linux: 3.13.0-139.188 -proposed tracker

[Kleber Sacilotto de Souza]

This bug is for tracking the <version to be filled> upload package. This bug will contain status and testing results related to that upload.

For an explanation of the tasks and the associated workflow see: https://wiki.ubuntu.com/Kernel/kernel-sru-workflow

backports: 1741610
derivatives:
-- swm properties --
boot-testing-requested: true
phase: Released
proposed-announcement-sent: true
proposed-testing-requested: true

[Launchpad Janitor (janitor)]

This bug was fixed in the package linux - 3.13.0-139.188

---------------
linux (3.13.0-139.188) trusty; urgency=low

  * linux: 3.13.0-139.188 -proposed tracker (LP: #1741609)

  * CVE-2017-5754
    - perf/x86: Correctly use FEATURE_PDCM
    - arch: Introduce smp_load_acquire(), smp_store_release()
    - mm, x86: Account for TLB flushes only when debugging
    - x86/mm: Clean up inconsistencies when flushing TLB ranges
    - x86/mm: Eliminate redundant page table walk during TLB range flushing
    - mm, x86: Revisit tlb_flushall_shift tuning for page flushes except on
      IvyBridge
    - x86/mm: Clean up the TLB flushing code
    - x86/mm: Rip out complicated, out-of-date, buggy TLB flushing
    - x86/mm: Fix missed global TLB flush stat
    - x86/mm: New tunable for single vs full TLB flush
    - x86/mm: Set TLB flush tunable to sane value (33)
    - x86/mm: Fix sparse 'tlb_single_page_flush_ceiling' warning and make the
      variable read-mostly
    - rcu: Provide counterpart to rcu_dereference() for non-RCU situations
    - rcu: Move lockless_dereference() out of rcupdate.h
    - x86/ldt: Make modify_ldt synchronous
    - x86/ldt: Correct LDT access in single stepping logic
    - x86/ldt: Correct FPU emulation access to LDT
    - x86/ldt: Further fix FPU emulation
    - x86/mm: Disable preemption during CR3 read+write
    - x86: Clean up cr4 manipulation
    - x86/mm: Add INVPCID helpers
    - x86/mm: Fix INVPCID asm constraint
    - x86/mm: Add a 'noinvpcid' boot option to turn off INVPCID
    - x86/mm: If INVPCID is available, use it to flush global mappings
    - mm/mmu_context, sched/core: Fix mmu_context.h assumption
    - sched/core: Add switch_mm_irqs_off() and use it in the scheduler
    - x86/mm: Build arch/x86/mm/tlb.c even on !SMP
    - x86/mm, sched/core: Uninline switch_mm()
    - x86/mm, sched/core: Turn off IRQs in switch_mm()
    - sched/core: Idle_task_exit() shouldn't use switch_mm_irqs_off()
    - x86/irq: Do not substract irq_tlb_count from irq_call_count
    - x86/vm86/32: Switch to flush_tlb_mm_range() in mark_screen_rdonly()
    - x86/mm: Remove flush_tlb() and flush_tlb_current_task()
    - x86/mm: Make flush_tlb_mm_range() more predictable
    - x86/mm: Reimplement flush_tlb_page() using flush_tlb_mm_range()
    - x86/mm: Remove the UP asm/tlbflush.h code, always use the (formerly) SMP
      code
    - x86/mm: Disable PCID on 32-bit kernels
    - x86/mm: Add the 'nopcid' boot option to turn off PCID
    - x86/mm: Enable CR4.PCIDE on supported systems
    - x86/mm/64: Fix reboot interaction with CR4.PCIDE
    - KAISER: Kernel Address Isolation
    - x86/mm/kaiser: re-enable vsyscalls
    - kaiser: user_map __kprobes_text too
    - kaiser: alloc_ldt_struct() use get_zeroed_page()
    - x86/alternatives: Cleanup DPRINTK macro
    - x86/alternatives: Add instruction padding
    - x86/alternatives: Make JMPs more robust
    - x86/alternatives: Use optimized NOPs for padding
    - kaiser: add "nokaiser" boot option, using ALTERNATIVE
    - x86, boot: Carve out early cmdline parsing function
    - x86/boot: Fix early command-line parsing when matching at end
    - x86/boot: Fix early command-line parsing when partial ...

[Brad Figg (brad-figg)]

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

[A. Denton (aquina)]

> - kaiser: add "nokaiser" boot option, using ALTERNATIVE

O.k. for clarification; does GRUB_CMDLINE_LINUX_DEFAULT="nokaiser" in /etc/default/grub disable kaiser at boot time then?

[A. Denton (aquina)]

I found the following thread, which explains some things. I'll post it here in case someone else was looking for an answer.

https://askubuntu.com/questions/991874/how-to-disable-page-table-isolation-to-regain-performance-lost-due-to-intel-cpu