- 26 Apr, 2017 3 commits
-
-
Paul E. McKenney authored
Commit f60d231a ("srcu: Crude control of expedited grace periods") introduced a per-srcu_struct atomic counter to track outstanding requests for grace periods. This works, but represents a memory-contention bottleneck. This commit therefore uses the srcu_node combining tree to remove this bottleneck. This commit adds new ->srcu_gp_seq_needed_exp fields to the srcu_data, srcu_node, and srcu_struct structures, which track the farthest-in-the-future grace period that must be expedited, which in turn requires that all nearer-term grace periods also be expedited. Requests for expediting start with the srcu_data structure, run up through the srcu_node tree, and end at the srcu_struct structure. Note that it may be necessary to expedite a grace period that just now started, and this is handled by a new srcu_funnel_exp_start() function, which is invoked when the grace period itself is already in its way, but when that grace period was not marked as expedited. A new srcu_get_delay() function returns zero if there is at least one expedited SRCU grace period in flight, or SRCU_INTERVAL otherwise. This function is used to calculate delays: Normal grace periods are allowed to extend in order to cover more requests with a given grace-period computation, which decreases per-request overhead. Signed-off-by:
Paul E. McKenney <paulmck@linux.vnet.ibm.com> Tested-by:
Mike Galbraith <efault@gmx.de>
-
Paul E. McKenney authored
In the past, SRCU was simple enough that there was little point in making the rcutorture writer stall messages print the SRCU grace-period number state. With the advent of Tree SRCU, this has changed. This commit therefore makes Classic, Tiny, and Tree SRCU report this state to rcutorture as needed. Signed-off-by:
-
Paul E. McKenney authored
The current Tree SRCU implementation schedules a workqueue for every srcu_data covered by a given leaf srcu_node structure having callbacks, even if only one of those srcu_data structures actually contains callbacks. This is clearly inefficient for workloads that don't feature callbacks everywhere all the time. This commit therefore adds an array of masks that are used by the leaf srcu_node structures to track exactly which srcu_data structures contain callbacks. Signed-off-by:
-
- 24 Apr, 2017 2 commits
-
-
Paul E. McKenney authored
SRCU is optional, and included only if there is a "select SRCU" in effect. However, we now have Tiny SRCU, so this commit defaults CONFIG_SRCU=y. Reported-by:
kbuild test robot <fengguang.wu@intel.com> Signed-off-by:
-
Paul E. McKenney authored
If the CONFIG_SRCU option is not selected, for example, when building arch/tile allnoconfig, the following build errors appear: kernel/rcu/tree.o: In function `srcu_online_cpu': tree.c:(.text+0x4248): multiple definition of `srcu_online_cpu' kernel/rcu/srcutree.o:srcutree.c:(.text+0x2120): first defined here kernel/rcu/tree.o: In function `srcu_offline_cpu': tree.c:(.text+0x4250): multiple definition of `srcu_offline_cpu' kernel/rcu/srcutree.o:srcutree.c:(.text+0x2160): first defined here The corresponding .config file shows CONFIG_TREE_SRCU=y, but no sign of CONFIG_SRCU, which fatally confuses SRCU's #ifdefs, resulting in the above errors. The reason this occurs is the folowing line in init/Kconfig's definition for TREE_SRCU: default y if !TINY_RCU && !CLASSIC_SRCU If CONFIG_CLASSIC_SRCU=n, as it will be in for allnoconfig, and if CONFIG_SMP=y, then we will get CONFIG_TREE_SRCU=y but no CONFIG_SRCU, as seen in the .config file, and which will result in the above errors. This error did not show up during rcutorture testing because rcutorture forces CONFIG_SRCU=y, as it must to prevent build errors in rcutorture.c. This commit therefore conditions TREE_SRCU (and TINY_SRCU, while it is at it) with SRCU, like this: default y if SRCU && !TINY_RCU && !CLASSIC_SRCU Reported-by:
Ingo Molnar <mingo@kernel.org> Signed-off-by:
-
- 23 Apr, 2017 1 commit
-
-
Ingo Molnar authored
