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: Link tests after GCC_NO_EXECUTABLES

Bernd Schmidt wrote:

I've now checked in the libstdc++-v3 patch, which is a reasonably obvious bug fix. I've never had an opinion on whether other targets ought to fail link tests, hence I'm not really interested in pursuing the other part of the patch. I'll leave that to you or Mark, if you wish.

I think the libstdc++-v3 change is indeed pretty obvious.

I still think that reverting Rask's patch is the right thing to do. I think the fundamental assumption that you can use the simulator target as way to probe the target is just wrong for bare-metal systems; what's available in simulation and might what might be available otherwise need not be the same. We're making assumptions about the target system when the only reliable way to do that is to have the person configuring the compiler tell us. I think that we sometimes try too hard to make configury smart; essentially, we're preferring wrong answers to making the user do some work.

But, I also don't have a compelling reason to push forward on that myself at this point.

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]