- 28 11月, 2017 9 次提交
-
-
由 Antoine Tenart 提交于
When an allocation in the txq_init path fails, the allocated buffers end-up being freed twice: in the txq_init error path, and in txq_deinit. This lead to issues as txq_deinit would work on already freed memory regions: kernel BUG at mm/slub.c:3915! Internal error: Oops - BUG: 0 [#1] PREEMPT SMP This patch fixes this by removing the txq_init own error path, as the txq_deinit function is always called on errors. This was introduced by TSO as way more buffers are allocated. Fixes: 186cd4d4 ("net: mvpp2: software tso support") Signed-off-by: NAntoine Tenart <antoine.tenart@free-electrons.com> Signed-off-by: NDavid S. Miller <davem@davemloft.net>
-
由 Petr Machata 提交于
The function mlxsw_sp_nexthop_rif_update() walks the list of nexthops associated with a RIF, and updates the corresponding entries in the switch. It is used in particular when a tunnel underlay netdevice moves to a different VRF, and all the nexthops are migrated over to a new RIF. The problem is that each nexthop holds a reference to its RIF, and that is not updated. So after the old RIF is gone, further activity on these nexthops (such as downing the underlay netdevice) dereferences a dangling pointer. Fix the issue by updating rif of impacted nexthops before calling mlxsw_sp_nexthop_rif_update(). Fixes: 0c5f1cd5 ("mlxsw: spectrum_router: Generalize __mlxsw_sp_ipip_entry_update_tunnel()") Signed-off-by: NPetr Machata <petrm@mellanox.com> Reviewed-by: NIdo Schimmel <idosch@mellanox.com> Signed-off-by: NJiri Pirko <jiri@mellanox.com> Signed-off-by: NDavid S. Miller <davem@davemloft.net>
-
由 Petr Machata 提交于
Some tunnels that are offloadable on their own can nonetheless be demoted to slow path if their local address is in conflict with that of another tunnel. When a route is formed for such a tunnel, mlxsw_sp_nexthop_ipip_init() fails to find the corresponding IPIP entry, and that triggers a FIB abort. Resolve the problem by not assuming that a tunnel for which mlxsw_sp_ipip_ops.can_offload() holds also automatically has an IPIP entry. Fixes: af641713 ("mlxsw: spectrum_router: Onload conflicting tunnels") Signed-off-by: NPetr Machata <petrm@mellanox.com> Reviewed-by: NIdo Schimmel <idosch@mellanox.com> Signed-off-by: NJiri Pirko <jiri@mellanox.com> Signed-off-by: NDavid S. Miller <davem@davemloft.net>
-
由 Petr Machata 提交于
The mlxsw driver currently doesn't offload GRE tunnels if they have the same local address and use the same underlay VRF. When such a situation arises, the tunnels in conflict are demoted to slow path. However, the current code only verifies this condition on tunnel creation and tunnel change, not when a tunnel is moved to a different VRF. When the tunnel has no bound device, underlay and overlay are the same. Thus moving a tunnel moves the underlay as well, and that can cause local address conflict. So modify mlxsw_sp_netdevice_ipip_ol_vrf_event() to check if there are any conflicting tunnels, and demote them if yes. Fixes: af641713 ("mlxsw: spectrum_router: Onload conflicting tunnels") Signed-off-by: NPetr Machata <petrm@mellanox.com> Reviewed-by: NIdo Schimmel <idosch@mellanox.com> Signed-off-by: NJiri Pirko <jiri@mellanox.com> Signed-off-by: NDavid S. Miller <davem@davemloft.net>
-
由 Petr Machata 提交于
When a new local route is added, an IPIP entry is looked up to determine whether the route should be offloaded as a tunnel decap or as a trap. That decision should take into account whether the tunnel netdevice in question is actually IFF_UP, and only install a decap offload if it is. Fixes: 0063587d ("mlxsw: spectrum: Support decap-only IP-in-IP tunnels") Signed-off-by: NPetr Machata <petrm@mellanox.com> Reviewed-by: NIdo Schimmel <idosch@mellanox.com> Signed-off-by: NJiri Pirko <jiri@mellanox.com> Signed-off-by: NDavid S. Miller <davem@davemloft.net>
-
由 Ahmad Fatoum 提交于
Signed-off-by: NAhmad Fatoum <ahmad@a3f.at> Tested-by: NAaron Brown <aaron.f.brown@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
由 Amritha Nambiar 提交于