Merge branch 'for-mingo' of git://git.kernel.org/pub/scm/linux/kernel/git/paulmck/linux-rcu into core/rcu Pull RCU updates from Paul E. McKenney: - Documentation updates. - Miscellaneous fixes. - Parallelize SRCU callback handling (plus overlapping patches). Signed-off-by:
-
- 21 Apr, 2017 29 commits
-
-
git://linux-nfs.org/~bfields/linuxLinus Torvalds authored
Pull nfsd bugfix from Bruce Fields: "Fix a 4.11 regression that triggers a BUG() on an attempt to use an unsupported NFSv4 compound op" * tag 'nfsd-4.11-2' of git://linux-nfs.org/~bfields/linux: nfsd: fix oops on unsupported operation
-
git://git.kernel.org/pub/scm/linux/kernel/git/davem/netLinus Torvalds authored
Pull networking fixes from David Miller: 1) Don't race in IPSEC dumps, from Yuejie Shi. 2) Verify lengths properly in IPSEC reqeusts, from Herbert Xu. 3) Fix out of bounds access in ipv6 segment routing code, from David Lebrun. 4) Don't write into the header of cloned SKBs in smsc95xx driver, from James Hughes. 5) Several other drivers have this bug too, fix them. From Eric Dumazet. 6) Fix access to uninitialized data in TC action cookie code, from Wolfgang Bumiller. 7) Fix double free in IPV6 segment routing, again from David Lebrun. 8) Don't let userspace set the RTF_PCPU flag, oops. From David Ahern. 9) Fix use after free in qrtr code, from Dan Carpenter. 10) Don't double-destroy devices in ip6mr code, from Nikolay Aleksandrov. 11) Don't pass out-of-range TX queue indices into drivers, from Tushar Dave. * git://git.kernel.org/pub/scm/linux/kernel/git/davem/net: (30 commits) netpoll: Check for skb->queue_mapping ip6mr: fix notification device destruction bpf, doc: update bpf maintainers entry net: qrtr: potential use after free in qrtr_sendmsg() bpf: Fix values type used in test_maps net: ipv6: RTF_PCPU should not be settable from userspace gso: Validate assumption of frag_list segementation kaweth: use skb_cow_head() to deal with cloned skbs ch9200: use skb_cow_head() to deal with cloned skbs lan78xx: use skb_cow_head() to deal with cloned skbs sr9700: use skb_cow_head() to deal with cloned skbs cx82310_eth: use skb_cow_head() to deal with cloned skbs smsc75xx: use skb_cow_head() to deal with cloned skbs ipv6: sr: fix double free of skb after handling invalid SRH MAINTAINERS: Add "B:" field for networking. net sched actions: allocate act cookie early qed: Fix issue in populating the PFC config paramters. qed: Fix possible system hang in the dcbnl-getdcbx() path. qed: Fix sending an invalid PFC error mask to MFW. qed: Fix possible error in populating max_tc field. ... -
Tushar Dave authored
Reducing real_num_tx_queues needs to be in sync with skb queue_mapping otherwise skbs with queue_mapping greater than real_num_tx_queues can be sent to the underlying driver and can result in kernel panic. One such event is running netconsole and enabling VF on the same device. Or running netconsole and changing number of tx queues via ethtool on same device. e.g. Unable to handle kernel NULL pointer dereference tsk->{mm,active_mm}->context = 0000000000001525 tsk->{mm,active_mm}->pgd = fff800130ff9a000 \|/ ____ \|/ "@'/ .. \`@" /_| \__/ |_\ \__U_/ kworker/48:1(475): Oops [#1] CPU: 48 PID: 475 Comm: kworker/48:1 Tainted: G OE 4.11.0-rc3-davem-net+ #7 Workqueue: events queue_process task: fff80013113299c0 task.stack: fff800131132c000 TSTATE: 0000004480e01600 TPC: 00000000103f9e3c TNPC: 00000000103f9e40 Y: 00000000 Tainted: G OE TPC: <ixgbe_xmit_frame_ring+0x7c/0x6c0 [ixgbe]> g0: 0000000000000000 g1: 0000000000003fff g2: 0000000000000000 g3: 0000000000000001 g4: fff80013113299c0 g5: fff8001fa6808000 g6: fff800131132c000 g7: 00000000000000c0 o0: fff8001fa760c460 o1: fff8001311329a50 o2: fff8001fa7607504 o3: 0000000000000003 o4: fff8001f96e63a40 o5: fff8001311d77ec0 sp: fff800131132f0e1 ret_pc: 000000000049ed94 RPC: <set_next_entity+0x34/0xb80> l0: 0000000000000000 l1: 0000000000000800 l2: 0000000000000000 l3: 0000000000000000 l4: 000b2aa30e34b10d l5: 0000000000000000 l6: 0000000000000000 l7: