Re: bug#7489: [coreutils] over aggressive threads in sort

[Jim Meyering]

Chen Guo wrote:

> Hi Professor Eggert,
> On Mon, Nov 29, 2010 at 11:16 AM, Paul Eggert <address@hidden> wrote:
>>  (for i in $(seq 12); do read line; echo $i; sleep .1; done
>>  cat > /dev/null) < fifo &
>>  (ulimit -t 1; ./sort in > fifo \
>>  || echo killed via $(env kill -l $(expr $? - 128)))
>
> I ran this 10 times or so on an i7 and couldn't trigger anything. Is
> seq 12 supposed to vary depending on the number of cores?

Hi Chen,

Here's a stand-alone command-line test using the sort in your PATH:

    rm -f in fifo
    seq 100000 > in
    mkfifo fifo
    (for i in $(seq 12); do read line; echo $i; sleep .1; done
      cat > /dev/null) < fifo &
    (ulimit -t 1; sort in > fifo \
      || echo killed via $(env kill -l $(expr $? - 128)))

The 12 x 0.1-second sleeps are to ensure that the busy-waiting
sort will accumulate more than 1.0 seconds of CPU time, and thus
surpass the CPU time limit imposed by "ulimit -t 1".

With more processes, then you may use a number smaller than 12.
Running on a 6-core i7, I see the "killed via XCPU" message
anywhere between the "1" and "4" output lines.  E.g.,

    1
    2
    killed via XCPU
    $ 3                                                                         
 :
    4
    5
    6
    7
    8
    9
    10
    11
    12