linux/drivers/gpu/drm/i915/i915_gem_context.c

949 lines
27 KiB
C
Raw Normal View History

drm/i915: preliminary context support Very basic code for context setup/destruction in the driver. Adds the file i915_gem_context.c This file implements HW context support. On gen5+ a HW context consists of an opaque GPU object which is referenced at times of context saves and restores. With RC6 enabled, the context is also referenced as the GPU enters and exists from RC6 (GPU has it's own internal power context, except on gen5). Though something like a context does exist for the media ring, the code only supports contexts for the render ring. In software, there is a distinction between contexts created by the user, and the default HW context. The default HW context is used by GPU clients that do not request setup of their own hardware context. The default context's state is never restored to help prevent programming errors. This would happen if a client ran and piggy-backed off another clients GPU state. The default context only exists to give the GPU some offset to load as the current to invoke a save of the context we actually care about. In fact, the code could likely be constructed, albeit in a more complicated fashion, to never use the default context, though that limits the driver's ability to swap out, and/or destroy other contexts. All other contexts are created as a request by the GPU client. These contexts store GPU state, and thus allow GPU clients to not re-emit state (and potentially query certain state) at any time. The kernel driver makes certain that the appropriate commands are inserted. There are 4 entry points into the contexts, init, fini, open, close. The names are self-explanatory except that init can be called during reset, and also during pm thaw/resume. As we expect our context to be preserved across these events, we do not reinitialize in this case. As Adam Jackson pointed out, The cutoff of 1MB where a HW context is considered too big is arbitrary. The reason for this is even though context sizes are increasing with every generation, they have yet to eclipse even 32k. If we somehow read back way more than that, it probably means BIOS has done something strange, or we're running on a platform that wasn't designed for this. v2: rename load/unload to init/fini (daniel) remove ILK support for get_size() (indirectly daniel) add HAS_HW_CONTEXTS macro to clarify supported platforms (daniel) added comments (Ben) Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
2012-06-04 21:42:42 +00:00
/*
* Copyright © 2011-2012 Intel Corporation
*
* Permission is hereby granted, free of charge, to any person obtaining a
* copy of this software and associated documentation files (the "Software"),
* to deal in the Software without restriction, including without limitation
* the rights to use, copy, modify, merge, publish, distribute, sublicense,
* and/or sell copies of the Software, and to permit persons to whom the
* Software is furnished to do so, subject to the following conditions:
*
* The above copyright notice and this permission notice (including the next
* paragraph) shall be included in all copies or substantial portions of the
* Software.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
* THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
* FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
* IN THE SOFTWARE.
*
* Authors:
* Ben Widawsky <ben@bwidawsk.net>
*
*/
/*
* This file implements HW context support. On gen5+ a HW context consists of an
* opaque GPU object which is referenced at times of context saves and restores.
* With RC6 enabled, the context is also referenced as the GPU enters and exists
* from RC6 (GPU has it's own internal power context, except on gen5). Though
* something like a context does exist for the media ring, the code only
* supports contexts for the render ring.
*
* In software, there is a distinction between contexts created by the user,
* and the default HW context. The default HW context is used by GPU clients
* that do not request setup of their own hardware context. The default
* context's state is never restored to help prevent programming errors. This
* would happen if a client ran and piggy-backed off another clients GPU state.
* The default context only exists to give the GPU some offset to load as the
* current to invoke a save of the context we actually care about. In fact, the
* code could likely be constructed, albeit in a more complicated fashion, to
* never use the default context, though that limits the driver's ability to
* swap out, and/or destroy other contexts.
*
* All other contexts are created as a request by the GPU client. These contexts
* store GPU state, and thus allow GPU clients to not re-emit state (and
* potentially query certain state) at any time. The kernel driver makes
* certain that the appropriate commands are inserted.
*
* The context life cycle is semi-complicated in that context BOs may live
* longer than the context itself because of the way the hardware, and object
* tracking works. Below is a very crude representation of the state machine
* describing the context life.
* refcount pincount active
* S0: initial state 0 0 0
* S1: context created 1 0 0
* S2: context is currently running 2 1 X
* S3: GPU referenced, but not current 2 0 1
* S4: context is current, but destroyed 1 1 0
* S5: like S3, but destroyed 1 0 1
*
* The most common (but not all) transitions:
* S0->S1: client creates a context
* S1->S2: client submits execbuf with context
* S2->S3: other clients submits execbuf with context
* S3->S1: context object was retired
* S3->S2: clients submits another execbuf
* S2->S4: context destroy called with current context
* S3->S5->S0: destroy path
* S4->S5->S0: destroy path on current context
*
* There are two confusing terms used above:
* The "current context" means the context which is currently running on the
* GPU. The GPU has loaded its state already and has stored away the gtt
drm/i915: preliminary context support Very basic code for context setup/destruction in the driver. Adds the file i915_gem_context.c This file implements HW context support. On gen5+ a HW context consists of an opaque GPU object which is referenced at times of context saves and restores. With RC6 enabled, the context is also referenced as the GPU enters and exists from RC6 (GPU has it's own internal power context, except on gen5). Though something like a context does exist for the media ring, the code only supports contexts for the render ring. In software, there is a distinction between contexts created by the user, and the default HW context. The default HW context is used by GPU clients that do not request setup of their own hardware context. The default context's state is never restored to help prevent programming errors. This would happen if a client ran and piggy-backed off another clients GPU state. The default context only exists to give the GPU some offset to load as the current to invoke a save of the context we actually care about. In fact, the code could likely be constructed, albeit in a more complicated fashion, to never use the default context, though that limits the driver's ability to swap out, and/or destroy other contexts. All other contexts are created as a request by the GPU client. These contexts store GPU state, and thus allow GPU clients to not re-emit state (and potentially query certain state) at any time. The kernel driver makes certain that the appropriate commands are inserted. There are 4 entry points into the contexts, init, fini, open, close. The names are self-explanatory except that init can be called during reset, and also during pm thaw/resume. As we expect our context to be preserved across these events, we do not reinitialize in this case. As Adam Jackson pointed out, The cutoff of 1MB where a HW context is considered too big is arbitrary. The reason for this is even though context sizes are increasing with every generation, they have yet to eclipse even 32k. If we somehow read back way more than that, it probably means BIOS has done something strange, or we're running on a platform that wasn't designed for this. v2: rename load/unload to init/fini (daniel) remove ILK support for get_size() (indirectly daniel) add HAS_HW_CONTEXTS macro to clarify supported platforms (daniel) added comments (Ben) Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
2012-06-04 21:42:42 +00:00
* offset of the BO. The GPU is not actively referencing the data at this
* offset, but it will on the next context switch. The only way to avoid this
* is to do a GPU reset.
*
* An "active context' is one which was previously the "current context" and is
* on the active list waiting for the next context switch to occur. Until this
* happens, the object must remain at the same gtt offset. It is therefore
* possible to destroy a context, but it is still active.
*
*/
#include <drm/drmP.h>
#include <drm/i915_drm.h>
drm/i915: preliminary context support Very basic code for context setup/destruction in the driver. Adds the file i915_gem_context.c This file implements HW context support. On gen5+ a HW context consists of an opaque GPU object which is referenced at times of context saves and restores. With RC6 enabled, the context is also referenced as the GPU enters and exists from RC6 (GPU has it's own internal power context, except on gen5). Though something like a context does exist for the media ring, the code only supports contexts for the render ring. In software, there is a distinction between contexts created by the user, and the default HW context. The default HW context is used by GPU clients that do not request setup of their own hardware context. The default context's state is never restored to help prevent programming errors. This would happen if a client ran and piggy-backed off another clients GPU state. The default context only exists to give the GPU some offset to load as the current to invoke a save of the context we actually care about. In fact, the code could likely be constructed, albeit in a more complicated fashion, to never use the default context, though that limits the driver's ability to swap out, and/or destroy other contexts. All other contexts are created as a request by the GPU client. These contexts store GPU state, and thus allow GPU clients to not re-emit state (and potentially query certain state) at any time. The kernel driver makes certain that the appropriate commands are inserted. There are 4 entry points into the contexts, init, fini, open, close. The names are self-explanatory except that init can be called during reset, and also during pm thaw/resume. As we expect our context to be preserved across these events, we do not reinitialize in this case. As Adam Jackson pointed out, The cutoff of 1MB where a HW context is considered too big is arbitrary. The reason for this is even though context sizes are increasing with every generation, they have yet to eclipse even 32k. If we somehow read back way more than that, it probably means BIOS has done something strange, or we're running on a platform that wasn't designed for this. v2: rename load/unload to init/fini (daniel) remove ILK support for get_size() (indirectly daniel) add HAS_HW_CONTEXTS macro to clarify supported platforms (daniel) added comments (Ben) Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
2012-06-04 21:42:42 +00:00
#include "i915_drv.h"
#include "i915_trace.h"
drm/i915: preliminary context support Very basic code for context setup/destruction in the driver. Adds the file i915_gem_context.c This file implements HW context support. On gen5+ a HW context consists of an opaque GPU object which is referenced at times of context saves and restores. With RC6 enabled, the context is also referenced as the GPU enters and exists from RC6 (GPU has it's own internal power context, except on gen5). Though something like a context does exist for the media ring, the code only supports contexts for the render ring. In software, there is a distinction between contexts created by the user, and the default HW context. The default HW context is used by GPU clients that do not request setup of their own hardware context. The default context's state is never restored to help prevent programming errors. This would happen if a client ran and piggy-backed off another clients GPU state. The default context only exists to give the GPU some offset to load as the current to invoke a save of the context we actually care about. In fact, the code could likely be constructed, albeit in a more complicated fashion, to never use the default context, though that limits the driver's ability to swap out, and/or destroy other contexts. All other contexts are created as a request by the GPU client. These contexts store GPU state, and thus allow GPU clients to not re-emit state (and potentially query certain state) at any time. The kernel driver makes certain that the appropriate commands are inserted. There are 4 entry points into the contexts, init, fini, open, close. The names are self-explanatory except that init can be called during reset, and also during pm thaw/resume. As we expect our context to be preserved across these events, we do not reinitialize in this case. As Adam Jackson pointed out, The cutoff of 1MB where a HW context is considered too big is arbitrary. The reason for this is even though context sizes are increasing with every generation, they have yet to eclipse even 32k. If we somehow read back way more than that, it probably means BIOS has done something strange, or we're running on a platform that wasn't designed for this. v2: rename load/unload to init/fini (daniel) remove ILK support for get_size() (indirectly daniel) add HAS_HW_CONTEXTS macro to clarify supported platforms (daniel) added comments (Ben) Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
2012-06-04 21:42:42 +00:00
/* This is a HW constraint. The value below is the largest known requirement
* I've seen in a spec to date, and that was a workaround for a non-shipping
* part. It should be safe to decrease this, but it's more future proof as is.
*/
#define GEN6_CONTEXT_ALIGN (64<<10)
#define GEN7_CONTEXT_ALIGN 4096
static size_t get_context_alignment(struct drm_device *dev)
{
if (IS_GEN6(dev))
return GEN6_CONTEXT_ALIGN;
return GEN7_CONTEXT_ALIGN;
}
drm/i915: preliminary context support Very basic code for context setup/destruction in the driver. Adds the file i915_gem_context.c This file implements HW context support. On gen5+ a HW context consists of an opaque GPU object which is referenced at times of context saves and restores. With RC6 enabled, the context is also referenced as the GPU enters and exists from RC6 (GPU has it's own internal power context, except on gen5). Though something like a context does exist for the media ring, the code only supports contexts for the render ring. In software, there is a distinction between contexts created by the user, and the default HW context. The default HW context is used by GPU clients that do not request setup of their own hardware context. The default context's state is never restored to help prevent programming errors. This would happen if a client ran and piggy-backed off another clients GPU state. The default context only exists to give the GPU some offset to load as the current to invoke a save of the context we actually care about. In fact, the code could likely be constructed, albeit in a more complicated fashion, to never use the default context, though that limits the driver's ability to swap out, and/or destroy other contexts. All other contexts are created as a request by the GPU client. These contexts store GPU state, and thus allow GPU clients to not re-emit state (and potentially query certain state) at any time. The kernel driver makes certain that the appropriate commands are inserted. There are 4 entry points into the contexts, init, fini, open, close. The names are self-explanatory except that init can be called during reset, and also during pm thaw/resume. As we expect our context to be preserved across these events, we do not reinitialize in this case. As Adam Jackson pointed out, The cutoff of 1MB where a HW context is considered too big is arbitrary. The reason for this is even though context sizes are increasing with every generation, they have yet to eclipse even 32k. If we somehow read back way more than that, it probably means BIOS has done something strange, or we're running on a platform that wasn't designed for this. v2: rename load/unload to init/fini (daniel) remove ILK support for get_size() (indirectly daniel) add HAS_HW_CONTEXTS macro to clarify supported platforms (daniel) added comments (Ben) Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
2012-06-04 21:42:42 +00:00
static int get_context_size(struct drm_device *dev)
{
struct drm_i915_private *dev_priv = dev->dev_private;
int ret;
u32 reg;
switch (INTEL_INFO(dev)->gen) {
case 6:
reg = I915_READ(CXT_SIZE);
ret = GEN6_CXT_TOTAL_SIZE(reg) * 64;
break;
case 7:
reg = I915_READ(GEN7_CXT_SIZE);
if (IS_HASWELL(dev))
ret = HSW_CXT_TOTAL_SIZE;
else
ret = GEN7_CXT_TOTAL_SIZE(reg) * 64;
drm/i915: preliminary context support Very basic code for context setup/destruction in the driver. Adds the file i915_gem_context.c This file implements HW context support. On gen5+ a HW context consists of an opaque GPU object which is referenced at times of context saves and restores. With RC6 enabled, the context is also referenced as the GPU enters and exists from RC6 (GPU has it's own internal power context, except on gen5). Though something like a context does exist for the media ring, the code only supports contexts for the render ring. In software, there is a distinction between contexts created by the user, and the default HW context. The default HW context is used by GPU clients that do not request setup of their own hardware context. The default context's state is never restored to help prevent programming errors. This would happen if a client ran and piggy-backed off another clients GPU state. The default context only exists to give the GPU some offset to load as the current to invoke a save of the context we actually care about. In fact, the code could likely be constructed, albeit in a more complicated fashion, to never use the default context, though that limits the driver's ability to swap out, and/or destroy other contexts. All other contexts are created as a request by the GPU client. These contexts store GPU state, and thus allow GPU clients to not re-emit state (and potentially query certain state) at any time. The kernel driver makes certain that the appropriate commands are inserted. There are 4 entry points into the contexts, init, fini, open, close. The names are self-explanatory except that init can be called during reset, and also during pm thaw/resume. As we expect our context to be preserved across these events, we do not reinitialize in this case. As Adam Jackson pointed out, The cutoff of 1MB where a HW context is considered too big is arbitrary. The reason for this is even though context sizes are increasing with every generation, they have yet to eclipse even 32k. If we somehow read back way more than that, it probably means BIOS has done something strange, or we're running on a platform that wasn't designed for this. v2: rename load/unload to init/fini (daniel) remove ILK support for get_size() (indirectly daniel) add HAS_HW_CONTEXTS macro to clarify supported platforms (daniel) added comments (Ben) Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
2012-06-04 21:42:42 +00:00
break;
case 8:
ret = GEN8_CXT_TOTAL_SIZE;
break;
drm/i915: preliminary context support Very basic code for context setup/destruction in the driver. Adds the file i915_gem_context.c This file implements HW context support. On gen5+ a HW context consists of an opaque GPU object which is referenced at times of context saves and restores. With RC6 enabled, the context is also referenced as the GPU enters and exists from RC6 (GPU has it's own internal power context, except on gen5). Though something like a context does exist for the media ring, the code only supports contexts for the render ring. In software, there is a distinction between contexts created by the user, and the default HW context. The default HW context is used by GPU clients that do not request setup of their own hardware context. The default context's state is never restored to help prevent programming errors. This would happen if a client ran and piggy-backed off another clients GPU state. The default context only exists to give the GPU some offset to load as the current to invoke a save of the context we actually care about. In fact, the code could likely be constructed, albeit in a more complicated fashion, to never use the default context, though that limits the driver's ability to swap out, and/or destroy other contexts. All other contexts are created as a request by the GPU client. These contexts store GPU state, and thus allow GPU clients to not re-emit state (and potentially query certain state) at any time. The kernel driver makes certain that the appropriate commands are inserted. There are 4 entry points into the contexts, init, fini, open, close. The names are self-explanatory except that init can be called during reset, and also during pm thaw/resume. As we expect our context to be preserved across these events, we do not reinitialize in this case. As Adam Jackson pointed out, The cutoff of 1MB where a HW context is considered too big is arbitrary. The reason for this is even though context sizes are increasing with every generation, they have yet to eclipse even 32k. If we somehow read back way more than that, it probably means BIOS has done something strange, or we're running on a platform that wasn't designed for this. v2: rename load/unload to init/fini (daniel) remove ILK support for get_size() (indirectly daniel) add HAS_HW_CONTEXTS macro to clarify supported platforms (daniel) added comments (Ben) Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
2012-06-04 21:42:42 +00:00
default:
BUG();
}
return ret;
}
void i915_gem_context_free(struct kref *ctx_ref)
{
struct intel_context *ctx = container_of(ctx_ref,
typeof(*ctx), ref);
trace_i915_context_free(ctx);
if (i915.enable_execlists)
intel_lr_context_free(ctx);
i915_ppgtt_put(ctx->ppgtt);
if (ctx->legacy_hw_ctx.rcs_state)
drm_gem_object_unreference(&ctx->legacy_hw_ctx.rcs_state->base);
list_del(&ctx->link);
kfree(ctx);
}
drm/i915/bdw: A bit more advanced LR context alloc/free Now that we have the ability to allocate our own context backing objects and we have multiplexed one of them per engine inside the context structs, we can finally allocate and free them correctly. Regarding the context size, reading the register to calculate the sizes can work, I think, however the docs are very clear about the actual context sizes on GEN8, so just hardcode that and use it. v2: Rebased on top of the Full PPGTT series. It is important to notice that at this point we have one global default context per engine, all of them using the aliasing PPGTT (as opposed to the single global default context we have with legacy HW contexts). v3: - Go back to one single global default context, this time with multiple backing objects inside. - Use different context sizes for non-render engines, as suggested by Damien (still hardcoded, since the information about the context size registers in the BSpec is, well, *lacking*). - Render ctx size is 20 (or 19) pages, but not 21 (caught by Damien). - Move default context backing object creation to intel_init_ring (so that we don't waste memory in rings that might not get initialized). v4: - Reuse the HW legacy context init/fini. - Create a separate free function. - Rename the functions with an intel_ preffix. v5: Several rebases to account for the changes in the previous patches. Signed-off-by: Ben Widawsky <ben@bwidawsk.net> (v1) Signed-off-by: Oscar Mateo <oscar.mateo@intel.com> Reviewed-by: Damien Lespiau <damien.lespiau@intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-24 16:04:14 +00:00
struct drm_i915_gem_object *
i915_gem_alloc_context_obj(struct drm_device *dev, size_t size)
{
struct drm_i915_gem_object *obj;
int ret;
obj = i915_gem_object_create_stolen(dev, size);
if (obj == NULL)
obj = i915_gem_alloc_object(dev, size);
if (obj == NULL)
return ERR_PTR(-ENOMEM);
/*
* Try to make the context utilize L3 as well as LLC.
*
* On VLV we don't have L3 controls in the PTEs so we
* shouldn't touch the cache level, especially as that
* would make the object snooped which might have a
* negative performance impact.
