Bug#659153: [Pkg-haskell-maintainers] Bug#659153: haskell-bindings-gpgme: FTBFS: Missing (or bad) header file: gpgme.h

2012-02-08 Thread Clint Adams
On Wed, Feb 08, 2012 at 01:50:22PM -0500, Aaron M. Ucko wrote: Builds of haskell-bindings-gpgme in minimal environments (such as on the autobuilders) have been failng: library.setup: Missing dependency on a foreign library: * Missing (or bad) header file: gpgme.h Please add a build

Bug#659153: [Pkg-haskell-maintainers] Bug#659153: haskell-bindings-gpgme: FTBFS: Missing (or bad) header file: gpgme.h

2012-02-08 Thread Aaron M. Ucko
Clint Adams cl...@debian.org writes: So is /usr/include/gpgme.h there on i386? I'd meant to check that; sorry for jumping to conclusions. gpgme.h is present, so the problem is evidently more subtle; I'll see if I can reproduce it with configure -v3, per the full message's suggestion. (I'm not

Bug#659153: [Pkg-haskell-maintainers] Bug#659153: haskell-bindings-gpgme: FTBFS: Missing (or bad) header file: gpgme.h

2012-02-08 Thread Aaron M. Ucko
u...@debian.org (Aaron M. Ucko) writes: gpgme.h is present, so the problem is evidently more subtle; I'll see if I can reproduce it with configure -v3, per the full message's suggestion. (I'm not involved with any autobuilders myself; I just investigate cases of packages showing up on amd64

Bug#659153: [Pkg-haskell-maintainers] Bug#659153: haskell-bindings-gpgme: FTBFS: Missing (or bad) header file: gpgme.h

2012-02-08 Thread Clint Adams
On Wed, Feb 08, 2012 at 03:48:40PM -0500, Aaron M. Ucko wrote: Ah, it looks like gpgme.h insists on seeing -D_FILE_OFFSET_BITS=64 for consistency, but only on architectures that default to 32-bit offsets; builds on amd64, ia64, kfreebsd-*, and s390x should all leave _FILE_OFFSET_BITS unset.

Bug#659153: [Pkg-haskell-maintainers] Bug#659153: haskell-bindings-gpgme: FTBFS: Missing (or bad) header file: gpgme.h

2012-02-08 Thread Aaron M. Ucko
Clint Adams cl...@debian.org writes: Great, thanks for diagnosing this. You're welcome. Meanwhile, a second look at the header indicates that it should be safe to supply _FILE_OFFSET_BITS unconditionally after all. -- Aaron M. Ucko, KB1CJC (amu at alum.mit.edu, ucko at debian.org)