-12-192-68/+109
| | | | | | | | | | | | | | | | | | | | | | | | This patch updates the Add Device mgmt command handler to use a hci_request to wait for HCI command completion before notifying user space of the mgmt command completion. To do this we need to add an extra hci_request parameter to the hci_conn_params_set function. Since this function has no other users besides mgmt.c it's moved there as a static function. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
| * Bluetooth: Fix Remove Device to wait for HCI before sending cmd_completeJohan Hedberg2014-12-191-22/+62
| | | | | | | | | | | | | | | | | | | | | | This patch updates the Remove Device mgmt command handler to use a hci_request to wait for HCI command completion before notifying user space of the mgmt command completion. This way we ensure that once the mgmt command returns all HCI commands triggered by it have also completed. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
| * Bluetooth: Add hci_request support for hci_update_background_scanJohan Hedberg2014-12-194-95/+105
| | | | | | | | | | | | | | | | | | | | | | | | | | Many places using hci_update_background_scan() try to synchronize whatever they're doing with the help of hci_request callbacks. However, since the hci_update_background_scan() function hasn't so far accepted a hci_request pointer any commands triggered by it have been left out by the synchronization. This patch modifies the API in a similar way as was done for hci_update_page_scan, i.e. there's a variant that takes a hci_request and another one that takes a hci_dev. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
| * Bluetooth: Split hci_request helpers to hci_request.[ch]Johan Hedberg2014-12-197-370/+453
| | | | | | | | | | | | | | | | | | | | None of the hci_request related things in net/bluetooth/hci_core.h are needed anywhere outside of the core bluetooth module. This patch creates a new net/bluetooth/hci_request.c file with its corresponding h-file and moves the functionality there from hci_core.c and hci_core.h. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
| * Bluetooth: Split hci_update_page_scan into two functionsJohan Hedberg2014-12-193-15/+22
| | | | | | | | | | | | | | | | | | | | | | To keep the parameter list and its semantics clear it makes sense to split the hci_update_page_scan function into two separate functions: one taking a hci_dev and another taking a hci_request. The one taking a hci_dev constructs its own hci_request and then calls the other function. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
| * nl802154: introduce support for cca settingsAlexander Aring2014-12-193-3/+71
| | | | | | | | | | | | | | This patch adds support for setting cca parameters via nl802154. Signed-off-by: Alexander Aring <alex.aring@gmail.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
| * ieee802154: rework cca settingAlexander Aring2014-12-195-9/+10
| | | | | | | | | | | | | | | | | | | | The current cca setting handle is a driver specific call. We need to introduce some 802.15.4 specific layer and mapping 802.15.4 cca modes to driver specific ones inside the 802.15.4 driver. This patch will add such 802.15.4 layer and mapping the cca settings to driver specific ones. Signed-off-by: Alexander Aring <alex.aring@gmail.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
* | geneve: Add Geneve GRO supportJoe Stringer2015-01-021-2/+95
| | | | | | | | | | | | | | | | | | This results in an approximately 30% increase in throughput when handling encapsulated bulk traffic. Signed-off-by: Joe Stringer <joestringer@nicira.com> Signed-off-by: Jesse Gross <jesse@nicira.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | net: Add Transparent Ethernet Bridging GRO support.Jesse Gross2015-01-021-0/+92
| | | | | | | | | | | | | | | | | | Currently the only tunnel protocol that supports GRO with encapsulated Ethernet is VXLAN. This pulls out the Ethernet code into a proper layer so that it can be used by other tunnel protocols such as GRE and Geneve. Signed-off-by: Jesse Gross <jesse@nicira.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | fib_trie: Add tracking value for suffix lengthAlexander Duyck2014-12-311-6/+116
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This change adds a tracking value for the maximum suffix length of all prefixes stored in any given tnode. With this value we can determine if we need to backtrace or not based on if the suffix is greater than the pos value. By doing this we can reduce the CPU overhead for lookups in the local table as many of the prefixes there are 32b long and have a suffix length of 0 meaning we can immediately backtrace to the root node without needing to test any of the nodes between it and where we ended up. Signed-off-by: Alexander Duyck <alexander.h.duyck@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | fib_trie: Remove checks for index >= tnode_child_length from tnode_get_childAlexander Duyck2014-12-311-9/+5
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | For some reason the compiler doesn't seem to understand that when we are in a loop that runs from tnode_child_length - 1 to 0 we don't expect the value of tn->bits to change. As such every call to tnode_get_child was rerunning tnode_chile_length which ended up consuming quite a bit of space in the resultant assembly code. I have gone though and verified that in all cases where tnode_get_child is used we are either winding though a fixed loop from tnode_child_length - 1 to 0, or are in a fastpath case where we are verifying the value by either checking for any remaining bits after shifting index by bits and testing for leaf, or by using tnode_child_length. size net/ipv4/fib_trie.o Before: text data bss dec hex filename 15506 376 8 15890 3e12 net/ipv4/fib_trie.o After: text data bss dec hex filename 14827 376 8 15211 3b6b net/ipv4/fib_trie.o Signed-off-by: Alexander Duyck <alexander.h.duyck@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | fib_trie: inflate/halve nodes in a more RCU friendly wayAlexander Duyck2014-12-311-121/+115
