Linux 5.1-rc7

Linux 5.1-rc7

[Linus Torvalds]

If rc6 was bigger than I wished, it really does seem to have been just
due to timing of pull requests. Because rc7 is tiny.

Just under half of the patch is various kinds of networking changes: a
mix of core networking, network drivers and some netfilter selftests.

The rest is mostly the usual architecture fixes, filesystems, and
other drivers (mostly rdma and gpu). And misc random changes
(Documentation, tracing, some mm fixlets).

But it's all pretty tiny. Plus about 30% of the patches are marked for
stable, so on the whole it really does feel like 5.1 is on target for
a regular release next weekend.

Knock wood.

           Linus

---

Achim Dahlhoff (1):
      dmaengine: sh: rcar-dmac: Fix glitch in dmaengine_tx_status

Alexander Potapenko (1):
      netfilter: conntrack: initialize ct->timeout

Alexey Kardashevskiy (2):
      powerpc/mm_iommu: Fix potential deadlock
      powerpc/mm_iommu: Allow pinning large regions

Amit Cohen (1):
      mlxsw: spectrum: Fix autoneg status in ethtool

Andrei Vagin (1):
      netfilter: fix nf_l4proto_log_invalid to log invalid packets

Andrey Ryabinin (2):
      mm/page_alloc.c: avoid potential NULL pointer dereference
      mm/page_alloc.c: fix never set ALLOC_NOFRAGMENT flag

Ard Biesheuvel (2):
      arm64/module: ftrace: deal with place relative nature of PLTs
      ARM: 8857/1: efi: enable CP15 DMB instructions before cleaning the cache

Arnd Bergmann (2):
      arch: add pidfd and io_uring syscalls everywhere
      s390: ctcm: fix ctcm_new_device error return code

Aurelien Jarno (1):
      MIPS: scall64-o32: Fix indirect syscall number load

Baolin Wang (1):
      gpio: eic: sprd: Fix incorrect irq type setting for the sync EIC

Bjorn Andersson (1):
      arm64: mm: Ensure tail of unaligned initrd is reserved

Borislav Petkov (1):
      x86/boot: Disable RSDP parsing temporarily

Christian König (1):
      drm/ttm: fix re-init of global structures

Colin Ian King (4):
      nfp: abm: fix spelling mistake "offseting" -> "offsetting"
      net: ipv6: addrlabel: fix spelling mistake "requewst" -> "request"
      firestream: fix spelling mistake "tramsitted" -> "transmitted"
      net: atheros: fix spelling mistake "underun" -> "underrun"

Corentin Labbe (2):
      dt-bindings: add an explanation for internal phy-mode
      Documentation: decnet: remove reference to CONFIG_DECNET_ROUTE_FWMARK

Dan Carpenter (1):
      netfilter: nf_tables: prevent shift wrap in nft_chain_parse_hook()

Dan Murphy (1):
      MAINTAINERS: LEDs: Add designated reviewer for LED subsystem

Daniel Gomez (2):
      spi: Micrel eth switch: declare missing of table
      spi: ST ST95HF NFC: declare missing of table

Dave Airlie (2):
      Revert "drm/i915/fbdev: Actually configure untiled displays"
      Revert "drm/virtio: drop prime import/export callbacks"

David Hildenbrand (1):
      mm/memory_hotplug.c: drop memory device reference after
find_memory_block()

Dirk Behme (1):
      dmaengine: sh: rcar-dmac: With cyclic DMA residue 0 is valid

Erez Alfasi (1):
      net/mlx5e: ethtool, Remove unsupported SFP EEPROM high pages query

Eric Dumazet (3):
      rxrpc: fix race condition in rxrpc_input_packet()
      net/rose: fix unbound loop in rose_loopback_timer()
      ipv4: add sanity checks in ipv4_link_failure()

Florian Fainelli (1):
      MAINTAINERS: BMIPS: Add internal Broadcom mailing list

Florian Westphal (6):
      selftests: netfilter: check icmp pkttoobig errors are set as related
      netfilter: conntrack: don't set related state for different outer address
      netfilter: ctnetlink: don't use conntrack/expect object addresses as id
      netfilter: nat: fix icmp id randomization
      netfilter: ebtables: CONFIG_COMPAT: drop a bogus WARN_ON
      netfilter: never get/set skb->tstamp

