<div dir="ltr">So it is. That's what I get for pushing this right before I left for work... patch sent to list.<br></div><div class="gmail_extra"><br><div class="gmail_quote">On Tue, Feb 23, 2016 at 1:19 PM, Jan Vesely <span dir="ltr"><<a href="mailto:jan.vesely@rutgers.edu" target="_blank">jan.vesely@rutgers.edu</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div class="HOEnZb"><div class="h5">On Mon, 2016-02-22 at 23:25 -0500, Jan Vesely wrote:<br>
> On Mon, 2016-02-22 at 20:36 -0600, Aaron Watry wrote:<br>
> > The scalar float/double function bodies are a direct copy/paste<br>
> > with usage of the CLC wrappers to vectorize them.<br>
> ><br>
> > This commit also adds in the FP_ILOGB0 and FP_ILOGBNAN macros which<br>
> > are<br>
> > equal to the results of ilogb(0.0f) and ilogb(float nan)<br>
> > respectively.<br>
> ><br>
> > v2: Add FP_ILOGB0 and FP_ILOGBNAN definitions<br>
> ><br>
> > Signed-off-by: Aaron Watry <<a href="mailto:awatry@gmail.com">awatry@gmail.com</a>><br>
> > CC: Jan Vesely <<a href="mailto:jan.vesely@rutgers.edu">jan.vesely@rutgers.edu</a>><br>
> > v1 Reviewed-by: Tom Stellard <<a href="mailto:thomas.stellard@amd.com">thomas.stellard@amd.com</a>><br>
> > ---<br>
> > Hope this looks ok.<br>
> ><br>
> > I've tested the macro definitions with some piglit tests locally<br>
> > which<br>
> > I'll send to the piglit list in a minute.<br>
><br>
> Reviewed-by: Jan Vesely <<a href="mailto:jan.vesely@rutgers.edu">jan.vesely@rutgers.edu</a>><br>
><br>
> Thanks.<br>
> Jan<br>
><br>
> ><br>
> > generic/include/clc/clc.h | 1 +<br>
> > generic/include/clc/float/definitions.h | 3 ++<br>
> > generic/include/clc/math/ilogb.h | 5 +++<br>
> > generic/include/clc/math/ilogb.inc | 1 +<br>
> > generic/lib/SOURCES | 1 +<br>
> > generic/lib/math/<a href="http://ilogb.cl" rel="noreferrer" target="_blank">ilogb.cl</a> | 57<br>
> > +++++++++++++++++++++++++++++++++<br>
> > 6 files changed, 68 insertions(+)<br>
> > create mode 100644 generic/include/clc/math/ilogb.h<br>
> > create mode 100644 generic/include/clc/math/ilogb.inc<br>
> > create mode 100644 generic/lib/math/<a href="http://ilogb.cl" rel="noreferrer" target="_blank">ilogb.cl</a><br>
> ><br>
> > diff --git a/generic/include/clc/clc.h b/generic/include/clc/clc.h<br>
> > index 4060ea1..b106923 100644<br>
> > --- a/generic/include/clc/clc.h<br>
> > +++ b/generic/include/clc/clc.h<br>
> > @@ -62,6 +62,7 @@<br>
> > #include <clc/math/half_rsqrt.h><br>
> > #include <clc/math/half_sqrt.h><br>
> > #include <clc/math/hypot.h><br>
> > +#include <clc/math/ilogb.h><br>
> > #include <clc/math/ldexp.h><br>
> > #include <clc/math/log.h><br>
> > #include <clc/math/log10.h><br>
> > diff --git a/generic/include/clc/float/definitions.h<br>
> > b/generic/include/clc/float/definitions.h<br>
> > index 329b623..6010ed2 100644<br>
> > --- a/generic/include/clc/float/definitions.h<br>
> > +++ b/generic/include/clc/float/definitions.h<br>
> > @@ -14,6 +14,9 @@<br>
> > #define FLT_MIN 0x1.0p-126f<br>
> > #define FLT_EPSILON 0x1.0p-23f<br>
> > <br>
> > +#define FP_ILOGB0 (-2147483647 - 1)<br>
> > +#define FP_ILOGBNAN (-2147483647 - 1)<br>
> > +<br>
> > #define M_E_F 0x1.5bf0a8p+1f<br>
> > #define M_LOG2E_F 0x1.715476p+0f<br>
> > #define M_LOG10E_F 0x1.bcb7b2p-2f<br>
> > diff --git a/generic/include/clc/math/ilogb.h<br>
> > b/generic/include/clc/math/ilogb.h<br>
> > new file mode 100644<br>
> > index 0000000..2bb9e9c<br>
> > --- /dev/null<br>
> > +++ b/generic/include/clc/math/ilogb.h<br>
> > @@ -0,0 +1,5 @@<br>
> > +#define __CLC_BODY <clc/math/ilogb.inc><br>
> > +<br>
> > +#include <clc/math/gentype.inc><br>
> > +<br>
> > +#undef __CLC_BODY<br>
> > diff --git a/generic/include/clc/math/ilogb.inc<br>
> > b/generic/include/clc/math/ilogb.inc<br>
> > new file mode 100644<br>
> > index 0000000..7f99fb4<br>
> > --- /dev/null<br>
> > +++ b/generic/include/clc/math/ilogb.inc<br>
> > @@ -0,0 +1 @@<br>
> > +_CLC_OVERLOAD _CLC_DECL __CLC_INTN ilogb(__CLC_GENTYPE x);<br>
> > diff --git a/generic/lib/SOURCES b/generic/lib/SOURCES<br>
> > index c3a5a8a..facb58b 100644<br>
> > --- a/generic/lib/SOURCES<br>
> > +++ b/generic/lib/SOURCES<br>