Adding cloud filters could fail for a number of reasons, unsupported filter fields for example, which fails during validation of fields itself. This will not result in admin command errors and converting the admin queue status to posix error code using i40e_aq_rc_to_posix would result in incorrect error values. If the failure was due to AQ error itself, reporting that correctly is handled in the inner function. Signed-off-by: NAmritha Nambiar <amritha.nambiar@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
由 Sasha Neftin 提交于
This is a follow on to commit b10effb9 ("fix buffer overrun while the I219 is processing DMA transactions") to address David Laights concerns about the use of "magic" numbers. So define masks as well as add additional code comments to give a better understanding of what needs to be done to avoid a buffer overrun. Signed-off-by: NSasha Neftin <sasha.neftin@intel.com> Reviewed-by: NAlexander H Duyck <alexander.h.duyck@intel.com> Reviewed-by: NDima Ruinskiy <dima.ruinskiy@intel.com> Reviewed-by: NRaanan Avargil <raanan.avargil@intel.com> Tested-by: NAaron Brown <aaron.f.brown@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
由 Gustavo A R Silva 提交于
sizeof when applied to a pointer typed expression gives the size of the pointer. The proper fix in this particular case is to code sizeof(*vfres) instead of sizeof(vfres). This issue was detected with the help of Coccinelle. Signed-off-by: NGustavo A R Silva <garsilva@embeddedor.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
- 25 11月, 2017 2 次提交
-
-
由 Sunil Goutham 提交于
Don't offload IP header checksum to NIC. This fixes a previous patch which enabled checksum offloading for both IPv4 and IPv6 packets. So L3 checksum offload was getting enabled for IPv6 pkts. And HW is dropping these pkts as it assumes the pkt is IPv4 when IP csum offload is set in the SQ descriptor. Fixes: 3a9024f5 ("net: thunderx: Enable TSO and checksum offloads for ipv6") Signed-off-by: NSunil Goutham <sgoutham@cavium.com> Signed-off-by: NAleksey Makarov <aleksey.makarov@auriga.com> Reviewed-by: NEric Dumazet <edumazet@google.com> Signed-off-by: NDavid S. Miller <davem@davemloft.net>
-
由 Zhu Yanjun 提交于
The function pci_unmap_page is obsolete. So it is replaced with the function dma_unmap_page. CC: Srinivas Eeda <srinivas.eeda@oracle.com> CC: Joe Jin <joe.jin@oracle.com> CC: Junxiao Bi <junxiao.bi@oracle.com> Signed-off-by: NZhu Yanjun <yanjun.zhu@oracle.com> Signed-off-by: NDavid S. Miller <davem@davemloft.net>
-
- 24 11月, 2017 1 次提交
-
-
由 Tobias Jakobi 提交于
Commit 6fa1ba61 partially implemented the new ethtool API, by replacing get_settings() with get_link_ksettings(). This breaks ethtool, since the userspace tool (according to the new API specs) never tries the legacy set() call, when the new get() call succeeds. All attempts to chance some setting from userspace result in: > Cannot set new settings: Operation not supported Implement the missing set() call. Signed-off-by: NTobias Jakobi <tjakobi@math.uni-bielefeld.de> Tested-by: NHolger Hoffstätte <holger@applied-asynchrony.com> Reviewed-by: NAndrew Lunn <andrew@lunn.ch> Signed-off-by: NDavid S. Miller <davem@davemloft.net>
-
- 22 11月, 2017 13 次提交
-
-
由 Brian King 提交于
The original issue being fixed in this patch was seen with the ixgbe driver, but the same issue exists with i40evf as well, as the code is very similar. read_barrier_depends is not sufficient to ensure loads following it are not speculatively loaded out of order by the CPU, which can result in stale data being loaded, causing potential system crashes. Cc: stable <stable@vger.kernel.org> Signed-off-by: NBrian King <brking@linux.vnet.ibm.com> Acked-by: NJesse Brandeburg <jesse.brandeburg@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
由 Brian King 提交于
The original issue being fixed in this patch was seen with the ixgbe driver, but the same issue exists with fm10k as well, as the code is very similar. read_barrier_depends is not sufficient to ensure loads following it are not speculatively loaded out of order by the CPU, which can result in stale data being loaded, causing potential system crashes. Cc: stable <stable@vger.kernel.org> Signed-off-by: NBrian King <brking@linux.vnet.ibm.com> Acked-by: NJesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
