[Mesa-stable] [Mesa-dev] [PATCH] gm107/ir: fix sign bit emission for FADD32I

Ilia Mirkin imirkin at alum.mit.edu
Mon Jul 4 11:59:57 UTC 2016


That flips the sign of the immediate. Why not flip 0x35, which is an
explicit neg modifier? I guess we mess with the immediate in the other
emitters, so r-b either way.

As an aside, how did you hit this? Should have gotten folded in...

On Jul 4, 2016 7:12 AM, "Samuel Pitoiset" <samuel.pitoiset at gmail.com> wrote:

> When emitting OP_SUB, the sign bit for FADD and FADD32I is not
> at the same position. It's at position 45 for FADD but 51 for FADD32I.
>
> This fixes the following piglit test:
> tests/spec/arb_fragment_program/fdo30337b.shader_test
>
> Signed-off-by: Samuel Pitoiset <samuel.pitoiset at gmail.com>
> Cc: <mesa-stable at lists.freedesktop.org>
> ---
>  src/gallium/drivers/nouveau/codegen/nv50_ir_emit_gm107.cpp | 9 ++++++---
>  1 file changed, 6 insertions(+), 3 deletions(-)
>
> diff --git a/src/gallium/drivers/nouveau/codegen/nv50_ir_emit_gm107.cpp
> b/src/gallium/drivers/nouveau/codegen/nv50_ir_emit_gm107.cpp
> index 2c5e8f6..f1ba27a 100644
> --- a/src/gallium/drivers/nouveau/codegen/nv50_ir_emit_gm107.cpp
> +++ b/src/gallium/drivers/nouveau/codegen/nv50_ir_emit_gm107.cpp
> @@ -1234,6 +1234,9 @@ CodeEmitterGM107::emitFADD()
>        emitABS(0x2e, insn->src(0));
>        emitNEG(0x2d, insn->src(1));
>        emitFMZ(0x2c, 1);
> +
> +      if (insn->op == OP_SUB)
> +         code[1] ^= 0x00002000;
>     } else {
>        emitInsn(0x08000000);
>        emitABS(0x39, insn->src(1));
> @@ -1243,10 +1246,10 @@ CodeEmitterGM107::emitFADD()
>        emitNEG(0x35, insn->src(1));
>        emitCC  (0x34);
>        emitIMMD(0x14, 32, insn->src(1));
> -   }
>
> -   if (insn->op == OP_SUB)
> -      code[1] ^= 0x00002000;
> +      if (insn->op == OP_SUB)
> +         code[1] ^= 0x00080000;
> +   }
>
>     emitGPR(0x08, insn->src(0));
>     emitGPR(0x00, insn->def(0));
> --
> 2.8.0
>
> _______________________________________________
> mesa-dev mailing list
> mesa-dev at lists.freedesktop.org
> https://lists.freedesktop.org/mailman/listinfo/mesa-dev
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.freedesktop.org/archives/mesa-stable/attachments/20160704/33be3222/attachment.html>


More information about the mesa-stable mailing list