mirror of
https://github.com/AuxXxilium/linux_dsm_epyc7002.git
synced 2024-12-23 21:32:21 +07:00
6acbea89ef
This patch adds per engine reset and recovery (TDR) support when GuC is used to submit workloads to GPU. In the case of i915 directly submission to ELSP, driver manages hang detection, recovery and resubmission. With GuC submission these tasks are shared between driver and GuC. i915 is still responsible for detecting a hang, and when it does it only requests GuC to reset that Engine. GuC internally manages acquiring forcewake and idling the engine before resetting it. Once the reset is successful, i915 takes over again and handles the resubmission. The scheduler in i915 knows which requests are pending so after resetting a engine, pending workloads/requests are resubmitted again. v2: s/i915_guc_request_engine_reset/i915_guc_reset_engine/ to match the non-guc function names. v3: Removed debug message about engine restarting from which request, since the new baseline do it regardless of submission mode. (Chris) v4: Rebase. v5: Do not pass unnecessary reporting flags to the fw (Jeff); tasklet_schedule(&execlists->irq_tasklet) handles the resubmit; rebase. v6: Rename the existing reset engine function and share a similar interface between guc and non-guc paths (Chris). Signed-off-by: Michel Thierry <michel.thierry@intel.com> Cc: Chris Wilson <chris@chris-wilson.co.uk> Link: https://patchwork.freedesktop.org/patch/msgid/20171031225309.10888-1-michel.thierry@intel.com Reviewed-by: Jeff McGee <jeff.mcgee@intel.com> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
384 lines
10 KiB
C
384 lines
10 KiB
C
/*
|
|
* Copyright © 2014-2017 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.
|
|
*
|
|
*/
|
|
|
|
#include "intel_guc.h"
|
|
#include "i915_drv.h"
|
|
#include "i915_guc_submission.h"
|
|
|
|
static void gen8_guc_raise_irq(struct intel_guc *guc)
|
|
{
|
|
struct drm_i915_private *dev_priv = guc_to_i915(guc);
|
|
|
|
I915_WRITE(GUC_SEND_INTERRUPT, GUC_SEND_TRIGGER);
|
|
}
|
|
|
|
static inline i915_reg_t guc_send_reg(struct intel_guc *guc, u32 i)
|
|
{
|
|
GEM_BUG_ON(!guc->send_regs.base);
|
|
GEM_BUG_ON(!guc->send_regs.count);
|
|
GEM_BUG_ON(i >= guc->send_regs.count);
|
|
|
|
return _MMIO(guc->send_regs.base + 4 * i);
|
|
}
|
|
|
|
void intel_guc_init_send_regs(struct intel_guc *guc)
|
|
{
|
|
struct drm_i915_private *dev_priv = guc_to_i915(guc);
|
|
enum forcewake_domains fw_domains = 0;
|
|
unsigned int i;
|
|
|
|
guc->send_regs.base = i915_mmio_reg_offset(SOFT_SCRATCH(0));
|
|
guc->send_regs.count = SOFT_SCRATCH_COUNT - 1;
|
|
|
|
for (i = 0; i < guc->send_regs.count; i++) {
|
|
fw_domains |= intel_uncore_forcewake_for_reg(dev_priv,
|
|
guc_send_reg(guc, i),
|
|
FW_REG_READ | FW_REG_WRITE);
|
|
}
|
|
guc->send_regs.fw_domains = fw_domains;
|
|
}
|
|
|
|
void intel_guc_init_early(struct intel_guc *guc)
|
|
{
|
|
intel_guc_ct_init_early(&guc->ct);
|
|
|
|
mutex_init(&guc->send_mutex);
|
|
guc->send = intel_guc_send_nop;
|
|
guc->notify = gen8_guc_raise_irq;
|
|
}
|
|
|
|
static u32 get_gt_type(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)
|
|
{
|
|
u32 gen = INTEL_GEN(dev_priv);
|
|
|
|
switch (gen) {
|
|
case 9:
|
|
return GUC_CORE_FAMILY_GEN9;
|
|
|
|
default:
|
|
MISSING_CASE(gen);
|
|
return GUC_CORE_FAMILY_UNKNOWN;
|
|
}
|
|
}
|
|
|
|
/*
|
|
* Initialise the GuC parameter block before starting the firmware
|
|
* transfer. These parameters are read by the firmware on startup
|
|
* and cannot be changed thereafter.