由 Brian King 提交于
The original issue being fixed in this patch was seen with the ixgbe driver, but the same issue exists with igb as well, as the code is very similar. read_barrier_depends is not sufficient to ensure loads following it are not speculatively loaded out of order by the CPU, which can result in stale data being loaded, causing potential system crashes. Cc: stable <stable@vger.kernel.org> Signed-off-by: NBrian King <brking@linux.vnet.ibm.com> Acked-by: NJesse Brandeburg <jesse.brandeburg@intel.com> Tested-by: NAaron Brown <aaron.f.brown@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
由 Brian King 提交于
The original issue being fixed in this patch was seen with the ixgbe driver, but the same issue exists with igbvf as well, as the code is very similar. read_barrier_depends is not sufficient to ensure loads following it are not speculatively loaded out of order by the CPU, which can result in stale data being loaded, causing potential system crashes. Cc: stable <stable@vger.kernel.org> Signed-off-by: NBrian King <brking@linux.vnet.ibm.com> Acked-by: NJesse Brandeburg <jesse.brandeburg@intel.com> Tested-by: NAaron Brown <aaron.f.brown@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
由 Brian King 提交于
The original issue being fixed in this patch was seen with the ixgbe driver, but the same issue exists with ixgbevf as well, as the code is very similar. read_barrier_depends is not sufficient to ensure loads following it are not speculatively loaded out of order by the CPU, which can result in stale data being loaded, causing potential system crashes. Cc: stable <stable@vger.kernel.org> Signed-off-by: NBrian King <brking@linux.vnet.ibm.com> Acked-by: NJesse Brandeburg <jesse.brandeburg@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
由 Brian King 提交于
The original issue being fixed in this patch was seen with the ixgbe driver, but the same issue exists with i40e as well, as the code is very similar. read_barrier_depends is not sufficient to ensure loads following it are not speculatively loaded out of order by the CPU, which can result in stale data being loaded, causing potential system crashes. Cc: stable <stable@vger.kernel.org> Signed-off-by: NBrian King <brking@linux.vnet.ibm.com> Acked-by: NJesse Brandeburg <jesse.brandeburg@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
由 Brian King 提交于
This patch fixes an issue seen on Power systems with ixgbe which results in skb list corruption and an eventual kernel oops. The following is what was observed: CPU 1 CPU2 ============================ ============================ 1: ixgbe_xmit_frame_ring ixgbe_clean_tx_irq 2: first->skb = skb eop_desc = tx_buffer->next_to_watch 3: ixgbe_tx_map read_barrier_depends() 4: wmb check adapter written status bit 5: first->next_to_watch = tx_desc napi_consume_skb(tx_buffer->skb ..); 6: writel(i, tx_ring->tail); The read_barrier_depends is insufficient to ensure that tx_buffer->skb does not get loaded prior to tx_buffer->next_to_watch, which then results in loading a stale skb pointer. This patch replaces the read_barrier_depends with smp_rmb to ensure loads are ordered with respect to the load of tx_buffer->next_to_watch. Cc: stable <stable@vger.kernel.org> Signed-off-by: NBrian King <brking@linux.vnet.ibm.com> Acked-by: NJesse Brandeburg <jesse.brandeburg@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
由 Alan Brady 提交于
After a reset we rebuild the VSIs which is going to clobber any promiscuous settings we had before reset. This makes it so that we restore the promiscuous settings we had before reset. Signed-off-by: NAlan Brady <alan.brady@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
由 Alan Brady 提交于
