2015-08-12 22:43:36 +08:00
|
|
|
/*
|
|
|
|
* Copyright © 2014 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:
|
|
|
|
* Vinit Azad <vinit.azad@intel.com>
|
|
|
|
* Ben Widawsky <ben@bwidawsk.net>
|
|
|
|
* Dave Gordon <david.s.gordon@intel.com>
|
|
|
|
* Alex Dai <yu.dai@intel.com>
|
|
|
|
*/
|
|
|
|
#include <linux/firmware.h>
|
|
|
|
#include "i915_drv.h"
|
|
|
|
#include "intel_guc.h"
|
|
|
|
|
|
|
|
/**
|
2015-10-20 07:10:54 +08:00
|
|
|
* DOC: GuC-specific firmware loader
|
2015-08-12 22:43:36 +08:00
|
|
|
*
|
|
|
|
* intel_guc:
|
|
|
|
* Top level structure of guc. It handles firmware loading and manages client
|
|
|
|
* pool and doorbells. intel_guc owns a i915_guc_client to replace the legacy
|
|
|
|
* ExecList submission.
|
|
|
|
*
|
|
|
|
* Firmware versioning:
|
|
|
|
* The firmware build process will generate a version header file with major and
|
|
|
|
* minor version defined. The versions are built into CSS header of firmware.
|
|
|
|
* i915 kernel driver set the minimal firmware version required per platform.
|
|
|
|
* The firmware installation package will install (symbolic link) proper version
|
|
|
|
* of firmware.
|
|
|
|
*
|
|
|
|
* GuC address space:
|
|
|
|
* GuC does not allow any gfx GGTT address that falls into range [0, WOPCM_TOP),
|
|
|
|
* which is reserved for Boot ROM, SRAM and WOPCM. Currently this top address is
|
|
|
|
* 512K. In order to exclude 0-512K address space from GGTT, all gfx objects
|
|
|
|
* used by GuC is pinned with PIN_OFFSET_BIAS along with size of WOPCM.
|
|
|
|
*
|
|
|
|
* Firmware log:
|
|
|
|
* Firmware log is enabled by setting i915.guc_log_level to non-negative level.
|
|
|
|
* Log data is printed out via reading debugfs i915_guc_log_dump. Reading from
|
|
|
|
* i915_guc_load_status will print out firmware loading status and scratch
|
|
|
|
* registers value.
|
|
|
|
*
|
|
|
|
*/
|
|
|
|
|
2016-08-10 23:16:46 +08:00
|
|
|
#define SKL_FW_MAJOR 6
|
|
|
|
#define SKL_FW_MINOR 1
|
|
|
|
|
|
|
|
#define BXT_FW_MAJOR 8
|
|
|
|
#define BXT_FW_MINOR 7
|
|
|
|
|
|
|
|
#define KBL_FW_MAJOR 9
|
|
|
|
#define KBL_FW_MINOR 14
|
|
|
|
|
|
|
|
#define GUC_FW_PATH(platform, major, minor) \
|
|
|
|
"i915/" __stringify(platform) "_guc_ver" __stringify(major) "_" __stringify(minor) ".bin"
|
|
|
|
|
|
|
|
#define I915_SKL_GUC_UCODE GUC_FW_PATH(skl, SKL_FW_MAJOR, SKL_FW_MINOR)
|
2015-08-12 22:43:36 +08:00
|
|
|
MODULE_FIRMWARE(I915_SKL_GUC_UCODE);
|
|
|
|
|
2016-08-10 23:16:46 +08:00
|
|
|
#define I915_BXT_GUC_UCODE GUC_FW_PATH(bxt, BXT_FW_MAJOR, BXT_FW_MINOR)
|
2016-05-06 18:42:53 +08:00
|
|
|
MODULE_FIRMWARE(I915_BXT_GUC_UCODE);
|
|
|
|
|
2016-08-10 23:16:46 +08:00
|
|
|
#define I915_KBL_GUC_UCODE GUC_FW_PATH(kbl, KBL_FW_MAJOR, KBL_FW_MINOR)
|
2016-07-01 00:37:52 +08:00
|
|
|
MODULE_FIRMWARE(I915_KBL_GUC_UCODE);
|
|
|
|
|
2015-08-12 22:43:36 +08:00
|
|
|
/* User-friendly representation of an enum */
|
|
|
|
const char *intel_guc_fw_status_repr(enum intel_guc_fw_status status)
|
|
|
|
{
|
|
|
|
switch (status) {
|
|
|
|
case GUC_FIRMWARE_FAIL:
|
|
|
|
return "FAIL";
|
|
|
|
case GUC_FIRMWARE_NONE:
|
|
|
|
return "NONE";
|
|
|
|
case GUC_FIRMWARE_PENDING:
|
|
|
|
return "PENDING";
|
|
|
|
case GUC_FIRMWARE_SUCCESS:
|
|
|
|
return "SUCCESS";
|
|
|
|
default:
|
|
|
|
return "UNKNOWN!";
|
|
|
|
}
|
|
|
|
};
|
|
|
|
|
2016-09-13 04:19:36 +08:00
|
|
|
static void guc_interrupts_release(struct drm_i915_private *dev_priv)
|
2015-08-12 22:43:42 +08:00
|
|
|
{
|
2016-03-16 19:00:36 +08:00
|
|
|
struct intel_engine_cs *engine;
|
drm/i915: Allocate intel_engine_cs structure only for the enabled engines
With the possibility of addition of many more number of rings in future,
the drm_i915_private structure could bloat as an array, of type
intel_engine_cs, is embedded inside it.
struct intel_engine_cs engine[I915_NUM_ENGINES];
Though this is still fine as generally there is only a single instance of
drm_i915_private structure used, but not all of the possible rings would be
enabled or active on most of the platforms. Some memory can be saved by
allocating intel_engine_cs structure only for the enabled/active engines.
Currently the engine/ring ID is kept static and dev_priv->engine[] is simply
indexed using the enums defined in intel_engine_id.
To save memory and continue using the static engine/ring IDs, 'engine' is
defined as an array of pointers.
struct intel_engine_cs *engine[I915_NUM_ENGINES];
dev_priv->engine[engine_ID] will be NULL for disabled engine instances.
There is a text size reduction of 928 bytes, from 1028200 to 1027272, for
i915.o file (but for i915.ko file text size remain same as 1193131 bytes).
v2:
- Remove the engine iterator field added in drm_i915_private structure,
instead pass a local iterator variable to the for_each_engine**
macros. (Chris)
- Do away with intel_engine_initialized() and instead directly use the
NULL pointer check on engine pointer. (Chris)
v3:
- Remove for_each_engine_id() macro, as the updated macro for_each_engine()
can be used in place of it. (Chris)
- Protect the access to Render engine Fault register with a NULL check, as
engine specific init is done later in Driver load sequence.
v4:
- Use !!dev_priv->engine[VCS] style for the engine check in getparam. (Chris)
- Kill the superfluous init_engine_lists().
v5:
- Cleanup the intel_engines_init() & intel_engines_setup(), with respect to
allocation of intel_engine_cs structure. (Chris)
v6:
- Rebase.
v7:
- Optimize the for_each_engine_masked() macro. (Chris)
- Change the type of 'iter' local variable to enum intel_engine_id. (Chris)
- Rebase.
v8: Rebase.
v9: Rebase.
v10:
- For index calculation use engine ID instead of pointer based arithmetic in
intel_engine_sync_index() as engine pointers are not contiguous now (Chris)
- For appropriateness, rename local enum variable 'iter' to 'id'. (Joonas)
- Use for_each_engine macro for cleanup in intel_engines_init() and remove
check for NULL engine pointer in cleanup() routines. (Joonas)
v11: Rebase.
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Akash Goel <akash.goel@intel.com>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/1476378888-7372-1-git-send-email-akash.goel@intel.com
2016-10-14 01:14:48 +08:00
|
|
|
enum intel_engine_id id;
|
2016-03-24 19:20:38 +08:00
|
|
|
int irqs;
|
2015-08-12 22:43:42 +08:00
|
|
|
|
drm/i915/guc: don't ever forward VBlank to the GuC
If a context waiting for VBlank were switched out, switching
in the next context and generating a CSB event in the process,
then the GuC would have to put the context back in the queue,
and then observe the subsequent VBlank interrupt so that it
could resubmit the suspended context.
However, we always set the CTX_CTRL_INHIBIT_SYN_CTX_SWITCH bit
in the RING_CONTEXT_CONTROL register, so this case cannot occur.