Frank Sorenson (1):
      cifs: do not attempt cifs operation on smb2+ rename error

Geert Uytterhoeven (1):
      gpio: Fix gpiochip_add_data_with_key() error path

Guy Levi (1):
      IB/mlx5: Fix scatter to CQE in DCT QP creation

Hangbin Liu (1):
      team: fix possible recursive locking when add slaves

Harry Pan (1):
      perf/x86/intel: Update KBL Package C-state events to also
include PC8/PC9/PC10 counters

Herbert Xu (2):
      crypto: xts - Fix atomic sleep when walking skcipher
      crypto: lrw - Fix atomic sleep when walking skcipher

Ido Schimmel (1):
      mlxsw: pci: Reincrease PCI reset timeout

Ilias Apalodimas (1):
      net: socionext: replace napi_alloc_frag with the netdev variant on init

Jacky Bai (1):
      Input: snvs_pwrkey - make it depend on ARCH_MXC

Jakub Kicinski (3):
      net/tls: fix refcount adjustment in fallback
      net/tls: avoid potential deadlock in tls_set_device_offload_rx()
      net/tls: don't leak IV and record seq when offload fails

Jann Horn (1):
      tracing: Fix buffer_ref pipe ops

Jason Gunthorpe (3):
      RDMA/mlx5: Do not allow the user to write to the clock page
      RDMA/mlx5: Use rdma_user_map_io for mapping BAR pages
      RDMA/ucontext: Fix regression with disassociate

Jeff Layton (5):
      nfsd: wake waiters blocked on file_lock before deleting it
      nfsd: wake blocked file lock waiters before sending callback
      ceph: only use d_name directly when parent is locked
      ceph: ensure d_name stability in ceph_dentry_hash()
      ceph: handle the case where a dentry has been renamed on outstanding req

Jens Axboe (2):
      io_uring: fail io_uring_register(2) on a dying io_uring instance
      io_uring: remove 'state' argument from io_{read,write} path

Jonas Karlman (2):
      drm: bridge: dw-hdmi: Fix overflow workaround for Rockchip SoCs
      drm/bridge: dw-hdmi: fix SCDC configuration for ddc-i2c-bus

Jonathan Neuschäfer (1):
      drm/sched: Fix description of drm_sched_stop

Josh Collier (1):
      IB/rdmavt: Fix frwr memory registration

Julian Anastasov (1):
      ipvs: do not schedule icmp errors from tunnels

Jérôme Glisse (2):
      cifs: fix page reference leak with readv/writev
      zram: pass down the bvec we need to read into in the work struct

Lijun Ou (1):
      RDMA/hns: Bugfix for mapping user db

Linus Torvalds (2):
      slip: make slhc_free() silently accept an error pointer
      Linux 5.1-rc7

Lucas Stach (3):
      gpu: ipu-v3: dp: fix CSC handling
      drm/imx: don't skip DP channel disable for background plane
      Input: synaptics-rmi4 - write config register values to the right offset

Maarten Lankhorst (2):
      drm/vc4: Fix memory leak during gpu reset.
      drm/vc4: Fix compilation error reported by kbuild test bot

Maxim Mikityanskiy (2):
      net/mlx5e: Fix use-after-free after xdp_return_frame
      net/mlx5e: Fix the max MTU check in case of XDP

Mel Gorman (2):
      mm: do not boost watermarks to avoid fragmentation for the
DISCONTIG memory model
      mm, page_alloc: always use a captured page regardless of compaction result

Miaohe Lin (1):
      net: vrf: Fix operation not supported when set vrf mac

Michael Ellerman (1):
      powerpc/mm/radix: Make Radix require HUGETLB_PAGE

Murphy Zhou (1):
      nfsd/nfsd3_proc_readdir: fix buffer count and page pointers

NeilBrown (1):
      sunrpc: don't mark uninitialised items as VALID.

