Question on mm: propagate error from stack expansion even for guard page

[shengyong]

å 2015/1/12 4:59, Linus Torvalds åé:
> Another week, another -rc.
> 
> Things have remained reasonably calm, although we also had a few
> last-minute MM regressions. Happily, most of them got fixed really
> quickly, with one remaining arm64 issue still pending.
> 
> So go forth and test some more. I'll be traveling for the next two
> weeks due to LCA, but I should have internet, and if things continue
> to be reasonably calm I don't think that my travel should be all that
> noticeable.  Finally the timing worked out, unlike several releases
> last year.
> 
> Anyway, the appended shortlog gives the details, but apart from the
> kgdb patches showing up as some uncommon work under kernel/debug/,
> things look fairly normal: mostly driver updates (gpu, pinctrl, hid,
> networking), architecture updates (mainly x86 this time, some minor
> arm[64] stuff), and some tooling fixes (mainly perf).
> 
>               Linus
> 
> ---
> 
> Aaron Plattner (1):
>       ALSA: hda - Add new GPU codec ID 0x10de0072 to snd-hda
> 
> Alex Deucher (4):
>       drm/radeon: KV has three PPLLs (v2)
>       drm/radeon: fix sad_count check for dce3
>       drm/radeon: properly filter DP1.2 4k modes on non-DP1.2 hw
>       drm/radeon: adjust default bapm settings for KV
> 
> Alex Thorlton (1):
>       sched: Fix KMALLOC_MAX_SIZE overflow during cpumask allocation
> 
> Alexandre Courbot (1):
>       drm/nouveau: fix missing return statement in nouveau_ttm_tt_unpopulate
> 
> Alexey Khoroshilov (1):
>       drm/radeon: do not leave queue acquired if timeout happens in
> kgd_hqd_destroy()
> 
> Andi Kleen (1):
>       perf/x86/uncore/hsw-ep: Handle systems with only two SBOXes
> 
> Andy Lutomirski (4):
>       x86_64, vdso: Fix the vdso address randomization algorithm
>       x86, vdso: Use asm volatile in __getcpu
>       perf: Move task_pt_regs sampling into arch code
>       perf/x86_64: Improve user regs sampling
> 
> Andy Shevchenko (1):
>       qla3xxx: don't allow never end busy loop
> 
> Anton Vorontsov (6):
>       kdb: Remove currently unused kdbtab_t->cmd_flags
>       kdb: Rename kdb_repeat_t to kdb_cmdflags_t, cmd_repeat to cmd_flags
>       kdb: Rename kdb_register_repeat() to kdb_register_flags()
>       kdb: Use KDB_REPEAT_* values as flags
>       kdb: Remove KDB_REPEAT_NONE flag
>       kdb: Add enable mask for groups of commands
> 
> Ard Biesheuvel (1):
>       arm64/efi: add missing call to early_ioremap_reset()
> 
> Ben Goz (5):
>       amdkfd: Fixing topology bug in building sysfs nodes
>       drm/amd: Fixing typos in kfd<->kgd interface
>       drm/amdkfd: Load mqd to hqd in non-HWS mode
>       drm/radeon: Assign VMID to PASID for IH in non-HWS mode
>       drm/amdkfd: unmap VMID<-->PASID when relesing VMID (non-HWS)
> 
> Ben Pfaff (1):
>       openvswitch: Consistently include VLAN header in flow and port stats.
> 
> Ben Skeggs (4):
>       drm/nouveau/bios: fix oops on pre-nv50 chipsets
>       drm/nouveau: wake up the card if necessary during gem callbacks
>       drm/nouveau/fb/ram/mcp77: subclass nouveau_ram
>       drm/nouveau/fb/ram/mcp77: use carveout reg to determine size
> 
> Benjamin Tissoires (1):
>       HID: logitech-hidpp: prefix the name with "Logitech"
> 
> BjÃrn Mork (1):
>       x86/build: Clean auto-generated processor feature files
> 
> Bruno PrÃmont (2):
>       drm/nouveau/nouveau: Do not BUG_ON(!spin_is_locked()) on UP
>       qla2xxx: fix busy wait regression
> 
> Chris Mason (2):
>       Btrfs: add more maintainers
>       Btrfs: don't delay inode ref updates during log replay
> 
> Chris Wilson (1):
>       mutex: Always clear owner field upon mutex_unlock()
> 
> Corey Minyard (1):
>       ipmi: Fix compile warning with tv_usec
> 
> Dan Carpenter (3):
>       Btrfs, scrub: uninitialized variable in scrub_extent_for_parity()
>       drm/radeon: integer underflow in radeon_cp_dispatch_texture()
>       HID: roccat: potential out of bounds in pyra_sysfs_write_settings()
> 
> Daniel Mack (1):
>       ALSA: snd-usb-caiaq: fix stream count check
> 
> Daniel Nicoletti (1):
>       HID: Allow HID_BATTERY_STRENGTH to be enabled
> 