Furthermore we don't use the GuC's internal scheduler or allow
it to auto-resubmit workloads. Consequently, the GuC doesn't
need to see VBlanks, and by sending them to it we may be waking
it up unnecessarily, which might reduce RC6 residency and
increase power consumption.
So this patch removes the setting of the GFC_FORWARD_VBLANK
field from the code that diverts interrupts towards the GuC.
(The code to direct interrupts to the host, OTOH, continues to
explicitly set the field to "never send VBlanks to the GuC".)
v3:
Remove the line of code completely (original set the field
to ALWAYS forward, v1 changed it to CONDITIONAL forwarding,
v2 explicitly set it to NEVER, v3 just doesn't touch it at
all, as we know it's already set to NEVER).
Signed-off-by: Dave Gordon <david.s.gordon@intel.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch> (previous version)
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/1466780277-23435-1-git-send-email-david.s.gordon@intel.com
2016-06-24 22:57:57 +08:00
|
|
|
/* tell all command streamers NOT to forward interrupts or vblank to GuC */
|
2015-08-12 22:43:42 +08:00
|
|
|
irqs = _MASKED_FIELD(GFX_FORWARD_VBLANK_MASK, GFX_FORWARD_VBLANK_NEVER);
|
|
|
|
irqs |= _MASKED_BIT_DISABLE(GFX_INTERRUPT_STEERING);
|
drm/i915: Allocate intel_engine_cs structure only for the enabled engines
With the possibility of addition of many more number of rings in future,
the drm_i915_private structure could bloat as an array, of type
intel_engine_cs, is embedded inside it.
struct intel_engine_cs engine[I915_NUM_ENGINES];
Though this is still fine as generally there is only a single instance of
drm_i915_private structure used, but not all of the possible rings would be
enabled or active on most of the platforms. Some memory can be saved by
allocating intel_engine_cs structure only for the enabled/active engines.
Currently the engine/ring ID is kept static and dev_priv->engine[] is simply
indexed using the enums defined in intel_engine_id.
To save memory and continue using the static engine/ring IDs, 'engine' is
defined as an array of pointers.
struct intel_engine_cs *engine[I915_NUM_ENGINES];
dev_priv->engine[engine_ID] will be NULL for disabled engine instances.
There is a text size reduction of 928 bytes, from 1028200 to 1027272, for
i915.o file (but for i915.ko file text size remain same as 1193131 bytes).
v2:
- Remove the engine iterator field added in drm_i915_private structure,
instead pass a local iterator variable to the for_each_engine**
macros. (Chris)
- Do away with intel_engine_initialized() and instead directly use the
NULL pointer check on engine pointer. (Chris)
v3:
- Remove for_each_engine_id() macro, as the updated macro for_each_engine()
can be used in place of it. (Chris)
- Protect the access to Render engine Fault register with a NULL check, as
engine specific init is done later in Driver load sequence.
v4:
- Use !!dev_priv->engine[VCS] style for the engine check in getparam. (Chris)
- Kill the superfluous init_engine_lists().
v5:
- Cleanup the intel_engines_init() & intel_engines_setup(), with respect to
allocation of intel_engine_cs structure. (Chris)
v6:
- Rebase.
v7:
- Optimize the for_each_engine_masked() macro. (Chris)
- Change the type of 'iter' local variable to enum intel_engine_id. (Chris)
- Rebase.
v8: Rebase.
v9: Rebase.
v10:
- For index calculation use engine ID instead of pointer based arithmetic in
intel_engine_sync_index() as engine pointers are not contiguous now (Chris)
- For appropriateness, rename local enum variable 'iter' to 'id'. (Joonas)
- Use for_each_engine macro for cleanup in intel_engines_init() and remove
check for NULL engine pointer in cleanup() routines. (Joonas)
v11: Rebase.
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Akash Goel <akash.goel@intel.com>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/1476378888-7372-1-git-send-email-akash.goel@intel.com
2016-10-14 01:14:48 +08:00
|
|
|
for_each_engine(engine, dev_priv, id)
|
2016-03-16 19:00:36 +08:00
|
|
|
I915_WRITE(RING_MODE_GEN7(engine), irqs);
|
2015-08-12 22:43:42 +08:00
|
|
|
|
|
|
|
/* route all GT interrupts to the host */
|
|
|
|
I915_WRITE(GUC_BCS_RCS_IER, 0);
|
|
|
|
I915_WRITE(GUC_VCS2_VCS1_IER, 0);
|
|
|
|
I915_WRITE(GUC_WD_VECS_IER, 0);
|
|
|
|
}
|
|
|
|
|
2016-09-13 04:19:36 +08:00
|
|
|
static void guc_interrupts_capture(struct drm_i915_private *dev_priv)
|
2015-08-12 22:43:42 +08:00
|
|
|
{
|
2016-03-16 19:00:36 +08:00
|
|
|
struct intel_engine_cs *engine;
|
drm/i915: Allocate intel_engine_cs structure only for the enabled engines
With the possibility of addition of many more number of rings in future,
the drm_i915_private structure could bloat as an array, of type
intel_engine_cs, is embedded inside it.
struct intel_engine_cs engine[I915_NUM_ENGINES];
Though this is still fine as generally there is only a single instance of
drm_i915_private structure used, but not all of the possible rings would be
enabled or active on most of the platforms. Some memory can be saved by
allocating intel_engine_cs structure only for the enabled/active engines.
Currently the engine/ring ID is kept static and dev_priv->engine[] is simply
indexed using the enums defined in intel_engine_id.
To save memory and continue using the static engine/ring IDs, 'engine' is
defined as an array of pointers.
struct intel_engine_cs *engine[I915_NUM_ENGINES];
dev_priv->engine[engine_ID] will be NULL for disabled engine instances.
There is a text size reduction of 928 bytes, from 1028200 to 1027272, for
i915.o file (but for i915.ko file text size remain same as 1193131 bytes).
v2:
- Remove the engine iterator field added in drm_i915_private structure,
instead pass a local iterator variable to the for_each_engine**
macros. (Chris)
- Do away with intel_engine_initialized() and instead directly use the
NULL pointer check on engine pointer. (Chris)
v3:
- Remove for_each_engine_id() macro, as the updated macro for_each_engine()
can be used in place of it. (Chris)
- Protect the access to Render engine Fault register with a NULL check, as
engine specific init is done later in Driver load sequence.
v4:
- Use !!dev_priv->engine[VCS] style for the engine check in getparam. (Chris)
- Kill the superfluous init_engine_lists().
v5:
- Cleanup the intel_engines_init() & intel_engines_setup(), with respect to
allocation of intel_engine_cs structure. (Chris)
v6:
- Rebase.
v7:
- Optimize the for_each_engine_masked() macro. (Chris)
- Change the type of 'iter' local variable to enum intel_engine_id. (Chris)
- Rebase.
v8: Rebase.
v9: Rebase.
v10:
- For index calculation use engine ID instead of pointer based arithmetic in
intel_engine_sync_index() as engine pointers are not contiguous now (Chris)
- For appropriateness, rename local enum variable 'iter' to 'id'. (Joonas)
- Use for_each_engine macro for cleanup in intel_engines_init() and remove
check for NULL engine pointer in cleanup() routines. (Joonas)
v11: Rebase.
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Akash Goel <akash.goel@intel.com>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/1476378888-7372-1-git-send-email-akash.goel@intel.com
2016-10-14 01:14:48 +08:00
|
|
|
enum intel_engine_id id;
|
2016-03-24 19:20:38 +08:00
|
|
|
int irqs;
|
2016-05-31 16:28:27 +08:00
|
|
|
u32 tmp;
|
2015-08-12 22:43:42 +08:00
|
|
|
|
drm/i915/guc: don't ever forward VBlank to the GuC
If a context waiting for VBlank were switched out, switching
in the next context and generating a CSB event in the process,
then the GuC would have to put the context back in the queue,
and then observe the subsequent VBlank interrupt so that it
could resubmit the suspended context.
However, we always set the CTX_CTRL_INHIBIT_SYN_CTX_SWITCH bit
in the RING_CONTEXT_CONTROL register, so this case cannot occur.
Furthermore we don't use the GuC's internal scheduler or allow
it to auto-resubmit workloads. Consequently, the GuC doesn't
need to see VBlanks, and by sending them to it we may be waking
it up unnecessarily, which might reduce RC6 residency and
increase power consumption.
So this patch removes the setting of the GFC_FORWARD_VBLANK
field from the code that diverts interrupts towards the GuC.