|
|
*/
|
|
void intel_guc_init_params(struct intel_guc *guc)
|
|
{
|
|
struct drm_i915_private *dev_priv = guc_to_i915(guc);
|
|
u32 params[GUC_CTL_MAX_DWORDS];
|
|
int i;
|
|
|
|
memset(params, 0, sizeof(params));
|
|
|
|
params[GUC_CTL_DEVICE_INFO] |=
|
|
(get_gt_type(dev_priv) << GUC_CTL_GT_TYPE_SHIFT) |
|
|
(get_core_family(dev_priv) << GUC_CTL_CORE_FAMILY_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;
|
|
|
|
params[GUC_CTL_LOG_PARAMS] = guc->log.flags;
|
|
|
|
if (i915_modparams.guc_log_level >= 0) {
|
|
params[GUC_CTL_DEBUG] =
|
|
i915_modparams.guc_log_level << GUC_LOG_VERBOSITY_SHIFT;
|
|
} else {
|
|
params[GUC_CTL_DEBUG] = GUC_LOG_DISABLED;
|
|
}
|
|
|
|
/* If GuC submission is enabled, set up additional parameters here */
|
|
if (i915_modparams.enable_guc_submission) {
|
|
u32 ads = guc_ggtt_offset(guc->ads_vma) >> PAGE_SHIFT;
|
|
u32 pgs = guc_ggtt_offset(dev_priv->guc.stage_desc_pool);
|
|
u32 ctx_in_16 = GUC_MAX_STAGE_DESCRIPTORS / 16;
|
|
|
|
params[GUC_CTL_DEBUG] |= ads << GUC_ADS_ADDR_SHIFT;
|
|
params[GUC_CTL_DEBUG] |= GUC_ADS_ENABLED;
|
|
|
|
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;
|
|
}
|
|
|
|
/*
|
|
* All SOFT_SCRATCH registers are in FORCEWAKE_BLITTER domain and
|
|
* they are power context saved so it's ok to release forcewake
|
|
* when we are done here and take it again at xfer time.
|
|
*/
|
|
intel_uncore_forcewake_get(dev_priv, FORCEWAKE_BLITTER);
|
|
|
|
I915_WRITE(SOFT_SCRATCH(0), 0);
|
|
|
|
for (i = 0; i < GUC_CTL_MAX_DWORDS; i++)
|
|
I915_WRITE(SOFT_SCRATCH(1 + i), params[i]);
|
|
|
|
intel_uncore_forcewake_put(dev_priv, FORCEWAKE_BLITTER);
|
|
}
|
|
|
|
int intel_guc_send_nop(struct intel_guc *guc, const u32 *action, u32 len)
|
|
{
|
|
WARN(1, "Unexpected send: action=%#x\n", *action);
|
|
return -ENODEV;
|
|
}
|
|
|
|
/*
|
|
* This function implements the MMIO based host to GuC interface.
|
|
*/
|
|
int intel_guc_send_mmio(struct intel_guc *guc, const u32 *action, u32 len)
|
|
{
|
|
struct drm_i915_private *dev_priv = guc_to_i915(guc);
|
|
u32 status;
|
|
int i;
|
|
int ret;
|
|
|
|
GEM_BUG_ON(!len);
|
|
GEM_BUG_ON(len > guc->send_regs.count);
|
|
|
|
/* If CT is available, we expect to use MMIO only during init/fini */
|
|
GEM_BUG_ON(HAS_GUC_CT(dev_priv) &&
|
|
*action != INTEL_GUC_ACTION_REGISTER_COMMAND_TRANSPORT_BUFFER &&
|
|
*action != INTEL_GUC_ACTION_DEREGISTER_COMMAND_TRANSPORT_BUFFER);
|
|
|
|
mutex_lock(&guc->send_mutex);
|
|
intel_uncore_forcewake_get(dev_priv, guc->send_regs.fw_domains);
|
|
|
|
for (i = 0; i < len; i++)
|
|
I915_WRITE(guc_send_reg(guc, i), action[i]);
|
|
|
|
POSTING_READ(guc_send_reg(guc, i - 1));
|
|
|
|
intel_guc_notify(guc);
|
|
|
|
/*
|
|
* No GuC command should ever take longer than 10ms.