fff8001fa7605028 i0: fff80013111a8a00 i1: fff80013155a0780 i2: 0000000000000000 i3: 0000000000000000 i4: 0000000000000000 i5: 0000000000100000 i6: fff800131132f1a1 i7: 00000000103fa4b0 I7: <ixgbe_xmit_frame+0x30/0xa0 [ixgbe]> Call Trace: [00000000103fa4b0] ixgbe_xmit_frame+0x30/0xa0 [ixgbe] [0000000000998c74] netpoll_start_xmit+0xf4/0x200 [0000000000998e10] queue_process+0x90/0x160 [0000000000485fa8] process_one_work+0x188/0x480 [0000000000486410] worker_thread+0x170/0x4c0 [000000000048c6b8] kthread+0xd8/0x120 [0000000000406064] ret_from_fork+0x1c/0x2c [0000000000000000] (null) Disabling lock debugging due to kernel taint Caller[00000000103fa4b0]: ixgbe_xmit_frame+0x30/0xa0 [ixgbe] Caller[0000000000998c74]: netpoll_start_xmit+0xf4/0x200 Caller[0000000000998e10]: queue_process+0x90/0x160 Caller[0000000000485fa8]: process_one_work+0x188/0x480 Caller[0000000000486410]: worker_thread+0x170/0x4c0 Caller[000000000048c6b8]: kthread+0xd8/0x120 Caller[0000000000406064]: ret_from_fork+0x1c/0x2c Caller[0000000000000000]: (null) Signed-off-by:Tushar Dave <tushar.n.dave@oracle.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Nikolay Aleksandrov authored
Andrey Konovalov reported a BUG caused by the ip6mr code which is caused because we call unregister_netdevice_many for a device that is already being destroyed. In IPv4's ipmr that has been resolved by two commits long time ago by introducing the "notify" parameter to the delete function and avoiding the unregister when called from a notifier, so let's do the same for ip6mr. The trace from Andrey: ------------[ cut here ]------------ kernel BUG at net/core/dev.c:6813! invalid opcode: 0000 [#1] SMP KASAN Modules linked in: CPU: 1 PID: 1165 Comm: kworker/u4:3 Not tainted 4.11.0-rc7+ #251 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Bochs 01/01/2011 Workqueue: netns cleanup_net task: ffff880069208000 task.stack: ffff8800692d8000 RIP: 0010:rollback_registered_many+0x348/0xeb0 net/core/dev.c:6813 RSP: 0018:ffff8800692de7f0 EFLAGS: 00010297 RAX: ffff880069208000 RBX: 0000000000000002 RCX: 0000000000000001 RDX: 0000000000000000 RSI: 0000000000000000 RDI: ffff88006af90569 RBP: ffff8800692de9f0 R08: ffff8800692dec60 R09: 0000000000000000 R10: 0000000000000006 R11: 0000000000000000 R12: ffff88006af90070 R13: ffff8800692debf0 R14: dffffc0000000000 R15: ffff88006af90000 FS: 0000000000000000(0000) GS:ffff88006cb00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 00007fe7e897d870 CR3: 00000000657e7000 CR4: 00000000000006e0 Call Trace: unregister_netdevice_many.part.105+0x87/0x440 net/core/dev.c:7881 unregister_netdevice_many+0xc8/0x120 net/core/dev.c:7880 ip6mr_device_event+0x362/0x3f0 net/ipv6/ip6mr.c:1346 notifier_call_chain+0x145/0x2f0 kernel/notifier.c:93 __raw_notifier_call_chain kernel/notifier.c:394 raw_notifier_call_chain+0x2d/0x40 kernel/notifier.c:401 call_netdevice_notifiers_info+0x51/0x90 net/core/dev.c:1647 call_netdevice_notifiers net/core/dev.c:1663 rollback_registered_many+0x919/0xeb0 net/core/dev.c:6841 unregister_netdevice_many.part.105+0x87/0x440 net/core/dev.c:7881 unregister_netdevice_many net/core/dev.c:7880 default_device_exit_batch+0x4fa/0x640 net/core/dev.c:8333 ops_exit_list.isra.4+0x100/0x150 net/core/net_namespace.c:144 cleanup_net+0x5a8/0xb40 net/core/net_namespace.c:463 process_one_work+0xc04/0x1c10 kernel/workqueue.c:2097 worker_thread+0x223/0x19c0 kernel/workqueue.c:2231 kthread+0x35e/0x430 kernel/kthread.c:231 ret_from_fork+0x31/0x40 arch/x86/entry/entry_64.S:430 Code: 3c 32 00 0f 85 70 0b 00 00 48 b8 00 02 00 00 00 00 ad de 49 89 47 78 e9 93 fe ff ff 49 8d 57 70 49 8d 5f 78 eb 9e e8 88 7a 14 fe <0f> 0b 48 8b 9d 28 fe ff ff e8 7a 7a 14 fe 48 b8 00 00 00 00 00 RIP: rollback_registered_many+0x348/0xeb0 RSP: ffff8800692de7f0 ---[ end trace e0b29c57e9b3292c ]--- Reported-by:
Andrey Konovalov <andreyknvl@google.com> Signed-off-by:
Nikolay Aleksandrov <nikolay@cumulusnetworks.com> Tested-by:
-
Daniel Borkmann authored