The current method for notifying clients of l2 parameters is broken because we fail to copy the new parameters to the client instance struct, we need to do the notification before the client 'open' function pointer gets called, and lastly we should set the l2 parameters when first adding a client instance. This patch first introduces the i40evf_client_get_params function to prevent code duplication in the i40evf_client_add_instance and the i40evf_notify_client_l2_params functions. We then fix the notify l2 params function to actually copy the parameters to client instance struct and do the same in the *_add_instance' function. Lastly this patch reorganizes the priority in which client tasks fire so that if the flag for notifying l2 params is set, it will trigger before the open because the client needs these new parameters as part of a client open task. Signed-off-by: NAlan Brady <alan.brady@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
由 Filip Sadowski 提交于
This patch allows detection of upcoming core reset in case NIC gets stuck while performing FLR reset. The i40e_pf_reset() function returns I40E_ERR_NOT_READY when global reset was detected. Signed-off-by: NFilip Sadowski <filip.sadowski@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
由 Amritha Nambiar 提交于
It is safe to remove the upper limit of 64 queues on a channel VSI. The upper bound is determined by the VSI's num_queue_pairs and gets validated when the queue mapping info through mqprio interface is subject to bound checking in the driver. Signed-off-by: NAmritha Nambiar <amritha.nambiar@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
由 Zijie Pan 提交于
num_mac should be increased only after the call to i40e_add_mac_filter(). Fixes: 5f527ba9 ("i40e: Limit the number of MAC and VLAN addresses that can be added for VFs") Signed-off-by: NZijie Pan <zijie.pan@6wind.com> Signed-off-by: NNicolas Dichtel <nicolas.dichtel@6wind.com> Reviewed-by: NTushar Dave <tushar.n.dave@oracle.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
由 Jacob Keller 提交于
Since commit 96a39aed ("i40e: Acquire NVM lock before reads on all devices") we've used the NVM lock to synchronize NVM reads even on devices which don't strictly need the lock. Doing so can cause a regression on older firmware prior to 1.5, especially when downgrading the firmware. Fix this by only grabbing the lock if we're running on an X722 device (which requires the lock as it uses the AdminQ to read the NVM), or if we're currently running 1.5 or newer firmware. Signed-off-by: NJacob Keller <jacob.e.keller@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
- 21 11月, 2017 3 次提交
-
-
由 Jakub Kicinski 提交于
Commit 0115552e ("nfp: remove false positive offloads in flower vxlan") missed adding kdoc for a new parameter of nfp_flower_add_offload(). Signed-off-by: NJakub Kicinski <jakub.kicinski@netronome.com> Reviewed-by: NSimon Horman <simon.horman@netronome.com> Signed-off-by: NDavid S. Miller <davem@davemloft.net>
-
由 Ido Schimmel 提交于
On some systems, when we unsplit a port we need to re-create two ports instead. On other systems, only one needs to be re-created. Do not try to create a port if during driver initialization it was assigned a negative module number, which is invalid. This avoids the following error during unsplit: [ 941.012478] mlxsw_spectrum 0000:01:00.0: Port 43: Failed to map module The error is harmless and caused by the fact that a local port is already mapped to module 0. Fixes: be94535f ("mlxsw: spectrum: Make split flow match firmware requirements") Signed-off-by: NIdo Schimmel <idosch@mellanox.com> Reviewed-by: NArkadi Sharshevsky <arkadis@mellanox.com> Signed-off-by: NJiri Pirko <jiri@mellanox.com> Signed-off-by: NDavid S. Miller <davem@davemloft.net>
-
由 Jakub Kicinski 提交于
With TC shared block changes we can't depend on correct netdev pointer being available in cls_bpf. Move the device validation to the driver. Core will only make sure that offloaded programs are always attached in the driver (or in HW by the driver). We trust that drivers which implement offload callbacks will perform necessary checks. Moving the checks to the driver is generally a useful thing, in practice the check should be against a switchdev instance, not a netdev, given that most ASICs will probably allow using the same program on many ports. Signed-off-by: NJakub Kicinski <jakub.kicinski@netronome.com> Reviewed-by: NQuentin Monnet <quentin.monnet@netronome.com> Acked-by: NAlexei Starovoitov <ast@kernel.org> Acked-by: NDaniel Borkmann <daniel@iogearbox.net> Acked-by: NJiri Pirko <jiri@mellanox.com> Signed-off-by: NDaniel Borkmann <daniel@iogearbox.net>