(The code to direct interrupts to the host, OTOH, continues to
explicitly set the field to "never send VBlanks to the GuC".)
v3:
Remove the line of code completely (original set the field
to ALWAYS forward, v1 changed it to CONDITIONAL forwarding,
v2 explicitly set it to NEVER, v3 just doesn't touch it at
all, as we know it's already set to NEVER).
Signed-off-by: Dave Gordon <david.s.gordon@intel.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch> (previous version)
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/1466780277-23435-1-git-send-email-david.s.gordon@intel.com
2016-06-24 22:57:57 +08:00
|
|
|
/* tell all command streamers to forward interrupts (but not vblank) to GuC */
|
|
|
|
irqs = _MASKED_BIT_ENABLE(GFX_INTERRUPT_STEERING);
|
drm/i915: Allocate intel_engine_cs structure only for the enabled engines
With the possibility of addition of many more number of rings in future,
the drm_i915_private structure could bloat as an array, of type
intel_engine_cs, is embedded inside it.
struct intel_engine_cs engine[I915_NUM_ENGINES];
Though this is still fine as generally there is only a single instance of
drm_i915_private structure used, but not all of the possible rings would be
enabled or active on most of the platforms. Some memory can be saved by
allocating intel_engine_cs structure only for the enabled/active engines.
Currently the engine/ring ID is kept static and dev_priv->engine[] is simply
indexed using the enums defined in intel_engine_id.
To save memory and continue using the static engine/ring IDs, 'engine' is
defined as an array of pointers.
struct intel_engine_cs *engine[I915_NUM_ENGINES];
dev_priv->engine[engine_ID] will be NULL for disabled engine instances.
There is a text size reduction of 928 bytes, from 1028200 to 1027272, for
i915.o file (but for i915.ko file text size remain same as 1193131 bytes).
v2:
- Remove the engine iterator field added in drm_i915_private structure,
instead pass a local iterator variable to the for_each_engine**
macros. (Chris)
- Do away with intel_engine_initialized() and instead directly use the
NULL pointer check on engine pointer. (Chris)
v3:
- Remove for_each_engine_id() macro, as the updated macro for_each_engine()
can be used in place of it. (Chris)
- Protect the access to Render engine Fault register with a NULL check, as
engine specific init is done later in Driver load sequence.
v4:
- Use !!dev_priv->engine[VCS] style for the engine check in getparam. (Chris)
- Kill the superfluous init_engine_lists().
v5:
- Cleanup the intel_engines_init() & intel_engines_setup(), with respect to
allocation of intel_engine_cs structure. (Chris)
v6:
- Rebase.
v7:
- Optimize the for_each_engine_masked() macro. (Chris)
- Change the type of 'iter' local variable to enum intel_engine_id. (Chris)
- Rebase.
v8: Rebase.
v9: Rebase.
v10:
- For index calculation use engine ID instead of pointer based arithmetic in
intel_engine_sync_index() as engine pointers are not contiguous now (Chris)
- For appropriateness, rename local enum variable 'iter' to 'id'. (Joonas)
- Use for_each_engine macro for cleanup in intel_engines_init() and remove
check for NULL engine pointer in cleanup() routines. (Joonas)
v11: Rebase.
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Akash Goel <akash.goel@intel.com>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/1476378888-7372-1-git-send-email-akash.goel@intel.com
2016-10-14 01:14:48 +08:00
|
|
|
for_each_engine(engine, dev_priv, id)
|
2016-03-16 19:00:36 +08:00
|
|
|
I915_WRITE(RING_MODE_GEN7(engine), irqs);
|
2015-08-12 22:43:42 +08:00
|
|
|
|
|
|
|
/* route USER_INTERRUPT to Host, all others are sent to GuC. */
|
|
|
|
irqs = GT_RENDER_USER_INTERRUPT << GEN8_RCS_IRQ_SHIFT |
|
|
|
|
GT_RENDER_USER_INTERRUPT << GEN8_BCS_IRQ_SHIFT;
|
|
|
|
/* These three registers have the same bit definitions */
|
|
|
|
I915_WRITE(GUC_BCS_RCS_IER, ~irqs);
|
|
|
|
I915_WRITE(GUC_VCS2_VCS1_IER, ~irqs);
|
|
|
|
I915_WRITE(GUC_WD_VECS_IER, ~irqs);
|
2016-05-31 16:28:27 +08:00
|
|
|
|
|
|
|
/*
|
2016-09-13 04:19:35 +08:00
|
|
|
* The REDIRECT_TO_GUC bit of the PMINTRMSK register directs all
|
|
|
|
* (unmasked) PM interrupts to the GuC. All other bits of this
|
|
|
|
* register *disable* generation of a specific interrupt.
|
|
|
|
*
|
|
|
|
* 'pm_intr_keep' indicates bits that are NOT to be set when
|
|
|
|
* writing to the PM interrupt mask register, i.e. interrupts
|
|
|
|
* that must not be disabled.
|
|
|
|
*
|
|
|
|
* If the GuC is handling these interrupts, then we must not let
|
|
|
|
* the PM code disable ANY interrupt that the GuC is expecting.
|
|
|
|
* So for each ENABLED (0) bit in this register, we must SET the
|
|
|
|
* bit in pm_intr_keep so that it's left enabled for the GuC.
|
|
|
|
*
|
|
|
|
* OTOH the REDIRECT_TO_GUC bit is initially SET in pm_intr_keep
|
|
|
|
* (so interrupts go to the DISPLAY unit at first); but here we
|
|
|
|
* need to CLEAR that bit, which will result in the register bit
|
|
|
|
* being left SET!
|
|
|
|
*/
|
2016-05-31 16:28:27 +08:00
|
|
|
tmp = I915_READ(GEN6_PMINTRMSK);
|
2016-09-13 04:19:35 +08:00
|
|
|
if (tmp & GEN8_PMINTR_REDIRECT_TO_GUC) {
|
|
|
|
dev_priv->rps.pm_intr_keep |= ~tmp;
|
|
|
|
dev_priv->rps.pm_intr_keep &= ~GEN8_PMINTR_REDIRECT_TO_GUC;
|
2016-05-31 16:28:27 +08:00
|
|
|
}
|
2015-08-12 22:43:42 +08:00
|
|
|
}
|
|
|
|
|
2015-08-12 22:43:36 +08:00
|
|
|
static u32 get_gttype(struct drm_i915_private *dev_priv)
|
|
|
|
{
|
|
|
|
/* XXX: GT type based on PCI device ID? field seems unused by fw */
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static u32 get_core_family(struct drm_i915_private *dev_priv)
|
|
|
|
{
|
2016-08-19 01:17:24 +08:00
|
|
|
u32 gen = INTEL_GEN(dev_priv);
|
|
|
|
|
|
|
|
switch (gen) {
|
2015-08-12 22:43:36 +08:00
|
|
|
case 9:
|
|
|
|
return GFXCORE_FAMILY_GEN9;
|
|
|
|
|
|
|
|
default:
|
2016-08-19 01:17:24 +08:00
|
|
|
WARN(1, "GEN%d does not support GuC operation!\n", gen);
|
2015-08-12 22:43:36 +08:00
|
|
|
return GFXCORE_FAMILY_UNKNOWN;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2016-09-13 04:19:36 +08:00
|
|
|
/*
|
|
|
|
* Initialise the GuC parameter block before starting the firmware
|
|
|
|
* transfer. These parameters are read by the firmware on startup
|
|
|
|
* and cannot be changed thereafter.
