> Bash on Linux seems to have problems starting subprocesses
> occasionally, the following program fails randomly:
> #!/bin/bash
> printf %5d 2
> i=3
> while /usr/bin/test $i -lt 10000 ; do
> j=3
> while /usr/bin/test $j -lt $i ; do
> if /usr/bin/test `expr $i % $j` -eq 0 ; then
> j=$i
> fi
> j=`expr $j + 2`
> done
> if /usr/bin/test $j -eq $i ; then
> printf %5d $i
> fi
> i=`expr $i + 2 `
> done
> typical output:
>
> # sh primes.sh
> 2 3 5 7 11 13 17 19 23 29 31 37 41 43 47 53 59 61 67 71 73 79 83 89 97 101 103 107 109 113 127 131 137 139 149 151 157 163 167 173 179 181 191 193 197 199 211 223 227 229 233 239 241 251 257 263 269 271 277 281 283 293 307 311 313 317 331 337 347 349 353 359 367 373 379 383 389 397 401 409 419 421 431 433 439 443 449 457 461 463 467 479 487 491 499 503 509 521 523 541 547 557 563 569 571 577 587 593 599 601 607 613 617 619 631 641 643 647 653 659 661 673 677 683 691 701 709 719 727 733 739 743 751 757 761 769 773 787 797 809 811 821 823 827 829 839 853 857 859primes.sh: Can't reopen pipe to command substitution (fd 4): No child processes
> /usr/bin/test: argument expected
> /usr/bin/test: argument expected
> primes.sh: Can't reopen pipe to command substitution (fd 4): No child processes
> /usr/bin/test: argument expected
> 877 881 883 887 907 911 919 929 937 941 947 953 967 971 977 983
Note that those errors are NOT bash errors, but errors from the
separate test program that you're calling.
That also leads to the obvious question: Why use an external test
program when bash has a faster one built in to it? Here's the above
script, with a common tweak to enable it to process the larger numbers
faster, and written to use bash's internal test command instead...
===8<=== CUT ===>8===
#!/bin/bash
printf %5d 2
declare -i i=3 j k=1 l
while [ $i -lt 10000 ]; do
j=3
l=`echo 2k$i v0.5+p | dc | cut -d . -f 1`
while [ $j -le $l ]; do
if [ $[$i%$j] -eq 0 ]; then
break
fi
j=$j+2
done
if [ $j -gt $l ]; then
printf %5d $i
k=$k+1
if [ $k -eq 15 ]; then
echo
k=0
fi
fi
i=$i+2
done
===8<=== CUT ===>8===
This version prints out the results much faster than yours does.
Incidentally, the 'tweak' referred to above is to recognise that any
number can only have a divisor greater than its square root if it also
has one smaller than its square root, so the inner loop only tests up
to and including the odd number nearest to the square root.
As an example of the speedup involved, when it's checking 9973 for
primality, it only needs to check for divisibility by one of 50
numbers to determine that, rather than checking for divisibility by
one of 4,986 numbers.
Incidentally, it determines that 9973 is prime.
Also, a second speedup is gained by moving the printing test outside
of the inner loop, and a third by declaring all numeric variables to
be numeric.
> I've tested this on 2.2.6 2.2.6ac1 2.2.9 and 2.2.10ac10 with
> bash versions 1.14 and 2.03. Machines tested were idle apart
> from the shellscript, with plenty of RAM.
This looks like a faulty test command to me - certainly, I was unable
to get the script to fail...
> This does not happen on 2.0.36, nor does this happen in bash on
> Solaris, Irix or Digital Unix.
Not sure what the problem is, but it's neither a kernel one nor a bash
one...
Best wishes from Riley.
+----------------------------------------------------------------------+
| There is something frustrating about the quality and speed of Linux |
| development, ie., the quality is too high and the speed is too high, |
| in other words, I can implement this XXXX feature, but I bet someone |
| else has already done so and is just about to release their patch. |
+----------------------------------------------------------------------+
* ftp://ftp.MemAlpha.cx/pub/rhw/Linux
* http://www.MemAlpha.cx/kernel.versions.html
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/