converted smbclient to use clientgen.c rather than clientutil.c

I did this when I saw yet another bug report complaining about
smbclient intermittently missing files. Rather than applying more
patches to smbclient it was better to move to the more robust
clientgen.c code.

The conversion wasn't perfect, I probably lost some features of
smbclient while doing it, but at least smbclient should be consistent
now. It if fails it should _always_ fail rather than giving people the
false impression of a reliable utility.

the tar stuff seems to work, but hasn't had much testing as I never
use it myself. I'm sure someone will find bugs in my conversion of
smbtar.c. It was quite tricky as it did a lot of its own SMB calls. It
now uses clientgen.c exclusively.

smbclient is still quite messy, but at least it doesn't build its own
SMB packets.

I haven't touched smbmount as I never use it. Mike, do you want to
convert smbmount to use clientgen.c?
(This used to be commit e14ca7765ace1b721dad8eca4a527a4e4a8f1ab8)

1 files changed, 1 insertions, 1 deletions

diff --git a/source3/smbd/dfree.c b/source3/smbd/dfree.c
index 020386645c..8cba8d0644 100644
--- a/source3/smbd/dfree.c
+++ b/source3/smbd/dfree.c
@@ -205,7 +205,7 @@ static SMB_BIG_UINT disk_free(char *path,SMB_BIG_UINT *bsize,SMB_BIG_UINT *dfree
 
 	/* FIXME : Any reason for this assumption ? */
 	if (*bsize < 256) {
-		DEBUG(5,("disk_free:Warning: bsize == %d < 256 . Changing to assumed correct bsize = 512\n",*bsize));
+		DEBUG(5,("disk_free:Warning: bsize == %d < 256 . Changing to assumed correct bsize = 512\n",(int)*bsize));
 		*bsize = 512;
 	}