Skip to contentSkip to conversation

HuC Loading Patches

PATCH In review anushasr 8 participants 43 messages Dec 8, 2016 – Dec 14, 2016

Subsystems (3)
lore
Share & tools
lore

anushasr

author
#0 lore raw

cover HuC Loading Patches

These patches add HuC loading support. The GuC is required to authenticate the HuC. The userspace patches that check for a fully loaded HuC firmware and use it can be found at:
https://lists.freedesktop.org/archives/libva/2016-September/004554.html
https://lists.freedesktop.org/archives/libva/2016-September/004555.html

More information regarding the HuC, batch commands that configure the HuC etc can be found at-
https://01.org/sites/default/files/documentation/intel-gfx-prm-osrc-skl-vol02a-commandreference-instructions-huc.pdf
https://www.x.org/docs/intel/CHV/intel-gfx-prm-osrc-chv-bsw-vol10-hevc.pdf

v2: rebased.
v3: rebased. Changed the code following the review comments.
v4: Added action_lock initialization fix provided by Arek
(Hiler Arkadiusz) to the first patch in the series- Make the
GuC fw loading helper functions general.
v5: rebased on top of drm-tip. The patch series is now in sync with GuC
code reorganization efforts by Arek-
https://patchwork.freedesktop.org/series/15896/
v6: rebased. Organize the code-move contents of intel_huc.h to intel_uc.h. Update functions intel_huc_load(),intel_huc_init() and intel_uc_fw_fetch() to accept dev_priv instead of dev.

Anusha Srivatsa (3):
drm/i915/huc: Add HuC fw loading support
drm/i915/huc: Add BXT HuC Loading Support
drm/i915/HuC: Add KBL huC loading Support

Peter Antoine (5):
drm/i915/guc: Make the GuC fw loading helper functions general drm/i915/huc: Unified css_header struct for GuC and HuC drm/i915/huc: Add debugfs for HuC loading status check drm/i915/huc: Support HuC authentication drm/i915/get_params: Add HuC status to getparams

 drivers/gpu/drm/i915/Makefile              |   1 +
 drivers/gpu/drm/i915/i915_debugfs.c        |  43 +++-
 drivers/gpu/drm/i915/i915_drv.c            |   8 +-
 drivers/gpu/drm/i915/i915_drv.h            |   3 +-
 drivers/gpu/drm/i915/i915_guc_reg.h        |   3 +
 drivers/gpu/drm/i915/i915_guc_submission.c |   4 +-
 drivers/gpu/drm/i915/intel_guc_fwif.h      |  22 ++-
 drivers/gpu/drm/i915/intel_guc_loader.c    | 199 ++++++++++---------
 drivers/gpu/drm/i915/intel_huc_loader.c    | 303 +++++++++++++++++++++++++++++
 drivers/gpu/drm/i915/intel_uc.c            |  61 ++++++
 drivers/gpu/drm/i915/intel_uc.h            |  68 +++++--
 include/uapi/drm/i915_drm.h                |   1 +
 12 files changed, 593 insertions(+), 123 deletions(-)
 create mode 100644 drivers/gpu/drm/i915/intel_huc_loader.c
Signature · 6 lines
2.7.4

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
author
#1 lore raw

PATCH 1/8 drm/i915/guc: Make the GuC fw loading helper functions general

From: Peter Antoine <[email protected]>

Rename some of the GuC fw loading code to make them more general. We will utilise them for HuC loading as well.
s/intel_guc_fw/intel_uc_fw/g s/GUC_FIRMWARE/UC_FIRMWARE/g

Struct intel_guc_fw is renamed to intel_uc_fw. Prefix of tts members, such as 'guc' or 'guc_fw' either is renamed to 'uc' or removed for same purpose.

v2: rebased on top of nightly.
reapplied the search/replace as upstream code as changed.
v3: rebased again on drm-nightly.
v4: removed G from messages in shared fw fetch function.
v5: rebased.
v7: rebased.
v8: rebased.
v9: rebased.
v10: rebased.
v11: rebased.
v12: rebased on top of drm-tip
v13: rebased.Updated dev to dev_priv in intel_guc_setup(), guc_fw_getch() and intel_guc_init().

---
 drivers/gpu/drm/i915/i915_debugfs.c        |  12 +--
 drivers/gpu/drm/i915/i915_guc_submission.c |   4 +-
 drivers/gpu/drm/i915/intel_guc_loader.c    | 157 +++++++++++++++--------------
 drivers/gpu/drm/i915/intel_uc.h            |  40 ++++----
 4 files changed, 108 insertions(+), 105 deletions(-)
drivers/gpu/drm/i915/i915_debugfs.c +6 -6
index a746130..0e5ef62 100644
@@ -2338,7 +2338,7 @@ static int i915_llc(struct seq_file *m, void *data)
 static int i915_guc_load_status_info(struct seq_file *m, void *data)
 {
 	struct drm_i915_private *dev_priv = node_to_i915(m->private);
-	struct intel_guc_fw *guc_fw = &dev_priv->guc.guc_fw;
+	struct intel_uc_fw *guc_fw = &dev_priv->guc.guc_fw;
 	u32 tmp, i;
 
 	if (!HAS_GUC_UCODE(dev_priv))
@@ -2346,15 +2346,15 @@ static int i915_guc_load_status_info(struct seq_file *m, void *data)
 
 	seq_printf(m, "GuC firmware status:\n");
 	seq_printf(m, "\tpath: %s\n",
-		guc_fw->guc_fw_path);
+		guc_fw->uc_fw_path);
 	seq_printf(m, "\tfetch: %s\n",
-		intel_guc_fw_status_repr(guc_fw->guc_fw_fetch_status));
+		intel_uc_fw_status_repr(guc_fw->fetch_status));
 	seq_printf(m, "\tload: %s\n",
-		intel_guc_fw_status_repr(guc_fw->guc_fw_load_status));
+		intel_uc_fw_status_repr(guc_fw->load_status));
 	seq_printf(m, "\tversion wanted: %d.%d\n",
-		guc_fw->guc_fw_major_wanted, guc_fw->guc_fw_minor_wanted);
+		guc_fw->major_ver_wanted, guc_fw->minor_ver_wanted);
 	seq_printf(m, "\tversion found: %d.%d\n",
-		guc_fw->guc_fw_major_found, guc_fw->guc_fw_minor_found);
+		guc_fw->major_ver_found, guc_fw->minor_ver_found);
 	seq_printf(m, "\theader: offset is %d; size = %d\n",
 		guc_fw->header_offset, guc_fw->header_size);
 	seq_printf(m, "\tuCode: offset is %d; size = %d\n",
drivers/gpu/drm/i915/i915_guc_submission.c +2 -2
index 7fa4e74..e156a4b 100644
@@ -1493,7 +1493,7 @@ int intel_guc_suspend(struct drm_i915_private *dev_priv)
 	struct i915_gem_context *ctx;
 	u32 data[3];
 
-	if (guc->guc_fw.guc_fw_load_status != GUC_FIRMWARE_SUCCESS)
+	if (guc->guc_fw.load_status != UC_FIRMWARE_SUCCESS)
 		return 0;
 
 	gen9_disable_guc_interrupts(dev_priv);
@@ -1520,7 +1520,7 @@ int intel_guc_resume(struct drm_i915_private *dev_priv)
 	struct i915_gem_context *ctx;
 	u32 data[3];
 
-	if (guc->guc_fw.guc_fw_load_status != GUC_FIRMWARE_SUCCESS)
+	if (guc->guc_fw.load_status != UC_FIRMWARE_SUCCESS)
 		return 0;
 
 	if (i915.guc_log_level >= 0)
drivers/gpu/drm/i915/intel_guc_loader.c +79 -78

342 lines not on this page · raw message

drivers/gpu/drm/i915/intel_uc.h +21 -19
index 11f5608..f9f598d 100644
@@ -91,29 +91,31 @@ struct i915_guc_client {
 	uint64_t submissions[I915_NUM_ENGINES];
 };
 
-enum intel_guc_fw_status {
-	GUC_FIRMWARE_FAIL = -1,
-	GUC_FIRMWARE_NONE = 0,
-	GUC_FIRMWARE_PENDING,
-	GUC_FIRMWARE_SUCCESS
+enum intel_uc_fw_status {
+	UC_FIRMWARE_FAIL = -1,
+	UC_FIRMWARE_NONE = 0,
+	UC_FIRMWARE_PENDING,
+	UC_FIRMWARE_SUCCESS
 };
 
 /*
  * This structure encapsulates all the data needed during the process
  * of fetching, caching, and loading the firmware image into the GuC.
  */
-struct intel_guc_fw {
-	const char *			guc_fw_path;
-	size_t				guc_fw_size;
-	struct drm_i915_gem_object *	guc_fw_obj;
-	enum intel_guc_fw_status	guc_fw_fetch_status;
-	enum intel_guc_fw_status	guc_fw_load_status;
-
-	uint16_t			guc_fw_major_wanted;
-	uint16_t			guc_fw_minor_wanted;
-	uint16_t			guc_fw_major_found;
-	uint16_t			guc_fw_minor_found;
-
+struct intel_uc_fw {
+	struct drm_device *uc_dev;
+	const char *uc_fw_path;
+	size_t uc_fw_size;
+	struct drm_i915_gem_object *uc_fw_obj;
+	enum intel_uc_fw_status fetch_status;
+	enum intel_uc_fw_status load_status;
+
+	uint16_t major_ver_wanted;
+	uint16_t minor_ver_wanted;
+	uint16_t major_ver_found;
+	uint16_t minor_ver_found;
+
+	uint32_t fw_type;
 	uint32_t header_size;
 	uint32_t header_offset;
 	uint32_t rsa_size;
@@ -139,7 +141,7 @@ struct intel_guc_log {
 };
 
 struct intel_guc {
-	struct intel_guc_fw guc_fw;
+	struct intel_uc_fw guc_fw;
 	struct intel_guc_log log;
 
 	/* intel_guc_recv interrupt related state */
@@ -181,7 +183,7 @@ int intel_guc_log_control(struct intel_guc *guc, u32 control_val);
 extern void intel_guc_init(struct drm_i915_private *dev_priv);
 extern int intel_guc_setup(struct drm_i915_private *dev_priv);
 extern void intel_guc_fini(struct drm_i915_private *dev_priv);
-extern const char *intel_guc_fw_status_repr(enum intel_guc_fw_status status);
+extern const char *intel_uc_fw_status_repr(enum intel_uc_fw_status status);
 extern int intel_guc_suspend(struct drm_i915_private *dev_priv);
 extern int intel_guc_resume(struct drm_i915_private *dev_priv);
 
Signature · 6 lines
2.7.4

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Reviewed-by on v1 1/8
#2 lore raw

Re: [PATCH 1/8] drm/i915/guc: Make the GuC fw loading helper functions general

Quoted 11 lines · anushasr

From: Peter Antoine <[email protected]>

Rename some of the GuC fw loading code to make them more general. We will utilise them for HuC loading as well.
s/intel_guc_fw/intel_uc_fw/g s/GUC_FIRMWARE/UC_FIRMWARE/g

Struct intel_guc_fw is renamed to intel_uc_fw. Prefix of tts members, such as 'guc' or 'guc_fw' either is renamed to 'uc' or removed for same purpose.

Quoted 13 lines

v2: rebased on top of nightly.
reapplied the search/replace as upstream code as changed.
v3: rebased again on drm-nightly.
v4: removed G from messages in shared fw fetch function.
v5: rebased.
v7: rebased.
v8: rebased.
v9: rebased.
v10: rebased.
v11: rebased.
v12: rebased on top of drm-tip
v13: rebased.Updated dev to dev_priv in intel_guc_setup(), guc_fw_getch() and intel_guc_init().

This is little bit misleading. It suggests that you did the change from dev to dev_priv, while it really happened externally. You just had to rebase and accomodate for that.

nontheless...

Quoted 3 lines
Signature · 6 lines
Cheers,
Arek
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
#3 lore raw

Re: [PATCH 1/8] drm/i915/guc: Make the GuC fw loading helper functions general

In reply to anushasr
Quoted 462 lines · anushasr

From: Peter Antoine <[email protected]>

Rename some of the GuC fw loading code to make them more general. We will utilise them for HuC loading as well.
s/intel_guc_fw/intel_uc_fw/g s/GUC_FIRMWARE/UC_FIRMWARE/g

Struct intel_guc_fw is renamed to intel_uc_fw. Prefix of tts members, such as 'guc' or 'guc_fw' either is renamed to 'uc' or removed for same purpose.

v2: rebased on top of nightly.
reapplied the search/replace as upstream code as changed.
v3: rebased again on drm-nightly.
v4: removed G from messages in shared fw fetch function.
v5: rebased.
v7: rebased.
v8: rebased.
v9: rebased.
v10: rebased.
v11: rebased.
v12: rebased on top of drm-tip
v13: rebased.Updated dev to dev_priv in intel_guc_setup(), guc_fw_getch() and intel_guc_init().

---
 drivers/gpu/drm/i915/i915_debugfs.c        |  12 +--
 drivers/gpu/drm/i915/i915_guc_submission.c |   4 +-
 drivers/gpu/drm/i915/intel_guc_loader.c    | 157 +++++++++++++++--------------
 drivers/gpu/drm/i915/intel_uc.h            |  40 ++++----
 4 files changed, 108 insertions(+), 105 deletions(-)
drivers/gpu/drm/i915/i915_debugfs.c +6 -6
index a746130..0e5ef62 100644
@@ -2338,7 +2338,7 @@ static int i915_llc(struct seq_file *m, void *data)
 static int i915_guc_load_status_info(struct seq_file *m, void *data)
 {
 	struct drm_i915_private *dev_priv = node_to_i915(m->private);
-	struct intel_guc_fw *guc_fw = &dev_priv->guc.guc_fw;
+	struct intel_uc_fw *guc_fw = &dev_priv->guc.guc_fw;
 	u32 tmp, i;
 
 	if (!HAS_GUC_UCODE(dev_priv))
@@ -2346,15 +2346,15 @@ static int i915_guc_load_status_info(struct seq_file *m, void *data)
 
 	seq_printf(m, "GuC firmware status:\n");
 	seq_printf(m, "\tpath: %s\n",
-		guc_fw->guc_fw_path);
+		guc_fw->uc_fw_path);
 	seq_printf(m, "\tfetch: %s\n",
-		intel_guc_fw_status_repr(guc_fw->guc_fw_fetch_status));
+		intel_uc_fw_status_repr(guc_fw->fetch_status));
 	seq_printf(m, "\tload: %s\n",
-		intel_guc_fw_status_repr(guc_fw->guc_fw_load_status));
+		intel_uc_fw_status_repr(guc_fw->load_status));
 	seq_printf(m, "\tversion wanted: %d.%d\n",
-		guc_fw->guc_fw_major_wanted, guc_fw->guc_fw_minor_wanted);
+		guc_fw->major_ver_wanted, guc_fw->minor_ver_wanted);
 	seq_printf(m, "\tversion found: %d.%d\n",
-		guc_fw->guc_fw_major_found, guc_fw->guc_fw_minor_found);
+		guc_fw->major_ver_found, guc_fw->minor_ver_found);
 	seq_printf(m, "\theader: offset is %d; size = %d\n",
 		guc_fw->header_offset, guc_fw->header_size);
 	seq_printf(m, "\tuCode: offset is %d; size = %d\n",
drivers/gpu/drm/i915/i915_guc_submission.c +2 -2
index 7fa4e74..e156a4b 100644
@@ -1493,7 +1493,7 @@ int intel_guc_suspend(struct drm_i915_private *dev_priv)
 	struct i915_gem_context *ctx;
 	u32 data[3];
 
-	if (guc->guc_fw.guc_fw_load_status != GUC_FIRMWARE_SUCCESS)
+	if (guc->guc_fw.load_status != UC_FIRMWARE_SUCCESS)
 		return 0;
 
 	gen9_disable_guc_interrupts(dev_priv);
@@ -1520,7 +1520,7 @@ int intel_guc_resume(struct drm_i915_private *dev_priv)
 	struct i915_gem_context *ctx;
 	u32 data[3];
 
-	if (guc->guc_fw.guc_fw_load_status != GUC_FIRMWARE_SUCCESS)
+	if (guc->guc_fw.load_status != UC_FIRMWARE_SUCCESS)
 		return 0;
 
 	if (i915.guc_log_level >= 0)
drivers/gpu/drm/i915/intel_guc_loader.c +79 -78
index 21db697..8f04f6e 100644
@@ -81,16 +81,16 @@ MODULE_FIRMWARE(I915_BXT_GUC_UCODE);
 MODULE_FIRMWARE(I915_KBL_GUC_UCODE);
 
 /* User-friendly representation of an enum */
-const char *intel_guc_fw_status_repr(enum intel_guc_fw_status status)
+const char *intel_uc_fw_status_repr(enum intel_uc_fw_status status)
 {
 	switch (status) {
-	case GUC_FIRMWARE_FAIL:
+	case UC_FIRMWARE_FAIL:
 		return "FAIL";
-	case GUC_FIRMWARE_NONE:
+	case UC_FIRMWARE_NONE:
 		return "NONE";
-	case GUC_FIRMWARE_PENDING:
+	case UC_FIRMWARE_PENDING:
 		return "PENDING";
-	case GUC_FIRMWARE_SUCCESS:
+	case UC_FIRMWARE_SUCCESS:
 		return "SUCCESS";
 	default:
 		return "UNKNOWN!";
@@ -278,7 +278,7 @@ static inline bool guc_ucode_response(struct drm_i915_private *dev_priv,
 static int guc_ucode_xfer_dma(struct drm_i915_private *dev_priv,
 			      struct i915_vma *vma)
 {
-	struct intel_guc_fw *guc_fw = &dev_priv->guc.guc_fw;
+	struct intel_uc_fw *guc_fw = &dev_priv->guc.guc_fw;
 	unsigned long offset;
 	struct sg_table *sg = vma->pages;
 	u32 status, rsa[UOS_RSA_SCRATCH_MAX_COUNT];
@@ -350,17 +350,17 @@ static u32 guc_wopcm_size(struct drm_i915_private *dev_priv)
  */
 static int guc_ucode_xfer(struct drm_i915_private *dev_priv)
 {
-	struct intel_guc_fw *guc_fw = &dev_priv->guc.guc_fw;
+	struct intel_uc_fw *guc_fw = &dev_priv->guc.guc_fw;
 	struct i915_vma *vma;
 	int ret;
 
-	ret = i915_gem_object_set_to_gtt_domain(guc_fw->guc_fw_obj, false);
+	ret = i915_gem_object_set_to_gtt_domain(guc_fw->uc_fw_obj, false);
 	if (ret) {
 		DRM_DEBUG_DRIVER("set-domain failed %d\n", ret);
 		return ret;
 	}
 
-	vma = i915_gem_object_ggtt_pin(guc_fw->guc_fw_obj, NULL, 0, 0, 0);
+	vma = i915_gem_object_ggtt_pin(guc_fw->uc_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);
@@ -450,14 +450,14 @@ static int guc_hw_reset(struct drm_i915_private *dev_priv)
  */
 int intel_guc_setup(struct drm_i915_private *dev_priv)
 {
-	struct intel_guc_fw *guc_fw = &dev_priv->guc.guc_fw;
-	const char *fw_path = guc_fw->guc_fw_path;
+	struct intel_uc_fw *guc_fw = &dev_priv->guc.guc_fw;
+	const char *fw_path = guc_fw->uc_fw_path;
 	int retries, ret, err;
 
 	DRM_DEBUG_DRIVER("GuC fw status: path %s, fetch %s, load %s\n",
 		fw_path,
-		intel_guc_fw_status_repr(guc_fw->guc_fw_fetch_status),
-		intel_guc_fw_status_repr(guc_fw->guc_fw_load_status));
+		intel_uc_fw_status_repr(guc_fw->fetch_status),
+		intel_uc_fw_status_repr(guc_fw->load_status));
 
 	/* Loading forbidden, or no firmware to load? */
 	if (!i915.enable_guc_loading) {
@@ -475,10 +475,10 @@ int intel_guc_setup(struct drm_i915_private *dev_priv)
 	}
 
 	/* Fetch failed, or already fetched but failed to load? */
-	if (guc_fw->guc_fw_fetch_status != GUC_FIRMWARE_SUCCESS) {
+	if (guc_fw->fetch_status != UC_FIRMWARE_SUCCESS) {
 		err = -EIO;
 		goto fail;
-	} else if (guc_fw->guc_fw_load_status == GUC_FIRMWARE_FAIL) {
+	} else if (guc_fw->load_status == UC_FIRMWARE_FAIL) {
 		err = -ENOEXEC;
 		goto fail;
 	}
@@ -486,11 +486,11 @@ int intel_guc_setup(struct drm_i915_private *dev_priv)
 	guc_interrupts_release(dev_priv);
 	gen9_reset_guc_interrupts(dev_priv);
 
-	guc_fw->guc_fw_load_status = GUC_FIRMWARE_PENDING;
+	guc_fw->load_status = UC_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));
+		intel_uc_fw_status_repr(guc_fw->fetch_status),
+		intel_uc_fw_status_repr(guc_fw->load_status));
 
 	err = i915_guc_submission_init(dev_priv);
 	if (err)
@@ -522,11 +522,11 @@ int intel_guc_setup(struct drm_i915_private *dev_priv)
 			 "retry %d more time(s)\n", err, retries);
 	}
 
-	guc_fw->guc_fw_load_status = GUC_FIRMWARE_SUCCESS;
+	guc_fw->load_status = UC_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));
+		intel_uc_fw_status_repr(guc_fw->fetch_status),
+		intel_uc_fw_status_repr(guc_fw->load_status));
 
 	if (i915.enable_guc_submission) {
 		if (i915.guc_log_level >= 0)
@@ -541,8 +541,8 @@ int intel_guc_setup(struct drm_i915_private *dev_priv)
 	return 0;
 
 fail:
-	if (guc_fw->guc_fw_load_status == GUC_FIRMWARE_PENDING)
-		guc_fw->guc_fw_load_status = GUC_FIRMWARE_FAIL;
+	if (guc_fw->load_status == UC_FIRMWARE_PENDING)
+		guc_fw->load_status = UC_FIRMWARE_FAIL;
 
 	guc_interrupts_release(dev_priv);
 	i915_guc_submission_disable(dev_priv);
@@ -587,8 +587,8 @@ int intel_guc_setup(struct drm_i915_private *dev_priv)
 	return ret;
 }
 
-static void guc_fw_fetch(struct drm_i915_private *dev_priv,
-			 struct intel_guc_fw *guc_fw)
+void intel_uc_fw_fetch(struct drm_i915_private *dev_priv,
+			 struct intel_uc_fw *uc_fw)
 {
 	struct pci_dev *pdev = dev_priv->drm.pdev;
 	struct drm_i915_gem_object *obj;
@@ -597,17 +597,17 @@ static void guc_fw_fetch(struct drm_i915_private *dev_priv,
 	size_t size;
 	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));
+	DRM_DEBUG_DRIVER("before requesting firmware: uC fw fetch status %s\n",
+		intel_uc_fw_status_repr(uc_fw->fetch_status));
 
-	err = request_firmware(&fw, guc_fw->guc_fw_path, &pdev->dev);
+	err = request_firmware(&fw, uc_fw->uc_fw_path, &pdev->dev);
 	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);
+	DRM_DEBUG_DRIVER("fetch uC fw from %s succeeded, fw %p\n",
+		uc_fw->uc_fw_path, fw);
 
 	/* Check the size of the blob before examining buffer contents */
 	if (fw->size < sizeof(struct guc_css_header)) {
@@ -618,36 +618,36 @@ static void guc_fw_fetch(struct drm_i915_private *dev_priv,
 	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 -
+	uc_fw->header_offset = 0;
+	uc_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)) {
+	if (uc_fw->header_size != sizeof(struct guc_css_header)) {
 		DRM_NOTE("CSS header definition mismatch\n");
 		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);
+	uc_fw->ucode_offset = uc_fw->header_offset + uc_fw->header_size;
+	uc_fw->ucode_size = (css->size_dw - css->header_size_dw) * sizeof(u32);
 
 	/* now RSA */
 	if (css->key_size_dw != UOS_RSA_SCRATCH_MAX_COUNT) {
 		DRM_NOTE("RSA key size is bad\n");
 		goto fail;
 	}
-	guc_fw->rsa_offset = guc_fw->ucode_offset + guc_fw->ucode_size;
-	guc_fw->rsa_size = css->key_size_dw * sizeof(u32);
+	uc_fw->rsa_offset = uc_fw->ucode_offset + uc_fw->ucode_size;
+	uc_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;
+	size = uc_fw->header_size + uc_fw->ucode_size + uc_fw->rsa_size;
 	if (fw->size < size) {
 		DRM_NOTE("Missing firmware components\n");
 		goto fail;
 	}
 
 	/* Header and uCode will be loaded to WOPCM. Size of the two. */
-	size = guc_fw->header_size + guc_fw->ucode_size;
+	size = uc_fw->header_size + uc_fw->ucode_size;
 	if (size > guc_wopcm_size(dev_priv)) {
 		DRM_NOTE("Firmware is too large to fit in WOPCM\n");
 		goto fail;
@@ -659,21 +659,21 @@ static void guc_fw_fetch(struct drm_i915_private *dev_priv,
 	 * TWO bytes each (i.e. u16), although all pointers and offsets are defined
 	 * in terms of bytes (u8).
 	 */
-	guc_fw->guc_fw_major_found = css->guc_sw_version >> 16;
-	guc_fw->guc_fw_minor_found = css->guc_sw_version & 0xFFFF;
-
-	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) {
-		DRM_NOTE("GuC firmware version %d.%d, required %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);
+	uc_fw->major_ver_found = css->guc_sw_version >> 16;
+	uc_fw->minor_ver_found = css->guc_sw_version & 0xFFFF;
+
+	if (uc_fw->major_ver_found != uc_fw->major_ver_wanted ||
+	    uc_fw->minor_ver_found < uc_fw->minor_ver_wanted) {
+		DRM_NOTE("uC firmware version %d.%d, required %d.%d\n",
+			uc_fw->major_ver_found, uc_fw->minor_ver_found,
+			uc_fw->major_ver_wanted, uc_fw->minor_ver_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);
+			uc_fw->major_ver_found, uc_fw->minor_ver_found,
+			uc_fw->major_ver_wanted, uc_fw->minor_ver_wanted);
 
 	mutex_lock(&dev_priv->drm.struct_mutex);
 	obj = i915_gem_object_create_from_data(dev_priv, fw->data, fw->size);
@@ -683,31 +683,31 @@ static void guc_fw_fetch(struct drm_i915_private *dev_priv,
 		goto fail;
 	}
 
-	guc_fw->guc_fw_obj = obj;
-	guc_fw->guc_fw_size = fw->size;
+	uc_fw->uc_fw_obj = obj;
+	uc_fw->uc_fw_size = fw->size;
 
-	DRM_DEBUG_DRIVER("GuC fw fetch status SUCCESS, obj %p\n",
-			guc_fw->guc_fw_obj);
+	DRM_DEBUG_DRIVER("uC fw fetch status SUCCESS, obj %p\n",
+			uc_fw->uc_fw_obj);
 
 	release_firmware(fw);
-	guc_fw->guc_fw_fetch_status = GUC_FIRMWARE_SUCCESS;
+	uc_fw->fetch_status = UC_FIRMWARE_SUCCESS;
 	return;
 
 fail:
-	DRM_WARN("Failed to fetch valid GuC firmware from %s (error %d)\n",
-		 guc_fw->guc_fw_path, err);
-	DRM_DEBUG_DRIVER("GuC fw fetch status FAIL; err %d, fw %p, obj %p\n",
-		err, fw, guc_fw->guc_fw_obj);
+	DRM_WARN("Failed to fetch valid uC firmware from %s (error %d)\n",
+		 uc_fw->uc_fw_path, err);
+	DRM_DEBUG_DRIVER("uC fw fetch status FAIL; err %d, fw %p, obj %p\n",
+		err, fw, uc_fw->uc_fw_obj);
 
 	mutex_lock(&dev_priv->drm.struct_mutex);
-	obj = guc_fw->guc_fw_obj;
+	obj = uc_fw->uc_fw_obj;
 	if (obj)
 		i915_gem_object_put(obj);
-	guc_fw->guc_fw_obj = NULL;
+	uc_fw->uc_fw_obj = NULL;
 	mutex_unlock(&dev_priv->drm.struct_mutex);
 
 	release_firmware(fw);		/* OK even if fw is NULL */
-	guc_fw->guc_fw_fetch_status = GUC_FIRMWARE_FAIL;
+	uc_fw->fetch_status = UC_FIRMWARE_FAIL;
 }
 
 /**
@@ -721,7 +721,7 @@ static void guc_fw_fetch(struct drm_i915_private *dev_priv,
  */
 void intel_guc_init(struct drm_i915_private *dev_priv)
 {
-	struct intel_guc_fw *guc_fw = &dev_priv->guc.guc_fw;
+	struct intel_uc_fw *guc_fw = &dev_priv->guc.guc_fw;
 	const char *fw_path;
 
 	if (!HAS_GUC(dev_priv)) {
@@ -739,23 +739,24 @@ void intel_guc_init(struct drm_i915_private *dev_priv)
 		fw_path = NULL;
 	} else if (IS_SKYLAKE(dev_priv)) {
 		fw_path = I915_SKL_GUC_UCODE;
-		guc_fw->guc_fw_major_wanted = SKL_FW_MAJOR;
-		guc_fw->guc_fw_minor_wanted = SKL_FW_MINOR;
+		guc_fw->major_ver_wanted = SKL_FW_MAJOR;
+		guc_fw->minor_ver_wanted = SKL_FW_MINOR;
 	} else if (IS_BROXTON(dev_priv)) {
 		fw_path = I915_BXT_GUC_UCODE;
-		guc_fw->guc_fw_major_wanted = BXT_FW_MAJOR;
-		guc_fw->guc_fw_minor_wanted = BXT_FW_MINOR;
+		guc_fw->major_ver_wanted = BXT_FW_MAJOR;
+		guc_fw->minor_ver_wanted = BXT_FW_MINOR;
 	} else if (IS_KABYLAKE(dev_priv)) {
 		fw_path = I915_KBL_GUC_UCODE;
-		guc_fw->guc_fw_major_wanted = KBL_FW_MAJOR;
-		guc_fw->guc_fw_minor_wanted = KBL_FW_MINOR;
+		guc_fw->major_ver_wanted = KBL_FW_MAJOR;
+		guc_fw->minor_ver_wanted = KBL_FW_MINOR;
 	} else {
 		fw_path = "";	/* unknown device */
 	}
 
-	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;
+	guc_fw->uc_dev = &dev_priv->drm;
+	guc_fw->uc_fw_path = fw_path;
+	guc_fw->fetch_status = UC_FIRMWARE_NONE;
+	guc_fw->load_status = UC_FIRMWARE_NONE;
 
 	/* Early (and silent) return if GuC loading is disabled */
 	if (!i915.enable_guc_loading)
@@ -765,9 +766,9 @@ void intel_guc_init(struct drm_i915_private *dev_priv)
 	if (*fw_path == '\0')
 		return;
 
-	guc_fw->guc_fw_fetch_status = GUC_FIRMWARE_PENDING;
+	guc_fw->fetch_status = UC_FIRMWARE_PENDING;
 	DRM_DEBUG_DRIVER("GuC firmware pending, path %s\n", fw_path);
-	guc_fw_fetch(dev_priv, guc_fw);
+	intel_uc_fw_fetch(dev_priv, guc_fw);
 	/* status must now be FAIL or SUCCESS */
 }
 
@@ -777,17 +778,17 @@ void intel_guc_init(struct drm_i915_private *dev_priv)
  */
 void intel_guc_fini(struct drm_i915_private *dev_priv)
 {
-	struct intel_guc_fw *guc_fw = &dev_priv->guc.guc_fw;
+	struct intel_uc_fw *guc_fw = &dev_priv->guc.guc_fw;
 
 	mutex_lock(&dev_priv->drm.struct_mutex);
 	guc_interrupts_release(dev_priv);
 	i915_guc_submission_disable(dev_priv);
 	i915_guc_submission_fini(dev_priv);
 
-	if (guc_fw->guc_fw_obj)
-		i915_gem_object_put(guc_fw->guc_fw_obj);
-	guc_fw->guc_fw_obj = NULL;
+	if (guc_fw->uc_fw_obj)
+		i915_gem_object_put(guc_fw->uc_fw_obj);
+	guc_fw->uc_fw_obj = NULL;
 	mutex_unlock(&dev_priv->drm.struct_mutex);
 
-	guc_fw->guc_fw_fetch_status = GUC_FIRMWARE_NONE;
+	guc_fw->fetch_status = UC_FIRMWARE_NONE;
 }
drivers/gpu/drm/i915/intel_uc.h +0 -0
index 11f5608..f9f598d 100644
@@ -91,29 +91,31 @@ struct i915_guc_client {
 	uint64_t submissions[I915_NUM_ENGINES];
 };
 

drivers/gpu/drm/i915/intel_uc.h struct i915_guc_client { · line 98

Quoted 10 lines
drivers/gpu/drm/i915/intel_uc.h +5 -5
-enum intel_guc_fw_status {
-	GUC_FIRMWARE_FAIL = -1,
-	GUC_FIRMWARE_NONE = 0,
-	GUC_FIRMWARE_PENDING,
-	GUC_FIRMWARE_SUCCESS
+enum intel_uc_fw_status {
+	UC_FIRMWARE_FAIL = -1,
+	UC_FIRMWARE_NONE = 0,
+	UC_FIRMWARE_PENDING,
+	UC_FIRMWARE_SUCCESS

Shouldn't we use INTEL_UC_ prefix here (in similar way as in GUC_ACTION)?

Quoted 10 lines
drivers/gpu/drm/i915/intel_uc.h +0 -4
 };
 
 /*
  * This structure encapsulates all the data needed during the process
  * of fetching, caching, and loading the firmware image into the GuC.
  */
-struct intel_guc_fw {
-	const char *			guc_fw_path;
-	size_t				guc_fw_size;
-	struct drm_i915_gem_object *	guc_fw_obj;

drivers/gpu/drm/i915/intel_uc.h struct i915_guc_client { · line 106

Quoted 10 lines
drivers/gpu/drm/i915/intel_uc.h +2 -8
-	enum intel_guc_fw_status	guc_fw_fetch_status;
-	enum intel_guc_fw_status	guc_fw_load_status;
-
-	uint16_t			guc_fw_major_wanted;
-	uint16_t			guc_fw_minor_wanted;
-	uint16_t			guc_fw_major_found;
-	uint16_t			guc_fw_minor_found;
-
+struct intel_uc_fw {
+	struct drm_device *uc_dev;

Why do you need this back pointer to drm_dev ? More likely all you need is pointer to dev_priv. But for dev_priv you can always use guc_to_i915().

drivers/gpu/drm/i915/intel_uc.h struct i915_guc_client { · line 109

Quoted 3 lines
drivers/gpu/drm/i915/intel_uc.h +3 -0
+	const char *uc_fw_path;
+	size_t uc_fw_size;
+	struct drm_i915_gem_object *uc_fw_obj;

All these uc_fw_ prefixes seem to be redundant as this struct is just about uc_fw.

drivers/gpu/drm/i915/intel_uc.h struct i915_guc_client { · line 118

Quoted 9 lines
drivers/gpu/drm/i915/intel_uc.h +9 -0
+	enum intel_uc_fw_status fetch_status;
+	enum intel_uc_fw_status load_status;
+
+	uint16_t major_ver_wanted;
+	uint16_t minor_ver_wanted;
+	uint16_t major_ver_found;
+	uint16_t minor_ver_found;
+
+	uint32_t fw_type;

What is the purpose of this field ?

drivers/gpu/drm/i915/intel_uc.h struct intel_guc_log { · line 144

Quoted 9 lines
drivers/gpu/drm/i915/intel_uc.h +1 -1
 	uint32_t header_size;
 	uint32_t header_offset;
 	uint32_t rsa_size;
@@ -139,7 +141,7 @@ struct intel_guc_log {
 };
 
 struct intel_guc {
-	struct intel_guc_fw guc_fw;
+	struct intel_uc_fw guc_fw;

Btw, guc_ prefix in field name here is also redundant.

Quoted 19 lines
drivers/gpu/drm/i915/intel_uc.h +1 -1
 	struct intel_guc_log log;
 
 	/* intel_guc_recv interrupt related state */
@@ -181,7 +183,7 @@ int intel_guc_log_control(struct intel_guc *guc, u32 control_val);
 extern void intel_guc_init(struct drm_i915_private *dev_priv);
 extern int intel_guc_setup(struct drm_i915_private *dev_priv);
 extern void intel_guc_fini(struct drm_i915_private *dev_priv);
-extern const char *intel_guc_fw_status_repr(enum intel_guc_fw_status status);
+extern const char *intel_uc_fw_status_repr(enum intel_uc_fw_status status);
 extern int intel_guc_suspend(struct drm_i915_private *dev_priv);
 extern int intel_guc_resume(struct drm_i915_private *dev_priv);
 
Signature · 6 lines
2.7.4

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

#4 lore raw

Re: [PATCH 1/8] drm/i915/guc: Make the GuC fw loading helper functions general

Quoted 493 lines · Michal Wajdeczko
Quoted 462 lines · anushasr

From: Peter Antoine <[email protected]>

Rename some of the GuC fw loading code to make them more general. We will utilise them for HuC loading as well.
s/intel_guc_fw/intel_uc_fw/g s/GUC_FIRMWARE/UC_FIRMWARE/g

Struct intel_guc_fw is renamed to intel_uc_fw. Prefix of tts members, such as 'guc' or 'guc_fw' either is renamed to 'uc' or removed for same purpose.

v2: rebased on top of nightly.
reapplied the search/replace as upstream code as changed.
v3: rebased again on drm-nightly.
v4: removed G from messages in shared fw fetch function.
v5: rebased.
v7: rebased.
v8: rebased.
v9: rebased.
v10: rebased.
v11: rebased.
v12: rebased on top of drm-tip
v13: rebased.Updated dev to dev_priv in intel_guc_setup(), guc_fw_getch() and intel_guc_init().

