This is the mail archive of the gcc-help@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: Mutually Exclusive switches


On 11/11/2010 6:00 AM, Brian Budge wrote:
On the one hand, it's nice for gcc to do what you tell it:  tune for
athlon-xp, and allow sse2 instructions.  On the other hand, it would
be nice to get a big bold warning that your arch does not support one
of the other command-line flags.  That's my opinion at least.  I would
guess that this is not a bug, unless gcc claims to already support
diagnoses of command-line argument flag conflicts.

Brian

On Thu, Nov 11, 2010 at 4:57 AM, tom<tompotts@madtom1999.bbmax.co.uk> wrote:
I've just found out the problem with a project is due to flag  -msse2 being
used along with -march=athlon-xp (32bit) which doesn't support sse2!
Does this count as a bug or crass stupidity or both?
Tom

gnu policy always has been to keep down the default noise level. With -v switch set, one might like to get comments about which switches over-ride earlier ones. Efforts to clean up gcc command line switching seem to be far between, some of the reasons being evident. I heard once of a project to get them streamlined for supporting my employer's products, but it may have faded away.
Personally, I get confused by the use of many redundant and conflicting switches. From reading and sometimes answering questions. my impression is most people fall in one of 2 camps. Either they prefer a sub-optimal set of switches, or they go overboard.


--
Tim Prince


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]