| | | | | | | | | | | | | | | | | | | | | | | | | | This change pulls the node_set_parent functionality out of put_child_reorg and instead leaves that to the function to take care of as well. By doing this we can fully construct the new cluster of tnodes and all of the pointers out of it before we start routing pointers into it. I am suspecting this will likely fix some concurency issues though I don't have a good test to show as such. Signed-off-by: Alexander Duyck <alexander.h.duyck@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | fib_trie: Push tnode flushing down to inflate/halveAlexander Duyck2014-12-311-53/+50
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This change pushes the tnode freeing down into the inflate and halve functions. It makes more sense here as we have a better grasp of what is going on and when a given cluster of nodes is ready to be freed. I believe this may address a bug in the freeing logic as well. For some reason if the freelist got to a certain size we would call synchronize_rcu(). I'm assuming that what they meant to do is call synchronize_rcu() after they had handed off that much memory via call_rcu(). As such that is what I have updated the behavior to be. Signed-off-by: Alexander Duyck <alexander.h.duyck@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | fib_trie: Push assignment of child to parent down into inflate/halveAlexander Duyck2014-12-311-83/+66
| | | | | | | | | | | | | | | | | | | | This change makes it so that the assignment of the tnode to the parent is handled directly within whatever function is currently handling the node be it inflate, halve, or resize. By doing this we can avoid some of the need to set NULL pointers in the tree while we are resizing the subnodes. Signed-off-by: Alexander Duyck <alexander.h.duyck@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | fib_trie: Add functions should_inflate and should_halveAlexander Duyck2014-12-311-86/+89
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This change pulls the logic for if we should inflate/halve the nodes out into separate functions. It also addresses what I believe is a bug where 1 full node is all that is needed to keep a node from ever being halved. Simple script to reproduce the issue: modprobe dummy; ifconfig dummy0 up for i in `seq 0 255`; do ifconfig dummy0:$i 10.0.${i}.1/24 up; done ifconfig dummy0:256 10.0.255.33/16 up for i in `seq 0 254`; do ifconfig dummy0:$i down; done Results from /proc/net/fib_triestat Before: Local: Aver depth: 3.00 Max depth: 4 Leaves: 17 Prefixes: 18 Internal nodes: 11 1: 8 2: 2 10: 1 Pointers: 1048 Null ptrs: 1021 Total size: 11 kB After: Local: Aver depth: 3.41 Max depth: 5 Leaves: 17 Prefixes: 18 Internal nodes: 12 1: 8 2: 3 3: 1 Pointers: 36 Null ptrs: 8 Total size: 3 kB Signed-off-by: Alexander Duyck <alexander.h.duyck@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | fib_trie: Move resize to after inflate/halveAlexander Duyck2014-12-311-157/+154
| | | | | | | | | | | | | | | | This change consists of a cut/paste of resize to behind inflate and halve so that I could remove the two function prototypes. Signed-off-by: Alexander Duyck <alexander.h.duyck@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | fib_trie: Push rcu_read_lock/unlock to callersAlexander Duyck2014-12-313-101/+85
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This change is to start cleaning up some of the rcu_read_lock/unlock handling. I realized while reviewing the code there are several spots that I don't believe are being handled correctly or are masking warnings by locally calling rcu_read_lock/unlock instead of calling them at the correct level. A common example is a call to fib_get_table followed by fib_table_lookup. The rcu_read_lock/unlock ought to wrap both but there are several spots where they were not wrapped. Signed-off-by: Alexander Duyck <alexander.h.duyck@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | fib_trie: Use unsigned long for anything dealing with a shift by bitsAlexander Duyck2014-12-311-27/+26
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This change makes it so that anything that can be shifted by, or compared to a value shifted by bits is updated to be an unsigned long. This is mostly a precaution against an insanely huge address space that somehow starts coming close to the 2^32 root node size which would require something like 1.5 billion addresses. I chose unsigned long instead of unsigned long long since I do not believe it is possible to allocate a 32 bit tnode on a 32 bit system as the memory consumed would be 16GB + 28B which exceeds the addressible space for any one process. Signed-off-by: Alexander Duyck <alexander.h.duyck@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | fib_trie: Update meaning of pos to represent unchecked bitsAlexander Duyck2014-12-311-113/+81
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This change moves the pos value to the other side of the "bits" field. By doing this it actually simplifies a significant amount of code in the trie. For example when halving a tree we know that the bit lost exists at oldnode->pos, and if we inflate the tree the new bit being add is at tn->pos. Previously to find those bits you would have to subtract pos and bits from the keylength or start with a value of (1 << 31) and then shift that. There are a number of spots throughout the code that benefit from this. In the case of the hot-path searches the main advantage is that we can drop 2 or more operations from the search path as we no longer need to compute the value for the index to be shifted by and can instead just use the raw pos value. In addition the tkey_extract_bits is now defunct and can be replaced by get_index since the two operations were doing the same thing, but now get_index does it much more quickly as it is only an xor and shift versus a pair of shifts and a subtraction. Signed-off-by: Alexander Duyck <alexander.h.duyck@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | fib_trie: Optimize fib_table_insertAlexander Duyck2014-12-311-125/+71