|
|
* Fast commands should still complete in 10us.
|
|
*/
|
|
ret = __intel_wait_for_register_fw(dev_priv,
|
|
guc_send_reg(guc, 0),
|
|
INTEL_GUC_RECV_MASK,
|
|
INTEL_GUC_RECV_MASK,
|
|
10, 10, &status);
|
|
if (status != INTEL_GUC_STATUS_SUCCESS) {
|
|
/*
|
|
* Either the GuC explicitly returned an error (which
|
|
* we convert to -EIO here) or no response at all was
|
|
* received within the timeout limit (-ETIMEDOUT)
|
|
*/
|
|
if (ret != -ETIMEDOUT)
|
|
ret = -EIO;
|
|
|
|
DRM_WARN("INTEL_GUC_SEND: Action 0x%X failed;"
|
|
" ret=%d status=0x%08X response=0x%08X\n",
|
|
action[0], ret, status, I915_READ(SOFT_SCRATCH(15)));
|
|
}
|
|
|
|
intel_uncore_forcewake_put(dev_priv, guc->send_regs.fw_domains);
|
|
mutex_unlock(&guc->send_mutex);
|
|
|
|
return ret;
|
|
}
|
|
|
|
int intel_guc_sample_forcewake(struct intel_guc *guc)
|
|
{
|
|
struct drm_i915_private *dev_priv = guc_to_i915(guc);
|
|
u32 action[2];
|
|
|
|
action[0] = INTEL_GUC_ACTION_SAMPLE_FORCEWAKE;
|
|
/* WaRsDisableCoarsePowerGating:skl,bxt */
|
|
if (!intel_rc6_enabled() ||
|
|
NEEDS_WaRsDisableCoarsePowerGating(dev_priv))
|
|
action[1] = 0;
|
|
else
|
|
/* bit 0 and 1 are for Render and Media domain separately */
|
|
action[1] = GUC_FORCEWAKE_RENDER | GUC_FORCEWAKE_MEDIA;
|
|
|
|
return intel_guc_send(guc, action, ARRAY_SIZE(action));
|
|
}
|
|
|
|
/**
|
|
* intel_guc_auth_huc() - Send action to GuC to authenticate HuC ucode
|
|
* @guc: intel_guc structure
|
|
* @rsa_offset: rsa offset w.r.t ggtt base of huc vma
|
|
*
|
|
* Triggers a HuC firmware authentication request to the GuC via intel_guc_send
|
|
* INTEL_GUC_ACTION_AUTHENTICATE_HUC interface. This function is invoked by
|
|
* intel_huc_auth().