Add various related files that have been missing under BPF entry covering essential parts of its infrastructure and also add myself as co-maintainer. Signed-off-by:
Daniel Borkmann <daniel@iogearbox.net> Acked-by:
Alexei Starovoitov <ast@kernel.org> Signed-off-by:
-
Dan Carpenter authored
If skb_pad() fails then it frees the skb so we should check for errors. Fixes: bdabad3e ("net: Add Qualcomm IPC router") Signed-off-by:
Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by:
-
David Miller authored
Maps of per-cpu type have their value element size adjusted to 8 if it is specified smaller during various map operations. This makes test_maps as a 32-bit binary fail, in fact the kernel writes past the end of the value's array on the user's stack. To be quite honest, I think the kernel should reject creation of a per-cpu map that doesn't have a value size of at least 8 if that's what the kernel is going to silently adjust to later. If the user passed something smaller, it is a sizeof() calcualtion based upon the type they will actually use (just like in this testcase code) in later calls to the map operations. Fixes: df570f57 ("samples/bpf: unit test for BPF_MAP_TYPE_PERCPU_ARRAY") Signed-off-by:
-
David Ahern authored
Andrey reported a fault in the IPv6 route code: kasan: GPF could be caused by NULL-ptr deref or user memory access general protection fault: 0000 [#1] SMP KASAN Modules linked in: CPU: 1 PID: 4035 Comm: a.out Not tainted 4.11.0-rc7+ #250 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Bochs 01/01/2011 task: ffff880069809600 task.stack: ffff880062dc8000 RIP: 0010:ip6_rt_cache_alloc+0xa6/0x560 net/ipv6/route.c:975 RSP: 0018:ffff880062dced30 EFLAGS: 00010206 RAX: dffffc0000000000 RBX: ffff8800670561c0 RCX: 0000000000000006 RDX: 0000000000000003 RSI: ffff880062dcfb28 RDI: 0000000000000018 RBP: ffff880062dced68 R08: 0000000000000001 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000000 R12: 0000000000000000 R13: ffff880062dcfb28 R14: dffffc0000000000 R15: 0000000000000000 FS: 00007feebe37e7c0(0000) GS:ffff88006cb00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 00000000205a0fe4 CR3: 000000006b5c9000 CR4: 00000000000006e0 Call Trace: ip6_pol_route+0x1512/0x1f20 net/ipv6/route.c:1128 ip6_pol_route_output+0x4c/0x60 net/ipv6/route.c:1212 ... Andrey's syzkaller program passes rtmsg.rtmsg_flags with the RTF_PCPU bit set. Flags passed to the kernel are blindly copied to the allocated rt6_info by ip6_route_info_create making a newly inserted route appear as though it is a per-cpu route. ip6_rt_cache_alloc sees the flag set and expects rt->dst.from to be set - which it is not since it is not really a per-cpu copy. The subsequent call to __ip6_dst_alloc then generates the fault. Fix by checking for the flag and failing with EINVAL. Fixes: d52d3997 ("ipv6: Create percpu rt6_info") Reported-by:
David Ahern <dsa@cumulusnetworks.com> Acked-by:
Martin KaFai Lau <kafai@fb.com> Tested-by:
-
Ilan Tayari authored
Commit 07b26c94 ("gso: Support partial splitting at the frag_list pointer") assumes that all SKBs in a frag_list (except maybe the last one) contain the same amount of GSO payload. This assumption is not always correct, resulting in the following warning message in the log: skb_segment: too many frags For example, mlx5 driver in Striding RQ mode creates some RX SKBs with one frag, and some with 2 frags. After GRO, the frag_list SKBs end up having different amounts of payload. If this frag_list SKB is then forwarded, the aforementioned assumption is violated. Validate the assumption, and fall back to software GSO if it not true. Change-Id: Ia03983f4a47b6534dd987d7a2aad96d54d46d212 Fixes: 07b26c94 ("gso: Support partial splitting at the frag_list pointer") Signed-off-by:
Ilan Tayari <ilant@mellanox.com> Signed-off-by:
Ilya Lesokhin <ilyal@mellanox.com> Signed-off-by:
-
David S. Miller authored
Eric Dumazet says: ==================== net: use skb_cow_head() to deal with cloned skbs James Hughes found an issue with smsc95xx driver. Same problematic code is found in other drivers. ==================== Signed-off-by: -
