Commit ee71434e authored by Aravind Iddamsetty's avatar Aravind Iddamsetty Committed by Daniele Ceraolo Spurio

drm/i915/mtl: Handle wopcm per-GT and limit calculations.

With MTL standalone media architecture the wopcm layout has changed,
with separate partitioning in WOPCM for the root GT GuC and the media
GT GuC. The size of WOPCM is 4MB with the lower 2MB reserved for the
media GT and the upper 2MB for the root GT.

Given that MTL has GuC deprivilege, the WOPCM registers are pre-locked
by the bios. Therefore, we can skip all the math for the partitioning
and just limit ourselves to sanity-checking the values.

v2: fix makefile file ordering (Jani)
v3: drop XELPM_SAMEDIA_WOPCM_SIZE, check huc instead of VDBOX (John)
v4: further clarify commit message, remove blank line (John)
Signed-off-by: default avatarAravind Iddamsetty <aravind.iddamsetty@intel.com>
Signed-off-by: default avatarDaniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Cc: Matt Roper <matthew.d.roper@intel.com>
Cc: John Harrison <john.c.harrison@intel.com>
Cc: Alan Previn <alan.previn.teres.alexis@intel.com>
Cc: Jani Nikula <jani.nikula@linux.intel.com>
Reviewed-by: default avatarJohn Harrison <John.C.Harrison@Intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20221108020600.3575467-5-daniele.ceraolospurio@intel.com
parent 00b4c0ef
...@@ -494,7 +494,7 @@ WOPCM ...@@ -494,7 +494,7 @@ WOPCM
WOPCM Layout WOPCM Layout
~~~~~~~~~~~~ ~~~~~~~~~~~~
.. kernel-doc:: drivers/gpu/drm/i915/intel_wopcm.c .. kernel-doc:: drivers/gpu/drm/i915/gt/intel_wopcm.c
:doc: WOPCM Layout :doc: WOPCM Layout
GuC GuC
......
...@@ -127,9 +127,11 @@ gt-y += \ ...@@ -127,9 +127,11 @@ gt-y += \
gt/intel_sseu.o \ gt/intel_sseu.o \
gt/intel_sseu_debugfs.o \ gt/intel_sseu_debugfs.o \
gt/intel_timeline.o \ gt/intel_timeline.o \
gt/intel_wopcm.o \
gt/intel_workarounds.o \ gt/intel_workarounds.o \
gt/shmem_utils.o \ gt/shmem_utils.o \
gt/sysfs_engines.o gt/sysfs_engines.o
# x86 intel-gtt module support # x86 intel-gtt module support
gt-$(CONFIG_X86) += gt/intel_ggtt_gmch.o gt-$(CONFIG_X86) += gt/intel_ggtt_gmch.o
# autogenerated null render state # autogenerated null render state
...@@ -183,8 +185,7 @@ i915-y += \ ...@@ -183,8 +185,7 @@ i915-y += \
i915_trace_points.o \ i915_trace_points.o \
i915_ttm_buddy_manager.o \ i915_ttm_buddy_manager.o \
i915_vma.o \ i915_vma.o \
i915_vma_resource.o \ i915_vma_resource.o
intel_wopcm.o
# general-purpose microcontroller (GuC) support # general-purpose microcontroller (GuC) support
i915-y += gt/uc/intel_uc.o \ i915-y += gt/uc/intel_uc.o \
......
...@@ -560,7 +560,7 @@ static int init_ggtt(struct i915_ggtt *ggtt) ...@@ -560,7 +560,7 @@ static int init_ggtt(struct i915_ggtt *ggtt)
* why. * why.
*/ */
ggtt->pin_bias = max_t(u32, I915_GTT_PAGE_SIZE, ggtt->pin_bias = max_t(u32, I915_GTT_PAGE_SIZE,
intel_wopcm_guc_size(&ggtt->vm.i915->wopcm)); intel_wopcm_guc_size(&ggtt->vm.gt->wopcm));
ret = intel_vgt_balloon(ggtt); ret = intel_vgt_balloon(ggtt);
if (ret) if (ret)
......
