yaxunl added inline comments. ================ Comment at: lib/Headers/opencl-c.h:17051 @@ +17050,3 @@ +#define CLK_SUCCESS 0 +#define CLK_ENQUEUE_FAILURE -101 +#define CLK_INVALID_QUEUE -102 ---------------- jprice wrote: > yaxunl wrote: > > Anastasia wrote: > > > Are those arbitrary taken values I am guessing? > > How about this? > > > > // The value are defined according to their order in the spec. > > #define CLK_SUCCESS 0 > > #define CLK_ENQUEUE_FAILURE -1 > > #define CLK_INVALID_QUEUE -2 > > #define CLK_INVALID_NDRANGE -3 > > #define CLK_INVALID_EVENT_WAIT_LIST -4 > > #define CLK_DEVICE_QUEUE_FULL -5 > > #define CLK_INVALID_ARG_SIZE -6 > > #define CLK_EVENT_ALLOCATION_FAILURE -8 > > #define CLK_OUT_OF_RESOURCES -9 > > > These values were taken from the table in section 2.10.6.1 in the SPIR 2.0 > provisional spec. Some of the choices seem arbitrary, while others were > chosen to match the value of the equivalent runtime API error. > > I imagine changing these values now might cause incompatibilities with > existing SPIR consumers? Then I guess we'd better keep it as is.
http://reviews.llvm.org/D18369 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits