30.11.2014, 19:56, Ilia Mirkin kirjoitti: > Indications are that no GF116's actually have a copy engine there, but > actually have the decompression engine. This engine can be made to do > copies, but that should be done separately. > > Unclear why this didn't turn up on all GF116's, but perhaps the > non-mobile ones came with enough VRAM to not trigger ttm migrations in > test scenarios.
FWIW, I experienced this on my non-mobile GF116 when keeping the system running for several days and running graphics-intensive games. I don't have the card at hand right now so don't know how much VRAM it had, though. Thanks for getting this figured out :) > Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=85465 > Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=59168 > Cc: [email protected] > Signed-off-by: Ilia Mirkin <[email protected]> > --- > nvkm/engine/device/nvc0.c | 1 - > 1 file changed, 1 deletion(-) > > diff --git a/nvkm/engine/device/nvc0.c b/nvkm/engine/device/nvc0.c > index cd05677..72a40f9 100644 > --- a/nvkm/engine/device/nvc0.c > +++ b/nvkm/engine/device/nvc0.c > @@ -218,7 +218,6 @@ nvc0_identify(struct nouveau_device *device) > device->oclass[NVDEV_ENGINE_BSP ] = &nvc0_bsp_oclass; > device->oclass[NVDEV_ENGINE_PPP ] = &nvc0_ppp_oclass; > device->oclass[NVDEV_ENGINE_COPY0 ] = &nvc0_copy0_oclass; > - device->oclass[NVDEV_ENGINE_COPY1 ] = &nvc0_copy1_oclass; > device->oclass[NVDEV_ENGINE_DISP ] = nva3_disp_oclass; > device->oclass[NVDEV_ENGINE_PERFMON] = &nvc0_perfmon_oclass; > break; > -- Anssi Hannula _______________________________________________ Nouveau mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/nouveau