...@@ -54,6 +54,7 @@ void intel_gt_common_init_early(struct intel_gt *gt) ...@@ -54,6 +54,7 @@ void intel_gt_common_init_early(struct intel_gt *gt)
seqcount_mutex_init(&gt->tlb.seqno, &gt->tlb.invalidate_lock); seqcount_mutex_init(&gt->tlb.seqno, &gt->tlb.invalidate_lock);
intel_gt_pm_init_early(gt); intel_gt_pm_init_early(gt);
intel_wopcm_init_early(&gt->wopcm);
intel_uc_init_early(&gt->uc); intel_uc_init_early(&gt->uc);
intel_rps_init_early(&gt->rps); intel_rps_init_early(&gt->rps);
} }
......
...@@ -31,6 +31,7 @@ ...@@ -31,6 +31,7 @@
#include "intel_migrate_types.h" #include "intel_migrate_types.h"
#include "intel_wakeref.h" #include "intel_wakeref.h"
#include "pxp/intel_pxp_types.h" #include "pxp/intel_pxp_types.h"
#include "intel_wopcm.h"
struct drm_i915_private; struct drm_i915_private;
struct i915_ggtt; struct i915_ggtt;
...@@ -101,6 +102,7 @@ struct intel_gt { ...@@ -101,6 +102,7 @@ struct intel_gt {
struct intel_uc uc; struct intel_uc uc;
struct intel_gsc gsc; struct intel_gsc gsc;
struct intel_wopcm wopcm;
struct { struct {
/* Serialize global tlb invalidations */ /* Serialize global tlb invalidations */
......
...@@ -64,9 +64,9 @@ ...@@ -64,9 +64,9 @@
#define GEN9_GUC_FW_RESERVED SZ_128K #define GEN9_GUC_FW_RESERVED SZ_128K
#define GEN9_GUC_WOPCM_OFFSET (GUC_WOPCM_RESERVED + GEN9_GUC_FW_RESERVED) #define GEN9_GUC_WOPCM_OFFSET (GUC_WOPCM_RESERVED + GEN9_GUC_FW_RESERVED)
static inline struct drm_i915_private *wopcm_to_i915(struct intel_wopcm *wopcm) static inline struct intel_gt *wopcm_to_gt(struct intel_wopcm *wopcm)
{ {
return container_of(wopcm, struct drm_i915_private, wopcm); return container_of(wopcm, struct intel_gt, wopcm);
} }
/** /**
...@@ -77,7 +77,8 @@ static inline struct drm_i915_private *wopcm_to_i915(struct intel_wopcm *wopcm) ...@@ -77,7 +77,8 @@ static inline struct drm_i915_private *wopcm_to_i915(struct intel_wopcm *wopcm)
*/ */
void intel_wopcm_init_early(struct intel_wopcm *wopcm) void intel_wopcm_init_early(struct intel_wopcm *wopcm)
{ {
struct drm_i915_private *i915 = wopcm_to_i915(wopcm); struct intel_gt *gt = wopcm_to_gt(wopcm);
struct drm_i915_private *i915 = gt->i915;
if (!HAS_GT_UC(i915)) if (!HAS_GT_UC(i915))
return; return;
...@@ -157,10 +158,11 @@ static bool check_hw_restrictions(struct drm_i915_private *i915, ...@@ -157,10 +158,11 @@ static bool check_hw_restrictions(struct drm_i915_private *i915,
return true; return true;
} }
static bool __check_layout(struct drm_i915_private *i915, u32 wopcm_size, static bool __check_layout(struct intel_gt *gt, u32 wopcm_size,
u32 guc_wopcm_base, u32 guc_wopcm_size, u32 guc_wopcm_base, u32 guc_wopcm_size,
u32 guc_fw_size, u32 huc_fw_size) u32 guc_fw_size, u32 huc_fw_size)
{ {
struct drm_i915_private *i915 = gt->i915;
const u32 ctx_rsvd = context_reserved_size(i915); const u32 ctx_rsvd = context_reserved_size(i915);
