Compiling on Solaris 7 w/ spro 2.0

Grant Beattie grant at grunta.com
Tue Sep 5 10:09:30 GMT 2000


On Tue, Sep 05, 2000 at 04:27:16AM -0500, Peter Samuelson wrote:

> > smbd/fileio.c, line 220: compiler error: no table entry for op REG
> > cc: iropt failed, iropt & cg passes skipped for smbd/fileio.c
> 
> Exact version of Samba?  The line number, by itself, means little.  My
> version of Samba doesn't seem to have anything suspicious on line 220
> of that file....

damn - sorry.. it was 2.0.7.

> Alternatively, give us just a little context of the line it barfed on
> so we can locate it:
> 
>   head -225 smbd/fileio.c | tail -10

216    if (lock_share_entry(fsp->conn, dev, inode, &token) == False) {
217      DEBUG(0,("write_file: failed to lock share mode entry for file %s.\n", fsp->fsp_name ));
218    }
219
220    num_share_modes = get_share_modes(fsp->conn, token, dev, inode,
&share_list);
221
222    for(i = 0; i < num_share_modes; i++) {
223      share_mode_entry *share_entry = &share_list[i];
224
225      /*

> Your compiler reports itself to be exactly 8 years old.  (Frankly, I
> find it hard to believe that such a compiler would generate sun4u code
> at all.)  When did you last evaluate gcc?  How much worse did it do at
> the time?  Just curious.

Not long ago, gcc 2.8.1, 2.95.2.

The compiler doesn't do anything with architecture specific code - that's
ld and the runtime linker's job. :)

It's old, but it's fast, generates damn good code and it works. I'm
in the process of negotiating a license for spro 6.0, but I don't see
a reason it shouldn't work with 2.0, as I believe it is ANSI compliant.

I did end up building it with gcc because I needed to test the
functionality, but I'm sure it will be faster with spro :)

g.




More information about the samba mailing list