-
- 20 11月, 2017 2 次提交
-
-
由 Christophe JAILLET 提交于
Some statements are not enough or too much indented. Fix it to improve readalbility. Signed-off-by: NChristophe JAILLET <christophe.jaillet@wanadoo.fr> Signed-off-by: NDavid S. Miller <davem@davemloft.net>
-
由 Netanel Belgazal 提交于
In rare cases, ena driver would reset and re-start the device, for example, in case of misbehaving application that causes transmit timeout The first step in the reset procedure is to stop the Tx traffic by calling ena_carrier_off(). After the driver have just started the device reset procedure, device happens to send an asynchronous notification (via AENQ) to the driver than there was a link change (to link-up state). This link change is mapped to a call to netif_carrier_on() which re-activates the Tx queues, violating the assumption of no tx traffic until device reset is completed, as the reset task might still be in the process of queues initialization, leading to an access to uninitialized memory. Signed-off-by: NNetanel Belgazal <netanel@amazon.com> Signed-off-by: NDavid S. Miller <davem@davemloft.net>
-
- 19 11月, 2017 2 次提交
-
-
由 Heiner Kallweit 提交于
Adjust the code to use the same green settings as in the latest vendor driver. Signed-off-by: NHeiner Kallweit <hkallweit1@gmail.com> Signed-off-by: NDavid S. Miller <davem@davemloft.net>
-
由 Heiner Kallweit 提交于
Name of functions rtl_w0w1_eri and rtl_w0w1_phy is somewhat misleading regarding order of arguments. One could assume that w0w1 means argument with bits to be reset comes before argument with bits to set. However this is not the case. So fix the order of arguments in several statements. In addition fix EEE advertisement. The current code resets the bits for 100BaseT and 1000BaseT EEE advertisement what is not what we want. I have a little of a hard time to find a proper "Fixes" line as the issue seems to have been there forever (at least it existed already when the driver was moved to the current place in 2011). The patch was tested on a Zotac Mini-PC with a RTL8111E-VL chip. Before the patch EEE was disabled, now it's properly advertised and works fine. Signed-off-by: NHeiner Kallweit <hkallweit1@gmail.com> Signed-off-by: NDavid S. Miller <davem@davemloft.net>
-
- 18 11月, 2017 1 次提交
-
-
This patch fixes the dma_mapping_error call to use the correct dma_addr which is inside the ibmvnic_vpd struct. Moreover, it fixes an uninitialized warning regarding a local dma_addr variable which is not used anymore. Fixes: 4e6759be ("ibmvnic: Feature implementation of VPD for the ibmvnic driver") Reported-by: NStephen Rothwell <sfr@canb.auug.org.au> Signed-off-by: NDesnes A. Nunes do Rosario <desnesn@linux.vnet.ibm.com> Signed-off-by: NDavid S. Miller <davem@davemloft.net>
-
- 17 11月, 2017 5 次提交
-
-
由 John Hurley 提交于
Pass information to the match offload on whether or not the repr is the ingress or egress dev. Only accept tunnel matches if repr is the egress dev. This means rules such as the following are successfully offloaded: tc .. add dev vxlan0 .. enc_dst_port 4789 .. action redirect dev nfp_p0 While rules such as the following are rejected: tc .. add dev nfp_p0 .. enc_dst_port 4789 .. action redirect dev vxlan0 Also reject non tunnel flows that are offloaded to an egress dev. Non tunnel matches assume that the offload dev is the ingress port and offload a match accordingly. Fixes: 611aec10 ("nfp: compile flower vxlan tunnel metadata match fields") Signed-off-by: NJohn Hurley <john.hurley@netronome.com> Reviewed-by: NJakub Kicinski <jakub.kicinski@netronome.com> Signed-off-by: NDavid S. Miller <davem@davemloft.net>
-
由 John Hurley 提交于