*/
if (INTEL_INFO(dev)->gen >= 7 && !IS_VALLEYVIEW(dev)) {
ret = i915_gem_object_set_cache_level(obj, I915_CACHE_L3_LLC);
/* Failure shouldn't ever happen this early */
if (WARN_ON(ret)) {
drm_gem_object_unreference(&obj->base);
return ERR_PTR(ret);
}
}
return obj;
}
static struct intel_context *
__create_hw_context(struct drm_device *dev,
struct drm_i915_file_private *file_priv)
{
struct drm_i915_private *dev_priv = dev->dev_private;
struct intel_context *ctx;
int ret;
ctx = kzalloc(sizeof(*ctx), GFP_KERNEL);
if (ctx == NULL)
return ERR_PTR(-ENOMEM);
kref_init(&ctx->ref);
list_add_tail(&ctx->link, &dev_priv->context_list);
if (dev_priv->hw_context_size) {
struct drm_i915_gem_object *obj =
i915_gem_alloc_context_obj(dev, dev_priv->hw_context_size);
if (IS_ERR(obj)) {
ret = PTR_ERR(obj);
goto err_out;
}
drm/i915: Emphasize that ctx->obj & ctx->is_initialized refer to the legacy rcs ctx We have already advanced that Logical Ring Contexts have their own kind of backing objects, but everything will be better explained in the Execlists series. For now, suffice it to say that the current backing object is only ever used with the render ring, so we're making this fact more explicit (which is a good reason on its own). As for the is_initialized flag, we only use to signify that the render state has been initialized (a.k.a. golden context, a.k.a. null context). It doesn't mean anything for the other engines, so make that distinction obvious. Done with the following Coccinelle patch (plus manual frobbing of the struct): @@ struct intel_context c; @@ - (c).obj + c.legacy_hw_ctx.rcs_state @@ struct intel_context *c; @@ - (c)->obj + c->legacy_hw_ctx.rcs_state @@ struct intel_context c; @@ - (c).is_initialized + c.legacy_hw_ctx.initialized @@ struct intel_context *c; @@ - (c)->is_initialized + c->legacy_hw_ctx.initialized This Execlists prep-work patch has been suggested by Chris Wilson and Daniel Vetter separately. Initially, it was two separate patches: drm/i915: Rename ctx->obj to ctx->rcs_state drm/i915: Make it obvious that ctx->id is merely a user handle Signed-off-by: Oscar Mateo <oscar.mateo@intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> [danvet: s/id/is_initialized/ to fix the subject and resolve a conflict in i915_gem_context_reset. Also introduce a new lctx local variable to avoid overtly long lines.] Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-03 15:27:59 +00:00
ctx->legacy_hw_ctx.rcs_state = obj;
}
/* Default context will never have a file_priv */
if (file_priv != NULL) {
ret = idr_alloc(&file_priv->context_idr, ctx,
drm/i915: Emphasize that ctx->id is merely a user handle This is an Execlists preparatory patch, since they make context ID become an overloaded term: - In the software, it was used to distinguish which context userspace was trying to use. - In the BSpec, the term is used to describe the 20-bits long field the hardware uses to it to discriminate the contexts that are submitted to the ELSP and inform the driver about their current status (via Context Switch Interrupts and Context Status Buffers). Initially, I tried to make the different meanings converge, but it proved impossible: - The software ctx->id is per-filp, while the hardware one needs to be globally unique. - Also, we multiplex several backing states objects per intel_context, and all of them need unique HW IDs. - I tried adding a per-filp ID and then composing the HW context ID as: ctx->id + file_priv->id + ring->id, but the fact that the hardware only uses 20-bits means we have to artificially limit the number of filps or contexts the userspace can create. The ctx->user_handle renaming bits are done with this Cocci patch (plus manual frobbing of the struct declaration): @@ struct intel_context c; @@ - (c).id + c.user_handle @@ struct intel_context *c; @@ - (c)->id + c->user_handle Also, while we are at it, s/DEFAULT_CONTEXT_ID/DEFAULT_CONTEXT_HANDLE and change the type to unsigned 32 bits. v2: s/handle/user_handle and change the type to uint32_t as suggested by Chris Wilson. Reviewed-by: Jesse Barnes <jbarnes@virtuousgeek.org> (v1) Signed-off-by: Oscar Mateo <oscar.mateo@intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-03 15:28:00 +00:00
DEFAULT_CONTEXT_HANDLE, 0, GFP_KERNEL);
if (ret < 0)
goto err_out;
} else
drm/i915: Emphasize that ctx->id is merely a user handle This is an Execlists preparatory patch, since they make context ID become an overloaded term: - In the software, it was used to distinguish which context userspace was trying to use. - In the BSpec, the term is used to describe the 20-bits long field the hardware uses to it to discriminate the contexts that are submitted to the ELSP and inform the driver about their current status (via Context Switch Interrupts and Context Status Buffers). Initially, I tried to make the different meanings converge, but it proved impossible: - The software ctx->id is per-filp, while the hardware one needs to be globally unique. - Also, we multiplex several backing states objects per intel_context, and all of them need unique HW IDs. - I tried adding a per-filp ID and then composing the HW context ID as: ctx->id + file_priv->id + ring->id, but the fact that the hardware only uses 20-bits means we have to artificially limit the number of filps or contexts the userspace can create. The ctx->user_handle renaming bits are done with this Cocci patch (plus manual frobbing of the struct declaration): @@ struct intel_context c; @@ - (c).id + c.user_handle @@ struct intel_context *c; @@ - (c)->id + c->user_handle Also, while we are at it, s/DEFAULT_CONTEXT_ID/DEFAULT_CONTEXT_HANDLE and change the type to unsigned 32 bits. v2: s/handle/user_handle and change the type to uint32_t as suggested by Chris Wilson. Reviewed-by: Jesse Barnes <jbarnes@virtuousgeek.org> (v1) Signed-off-by: Oscar Mateo <oscar.mateo@intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-03 15:28:00 +00:00
ret = DEFAULT_CONTEXT_HANDLE;
ctx->file_priv = file_priv;
drm/i915: Emphasize that ctx->id is merely a user handle This is an Execlists preparatory patch, since they make context ID become an overloaded term: - In the software, it was used to distinguish which context userspace was trying to use. - In the BSpec, the term is used to describe the 20-bits long field the hardware uses to it to discriminate the contexts that are submitted to the ELSP and inform the driver about their current status (via Context Switch Interrupts and Context Status Buffers). Initially, I tried to make the different meanings converge, but it proved impossible: - The software ctx->id is per-filp, while the hardware one needs to be globally unique. - Also, we multiplex several backing states objects per intel_context, and all of them need unique HW IDs. - I tried adding a per-filp ID and then composing the HW context ID as: ctx->id + file_priv->id + ring->id, but the fact that the hardware only uses 20-bits means we have to artificially limit the number of filps or contexts the userspace can create. The ctx->user_handle renaming bits are done with this Cocci patch (plus manual frobbing of the struct declaration): @@ struct intel_context c; @@ - (c).id + c.user_handle @@ struct intel_context *c; @@ - (c)->id + c->user_handle Also, while we are at it, s/DEFAULT_CONTEXT_ID/DEFAULT_CONTEXT_HANDLE and change the type to unsigned 32 bits. v2: s/handle/user_handle and change the type to uint32_t as suggested by Chris Wilson. Reviewed-by: Jesse Barnes <jbarnes@virtuousgeek.org> (v1) Signed-off-by: Oscar Mateo <oscar.mateo@intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-03 15:28:00 +00:00
ctx->user_handle = ret;
drm/i915: Do remaps for all contexts On both Ivybridge and Haswell, row remapping information is saved and restored with context. This means, we never actually properly supported the l3 remapping because our sysfs interface is asynchronous (and not tied to any context), and the known faulty HW would be reused by the next context to run. Not that due to the asynchronous nature of the sysfs entry, there is no point modifying the registers for the existing context. Instead we set a flag for all contexts to load the correct remapping information on the next run. Interested clients can use debugfs to determine whether or not the row has been remapped. One could propose at this point that we just do the remapping in the kernel. I guess since we have to maintain the sysfs interface anyway, I'm not sure how useful it is, and I do like keeping the policy in userspace; (it wasn't my original decision to make the interface the way it is, so I'm not attached). v2: Force a context switch when we have a remap on the next switch. (Ville) Don't let userspace use the interface with disabled contexts. v3: Don't force a context switch, just let it nop Improper context slice remap initialization, 1<<1 instead of 1<<i, but I rewrote it to avoid a second round of confusion. Error print moved to error path (All Ville) Added a comment on why the slice remap initialization happens. CC: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Ben Widawsky <ben@bwidawsk.net> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2013-09-19 02:03:18 +00:00
/* NB: Mark all slices as needing a remap so that when the context first
* loads it will restore whatever remap state already exists. If there
* is no remap info, it will be a NOP. */
ctx->remap_slice = (1 << NUM_L3_SLICES(dev)) - 1;
ctx->hang_stats.ban_period_seconds = DRM_I915_CTX_BAN_PERIOD;
return ctx;
err_out:
i915_gem_context_unreference(ctx);
return ERR_PTR(ret);
}
drm/i915: preliminary context support Very basic code for context setup/destruction in the driver. Adds the file i915_gem_context.c This file implements HW context support. On gen5+ a HW context consists of an opaque GPU object which is referenced at times of context saves and restores. With RC6 enabled, the context is also referenced as the GPU enters and exists from RC6 (GPU has it's own internal power context, except on gen5). Though something like a context does exist for the media ring, the code only supports contexts for the render ring. In software, there is a distinction between contexts created by the user, and the default HW context. The default HW context is used by GPU clients that do not request setup of their own hardware context. The default context's state is never restored to help prevent programming errors. This would happen if a client ran and piggy-backed off another clients GPU state. The default context only exists to give the GPU some offset to load as the current to invoke a save of the context we actually care about. In fact, the code could likely be constructed, albeit in a more complicated fashion, to never use the default context, though that limits the driver's ability to swap out, and/or destroy other contexts. All other contexts are created as a request by the GPU client. These contexts store GPU state, and thus allow GPU clients to not re-emit state (and potentially query certain state) at any time. The kernel driver makes certain that the appropriate commands are inserted. There are 4 entry points into the contexts, init, fini, open, close. The names are self-explanatory except that init can be called during reset, and also during pm thaw/resume. As we expect our context to be preserved across these events, we do not reinitialize in this case. As Adam Jackson pointed out, The cutoff of 1MB where a HW context is considered too big is arbitrary. The reason for this is even though context sizes are increasing with every generation, they have yet to eclipse even 32k. If we somehow read back way more than that, it probably means BIOS has done something strange, or we're running on a platform that wasn't designed for this. v2: rename load/unload to init/fini (daniel) remove ILK support for get_size() (indirectly daniel) add HAS_HW_CONTEXTS macro to clarify supported platforms (daniel) added comments (Ben) Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
2012-06-04 21:42:42 +00:00
/**
* The default context needs to exist per ring that uses contexts. It stores the
* context state of the GPU for applications that don't utilize HW contexts, as
* well as an idle case.
*/
static struct intel_context *
i915_gem_create_context(struct drm_device *dev,
struct drm_i915_file_private *file_priv)
drm/i915: preliminary context support Very basic code for context setup/destruction in the driver. Adds the file i915_gem_context.c This file implements HW context support. On gen5+ a HW context consists of an opaque GPU object which is referenced at times of context saves and restores. With RC6 enabled, the context is also referenced as the GPU enters and exists from RC6 (GPU has it's own internal power context, except on gen5). Though something like a context does exist for the media ring, the code only supports contexts for the render ring. In software, there is a distinction between contexts created by the user, and the default HW context. The default HW context is used by GPU clients that do not request setup of their own hardware context. The default context's state is never restored to help prevent programming errors. This would happen if a client ran and piggy-backed off another clients GPU state. The default context only exists to give the GPU some offset to load as the current to invoke a save of the context we actually care about. In fact, the code could likely be constructed, albeit in a more complicated fashion, to never use the default context, though that limits the driver's ability to swap out, and/or destroy other contexts. All other contexts are created as a request by the GPU client. These contexts store GPU state, and thus allow GPU clients to not re-emit state (and potentially query certain state) at any time. The kernel driver makes certain that the appropriate commands are inserted. There are 4 entry points into the contexts, init, fini, open, close. The names are self-explanatory except that init can be called during reset, and also during pm thaw/resume. As we expect our context to be preserved across these events, we do not reinitialize in this case. As Adam Jackson pointed out, The cutoff of 1MB where a HW context is considered too big is arbitrary. The reason for this is even though context sizes are increasing with every generation, they have yet to eclipse even 32k. If we somehow read back way more than that, it probably means BIOS has done something strange, or we're running on a platform that wasn't designed for this. v2: rename load/unload to init/fini (daniel) remove ILK support for get_size() (indirectly daniel) add HAS_HW_CONTEXTS macro to clarify supported platforms (daniel) added comments (Ben) Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
2012-06-04 21:42:42 +00:00
{
drm/i915: Always pin the default context Through a twisty and circuituous path it is possible to currently trick the code into creating a default context and forgetting to pin it immediately into the GGTT. (This requires a system using contexts without an aliasing ppgtt, which is currently restricted to Baytrails machines manually specifying a module parameter to force enable contexts, or on Sandybridge and later that manually disable the aliasing ppgtt.) The consequence is that during module unload we attempt to unpin the default context twice and encounter a BUG remonstrating that we attempt to unpin an unbound object. [ 161.002869] Kernel BUG at f84861f8 [verbose debug info unavailable] [ 161.002875] invalid opcode: 0000 [#1] SMP [ 161.002882] Modules linked in: coretemp kvm_intel kvm crc32_pclmul aesni_intel aes_i586 xts lrw gf128mul ablk_helper cryptd hid_sensor_accel_3d hid_sensor_gyro_3d hid_sensor_magn_3d hid_sensor_trigger industrialio_triggered_buffer kfifo_buf industrialio hid_sensor_iio_common snd_hda_codec_hdmi snd_hda_codec_realtek snd_hda_intel snd_hda_codec snd_hwdep snd_pcm snd_page_alloc snd_seq_midi snd_seq_midi_event dm_multipath scsi_dh asix ppdev usbnet snd_rawmidi mii hid_sensor_hub microcode snd_seq rfcomm bnep snd_seq_device bluetooth snd_timer snd parport_pc binfmt_misc soundcore dw_dmac_pci dw_dmac_core mac_hid lp parport dm_mirror dm_region_hash dm_log hid_generic usbhid hid i915(O-) drm_kms_helper(O) igb dca ptp pps_core i2c_algo_bit drm(O) ahci libahci video [ 161.002991] CPU: 0 PID: 2114 Comm: rmmod Tainted: G W O 3.13.0-rc8+ #2 [ 161.002997] Hardware name: NEXCOM VTC1010/Aptio CRB, BIOS 5.6.5 09/24/2013 [ 161.003004] task: dbdd6800 ti: dbe0e000 task.ti: dbe0e000 [ 161.003010] EIP: 0060:[<f84861f8>] EFLAGS: 00010246 CPU: 0 [ 161.003044] EIP is at i915_gem_object_ggtt_unpin+0x88/0x90 [i915] [ 161.003050] EAX: dfce3840 EBX: 00000000 ECX: dfafd690 EDX: dfce3874 [ 161.003056] ESI: c0086b40 EDI: df962e00 EBP: dbe0fe1c ESP: dbe0fe0c [ 161.003062] DS: 007b ES: 007b FS: 00d8 GS: 00e0 SS: 0068 [ 161.003068] CR0: 8005003b CR2: b7718000 CR3: 1bec0000 CR4: 001007f0 [ 161.003076] Stack: [ 161.003081] 00afc014 00000004 c0086b40 dfafc000 dbe0fe38 f8487e5a dfaa5400 c0086b40 [ 161.003099] dfafc000 dfaa5400 dfaa5414 dbe0fe58 f84741aa 00000000 f89c34b9 dfaa5414 [ 161.003117] dfaa5400 dfaa5400 f644b000 dbe0fe6c f89a5443 dfaa5400 f8505000 f644b000 [ 161.003134] Call Trace: [ 161.003169] [<f8487e5a>] i915_gem_context_fini+0xba/0x1c0 [i915] [ 161.003202] [<f84741aa>] i915_driver_unload+0x1fa/0x2f0 [i915] [ 161.003232] [<f89a5443>] drm_dev_unregister+0x23/0x90 [drm] [ 161.003259] [<f89a54ed>] drm_put_dev+0x3d/0x70 [drm] [ 161.003294] [<f8470615>] i915_pci_remove+0x15/0x20 [i915] [ 161.003306] [<c1338a6f>] pci_device_remove+0x2f/0xa0 [ 161.003317] [<c140c871>] __device_release_driver+0x61/0xc0 [ 161.003328] [<c140d12f>] driver_detach+0x8f/0xa0 [ 161.003341] [<c140c54f>] bus_remove_driver+0x4f/0xc0 [ 161.003353] [<c140d708>] driver_unregister+0x28/0x60 [ 161.003362] [<c10cee42>] ? stop_cpus+0x32/0x40 [ 161.003372] [<c10bd510>] ? module_refcount+0x90/0x90 [ 161.003383] [<c13378c5>] pci_unregister_driver+0x15/0x60 [ 161.003413] [<f89a739f>] drm_pci_exit+0x9f/0xb0 [drm] [ 161.003458] [<f84e624a>] i915_exit+0x1b/0x1d [i915] [ 161.003468] [<c10bf8a8>] SyS_delete_module+0x158/0x1f0 [ 161.003480] [<c1173d5d>] ? ____fput+0xd/0x10 [ 161.003488] [<c106f0fe>] ? task_work_run+0x7e/0xb0 [ 161.003499] [<c165a68d>] sysenter_do_call+0x12/0x28 [ 161.003505] Code: 0f b6 4d f3 8d 51 0f 83 e1 f0 83 e2 0f 09 d1 84 d2 88 48 54 75 07 80 a7 91 00 00 00 7f 83 c4 04 5b 5e 5f 5d c3 8d b6 00 00 00 00 <0f> 0b 8d b6 00 00 00 00 55 89 e5 57 56 53 83 ec 64 3e 8d 74 26 [ 161.003586] EIP: [<f84861f8>] i915_gem_object_ggtt_unpin+0x88/0x90 [i915] SS:ESP 0068:dbe0fe0c v2: Rename the local variable (is_default_ctx) to avoid confusion with the function is_default_ctx(). And correct Jesse's email address. Reported-by: Jesse Barnes <jbarnes@virtuousgeek.org> Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=73985 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Jesse Barnes <jbarnes@virtuousgeek.org> Cc: Ben Widawsky <benjamin.widawsky@intel.com> Tested-by: Jesse Barnes <jbarnes@virtuousgeek.org> Reviewed-by: Ben Widawsky <benjamin.widawsky@intel.com> [danvet: Fix up the rebase fail from my first attempt, thankfully pointed out by Ville.] Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-01-23 19:40:02 +00:00
const bool is_global_default_ctx = file_priv == NULL;
struct intel_context *ctx;
int ret = 0;
BUG_ON(!mutex_is_locked(&dev->struct_mutex));
ctx = __create_hw_context(dev, file_priv);
if (IS_ERR(ctx))
return ctx;
drm/i915: Emphasize that ctx->obj & ctx->is_initialized refer to the legacy rcs ctx We have already advanced that Logical Ring Contexts have their own kind of backing objects, but everything will be better explained in the Execlists series. For now, suffice it to say that the current backing object is only ever used with the render ring, so we're making this fact more explicit (which is a good reason on its own). As for the is_initialized flag, we only use to signify that the render state has been initialized (a.k.a. golden context, a.k.a. null context). It doesn't mean anything for the other engines, so make that distinction obvious. Done with the following Coccinelle patch (plus manual frobbing of the struct): @@ struct intel_context c; @@ - (c).obj + c.legacy_hw_ctx.rcs_state @@ struct intel_context *c; @@ - (c)->obj + c->legacy_hw_ctx.rcs_state @@ struct intel_context c; @@ - (c).is_initialized + c.legacy_hw_ctx.initialized @@ struct intel_context *c; @@ - (c)->is_initialized + c->legacy_hw_ctx.initialized This Execlists prep-work patch has been suggested by Chris Wilson and Daniel Vetter separately. Initially, it was two separate patches: drm/i915: Rename ctx->obj to ctx->rcs_state drm/i915: Make it obvious that ctx->id is merely a user handle Signed-off-by: Oscar Mateo <oscar.mateo@intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> [danvet: s/id/is_initialized/ to fix the subject and resolve a conflict in i915_gem_context_reset. Also introduce a new lctx local variable to avoid overtly long lines.] Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-03 15:27:59 +00:00
if (is_global_default_ctx && ctx->legacy_hw_ctx.rcs_state) {
drm/i915: Always pin the default context Through a twisty and circuituous path it is possible to currently trick the code into creating a default context and forgetting to pin it immediately into the GGTT. (This requires a system using contexts without an aliasing ppgtt, which is currently restricted to Baytrails machines manually specifying a module parameter to force enable contexts, or on Sandybridge and later that manually disable the aliasing ppgtt.) The consequence is that during module unload we attempt to unpin the default context twice and encounter a BUG remonstrating that we attempt to unpin an unbound object. [ 161.002869] Kernel BUG at f84861f8 [verbose debug info unavailable] [ 161.002875] invalid opcode: 0000 [#1] SMP [ 161.002882] Modules linked in: coretemp kvm_intel kvm crc32_pclmul aesni_intel aes_i586 xts lrw gf128mul ablk_helper cryptd hid_sensor_accel_3d hid_sensor_gyro_3d hid_sensor_magn_3d hid_sensor_trigger industrialio_triggered_buffer kfifo_buf industrialio hid_sensor_iio_common snd_hda_codec_hdmi snd_hda_codec_realtek snd_hda_intel snd_hda_codec snd_hwdep snd_pcm snd_page_alloc snd_seq_midi snd_seq_midi_event dm_multipath scsi_dh asix ppdev usbnet snd_rawmidi mii hid_sensor_hub microcode snd_seq rfcomm bnep snd_seq_device bluetooth snd_timer snd parport_pc binfmt_misc soundcore dw_dmac_pci dw_dmac_core mac_hid lp parport dm_mirror dm_region_hash dm_log hid_generic usbhid hid i915(O-) drm_kms_helper(O) igb dca ptp pps_core i2c_algo_bit drm(O) ahci libahci video [ 161.002991] CPU: 0 PID: 2114 Comm: rmmod Tainted: G W O 3.13.0-rc8+ #2 [ 161.002997] Hardware name: NEXCOM VTC1010/Aptio CRB, BIOS 5.6.5 09/24/2013 [ 161.003004] task: dbdd6800 ti: dbe0e000 task.ti: dbe0e000 [ 161.003010] EIP: 0060:[<f84861f8>] EFLAGS: 00010246 CPU: 0 [ 161.003044] EIP is at i915_gem_object_ggtt_unpin+0x88/0x90 [i915] [ 161.003050] EAX: dfce3840 EBX: 00000000 ECX: dfafd690 EDX: dfce3874 [ 161.003056] ESI: c0086b40 EDI: df962e00 EBP: dbe0fe1c ESP: dbe0fe0c [ 161.003062] DS: 007b ES: 007b FS: 00d8 GS: 00e0 SS: 0068 [ 161.003068] CR0: 8005003b CR2: b7718000 CR3: 1bec0000 CR4: 001007f0 [ 161.003076] Stack: [ 161.003081] 00afc014 00000004 c0086b40 dfafc000 dbe0fe38 f8487e5a dfaa5400 c0086b40 [ 161.003099] dfafc000 dfaa5400 dfaa5414 dbe0fe58 f84741aa 00000000 f89c34b9 dfaa5414 [ 161.003117] dfaa5400 dfaa5400 f644b000 dbe0fe6c f89a5443 dfaa5400 f8505000 f644b000 [ 161.003134] Call Trace: [ 161.003169] [<f8487e5a>] i915_gem_context_fini+0xba/0x1c0 [i915] [ 161.003202] [<f84741aa>] i915_driver_unload+0x1fa/0x2f0 [i915] [ 161.003232] [<f89a5443>] drm_dev_unregister+0x23/0x90 [drm] [ 161.003259] [<f89a54ed>] drm_put_dev+0x3d/0x70 [drm] [ 161.003294] [<f8470615>] i915_pci_remove+0x15/0x20 [i915] [ 161.003306] [<c1338a6f>] pci_device_remove+0x2f/0xa0 [ 161.003317] [<c140c871>] __device_release_driver+0x61/0xc0 [ 161.003328] [<c140d12f>] driver_detach+0x8f/0xa0 [ 161.003341] [<c140c54f>] bus_remove_driver+0x4f/0xc0 [ 161.003353] [<c140d708>] driver_unregister+0x28/0x60 [ 161.003362] [<c10cee42>] ? stop_cpus+0x32/0x40 [ 161.003372] [<c10bd510>] ? module_refcount+0x90/0x90 [ 161.003383] [<c13378c5>] pci_unregister_driver+0x15/0x60 [ 161.003413] [<f89a739f>] drm_pci_exit+0x9f/0xb0 [drm] [ 161.003458] [<f84e624a>] i915_exit+0x1b/0x1d [i915] [ 161.003468] [<c10bf8a8>] SyS_delete_module+0x158/0x1f0 [ 161.003480] [<c1173d5d>] ? ____fput+0xd/0x10 [ 161.003488] [<c106f0fe>] ? task_work_run+0x7e/0xb0 [ 161.003499] [<c165a68d>] sysenter_do_call+0x12/0x28 [ 161.003505] Code: 0f b6 4d f3 8d 51 0f 83 e1 f0 83 e2 0f 09 d1 84 d2 88 48 54 75 07 80 a7 91 00 00 00 7f 83 c4 04 5b 5e 5f 5d c3 8d b6 00 00 00 00 <0f> 0b 8d b6 00 00 00 00 55 89 e5 57 56 53 83 ec 64 3e 8d 74 26 [ 161.003586] EIP: [<f84861f8>] i915_gem_object_ggtt_unpin+0x88/0x90 [i915] SS:ESP 0068:dbe0fe0c v2: Rename the local variable (is_default_ctx) to avoid confusion with the function is_default_ctx(). And correct Jesse's email address. Reported-by: Jesse Barnes <jbarnes@virtuousgeek.org> Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=73985 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Jesse Barnes <jbarnes@virtuousgeek.org> Cc: Ben Widawsky <benjamin.widawsky@intel.com> Tested-by: Jesse Barnes <jbarnes@virtuousgeek.org> Reviewed-by: Ben Widawsky <benjamin.widawsky@intel.com> [danvet: Fix up the rebase fail from my first attempt, thankfully pointed out by Ville.] Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-01-23 19:40:02 +00:00
/* We may need to do things with the shrinker which
* require us to immediately switch back to the default
* context. This can cause a problem as pinning the
* default context also requires GTT space which may not
* be available. To avoid this we always pin the default
* context.