Eric Dumazet authored
We can use skb_cow_head() to properly deal with clones, especially the ones coming from TCP stack that allow their head being modified. This avoids a copy. Signed-off-by:
Eric Dumazet <edumazet@google.com> Cc: James Hughes <james.hughes@raspberrypi.org> Signed-off-by:
-
Eric Dumazet authored
We need to ensure there is enough headroom to push extra header, but we also need to check if we are allowed to change headers. skb_cow_head() is the proper helper to deal with this. Fixes: 4a476bd6 ("usbnet: New driver for QinHeng CH9200 devices") Signed-off-by:
-
Eric Dumazet authored
We need to ensure there is enough headroom to push extra header, but we also need to check if we are allowed to change headers. skb_cow_head() is the proper helper to deal with this. Fixes: 55d7de9d ("Microchip's LAN7800 family USB 2/3 to 10/100/1000 Ethernet device driver") Signed-off-by:
-
Eric Dumazet authored
We need to ensure there is enough headroom to push extra header, but we also need to check if we are allowed to change headers. skb_cow_head() is the proper helper to deal with this. Fixes: c9b37458 ("USB2NET : SR9700 : One chip USB 1.1 USB2NET SR9700Device Driver Support") Signed-off-by:
-
Eric Dumazet authored
We need to ensure there is enough headroom to push extra header, but we also need to check if we are allowed to change headers. skb_cow_head() is the proper helper to deal with this. Fixes: cc28a20e ("introduce cx82310_eth: Conexant CX82310-based ADSL router USB ethernet driver") Signed-off-by:
-
Eric Dumazet authored
We need to ensure there is enough headroom to push extra header, but we also need to check if we are allowed to change headers. skb_cow_head() is the proper helper to deal with this. Fixes: d0cad871 ("smsc75xx: SMSC LAN75xx USB gigabit ethernet adapter driver") Signed-off-by:
-
David Lebrun authored
The icmpv6_param_prob() function already does a kfree_skb(), this patch removes the duplicate one. Fixes: 1ababeba ("ipv6: implement dataplane support for rthdr type 4 (Segment Routing Header)") Reported-by:
David Lebrun <david.lebrun@uclouvain.be> Signed-off-by:
-
git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linuxLinus Torvalds authored
Pull powerpc fixes from Michael Ellerman: "Just two fixes. The first fixes kprobing a stdu, and is marked for stable as it's been broken for ~ever. In hindsight this could have gone in next. The other is a fix for a change we merged this cycle, where if we take a certain exception when the kernel is running relocated (currently only used for kdump), we checkstop the box. Thanks to Ravi Bangoria" * tag 'powerpc-4.11-8' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux: powerpc/64: Fix HMI exception on LE with CONFIG_RELOCATABLE=y powerpc/kprobe: Fix oops when kprobed on 'stdu' instruction
-
git://git.kernel.org/pub/scm/linux/kernel/git/helgaas/pciLinus Torvalds authored
Pull PCI fix from Bjorn Helgaas: "Sorry this is so late. It's been in -next for over a week, but I forgot to send it on until now. A single fix to the DT binding of the HiSilicon PCIe host support" * tag 'pci-v4.11-fixes-5' of git://git.kernel.org/pub/scm/linux/kernel/git/helgaas/pci: PCI: hisi: Fix DT binding (hisi-pcie-almost-ecam)
-
git://git.kernel.dk/linux-blockLinus Torvalds authored
Pull block layer fixes from Jens Axboe: "A couple of last minute fixes for regressions in this cycle. More specifically: - Two patches from Andy, adjusting the NVMe APST quirks to avoid some issues specific to one Toshiba drive, and some variant of Samsung on two specific Dell laptops. - A fix for mtip32xx, turning off mq scheduling on that device. We have a real fix for this, but it's too late in the cycle. Thankfully we already have a NO_SCHED flag we can apply here. A prep patch for this is ensuring that we honor the NO_SCHED flag when attempting to online switch schedulers, previsouly we only did so for drive load time. From Ming. - Fixing an oops in blk-mq polling with scheduling attached. This one is easily reproducible, it would be a shame to release 4.11 with that issue. From me. I'd prefer not having to send in patches at this point in time, but the above are all things that have regressed in this cycle and the fixes are relatively straight forward" * 'for-linus' of git://git.kernel.dk/linux-block: blk-mq: fix potential oops with polling and blk-mq scheduler nvme: Quirk APST off on "THNSF5256GPUK TOSHIBA" nvme: Adjust the Samsung APST quirk mtip32xx: pass BLK_MQ_F_NO_SCHED block: respect BLK_MQ_F_NO_SCHED -
