linux-kernel - [PATCH 0/9] restart_block: Prepare the ground for dumping timeout

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [thread-next>] [day] [month] [year] [list]

Message-Id: <20190909102340.8592-1-dima@arista.com>
Date:   Mon,  9 Sep 2019 11:23:31 +0100
From:   Dmitry Safonov <dima@...sta.com>
To:     linux-kernel@...r.kernel.org
Cc:     Dmitry Safonov <0x7f454c46@...il.com>,
        Dmitry Safonov <dima@...sta.com>,
        Adrian Reber <adrian@...as.de>,
        Alexander Viro <viro@...iv.linux.org.uk>,
        Andrei Vagin <avagin@...nvz.org>,
        Andy Lutomirski <luto@...nel.org>,
        Cyrill Gorcunov <gorcunov@...nvz.org>,
        Ingo Molnar <mingo@...hat.com>,
        Oleg Nesterov <oleg@...hat.com>,
        Pavel Emelyanov <xemul@...tuozzo.com>,
        Thomas Gleixner <tglx@...utronix.de>,
        containers@...ts.linux-foundation.org,
        linux-fsdevel@...r.kernel.org
Subject: [PATCH 0/9] restart_block: Prepare the ground for dumping timeout

Hi,

I'm trying to address an issue in CRIU (Checkpoint Restore In Userspace)
about timed syscalls restart. It's not possible to use restart_syscall()
as the majority of applications does, as after restore the kernel doesn't
know anything about a syscall that may have been interrupted on
checkpoint. That's because the tasks are re-created from scratch and so
there isn't task_struct::restart_block set on a new task.

As a preparation, unify timeouts for different syscalls in
restart_block.

On contrary, I'm struggling with patches that introduce the new ptrace()
request API. I'll speak about difficulties of designing new ptrace
operation on Containers Microconference at Plumbers [with a hope to
find the sensible solution].

Cc: Adrian Reber <adrian@...as.de>
Cc: Alexander Viro <viro@...iv.linux.org.uk>
Cc: Andrei Vagin <avagin@...nvz.org>
Cc: Andy Lutomirski <luto@...nel.org>
Cc: Cyrill Gorcunov <gorcunov@...nvz.org>
Cc: Dmitry Safonov <0x7f454c46@...il.com>
Cc: Ingo Molnar <mingo@...hat.com>
Cc: Oleg Nesterov <oleg@...hat.com>
Cc: Pavel Emelyanov <xemul@...tuozzo.com>
Cc: Thomas Gleixner <tglx@...utronix.de>
Cc: containers@...ts.linux-foundation.org
Cc: linux-fsdevel@...r.kernel.org

Dmitry Safonov (9):
  futex: Remove unused uaddr2 in restart_block
  restart_block: Prevent userspace set part of the block
  select: Convert __esimate_accuracy() to ktime_t
  select: Micro-optimise __estimate_accuracy()
  select: Convert select_estimate_accuracy() to take ktime_t
  select: Extract common code into do_sys_ppoll()
  select: Use ktime_t in do_sys_poll() and do_poll()
  select/restart_block: Convert poll's timeout to u64
  restart_block: Make common timeout

 fs/eventpoll.c                 |   4 +-
 fs/select.c                    | 214 ++++++++++++---------------------
 include/linux/poll.h           |   2 +-
 include/linux/restart_block.h  |  11 +-
 kernel/futex.c                 |  14 +--
 kernel/time/alarmtimer.c       |   6 +-
 kernel/time/hrtimer.c          |  14 ++-
 kernel/time/posix-cpu-timers.c |  10 +-
 kernel/time/posix-stubs.c      |   8 +-
 kernel/time/posix-timers.c     |   8 +-
 10 files changed, 115 insertions(+), 176 deletions(-)

-- 
2.23.0