---
 drivers/gpu/drm/i915/i915_debugfs.c        |  12 +--
 drivers/gpu/drm/i915/i915_guc_submission.c |   4 +-
 drivers/gpu/drm/i915/intel_guc_loader.c    | 157 +++++++++++++++--------------
 drivers/gpu/drm/i915/intel_uc.h            |  40 ++++----
 4 files changed, 108 insertions(+), 105 deletions(-)
drivers/gpu/drm/i915/i915_debugfs.c +6 -6
index a746130..0e5ef62 100644
@@ -2338,7 +2338,7 @@ static int i915_llc(struct seq_file *m, void *data)
 static int i915_guc_load_status_info(struct seq_file *m, void *data)
 {
 	struct drm_i915_private *dev_priv = node_to_i915(m->private);
-	struct intel_guc_fw *guc_fw = &dev_priv->guc.guc_fw;
+	struct intel_uc_fw *guc_fw = &dev_priv->guc.guc_fw;
 	u32 tmp, i;
 
 	if (!HAS_GUC_UCODE(dev_priv))
@@ -2346,15 +2346,15 @@ static int i915_guc_load_status_info(struct seq_file *m, void *data)
 
 	seq_printf(m, "GuC firmware status:\n");
 	seq_printf(m, "\tpath: %s\n",
-		guc_fw->guc_fw_path);
+		guc_fw->uc_fw_path);
 	seq_printf(m, "\tfetch: %s\n",
-		intel_guc_fw_status_repr(guc_fw->guc_fw_fetch_status));
+		intel_uc_fw_status_repr(guc_fw->fetch_status));
 	seq_printf(m, "\tload: %s\n",
-		intel_guc_fw_status_repr(guc_fw->guc_fw_load_status));
+		intel_uc_fw_status_repr(guc_fw->load_status));
 	seq_printf(m, "\tversion wanted: %d.%d\n",
-		guc_fw->guc_fw_major_wanted, guc_fw->guc_fw_minor_wanted);
+		guc_fw->major_ver_wanted, guc_fw->minor_ver_wanted);
 	seq_printf(m, "\tversion found: %d.%d\n",
-		guc_fw->guc_fw_major_found, guc_fw->guc_fw_minor_found);
+		guc_fw->major_ver_found, guc_fw->minor_ver_found);
 	seq_printf(m, "\theader: offset is %d; size = %d\n",
 		guc_fw->header_offset, guc_fw->header_size);
 	seq_printf(m, "\tuCode: offset is %d; size = %d\n",
drivers/gpu/drm/i915/i915_guc_submission.c +2 -2
index 7fa4e74..e156a4b 100644
@@ -1493,7 +1493,7 @@ int intel_guc_suspend(struct drm_i915_private *dev_priv)
 	struct i915_gem_context *ctx;
 	u32 data[3];
 
-	if (guc->guc_fw.guc_fw_load_status != GUC_FIRMWARE_SUCCESS)
+	if (guc->guc_fw.load_status != UC_FIRMWARE_SUCCESS)
 		return 0;
 
 	gen9_disable_guc_interrupts(dev_priv);
@@ -1520,7 +1520,7 @@ int intel_guc_resume(struct drm_i915_private *dev_priv)
 	struct i915_gem_context *ctx;
 	u32 data[3];
 
-	if (guc->guc_fw.guc_fw_load_status != GUC_FIRMWARE_SUCCESS)
+	if (guc->guc_fw.load_status != UC_FIRMWARE_SUCCESS)
 		return 0;
 
 	if (i915.guc_log_level >= 0)
drivers/gpu/drm/i915/intel_guc_loader.c +79 -78
index 21db697..8f04f6e 100644
@@ -81,16 +81,16 @@ MODULE_FIRMWARE(I915_BXT_GUC_UCODE);
 MODULE_FIRMWARE(I915_KBL_GUC_UCODE);
 
 /* User-friendly representation of an enum */
-const char *intel_guc_fw_status_repr(enum intel_guc_fw_status status)
+const char *intel_uc_fw_status_repr(enum intel_uc_fw_status status)
 {
 	switch (status) {
-	case GUC_FIRMWARE_FAIL:
+	case UC_FIRMWARE_FAIL:
 		return "FAIL";
-	case GUC_FIRMWARE_NONE:
+	case UC_FIRMWARE_NONE:
 		return "NONE";
-	case GUC_FIRMWARE_PENDING:
+	case UC_FIRMWARE_PENDING:
 		return "PENDING";
-	case GUC_FIRMWARE_SUCCESS:
+	case UC_FIRMWARE_SUCCESS:
 		return "SUCCESS";
 	default:
 		return "UNKNOWN!";
@@ -278,7 +278,7 @@ static inline bool guc_ucode_response(struct drm_i915_private *dev_priv,
 static int guc_ucode_xfer_dma(struct drm_i915_private *dev_priv,
 			      struct i915_vma *vma)
 {
-	struct intel_guc_fw *guc_fw = &dev_priv->guc.guc_fw;
+	struct intel_uc_fw *guc_fw = &dev_priv->guc.guc_fw;
 	unsigned long offset;
 	struct sg_table *sg = vma->pages;
 	u32 status, rsa[UOS_RSA_SCRATCH_MAX_COUNT];
@@ -350,17 +350,17 @@ static u32 guc_wopcm_size(struct drm_i915_private *dev_priv)
  */
 static int guc_ucode_xfer(struct drm_i915_private *dev_priv)
 {
-	struct intel_guc_fw *guc_fw = &dev_priv->guc.guc_fw;
+	struct intel_uc_fw *guc_fw = &dev_priv->guc.guc_fw;
 	struct i915_vma *vma;
 	int ret;
 
-	ret = i915_gem_object_set_to_gtt_domain(guc_fw->guc_fw_obj, false);
+	ret = i915_gem_object_set_to_gtt_domain(guc_fw->uc_fw_obj, false);
 	if (ret) {
 		DRM_DEBUG_DRIVER("set-domain failed %d\n", ret);
 		return ret;
 	}
 
-	vma = i915_gem_object_ggtt_pin(guc_fw->guc_fw_obj, NULL, 0, 0, 0);
+	vma = i915_gem_object_ggtt_pin(guc_fw->uc_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);
@@ -450,14 +450,14 @@ static int guc_hw_reset(struct drm_i915_private *dev_priv)
  */
 int intel_guc_setup(struct drm_i915_private *dev_priv)
 {
-	struct intel_guc_fw *guc_fw = &dev_priv->guc.guc_fw;
-	const char *fw_path = guc_fw->guc_fw_path;
+	struct intel_uc_fw *guc_fw = &dev_priv->guc.guc_fw;
+	const char *fw_path = guc_fw->uc_fw_path;
 	int retries, ret, err;
 
 	DRM_DEBUG_DRIVER("GuC fw status: path %s, fetch %s, load %s\n",
 		fw_path,
-		intel_guc_fw_status_repr(guc_fw->guc_fw_fetch_status),
-		intel_guc_fw_status_repr(guc_fw->guc_fw_load_status));
+		intel_uc_fw_status_repr(guc_fw->fetch_status),
+		intel_uc_fw_status_repr(guc_fw->load_status));
 
 	/* Loading forbidden, or no firmware to load? */
 	if (!i915.enable_guc_loading) {
@@ -475,10 +475,10 @@ int intel_guc_setup(struct drm_i915_private *dev_priv)
 	}
 
 	/* Fetch failed, or already fetched but failed to load? */
-	if (guc_fw->guc_fw_fetch_status != GUC_FIRMWARE_SUCCESS) {
+	if (guc_fw->fetch_status != UC_FIRMWARE_SUCCESS) {
 		err = -EIO;
 		goto fail;
-	} else if (guc_fw->guc_fw_load_status == GUC_FIRMWARE_FAIL) {
+	} else if (guc_fw->load_status == UC_FIRMWARE_FAIL) {
 		err = -ENOEXEC;
 		goto fail;
 	}
@@ -486,11 +486,11 @@ int intel_guc_setup(struct drm_i915_private *dev_priv)
 	guc_interrupts_release(dev_priv);
 	gen9_reset_guc_interrupts(dev_priv);
 
-	guc_fw->guc_fw_load_status = GUC_FIRMWARE_PENDING;
+	guc_fw->load_status = UC_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));
+		intel_uc_fw_status_repr(guc_fw->fetch_status),
+		intel_uc_fw_status_repr(guc_fw->load_status));
 
 	err = i915_guc_submission_init(dev_priv);
 	if (err)
@@ -522,11 +522,11 @@ int intel_guc_setup(struct drm_i915_private *dev_priv)
 			 "retry %d more time(s)\n", err, retries);
 	}
 
-	guc_fw->guc_fw_load_status = GUC_FIRMWARE_SUCCESS;
+	guc_fw->load_status = UC_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));
+		intel_uc_fw_status_repr(guc_fw->fetch_status),
+		intel_uc_fw_status_repr(guc_fw->load_status));
 
 	if (i915.enable_guc_submission) {
 		if (i915.guc_log_level >= 0)
@@ -541,8 +541,8 @@ int intel_guc_setup(struct drm_i915_private *dev_priv)
 	return 0;
 
 fail:
-	if (guc_fw->guc_fw_load_status == GUC_FIRMWARE_PENDING)
-		guc_fw->guc_fw_load_status = GUC_FIRMWARE_FAIL;
+	if (guc_fw->load_status == UC_FIRMWARE_PENDING)
+		guc_fw->load_status = UC_FIRMWARE_FAIL;
 
 	guc_interrupts_release(dev_priv);
 	i915_guc_submission_disable(dev_priv);
@@ -587,8 +587,8 @@ int intel_guc_setup(struct drm_i915_private *dev_priv)
 	return ret;
 }
 
-static void guc_fw_fetch(struct drm_i915_private *dev_priv,
-			 struct intel_guc_fw *guc_fw)
+void intel_uc_fw_fetch(struct drm_i915_private *dev_priv,
+			 struct intel_uc_fw *uc_fw)
 {
 	struct pci_dev *pdev = dev_priv->drm.pdev;
 	struct drm_i915_gem_object *obj;
@@ -597,17 +597,17 @@ static void guc_fw_fetch(struct drm_i915_private *dev_priv,
 	size_t size;
 	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));
+	DRM_DEBUG_DRIVER("before requesting firmware: uC fw fetch status %s\n",
+		intel_uc_fw_status_repr(uc_fw->fetch_status));
 
-	err = request_firmware(&fw, guc_fw->guc_fw_path, &pdev->dev);
+	err = request_firmware(&fw, uc_fw->uc_fw_path, &pdev->dev);
 	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);
+	DRM_DEBUG_DRIVER("fetch uC fw from %s succeeded, fw %p\n",
+		uc_fw->uc_fw_path, fw);
 
 	/* Check the size of the blob before examining buffer contents */
 	if (fw->size < sizeof(struct guc_css_header)) {
@@ -618,36 +618,36 @@ static void guc_fw_fetch(struct drm_i915_private *dev_priv,
 	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 -
+	uc_fw->header_offset = 0;
+	uc_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)) {
+	if (uc_fw->header_size != sizeof(struct guc_css_header)) {
 		DRM_NOTE("CSS header definition mismatch\n");
 		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);
+	uc_fw->ucode_offset = uc_fw->header_offset + uc_fw->header_size;
+	uc_fw->ucode_size = (css->size_dw - css->header_size_dw) * sizeof(u32);
 
 	/* now RSA */
 	if (css->key_size_dw != UOS_RSA_SCRATCH_MAX_COUNT) {
 		DRM_NOTE("RSA key size is bad\n");
 		goto fail;
 	}
-	guc_fw->rsa_offset = guc_fw->ucode_offset + guc_fw->ucode_size;
-	guc_fw->rsa_size = css->key_size_dw * sizeof(u32);
+	uc_fw->rsa_offset = uc_fw->ucode_offset + uc_fw->ucode_size;
+	uc_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;
+	size = uc_fw->header_size + uc_fw->ucode_size + uc_fw->rsa_size;
 	if (fw->size < size) {
 		DRM_NOTE("Missing firmware components\n");
 		goto fail;
 	}
 
 	/* Header and uCode will be loaded to WOPCM. Size of the two. */
-	size = guc_fw->header_size + guc_fw->ucode_size;
+	size = uc_fw->header_size + uc_fw->ucode_size;
 	if (size > guc_wopcm_size(dev_priv)) {
 		DRM_NOTE("Firmware is too large to fit in WOPCM\n");
 		goto fail;
@@ -659,21 +659,21 @@ static void guc_fw_fetch(struct drm_i915_private *dev_priv,
 	 * TWO bytes each (i.e. u16), although all pointers and offsets are defined
 	 * in terms of bytes (u8).
 	 */
-	guc_fw->guc_fw_major_found = css->guc_sw_version >> 16;
-	guc_fw->guc_fw_minor_found = css->guc_sw_version & 0xFFFF;
-
-	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) {
-		DRM_NOTE("GuC firmware version %d.%d, required %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);
+	uc_fw->major_ver_found = css->guc_sw_version >> 16;
+	uc_fw->minor_ver_found = css->guc_sw_version & 0xFFFF;
+
+	if (uc_fw->major_ver_found != uc_fw->major_ver_wanted ||
+	    uc_fw->minor_ver_found < uc_fw->minor_ver_wanted) {
+		DRM_NOTE("uC firmware version %d.%d, required %d.%d\n",
+			uc_fw->major_ver_found, uc_fw->minor_ver_found,
+			uc_fw->major_ver_wanted, uc_fw->minor_ver_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);
+			uc_fw->major_ver_found, uc_fw->minor_ver_found,
+			uc_fw->major_ver_wanted, uc_fw->minor_ver_wanted);
 
 	mutex_lock(&dev_priv->drm.struct_mutex);
 	obj = i915_gem_object_create_from_data(dev_priv, fw->data, fw->size);
@@ -683,31 +683,31 @@ static void guc_fw_fetch(struct drm_i915_private *dev_priv,
 		goto fail;
 	}
 
-	guc_fw->guc_fw_obj = obj;
-	guc_fw->guc_fw_size = fw->size;
+	uc_fw->uc_fw_obj = obj;
+	uc_fw->uc_fw_size = fw->size;
 
-	DRM_DEBUG_DRIVER("GuC fw fetch status SUCCESS, obj %p\n",
-			guc_fw->guc_fw_obj);
+	DRM_DEBUG_DRIVER("uC fw fetch status SUCCESS, obj %p\n",
+			uc_fw->uc_fw_obj);
 
 	release_firmware(fw);
-	guc_fw->guc_fw_fetch_status = GUC_FIRMWARE_SUCCESS;
+	uc_fw->fetch_status = UC_FIRMWARE_SUCCESS;
 	return;
 
 fail:
-	DRM_WARN("Failed to fetch valid GuC firmware from %s (error %d)\n",
-		 guc_fw->guc_fw_path, err);
-	DRM_DEBUG_DRIVER("GuC fw fetch status FAIL; err %d, fw %p, obj %p\n",
-		err, fw, guc_fw->guc_fw_obj);
+	DRM_WARN("Failed to fetch valid uC firmware from %s (error %d)\n",
+		 uc_fw->uc_fw_path, err);
+	DRM_DEBUG_DRIVER("uC fw fetch status FAIL; err %d, fw %p, obj %p\n",
+		err, fw, uc_fw->uc_fw_obj);
 
 	mutex_lock(&dev_priv->drm.struct_mutex);
-	obj = guc_fw->guc_fw_obj;
+	obj = uc_fw->uc_fw_obj;
 	if (obj)
 		i915_gem_object_put(obj);
-	guc_fw->guc_fw_obj = NULL;
+	uc_fw->uc_fw_obj = NULL;
 	mutex_unlock(&dev_priv->drm.struct_mutex);
 
 	release_firmware(fw);		/* OK even if fw is NULL */
-	guc_fw->guc_fw_fetch_status = GUC_FIRMWARE_FAIL;
+	uc_fw->fetch_status = UC_FIRMWARE_FAIL;
 }
 
 /**
@@ -721,7 +721,7 @@ static void guc_fw_fetch(struct drm_i915_private *dev_priv,
  */
 void intel_guc_init(struct drm_i915_private *dev_priv)
 {
-	struct intel_guc_fw *guc_fw = &dev_priv->guc.guc_fw;
+	struct intel_uc_fw *guc_fw = &dev_priv->guc.guc_fw;
 	const char *fw_path;
 
 	if (!HAS_GUC(dev_priv)) {
@@ -739,23 +739,24 @@ void intel_guc_init(struct drm_i915_private *dev_priv)
 		fw_path = NULL;
 	} else if (IS_SKYLAKE(dev_priv)) {
 		fw_path = I915_SKL_GUC_UCODE;
-		guc_fw->guc_fw_major_wanted = SKL_FW_MAJOR;
-		guc_fw->guc_fw_minor_wanted = SKL_FW_MINOR;
+		guc_fw->major_ver_wanted = SKL_FW_MAJOR;
+		guc_fw->minor_ver_wanted = SKL_FW_MINOR;
 	} else if (IS_BROXTON(dev_priv)) {
 		fw_path = I915_BXT_GUC_UCODE;
-		guc_fw->guc_fw_major_wanted = BXT_FW_MAJOR;
-		guc_fw->guc_fw_minor_wanted = BXT_FW_MINOR;
+		guc_fw->major_ver_wanted = BXT_FW_MAJOR;
+		guc_fw->minor_ver_wanted = BXT_FW_MINOR;
 	} else if (IS_KABYLAKE(dev_priv)) {
 		fw_path = I915_KBL_GUC_UCODE;
-		guc_fw->guc_fw_major_wanted = KBL_FW_MAJOR;
-		guc_fw->guc_fw_minor_wanted = KBL_FW_MINOR;
+		guc_fw->major_ver_wanted = KBL_FW_MAJOR;
+		guc_fw->minor_ver_wanted = KBL_FW_MINOR;
 	} else {
 		fw_path = "";	/* unknown device */
 	}
 
-	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;
+	guc_fw->uc_dev = &dev_priv->drm;
+	guc_fw->uc_fw_path = fw_path;
+	guc_fw->fetch_status = UC_FIRMWARE_NONE;
+	guc_fw->load_status = UC_FIRMWARE_NONE;
 
 	/* Early (and silent) return if GuC loading is disabled */
 	if (!i915.enable_guc_loading)
@@ -765,9 +766,9 @@ void intel_guc_init(struct drm_i915_private *dev_priv)
 	if (*fw_path == '\0')
 		return;
 
-	guc_fw->guc_fw_fetch_status = GUC_FIRMWARE_PENDING;
+	guc_fw->fetch_status = UC_FIRMWARE_PENDING;
 	DRM_DEBUG_DRIVER("GuC firmware pending, path %s\n", fw_path);
-	guc_fw_fetch(dev_priv, guc_fw);
+	intel_uc_fw_fetch(dev_priv, guc_fw);
 	/* status must now be FAIL or SUCCESS */
 }
 
@@ -777,17 +778,17 @@ void intel_guc_init(struct drm_i915_private *dev_priv)
  */
 void intel_guc_fini(struct drm_i915_private *dev_priv)
 {
-	struct intel_guc_fw *guc_fw = &dev_priv->guc.guc_fw;
+	struct intel_uc_fw *guc_fw = &dev_priv->guc.guc_fw;
 
 	mutex_lock(&dev_priv->drm.struct_mutex);
 	guc_interrupts_release(dev_priv);
 	i915_guc_submission_disable(dev_priv);
 	i915_guc_submission_fini(dev_priv);
 
-	if (guc_fw->guc_fw_obj)
-		i915_gem_object_put(guc_fw->guc_fw_obj);
-	guc_fw->guc_fw_obj = NULL;
+	if (guc_fw->uc_fw_obj)
+		i915_gem_object_put(guc_fw->uc_fw_obj);
+	guc_fw->uc_fw_obj = NULL;
 	mutex_unlock(&dev_priv->drm.struct_mutex);
 
-	guc_fw->guc_fw_fetch_status = GUC_FIRMWARE_NONE;
+	guc_fw->fetch_status = UC_FIRMWARE_NONE;
 }
drivers/gpu/drm/i915/intel_uc.h +0 -0
index 11f5608..f9f598d 100644
@@ -91,29 +91,31 @@ struct i915_guc_client {
 	uint64_t submissions[I915_NUM_ENGINES];
 };
 

drivers/gpu/drm/i915/intel_uc.h struct i915_guc_client { · line 98

Quoted 10 lines
drivers/gpu/drm/i915/intel_uc.h +5 -5
-enum intel_guc_fw_status {
-	GUC_FIRMWARE_FAIL = -1,
-	GUC_FIRMWARE_NONE = 0,
-	GUC_FIRMWARE_PENDING,
-	GUC_FIRMWARE_SUCCESS
+enum intel_uc_fw_status {
+	UC_FIRMWARE_FAIL = -1,
+	UC_FIRMWARE_NONE = 0,
+	UC_FIRMWARE_PENDING,
+	UC_FIRMWARE_SUCCESS

Shouldn't we use INTEL_UC_ prefix here (in similar way as in GUC_ACTION)?

Quoted 16 lines
drivers/gpu/drm/i915/intel_uc.h +0 -10
 };
 
 /*
  * This structure encapsulates all the data needed during the process
  * of fetching, caching, and loading the firmware image into the GuC.
  */
-struct intel_guc_fw {
-	const char *			guc_fw_path;
-	size_t				guc_fw_size;
-	struct drm_i915_gem_object *	guc_fw_obj;
-	enum intel_guc_fw_status	guc_fw_fetch_status;
-	enum intel_guc_fw_status	guc_fw_load_status;
-
-	uint16_t			guc_fw_major_wanted;
-	uint16_t			guc_fw_minor_wanted;
-	uint16_t			guc_fw_major_found;
Quoted 8 lines

drivers/gpu/drm/i915/intel_uc.h struct i915_guc_client { · line 106

Quoted 4 lines
drivers/gpu/drm/i915/intel_uc.h +2 -2
-	uint16_t			guc_fw_minor_found;
-
+struct intel_uc_fw {
+	struct drm_device *uc_dev;

Why do you need this back pointer to drm_dev ? More likely all you need is pointer to dev_priv. But for dev_priv you can always use guc_to_i915().

It seems like a left over from the old days. I have a series that refactors all things GuC that I'll push here soon, including the removal of this field.

Since this series advertises as a simple renaming I didn't force it here.

Quoted 7 lines

drivers/gpu/drm/i915/intel_uc.h struct i915_guc_client { · line 109

Quoted 3 lines
drivers/gpu/drm/i915/intel_uc.h +3 -0
+	const char *uc_fw_path;
+	size_t uc_fw_size;
+	struct drm_i915_gem_object *uc_fw_obj;

All these uc_fw_ prefixes seem to be redundant as this struct is just about uc_fw.

Quoted 11 lines

drivers/gpu/drm/i915/intel_uc.h struct i915_guc_client { · line 118

Quoted 9 lines
drivers/gpu/drm/i915/intel_uc.h +9 -0
+	enum intel_uc_fw_status fetch_status;
+	enum intel_uc_fw_status load_status;
+
+	uint16_t major_ver_wanted;
+	uint16_t minor_ver_wanted;
+	uint16_t major_ver_found;
+	uint16_t minor_ver_found;
+
+	uint32_t fw_type;

What is the purpose of this field ?

It is used in the patch 2, to set {major,minor}_ver_found depending on which firmware we pass to intel_uc_fw_fetch()

Quoted 37 lines

drivers/gpu/drm/i915/intel_uc.h struct intel_guc_log { · line 144

Quoted 10 lines
drivers/gpu/drm/i915/intel_uc.h +1 -1
 	uint32_t header_size;
 	uint32_t header_offset;
 	uint32_t rsa_size;
@@ -139,7 +141,7 @@ struct intel_guc_log {
 };
 
 struct intel_guc {
-	struct intel_guc_fw guc_fw;
+	struct intel_uc_fw guc_fw;

Btw, guc_ prefix in field name here is also redundant.

Quoted 19 lines
drivers/gpu/drm/i915/intel_uc.h +1 -1
 	struct intel_guc_log log;
 
 	/* intel_guc_recv interrupt related state */
@@ -181,7 +183,7 @@ int intel_guc_log_control(struct intel_guc *guc, u32 control_val);
 extern void intel_guc_init(struct drm_i915_private *dev_priv);
 extern int intel_guc_setup(struct drm_i915_private *dev_priv);
 extern void intel_guc_fini(struct drm_i915_private *dev_priv);
-extern const char *intel_guc_fw_status_repr(enum intel_guc_fw_status status);
+extern const char *intel_uc_fw_status_repr(enum intel_uc_fw_status status);
 extern int intel_guc_suspend(struct drm_i915_private *dev_priv);
 extern int intel_guc_resume(struct drm_i915_private *dev_priv);
 
Signature · 6 lines
2.7.4

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Signature · 6 lines
Cheers,
Arek
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
#5 lore raw

Re: [PATCH 1/8] drm/i915/guc: Make the GuC fw loading helper functions general

Quoted 503 lines · Arkadiusz Hiler
Quoted 493 lines · Michal Wajdeczko
Quoted 462 lines · anushasr

From: Peter Antoine <[email protected]>

Rename some of the GuC fw loading code to make them more general. We will utilise them for HuC loading as well.
s/intel_guc_fw/intel_uc_fw/g s/GUC_FIRMWARE/UC_FIRMWARE/g

Struct intel_guc_fw is renamed to intel_uc_fw. Prefix of tts members, such as 'guc' or 'guc_fw' either is renamed to 'uc' or removed for same purpose.

v2: rebased on top of nightly.
reapplied the search/replace as upstream code as changed.
v3: rebased again on drm-nightly.
v4: removed G from messages in shared fw fetch function.
v5: rebased.
v7: rebased.
v8: rebased.
v9: rebased.
v10: rebased.
v11: rebased.
v12: rebased on top of drm-tip
v13: rebased.Updated dev to dev_priv in intel_guc_setup(), guc_fw_getch() and intel_guc_init().

---
 drivers/gpu/drm/i915/i915_debugfs.c        |  12 +--
 drivers/gpu/drm/i915/i915_guc_submission.c |   4 +-
 drivers/gpu/drm/i915/intel_guc_loader.c    | 157 +++++++++++++++--------------
 drivers/gpu/drm/i915/intel_uc.h            |  40 ++++----
 4 files changed, 108 insertions(+), 105 deletions(-)
drivers/gpu/drm/i915/i915_debugfs.c +6 -6
index a746130..0e5ef62 100644
@@ -2338,7 +2338,7 @@ static int i915_llc(struct seq_file *m, void *data)
 static int i915_guc_load_status_info(struct seq_file *m, void *data)
 {
 	struct drm_i915_private *dev_priv = node_to_i915(m->private);
-	struct intel_guc_fw *guc_fw = &dev_priv->guc.guc_fw;
+	struct intel_uc_fw *guc_fw = &dev_priv->guc.guc_fw;
 	u32 tmp, i;
 
 	if (!HAS_GUC_UCODE(dev_priv))
@@ -2346,15 +2346,15 @@ static int i915_guc_load_status_info(struct seq_file *m, void *data)
 
 	seq_printf(m, "GuC firmware status:\n");
 	seq_printf(m, "\tpath: %s\n",
-		guc_fw->guc_fw_path);
+		guc_fw->uc_fw_path);
 	seq_printf(m, "\tfetch: %s\n",
-		intel_guc_fw_status_repr(guc_fw->guc_fw_fetch_status));
+		intel_uc_fw_status_repr(guc_fw->fetch_status));
 	seq_printf(m, "\tload: %s\n",
-		intel_guc_fw_status_repr(guc_fw->guc_fw_load_status));
+		intel_uc_fw_status_repr(guc_fw->load_status));
 	seq_printf(m, "\tversion wanted: %d.%d\n",
-		guc_fw->guc_fw_major_wanted, guc_fw->guc_fw_minor_wanted);
+		guc_fw->major_ver_wanted, guc_fw->minor_ver_wanted);
 	seq_printf(m, "\tversion found: %d.%d\n",
-		guc_fw->guc_fw_major_found, guc_fw->guc_fw_minor_found);
+		guc_fw->major_ver_found, guc_fw->minor_ver_found);
 	seq_printf(m, "\theader: offset is %d; size = %d\n",
 		guc_fw->header_offset, guc_fw->header_size);
 	seq_printf(m, "\tuCode: offset is %d; size = %d\n",
drivers/gpu/drm/i915/i915_guc_submission.c +2 -2
index 7fa4e74..e156a4b 100644
@@ -1493,7 +1493,7 @@ int intel_guc_suspend(struct drm_i915_private *dev_priv)
 	struct i915_gem_context *ctx;
 	u32 data[3];
 
-	if (guc->guc_fw.guc_fw_load_status != GUC_FIRMWARE_SUCCESS)
+	if (guc->guc_fw.load_status != UC_FIRMWARE_SUCCESS)
 		return 0;
 
 	gen9_disable_guc_interrupts(dev_priv);
@@ -1520,7 +1520,7 @@ int intel_guc_resume(struct drm_i915_private *dev_priv)
 	struct i915_gem_context *ctx;
 	u32 data[3];
 
-	if (guc->guc_fw.guc_fw_load_status != GUC_FIRMWARE_SUCCESS)
+	if (guc->guc_fw.load_status != UC_FIRMWARE_SUCCESS)
 		return 0;
 
 	if (i915.guc_log_level >= 0)
drivers/gpu/drm/i915/intel_guc_loader.c +79 -78
index 21db697..8f04f6e 100644
@@ -81,16 +81,16 @@ MODULE_FIRMWARE(I915_BXT_GUC_UCODE);
 MODULE_FIRMWARE(I915_KBL_GUC_UCODE);
 
 /* User-friendly representation of an enum */
-const char *intel_guc_fw_status_repr(enum intel_guc_fw_status status)
+const char *intel_uc_fw_status_repr(enum intel_uc_fw_status status)
 {
 	switch (status) {
-	case GUC_FIRMWARE_FAIL:
+	case UC_FIRMWARE_FAIL:
 		return "FAIL";
-	case GUC_FIRMWARE_NONE:
+	case UC_FIRMWARE_NONE:
 		return "NONE";
-	case GUC_FIRMWARE_PENDING:
+	case UC_FIRMWARE_PENDING:
 		return "PENDING";
-	case GUC_FIRMWARE_SUCCESS:
+	case UC_FIRMWARE_SUCCESS:
 		return "SUCCESS";
 	default:
 		return "UNKNOWN!";
@@ -278,7 +278,7 @@ static inline bool guc_ucode_response(struct drm_i915_private *dev_priv,
 static int guc_ucode_xfer_dma(struct drm_i915_private *dev_priv,
 			      struct i915_vma *vma)
 {
-	struct intel_guc_fw *guc_fw = &dev_priv->guc.guc_fw;
+	struct intel_uc_fw *guc_fw = &dev_priv->guc.guc_fw;
 	unsigned long offset;
 	struct sg_table *sg = vma->pages;
 	u32 status, rsa[UOS_RSA_SCRATCH_MAX_COUNT];
@@ -350,17 +350,17 @@ static u32 guc_wopcm_size(struct drm_i915_private *dev_priv)
  */
 static int guc_ucode_xfer(struct drm_i915_private *dev_priv)
 {
-	struct intel_guc_fw *guc_fw = &dev_priv->guc.guc_fw;
+	struct intel_uc_fw *guc_fw = &dev_priv->guc.guc_fw;
 	struct i915_vma *vma;
 	int ret;
 
-	ret = i915_gem_object_set_to_gtt_domain(guc_fw->guc_fw_obj, false);
+	ret = i915_gem_object_set_to_gtt_domain(guc_fw->uc_fw_obj, false);
 	if (ret) {
 		DRM_DEBUG_DRIVER("set-domain failed %d\n", ret);
 		return ret;
 	}
 
-	vma = i915_gem_object_ggtt_pin(guc_fw->guc_fw_obj, NULL, 0, 0, 0);
+	vma = i915_gem_object_ggtt_pin(guc_fw->uc_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);
@@ -450,14 +450,14 @@ static int guc_hw_reset(struct drm_i915_private *dev_priv)
  */
 int intel_guc_setup(struct drm_i915_private *dev_priv)
 {
-	struct intel_guc_fw *guc_fw = &dev_priv->guc.guc_fw;
-	const char *fw_path = guc_fw->guc_fw_path;
+	struct intel_uc_fw *guc_fw = &dev_priv->guc.guc_fw;
+	const char *fw_path = guc_fw->uc_fw_path;
 	int retries, ret, err;
 
 	DRM_DEBUG_DRIVER("GuC fw status: path %s, fetch %s, load %s\n",
 		fw_path,
-		intel_guc_fw_status_repr(guc_fw->guc_fw_fetch_status),
-		intel_guc_fw_status_repr(guc_fw->guc_fw_load_status));
+		intel_uc_fw_status_repr(guc_fw->fetch_status),
+		intel_uc_fw_status_repr(guc_fw->load_status));
 
 	/* Loading forbidden, or no firmware to load? */
 	if (!i915.enable_guc_loading) {
@@ -475,10 +475,10 @@ int intel_guc_setup(struct drm_i915_private *dev_priv)
 	}
 
 	/* Fetch failed, or already fetched but failed to load? */
-	if (guc_fw->guc_fw_fetch_status != GUC_FIRMWARE_SUCCESS) {
+	if (guc_fw->fetch_status != UC_FIRMWARE_SUCCESS) {
 		err = -EIO;
 		goto fail;
-	} else if (guc_fw->guc_fw_load_status == GUC_FIRMWARE_FAIL) {
+	} else if (guc_fw->load_status == UC_FIRMWARE_FAIL) {
 		err = -ENOEXEC;
 		goto fail;
 	}
@@ -486,11 +486,11 @@ int intel_guc_setup(struct drm_i915_private *dev_priv)
 	guc_interrupts_release(dev_priv);
 	gen9_reset_guc_interrupts(dev_priv);
 
-	guc_fw->guc_fw_load_status = GUC_FIRMWARE_PENDING;
+	guc_fw->load_status = UC_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));
+		intel_uc_fw_status_repr(guc_fw->fetch_status),
+		intel_uc_fw_status_repr(guc_fw->load_status));
 
 	err = i915_guc_submission_init(dev_priv);
 	if (err)
@@ -522,11 +522,11 @@ int intel_guc_setup(struct drm_i915_private *dev_priv)
 			 "retry %d more time(s)\n", err, retries);
 	}
 
-	guc_fw->guc_fw_load_status = GUC_FIRMWARE_SUCCESS;
+	guc_fw->load_status = UC_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));
+		intel_uc_fw_status_repr(guc_fw->fetch_status),
+		intel_uc_fw_status_repr(guc_fw->load_status));
 
 	if (i915.enable_guc_submission) {
 		if (i915.guc_log_level >= 0)
@@ -541,8 +541,8 @@ int intel_guc_setup(struct drm_i915_private *dev_priv)
 	return 0;
 
 fail:
-	if (guc_fw->guc_fw_load_status == GUC_FIRMWARE_PENDING)
-		guc_fw->guc_fw_load_status = GUC_FIRMWARE_FAIL;
+	if (guc_fw->load_status == UC_FIRMWARE_PENDING)
+		guc_fw->load_status = UC_FIRMWARE_FAIL;
 
 	guc_interrupts_release(dev_priv);
 	i915_guc_submission_disable(dev_priv);
@@ -587,8 +587,8 @@ int intel_guc_setup(struct drm_i915_private *dev_priv)
 	return ret;
 }
 
-static void guc_fw_fetch(struct drm_i915_private *dev_priv,
-			 struct intel_guc_fw *guc_fw)
+void intel_uc_fw_fetch(struct drm_i915_private *dev_priv,
+			 struct intel_uc_fw *uc_fw)
 {
 	struct pci_dev *pdev = dev_priv->drm.pdev;
 	struct drm_i915_gem_object *obj;
@@ -597,17 +597,17 @@ static void guc_fw_fetch(struct drm_i915_private *dev_priv,
 	size_t size;
 	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));
+	DRM_DEBUG_DRIVER("before requesting firmware: uC fw fetch status %s\n",
+		intel_uc_fw_status_repr(uc_fw->fetch_status));
 
-	err = request_firmware(&fw, guc_fw->guc_fw_path, &pdev->dev);
+	err = request_firmware(&fw, uc_fw->uc_fw_path, &pdev->dev);
 	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);
+	DRM_DEBUG_DRIVER("fetch uC fw from %s succeeded, fw %p\n",
+		uc_fw->uc_fw_path, fw);
 
 	/* Check the size of the blob before examining buffer contents */
 	if (fw->size < sizeof(struct guc_css_header)) {
@@ -618,36 +618,36 @@ static void guc_fw_fetch(struct drm_i915_private *dev_priv,
 	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 -
+	uc_fw->header_offset = 0;
+	uc_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)) {
+	if (uc_fw->header_size != sizeof(struct guc_css_header)) {
 		DRM_NOTE("CSS header definition mismatch\n");
 		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);
+	uc_fw->ucode_offset = uc_fw->header_offset + uc_fw->header_size;
+	uc_fw->ucode_size = (css->size_dw - css->header_size_dw) * sizeof(u32);
 
 	/* now RSA */
 	if (css->key_size_dw != UOS_RSA_SCRATCH_MAX_COUNT) {
 		DRM_NOTE("RSA key size is bad\n");
 		goto fail;
 	}
-	guc_fw->rsa_offset = guc_fw->ucode_offset + guc_fw->ucode_size;
-	guc_fw->rsa_size = css->key_size_dw * sizeof(u32);
+	uc_fw->rsa_offset = uc_fw->ucode_offset + uc_fw->ucode_size;
+	uc_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;
+	size = uc_fw->header_size + uc_fw->ucode_size + uc_fw->rsa_size;
 	if (fw->size < size) {
 		DRM_NOTE("Missing firmware components\n");
 		goto fail;
 	}
 
 	/* Header and uCode will be loaded to WOPCM. Size of the two. */
-	size = guc_fw->header_size + guc_fw->ucode_size;
+	size = uc_fw->header_size + uc_fw->ucode_size;
 	if (size > guc_wopcm_size(dev_priv)) {
 		DRM_NOTE("Firmware is too large to fit in WOPCM\n");
 		goto fail;
@@ -659,21 +659,21 @@ static void guc_fw_fetch(struct drm_i915_private *dev_priv,
 	 * TWO bytes each (i.e. u16), although all pointers and offsets are defined
 	 * in terms of bytes (u8).
 	 */
-	guc_fw->guc_fw_major_found = css->guc_sw_version >> 16;
-	guc_fw->guc_fw_minor_found = css->guc_sw_version & 0xFFFF;
-
-	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) {
-		DRM_NOTE("GuC firmware version %d.%d, required %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);
+	uc_fw->major_ver_found = css->guc_sw_version >> 16;
+	uc_fw->minor_ver_found = css->guc_sw_version & 0xFFFF;
+
+	if (uc_fw->major_ver_found != uc_fw->major_ver_wanted ||
+	    uc_fw->minor_ver_found < uc_fw->minor_ver_wanted) {
+		DRM_NOTE("uC firmware version %d.%d, required %d.%d\n",
+			uc_fw->major_ver_found, uc_fw->minor_ver_found,
+			uc_fw->major_ver_wanted, uc_fw->minor_ver_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);
+			uc_fw->major_ver_found, uc_fw->minor_ver_found,
+			uc_fw->major_ver_wanted, uc_fw->minor_ver_wanted);
 
 	mutex_lock(&dev_priv->drm.struct_mutex);
 	obj = i915_gem_object_create_from_data(dev_priv, fw->data, fw->size);
@@ -683,31 +683,31 @@ static void guc_fw_fetch(struct drm_i915_private *dev_priv,
 		goto fail;
 	}
 
-	guc_fw->guc_fw_obj = obj;
-	guc_fw->guc_fw_size = fw->size;
+	uc_fw->uc_fw_obj = obj;
+	uc_fw->uc_fw_size = fw->size;
 
-	DRM_DEBUG_DRIVER("GuC fw fetch status SUCCESS, obj %p\n",
-			guc_fw->guc_fw_obj);
+	DRM_DEBUG_DRIVER("uC fw fetch status SUCCESS, obj %p\n",
+			uc_fw->uc_fw_obj);
 
 	release_firmware(fw);
-	guc_fw->guc_fw_fetch_status = GUC_FIRMWARE_SUCCESS;
+	uc_fw->fetch_status = UC_FIRMWARE_SUCCESS;
 	return;
 
 fail:
-	DRM_WARN("Failed to fetch valid GuC firmware from %s (error %d)\n",
-		 guc_fw->guc_fw_path, err);
-	DRM_DEBUG_DRIVER("GuC fw fetch status FAIL; err %d, fw %p, obj %p\n",
-		err, fw, guc_fw->guc_fw_obj);
+	DRM_WARN("Failed to fetch valid uC firmware from %s (error %d)\n",
+		 uc_fw->uc_fw_path, err);
+	DRM_DEBUG_DRIVER("uC fw fetch status FAIL; err %d, fw %p, obj %p\n",
+		err, fw, uc_fw->uc_fw_obj);
 
 	mutex_lock(&dev_priv->drm.struct_mutex);
-	obj = guc_fw->guc_fw_obj;
+	obj = uc_fw->uc_fw_obj;
 	if (obj)
 		i915_gem_object_put(obj);
-	guc_fw->guc_fw_obj = NULL;
+	uc_fw->uc_fw_obj = NULL;
 	mutex_unlock(&dev_priv->drm.struct_mutex);
 
 	release_firmware(fw);		/* OK even if fw is NULL */
-	guc_fw->guc_fw_fetch_status = GUC_FIRMWARE_FAIL;
+	uc_fw->fetch_status = UC_FIRMWARE_FAIL;
 }
 
 /**
@@ -721,7 +721,7 @@ static void guc_fw_fetch(struct drm_i915_private *dev_priv,
  */
 void intel_guc_init(struct drm_i915_private *dev_priv)
 {
-	struct intel_guc_fw *guc_fw = &dev_priv->guc.guc_fw;
+	struct intel_uc_fw *guc_fw = &dev_priv->guc.guc_fw;
 	const char *fw_path;
 
 	if (!HAS_GUC(dev_priv)) {
@@ -739,23 +739,24 @@ void intel_guc_init(struct drm_i915_private *dev_priv)
 		fw_path = NULL;
 	} else if (IS_SKYLAKE(dev_priv)) {
 		fw_path = I915_SKL_GUC_UCODE;
-		guc_fw->guc_fw_major_wanted = SKL_FW_MAJOR;
-		guc_fw->guc_fw_minor_wanted = SKL_FW_MINOR;
+		guc_fw->major_ver_wanted = SKL_FW_MAJOR;
+		guc_fw->minor_ver_wanted = SKL_FW_MINOR;
 	} else if (IS_BROXTON(dev_priv)) {
 		fw_path = I915_BXT_GUC_UCODE;
-		guc_fw->guc_fw_major_wanted = BXT_FW_MAJOR;
-		guc_fw->guc_fw_minor_wanted = BXT_FW_MINOR;
+		guc_fw->major_ver_wanted = BXT_FW_MAJOR;
+		guc_fw->minor_ver_wanted = BXT_FW_MINOR;
 	} else if (IS_KABYLAKE(dev_priv)) {
 		fw_path = I915_KBL_GUC_UCODE;
-		guc_fw->guc_fw_major_wanted = KBL_FW_MAJOR;
-		guc_fw->guc_fw_minor_wanted = KBL_FW_MINOR;
+		guc_fw->major_ver_wanted = KBL_FW_MAJOR;
+		guc_fw->minor_ver_wanted = KBL_FW_MINOR;
 	} else {
 		fw_path = "";	/* unknown device */
 	}
 
-	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;
+	guc_fw->uc_dev = &dev_priv->drm;
+	guc_fw->uc_fw_path = fw_path;
+	guc_fw->fetch_status = UC_FIRMWARE_NONE;
+	guc_fw->load_status = UC_FIRMWARE_NONE;
 
 	/* Early (and silent) return if GuC loading is disabled */
 	if (!i915.enable_guc_loading)
@@ -765,9 +766,9 @@ void intel_guc_init(struct drm_i915_private *dev_priv)
 	if (*fw_path == '\0')
 		return;
 
-	guc_fw->guc_fw_fetch_status = GUC_FIRMWARE_PENDING;
+	guc_fw->fetch_status = UC_FIRMWARE_PENDING;
 	DRM_DEBUG_DRIVER("GuC firmware pending, path %s\n", fw_path);
-	guc_fw_fetch(dev_priv, guc_fw);
+	intel_uc_fw_fetch(dev_priv, guc_fw);
 	/* status must now be FAIL or SUCCESS */
 }
 
@@ -777,17 +778,17 @@ void intel_guc_init(struct drm_i915_private *dev_priv)
  */
 void intel_guc_fini(struct drm_i915_private *dev_priv)
 {
-	struct intel_guc_fw *guc_fw = &dev_priv->guc.guc_fw;
+	struct intel_uc_fw *guc_fw = &dev_priv->guc.guc_fw;
 
 	mutex_lock(&dev_priv->drm.struct_mutex);
 	guc_interrupts_release(dev_priv);
 	i915_guc_submission_disable(dev_priv);
 	i915_guc_submission_fini(dev_priv);
 
-	if (guc_fw->guc_fw_obj)
-		i915_gem_object_put(guc_fw->guc_fw_obj);
-	guc_fw->guc_fw_obj = NULL;
+	if (guc_fw->uc_fw_obj)
+		i915_gem_object_put(guc_fw->uc_fw_obj);
+	guc_fw->uc_fw_obj = NULL;
 	mutex_unlock(&dev_priv->drm.struct_mutex);
 
-	guc_fw->guc_fw_fetch_status = GUC_FIRMWARE_NONE;
+	guc_fw->fetch_status = UC_FIRMWARE_NONE;
 }
drivers/gpu/drm/i915/intel_uc.h +0 -0
index 11f5608..f9f598d 100644
@@ -91,29 +91,31 @@ struct i915_guc_client {
 	uint64_t submissions[I915_NUM_ENGINES];
 };
 

drivers/gpu/drm/i915/intel_uc.h struct i915_guc_client { · line 98

Quoted 10 lines
drivers/gpu/drm/i915/intel_uc.h +5 -5
-enum intel_guc_fw_status {
-	GUC_FIRMWARE_FAIL = -1,
-	GUC_FIRMWARE_NONE = 0,
-	GUC_FIRMWARE_PENDING,
-	GUC_FIRMWARE_SUCCESS
+enum intel_uc_fw_status {
+	UC_FIRMWARE_FAIL = -1,
+	UC_FIRMWARE_NONE = 0,
+	UC_FIRMWARE_PENDING,
+	UC_FIRMWARE_SUCCESS

Shouldn't we use INTEL_UC_ prefix here (in similar way as in GUC_ACTION)?