Nikolay Borisov (1):
      btrfs: Switch memory allocations in async csum calculation path
to kvmalloc

Pan Bian (1):
      Input: synaptics-rmi4 - fix possible double free

Paul Kocialkowski (4):
      drm/sun4i: Add missing drm_atomic_helper_shutdown at driver unbind
      drm/sun4i: Set device driver data at bind time for use in unbind
      drm/sun4i: Fix component unbinding and component master deletion
      drm/sun4i: Unbind components before releasing DRM and memory

Peter Zijlstra (1):
      trace: Fix preempt_enable_no_resched() abuse

Petr Machata (1):
      mlxsw: spectrum: Put MC TCs into DWRR mode

Petr Štetiar (2):
      MIPS: perf: ath79: Fix perfcount IRQ assignment
      of_net: Fix residues after of_get_nvmem_mac_address removal

Po-Hsu Lin (2):
      selftests/net: correct the return value for run_netsocktests
      selftests/net: correct the return value for run_afpackettests

Qian Cai (1):
      x86/mm: Fix a crash with kmemleak_scan()

Ronnie Sahlberg (1):
      cifs: fix memory leak in SMB2_read

Russell King (1):
      ARM: fix function graph tracer and unwinder dependencies

Shun-Chih Yu (1):
      dmaengine: mediatek-cqdma: fix wrong register usage in mtk_cqdma_start

Stefan Bühler (3):
      io_uring: fix race condition reading SQ entries
      io_uring: fix race condition when sq threads goes sleeping
      io_uring: fix poll full SQ detection

Stefan Wahren (1):
      dmaengine: bcm2835: Avoid GFP_KERNEL in device_prep_slave_sg

Su Bao Cheng (1):
      stmmac: pci: Adjust IOT2000 matching

Tao Ren (1):
      net/ncsi: handle overflow when incrementing mac address

Thomas Hellstrom (1):
      drm/vmwgfx: Fix dma API layer violation

Tigran Tadevosyan (1):
      ARM: 8856/1: NOMMU: Fix CCR register faulty initialization when
MPU is disabled

Trond Myklebust (1):
      nfsd: Don't release the callback slot unless it was actually held

Uladzislau Rezki (Sony) (1):
      lib/test_vmalloc.c: do not create cpumask_t variable on stack

Ville Syrjälä (2):
      drm/i915: Do not enable FEC without DSC
      drm/i915: Restore correct bxt_ddi_phy_calc_lane_lat_optim_mask()
calculation

Vinod Koul (1):
      net: stmmac: move stmmac_check_ether_addr() to driver probe

Wenwen Wang (1):
      tracing: Fix a memory leak by early error exit in trace_pid_write()

Xie XiuQi (1):
      sched/numa: Fix a possible divide-by-zero

Yan, Zheng (1):
      ceph: fix ci->i_head_snapc leak

YueHaibing (3):
      net: stmmac: Use bfsize1 in ndesc_init_rx_desc
      lib/Kconfig.debug: fix build error without CONFIG_BLOCK
      fs/proc/proc_sysctl.c: Fix a NULL pointer dereference

ZhangXiaoxu (1):
      ipv4: set the tcp_min_rtt_wlen range from 0 to one day

Zhu Yanjun (1):
      net: rds: exchange of 8K and 1M pool

v5.1-rc7 fails to build on s390x (Re: Linux 5.1-rc7)

[Michal Kubecek]

v5.1-rc7 fails to build on s390x due to

		vmf->page = ZERO_PAGE(vmf->vm_start);

from commit 67f269b37f9b ("RDMA/ucontext: Fix regression with
disassociate"). This is not a problem on x86_64 where ZERO_PAGE()
doesn't use its argument but s390 version does.

I suppose the line should rather read

		vmf->page = ZERO_PAGE(vmf->vma->vm_start);

but I'm just guessing.

Michal Kubecek

Re: v5.1-rc7 fails to build on s390x (Re: Linux 5.1-rc7)

[Kirill A. Shutemov]

On Mon, Apr 29, 2019 at 07:21:36AM +0200, Michal Kubecek wrote:
> v5.1-rc7 fails to build on s390x due to
> 
> 		vmf->page = ZERO_PAGE(vmf->vm_start);
> 
> from commit 67f269b37f9b ("RDMA/ucontext: Fix regression with
> disassociate"). This is not a problem on x86_64 where ZERO_PAGE()
> doesn't use its argument but s390 version does.
> 
> I suppose the line should rather read
> 
> 		vmf->page = ZERO_PAGE(vmf->vma->vm_start);
> 
> but I'm just guessing.

I belive it has to be

		vmf->page = ZERO_PAGE(vmf->address);

-- 
 Kirill A. Shutemov