|
|
|
|
*/
|
|
|
|
static void guc_params_init(struct drm_i915_private *dev_priv)
|
2015-08-12 22:43:36 +08:00
|
|
|
{
|
|
|
|
struct intel_guc *guc = &dev_priv->guc;
|
|
|
|
u32 params[GUC_CTL_MAX_DWORDS];
|
|
|
|
int i;
|
|
|
|
|
|
|
|
memset(¶ms, 0, sizeof(params));
|
|
|
|
|
|
|
|
params[GUC_CTL_DEVICE_INFO] |=
|
|
|
|
(get_gttype(dev_priv) << GUC_CTL_GTTYPE_SHIFT) |
|
|
|
|
(get_core_family(dev_priv) << GUC_CTL_COREFAMILY_SHIFT);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* GuC ARAT increment is 10 ns. GuC default scheduler quantum is one
|
|
|
|
* second. This ARAR is calculated by:
|
|
|
|
* Scheduler-Quantum-in-ns / ARAT-increment-in-ns = 1000000000 / 10
|
|
|
|
*/
|
|
|
|
params[GUC_CTL_ARAT_HIGH] = 0;
|
|
|
|
params[GUC_CTL_ARAT_LOW] = 100000000;
|
|
|
|
|
|
|
|
params[GUC_CTL_WA] |= GUC_CTL_WA_UK_BY_DRIVER;
|
|
|
|
|
|
|
|
params[GUC_CTL_FEATURE] |= GUC_CTL_DISABLE_SCHEDULER |
|
|
|
|
GUC_CTL_VCS2_ENABLED;
|
|
|
|
|
2016-10-13 00:24:29 +08:00
|
|
|
params[GUC_CTL_LOG_PARAMS] = guc->log.flags;
|
2016-10-13 00:24:27 +08:00
|
|
|
|
2015-08-12 22:43:36 +08:00
|
|
|
if (i915.guc_log_level >= 0) {
|
|
|
|
params[GUC_CTL_DEBUG] =
|
|
|
|
i915.guc_log_level << GUC_LOG_VERBOSITY_SHIFT;
|
2016-10-13 00:24:27 +08:00
|
|
|
} else
|
|
|
|
params[GUC_CTL_DEBUG] = GUC_LOG_DISABLED;
|
2015-08-12 22:43:36 +08:00
|
|
|
|
2016-08-15 17:48:51 +08:00
|
|
|
if (guc->ads_vma) {
|
2016-08-15 17:49:07 +08:00
|
|
|
u32 ads = i915_ggtt_offset(guc->ads_vma) >> PAGE_SHIFT;
|
2015-12-19 04:00:12 +08:00
|
|
|
params[GUC_CTL_DEBUG] |= ads << GUC_ADS_ADDR_SHIFT;
|
|
|
|
params[GUC_CTL_DEBUG] |= GUC_ADS_ENABLED;
|
|
|
|
}
|
|
|
|
|
2015-08-12 22:43:39 +08:00
|
|
|
/* If GuC submission is enabled, set up additional parameters here */
|
|
|
|
if (i915.enable_guc_submission) {
|
2016-08-15 17:49:07 +08:00
|
|
|
u32 pgs = i915_ggtt_offset(dev_priv->guc.ctx_pool_vma);
|
2015-08-12 22:43:39 +08:00
|
|
|
u32 ctx_in_16 = GUC_MAX_GPU_CONTEXTS / 16;
|
|
|
|
|
|
|
|
pgs >>= PAGE_SHIFT;
|
|
|
|
params[GUC_CTL_CTXINFO] = (pgs << GUC_CTL_BASE_ADDR_SHIFT) |
|
|
|
|
(ctx_in_16 << GUC_CTL_CTXNUM_IN16_SHIFT);
|
|
|
|
|
|
|
|
params[GUC_CTL_FEATURE] |= GUC_CTL_KERNEL_SUBMISSIONS;
|
|
|
|
|
|
|
|
/* Unmask this bit to enable the GuC's internal scheduler */
|
|
|
|
params[GUC_CTL_FEATURE] &= ~GUC_CTL_DISABLE_SCHEDULER;
|
|
|
|
}
|
|
|
|
|
2015-08-12 22:43:36 +08:00
|
|
|
I915_WRITE(SOFT_SCRATCH(0), 0);
|
|
|
|
|
|
|
|
for (i = 0; i < GUC_CTL_MAX_DWORDS; i++)
|
|
|
|
I915_WRITE(SOFT_SCRATCH(1 + i), params[i]);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Read the GuC status register (GUC_STATUS) and store it in the
|
|
|
|
* specified location; then return a boolean indicating whether
|
|
|
|
* the value matches either of two values representing completion
|
|
|
|
* of the GuC boot process.
|
|
|
|
*
|
2016-02-11 18:27:31 +08:00
|
|
|
* This is used for polling the GuC status in a wait_for()
|
2015-08-12 22:43:36 +08:00
|
|
|
* loop below.
|
|
|
|
*/
|
|
|
|
static inline bool guc_ucode_response(struct drm_i915_private *dev_priv,
|
|
|
|
u32 *status)
|
|
|
|
{
|
|
|
|
u32 val = I915_READ(GUC_STATUS);
|
2015-09-23 04:48:40 +08:00
|
|
|
u32 uk_val = val & GS_UKERNEL_MASK;
|
2015-08-12 22:43:36 +08:00
|
|
|
*status = val;
|
2015-09-23 04:48:40 +08:00
|
|
|
return (uk_val == GS_UKERNEL_READY ||
|
|
|
|
((val & GS_MIA_CORE_STATE) && uk_val == GS_UKERNEL_LAPIC_DONE));
|
2015-08-12 22:43:36 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Transfer the firmware image to RAM for execution by the microcontroller.
|
|
|
|
*
|
|
|
|
* Architecturally, the DMA engine is bidirectional, and can potentially even
|
|
|
|
* transfer between GTT locations. This functionality is left out of the API
|
|
|
|
* for now as there is no need for it.
|
|
|
|
*
|
|
|
|
* Note that GuC needs the CSS header plus uKernel code to be copied by the
|
|
|
|
* DMA engine in one operation, whereas the RSA signature is loaded via MMIO.
|
|
|
|
*/
|
2016-08-15 17:49:06 +08:00
|
|
|
static int guc_ucode_xfer_dma(struct drm_i915_private *dev_priv,
|
|
|
|
struct i915_vma *vma)
|
2015-08-12 22:43:36 +08:00
|
|
|
{
|
|
|
|
struct intel_guc_fw *guc_fw = &dev_priv->guc.guc_fw;
|
|
|
|
unsigned long offset;
|
2016-08-15 17:49:06 +08:00
|
|
|
struct sg_table *sg = vma->pages;
|
2015-10-20 07:10:54 +08:00
|
|
|
u32 status, rsa[UOS_RSA_SCRATCH_MAX_COUNT];
|
2015-08-12 22:43:36 +08:00
|
|
|
int i, ret = 0;
|
|
|
|
|
2015-10-20 07:10:54 +08:00
|
|
|
/* where RSA signature starts */
|
|
|
|
offset = guc_fw->rsa_offset;
|
2015-08-12 22:43:36 +08:00
|
|
|
|
|
|
|
/* Copy RSA signature from the fw image to HW for verification */
|
2015-10-20 07:10:54 +08:00
|
|
|
sg_pcopy_to_buffer(sg->sgl, sg->nents, rsa, sizeof(rsa), offset);
|
|
|
|
for (i = 0; i < UOS_RSA_SCRATCH_MAX_COUNT; i++)
|
2015-09-19 01:03:24 +08:00
|
|
|
I915_WRITE(UOS_RSA_SCRATCH(i), rsa[i]);
|
2015-08-12 22:43:36 +08:00
|
|
|
|
2015-10-20 07:10:54 +08:00
|
|
|
/* The header plus uCode will be copied to WOPCM via DMA, excluding any
|
|
|
|
* other components */
|
|
|
|
I915_WRITE(DMA_COPY_SIZE, guc_fw->header_size + guc_fw->ucode_size);
|
|
|
|
|
2015-08-12 22:43:36 +08:00
|
|
|
/* Set the source address for the new blob */
|
2016-08-15 17:49:07 +08:00
|
|
|
offset = i915_ggtt_offset(vma) + guc_fw->header_offset;
|
2015-08-12 22:43:36 +08:00
|
|
|
I915_WRITE(DMA_ADDR_0_LOW, lower_32_bits(offset));
|
|
|
|
I915_WRITE(DMA_ADDR_0_HIGH, upper_32_bits(offset) & 0xFFFF);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Set the DMA destination. Current uCode expects the code to be
|
|
|
|
* loaded at 8k; locations below this are used for the stack.
|
|
|
|
*/
|
|
|
|
I915_WRITE(DMA_ADDR_1_LOW, 0x2000);
|
|
|
|
I915_WRITE(DMA_ADDR_1_HIGH, DMA_ADDRESS_SPACE_WOPCM);
|
|
|
|
|
|
|
|
/* Finally start the DMA */
|
|
|
|
I915_WRITE(DMA_CTRL, _MASKED_BIT_ENABLE(UOS_MOVE | START_DMA));
|
|
|
|
|
|
|
|
/*
|
2016-02-11 18:27:31 +08:00
|
|
|
* Wait for the DMA to complete & the GuC to start up.
|
2015-08-12 22:43:36 +08:00
|
|
|
* NB: Docs recommend not using the interrupt for completion.