Register a callback for offloading flows that have a repr as their egress device. The new egdev_register function is added to net-next for the 4.15 release. Signed-off-by: NJohn Hurley <john.hurley@netronome.com> Reviewed-by: NJakub Kicinski <jakub.kicinski@netronome.com> Signed-off-by: NDavid S. Miller <davem@davemloft.net>
-
由 Dirk van der Merwe 提交于
The PF netdev is used for data transfer for reprs, so reprs inherit the maximum MTU settings of the PF netdev. Fixes: 5de73ee4 ("nfp: general representor implementation") Signed-off-by: NDirk van der Merwe <dirk.vandermerwe@netronome.com> Reviewed-by: NJakub Kicinski <jakub.kicinski@netronome.com> Signed-off-by: NDavid S. Miller <davem@davemloft.net>
-
由 Pieter Jansen van Vuuren 提交于
Correct typo in vlan receive MAC stats. Previously the MAC statistics reported in ethtool for vlan receive contained a typo resulting in ethtool reporting rx_vlan_reveive_ok instead of rx_vlan_received_ok. Fixes: a5950182 ("nfp: map mac_stats and vf_cfg BARs") Fixes: 098ce840 ("nfp: report MAC statistics in ethtool") Reported-by: NBrendan Galloway <brendan.galloway@netronome.com> Signed-off-by: NPieter Jansen van Vuuren <pieter.jansenvanvuuren@netronome.com> Reviewed-by: NJakub Kicinski <jakub.kicinski@netronome.com> Reviewed-by: NSimon Horman <simon.horman@netronome.com> Signed-off-by: NDavid S. Miller <davem@davemloft.net>
-
由 Pieter Jansen van Vuuren 提交于
Hardware has no notion of new or last mask id, instead it makes use of the message type (i.e. add flow or del flow) in combination with a single bit in metadata flags to determine when to add or delete a mask id. Previously we made use of the new or last flags to indicate that a new mask should be allocated or deallocated, respectively. This incorrect behaviour is fixed by making use single bit in metadata flags to indicate mask allocation or deallocation. Fixes: 43f84b72 ("nfp: add metadata to each flow offload") Signed-off-by: NPieter Jansen van Vuuren <pieter.jansenvanvuuren@netronome.com> Reviewed-by: NJakub Kicinski <jakub.kicinski@netronome.com> Signed-off-by: NDavid S. Miller <davem@davemloft.net>
-
- 16 11月, 2017 2 次提交
-
-
由 Huacai Chen 提交于
This patch try to fix the building error on MIPS. The reason is MIPS has already defined the LONG macro, which conflicts with the LONG enum in drivers/net/ethernet/fealnx.c. Signed-off-by: NHuacai Chen <chenhc@lemote.com> Signed-off-by: NDavid S. Miller <davem@davemloft.net>
-
由 Mel Gorman 提交于
As the page free path makes no distinction between cache hot and cold pages, there is no real useful ordering of pages in the free list that allocation requests can take advantage of. Juding from the users of __GFP_COLD, it is likely that a number of them are the result of copying other sites instead of actually measuring the impact. Remove the __GFP_COLD parameter which simplifies a number of paths in the page allocator. This is potentially controversial but bear in mind that the size of the per-cpu pagelists versus modern cache sizes means that the whole per-cpu list can often fit in the L3 cache. Hence, there is only a potential benefit for microbenchmarks that alloc/free pages in a tight loop. It's even worse when THP is taken into account which has little or no chance of getting a cache-hot page as the per-cpu list is bypassed and the zeroing of multiple pages will thrash the cache anyway. The truncate microbenchmarks are not shown as this patch affects the allocation path and not the free path. A page fault microbenchmark was tested but it showed no sigificant difference which is not surprising given that the __GFP_COLD branches are a miniscule percentage of the fault path. Link: http://lkml.kernel.org/r/20171018075952.10627-9-mgorman@techsingularity.netSigned-off-by: NMel Gorman <mgorman@techsingularity.net> Acked-by: NVlastimil Babka <vbabka@suse.cz> Cc: Andi Kleen <ak@linux.intel.com> Cc: Dave Chinner <david@fromorbit.com> Cc: Dave Hansen <dave.hansen@intel.com> Cc: Jan Kara <jack@suse.cz> Cc: Johannes Weiner <hannes@cmpxchg.org> Signed-off-by: NAndrew Morton <akpm@linux-foundation.org> Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
-