Quoted 16 lines
drivers/gpu/drm/i915/intel_uc.h +0 -10
 };
 
 /*
  * This structure encapsulates all the data needed during the process
  * of fetching, caching, and loading the firmware image into the GuC.
  */
-struct intel_guc_fw {
-	const char *			guc_fw_path;
-	size_t				guc_fw_size;
-	struct drm_i915_gem_object *	guc_fw_obj;
-	enum intel_guc_fw_status	guc_fw_fetch_status;
-	enum intel_guc_fw_status	guc_fw_load_status;
-
-	uint16_t			guc_fw_major_wanted;
-	uint16_t			guc_fw_minor_wanted;
-	uint16_t			guc_fw_major_found;
Quoted 8 lines

drivers/gpu/drm/i915/intel_uc.h struct i915_guc_client { · line 106

Quoted 4 lines
drivers/gpu/drm/i915/intel_uc.h +2 -2
-	uint16_t			guc_fw_minor_found;
-
+struct intel_uc_fw {
+	struct drm_device *uc_dev;

Why do you need this back pointer to drm_dev ? More likely all you need is pointer to dev_priv. But for dev_priv you can always use guc_to_i915().

Quoted 6 lines

It seems like a left over from the old days. I have a series that refactors all things GuC that I'll push here soon, including the removal of this field.

Since this series advertises as a simple renaming I didn't force it here.

If you want to treat this as simple 'renaming' then 'adding' unused field is even more undesired and shall be avoided.

Quoted 20 lines
Quoted 8 lines

drivers/gpu/drm/i915/intel_uc.h struct i915_guc_client { · line 109

Quoted 4 lines
drivers/gpu/drm/i915/intel_uc.h +3 -0
+	const char *uc_fw_path;
+	size_t uc_fw_size;
+	struct drm_i915_gem_object *uc_fw_obj;

All these uc_fw_ prefixes seem to be redundant as this struct is just about uc_fw.

Quoted 11 lines

drivers/gpu/drm/i915/intel_uc.h struct i915_guc_client { · line 118

Quoted 9 lines
drivers/gpu/drm/i915/intel_uc.h +9 -0
+	enum intel_uc_fw_status fetch_status;
+	enum intel_uc_fw_status load_status;
+
+	uint16_t major_ver_wanted;
+	uint16_t minor_ver_wanted;
+	uint16_t major_ver_found;
+	uint16_t minor_ver_found;
+
+	uint32_t fw_type;

What is the purpose of this field ?

Quoted 2 lines

It is used in the patch 2, to set {major,minor}_ver_found depending on which firmware we pass to intel_uc_fw_fetch()

Then maybe is shall be moved to patch 2 ?

Quoted 30 lines
Quoted 30 lines

drivers/gpu/drm/i915/intel_uc.h struct intel_guc_log { · line 144

Quoted 11 lines
drivers/gpu/drm/i915/intel_uc.h +1 -1
 	uint32_t header_size;
 	uint32_t header_offset;
 	uint32_t rsa_size;
@@ -139,7 +141,7 @@ struct intel_guc_log {
 };
 
 struct intel_guc {
-	struct intel_guc_fw guc_fw;
+	struct intel_uc_fw guc_fw;

Btw, guc_ prefix in field name here is also redundant.

Quoted 15 lines
drivers/gpu/drm/i915/intel_uc.h +1 -1
 	struct intel_guc_log log;
 
 	/* intel_guc_recv interrupt related state */
@@ -181,7 +183,7 @@ int intel_guc_log_control(struct intel_guc *guc, u32 control_val);
 extern void intel_guc_init(struct drm_i915_private *dev_priv);
 extern int intel_guc_setup(struct drm_i915_private *dev_priv);
 extern void intel_guc_fini(struct drm_i915_private *dev_priv);
-extern const char *intel_guc_fw_status_repr(enum intel_guc_fw_status status);
+extern const char *intel_uc_fw_status_repr(enum intel_uc_fw_status status);
 extern int intel_guc_suspend(struct drm_i915_private *dev_priv);
 extern int intel_guc_resume(struct drm_i915_private *dev_priv);
 
Signature · 1 line
2.7.4
Quoted 16 lines
Quoted 8 lines
Quoted 4 lines

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Signature · 6 lines
Cheers,
Arek
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

#6 lore raw

Re: [PATCH 1/8] drm/i915/guc: Make the GuC fw loading helper functions general

Quoted 534 lines · Michal Wajdeczko
Quoted 503 lines · Arkadiusz Hiler
Quoted 493 lines · Michal Wajdeczko
Quoted 462 lines · anushasr

From: Peter Antoine <[email protected]>

Rename some of the GuC fw loading code to make them more general. We will utilise them for HuC loading as well.
s/intel_guc_fw/intel_uc_fw/g s/GUC_FIRMWARE/UC_FIRMWARE/g

Struct intel_guc_fw is renamed to intel_uc_fw. Prefix of tts members, such as 'guc' or 'guc_fw' either is renamed to 'uc' or removed for same purpose.

v2: rebased on top of nightly.
reapplied the search/replace as upstream code as changed.
v3: rebased again on drm-nightly.
v4: removed G from messages in shared fw fetch function.
v5: rebased.
v7: rebased.
v8: rebased.
v9: rebased.
v10: rebased.
v11: rebased.
v12: rebased on top of drm-tip
v13: rebased.Updated dev to dev_priv in intel_guc_setup(), guc_fw_getch() and intel_guc_init().

---
 drivers/gpu/drm/i915/i915_debugfs.c        |  12 +--
 drivers/gpu/drm/i915/i915_guc_submission.c |   4 +-
 drivers/gpu/drm/i915/intel_guc_loader.c    | 157 +++++++++++++++--------------
 drivers/gpu/drm/i915/intel_uc.h            |  40 ++++----
 4 files changed, 108 insertions(+), 105 deletions(-)
drivers/gpu/drm/i915/i915_debugfs.c +6 -6
index a746130..0e5ef62 100644
@@ -2338,7 +2338,7 @@ static int i915_llc(struct seq_file *m, void *data)
 static int i915_guc_load_status_info(struct seq_file *m, void *data)
 {
 	struct drm_i915_private *dev_priv = node_to_i915(m->private);
-	struct intel_guc_fw *guc_fw = &dev_priv->guc.guc_fw;
+	struct intel_uc_fw *guc_fw = &dev_priv->guc.guc_fw;
 	u32 tmp, i;
 
 	if (!HAS_GUC_UCODE(dev_priv))
@@ -2346,15 +2346,15 @@ static int i915_guc_load_status_info(struct seq_file *m, void *data)
 
 	seq_printf(m, "GuC firmware status:\n");
 	seq_printf(m, "\tpath: %s\n",
-		guc_fw->guc_fw_path);
+		guc_fw->uc_fw_path);
 	seq_printf(m, "\tfetch: %s\n",
-		intel_guc_fw_status_repr(guc_fw->guc_fw_fetch_status));
+		intel_uc_fw_status_repr(guc_fw->fetch_status));
 	seq_printf(m, "\tload: %s\n",
-		intel_guc_fw_status_repr(guc_fw->guc_fw_load_status));
+		intel_uc_fw_status_repr(guc_fw->load_status));
 	seq_printf(m, "\tversion wanted: %d.%d\n",
-		guc_fw->guc_fw_major_wanted, guc_fw->guc_fw_minor_wanted);
+		guc_fw->major_ver_wanted, guc_fw->minor_ver_wanted);
 	seq_printf(m, "\tversion found: %d.%d\n",
-		guc_fw->guc_fw_major_found, guc_fw->guc_fw_minor_found);
+		guc_fw->major_ver_found, guc_fw->minor_ver_found);
 	seq_printf(m, "\theader: offset is %d; size = %d\n",
 		guc_fw->header_offset, guc_fw->header_size);
 	seq_printf(m, "\tuCode: offset is %d; size = %d\n",
drivers/gpu/drm/i915/i915_guc_submission.c +2 -2
index 7fa4e74..e156a4b 100644
@@ -1493,7 +1493,7 @@ int intel_guc_suspend(struct drm_i915_private *dev_priv)
 	struct i915_gem_context *ctx;
 	u32 data[3];
 
-	if (guc->guc_fw.guc_fw_load_status != GUC_FIRMWARE_SUCCESS)
+	if (guc->guc_fw.load_status != UC_FIRMWARE_SUCCESS)
 		return 0;
 
 	gen9_disable_guc_interrupts(dev_priv);
@@ -1520,7 +1520,7 @@ int intel_guc_resume(struct drm_i915_private *dev_priv)
 	struct i915_gem_context *ctx;
 	u32 data[3];
 
-	if (guc->guc_fw.guc_fw_load_status != GUC_FIRMWARE_SUCCESS)
+	if (guc->guc_fw.load_status != UC_FIRMWARE_SUCCESS)
 		return 0;
 
 	if (i915.guc_log_level >= 0)
drivers/gpu/drm/i915/intel_guc_loader.c +79 -78
index 21db697..8f04f6e 100644
@@ -81,16 +81,16 @@ MODULE_FIRMWARE(I915_BXT_GUC_UCODE);
 MODULE_FIRMWARE(I915_KBL_GUC_UCODE);
 
 /* User-friendly representation of an enum */
-const char *intel_guc_fw_status_repr(enum intel_guc_fw_status status)
+const char *intel_uc_fw_status_repr(enum intel_uc_fw_status status)
 {
 	switch (status) {
-	case GUC_FIRMWARE_FAIL:
+	case UC_FIRMWARE_FAIL:
 		return "FAIL";
-	case GUC_FIRMWARE_NONE:
+	case UC_FIRMWARE_NONE:
 		return "NONE";
-	case GUC_FIRMWARE_PENDING:
+	case UC_FIRMWARE_PENDING:
 		return "PENDING";
-	case GUC_FIRMWARE_SUCCESS:
+	case UC_FIRMWARE_SUCCESS:
 		return "SUCCESS";
 	default:
 		return "UNKNOWN!";
@@ -278,7 +278,7 @@ static inline bool guc_ucode_response(struct drm_i915_private *dev_priv,
 static int guc_ucode_xfer_dma(struct drm_i915_private *dev_priv,
 			      struct i915_vma *vma)
 {
-	struct intel_guc_fw *guc_fw = &dev_priv->guc.guc_fw;
+	struct intel_uc_fw *guc_fw = &dev_priv->guc.guc_fw;
 	unsigned long offset;
 	struct sg_table *sg = vma->pages;
 	u32 status, rsa[UOS_RSA_SCRATCH_MAX_COUNT];
@@ -350,17 +350,17 @@ static u32 guc_wopcm_size(struct drm_i915_private *dev_priv)
  */
 static int guc_ucode_xfer(struct drm_i915_private *dev_priv)
 {
-	struct intel_guc_fw *guc_fw = &dev_priv->guc.guc_fw;
+	struct intel_uc_fw *guc_fw = &dev_priv->guc.guc_fw;
 	struct i915_vma *vma;
 	int ret;
 
-	ret = i915_gem_object_set_to_gtt_domain(guc_fw->guc_fw_obj, false);
+	ret = i915_gem_object_set_to_gtt_domain(guc_fw->uc_fw_obj, false);
 	if (ret) {
 		DRM_DEBUG_DRIVER("set-domain failed %d\n", ret);
 		return ret;
 	}
 
-	vma = i915_gem_object_ggtt_pin(guc_fw->guc_fw_obj, NULL, 0, 0, 0);
+	vma = i915_gem_object_ggtt_pin(guc_fw->uc_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);
@@ -450,14 +450,14 @@ static int guc_hw_reset(struct drm_i915_private *dev_priv)
  */
 int intel_guc_setup(struct drm_i915_private *dev_priv)
 {
-	struct intel_guc_fw *guc_fw = &dev_priv->guc.guc_fw;
-	const char *fw_path = guc_fw->guc_fw_path;
+	struct intel_uc_fw *guc_fw = &dev_priv->guc.guc_fw;
+	const char *fw_path = guc_fw->uc_fw_path;
 	int retries, ret, err;
 
 	DRM_DEBUG_DRIVER("GuC fw status: path %s, fetch %s, load %s\n",
 		fw_path,
-		intel_guc_fw_status_repr(guc_fw->guc_fw_fetch_status),
-		intel_guc_fw_status_repr(guc_fw->guc_fw_load_status));
+		intel_uc_fw_status_repr(guc_fw->fetch_status),
+		intel_uc_fw_status_repr(guc_fw->load_status));
 
 	/* Loading forbidden, or no firmware to load? */
 	if (!i915.enable_guc_loading) {
@@ -475,10 +475,10 @@ int intel_guc_setup(struct drm_i915_private *dev_priv)
 	}
 
 	/* Fetch failed, or already fetched but failed to load? */
-	if (guc_fw->guc_fw_fetch_status != GUC_FIRMWARE_SUCCESS) {
+	if (guc_fw->fetch_status != UC_FIRMWARE_SUCCESS) {
 		err = -EIO;
 		goto fail;
-	} else if (guc_fw->guc_fw_load_status == GUC_FIRMWARE_FAIL) {
+	} else if (guc_fw->load_status == UC_FIRMWARE_FAIL) {
 		err = -ENOEXEC;
 		goto fail;
 	}
@@ -486,11 +486,11 @@ int intel_guc_setup(struct drm_i915_private *dev_priv)
 	guc_interrupts_release(dev_priv);
 	gen9_reset_guc_interrupts(dev_priv);
 
-	guc_fw->guc_fw_load_status = GUC_FIRMWARE_PENDING;
+	guc_fw->load_status = UC_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));
+		intel_uc_fw_status_repr(guc_fw->fetch_status),
+		intel_uc_fw_status_repr(guc_fw->load_status));
 
 	err = i915_guc_submission_init(dev_priv);
 	if (err)
@@ -522,11 +522,11 @@ int intel_guc_setup(struct drm_i915_private *dev_priv)
 			 "retry %d more time(s)\n", err, retries);
 	}
 
-	guc_fw->guc_fw_load_status = GUC_FIRMWARE_SUCCESS;
+	guc_fw->load_status = UC_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));
+		intel_uc_fw_status_repr(guc_fw->fetch_status),
+		intel_uc_fw_status_repr(guc_fw->load_status));
 
 	if (i915.enable_guc_submission) {
 		if (i915.guc_log_level >= 0)
@@ -541,8 +541,8 @@ int intel_guc_setup(struct drm_i915_private *dev_priv)
 	return 0;
 
 fail:
-	if (guc_fw->guc_fw_load_status == GUC_FIRMWARE_PENDING)
-		guc_fw->guc_fw_load_status = GUC_FIRMWARE_FAIL;
+	if (guc_fw->load_status == UC_FIRMWARE_PENDING)
+		guc_fw->load_status = UC_FIRMWARE_FAIL;
 
 	guc_interrupts_release(dev_priv);
 	i915_guc_submission_disable(dev_priv);
@@ -587,8 +587,8 @@ int intel_guc_setup(struct drm_i915_private *dev_priv)
 	return ret;
 }
 
-static void guc_fw_fetch(struct drm_i915_private *dev_priv,
-			 struct intel_guc_fw *guc_fw)
+void intel_uc_fw_fetch(struct drm_i915_private *dev_priv,
+			 struct intel_uc_fw *uc_fw)
 {
 	struct pci_dev *pdev = dev_priv->drm.pdev;
 	struct drm_i915_gem_object *obj;
@@ -597,17 +597,17 @@ static void guc_fw_fetch(struct drm_i915_private *dev_priv,
 	size_t size;
 	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));
+	DRM_DEBUG_DRIVER("before requesting firmware: uC fw fetch status %s\n",
+		intel_uc_fw_status_repr(uc_fw->fetch_status));
 
-	err = request_firmware(&fw, guc_fw->guc_fw_path, &pdev->dev);
+	err = request_firmware(&fw, uc_fw->uc_fw_path, &pdev->dev);
 	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);
+	DRM_DEBUG_DRIVER("fetch uC fw from %s succeeded, fw %p\n",
+		uc_fw->uc_fw_path, fw);
 
 	/* Check the size of the blob before examining buffer contents */
 	if (fw->size < sizeof(struct guc_css_header)) {
@@ -618,36 +618,36 @@ static void guc_fw_fetch(struct drm_i915_private *dev_priv,
 	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 -
+	uc_fw->header_offset = 0;
+	uc_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)) {
+	if (uc_fw->header_size != sizeof(struct guc_css_header)) {
 		DRM_NOTE("CSS header definition mismatch\n");
 		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);
+	uc_fw->ucode_offset = uc_fw->header_offset + uc_fw->header_size;
+	uc_fw->ucode_size = (css->size_dw - css->header_size_dw) * sizeof(u32);
 
 	/* now RSA */
 	if (css->key_size_dw != UOS_RSA_SCRATCH_MAX_COUNT) {
 		DRM_NOTE("RSA key size is bad\n");
 		goto fail;
 	}
-	guc_fw->rsa_offset = guc_fw->ucode_offset + guc_fw->ucode_size;
-	guc_fw->rsa_size = css->key_size_dw * sizeof(u32);
+	uc_fw->rsa_offset = uc_fw->ucode_offset + uc_fw->ucode_size;
+	uc_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;
+	size = uc_fw->header_size + uc_fw->ucode_size + uc_fw->rsa_size;
 	if (fw->size < size) {
 		DRM_NOTE("Missing firmware components\n");
 		goto fail;
 	}
 
 	/* Header and uCode will be loaded to WOPCM. Size of the two. */
-	size = guc_fw->header_size + guc_fw->ucode_size;
+	size = uc_fw->header_size + uc_fw->ucode_size;
 	if (size > guc_wopcm_size(dev_priv)) {
 		DRM_NOTE("Firmware is too large to fit in WOPCM\n");
 		goto fail;
@@ -659,21 +659,21 @@ static void guc_fw_fetch(struct drm_i915_private *dev_priv,
 	 * TWO bytes each (i.e. u16), although all pointers and offsets are defined
 	 * in terms of bytes (u8).
 	 */
-	guc_fw->guc_fw_major_found = css->guc_sw_version >> 16;
-	guc_fw->guc_fw_minor_found = css->guc_sw_version & 0xFFFF;
-
-	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) {
-		DRM_NOTE("GuC firmware version %d.%d, required %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);
+	uc_fw->major_ver_found = css->guc_sw_version >> 16;
+	uc_fw->minor_ver_found = css->guc_sw_version & 0xFFFF;
+
+	if (uc_fw->major_ver_found != uc_fw->major_ver_wanted ||
+	    uc_fw->minor_ver_found < uc_fw->minor_ver_wanted) {
+		DRM_NOTE("uC firmware version %d.%d, required %d.%d\n",
+			uc_fw->major_ver_found, uc_fw->minor_ver_found,
+			uc_fw->major_ver_wanted, uc_fw->minor_ver_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);
+			uc_fw->major_ver_found, uc_fw->minor_ver_found,
+			uc_fw->major_ver_wanted, uc_fw->minor_ver_wanted);
 
 	mutex_lock(&dev_priv->drm.struct_mutex);
 	obj = i915_gem_object_create_from_data(dev_priv, fw->data, fw->size);
@@ -683,31 +683,31 @@ static void guc_fw_fetch(struct drm_i915_private *dev_priv,
 		goto fail;
 	}
 
-	guc_fw->guc_fw_obj = obj;
-	guc_fw->guc_fw_size = fw->size;
+	uc_fw->uc_fw_obj = obj;
+	uc_fw->uc_fw_size = fw->size;
 
-	DRM_DEBUG_DRIVER("GuC fw fetch status SUCCESS, obj %p\n",
-			guc_fw->guc_fw_obj);
+	DRM_DEBUG_DRIVER("uC fw fetch status SUCCESS, obj %p\n",
+			uc_fw->uc_fw_obj);
 
 	release_firmware(fw);
-	guc_fw->guc_fw_fetch_status = GUC_FIRMWARE_SUCCESS;
+	uc_fw->fetch_status = UC_FIRMWARE_SUCCESS;
 	return;
 
 fail:
-	DRM_WARN("Failed to fetch valid GuC firmware from %s (error %d)\n",
-		 guc_fw->guc_fw_path, err);
-	DRM_DEBUG_DRIVER("GuC fw fetch status FAIL; err %d, fw %p, obj %p\n",
-		err, fw, guc_fw->guc_fw_obj);
+	DRM_WARN("Failed to fetch valid uC firmware from %s (error %d)\n",
+		 uc_fw->uc_fw_path, err);
+	DRM_DEBUG_DRIVER("uC fw fetch status FAIL; err %d, fw %p, obj %p\n",
+		err, fw, uc_fw->uc_fw_obj);
 
 	mutex_lock(&dev_priv->drm.struct_mutex);
-	obj = guc_fw->guc_fw_obj;
+	obj = uc_fw->uc_fw_obj;
 	if (obj)
 		i915_gem_object_put(obj);
-	guc_fw->guc_fw_obj = NULL;
+	uc_fw->uc_fw_obj = NULL;
 	mutex_unlock(&dev_priv->drm.struct_mutex);
 
 	release_firmware(fw);		/* OK even if fw is NULL */
-	guc_fw->guc_fw_fetch_status = GUC_FIRMWARE_FAIL;
+	uc_fw->fetch_status = UC_FIRMWARE_FAIL;
 }
 
 /**
@@ -721,7 +721,7 @@ static void guc_fw_fetch(struct drm_i915_private *dev_priv,
  */
 void intel_guc_init(struct drm_i915_private *dev_priv)
 {
-	struct intel_guc_fw *guc_fw = &dev_priv->guc.guc_fw;
+	struct intel_uc_fw *guc_fw = &dev_priv->guc.guc_fw;
 	const char *fw_path;
 
 	if (!HAS_GUC(dev_priv)) {
@@ -739,23 +739,24 @@ void intel_guc_init(struct drm_i915_private *dev_priv)
 		fw_path = NULL;
 	} else if (IS_SKYLAKE(dev_priv)) {
 		fw_path = I915_SKL_GUC_UCODE;
-		guc_fw->guc_fw_major_wanted = SKL_FW_MAJOR;
-		guc_fw->guc_fw_minor_wanted = SKL_FW_MINOR;
+		guc_fw->major_ver_wanted = SKL_FW_MAJOR;
+		guc_fw->minor_ver_wanted = SKL_FW_MINOR;
 	} else if (IS_BROXTON(dev_priv)) {
 		fw_path = I915_BXT_GUC_UCODE;
-		guc_fw->guc_fw_major_wanted = BXT_FW_MAJOR;
-		guc_fw->guc_fw_minor_wanted = BXT_FW_MINOR;
+		guc_fw->major_ver_wanted = BXT_FW_MAJOR;
+		guc_fw->minor_ver_wanted = BXT_FW_MINOR;
 	} else if (IS_KABYLAKE(dev_priv)) {
 		fw_path = I915_KBL_GUC_UCODE;
-		guc_fw->guc_fw_major_wanted = KBL_FW_MAJOR;
-		guc_fw->guc_fw_minor_wanted = KBL_FW_MINOR;
+		guc_fw->major_ver_wanted = KBL_FW_MAJOR;
+		guc_fw->minor_ver_wanted = KBL_FW_MINOR;
 	} else {
 		fw_path = "";	/* unknown device */
 	}
 
-	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;
+	guc_fw->uc_dev = &dev_priv->drm;
+	guc_fw->uc_fw_path = fw_path;
+	guc_fw->fetch_status = UC_FIRMWARE_NONE;
+	guc_fw->load_status = UC_FIRMWARE_NONE;
 
 	/* Early (and silent) return if GuC loading is disabled */
 	if (!i915.enable_guc_loading)
@@ -765,9 +766,9 @@ void intel_guc_init(struct drm_i915_private *dev_priv)
 	if (*fw_path == '\0')
 		return;
 
-	guc_fw->guc_fw_fetch_status = GUC_FIRMWARE_PENDING;
+	guc_fw->fetch_status = UC_FIRMWARE_PENDING;
 	DRM_DEBUG_DRIVER("GuC firmware pending, path %s\n", fw_path);
-	guc_fw_fetch(dev_priv, guc_fw);
+	intel_uc_fw_fetch(dev_priv, guc_fw);
 	/* status must now be FAIL or SUCCESS */
 }
 
@@ -777,17 +778,17 @@ void intel_guc_init(struct drm_i915_private *dev_priv)
  */
 void intel_guc_fini(struct drm_i915_private *dev_priv)
 {
-	struct intel_guc_fw *guc_fw = &dev_priv->guc.guc_fw;
+	struct intel_uc_fw *guc_fw = &dev_priv->guc.guc_fw;
 
 	mutex_lock(&dev_priv->drm.struct_mutex);
 	guc_interrupts_release(dev_priv);
 	i915_guc_submission_disable(dev_priv);
 	i915_guc_submission_fini(dev_priv);
 
-	if (guc_fw->guc_fw_obj)
-		i915_gem_object_put(guc_fw->guc_fw_obj);
-	guc_fw->guc_fw_obj = NULL;
+	if (guc_fw->uc_fw_obj)
+		i915_gem_object_put(guc_fw->uc_fw_obj);
+	guc_fw->uc_fw_obj = NULL;
 	mutex_unlock(&dev_priv->drm.struct_mutex);
 
-	guc_fw->guc_fw_fetch_status = GUC_FIRMWARE_NONE;
+	guc_fw->fetch_status = UC_FIRMWARE_NONE;
 }
drivers/gpu/drm/i915/intel_uc.h +0 -0
index 11f5608..f9f598d 100644
@@ -91,29 +91,31 @@ struct i915_guc_client {
 	uint64_t submissions[I915_NUM_ENGINES];
 };
 

drivers/gpu/drm/i915/intel_uc.h struct i915_guc_client { · line 98

Quoted 10 lines
drivers/gpu/drm/i915/intel_uc.h +5 -5
-enum intel_guc_fw_status {
-	GUC_FIRMWARE_FAIL = -1,
-	GUC_FIRMWARE_NONE = 0,
-	GUC_FIRMWARE_PENDING,
-	GUC_FIRMWARE_SUCCESS
+enum intel_uc_fw_status {
+	UC_FIRMWARE_FAIL = -1,
+	UC_FIRMWARE_NONE = 0,
+	UC_FIRMWARE_PENDING,
+	UC_FIRMWARE_SUCCESS

Shouldn't we use INTEL_UC_ prefix here (in similar way as in GUC_ACTION)?

Quoted 16 lines
drivers/gpu/drm/i915/intel_uc.h +0 -10
 };
 
 /*
  * This structure encapsulates all the data needed during the process
  * of fetching, caching, and loading the firmware image into the GuC.
  */
-struct intel_guc_fw {
-	const char *			guc_fw_path;
-	size_t				guc_fw_size;
-	struct drm_i915_gem_object *	guc_fw_obj;
-	enum intel_guc_fw_status	guc_fw_fetch_status;
-	enum intel_guc_fw_status	guc_fw_load_status;
-
-	uint16_t			guc_fw_major_wanted;
-	uint16_t			guc_fw_minor_wanted;
-	uint16_t			guc_fw_major_found;
Quoted 8 lines

drivers/gpu/drm/i915/intel_uc.h struct i915_guc_client { · line 106

Quoted 4 lines
drivers/gpu/drm/i915/intel_uc.h +2 -2
-	uint16_t			guc_fw_minor_found;
-
+struct intel_uc_fw {
+	struct drm_device *uc_dev;

Why do you need this back pointer to drm_dev ? More likely all you need is pointer to dev_priv. But for dev_priv you can always use guc_to_i915().

Quoted 6 lines

It seems like a left over from the old days. I have a series that refactors all things GuC that I'll push here soon, including the removal of this field.

Since this series advertises as a simple renaming I didn't force it here.

If you want to treat this as simple 'renaming' then 'adding' unused field is even more undesired and shall be avoided.

Quoted 20 lines
Quoted 19 lines

drivers/gpu/drm/i915/intel_uc.h struct i915_guc_client { · line 109

Quoted 4 lines
drivers/gpu/drm/i915/intel_uc.h +3 -0
+	const char *uc_fw_path;
+	size_t uc_fw_size;
+	struct drm_i915_gem_object *uc_fw_obj;

All these uc_fw_ prefixes seem to be redundant as this struct is just about uc_fw.

drivers/gpu/drm/i915/intel_uc.h struct i915_guc_client { · line 118

Quoted 9 lines
drivers/gpu/drm/i915/intel_uc.h +9 -0
+	enum intel_uc_fw_status fetch_status;
+	enum intel_uc_fw_status load_status;
+
+	uint16_t major_ver_wanted;
+	uint16_t minor_ver_wanted;
+	uint16_t major_ver_found;
+	uint16_t minor_ver_found;
+
+	uint32_t fw_type;

What is the purpose of this field ?

Quoted 4 lines
Quoted 2 lines

It is used in the patch 2, to set {major,minor}_ver_found depending on which firmware we pass to intel_uc_fw_fetch()

Then maybe is shall be moved to patch 2 ?

Sound reasonable as that's the first usage, although is also fits here along making te structs general. I have no strong opinion on that though.

Quoted 47 lines
Quoted 47 lines
Quoted 39 lines

drivers/gpu/drm/i915/intel_uc.h struct intel_guc_log { · line 144

Quoted 12 lines
drivers/gpu/drm/i915/intel_uc.h +1 -1
 	uint32_t header_size;
 	uint32_t header_offset;
 	uint32_t rsa_size;
@@ -139,7 +141,7 @@ struct intel_guc_log {
 };
 
 struct intel_guc {
-	struct intel_guc_fw guc_fw;
+	struct intel_uc_fw guc_fw;

Btw, guc_ prefix in field name here is also redundant.

Quoted 19 lines
drivers/gpu/drm/i915/intel_uc.h +1 -1
 	struct intel_guc_log log;
 
 	/* intel_guc_recv interrupt related state */
@@ -181,7 +183,7 @@ int intel_guc_log_control(struct intel_guc *guc, u32 control_val);
 extern void intel_guc_init(struct drm_i915_private *dev_priv);
 extern int intel_guc_setup(struct drm_i915_private *dev_priv);
 extern void intel_guc_fini(struct drm_i915_private *dev_priv);
-extern const char *intel_guc_fw_status_repr(enum intel_guc_fw_status status);
+extern const char *intel_uc_fw_status_repr(enum intel_uc_fw_status status);
 extern int intel_guc_suspend(struct drm_i915_private *dev_priv);
 extern int intel_guc_resume(struct drm_i915_private *dev_priv);
 
Signature · 6 lines
2.7.4

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Signature · 6 lines
Cheers,
Arek
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Signature · 6 lines
Cheers,
Arek
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
#7 lore raw

Re: [PATCH 1/8] drm/i915/guc: Make the GuC fw loading helper functions general

Quoted 488 lines

-----Original Message-----
From: Intel-gfx [mailto:[email protected]] On Behalf Of Arkadiusz Hiler
Sent: Friday, December 9, 2016 5:09 AM
To: Michal Wajdeczko <[email protected]>
Cc: [email protected]; Alex Dai <[email protected]>; Peter Antoine <[email protected]>
Subject: Re: [Intel-gfx] [PATCH 1/8] drm/i915/guc: Make the GuC fw loading helper functions general

Quoted 28 lines · Michal Wajdeczko
Quoted 27 lines · Arkadiusz Hiler
Quoted 26 lines · Michal Wajdeczko
Quoted 25 lines · anushasr

From: Peter Antoine <[email protected]>

Rename some of the GuC fw loading code to make them more general. We will utilise them for HuC loading as well.
s/intel_guc_fw/intel_uc_fw/g s/GUC_FIRMWARE/UC_FIRMWARE/g

Struct intel_guc_fw is renamed to intel_uc_fw. Prefix of tts members, such as 'guc' or 'guc_fw' either is renamed to 'uc' or removed for same purpose.

v2: rebased on top of nightly.
reapplied the search/replace as upstream code as changed.
v3: rebased again on drm-nightly.
v4: removed G from messages in shared fw fetch function.
v5: rebased.
v7: rebased.
v8: rebased.
v9: rebased.
v10: rebased.
v11: rebased.
v12: rebased on top of drm-tip
v13: rebased.Updated dev to dev_priv in intel_guc_setup(), guc_fw_getch() and intel_guc_init().

Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
---
 drivers/gpu/drm/i915/i915_debugfs.c        |  12 +--
 drivers/gpu/drm/i915/i915_guc_submission.c |   4 +-
 drivers/gpu/drm/i915/intel_guc_loader.c    | 157 +++++++++++++++-------

-------

Quoted 35 lines
Quoted 20 lines
Quoted 20 lines
Quoted 20 lines
 drivers/gpu/drm/i915/intel_uc.h            |  40 ++++----
 4 files changed, 108 insertions(+), 105 deletions(-)
drivers/gpu/drm/i915/i915_debugfs.c +0 -0
index a746130..0e5ef62 100644
@@ -2338,7 +2338,7 @@ static int i915_llc(struct seq_file *m,

void *data) static int i915_guc_load_status_info(struct seq_file *m, void *data) {

drivers/gpu/drm/i915/i915_debugfs.c +1 -1
 	struct drm_i915_private *dev_priv = node_to_i915(m->private);
-	struct intel_guc_fw *guc_fw = &dev_priv->guc.guc_fw;
+	struct intel_uc_fw *guc_fw = &dev_priv->guc.guc_fw;
 	u32 tmp, i;
 
 	if (!HAS_GUC_UCODE(dev_priv))
@@ -2346,15 +2346,15 @@ static int
 i915_guc_load_status_info(struct seq_file *m, void *data)
 
Quoted 12 lines
Quoted 4 lines
Quoted 4 lines
drivers/gpu/drm/i915/i915_debugfs.c +1 -1
 	seq_printf(m, "GuC firmware status:\n");
 	seq_printf(m, "\tpath: %s\n",
-		guc_fw->guc_fw_path);
+		guc_fw->uc_fw_path);
Quoted 8 lines

drivers/gpu/drm/i915/i915_debugfs.c static int i915_guc_load_status_info(struct seq_file *m, void *data)

Quoted 8 lines
drivers/gpu/drm/i915/i915_debugfs.c +2 -3
 	seq_printf(m, "\tfetch: %s\n",
-		intel_guc_fw_status_repr(guc_fw->guc_fw_fetch_status));
+		intel_uc_fw_status_repr(guc_fw->fetch_status));
 	seq_printf(m, "\tload: %s\n",
-		intel_guc_fw_status_repr(guc_fw->guc_fw_load_status));
+		intel_uc_fw_status_repr(guc_fw->load_status));
 	seq_printf(m, "\tversion wanted: %d.%d\n",
-		guc_fw->guc_fw_major_wanted, guc_fw-

guc_fw_minor_wanted);

Quoted 1 line
Quoted 1 line

drivers/gpu/drm/i915/i915_debugfs.c static int · line 2356

Quoted 1 line
drivers/gpu/drm/i915/i915_debugfs.c +1 -0
+		guc_fw->major_ver_wanted, guc_fw-

minor_ver_wanted);

Quoted 12 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
drivers/gpu/drm/i915/i915_debugfs.c +1 -1
 	seq_printf(m, "\tversion found: %d.%d\n",
-		guc_fw->guc_fw_major_found, guc_fw->guc_fw_minor_found);
+		guc_fw->major_ver_found, guc_fw->minor_ver_found);
 	seq_printf(m, "\theader: offset is %d; size = %d\n",
Quoted 8 lines
Quoted 8 lines

drivers/gpu/drm/i915/i915_guc_submission.c int intel_guc_suspend(struct drm_i915_private

Quoted 8 lines
drivers/gpu/drm/i915/i915_debugfs.c +0 -0
 		guc_fw->header_offset, guc_fw->header_size);

seq_printf(m, "\tuCode: offset is %d; size = %d\n", diff --git
a/drivers/gpu/drm/i915/i915_guc_submission.c b/drivers/gpu/drm/i915/i915_guc_submission.c
index 7fa4e74..e156a4b 100644

drivers/gpu/drm/i915/i915_guc_submission.c +0 -0
@@ -1493,7 +1493,7 @@ int intel_guc_suspend(struct drm_i915_private

*dev_priv)

Quoted 9 lines
Quoted 9 lines
Quoted 9 lines

drivers/gpu/drm/i915/i915_guc_submission.c int intel_guc_resume(struct drm_i915_private

Quoted 9 lines
drivers/gpu/drm/i915/i915_guc_submission.c +1 -1
 	struct i915_gem_context *ctx;
 	u32 data[3];
 
-	if (guc->guc_fw.guc_fw_load_status != GUC_FIRMWARE_SUCCESS)
+	if (guc->guc_fw.load_status != UC_FIRMWARE_SUCCESS)
 		return 0;
 
 	gen9_disable_guc_interrupts(dev_priv);
@@ -1520,7 +1520,7 @@ int intel_guc_resume(struct drm_i915_private

*dev_priv)

Quoted 30 lines
Quoted 30 lines
Quoted 30 lines
Quoted 30 lines
drivers/gpu/drm/i915/i915_guc_submission.c +1 -1
 	struct i915_gem_context *ctx;
 	u32 data[3];
 
-	if (guc->guc_fw.guc_fw_load_status != GUC_FIRMWARE_SUCCESS)
+	if (guc->guc_fw.load_status != UC_FIRMWARE_SUCCESS)
 		return 0;
 
 	if (i915.guc_log_level >= 0)
drivers/gpu/drm/i915/intel_guc_loader.c +5 -3
index 21db697..8f04f6e 100644
@@ -81,16 +81,16 @@ MODULE_FIRMWARE(I915_BXT_GUC_UCODE);
 MODULE_FIRMWARE(I915_KBL_GUC_UCODE);
 
 /* User-friendly representation of an enum */ -const char
 *intel_guc_fw_status_repr(enum intel_guc_fw_status status)
+const char *intel_uc_fw_status_repr(enum intel_uc_fw_status
+status)
 {
 	switch (status) {
-	case GUC_FIRMWARE_FAIL:
+	case UC_FIRMWARE_FAIL:
 		return "FAIL";
-	case GUC_FIRMWARE_NONE:
+	case UC_FIRMWARE_NONE:
 		return "NONE";
-	case GUC_FIRMWARE_PENDING:
+	case UC_FIRMWARE_PENDING:
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
drivers/gpu/drm/i915/intel_guc_loader.c +1 -1
 		return "PENDING";
-	case GUC_FIRMWARE_SUCCESS:
+	case UC_FIRMWARE_SUCCESS:
 		return "SUCCESS";
 	default:
 		return "UNKNOWN!";
@@ -278,7 +278,7 @@ static inline bool guc_ucode_response(struct

drm_i915_private *dev_priv, static int guc_ucode_xfer_dma(struct

drm_i915_private *dev_priv,

Quoted 23 lines
Quoted 23 lines
Quoted 23 lines
Quoted 23 lines
drivers/gpu/drm/i915/intel_guc_loader.c +1 -1
 			      struct i915_vma *vma)
 {
-	struct intel_guc_fw *guc_fw = &dev_priv->guc.guc_fw;
+	struct intel_uc_fw *guc_fw = &dev_priv->guc.guc_fw;
 	unsigned long offset;
 	struct sg_table *sg = vma->pages;
 	u32 status, rsa[UOS_RSA_SCRATCH_MAX_COUNT]; @@ -350,17 +350,17

@@ static u32 guc_wopcm_size(struct drm_i915_private *dev_priv)
*/
static int guc_ucode_xfer(struct drm_i915_private *dev_priv) {

drivers/gpu/drm/i915/intel_guc_loader.c +3 -2
-	struct intel_guc_fw *guc_fw = &dev_priv->guc.guc_fw;
+	struct intel_uc_fw *guc_fw = &dev_priv->guc.guc_fw;
 	struct i915_vma *vma;
 	int ret;
 
-	ret = i915_gem_object_set_to_gtt_domain(guc_fw->guc_fw_obj, false);
+	ret = i915_gem_object_set_to_gtt_domain(guc_fw->uc_fw_obj,
+false);
 	if (ret) {
 		DRM_DEBUG_DRIVER("set-domain failed %d\n", ret);
 		return ret;
 	}
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines

drivers/gpu/drm/i915/intel_guc_loader.c

Quoted 2 lines
drivers/gpu/drm/i915/intel_guc_loader.c +1 -1
-	vma = i915_gem_object_ggtt_pin(guc_fw->guc_fw_obj, NULL, 0, 0, 0);
+	vma = i915_gem_object_ggtt_pin(guc_fw->uc_fw_obj, NULL, 0,

0,

Quoted 5 lines
Quoted 5 lines
Quoted 5 lines

drivers/gpu/drm/i915/intel_guc_loader.c static int guc_hw_reset(struct

Quoted 5 lines

+0);

drivers/gpu/drm/i915/intel_guc_loader.c +0 -0
 	if (IS_ERR(vma)) {
 		DRM_DEBUG_DRIVER("pin failed %d\n", (int)PTR_ERR(vma));
 		return PTR_ERR(vma);
@@ -450,14 +450,14 @@ static int guc_hw_reset(struct

drm_i915_private *dev_priv)

Quoted 20 lines
Quoted 20 lines
Quoted 20 lines
Quoted 20 lines
drivers/gpu/drm/i915/intel_guc_loader.c +4 -4
  */
 int intel_guc_setup(struct drm_i915_private *dev_priv)  {
-	struct intel_guc_fw *guc_fw = &dev_priv->guc.guc_fw;
-	const char *fw_path = guc_fw->guc_fw_path;
+	struct intel_uc_fw *guc_fw = &dev_priv->guc.guc_fw;
+	const char *fw_path = guc_fw->uc_fw_path;
 	int retries, ret, err;
 
 	DRM_DEBUG_DRIVER("GuC fw status: path %s, fetch %s, load %s\n",
 		fw_path,
-		intel_guc_fw_status_repr(guc_fw->guc_fw_fetch_status),
-		intel_guc_fw_status_repr(guc_fw->guc_fw_load_status));
+		intel_uc_fw_status_repr(guc_fw->fetch_status),
+		intel_uc_fw_status_repr(guc_fw->load_status));
 
 	/* Loading forbidden, or no firmware to load? */
 	if (!i915.enable_guc_loading) { @@ -475,10 +475,10 @@ int
 intel_guc_setup(struct drm_i915_private *dev_priv)

}

Quoted 11 lines
Quoted 11 lines
Quoted 11 lines

drivers/gpu/drm/i915/intel_guc_loader.c int intel_guc_setup(struct drm_i915_private

Quoted 11 lines
drivers/gpu/drm/i915/intel_guc_loader.c +2 -2
 	/* Fetch failed, or already fetched but failed to load? */
-	if (guc_fw->guc_fw_fetch_status != GUC_FIRMWARE_SUCCESS) {
+	if (guc_fw->fetch_status != UC_FIRMWARE_SUCCESS) {
 		err = -EIO;
 		goto fail;
-	} else if (guc_fw->guc_fw_load_status == GUC_FIRMWARE_FAIL) {
+	} else if (guc_fw->load_status == UC_FIRMWARE_FAIL) {
 		err = -ENOEXEC;
 		goto fail;
 	}
@@ -486,11 +486,11 @@ int intel_guc_setup(struct drm_i915_private

*dev_priv)

Quoted 12 lines
Quoted 12 lines
Quoted 12 lines
Quoted 12 lines
drivers/gpu/drm/i915/intel_guc_loader.c +3 -3
 	guc_interrupts_release(dev_priv);
 	gen9_reset_guc_interrupts(dev_priv);
 
-	guc_fw->guc_fw_load_status = GUC_FIRMWARE_PENDING;
+	guc_fw->load_status = UC_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));
+		intel_uc_fw_status_repr(guc_fw->fetch_status),
+		intel_uc_fw_status_repr(guc_fw->load_status));
 
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines

drivers/gpu/drm/i915/intel_guc_loader.c int intel_guc_setup(struct drm_i915_private

Quoted 3 lines
drivers/gpu/drm/i915/intel_guc_loader.c +0 -0
 	err = i915_guc_submission_init(dev_priv);
 	if (err)
@@ -522,11 +522,11 @@ int intel_guc_setup(struct drm_i915_private

*dev_priv)

Quoted 12 lines
Quoted 12 lines
Quoted 12 lines
Quoted 12 lines
drivers/gpu/drm/i915/intel_guc_loader.c +3 -3
 			 "retry %d more time(s)\n", err, retries);
 	}
 
-	guc_fw->guc_fw_load_status = GUC_FIRMWARE_SUCCESS;
+	guc_fw->load_status = UC_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));
+		intel_uc_fw_status_repr(guc_fw->fetch_status),
+		intel_uc_fw_status_repr(guc_fw->load_status));
 
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines

drivers/gpu/drm/i915/intel_guc_loader.c int intel_guc_setup(struct drm_i915_private

Quoted 3 lines
drivers/gpu/drm/i915/intel_guc_loader.c +0 -0
 	if (i915.enable_guc_submission) {
 		if (i915.guc_log_level >= 0)
@@ -541,8 +541,8 @@ int intel_guc_setup(struct drm_i915_private

*dev_priv)

Quoted 11 lines
Quoted 11 lines
Quoted 11 lines

drivers/gpu/drm/i915/intel_guc_loader.c int intel_guc_setup(struct drm_i915_private

Quoted 11 lines
drivers/gpu/drm/i915/intel_guc_loader.c +2 -2
 	return 0;
 
 fail:
-	if (guc_fw->guc_fw_load_status == GUC_FIRMWARE_PENDING)
-		guc_fw->guc_fw_load_status = GUC_FIRMWARE_FAIL;
+	if (guc_fw->load_status == UC_FIRMWARE_PENDING)
+		guc_fw->load_status = UC_FIRMWARE_FAIL;
 
 	guc_interrupts_release(dev_priv);
 	i915_guc_submission_disable(dev_priv);
@@ -587,8 +587,8 @@ int intel_guc_setup(struct drm_i915_private

*dev_priv)

Quoted 14 lines
Quoted 14 lines
Quoted 14 lines
Quoted 14 lines
drivers/gpu/drm/i915/intel_guc_loader.c +2 -2
 	return ret;
 }
 
-static void guc_fw_fetch(struct drm_i915_private *dev_priv,
-			 struct intel_guc_fw *guc_fw)
+void intel_uc_fw_fetch(struct drm_i915_private *dev_priv,
+			 struct intel_uc_fw *uc_fw)
 {
 	struct pci_dev *pdev = dev_priv->drm.pdev;
 	struct drm_i915_gem_object *obj; @@ -597,17 +597,17 @@ static

void guc_fw_fetch(struct drm_i915_private *dev_priv,

drivers/gpu/drm/i915/intel_guc_loader.c +0 -0
 	size_t size;
 	int err;
Quoted 1 line
Quoted 1 line
Quoted 1 line

drivers/gpu/drm/i915/intel_guc_loader.c

Quoted 1 line
drivers/gpu/drm/i915/intel_guc_loader.c +0 -1
-	DRM_DEBUG_DRIVER("before requesting firmware: GuC fw fetch status

%s\n",

Quoted 2 lines
Quoted 2 lines
Quoted 2 lines

drivers/gpu/drm/i915/intel_guc_loader.c

Quoted 2 lines
drivers/gpu/drm/i915/intel_guc_loader.c +1 -1
-		intel_guc_fw_status_repr(guc_fw->guc_fw_fetch_status));
+	DRM_DEBUG_DRIVER("before requesting firmware: uC fw fetch

status %s\n",

Quoted 9 lines
Quoted 9 lines
Quoted 9 lines
Quoted 9 lines
drivers/gpu/drm/i915/intel_guc_loader.c +2 -1
+		intel_uc_fw_status_repr(uc_fw->fetch_status));
 
-	err = request_firmware(&fw, guc_fw->guc_fw_path, &pdev->dev);
+	err = request_firmware(&fw, uc_fw->uc_fw_path, &pdev->dev);
 	if (err)
 		goto fail;
 	if (!fw)
 		goto fail;
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines

drivers/gpu/drm/i915/intel_guc_loader.c static void guc_fw_fetch(struct drm_i915_private *dev_priv,

Quoted 3 lines
drivers/gpu/drm/i915/intel_guc_loader.c +1 -2
-	DRM_DEBUG_DRIVER("fetch GuC fw from %s succeeded, fw %p\n",
-		guc_fw->guc_fw_path, fw);
+	DRM_DEBUG_DRIVER("fetch uC fw from %s succeeded, fw

%p\n",

Quoted 21 lines
Quoted 21 lines
Quoted 21 lines
Quoted 21 lines
drivers/gpu/drm/i915/intel_guc_loader.c +6 -3
+		uc_fw->uc_fw_path, fw);
 
 	/* Check the size of the blob before examining buffer contents */
 	if (fw->size < sizeof(struct guc_css_header)) { @@ -618,36
+618,36 @@ static void guc_fw_fetch(struct drm_i915_private *dev_priv,
 	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 -
+	uc_fw->header_offset = 0;
+	uc_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)) {
+	if (uc_fw->header_size != sizeof(struct guc_css_header)) {
 		DRM_NOTE("CSS header definition mismatch\n");
 		goto fail;
 	}
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines

drivers/gpu/drm/i915/intel_guc_loader.c static void guc_fw_fetch(struct drm_i915_private *dev_priv,

Quoted 3 lines
drivers/gpu/drm/i915/intel_guc_loader.c +0 -2
 	/* 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);

Quoted 24 lines
Quoted 1 line
Quoted 1 line

drivers/gpu/drm/i915/intel_guc_loader.c

Quoted 1 line
drivers/gpu/drm/i915/intel_guc_loader.c +1 -0
+	uc_fw->ucode_offset = uc_fw->header_offset + uc_fw-

header_size;

Quoted 22 lines
Quoted 22 lines
Quoted 22 lines
drivers/gpu/drm/i915/intel_guc_loader.c +6 -3
+	uc_fw->ucode_size = (css->size_dw - css->header_size_dw) *
+sizeof(u32);
 
 	/* now RSA */
 	if (css->key_size_dw != UOS_RSA_SCRATCH_MAX_COUNT) {
 		DRM_NOTE("RSA key size is bad\n");
 		goto fail;
 	}
-	guc_fw->rsa_offset = guc_fw->ucode_offset + guc_fw->ucode_size;
-	guc_fw->rsa_size = css->key_size_dw * sizeof(u32);
+	uc_fw->rsa_offset = uc_fw->ucode_offset + uc_fw->ucode_size;
+	uc_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;
+	size = uc_fw->header_size + uc_fw->ucode_size +
+uc_fw->rsa_size;
 	if (fw->size < size) {
 		DRM_NOTE("Missing firmware components\n");
 		goto fail;
 	}
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines

drivers/gpu/drm/i915/intel_guc_loader.c static void guc_fw_fetch(struct

Quoted 7 lines
drivers/gpu/drm/i915/intel_guc_loader.c +1 -1
 	/* Header and uCode will be loaded to WOPCM. Size of the two. */
-	size = guc_fw->header_size + guc_fw->ucode_size;
+	size = uc_fw->header_size + uc_fw->ucode_size;
 	if (size > guc_wopcm_size(dev_priv)) {
 		DRM_NOTE("Firmware is too large to fit in WOPCM\n");
 		goto fail;
@@ -659,21 +659,21 @@ static void guc_fw_fetch(struct

drm_i915_private *dev_priv,

Quoted 11 lines
Quoted 10 lines
Quoted 10 lines

drivers/gpu/drm/i915/intel_guc_loader.c static void guc_fw_fetch(struct drm_i915_private *dev_priv,

Quoted 10 lines
drivers/gpu/drm/i915/intel_guc_loader.c +0 -7
 	 * TWO bytes each (i.e. u16), although all pointers and offsets are defined
 	 * in terms of bytes (u8).
 	 */
-	guc_fw->guc_fw_major_found = css->guc_sw_version >> 16;
-	guc_fw->guc_fw_minor_found = css->guc_sw_version & 0xFFFF;
-
-	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) {
-		DRM_NOTE("GuC firmware version %d.%d, required %d.%d\n",
-			guc_fw->guc_fw_major_found, guc_fw-

guc_fw_minor_found,

Quoted 8 lines
Quoted 1 line
Quoted 1 line

drivers/gpu/drm/i915/intel_guc_loader.c static void guc_fw_fetch(struct · line 669

Quoted 1 line
drivers/gpu/drm/i915/intel_guc_loader.c +0 -1
-			guc_fw->guc_fw_major_wanted, guc_fw-

guc_fw_minor_wanted);

Quoted 6 lines
Quoted 6 lines

drivers/gpu/drm/i915/intel_guc_loader.c static void guc_fw_fetch(struct drm_i915_private *dev_priv,

Quoted 6 lines
drivers/gpu/drm/i915/intel_guc_loader.c +6 -0
+	uc_fw->major_ver_found = css->guc_sw_version >> 16;
+	uc_fw->minor_ver_found = css->guc_sw_version & 0xFFFF;
+
+	if (uc_fw->major_ver_found != uc_fw->major_ver_wanted ||
+	    uc_fw->minor_ver_found < uc_fw->minor_ver_wanted) {
+		DRM_NOTE("uC firmware version %d.%d, required

%d.%d\n",

Quoted 9 lines
Quoted 1 line
Quoted 1 line

drivers/gpu/drm/i915/intel_guc_loader.c static void guc_fw_fetch(struct · line 668

Quoted 1 line
drivers/gpu/drm/i915/intel_guc_loader.c +1 -0
+			uc_fw->major_ver_found, uc_fw-

minor_ver_found,

Quoted 1 line
Quoted 1 line

drivers/gpu/drm/i915/intel_guc_loader.c static void guc_fw_fetch(struct · line 669

Quoted 1 line
drivers/gpu/drm/i915/intel_guc_loader.c +1 -0
+			uc_fw->major_ver_wanted, uc_fw-

minor_ver_wanted);

Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
drivers/gpu/drm/i915/intel_guc_loader.c +0 -0
 		err = -ENOEXEC;
 		goto fail;
 	}
 
 	DRM_DEBUG_DRIVER("firmware version %d.%d OK (minimum

%d.%d)\n",

Quoted 12 lines
Quoted 1 line
Quoted 1 line

drivers/gpu/drm/i915/intel_guc_loader.c static void guc_fw_fetch(struct · line 675

Quoted 1 line
drivers/gpu/drm/i915/intel_guc_loader.c +0 -1
-			guc_fw->guc_fw_major_found, guc_fw-

guc_fw_minor_found,

Quoted 1 line
Quoted 1 line

drivers/gpu/drm/i915/intel_guc_loader.c static void guc_fw_fetch(struct · line 676

Quoted 1 line
drivers/gpu/drm/i915/intel_guc_loader.c +0 -1
-			guc_fw->guc_fw_major_wanted, guc_fw-

guc_fw_minor_wanted);

Quoted 1 line
Quoted 1 line

drivers/gpu/drm/i915/intel_guc_loader.c static void guc_fw_fetch(struct · line 675

Quoted 1 line
drivers/gpu/drm/i915/intel_guc_loader.c +1 -0
+			uc_fw->major_ver_found, uc_fw-

minor_ver_found,

Quoted 1 line
Quoted 1 line

drivers/gpu/drm/i915/intel_guc_loader.c static void guc_fw_fetch(struct · line 676

Quoted 1 line
drivers/gpu/drm/i915/intel_guc_loader.c +1 -0
+			uc_fw->major_ver_wanted, uc_fw-

minor_ver_wanted);

Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
drivers/gpu/drm/i915/intel_guc_loader.c +0 -0
 	mutex_lock(&dev_priv->drm.struct_mutex);
 	obj = i915_gem_object_create_from_data(dev_priv, fw->data,

fw->size); @@ -683,31 +683,31 @@ static void guc_fw_fetch(struct

drm_i915_private *dev_priv,

Quoted 18 lines
Quoted 18 lines
Quoted 18 lines
Quoted 18 lines
drivers/gpu/drm/i915/intel_guc_loader.c +5 -5
 		goto fail;
 	}
 
-	guc_fw->guc_fw_obj = obj;
-	guc_fw->guc_fw_size = fw->size;
+	uc_fw->uc_fw_obj = obj;
+	uc_fw->uc_fw_size = fw->size;
 
-	DRM_DEBUG_DRIVER("GuC fw fetch status SUCCESS, obj %p\n",
-			guc_fw->guc_fw_obj);
+	DRM_DEBUG_DRIVER("uC fw fetch status SUCCESS, obj %p\n",
+			uc_fw->uc_fw_obj);
 
 	release_firmware(fw);
-	guc_fw->guc_fw_fetch_status = GUC_FIRMWARE_SUCCESS;
+	uc_fw->fetch_status = UC_FIRMWARE_SUCCESS;
 	return;
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines

drivers/gpu/drm/i915/intel_guc_loader.c static void guc_fw_fetch(struct drm_i915_private *dev_priv,

Quoted 4 lines

fail:

drivers/gpu/drm/i915/intel_guc_loader.c +0 -3
-	DRM_WARN("Failed to fetch valid GuC firmware from %s (error %d)\n",
-		 guc_fw->guc_fw_path, err);
-	DRM_DEBUG_DRIVER("GuC fw fetch status FAIL; err %d, fw %p, obj

%p\n",

Quoted 2 lines
Quoted 2 lines
Quoted 2 lines

drivers/gpu/drm/i915/intel_guc_loader.c

Quoted 2 lines
drivers/gpu/drm/i915/intel_guc_loader.c +1 -1
-		err, fw, guc_fw->guc_fw_obj);
+	DRM_WARN("Failed to fetch valid uC firmware from %s (error

%d)\n",

Quoted 2 lines
Quoted 2 lines
Quoted 2 lines

drivers/gpu/drm/i915/intel_guc_loader.c

Quoted 2 lines
drivers/gpu/drm/i915/intel_guc_loader.c +2 -0
+		 uc_fw->uc_fw_path, err);
+	DRM_DEBUG_DRIVER("uC fw fetch status FAIL; err %d, fw %p,

obj %p\n",

Quoted 11 lines
Quoted 11 lines
Quoted 11 lines
Quoted 11 lines
drivers/gpu/drm/i915/intel_guc_loader.c +3 -2
+		err, fw, uc_fw->uc_fw_obj);
 
 	mutex_lock(&dev_priv->drm.struct_mutex);
-	obj = guc_fw->guc_fw_obj;
+	obj = uc_fw->uc_fw_obj;
 	if (obj)
 		i915_gem_object_put(obj);
-	guc_fw->guc_fw_obj = NULL;
+	uc_fw->uc_fw_obj = NULL;
 	mutex_unlock(&dev_priv->drm.struct_mutex);
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines

drivers/gpu/drm/i915/intel_guc_loader.c static void guc_fw_fetch(struct drm_i915_private

Quoted 7 lines
drivers/gpu/drm/i915/intel_guc_loader.c +1 -1
 	release_firmware(fw);		/* OK even if fw is NULL */
-	guc_fw->guc_fw_fetch_status = GUC_FIRMWARE_FAIL;
+	uc_fw->fetch_status = UC_FIRMWARE_FAIL;
 }
 
 /**
@@ -721,7 +721,7 @@ static void guc_fw_fetch(struct drm_i915_private

*dev_priv,

Quoted 8 lines
Quoted 8 lines
Quoted 8 lines

drivers/gpu/drm/i915/intel_guc_loader.c void intel_guc_init(struct drm_i915_private

Quoted 8 lines
drivers/gpu/drm/i915/intel_guc_loader.c +1 -1
  */
 void intel_guc_init(struct drm_i915_private *dev_priv)  {
-	struct intel_guc_fw *guc_fw = &dev_priv->guc.guc_fw;
+	struct intel_uc_fw *guc_fw = &dev_priv->guc.guc_fw;
 	const char *fw_path;
 
 	if (!HAS_GUC(dev_priv)) {
@@ -739,23 +739,24 @@ void intel_guc_init(struct drm_i915_private

*dev_priv)

Quoted 31 lines
Quoted 31 lines
Quoted 31 lines
Quoted 31 lines
drivers/gpu/drm/i915/intel_guc_loader.c +10 -9
 		fw_path = NULL;
 	} else if (IS_SKYLAKE(dev_priv)) {
 		fw_path = I915_SKL_GUC_UCODE;
-		guc_fw->guc_fw_major_wanted = SKL_FW_MAJOR;
-		guc_fw->guc_fw_minor_wanted = SKL_FW_MINOR;
+		guc_fw->major_ver_wanted = SKL_FW_MAJOR;
+		guc_fw->minor_ver_wanted = SKL_FW_MINOR;
 	} else if (IS_BROXTON(dev_priv)) {
 		fw_path = I915_BXT_GUC_UCODE;
-		guc_fw->guc_fw_major_wanted = BXT_FW_MAJOR;
-		guc_fw->guc_fw_minor_wanted = BXT_FW_MINOR;
+		guc_fw->major_ver_wanted = BXT_FW_MAJOR;
+		guc_fw->minor_ver_wanted = BXT_FW_MINOR;
 	} else if (IS_KABYLAKE(dev_priv)) {
 		fw_path = I915_KBL_GUC_UCODE;
-		guc_fw->guc_fw_major_wanted = KBL_FW_MAJOR;
-		guc_fw->guc_fw_minor_wanted = KBL_FW_MINOR;
+		guc_fw->major_ver_wanted = KBL_FW_MAJOR;
+		guc_fw->minor_ver_wanted = KBL_FW_MINOR;
 	} else {
 		fw_path = "";	/* unknown device */
 	}
 
-	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;
+	guc_fw->uc_dev = &dev_priv->drm;
+	guc_fw->uc_fw_path = fw_path;
+	guc_fw->fetch_status = UC_FIRMWARE_NONE;
+	guc_fw->load_status = UC_FIRMWARE_NONE;
 
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines

drivers/gpu/drm/i915/intel_guc_loader.c void intel_guc_init(struct drm_i915_private

Quoted 3 lines
drivers/gpu/drm/i915/intel_guc_loader.c +0 -0
 	/* Early (and silent) return if GuC loading is disabled */
 	if (!i915.enable_guc_loading)
@@ -765,9 +766,9 @@ void intel_guc_init(struct drm_i915_private

*dev_priv)

Quoted 10 lines
Quoted 10 lines
Quoted 10 lines
Quoted 10 lines
drivers/gpu/drm/i915/intel_guc_loader.c +2 -2
 	if (*fw_path == '\0')
 		return;
 
-	guc_fw->guc_fw_fetch_status = GUC_FIRMWARE_PENDING;
+	guc_fw->fetch_status = UC_FIRMWARE_PENDING;
 	DRM_DEBUG_DRIVER("GuC firmware pending, path %s\n", fw_path);
-	guc_fw_fetch(dev_priv, guc_fw);
+	intel_uc_fw_fetch(dev_priv, guc_fw);
 	/* status must now be FAIL or SUCCESS */  }
 
Quoted 42 lines
Quoted 1 line
Quoted 1 line
Quoted 1 line

drivers/gpu/drm/i915/intel_guc_loader.c void intel_guc_init(struct drm_i915_private

Quoted 1 line
drivers/gpu/drm/i915/intel_guc_loader.c +0 -0
@@ -777,17 +778,17 @@ void intel_guc_init(struct drm_i915_private

*dev_priv)

Quoted 40 lines
Quoted 40 lines
Quoted 40 lines
Quoted 28 lines
drivers/gpu/drm/i915/intel_guc_loader.c +5 -5
  */
 void intel_guc_fini(struct drm_i915_private *dev_priv)  {
-	struct intel_guc_fw *guc_fw = &dev_priv->guc.guc_fw;
+	struct intel_uc_fw *guc_fw = &dev_priv->guc.guc_fw;
 
 	mutex_lock(&dev_priv->drm.struct_mutex);
 	guc_interrupts_release(dev_priv);
 	i915_guc_submission_disable(dev_priv);
 	i915_guc_submission_fini(dev_priv);
 
-	if (guc_fw->guc_fw_obj)
-		i915_gem_object_put(guc_fw->guc_fw_obj);
-	guc_fw->guc_fw_obj = NULL;
+	if (guc_fw->uc_fw_obj)
+		i915_gem_object_put(guc_fw->uc_fw_obj);
+	guc_fw->uc_fw_obj = NULL;
 	mutex_unlock(&dev_priv->drm.struct_mutex);
 
-	guc_fw->guc_fw_fetch_status = GUC_FIRMWARE_NONE;
+	guc_fw->fetch_status = UC_FIRMWARE_NONE;
 }
(quoted diff) +0 -0

b/drivers/gpu/drm/i915/intel_uc.h index 11f5608..f9f598d 100644

drivers/gpu/drm/i915/intel_uc.h +0 -0
@@ -91,29 +91,31 @@ struct i915_guc_client {
 	uint64_t submissions[I915_NUM_ENGINES];  };
 

drivers/gpu/drm/i915/intel_uc.h struct i915_guc_client { · line 97

Quoted 10 lines
drivers/gpu/drm/i915/intel_uc.h +5 -5
-enum intel_guc_fw_status {
-	GUC_FIRMWARE_FAIL = -1,
-	GUC_FIRMWARE_NONE = 0,
-	GUC_FIRMWARE_PENDING,
-	GUC_FIRMWARE_SUCCESS
+enum intel_uc_fw_status {
+	UC_FIRMWARE_FAIL = -1,
+	UC_FIRMWARE_NONE = 0,
+	UC_FIRMWARE_PENDING,
+	UC_FIRMWARE_SUCCESS

Shouldn't we use INTEL_UC_ prefix here (in similar way as in GUC_ACTION)?

Will do. Thanks for your review.
-Anusha

Quoted 30 lines
Quoted 30 lines
Quoted 30 lines
Quoted 17 lines
Quoted 17 lines
drivers/gpu/drm/i915/intel_uc.h +0 -10
 };
 
 /*
  * This structure encapsulates all the data needed during the process
  * of fetching, caching, and loading the firmware image into the GuC.
  */
-struct intel_guc_fw {
-	const char *			guc_fw_path;
-	size_t				guc_fw_size;
-	struct drm_i915_gem_object *	guc_fw_obj;
-	enum intel_guc_fw_status	guc_fw_fetch_status;
-	enum intel_guc_fw_status	guc_fw_load_status;
-
-	uint16_t			guc_fw_major_wanted;
-	uint16_t			guc_fw_minor_wanted;
-	uint16_t			guc_fw_major_found;
Quoted 8 lines

drivers/gpu/drm/i915/intel_uc.h struct i915_guc_client { · line 106

Quoted 4 lines
drivers/gpu/drm/i915/intel_uc.h +2 -2
-	uint16_t			guc_fw_minor_found;
-
+struct intel_uc_fw {
+	struct drm_device *uc_dev;

Why do you need this back pointer to drm_dev ? More likely all you need is pointer to dev_priv. But for dev_priv you can always use guc_to_i915().

It seems like a left over from the old days. I have a series that refactors all things GuC that I'll push here soon, including the removal of this field.

Quoted 5 lines
Quoted 5 lines
Quoted 2 lines

Since this series advertises as a simple renaming I didn't force it here.

If you want to treat this as simple 'renaming' then 'adding' unused field is even more undesired and shall be avoided.

Michal, I don’t think I follow you completely here. Do you want me to rename it or wait for Arek's series?

Quoted 21 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines

drivers/gpu/drm/i915/intel_uc.h struct i915_guc_client { · line 109

Quoted 4 lines
drivers/gpu/drm/i915/intel_uc.h +3 -0
+	const char *uc_fw_path;
+	size_t uc_fw_size;
+	struct drm_i915_gem_object *uc_fw_obj;

All these uc_fw_ prefixes seem to be redundant as this struct is just about

uc_fw.

Quoted 14 lines
Quoted 14 lines
Quoted 13 lines

drivers/gpu/drm/i915/intel_uc.h struct i915_guc_client { · line 118

Quoted 11 lines
drivers/gpu/drm/i915/intel_uc.h +9 -0
+	enum intel_uc_fw_status fetch_status;
+	enum intel_uc_fw_status load_status;
+
+	uint16_t major_ver_wanted;
+	uint16_t minor_ver_wanted;
+	uint16_t major_ver_found;
+	uint16_t minor_ver_found;
+
+	uint32_t fw_type;

What is the purpose of this field ?

Quoted 8 lines
Quoted 4 lines
Quoted 2 lines

It is used in the patch 2, to set {major,minor}_ver_found depending on which firmware we pass to intel_uc_fw_fetch()

Then maybe is shall be moved to patch 2 ?

Sound reasonable as that's the first usage, although is also fits here along making te structs general. I have no strong opinion on that though.

Will put it in patch 2 then. Thanks.

Quoted 49 lines
Quoted 41 lines
Quoted 41 lines
Quoted 40 lines

drivers/gpu/drm/i915/intel_uc.h struct intel_guc_log { }; · line 143

Quoted 10 lines
drivers/gpu/drm/i915/intel_uc.h +1 -1
 	uint32_t header_size;
 	uint32_t header_offset;
 	uint32_t rsa_size;
@@ -139,7 +141,7 @@ struct intel_guc_log {  };
 
 struct intel_guc {
-	struct intel_guc_fw guc_fw;
+	struct intel_uc_fw guc_fw;

Btw, guc_ prefix in field name here is also redundant.

Quoted 15 lines
drivers/gpu/drm/i915/intel_uc.h +0 -0
 	struct intel_guc_log log;
 
 	/* intel_guc_recv interrupt related state */ @@ -181,7 +183,7

@@ int intel_guc_log_control(struct intel_guc *guc, u32 control_val); extern void intel_guc_init(struct drm_i915_private *dev_priv); extern int intel_guc_setup(struct drm_i915_private *dev_priv); extern void intel_guc_fini(struct drm_i915_private *dev_priv); -extern const char *intel_guc_fw_status_repr(enum intel_guc_fw_status status);

drivers/gpu/drm/i915/intel_uc.h +2 -0
+extern const char *intel_uc_fw_status_repr(enum
+intel_uc_fw_status status);
 extern int intel_guc_suspend(struct drm_i915_private

*dev_priv); extern int intel_guc_resume(struct drm_i915_private *dev_priv);

Quoted 7 lines

--
2.7.4

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Quoted 7 lines
Quoted 7 lines

--
Cheers,
Arek
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Quoted 7 lines

--
Cheers,
Arek
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

author
#8 lore raw

PATCH 2/8 drm/i915/huc: Unified css_header struct for GuC and HuC

In reply to anushasr

From: Peter Antoine <[email protected]>

HuC firmware css header has almost exactly same definition as GuC firmware except for the sw_version. Also, add a new member fw_type into intel_uc_fw to indicate what kind of fw it is. So, the loader will pull right sw_version from header.

v2: rebased on-top of drm-intel-nightly
v3: rebased on-top of drm-intel-nightly (again).
v4: rebased + spaces.
v7: rebased.
v8: rebased.
v9: rebased. Rename device_id to guc_branch_client_version, make guc_sw_version a union. <Jeff Mcgee>. Put UC_FW_TYPE_GUC and UC_FW_TYPE_HUC into an enum.
v10: rebased.
v11: rebased.
v12: rebased on top of drm-tip.
v13: rebased.Update dev to dev_priv in intel_uc_fw_fetch

---
 drivers/gpu/drm/i915/intel_guc_fwif.h   | 21 +++++++++++++----
 drivers/gpu/drm/i915/intel_guc_loader.c | 41 ++++++++++++++++++++++-----------
 drivers/gpu/drm/i915/intel_uc.h         |  5 ++++
 3 files changed, 50 insertions(+), 17 deletions(-)
drivers/gpu/drm/i915/intel_guc_fwif.h +17 -4
index 3202b32..c1e7faf 100644
@@ -145,7 +145,7 @@
  * The GuC firmware layout looks like this:
  *
  *     +-------------------------------+
- *     |        guc_css_header         |
+ *     |         uc_css_header         |
  *     |                               |
  *     | contains major/minor version  |
  *     +-------------------------------+
@@ -172,9 +172,16 @@
  * 3. Length info of each component can be found in header, in dwords.
  * 4. Modulus and exponent key are not required by driver. They may not appear
  *    in fw. So driver will load a truncated firmware in this case.
+ *
+ * HuC firmware layout is same as GuC firmware.
+ *
+ * HuC firmware css header is different. However, the only difference is where
+ * the version information is saved. The uc_css_header is unified to support
+ * both. Driver should get HuC version from uc_css_header.huc_sw_version, while
+ * uc_css_header.guc_sw_version for GuC.
  */
 
-struct guc_css_header {
+struct uc_css_header {
 	uint32_t module_type;
 	/* header_size includes all non-uCode bits, including css_header, rsa
 	 * key, modulus key and exponent data. */
@@ -205,8 +212,14 @@ struct guc_css_header {
 
 	char username[8];
 	char buildnumber[12];
-	uint32_t device_id;
-	uint32_t guc_sw_version;
+	union {
+		uint32_t guc_branch_client_version;
+		uint32_t huc_sw_version;
+	};
+	union {
+		uint32_t guc_sw_version;
+		uint32_t huc_reserved;
+	};
 	uint32_t prod_preprod_fw;
 	uint32_t reserved[12];
 	uint32_t header_info;
drivers/gpu/drm/i915/intel_guc_loader.c +28 -13

75 lines not on this page · raw message

drivers/gpu/drm/i915/intel_uc.h +5 -0
index f9f598d..be89f0b 100644
@@ -98,6 +98,11 @@ enum intel_uc_fw_status {
 	UC_FIRMWARE_SUCCESS
 };
 
+enum {
+	UC_FW_TYPE_GUC,
+	UC_FW_TYPE_HUC
+};
+
 /*
  * This structure encapsulates all the data needed during the process
  * of fetching, caching, and loading the firmware image into the GuC.
Signature · 6 lines
2.7.4

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Reviewed-by on v1 2/8
#9 lore raw

Re: [PATCH 2/8] drm/i915/huc: Unified css_header struct for GuC and HuC

Quoted 20 lines · anushasr

From: Peter Antoine <[email protected]>

HuC firmware css header has almost exactly same definition as GuC firmware except for the sw_version. Also, add a new member fw_type into intel_uc_fw to indicate what kind of fw it is. So, the loader will pull right sw_version from header.

v2: rebased on-top of drm-intel-nightly
v3: rebased on-top of drm-intel-nightly (again).
v4: rebased + spaces.
v7: rebased.
v8: rebased.
v9: rebased. Rename device_id to guc_branch_client_version, make guc_sw_version a union. <Jeff Mcgee>. Put UC_FW_TYPE_GUC and UC_FW_TYPE_HUC into an enum.
v10: rebased.
v11: rebased.
v12: rebased on top of drm-tip.
v13: rebased.Update dev to dev_priv in intel_uc_fw_fetch

Quoted 4 lines
Quoted 5 lines
---
 drivers/gpu/drm/i915/intel_guc_fwif.h   | 21 +++++++++++++----
 drivers/gpu/drm/i915/intel_guc_loader.c | 41 ++++++++++++++++++++++-----------
 drivers/gpu/drm/i915/intel_uc.h         |  5 ++++
 3 files changed, 50 insertions(+), 17 deletions(-)
Signature · 6 lines
Cheers,
Arek
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
#10 lore raw

Re: [PATCH 2/8] drm/i915/huc: Unified css_header struct for GuC and HuC

In reply to anushasr
Quoted 48 lines · anushasr

From: Peter Antoine <[email protected]>

HuC firmware css header has almost exactly same definition as GuC firmware except for the sw_version. Also, add a new member fw_type into intel_uc_fw to indicate what kind of fw it is. So, the loader will pull right sw_version from header.

v2: rebased on-top of drm-intel-nightly
v3: rebased on-top of drm-intel-nightly (again).
v4: rebased + spaces.
v7: rebased.
v8: rebased.
v9: rebased. Rename device_id to guc_branch_client_version, make guc_sw_version a union. <Jeff Mcgee>. Put UC_FW_TYPE_GUC and UC_FW_TYPE_HUC into an enum.
v10: rebased.
v11: rebased.
v12: rebased on top of drm-tip.
v13: rebased.Update dev to dev_priv in intel_uc_fw_fetch

---
 drivers/gpu/drm/i915/intel_guc_fwif.h   | 21 +++++++++++++----
 drivers/gpu/drm/i915/intel_guc_loader.c | 41 ++++++++++++++++++++++-----------
 drivers/gpu/drm/i915/intel_uc.h         |  5 ++++
 3 files changed, 50 insertions(+), 17 deletions(-)
drivers/gpu/drm/i915/intel_guc_fwif.h +2 -1
index 3202b32..c1e7faf 100644
@@ -145,7 +145,7 @@
  * The GuC firmware layout looks like this:
  *
  *     +-------------------------------+
- *     |        guc_css_header         |
+ *     |         uc_css_header         |
  *     |                               |
  *     | contains major/minor version  |
  *     +-------------------------------+
@@ -172,9 +172,16 @@
  * 3. Length info of each component can be found in header, in dwords.
  * 4. Modulus and exponent key are not required by driver. They may not appear
  *    in fw. So driver will load a truncated firmware in this case.
+ *

drivers/gpu/drm/i915/intel_guc_fwif.h line 184

Quoted 10 lines
drivers/gpu/drm/i915/intel_guc_fwif.h +7 -1
+ * HuC firmware layout is same as GuC firmware.
+ *
+ * HuC firmware css header is different. However, the only difference is where
+ * the version information is saved. The uc_css_header is unified to support
+ * both. Driver should get HuC version from uc_css_header.huc_sw_version, while
+ * uc_css_header.guc_sw_version for GuC.
  */
 
-struct guc_css_header {
+struct uc_css_header {

Hmm, I'm wondering why we don't use "intel_uc_" prefix for structs defined here. It seems that only enums are defined with intel_ prefix.

Also, it looks that this struct definition is very similar to the intel_css_header defined in intel_csr.c. Are there any plans to unify them all ?

drivers/gpu/drm/i915/intel_guc_fwif.h line 185

Quoted 1 line
drivers/gpu/drm/i915/intel_guc_fwif.h +0 -0
 	uint32_t module_type;

What values are used here? Are they the same as used in fw_type?

