netdev - Re: Bisected softirq accounting issue in v4.11-rc1~170^2~28

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-prev] [thread-next>] [day] [month] [year] [list]
Date:   Tue, 28 Mar 2017 18:34:52 +0800
From:   Wanpeng Li <kernellwp@...il.com>
To:     Jesper Dangaard Brouer <brouer@...hat.com>
Cc:     Frederic Weisbecker <fweisbec@...il.com>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        "netdev@...r.kernel.org" <netdev@...r.kernel.org>,
        linux-mm <linux-mm@...ck.org>,
        Mel Gorman <mgorman@...hsingularity.net>,
        Tariq Toukan <tariqt@...lanox.com>,
        Tariq Toukan <ttoukan.linux@...il.com>,
        Peter Zijlstra <peterz@...radead.org>,
        Rik van Riel <riel@...hat.com>,
        Thomas Gleixner <tglx@...utronix.de>,
        Ingo Molnar <mingo@...nel.org>
Subject: Re: Bisected softirq accounting issue in v4.11-rc1~170^2~28

2017-03-28 16:14 GMT+08:00 Jesper Dangaard Brouer <brouer@...hat.com>:
>
> (While evaluating some changes to the page allocator) I ran into an
> issue with ksoftirqd getting too much CPU sched time.
>
> I bisected the problem to
>  a499a5a14dbd ("sched/cputime: Increment kcpustat directly on irqtime account")
>
>  a499a5a14dbd1d0315a96fc62a8798059325e9e6 is the first bad commit
>  commit a499a5a14dbd1d0315a96fc62a8798059325e9e6
>  Author: Frederic Weisbecker <fweisbec@...il.com>
>  Date:   Tue Jan 31 04:09:32 2017 +0100
>
>     sched/cputime: Increment kcpustat directly on irqtime account
>
>     The irqtime is accounted is nsecs and stored in
>     cpu_irq_time.hardirq_time and cpu_irq_time.softirq_time. Once the
>     accumulated amount reaches a new jiffy, this one gets accounted to the
>     kcpustat.
>
>     This was necessary when kcpustat was stored in cputime_t, which could at
>     worst have jiffies granularity. But now kcpustat is stored in nsecs
>     so this whole discretization game with temporary irqtime storage has
>     become unnecessary.
>
>     We can now directly account the irqtime to the kcpustat.
>
>     Signed-off-by: Frederic Weisbecker <fweisbec@...il.com>
>     Cc: Benjamin Herrenschmidt <benh@...nel.crashing.org>
>     Cc: Fenghua Yu <fenghua.yu@...el.com>
>     Cc: Heiko Carstens <heiko.carstens@...ibm.com>
>     Cc: Linus Torvalds <torvalds@...ux-foundation.org>
>     Cc: Martin Schwidefsky <schwidefsky@...ibm.com>
>     Cc: Michael Ellerman <mpe@...erman.id.au>
>     Cc: Paul Mackerras <paulus@...ba.org>
>     Cc: Peter Zijlstra <peterz@...radead.org>
>     Cc: Rik van Riel <riel@...hat.com>
>     Cc: Stanislaw Gruszka <sgruszka@...hat.com>
>     Cc: Thomas Gleixner <tglx@...utronix.de>
>     Cc: Tony Luck <tony.luck@...el.com>
>     Cc: Wanpeng Li <wanpeng.li@...mail.com>
>     Link: http://lkml.kernel.org/r/1485832191-26889-17-git-send-email-fweisbec@gmail.com
>     Signed-off-by: Ingo Molnar <mingo@...nel.org>
>
> The reproducer is running a userspace udp_sink[1] program, and taskset
> pinning the process to the same CPU as softirq RX is running on, and
> starting a UDP flood with pktgen (tool part of kernel tree:
> samples/pktgen/pktgen_sample03_burst_single_flow.sh).
>
> [1] udp_sink
>  https://github.com/netoptimizer/network-testing/blob/master/src/udp_sink.c
>
> The expected results (after commit 4cd13c21b207 ("softirq: Let
> ksoftirqd do its job")) is that the scheduler split the CPU time 50/50
> between udp_sink and ksoftirqd.
>
> After this commit, the udp_sink program does not get any sched CPU
> time, and no packets are delivered to userspace.  (All packets are
> dropped by softirq due to a full socket queue, nstat UdpRcvbufErrors).
>
> A related symptom is that ksoftirqd no longer get accounted in top.
>
> $ grep CONFIG_IRQ_TIME_ACCOUNTING .config
> CONFIG_IRQ_TIME_ACCOUNTING=y
>
> Full .config uploaded here[2]:
>  [2] http://people.netfilter.org/hawk/kconfig/config02-bisect-softirq-a499a5a14dbd

void irqtime_account_irq(struct task_struct *curr)
{
    struct irqtime *irqtime = this_cpu_ptr(&cpu_irqtime);
    u64 *cpustat = kcpustat_this_cpu->cpustat;
    s64 delta;
    int cpu;

    if (!sched_clock_irqtime)
        return;

    cpu = smp_processor_id();
    delta = sched_clock_cpu(cpu) - irqtime->irq_start_time;

sched_clock_cpu(cpu) should be converted from cputime to ns. However,
cputime_to_nsecs() API is removed by
https://lkml.org/lkml/2017/1/22/230 for generic usage, so could you
try the below patch just for testing?

diff --git a/include/linux/sched/cputime.h b/include/linux/sched/cputime.h
index 4c5b973..166efba 100644
--- a/include/linux/sched/cputime.h
+++ b/include/linux/sched/cputime.h
@@ -7,14 +7,14 @@
  * cputime accounting APIs:
  */

-#ifdef CONFIG_VIRT_CPU_ACCOUNTING_NATIVE
-#include <asm/cputime.h>
+#define cputime_div(__ct, divisor) div_u64((__force u64)__ct, divisor)
+#define cputime_to_usecs(__ct) \
+    cputime_div(__ct, NSEC_PER_USEC)

 #ifndef cputime_to_nsecs
 # define cputime_to_nsecs(__ct)    \
     (cputime_to_usecs(__ct) * NSEC_PER_USEC)
 #endif
-#endif /* CONFIG_VIRT_CPU_ACCOUNTING_NATIVE */

 #ifdef CONFIG_VIRT_CPU_ACCOUNTING_GEN
 extern void task_cputime(struct task_struct *t,
diff --git a/kernel/sched/cputime.c b/kernel/sched/cputime.c
index f3778e2b..68064d1 100644
--- a/kernel/sched/cputime.c
+++ b/kernel/sched/cputime.c
@@ -49,7 +49,7 @@ void irqtime_account_irq(struct task_struct *curr)
         return;

     cpu = smp_processor_id();
-    delta = sched_clock_cpu(cpu) - irqtime->irq_start_time;
+    delta = cputime_to_nsecs(sched_clock_cpu(cpu)) - irqtime->irq_start_time;
     irqtime->irq_start_time += delta;

     u64_stats_update_begin(&irqtime->sync);

Regards,
Wanpeng Li