u32 size; u32 size;
...@@ -181,6 +183,7 @@ static bool __check_layout(struct drm_i915_private *i915, u32 wopcm_size, ...@@ -181,6 +183,7 @@ static bool __check_layout(struct drm_i915_private *i915, u32 wopcm_size,
return false; return false;
} }
if (intel_uc_supports_huc(&gt->uc)) {
size = huc_fw_size + WOPCM_RESERVED_SIZE; size = huc_fw_size + WOPCM_RESERVED_SIZE;
if (unlikely(guc_wopcm_base < size)) { if (unlikely(guc_wopcm_base < size)) {
drm_err(&i915->drm, "WOPCM: no space for %s: %uK < %uK\n", drm_err(&i915->drm, "WOPCM: no space for %s: %uK < %uK\n",
...@@ -188,6 +191,7 @@ static bool __check_layout(struct drm_i915_private *i915, u32 wopcm_size, ...@@ -188,6 +191,7 @@ static bool __check_layout(struct drm_i915_private *i915, u32 wopcm_size,
guc_wopcm_base / SZ_1K, size / SZ_1K); guc_wopcm_base / SZ_1K, size / SZ_1K);
return false; return false;
} }
}
return check_hw_restrictions(i915, guc_wopcm_base, guc_wopcm_size, return check_hw_restrictions(i915, guc_wopcm_base, guc_wopcm_size,
huc_fw_size); huc_fw_size);
...@@ -228,8 +232,8 @@ static bool __wopcm_regs_writable(struct intel_uncore *uncore) ...@@ -228,8 +232,8 @@ static bool __wopcm_regs_writable(struct intel_uncore *uncore)
*/ */
void intel_wopcm_init(struct intel_wopcm *wopcm) void intel_wopcm_init(struct intel_wopcm *wopcm)
{ {
struct drm_i915_private *i915 = wopcm_to_i915(wopcm); struct intel_gt *gt = wopcm_to_gt(wopcm);
struct intel_gt *gt = to_gt(i915); struct drm_i915_private *i915 = gt->i915;
u32 guc_fw_size = intel_uc_fw_get_upload_size(&gt->uc.guc.fw); u32 guc_fw_size = intel_uc_fw_get_upload_size(&gt->uc.guc.fw);
u32 huc_fw_size = intel_uc_fw_get_upload_size(&gt->uc.huc.fw); u32 huc_fw_size = intel_uc_fw_get_upload_size(&gt->uc.huc.fw);
u32 ctx_rsvd = context_reserved_size(i915); u32 ctx_rsvd = context_reserved_size(i915);
...@@ -274,6 +278,19 @@ void intel_wopcm_init(struct intel_wopcm *wopcm) ...@@ -274,6 +278,19 @@ void intel_wopcm_init(struct intel_wopcm *wopcm)
goto check; goto check;
} }
/*
* On platforms with a media GT, the WOPCM is partitioned between the
* two GTs, so we would have to take that into account when doing the
* math below. There is also a new section reserved for the GSC context
* that would have to be factored in. However, all platforms with a
* media GT also have GuC depriv enabled, so the WOPCM regs are
* pre-locked and therefore we don't have to do the math ourselves.
*/
if (unlikely(i915->media_gt)) {
drm_err(&i915->drm, "Unlocked WOPCM regs with media GT\n");
return;
}
/* /*
* Aligned value of guc_wopcm_base will determine available WOPCM space * Aligned value of guc_wopcm_base will determine available WOPCM space
* for HuC firmware and mandatory reserved area. * for HuC firmware and mandatory reserved area.
