[Pkg-ofed-devel] Bug#832625: infinipath-psm: FTBFS on i386: %ld vs. size_t

[Aaron M. Ucko]

Source: infinipath-psm
Version: 3.3+19.g67c0807.open-2
Severity: important
Justification: fails to build from source

Thanks for taking care of #807149.  The i386 build now makes some
progress, but still fails:

  In file included from /«BUILDDIR»/infinipath-psm-3.3+19.g67c0807.open/include/ipath_user.h:63:0,
                   from /«BUILDDIR»/infinipath-psm-3.3+19.g67c0807.open/psm_user.h:45,
                   from am_reqrep_shmem.c:40:
  am_reqrep_shmem.c: In function 'psmi_shm_attach':
  /«BUILDDIR»/infinipath-psm-3.3+19.g67c0807.open/include/ipath_udebug.h:79:28: error: format '%ld' expects argument of type 'long int', but argument 6 has type 'size_t {aka unsigned int}' [-Werror=format=]
      fprintf(__ipath_dbgout, "%s%s: " fmt, __ipath_mylabel, __func__, \
                              ^
  /«BUILDDIR»/infinipath-psm-3.3+19.g67c0807.open/include/ipath_udebug.h:95:31: note: in expansion of macro '__IPATH_DBG_WHICH'
   #define _IPATH_PRDBG(fmt,...) __IPATH_DBG_WHICH(__IPATH_PROCDBG,fmt,##__VA_ARGS__)
                                 ^
  am_reqrep_shmem.c:544:2: note: in expansion of macro '_IPATH_PRDBG'
    _IPATH_PRDBG("Mapped and initialized shm object control page at %p,"
    ^

Although size_t is de facto equivalent to unsigned long on all Debian
architectures, on i386 (and IIRC 32-bit architectures in general), it
is formally equivalent to unsigned int rather than unsigned long.  To
account for this possibility, you can use the 'z' (siZe_t) width
specifier: %zd or %zu.

Could you please take a look?

Thanks!