Quoted 6 lines
drivers/gpu/drm/i915/intel_guc_fwif.h +0 -0
 	/* header_size includes all non-uCode bits, including css_header, rsa
 	 * key, modulus key and exponent data. */
@@ -205,8 +212,14 @@ struct guc_css_header {
 
 	char username[8];
 	char buildnumber[12];

drivers/gpu/drm/i915/intel_guc_fwif.h struct guc_css_header { · line 222

Quoted 10 lines
drivers/gpu/drm/i915/intel_guc_fwif.h +8 -2
-	uint32_t device_id;
-	uint32_t guc_sw_version;
+	union {
+		uint32_t guc_branch_client_version;
+		uint32_t huc_sw_version;
+	};
+	union {
+		uint32_t guc_sw_version;
+		uint32_t huc_reserved;
+	};

Maybe to make this a little easier to read we can use:

union {

  struct {
    uint32_t branch_client_version;
    uint32_t sw_version;
  } guc;
  struct {
    uint32_t sw_version;
    unit32_t reserved;

} huc;
};

Quoted 90 lines
drivers/gpu/drm/i915/intel_guc_fwif.h +0 -0
 	uint32_t prod_preprod_fw;
 	uint32_t reserved[12];
 	uint32_t header_info;
drivers/gpu/drm/i915/intel_guc_loader.c +28 -13
index 8f04f6e..26a184f 100644
@@ -593,7 +593,7 @@ void intel_uc_fw_fetch(struct drm_i915_private *dev_priv,
 	struct pci_dev *pdev = dev_priv->drm.pdev;
 	struct drm_i915_gem_object *obj;
 	const struct firmware *fw = NULL;
-	struct guc_css_header *css;
+	struct uc_css_header *css;
 	size_t size;
 	int err;
 
@@ -610,19 +610,19 @@ void intel_uc_fw_fetch(struct drm_i915_private *dev_priv,
 		uc_fw->uc_fw_path, fw);
 
 	/* Check the size of the blob before examining buffer contents */
-	if (fw->size < sizeof(struct guc_css_header)) {
+	if (fw->size < sizeof(struct uc_css_header)) {
 		DRM_NOTE("Firmware header is missing\n");
 		goto fail;
 	}
 
-	css = (struct guc_css_header *)fw->data;
+	css = (struct uc_css_header *)fw->data;
 
 	/* Firmware bits always start from header */
 	uc_fw->header_offset = 0;
 	uc_fw->header_size = (css->header_size_dw - css->modulus_size_dw -
 		css->key_size_dw - css->exponent_size_dw) * sizeof(u32);
 
-	if (uc_fw->header_size != sizeof(struct guc_css_header)) {
+	if (uc_fw->header_size != sizeof(struct uc_css_header)) {
 		DRM_NOTE("CSS header definition mismatch\n");
 		goto fail;
 	}
@@ -646,21 +646,36 @@ void intel_uc_fw_fetch(struct drm_i915_private *dev_priv,
 		goto fail;
 	}
 
-	/* Header and uCode will be loaded to WOPCM. Size of the two. */
-	size = uc_fw->header_size + uc_fw->ucode_size;
-	if (size > guc_wopcm_size(dev_priv)) {
-		DRM_NOTE("Firmware is too large to fit in WOPCM\n");
-		goto fail;
-	}
-
 	/*
 	 * 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).
 	 */
-	uc_fw->major_ver_found = css->guc_sw_version >> 16;
-	uc_fw->minor_ver_found = css->guc_sw_version & 0xFFFF;
+	switch (uc_fw->fw_type) {
+	case UC_FW_TYPE_GUC:
+		/* Header and uCode will be loaded to WOPCM. Size of the two. */
+		size = uc_fw->header_size + uc_fw->ucode_size;
+
+		/* Top 32k of WOPCM is reserved (8K stack + 24k RC6 context). */
+		if (size > guc_wopcm_size(dev_priv)) {
+			DRM_ERROR("Firmware is too large to fit in WOPCM\n");
+			goto fail;
+		}
+		uc_fw->major_ver_found = css->guc_sw_version >> 16;
+		uc_fw->minor_ver_found = css->guc_sw_version & 0xFFFF;
+		break;
+
+	case UC_FW_TYPE_HUC:
+		uc_fw->major_ver_found = css->huc_sw_version >> 16;
+		uc_fw->minor_ver_found = css->huc_sw_version & 0xFFFF;
+		break;
+
+	default:
+		DRM_ERROR("Unknown firmware type %d\n", uc_fw->fw_type);
+		err = -ENOEXEC;
+		goto fail;
+	}
 
 	if (uc_fw->major_ver_found != uc_fw->major_ver_wanted ||
 	    uc_fw->minor_ver_found < uc_fw->minor_ver_wanted) {
drivers/gpu/drm/i915/intel_uc.h +0 -0
index f9f598d..be89f0b 100644

drivers/gpu/drm/i915/intel_uc.h enum intel_uc_fw_status { · line 103

Quoted 6 lines
drivers/gpu/drm/i915/intel_uc.h +3 -0
@@ -98,6 +98,11 @@ enum intel_uc_fw_status {
 	UC_FIRMWARE_SUCCESS
 };
 
+enum {
+	UC_FW_TYPE_GUC,
+	UC_FW_TYPE_HUC

Can we have INTEL_ prefix here?

Quoted 12 lines
drivers/gpu/drm/i915/intel_uc.h +2 -0
+};
+
 /*
  * This structure encapsulates all the data needed during the process
  * of fetching, caching, and loading the firmware image into the GuC.
Signature · 6 lines
2.7.4

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

#11 lore raw

Re: [PATCH 2/8] drm/i915/huc: Unified css_header struct for GuC and HuC

Quoted 74 lines

-----Original Message-----
From: Michal Wajdeczko [mailto:[email protected]]
Sent: Friday, December 9, 2016 3:56 AM
To: Srivatsa, Anusha <[email protected]>
Cc: [email protected]; Alex Dai <[email protected]>; Peter Antoine <[email protected]>
Subject: Re: [Intel-gfx] [PATCH 2/8] drm/i915/huc: Unified css_header struct for GuC and HuC

Quoted 20 lines · anushasr

From: Peter Antoine <[email protected]>

HuC firmware css header has almost exactly same definition as GuC firmware except for the sw_version. Also, add a new member fw_type into intel_uc_fw to indicate what kind of fw it is. So, the loader will pull right sw_version from header.

v2: rebased on-top of drm-intel-nightly
v3: rebased on-top of drm-intel-nightly (again).
v4: rebased + spaces.
v7: rebased.
v8: rebased.
v9: rebased. Rename device_id to guc_branch_client_version, make guc_sw_version a union. <Jeff Mcgee>. Put UC_FW_TYPE_GUC and UC_FW_TYPE_HUC into an enum.
v10: rebased.
v11: rebased.
v12: rebased on top of drm-tip.
v13: rebased.Update dev to dev_priv in intel_uc_fw_fetch

Quoted 7 lines
---
 drivers/gpu/drm/i915/intel_guc_fwif.h   | 21 +++++++++++++----
 drivers/gpu/drm/i915/intel_guc_loader.c | 41 ++++++++++++++++++++++-------

----

Quoted 18 lines
 drivers/gpu/drm/i915/intel_uc.h         |  5 ++++
 3 files changed, 50 insertions(+), 17 deletions(-)
drivers/gpu/drm/i915/intel_guc_fwif.h +1 -1
index 3202b32..c1e7faf 100644
@@ -145,7 +145,7 @@
  * The GuC firmware layout looks like this:
  *
  *     +-------------------------------+
- *     |        guc_css_header         |
+ *     |         uc_css_header         |
  *     |                               |
  *     | contains major/minor version  |
  *     +-------------------------------+

drivers/gpu/drm/i915/intel_guc_fwif.h line 173

Quoted 2 lines
drivers/gpu/drm/i915/intel_guc_fwif.h +0 -0
@@ -172,9 +172,16 @@
  * 3. Length info of each component can be found in header, in dwords.
  * 4. Modulus and exponent key are not required by driver. They may not

appear

Quoted 13 lines
drivers/gpu/drm/i915/intel_guc_fwif.h +10 -0
  *    in fw. So driver will load a truncated firmware in this case.
+ *
+ * HuC firmware layout is same as GuC firmware.
+ *
+ * HuC firmware css header is different. However, the only difference
+ is where
+ * the version information is saved. The uc_css_header is unified to
+ support
+ * both. Driver should get HuC version from
+ uc_css_header.huc_sw_version, while
+ * uc_css_header.guc_sw_version for GuC.
  */
 
Quoted 5 lines

drivers/gpu/drm/i915/intel_guc_fwif.h line 184

Quoted 2 lines
drivers/gpu/drm/i915/intel_guc_fwif.h +1 -1
-struct guc_css_header {
+struct uc_css_header {

Hmm, I'm wondering why we don't use "intel_uc_" prefix for structs defined here. It seems that only enums are defined with intel_ prefix.

If we rename it to intel_uc_css_header, wont it be confused with intel_css_header in intel_csr.c? Unles we want to combine them....

Quoted 3 lines

Also, it looks that this struct definition is very similar to the intel_css_header defined in intel_csr.c. Are there any plans to unify them all ?

No idea about any plans regarding unifying the struct in intel_csr.c and this struct. Arek, any idea?

Quoted 32 lines
Quoted 1 line

uint32_t module_type;

What values are used here? Are they the same as used in fw_type?

Quoted 6 lines
drivers/gpu/drm/i915/intel_guc_fwif.h +0 -0
 	/* header_size includes all non-uCode bits, including css_header, rsa
 	 * key, modulus key and exponent data. */ @@ -205,8 +212,14 @@

struct guc_css_header {

drivers/gpu/drm/i915/intel_guc_fwif.h +0 -0
 	char username[8];
 	char buildnumber[12];

drivers/gpu/drm/i915/intel_guc_fwif.h struct guc_css_header { · line 222

Quoted 10 lines
drivers/gpu/drm/i915/intel_guc_fwif.h +8 -2
-	uint32_t device_id;
-	uint32_t guc_sw_version;
+	union {
+		uint32_t guc_branch_client_version;
+		uint32_t huc_sw_version;
+	};
+	union {
+		uint32_t guc_sw_version;
+		uint32_t huc_reserved;
+	};

Maybe to make this a little easier to read we can use:

union {

 struct {
   uint32_t branch_client_version;
   uint32_t sw_version;
 } guc;
 struct {
   uint32_t sw_version;
   unit32_t reserved;
Quoted 2 lines

} huc;
};

Yes. Will do.

Quoted 102 lines

drivers/gpu/drm/i915/intel_guc_loader.c void intel_uc_fw_fetch(struct drm_i915_private

Quoted 10 lines
drivers/gpu/drm/i915/intel_guc_fwif.h +0 -0
 	uint32_t prod_preprod_fw;
 	uint32_t reserved[12];
 	uint32_t header_info;
drivers/gpu/drm/i915/intel_guc_loader.c +0 -0
index 8f04f6e..26a184f 100644
@@ -593,7 +593,7 @@ void intel_uc_fw_fetch(struct drm_i915_private

*dev_priv,

drivers/gpu/drm/i915/intel_guc_loader.c void intel_uc_fw_fetch(struct drm_i915_private

Quoted 9 lines
drivers/gpu/drm/i915/intel_guc_loader.c +1 -1
 	struct pci_dev *pdev = dev_priv->drm.pdev;
 	struct drm_i915_gem_object *obj;
 	const struct firmware *fw = NULL;
-	struct guc_css_header *css;
+	struct uc_css_header *css;
 	size_t size;
 	int err;
 
@@ -610,19 +610,19 @@ void intel_uc_fw_fetch(struct drm_i915_private

*dev_priv,

drivers/gpu/drm/i915/intel_guc_loader.c void intel_uc_fw_fetch(struct drm_i915_private

Quoted 23 lines
drivers/gpu/drm/i915/intel_guc_loader.c +3 -3
 		uc_fw->uc_fw_path, fw);
 
 	/* Check the size of the blob before examining buffer contents */
-	if (fw->size < sizeof(struct guc_css_header)) {
+	if (fw->size < sizeof(struct uc_css_header)) {
 		DRM_NOTE("Firmware header is missing\n");
 		goto fail;
 	}
 
-	css = (struct guc_css_header *)fw->data;
+	css = (struct uc_css_header *)fw->data;
 
 	/* Firmware bits always start from header */
 	uc_fw->header_offset = 0;
 	uc_fw->header_size = (css->header_size_dw - css->modulus_size_dw -
 		css->key_size_dw - css->exponent_size_dw) * sizeof(u32);
 
-	if (uc_fw->header_size != sizeof(struct guc_css_header)) {
+	if (uc_fw->header_size != sizeof(struct uc_css_header)) {
 		DRM_NOTE("CSS header definition mismatch\n");
 		goto fail;
 	}
@@ -646,21 +646,36 @@ void intel_uc_fw_fetch(struct drm_i915_private

*dev_priv,

drivers/gpu/drm/i915/intel_guc_loader.c void intel_uc_fw_fetch(struct drm_i915_private *dev_priv,

Quoted 12 lines
drivers/gpu/drm/i915/intel_guc_loader.c +0 -7
 		goto fail;
 	}
 
-	/* Header and uCode will be loaded to WOPCM. Size of the two. */
-	size = uc_fw->header_size + uc_fw->ucode_size;
-	if (size > guc_wopcm_size(dev_priv)) {
-		DRM_NOTE("Firmware is too large to fit in WOPCM\n");
-		goto fail;
-	}
-
 	/*
 	 * The GuC firmware image has the version number embedded at a well-

known

drivers/gpu/drm/i915/intel_guc_loader.c void intel_uc_fw_fetch(struct drm_i915_private *dev_priv,

Quoted 9 lines
drivers/gpu/drm/i915/intel_guc_loader.c +3 -2
 	 * 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).
 	 */
-	uc_fw->major_ver_found = css->guc_sw_version >> 16;
-	uc_fw->minor_ver_found = css->guc_sw_version & 0xFFFF;
+	switch (uc_fw->fw_type) {
+	case UC_FW_TYPE_GUC:
+		/* Header and uCode will be loaded to WOPCM. Size of the two.

*/

drivers/gpu/drm/i915/intel_guc_loader.c void intel_uc_fw_fetch(struct drm_i915_private *dev_priv,

Quoted 3 lines
drivers/gpu/drm/i915/intel_guc_loader.c +3 -0
+		size = uc_fw->header_size + uc_fw->ucode_size;
+
+		/* Top 32k of WOPCM is reserved (8K stack + 24k RC6 context).

*/

drivers/gpu/drm/i915/intel_guc_loader.c void intel_uc_fw_fetch(struct drm_i915_private · line 662

Quoted 2 lines
drivers/gpu/drm/i915/intel_guc_loader.c +2 -0
+		if (size > guc_wopcm_size(dev_priv)) {
+			DRM_ERROR("Firmware is too large to fit in

WOPCM\n");

Quoted 27 lines
drivers/gpu/drm/i915/intel_guc_loader.c +16 -0
+			goto fail;
+		}
+		uc_fw->major_ver_found = css->guc_sw_version >> 16;
+		uc_fw->minor_ver_found = css->guc_sw_version & 0xFFFF;
+		break;
+
+	case UC_FW_TYPE_HUC:
+		uc_fw->major_ver_found = css->huc_sw_version >> 16;
+		uc_fw->minor_ver_found = css->huc_sw_version & 0xFFFF;
+		break;
+
+	default:
+		DRM_ERROR("Unknown firmware type %d\n", uc_fw->fw_type);
+		err = -ENOEXEC;
+		goto fail;
+	}
 
 	if (uc_fw->major_ver_found != uc_fw->major_ver_wanted ||
 	    uc_fw->minor_ver_found < uc_fw->minor_ver_wanted) { diff --git

a/drivers/gpu/drm/i915/intel_uc.h b/drivers/gpu/drm/i915/intel_uc.h index f9f598d..be89f0b 100644

drivers/gpu/drm/i915/intel_uc.h +0 -0
@@ -98,6 +98,11 @@ enum intel_uc_fw_status {
 	UC_FIRMWARE_SUCCESS
 };
 
Quoted 8 lines

drivers/gpu/drm/i915/intel_uc.h enum intel_uc_fw_status { · line 103

Quoted 3 lines
drivers/gpu/drm/i915/intel_uc.h +3 -0
+enum {
+	UC_FW_TYPE_GUC,
+	UC_FW_TYPE_HUC

Can we have INTEL_ prefix here?

Quoted 2 lines
drivers/gpu/drm/i915/intel_uc.h +1 -0
+};

Yes. Thanks Michal.

Anusha

Quoted 10 lines
Quoted 10 lines

/*
* This structure encapsulates all the data needed during the process
* of fetching, caching, and loading the firmware image into the GuC.
--
2.7.4

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

#12 lore raw

Re: [PATCH 2/8] drm/i915/huc: Unified css_header struct for GuC and HuC

Quoted 79 lines · Anusha
Quoted 74 lines

-----Original Message-----
From: Michal Wajdeczko [mailto:[email protected]]
Sent: Friday, December 9, 2016 3:56 AM
To: Srivatsa, Anusha <[email protected]>
Cc: [email protected]; Alex Dai <[email protected]>; Peter Antoine <[email protected]>
Subject: Re: [Intel-gfx] [PATCH 2/8] drm/i915/huc: Unified css_header struct for GuC and HuC

Quoted 20 lines · anushasr

From: Peter Antoine <[email protected]>

HuC firmware css header has almost exactly same definition as GuC firmware except for the sw_version. Also, add a new member fw_type into intel_uc_fw to indicate what kind of fw it is. So, the loader will pull right sw_version from header.

v2: rebased on-top of drm-intel-nightly
v3: rebased on-top of drm-intel-nightly (again).
v4: rebased + spaces.
v7: rebased.
v8: rebased.
v9: rebased. Rename device_id to guc_branch_client_version, make guc_sw_version a union. <Jeff Mcgee>. Put UC_FW_TYPE_GUC and UC_FW_TYPE_HUC into an enum.
v10: rebased.
v11: rebased.
v12: rebased on top of drm-tip.
v13: rebased.Update dev to dev_priv in intel_uc_fw_fetch

Quoted 7 lines
---
 drivers/gpu/drm/i915/intel_guc_fwif.h   | 21 +++++++++++++----
 drivers/gpu/drm/i915/intel_guc_loader.c | 41 ++++++++++++++++++++++-------

----

Quoted 18 lines
 drivers/gpu/drm/i915/intel_uc.h         |  5 ++++
 3 files changed, 50 insertions(+), 17 deletions(-)
drivers/gpu/drm/i915/intel_guc_fwif.h +1 -1
index 3202b32..c1e7faf 100644
@@ -145,7 +145,7 @@
  * The GuC firmware layout looks like this:
  *
  *     +-------------------------------+
- *     |        guc_css_header         |
+ *     |         uc_css_header         |
  *     |                               |
  *     | contains major/minor version  |
  *     +-------------------------------+

drivers/gpu/drm/i915/intel_guc_fwif.h line 173

Quoted 2 lines
drivers/gpu/drm/i915/intel_guc_fwif.h +0 -0
@@ -172,9 +172,16 @@
  * 3. Length info of each component can be found in header, in dwords.
  * 4. Modulus and exponent key are not required by driver. They may not

appear

Quoted 13 lines
drivers/gpu/drm/i915/intel_guc_fwif.h +10 -0
  *    in fw. So driver will load a truncated firmware in this case.
+ *
+ * HuC firmware layout is same as GuC firmware.
+ *
+ * HuC firmware css header is different. However, the only difference
+ is where
+ * the version information is saved. The uc_css_header is unified to
+ support
+ * both. Driver should get HuC version from
+ uc_css_header.huc_sw_version, while
+ * uc_css_header.guc_sw_version for GuC.
  */
 
Quoted 5 lines

drivers/gpu/drm/i915/intel_guc_fwif.h line 184

Quoted 2 lines
drivers/gpu/drm/i915/intel_guc_fwif.h +1 -1
-struct guc_css_header {
+struct uc_css_header {

Hmm, I'm wondering why we don't use "intel_uc_" prefix for structs defined here. It seems that only enums are defined with intel_ prefix.

Quoted 6 lines

If we rename it to intel_uc_css_header, wont it be confused with intel_css_header in intel_csr.c? Unles we want to combine them....

Quoted 3 lines

Also, it looks that this struct definition is very similar to the intel_css_header defined in intel_csr.c. Are there any plans to unify them all ?

No idea about any plans regarding unifying the struct in intel_csr.c and this struct. Arek, any idea?

Currently not, but that seems like idea worth exploring. I'll look into it this week and if it turns sensible to do, I'll add it to my GuC cleanup series.

Quoted 3 lines
Quoted 3 lines
Quoted 1 line

uint32_t module_type;

What values are used here? Are they the same as used in fw_type?

module_type is just a part of HuC/GuC's firmware header (css). You can read it from the blobs:

module_type for guc == 0006 0000 00a1 0000 (v6, v4) module_type for huc == 0006 0000 00a1 0000 (v1)

It's kind of magic value to determine it's a firmware (I am not aware of other values than the one above, but that might change in the future). We do not check for that exact field though. We do not use it at all, but that, as well, might change in the future.

fw_type provides means to distingush between HuC/GuC FW.

Quoted 158 lines
Quoted 28 lines
Quoted 7 lines
drivers/gpu/drm/i915/intel_guc_fwif.h +0 -0
 	/* header_size includes all non-uCode bits, including css_header, rsa
 	 * key, modulus key and exponent data. */ @@ -205,8 +212,14 @@

struct guc_css_header {

drivers/gpu/drm/i915/intel_guc_fwif.h +0 -0
 	char username[8];
 	char buildnumber[12];

drivers/gpu/drm/i915/intel_guc_fwif.h struct guc_css_header { · line 222

Quoted 10 lines
drivers/gpu/drm/i915/intel_guc_fwif.h +8 -2
-	uint32_t device_id;
-	uint32_t guc_sw_version;
+	union {
+		uint32_t guc_branch_client_version;
+		uint32_t huc_sw_version;
+	};
+	union {
+		uint32_t guc_sw_version;
+		uint32_t huc_reserved;
+	};

Maybe to make this a little easier to read we can use:

union {

 struct {
   uint32_t branch_client_version;
   uint32_t sw_version;
 } guc;
 struct {
   uint32_t sw_version;
   unit32_t reserved;
Quoted 2 lines

} huc;
};

Yes. Will do.

Quoted 102 lines

drivers/gpu/drm/i915/intel_guc_loader.c void intel_uc_fw_fetch(struct drm_i915_private

Quoted 10 lines
drivers/gpu/drm/i915/intel_guc_fwif.h +0 -0
 	uint32_t prod_preprod_fw;
 	uint32_t reserved[12];
 	uint32_t header_info;
drivers/gpu/drm/i915/intel_guc_loader.c +0 -0
index 8f04f6e..26a184f 100644
@@ -593,7 +593,7 @@ void intel_uc_fw_fetch(struct drm_i915_private

*dev_priv,

drivers/gpu/drm/i915/intel_guc_loader.c void intel_uc_fw_fetch(struct drm_i915_private

Quoted 9 lines
drivers/gpu/drm/i915/intel_guc_loader.c +1 -1
 	struct pci_dev *pdev = dev_priv->drm.pdev;
 	struct drm_i915_gem_object *obj;
 	const struct firmware *fw = NULL;
-	struct guc_css_header *css;
+	struct uc_css_header *css;
 	size_t size;
 	int err;
 
@@ -610,19 +610,19 @@ void intel_uc_fw_fetch(struct drm_i915_private

*dev_priv,

drivers/gpu/drm/i915/intel_guc_loader.c void intel_uc_fw_fetch(struct drm_i915_private

Quoted 23 lines
drivers/gpu/drm/i915/intel_guc_loader.c +3 -3
 		uc_fw->uc_fw_path, fw);
 
 	/* Check the size of the blob before examining buffer contents */
-	if (fw->size < sizeof(struct guc_css_header)) {
+	if (fw->size < sizeof(struct uc_css_header)) {
 		DRM_NOTE("Firmware header is missing\n");
 		goto fail;
 	}
 
-	css = (struct guc_css_header *)fw->data;
+	css = (struct uc_css_header *)fw->data;
 
 	/* Firmware bits always start from header */
 	uc_fw->header_offset = 0;
 	uc_fw->header_size = (css->header_size_dw - css->modulus_size_dw -
 		css->key_size_dw - css->exponent_size_dw) * sizeof(u32);
 
-	if (uc_fw->header_size != sizeof(struct guc_css_header)) {
+	if (uc_fw->header_size != sizeof(struct uc_css_header)) {
 		DRM_NOTE("CSS header definition mismatch\n");
 		goto fail;
 	}
@@ -646,21 +646,36 @@ void intel_uc_fw_fetch(struct drm_i915_private

*dev_priv,

drivers/gpu/drm/i915/intel_guc_loader.c void intel_uc_fw_fetch(struct drm_i915_private *dev_priv,

Quoted 12 lines
drivers/gpu/drm/i915/intel_guc_loader.c +0 -7
 		goto fail;
 	}
 
-	/* Header and uCode will be loaded to WOPCM. Size of the two. */
-	size = uc_fw->header_size + uc_fw->ucode_size;
-	if (size > guc_wopcm_size(dev_priv)) {
-		DRM_NOTE("Firmware is too large to fit in WOPCM\n");
-		goto fail;
-	}
-
 	/*
 	 * The GuC firmware image has the version number embedded at a well-

known

drivers/gpu/drm/i915/intel_guc_loader.c void intel_uc_fw_fetch(struct drm_i915_private *dev_priv,

Quoted 9 lines
drivers/gpu/drm/i915/intel_guc_loader.c +3 -2
 	 * 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).
 	 */
-	uc_fw->major_ver_found = css->guc_sw_version >> 16;
-	uc_fw->minor_ver_found = css->guc_sw_version & 0xFFFF;
+	switch (uc_fw->fw_type) {
+	case UC_FW_TYPE_GUC:
+		/* Header and uCode will be loaded to WOPCM. Size of the two.

*/

drivers/gpu/drm/i915/intel_guc_loader.c void intel_uc_fw_fetch(struct drm_i915_private *dev_priv,

Quoted 3 lines
drivers/gpu/drm/i915/intel_guc_loader.c +3 -0
+		size = uc_fw->header_size + uc_fw->ucode_size;
+
+		/* Top 32k of WOPCM is reserved (8K stack + 24k RC6 context).

*/

drivers/gpu/drm/i915/intel_guc_loader.c void intel_uc_fw_fetch(struct drm_i915_private · line 662

Quoted 2 lines
drivers/gpu/drm/i915/intel_guc_loader.c +2 -0
+		if (size > guc_wopcm_size(dev_priv)) {
+			DRM_ERROR("Firmware is too large to fit in

WOPCM\n");

Quoted 27 lines
drivers/gpu/drm/i915/intel_guc_loader.c +16 -0
+			goto fail;
+		}
+		uc_fw->major_ver_found = css->guc_sw_version >> 16;
+		uc_fw->minor_ver_found = css->guc_sw_version & 0xFFFF;
+		break;
+
+	case UC_FW_TYPE_HUC:
+		uc_fw->major_ver_found = css->huc_sw_version >> 16;
+		uc_fw->minor_ver_found = css->huc_sw_version & 0xFFFF;
+		break;
+
+	default:
+		DRM_ERROR("Unknown firmware type %d\n", uc_fw->fw_type);
+		err = -ENOEXEC;
+		goto fail;
+	}
 
 	if (uc_fw->major_ver_found != uc_fw->major_ver_wanted ||
 	    uc_fw->minor_ver_found < uc_fw->minor_ver_wanted) { diff --git

a/drivers/gpu/drm/i915/intel_uc.h b/drivers/gpu/drm/i915/intel_uc.h index f9f598d..be89f0b 100644

drivers/gpu/drm/i915/intel_uc.h +0 -0
@@ -98,6 +98,11 @@ enum intel_uc_fw_status {
 	UC_FIRMWARE_SUCCESS
 };
 
Quoted 8 lines

drivers/gpu/drm/i915/intel_uc.h enum intel_uc_fw_status { · line 103

Quoted 3 lines
drivers/gpu/drm/i915/intel_uc.h +3 -0
+enum {
+	UC_FW_TYPE_GUC,
+	UC_FW_TYPE_HUC

Can we have INTEL_ prefix here?

Quoted 2 lines
drivers/gpu/drm/i915/intel_uc.h +1 -0
+};

Yes. Thanks Michal.

Anusha

Quoted 10 lines
Quoted 10 lines

/*
* This structure encapsulates all the data needed during the process
* of fetching, caching, and loading the firmware image into the GuC.
--
2.7.4

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Signature · 6 lines
Cheers,
Arek
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
author
#13 lore raw

PATCH 3/8 drm/i915/huc: Add HuC fw loading support

In reply to anushasr

From: Anusha Srivatsa <[email protected]>

The HuC loading process is similar to GuC. The intel_uc_fw_fetch() is used for both cases.

HuC loading needs to be before GuC loading. The WOPCM setting must be done early before loading any of them.

v2: rebased on-top of drm-intel-nightly.
removed if(HAS_GUC()) before the guc call. (D.Gordon) update huc_version number of format.
v3: rebased to drm-intel-nightly, changed the file name format to
match the one in the huc package.
Changed dev->dev_private to to_i915()
v4: moved function back to where it was.
change wait_for_atomic to wait_for.
v5: rebased + comment changes.
v7: rebased.
v8: rebased.
v9: rebased. Changed the year in the copyright message to reflect the right year.Correct the comments,remove the unwanted WARN message, replace drm_gem_object_unreference() with i915_gem_object_put().Make the prototypes in intel_huc.h non-extern.
v10: rebased. Update the file construction done by HuC. It is similar to GuC.Adopted the approach used in-
https://patchwork.freedesktop.org/patch/104355/ <Tvrtko Ursulin>
v11: Fix warnings remove old declaration
v12: Change dev to dev_priv in macro definition.
Corrected comments.
v13: rebased.
v14: rebased on top of drm-tip
v15: rebased. Updated functions intel_huc_load(),intel_huc_init() and intel_uc_fw_fetch() to accept dev_priv instead of dev. Moved contents of intel_huc.h to intel_uc.h

---
 drivers/gpu/drm/i915/Makefile           |   1 +
 drivers/gpu/drm/i915/i915_drv.c         |   4 +-
 drivers/gpu/drm/i915/i915_drv.h         |   3 +-
 drivers/gpu/drm/i915/i915_guc_reg.h     |   3 +
 drivers/gpu/drm/i915/intel_guc_loader.c |   7 +-
 drivers/gpu/drm/i915/intel_huc_loader.c | 264 ++++++++++++++++++++++++++++++++
 drivers/gpu/drm/i915/intel_uc.h         |  22 +++
 7 files changed, 299 insertions(+), 5 deletions(-)
 create mode 100644 drivers/gpu/drm/i915/intel_huc_loader.c
drivers/gpu/drm/i915/Makefile +1 -0
index 3c30916..01d4f4b 100644
@@ -57,6 +57,7 @@ i915-y += i915_cmd_parser.o \
 # general-purpose microcontroller (GuC) support
 i915-y += intel_uc.o \
 	  intel_guc_loader.o \
+	  intel_huc_loader.o \
 	  i915_guc_submission.o
 
 # autogenerated null render state
drivers/gpu/drm/i915/i915_drv.c +3 -1
index 6428588..85a47c2 100644
@@ -600,6 +600,7 @@ static int i915_load_modeset_init(struct drm_device *dev)
 	if (ret)
 		goto cleanup_irq;
 
+	intel_huc_init(dev_priv);
 	intel_guc_init(dev_priv);
 
 	ret = i915_gem_init(dev_priv);
@@ -627,6 +628,7 @@ static int i915_load_modeset_init(struct drm_device *dev)
 		DRM_ERROR("failed to idle hardware; continuing to unload!\n");
 	i915_gem_fini(dev_priv);
 cleanup_irq:
+	intel_huc_fini(dev);
 	intel_guc_fini(dev_priv);
 	drm_irq_uninstall(dev);
 	intel_teardown_gmbus(dev_priv);
@@ -1313,7 +1315,7 @@ void i915_driver_unload(struct drm_device *dev)
 
 	/* Flush any outstanding unpin_work. */
 	drain_workqueue(dev_priv->wq);
-
+	intel_huc_fini(dev);
 	intel_guc_fini(dev_priv);
 	i915_gem_fini(dev_priv);
 	intel_fbc_cleanup_cfb(dev_priv);
drivers/gpu/drm/i915/i915_drv.h +2 -1
index 1480e73..0371ca4 100644
@@ -2036,6 +2036,7 @@ struct drm_i915_private {
 
 	struct intel_gvt *gvt;
 
+	struct intel_huc huc;
 	struct intel_guc guc;
 
 	struct intel_csr csr;
@@ -2810,7 +2811,7 @@ intel_info(const struct drm_i915_private *dev_priv)
 #define HAS_GUC(dev_priv)	((dev_priv)->info.has_guc)
 #define HAS_GUC_UCODE(dev_priv)	(HAS_GUC(dev_priv))
 #define HAS_GUC_SCHED(dev_priv)	(HAS_GUC(dev_priv))
-
+#define HAS_HUC_UCODE(dev_priv)	(HAS_GUC(dev_priv))
 #define HAS_RESOURCE_STREAMER(dev_priv) ((dev_priv)->info.has_resource_streamer)
 
 #define HAS_POOLED_EU(dev_priv)	((dev_priv)->info.has_pooled_eu)
drivers/gpu/drm/i915/i915_guc_reg.h +3 -0
index 5e638fc..f9829f6 100644
@@ -61,9 +61,12 @@
 #define   DMA_ADDRESS_SPACE_GTT		  (8 << 16)
 #define DMA_COPY_SIZE			_MMIO(0xc310)
 #define DMA_CTRL			_MMIO(0xc314)
+#define   HUC_UKERNEL			  (1<<9)
 #define   UOS_MOVE			  (1<<4)
 #define   START_DMA			  (1<<0)
 #define DMA_GUC_WOPCM_OFFSET		_MMIO(0xc340)
+#define   HUC_LOADING_AGENT_VCR		  (0<<1)
+#define   HUC_LOADING_AGENT_GUC		  (1<<1)
 #define   GUC_WOPCM_OFFSET_VALUE	  0x80000	/* 512KB */
 #define GUC_MAX_IDLE_COUNT		_MMIO(0xC3E4)
 
drivers/gpu/drm/i915/intel_guc_loader.c +4 -3
index 26a184f..b971351 100644
@@ -309,8 +309,8 @@ static int guc_ucode_xfer_dma(struct drm_i915_private *dev_priv,
 	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));
-
+	I915_WRITE(DMA_CTRL, _MASKED_BIT_ENABLE(UOS_MOVE | START_DMA) |
+		_MASKED_BIT_DISABLE(HUC_UKERNEL));
 	/*
 	 * Wait for the DMA to complete & the GuC to start up.
 	 * NB: Docs recommend not using the interrupt for completion.
@@ -334,7 +334,7 @@ static int guc_ucode_xfer_dma(struct drm_i915_private *dev_priv,
 	return ret;
 }
 
-static u32 guc_wopcm_size(struct drm_i915_private *dev_priv)
+u32 guc_wopcm_size(struct drm_i915_private *dev_priv)
 {
 	u32 wopcm_size = GUC_WOPCM_TOP;
 
@@ -511,6 +511,7 @@ int intel_guc_setup(struct drm_i915_private *dev_priv)
 		if (err)
 			goto fail;
 
+		intel_huc_load(dev_priv);
 		err = guc_ucode_xfer(dev_priv);
 		if (!err)
 			break;
drivers/gpu/drm/i915/intel_huc_loader.c new file +264 -0

264 lines not on this page · raw message

drivers/gpu/drm/i915/intel_uc.h +22 -0
index be89f0b..ac92946 100644
@@ -24,6 +24,12 @@
 #ifndef _INTEL_UC_H_
 #define _INTEL_UC_H_
 
+#ifndef _INTEL_HUC_H_
+#define _INTEL_HUC_H_
+
+#define HUC_STATUS2             _MMIO(0xD3B0)
+#define   HUC_FW_VERIFIED       (1<<7)
+
 #include "intel_guc_fwif.h"
 #include "i915_guc_reg.h"
 #include "intel_ringbuffer.h"
@@ -175,6 +181,13 @@ struct intel_guc {
 	struct mutex send_mutex;
 };
 
+struct intel_huc {
+	/* Generic uC firmware management */
+	struct intel_uc_fw huc_fw;
+
+	/* HuC-specific additions */
+};
+
 /* intel_uc.c */
 void intel_uc_init_early(struct drm_i915_private *dev_priv);
 bool intel_guc_recv(struct drm_i915_private *dev_priv, u32 *status);
@@ -191,6 +204,9 @@ extern void intel_guc_fini(struct drm_i915_private *dev_priv);
 extern const char *intel_uc_fw_status_repr(enum intel_uc_fw_status status);
 extern int intel_guc_suspend(struct drm_i915_private *dev_priv);
 extern int intel_guc_resume(struct drm_i915_private *dev_priv);
+void intel_uc_fw_fetch(struct drm_i915_private *dev_priv,
+	struct intel_uc_fw *uc_fw);
+u32 guc_wopcm_size(struct drm_i915_private *dev_priv);
 
 /* i915_guc_submission.c */
 int i915_guc_submission_init(struct drm_i915_private *dev_priv);
@@ -205,4 +221,10 @@ void i915_guc_register(struct drm_i915_private *dev_priv);
 void i915_guc_unregister(struct drm_i915_private *dev_priv);
 int i915_guc_log_control(struct drm_i915_private *dev_priv, u64 control_val);
 
+/* intel_huc_loader.c */
+void intel_huc_init(struct drm_i915_private *dev_priv);
+void intel_huc_fini(struct drm_device *dev);
+int intel_huc_load(struct drm_i915_private *dev_priv);
+
+#endif
 #endif
Signature · 6 lines
2.7.4

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
#14 lore raw

Re: [PATCH 3/8] drm/i915/huc: Add HuC fw loading support

Quoted 208 lines · anushasr

From: Anusha Srivatsa <[email protected]>

The HuC loading process is similar to GuC. The intel_uc_fw_fetch() is used for both cases.