|
|
|
|
* Measurements indicate this should take no more than 20ms, so a
|
|
|
|
* timeout here indicates that the GuC has failed and is unusable.
|
|
|
|
* (Higher levels of the driver will attempt to fall back to
|
|
|
|
* execlist mode if this happens.)
|
|
|
|
*/
|
2016-02-11 18:27:31 +08:00
|
|
|
ret = wait_for(guc_ucode_response(dev_priv, &status), 100);
|
2015-08-12 22:43:36 +08:00
|
|
|
|
|
|
|
DRM_DEBUG_DRIVER("DMA status 0x%x, GuC status 0x%x\n",
|
|
|
|
I915_READ(DMA_CTRL), status);
|
|
|
|
|
|
|
|
if ((status & GS_BOOTROM_MASK) == GS_BOOTROM_RSA_FAILED) {
|
|
|
|
DRM_ERROR("GuC firmware signature verification failed\n");
|
|
|
|
ret = -ENOEXEC;
|
|
|
|
}
|
|
|
|
|
|
|
|
DRM_DEBUG_DRIVER("returning %d\n", ret);
|
|
|
|
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2016-05-17 22:12:45 +08:00
|
|
|
static u32 guc_wopcm_size(struct drm_i915_private *dev_priv)
|
|
|
|
{
|
|
|
|
u32 wopcm_size = GUC_WOPCM_TOP;
|
|
|
|
|
|
|
|
/* On BXT, the top of WOPCM is reserved for RC6 context */
|
|
|
|
if (IS_BROXTON(dev_priv))
|
|
|
|
wopcm_size -= BXT_GUC_WOPCM_RC6_RESERVED;
|
|
|
|
|
|
|
|
return wopcm_size;
|
|
|
|
}
|
|
|
|
|
2015-08-12 22:43:36 +08:00
|
|
|
/*
|
|
|
|
* Load the GuC firmware blob into the MinuteIA.
|
|
|
|
*/
|
|
|
|
static int guc_ucode_xfer(struct drm_i915_private *dev_priv)
|
|
|
|
{
|
|
|
|
struct intel_guc_fw *guc_fw = &dev_priv->guc.guc_fw;
|
2016-08-15 17:49:06 +08:00
|
|
|
struct i915_vma *vma;
|
2015-08-12 22:43:36 +08:00
|
|
|
int ret;
|
|
|
|
|
|
|
|
ret = i915_gem_object_set_to_gtt_domain(guc_fw->guc_fw_obj, false);
|
|
|
|
if (ret) {
|
|
|
|
DRM_DEBUG_DRIVER("set-domain failed %d\n", ret);
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2016-08-15 17:49:06 +08:00
|
|
|
vma = i915_gem_object_ggtt_pin(guc_fw->guc_fw_obj, NULL, 0, 0, 0);
|
|
|
|
if (IS_ERR(vma)) {
|
|
|
|
DRM_DEBUG_DRIVER("pin failed %d\n", (int)PTR_ERR(vma));
|
|
|
|
return PTR_ERR(vma);
|
2015-08-12 22:43:36 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
/* Invalidate GuC TLB to let GuC take the latest updates to GTT. */
|
|
|
|
I915_WRITE(GEN8_GTCR, GEN8_GTCR_INVALIDATE);
|
|
|
|
|
|
|
|
intel_uncore_forcewake_get(dev_priv, FORCEWAKE_ALL);
|
|
|
|
|
|
|
|
/* init WOPCM */
|
2016-05-17 22:12:45 +08:00
|
|
|
I915_WRITE(GUC_WOPCM_SIZE, guc_wopcm_size(dev_priv));
|
2015-08-12 22:43:36 +08:00
|
|
|
I915_WRITE(DMA_GUC_WOPCM_OFFSET, GUC_WOPCM_OFFSET_VALUE);
|
|
|
|
|
|
|
|
/* Enable MIA caching. GuC clock gating is disabled. */
|
|
|
|
I915_WRITE(GUC_SHIM_CONTROL, GUC_SHIM_CONTROL_VALUE);
|
|
|
|
|
2016-09-16 21:59:44 +08:00
|
|
|
/* WaDisableMinuteIaClockGating:bxt */
|
2016-10-13 18:03:04 +08:00
|
|
|
if (IS_BXT_REVID(dev_priv, 0, BXT_REVID_A1)) {
|
2015-09-08 17:31:53 +08:00
|
|
|
I915_WRITE(GUC_SHIM_CONTROL, (I915_READ(GUC_SHIM_CONTROL) &
|
|
|
|
~GUC_ENABLE_MIA_CLOCK_GATING));
|
|
|
|
}
|
|
|
|
|
2016-09-26 20:07:51 +08:00
|
|
|
/* WaC6DisallowByGfxPause:bxt */
|
2016-10-13 18:03:04 +08:00
|
|
|
if (IS_BXT_REVID(dev_priv, 0, BXT_REVID_B0))
|
2016-07-20 18:00:25 +08:00
|
|
|
I915_WRITE(GEN6_GFXPAUSE, 0x30FFF);
|
2015-08-12 22:43:36 +08:00
|
|
|
|
2016-10-13 18:03:04 +08:00
|
|
|
if (IS_BROXTON(dev_priv))
|
2015-08-12 22:43:36 +08:00
|
|
|
I915_WRITE(GEN9LP_GT_PM_CONFIG, GT_DOORBELL_ENABLE);
|
|
|
|
else
|
|
|
|
I915_WRITE(GEN9_GT_PM_CONFIG, GT_DOORBELL_ENABLE);
|
|
|
|
|
2016-10-13 18:03:10 +08:00
|
|
|
if (IS_GEN9(dev_priv)) {
|
2015-08-12 22:43:36 +08:00
|
|
|
/* DOP Clock Gating Enable for GuC clocks */
|
|
|
|
I915_WRITE(GEN7_MISCCPCTL, (GEN8_DOP_CLOCK_GATE_GUC_ENABLE |
|
|
|
|
I915_READ(GEN7_MISCCPCTL)));
|
|
|
|
|
2016-09-13 04:19:36 +08:00
|
|
|
/* allows for 5us (in 10ns units) before GT can go to RC6 */
|
2015-08-12 22:43:36 +08:00
|
|
|
I915_WRITE(GUC_ARAT_C6DIS, 0x1FF);
|
|
|
|
}
|
|
|
|
|
2016-09-13 04:19:36 +08:00
|
|
|
guc_params_init(dev_priv);
|
2015-08-12 22:43:36 +08:00
|
|
|
|
2016-08-15 17:49:06 +08:00
|
|
|
ret = guc_ucode_xfer_dma(dev_priv, vma);
|
2015-08-12 22:43:36 +08:00
|
|
|
|
|
|
|
intel_uncore_forcewake_put(dev_priv, FORCEWAKE_ALL);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* We keep the object pages for reuse during resume. But we can unpin it
|
|
|
|
* now that DMA has completed, so it doesn't continue to take up space.
|
|
|
|
*/
|
2016-08-15 17:49:06 +08:00
|
|
|
i915_vma_unpin(vma);
|
2015-08-12 22:43:36 +08:00
|
|
|
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2016-09-13 04:19:36 +08:00
|
|
|
static int guc_hw_reset(struct drm_i915_private *dev_priv)
|
2016-04-05 01:50:56 +08:00
|
|
|
{
|
|
|
|
int ret;
|
|
|
|
u32 guc_status;
|
|
|
|
|
|
|
|
ret = intel_guc_reset(dev_priv);
|
|
|
|
if (ret) {
|
|
|
|
DRM_ERROR("GuC reset failed, ret = %d\n", ret);
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
guc_status = I915_READ(GUC_STATUS);
|
|
|
|
WARN(!(guc_status & GS_MIA_IN_RESET),
|
|
|
|
"GuC status: 0x%x, MIA core expected to be in reset\n", guc_status);
|
|
|
|
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2015-08-12 22:43:36 +08:00
|
|
|
/**
|
2016-05-13 22:36:29 +08:00
|
|
|
* intel_guc_setup() - finish preparing the GuC for activity
|
2015-08-12 22:43:36 +08:00
|
|
|
* @dev: drm device
|
|
|
|
*
|
|
|
|
* Called from gem_init_hw() during driver loading and also after a GPU reset.
|
|
|
|
*
|
2016-05-13 22:36:29 +08:00
|
|
|
* The main action required here it to load the GuC uCode into the device.