> Daniel Thompson (3):
>       kdb: Categorize kdb commands (similar to SysRq categorization)
>       kdb: Allow access to sensitive commands to be restricted by default
>       kgdb: timeout if secondary CPUs ignore the roundup
> 
> David Drysdale (1):
>       vfs: renumber FMODE_NONOTIFY and add to uniqueness check
> 
> Doug Anderson (2):
>       pinctrl: rockchip: Handle wakeup pins
>       pinctrl: rockchip: Fix enable/disable/mask/unmask
> 
> Fabian Frederick (1):
>       kernel/debug/debug_core.c: Logging clean-up
> 
> Felipe Balbi (1):
>       net: ethernet: cpsw: fix hangs with interrupts
> 
> Filipe Manana (1):
>       Btrfs: correctly get tree level in tree_backref_for_extent
> 
> Francesco VIRLINZI (1):
>       pinctrl: st: avoid multiple mutex lock
> 
> Giedrius StatkeviÄius (1):
>       HID: Add a new id 0x501a for Genius MousePen i608X
> 
> Govindarajulu Varadarajan (1):
>       enic: free all rq buffs when allocation fails
> 
> Grygorii Strashko (1):
>       ARM: 8253/1: mm: use phys_addr_t type in map_lowmem() for kernel
> mem region
> 
> Hanjun Guo (2):
>       ACPI / processor: Convert apic_id to phys_id to make it arch agnostic
>       ACPI / processor: Rename acpi_(un)map_lsapic() to acpi_(un)map_cpu()
> 
> Hans de Goede (1):
>       ACPI / video: Add disable_native_backlight quirk for Dell XPS15 L521X
> 
> Henrik Rydberg (1):
>       MAINTAINERS: update rydberg's addresses
> 
> Herbert Xu (1):
>       tcp: Do not apply TSO segment limit to non-TSO packets
> 
> Ilia Mirkin (1):
>       drm/nv4c/mc: disable msi
> 
> Ilya Dryomov (2):
>       ceph: use %zu for len in ceph_fill_inline_data()
>       libceph: fix sparse endianness warnings
> 
> Imre Deak (1):
>       drm/i915: add missing rpm ref to i915_gem_pwrite_ioctl
> 
> J. Bruce Fields (1):
>       rpc: fix xdr_truncate_encode to handle buffer ending on page boundary
> 
> Jack Morgenstein (1):
>       net/mlx4_core: Fix error flow in mlx4_init_hca()
> 
> Jakub Wilk (1):
>       ext4: remove spurious KERN_INFO from ext4_warning call
> 
> Jan Beulich (1):
>       x86: Fix step size adjustment during initial memory mapping
> 
> Jeff Layton (1):
>       nfsd: fix fi_delegees leak when fi_had_conflict returns true
> 
> Joe Perches (1):
>       i40e: Fix possible memory leak in i40e_dbg_dump_desc
> 
> Joerg Roedel (2):
>       iommu/vt-d: Fix dmar_domain leak in iommu_attach_device
>       iommu/vt-d: Remove dead code in device_notifier
> 
> Johannes Berg (1):
>       Revert "mac80211: Fix accounting of the tailroom-needed counter"
> 
> Johannes Weiner (2):
>       mm: protect set_page_dirty() from ongoing truncation
>       mm: memcontrol: switch soft limit default back to infinity
> 
> John W. Linville (1):
>       e100: fix typo in MDI/MDI-X eeprom check in e100_phy_init
> 
> Joonsoo Kim (1):
>       mm/debug_pagealloc: remove obsolete Kconfig options
> 
> Josef Bacik (1):
>       Btrfs: abort transaction if we don't find the block group
> 
> Joseph Qi (1):
>       ocfs2: remove bogus check in dlm_process_recovery_data
> 
> Julia Lawall (4):
>       net: Xilinx: fix error return code
>       myri10ge: fix error return code
>       net: sun4i-emac: fix error return code
>       net: axienet: fix error return code
> 
> Karl Relton (1):
>       HID: add battery quirk for
> USB_DEVICE_ID_APPLE_ALU_WIRELESS_2011_ISO keyboard
> 
> Kirill Smelkov (1):
>       tools/liblockdep: Fix debug_check thinko in mutex destroy
> 
> Konstantin Khlebnikov (2):
>       mm: prevent endless growth of anon_vma hierarchy
>       mm: fix corner case in anon_vma endless growing prevention
> 
> Kostya Belezko (1):
>       Altera TSE: Add missing phydev
> 
> Kristian Evensen (1):
>       qmi_wwan: Set random MAC on devices with buggy fw
> 
> Laurent Pinchart (1):
>       iommu/ipmmu-vmsa: Change IOMMU_EXEC to IOMMU_NOEXEC
> 
> Linus LÃssing (4):
>       batman-adv: fix delayed foreign originator recognition
>       batman-adv: fix counter for multicast supporting nodes
>       batman-adv: fix multicast counter when purging originators
>       batman-adv: fix potential TT client + orig-node memory leak
> 
> Linus Torvalds (3):
>       mm: propagate error from stack expansion even for guard page
hi,

