On Mon, Jan 16, 2006 at 05:52:34PM -0500, John David Anglin wrote: > > > The > > > only reason they would need rebuilt after a glibc bug fix would be if the > > > glibc ABI changed in the process, and that would be Very Bad<tm>. > > > > Yeah, hopefully that doesn't happen. > > As I have previously mentioned, there is a pending ABI issue affecting > the passing of complex and vector floating point numbers on the PA > (floats in particular on the 32-bit port, and both floats and doubles > on the 64-bit port). It turns out we are using the wrong registers > to pass these types.
sorry - I forgot about that. Anyay, it's not a disaster. > I'm considering fixing this in GCC 4.1 (patch under test this weekend). > We already have an ABI change under HP-UX in GCC 4.1 (dwarf2 exception > support). I've also observered that there is an incompatibility in > libgcc under linux between 4.0 and 4.1. I don't know if this is an ABI > change or not. For these reasons, now is the time to fix these issue. > > If anybody feels strongly that the status quo should be maintained > please let me know. I'm going to bump the libgcc_s versions. Debian "testing" (etch) is officially using gcc 4.0, right? Does Debian promote binaries from "unstable" (using gcc 4.1) to "testing"? If this ABI roll goes through, I can help track down debian packages that need to be rebuilt if given an existing sample file to look at. grant -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