|
2015-08-12 22:43:36 +08:00
|
|
|
* The firmware image should have already been fetched into memory by the
|
2016-05-13 22:36:29 +08:00
|
|
|
* earlier call to intel_guc_init(), so here we need only check that worked,
|
|
|
|
* and then transfer the image to the h/w.
|
2015-08-12 22:43:36 +08:00
|
|
|
*
|
|
|
|
* Return: non-zero code on error
|
|
|
|
*/
|
2016-05-13 22:36:29 +08:00
|
|
|
int intel_guc_setup(struct drm_device *dev)
|
2015-08-12 22:43:36 +08:00
|
|
|
{
|
2016-07-04 18:34:36 +08:00
|
|
|
struct drm_i915_private *dev_priv = to_i915(dev);
|
2015-08-12 22:43:36 +08:00
|
|
|
struct intel_guc_fw *guc_fw = &dev_priv->guc.guc_fw;
|
2016-05-20 18:42:42 +08:00
|
|
|
const char *fw_path = guc_fw->guc_fw_path;
|
|
|
|
int retries, ret, err;
|
2015-08-12 22:43:36 +08:00
|
|
|
|
2016-05-20 18:42:42 +08:00
|
|
|
DRM_DEBUG_DRIVER("GuC fw status: path %s, fetch %s, load %s\n",
|
|
|
|
fw_path,
|
2015-08-12 22:43:36 +08:00
|
|
|
intel_guc_fw_status_repr(guc_fw->guc_fw_fetch_status),
|
|
|
|
intel_guc_fw_status_repr(guc_fw->guc_fw_load_status));
|
|
|
|
|
2016-05-20 18:42:42 +08:00
|
|
|
/* Loading forbidden, or no firmware to load? */
|
|
|
|
if (!i915.enable_guc_loading) {
|
|
|
|
err = 0;
|
|
|
|
goto fail;
|
2016-06-07 16:14:49 +08:00
|
|
|
} else if (fw_path == NULL) {
|
|
|
|
/* Device is known to have no uCode (e.g. no GuC) */
|
|
|
|
err = -ENXIO;
|
|
|
|
goto fail;
|
|
|
|
} else if (*fw_path == '\0') {
|
|
|
|
/* Device has a GuC but we don't know what f/w to load? */
|
2016-08-19 01:17:24 +08:00
|
|
|
WARN(1, "No GuC firmware known for this platform!\n");
|
2016-05-20 18:42:42 +08:00
|
|
|
err = -ENODEV;
|
|
|
|
goto fail;
|
|
|
|
}
|
2015-08-12 22:43:36 +08:00
|
|
|
|
2016-05-20 18:42:42 +08:00
|
|
|
/* Fetch failed, or already fetched but failed to load? */
|
|
|
|
if (guc_fw->guc_fw_fetch_status != GUC_FIRMWARE_SUCCESS) {
|
2015-08-12 22:43:36 +08:00
|
|
|
err = -EIO;
|
|
|
|
goto fail;
|
2016-05-20 18:42:42 +08:00
|
|
|
} else if (guc_fw->guc_fw_load_status == GUC_FIRMWARE_FAIL) {
|
|
|
|
err = -ENOEXEC;
|
2015-08-12 22:43:36 +08:00
|
|
|
goto fail;
|
|
|
|
}
|
|
|
|
|
2016-09-13 04:19:36 +08:00
|
|
|
guc_interrupts_release(dev_priv);
|
2016-10-13 00:24:31 +08:00
|
|
|
gen9_reset_guc_interrupts(dev_priv);
|
2016-05-20 18:42:42 +08:00
|
|
|
|
|
|
|
guc_fw->guc_fw_load_status = GUC_FIRMWARE_PENDING;
|
|
|
|
|
|
|
|
DRM_DEBUG_DRIVER("GuC fw status: fetch %s, load %s\n",
|
|
|
|
intel_guc_fw_status_repr(guc_fw->guc_fw_fetch_status),
|
|
|
|
intel_guc_fw_status_repr(guc_fw->guc_fw_load_status));
|
|
|
|
|
2016-06-11 01:29:26 +08:00
|
|
|
err = i915_guc_submission_init(dev_priv);
|
2015-08-12 22:43:39 +08:00
|
|
|
if (err)
|
|
|
|
goto fail;
|
|
|
|
|
2016-04-05 01:50:56 +08:00
|
|
|
/*
|
|
|
|
* WaEnableuKernelHeaderValidFix:skl,bxt
|
|
|
|
* For BXT, this is only upto B0 but below WA is required for later
|
|
|
|
* steppings also so this is extended as well.
|
|
|
|
*/
|
|
|
|
/* WaEnableGuCBootHashCheckNotSet:skl,bxt */
|
2016-04-05 01:50:57 +08:00
|
|
|
for (retries = 3; ; ) {
|
|
|
|
/*
|
|
|
|
* Always reset the GuC just before (re)loading, so
|
|
|
|
* that the state and timing are fairly predictable
|
|
|
|
*/
|
2016-09-13 04:19:36 +08:00
|
|
|
err = guc_hw_reset(dev_priv);
|
2016-08-19 01:17:24 +08:00
|
|
|
if (err)
|
2016-04-05 01:50:56 +08:00
|
|
|
goto fail;
|
2016-04-05 01:50:57 +08:00
|
|
|
|
|
|
|
err = guc_ucode_xfer(dev_priv);
|
|
|
|
if (!err)
|
|
|
|
break;
|
|
|
|
|
|
|
|
if (--retries == 0)
|
|
|
|
goto fail;
|
|
|
|
|
2016-05-20 18:42:42 +08:00
|
|
|
DRM_INFO("GuC fw load failed: %d; will reset and "
|
|
|
|
"retry %d more time(s)\n", err, retries);
|
2016-04-05 01:50:56 +08:00
|
|
|
}
|
2015-08-12 22:43:36 +08:00
|
|
|
|
|
|
|
guc_fw->guc_fw_load_status = GUC_FIRMWARE_SUCCESS;
|
|
|
|
|
|
|
|
DRM_DEBUG_DRIVER("GuC fw status: fetch %s, load %s\n",
|
|
|
|
intel_guc_fw_status_repr(guc_fw->guc_fw_fetch_status),
|
|
|
|
intel_guc_fw_status_repr(guc_fw->guc_fw_load_status));
|
|
|
|
|
2015-08-12 22:43:41 +08:00
|
|
|
if (i915.enable_guc_submission) {
|
2016-10-13 00:24:31 +08:00
|
|
|
if (i915.guc_log_level >= 0)
|
|
|
|
gen9_enable_guc_interrupts(dev_priv);
|
|
|
|
|
2016-06-11 01:29:26 +08:00
|
|
|
err = i915_guc_submission_enable(dev_priv);
|
2015-08-12 22:43:41 +08:00
|
|
|
if (err)
|
|
|
|
goto fail;
|
2016-09-13 04:19:36 +08:00
|
|
|
guc_interrupts_capture(dev_priv);
|
2015-08-12 22:43:41 +08:00
|
|
|
}
|
|
|
|
|
2015-08-12 22:43:36 +08:00
|
|
|
return 0;
|
|
|
|
|
|
|
|
fail:
|
|
|
|
if (guc_fw->guc_fw_load_status == GUC_FIRMWARE_PENDING)
|
|
|
|
guc_fw->guc_fw_load_status = GUC_FIRMWARE_FAIL;
|
|
|
|
|
2016-09-13 04:19:36 +08:00
|
|
|
guc_interrupts_release(dev_priv);
|
2016-06-11 01:29:26 +08:00
|
|
|
i915_guc_submission_disable(dev_priv);
|
|
|
|
i915_guc_submission_fini(dev_priv);
|
2015-08-12 22:43:41 +08:00
|
|
|
|
2016-05-20 18:42:42 +08:00
|
|
|
/*
|
|
|
|
* We've failed to load the firmware :(
|
|
|
|
*
|
|
|
|
* Decide whether to disable GuC submission and fall back to
|
|
|
|
* execlist mode, and whether to hide the error by returning
|
|
|
|
* zero or to return -EIO, which the caller will treat as a
|
|
|
|
* nonfatal error (i.e. it doesn't prevent driver load, but
|
|
|
|
* marks the GPU as wedged until reset).