*/
drm/i915: Emphasize that ctx->obj & ctx->is_initialized refer to the legacy rcs ctx We have already advanced that Logical Ring Contexts have their own kind of backing objects, but everything will be better explained in the Execlists series. For now, suffice it to say that the current backing object is only ever used with the render ring, so we're making this fact more explicit (which is a good reason on its own). As for the is_initialized flag, we only use to signify that the render state has been initialized (a.k.a. golden context, a.k.a. null context). It doesn't mean anything for the other engines, so make that distinction obvious. Done with the following Coccinelle patch (plus manual frobbing of the struct): @@ struct intel_context c; @@ - (c).obj + c.legacy_hw_ctx.rcs_state @@ struct intel_context *c; @@ - (c)->obj + c->legacy_hw_ctx.rcs_state @@ struct intel_context c; @@ - (c).is_initialized + c.legacy_hw_ctx.initialized @@ struct intel_context *c; @@ - (c)->is_initialized + c->legacy_hw_ctx.initialized This Execlists prep-work patch has been suggested by Chris Wilson and Daniel Vetter separately. Initially, it was two separate patches: drm/i915: Rename ctx->obj to ctx->rcs_state drm/i915: Make it obvious that ctx->id is merely a user handle Signed-off-by: Oscar Mateo <oscar.mateo@intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> [danvet: s/id/is_initialized/ to fix the subject and resolve a conflict in i915_gem_context_reset. Also introduce a new lctx local variable to avoid overtly long lines.] Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-03 15:27:59 +00:00
ret = i915_gem_obj_ggtt_pin(ctx->legacy_hw_ctx.rcs_state,
get_context_alignment(dev), 0);
drm/i915: Always pin the default context Through a twisty and circuituous path it is possible to currently trick the code into creating a default context and forgetting to pin it immediately into the GGTT. (This requires a system using contexts without an aliasing ppgtt, which is currently restricted to Baytrails machines manually specifying a module parameter to force enable contexts, or on Sandybridge and later that manually disable the aliasing ppgtt.) The consequence is that during module unload we attempt to unpin the default context twice and encounter a BUG remonstrating that we attempt to unpin an unbound object. [ 161.002869] Kernel BUG at f84861f8 [verbose debug info unavailable] [ 161.002875] invalid opcode: 0000 [#1] SMP [ 161.002882] Modules linked in: coretemp kvm_intel kvm crc32_pclmul aesni_intel aes_i586 xts lrw gf128mul ablk_helper cryptd hid_sensor_accel_3d hid_sensor_gyro_3d hid_sensor_magn_3d hid_sensor_trigger industrialio_triggered_buffer kfifo_buf industrialio hid_sensor_iio_common snd_hda_codec_hdmi snd_hda_codec_realtek snd_hda_intel snd_hda_codec snd_hwdep snd_pcm snd_page_alloc snd_seq_midi snd_seq_midi_event dm_multipath scsi_dh asix ppdev usbnet snd_rawmidi mii hid_sensor_hub microcode snd_seq rfcomm bnep snd_seq_device bluetooth snd_timer snd parport_pc binfmt_misc soundcore dw_dmac_pci dw_dmac_core mac_hid lp parport dm_mirror dm_region_hash dm_log hid_generic usbhid hid i915(O-) drm_kms_helper(O) igb dca ptp pps_core i2c_algo_bit drm(O) ahci libahci video [ 161.002991] CPU: 0 PID: 2114 Comm: rmmod Tainted: G W O 3.13.0-rc8+ #2 [ 161.002997] Hardware name: NEXCOM VTC1010/Aptio CRB, BIOS 5.6.5 09/24/2013 [ 161.003004] task: dbdd6800 ti: dbe0e000 task.ti: dbe0e000 [ 161.003010] EIP: 0060:[<f84861f8>] EFLAGS: 00010246 CPU: 0 [ 161.003044] EIP is at i915_gem_object_ggtt_unpin+0x88/0x90 [i915] [ 161.003050] EAX: dfce3840 EBX: 00000000 ECX: dfafd690 EDX: dfce3874 [ 161.003056] ESI: c0086b40 EDI: df962e00 EBP: dbe0fe1c ESP: dbe0fe0c [ 161.003062] DS: 007b ES: 007b FS: 00d8 GS: 00e0 SS: 0068 [ 161.003068] CR0: 8005003b CR2: b7718000 CR3: 1bec0000 CR4: 001007f0 [ 161.003076] Stack: [ 161.003081] 00afc014 00000004 c0086b40 dfafc000 dbe0fe38 f8487e5a dfaa5400 c0086b40 [ 161.003099] dfafc000 dfaa5400 dfaa5414 dbe0fe58 f84741aa 00000000 f89c34b9 dfaa5414 [ 161.003117] dfaa5400 dfaa5400 f644b000 dbe0fe6c f89a5443 dfaa5400 f8505000 f644b000 [ 161.003134] Call Trace: [ 161.003169] [<f8487e5a>] i915_gem_context_fini+0xba/0x1c0 [i915] [ 161.003202] [<f84741aa>] i915_driver_unload+0x1fa/0x2f0 [i915] [ 161.003232] [<f89a5443>] drm_dev_unregister+0x23/0x90 [drm] [ 161.003259] [<f89a54ed>] drm_put_dev+0x3d/0x70 [drm] [ 161.003294] [<f8470615>] i915_pci_remove+0x15/0x20 [i915] [ 161.003306] [<c1338a6f>] pci_device_remove+0x2f/0xa0 [ 161.003317] [<c140c871>] __device_release_driver+0x61/0xc0 [ 161.003328] [<c140d12f>] driver_detach+0x8f/0xa0 [ 161.003341] [<c140c54f>] bus_remove_driver+0x4f/0xc0 [ 161.003353] [<c140d708>] driver_unregister+0x28/0x60 [ 161.003362] [<c10cee42>] ? stop_cpus+0x32/0x40 [ 161.003372] [<c10bd510>] ? module_refcount+0x90/0x90 [ 161.003383] [<c13378c5>] pci_unregister_driver+0x15/0x60 [ 161.003413] [<f89a739f>] drm_pci_exit+0x9f/0xb0 [drm] [ 161.003458] [<f84e624a>] i915_exit+0x1b/0x1d [i915] [ 161.003468] [<c10bf8a8>] SyS_delete_module+0x158/0x1f0 [ 161.003480] [<c1173d5d>] ? ____fput+0xd/0x10 [ 161.003488] [<c106f0fe>] ? task_work_run+0x7e/0xb0 [ 161.003499] [<c165a68d>] sysenter_do_call+0x12/0x28 [ 161.003505] Code: 0f b6 4d f3 8d 51 0f 83 e1 f0 83 e2 0f 09 d1 84 d2 88 48 54 75 07 80 a7 91 00 00 00 7f 83 c4 04 5b 5e 5f 5d c3 8d b6 00 00 00 00 <0f> 0b 8d b6 00 00 00 00 55 89 e5 57 56 53 83 ec 64 3e 8d 74 26 [ 161.003586] EIP: [<f84861f8>] i915_gem_object_ggtt_unpin+0x88/0x90 [i915] SS:ESP 0068:dbe0fe0c v2: Rename the local variable (is_default_ctx) to avoid confusion with the function is_default_ctx(). And correct Jesse's email address. Reported-by: Jesse Barnes <jbarnes@virtuousgeek.org> Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=73985 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Jesse Barnes <jbarnes@virtuousgeek.org> Cc: Ben Widawsky <benjamin.widawsky@intel.com> Tested-by: Jesse Barnes <jbarnes@virtuousgeek.org> Reviewed-by: Ben Widawsky <benjamin.widawsky@intel.com> [danvet: Fix up the rebase fail from my first attempt, thankfully pointed out by Ville.] Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-01-23 19:40:02 +00:00
if (ret) {
DRM_DEBUG_DRIVER("Couldn't pin %d\n", ret);
goto err_destroy;
}
}
if (USES_FULL_PPGTT(dev)) {
struct i915_hw_ppgtt *ppgtt = i915_ppgtt_create(dev, file_priv);
if (IS_ERR_OR_NULL(ppgtt)) {
DRM_DEBUG_DRIVER("PPGTT setup failed (%ld)\n",
PTR_ERR(ppgtt));
ret = PTR_ERR(ppgtt);
drm/i915: Always pin the default context Through a twisty and circuituous path it is possible to currently trick the code into creating a default context and forgetting to pin it immediately into the GGTT. (This requires a system using contexts without an aliasing ppgtt, which is currently restricted to Baytrails machines manually specifying a module parameter to force enable contexts, or on Sandybridge and later that manually disable the aliasing ppgtt.) The consequence is that during module unload we attempt to unpin the default context twice and encounter a BUG remonstrating that we attempt to unpin an unbound object. [ 161.002869] Kernel BUG at f84861f8 [verbose debug info unavailable] [ 161.002875] invalid opcode: 0000 [#1] SMP [ 161.002882] Modules linked in: coretemp kvm_intel kvm crc32_pclmul aesni_intel aes_i586 xts lrw gf128mul ablk_helper cryptd hid_sensor_accel_3d hid_sensor_gyro_3d hid_sensor_magn_3d hid_sensor_trigger industrialio_triggered_buffer kfifo_buf industrialio hid_sensor_iio_common snd_hda_codec_hdmi snd_hda_codec_realtek snd_hda_intel snd_hda_codec snd_hwdep snd_pcm snd_page_alloc snd_seq_midi snd_seq_midi_event dm_multipath scsi_dh asix ppdev usbnet snd_rawmidi mii hid_sensor_hub microcode snd_seq rfcomm bnep snd_seq_device bluetooth snd_timer snd parport_pc binfmt_misc soundcore dw_dmac_pci dw_dmac_core mac_hid lp parport dm_mirror dm_region_hash dm_log hid_generic usbhid hid i915(O-) drm_kms_helper(O) igb dca ptp pps_core i2c_algo_bit drm(O) ahci libahci video [ 161.002991] CPU: 0 PID: 2114 Comm: rmmod Tainted: G W O 3.13.0-rc8+ #2 [ 161.002997] Hardware name: NEXCOM VTC1010/Aptio CRB, BIOS 5.6.5 09/24/2013 [ 161.003004] task: dbdd6800 ti: dbe0e000 task.ti: dbe0e000 [ 161.003010] EIP: 0060:[<f84861f8>] EFLAGS: 00010246 CPU: 0 [ 161.003044] EIP is at i915_gem_object_ggtt_unpin+0x88/0x90 [i915] [ 161.003050] EAX: dfce3840 EBX: 00000000 ECX: dfafd690 EDX: dfce3874 [ 161.003056] ESI: c0086b40 EDI: df962e00 EBP: dbe0fe1c ESP: dbe0fe0c [ 161.003062] DS: 007b ES: 007b FS: 00d8 GS: 00e0 SS: 0068 [ 161.003068] CR0: 8005003b CR2: b7718000 CR3: 1bec0000 CR4: 001007f0 [ 161.003076] Stack: [ 161.003081] 00afc014 00000004 c0086b40 dfafc000 dbe0fe38 f8487e5a dfaa5400 c0086b40 [ 161.003099] dfafc000 dfaa5400 dfaa5414 dbe0fe58 f84741aa 00000000 f89c34b9 dfaa5414 [ 161.003117] dfaa5400 dfaa5400 f644b000 dbe0fe6c f89a5443 dfaa5400 f8505000 f644b000 [ 161.003134] Call Trace: [ 161.003169] [<f8487e5a>] i915_gem_context_fini+0xba/0x1c0 [i915] [ 161.003202] [<f84741aa>] i915_driver_unload+0x1fa/0x2f0 [i915] [ 161.003232] [<f89a5443>] drm_dev_unregister+0x23/0x90 [drm] [ 161.003259] [<f89a54ed>] drm_put_dev+0x3d/0x70 [drm] [ 161.003294] [<f8470615>] i915_pci_remove+0x15/0x20 [i915] [ 161.003306] [<c1338a6f>] pci_device_remove+0x2f/0xa0 [ 161.003317] [<c140c871>] __device_release_driver+0x61/0xc0 [ 161.003328] [<c140d12f>] driver_detach+0x8f/0xa0 [ 161.003341] [<c140c54f>] bus_remove_driver+0x4f/0xc0 [ 161.003353] [<c140d708>] driver_unregister+0x28/0x60 [ 161.003362] [<c10cee42>] ? stop_cpus+0x32/0x40 [ 161.003372] [<c10bd510>] ? module_refcount+0x90/0x90 [ 161.003383] [<c13378c5>] pci_unregister_driver+0x15/0x60 [ 161.003413] [<f89a739f>] drm_pci_exit+0x9f/0xb0 [drm] [ 161.003458] [<f84e624a>] i915_exit+0x1b/0x1d [i915] [ 161.003468] [<c10bf8a8>] SyS_delete_module+0x158/0x1f0 [ 161.003480] [<c1173d5d>] ? ____fput+0xd/0x10 [ 161.003488] [<c106f0fe>] ? task_work_run+0x7e/0xb0 [ 161.003499] [<c165a68d>] sysenter_do_call+0x12/0x28 [ 161.003505] Code: 0f b6 4d f3 8d 51 0f 83 e1 f0 83 e2 0f 09 d1 84 d2 88 48 54 75 07 80 a7 91 00 00 00 7f 83 c4 04 5b 5e 5f 5d c3 8d b6 00 00 00 00 <0f> 0b 8d b6 00 00 00 00 55 89 e5 57 56 53 83 ec 64 3e 8d 74 26 [ 161.003586] EIP: [<f84861f8>] i915_gem_object_ggtt_unpin+0x88/0x90 [i915] SS:ESP 0068:dbe0fe0c v2: Rename the local variable (is_default_ctx) to avoid confusion with the function is_default_ctx(). And correct Jesse's email address. Reported-by: Jesse Barnes <jbarnes@virtuousgeek.org> Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=73985 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Jesse Barnes <jbarnes@virtuousgeek.org> Cc: Ben Widawsky <benjamin.widawsky@intel.com> Tested-by: Jesse Barnes <jbarnes@virtuousgeek.org> Reviewed-by: Ben Widawsky <benjamin.widawsky@intel.com> [danvet: Fix up the rebase fail from my first attempt, thankfully pointed out by Ville.] Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-01-23 19:40:02 +00:00
goto err_unpin;
}
ctx->ppgtt = ppgtt;
}
trace_i915_context_create(ctx);
return ctx;
drm/i915: Always pin the default context Through a twisty and circuituous path it is possible to currently trick the code into creating a default context and forgetting to pin it immediately into the GGTT. (This requires a system using contexts without an aliasing ppgtt, which is currently restricted to Baytrails machines manually specifying a module parameter to force enable contexts, or on Sandybridge and later that manually disable the aliasing ppgtt.) The consequence is that during module unload we attempt to unpin the default context twice and encounter a BUG remonstrating that we attempt to unpin an unbound object. [ 161.002869] Kernel BUG at f84861f8 [verbose debug info unavailable] [ 161.002875] invalid opcode: 0000 [#1] SMP [ 161.002882] Modules linked in: coretemp kvm_intel kvm crc32_pclmul aesni_intel aes_i586 xts lrw gf128mul ablk_helper cryptd hid_sensor_accel_3d hid_sensor_gyro_3d hid_sensor_magn_3d hid_sensor_trigger industrialio_triggered_buffer kfifo_buf industrialio hid_sensor_iio_common snd_hda_codec_hdmi snd_hda_codec_realtek snd_hda_intel snd_hda_codec snd_hwdep snd_pcm snd_page_alloc snd_seq_midi snd_seq_midi_event dm_multipath scsi_dh asix ppdev usbnet snd_rawmidi mii hid_sensor_hub microcode snd_seq rfcomm bnep snd_seq_device bluetooth snd_timer snd parport_pc binfmt_misc soundcore dw_dmac_pci dw_dmac_core mac_hid lp parport dm_mirror dm_region_hash dm_log hid_generic usbhid hid i915(O-) drm_kms_helper(O) igb dca ptp pps_core i2c_algo_bit drm(O) ahci libahci video [ 161.002991] CPU: 0 PID: 2114 Comm: rmmod Tainted: G W O 3.13.0-rc8+ #2 [ 161.002997] Hardware name: NEXCOM VTC1010/Aptio CRB, BIOS 5.6.5 09/24/2013 [ 161.003004] task: dbdd6800 ti: dbe0e000 task.ti: dbe0e000 [ 161.003010] EIP: 0060:[<f84861f8>] EFLAGS: 00010246 CPU: 0 [ 161.003044] EIP is at i915_gem_object_ggtt_unpin+0x88/0x90 [i915] [ 161.003050] EAX: dfce3840 EBX: 00000000 ECX: dfafd690 EDX: dfce3874 [ 161.003056] ESI: c0086b40 EDI: df962e00 EBP: dbe0fe1c ESP: dbe0fe0c [ 161.003062] DS: 007b ES: 007b FS: 00d8 GS: 00e0 SS: 0068 [ 161.003068] CR0: 8005003b CR2: b7718000 CR3: 1bec0000 CR4: 001007f0 [ 161.003076] Stack: [ 161.003081] 00afc014 00000004 c0086b40 dfafc000 dbe0fe38 f8487e5a dfaa5400 c0086b40 [ 161.003099] dfafc000 dfaa5400 dfaa5414 dbe0fe58 f84741aa 00000000 f89c34b9 dfaa5414 [ 161.003117] dfaa5400 dfaa5400 f644b000 dbe0fe6c f89a5443 dfaa5400 f8505000 f644b000 [ 161.003134] Call Trace: [ 161.003169] [<f8487e5a>] i915_gem_context_fini+0xba/0x1c0 [i915] [ 161.003202] [<f84741aa>] i915_driver_unload+0x1fa/0x2f0 [i915] [ 161.003232] [<f89a5443>] drm_dev_unregister+0x23/0x90 [drm] [ 161.003259] [<f89a54ed>] drm_put_dev+0x3d/0x70 [drm] [ 161.003294] [<f8470615>] i915_pci_remove+0x15/0x20 [i915] [ 161.003306] [<c1338a6f>] pci_device_remove+0x2f/0xa0 [ 161.003317] [<c140c871>] __device_release_driver+0x61/0xc0 [ 161.003328] [<c140d12f>] driver_detach+0x8f/0xa0 [ 161.003341] [<c140c54f>] bus_remove_driver+0x4f/0xc0 [ 161.003353] [<c140d708>] driver_unregister+0x28/0x60 [ 161.003362] [<c10cee42>] ? stop_cpus+0x32/0x40 [ 161.003372] [<c10bd510>] ? module_refcount+0x90/0x90 [ 161.003383] [<c13378c5>] pci_unregister_driver+0x15/0x60 [ 161.003413] [<f89a739f>] drm_pci_exit+0x9f/0xb0 [drm] [ 161.003458] [<f84e624a>] i915_exit+0x1b/0x1d [i915] [ 161.003468] [<c10bf8a8>] SyS_delete_module+0x158/0x1f0 [ 161.003480] [<c1173d5d>] ? ____fput+0xd/0x10 [ 161.003488] [<c106f0fe>] ? task_work_run+0x7e/0xb0 [ 161.003499] [<c165a68d>] sysenter_do_call+0x12/0x28 [ 161.003505] Code: 0f b6 4d f3 8d 51 0f 83 e1 f0 83 e2 0f 09 d1 84 d2 88 48 54 75 07 80 a7 91 00 00 00 7f 83 c4 04 5b 5e 5f 5d c3 8d b6 00 00 00 00 <0f> 0b 8d b6 00 00 00 00 55 89 e5 57 56 53 83 ec 64 3e 8d 74 26 [ 161.003586] EIP: [<f84861f8>] i915_gem_object_ggtt_unpin+0x88/0x90 [i915] SS:ESP 0068:dbe0fe0c v2: Rename the local variable (is_default_ctx) to avoid confusion with the function is_default_ctx(). And correct Jesse's email address. Reported-by: Jesse Barnes <jbarnes@virtuousgeek.org> Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=73985 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Jesse Barnes <jbarnes@virtuousgeek.org> Cc: Ben Widawsky <benjamin.widawsky@intel.com> Tested-by: Jesse Barnes <jbarnes@virtuousgeek.org> Reviewed-by: Ben Widawsky <benjamin.widawsky@intel.com> [danvet: Fix up the rebase fail from my first attempt, thankfully pointed out by Ville.] Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-01-23 19:40:02 +00:00
err_unpin:
drm/i915: Emphasize that ctx->obj & ctx->is_initialized refer to the legacy rcs ctx We have already advanced that Logical Ring Contexts have their own kind of backing objects, but everything will be better explained in the Execlists series. For now, suffice it to say that the current backing object is only ever used with the render ring, so we're making this fact more explicit (which is a good reason on its own). As for the is_initialized flag, we only use to signify that the render state has been initialized (a.k.a. golden context, a.k.a. null context). It doesn't mean anything for the other engines, so make that distinction obvious. Done with the following Coccinelle patch (plus manual frobbing of the struct): @@ struct intel_context c; @@ - (c).obj + c.legacy_hw_ctx.rcs_state @@ struct intel_context *c; @@ - (c)->obj + c->legacy_hw_ctx.rcs_state @@ struct intel_context c; @@ - (c).is_initialized + c.legacy_hw_ctx.initialized @@ struct intel_context *c; @@ - (c)->is_initialized + c->legacy_hw_ctx.initialized This Execlists prep-work patch has been suggested by Chris Wilson and Daniel Vetter separately. Initially, it was two separate patches: drm/i915: Rename ctx->obj to ctx->rcs_state drm/i915: Make it obvious that ctx->id is merely a user handle Signed-off-by: Oscar Mateo <oscar.mateo@intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> [danvet: s/id/is_initialized/ to fix the subject and resolve a conflict in i915_gem_context_reset. Also introduce a new lctx local variable to avoid overtly long lines.] Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-03 15:27:59 +00:00
if (is_global_default_ctx && ctx->legacy_hw_ctx.rcs_state)
i915_gem_object_ggtt_unpin(ctx->legacy_hw_ctx.rcs_state);
err_destroy:
i915_gem_context_unreference(ctx);
return ERR_PTR(ret);
drm/i915: preliminary context support Very basic code for context setup/destruction in the driver. Adds the file i915_gem_context.c This file implements HW context support. On gen5+ a HW context consists of an opaque GPU object which is referenced at times of context saves and restores. With RC6 enabled, the context is also referenced as the GPU enters and exists from RC6 (GPU has it's own internal power context, except on gen5). Though something like a context does exist for the media ring, the code only supports contexts for the render ring. In software, there is a distinction between contexts created by the user, and the default HW context. The default HW context is used by GPU clients that do not request setup of their own hardware context. The default context's state is never restored to help prevent programming errors. This would happen if a client ran and piggy-backed off another clients GPU state. The default context only exists to give the GPU some offset to load as the current to invoke a save of the context we actually care about. In fact, the code could likely be constructed, albeit in a more complicated fashion, to never use the default context, though that limits the driver's ability to swap out, and/or destroy other contexts. All other contexts are created as a request by the GPU client. These contexts store GPU state, and thus allow GPU clients to not re-emit state (and potentially query certain state) at any time. The kernel driver makes certain that the appropriate commands are inserted. There are 4 entry points into the contexts, init, fini, open, close. The names are self-explanatory except that init can be called during reset, and also during pm thaw/resume. As we expect our context to be preserved across these events, we do not reinitialize in this case. As Adam Jackson pointed out, The cutoff of 1MB where a HW context is considered too big is arbitrary. The reason for this is even though context sizes are increasing with every generation, they have yet to eclipse even 32k. If we somehow read back way more than that, it probably means BIOS has done something strange, or we're running on a platform that wasn't designed for this. v2: rename load/unload to init/fini (daniel) remove ILK support for get_size() (indirectly daniel) add HAS_HW_CONTEXTS macro to clarify supported platforms (daniel) added comments (Ben) Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
2012-06-04 21:42:42 +00:00
}
void i915_gem_context_reset(struct drm_device *dev)
{
struct drm_i915_private *dev_priv = dev->dev_private;
int i;
if (i915.enable_execlists) {
struct intel_context *ctx;
list_for_each_entry(ctx, &dev_priv->context_list, link) {
intel_lr_context_reset(dev, ctx);
}
return;
}
for (i = 0; i < I915_NUM_RINGS; i++) {
struct intel_engine_cs *ring = &dev_priv->ring[i];
drm/i915: Emphasize that ctx->obj & ctx->is_initialized refer to the legacy rcs ctx We have already advanced that Logical Ring Contexts have their own kind of backing objects, but everything will be better explained in the Execlists series. For now, suffice it to say that the current backing object is only ever used with the render ring, so we're making this fact more explicit (which is a good reason on its own). As for the is_initialized flag, we only use to signify that the render state has been initialized (a.k.a. golden context, a.k.a. null context). It doesn't mean anything for the other engines, so make that distinction obvious. Done with the following Coccinelle patch (plus manual frobbing of the struct): @@ struct intel_context c; @@ - (c).obj + c.legacy_hw_ctx.rcs_state @@ struct intel_context *c; @@ - (c)->obj + c->legacy_hw_ctx.rcs_state @@ struct intel_context c; @@ - (c).is_initialized + c.legacy_hw_ctx.initialized @@ struct intel_context *c; @@ - (c)->is_initialized + c->legacy_hw_ctx.initialized This Execlists prep-work patch has been suggested by Chris Wilson and Daniel Vetter separately. Initially, it was two separate patches: drm/i915: Rename ctx->obj to ctx->rcs_state drm/i915: Make it obvious that ctx->id is merely a user handle Signed-off-by: Oscar Mateo <oscar.mateo@intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> [danvet: s/id/is_initialized/ to fix the subject and resolve a conflict in i915_gem_context_reset. Also introduce a new lctx local variable to avoid overtly long lines.] Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-03 15:27:59 +00:00