HuC loading needs to be before GuC loading. The WOPCM setting must be done early before loading any of them.

v2: rebased on-top of drm-intel-nightly.
removed if(HAS_GUC()) before the guc call. (D.Gordon) update huc_version number of format.
v3: rebased to drm-intel-nightly, changed the file name format to
match the one in the huc package.
Changed dev->dev_private to to_i915()
v4: moved function back to where it was.
change wait_for_atomic to wait_for.
v5: rebased + comment changes.
v7: rebased.
v8: rebased.
v9: rebased. Changed the year in the copyright message to reflect the right year.Correct the comments,remove the unwanted WARN message, replace drm_gem_object_unreference() with i915_gem_object_put().Make the prototypes in intel_huc.h non-extern.
v10: rebased. Update the file construction done by HuC. It is similar to GuC.Adopted the approach used in-
https://patchwork.freedesktop.org/patch/104355/ <Tvrtko Ursulin>
v11: Fix warnings remove old declaration
v12: Change dev to dev_priv in macro definition.
Corrected comments.
v13: rebased.
v14: rebased on top of drm-tip
v15: rebased. Updated functions intel_huc_load(),intel_huc_init() and intel_uc_fw_fetch() to accept dev_priv instead of dev. Moved contents of intel_huc.h to intel_uc.h

---
 drivers/gpu/drm/i915/Makefile           |   1 +
 drivers/gpu/drm/i915/i915_drv.c         |   4 +-
 drivers/gpu/drm/i915/i915_drv.h         |   3 +-
 drivers/gpu/drm/i915/i915_guc_reg.h     |   3 +
 drivers/gpu/drm/i915/intel_guc_loader.c |   7 +-
 drivers/gpu/drm/i915/intel_huc_loader.c | 264 ++++++++++++++++++++++++++++++++
 drivers/gpu/drm/i915/intel_uc.h         |  22 +++
 7 files changed, 299 insertions(+), 5 deletions(-)
 create mode 100644 drivers/gpu/drm/i915/intel_huc_loader.c
drivers/gpu/drm/i915/Makefile +1 -0
index 3c30916..01d4f4b 100644
@@ -57,6 +57,7 @@ i915-y += i915_cmd_parser.o \
 # general-purpose microcontroller (GuC) support
 i915-y += intel_uc.o \
 	  intel_guc_loader.o \
+	  intel_huc_loader.o \
 	  i915_guc_submission.o
 
 # autogenerated null render state
drivers/gpu/drm/i915/i915_drv.c +3 -1
index 6428588..85a47c2 100644
@@ -600,6 +600,7 @@ static int i915_load_modeset_init(struct drm_device *dev)
 	if (ret)
 		goto cleanup_irq;
 
+	intel_huc_init(dev_priv);
 	intel_guc_init(dev_priv);
 
 	ret = i915_gem_init(dev_priv);
@@ -627,6 +628,7 @@ static int i915_load_modeset_init(struct drm_device *dev)
 		DRM_ERROR("failed to idle hardware; continuing to unload!\n");
 	i915_gem_fini(dev_priv);
 cleanup_irq:
+	intel_huc_fini(dev);
 	intel_guc_fini(dev_priv);
 	drm_irq_uninstall(dev);
 	intel_teardown_gmbus(dev_priv);
@@ -1313,7 +1315,7 @@ void i915_driver_unload(struct drm_device *dev)
 
 	/* Flush any outstanding unpin_work. */
 	drain_workqueue(dev_priv->wq);
-
+	intel_huc_fini(dev);
 	intel_guc_fini(dev_priv);
 	i915_gem_fini(dev_priv);
 	intel_fbc_cleanup_cfb(dev_priv);
drivers/gpu/drm/i915/i915_drv.h +2 -1
index 1480e73..0371ca4 100644
@@ -2036,6 +2036,7 @@ struct drm_i915_private {
 
 	struct intel_gvt *gvt;
 
+	struct intel_huc huc;
 	struct intel_guc guc;
 
 	struct intel_csr csr;
@@ -2810,7 +2811,7 @@ intel_info(const struct drm_i915_private *dev_priv)
 #define HAS_GUC(dev_priv)	((dev_priv)->info.has_guc)
 #define HAS_GUC_UCODE(dev_priv)	(HAS_GUC(dev_priv))
 #define HAS_GUC_SCHED(dev_priv)	(HAS_GUC(dev_priv))
-
+#define HAS_HUC_UCODE(dev_priv)	(HAS_GUC(dev_priv))
 #define HAS_RESOURCE_STREAMER(dev_priv) ((dev_priv)->info.has_resource_streamer)
 
 #define HAS_POOLED_EU(dev_priv)	((dev_priv)->info.has_pooled_eu)
drivers/gpu/drm/i915/i915_guc_reg.h +3 -0
index 5e638fc..f9829f6 100644
@@ -61,9 +61,12 @@
 #define   DMA_ADDRESS_SPACE_GTT		  (8 << 16)
 #define DMA_COPY_SIZE			_MMIO(0xc310)
 #define DMA_CTRL			_MMIO(0xc314)
+#define   HUC_UKERNEL			  (1<<9)
 #define   UOS_MOVE			  (1<<4)
 #define   START_DMA			  (1<<0)
 #define DMA_GUC_WOPCM_OFFSET		_MMIO(0xc340)
+#define   HUC_LOADING_AGENT_VCR		  (0<<1)
+#define   HUC_LOADING_AGENT_GUC		  (1<<1)
 #define   GUC_WOPCM_OFFSET_VALUE	  0x80000	/* 512KB */
 #define GUC_MAX_IDLE_COUNT		_MMIO(0xC3E4)
 
drivers/gpu/drm/i915/intel_guc_loader.c +4 -3
index 26a184f..b971351 100644
@@ -309,8 +309,8 @@ static int guc_ucode_xfer_dma(struct drm_i915_private *dev_priv,
 	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));
-
+	I915_WRITE(DMA_CTRL, _MASKED_BIT_ENABLE(UOS_MOVE | START_DMA) |
+		_MASKED_BIT_DISABLE(HUC_UKERNEL));
 	/*
 	 * Wait for the DMA to complete & the GuC to start up.
 	 * NB: Docs recommend not using the interrupt for completion.
@@ -334,7 +334,7 @@ static int guc_ucode_xfer_dma(struct drm_i915_private *dev_priv,
 	return ret;
 }
 
-static u32 guc_wopcm_size(struct drm_i915_private *dev_priv)
+u32 guc_wopcm_size(struct drm_i915_private *dev_priv)
 {
 	u32 wopcm_size = GUC_WOPCM_TOP;
 
@@ -511,6 +511,7 @@ int intel_guc_setup(struct drm_i915_private *dev_priv)
 		if (err)
 			goto fail;
 
+		intel_huc_load(dev_priv);
 		err = guc_ucode_xfer(dev_priv);
 		if (!err)
 			break;
drivers/gpu/drm/i915/intel_huc_loader.c new file +40 -0
new file mode 100644
index 0000000..e0efd1c
@@ -0,0 +1,264 @@
+/*
+ * Copyright © 2016 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 <linux/firmware.h>
+#include "i915_drv.h"
+#include "intel_uc.h"
+
+/**
+ * DOC: HuC Firmware
+ *
+ * Motivation:
+ * GEN9 introduces a new dedicated firmware for usage in media HEVC (High
+ * Efficiency Video Coding) operations. Userspace can use the firmware
+ * capabilities by adding HuC specific commands to batch buffers.
+ *
+ * Implementation:
+ * The same firmware loader is used as the GuC. However, the actual
+ * loading to HW is deferred until GEM initialization is done.
+ *
+ * Note that HuC firmware loading must be done before GuC loading.

drivers/gpu/drm/i915/intel_huc_loader.c line 50

Quoted 10 lines
drivers/gpu/drm/i915/intel_huc_loader.c new file +10 -0
+ */
+
+#define SKL_FW_MAJOR 01
+#define SKL_FW_MINOR 07
+#define SKL_BLD_NUM 1398
+
+#define HUC_FW_PATH(platform, major, minor, bld_num) \
+	"i915/" __stringify(platform) "_huc_ver" __stringify(major) "_" \
+	__stringify(minor) "_" __stringify(bld_num) ".bin"
+

Generally the patch looks well, but I'll withold my r-b till we have a consensus on what parts of the huc version should we lock the kernel.

Quoted 273 lines
drivers/gpu/drm/i915/intel_huc_loader.c new file +214 -0
+#define I915_SKL_HUC_UCODE HUC_FW_PATH(skl, SKL_FW_MAJOR, \
+	SKL_FW_MINOR, SKL_BLD_NUM)
+MODULE_FIRMWARE(I915_SKL_HUC_UCODE);
+
+/**
+ * huc_ucode_xfer() - DMA's the firmware
+ * @dev_priv: the drm device
+ *
+ * This function takes the gem object containing the firmware, sets up the DMA
+ * engine MMIO, triggers the DMA operation and waits for it to finish.
+ *
+ * Transfer the firmware image to RAM for execution by the microcontroller.
+ *
+ * Return: 0 on success, non-zero on failure
+ */
+
+static int huc_ucode_xfer(struct drm_i915_private *dev_priv)
+{
+	struct intel_uc_fw *huc_fw = &dev_priv->huc.huc_fw;
+	struct i915_vma *vma;
+	unsigned long offset = 0;
+	u32 size;
+	int ret;
+
+	ret = i915_gem_object_set_to_gtt_domain(huc_fw->uc_fw_obj, false);
+	if (ret) {
+		DRM_DEBUG_DRIVER("set-domain failed %d\n", ret);
+		return ret;
+	}
+
+	vma = i915_gem_object_ggtt_pin(huc_fw->uc_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);
+	}
+
+	/* 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 */
+	I915_WRITE(GUC_WOPCM_SIZE, guc_wopcm_size(dev_priv));
+	I915_WRITE(DMA_GUC_WOPCM_OFFSET, GUC_WOPCM_OFFSET_VALUE |
+			HUC_LOADING_AGENT_GUC);
+
+	/* Set the source address for the uCode */
+	offset = i915_ggtt_offset(vma) + huc_fw->header_offset;
+	I915_WRITE(DMA_ADDR_0_LOW, lower_32_bits(offset));
+	I915_WRITE(DMA_ADDR_0_HIGH, upper_32_bits(offset) & 0xFFFF);
+
+	/* Hardware doesn't look at destination address for HuC. Set it to 0,
+	 * but still program the correct address space.
+	 */
+	I915_WRITE(DMA_ADDR_1_LOW, 0);
+	I915_WRITE(DMA_ADDR_1_HIGH, DMA_ADDRESS_SPACE_WOPCM);
+
+	size = huc_fw->header_size + huc_fw->ucode_size;
+	I915_WRITE(DMA_COPY_SIZE, size);
+
+	/* Start the DMA */
+	I915_WRITE(DMA_CTRL, _MASKED_BIT_ENABLE(HUC_UKERNEL | START_DMA));
+
+	/* Wait for DMA to finish */
+	ret = wait_for((I915_READ(DMA_CTRL) & START_DMA) == 0, 100);
+
+	DRM_DEBUG_DRIVER("HuC DMA transfer wait over with ret %d\n", ret);
+
+	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.
+	 */
+	i915_vma_unpin(vma);
+
+	return ret;
+}
+
+/**
+ * intel_huc_init() - initiate HuC firmware loading request
+ * @dev: the drm device
+ *
+ * Called early during driver load, but after GEM is initialised. The loading
+ * will continue only when driver explicitly specify firmware name and version.
+ * All other cases are considered as UC_FIRMWARE_NONE either because HW is not
+ * capable or driver yet support it. And there will be no error message for
+ * UC_FIRMWARE_NONE cases.
+ *
+ * The DMA-copying to HW is done later when intel_huc_load() is called.
+ */
+void intel_huc_init(struct drm_i915_private *dev_priv)
+{
+	struct intel_huc *huc = &dev_priv->huc;
+	struct intel_uc_fw *huc_fw = &huc->huc_fw;
+	const char *fw_path = NULL;
+
+	huc_fw->uc_fw_path = NULL;
+	huc_fw->fetch_status = UC_FIRMWARE_NONE;
+	huc_fw->load_status = UC_FIRMWARE_NONE;
+	huc_fw->fw_type = UC_FW_TYPE_HUC;
+
+	if (!HAS_HUC_UCODE(dev_priv))
+		return;
+
+	if (IS_SKYLAKE(dev_priv)) {
+		fw_path = I915_SKL_HUC_UCODE;
+		huc_fw->major_ver_wanted = SKL_FW_MAJOR;
+		huc_fw->minor_ver_wanted = SKL_FW_MINOR;
+	}
+
+	huc_fw->uc_fw_path = fw_path;
+	huc_fw->fetch_status = UC_FIRMWARE_PENDING;
+
+	DRM_DEBUG_DRIVER("HuC firmware pending, path %s\n", fw_path);
+
+	intel_uc_fw_fetch(dev_priv, huc_fw);
+}
+
+/**
+ * intel_huc_load() - load HuC uCode to device
+ * @dev: the drm device
+ *
+ * Called from gem_init_hw() during driver loading and also after a GPU reset.
+ * Be note that HuC loading must be done before GuC loading.
+ *
+ * The firmware image should have already been fetched into memory by the
+ * earlier call to intel_huc_init(), so here we need only check that
+ * is succeeded, and then transfer the image to the h/w.
+ *
+ * Return:	non-zero code on error
+ */
+int intel_huc_load(struct drm_i915_private *dev_priv)
+{
+	struct intel_uc_fw *huc_fw = &dev_priv->huc.huc_fw;
+	int err;
+
+	if (huc_fw->fetch_status == UC_FIRMWARE_NONE)
+		return 0;
+
+	DRM_DEBUG_DRIVER("%s fw status: fetch %s, load %s\n",
+		huc_fw->uc_fw_path,
+		intel_uc_fw_status_repr(huc_fw->fetch_status),
+		intel_uc_fw_status_repr(huc_fw->load_status));
+
+	if (huc_fw->fetch_status == UC_FIRMWARE_SUCCESS &&
+	    huc_fw->load_status == UC_FIRMWARE_FAIL)
+		return -ENOEXEC;
+
+	huc_fw->load_status = UC_FIRMWARE_PENDING;
+
+	switch (huc_fw->fetch_status) {
+	case UC_FIRMWARE_FAIL:
+		/* something went wrong :( */
+		err = -EIO;
+		goto fail;
+
+	case UC_FIRMWARE_NONE:
+	case UC_FIRMWARE_PENDING:
+	default:
+		/* "can't happen" */
+		WARN_ONCE(1, "HuC fw %s invalid fetch_status %s [%d]\n",
+			huc_fw->uc_fw_path,
+			intel_uc_fw_status_repr(huc_fw->fetch_status),
+			huc_fw->fetch_status);
+		err = -ENXIO;
+		goto fail;
+
+	case UC_FIRMWARE_SUCCESS:
+		break;
+	}
+
+	err = huc_ucode_xfer(dev_priv);
+	if (err)
+		goto fail;
+
+	huc_fw->load_status = UC_FIRMWARE_SUCCESS;
+
+	DRM_DEBUG_DRIVER("%s fw status: fetch %s, load %s\n",
+		huc_fw->uc_fw_path,
+		intel_uc_fw_status_repr(huc_fw->fetch_status),
+		intel_uc_fw_status_repr(huc_fw->load_status));
+
+	return 0;
+
+fail:
+	if (huc_fw->load_status == UC_FIRMWARE_PENDING)
+		huc_fw->load_status = UC_FIRMWARE_FAIL;
+
+	DRM_ERROR("Failed to complete HuC uCode load with ret %d\n", err);
+
+	return err;
+}
+
+/**
+ * intel_huc_fini() - clean up resources allocated for HuC
+ * @dev: the drm device
+ *
+ * Cleans up by releasing the huc firmware GEM obj.
+ */
+void intel_huc_fini(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = to_i915(dev);
+	struct intel_uc_fw *huc_fw = &dev_priv->huc.huc_fw;
+
+	mutex_lock(&dev->struct_mutex);
+	if (huc_fw->uc_fw_obj)
+		i915_gem_object_put(huc_fw->uc_fw_obj);
+	huc_fw->uc_fw_obj = NULL;
+	mutex_unlock(&dev->struct_mutex);
+
+	huc_fw->fetch_status = UC_FIRMWARE_NONE;
+}
+
drivers/gpu/drm/i915/intel_uc.h +22 -0
index be89f0b..ac92946 100644
@@ -24,6 +24,12 @@
 #ifndef _INTEL_UC_H_
 #define _INTEL_UC_H_
 
+#ifndef _INTEL_HUC_H_
+#define _INTEL_HUC_H_
+
+#define HUC_STATUS2             _MMIO(0xD3B0)
+#define   HUC_FW_VERIFIED       (1<<7)
+
 #include "intel_guc_fwif.h"
 #include "i915_guc_reg.h"
 #include "intel_ringbuffer.h"
@@ -175,6 +181,13 @@ struct intel_guc {
 	struct mutex send_mutex;
 };
 
+struct intel_huc {
+	/* Generic uC firmware management */
+	struct intel_uc_fw huc_fw;
+
+	/* HuC-specific additions */
+};
+
 /* intel_uc.c */
 void intel_uc_init_early(struct drm_i915_private *dev_priv);
 bool intel_guc_recv(struct drm_i915_private *dev_priv, u32 *status);
@@ -191,6 +204,9 @@ extern void intel_guc_fini(struct drm_i915_private *dev_priv);
 extern const char *intel_uc_fw_status_repr(enum intel_uc_fw_status status);
 extern int intel_guc_suspend(struct drm_i915_private *dev_priv);
 extern int intel_guc_resume(struct drm_i915_private *dev_priv);
+void intel_uc_fw_fetch(struct drm_i915_private *dev_priv,
+	struct intel_uc_fw *uc_fw);
+u32 guc_wopcm_size(struct drm_i915_private *dev_priv);
 
 /* i915_guc_submission.c */
 int i915_guc_submission_init(struct drm_i915_private *dev_priv);
@@ -205,4 +221,10 @@ void i915_guc_register(struct drm_i915_private *dev_priv);
 void i915_guc_unregister(struct drm_i915_private *dev_priv);
 int i915_guc_log_control(struct drm_i915_private *dev_priv, u64 control_val);
 
+/* intel_huc_loader.c */
+void intel_huc_init(struct drm_i915_private *dev_priv);
+void intel_huc_fini(struct drm_device *dev);
+int intel_huc_load(struct drm_i915_private *dev_priv);
+
+#endif
 #endif
Signature · 6 lines
2.7.4

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Signature · 6 lines
Cheers,
Arek
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
#15 lore raw

Re: [PATCH 3/8] drm/i915/huc: Add HuC fw loading support

Quoted 11 lines · Arkadiusz Hiler

drivers/gpu/drm/i915/intel_guc_loader.c int intel_guc_setup(struct drm_i915_private *dev_priv) · line 514

Quoted 10 lines · anushasr
drivers/gpu/drm/i915/intel_guc_loader.c +2 -1
-static u32 guc_wopcm_size(struct drm_i915_private *dev_priv)
+u32 guc_wopcm_size(struct drm_i915_private *dev_priv)
 {
 	u32 wopcm_size = GUC_WOPCM_TOP;
@@ -511,6 +511,7 @@ int intel_guc_setup(struct drm_i915_private *dev_priv)
 		if (err)
 			goto fail;
 
+		intel_huc_load(dev_priv);

We don't need error handling? That would simplify a lot of our code! -Chris

Signature · 5 lines
Chris Wilson, Intel Open Source Technology Centre
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
#16 lore raw

Re: [PATCH 3/8] drm/i915/huc: Add HuC fw loading support

Quoted 7 lines · Chris Wilson
Quoted 6 lines · Arkadiusz Hiler
Quoted 5 lines · anushasr
drivers/gpu/drm/i915/intel_guc_loader.c +1 -1
-static u32 guc_wopcm_size(struct drm_i915_private *dev_priv)
+u32 guc_wopcm_size(struct drm_i915_private *dev_priv)
 {
 	u32 wopcm_size = GUC_WOPCM_TOP;
Quoted 8 lines
Quoted 5 lines

drivers/gpu/drm/i915/intel_guc_loader.c int intel_guc_setup(struct drm_i915_private *dev_priv) · line 514

Quoted 5 lines
drivers/gpu/drm/i915/intel_guc_loader.c +1 -0
@@ -511,6 +511,7 @@ int intel_guc_setup(struct drm_i915_private *dev_priv)
 		if (err)
 			goto fail;
 
+		intel_huc_load(dev_priv);

We don't need error handling? That would simplify a lot of our code! -Chris

With this patch series on this specific piece of code - not really. HuC support it intorduce is _best-eforrty_.

If the function would report error we would not act on it in anyway other than logging the fail (which the function already does for us).

As Anusha discussed here, there will be some code reorganization due to introduction of i915.enable_huc and deprecation of enable_guc_loading. Once we want to have enable_huc=2, there is a reason to change the signature and report errors.

Signature · 6 lines
Cheers,
Arek
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
#17 lore raw

Re: [PATCH 3/8] drm/i915/huc: Add HuC fw loading support

Quoted 20 lines · Arkadiusz Hiler
Quoted 7 lines · Chris Wilson
Quoted 6 lines · Arkadiusz Hiler
Quoted 5 lines · anushasr
drivers/gpu/drm/i915/intel_guc_loader.c +1 -1
-static u32 guc_wopcm_size(struct drm_i915_private *dev_priv)
+u32 guc_wopcm_size(struct drm_i915_private *dev_priv)
 {
 	u32 wopcm_size = GUC_WOPCM_TOP;
Quoted 8 lines
Quoted 5 lines

drivers/gpu/drm/i915/intel_guc_loader.c int intel_guc_setup(struct drm_i915_private *dev_priv) · line 514

Quoted 5 lines
drivers/gpu/drm/i915/intel_guc_loader.c +1 -0
@@ -511,6 +511,7 @@ int intel_guc_setup(struct drm_i915_private *dev_priv)
 		if (err)
 			goto fail;
 
+		intel_huc_load(dev_priv);

We don't need error handling? That would simplify a lot of our code! -Chris

With this patch series on this specific piece of code - not really. HuC support it intorduce is _best-eforrty_.

Quoted 7 lines

If the function would report error we would not act on it in anyway other than logging the fail (which the function already does for us).

As Anusha discussed here, there will be some code reorganization due to introduction of i915.enable_huc and deprecation of enable_guc_loading. Once we want to have enable_huc=2, there is a reason to change the signature and report errors.

Ergh, I've got signatures of auth and load mixed up. But most of the statement above still semi-holds.

Signature · 6 lines
Cheers,
Arek
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
#18 lore raw

Re: [PATCH 3/8] drm/i915/huc: Add HuC fw loading support

In reply to anushasr
Quoted 202 lines · anushasr

From: Anusha Srivatsa <[email protected]>

The HuC loading process is similar to GuC. The intel_uc_fw_fetch() is used for both cases.

HuC loading needs to be before GuC loading. The WOPCM setting must be done early before loading any of them.

v2: rebased on-top of drm-intel-nightly.
removed if(HAS_GUC()) before the guc call. (D.Gordon) update huc_version number of format.
v3: rebased to drm-intel-nightly, changed the file name format to
match the one in the huc package.
Changed dev->dev_private to to_i915()
v4: moved function back to where it was.
change wait_for_atomic to wait_for.
v5: rebased + comment changes.
v7: rebased.
v8: rebased.
v9: rebased. Changed the year in the copyright message to reflect the right year.Correct the comments,remove the unwanted WARN message, replace drm_gem_object_unreference() with i915_gem_object_put().Make the prototypes in intel_huc.h non-extern.
v10: rebased. Update the file construction done by HuC. It is similar to GuC.Adopted the approach used in-
https://patchwork.freedesktop.org/patch/104355/ <Tvrtko Ursulin>
v11: Fix warnings remove old declaration
v12: Change dev to dev_priv in macro definition.
Corrected comments.
v13: rebased.
v14: rebased on top of drm-tip
v15: rebased. Updated functions intel_huc_load(),intel_huc_init() and intel_uc_fw_fetch() to accept dev_priv instead of dev. Moved contents of intel_huc.h to intel_uc.h

---
 drivers/gpu/drm/i915/Makefile           |   1 +
 drivers/gpu/drm/i915/i915_drv.c         |   4 +-
 drivers/gpu/drm/i915/i915_drv.h         |   3 +-
 drivers/gpu/drm/i915/i915_guc_reg.h     |   3 +
 drivers/gpu/drm/i915/intel_guc_loader.c |   7 +-
 drivers/gpu/drm/i915/intel_huc_loader.c | 264 ++++++++++++++++++++++++++++++++
 drivers/gpu/drm/i915/intel_uc.h         |  22 +++
 7 files changed, 299 insertions(+), 5 deletions(-)
 create mode 100644 drivers/gpu/drm/i915/intel_huc_loader.c
drivers/gpu/drm/i915/Makefile +1 -0
index 3c30916..01d4f4b 100644
@@ -57,6 +57,7 @@ i915-y += i915_cmd_parser.o \
 # general-purpose microcontroller (GuC) support
 i915-y += intel_uc.o \
 	  intel_guc_loader.o \
+	  intel_huc_loader.o \
 	  i915_guc_submission.o
 
 # autogenerated null render state
drivers/gpu/drm/i915/i915_drv.c +3 -1
index 6428588..85a47c2 100644
@@ -600,6 +600,7 @@ static int i915_load_modeset_init(struct drm_device *dev)
 	if (ret)
 		goto cleanup_irq;
 
+	intel_huc_init(dev_priv);
 	intel_guc_init(dev_priv);
 
 	ret = i915_gem_init(dev_priv);
@@ -627,6 +628,7 @@ static int i915_load_modeset_init(struct drm_device *dev)
 		DRM_ERROR("failed to idle hardware; continuing to unload!\n");
 	i915_gem_fini(dev_priv);
 cleanup_irq:
+	intel_huc_fini(dev);
 	intel_guc_fini(dev_priv);
 	drm_irq_uninstall(dev);
 	intel_teardown_gmbus(dev_priv);
@@ -1313,7 +1315,7 @@ void i915_driver_unload(struct drm_device *dev)
 
 	/* Flush any outstanding unpin_work. */
 	drain_workqueue(dev_priv->wq);
-
+	intel_huc_fini(dev);
 	intel_guc_fini(dev_priv);
 	i915_gem_fini(dev_priv);
 	intel_fbc_cleanup_cfb(dev_priv);
drivers/gpu/drm/i915/i915_drv.h +2 -1
index 1480e73..0371ca4 100644
@@ -2036,6 +2036,7 @@ struct drm_i915_private {
 
 	struct intel_gvt *gvt;
 
+	struct intel_huc huc;
 	struct intel_guc guc;
 
 	struct intel_csr csr;
@@ -2810,7 +2811,7 @@ intel_info(const struct drm_i915_private *dev_priv)
 #define HAS_GUC(dev_priv)	((dev_priv)->info.has_guc)
 #define HAS_GUC_UCODE(dev_priv)	(HAS_GUC(dev_priv))
 #define HAS_GUC_SCHED(dev_priv)	(HAS_GUC(dev_priv))
-
+#define HAS_HUC_UCODE(dev_priv)	(HAS_GUC(dev_priv))
 #define HAS_RESOURCE_STREAMER(dev_priv) ((dev_priv)->info.has_resource_streamer)
 
 #define HAS_POOLED_EU(dev_priv)	((dev_priv)->info.has_pooled_eu)
drivers/gpu/drm/i915/i915_guc_reg.h +3 -0
index 5e638fc..f9829f6 100644
@@ -61,9 +61,12 @@
 #define   DMA_ADDRESS_SPACE_GTT		  (8 << 16)
 #define DMA_COPY_SIZE			_MMIO(0xc310)
 #define DMA_CTRL			_MMIO(0xc314)
+#define   HUC_UKERNEL			  (1<<9)
 #define   UOS_MOVE			  (1<<4)
 #define   START_DMA			  (1<<0)
 #define DMA_GUC_WOPCM_OFFSET		_MMIO(0xc340)
+#define   HUC_LOADING_AGENT_VCR		  (0<<1)
+#define   HUC_LOADING_AGENT_GUC		  (1<<1)
 #define   GUC_WOPCM_OFFSET_VALUE	  0x80000	/* 512KB */
 #define GUC_MAX_IDLE_COUNT		_MMIO(0xC3E4)
 
drivers/gpu/drm/i915/intel_guc_loader.c +4 -3
index 26a184f..b971351 100644
@@ -309,8 +309,8 @@ static int guc_ucode_xfer_dma(struct drm_i915_private *dev_priv,
 	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));
-
+	I915_WRITE(DMA_CTRL, _MASKED_BIT_ENABLE(UOS_MOVE | START_DMA) |
+		_MASKED_BIT_DISABLE(HUC_UKERNEL));
 	/*
 	 * Wait for the DMA to complete & the GuC to start up.
 	 * NB: Docs recommend not using the interrupt for completion.
@@ -334,7 +334,7 @@ static int guc_ucode_xfer_dma(struct drm_i915_private *dev_priv,
 	return ret;
 }
 
-static u32 guc_wopcm_size(struct drm_i915_private *dev_priv)
+u32 guc_wopcm_size(struct drm_i915_private *dev_priv)
 {
 	u32 wopcm_size = GUC_WOPCM_TOP;
 
@@ -511,6 +511,7 @@ int intel_guc_setup(struct drm_i915_private *dev_priv)
 		if (err)
 			goto fail;
 
+		intel_huc_load(dev_priv);
 		err = guc_ucode_xfer(dev_priv);
 		if (!err)
 			break;
drivers/gpu/drm/i915/intel_huc_loader.c new file +34 -0
new file mode 100644
index 0000000..e0efd1c
@@ -0,0 +1,264 @@
+/*
+ * Copyright © 2016 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 <linux/firmware.h>
+#include "i915_drv.h"
+#include "intel_uc.h"
+
+/**
+ * DOC: HuC Firmware
+ *
+ * Motivation:
+ * GEN9 introduces a new dedicated firmware for usage in media HEVC (High
+ * Efficiency Video Coding) operations. Userspace can use the firmware
+ * capabilities by adding HuC specific commands to batch buffers.

drivers/gpu/drm/i915/intel_huc_loader.c line 44

Quoted 10 lines
drivers/gpu/drm/i915/intel_huc_loader.c new file +10 -0
+ *
+ * Implementation:
+ * The same firmware loader is used as the GuC. However, the actual
+ * loading to HW is deferred until GEM initialization is done.
+ *
+ * Note that HuC firmware loading must be done before GuC loading.
+ */
+
+#define SKL_FW_MAJOR 01
+#define SKL_FW_MINOR 07

Can we use SKL_HUC_FW_ prefix to distinguish these macros from similar defined in intel_guc_loader.c ?

Quoted 78 lines
drivers/gpu/drm/i915/intel_huc_loader.c new file +78 -0
+#define SKL_BLD_NUM 1398
+
+#define HUC_FW_PATH(platform, major, minor, bld_num) \
+	"i915/" __stringify(platform) "_huc_ver" __stringify(major) "_" \
+	__stringify(minor) "_" __stringify(bld_num) ".bin"
+
+#define I915_SKL_HUC_UCODE HUC_FW_PATH(skl, SKL_FW_MAJOR, \
+	SKL_FW_MINOR, SKL_BLD_NUM)
+MODULE_FIRMWARE(I915_SKL_HUC_UCODE);
+
+/**
+ * huc_ucode_xfer() - DMA's the firmware
+ * @dev_priv: the drm device
+ *
+ * This function takes the gem object containing the firmware, sets up the DMA
+ * engine MMIO, triggers the DMA operation and waits for it to finish.
+ *
+ * Transfer the firmware image to RAM for execution by the microcontroller.
+ *
+ * Return: 0 on success, non-zero on failure
+ */
+
+static int huc_ucode_xfer(struct drm_i915_private *dev_priv)
+{
+	struct intel_uc_fw *huc_fw = &dev_priv->huc.huc_fw;
+	struct i915_vma *vma;
+	unsigned long offset = 0;
+	u32 size;
+	int ret;
+
+	ret = i915_gem_object_set_to_gtt_domain(huc_fw->uc_fw_obj, false);
+	if (ret) {
+		DRM_DEBUG_DRIVER("set-domain failed %d\n", ret);
+		return ret;
+	}
+
+	vma = i915_gem_object_ggtt_pin(huc_fw->uc_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);
+	}
+
+	/* 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 */
+	I915_WRITE(GUC_WOPCM_SIZE, guc_wopcm_size(dev_priv));
+	I915_WRITE(DMA_GUC_WOPCM_OFFSET, GUC_WOPCM_OFFSET_VALUE |
+			HUC_LOADING_AGENT_GUC);
+
+	/* Set the source address for the uCode */
+	offset = i915_ggtt_offset(vma) + huc_fw->header_offset;
+	I915_WRITE(DMA_ADDR_0_LOW, lower_32_bits(offset));
+	I915_WRITE(DMA_ADDR_0_HIGH, upper_32_bits(offset) & 0xFFFF);
+
+	/* Hardware doesn't look at destination address for HuC. Set it to 0,
+	 * but still program the correct address space.
+	 */
+	I915_WRITE(DMA_ADDR_1_LOW, 0);
+	I915_WRITE(DMA_ADDR_1_HIGH, DMA_ADDRESS_SPACE_WOPCM);
+
+	size = huc_fw->header_size + huc_fw->ucode_size;
+	I915_WRITE(DMA_COPY_SIZE, size);
+
+	/* Start the DMA */
+	I915_WRITE(DMA_CTRL, _MASKED_BIT_ENABLE(HUC_UKERNEL | START_DMA));
+
+	/* Wait for DMA to finish */
+	ret = wait_for((I915_READ(DMA_CTRL) & START_DMA) == 0, 100);
+
+	DRM_DEBUG_DRIVER("HuC DMA transfer wait over with ret %d\n", ret);
+
+	intel_uncore_forcewake_put(dev_priv, FORCEWAKE_ALL);
+
+	/*
+	 * We keep the object pages for reuse during resume. But we can unpin it

drivers/gpu/drm/i915/intel_huc_loader.c line 132

Quoted 10 lines
drivers/gpu/drm/i915/intel_huc_loader.c new file +10 -0
+	 * now that DMA has completed, so it doesn't continue to take up space.
+	 */
+	i915_vma_unpin(vma);
+
+	return ret;
+}
+
+/**
+ * intel_huc_init() - initiate HuC firmware loading request
+ * @dev: the drm device

Mismatched param name.

drivers/gpu/drm/i915/intel_huc_loader.c line 144

Quoted 12 lines
drivers/gpu/drm/i915/intel_huc_loader.c new file +12 -0
+ *
+ * Called early during driver load, but after GEM is initialised. The loading
+ * will continue only when driver explicitly specify firmware name and version.
+ * All other cases are considered as UC_FIRMWARE_NONE either because HW is not
+ * capable or driver yet support it. And there will be no error message for
+ * UC_FIRMWARE_NONE cases.
+ *
+ * The DMA-copying to HW is done later when intel_huc_load() is called.
+ */
+void intel_huc_init(struct drm_i915_private *dev_priv)
+{
+	struct intel_huc *huc = &dev_priv->huc;

Maybe *huc shall be passed as only param (to match intel_huc function name). Then dev_priv could be recreated from huc_to_i915().

Quoted 18 lines
drivers/gpu/drm/i915/intel_huc_loader.c new file +18 -0
+	struct intel_uc_fw *huc_fw = &huc->huc_fw;
+	const char *fw_path = NULL;
+
+	huc_fw->uc_fw_path = NULL;
+	huc_fw->fetch_status = UC_FIRMWARE_NONE;
+	huc_fw->load_status = UC_FIRMWARE_NONE;
+	huc_fw->fw_type = UC_FW_TYPE_HUC;
+
+	if (!HAS_HUC_UCODE(dev_priv))
+		return;
+
+	if (IS_SKYLAKE(dev_priv)) {
+		fw_path = I915_SKL_HUC_UCODE;
+		huc_fw->major_ver_wanted = SKL_FW_MAJOR;
+		huc_fw->minor_ver_wanted = SKL_FW_MINOR;
+	}
+
+	huc_fw->uc_fw_path = fw_path;

drivers/gpu/drm/i915/intel_huc_loader.c line 172

Quoted 10 lines
drivers/gpu/drm/i915/intel_huc_loader.c new file +10 -0
+	huc_fw->fetch_status = UC_FIRMWARE_PENDING;
+
+	DRM_DEBUG_DRIVER("HuC firmware pending, path %s\n", fw_path);
+
+	intel_uc_fw_fetch(dev_priv, huc_fw);
+}
+
+/**
+ * intel_huc_load() - load HuC uCode to device
+ * @dev: the drm device

Mismatched param name.

Quoted 97 lines
drivers/gpu/drm/i915/intel_huc_loader.c new file +92 -0
+ *
+ * Called from gem_init_hw() during driver loading and also after a GPU reset.
+ * Be note that HuC loading must be done before GuC loading.
+ *
+ * The firmware image should have already been fetched into memory by the
+ * earlier call to intel_huc_init(), so here we need only check that
+ * is succeeded, and then transfer the image to the h/w.
+ *
+ * Return:	non-zero code on error
+ */
+int intel_huc_load(struct drm_i915_private *dev_priv)
+{
+	struct intel_uc_fw *huc_fw = &dev_priv->huc.huc_fw;
+	int err;
+
+	if (huc_fw->fetch_status == UC_FIRMWARE_NONE)
+		return 0;
+
+	DRM_DEBUG_DRIVER("%s fw status: fetch %s, load %s\n",
+		huc_fw->uc_fw_path,
+		intel_uc_fw_status_repr(huc_fw->fetch_status),
+		intel_uc_fw_status_repr(huc_fw->load_status));
+
+	if (huc_fw->fetch_status == UC_FIRMWARE_SUCCESS &&
+	    huc_fw->load_status == UC_FIRMWARE_FAIL)
+		return -ENOEXEC;
+
+	huc_fw->load_status = UC_FIRMWARE_PENDING;
+
+	switch (huc_fw->fetch_status) {
+	case UC_FIRMWARE_FAIL:
+		/* something went wrong :( */
+		err = -EIO;
+		goto fail;
+
+	case UC_FIRMWARE_NONE:
+	case UC_FIRMWARE_PENDING:
+	default:
+		/* "can't happen" */
+		WARN_ONCE(1, "HuC fw %s invalid fetch_status %s [%d]\n",
+			huc_fw->uc_fw_path,
+			intel_uc_fw_status_repr(huc_fw->fetch_status),
+			huc_fw->fetch_status);
+		err = -ENXIO;
+		goto fail;
+
+	case UC_FIRMWARE_SUCCESS:
+		break;
+	}
+
+	err = huc_ucode_xfer(dev_priv);
+	if (err)
+		goto fail;
+
+	huc_fw->load_status = UC_FIRMWARE_SUCCESS;
+
+	DRM_DEBUG_DRIVER("%s fw status: fetch %s, load %s\n",
+		huc_fw->uc_fw_path,
+		intel_uc_fw_status_repr(huc_fw->fetch_status),
+		intel_uc_fw_status_repr(huc_fw->load_status));
+
+	return 0;
+
+fail:
+	if (huc_fw->load_status == UC_FIRMWARE_PENDING)
+		huc_fw->load_status = UC_FIRMWARE_FAIL;
+
+	DRM_ERROR("Failed to complete HuC uCode load with ret %d\n", err);
+
+	return err;
+}
+
+/**
+ * intel_huc_fini() - clean up resources allocated for HuC
+ * @dev: the drm device
+ *
+ * Cleans up by releasing the huc firmware GEM obj.
+ */
+void intel_huc_fini(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = to_i915(dev);
+	struct intel_uc_fw *huc_fw = &dev_priv->huc.huc_fw;
+
+	mutex_lock(&dev->struct_mutex);
+	if (huc_fw->uc_fw_obj)
+		i915_gem_object_put(huc_fw->uc_fw_obj);
+	huc_fw->uc_fw_obj = NULL;
+	mutex_unlock(&dev->struct_mutex);
+
+	huc_fw->fetch_status = UC_FIRMWARE_NONE;
+}
+
drivers/gpu/drm/i915/intel_uc.h +0 -0
index be89f0b..ac92946 100644

drivers/gpu/drm/i915/intel_uc.h line 28

Quoted 5 lines
drivers/gpu/drm/i915/intel_uc.h +2 -0
@@ -24,6 +24,12 @@
 #ifndef _INTEL_UC_H_
 #define _INTEL_UC_H_
 
+#ifndef _INTEL_HUC_H_
+#define _INTEL_HUC_H_

Typo ? This is still intel_uc.h file, right ?

drivers/gpu/drm/i915/intel_uc.h line 32

Quoted 4 lines
drivers/gpu/drm/i915/intel_uc.h +4 -0
+
+#define HUC_STATUS2             _MMIO(0xD3B0)
+#define   HUC_FW_VERIFIED       (1<<7)
+

Is it correct place for these defs?
What about i915_guc_reg.h ?

Quoted 18 lines
drivers/gpu/drm/i915/intel_uc.h +7 -0
 #include "intel_guc_fwif.h"
 #include "i915_guc_reg.h"
 #include "intel_ringbuffer.h"
@@ -175,6 +181,13 @@ struct intel_guc {
 	struct mutex send_mutex;
 };
 
+struct intel_huc {
+	/* Generic uC firmware management */
+	struct intel_uc_fw huc_fw;
+
+	/* HuC-specific additions */
+};
+
 /* intel_uc.c */
 void intel_uc_init_early(struct drm_i915_private *dev_priv);
 bool intel_guc_recv(struct drm_i915_private *dev_priv, u32 *status);

drivers/gpu/drm/i915/intel_uc.h extern void intel_guc_fini(struct drm_i915_private *dev_priv); · line 209

Quoted 6 lines
drivers/gpu/drm/i915/intel_uc.h +3 -0
@@ -191,6 +204,9 @@ extern void intel_guc_fini(struct drm_i915_private *dev_priv);
 extern const char *intel_uc_fw_status_repr(enum intel_uc_fw_status status);
 extern int intel_guc_suspend(struct drm_i915_private *dev_priv);
 extern int intel_guc_resume(struct drm_i915_private *dev_priv);
+void intel_uc_fw_fetch(struct drm_i915_private *dev_priv,
+	struct intel_uc_fw *uc_fw);
+u32 guc_wopcm_size(struct drm_i915_private *dev_priv);

All other public functions have intel_ prefix.

Quoted 21 lines
drivers/gpu/drm/i915/intel_uc.h +6 -0
 /* i915_guc_submission.c */
 int i915_guc_submission_init(struct drm_i915_private *dev_priv);
@@ -205,4 +221,10 @@ void i915_guc_register(struct drm_i915_private *dev_priv);
 void i915_guc_unregister(struct drm_i915_private *dev_priv);
 int i915_guc_log_control(struct drm_i915_private *dev_priv, u64 control_val);
 
+/* intel_huc_loader.c */
+void intel_huc_init(struct drm_i915_private *dev_priv);
+void intel_huc_fini(struct drm_device *dev);
+int intel_huc_load(struct drm_i915_private *dev_priv);
+
+#endif
 #endif
Signature · 6 lines
2.7.4

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

#19 lore raw

Re: [PATCH 3/8] drm/i915/huc: Add HuC fw loading support

Quoted 245 lines

-----Original Message-----
From: Michal Wajdeczko [mailto:[email protected]]
Sent: Friday, December 9, 2016 4:18 AM
To: Srivatsa, Anusha <[email protected]>
Cc: [email protected]; Alex Dai <[email protected]>; Peter Antoine <[email protected]>
Subject: Re: [Intel-gfx] [PATCH 3/8] drm/i915/huc: Add HuC fw loading support

Quoted 40 lines · anushasr

From: Anusha Srivatsa <[email protected]>

The HuC loading process is similar to GuC. The intel_uc_fw_fetch() is used for both cases.