git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pmLinus Torvalds authored
Pull ACPI build fix from Rafael Wysocki: "This avoids a false-positive build warning from the compiler. Specifics: - Avoid a false-positive warning regarding a variable that may not be initialized that started to trigger after a previous general build fix (Arnd Bergmann)" * tag 'acpi-4.11-final' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm: ACPI / power: Avoid maybe-uninitialized warning -
git://git.kernel.org/pub/scm/linux/kernel/git/ulfh/mmcLinus Torvalds authored
Pull MMC fixes from Ulf Hansson: "MMC core: - kmalloc sdio scratch buffer to make it DMA-friendly MMC host: - dw_mmc: Fix behaviour for SDIO IRQs when runtime PM is used - sdhci-esdhc-imx: Correct pad I/O drive strength for UHS-DDR50 cards" * tag 'mmc-v4.11-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/ulfh/mmc: mmc: sdhci-esdhc-imx: increase the pad I/O drive strength for DDR50 card mmc: dw_mmc: Don't allow Runtime PM for SDIO cards mmc: sdio: fix alignment issue in struct sdio_func -
git://git.kernel.org/pub/scm/linux/kernel/git/dtor/inputLinus Torvalds authored
Pull input fixlet from Dmitry Torokhov: "An update to Elan PS/2 driver to allow working on yet another Lifebook" * 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input: Input: elantech - add Fujitsu Lifebook E547 to force crc_enabled
-
David S. Miller authored
We want people to report bugs to the netdev list. Signed-off-by: -
Paul E. McKenney authored
doc.2017.04.12a: Documentation updates fixes.2017.04.19a: Miscellaneous fixes srcu.2017.04.21a: Parallelize SRCU callback handling
-
Paul E. McKenney authored
Currently, a call to schedule() acts as a Tasks RCU quiescent state only if a context switch actually takes place. However, just the call to schedule() guarantees that the calling task has moved off of whatever tracing trampoline that it might have been one previously. This commit therefore plumbs schedule()'s "preempt" parameter into rcu_note_context_switch(), which then records the Tasks RCU quiescent state, but only if this call to schedule() was -not- due to a preemption. To avoid adding overhead to the common-case context-switch path, this commit hides the rcu_note_context_switch() check under an existing non-common-case check. Suggested-by:
Steven Rostedt <rostedt@goodmis.org> Signed-off-by:
-
Paul E. McKenney authored
Although Tree SRCU does reduce delays when there is at least one synchronize_srcu_expedited() invocation pending, srcu_schedule_cbs_snp() still waits for SRCU_INTERVAL before invoking callbacks. Since synchronize_srcu_expedited() now posts a callback and waits for that callback to do a wakeup, this destroys the expedited nature of synchronize_srcu_expedited(). This destruction became apparent to Marc Zyngier in the guise of a guest-OS bootup slowdown from five seconds to no fewer than forty seconds. This commit therefore invokes callbacks immediately at the end of the grace period when there is at least one synchronize_srcu_expedited() invocation pending. This brought Marc's guest-OS bootup times back into the realm of reason. Reported-by:
Marc Zyngier <marc.zyngier@arm.com> Signed-off-by:
-
Paul E. McKenney authored