struct intel_context *lctx = ring->last_context;
if (lctx) {
if (lctx->legacy_hw_ctx.rcs_state && i == RCS)
i915_gem_object_ggtt_unpin(lctx->legacy_hw_ctx.rcs_state);
i915_gem_context_unreference(lctx);
ring->last_context = NULL;
}
}
}
int i915_gem_context_init(struct drm_device *dev)
drm/i915: preliminary context support Very basic code for context setup/destruction in the driver. Adds the file i915_gem_context.c This file implements HW context support. On gen5+ a HW context consists of an opaque GPU object which is referenced at times of context saves and restores. With RC6 enabled, the context is also referenced as the GPU enters and exists from RC6 (GPU has it's own internal power context, except on gen5). Though something like a context does exist for the media ring, the code only supports contexts for the render ring. In software, there is a distinction between contexts created by the user, and the default HW context. The default HW context is used by GPU clients that do not request setup of their own hardware context. The default context's state is never restored to help prevent programming errors. This would happen if a client ran and piggy-backed off another clients GPU state. The default context only exists to give the GPU some offset to load as the current to invoke a save of the context we actually care about. In fact, the code could likely be constructed, albeit in a more complicated fashion, to never use the default context, though that limits the driver's ability to swap out, and/or destroy other contexts. All other contexts are created as a request by the GPU client. These contexts store GPU state, and thus allow GPU clients to not re-emit state (and potentially query certain state) at any time. The kernel driver makes certain that the appropriate commands are inserted. There are 4 entry points into the contexts, init, fini, open, close. The names are self-explanatory except that init can be called during reset, and also during pm thaw/resume. As we expect our context to be preserved across these events, we do not reinitialize in this case. As Adam Jackson pointed out, The cutoff of 1MB where a HW context is considered too big is arbitrary. The reason for this is even though context sizes are increasing with every generation, they have yet to eclipse even 32k. If we somehow read back way more than that, it probably means BIOS has done something strange, or we're running on a platform that wasn't designed for this. v2: rename load/unload to init/fini (daniel) remove ILK support for get_size() (indirectly daniel) add HAS_HW_CONTEXTS macro to clarify supported platforms (daniel) added comments (Ben) Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
2012-06-04 21:42:42 +00:00
{
struct drm_i915_private *dev_priv = dev->dev_private;
struct intel_context *ctx;
int i;
drm/i915: preliminary context support Very basic code for context setup/destruction in the driver. Adds the file i915_gem_context.c This file implements HW context support. On gen5+ a HW context consists of an opaque GPU object which is referenced at times of context saves and restores. With RC6 enabled, the context is also referenced as the GPU enters and exists from RC6 (GPU has it's own internal power context, except on gen5). Though something like a context does exist for the media ring, the code only supports contexts for the render ring. In software, there is a distinction between contexts created by the user, and the default HW context. The default HW context is used by GPU clients that do not request setup of their own hardware context. The default context's state is never restored to help prevent programming errors. This would happen if a client ran and piggy-backed off another clients GPU state. The default context only exists to give the GPU some offset to load as the current to invoke a save of the context we actually care about. In fact, the code could likely be constructed, albeit in a more complicated fashion, to never use the default context, though that limits the driver's ability to swap out, and/or destroy other contexts. All other contexts are created as a request by the GPU client. These contexts store GPU state, and thus allow GPU clients to not re-emit state (and potentially query certain state) at any time. The kernel driver makes certain that the appropriate commands are inserted. There are 4 entry points into the contexts, init, fini, open, close. The names are self-explanatory except that init can be called during reset, and also during pm thaw/resume. As we expect our context to be preserved across these events, we do not reinitialize in this case. As Adam Jackson pointed out, The cutoff of 1MB where a HW context is considered too big is arbitrary. The reason for this is even though context sizes are increasing with every generation, they have yet to eclipse even 32k. If we somehow read back way more than that, it probably means BIOS has done something strange, or we're running on a platform that wasn't designed for this. v2: rename load/unload to init/fini (daniel) remove ILK support for get_size() (indirectly daniel) add HAS_HW_CONTEXTS macro to clarify supported platforms (daniel) added comments (Ben) Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
2012-06-04 21:42:42 +00:00
/* Init should only be called once per module load. Eventually the
* restriction on the context_disabled check can be loosened. */
if (WARN_ON(dev_priv->ring[RCS].default_context))
return 0;
drm/i915: preliminary context support Very basic code for context setup/destruction in the driver. Adds the file i915_gem_context.c This file implements HW context support. On gen5+ a HW context consists of an opaque GPU object which is referenced at times of context saves and restores. With RC6 enabled, the context is also referenced as the GPU enters and exists from RC6 (GPU has it's own internal power context, except on gen5). Though something like a context does exist for the media ring, the code only supports contexts for the render ring. In software, there is a distinction between contexts created by the user, and the default HW context. The default HW context is used by GPU clients that do not request setup of their own hardware context. The default context's state is never restored to help prevent programming errors. This would happen if a client ran and piggy-backed off another clients GPU state. The default context only exists to give the GPU some offset to load as the current to invoke a save of the context we actually care about. In fact, the code could likely be constructed, albeit in a more complicated fashion, to never use the default context, though that limits the driver's ability to swap out, and/or destroy other contexts. All other contexts are created as a request by the GPU client. These contexts store GPU state, and thus allow GPU clients to not re-emit state (and potentially query certain state) at any time. The kernel driver makes certain that the appropriate commands are inserted. There are 4 entry points into the contexts, init, fini, open, close. The names are self-explanatory except that init can be called during reset, and also during pm thaw/resume. As we expect our context to be preserved across these events, we do not reinitialize in this case. As Adam Jackson pointed out, The cutoff of 1MB where a HW context is considered too big is arbitrary. The reason for this is even though context sizes are increasing with every generation, they have yet to eclipse even 32k. If we somehow read back way more than that, it probably means BIOS has done something strange, or we're running on a platform that wasn't designed for this. v2: rename load/unload to init/fini (daniel) remove ILK support for get_size() (indirectly daniel) add HAS_HW_CONTEXTS macro to clarify supported platforms (daniel) added comments (Ben) Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
2012-06-04 21:42:42 +00:00
if (i915.enable_execlists) {
/* NB: intentionally left blank. We will allocate our own
* backing objects as we need them, thank you very much */
dev_priv->hw_context_size = 0;
} else if (HAS_HW_CONTEXTS(dev)) {
dev_priv->hw_context_size = round_up(get_context_size(dev), 4096);
if (dev_priv->hw_context_size > (1<<20)) {
DRM_DEBUG_DRIVER("Disabling HW Contexts; invalid size %d\n",
dev_priv->hw_context_size);
dev_priv->hw_context_size = 0;
}
drm/i915: preliminary context support Very basic code for context setup/destruction in the driver. Adds the file i915_gem_context.c This file implements HW context support. On gen5+ a HW context consists of an opaque GPU object which is referenced at times of context saves and restores. With RC6 enabled, the context is also referenced as the GPU enters and exists from RC6 (GPU has it's own internal power context, except on gen5). Though something like a context does exist for the media ring, the code only supports contexts for the render ring. In software, there is a distinction between contexts created by the user, and the default HW context. The default HW context is used by GPU clients that do not request setup of their own hardware context. The default context's state is never restored to help prevent programming errors. This would happen if a client ran and piggy-backed off another clients GPU state. The default context only exists to give the GPU some offset to load as the current to invoke a save of the context we actually care about. In fact, the code could likely be constructed, albeit in a more complicated fashion, to never use the default context, though that limits the driver's ability to swap out, and/or destroy other contexts. All other contexts are created as a request by the GPU client. These contexts store GPU state, and thus allow GPU clients to not re-emit state (and potentially query certain state) at any time. The kernel driver makes certain that the appropriate commands are inserted. There are 4 entry points into the contexts, init, fini, open, close. The names are self-explanatory except that init can be called during reset, and also during pm thaw/resume. As we expect our context to be preserved across these events, we do not reinitialize in this case. As Adam Jackson pointed out, The cutoff of 1MB where a HW context is considered too big is arbitrary. The reason for this is even though context sizes are increasing with every generation, they have yet to eclipse even 32k. If we somehow read back way more than that, it probably means BIOS has done something strange, or we're running on a platform that wasn't designed for this. v2: rename load/unload to init/fini (daniel) remove ILK support for get_size() (indirectly daniel) add HAS_HW_CONTEXTS macro to clarify supported platforms (daniel) added comments (Ben) Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
2012-06-04 21:42:42 +00:00
}
ctx = i915_gem_create_context(dev, NULL);
if (IS_ERR(ctx)) {
DRM_ERROR("Failed to create default global context (error %ld)\n",
PTR_ERR(ctx));
return PTR_ERR(ctx);
drm/i915: preliminary context support Very basic code for context setup/destruction in the driver. Adds the file i915_gem_context.c This file implements HW context support. On gen5+ a HW context consists of an opaque GPU object which is referenced at times of context saves and restores. With RC6 enabled, the context is also referenced as the GPU enters and exists from RC6 (GPU has it's own internal power context, except on gen5). Though something like a context does exist for the media ring, the code only supports contexts for the render ring. In software, there is a distinction between contexts created by the user, and the default HW context. The default HW context is used by GPU clients that do not request setup of their own hardware context. The default context's state is never restored to help prevent programming errors. This would happen if a client ran and piggy-backed off another clients GPU state. The default context only exists to give the GPU some offset to load as the current to invoke a save of the context we actually care about. In fact, the code could likely be constructed, albeit in a more complicated fashion, to never use the default context, though that limits the driver's ability to swap out, and/or destroy other contexts. All other contexts are created as a request by the GPU client. These contexts store GPU state, and thus allow GPU clients to not re-emit state (and potentially query certain state) at any time. The kernel driver makes certain that the appropriate commands are inserted. There are 4 entry points into the contexts, init, fini, open, close. The names are self-explanatory except that init can be called during reset, and also during pm thaw/resume. As we expect our context to be preserved across these events, we do not reinitialize in this case. As Adam Jackson pointed out, The cutoff of 1MB where a HW context is considered too big is arbitrary. The reason for this is even though context sizes are increasing with every generation, they have yet to eclipse even 32k. If we somehow read back way more than that, it probably means BIOS has done something strange, or we're running on a platform that wasn't designed for this. v2: rename load/unload to init/fini (daniel) remove ILK support for get_size() (indirectly daniel) add HAS_HW_CONTEXTS macro to clarify supported platforms (daniel) added comments (Ben) Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
2012-06-04 21:42:42 +00:00
}
for (i = 0; i < I915_NUM_RINGS; i++) {
struct intel_engine_cs *ring = &dev_priv->ring[i];
/* NB: RCS will hold a ref for all rings */
ring->default_context = ctx;
}
DRM_DEBUG_DRIVER("%s context support initialized\n",
i915.enable_execlists ? "LR" :
dev_priv->hw_context_size ? "HW" : "fake");
return 0;
drm/i915: preliminary context support Very basic code for context setup/destruction in the driver. Adds the file i915_gem_context.c This file implements HW context support. On gen5+ a HW context consists of an opaque GPU object which is referenced at times of context saves and restores. With RC6 enabled, the context is also referenced as the GPU enters and exists from RC6 (GPU has it's own internal power context, except on gen5). Though something like a context does exist for the media ring, the code only supports contexts for the render ring. In software, there is a distinction between contexts created by the user, and the default HW context. The default HW context is used by GPU clients that do not request setup of their own hardware context. The default context's state is never restored to help prevent programming errors. This would happen if a client ran and piggy-backed off another clients GPU state. The default context only exists to give the GPU some offset to load as the current to invoke a save of the context we actually care about. In fact, the code could likely be constructed, albeit in a more complicated fashion, to never use the default context, though that limits the driver's ability to swap out, and/or destroy other contexts. All other contexts are created as a request by the GPU client. These contexts store GPU state, and thus allow GPU clients to not re-emit state (and potentially query certain state) at any time. The kernel driver makes certain that the appropriate commands are inserted. There are 4 entry points into the contexts, init, fini, open, close. The names are self-explanatory except that init can be called during reset, and also during pm thaw/resume. As we expect our context to be preserved across these events, we do not reinitialize in this case. As Adam Jackson pointed out, The cutoff of 1MB where a HW context is considered too big is arbitrary. The reason for this is even though context sizes are increasing with every generation, they have yet to eclipse even 32k. If we somehow read back way more than that, it probably means BIOS has done something strange, or we're running on a platform that wasn't designed for this. v2: rename load/unload to init/fini (daniel) remove ILK support for get_size() (indirectly daniel) add HAS_HW_CONTEXTS macro to clarify supported platforms (daniel) added comments (Ben) Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
2012-06-04 21:42:42 +00:00
}
void i915_gem_context_fini(struct drm_device *dev)
{
struct drm_i915_private *dev_priv = dev->dev_private;
struct intel_context *dctx = dev_priv->ring[RCS].default_context;
int i;
drm/i915: preliminary context support Very basic code for context setup/destruction in the driver. Adds the file i915_gem_context.c This file implements HW context support. On gen5+ a HW context consists of an opaque GPU object which is referenced at times of context saves and restores. With RC6 enabled, the context is also referenced as the GPU enters and exists from RC6 (GPU has it's own internal power context, except on gen5). Though something like a context does exist for the media ring, the code only supports contexts for the render ring. In software, there is a distinction between contexts created by the user, and the default HW context. The default HW context is used by GPU clients that do not request setup of their own hardware context. The default context's state is never restored to help prevent programming errors. This would happen if a client ran and piggy-backed off another clients GPU state. The default context only exists to give the GPU some offset to load as the current to invoke a save of the context we actually care about. In fact, the code could likely be constructed, albeit in a more complicated fashion, to never use the default context, though that limits the driver's ability to swap out, and/or destroy other contexts. All other contexts are created as a request by the GPU client. These contexts store GPU state, and thus allow GPU clients to not re-emit state (and potentially query certain state) at any time. The kernel driver makes certain that the appropriate commands are inserted. There are 4 entry points into the contexts, init, fini, open, close. The names are self-explanatory except that init can be called during reset, and also during pm thaw/resume. As we expect our context to be preserved across these events, we do not reinitialize in this case. As Adam Jackson pointed out, The cutoff of 1MB where a HW context is considered too big is arbitrary. The reason for this is even though context sizes are increasing with every generation, they have yet to eclipse even 32k. If we somehow read back way more than that, it probably means BIOS has done something strange, or we're running on a platform that wasn't designed for this. v2: rename load/unload to init/fini (daniel) remove ILK support for get_size() (indirectly daniel) add HAS_HW_CONTEXTS macro to clarify supported platforms (daniel) added comments (Ben) Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
2012-06-04 21:42:42 +00:00
drm/i915: Emphasize that ctx->obj & ctx->is_initialized refer to the legacy rcs ctx We have already advanced that Logical Ring Contexts have their own kind of backing objects, but everything will be better explained in the Execlists series. For now, suffice it to say that the current backing object is only ever used with the render ring, so we're making this fact more explicit (which is a good reason on its own). As for the is_initialized flag, we only use to signify that the render state has been initialized (a.k.a. golden context, a.k.a. null context). It doesn't mean anything for the other engines, so make that distinction obvious. Done with the following Coccinelle patch (plus manual frobbing of the struct): @@ struct intel_context c; @@ - (c).obj + c.legacy_hw_ctx.rcs_state @@ struct intel_context *c; @@ - (c)->obj + c->legacy_hw_ctx.rcs_state @@ struct intel_context c; @@ - (c).is_initialized + c.legacy_hw_ctx.initialized @@ struct intel_context *c; @@ - (c)->is_initialized + c->legacy_hw_ctx.initialized This Execlists prep-work patch has been suggested by Chris Wilson and Daniel Vetter separately. Initially, it was two separate patches: drm/i915: Rename ctx->obj to ctx->rcs_state drm/i915: Make it obvious that ctx->id is merely a user handle Signed-off-by: Oscar Mateo <oscar.mateo@intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> [danvet: s/id/is_initialized/ to fix the subject and resolve a conflict in i915_gem_context_reset. Also introduce a new lctx local variable to avoid overtly long lines.] Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-03 15:27:59 +00:00
if (dctx->legacy_hw_ctx.rcs_state) {
/* The only known way to stop the gpu from accessing the hw context is
* to reset it. Do this as the very last operation to avoid confusing
* other code, leading to spurious errors. */
intel_gpu_reset(dev);
/* When default context is created and switched to, base object refcount
* will be 2 (+1 from object creation and +1 from do_switch()).