...@@ -295,7 +312,7 @@ void intel_wopcm_init(struct intel_wopcm *wopcm) ...@@ -295,7 +312,7 @@ void intel_wopcm_init(struct intel_wopcm *wopcm)
guc_wopcm_base / SZ_1K, guc_wopcm_size / SZ_1K); guc_wopcm_base / SZ_1K, guc_wopcm_size / SZ_1K);
check: check:
if (__check_layout(i915, wopcm_size, guc_wopcm_base, guc_wopcm_size, if (__check_layout(gt, wopcm_size, guc_wopcm_base, guc_wopcm_size,
guc_fw_size, huc_fw_size)) { guc_fw_size, huc_fw_size)) {
wopcm->guc.base = guc_wopcm_base; wopcm->guc.base = guc_wopcm_base;
wopcm->guc.size = guc_wopcm_size; wopcm->guc.size = guc_wopcm_size;
......
...@@ -357,8 +357,8 @@ static int uc_init_wopcm(struct intel_uc *uc) ...@@ -357,8 +357,8 @@ static int uc_init_wopcm(struct intel_uc *uc)
{ {
struct intel_gt *gt = uc_to_gt(uc); struct intel_gt *gt = uc_to_gt(uc);
struct intel_uncore *uncore = gt->uncore; struct intel_uncore *uncore = gt->uncore;
u32 base = intel_wopcm_guc_base(&gt->i915->wopcm); u32 base = intel_wopcm_guc_base(&gt->wopcm);
u32 size = intel_wopcm_guc_size(&gt->i915->wopcm); u32 size = intel_wopcm_guc_size(&gt->wopcm);
u32 huc_agent = intel_uc_uses_huc(uc) ? HUC_LOADING_AGENT_GUC : 0; u32 huc_agent = intel_uc_uses_huc(uc) ? HUC_LOADING_AGENT_GUC : 0;
u32 mask; u32 mask;
int err; int err;
......
...@@ -478,10 +478,11 @@ static int check_gsc_manifest(const struct firmware *fw, ...@@ -478,10 +478,11 @@ static int check_gsc_manifest(const struct firmware *fw,
return 0; return 0;
} }
static int check_ccs_header(struct drm_i915_private *i915, static int check_ccs_header(struct intel_gt *gt,
const struct firmware *fw, const struct firmware *fw,
struct intel_uc_fw *uc_fw) struct intel_uc_fw *uc_fw)
{ {
struct drm_i915_private *i915 = gt->i915;
struct uc_css_header *css; struct uc_css_header *css;
size_t size; size_t size;
...@@ -523,10 +524,10 @@ static int check_ccs_header(struct drm_i915_private *i915, ...@@ -523,10 +524,10 @@ static int check_ccs_header(struct drm_i915_private *i915,
/* Sanity check whether this fw is not larger than whole WOPCM memory */ /* Sanity check whether this fw is not larger than whole WOPCM memory */
size = __intel_uc_fw_get_upload_size(uc_fw); size = __intel_uc_fw_get_upload_size(uc_fw);
if (unlikely(size >= i915->wopcm.size)) { if (unlikely(size >= gt->wopcm.size)) {
drm_warn(&i915->drm, "%s firmware %s: invalid size: %zu > %zu\n", drm_warn(&i915->drm, "%s firmware %s: invalid size: %zu > %zu\n",
intel_uc_fw_type_repr(uc_fw->type), uc_fw->file_selected.path, intel_uc_fw_type_repr(uc_fw->type), uc_fw->file_selected.path,
size, (size_t)i915->wopcm.size); size, (size_t)gt->wopcm.size);
return -E2BIG; return -E2BIG;
} }
...@@ -554,7 +555,8 @@ static int check_ccs_header(struct drm_i915_private *i915, ...@@ -554,7 +555,8 @@ static int check_ccs_header(struct drm_i915_private *i915,
*/ */
int intel_uc_fw_fetch(struct intel_uc_fw *uc_fw) int intel_uc_fw_fetch(struct intel_uc_fw *uc_fw)