Peter Zijlstra proposed using SRCU to reduce mmap_sem contention [1,2], however, there are workloads that could result in a high volume of concurrent invocations of call_srcu(), which with current SRCU would result in excessive lock contention on the srcu_struct structure's ->queue_lock, which protects SRCU's callback lists. This commit therefore moves SRCU to per-CPU callback lists, thus greatly reducing contention. Because a given SRCU instance no longer has a single centralized callback list, starting grace periods and invoking callbacks are both more complex than in the single-list Classic SRCU implementation. Starting grace periods and handling callbacks are now handled using an srcu_node tree that is in some ways similar to the rcu_node trees used by RCU-bh, RCU-preempt, and RCU-sched (for example, the srcu_node tree shape is controlled by exactly the same Kconfig options and boot parameters that control the shape of the rcu_node tree). In addition, the old per-CPU srcu_array structure is now named srcu_data and contains an rcu_segcblist structure named ->srcu_cblist for its callbacks (and a spinlock to protect this). The srcu_struct gets an srcu_gp_seq that is used to associate callback segments with the corresponding completion-time grace-period number. These completion-time grace-period numbers are propagated up the srcu_node tree so that the grace-period workqueue handler can determine whether additional grace periods are needed on the one hand and where to look for callbacks that are ready to be invoked. The srcu_barrier() function must now wait on all instances of the per-CPU ->srcu_cblist. Because each ->srcu_cblist is protected by ->lock, srcu_barrier() can remotely add the needed callbacks. In theory, it could also remotely start grace periods, but in practice doing so is complex and racy. And interestingly enough, it is never necessary for srcu_barrier() to start a grace period because srcu_barrier() only enqueues a callback when a callback is already present--and it turns out that a grace period has to have already been started for this pre-existing callback. Furthermore, it is only the callback that srcu_barrier() needs to wait on, not any particular grace period. Therefore, a new rcu_segcblist_entrain() function enqueues the srcu_barrier() function's callback into the same segment occupied by the last pre-existing callback in the list. The special case where all the pre-existing callbacks are on a different list (because they are in the process of being invoked) is handled by enqueuing srcu_barrier()'s callback into the RCU_DONE_TAIL segment, relying on the done-callbacks check that takes place after all callbacks are inovked. Note that the readers use the same algorithm as before. Note that there is a separate srcu_idx that tells the readers what counter to increment. This unfortunately cannot be combined with srcu_gp_seq because they need to be incremented at different times. This commit introduces some ugly #ifdefs in rcutorture. These will go away when I feel good enough about Tree SRCU to ditch Classic SRCU. Some crude performance comparisons, courtesy of a quickly hacked rcuperf asynchronous-grace-period capability: Callback Queuing Overhead ------------------------- # CPUS Classic SRCU Tree SRCU ------ ------------ --------- 2 0.349 us 0.342 us 16 31.66 us 0.4 us 41 --------- 0.417 us The times are the 90th percentiles, a statistic that was chosen to reject the overheads of the occasional srcu_barrier() call needed to avoid OOMing the test machine. The rcuperf test hangs when running Classic SRCU at 41 CPUs, hence the line of dashes. Despite the hacks to both the rcuperf code and that statistics, this is a convincing demonstration of Tree SRCU's performance and scalability advantages. [1] https://lwn.net/Articles/309030/ [2] https://patchwork.kernel.org/patch/5108281/Signed-off-by:
-
Paul E. McKenney authored
Parallelizing SRCU callback handling will increase the size of srcu_struct, which will move the kvm structure's kvm_arch field out of reach of powerpc's current assembly code, which will result in the following sort of build error: arch/powerpc/kvm/book3s_hv_rmhandlers.S:617: Error: operand out of range (0x000000000000b328 is not between 0xffffffffffff8000 and 0x0000000000007fff) This commit moves the srcu_struct fields in the kvm structure to follow the kvm_arch field, which will allow powerpc's assembly code to continue to be able to reach the kvm_arch field. Reported-by:
Stephen Rothwell <sfr@canb.auug.org.au> Reported-by:
Michael Ellerman <michaele@au1.ibm.com> Reported-by:
Paolo Bonzini <pbonzini@redhat.com> Signed-off-by:
Michael Ellerman <mpe@ellerman.id.au> Acked-by:
-
- 20 Apr, 2017 5 commits
-
-
Linus Torvalds authored
Merge two mm fixes from Andrew Morton. * emailed patches from Andrew Morton <akpm@linux-foundation.org>: mm: prevent NR_ISOLATE_* stats from going negative Revert "mm, page_alloc: only use per-cpu allocator for irq-safe requests"
-
Rabin Vincent authored
Commit 6afcf8ef ("mm, compaction: fix NR_ISOLATED_* stats for pfn based migration") moved the dec_node_page_state() call (along with the page_is_file_cache() call) to after putback_lru_page(). But page_is_file_cache() can change after putback_lru_page() is called, so it should be called before putback_lru_page(), as it was before that patch, to prevent NR_ISOLATE_* stats from going negative. Without this fix, non-CONFIG_SMP kernels end up hanging in the while(too_many_isolated()) { congestion_wait() } loop in shrink_active_list() due to the negative stats. Mem-Info: active_anon:32567 inactive_anon:121 isolated_anon:1 active_file:6066 inactive_file:6639 isolated_file:4294967295 ^^^^^^^^^^ unevictable:0 dirty:115 writeback:0 unstable:0 slab_reclaimable:2086 slab_unreclaimable:3167 mapped:3398 shmem:18366 pagetables:1145 bounce:0 free:1798 free_pcp:13 free_cma:0 Fixes: 6afcf8ef ("mm, compaction: fix NR_ISOLATED_* stats for pfn based migration") Link: http://lkml.kernel.org/r/1492683865-27549-1-git-send-email-rabin.vincent@axis.comSigned-off-by:
Rabin Vincent <rabinv@axis.com> Acked-by:
Michal Hocko <mhocko@suse.com> Cc: Ming Ling <ming.ling@spreadtrum.com> Cc: Minchan Kim <minchan@kernel.org> Cc: Vlastimil Babka <vbabka@suse.cz> Cc: <stable@vger.kernel.org> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
Mel Gorman authored
This reverts commit 374ad05a. While the patch worked great for userspace allocations, the fact that softirq loses the per-cpu allocator caused problems. It needs to be redone taking into account that a separate list is needed for hard/soft IRQs or alternatively find a cheap way of detecting reentry due to an interrupt. Both are possible but sufficiently tricky that it shouldn't be rushed. Jesper had one method for allowing softirqs but reported that the cost was high enough that it performed similarly to a plain revert. His figures for netperf TCP_STREAM were as follows Baseline v4.10.0 : 60316 Mbit/s Current 4.11.0-rc6: 47491 Mbit/s Jesper's patch : 60662 Mbit/s This patch : 60106 Mbit/s As this is a regression, I wish to revert to noirq allocator for now and go back to the drawing board. Link: http://lkml.kernel.org/r/20170415145350.ixy7vtrzdzve57mh@techsingularity.netSigned-off-by:
Mel Gorman <mgorman@techsingularity.net> Reported-by:
Tariq Toukan <ttoukan.linux@gmail.com> Acked-by:
Jesper Dangaard Brouer <brouer@redhat.com> Signed-off-by:
-
Jens Axboe authored
If we have a scheduler attached, blk_mq_tag_to_rq() on the scheduled tags will return NULL if a request is no longer in flight. This is different than using the normal tags, where it will always return the fixed request. Check for this condition for polling, in case we happen to enter polling for a completed request. The request address remains valid, so this check and return should be perfectly safe. Fixes: bd166ef1 ("blk-mq-sched: add framework for MQ capable IO schedulers") Tested-by:
Stephen Bates <sbates@raithlin.com> Signed-off-by:
Jens Axboe <axboe@fb.com>
-
Andy Lutomirski authored
There's a report that it malfunctions with APST on. See https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1678184 Cc: Kai-Heng Feng <kai.heng.feng@canonical.com> Signed-off-by:
Andy Lutomirski <luto@kernel.org> Signed-off-by:
-