* i915_gem_context_fini() will be called after gpu_idle() has switched
* to default context. So we need to unreference the base object once
* to offset the do_switch part, so that i915_gem_context_unreference()
* can then free the base object correctly. */
WARN_ON(!dev_priv->ring[RCS].last_context);
if (dev_priv->ring[RCS].last_context == dctx) {
/* Fake switch to NULL context */
drm/i915: Emphasize that ctx->obj & ctx->is_initialized refer to the legacy rcs ctx We have already advanced that Logical Ring Contexts have their own kind of backing objects, but everything will be better explained in the Execlists series. For now, suffice it to say that the current backing object is only ever used with the render ring, so we're making this fact more explicit (which is a good reason on its own). As for the is_initialized flag, we only use to signify that the render state has been initialized (a.k.a. golden context, a.k.a. null context). It doesn't mean anything for the other engines, so make that distinction obvious. Done with the following Coccinelle patch (plus manual frobbing of the struct): @@ struct intel_context c; @@ - (c).obj + c.legacy_hw_ctx.rcs_state @@ struct intel_context *c; @@ - (c)->obj + c->legacy_hw_ctx.rcs_state @@ struct intel_context c; @@ - (c).is_initialized + c.legacy_hw_ctx.initialized @@ struct intel_context *c; @@ - (c)->is_initialized + c->legacy_hw_ctx.initialized This Execlists prep-work patch has been suggested by Chris Wilson and Daniel Vetter separately. Initially, it was two separate patches: drm/i915: Rename ctx->obj to ctx->rcs_state drm/i915: Make it obvious that ctx->id is merely a user handle Signed-off-by: Oscar Mateo <oscar.mateo@intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> [danvet: s/id/is_initialized/ to fix the subject and resolve a conflict in i915_gem_context_reset. Also introduce a new lctx local variable to avoid overtly long lines.] Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-03 15:27:59 +00:00
WARN_ON(dctx->legacy_hw_ctx.rcs_state->active);
i915_gem_object_ggtt_unpin(dctx->legacy_hw_ctx.rcs_state);
i915_gem_context_unreference(dctx);
dev_priv->ring[RCS].last_context = NULL;
}
drm/i915: Emphasize that ctx->obj & ctx->is_initialized refer to the legacy rcs ctx We have already advanced that Logical Ring Contexts have their own kind of backing objects, but everything will be better explained in the Execlists series. For now, suffice it to say that the current backing object is only ever used with the render ring, so we're making this fact more explicit (which is a good reason on its own). As for the is_initialized flag, we only use to signify that the render state has been initialized (a.k.a. golden context, a.k.a. null context). It doesn't mean anything for the other engines, so make that distinction obvious. Done with the following Coccinelle patch (plus manual frobbing of the struct): @@ struct intel_context c; @@ - (c).obj + c.legacy_hw_ctx.rcs_state @@ struct intel_context *c; @@ - (c)->obj + c->legacy_hw_ctx.rcs_state @@ struct intel_context c; @@ - (c).is_initialized + c.legacy_hw_ctx.initialized @@ struct intel_context *c; @@ - (c)->is_initialized + c->legacy_hw_ctx.initialized This Execlists prep-work patch has been suggested by Chris Wilson and Daniel Vetter separately. Initially, it was two separate patches: drm/i915: Rename ctx->obj to ctx->rcs_state drm/i915: Make it obvious that ctx->id is merely a user handle Signed-off-by: Oscar Mateo <oscar.mateo@intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> [danvet: s/id/is_initialized/ to fix the subject and resolve a conflict in i915_gem_context_reset. Also introduce a new lctx local variable to avoid overtly long lines.] Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-03 15:27:59 +00:00
i915_gem_object_ggtt_unpin(dctx->legacy_hw_ctx.rcs_state);
}
for (i = 0; i < I915_NUM_RINGS; i++) {
struct intel_engine_cs *ring = &dev_priv->ring[i];
if (ring->last_context)
i915_gem_context_unreference(ring->last_context);
ring->default_context = NULL;
ring->last_context = NULL;
}
i915_gem_context_unreference(dctx);
drm/i915: preliminary context support Very basic code for context setup/destruction in the driver. Adds the file i915_gem_context.c This file implements HW context support. On gen5+ a HW context consists of an opaque GPU object which is referenced at times of context saves and restores. With RC6 enabled, the context is also referenced as the GPU enters and exists from RC6 (GPU has it's own internal power context, except on gen5). Though something like a context does exist for the media ring, the code only supports contexts for the render ring. In software, there is a distinction between contexts created by the user, and the default HW context. The default HW context is used by GPU clients that do not request setup of their own hardware context. The default context's state is never restored to help prevent programming errors. This would happen if a client ran and piggy-backed off another clients GPU state. The default context only exists to give the GPU some offset to load as the current to invoke a save of the context we actually care about. In fact, the code could likely be constructed, albeit in a more complicated fashion, to never use the default context, though that limits the driver's ability to swap out, and/or destroy other contexts. All other contexts are created as a request by the GPU client. These contexts store GPU state, and thus allow GPU clients to not re-emit state (and potentially query certain state) at any time. The kernel driver makes certain that the appropriate commands are inserted. There are 4 entry points into the contexts, init, fini, open, close. The names are self-explanatory except that init can be called during reset, and also during pm thaw/resume. As we expect our context to be preserved across these events, we do not reinitialize in this case. As Adam Jackson pointed out, The cutoff of 1MB where a HW context is considered too big is arbitrary. The reason for this is even though context sizes are increasing with every generation, they have yet to eclipse even 32k. If we somehow read back way more than that, it probably means BIOS has done something strange, or we're running on a platform that wasn't designed for this. v2: rename load/unload to init/fini (daniel) remove ILK support for get_size() (indirectly daniel) add HAS_HW_CONTEXTS macro to clarify supported platforms (daniel) added comments (Ben) Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
2012-06-04 21:42:42 +00:00
}
int i915_gem_context_enable(struct intel_engine_cs *ring)
{
int ret;
if (i915.enable_execlists) {
if (ring->init_context == NULL)
return 0;
ret = ring->init_context(ring, ring->default_context);
} else
ret = i915_switch_context(ring, ring->default_context);
if (ret) {
DRM_ERROR("ring init context: %d\n", ret);
return ret;
}
return 0;
}
static int context_idr_cleanup(int id, void *p, void *data)
{
struct intel_context *ctx = p;
i915_gem_context_unreference(ctx);
return 0;
drm/i915: preliminary context support Very basic code for context setup/destruction in the driver. Adds the file i915_gem_context.c This file implements HW context support. On gen5+ a HW context consists of an opaque GPU object which is referenced at times of context saves and restores. With RC6 enabled, the context is also referenced as the GPU enters and exists from RC6 (GPU has it's own internal power context, except on gen5). Though something like a context does exist for the media ring, the code only supports contexts for the render ring. In software, there is a distinction between contexts created by the user, and the default HW context. The default HW context is used by GPU clients that do not request setup of their own hardware context. The default context's state is never restored to help prevent programming errors. This would happen if a client ran and piggy-backed off another clients GPU state. The default context only exists to give the GPU some offset to load as the current to invoke a save of the context we actually care about. In fact, the code could likely be constructed, albeit in a more complicated fashion, to never use the default context, though that limits the driver's ability to swap out, and/or destroy other contexts. All other contexts are created as a request by the GPU client. These contexts store GPU state, and thus allow GPU clients to not re-emit state (and potentially query certain state) at any time. The kernel driver makes certain that the appropriate commands are inserted. There are 4 entry points into the contexts, init, fini, open, close. The names are self-explanatory except that init can be called during reset, and also during pm thaw/resume. As we expect our context to be preserved across these events, we do not reinitialize in this case. As Adam Jackson pointed out, The cutoff of 1MB where a HW context is considered too big is arbitrary. The reason for this is even though context sizes are increasing with every generation, they have yet to eclipse even 32k. If we somehow read back way more than that, it probably means BIOS has done something strange, or we're running on a platform that wasn't designed for this. v2: rename load/unload to init/fini (daniel) remove ILK support for get_size() (indirectly daniel) add HAS_HW_CONTEXTS macro to clarify supported platforms (daniel) added comments (Ben) Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
2012-06-04 21:42:42 +00:00
}
int i915_gem_context_open(struct drm_device *dev, struct drm_file *file)
{
struct drm_i915_file_private *file_priv = file->driver_priv;
struct intel_context *ctx;
idr_init(&file_priv->context_idr);
mutex_lock(&dev->struct_mutex);
ctx = i915_gem_create_context(dev, file_priv);
mutex_unlock(&dev->struct_mutex);
if (IS_ERR(ctx)) {
idr_destroy(&file_priv->context_idr);
return PTR_ERR(ctx);
}
return 0;
}
drm/i915: preliminary context support Very basic code for context setup/destruction in the driver. Adds the file i915_gem_context.c This file implements HW context support. On gen5+ a HW context consists of an opaque GPU object which is referenced at times of context saves and restores. With RC6 enabled, the context is also referenced as the GPU enters and exists from RC6 (GPU has it's own internal power context, except on gen5). Though something like a context does exist for the media ring, the code only supports contexts for the render ring. In software, there is a distinction between contexts created by the user, and the default HW context. The default HW context is used by GPU clients that do not request setup of their own hardware context. The default context's state is never restored to help prevent programming errors. This would happen if a client ran and piggy-backed off another clients GPU state. The default context only exists to give the GPU some offset to load as the current to invoke a save of the context we actually care about. In fact, the code could likely be constructed, albeit in a more complicated fashion, to never use the default context, though that limits the driver's ability to swap out, and/or destroy other contexts. All other contexts are created as a request by the GPU client. These contexts store GPU state, and thus allow GPU clients to not re-emit state (and potentially query certain state) at any time. The kernel driver makes certain that the appropriate commands are inserted. There are 4 entry points into the contexts, init, fini, open, close. The names are self-explanatory except that init can be called during reset, and also during pm thaw/resume. As we expect our context to be preserved across these events, we do not reinitialize in this case. As Adam Jackson pointed out, The cutoff of 1MB where a HW context is considered too big is arbitrary. The reason for this is even though context sizes are increasing with every generation, they have yet to eclipse even 32k. If we somehow read back way more than that, it probably means BIOS has done something strange, or we're running on a platform that wasn't designed for this. v2: rename load/unload to init/fini (daniel) remove ILK support for get_size() (indirectly daniel) add HAS_HW_CONTEXTS macro to clarify supported platforms (daniel) added comments (Ben) Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
2012-06-04 21:42:42 +00:00
void i915_gem_context_close(struct drm_device *dev, struct drm_file *file)
{
struct drm_i915_file_private *file_priv = file->driver_priv;
drm/i915: preliminary context support Very basic code for context setup/destruction in the driver. Adds the file i915_gem_context.c This file implements HW context support. On gen5+ a HW context consists of an opaque GPU object which is referenced at times of context saves and restores. With RC6 enabled, the context is also referenced as the GPU enters and exists from RC6 (GPU has it's own internal power context, except on gen5). Though something like a context does exist for the media ring, the code only supports contexts for the render ring. In software, there is a distinction between contexts created by the user, and the default HW context. The default HW context is used by GPU clients that do not request setup of their own hardware context. The default context's state is never restored to help prevent programming errors. This would happen if a client ran and piggy-backed off another clients GPU state. The default context only exists to give the GPU some offset to load as the current to invoke a save of the context we actually care about. In fact, the code could likely be constructed, albeit in a more complicated fashion, to never use the default context, though that limits the driver's ability to swap out, and/or destroy other contexts. All other contexts are created as a request by the GPU client. These contexts store GPU state, and thus allow GPU clients to not re-emit state (and potentially query certain state) at any time. The kernel driver makes certain that the appropriate commands are inserted. There are 4 entry points into the contexts, init, fini, open, close. The names are self-explanatory except that init can be called during reset, and also during pm thaw/resume. As we expect our context to be preserved across these events, we do not reinitialize in this case. As Adam Jackson pointed out, The cutoff of 1MB where a HW context is considered too big is arbitrary. The reason for this is even though context sizes are increasing with every generation, they have yet to eclipse even 32k. If we somehow read back way more than that, it probably means BIOS has done something strange, or we're running on a platform that wasn't designed for this. v2: rename load/unload to init/fini (daniel) remove ILK support for get_size() (indirectly daniel) add HAS_HW_CONTEXTS macro to clarify supported platforms (daniel) added comments (Ben) Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
2012-06-04 21:42:42 +00:00
idr_for_each(&file_priv->context_idr, context_idr_cleanup, NULL);
idr_destroy(&file_priv->context_idr);
}
struct intel_context *
i915_gem_context_get(struct drm_i915_file_private *file_priv, u32 id)
{
struct intel_context *ctx;
ctx = (struct intel_context *)idr_find(&file_priv->context_idr, id);
if (!ctx)
return ERR_PTR(-ENOENT);
return ctx;
drm/i915: preliminary context support Very basic code for context setup/destruction in the driver. Adds the file i915_gem_context.c This file implements HW context support. On gen5+ a HW context consists of an opaque GPU object which is referenced at times of context saves and restores. With RC6 enabled, the context is also referenced as the GPU enters and exists from RC6 (GPU has it's own internal power context, except on gen5). Though something like a context does exist for the media ring, the code only supports contexts for the render ring. In software, there is a distinction between contexts created by the user, and the default HW context. The default HW context is used by GPU clients that do not request setup of their own hardware context. The default context's state is never restored to help prevent programming errors. This would happen if a client ran and piggy-backed off another clients GPU state. The default context only exists to give the GPU some offset to load as the current to invoke a save of the context we actually care about. In fact, the code could likely be constructed, albeit in a more complicated fashion, to never use the default context, though that limits the driver's ability to swap out, and/or destroy other contexts. All other contexts are created as a request by the GPU client. These contexts store GPU state, and thus allow GPU clients to not re-emit state (and potentially query certain state) at any time. The kernel driver makes certain that the appropriate commands are inserted. There are 4 entry points into the contexts, init, fini, open, close. The names are self-explanatory except that init can be called during reset, and also during pm thaw/resume. As we expect our context to be preserved across these events, we do not reinitialize in this case. As Adam Jackson pointed out, The cutoff of 1MB where a HW context is considered too big is arbitrary. The reason for this is even though context sizes are increasing with every generation, they have yet to eclipse even 32k. If we somehow read back way more than that, it probably means BIOS has done something strange, or we're running on a platform that wasn't designed for this. v2: rename load/unload to init/fini (daniel) remove ILK support for get_size() (indirectly daniel) add HAS_HW_CONTEXTS macro to clarify supported platforms (daniel) added comments (Ben) Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
2012-06-04 21:42:42 +00:00
}
static inline int
mi_set_context(struct intel_engine_cs *ring,
struct intel_context *new_context,
u32 hw_flags)
{
u32 flags = hw_flags | MI_MM_SPACE_GTT;
drm/i915: Disable PSMI sleep messages on all rings around context switches There exists a current workaround to prevent a hang on context switch should the ring go to sleep in the middle of the restore, WaProgramMiArbOnOffAroundMiSetContext (applicable to all gen7+). In spite of disabling arbitration (which prevents the ring from powering down during the critical section) we were still hitting hangs that had the hallmarks of the known erratum. That is we are still seeing hangs "on the last instruction in the context restore". By comparing -nightly (broken) with requests (working), we were able to deduce that it was the semaphore LRI cross-talk that reproduced the original failure. The key was that requests implemented deferred semaphore signalling, and disabling that, i.e. emitting the semaphore signal to every other ring after every batch restored the frequent hang. Explicitly disabling PSMI sleep on the RCS ring was insufficient, all the rings had to be awake to prevent the hangs. Fortunately, we can reduce the wakelock to the MI_SET_CONTEXT operation itself, and so should be able to limit the extra power implications. Since the MI_ARB_ON_OFF workaround is listed for all gen7 and above products, we should apply this extra hammer for all of the same platforms despite so far that we have only been able to reproduce the hang on certain ivb and hsw models. The last question is whether we want to always use the extra hammer or only when we know semaphores are in operation. At the moment, we only use LRI on non-RCS rings for semaphores, but that may change in the future with the possibility of reintroducing this bug under subtle conditions. v2: Make it explicit that the PSMI LRI are an extension to the original workaround for the other rings. v3: Bikeshedding variable names and whitespacing Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=80660 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=83677 Cc: Simon Farnsworth <simon@farnz.org.uk> Cc: Daniel Vetter <daniel@ffwll.ch> Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Tested-by: Peter Frühberger <fritsch@xbmc.org> Reviewed-by: Daniel Vetter <daniel@ffwll.ch> Cc: stable@vger.kernel.org Signed-off-by: Jani Nikula <jani.nikula@intel.com>
2014-12-16 10:02:27 +00:00
const int num_rings =
/* Use an extended w/a on ivb+ if signalling from other rings */
i915_semaphore_is_enabled(ring->dev) ?
hweight32(INTEL_INFO(ring->dev)->ring_mask) - 1 :
0;
int len, i, ret;
/* w/a: If Flush TLB Invalidation Mode is enabled, driver must do a TLB
* invalidation prior to MI_SET_CONTEXT. On GEN6 we don't set the value
* explicitly, so we rely on the value at ring init, stored in
* itlb_before_ctx_switch.
*/
if (IS_GEN6(ring->dev)) {
ret = ring->flush(ring, I915_GEM_GPU_DOMAINS, 0);
if (ret)
return ret;
}
/* These flags are for resource streamer on HSW+ */
if (!IS_HASWELL(ring->dev) && INTEL_INFO(ring->dev)->gen < 8)
flags |= (MI_SAVE_EXT_STATE_EN | MI_RESTORE_EXT_STATE_EN);
drm/i915: Disable PSMI sleep messages on all rings around context switches There exists a current workaround to prevent a hang on context switch should the ring go to sleep in the middle of the restore, WaProgramMiArbOnOffAroundMiSetContext (applicable to all gen7+). In spite of disabling arbitration (which prevents the ring from powering down during the critical section) we were still hitting hangs that had the hallmarks of the known erratum. That is we are still seeing hangs "on the last instruction in the context restore". By comparing -nightly (broken) with requests (working), we were able to deduce that it was the semaphore LRI cross-talk that reproduced the original failure. The key was that requests implemented deferred semaphore signalling, and disabling that, i.e. emitting the semaphore signal to every other ring after every batch restored the frequent hang. Explicitly disabling PSMI sleep on the RCS ring was insufficient, all the rings had to be awake to prevent the hangs. Fortunately, we can reduce the wakelock to the MI_SET_CONTEXT operation itself, and so should be able to limit the extra power implications. Since the MI_ARB_ON_OFF workaround is listed for all gen7 and above products, we should apply this extra hammer for all of the same platforms despite so far that we have only been able to reproduce the hang on certain ivb and hsw models. The last question is whether we want to always use the extra hammer or only when we know semaphores are in operation. At the moment, we only use LRI on non-RCS rings for semaphores, but that may change in the future with the possibility of reintroducing this bug under subtle conditions. v2: Make it explicit that the PSMI LRI are an extension to the original workaround for the other rings. v3: Bikeshedding variable names and whitespacing Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=80660 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=83677 Cc: Simon Farnsworth <simon@farnz.org.uk> Cc: Daniel Vetter <daniel@ffwll.ch> Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Tested-by: Peter Frühberger <fritsch@xbmc.org> Reviewed-by: Daniel Vetter <daniel@ffwll.ch> Cc: stable@vger.kernel.org Signed-off-by: Jani Nikula <jani.nikula@intel.com>
2014-12-16 10:02:27 +00:00
len = 4;
if (INTEL_INFO(ring->dev)->gen >= 7)
len += 2 + (num_rings ? 4*num_rings + 2 : 0);
ret = intel_ring_begin(ring, len);
if (ret)
return ret;
/* WaProgramMiArbOnOffAroundMiSetContext:ivb,vlv,hsw,bdw,chv */
drm/i915: Disable PSMI sleep messages on all rings around context switches There exists a current workaround to prevent a hang on context switch should the ring go to sleep in the middle of the restore, WaProgramMiArbOnOffAroundMiSetContext (applicable to all gen7+). In spite of disabling arbitration (which prevents the ring from powering down during the critical section) we were still hitting hangs that had the hallmarks of the known erratum. That is we are still seeing hangs "on the last instruction in the context restore". By comparing -nightly (broken) with requests (working), we were able to deduce that it was the semaphore LRI cross-talk that reproduced the original failure. The key was that requests implemented deferred semaphore signalling, and disabling that, i.e. emitting the semaphore signal to every other ring after every batch restored the frequent hang. Explicitly disabling PSMI sleep on the RCS ring was insufficient, all the rings had to be awake to prevent the hangs. Fortunately, we can reduce the wakelock to the MI_SET_CONTEXT operation itself, and so should be able to limit the extra power implications. Since the MI_ARB_ON_OFF workaround is listed for all gen7 and above products, we should apply this extra hammer for all of the same platforms despite so far that we have only been able to reproduce the hang on certain ivb and hsw models. The last question is whether we want to always use the extra hammer or only when we know semaphores are in operation. At the moment, we only use LRI on non-RCS rings for semaphores, but that may change in the future with the possibility of reintroducing this bug under subtle conditions. v2: Make it explicit that the PSMI LRI are an extension to the original workaround for the other rings. v3: Bikeshedding variable names and whitespacing Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=80660 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=83677 Cc: Simon Farnsworth <simon@farnz.org.uk> Cc: Daniel Vetter <daniel@ffwll.ch> Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Tested-by: Peter Frühberger <fritsch@xbmc.org> Reviewed-by: Daniel Vetter <daniel@ffwll.ch> Cc: stable@vger.kernel.org Signed-off-by: Jani Nikula <jani.nikula@intel.com>
2014-12-16 10:02:27 +00:00
if (INTEL_INFO(ring->dev)->gen >= 7) {
intel_ring_emit(ring, MI_ARB_ON_OFF | MI_ARB_DISABLE);
drm/i915: Disable PSMI sleep messages on all rings around context switches There exists a current workaround to prevent a hang on context switch should the ring go to sleep in the middle of the restore, WaProgramMiArbOnOffAroundMiSetContext (applicable to all gen7+). In spite of disabling arbitration (which prevents the ring from powering down during the critical section) we were still hitting hangs that had the hallmarks of the known erratum. That is we are still seeing hangs "on the last instruction in the context restore". By comparing -nightly (broken) with requests (working), we were able to deduce that it was the semaphore LRI cross-talk that reproduced the original failure. The key was that requests implemented deferred semaphore signalling, and disabling that, i.e. emitting the semaphore signal to every other ring after every batch restored the frequent hang. Explicitly disabling PSMI sleep on the RCS ring was insufficient, all the rings had to be awake to prevent the hangs. Fortunately, we can reduce the wakelock to the MI_SET_CONTEXT operation itself, and so should be able to limit the extra power implications. Since the MI_ARB_ON_OFF workaround is listed for all gen7 and above products, we should apply this extra hammer for all of the same platforms despite so far that we have only been able to reproduce the hang on certain ivb and hsw models. The last question is whether we want to always use the extra hammer or only when we know semaphores are in operation. At the moment, we only use LRI on non-RCS rings for semaphores, but that may change in the future with the possibility of reintroducing this bug under subtle conditions. v2: Make it explicit that the PSMI LRI are an extension to the original workaround for the other rings. v3: Bikeshedding variable names and whitespacing Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=80660 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=83677 Cc: Simon Farnsworth <simon@farnz.org.uk> Cc: Daniel Vetter <daniel@ffwll.ch> Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Tested-by: Peter Frühberger <fritsch@xbmc.org> Reviewed-by: Daniel Vetter <daniel@ffwll.ch> Cc: stable@vger.kernel.org Signed-off-by: Jani Nikula <jani.nikula@intel.com>
2014-12-16 10:02:27 +00:00
if (num_rings) {
struct intel_engine_cs *signaller;
intel_ring_emit(ring, MI_LOAD_REGISTER_IMM(num_rings));
for_each_ring(signaller, to_i915(ring->dev), i) {
if (signaller == ring)
continue;
intel_ring_emit(ring, RING_PSMI_CTL(signaller->mmio_base));
intel_ring_emit(ring, _MASKED_BIT_ENABLE(GEN6_PSMI_SLEEP_MSG_DISABLE));
}
}
}
intel_ring_emit(ring, MI_NOOP);
intel_ring_emit(ring, MI_SET_CONTEXT);
drm/i915: Emphasize that ctx->obj & ctx->is_initialized refer to the legacy rcs ctx We have already advanced that Logical Ring Contexts have their own kind of backing objects, but everything will be better explained in the Execlists series. For now, suffice it to say that the current backing object is only ever used with the render ring, so we're making this fact more explicit (which is a good reason on its own). As for the is_initialized flag, we only use to signify that the render state has been initialized (a.k.a. golden context, a.k.a. null context). It doesn't mean anything for the other engines, so make that distinction obvious. Done with the following Coccinelle patch (plus manual frobbing of the struct): @@ struct intel_context c; @@ - (c).obj + c.legacy_hw_ctx.rcs_state @@ struct intel_context *c; @@ - (c)->obj + c->legacy_hw_ctx.rcs_state @@ struct intel_context c; @@ - (c).is_initialized + c.legacy_hw_ctx.initialized @@ struct intel_context *c; @@ - (c)->is_initialized + c->legacy_hw_ctx.initialized This Execlists prep-work patch has been suggested by Chris Wilson and Daniel Vetter separately. Initially, it was two separate patches: drm/i915: Rename ctx->obj to ctx->rcs_state drm/i915: Make it obvious that ctx->id is merely a user handle Signed-off-by: Oscar Mateo <oscar.mateo@intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> [danvet: s/id/is_initialized/ to fix the subject and resolve a conflict in i915_gem_context_reset. Also introduce a new lctx local variable to avoid overtly long lines.] Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-03 15:27:59 +00:00
intel_ring_emit(ring, i915_gem_obj_ggtt_offset(new_context->legacy_hw_ctx.rcs_state) |
flags);
/*
* w/a: MI_SET_CONTEXT must always be followed by MI_NOOP
* WaMiSetContext_Hang:snb,ivb,vlv
*/
intel_ring_emit(ring, MI_NOOP);
drm/i915: Disable PSMI sleep messages on all rings around context switches There exists a current workaround to prevent a hang on context switch should the ring go to sleep in the middle of the restore, WaProgramMiArbOnOffAroundMiSetContext (applicable to all gen7+). In spite of disabling arbitration (which prevents the ring from powering down during the critical section) we were still hitting hangs that had the hallmarks of the known erratum. That is we are still seeing hangs "on the last instruction in the context restore". By comparing -nightly (broken) with requests (working), we were able to deduce that it was the semaphore LRI cross-talk that reproduced the original failure. The key was that requests implemented deferred semaphore signalling, and disabling that, i.e. emitting the semaphore signal to every other ring after every batch restored the frequent hang. Explicitly disabling PSMI sleep on the RCS ring was insufficient, all the rings had to be awake to prevent the hangs. Fortunately, we can reduce the wakelock to the MI_SET_CONTEXT operation itself, and so should be able to limit the extra power implications. Since the MI_ARB_ON_OFF workaround is listed for all gen7 and above products, we should apply this extra hammer for all of the same platforms despite so far that we have only been able to reproduce the hang on certain ivb and hsw models. The last question is whether we want to always use the extra hammer or only when we know semaphores are in operation. At the moment, we only use LRI on non-RCS rings for semaphores, but that may change in the future with the possibility of reintroducing this bug under subtle conditions. v2: Make it explicit that the PSMI LRI are an extension to the original workaround for the other rings. v3: Bikeshedding variable names and whitespacing Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=80660 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=83677 Cc: Simon Farnsworth <simon@farnz.org.uk> Cc: Daniel Vetter <daniel@ffwll.ch> Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Tested-by: Peter Frühberger <fritsch@xbmc.org> Reviewed-by: Daniel Vetter <daniel@ffwll.ch> Cc: stable@vger.kernel.org Signed-off-by: Jani Nikula <jani.nikula@intel.com>
2014-12-16 10:02:27 +00:00
if (INTEL_INFO(ring->dev)->gen >= 7) {
if (num_rings) {
struct intel_engine_cs *signaller;
intel_ring_emit(ring, MI_LOAD_REGISTER_IMM(num_rings));
for_each_ring(signaller, to_i915(ring->dev), i) {
if (signaller == ring)
continue;
intel_ring_emit(ring, RING_PSMI_CTL(signaller->mmio_base));
intel_ring_emit(ring, _MASKED_BIT_DISABLE(GEN6_PSMI_SLEEP_MSG_DISABLE));
}
}
intel_ring_emit(ring, MI_ARB_ON_OFF | MI_ARB_ENABLE);
drm/i915: Disable PSMI sleep messages on all rings around context switches There exists a current workaround to prevent a hang on context switch should the ring go to sleep in the middle of the restore, WaProgramMiArbOnOffAroundMiSetContext (applicable to all gen7+). In spite of disabling arbitration (which prevents the ring from powering down during the critical section) we were still hitting hangs that had the hallmarks of the known erratum. That is we are still seeing hangs "on the last instruction in the context restore". By comparing -nightly (broken) with requests (working), we were able to deduce that it was the semaphore LRI cross-talk that reproduced the original failure. The key was that requests implemented deferred semaphore signalling, and disabling that, i.e. emitting the semaphore signal to every other ring after every batch restored the frequent hang. Explicitly disabling PSMI sleep on the RCS ring was insufficient, all the rings had to be awake to prevent the hangs. Fortunately, we can reduce the wakelock to the MI_SET_CONTEXT operation itself, and so should be able to limit the extra power implications. Since the MI_ARB_ON_OFF workaround is listed for all gen7 and above products, we should apply this extra hammer for all of the same platforms despite so far that we have only been able to reproduce the hang on certain ivb and hsw models. The last question is whether we want to always use the extra hammer or only when we know semaphores are in operation. At the moment, we only use LRI on non-RCS rings for semaphores, but that may change in the future with the possibility of reintroducing this bug under subtle conditions. v2: Make it explicit that the PSMI LRI are an extension to the original workaround for the other rings. v3: Bikeshedding variable names and whitespacing Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=80660 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=83677 Cc: Simon Farnsworth <simon@farnz.org.uk> Cc: Daniel Vetter <daniel@ffwll.ch> Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Tested-by: Peter Frühberger <fritsch@xbmc.org> Reviewed-by: Daniel Vetter <daniel@ffwll.ch> Cc: stable@vger.kernel.org Signed-off-by: Jani Nikula <jani.nikula@intel.com>
2014-12-16 10:02:27 +00:00
}
intel_ring_advance(ring);
return ret;
}
static inline bool should_skip_switch(struct intel_engine_cs *ring,
struct intel_context *from,
struct intel_context *to)
{
drm/i915: Track page table reload need This patch was formerly known as, "Force pd restore when PDEs change, gen6-7." I had to change the name because it is needed for GEN8 too. The real issue this is trying to solve is when a new object is mapped into the current address space. The GPU does not snoop the new mapping so we must do the gen specific action to reload the page tables. GEN8 and GEN7 do differ in the way they load page tables for the RCS. GEN8 does so with the context restore, while GEN7 requires the proper load commands in the command streamer. Non-render is similar for both. Caveat for GEN7 The docs say you cannot change the PDEs of a currently running context. We never map new PDEs of a running context, and expect them to be present - so I think this is okay. (We can unmap, but this should also be okay since we only unmap unreferenced objects that the GPU shouldn't be tryingto va->pa xlate.) The MI_SET_CONTEXT command does have a flag to signal that even if the context is the same, force a reload. It's unclear exactly what this does, but I have a hunch it's the right thing to do. The logic assumes that we always emit a context switch after mapping new PDEs, and before we submit a batch. This is the case today, and has been the case since the inception of hardware contexts. A note in the comment let's the user know. It's not just for gen8. If the current context has mappings change, we need a context reload to switch v2: Rebased after ppgtt clean up patches. Split the warning for aliasing and true ppgtt options. And do not break aliasing ppgtt, where to->ppgtt is always null. v3: Invalidate PPGTT TLBs inside alloc_va_range. v4: Rename ppgtt_invalidate_tlbs to mark_tlbs_dirty and move pd_dirty_rings from i915_address_space to i915_hw_ppgtt. Fixes when neither ctx->ppgtt and aliasing_ppgtt exist. v5: Removed references to teardown_va_range. v6: Updated needs_pd_load_pre/post. v7: Fix pd_dirty_rings check in needs_pd_load_post, and update/move comment about updated PDEs to object_pin/bind (Mika). Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com> Signed-off-by: Ben Widawsky <ben@bwidawsk.net> Signed-off-by: Michel Thierry <michel.thierry@intel.com> (v2+) Reviewed-by: Mika Kuoppala <mika.kuoppala@intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2015-03-19 12:53:28 +00:00
if (to->remap_slice)
return false;
if (to->ppgtt && from == to &&
!(intel_ring_flag(ring) & to->ppgtt->pd_dirty_rings))
return true;
return false;
}
static bool
needs_pd_load_pre(struct intel_engine_cs *ring, struct intel_context *to)
{
struct drm_i915_private *dev_priv = ring->dev->dev_private;
if (!to->ppgtt)
return false;
if (INTEL_INFO(ring->dev)->gen < 8)
return true;
if (ring != &dev_priv->ring[RCS])
return true;
return false;
}
static bool
needs_pd_load_post(struct intel_engine_cs *ring, struct intel_context *to,
u32 hw_flags)
{
struct drm_i915_private *dev_priv = ring->dev->dev_private;
if (!to->ppgtt)
return false;
if (!IS_GEN8(ring->dev))
return false;
if (ring != &dev_priv->ring[RCS])
return false;
if (hw_flags & MI_RESTORE_INHIBIT)
return true;
return false;
}
static int do_switch(struct intel_engine_cs *ring,
struct intel_context *to)
{
drm/i915: Create bind/unbind abstraction for VMAs To sum up what goes on here, we abstract the vma binding, similarly to the previous object binding. This helps for distinguishing legacy binding, versus modern binding. To keep the code churn as minimal as possible, I am leaving in insert_entries(). It serves as the per platform pte writing basically. bind_vma and insert_entries do share a lot of similarities, and I did have designs to combine the two, but as mentioned already... too much churn in an already massive patchset. What follows are the 3 commits which existed discretely in the original submissions. Upon rebasing on Broadwell support, it became clear that separation was not good, and only made for more error prone code. Below are the 3 commit messages with all their history. drm/i915: Add bind/unbind object functions to VMA drm/i915: Use the new vm [un]bind functions drm/i915: reduce vm->insert_entries() usage drm/i915: Add bind/unbind object functions to VMA As we plumb the code with more VM information, it has become more obvious that the easiest way to deal with bind and unbind is to simply put the function pointers in the vm, and let those choose the correct way to handle the page table updates. This change allows many places in the code to simply be vm->bind, and not have to worry about distinguishing PPGTT vs GGTT. Notice that this patch has no impact on functionality. I've decided to save the actual change until the next patch because I think it's easier to review that way. I'm happy to squash the two, or let Daniel do it on merge. v2: Make ggtt handle the quirky aliasing ppgtt Add flags to bind object to support above Don't ever call bind/unbind directly for PPGTT until we have real, full PPGTT (use NULLs to assert this) Make sure we rebind the ggtt if there already is a ggtt binding. This happens on set cache levels. Use VMA for bind/unbind (Daniel, Ben) v3: Reorganize ggtt_vma_bind to be more concise and easier to read (Ville). Change logic in unbind to only unbind ggtt when there is a global mapping, and to remove a redundant check if the aliasing ppgtt exists. v4: Make the bind function a bit smarter about the cache levels to avoid unnecessary multiple remaps. "I accept it is a wart, I think unifying the pin_vma / bind_vma could be unified later" (Chris) Removed the git notes, and put version info here. (Daniel) v5: Update the comment to not suck (Chris) v6: Move bind/unbind to the VMA. It makes more sense in the VMA structure (always has, but I was previously lazy). With this change, it will allow us to keep a distinct insert_entries. Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Signed-off-by: Ben Widawsky <ben@bwidawsk.net> drm/i915: Use the new vm [un]bind functions Building on the last patch which created the new function pointers in the VM for bind/unbind, here we actually put those new function pointers to use. Split out as a separate patch to aid in review. I'm fine with squashing into the previous patch if people request it. v2: Updated to address the smart ggtt which can do aliasing as needed Make sure we bind to global gtt when mappable and fenceable. I thought we could get away without this initialy, but we cannot. v3: Make the global GTT binding explicitly use the ggtt VM for bind_vma(). While at it, use the new ggtt_vma helper (Chris) At this point the original mailing list thread diverges. ie. v4^: use target_obj instead of obj for gen6 relocate_entry vma->bind_vma() can be called safely during pin. So simply do that instead of the complicated conditionals. Don't restore PPGTT bound objects on resume path Bug fix in resume path for globally bound Bos Properly handle secure dispatch Rebased on vma bind/unbind conversion Signed-off-by: Ben Widawsky <ben@bwidawsk.net> drm/i915: reduce vm->insert_entries() usage FKA: drm/i915: eliminate vm->insert_entries() With bind/unbind function pointers in place, we no longer need insert_entries. We could, and want, to remove clear_range, however it's not totally easy at this point. Since it's used in a couple of place still that don't only deal in objects: setup, ppgtt init, and restore gtt mappings. v2: Don't actually remove insert_entries, just limit its usage. It will be useful when we introduce gen8. It will always be called from the vma bind/unbind. Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> (v1) Signed-off-by: Ben Widawsky <ben@bwidawsk.net> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2013-12-06 22:10:56 +00:00
struct drm_i915_private *dev_priv = ring->dev->dev_private;
struct intel_context *from = ring->last_context;
u32 hw_flags = 0;
bool uninitialized = false;
drm/i915: Do remaps for all contexts On both Ivybridge and Haswell, row remapping information is saved and restored with context. This means, we never actually properly supported the l3 remapping because our sysfs interface is asynchronous (and not tied to any context), and the known faulty HW would be reused by the next context to run. Not that due to the asynchronous nature of the sysfs entry, there is no point modifying the registers for the existing context. Instead we set a flag for all contexts to load the correct remapping information on the next run. Interested clients can use debugfs to determine whether or not the row has been remapped. One could propose at this point that we just do the remapping in the kernel. I guess since we have to maintain the sysfs interface anyway, I'm not sure how useful it is, and I do like keeping the policy in userspace; (it wasn't my original decision to make the interface the way it is, so I'm not attached). v2: Force a context switch when we have a remap on the next switch. (Ville) Don't let userspace use the interface with disabled contexts. v3: Don't force a context switch, just let it nop Improper context slice remap initialization, 1<<1 instead of 1<<i, but I rewrote it to avoid a second round of confusion. Error print moved to error path (All Ville) Added a comment on why the slice remap initialization happens. CC: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Ben Widawsky <ben@bwidawsk.net> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2013-09-19 02:03:18 +00:00
int ret, i;
if (from != NULL && ring == &dev_priv->ring[RCS]) {
drm/i915: Emphasize that ctx->obj & ctx->is_initialized refer to the legacy rcs ctx We have already advanced that Logical Ring Contexts have their own kind of backing objects, but everything will be better explained in the Execlists series. For now, suffice it to say that the current backing object is only ever used with the render ring, so we're making this fact more explicit (which is a good reason on its own). As for the is_initialized flag, we only use to signify that the render state has been initialized (a.k.a. golden context, a.k.a. null context). It doesn't mean anything for the other engines, so make that distinction obvious. Done with the following Coccinelle patch (plus manual frobbing of the struct): @@ struct intel_context c; @@ - (c).obj + c.legacy_hw_ctx.rcs_state @@ struct intel_context *c; @@ - (c)->obj + c->legacy_hw_ctx.rcs_state @@ struct intel_context c; @@ - (c).is_initialized + c.legacy_hw_ctx.initialized @@ struct intel_context *c; @@ - (c)->is_initialized + c->legacy_hw_ctx.initialized This Execlists prep-work patch has been suggested by Chris Wilson and Daniel Vetter separately. Initially, it was two separate patches: drm/i915: Rename ctx->obj to ctx->rcs_state drm/i915: Make it obvious that ctx->id is merely a user handle Signed-off-by: Oscar Mateo <oscar.mateo@intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> [danvet: s/id/is_initialized/ to fix the subject and resolve a conflict in i915_gem_context_reset. Also introduce a new lctx local variable to avoid overtly long lines.] Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-03 15:27:59 +00:00
BUG_ON(from->legacy_hw_ctx.rcs_state == NULL);
BUG_ON(!i915_gem_obj_is_pinned(from->legacy_hw_ctx.rcs_state));
}
if (should_skip_switch(ring, from, to))
return 0;
/* Trying to pin first makes error handling easier. */
if (ring == &dev_priv->ring[RCS]) {
drm/i915: Emphasize that ctx->obj & ctx->is_initialized refer to the legacy rcs ctx We have already advanced that Logical Ring Contexts have their own kind of backing objects, but everything will be better explained in the Execlists series. For now, suffice it to say that the current backing object is only ever used with the render ring, so we're making this fact more explicit (which is a good reason on its own). As for the is_initialized flag, we only use to signify that the render state has been initialized (a.k.a. golden context, a.k.a. null context). It doesn't mean anything for the other engines, so make that distinction obvious. Done with the following Coccinelle patch (plus manual frobbing of the struct): @@ struct intel_context c; @@ - (c).obj + c.legacy_hw_ctx.rcs_state @@ struct intel_context *c; @@ - (c)->obj + c->legacy_hw_ctx.rcs_state @@ struct intel_context c; @@ - (c).is_initialized + c.legacy_hw_ctx.initialized @@ struct intel_context *c; @@ - (c)->is_initialized + c->legacy_hw_ctx.initialized This Execlists prep-work patch has been suggested by Chris Wilson and Daniel Vetter separately. Initially, it was two separate patches: drm/i915: Rename ctx->obj to ctx->rcs_state drm/i915: Make it obvious that ctx->id is merely a user handle Signed-off-by: Oscar Mateo <oscar.mateo@intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> [danvet: s/id/is_initialized/ to fix the subject and resolve a conflict in i915_gem_context_reset. Also introduce a new lctx local variable to avoid overtly long lines.] Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-03 15:27:59 +00:00
ret = i915_gem_obj_ggtt_pin(to->legacy_hw_ctx.rcs_state,
get_context_alignment(ring->dev), 0);
if (ret)
return ret;
}
drm/i915: Fix use-after-free in do_switch So apparently under ridiculous amounts of memory pressure we can get into trouble in do_switch when we try to move the old hw context backing storage object onto the active lists. With list debugging enabled that usually results in us chasing a poisoned pointer - which means we've hit upon a vma that has been removed from all lrus with list_del (and then deallocated, so it's a real use-after free). Ian Lister has done some great callchain chasing and noticed that we can reenter do_switch: i915_gem_do_execbuffer() i915_switch_context() do_switch() from = ring->last_context; i915_gem_object_pin() i915_gem_object_bind_to_gtt() ret = drm_mm_insert_node_in_range_generic(); // If the above call fails then it will try i915_gem_evict_something() // If that fails it will call i915_gem_evict_everything() ... i915_gem_evict_everything() i915_gpu_idle() i915_switch_context(DEFAULT_CONTEXT) Like with everything else where the shrinker or eviction code can invalidate pointers we need to reload relevant state. Note that there's no need to recheck whether a context switch is still required because: - Doing a switch to the same context is harmless (besides wasting a bit of energy). - This can only happen with the default context. But since that one's pinned we'll never call down into evict_everything under normal circumstances. Note that there's a little driver bringup fun involved namely that we could recourse into do_switch for the initial switch. Atm we're fine since we assign the context pointer only after the call to do_switch at driver load or resume time. And in the gpu reset case we skip the entire setup sequence (which might be a bug on its own, but definitely not this one here). Cc'ing stable since apparently ChromeOS guys are seeing this in the wild (and not just on artificial stress tests), see the reference. Note that in upstream code doesn't calle evict_everything directly from evict_something, that's an extension in this product branch. But we can still hit upon this bug (and apparently we do, see the linked backtraces). I've noticed this while trying to construct a testcase for this bug and utterly failed to provoke it. It looks like we need to driver the system squarly into the lowmem wall and provoke the shrinker to evict the context object by doing the last-ditch evict_everything call. Aside: There's currently no means to get a badly-fragmenting hw context object away from a bad spot in the upstream code. We should fix this by at least adding some code to evict_something to handle hw contexts. References: https://code.google.com/p/chromium/issues/detail?id=248191 Reported-by: Ian Lister <ian.lister@intel.com> Cc: Ian Lister <ian.lister@intel.com> Cc: stable@vger.kernel.org Cc: Ben Widawsky <benjamin.widawsky@intel.com> Cc: Stéphane Marchesin <marcheu@chromium.org> Cc: Bloomfield, Jon <jon.bloomfield@intel.com> Tested-by: Rafael Barbalho <rafael.barbalho@intel.com> Reviewed-by: Ian Lister <ian.lister@intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2013-12-05 14:42:34 +00:00
/*
* Pin can switch back to the default context if we end up calling into
* evict_everything - as a last ditch gtt defrag effort that also
* switches to the default context. Hence we need to reload from here.