|
|
|
|
*/
|
|
|
|
if (i915.enable_guc_loading > 1) {
|
|
|
|
ret = -EIO;
|
|
|
|
} else if (i915.enable_guc_submission > 1) {
|
|
|
|
ret = -EIO;
|
|
|
|
} else {
|
|
|
|
ret = 0;
|
|
|
|
}
|
|
|
|
|
2016-06-11 00:21:25 +08:00
|
|
|
if (err == 0 && !HAS_GUC_UCODE(dev))
|
|
|
|
; /* Don't mention the GuC! */
|
|
|
|
else if (err == 0)
|
2016-05-20 18:42:42 +08:00
|
|
|
DRM_INFO("GuC firmware load skipped\n");
|
2016-06-11 00:21:25 +08:00
|
|
|
else if (ret != -EIO)
|
2016-08-19 01:17:24 +08:00
|
|
|
DRM_NOTE("GuC firmware load failed: %d\n", err);
|
2016-06-11 00:21:25 +08:00
|
|
|
else
|
2016-08-19 01:17:24 +08:00
|
|
|
DRM_WARN("GuC firmware load failed: %d\n", err);
|
2016-05-20 18:42:42 +08:00
|
|
|
|
|
|
|
if (i915.enable_guc_submission) {
|
|
|
|
if (fw_path == NULL)
|
|
|
|
DRM_INFO("GuC submission without firmware not supported\n");
|
|
|
|
if (ret == 0)
|
2016-08-19 01:17:24 +08:00
|
|
|
DRM_NOTE("Falling back from GuC submission to execlist mode\n");
|
2016-05-20 18:42:42 +08:00
|
|
|
else
|
|
|
|
DRM_ERROR("GuC init failed: %d\n", ret);
|
|
|
|
}
|
|
|
|
i915.enable_guc_submission = 0;
|
|
|
|
|
|
|
|
return ret;
|
2015-08-12 22:43:36 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
static void guc_fw_fetch(struct drm_device *dev, struct intel_guc_fw *guc_fw)
|
|
|
|
{
|
2016-08-22 18:32:44 +08:00
|
|
|
struct pci_dev *pdev = dev->pdev;
|
2015-08-12 22:43:36 +08:00
|
|
|
struct drm_i915_gem_object *obj;
|
|
|
|
const struct firmware *fw;
|
2015-10-20 07:10:54 +08:00
|
|
|
struct guc_css_header *css;
|
|
|
|
size_t size;
|
2015-08-12 22:43:36 +08:00
|
|
|
int err;
|
|
|
|
|
|
|
|
DRM_DEBUG_DRIVER("before requesting firmware: GuC fw fetch status %s\n",
|
|
|
|
intel_guc_fw_status_repr(guc_fw->guc_fw_fetch_status));
|
|
|
|
|
2016-08-22 18:32:44 +08:00
|
|
|
err = request_firmware(&fw, guc_fw->guc_fw_path, &pdev->dev);
|
2015-08-12 22:43:36 +08:00
|
|
|
if (err)
|
|
|
|
goto fail;
|
|
|
|
if (!fw)
|
|
|
|
goto fail;
|
|
|
|
|
|
|
|
DRM_DEBUG_DRIVER("fetch GuC fw from %s succeeded, fw %p\n",
|
|
|
|
guc_fw->guc_fw_path, fw);
|
|
|
|
|
2015-10-20 07:10:54 +08:00
|
|
|
/* Check the size of the blob before examining buffer contents */
|
|
|
|
if (fw->size < sizeof(struct guc_css_header)) {
|
2016-08-19 01:17:24 +08:00
|
|
|
DRM_NOTE("Firmware header is missing\n");
|
2015-08-12 22:43:36 +08:00
|
|
|
goto fail;
|
2015-10-20 07:10:54 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
css = (struct guc_css_header *)fw->data;
|
|
|
|
|
|
|
|
/* Firmware bits always start from header */
|
|
|
|
guc_fw->header_offset = 0;
|
|
|
|
guc_fw->header_size = (css->header_size_dw - css->modulus_size_dw -
|
|
|
|
css->key_size_dw - css->exponent_size_dw) * sizeof(u32);
|
|
|
|
|
|
|
|
if (guc_fw->header_size != sizeof(struct guc_css_header)) {
|
2016-08-19 01:17:24 +08:00
|
|
|
DRM_NOTE("CSS header definition mismatch\n");
|
2015-10-20 07:10:54 +08:00
|
|
|
goto fail;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* then, uCode */
|
|
|
|
guc_fw->ucode_offset = guc_fw->header_offset + guc_fw->header_size;
|
|
|
|
guc_fw->ucode_size = (css->size_dw - css->header_size_dw) * sizeof(u32);
|
|
|
|
|
|
|
|
/* now RSA */
|
|
|
|
if (css->key_size_dw != UOS_RSA_SCRATCH_MAX_COUNT) {
|
2016-08-19 01:17:24 +08:00
|
|
|
DRM_NOTE("RSA key size is bad\n");
|
2015-10-20 07:10:54 +08:00
|
|
|
goto fail;
|
|
|
|
}
|
|
|
|
guc_fw->rsa_offset = guc_fw->ucode_offset + guc_fw->ucode_size;
|
|
|
|
guc_fw->rsa_size = css->key_size_dw * sizeof(u32);
|
|
|
|
|
|
|
|
/* At least, it should have header, uCode and RSA. Size of all three. */
|
|
|
|
size = guc_fw->header_size + guc_fw->ucode_size + guc_fw->rsa_size;
|
|
|
|
if (fw->size < size) {
|
2016-08-19 01:17:24 +08:00
|
|
|
DRM_NOTE("Missing firmware components\n");
|
2015-10-20 07:10:54 +08:00
|
|
|
goto fail;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Header and uCode will be loaded to WOPCM. Size of the two. */
|
|
|
|
size = guc_fw->header_size + guc_fw->ucode_size;
|
2016-07-04 18:34:37 +08:00
|
|
|
if (size > guc_wopcm_size(to_i915(dev))) {
|
2016-08-19 01:17:24 +08:00
|
|
|
DRM_NOTE("Firmware is too large to fit in WOPCM\n");
|
2015-10-20 07:10:54 +08:00
|
|
|
goto fail;
|
|
|
|
}
|
2015-08-12 22:43:36 +08:00
|
|
|
|
|
|
|
/*
|
|
|
|
* The GuC firmware image has the version number embedded at a well-known
|
|
|
|
* offset within the firmware blob; note that major / minor version are
|
|
|
|
* TWO bytes each (i.e. u16), although all pointers and offsets are defined
|
|
|
|
* in terms of bytes (u8).
|
|
|
|
*/
|
2015-10-20 07:10:54 +08:00
|
|
|
guc_fw->guc_fw_major_found = css->guc_sw_version >> 16;
|
|
|
|
guc_fw->guc_fw_minor_found = css->guc_sw_version & 0xFFFF;
|
2015-08-12 22:43:36 +08:00
|
|
|
|
|
|
|
if (guc_fw->guc_fw_major_found != guc_fw->guc_fw_major_wanted ||
|
|
|
|
guc_fw->guc_fw_minor_found < guc_fw->guc_fw_minor_wanted) {
|
2016-08-19 01:17:24 +08:00
|
|
|
DRM_NOTE("GuC firmware version %d.%d, required %d.%d\n",
|
2015-08-12 22:43:36 +08:00
|
|
|
guc_fw->guc_fw_major_found, guc_fw->guc_fw_minor_found,
|
|
|
|
guc_fw->guc_fw_major_wanted, guc_fw->guc_fw_minor_wanted);
|
|
|
|
err = -ENOEXEC;
|
|
|
|
goto fail;
|
|
|
|
}
|
|
|
|
|
|
|
|
DRM_DEBUG_DRIVER("firmware version %d.%d OK (minimum %d.%d)\n",
|
|
|
|
guc_fw->guc_fw_major_found, guc_fw->guc_fw_minor_found,
|
|
|
|
guc_fw->guc_fw_major_wanted, guc_fw->guc_fw_minor_wanted);
|
|
|
|
|
2015-11-04 05:42:31 +08:00
|
|
|
mutex_lock(&dev->struct_mutex);
|
2015-08-12 22:43:36 +08:00
|
|
|
obj = i915_gem_object_create_from_data(dev, fw->data, fw->size);
|
2015-11-04 05:42:31 +08:00
|
|
|
mutex_unlock(&dev->struct_mutex);
|
2015-08-12 22:43:36 +08:00
|
|
|
if (IS_ERR_OR_NULL(obj)) {
|
|
|
|
err = obj ? PTR_ERR(obj) : -ENOMEM;
|
|
|
|
goto fail;
|
|
|
|
}
|
|
|
|
|
|
|
|
guc_fw->guc_fw_obj = obj;
|
|
|
|
guc_fw->guc_fw_size = fw->size;
|
|
|
|
|
|
|
|
DRM_DEBUG_DRIVER("GuC fw fetch status SUCCESS, obj %p\n",
|
|
|
|
guc_fw->guc_fw_obj);
|
|
|
|
|
|
|
|
release_firmware(fw);
|
|
|
|
guc_fw->guc_fw_fetch_status = GUC_FIRMWARE_SUCCESS;
|
|
|
|
return;
|
|
|
|
|
|
|
|
fail:
|
2016-08-19 01:17:24 +08:00
|
|
|
DRM_WARN("Failed to fetch valid GuC firmware from %s (error %d)\n",
|
|
|
|
guc_fw->guc_fw_path, err);
|
2015-08-12 22:43:36 +08:00
|
|
|
DRM_DEBUG_DRIVER("GuC fw fetch status FAIL; err %d, fw %p, obj %p\n",
|
|
|
|
err, fw, guc_fw->guc_fw_obj);
|
|
|
|
|
2016-01-14 03:01:50 +08:00
|
|
|
mutex_lock(&dev->struct_mutex);
|
2015-08-12 22:43:36 +08:00
|
|
|
obj = guc_fw->guc_fw_obj;
|
|
|
|
if (obj)
|
2016-07-20 20:31:53 +08:00
|
|
|
i915_gem_object_put(obj);
|
2015-08-12 22:43:36 +08:00
|
|
|
guc_fw->guc_fw_obj = NULL;
|
2016-01-14 03:01:50 +08:00
|
|
|
mutex_unlock(&dev->struct_mutex);
|
2015-08-12 22:43:36 +08:00
|
|
|
|
|
|
|
release_firmware(fw); /* OK even if fw is NULL */
|
|
|
|
guc_fw->guc_fw_fetch_status = GUC_FIRMWARE_FAIL;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
2016-05-13 22:36:29 +08:00
|
|
|
* intel_guc_init() - define parameters and fetch firmware
|
2015-08-12 22:43:36 +08:00
|
|
|
* @dev: drm device
|
|
|
|
*
|
|
|
|
* Called early during driver load, but after GEM is initialised.
