On 08/26/2009 10:50 PM, Pete Zaitcev wrote:
On Wed, 26 Aug 2009 22:26:50 -0400, Jeff Garzik<jeff@xxxxxxxxxx> wrote:
On 08/26/2009 10:06 PM, Pete Zaitcev wrote:
Signed-off-by: Pete Zaitcev<zaitcev@xxxxxxxxxx>
applied a modified version, see attached... after that short IRC
discussion, I think an array with a range check would be nicer than
coding it (though the C compiler should make a nice, efficient table
from 'switch' statement for us...)
I don't think it's enough. Look:
+static const char *cld_errlist[] =
+{
+ [CLE_OK] = "Success",
+ [CLE_SESS_EXISTS] = "Session exists",
When you do indexed initializations, it's possible to create
empty array elements (currently there are none because indexes
are contiguous). So, the following needs a NULL check:
+ return cld_errlist[ecode];
+}
True in general, but do you actually see a hole?
It seems to me that our error list will always be sequential, without
holes. Even if we obsolete an error, we still need to keep it around in
the headers and cld_errlist[]. And new error codes will go on the end
of the list...
Jeff
--
To unsubscribe from this list: send the line "unsubscribe hail-devel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