HuC loading needs to be before GuC loading. The WOPCM setting must be done early before loading any of them.

v2: rebased on-top of drm-intel-nightly.
removed if(HAS_GUC()) before the guc call. (D.Gordon) update huc_version number of format.
v3: rebased to drm-intel-nightly, changed the file name format to
match the one in the huc package.
Changed dev->dev_private to to_i915()
v4: moved function back to where it was.
change wait_for_atomic to wait_for.
v5: rebased + comment changes.
v7: rebased.
v8: rebased.
v9: rebased. Changed the year in the copyright message to reflect the right year.Correct the comments,remove the unwanted WARN message, replace drm_gem_object_unreference() with i915_gem_object_put().Make the prototypes in intel_huc.h non-extern.
v10: rebased. Update the file construction done by HuC. It is similar to GuC.Adopted the approach used in-
https://patchwork.freedesktop.org/patch/104355/ <Tvrtko Ursulin>
v11: Fix warnings remove old declaration
v12: Change dev to dev_priv in macro definition.
Corrected comments.
v13: rebased.
v14: rebased on top of drm-tip
v15: rebased. Updated functions intel_huc_load(),intel_huc_init() and intel_uc_fw_fetch() to accept dev_priv instead of dev. Moved contents of intel_huc.h to intel_uc.h

Quoted 7 lines
---
 drivers/gpu/drm/i915/Makefile           |   1 +
 drivers/gpu/drm/i915/i915_drv.c         |   4 +-
 drivers/gpu/drm/i915/i915_drv.h         |   3 +-
 drivers/gpu/drm/i915/i915_guc_reg.h     |   3 +
 drivers/gpu/drm/i915/intel_guc_loader.c |   7 +-
 drivers/gpu/drm/i915/intel_huc_loader.c | 264

++++++++++++++++++++++++++++++++

drivers/gpu/drm/i915/i915_drv.c static int i915_load_modeset_init(struct drm_device

Quoted 20 lines
 drivers/gpu/drm/i915/intel_uc.h         |  22 +++
 7 files changed, 299 insertions(+), 5 deletions(-)  create mode

100644 drivers/gpu/drm/i915/intel_huc_loader.c

(quoted diff) +0 -0

b/drivers/gpu/drm/i915/Makefile index 3c30916..01d4f4b 100644

drivers/gpu/drm/i915/Makefile +1 -0
@@ -57,6 +57,7 @@ i915-y += i915_cmd_parser.o \  # general-purpose
 microcontroller (GuC) support  i915-y += intel_uc.o \
 	  intel_guc_loader.o \
+	  intel_huc_loader.o \
 	  i915_guc_submission.o
 
 # autogenerated null render state
(quoted diff) +0 -0

b/drivers/gpu/drm/i915/i915_drv.c index 6428588..85a47c2 100644

drivers/gpu/drm/i915/i915_drv.c +0 -0
@@ -600,6 +600,7 @@ static int i915_load_modeset_init(struct drm_device

*dev)

drivers/gpu/drm/i915/i915_drv.c static int i915_load_modeset_init(struct drm_device

Quoted 8 lines
drivers/gpu/drm/i915/i915_drv.c +1 -0
 	if (ret)
 		goto cleanup_irq;
 
+	intel_huc_init(dev_priv);
 	intel_guc_init(dev_priv);
 
 	ret = i915_gem_init(dev_priv);
@@ -627,6 +628,7 @@ static int i915_load_modeset_init(struct drm_device

*dev)

drivers/gpu/drm/i915/intel_guc_loader.c static int guc_ucode_xfer_dma(struct drm_i915_private

Quoted 62 lines
drivers/gpu/drm/i915/i915_drv.c +2 -1
 		DRM_ERROR("failed to idle hardware; continuing to unload!\n");
 	i915_gem_fini(dev_priv);
 cleanup_irq:
+	intel_huc_fini(dev);
 	intel_guc_fini(dev_priv);
 	drm_irq_uninstall(dev);
 	intel_teardown_gmbus(dev_priv);
@@ -1313,7 +1315,7 @@ void i915_driver_unload(struct drm_device *dev)
 
 	/* Flush any outstanding unpin_work. */
 	drain_workqueue(dev_priv->wq);
-
+	intel_huc_fini(dev);
 	intel_guc_fini(dev_priv);
 	i915_gem_fini(dev_priv);
 	intel_fbc_cleanup_cfb(dev_priv);
(quoted diff) +0 -0

b/drivers/gpu/drm/i915/i915_drv.h index 1480e73..0371ca4 100644

drivers/gpu/drm/i915/i915_drv.h +2 -1
@@ -2036,6 +2036,7 @@ struct drm_i915_private {
 
 	struct intel_gvt *gvt;
 
+	struct intel_huc huc;
 	struct intel_guc guc;
 
 	struct intel_csr csr;
@@ -2810,7 +2811,7 @@ intel_info(const struct drm_i915_private *dev_priv)
 #define HAS_GUC(dev_priv)	((dev_priv)->info.has_guc)
 #define HAS_GUC_UCODE(dev_priv)	(HAS_GUC(dev_priv))
 #define HAS_GUC_SCHED(dev_priv)	(HAS_GUC(dev_priv))
-
+#define HAS_HUC_UCODE(dev_priv)	(HAS_GUC(dev_priv))
 #define HAS_RESOURCE_STREAMER(dev_priv)
 ((dev_priv)->info.has_resource_streamer)
 

#define HAS_POOLED_EU(dev_priv) ((dev_priv)->info.has_pooled_eu)

drivers/gpu/drm/i915/i915_guc_reg.h +3 -0
index 5e638fc..f9829f6 100644
@@ -61,9 +61,12 @@
 #define   DMA_ADDRESS_SPACE_GTT		  (8 << 16)
 #define DMA_COPY_SIZE			_MMIO(0xc310)
 #define DMA_CTRL			_MMIO(0xc314)
+#define   HUC_UKERNEL			  (1<<9)
 #define   UOS_MOVE			  (1<<4)
 #define   START_DMA			  (1<<0)
 #define DMA_GUC_WOPCM_OFFSET		_MMIO(0xc340)
+#define   HUC_LOADING_AGENT_VCR		  (0<<1)
+#define   HUC_LOADING_AGENT_GUC		  (1<<1)
 #define   GUC_WOPCM_OFFSET_VALUE	  0x80000	/* 512KB */
 #define GUC_MAX_IDLE_COUNT		_MMIO(0xC3E4)
 
drivers/gpu/drm/i915/intel_guc_loader.c +0 -0
index 26a184f..b971351 100644
@@ -309,8 +309,8 @@ static int guc_ucode_xfer_dma(struct drm_i915_private

*dev_priv,

drivers/gpu/drm/i915/intel_guc_loader.c static int guc_ucode_xfer_dma(struct drm_i915_private *dev_priv,

Quoted 4 lines
drivers/gpu/drm/i915/intel_guc_loader.c +0 -1
 	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));

drivers/gpu/drm/i915/intel_guc_loader.c static int guc_ucode_xfer_dma(struct drm_i915_private · line 312

Quoted 2 lines
drivers/gpu/drm/i915/intel_guc_loader.c +1 -1
-
+	I915_WRITE(DMA_CTRL, _MASKED_BIT_ENABLE(UOS_MOVE |

START_DMA) |

drivers/gpu/drm/i915/intel_guc_loader.c static int guc_ucode_xfer_dma(struct drm_i915_private

Quoted 5 lines
drivers/gpu/drm/i915/intel_guc_loader.c +1 -0
+		_MASKED_BIT_DISABLE(HUC_UKERNEL));
 	/*
 	 * Wait for the DMA to complete & the GuC to start up.
 	 * NB: Docs recommend not using the interrupt for completion.
@@ -334,7 +334,7 @@ static int guc_ucode_xfer_dma(struct drm_i915_private

*dev_priv,

Quoted 39 lines
drivers/gpu/drm/i915/intel_guc_loader.c +2 -1
 	return ret;
 }
 
-static u32 guc_wopcm_size(struct drm_i915_private *dev_priv)
+u32 guc_wopcm_size(struct drm_i915_private *dev_priv)
 {
 	u32 wopcm_size = GUC_WOPCM_TOP;
 
@@ -511,6 +511,7 @@ int intel_guc_setup(struct drm_i915_private *dev_priv)
 		if (err)
 			goto fail;
 
+		intel_huc_load(dev_priv);
 		err = guc_ucode_xfer(dev_priv);
 		if (!err)
 			break;
drivers/gpu/drm/i915/intel_huc_loader.c new file +16 -0
new file mode 100644
index 0000000..e0efd1c
@@ -0,0 +1,264 @@
+/*
+ * Copyright (c) 2016 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

drivers/gpu/drm/i915/intel_huc_loader.c line 26

Quoted 10 lines
drivers/gpu/drm/i915/intel_huc_loader.c new file +10 -0
+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

drivers/gpu/drm/i915/intel_huc_loader.c line 28

Quoted 2 lines
drivers/gpu/drm/i915/intel_huc_loader.c new file +2 -0
+SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM,

DAMAGES

Quoted 26 lines
drivers/gpu/drm/i915/intel_huc_loader.c new file +26 -0
+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 <linux/firmware.h>
+#include "i915_drv.h"
+#include "intel_uc.h"
+
+/**
+ * DOC: HuC Firmware
+ *
+ * Motivation:
+ * GEN9 introduces a new dedicated firmware for usage in media HEVC
+(High
+ * Efficiency Video Coding) operations. Userspace can use the
+firmware
+ * capabilities by adding HuC specific commands to batch buffers.
+ *
+ * Implementation:
+ * The same firmware loader is used as the GuC. However, the actual
+ * loading to HW is deferred until GEM initialization is done.
+ *
Quoted 8 lines

drivers/gpu/drm/i915/intel_huc_loader.c line 44

Quoted 5 lines
drivers/gpu/drm/i915/intel_huc_loader.c new file +5 -0
+ * Note that HuC firmware loading must be done before GuC loading.
+ */
+
+#define SKL_FW_MAJOR 01
+#define SKL_FW_MINOR 07

Can we use SKL_HUC_FW_ prefix to distinguish these macros from similar defined in intel_guc_loader.c ?

Sure

Quoted 102 lines
Quoted 40 lines
drivers/gpu/drm/i915/intel_huc_loader.c new file +39 -0
+#define SKL_BLD_NUM 1398
+
+#define HUC_FW_PATH(platform, major, minor, bld_num) \
+	"i915/" __stringify(platform) "_huc_ver" __stringify(major) "_" \
+	__stringify(minor) "_" __stringify(bld_num) ".bin"
+
+#define I915_SKL_HUC_UCODE HUC_FW_PATH(skl, SKL_FW_MAJOR, \
+	SKL_FW_MINOR, SKL_BLD_NUM)
+MODULE_FIRMWARE(I915_SKL_HUC_UCODE);
+
+/**
+ * huc_ucode_xfer() - DMA's the firmware
+ * @dev_priv: the drm device
+ *
+ * This function takes the gem object containing the firmware, sets
+up the DMA
+ * engine MMIO, triggers the DMA operation and waits for it to finish.
+ *
+ * Transfer the firmware image to RAM for execution by the microcontroller.
+ *
+ * Return: 0 on success, non-zero on failure  */
+
+static int huc_ucode_xfer(struct drm_i915_private *dev_priv) {
+	struct intel_uc_fw *huc_fw = &dev_priv->huc.huc_fw;
+	struct i915_vma *vma;
+	unsigned long offset = 0;
+	u32 size;
+	int ret;
+
+	ret = i915_gem_object_set_to_gtt_domain(huc_fw->uc_fw_obj, false);
+	if (ret) {
+		DRM_DEBUG_DRIVER("set-domain failed %d\n", ret);
+		return ret;
+	}
+
+	vma = i915_gem_object_ggtt_pin(huc_fw->uc_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);

drivers/gpu/drm/i915/intel_huc_loader.c

Quoted 10 lines
drivers/gpu/drm/i915/intel_huc_loader.c new file +10 -0
+	}
+
+	/* 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 */
+	I915_WRITE(GUC_WOPCM_SIZE, guc_wopcm_size(dev_priv));
+	I915_WRITE(DMA_GUC_WOPCM_OFFSET,

GUC_WOPCM_OFFSET_VALUE |

Quoted 8 lines
drivers/gpu/drm/i915/intel_huc_loader.c new file +8 -0
+			HUC_LOADING_AGENT_GUC);
+
+	/* Set the source address for the uCode */
+	offset = i915_ggtt_offset(vma) + huc_fw->header_offset;
+	I915_WRITE(DMA_ADDR_0_LOW, lower_32_bits(offset));
+	I915_WRITE(DMA_ADDR_0_HIGH, upper_32_bits(offset) & 0xFFFF);
+
+	/* Hardware doesn't look at destination address for HuC. Set it to 0,

drivers/gpu/drm/i915/intel_huc_loader.c

Quoted 10 lines
drivers/gpu/drm/i915/intel_huc_loader.c new file +10 -0
+	 * but still program the correct address space.
+	 */
+	I915_WRITE(DMA_ADDR_1_LOW, 0);
+	I915_WRITE(DMA_ADDR_1_HIGH, DMA_ADDRESS_SPACE_WOPCM);
+
+	size = huc_fw->header_size + huc_fw->ucode_size;
+	I915_WRITE(DMA_COPY_SIZE, size);
+
+	/* Start the DMA */
+	I915_WRITE(DMA_CTRL, _MASKED_BIT_ENABLE(HUC_UKERNEL |

START_DMA));

Quoted 10 lines
drivers/gpu/drm/i915/intel_huc_loader.c new file +10 -0
+
+	/* Wait for DMA to finish */
+	ret = wait_for((I915_READ(DMA_CTRL) & START_DMA) == 0, 100);
+
+	DRM_DEBUG_DRIVER("HuC DMA transfer wait over with ret %d\n", ret);
+
+	intel_uncore_forcewake_put(dev_priv, FORCEWAKE_ALL);
+
+	/*
+	 * We keep the object pages for reuse during resume. But we can unpin it

drivers/gpu/drm/i915/intel_huc_loader.c line 132

Quoted 10 lines
drivers/gpu/drm/i915/intel_huc_loader.c new file +10 -0
+	 * now that DMA has completed, so it doesn't continue to take up space.
+	 */
+	i915_vma_unpin(vma);
+
+	return ret;
+}
+
+/**
+ * intel_huc_init() - initiate HuC firmware loading request
+ * @dev: the drm device

Mismatched param name.

Quoted 9 lines
drivers/gpu/drm/i915/intel_huc_loader.c new file +9 -0
+ *
+ * Called early during driver load, but after GEM is initialised. The
+loading
+ * will continue only when driver explicitly specify firmware name and version.
+ * All other cases are considered as UC_FIRMWARE_NONE either because
+HW is not
+ * capable or driver yet support it. And there will be no error
+message for
+ * UC_FIRMWARE_NONE cases.
Quoted 8 lines

drivers/gpu/drm/i915/intel_huc_loader.c line 144

Quoted 5 lines
drivers/gpu/drm/i915/intel_huc_loader.c new file +5 -0
+ *
+ * The DMA-copying to HW is done later when intel_huc_load() is called.
+ */
+void intel_huc_init(struct drm_i915_private *dev_priv) {
+	struct intel_huc *huc = &dev_priv->huc;

Maybe *huc shall be passed as only param (to match intel_huc function name). Then dev_priv could be recreated from huc_to_i915().

Why? Can you elaborate?
-anusha

Quoted 130 lines
Quoted 18 lines
drivers/gpu/drm/i915/intel_huc_loader.c new file +17 -0
+	struct intel_uc_fw *huc_fw = &huc->huc_fw;
+	const char *fw_path = NULL;
+
+	huc_fw->uc_fw_path = NULL;
+	huc_fw->fetch_status = UC_FIRMWARE_NONE;
+	huc_fw->load_status = UC_FIRMWARE_NONE;
+	huc_fw->fw_type = UC_FW_TYPE_HUC;
+
+	if (!HAS_HUC_UCODE(dev_priv))
+		return;
+
+	if (IS_SKYLAKE(dev_priv)) {
+		fw_path = I915_SKL_HUC_UCODE;
+		huc_fw->major_ver_wanted = SKL_FW_MAJOR;
+		huc_fw->minor_ver_wanted = SKL_FW_MINOR;
+	}
+

drivers/gpu/drm/i915/intel_huc_loader.c line 172

Quoted 10 lines
drivers/gpu/drm/i915/intel_huc_loader.c new file +10 -0
+	huc_fw->uc_fw_path = fw_path;
+	huc_fw->fetch_status = UC_FIRMWARE_PENDING;
+
+	DRM_DEBUG_DRIVER("HuC firmware pending, path %s\n", fw_path);
+
+	intel_uc_fw_fetch(dev_priv, huc_fw); }
+
+/**
+ * intel_huc_load() - load HuC uCode to device
+ * @dev: the drm device

Mismatched param name.

Quoted 98 lines
drivers/gpu/drm/i915/intel_huc_loader.c new file +90 -0
+ *
+ * Called from gem_init_hw() during driver loading and also after a GPU reset.
+ * Be note that HuC loading must be done before GuC loading.
+ *
+ * The firmware image should have already been fetched into memory by
+the
+ * earlier call to intel_huc_init(), so here we need only check that
+ * is succeeded, and then transfer the image to the h/w.
+ *
+ * Return:	non-zero code on error
+ */
+int intel_huc_load(struct drm_i915_private *dev_priv) {
+	struct intel_uc_fw *huc_fw = &dev_priv->huc.huc_fw;
+	int err;
+
+	if (huc_fw->fetch_status == UC_FIRMWARE_NONE)
+		return 0;
+
+	DRM_DEBUG_DRIVER("%s fw status: fetch %s, load %s\n",
+		huc_fw->uc_fw_path,
+		intel_uc_fw_status_repr(huc_fw->fetch_status),
+		intel_uc_fw_status_repr(huc_fw->load_status));
+
+	if (huc_fw->fetch_status == UC_FIRMWARE_SUCCESS &&
+	    huc_fw->load_status == UC_FIRMWARE_FAIL)
+		return -ENOEXEC;
+
+	huc_fw->load_status = UC_FIRMWARE_PENDING;
+
+	switch (huc_fw->fetch_status) {
+	case UC_FIRMWARE_FAIL:
+		/* something went wrong :( */
+		err = -EIO;
+		goto fail;
+
+	case UC_FIRMWARE_NONE:
+	case UC_FIRMWARE_PENDING:
+	default:
+		/* "can't happen" */
+		WARN_ONCE(1, "HuC fw %s invalid fetch_status %s [%d]\n",
+			huc_fw->uc_fw_path,
+			intel_uc_fw_status_repr(huc_fw->fetch_status),
+			huc_fw->fetch_status);
+		err = -ENXIO;
+		goto fail;
+
+	case UC_FIRMWARE_SUCCESS:
+		break;
+	}
+
+	err = huc_ucode_xfer(dev_priv);
+	if (err)
+		goto fail;
+
+	huc_fw->load_status = UC_FIRMWARE_SUCCESS;
+
+	DRM_DEBUG_DRIVER("%s fw status: fetch %s, load %s\n",
+		huc_fw->uc_fw_path,
+		intel_uc_fw_status_repr(huc_fw->fetch_status),
+		intel_uc_fw_status_repr(huc_fw->load_status));
+
+	return 0;
+
+fail:
+	if (huc_fw->load_status == UC_FIRMWARE_PENDING)
+		huc_fw->load_status = UC_FIRMWARE_FAIL;
+
+	DRM_ERROR("Failed to complete HuC uCode load with ret %d\n", err);
+
+	return err;
+}
+
+/**
+ * intel_huc_fini() - clean up resources allocated for HuC
+ * @dev: the drm device
+ *
+ * Cleans up by releasing the huc firmware GEM obj.
+ */
+void intel_huc_fini(struct drm_device *dev) {
+	struct drm_i915_private *dev_priv = to_i915(dev);
+	struct intel_uc_fw *huc_fw = &dev_priv->huc.huc_fw;
+
+	mutex_lock(&dev->struct_mutex);
+	if (huc_fw->uc_fw_obj)
+		i915_gem_object_put(huc_fw->uc_fw_obj);
+	huc_fw->uc_fw_obj = NULL;
+	mutex_unlock(&dev->struct_mutex);
+
+	huc_fw->fetch_status = UC_FIRMWARE_NONE; }
+
(quoted diff) +0 -0

b/drivers/gpu/drm/i915/intel_uc.h index be89f0b..ac92946 100644

drivers/gpu/drm/i915/intel_uc.h +0 -0
@@ -24,6 +24,12 @@
 #ifndef _INTEL_UC_H_
 #define _INTEL_UC_H_
 
Quoted 4 lines

drivers/gpu/drm/i915/intel_uc.h line 28

Quoted 2 lines
drivers/gpu/drm/i915/intel_uc.h +2 -0
+#ifndef _INTEL_HUC_H_
+#define _INTEL_HUC_H_

Typo ? This is still intel_uc.h file, right ?

Yes it is intel_uc.h. Initially the above two initializations were in intel_huc.h. But now its contents are moved to intel_uc.h. One common file for guc and huc declarations.....

Quoted 8 lines

drivers/gpu/drm/i915/intel_uc.h line 32

Quoted 5 lines
drivers/gpu/drm/i915/intel_uc.h +4 -0
+
+#define HUC_STATUS2             _MMIO(0xD3B0)
+#define   HUC_FW_VERIFIED       (1<<7)
+

Is it correct place for these defs?
What about i915_guc_reg.h ?

This was also initially in intel_huc.h.

Quoted 20 lines
Quoted 20 lines
drivers/gpu/drm/i915/intel_uc.h +7 -0
 #include "intel_guc_fwif.h"
 #include "i915_guc_reg.h"
 #include "intel_ringbuffer.h"
@@ -175,6 +181,13 @@ struct intel_guc {
 	struct mutex send_mutex;
 };
 
+struct intel_huc {
+	/* Generic uC firmware management */
+	struct intel_uc_fw huc_fw;
+
+	/* HuC-specific additions */
+};
+
 /* intel_uc.c */
 void intel_uc_init_early(struct drm_i915_private *dev_priv);  bool

intel_guc_recv(struct drm_i915_private *dev_priv, u32 *status); @@ -191,6 +204,9 @@ extern void intel_guc_fini(struct drm_i915_private *dev_priv); extern const char *intel_uc_fw_status_repr(enum

Quoted 8 lines

drivers/gpu/drm/i915/intel_uc.h extern void intel_guc_fini(struct drm_i915_private *dev_priv); · line 209

Quoted 6 lines

intel_uc_fw_status status); extern int intel_guc_suspend(struct drm_i915_private *dev_priv); extern int intel_guc_resume(struct drm_i915_private *dev_priv);

drivers/gpu/drm/i915/intel_uc.h +3 -0
+void intel_uc_fw_fetch(struct drm_i915_private *dev_priv,
+	struct intel_uc_fw *uc_fw);
+u32 guc_wopcm_size(struct drm_i915_private *dev_priv);

All other public functions have intel_ prefix.

Got it. Will change. Thanks

Cheers,
Anusha

Quoted 19 lines
Quoted 19 lines

/* i915_guc_submission.c */
int i915_guc_submission_init(struct drm_i915_private *dev_priv); @@
-205,4 +221,10 @@ void i915_guc_register(struct drm_i915_private *dev_priv); void i915_guc_unregister(struct drm_i915_private *dev_priv); int i915_guc_log_control(struct drm_i915_private *dev_priv, u64 control_val);

drivers/gpu/drm/i915/intel_uc.h +6 -1
+/* intel_huc_loader.c */
+void intel_huc_init(struct drm_i915_private *dev_priv); void
+intel_huc_fini(struct drm_device *dev); int intel_huc_load(struct
+drm_i915_private *dev_priv);
+
+#endif
 #endif
--

2.7.4

Quoted 4 lines
Quoted 4 lines

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

#20 lore raw

Re: [PATCH 3/8] drm/i915/huc: Add HuC fw loading support

Quoted 498 lines · Anusha
Quoted 245 lines

-----Original Message-----
From: Michal Wajdeczko [mailto:[email protected]]
Sent: Friday, December 9, 2016 4:18 AM
To: Srivatsa, Anusha <[email protected]>
Cc: [email protected]; Alex Dai <[email protected]>; Peter Antoine <[email protected]>
Subject: Re: [Intel-gfx] [PATCH 3/8] drm/i915/huc: Add HuC fw loading support

Quoted 40 lines · anushasr

From: Anusha Srivatsa <[email protected]>

The HuC loading process is similar to GuC. The intel_uc_fw_fetch() is used for both cases.

HuC loading needs to be before GuC loading. The WOPCM setting must be done early before loading any of them.

v2: rebased on-top of drm-intel-nightly.
removed if(HAS_GUC()) before the guc call. (D.Gordon) update huc_version number of format.
v3: rebased to drm-intel-nightly, changed the file name format to
match the one in the huc package.
Changed dev->dev_private to to_i915()
v4: moved function back to where it was.
change wait_for_atomic to wait_for.
v5: rebased + comment changes.
v7: rebased.
v8: rebased.
v9: rebased. Changed the year in the copyright message to reflect the right year.Correct the comments,remove the unwanted WARN message, replace drm_gem_object_unreference() with i915_gem_object_put().Make the prototypes in intel_huc.h non-extern.
v10: rebased. Update the file construction done by HuC. It is similar to GuC.Adopted the approach used in-
https://patchwork.freedesktop.org/patch/104355/ <Tvrtko Ursulin>
v11: Fix warnings remove old declaration
v12: Change dev to dev_priv in macro definition.
Corrected comments.
v13: rebased.
v14: rebased on top of drm-tip
v15: rebased. Updated functions intel_huc_load(),intel_huc_init() and intel_uc_fw_fetch() to accept dev_priv instead of dev. Moved contents of intel_huc.h to intel_uc.h

Quoted 7 lines
---
 drivers/gpu/drm/i915/Makefile           |   1 +
 drivers/gpu/drm/i915/i915_drv.c         |   4 +-
 drivers/gpu/drm/i915/i915_drv.h         |   3 +-
 drivers/gpu/drm/i915/i915_guc_reg.h     |   3 +
 drivers/gpu/drm/i915/intel_guc_loader.c |   7 +-
 drivers/gpu/drm/i915/intel_huc_loader.c | 264

++++++++++++++++++++++++++++++++

drivers/gpu/drm/i915/i915_drv.c static int i915_load_modeset_init(struct drm_device

Quoted 20 lines
 drivers/gpu/drm/i915/intel_uc.h         |  22 +++
 7 files changed, 299 insertions(+), 5 deletions(-)  create mode

100644 drivers/gpu/drm/i915/intel_huc_loader.c

(quoted diff) +0 -0

b/drivers/gpu/drm/i915/Makefile index 3c30916..01d4f4b 100644

drivers/gpu/drm/i915/Makefile +1 -0
@@ -57,6 +57,7 @@ i915-y += i915_cmd_parser.o \  # general-purpose
 microcontroller (GuC) support  i915-y += intel_uc.o \
 	  intel_guc_loader.o \
+	  intel_huc_loader.o \
 	  i915_guc_submission.o
 
 # autogenerated null render state
(quoted diff) +0 -0

b/drivers/gpu/drm/i915/i915_drv.c index 6428588..85a47c2 100644

drivers/gpu/drm/i915/i915_drv.c +0 -0
@@ -600,6 +600,7 @@ static int i915_load_modeset_init(struct drm_device

*dev)

drivers/gpu/drm/i915/i915_drv.c static int i915_load_modeset_init(struct drm_device

Quoted 8 lines
drivers/gpu/drm/i915/i915_drv.c +1 -0
 	if (ret)
 		goto cleanup_irq;
 
+	intel_huc_init(dev_priv);
 	intel_guc_init(dev_priv);
 
 	ret = i915_gem_init(dev_priv);
@@ -627,6 +628,7 @@ static int i915_load_modeset_init(struct drm_device

*dev)

drivers/gpu/drm/i915/intel_guc_loader.c static int guc_ucode_xfer_dma(struct drm_i915_private

Quoted 62 lines
drivers/gpu/drm/i915/i915_drv.c +2 -1
 		DRM_ERROR("failed to idle hardware; continuing to unload!\n");
 	i915_gem_fini(dev_priv);
 cleanup_irq:
+	intel_huc_fini(dev);
 	intel_guc_fini(dev_priv);
 	drm_irq_uninstall(dev);
 	intel_teardown_gmbus(dev_priv);
@@ -1313,7 +1315,7 @@ void i915_driver_unload(struct drm_device *dev)
 
 	/* Flush any outstanding unpin_work. */
 	drain_workqueue(dev_priv->wq);
-
+	intel_huc_fini(dev);
 	intel_guc_fini(dev_priv);
 	i915_gem_fini(dev_priv);
 	intel_fbc_cleanup_cfb(dev_priv);
(quoted diff) +0 -0

b/drivers/gpu/drm/i915/i915_drv.h index 1480e73..0371ca4 100644

drivers/gpu/drm/i915/i915_drv.h +2 -1
@@ -2036,6 +2036,7 @@ struct drm_i915_private {
 
 	struct intel_gvt *gvt;
 
+	struct intel_huc huc;
 	struct intel_guc guc;
 
 	struct intel_csr csr;
@@ -2810,7 +2811,7 @@ intel_info(const struct drm_i915_private *dev_priv)
 #define HAS_GUC(dev_priv)	((dev_priv)->info.has_guc)
 #define HAS_GUC_UCODE(dev_priv)	(HAS_GUC(dev_priv))
 #define HAS_GUC_SCHED(dev_priv)	(HAS_GUC(dev_priv))
-
+#define HAS_HUC_UCODE(dev_priv)	(HAS_GUC(dev_priv))
 #define HAS_RESOURCE_STREAMER(dev_priv)
 ((dev_priv)->info.has_resource_streamer)
 

#define HAS_POOLED_EU(dev_priv) ((dev_priv)->info.has_pooled_eu)

drivers/gpu/drm/i915/i915_guc_reg.h +3 -0
index 5e638fc..f9829f6 100644
@@ -61,9 +61,12 @@
 #define   DMA_ADDRESS_SPACE_GTT		  (8 << 16)
 #define DMA_COPY_SIZE			_MMIO(0xc310)
 #define DMA_CTRL			_MMIO(0xc314)
+#define   HUC_UKERNEL			  (1<<9)
 #define   UOS_MOVE			  (1<<4)
 #define   START_DMA			  (1<<0)
 #define DMA_GUC_WOPCM_OFFSET		_MMIO(0xc340)
+#define   HUC_LOADING_AGENT_VCR		  (0<<1)
+#define   HUC_LOADING_AGENT_GUC		  (1<<1)
 #define   GUC_WOPCM_OFFSET_VALUE	  0x80000	/* 512KB */
 #define GUC_MAX_IDLE_COUNT		_MMIO(0xC3E4)
 
drivers/gpu/drm/i915/intel_guc_loader.c +0 -0
index 26a184f..b971351 100644
@@ -309,8 +309,8 @@ static int guc_ucode_xfer_dma(struct drm_i915_private

*dev_priv,

drivers/gpu/drm/i915/intel_guc_loader.c static int guc_ucode_xfer_dma(struct drm_i915_private *dev_priv,

Quoted 4 lines
drivers/gpu/drm/i915/intel_guc_loader.c +0 -1
 	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));

drivers/gpu/drm/i915/intel_guc_loader.c static int guc_ucode_xfer_dma(struct drm_i915_private · line 312

Quoted 2 lines
drivers/gpu/drm/i915/intel_guc_loader.c +1 -1
-
+	I915_WRITE(DMA_CTRL, _MASKED_BIT_ENABLE(UOS_MOVE |

START_DMA) |

drivers/gpu/drm/i915/intel_guc_loader.c static int guc_ucode_xfer_dma(struct drm_i915_private

Quoted 5 lines
drivers/gpu/drm/i915/intel_guc_loader.c +1 -0
+		_MASKED_BIT_DISABLE(HUC_UKERNEL));
 	/*
 	 * Wait for the DMA to complete & the GuC to start up.
 	 * NB: Docs recommend not using the interrupt for completion.
@@ -334,7 +334,7 @@ static int guc_ucode_xfer_dma(struct drm_i915_private

*dev_priv,

Quoted 39 lines
drivers/gpu/drm/i915/intel_guc_loader.c +2 -1
 	return ret;
 }
 
-static u32 guc_wopcm_size(struct drm_i915_private *dev_priv)
+u32 guc_wopcm_size(struct drm_i915_private *dev_priv)
 {
 	u32 wopcm_size = GUC_WOPCM_TOP;
 
@@ -511,6 +511,7 @@ int intel_guc_setup(struct drm_i915_private *dev_priv)
 		if (err)
 			goto fail;
 
+		intel_huc_load(dev_priv);
 		err = guc_ucode_xfer(dev_priv);
 		if (!err)
 			break;
drivers/gpu/drm/i915/intel_huc_loader.c new file +16 -0
new file mode 100644
index 0000000..e0efd1c
@@ -0,0 +1,264 @@
+/*
+ * Copyright (c) 2016 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

drivers/gpu/drm/i915/intel_huc_loader.c line 26

Quoted 10 lines
drivers/gpu/drm/i915/intel_huc_loader.c new file +10 -0
+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

drivers/gpu/drm/i915/intel_huc_loader.c line 28

Quoted 2 lines
drivers/gpu/drm/i915/intel_huc_loader.c new file +2 -0
+SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM,

DAMAGES

Quoted 26 lines
drivers/gpu/drm/i915/intel_huc_loader.c new file +26 -0
+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 <linux/firmware.h>
+#include "i915_drv.h"
+#include "intel_uc.h"
+
+/**
+ * DOC: HuC Firmware
+ *
+ * Motivation:
+ * GEN9 introduces a new dedicated firmware for usage in media HEVC
+(High
+ * Efficiency Video Coding) operations. Userspace can use the
+firmware
+ * capabilities by adding HuC specific commands to batch buffers.
+ *
+ * Implementation:
+ * The same firmware loader is used as the GuC. However, the actual
+ * loading to HW is deferred until GEM initialization is done.
+ *
Quoted 8 lines

drivers/gpu/drm/i915/intel_huc_loader.c line 44

Quoted 5 lines
drivers/gpu/drm/i915/intel_huc_loader.c new file +5 -0
+ * Note that HuC firmware loading must be done before GuC loading.
+ */
+
+#define SKL_FW_MAJOR 01
+#define SKL_FW_MINOR 07

Can we use SKL_HUC_FW_ prefix to distinguish these macros from similar defined in intel_guc_loader.c ?

Sure

Quoted 102 lines
Quoted 40 lines
drivers/gpu/drm/i915/intel_huc_loader.c new file +39 -0
+#define SKL_BLD_NUM 1398
+
+#define HUC_FW_PATH(platform, major, minor, bld_num) \
+	"i915/" __stringify(platform) "_huc_ver" __stringify(major) "_" \
+	__stringify(minor) "_" __stringify(bld_num) ".bin"
+
+#define I915_SKL_HUC_UCODE HUC_FW_PATH(skl, SKL_FW_MAJOR, \
+	SKL_FW_MINOR, SKL_BLD_NUM)
+MODULE_FIRMWARE(I915_SKL_HUC_UCODE);
+
+/**
+ * huc_ucode_xfer() - DMA's the firmware
+ * @dev_priv: the drm device
+ *
+ * This function takes the gem object containing the firmware, sets
+up the DMA
+ * engine MMIO, triggers the DMA operation and waits for it to finish.
+ *
+ * Transfer the firmware image to RAM for execution by the microcontroller.
+ *
+ * Return: 0 on success, non-zero on failure  */
+
+static int huc_ucode_xfer(struct drm_i915_private *dev_priv) {
+	struct intel_uc_fw *huc_fw = &dev_priv->huc.huc_fw;
+	struct i915_vma *vma;
+	unsigned long offset = 0;
+	u32 size;
+	int ret;
+
+	ret = i915_gem_object_set_to_gtt_domain(huc_fw->uc_fw_obj, false);
+	if (ret) {
+		DRM_DEBUG_DRIVER("set-domain failed %d\n", ret);
+		return ret;
+	}
+
+	vma = i915_gem_object_ggtt_pin(huc_fw->uc_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);

drivers/gpu/drm/i915/intel_huc_loader.c

Quoted 10 lines
drivers/gpu/drm/i915/intel_huc_loader.c new file +10 -0
+	}
+
+	/* 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 */
+	I915_WRITE(GUC_WOPCM_SIZE, guc_wopcm_size(dev_priv));
+	I915_WRITE(DMA_GUC_WOPCM_OFFSET,

GUC_WOPCM_OFFSET_VALUE |

Quoted 8 lines
drivers/gpu/drm/i915/intel_huc_loader.c new file +8 -0
+			HUC_LOADING_AGENT_GUC);
+
+	/* Set the source address for the uCode */
+	offset = i915_ggtt_offset(vma) + huc_fw->header_offset;
+	I915_WRITE(DMA_ADDR_0_LOW, lower_32_bits(offset));
+	I915_WRITE(DMA_ADDR_0_HIGH, upper_32_bits(offset) & 0xFFFF);
+
+	/* Hardware doesn't look at destination address for HuC. Set it to 0,

drivers/gpu/drm/i915/intel_huc_loader.c

Quoted 10 lines
drivers/gpu/drm/i915/intel_huc_loader.c new file +10 -0
+	 * but still program the correct address space.
+	 */
+	I915_WRITE(DMA_ADDR_1_LOW, 0);
+	I915_WRITE(DMA_ADDR_1_HIGH, DMA_ADDRESS_SPACE_WOPCM);
+
+	size = huc_fw->header_size + huc_fw->ucode_size;
+	I915_WRITE(DMA_COPY_SIZE, size);
+
+	/* Start the DMA */
+	I915_WRITE(DMA_CTRL, _MASKED_BIT_ENABLE(HUC_UKERNEL |

START_DMA));

Quoted 10 lines
drivers/gpu/drm/i915/intel_huc_loader.c new file +10 -0
+
+	/* Wait for DMA to finish */
+	ret = wait_for((I915_READ(DMA_CTRL) & START_DMA) == 0, 100);
+
+	DRM_DEBUG_DRIVER("HuC DMA transfer wait over with ret %d\n", ret);
+
+	intel_uncore_forcewake_put(dev_priv, FORCEWAKE_ALL);
+
+	/*
+	 * We keep the object pages for reuse during resume. But we can unpin it

drivers/gpu/drm/i915/intel_huc_loader.c line 132

Quoted 10 lines
drivers/gpu/drm/i915/intel_huc_loader.c new file +10 -0
+	 * now that DMA has completed, so it doesn't continue to take up space.
+	 */
+	i915_vma_unpin(vma);
+
+	return ret;
+}
+
+/**
+ * intel_huc_init() - initiate HuC firmware loading request
+ * @dev: the drm device

Mismatched param name.

Quoted 9 lines
drivers/gpu/drm/i915/intel_huc_loader.c new file +9 -0
+ *
+ * Called early during driver load, but after GEM is initialised. The
+loading
+ * will continue only when driver explicitly specify firmware name and version.
+ * All other cases are considered as UC_FIRMWARE_NONE either because
+HW is not
+ * capable or driver yet support it. And there will be no error
+message for
+ * UC_FIRMWARE_NONE cases.
Quoted 8 lines

drivers/gpu/drm/i915/intel_huc_loader.c line 144

Quoted 5 lines
drivers/gpu/drm/i915/intel_huc_loader.c new file +5 -0
+ *
+ * The DMA-copying to HW is done later when intel_huc_load() is called.
+ */
+void intel_huc_init(struct drm_i915_private *dev_priv) {
+	struct intel_huc *huc = &dev_priv->huc;

Maybe *huc shall be passed as only param (to match intel_huc function name). Then dev_priv could be recreated from huc_to_i915().

Why? Can you elaborate?
-anusha

Quoted 130 lines
Quoted 18 lines
drivers/gpu/drm/i915/intel_huc_loader.c new file +17 -0
+	struct intel_uc_fw *huc_fw = &huc->huc_fw;
+	const char *fw_path = NULL;
+
+	huc_fw->uc_fw_path = NULL;
+	huc_fw->fetch_status = UC_FIRMWARE_NONE;
+	huc_fw->load_status = UC_FIRMWARE_NONE;
+	huc_fw->fw_type = UC_FW_TYPE_HUC;
+
+	if (!HAS_HUC_UCODE(dev_priv))
+		return;
+
+	if (IS_SKYLAKE(dev_priv)) {
+		fw_path = I915_SKL_HUC_UCODE;
+		huc_fw->major_ver_wanted = SKL_FW_MAJOR;
+		huc_fw->minor_ver_wanted = SKL_FW_MINOR;
+	}
+

drivers/gpu/drm/i915/intel_huc_loader.c line 172

Quoted 10 lines
drivers/gpu/drm/i915/intel_huc_loader.c new file +10 -0
+	huc_fw->uc_fw_path = fw_path;
+	huc_fw->fetch_status = UC_FIRMWARE_PENDING;
+
+	DRM_DEBUG_DRIVER("HuC firmware pending, path %s\n", fw_path);
+
+	intel_uc_fw_fetch(dev_priv, huc_fw); }
+
+/**
+ * intel_huc_load() - load HuC uCode to device
+ * @dev: the drm device

Mismatched param name.