|
|
|
|
*
|
|
|
|
* The firmware will be transferred to the GuC's memory later,
|
2016-05-13 22:36:29 +08:00
|
|
|
* when intel_guc_setup() is called.
|
2015-08-12 22:43:36 +08:00
|
|
|
*/
|
2016-05-13 22:36:29 +08:00
|
|
|
void intel_guc_init(struct drm_device *dev)
|
2015-08-12 22:43:36 +08:00
|
|
|
{
|
2016-07-04 18:34:36 +08:00
|
|
|
struct drm_i915_private *dev_priv = to_i915(dev);
|
2015-08-12 22:43:36 +08:00
|
|
|
struct intel_guc_fw *guc_fw = &dev_priv->guc.guc_fw;
|
|
|
|
const char *fw_path;
|
|
|
|
|
2016-10-15 07:47:05 +08:00
|
|
|
if (!HAS_GUC(dev)) {
|
|
|
|
i915.enable_guc_loading = 0;
|
|
|
|
i915.enable_guc_submission = 0;
|
|
|
|
} else {
|
|
|
|
/* A negative value means "use platform default" */
|
|
|
|
if (i915.enable_guc_loading < 0)
|
|
|
|
i915.enable_guc_loading = HAS_GUC_UCODE(dev);
|
|
|
|
if (i915.enable_guc_submission < 0)
|
|
|
|
i915.enable_guc_submission = HAS_GUC_SCHED(dev);
|
|
|
|
}
|
2015-08-12 22:43:36 +08:00
|
|
|
|
|
|
|
if (!HAS_GUC_UCODE(dev)) {
|
|
|
|
fw_path = NULL;
|
2016-10-13 18:03:03 +08:00
|
|
|
} else if (IS_SKYLAKE(dev_priv)) {
|
2015-08-12 22:43:36 +08:00
|
|
|
fw_path = I915_SKL_GUC_UCODE;
|
2016-08-10 23:16:46 +08:00
|
|
|
guc_fw->guc_fw_major_wanted = SKL_FW_MAJOR;
|
|
|
|
guc_fw->guc_fw_minor_wanted = SKL_FW_MINOR;
|
2016-10-13 18:03:04 +08:00
|
|
|
} else if (IS_BROXTON(dev_priv)) {
|
2016-05-06 18:42:53 +08:00
|
|
|
fw_path = I915_BXT_GUC_UCODE;
|
2016-08-10 23:16:46 +08:00
|
|
|
guc_fw->guc_fw_major_wanted = BXT_FW_MAJOR;
|
|
|
|
guc_fw->guc_fw_minor_wanted = BXT_FW_MINOR;
|
2016-10-13 18:03:02 +08:00
|
|
|
} else if (IS_KABYLAKE(dev_priv)) {
|
2016-07-01 00:37:52 +08:00
|
|
|
fw_path = I915_KBL_GUC_UCODE;
|
2016-08-10 23:16:46 +08:00
|
|
|
guc_fw->guc_fw_major_wanted = KBL_FW_MAJOR;
|
|
|
|
guc_fw->guc_fw_minor_wanted = KBL_FW_MINOR;
|
2015-08-12 22:43:36 +08:00
|
|
|
} else {
|
|
|
|
fw_path = ""; /* unknown device */
|
|
|
|
}
|
|
|
|
|
|
|
|
guc_fw->guc_dev = dev;
|
|
|
|
guc_fw->guc_fw_path = fw_path;
|
|
|
|
guc_fw->guc_fw_fetch_status = GUC_FIRMWARE_NONE;
|
|
|
|
guc_fw->guc_fw_load_status = GUC_FIRMWARE_NONE;
|
|
|
|
|
2016-05-20 18:42:42 +08:00
|
|
|
/* Early (and silent) return if GuC loading is disabled */
|
|
|
|
if (!i915.enable_guc_loading)
|
|
|
|
return;
|
2015-08-12 22:43:36 +08:00
|
|
|
if (fw_path == NULL)
|
|
|
|
return;
|
2016-05-20 18:42:42 +08:00
|
|
|
if (*fw_path == '\0')
|
2015-08-12 22:43:36 +08:00
|
|
|
return;
|
|
|
|
|
|
|
|
guc_fw->guc_fw_fetch_status = GUC_FIRMWARE_PENDING;
|
|
|
|
DRM_DEBUG_DRIVER("GuC firmware pending, path %s\n", fw_path);
|
|
|
|
guc_fw_fetch(dev, guc_fw);
|
|
|
|
/* status must now be FAIL or SUCCESS */
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
2016-05-13 22:36:29 +08:00
|
|
|
* intel_guc_fini() - clean up all allocated resources
|
2015-08-12 22:43:36 +08:00
|
|
|
* @dev: drm device
|
|
|
|
*/
|
2016-05-13 22:36:29 +08:00
|
|
|
void intel_guc_fini(struct drm_device *dev)
|
2015-08-12 22:43:36 +08:00
|
|
|
{
|
2016-07-04 18:34:36 +08:00
|
|
|
struct drm_i915_private *dev_priv = to_i915(dev);
|
2015-08-12 22:43:36 +08:00
|
|
|
struct intel_guc_fw *guc_fw = &dev_priv->guc.guc_fw;
|
|
|
|
|
2016-01-14 03:01:50 +08:00
|
|
|
mutex_lock(&dev->struct_mutex);
|
2016-09-13 04:19:36 +08:00
|
|
|
guc_interrupts_release(dev_priv);
|
2016-06-11 01:29:26 +08:00
|
|
|
i915_guc_submission_disable(dev_priv);
|
|
|
|
i915_guc_submission_fini(dev_priv);
|
2015-08-12 22:43:39 +08:00
|
|
|
|
2015-08-12 22:43:36 +08:00
|
|
|
if (guc_fw->guc_fw_obj)
|
2016-07-20 20:31:53 +08:00
|
|
|
i915_gem_object_put(guc_fw->guc_fw_obj);
|
2015-08-12 22:43:36 +08:00
|
|
|
guc_fw->guc_fw_obj = NULL;
|
2015-11-04 05:42:31 +08:00
|
|
|
mutex_unlock(&dev->struct_mutex);
|
2015-08-12 22:43:36 +08:00
|
|
|
|
|
|
|
guc_fw->guc_fw_fetch_status = GUC_FIRMWARE_NONE;
|
|
|
|
}
|