{ {
struct drm_i915_private *i915 = __uc_fw_to_gt(uc_fw)->i915; struct intel_gt *gt = __uc_fw_to_gt(uc_fw);
struct drm_i915_private *i915 = gt->i915;
struct intel_uc_fw_file file_ideal; struct intel_uc_fw_file file_ideal;
struct device *dev = i915->drm.dev; struct device *dev = i915->drm.dev;
struct drm_i915_gem_object *obj; struct drm_i915_gem_object *obj;
...@@ -562,7 +564,7 @@ int intel_uc_fw_fetch(struct intel_uc_fw *uc_fw) ...@@ -562,7 +564,7 @@ int intel_uc_fw_fetch(struct intel_uc_fw *uc_fw)
bool old_ver = false; bool old_ver = false;
int err; int err;
GEM_BUG_ON(!i915->wopcm.size); GEM_BUG_ON(!gt->wopcm.size);
GEM_BUG_ON(!intel_uc_fw_is_enabled(uc_fw)); GEM_BUG_ON(!intel_uc_fw_is_enabled(uc_fw));
err = i915_inject_probe_error(i915, -ENXIO); err = i915_inject_probe_error(i915, -ENXIO);
...@@ -615,7 +617,7 @@ int intel_uc_fw_fetch(struct intel_uc_fw *uc_fw) ...@@ -615,7 +617,7 @@ int intel_uc_fw_fetch(struct intel_uc_fw *uc_fw)
if (uc_fw->loaded_via_gsc) if (uc_fw->loaded_via_gsc)
err = check_gsc_manifest(fw, uc_fw); err = check_gsc_manifest(fw, uc_fw);
else else
err = check_ccs_header(i915, fw, uc_fw); err = check_ccs_header(gt, fw, uc_fw);
if (err) if (err)
goto fail; goto fail;
......
...@@ -372,8 +372,6 @@ static int i915_driver_early_probe(struct drm_i915_private *dev_priv) ...@@ -372,8 +372,6 @@ static int i915_driver_early_probe(struct drm_i915_private *dev_priv)
if (ret) if (ret)
goto err_ttm; goto err_ttm;
intel_wopcm_init_early(&dev_priv->wopcm);
ret = intel_root_gt_init_early(dev_priv); ret = intel_root_gt_init_early(dev_priv);
if (ret < 0) if (ret < 0)
goto err_rootgt; goto err_rootgt;
......
...@@ -62,7 +62,6 @@ ...@@ -62,7 +62,6 @@
#include "intel_runtime_pm.h" #include "intel_runtime_pm.h"
#include "intel_step.h" #include "intel_step.h"
#include "intel_uncore.h" #include "intel_uncore.h"
#include "intel_wopcm.h"
struct drm_i915_clock_gating_funcs; struct drm_i915_clock_gating_funcs;
struct drm_i915_gem_object; struct drm_i915_gem_object;
...@@ -235,8 +234,6 @@ struct drm_i915_private { ...@@ -235,8 +234,6 @@ struct drm_i915_private {
struct intel_gvt *gvt; struct intel_gvt *gvt;
struct intel_wopcm wopcm;
struct pci_dev *bridge_dev; struct pci_dev *bridge_dev;
struct rb_root uabi_engines; struct rb_root uabi_engines;
......
...@@ -1140,9 +1140,10 @@ int i915_gem_init(struct drm_i915_private *dev_priv) ...@@ -1140,9 +1140,10 @@ int i915_gem_init(struct drm_i915_private *dev_priv)
if (ret) if (ret)
return ret; return ret;
for_each_gt(gt, dev_priv, i) for_each_gt(gt, dev_priv, i) {
intel_uc_fetch_firmwares(&gt->uc); intel_uc_fetch_firmwares(&gt->uc);
intel_wopcm_init(&dev_priv->wopcm); intel_wopcm_init(&gt->wopcm);
}
ret = i915_init_ggtt(dev_priv); ret = i915_init_ggtt(dev_priv);
if (ret) { if (ret) {
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment