On 06/04/17 12:15, Rodrigo Vivi wrote:
From: Ben Widawsky <benjamin.widawsky@xxxxxxxxx>
The docs are not yet correct, so I cannot provide a reference to it. In the
current docs, the size is actually smaller than SKL. This seems unlikely given
that in another part of the docs there are clearly more engines stored within
the context image.
!UPSTREAM: I got this number from the Windows driver, but we should add the right
reference to the docs when we have it..
v2: (Rodrigo) Fixup the missing break identified by Tvrtko.
Cc: Tvrtko Ursulin <tvrtko.ursulin@xxxxxxxxx>
Signed-off-by: Ben Widawsky <benjamin.widawsky@xxxxxxxxx>
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@xxxxxxxxx>
---
drivers/gpu/drm/i915/intel_lrc.c | 14 ++++++++++++--
1 file changed, 12 insertions(+), 2 deletions(-)
diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index 23e2bed..058d5f2 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -138,6 +138,7 @@
#include "i915_drv.h"
#include "intel_mocs.h"
+#define GEN10_LR_CONTEXT_RENDER_SIZE ((1 + 33) * PAGE_SIZE)
1+33 looks really big compared to what's in the specs. I agree that the
specs are unclear, but something around 1+18 should still be a safe
setting. Unfortunately I don't have a setup to verify, so I'm happy to
merge this as it is as long as we remember to come and re-assess it
(maybe add a TODO?)
Thanks,
Daniele
#define GEN9_LR_CONTEXT_RENDER_SIZE (22 * PAGE_SIZE)
#define GEN8_LR_CONTEXT_RENDER_SIZE (20 * PAGE_SIZE)
#define GEN8_LR_CONTEXT_OTHER_SIZE (2 * PAGE_SIZE)
@@ -1930,10 +1931,19 @@ uint32_t intel_lr_context_size(struct intel_engine_cs *engine)
switch (engine->id) {
case RCS:
- if (INTEL_GEN(engine->i915) >= 9)
+ switch (INTEL_GEN(engine->i915)) {
+ default:
+ DRM_ERROR("Unknown context size for GEN\n");
+ case 10:
+ ret = GEN10_LR_CONTEXT_RENDER_SIZE;
+ break;
+ case 9:
ret = GEN9_LR_CONTEXT_RENDER_SIZE;
- else
+ break;
+ case 8:
ret = GEN8_LR_CONTEXT_RENDER_SIZE;
+ break;
+ }
break;
case VCS:
case BCS:
_______________________________________________
Intel-gfx mailing list
Intel-gfx@xxxxxxxxxxxxxxxxxxxxx
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