> > @@ -90,6 +90,7 @@ math/<a href="http://frexp.cl" rel="noreferrer" target="_blank">frexp.cl</a><br>
> > math/<a href="http://half_rsqrt.cl" rel="noreferrer" target="_blank">half_rsqrt.cl</a><br>
> > math/<a href="http://half_sqrt.cl" rel="noreferrer" target="_blank">half_sqrt.cl</a><br>
> > math/<a href="http://hypot.cl" rel="noreferrer" target="_blank">hypot.cl</a><br>
> > +math/<a href="http://ilogb.cl" rel="noreferrer" target="_blank">ilogb.cl</a><br>
> > math/<a href="http://clc_ldexp.cl" rel="noreferrer" target="_blank">clc_ldexp.cl</a><br>
> > math/<a href="http://ldexp.cl" rel="noreferrer" target="_blank">ldexp.cl</a><br>
> > math/<a href="http://log.cl" rel="noreferrer" target="_blank">log.cl</a><br>
> > diff --git a/generic/lib/math/<a href="http://ilogb.cl" rel="noreferrer" target="_blank">ilogb.cl</a> b/generic/lib/math/<a href="http://ilogb.cl" rel="noreferrer" target="_blank">ilogb.cl</a><br>
> > new file mode 100644<br>
> > index 0000000..b783b7e<br>
> > --- /dev/null<br>
> > +++ b/generic/lib/math/<a href="http://ilogb.cl" rel="noreferrer" target="_blank">ilogb.cl</a><br>
> > @@ -0,0 +1,57 @@<br>
> > +/*<br>
> > + * Copyright (c) 2015 Advanced Micro Devices, Inc.<br>
> > + * Copyright (c) 2016 Aaron Watry<br>
> > + *<br>
> > + * Permission is hereby granted, free of charge, to any person<br>
> > obtaining a copy<br>
> > + * of this software and associated documentation files (the<br>
> > "Software"), to deal<br>
> > + * in the Software without restriction, including without<br>
> > limitation<br>
> > the rights<br>
> > + * to use, copy, modify, merge, publish, distribute, sublicense,<br>
> > and/or sell<br>
> > + * copies of the Software, and to permit persons to whom the<br>
> > Software is<br>
> > + * furnished to do so, subject to the following conditions:<br>
> > + *<br>
> > + * The above copyright notice and this permission notice shall be<br>
> > included in<br>
> > + * all copies or substantial portions of the Software.<br>
> > + *<br>
> > + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,<br>
> > EXPRESS OR<br>
> > + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF<br>
> > MERCHANTABILITY,<br>
> > + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO<br>
> > EVENT<br>
> > SHALL THE<br>
> > + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES<br>
> > OR<br>
> > OTHER<br>
> > + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,<br>
> > ARISING FROM,<br>
> > + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER<br>
> > DEALINGS IN<br>
> > + * THE SOFTWARE.<br>
> > + */<br>
> > +<br>
> > +#include <clc/clc.h><br>
> > +#include "../clcmacro.h"<br>
> > +#include "math.h"<br>
> > +<br>
> > +_CLC_OVERLOAD _CLC_DEF int ilogb(float x) {<br>
> > + uint ux = as_uint(x);<br>
> > + uint ax = ux & EXSIGNBIT_SP32;<br>
> > + int rs = -118 - (int) clz(ux & MANTBITS_SP32);<br>
> > + int r = (int) (ax >> EXPSHIFTBITS_SP32) - EXPBIAS_SP32;<br>
> > + r = ax < 0x00800000U ? rs : r;<br>
> > + r = ax > EXPBITS_SP32 | ax == 0 ? 0x80000000 : r;<br>
> > + r = ax == EXPBITS_SP32 ? 0x7fffffff : r;<br>
> > + return r;<br>
> > +}<br>
> > +<br>
> > +_CLC_UNARY_VECTORIZE(_CLC_OVERLOAD _CLC_DEF, int, ilogb, float);<br>
> > +<br>
> > +#ifdef cl_khr_fp64<br>
> > +#pragma OPENCL EXTENSION cl_khr_fp64 : enable<br>
> > +<br>
> > +_CLC_OVERLOAD _CLC_DEF ilogb(double x) {<br>
<br>
</div></div>looks like I was too fast in reviewing. Return type specifier is<br>
missing here.<br>
<br>
Jan<br>
<div class="HOEnZb"><div class="h5"><br>
> > + ulong ux = as_ulong(x);<br>
> > + ulong ax = ux & ~SIGNBIT_DP64;<br>
> > + int r = (int) (ax >> EXPSHIFTBITS_DP64) - EXPBIAS_DP64;<br>
> > + int rs = -1011 - (int) clz(ax & MANTBITS_DP64);<br>
> > + r = ax < 0x0010000000000000UL ? rs : r;<br>
> > + r = ax > 0x7ff0000000000000UL | ax == 0UL ? 0x80000000 : r;<br>
> > + r = ax == 0x7ff0000000000000UL ? 0x7fffffff : r;<br>
> > + return r;<br>
> > +}<br>
> > +<br>
> > +_CLC_UNARY_VECTORIZE(_CLC_OVERLOAD _CLC_DEF, int, ilogb, double);<br>
> > +<br>
> > +#endif // cl_khr_fp64<br>
</div></div><span class="HOEnZb"><font color="#888888">--<br>
Jan Vesely <<a href="mailto:jan.vesely@rutgers.edu">jan.vesely@rutgers.edu</a>><br>
</font></span></blockquote></div><br></div>