3.0.1pre1 on Red Hat 7.3 glibc-2.2.5-42 smbclient is much slower due to charset conversion. On the same host I use smbclient from 3.0 and from 2.2.7. I test with (note that the output of the command would be 49 lines long) time smbclient -s /tmp/file.conf -L host -Uunknown% >/dev/null 2>&1 First I try with file.conf with just "workgroup = workgroup". 2.2.7 takes about 0.300, 3.0.0 takes about 0.320 Then I try adding for 2.2.7: character set = iso8859-1 client code page = 850 for 3.0.0: unix charset = iso8859-15 display charset = iso8859-15 dos charset = cp850 time for 2.2.7 is almost unchanged, time for 3.0.0 is about 0.950. I understand the need to support unicode internally, but is there no way to make this better? I have many scripts that suddenly slow down. Thanks! $ ldd /usr/bin/smbclient libreadline.so.4 => /usr/lib/libreadline.so.4 (0x40013000) libncurses.so.5 => /usr/lib/libncurses.so.5 (0x4003a000) libcrypt.so.1 => /lib/libcrypt.so.1 (0x40087000) libresolv.so.2 => /lib/libresolv.so.2 (0x400b4000) libnsl.so.1 => /lib/libnsl.so.1 (0x400c4000) libdl.so.2 => /lib/libdl.so.2 (0x400d8000) libpopt.so.0 => /usr/lib/libpopt.so.0 (0x400db000) libc.so.6 => /lib/libc.so.6 (0x400e3000) /lib/ld-linux.so.2 => /lib/ld-linux.so.2 (0x40000000)
Startup time, particlularly when required for correctneess, it's really much of a concern. If your scripts are dramaticly slower, consier issuing multiple commands to each smbclient. By specifying a charset, you cause us to use iconv etc, which do take time to initialise.
iconv performance tuning is ongoing. Much work done since 3.0.0. Marking as fixed just so we can get some more feedback.
sorry for the same, cleaning up the database to prevent unecessary reopens of bugs.