| | | | | | | | | | | | | | | | | | This patch updates the fib_table_insert function to take advantage of the changes made to improve the performance of fib_table_lookup. As a result the code should be smaller and run faster then the original. Signed-off-by: Alexander Duyck <alexander.h.duyck@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | fib_trie: Optimize fib_find_nodeAlexander Duyck2014-12-311-15/+21
| | | | | | | | | | | | | | | | | | This patch makes use of the same features I made use of for fib_table_lookup to streamline fib_find_node. The resultant code should be smaller and run faster than the original. Signed-off-by: Alexander Duyck <alexander.h.duyck@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | fib_trie: Optimize fib_table_lookup to avoid wasting time on loops/variablesAlexander Duyck2014-12-311-157/+93
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This patch is meant to reduce the complexity of fib_table_lookup by reducing the number of variables to the bare minimum while still keeping the same if not improved functionality versus the original. Most of this change was started off by the desire to rid the function of chopped_off and current_prefix_length as they actually added very little to the function since they only applied when computing the cindex. I was able to replace them mostly with just a check for the prefix match. As long as the prefix between the key and the node being tested was the same we know we can search the tnode fully versus just testing cindex 0. The second portion of the change ended up being a massive reordering. Originally the calls to check_leaf were up near the start of the loop, and the backtracing and descending into lower levels of tnodes was later. This didn't make much sense as the structure of the tree means the leaves are always the last thing to be tested. As such I reordered things so that we instead have a loop that will delve into the tree and only exit when we have either found a leaf or we have exhausted the tree. The advantage of rearranging things like this is that we can fully inline check_leaf since there is now only one reference to it in the function. Signed-off-by: Alexander Duyck <alexander.h.duyck@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | fib_trie: Merge leaf into tnodeAlexander Duyck2014-12-311-182/+140
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This change makes it so that leaf and tnode are the same struct. As a result there is no need for rt_trie_node anymore since everyting can be merged into tnode. On 32b systems this results in the leaf being 4 bytes larger, however I don't know if that is really an issue as this and an eariler patch that added bits & pos have increased the size from 20 to 28. If I am not mistaken slub/slab allocate on power of 2 sizes so 20 was likely being rounded up to 32 anyway. Signed-off-by: Alexander Duyck <alexander.h.duyck@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | fib_trie: Merge tnode_free and leaf_free into node_freeAlexander Duyck2014-12-311-50/+40
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Both the leaf and the tnode had an rcu_head in them, but they had them in slightly different places. Since we now have them in the same spot and know that any node with bits == 0 is a leaf and the rest are either vmalloc or kmalloc tnodes depending on the value of bits it makes it easy to combine the functions and reduce overhead. In addition I have taken advantage of the rcu_head pointer to go ahead and put together a simple linked list instead of using the tnode pointer as this way we can merge either type of structure for freeing. Signed-off-by: Alexander Duyck <alexander.h.duyck@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | fib_trie: Make leaf and tnode more uniformAlexander Duyck2014-12-311-110/+82
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This change makes some fundamental changes to the way leaves and tnodes are constructed. The big differences are: 1. Leaves now populate pos and bits indicating their full key size. 2. Trie nodes now mask out their lower bits to be consistent with the leaf 3. Both structures have been reordered so that rt_trie_node now consisists of a much larger region including the pos, bits, and rcu portions of the tnode structure. On 32b systems this will result in the leaf being 4B larger as the pos and bits values were added to a hole created by the key as it was only 4B in length. Signed-off-by: Alexander Duyck <alexander.h.duyck@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | fib_trie: Update usage stats to be percpu instead of global variablesAlexander Duyck2014-12-312-21/+49
| | | | | | | | | | | | | | | | | | | | | | | | | | | | The trie usage stats were currently being shared by all threads that were calling fib_table_lookup. As a result when multiple threads were performing lookups simultaneously the trie would begin to cache bounce between those threads. In order to prevent this I have updated the usage stats to use a set of percpu variables. By doing this we should be able to avoid the cache bouncing and still make use of these stats. Signed-off-by: Alexander Duyck <alexander.h.duyck@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
generated by cgit v1.2.3 (git 2.25.1) at 2024-09-02 19:05:19 +0000