Re: [Mesa-dev] [PATCH] nvc0/ir: remove wrong assert in emitUADD()

2016-07-23 Thread Ilia Mirkin
On Jul 23, 2016 5:45 AM, "Samuel Pitoiset" wrote: > > > > On 07/23/2016 02:53 AM, Ilia Mirkin wrote: >> >> No, it's not. But if the real op is OP_SUB, then the logic is a bit >> different, since there's an implied extra neg on src(1). > > > Right, it's legal but for OP_SUB only. Conversely, it's

Re: [Mesa-dev] [PATCH] nvc0/ir: remove wrong assert in emitUADD()

2016-07-23 Thread Samuel Pitoiset
On 07/23/2016 02:53 AM, Ilia Mirkin wrote: No, it's not. But if the real op is OP_SUB, then the logic is a bit different, since there's an implied extra neg on src(1). Right, it's legal but for OP_SUB only. On Fri, Jul 22, 2016 at 7:47 PM, Samuel Pitoiset wrote: It's actually legal to h

Re: [Mesa-dev] [PATCH] nvc0/ir: remove wrong assert in emitUADD()

2016-07-22 Thread Ilia Mirkin
No, it's not. But if the real op is OP_SUB, then the logic is a bit different, since there's an implied extra neg on src(1). On Fri, Jul 22, 2016 at 7:47 PM, Samuel Pitoiset wrote: > It's actually legal to have neg modifiers on both sources. > > Signed-off-by: Samuel Pitoiset > --- > src/galli

[Mesa-dev] [PATCH] nvc0/ir: remove wrong assert in emitUADD()

2016-07-22 Thread Samuel Pitoiset
It's actually legal to have neg modifiers on both sources. Signed-off-by: Samuel Pitoiset --- src/gallium/drivers/nouveau/codegen/nv50_ir_emit_nvc0.cpp | 1 - 1 file changed, 1 deletion(-) diff --git a/src/gallium/drivers/nouveau/codegen/nv50_ir_emit_nvc0.cpp b/src/gallium/drivers/nouveau/code