Re: [Beignet] [PATCH 1/2] Backend: Add optimization for negtive modifier

On Wed, May 17, 2017 at 1:20 AM, rander.wang wrote: > LLVM transform Mad(a, -b, c) to > Add b, -b, 0 > Mad val, a, b, c I think you mean that LLVM translates > Add b, -b, 0 > Mad val, a, b, c to > Mad(a, -b, c) As it is

Re: [Beignet] [PATCH] utests: added for optimization negtiveAdd

In the patch title and the code: negtive -> negative ___ Beignet mailing list Beignet@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/beignet

Re: [Beignet] [PATCH V2] GBE: try to avoid bank conflict in register allocator.

On Wed, Apr 27, 2016 at 12:43 AM, Ruiling Song wrote: > v2: > fix build error. Some documentation or description about this would be very welcome. I cannot honestly imagine anyone being able to review this without it, and I think I know what this patch is doing. :) Some

Re: [Beignet] [PATCH V2 2/2] utests: add an utest for mix

On Wed, Oct 21, 2015 at 8:21 PM, Pan Xiuli wrote: > Add a testcase for compiler mix. Since mix will have > error, we take err limit as 1e-3 and print the max err. I don't know what OpenCL's spec says about this issue, but you should be aware that Gen's LRP instruction does

Re: [Beignet] [PATCH 3/8] Backend: Add gen8 instruction field for special accumulator.

On Tue, Sep 15, 2015 at 4:15 AM, wrote: > From: Junyan He > > The madm and invm function need to set accumulator id in the > instruction. On BDW, the write mask of the dst and channel > mask of src are reinterpreted for acc2~acc9 selection. > >

Re: [Beignet] [PATCH 5/8] Backend: Add the MADM function to gen8 encoder.

On Tue, Sep 15, 2015 at 4:15 AM, wrote: > From: Junyan He > > Signed-off-by: Junyan He > --- > backend/src/backend/gen8_encoder.cpp | 56 > > backend/src/backend/gen8_encoder.hpp |

Re: [Beignet] [PATCH 6/8] Backend: Implement FDIV64 on BDW.

On Tue, Sep 15, 2015 at 4:15 AM, wrote: > From: Junyan He > > According to the document, we use a set of instructions > to implement double type division. > > Signed-off-by: Junyan He > --- >

Re: [Beignet] [PATCH] libocl: fix degrees function precision issue.

On Thu, Aug 6, 2015 at 12:57 AM, xionghu@intel.com wrote: From: Luo Xionghu xionghu@intel.com should define and use M_180_PI_F directly instead of using 180/M_PI_F. Signed-off-by: Luo Xionghu xionghu@intel.com --- backend/src/libocl/include/ocl_float.h | 1 +

Re: [Beignet] [PATCH 1/8] Backend: Add half float as a new type.

On Thu, May 21, 2015 at 1:25 AM, junyan...@inbox.com wrote: From: Junyan He junyan...@linux.intel.com Because the CPU of X86 does not support half float instructions, there is no support for half float operations. So we introduce the half class to handle the operations for half float using

Re: [Beignet] [PATCH 6/8] Backend: Add half float ASM output support.

On Thu, May 21, 2015 at 1:25 AM, junyan...@inbox.com wrote: From: Junyan He junyan...@linux.intel.com Signed-off-by: Junyan He junyan...@linux.intel.com --- backend/src/backend/gen/gen_mesa_disasm.c | 83 +-- 1 file changed, 78 insertions(+), 5 deletions(-)

Re: [Beignet] [PATCH 6/7] replace mad with llvm intrinsic.

On Mon, Mar 9, 2015 at 10:59 PM, xionghu@intel.com wrote: From: Luo Xionghu xionghu@intel.com translate native mad to llvm.fma. I'm not sure that it matters for this patch, but do we know if Gen's MAD instruction is a fused-multiply-add? That is, does it not do an intermediate

Re: [Beignet] double precision support

On Tue, Mar 10, 2015 at 2:19 AM, Zhigang Gong zhigang.g...@linux.intel.com wrote: 2. The double support is not fully supported. For example, all the math functions and even the divide instruction is not supported. You're right that the hardware doesn't natively do most of the math operations

Re: [Beignet] [PATCH 6/7] replace mad with llvm intrinsic.

On Tue, Mar 10, 2015 at 6:55 PM, Song, Ruiling ruiling.s...@intel.com wrote: I'm not sure that it matters for this patch, but do we know if Gen's MAD instruction is a fused-multiply-add? That is, does it not do an intermediate rounding step after the multiply? I also have such kind of concern,

Re: [Beignet] [PATCH] libocl: refine implementation of sign().

On Wed, Jan 28, 2015 at 11:18 PM, Ruiling Song ruiling.s...@intel.com wrote: Avoid if-branching. Signed-off-by: Ruiling Song ruiling.s...@intel.com --- backend/src/libocl/tmpl/ocl_common.tmpl.cl | 16 +--- 1 file changed, 9 insertions(+), 7 deletions(-) diff --git