the commit `mm: propagate error from stack expansion even for guard page'
changes the return value of check_stack_guard_page(). In do_anonymous_page(),
it only returns VM_FAULT_SIGBUS if check_stack_guard_page() gets error, and
this leads to the userspace tasks get `Bus error'.

But in the situation,
    int main(void)
    {
        main();
        return 0;
    }

the infinite recursion exceeds the stack limit, and it causes `segmentaion
fault' in older kernels. So shall we distinguish different return values
of check_stack_guard_page()?

thanks,
Sheng Yong

>       mm: Don't count the stack guard page towards RLIMIT_STACK
>       linux 3.19-rc4
> 
> Luca Abeni (2):
>       sched/deadline: Fix migration of SCHED_DEADLINE tasks
>       sched/deadline: Avoid double-accounting in case of missed deadlines
> 
> Maor Gottlieb (1):
>       net/mlx4_core: Correcly update the mtt's offset in the MR re-reg flow
> 
> Mark Rutland (3):
>       arm64: Correct __NR_compat_syscalls for bpf
>       arm64: sanity checks: add missing AArch32 registers
>       ARM: 8249/1: mm: dump: don't skip regions
> 
> Martin HundebÃll (1):
>       batman-adv: fix lock class for decoding hash in network-coding.c
> 
> Masami Hiramatsu (1):
>       perf probe: Fix to fall back to find probe point in symbols
> 
> Mathias Krause (1):
>       crypto: aesni - fix "by8" variant for 128 bit keys
> 
> Michael S. Tsirkin (4):
>       virtio: make del_vqs idempotent
>       virtio_pci: device-specific release callback
>       virtio_pci: document why we defer kfree
>       vhost/net: length miscalculation
> 
> Michal Marek (2):
>       kbuild: Fix removal of the debian/ directory
>       Makefile: include arch/*/include/generated/uapi before .../generated
> 
> Mika Westerberg (1):
>       HID: i2c-hid: Do not free buffers in i2c_hid_stop()
> 
> Namhyung Kim (8):
>       perf report: Show progress bar for output resorting
>       perf ui/tui: Print backtrace symbols when segfault occurs
>       perf callchain: Append callchains only when requested
>       perf probe: Fix crash in dwarf_getcfi_elf
>       perf diff: Fix to sort by baseline field by default
>       perf hists: Fix children sort key behavior
>       perf callchain: Free callchains when hist entries are deleted
>       perf hists browser: Fix segfault when showing callchain
> 
> Oded Gabbay (8):
>       amdkfd: Fix accounting of device queues
>       amdkfd: Remove duplicate include
>       drivers: Move iommu/ before gpu/ in Makefile
>       drm: Put amdkfd before radeon in drm Makefile
>       drm/radeon: Init amdkfd only if it was compiled
>       drm/amdkfd: Do copy_to/from_user in general kfd_ioctl()
>       drm/amdkfd: reformat IOCTL definitions to drm-style
>       drm/amdkfd: rewrite kfd_ioctl() according to drm_ioctl()
> 
> Oleg Nesterov (2):
>       exit: fix race between wait_consider_task() and wait_task_zombie()
>       arch/blackfin/mach-bf533/boards/stamp.c: add linux/delay.h
> 
> Palik, Imre (1):
>       xen-netback: fixing the propagation of the transmit shaper timeout
> 
> Patrice CHOTARD (1):
>       pinctrl: st: Add irq_disable hook to st_gpio_irqchip
> 
> Paul Walmsley (4):
>       arm64: fix missing asm/pgtable-hwdef.h include in asm/processor.h
>       arm64: fix missing linux/bug.h include in asm/arch_timer.h