*/
from = ring->last_context;
if (needs_pd_load_pre(ring, to)) {
/* Older GENs and non render rings still want the load first,
* "PP_DCLV followed by PP_DIR_BASE register through Load
* Register Immediate commands in Ring Buffer before submitting
* a context."*/
trace_switch_mm(ring, to);
ret = to->ppgtt->switch_mm(to->ppgtt, ring);
if (ret)
goto unpin_out;
drm/i915: Track page table reload need This patch was formerly known as, "Force pd restore when PDEs change, gen6-7." I had to change the name because it is needed for GEN8 too. The real issue this is trying to solve is when a new object is mapped into the current address space. The GPU does not snoop the new mapping so we must do the gen specific action to reload the page tables. GEN8 and GEN7 do differ in the way they load page tables for the RCS. GEN8 does so with the context restore, while GEN7 requires the proper load commands in the command streamer. Non-render is similar for both. Caveat for GEN7 The docs say you cannot change the PDEs of a currently running context. We never map new PDEs of a running context, and expect them to be present - so I think this is okay. (We can unmap, but this should also be okay since we only unmap unreferenced objects that the GPU shouldn't be tryingto va->pa xlate.) The MI_SET_CONTEXT command does have a flag to signal that even if the context is the same, force a reload. It's unclear exactly what this does, but I have a hunch it's the right thing to do. The logic assumes that we always emit a context switch after mapping new PDEs, and before we submit a batch. This is the case today, and has been the case since the inception of hardware contexts. A note in the comment let's the user know. It's not just for gen8. If the current context has mappings change, we need a context reload to switch v2: Rebased after ppgtt clean up patches. Split the warning for aliasing and true ppgtt options. And do not break aliasing ppgtt, where to->ppgtt is always null. v3: Invalidate PPGTT TLBs inside alloc_va_range. v4: Rename ppgtt_invalidate_tlbs to mark_tlbs_dirty and move pd_dirty_rings from i915_address_space to i915_hw_ppgtt. Fixes when neither ctx->ppgtt and aliasing_ppgtt exist. v5: Removed references to teardown_va_range. v6: Updated needs_pd_load_pre/post. v7: Fix pd_dirty_rings check in needs_pd_load_post, and update/move comment about updated PDEs to object_pin/bind (Mika). Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com> Signed-off-by: Ben Widawsky <ben@bwidawsk.net> Signed-off-by: Michel Thierry <michel.thierry@intel.com> (v2+) Reviewed-by: Mika Kuoppala <mika.kuoppala@intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2015-03-19 12:53:28 +00:00
/* Doing a PD load always reloads the page dirs */
to->ppgtt->pd_dirty_rings &= ~intel_ring_flag(ring);
}
if (ring != &dev_priv->ring[RCS]) {
if (from)
i915_gem_context_unreference(from);
goto done;
}
drm/i915: Fix use-after-free in do_switch So apparently under ridiculous amounts of memory pressure we can get into trouble in do_switch when we try to move the old hw context backing storage object onto the active lists. With list debugging enabled that usually results in us chasing a poisoned pointer - which means we've hit upon a vma that has been removed from all lrus with list_del (and then deallocated, so it's a real use-after free). Ian Lister has done some great callchain chasing and noticed that we can reenter do_switch: i915_gem_do_execbuffer() i915_switch_context() do_switch() from = ring->last_context; i915_gem_object_pin() i915_gem_object_bind_to_gtt() ret = drm_mm_insert_node_in_range_generic(); // If the above call fails then it will try i915_gem_evict_something() // If that fails it will call i915_gem_evict_everything() ... i915_gem_evict_everything() i915_gpu_idle() i915_switch_context(DEFAULT_CONTEXT) Like with everything else where the shrinker or eviction code can invalidate pointers we need to reload relevant state. Note that there's no need to recheck whether a context switch is still required because: - Doing a switch to the same context is harmless (besides wasting a bit of energy). - This can only happen with the default context. But since that one's pinned we'll never call down into evict_everything under normal circumstances. Note that there's a little driver bringup fun involved namely that we could recourse into do_switch for the initial switch. Atm we're fine since we assign the context pointer only after the call to do_switch at driver load or resume time. And in the gpu reset case we skip the entire setup sequence (which might be a bug on its own, but definitely not this one here). Cc'ing stable since apparently ChromeOS guys are seeing this in the wild (and not just on artificial stress tests), see the reference. Note that in upstream code doesn't calle evict_everything directly from evict_something, that's an extension in this product branch. But we can still hit upon this bug (and apparently we do, see the linked backtraces). I've noticed this while trying to construct a testcase for this bug and utterly failed to provoke it. It looks like we need to driver the system squarly into the lowmem wall and provoke the shrinker to evict the context object by doing the last-ditch evict_everything call. Aside: There's currently no means to get a badly-fragmenting hw context object away from a bad spot in the upstream code. We should fix this by at least adding some code to evict_something to handle hw contexts. References: https://code.google.com/p/chromium/issues/detail?id=248191 Reported-by: Ian Lister <ian.lister@intel.com> Cc: Ian Lister <ian.lister@intel.com> Cc: stable@vger.kernel.org Cc: Ben Widawsky <benjamin.widawsky@intel.com> Cc: Stéphane Marchesin <marcheu@chromium.org> Cc: Bloomfield, Jon <jon.bloomfield@intel.com> Tested-by: Rafael Barbalho <rafael.barbalho@intel.com> Reviewed-by: Ian Lister <ian.lister@intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2013-12-05 14:42:34 +00:00
/*
* Clear this page out of any CPU caches for coherent swap-in/out. Note
* that thanks to write = false in this call and us not setting any gpu
* write domains when putting a context object onto the active list
* (when switching away from it), this won't block.
drm/i915: Fix use-after-free in do_switch So apparently under ridiculous amounts of memory pressure we can get into trouble in do_switch when we try to move the old hw context backing storage object onto the active lists. With list debugging enabled that usually results in us chasing a poisoned pointer - which means we've hit upon a vma that has been removed from all lrus with list_del (and then deallocated, so it's a real use-after free). Ian Lister has done some great callchain chasing and noticed that we can reenter do_switch: i915_gem_do_execbuffer() i915_switch_context() do_switch() from = ring->last_context; i915_gem_object_pin() i915_gem_object_bind_to_gtt() ret = drm_mm_insert_node_in_range_generic(); // If the above call fails then it will try i915_gem_evict_something() // If that fails it will call i915_gem_evict_everything() ... i915_gem_evict_everything() i915_gpu_idle() i915_switch_context(DEFAULT_CONTEXT) Like with everything else where the shrinker or eviction code can invalidate pointers we need to reload relevant state. Note that there's no need to recheck whether a context switch is still required because: - Doing a switch to the same context is harmless (besides wasting a bit of energy). - This can only happen with the default context. But since that one's pinned we'll never call down into evict_everything under normal circumstances. Note that there's a little driver bringup fun involved namely that we could recourse into do_switch for the initial switch. Atm we're fine since we assign the context pointer only after the call to do_switch at driver load or resume time. And in the gpu reset case we skip the entire setup sequence (which might be a bug on its own, but definitely not this one here). Cc'ing stable since apparently ChromeOS guys are seeing this in the wild (and not just on artificial stress tests), see the reference. Note that in upstream code doesn't calle evict_everything directly from evict_something, that's an extension in this product branch. But we can still hit upon this bug (and apparently we do, see the linked backtraces). I've noticed this while trying to construct a testcase for this bug and utterly failed to provoke it. It looks like we need to driver the system squarly into the lowmem wall and provoke the shrinker to evict the context object by doing the last-ditch evict_everything call. Aside: There's currently no means to get a badly-fragmenting hw context object away from a bad spot in the upstream code. We should fix this by at least adding some code to evict_something to handle hw contexts. References: https://code.google.com/p/chromium/issues/detail?id=248191 Reported-by: Ian Lister <ian.lister@intel.com> Cc: Ian Lister <ian.lister@intel.com> Cc: stable@vger.kernel.org Cc: Ben Widawsky <benjamin.widawsky@intel.com> Cc: Stéphane Marchesin <marcheu@chromium.org> Cc: Bloomfield, Jon <jon.bloomfield@intel.com> Tested-by: Rafael Barbalho <rafael.barbalho@intel.com> Reviewed-by: Ian Lister <ian.lister@intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2013-12-05 14:42:34 +00:00
*
* XXX: We need a real interface to do this instead of trickery.
*/
drm/i915: Emphasize that ctx->obj & ctx->is_initialized refer to the legacy rcs ctx We have already advanced that Logical Ring Contexts have their own kind of backing objects, but everything will be better explained in the Execlists series. For now, suffice it to say that the current backing object is only ever used with the render ring, so we're making this fact more explicit (which is a good reason on its own). As for the is_initialized flag, we only use to signify that the render state has been initialized (a.k.a. golden context, a.k.a. null context). It doesn't mean anything for the other engines, so make that distinction obvious. Done with the following Coccinelle patch (plus manual frobbing of the struct): @@ struct intel_context c; @@ - (c).obj + c.legacy_hw_ctx.rcs_state @@ struct intel_context *c; @@ - (c)->obj + c->legacy_hw_ctx.rcs_state @@ struct intel_context c; @@ - (c).is_initialized + c.legacy_hw_ctx.initialized @@ struct intel_context *c; @@ - (c)->is_initialized + c->legacy_hw_ctx.initialized This Execlists prep-work patch has been suggested by Chris Wilson and Daniel Vetter separately. Initially, it was two separate patches: drm/i915: Rename ctx->obj to ctx->rcs_state drm/i915: Make it obvious that ctx->id is merely a user handle Signed-off-by: Oscar Mateo <oscar.mateo@intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> [danvet: s/id/is_initialized/ to fix the subject and resolve a conflict in i915_gem_context_reset. Also introduce a new lctx local variable to avoid overtly long lines.] Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-03 15:27:59 +00:00
ret = i915_gem_object_set_to_gtt_domain(to->legacy_hw_ctx.rcs_state, false);
if (ret)
goto unpin_out;
if (!to->legacy_hw_ctx.initialized) {
hw_flags |= MI_RESTORE_INHIBIT;
/* NB: If we inhibit the restore, the context is not allowed to
* die because future work may end up depending on valid address
* space. This means we must enforce that a page table load
* occur when this occurs. */
} else if (to->ppgtt &&
(intel_ring_flag(ring) & to->ppgtt->pd_dirty_rings)) {
drm/i915: Track page table reload need This patch was formerly known as, "Force pd restore when PDEs change, gen6-7." I had to change the name because it is needed for GEN8 too. The real issue this is trying to solve is when a new object is mapped into the current address space. The GPU does not snoop the new mapping so we must do the gen specific action to reload the page tables. GEN8 and GEN7 do differ in the way they load page tables for the RCS. GEN8 does so with the context restore, while GEN7 requires the proper load commands in the command streamer. Non-render is similar for both. Caveat for GEN7 The docs say you cannot change the PDEs of a currently running context. We never map new PDEs of a running context, and expect them to be present - so I think this is okay. (We can unmap, but this should also be okay since we only unmap unreferenced objects that the GPU shouldn't be tryingto va->pa xlate.) The MI_SET_CONTEXT command does have a flag to signal that even if the context is the same, force a reload. It's unclear exactly what this does, but I have a hunch it's the right thing to do. The logic assumes that we always emit a context switch after mapping new PDEs, and before we submit a batch. This is the case today, and has been the case since the inception of hardware contexts. A note in the comment let's the user know. It's not just for gen8. If the current context has mappings change, we need a context reload to switch v2: Rebased after ppgtt clean up patches. Split the warning for aliasing and true ppgtt options. And do not break aliasing ppgtt, where to->ppgtt is always null. v3: Invalidate PPGTT TLBs inside alloc_va_range. v4: Rename ppgtt_invalidate_tlbs to mark_tlbs_dirty and move pd_dirty_rings from i915_address_space to i915_hw_ppgtt. Fixes when neither ctx->ppgtt and aliasing_ppgtt exist. v5: Removed references to teardown_va_range. v6: Updated needs_pd_load_pre/post. v7: Fix pd_dirty_rings check in needs_pd_load_post, and update/move comment about updated PDEs to object_pin/bind (Mika). Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com> Signed-off-by: Ben Widawsky <ben@bwidawsk.net> Signed-off-by: Michel Thierry <michel.thierry@intel.com> (v2+) Reviewed-by: Mika Kuoppala <mika.kuoppala@intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2015-03-19 12:53:28 +00:00
hw_flags |= MI_FORCE_RESTORE;
to->ppgtt->pd_dirty_rings &= ~intel_ring_flag(ring);
}
/* We should never emit switch_mm more than once */
WARN_ON(needs_pd_load_pre(ring, to) &&
needs_pd_load_post(ring, to, hw_flags));
ret = mi_set_context(ring, to, hw_flags);
if (ret)
goto unpin_out;
/* GEN8 does *not* require an explicit reload if the PDPs have been
* setup, and we do not wish to move them.
*/
if (needs_pd_load_post(ring, to, hw_flags)) {
trace_switch_mm(ring, to);
ret = to->ppgtt->switch_mm(to->ppgtt, ring);
/* The hardware context switch is emitted, but we haven't
* actually changed the state - so it's probably safe to bail
* here. Still, let the user know something dangerous has
* happened.
*/
if (ret) {
DRM_ERROR("Failed to change address space on context switch\n");
goto unpin_out;
}
}
drm/i915: Do remaps for all contexts On both Ivybridge and Haswell, row remapping information is saved and restored with context. This means, we never actually properly supported the l3 remapping because our sysfs interface is asynchronous (and not tied to any context), and the known faulty HW would be reused by the next context to run. Not that due to the asynchronous nature of the sysfs entry, there is no point modifying the registers for the existing context. Instead we set a flag for all contexts to load the correct remapping information on the next run. Interested clients can use debugfs to determine whether or not the row has been remapped. One could propose at this point that we just do the remapping in the kernel. I guess since we have to maintain the sysfs interface anyway, I'm not sure how useful it is, and I do like keeping the policy in userspace; (it wasn't my original decision to make the interface the way it is, so I'm not attached). v2: Force a context switch when we have a remap on the next switch. (Ville) Don't let userspace use the interface with disabled contexts. v3: Don't force a context switch, just let it nop Improper context slice remap initialization, 1<<1 instead of 1<<i, but I rewrote it to avoid a second round of confusion. Error print moved to error path (All Ville) Added a comment on why the slice remap initialization happens. CC: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Ben Widawsky <ben@bwidawsk.net> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2013-09-19 02:03:18 +00:00
for (i = 0; i < MAX_L3_SLICES; i++) {
if (!(to->remap_slice & (1<<i)))
continue;
ret = i915_gem_l3_remap(ring, i);
/* If it failed, try again next round */
if (ret)
DRM_DEBUG_DRIVER("L3 remapping failed\n");
else
to->remap_slice &= ~(1<<i);
}
/* The backing object for the context is done after switching to the
* *next* context. Therefore we cannot retire the previous context until
* the next context has already started running. In fact, the below code
* is a bit suboptimal because the retiring can occur simply after the
* MI_SET_CONTEXT instead of when the next seqno has completed.
