As GVT-g workload scheduler may submit workloads from different VMs to i915, some VM may use 32bit PPGTT addressing mode, while some VMs may use 48bit addressing mode, the context addressing mode bit in the context descriptor has to be aligned with guest workload. If the to-be-submitted context is a GVT context. Signed-off-by: Zhi Wang <zhi.a.wang@xxxxxxxxx> --- drivers/gpu/drm/i915/intel_lrc.c | 10 +++++++++- 1 file changed, 9 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c index da97bc5..48e8ca2 100644 --- a/drivers/gpu/drm/i915/intel_lrc.c +++ b/drivers/gpu/drm/i915/intel_lrc.c @@ -302,7 +302,8 @@ logical_ring_init_platform_invariants(struct intel_engine_cs *ring) * which remains valid until the context is unpinned. * * This is what a descriptor looks like, from LSB to MSB: - * bits 0-11: flags, GEN8_CTX_* (cached in ctx_desc_template) + * bits 0-2,3-11: flags, GEN8_CTX_* (cached in ctx_desc_template) + * bits 3: Context addressing mode * bits 12-31: LRCA, GTT address of (the HWSP of) this context * bits 32-51: ctx ID, a globally unique tag (the LRCA again!) * bits 52-63: reserved, may encode the engine ID (for GuC) @@ -317,6 +318,13 @@ intel_lr_context_descriptor_update(struct intel_context *ctx, LRC_PPHWSP_PN * PAGE_SIZE; desc = ring->ctx_desc_template; /* bits 0-11 */ + + if (!ctx->gvt_context) + desc |= GEN8_CTX_ADDRESSING_MODE(dev) << /* bit 3 */ + GEN8_CTX_ADDRESSING_MODE_SHIFT; + else + desc |= ctx->gvt_context_addressing_mode[ring->id]; + desc |= lrca; /* bits 12-31 */ desc |= (lrca >> PAGE_SHIFT) << GEN8_CTX_ID_SHIFT; /* bits 32-51 */ -- 1.9.1 _______________________________________________ Intel-gfx mailing list Intel-gfx@xxxxxxxxxxxxxxxxxxxxx http://lists.freedesktop.org/mailman/listinfo/intel-gfx
