On Wed, Feb 08, 2017 at 05:15:49PM +0200, Matan Barak (External) wrote: > On 08/02/2017 17:04, Leon Romanovsky wrote: > > Remove references to private kernel header and defines from exported > > ib_user_verb.h file. > > > > The code snippet below is used to reproduce the issue: > > > > #include > > #include > > > > int main(void) > > { > > printf("IB_USER_VERBS_ABI_VERSION = %d\n", IB_USER_VERBS_ABI_VERSION); > > return 0; > > } > > > > It fails during compilation phase with an error: > > ➜ /tmp gcc main.c > > main.c:2:31: fatal error: rdma/ib_user_verb.h: No such file or directory > > #include > > ^ > > compilation terminated. > > > > Fixes: 189aba99e700 ("IB/uverbs: Extend modify_qp and support packet pacing") > > CC: Bodong Wang > > CC: Matan Barak > > CC: Christoph Hellwig > > Tested-by: Slava Shwartsman > > Signed-off-by: Leon Romanovsky > > --- > > Changelog v0 -> v1 > > * Document the origin of hard-coded values. > > --- > > include/uapi/rdma/ib_user_verbs.h | 11 ++++++++--- > > 1 file changed, 8 insertions(+), 3 deletions(-) > > > > diff --git a/include/uapi/rdma/ib_user_verbs.h b/include/uapi/rdma/ib_user_verbs.h > > index dfdfe4e92d31..f4f87cff6dc6 100644 > > --- a/include/uapi/rdma/ib_user_verbs.h > > +++ b/include/uapi/rdma/ib_user_verbs.h > > @@ -37,7 +37,6 @@ > > #define IB_USER_VERBS_H > > > > #include > > -#include > > > > /* > > * Increment this value if any changes that break userspace ABI > > @@ -548,11 +547,17 @@ enum { > > }; > > > > enum { > > - IB_USER_LEGACY_LAST_QP_ATTR_MASK = IB_QP_DEST_QPN > > + /* > > + * This value is equal to IB_QP_DEST_QPN. > > + */ > > + IB_USER_LEGACY_LAST_QP_ATTR_MASK = 1ULL << 20, > > }; > > > > enum { > > - IB_USER_LAST_QP_ATTR_MASK = IB_QP_RATE_LIMIT > > + /* > > + * This value is equal to IB_QP_RATE_LIMIT. > > + */ > > + IB_USER_LAST_QP_ATTR_MASK = 1ULL << 25, > > }; > > > > struct ib_uverbs_ex_create_qp { > > -- > > 2.11.1 > > > > I tried to write this earlier, but for some reason my mailer didn't really > send it. > IMHO, since these enum values are actually passed between user-space and > kernel (attr_mask), it's acceptable to expose all enum values. > Otherwise, user-space should define all these symbols by by itself (as done > today), so why bother introduce only this explicit symbol? Mainly because it is -rc7 fix and it is MUST be applied as soon as possible, currently IB stack is broken. > -- > To unsubscribe from this list: send the line "unsubscribe linux-rdma" in > the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org > More majordomo info at http://vger.kernel.org/majordomo-info.html