*/
if (from != NULL) {
drm/i915: Emphasize that ctx->obj & ctx->is_initialized refer to the legacy rcs ctx We have already advanced that Logical Ring Contexts have their own kind of backing objects, but everything will be better explained in the Execlists series. For now, suffice it to say that the current backing object is only ever used with the render ring, so we're making this fact more explicit (which is a good reason on its own). As for the is_initialized flag, we only use to signify that the render state has been initialized (a.k.a. golden context, a.k.a. null context). It doesn't mean anything for the other engines, so make that distinction obvious. Done with the following Coccinelle patch (plus manual frobbing of the struct): @@ struct intel_context c; @@ - (c).obj + c.legacy_hw_ctx.rcs_state @@ struct intel_context *c; @@ - (c)->obj + c->legacy_hw_ctx.rcs_state @@ struct intel_context c; @@ - (c).is_initialized + c.legacy_hw_ctx.initialized @@ struct intel_context *c; @@ - (c)->is_initialized + c->legacy_hw_ctx.initialized This Execlists prep-work patch has been suggested by Chris Wilson and Daniel Vetter separately. Initially, it was two separate patches: drm/i915: Rename ctx->obj to ctx->rcs_state drm/i915: Make it obvious that ctx->id is merely a user handle Signed-off-by: Oscar Mateo <oscar.mateo@intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> [danvet: s/id/is_initialized/ to fix the subject and resolve a conflict in i915_gem_context_reset. Also introduce a new lctx local variable to avoid overtly long lines.] Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-03 15:27:59 +00:00
from->legacy_hw_ctx.rcs_state->base.read_domains = I915_GEM_DOMAIN_INSTRUCTION;
i915_vma_move_to_active(i915_gem_obj_to_ggtt(from->legacy_hw_ctx.rcs_state), ring);
/* As long as MI_SET_CONTEXT is serializing, ie. it flushes the
* whole damn pipeline, we don't need to explicitly mark the
* object dirty. The only exception is that the context must be
* correct in case the object gets swapped out. Ideally we'd be
* able to defer doing this until we know the object would be
* swapped, but there is no way to do that yet.
*/
drm/i915: Emphasize that ctx->obj & ctx->is_initialized refer to the legacy rcs ctx We have already advanced that Logical Ring Contexts have their own kind of backing objects, but everything will be better explained in the Execlists series. For now, suffice it to say that the current backing object is only ever used with the render ring, so we're making this fact more explicit (which is a good reason on its own). As for the is_initialized flag, we only use to signify that the render state has been initialized (a.k.a. golden context, a.k.a. null context). It doesn't mean anything for the other engines, so make that distinction obvious. Done with the following Coccinelle patch (plus manual frobbing of the struct): @@ struct intel_context c; @@ - (c).obj + c.legacy_hw_ctx.rcs_state @@ struct intel_context *c; @@ - (c)->obj + c->legacy_hw_ctx.rcs_state @@ struct intel_context c; @@ - (c).is_initialized + c.legacy_hw_ctx.initialized @@ struct intel_context *c; @@ - (c)->is_initialized + c->legacy_hw_ctx.initialized This Execlists prep-work patch has been suggested by Chris Wilson and Daniel Vetter separately. Initially, it was two separate patches: drm/i915: Rename ctx->obj to ctx->rcs_state drm/i915: Make it obvious that ctx->id is merely a user handle Signed-off-by: Oscar Mateo <oscar.mateo@intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> [danvet: s/id/is_initialized/ to fix the subject and resolve a conflict in i915_gem_context_reset. Also introduce a new lctx local variable to avoid overtly long lines.] Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-03 15:27:59 +00:00
from->legacy_hw_ctx.rcs_state->dirty = 1;
drm/i915: Do not add an interrupt for a context switch We use the request to ensure we hold a reference to the context for the duration that it remains in use by the ring. Each request only holds a reference to the current context, hence we emit a request after switching contexts with the final reference to the old context. However, the extra interrupt caused by that request is not useful (no timing critical function will wait for the context object), instead the overhead of servicing the IRQ shows up in some (lightweight) benchmarks. In order to keep the useful property of using the request to manage the context lifetime, we want to add a dummy request that is associated with the interrupt from the subsequent real request following the batch. The extra interrupt was added as a side-effect of using i915_add_request() in commit 112522f6789581824903f6f72082b5b841a7f0f9 Author: Chris Wilson <chris@chris-wilson.co.uk> Date: Thu May 2 16:48:07 2013 +0300 drm/i915: put context upon switching v2: Daniel convinced me that the request here was solely for context lifetime tracking and that we have the active ref to keep the object alive whilst the MI_SET_CONTEXT. So the only concern then is which context should get the blame for MI_SET_CONTEXT failing. The old scheme added a request for the old context so that any hang upto and including the switch away would mark the old context as guilty. Now any hang here implicates the new context. However since we have already gone through a complete flush with the last context in its last request, and all that lies in no-man's-land is an invalidate flush and the MI_SET_CONTEXT, we should be safe in not unduly placing blame on the new context. Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Ben Widawsky <ben@bwidawsk.net> Cc: Paulo Zanoni <paulo.r.zanoni@intel.com> Reviewed-by: Damien Lespiau <damien.lespiau@intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2013-08-26 22:50:53 +00:00
/* obj is kept alive until the next request by its active ref */
drm/i915: Emphasize that ctx->obj & ctx->is_initialized refer to the legacy rcs ctx We have already advanced that Logical Ring Contexts have their own kind of backing objects, but everything will be better explained in the Execlists series. For now, suffice it to say that the current backing object is only ever used with the render ring, so we're making this fact more explicit (which is a good reason on its own). As for the is_initialized flag, we only use to signify that the render state has been initialized (a.k.a. golden context, a.k.a. null context). It doesn't mean anything for the other engines, so make that distinction obvious. Done with the following Coccinelle patch (plus manual frobbing of the struct): @@ struct intel_context c; @@ - (c).obj + c.legacy_hw_ctx.rcs_state @@ struct intel_context *c; @@ - (c)->obj + c->legacy_hw_ctx.rcs_state @@ struct intel_context c; @@ - (c).is_initialized + c.legacy_hw_ctx.initialized @@ struct intel_context *c; @@ - (c)->is_initialized + c->legacy_hw_ctx.initialized This Execlists prep-work patch has been suggested by Chris Wilson and Daniel Vetter separately. Initially, it was two separate patches: drm/i915: Rename ctx->obj to ctx->rcs_state drm/i915: Make it obvious that ctx->id is merely a user handle Signed-off-by: Oscar Mateo <oscar.mateo@intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> [danvet: s/id/is_initialized/ to fix the subject and resolve a conflict in i915_gem_context_reset. Also introduce a new lctx local variable to avoid overtly long lines.] Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-03 15:27:59 +00:00
i915_gem_object_ggtt_unpin(from->legacy_hw_ctx.rcs_state);
i915_gem_context_unreference(from);
}
uninitialized = !to->legacy_hw_ctx.initialized;
drm/i915: Emphasize that ctx->obj & ctx->is_initialized refer to the legacy rcs ctx We have already advanced that Logical Ring Contexts have their own kind of backing objects, but everything will be better explained in the Execlists series. For now, suffice it to say that the current backing object is only ever used with the render ring, so we're making this fact more explicit (which is a good reason on its own). As for the is_initialized flag, we only use to signify that the render state has been initialized (a.k.a. golden context, a.k.a. null context). It doesn't mean anything for the other engines, so make that distinction obvious. Done with the following Coccinelle patch (plus manual frobbing of the struct): @@ struct intel_context c; @@ - (c).obj + c.legacy_hw_ctx.rcs_state @@ struct intel_context *c; @@ - (c)->obj + c->legacy_hw_ctx.rcs_state @@ struct intel_context c; @@ - (c).is_initialized + c.legacy_hw_ctx.initialized @@ struct intel_context *c; @@ - (c)->is_initialized + c->legacy_hw_ctx.initialized This Execlists prep-work patch has been suggested by Chris Wilson and Daniel Vetter separately. Initially, it was two separate patches: drm/i915: Rename ctx->obj to ctx->rcs_state drm/i915: Make it obvious that ctx->id is merely a user handle Signed-off-by: Oscar Mateo <oscar.mateo@intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> [danvet: s/id/is_initialized/ to fix the subject and resolve a conflict in i915_gem_context_reset. Also introduce a new lctx local variable to avoid overtly long lines.] Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-03 15:27:59 +00:00
to->legacy_hw_ctx.initialized = true;
done:
i915_gem_context_reference(to);
ring->last_context = to;
if (uninitialized) {
if (ring->init_context) {
ret = ring->init_context(ring, to);
if (ret)
DRM_ERROR("ring init context: %d\n", ret);
}
}
return 0;
unpin_out:
if (ring->id == RCS)
drm/i915: Emphasize that ctx->obj & ctx->is_initialized refer to the legacy rcs ctx We have already advanced that Logical Ring Contexts have their own kind of backing objects, but everything will be better explained in the Execlists series. For now, suffice it to say that the current backing object is only ever used with the render ring, so we're making this fact more explicit (which is a good reason on its own). As for the is_initialized flag, we only use to signify that the render state has been initialized (a.k.a. golden context, a.k.a. null context). It doesn't mean anything for the other engines, so make that distinction obvious. Done with the following Coccinelle patch (plus manual frobbing of the struct): @@ struct intel_context c; @@ - (c).obj + c.legacy_hw_ctx.rcs_state @@ struct intel_context *c; @@ - (c)->obj + c->legacy_hw_ctx.rcs_state @@ struct intel_context c; @@ - (c).is_initialized + c.legacy_hw_ctx.initialized @@ struct intel_context *c; @@ - (c)->is_initialized + c->legacy_hw_ctx.initialized This Execlists prep-work patch has been suggested by Chris Wilson and Daniel Vetter separately. Initially, it was two separate patches: drm/i915: Rename ctx->obj to ctx->rcs_state drm/i915: Make it obvious that ctx->id is merely a user handle Signed-off-by: Oscar Mateo <oscar.mateo@intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> [danvet: s/id/is_initialized/ to fix the subject and resolve a conflict in i915_gem_context_reset. Also introduce a new lctx local variable to avoid overtly long lines.] Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-03 15:27:59 +00:00
i915_gem_object_ggtt_unpin(to->legacy_hw_ctx.rcs_state);
return ret;
}
/**
* i915_switch_context() - perform a GPU context switch.
* @ring: ring for which we'll execute the context switch
* @to: the context to switch to
*
* The context life cycle is simple. The context refcount is incremented and
* decremented by 1 and create and destroy. If the context is in use by the GPU,
* it will have a refcount > 1. This allows us to destroy the context abstract
* object while letting the normal object tracking destroy the backing BO.
*
* This function should not be used in execlists mode. Instead the context is
* switched by writing to the ELSP and requests keep a reference to their
* context.
*/
int i915_switch_context(struct intel_engine_cs *ring,
struct intel_context *to)
{
struct drm_i915_private *dev_priv = ring->dev->dev_private;
WARN_ON(i915.enable_execlists);
WARN_ON(!mutex_is_locked(&dev_priv->dev->struct_mutex));
drm/i915: Emphasize that ctx->obj & ctx->is_initialized refer to the legacy rcs ctx We have already advanced that Logical Ring Contexts have their own kind of backing objects, but everything will be better explained in the Execlists series. For now, suffice it to say that the current backing object is only ever used with the render ring, so we're making this fact more explicit (which is a good reason on its own). As for the is_initialized flag, we only use to signify that the render state has been initialized (a.k.a. golden context, a.k.a. null context). It doesn't mean anything for the other engines, so make that distinction obvious. Done with the following Coccinelle patch (plus manual frobbing of the struct): @@ struct intel_context c; @@ - (c).obj + c.legacy_hw_ctx.rcs_state @@ struct intel_context *c; @@ - (c)->obj + c->legacy_hw_ctx.rcs_state @@ struct intel_context c; @@ - (c).is_initialized + c.legacy_hw_ctx.initialized @@ struct intel_context *c; @@ - (c)->is_initialized + c->legacy_hw_ctx.initialized This Execlists prep-work patch has been suggested by Chris Wilson and Daniel Vetter separately. Initially, it was two separate patches: drm/i915: Rename ctx->obj to ctx->rcs_state drm/i915: Make it obvious that ctx->id is merely a user handle Signed-off-by: Oscar Mateo <oscar.mateo@intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> [danvet: s/id/is_initialized/ to fix the subject and resolve a conflict in i915_gem_context_reset. Also introduce a new lctx local variable to avoid overtly long lines.] Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-03 15:27:59 +00:00
if (to->legacy_hw_ctx.rcs_state == NULL) { /* We have the fake context */
if (to != ring->last_context) {
i915_gem_context_reference(to);
if (ring->last_context)
i915_gem_context_unreference(ring->last_context);
ring->last_context = to;
}
return 0;
}
return do_switch(ring, to);
}
static bool contexts_enabled(struct drm_device *dev)
{
return i915.enable_execlists || to_i915(dev)->hw_context_size;
}
int i915_gem_context_create_ioctl(struct drm_device *dev, void *data,
struct drm_file *file)
{
struct drm_i915_gem_context_create *args = data;
struct drm_i915_file_private *file_priv = file->driver_priv;
struct intel_context *ctx;
int ret;
if (!contexts_enabled(dev))
return -ENODEV;
ret = i915_mutex_lock_interruptible(dev);
if (ret)
return ret;
ctx = i915_gem_create_context(dev, file_priv);
mutex_unlock(&dev->struct_mutex);
if (IS_ERR(ctx))
return PTR_ERR(ctx);
drm/i915: Emphasize that ctx->id is merely a user handle This is an Execlists preparatory patch, since they make context ID become an overloaded term: - In the software, it was used to distinguish which context userspace was trying to use. - In the BSpec, the term is used to describe the 20-bits long field the hardware uses to it to discriminate the contexts that are submitted to the ELSP and inform the driver about their current status (via Context Switch Interrupts and Context Status Buffers). Initially, I tried to make the different meanings converge, but it proved impossible: - The software ctx->id is per-filp, while the hardware one needs to be globally unique. - Also, we multiplex several backing states objects per intel_context, and all of them need unique HW IDs. - I tried adding a per-filp ID and then composing the HW context ID as: ctx->id + file_priv->id + ring->id, but the fact that the hardware only uses 20-bits means we have to artificially limit the number of filps or contexts the userspace can create. The ctx->user_handle renaming bits are done with this Cocci patch (plus manual frobbing of the struct declaration): @@ struct intel_context c; @@ - (c).id + c.user_handle @@ struct intel_context *c; @@ - (c)->id + c->user_handle Also, while we are at it, s/DEFAULT_CONTEXT_ID/DEFAULT_CONTEXT_HANDLE and change the type to unsigned 32 bits. v2: s/handle/user_handle and change the type to uint32_t as suggested by Chris Wilson. Reviewed-by: Jesse Barnes <jbarnes@virtuousgeek.org> (v1) Signed-off-by: Oscar Mateo <oscar.mateo@intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-03 15:28:00 +00:00
args->ctx_id = ctx->user_handle;
DRM_DEBUG_DRIVER("HW context %d created\n", args->ctx_id);
return 0;
}
int i915_gem_context_destroy_ioctl(struct drm_device *dev, void *data,
struct drm_file *file)
{
struct drm_i915_gem_context_destroy *args = data;
struct drm_i915_file_private *file_priv = file->driver_priv;
struct intel_context *ctx;
int ret;
drm/i915: Emphasize that ctx->id is merely a user handle This is an Execlists preparatory patch, since they make context ID become an overloaded term: - In the software, it was used to distinguish which context userspace was trying to use. - In the BSpec, the term is used to describe the 20-bits long field the hardware uses to it to discriminate the contexts that are submitted to the ELSP and inform the driver about their current status (via Context Switch Interrupts and Context Status Buffers). Initially, I tried to make the different meanings converge, but it proved impossible: - The software ctx->id is per-filp, while the hardware one needs to be globally unique. - Also, we multiplex several backing states objects per intel_context, and all of them need unique HW IDs. - I tried adding a per-filp ID and then composing the HW context ID as: ctx->id + file_priv->id + ring->id, but the fact that the hardware only uses 20-bits means we have to artificially limit the number of filps or contexts the userspace can create. The ctx->user_handle renaming bits are done with this Cocci patch (plus manual frobbing of the struct declaration): @@ struct intel_context c; @@ - (c).id + c.user_handle @@ struct intel_context *c; @@ - (c)->id + c->user_handle Also, while we are at it, s/DEFAULT_CONTEXT_ID/DEFAULT_CONTEXT_HANDLE and change the type to unsigned 32 bits. v2: s/handle/user_handle and change the type to uint32_t as suggested by Chris Wilson. Reviewed-by: Jesse Barnes <jbarnes@virtuousgeek.org> (v1) Signed-off-by: Oscar Mateo <oscar.mateo@intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-03 15:28:00 +00:00
if (args->ctx_id == DEFAULT_CONTEXT_HANDLE)
return -ENOENT;
ret = i915_mutex_lock_interruptible(dev);
if (ret)
return ret;
ctx = i915_gem_context_get(file_priv, args->ctx_id);
if (IS_ERR(ctx)) {
mutex_unlock(&dev->struct_mutex);
return PTR_ERR(ctx);
}
drm/i915: Emphasize that ctx->id is merely a user handle This is an Execlists preparatory patch, since they make context ID become an overloaded term: - In the software, it was used to distinguish which context userspace was trying to use. - In the BSpec, the term is used to describe the 20-bits long field the hardware uses to it to discriminate the contexts that are submitted to the ELSP and inform the driver about their current status (via Context Switch Interrupts and Context Status Buffers). Initially, I tried to make the different meanings converge, but it proved impossible: - The software ctx->id is per-filp, while the hardware one needs to be globally unique. - Also, we multiplex several backing states objects per intel_context, and all of them need unique HW IDs. - I tried adding a per-filp ID and then composing the HW context ID as: ctx->id + file_priv->id + ring->id, but the fact that the hardware only uses 20-bits means we have to artificially limit the number of filps or contexts the userspace can create. The ctx->user_handle renaming bits are done with this Cocci patch (plus manual frobbing of the struct declaration): @@ struct intel_context c; @@ - (c).id + c.user_handle @@ struct intel_context *c; @@ - (c)->id + c->user_handle Also, while we are at it, s/DEFAULT_CONTEXT_ID/DEFAULT_CONTEXT_HANDLE and change the type to unsigned 32 bits. v2: s/handle/user_handle and change the type to uint32_t as suggested by Chris Wilson. Reviewed-by: Jesse Barnes <jbarnes@virtuousgeek.org> (v1) Signed-off-by: Oscar Mateo <oscar.mateo@intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-07-03 15:28:00 +00:00
idr_remove(&ctx->file_priv->context_idr, ctx->user_handle);
i915_gem_context_unreference(ctx);
mutex_unlock(&dev->struct_mutex);
DRM_DEBUG_DRIVER("HW context %d destroyed\n", args->ctx_id);
return 0;
}
int i915_gem_context_getparam_ioctl(struct drm_device *dev, void *data,
struct drm_file *file)
{
struct drm_i915_file_private *file_priv = file->driver_priv;
struct drm_i915_gem_context_param *args = data;
struct intel_context *ctx;
int ret;
ret = i915_mutex_lock_interruptible(dev);
if (ret)
return ret;
ctx = i915_gem_context_get(file_priv, args->ctx_id);
if (IS_ERR(ctx)) {
mutex_unlock(&dev->struct_mutex);
return PTR_ERR(ctx);
}
args->size = 0;
switch (args->param) {
case I915_CONTEXT_PARAM_BAN_PERIOD:
args->value = ctx->hang_stats.ban_period_seconds;
break;
case I915_CONTEXT_PARAM_NO_ZEROMAP:
args->value = ctx->flags & CONTEXT_NO_ZEROMAP;
break;
default:
ret = -EINVAL;
break;
}
mutex_unlock(&dev->struct_mutex);
return ret;
}
int i915_gem_context_setparam_ioctl(struct drm_device *dev, void *data,
struct drm_file *file)
{
struct drm_i915_file_private *file_priv = file->driver_priv;
struct drm_i915_gem_context_param *args = data;
struct intel_context *ctx;
int ret;
ret = i915_mutex_lock_interruptible(dev);
if (ret)
return ret;
ctx = i915_gem_context_get(file_priv, args->ctx_id);
if (IS_ERR(ctx)) {
mutex_unlock(&dev->struct_mutex);
return PTR_ERR(ctx);
}
switch (args->param) {
case I915_CONTEXT_PARAM_BAN_PERIOD:
if (args->size)
ret = -EINVAL;
else if (args->value < ctx->hang_stats.ban_period_seconds &&
!capable(CAP_SYS_ADMIN))
ret = -EPERM;
else
ctx->hang_stats.ban_period_seconds = args->value;
break;
case I915_CONTEXT_PARAM_NO_ZEROMAP:
if (args->size) {
ret = -EINVAL;
} else {
ctx->flags &= ~CONTEXT_NO_ZEROMAP;
ctx->flags |= args->value ? CONTEXT_NO_ZEROMAP : 0;
}
break;
default:
ret = -EINVAL;
break;
}
mutex_unlock(&dev->struct_mutex);
return ret;
}