This patch massively reduces the amount of heap allocations at
the interface/command layer. Most commands with minimal output
should not allocate memory from the heap at all. Things like
repeatedly polling status, currentsong, and volume changes
should be faster as a result, and more importantly, not a source
of memory fragmentation.
These changes should be safe in that there's no way for a
remote-client to corrupt memory or otherwise do bad stuff to
MPD, but an extra set of eyes to review would be good. Of
course there's never any warranty :)
No longer do we use FILE * structures in the interface, which means
we don't have to allocate any new memory for most connections.
Now, before you go on about losing the buffering that FILE *
+implies+, remember that myfprintf() never took advantage of
any of the stdio buffering features.
To reduce the diff and make bugs easier to spot in the diff,
I've kept myfprintf in places where we write to files (and not
network interfaces). Expect myfprintf to go away entirely soon
(we'll use fprintf for writing regular files).
git-svn-id: https://svn.musicpd.org/mpd/trunk@4483 09075e82-0dd4-0310-85a5-a0d7c8717e4f
These are just warnings from sparse, but it makes the output
easier to read. I ran this through a quick perl script, but
of course verified the output by looking at the diff and making
sure the thing still compiles.
here's the quick perl script I wrote to generate this patch:
----------- 8< -----------
use Tie::File;
defined(my $pid = open my $fh, '-|') or die $!;
if (!$pid) {
open STDERR, '>&STDOUT' or die $!;
exec 'sparse', @ARGV or die $!;
}
my $na = 'warning: non-ANSI function declaration of function';
while (<$fh>) {
print STDERR $_;
if (/^(.+?\.[ch]):(\d+):(\d+): $na '(\w+)'/o) {
my ($f, $l, $pos, $func) = ($1, $2, $3, $4);
$l--;
tie my @x, 'Tie::File', $f or die "$!: $f";
print '-', $x[$l], "\n";
$x[$l] =~ s/\b($func\s*)\(\s*\)/$1(void)/;
print '+', $x[$l], "\n";
untie @x;
}
}
git-svn-id: https://svn.musicpd.org/mpd/trunk@4378 09075e82-0dd4-0310-85a5-a0d7c8717e4f
intelligently use memmove, when inserting nodes in a sorted list
git-svn-id: https://svn.musicpd.org/mpd/trunk@2677 09075e82-0dd4-0310-85a5-a0d7c8717e4f
this code needs some serious testing:
Note:
The song name optimization i think is worth it, saves about 200k of ram on my syste, however, having to create directory names iteratively each time we print probably isn't worth the cpu. We only save about 10k of ram for the computer todo alot more work, and the code maybe a little messier
git-svn-id: https://svn.musicpd.org/mpd/trunk@2604 09075e82-0dd4-0310-85a5-a0d7c8717e4f
print functions. this way we can always know wtf is going on!
also, remove some places where we were using fprintf and printf instead of
myfprintf
git-svn-id: https://svn.musicpd.org/mpd/trunk@734 09075e82-0dd4-0310-85a5-a0d7c8717e4f
Here, i've made readDirectoryDB detect when stuff is deleted, added and updated.
So after a update, and we call redDirectoryDB, we update the db instead of
just adding stuff w/o "updating" and "deleting" stuff as neccessary.
git-svn-id: https://svn.musicpd.org/mpd/trunk@659 09075e82-0dd4-0310-85a5-a0d7c8717e4f
Add initial stuff for AAC support, now we just need to write the decoder
git-svn-id: https://svn.musicpd.org/mpd/trunk@264 09075e82-0dd4-0310-85a5-a0d7c8717e4f
also, if we can't get the time, then don't add the song to the db!
git-svn-id: https://svn.musicpd.org/mpd/trunk@236 09075e82-0dd4-0310-85a5-a0d7c8717e4f
