linux: 4.13.0-37.42 -proposed tracker

Task status switched to Incomplete, since there is a potential regression bug 1755073
And there is a deployment issue for an Artful i386 node, bug 1755149

4.13.0-37.42 - generic
Regression test CMPL.

Issue to note in amd64:
  hwclock - unable to change time back to 19th centry on lodygin (bug 1714229)
  libhugetlbfs - failed 7 (Heap did not shrink, small_const is not hugepage), killed by signal 7, bad config 1
  monotonic_time - lodygin constantly failed with tsc test (bug 1732620)
  ubuntu_bpf - test_maps failed (bug 1750724)
  ubuntu_kvm_unit_tests - 22 failed on amaura, 15 failed on daedalus, 3 failed on lodygin, 3 failed on michael, 21 failed on pepe
  ubuntu_lxc - lxc-test-ubuntu failed (Failed to start networking in ubuntu-cloud container) failed on daedalus and michael
  ubuntu_unionmount_overlayfs_suite - timed out on michael, passed after retest
  xfstests - timed out on ext4 generic/430 test with repo from upstream

Issue to note in arm64:
  hwclock - issue for HP m400 (bug 1716603)
  libhugetlbfs - failed 1 killed by signal 1 bad config 1
  ubuntu_bpf - failed to build on ARM64 (bug 1675395)
  ubuntu_kernel_selftests - breakpoint test failed to build (bug 1680507)
  ubuntu_zfs_xfs_generic - failed with test 079 (bug 1732370)

Issue to note in i386:
  libhugetlbfs - failed 4 killed by signal 4 bad config 1
  ubuntu_bpf - test_verifier failed on Artful i386 (bug 1755101) test_maps failed (bug 1750724)
  ubuntu_kvm_unit_tests - 14 failed on pepe (vmexit*, msr, syscall, tsc, hyperv_clock)

Issue to note in ppc64le:
  libhugetlbfs - failed 7 killed by signal 6
  ubuntu_kvm_smoke_test - uvtool isseu (bug 1755118)

Issue to note in s390x (zVM):
  aio_dio_bugs - event res -22 (bug 1730895)
  libhugetlbfs - failed 11 killed by signal 7 bad config 1
  scrashme - Test failed to build (bug 1689240)
  ubuntu_bpf - failed to build on s390x (bug 1711299)
  ubuntu_kvm_smoke_test - uvtool-libvirt needs arch specific template (bug 1729854)

Note: s390x zKVM was not tested in this cycle, as the DNS server (jiufeng) is down.

bug 1755073 is not considered to be a blocker for the promotion of this package, so I'll set 'Regression Testing' as 'Fix Released'.

Hardware Certification have begun testing this -proposed kernel.

Hardware Certification have completed testing this -proposed kernel. No regressions were observed, results are available here: http://people.canonical.com/~hwcert/sru-testing/artful/4.13.0-37.42/artful-proposed-published.html

This bug was fixed in the package linux - 4.13.0-37.42

---------------
linux (4.13.0-37.42) artful; urgency=medium

  * linux: 4.13.0-37.42 -proposed tracker (LP: #1751798)

  * CVE-2017-5715 // CVE-2017-5753 // CVE-2017-5754
    - arm64: Add ASM_BUG()
    - arm64: consistently use bl for C exception entry
    - arm64: move non-entry code out of .entry.text
    - arm64: unwind: avoid percpu indirection for irq stack
    - arm64: unwind: disregard frame.sp when validating frame pointer
    - arm64: mm: Fix set_memory_valid() declaration
    - arm64: Convert __inval_cache_range() to area-based
    - arm64: Expose DC CVAP to userspace
    - arm64: Handle trapped DC CVAP
    - arm64: Implement pmem API support
    - arm64: uaccess: Implement *_flushcache variants
    - arm64/vdso: Support mremap() for vDSO
    - arm64: unwind: reference pt_regs via embedded stack frame
    - arm64: unwind: remove sp from struct stackframe
    - arm64: uaccess: Add the uaccess_flushcache.c file
    - arm64: fix pmem interface definition
    - arm64: compat: Remove leftover variable declaration
    - fork: allow arch-override of VMAP stack alignment
    - arm64: kernel: remove {THREAD,IRQ_STACK}_START_SP
    - arm64: factor out PAGE_* and CONT_* definitions
    - arm64: clean up THREAD_* definitions
    - arm64: clean up irq stack definitions
    - arm64: move SEGMENT_ALIGN to <asm/memory.h>
    - efi/arm64: add EFI_KIMG_ALIGN
    - arm64: factor out entry stack manipulation
    - arm64: assembler: allow adr_this_cpu to use the stack pointer
    - arm64: use an irq stack pointer
    - arm64: add basic VMAP_STACK support
    - arm64: add on_accessible_stack()
    - arm64: add VMAP_STACK overflow detection
    - arm64: Convert pte handling from inline asm to using (cmp)xchg
    - kvm: arm64: Convert kvm_set_s2pte_readonly() from inline asm to cmpxchg()
    - arm64: Move PTE_RDONLY bit handling out of set_pte_at()
    - arm64: Ignore hardware dirty bit updates in ptep_set_wrprotect()
    - arm64: Remove the !CONFIG_ARM64_HW_AFDBM alternative code paths
    - arm64: introduce separated bits for mm_context_t flags
    - arm64: cleanup {COMPAT_,}SET_PERSONALITY() macro
    - KVM: arm/arm64: Fix guest external abort matching
    - KVM: arm/arm64: vgic: constify seq_operations and file_operations
    - KVM: arm/arm64: vITS: Drop its_ite->lpi field
    - KVM: arm/arm64: Extract GICv3 max APRn index calculation
    - KVM: arm/arm64: Support uaccess of GICC_APRn
    - arm64: Use larger stacks when KASAN is selected
    - arm64: Define cputype macros for Falkor CPU
    - arm64: SW PAN: Point saved ttbr0 at the zero page when switching to init_mm
    - arm64: SW PAN: Update saved ttbr0 value on enter_lazy_tlb
    - x86/syscalls: Check address limit on user-mode return
    - arm/syscalls: Check address limit on user-mode return
    - arm64/syscalls: Check address limit on user-mode return
    - Revert "arm/syscalls: Check address limit on user-mode return"
    - syscalls: Use CHECK_DATA_CORRUPTION for addr_limit_user_check
    - arm/syscalls: Optimize address limit check
    - arm64/syscalls: Move address limit check in loop
    - futex: Remove duplicated code and fix undefined...

The package has been published and the bug is being set to Fix Released