Re: [Monotone-devel] oprofile data for mtn 0.37.

[Zack Weinberg]

On Wed, Mar 12, 2008 at 5:46 PM, Jack Lloyd <address@hidden> wrote:
> On Wed, Mar 12, 2008 at 05:19:50PM -0400, Zack Weinberg wrote:
>  > It occurs to me that we (monotone) create and destroy one Botan::Pipe
>  > and at least one subclass of Botan::Filter for every call to most of
>  > the functions in transform.hh.  If there's a global lock around
>  > allocating the memory for that, that could explain it.
>
>  Yup. At least one and actually probably several.
[...]

Given this, it occurred to me to try to cache constructed Pipe
objects, so that we avoid both the locking overhead and all other
construction overhead, in particular heap-allocation of memory blocks
&c.

Unfortunately, I can't make this work.  My code uses the Botan APIs in
the same way as this test program:

#include <string>
#include <iostream>
#include <botan/botan.h>

using std::string;
using std::cin;
using std::cout;
using Botan::Pipe;
using Botan::Hash_Filter;
using Botan::Hex_Encoder;

int
main(void)
{
  Pipe p(new Hash_Filter("SHA-160"), new Hex_Encoder);
  string buf;

  for (;;)
    {
      getline(cin, buf);
      if (cin.eof()) break;
      cout << '[' << buf << ']' << '\n';
      p.process_msg(buf);
      cout << '|' << (p.read_all_as_string()) << '|' << '\n';
    }
}

which only produces the SHA1 hash of the first line of input; the
intent is to hash each line separately.

$ cat test.txt
the quick brown fox jumps over the lazy dog
pack my box with five dozen liquor jugs
waltz, nymph, for quick jigs vex Bud
$ ./a.out < test.txt
[the quick brown fox jumps over the lazy dog]
|16312751EF9307C3FD1AFBCB993CDC80464BA0F1|
[pack my box with five dozen liquor jugs]
||
[waltz, nymph, for quick jigs vex Bud]
||

What am I doing wrong?
zw