>       arm64: fix missing asm/alternative.h include in kernel/module.c
>       arm64: fix missing asm/io.h include in kernel/smp_spin_table.c
> 
> Peter Wu (3):
>       HID: logitech-dj: check report length
>       HID: logitech-hidpp: check WTP report length
>       HID: logitech-hidpp: avoid unintended fall-through
> 
> Peter Zijlstra (2):
>       x86: Fix off-by-one in instruction decoder
>       sched, fanotify: Deal with nested sleeps
> 
> Pierre Moreau (1):
>       drm/nouveau/fb/ram/mcp77: enable NISO poller
> 
> Pranith Kumar (1):
>       assoc_array: Include rcupdate.h for call_rcu() definition
> 
> Rafael J. Wysocki (1):
>       ACPI / PM: Fix PM initialization for devices that are not present
> 
> Russell King (1):
>       ARM: wire up execveat syscall
> 
> Sasha Levin (3):
>       amdkfd: actually allocate longs for the pasid bitmask
>       KEYS: close race between key lookup and freeing
>       virtio_pci: defer kfree until release callback
> 
> Simon Wunderlich (1):
>       batman-adv: fix and simplify condition when bonding should be used
> 
> Sven KÃhler (1):
>       drm/nouveau/device: Add support for GK208B, resolves bug 86935
> 
> Sylvain BERTRAND (1):
>       x86: Fix mkcapflags.sh bash-ism
> 
> Taesoo Kim (1):
>       perf list: Fix --raw-dump option
> 
> Takashi Iwai (1):
>       ALSA: hda - Fix wrong gpio_dir & gpio_mask hint setups for IDT/STAC codecs
> 
> Takashi Sakamoto (1):
>       ALSA: fireworks: fix an endianness bug for transaction length
> 
> Tetsuo Handa (1):
>       sched/fair: Fix RCU stall upon -ENOMEM in sched_create_group()
> 
> Theodore Ts'o (2):
>       ext4: prevent online resize with backup superblock
>       Revert "ext4: fix suboptimal seek_{data,hole} extents traversial"
> 
> Tobias Klauser (1):
>       arm64: Remove unused prepare_to_copy()
> 
> Todd Fujinaka (1):
>       igb: Remove unneeded FIXME
> 
> Victor Kamensky (1):
>       ARM: 8275/1: mm: fix PMD_SECT_RDONLY undeclared compile error
> 
> Ville SyrjÃlà (3):
>       drm/i915: Kill check_power_well() calls
>       drm/i915: Don't call intel_prepare_page_flip() multiple times on gen2-4
>       Revert "drm/i915: Preserve VGACNTR bits from the BIOS"
> 
> Vinson Lee (1):
>       crypto: sha-mb - Add avx2_supported check.
> 
> Vladimir Davydov (1):
>       memcg: fix destination cgroup leak on task charges migration
> 
> Vlastimil Babka (1):
>       mm, vmscan: prevent kswapd livelock due to pfmemalloc-throttled
> process being killed
> 
> Wang Shilong (1):
>       Btrfs: call inode_dec_link_count() on mkdir error path
> 
> Wei Yang (1):
>       vfio-pci: Fix the check on pci device type in vfio_pci_probe()
> 
> Wolfram Sang (1):
>       iommu/rockchip: Drop owner assignment from platform_drivers
> 
> Xue jiufei (1):
>       ocfs2: fix the wrong directory passed to
> ocfs2_lookup_ino_from_name() when link file
> 
> Yongjian Xu (1):
>       qlcnic: Fix return value in qlcnic_probe()
> 
> Yuyang Du (1):
>       sched: Fix odd values in effective_load() calculations
> 
> hayeswang (1):
>       r8152: support ndo_features_check
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@xxxxxxxxxxxxxxx
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/
> 
> .
> 

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/