| Message ID | 20170522141722.50502-1-jarod@redhat.com (mailing list archive) |
|---|---|
| State | Accepted |
| Headers | show |
On 5/22/2017 10:17 AM, Jarod Wilson wrote: > Only ntohll was being checked to see if it wasn't defined, and was then > redefining htonll as well as ntohll. This was causing some problems for > the compile of the opa-ff package. Simple enough to rearrange this code a > bit such that htonll and ntohll are handled entirely independent of one > another. It used to be the case that they were treated independently. > Reported-by: Honggang Li <honli@redhat.com> > Signed-off-by: Jarod Wilson <jarod@redhat.com> Reviewed-by: Hal Rosenstock <hal@mellanox.com> -- To unsubscribe from this list: send the line "unsubscribe linux-rdma" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Mon, May 22, 2017 at 10:17:22AM -0400, Jarod Wilson wrote: > Only ntohll was being checked to see if it wasn't defined, and was then > redefining htonll as well as ntohll. This was causing some problems for > the compile of the opa-ff package. Simple enough to rearrange this code a > bit such that htonll and ntohll are handled entirely independent of one > another. What problem did this cause? > +/* Users should use the glibc functions directly, not these wrappers */ > #ifndef ntohll > -#undef htonll > #undef ntohll > -/* Users should use the glibc functions directly, not these wrappers */ > -static inline __attribute__((deprecated)) uint64_t htonll(uint64_t x) { return htobe64(x); } > static inline __attribute__((deprecated)) uint64_t ntohll(uint64_t x) { return be64toh(x); } > -#define htonll htonll > #define ntohll ntohll > #endif > +#ifndef htonll > +#undef htonll Both these undefs are not needed anymore. Seems fine to me, if not weird that something would have a problem here. Jason -- To unsubscribe from this list: send the line "unsubscribe linux-rdma" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On 2017-05-24 12:33 PM, Jason Gunthorpe wrote: > On Mon, May 22, 2017 at 10:17:22AM -0400, Jarod Wilson wrote: >> Only ntohll was being checked to see if it wasn't defined, and was then >> redefining htonll as well as ntohll. This was causing some problems for >> the compile of the opa-ff package. Simple enough to rearrange this code a >> bit such that htonll and ntohll are handled entirely independent of one >> another. > > What problem did this cause? > >> +/* Users should use the glibc functions directly, not these wrappers */ >> #ifndef ntohll >> -#undef htonll >> #undef ntohll >> -/* Users should use the glibc functions directly, not these wrappers */ >> -static inline __attribute__((deprecated)) uint64_t htonll(uint64_t x) { return htobe64(x); } >> static inline __attribute__((deprecated)) uint64_t ntohll(uint64_t x) { return be64toh(x); } >> -#define htonll htonll >> #define ntohll ntohll >> #endif >> +#ifndef htonll >> +#undef htonll > > Both these undefs are not needed anymore. > > Seems fine to me, if not weird that something would have a problem > here. My understanding from Honggang is that this was necessary for him to get opa-ff and infiniband-diags to compile (along with an update to our libibmad package). He should be able to fill in the blanks with further details.
On Mon, May 22, 2017 at 10:17:22AM -0400, Jarod Wilson wrote: > Only ntohll was being checked to see if it wasn't defined, and was then > redefining htonll as well as ntohll. This was causing some problems for > the compile of the opa-ff package. Simple enough to rearrange this code a > bit such that htonll and ntohll are handled entirely independent of one > another. > > Reported-by: Honggang Li <honli@redhat.com> > Signed-off-by: Jarod Wilson <jarod@redhat.com> > --- > libibumad/umad.h | 10 ++++++---- > 1 file changed, 6 insertions(+), 4 deletions(-) > Thanks, applied.
diff --git a/libibumad/umad.h b/libibumad/umad.h index 81811380..479165a8 100644 --- a/libibumad/umad.h +++ b/libibumad/umad.h @@ -247,15 +247,17 @@ static inline void umad_free(void *umad) free(umad); } +/* Users should use the glibc functions directly, not these wrappers */ #ifndef ntohll -#undef htonll #undef ntohll -/* Users should use the glibc functions directly, not these wrappers */ -static inline __attribute__((deprecated)) uint64_t htonll(uint64_t x) { return htobe64(x); } static inline __attribute__((deprecated)) uint64_t ntohll(uint64_t x) { return be64toh(x); } -#define htonll htonll #define ntohll ntohll #endif +#ifndef htonll +#undef htonll +static inline __attribute__((deprecated)) uint64_t htonll(uint64_t x) { return htobe64(x); } +#define htonll htonll +#endif END_C_DECLS #endif /* _UMAD_H */
Only ntohll was being checked to see if it wasn't defined, and was then redefining htonll as well as ntohll. This was causing some problems for the compile of the opa-ff package. Simple enough to rearrange this code a bit such that htonll and ntohll are handled entirely independent of one another. Reported-by: Honggang Li <honli@redhat.com> Signed-off-by: Jarod Wilson <jarod@redhat.com> --- libibumad/umad.h | 10 ++++++---- 1 file changed, 6 insertions(+), 4 deletions(-)