This is the mail archive of the 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: [PATCH] SEE optimization removal

Joseph S. Myers wrote:

>> The patch would be OK, except that we want to continue to accept old
>> optimization options (just ignoring them) to avoid breaking existing
>> Makefiles. 

> Although we've done this for some options (in general, it makes sense for 
> optimization options as they don't affect the semantics of the compiled 
> code), we haven't done it consistently.  

> think).  So I don't know if we actually have a general policy here.

I think we do, and if we don't, we should.

In particular, we should have the policy that optimization options that
do not affect semantics, once present in a release, are never removed.
Backwards compatibility is an important principle, and this case of
backwards compatibility has almost no cost to us.

I agree that if we remove an option with observable semantics (e.g.,
-fvolatile), we should reject the option, ideally with a helpful
message.  There, it's desirable to prevent old Makefiles from working,
so that we're not silently ignoring user intent.


Mark Mitchell
(650) 331-3385 x713

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