|
|
*
|
|
* Return: non-zero code on error
|
|
*/
|
|
int intel_guc_auth_huc(struct intel_guc *guc, u32 rsa_offset)
|
|
{
|
|
u32 action[] = {
|
|
INTEL_GUC_ACTION_AUTHENTICATE_HUC,
|
|
rsa_offset
|
|
};
|
|
|
|
return intel_guc_send(guc, action, ARRAY_SIZE(action));
|
|
}
|
|
|
|
/**
|
|
* intel_guc_suspend() - notify GuC entering suspend state
|
|
* @dev_priv: i915 device private
|
|
*/
|
|
int intel_guc_suspend(struct drm_i915_private *dev_priv)
|
|
{
|
|
struct intel_guc *guc = &dev_priv->guc;
|
|
u32 data[3];
|
|
|
|
if (guc->fw.load_status != INTEL_UC_FIRMWARE_SUCCESS)
|
|
return 0;
|
|
|
|
gen9_disable_guc_interrupts(dev_priv);
|
|
|
|
data[0] = INTEL_GUC_ACTION_ENTER_S_STATE;
|
|
/* any value greater than GUC_POWER_D0 */
|
|
data[1] = GUC_POWER_D1;
|
|
data[2] = guc_ggtt_offset(guc->shared_data);
|
|
|
|
return intel_guc_send(guc, data, ARRAY_SIZE(data));
|
|
}
|
|
|
|
/**
|
|
* intel_guc_reset_engine() - ask GuC to reset an engine
|
|
* @guc: intel_guc structure
|
|
* @engine: engine to be reset
|
|
*/
|
|
int intel_guc_reset_engine(struct intel_guc *guc,
|
|
struct intel_engine_cs *engine)
|
|
{
|
|
u32 data[7];
|
|
|
|
GEM_BUG_ON(!guc->execbuf_client);
|
|
|
|
data[0] = INTEL_GUC_ACTION_REQUEST_ENGINE_RESET;
|
|
data[1] = engine->guc_id;
|
|
data[2] = 0;
|
|
data[3] = 0;
|
|
data[4] = 0;
|
|
data[5] = guc->execbuf_client->stage_id;
|
|
data[6] = guc_ggtt_offset(guc->shared_data);
|
|
|
|
return intel_guc_send(guc, data, ARRAY_SIZE(data));
|
|
}
|
|
|
|
/**
|
|
* intel_guc_resume() - notify GuC resuming from suspend state
|
|
* @dev_priv: i915 device private
|
|
*/
|
|
int intel_guc_resume(struct drm_i915_private *dev_priv)
|
|
{
|
|
struct intel_guc *guc = &dev_priv->guc;
|
|
u32 data[3];
|
|
|
|
if (guc->fw.load_status != INTEL_UC_FIRMWARE_SUCCESS)
|
|
return 0;
|
|
|
|
if (i915_modparams.guc_log_level >= 0)
|
|
gen9_enable_guc_interrupts(dev_priv);
|
|
|
|
data[0] = INTEL_GUC_ACTION_EXIT_S_STATE;
|
|
data[1] = GUC_POWER_D0;
|
|
data[2] = guc_ggtt_offset(guc->shared_data);
|
|
|
|
return intel_guc_send(guc, data, ARRAY_SIZE(data));
|
|
}
|
|
|
|
/**
|
|
* intel_guc_allocate_vma() - Allocate a GGTT VMA for GuC usage
|
|
* @guc: the guc
|
|
* @size: size of area to allocate (both virtual space and memory)
|
|
*
|
|
* This is a wrapper to create an object for use with the GuC. In order to
|
|
* use it inside the GuC, an object needs to be pinned lifetime, so we allocate
|
|
* both some backing storage and a range inside the Global GTT. We must pin
|
|
* it in the GGTT somewhere other than than [0, GUC_WOPCM_TOP) because that
|
|
* range is reserved inside GuC.
|
|
*
|
|
* Return: A i915_vma if successful, otherwise an ERR_PTR.
|
|
*/
|
|
struct i915_vma *intel_guc_allocate_vma(struct intel_guc *guc, u32 size)
|
|
{
|
|
struct drm_i915_private *dev_priv = guc_to_i915(guc);
|
|
struct drm_i915_gem_object *obj;
|
|
struct i915_vma *vma;
|
|
int ret;
|
|
|
|
obj = i915_gem_object_create(dev_priv, size);
|
|
if (IS_ERR(obj))
|
|
return ERR_CAST(obj);
|
|
|
|
vma = i915_vma_instance(obj, &dev_priv->ggtt.base, NULL);
|
|
if (IS_ERR(vma))
|
|
goto err;
|
|
|
|
ret = i915_vma_pin(vma, 0, PAGE_SIZE,
|
|
PIN_GLOBAL | PIN_OFFSET_BIAS | GUC_WOPCM_TOP);
|
|
if (ret) {
|
|
vma = ERR_PTR(ret);
|
|
goto err;
|
|
}
|
|
|
|
return vma;
|
|
|
|
err:
|
|
i915_gem_object_put(obj);
|
|
return vma;
|
|
}
|
|
|
|
u32 intel_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_GEN9_LP(dev_priv))
|
|
wopcm_size -= BXT_GUC_WOPCM_RC6_RESERVED;
|
|
|
|
return wopcm_size;
|
|
}
|