Quoted 98 lines
drivers/gpu/drm/i915/intel_huc_loader.c new file +90 -0
+ *
+ * Called from gem_init_hw() during driver loading and also after a GPU reset.
+ * Be note that HuC loading must be done before GuC loading.
+ *
+ * The firmware image should have already been fetched into memory by
+the
+ * earlier call to intel_huc_init(), so here we need only check that
+ * is succeeded, and then transfer the image to the h/w.
+ *
+ * Return:	non-zero code on error
+ */
+int intel_huc_load(struct drm_i915_private *dev_priv) {
+	struct intel_uc_fw *huc_fw = &dev_priv->huc.huc_fw;
+	int err;
+
+	if (huc_fw->fetch_status == UC_FIRMWARE_NONE)
+		return 0;
+
+	DRM_DEBUG_DRIVER("%s fw status: fetch %s, load %s\n",
+		huc_fw->uc_fw_path,
+		intel_uc_fw_status_repr(huc_fw->fetch_status),
+		intel_uc_fw_status_repr(huc_fw->load_status));
+
+	if (huc_fw->fetch_status == UC_FIRMWARE_SUCCESS &&
+	    huc_fw->load_status == UC_FIRMWARE_FAIL)
+		return -ENOEXEC;
+
+	huc_fw->load_status = UC_FIRMWARE_PENDING;
+
+	switch (huc_fw->fetch_status) {
+	case UC_FIRMWARE_FAIL:
+		/* something went wrong :( */
+		err = -EIO;
+		goto fail;
+
+	case UC_FIRMWARE_NONE:
+	case UC_FIRMWARE_PENDING:
+	default:
+		/* "can't happen" */
+		WARN_ONCE(1, "HuC fw %s invalid fetch_status %s [%d]\n",
+			huc_fw->uc_fw_path,
+			intel_uc_fw_status_repr(huc_fw->fetch_status),
+			huc_fw->fetch_status);
+		err = -ENXIO;
+		goto fail;
+
+	case UC_FIRMWARE_SUCCESS:
+		break;
+	}
+
+	err = huc_ucode_xfer(dev_priv);
+	if (err)
+		goto fail;
+
+	huc_fw->load_status = UC_FIRMWARE_SUCCESS;
+
+	DRM_DEBUG_DRIVER("%s fw status: fetch %s, load %s\n",
+		huc_fw->uc_fw_path,
+		intel_uc_fw_status_repr(huc_fw->fetch_status),
+		intel_uc_fw_status_repr(huc_fw->load_status));
+
+	return 0;
+
+fail:
+	if (huc_fw->load_status == UC_FIRMWARE_PENDING)
+		huc_fw->load_status = UC_FIRMWARE_FAIL;
+
+	DRM_ERROR("Failed to complete HuC uCode load with ret %d\n", err);
+
+	return err;
+}
+
+/**
+ * intel_huc_fini() - clean up resources allocated for HuC
+ * @dev: the drm device
+ *
+ * Cleans up by releasing the huc firmware GEM obj.
+ */
+void intel_huc_fini(struct drm_device *dev) {
+	struct drm_i915_private *dev_priv = to_i915(dev);
+	struct intel_uc_fw *huc_fw = &dev_priv->huc.huc_fw;
+
+	mutex_lock(&dev->struct_mutex);
+	if (huc_fw->uc_fw_obj)
+		i915_gem_object_put(huc_fw->uc_fw_obj);
+	huc_fw->uc_fw_obj = NULL;
+	mutex_unlock(&dev->struct_mutex);
+
+	huc_fw->fetch_status = UC_FIRMWARE_NONE; }
+
(quoted diff) +0 -0

b/drivers/gpu/drm/i915/intel_uc.h index be89f0b..ac92946 100644

drivers/gpu/drm/i915/intel_uc.h +0 -0
@@ -24,6 +24,12 @@
 #ifndef _INTEL_UC_H_
 #define _INTEL_UC_H_
 
Quoted 6 lines
Quoted 4 lines

drivers/gpu/drm/i915/intel_uc.h line 28

Quoted 2 lines
drivers/gpu/drm/i915/intel_uc.h +2 -0
+#ifndef _INTEL_HUC_H_
+#define _INTEL_HUC_H_

Typo ? This is still intel_uc.h file, right ?

Yes it is intel_uc.h. Initially the above two initializations were in intel_huc.h. But now its contents are moved to intel_uc.h. One common file for guc and huc declarations.....

But those are guards that assure us that the content of header file is included one. The name should come strictly from the file name. One guard is enough. So drop the _INTEL_HUC_H_ and just leave the _INTEL_UC_H_

Quoted 68 lines
Quoted 8 lines

drivers/gpu/drm/i915/intel_uc.h line 32

Quoted 5 lines
drivers/gpu/drm/i915/intel_uc.h +4 -0
+
+#define HUC_STATUS2             _MMIO(0xD3B0)
+#define   HUC_FW_VERIFIED       (1<<7)
+

Is it correct place for these defs?
What about i915_guc_reg.h ?

This was also initially in intel_huc.h.

Quoted 20 lines
Quoted 20 lines
drivers/gpu/drm/i915/intel_uc.h +7 -0
 #include "intel_guc_fwif.h"
 #include "i915_guc_reg.h"
 #include "intel_ringbuffer.h"
@@ -175,6 +181,13 @@ struct intel_guc {
 	struct mutex send_mutex;
 };
 
+struct intel_huc {
+	/* Generic uC firmware management */
+	struct intel_uc_fw huc_fw;
+
+	/* HuC-specific additions */
+};
+
 /* intel_uc.c */
 void intel_uc_init_early(struct drm_i915_private *dev_priv);  bool

intel_guc_recv(struct drm_i915_private *dev_priv, u32 *status); @@ -191,6 +204,9 @@ extern void intel_guc_fini(struct drm_i915_private *dev_priv); extern const char *intel_uc_fw_status_repr(enum

Quoted 8 lines

drivers/gpu/drm/i915/intel_uc.h extern void intel_guc_fini(struct drm_i915_private *dev_priv); · line 209

Quoted 6 lines

intel_uc_fw_status status); extern int intel_guc_suspend(struct drm_i915_private *dev_priv); extern int intel_guc_resume(struct drm_i915_private *dev_priv);

drivers/gpu/drm/i915/intel_uc.h +3 -0
+void intel_uc_fw_fetch(struct drm_i915_private *dev_priv,
+	struct intel_uc_fw *uc_fw);
+u32 guc_wopcm_size(struct drm_i915_private *dev_priv);

All other public functions have intel_ prefix.

Got it. Will change. Thanks

Cheers,
Anusha

Quoted 19 lines
Quoted 19 lines

/* i915_guc_submission.c */
int i915_guc_submission_init(struct drm_i915_private *dev_priv); @@
-205,4 +221,10 @@ void i915_guc_register(struct drm_i915_private *dev_priv); void i915_guc_unregister(struct drm_i915_private *dev_priv); int i915_guc_log_control(struct drm_i915_private *dev_priv, u64 control_val);

drivers/gpu/drm/i915/intel_uc.h +6 -1
+/* intel_huc_loader.c */
+void intel_huc_init(struct drm_i915_private *dev_priv); void
+intel_huc_fini(struct drm_device *dev); int intel_huc_load(struct
+drm_i915_private *dev_priv);
+
+#endif
 #endif
--

2.7.4

Quoted 4 lines
Quoted 4 lines

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Signature · 6 lines
Cheers,
Arek
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
#21 lore raw

Re: [PATCH 3/8] drm/i915/huc: Add HuC fw loading support

In reply to anushasr

Hi all,

Executive decision required below:

On 08/12/2016 23:02, anushasr wrote:

[snip]

Quoted 17 lines
drivers/gpu/drm/i915/intel_huc_loader.c new file +17 -0
+
+/**
+ * DOC: HuC Firmware
+ *
+ * Motivation:
+ * GEN9 introduces a new dedicated firmware for usage in media HEVC (High
+ * Efficiency Video Coding) operations. Userspace can use the firmware
+ * capabilities by adding HuC specific commands to batch buffers.
+ *
+ * Implementation:
+ * The same firmware loader is used as the GuC. However, the actual
+ * loading to HW is deferred until GEM initialization is done.
+ *
+ * Note that HuC firmware loading must be done before GuC loading.
+ */
+
+#define SKL_FW_MAJOR 01

drivers/gpu/drm/i915/intel_huc_loader.c line 53

Quoted 10 lines
drivers/gpu/drm/i915/intel_huc_loader.c new file +10 -0
+#define SKL_FW_MINOR 07
+#define SKL_BLD_NUM 1398
+
+#define HUC_FW_PATH(platform, major, minor, bld_num) \
+	"i915/" __stringify(platform) "_huc_ver" __stringify(major) "_" \
+	__stringify(minor) "_" __stringify(bld_num) ".bin"
+
+#define I915_SKL_HUC_UCODE HUC_FW_PATH(skl, SKL_FW_MAJOR, \
+	SKL_FW_MINOR, SKL_BLD_NUM)
+MODULE_FIRMWARE(I915_SKL_HUC_UCODE);

Daniel & Jani, I understand in the GuC firmware discussions you were very much in the favour of encoding the full name (major and minor included) as the fw firmware?

Argument was that we want to in effect claim support only for one validated firmware binary with one version of i915.

In the case of the HuC we have a very similar situation with two key differences.

First, there is a build number in the file name as provided by the firmware team. We know that it will happen that only the build number changes with some fixes and the minor stays the same. And we know that the major indicates the interface compatibility.

Secondly, from all I can see, there are no interactions between the driver and the HuC firmware apart from driver loading it and thats it.

In the light of that I was advocating only using the major in the driver request fw name in order to improve usability both for developers (easier to test with different firmwares), and for users (be it distributions or end users - easier to upgrade the HuC firmware in case of codec issues by not having to patch and recompile the kernel).

Since I understand this topic has been beaten to death in the past and there are strong opinions on it, could you just okay (or not) the current proposal (as in posted patches) which encodes major, minor and build number in the fw name?

Regards,

Tvrtko
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

#22 lore raw

Re: [PATCH 3/8] drm/i915/huc: Add HuC fw loading support

Quoted 60 lines · Tvrtko Ursulin

Hi all,

Executive decision required below:

On 08/12/2016 23:02, anushasr wrote:

[snip]

Quoted 17 lines
drivers/gpu/drm/i915/intel_huc_loader.c new file +17 -0
+
+/**
+ * DOC: HuC Firmware
+ *
+ * Motivation:
+ * GEN9 introduces a new dedicated firmware for usage in media HEVC (High
+ * Efficiency Video Coding) operations. Userspace can use the firmware
+ * capabilities by adding HuC specific commands to batch buffers.
+ *
+ * Implementation:
+ * The same firmware loader is used as the GuC. However, the actual
+ * loading to HW is deferred until GEM initialization is done.
+ *
+ * Note that HuC firmware loading must be done before GuC loading.
+ */
+
+#define SKL_FW_MAJOR 01

drivers/gpu/drm/i915/intel_huc_loader.c line 53

Quoted 10 lines
drivers/gpu/drm/i915/intel_huc_loader.c new file +10 -0
+#define SKL_FW_MINOR 07
+#define SKL_BLD_NUM 1398
+
+#define HUC_FW_PATH(platform, major, minor, bld_num) \
+	"i915/" __stringify(platform) "_huc_ver" __stringify(major) "_" \
+	__stringify(minor) "_" __stringify(bld_num) ".bin"
+
+#define I915_SKL_HUC_UCODE HUC_FW_PATH(skl, SKL_FW_MAJOR, \
+	SKL_FW_MINOR, SKL_BLD_NUM)
+MODULE_FIRMWARE(I915_SKL_HUC_UCODE);

Daniel & Jani, I understand in the GuC firmware discussions you were very much in the favour of encoding the full name (major and minor included) as the fw firmware?

Argument was that we want to in effect claim support only for one validated firmware binary with one version of i915.

In the case of the HuC we have a very similar situation with two key differences.

First, there is a build number in the file name as provided by the firmware team. We know that it will happen that only the build number changes with some fixes and the minor stays the same. And we know that the major indicates the interface compatibility.

Secondly, from all I can see, there are no interactions between the driver and the HuC firmware apart from driver loading it and thats it.

In the light of that I was advocating only using the major in the driver request fw name in order to improve usability both for developers (easier to test with different firmwares), and for users (be it distributions or end users - easier to upgrade the HuC firmware in case of codec issues by not having to patch and recompile the kernel).

Quoted 4 lines

Since I understand this topic has been beaten to death in the past and there are strong opinions on it, could you just okay (or not) the current proposal (as in posted patches) which encodes major, minor and build number in the fw name?

The question is the same as it ever was, can you absolutely guarantee a new firmware version (even if just a build number bump) will not cause regressions? Note that we'll probably only have resources to test the latest kernel against the latest firmware, but accepting future firmware versions means even stable kernels will start using the new firmware versions after linux-firmware updates. We also can't easily retroactively prevent this from happening even if we find out that there will be breakage.

I still think we should only accept one firmware version. If we (or someone else) has the resources to test against older kernels, commits to enable newer firmware versions can be backported to stable.

I would love to be able to look at the firmware sources and say, oh, there are no interactions with the kernel whatsoever, but you know how that is. I don't know if there are interactions. I don't know if future blobs will have interactions, and break everything if the kernel doesn't do something the black box expects.

If you want to help developers test various firmware versions, I suggest the same thing I suggested the last time: add an _unsafe module parameter to specify the firmware filename/version to load.

BR,
Jani.

Signature · 5 lines
Jani Nikula, Intel Open Source Technology Center
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
#23 lore raw

Re: [PATCH 3/8] drm/i915/huc: Add HuC fw loading support

Quoted 83 lines

From: Intel-gfx [mailto:[email protected]] On Behalf Of Jani Nikula
Sent: Wednesday, December 14, 2016 7:20 AM
To: Tvrtko Ursulin <[email protected]>; Srivatsa, Anusha <[email protected]>; [email protected]
Subject: Re: [Intel-gfx] [PATCH 3/8] drm/i915/huc: Add HuC fw loading support

Quoted 62 lines · Tvrtko Ursulin

Hi all,

Executive decision required below:

On 08/12/2016 23:02, anushasr wrote:

[snip]

Quoted 19 lines
drivers/gpu/drm/i915/intel_huc_loader.c new file +19 -0
+
+/**
+ * DOC: HuC Firmware
+ *
+ * Motivation:
+ * GEN9 introduces a new dedicated firmware for usage in media HEVC
+(High
+ * Efficiency Video Coding) operations. Userspace can use the
+firmware
+ * capabilities by adding HuC specific commands to batch buffers.
+ *
+ * Implementation:
+ * The same firmware loader is used as the GuC. However, the actual
+ * loading to HW is deferred until GEM initialization is done.
+ *
+ * Note that HuC firmware loading must be done before GuC loading.
+ */
+
+#define SKL_FW_MAJOR 01

drivers/gpu/drm/i915/intel_huc_loader.c line 53

Quoted 10 lines
drivers/gpu/drm/i915/intel_huc_loader.c new file +10 -0
+#define SKL_FW_MINOR 07
+#define SKL_BLD_NUM 1398
+
+#define HUC_FW_PATH(platform, major, minor, bld_num) \
+	"i915/" __stringify(platform) "_huc_ver" __stringify(major) "_" \
+	__stringify(minor) "_" __stringify(bld_num) ".bin"
+
+#define I915_SKL_HUC_UCODE HUC_FW_PATH(skl, SKL_FW_MAJOR, \
+	SKL_FW_MINOR, SKL_BLD_NUM)
+MODULE_FIRMWARE(I915_SKL_HUC_UCODE);

Daniel & Jani, I understand in the GuC firmware discussions you were very much in the favour of encoding the full name (major and minor included) as the fw firmware?

Argument was that we want to in effect claim support only for one validated firmware binary with one version of i915.

In the case of the HuC we have a very similar situation with two key differences.

First, there is a build number in the file name as provided by the firmware team. We know that it will happen that only the build number changes with some fixes and the minor stays the same. And we know that the major indicates the interface compatibility.

Secondly, from all I can see, there are no interactions between the driver and the HuC firmware apart from driver loading it and thats it.

In the light of that I was advocating only using the major in the driver request fw name in order to improve usability both for developers (easier to test with different firmwares), and for users (be it distributions or end users - easier to upgrade the HuC firmware in case of codec issues by not having to patch and recompile the kernel).

Quoted 4 lines

Since I understand this topic has been beaten to death in the past and there are strong opinions on it, could you just okay (or not) the current proposal (as in posted patches) which encodes major, minor and build number in the fw name?

The question is the same as it ever was, can you absolutely guarantee a new firmware version (even if just a build number bump) will not cause regressions? Note that we'll probably only have resources to test the latest kernel against the latest firmware, but accepting future firmware versions means even stable kernels will start using the new firmware versions after linux-firmware updates. We also can't easily retroactively prevent this from happening even if we find out that there will be breakage.

Quoted 4 lines

I still think we should only accept one firmware version. If we (or someone else) has the resources to test against older kernels, commits to enable newer firmware versions can be backported to stable.

I agree with Jani's position here. We have to keep some controls to the combinations of FW and kernels so we know what works. At least until FW matures in the future.

Quoted 14 lines

I would love to be able to look at the firmware sources and say, oh, there are no interactions with the kernel whatsoever, but you know how that is. I don't know if there are interactions. I don't know if future blobs will have interactions, and break everything if the kernel doesn't do something the black box expects.

If you want to help developers test various firmware versions, I suggest the same thing I suggested the last time: add an _unsafe module parameter to specify the firmware filename/version to load.

BR,
Jani.

Quoted 6 lines

--
Jani Nikula, Intel Open Source Technology Center
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

author
#24 lore raw

PATCH 4/8 drm/i915/huc: Add BXT HuC Loading Support

In reply to anushasr

From: Anusha Srivatsa <[email protected]>

This patch adds the HuC Loading for the BXT by using the updated file construction.

Version 1.7 of the HuC firmware.

v2: rebased.
v3: rebased on top of drm-tip
v4: rebased.

---
 drivers/gpu/drm/i915/intel_huc_loader.c | 11 +++++++++++
 1 file changed, 11 insertions(+)
drivers/gpu/drm/i915/intel_huc_loader.c +11 -0
index e0efd1c..a7a1a7f 100644
@@ -40,6 +40,10 @@
  * Note that HuC firmware loading must be done before GuC loading.
  */
 
+#define BXT_FW_MAJOR 01
+#define BXT_FW_MINOR 07
+#define BXT_BLD_NUM 1398
+
 #define SKL_FW_MAJOR 01
 #define SKL_FW_MINOR 07
 #define SKL_BLD_NUM 1398
@@ -52,6 +56,9 @@
 	SKL_FW_MINOR, SKL_BLD_NUM)
 MODULE_FIRMWARE(I915_SKL_HUC_UCODE);
 
+#define I915_BXT_HUC_UCODE HUC_FW_PATH(bxt, BXT_FW_MAJOR, \
+	BXT_FW_MINOR, BXT_BLD_NUM)
+MODULE_FIRMWARE(I915_BXT_HUC_UCODE);
 /**
  * huc_ucode_xfer() - DMA's the firmware
  * @dev_priv: the drm device
@@ -157,6 +164,10 @@ void intel_huc_init(struct drm_i915_private *dev_priv)
 		fw_path = I915_SKL_HUC_UCODE;
 		huc_fw->major_ver_wanted = SKL_FW_MAJOR;
 		huc_fw->minor_ver_wanted = SKL_FW_MINOR;
+	} else if (IS_BROXTON(dev_priv)) {
+		fw_path = I915_BXT_HUC_UCODE;
+		huc_fw->major_ver_wanted = BXT_FW_MAJOR;
+		huc_fw->minor_ver_wanted = BXT_FW_MINOR;
 	}
 
 	huc_fw->uc_fw_path = fw_path;
Signature · 6 lines
2.7.4

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
author
#25 lore raw

PATCH 5/8 drm/i915/HuC: Add KBL huC loading Support

In reply to anushasr

From: Anusha Srivatsa <[email protected]>

This patch adds the support to load HuC on KBL
Version 2.0

v2: rebased.
v3: rebased on top of drm-tip
v4: rebased.

---
 drivers/gpu/drm/i915/intel_huc_loader.c | 16 ++++++++++++++++
 1 file changed, 16 insertions(+)
drivers/gpu/drm/i915/intel_huc_loader.c +16 -0
index a7a1a7f..96fc727 100644
@@ -48,6 +48,10 @@
 #define SKL_FW_MINOR 07
 #define SKL_BLD_NUM 1398
 
+#define KBL_FW_MAJOR 02
+#define KBL_FW_MINOR 00
+#define KBL_BLD_NUM 1810
+
 #define HUC_FW_PATH(platform, major, minor, bld_num) \
 	"i915/" __stringify(platform) "_huc_ver" __stringify(major) "_" \
 	__stringify(minor) "_" __stringify(bld_num) ".bin"
@@ -59,6 +63,11 @@ MODULE_FIRMWARE(I915_SKL_HUC_UCODE);
 #define I915_BXT_HUC_UCODE HUC_FW_PATH(bxt, BXT_FW_MAJOR, \
 	BXT_FW_MINOR, BXT_BLD_NUM)
 MODULE_FIRMWARE(I915_BXT_HUC_UCODE);
+
+#define I915_KBL_HUC_UCODE HUC_FW_PATH(kbl, KBL_FW_MAJOR, \
+	KBL_FW_MINOR, KBL_BLD_NUM)
+MODULE_FIRMWARE(I915_KBL_HUC_UCODE);
+
 /**
  * huc_ucode_xfer() - DMA's the firmware
  * @dev_priv: the drm device
@@ -168,8 +177,15 @@ void intel_huc_init(struct drm_i915_private *dev_priv)
 		fw_path = I915_BXT_HUC_UCODE;
 		huc_fw->major_ver_wanted = BXT_FW_MAJOR;
 		huc_fw->minor_ver_wanted = BXT_FW_MINOR;
+	} else if (IS_KABYLAKE(dev_priv)) {
+		fw_path = I915_KBL_HUC_UCODE;
+		huc_fw->major_ver_wanted = KBL_FW_MAJOR;
+		huc_fw->minor_ver_wanted = KBL_FW_MINOR;
 	}
 
+	if (fw_path == NULL)
+		return;
+
 	huc_fw->uc_fw_path = fw_path;
 	huc_fw->fetch_status = UC_FIRMWARE_PENDING;
 
Signature · 6 lines
2.7.4

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
author
#26 lore raw

PATCH 6/8 drm/i915/huc: Add debugfs for HuC loading status check

In reply to anushasr

From: Peter Antoine <[email protected]>

Add debugfs entry for HuC loading status check.

v2: rebase on-top of drm-intel-nightly.
v3: rebased again.
v7: rebased.
v8: rebased.
v9: rebased.
v10: rebased.
v11: rebased on top of drm-tip
v12: rebased.

---
 drivers/gpu/drm/i915/i915_debugfs.c | 31 +++++++++++++++++++++++++++++++
 1 file changed, 31 insertions(+)
drivers/gpu/drm/i915/i915_debugfs.c +31 -0
index 0e5ef62..639b66d 100644
@@ -2335,6 +2335,36 @@ static int i915_llc(struct seq_file *m, void *data)
 	return 0;
 }
 
+static int i915_huc_load_status_info(struct seq_file *m, void *data)
+{
+	struct drm_i915_private *dev_priv = node_to_i915(m->private);
+	struct intel_uc_fw *huc_fw = &dev_priv->huc.huc_fw;
+
+	if (!HAS_HUC_UCODE(dev_priv))
+		return 0;
+
+	seq_puts(m, "HuC firmware status:\n");
+	seq_printf(m, "\tpath: %s\n", huc_fw->uc_fw_path);
+	seq_printf(m, "\tfetch: %s\n",
+		intel_uc_fw_status_repr(huc_fw->fetch_status));
+	seq_printf(m, "\tload: %s\n",
+		intel_uc_fw_status_repr(huc_fw->load_status));
+	seq_printf(m, "\tversion wanted: %d.%d\n",
+		huc_fw->major_ver_wanted, huc_fw->minor_ver_wanted);
+	seq_printf(m, "\tversion found: %d.%d\n",
+		huc_fw->major_ver_found, huc_fw->minor_ver_found);
+	seq_printf(m, "\theader: offset is %d; size = %d\n",
+		huc_fw->header_offset, huc_fw->header_size);
+	seq_printf(m, "\tuCode: offset is %d; size = %d\n",
+		huc_fw->ucode_offset, huc_fw->ucode_size);
+	seq_printf(m, "\tRSA: offset is %d; size = %d\n",
+		huc_fw->rsa_offset, huc_fw->rsa_size);
+
+	seq_printf(m, "\nHuC status 0x%08x:\n", I915_READ(HUC_STATUS2));
+
+	return 0;
+}
+
 static int i915_guc_load_status_info(struct seq_file *m, void *data)
 {
 	struct drm_i915_private *dev_priv = node_to_i915(m->private);
@@ -5403,6 +5433,7 @@ static const struct drm_info_list i915_debugfs_list[] = {
 	{"i915_guc_info", i915_guc_info, 0},
 	{"i915_guc_load_status", i915_guc_load_status_info, 0},
 	{"i915_guc_log_dump", i915_guc_log_dump, 0},
+	{"i915_huc_load_status", i915_huc_load_status_info, 0},
 	{"i915_frequency_info", i915_frequency_info, 0},
 	{"i915_hangcheck_info", i915_hangcheck_info, 0},
 	{"i915_drpc_info", i915_drpc_info, 0},
Signature · 6 lines
2.7.4

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
author
#27 lore raw

PATCH 7/8 drm/i915/huc: Support HuC authentication

In reply to anushasr

From: Peter Antoine <[email protected]>

The HuC authentication is done by host2guc call. The HuC RSA keys are sent to GuC for authentication.

v2: rebased on top of drm-intel-nightly.
changed name format and upped version 1.7.
v3: rebased on top of drm-intel-nightly.
v4: changed wait_for_automic to wait_for
v5: rebased.
v7: rebased.
v8: rebased.
v9: rebased. Rename intel_huc_auh() to intel_guc_auth_huc() and place the prototype in intel_guc.h,correct the comments.
v10: rebased.
v11: rebased.
v12: rebased on top of drm-tip
v13: rebased. Moved intel_guc_auth_huc from i915_guc_submission.c to intel_uc.c.Update dev to dev_priv in intel_guc_auth_huc(). Renamed HOST2GUC_ACTION_AUTHENTICATE_HUC TO INTEL_GUC_ACTION_ AUTHENTICATE_HUC

---
 drivers/gpu/drm/i915/intel_guc_fwif.h   |  1 +
 drivers/gpu/drm/i915/intel_guc_loader.c |  2 ++
 drivers/gpu/drm/i915/intel_uc.c         | 61 +++++++++++++++++++++++++++++++++
 drivers/gpu/drm/i915/intel_uc.h         |  1 +
 4 files changed, 65 insertions(+)
drivers/gpu/drm/i915/intel_guc_fwif.h +1 -0
index c1e7faf..94a974d 100644
@@ -504,6 +504,7 @@ enum intel_guc_action {
 	INTEL_GUC_ACTION_EXIT_S_STATE = 0x502,
 	INTEL_GUC_ACTION_SLPC_REQUEST = 0x3003,
 	INTEL_GUC_ACTION_UK_LOG_ENABLE_LOGGING = 0x0E000,
+	INTEL_GUC_ACTION_AUTHENTICATE_HUC = 0x4000,
 	INTEL_GUC_ACTION_LIMIT
 };
 
drivers/gpu/drm/i915/intel_guc_loader.c +2 -0
index b971351..89d092b 100644
@@ -529,6 +529,8 @@ int intel_guc_setup(struct drm_i915_private *dev_priv)
 		intel_uc_fw_status_repr(guc_fw->fetch_status),
 		intel_uc_fw_status_repr(guc_fw->load_status));
 
+	intel_guc_auth_huc(dev_priv);
+
 	if (i915.enable_guc_submission) {
 		if (i915.guc_log_level >= 0)
 			gen9_enable_guc_interrupts(dev_priv);
drivers/gpu/drm/i915/intel_uc.c +61 -0

64 lines not on this page · raw message

drivers/gpu/drm/i915/intel_uc.h +1 -0
index ac92946..1db8bc2 100644
@@ -196,6 +196,7 @@ int intel_guc_sample_forcewake(struct intel_guc *guc);
 int intel_guc_log_flush_complete(struct intel_guc *guc);
 int intel_guc_log_flush(struct intel_guc *guc);
 int intel_guc_log_control(struct intel_guc *guc, u32 control_val);
+void intel_guc_auth_huc(struct drm_i915_private *dev_priv);
 
 /* intel_guc_loader.c */
 extern void intel_guc_init(struct drm_i915_private *dev_priv);
Signature · 6 lines
2.7.4

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Reviewed-by on v1 7/8
#28 lore raw

Re: [PATCH 7/8] drm/i915/huc: Support HuC authentication

Quoted 22 lines · anushasr

From: Peter Antoine <[email protected]>

The HuC authentication is done by host2guc call. The HuC RSA keys are sent to GuC for authentication.

v2: rebased on top of drm-intel-nightly.
changed name format and upped version 1.7.
v3: rebased on top of drm-intel-nightly.
v4: changed wait_for_automic to wait_for
v5: rebased.
v7: rebased.
v8: rebased.
v9: rebased. Rename intel_huc_auh() to intel_guc_auth_huc() and place the prototype in intel_guc.h,correct the comments.
v10: rebased.
v11: rebased.
v12: rebased on top of drm-tip
v13: rebased. Moved intel_guc_auth_huc from i915_guc_submission.c to intel_uc.c.Update dev to dev_priv in intel_guc_auth_huc(). Renamed HOST2GUC_ACTION_AUTHENTICATE_HUC TO INTEL_GUC_ACTION_ AUTHENTICATE_HUC

Quoted 4 lines
Quoted 120 lines
---
 drivers/gpu/drm/i915/intel_guc_fwif.h   |  1 +
 drivers/gpu/drm/i915/intel_guc_loader.c |  2 ++
 drivers/gpu/drm/i915/intel_uc.c         | 61 +++++++++++++++++++++++++++++++++
 drivers/gpu/drm/i915/intel_uc.h         |  1 +
 4 files changed, 65 insertions(+)
drivers/gpu/drm/i915/intel_guc_fwif.h +1 -0
index c1e7faf..94a974d 100644
@@ -504,6 +504,7 @@ enum intel_guc_action {
 	INTEL_GUC_ACTION_EXIT_S_STATE = 0x502,
 	INTEL_GUC_ACTION_SLPC_REQUEST = 0x3003,
 	INTEL_GUC_ACTION_UK_LOG_ENABLE_LOGGING = 0x0E000,
+	INTEL_GUC_ACTION_AUTHENTICATE_HUC = 0x4000,
 	INTEL_GUC_ACTION_LIMIT
 };
 
drivers/gpu/drm/i915/intel_guc_loader.c +2 -0
index b971351..89d092b 100644
@@ -529,6 +529,8 @@ int intel_guc_setup(struct drm_i915_private *dev_priv)
 		intel_uc_fw_status_repr(guc_fw->fetch_status),
 		intel_uc_fw_status_repr(guc_fw->load_status));
 
+	intel_guc_auth_huc(dev_priv);
+
 	if (i915.enable_guc_submission) {
 		if (i915.guc_log_level >= 0)
 			gen9_enable_guc_interrupts(dev_priv);
drivers/gpu/drm/i915/intel_uc.c +61 -0
index 8ae6795..445b9ad 100644
@@ -138,3 +138,64 @@ int intel_guc_log_control(struct intel_guc *guc, u32 control_val)
 
 	return intel_guc_send(guc, action, ARRAY_SIZE(action));
 }
+
+/**
+ * intel_guc_auth_huc() - authenticate ucode
+ * @dev: the drm device
+ *
+ * Triggers a HuC fw authentication request to the GuC via host-2-guc
+ * interface.
+ */
+void intel_guc_auth_huc(struct drm_i915_private *dev_priv)
+{
+	struct intel_guc *guc = &dev_priv->guc;
+	struct intel_huc *huc = &dev_priv->huc;
+	struct i915_vma *vma;
+	int ret;
+	u32 data[2];
+
+	/* Bypass the case where there is no HuC firmware */
+	if (huc->huc_fw.fetch_status == UC_FIRMWARE_NONE ||
+		huc->huc_fw.load_status == UC_FIRMWARE_NONE)
+		return;
+
+	if (guc->guc_fw.load_status != UC_FIRMWARE_SUCCESS) {
+		DRM_ERROR("HuC: GuC fw wasn't loaded. Can't authenticate");
+		return;
+	}
+
+	if (huc->huc_fw.load_status != UC_FIRMWARE_SUCCESS) {
+		DRM_ERROR("HuC: fw wasn't loaded. Nothing to authenticate");
+		return;
+	}
+
+	vma = i915_gem_object_ggtt_pin(huc->huc_fw.uc_fw_obj, NULL, 0, 0, 0);
+	if (IS_ERR(vma)) {
+		DRM_DEBUG_DRIVER("pin failed %d\n", (int)PTR_ERR(vma));
+		return;
+	}
+
+
+	/* Invalidate GuC TLB to let GuC take the latest updates to GTT. */
+	I915_WRITE(GEN8_GTCR, GEN8_GTCR_INVALIDATE);
+
+	/* Specify auth action and where public signature is. */
+	data[0] = INTEL_GUC_ACTION_AUTHENTICATE_HUC;
+	data[1] = i915_ggtt_offset(vma) + huc->huc_fw.rsa_offset;
+
+	ret = intel_guc_send(guc, data, ARRAY_SIZE(data));
+	if (ret) {
+		DRM_ERROR("HuC: GuC did not ack Auth request\n");
+		goto out;
+	}
+
+	/* Check authentication status, it should be done by now */
+	ret = wait_for((I915_READ(HUC_STATUS2) & HUC_FW_VERIFIED) > 0, 50);
+	if (ret) {
+		DRM_ERROR("HuC: Authentication failed\n");
+		goto out;
+	}
+
+out:
+	i915_vma_unpin(vma);
+}
drivers/gpu/drm/i915/intel_uc.h +1 -0
index ac92946..1db8bc2 100644
@@ -196,6 +196,7 @@ int intel_guc_sample_forcewake(struct intel_guc *guc);
 int intel_guc_log_flush_complete(struct intel_guc *guc);
 int intel_guc_log_flush(struct intel_guc *guc);
 int intel_guc_log_control(struct intel_guc *guc, u32 control_val);
+void intel_guc_auth_huc(struct drm_i915_private *dev_priv);
 
 /* intel_guc_loader.c */
 extern void intel_guc_init(struct drm_i915_private *dev_priv);
Signature · 6 lines
2.7.4

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Signature · 6 lines
Cheers,
Arek
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
#29 lore raw

Re: [PATCH 7/8] drm/i915/huc: Support HuC authentication

In reply to anushasr
Quoted 63 lines · anushasr

From: Peter Antoine <[email protected]>

The HuC authentication is done by host2guc call. The HuC RSA keys are sent to GuC for authentication.

v2: rebased on top of drm-intel-nightly.
changed name format and upped version 1.7.
v3: rebased on top of drm-intel-nightly.
v4: changed wait_for_automic to wait_for
v5: rebased.
v7: rebased.
v8: rebased.
v9: rebased. Rename intel_huc_auh() to intel_guc_auth_huc() and place the prototype in intel_guc.h,correct the comments.
v10: rebased.
v11: rebased.
v12: rebased on top of drm-tip
v13: rebased. Moved intel_guc_auth_huc from i915_guc_submission.c to intel_uc.c.Update dev to dev_priv in intel_guc_auth_huc(). Renamed HOST2GUC_ACTION_AUTHENTICATE_HUC TO INTEL_GUC_ACTION_ AUTHENTICATE_HUC

---
 drivers/gpu/drm/i915/intel_guc_fwif.h   |  1 +
 drivers/gpu/drm/i915/intel_guc_loader.c |  2 ++
 drivers/gpu/drm/i915/intel_uc.c         | 61 +++++++++++++++++++++++++++++++++
 drivers/gpu/drm/i915/intel_uc.h         |  1 +
 4 files changed, 65 insertions(+)
drivers/gpu/drm/i915/intel_guc_fwif.h +1 -0
index c1e7faf..94a974d 100644
@@ -504,6 +504,7 @@ enum intel_guc_action {
 	INTEL_GUC_ACTION_EXIT_S_STATE = 0x502,
 	INTEL_GUC_ACTION_SLPC_REQUEST = 0x3003,
 	INTEL_GUC_ACTION_UK_LOG_ENABLE_LOGGING = 0x0E000,
+	INTEL_GUC_ACTION_AUTHENTICATE_HUC = 0x4000,
 	INTEL_GUC_ACTION_LIMIT
 };
 
drivers/gpu/drm/i915/intel_guc_loader.c +2 -0
index b971351..89d092b 100644
@@ -529,6 +529,8 @@ int intel_guc_setup(struct drm_i915_private *dev_priv)
 		intel_uc_fw_status_repr(guc_fw->fetch_status),
 		intel_uc_fw_status_repr(guc_fw->load_status));
 
+	intel_guc_auth_huc(dev_priv);
+
 	if (i915.enable_guc_submission) {
 		if (i915.guc_log_level >= 0)
 			gen9_enable_guc_interrupts(dev_priv);
drivers/gpu/drm/i915/intel_uc.c +0 -0
index 8ae6795..445b9ad 100644

drivers/gpu/drm/i915/intel_uc.c int intel_guc_log_control(struct intel_guc *guc, u32 control_val) · line 144

Quoted 7 lines
drivers/gpu/drm/i915/intel_uc.c +4 -0
@@ -138,3 +138,64 @@ int intel_guc_log_control(struct intel_guc *guc, u32 control_val)
 
 	return intel_guc_send(guc, action, ARRAY_SIZE(action));
 }
+
+/**
+ * intel_guc_auth_huc() - authenticate ucode
+ * @dev: the drm device

Mismatched param name.

drivers/gpu/drm/i915/intel_uc.c int intel_guc_log_control(struct intel_guc *guc, u32 control_val) · line 149

Quoted 5 lines
drivers/gpu/drm/i915/intel_uc.c +5 -0
+ *
+ * Triggers a HuC fw authentication request to the GuC via host-2-guc
+ * interface.
+ */
+void intel_guc_auth_huc(struct drm_i915_private *dev_priv)

Can we use *guc as the first param to match other intel_guc functions?

Quoted 4 lines
drivers/gpu/drm/i915/intel_uc.c +4 -0
+{
+	struct intel_guc *guc = &dev_priv->guc;
+	struct intel_huc *huc = &dev_priv->huc;
+	struct i915_vma *vma;

drivers/gpu/drm/i915/intel_uc.c int intel_guc_log_control(struct intel_guc *guc, u32 control_val) · line 163

Quoted 10 lines
drivers/gpu/drm/i915/intel_uc.c +10 -0
+	int ret;
+	u32 data[2];
+
+	/* Bypass the case where there is no HuC firmware */
+	if (huc->huc_fw.fetch_status == UC_FIRMWARE_NONE ||
+		huc->huc_fw.load_status == UC_FIRMWARE_NONE)
+		return;
+
+	if (guc->guc_fw.load_status != UC_FIRMWARE_SUCCESS) {
+		DRM_ERROR("HuC: GuC fw wasn't loaded. Can't authenticate");

Hmm, this looks like late handling of earlier error.
Note that other functions in this file assume that Guc is working fine.

Quoted 13 lines
drivers/gpu/drm/i915/intel_uc.c +13 -0
+		return;
+	}
+
+	if (huc->huc_fw.load_status != UC_FIRMWARE_SUCCESS) {
+		DRM_ERROR("HuC: fw wasn't loaded. Nothing to authenticate");
+		return;
+	}
+
+	vma = i915_gem_object_ggtt_pin(huc->huc_fw.uc_fw_obj, NULL, 0, 0, 0);
+	if (IS_ERR(vma)) {
+		DRM_DEBUG_DRIVER("pin failed %d\n", (int)PTR_ERR(vma));
+		return;
+	}

drivers/gpu/drm/i915/intel_uc.c int intel_guc_log_control(struct intel_guc *guc, u32 control_val) · line 186

Quoted 10 lines
drivers/gpu/drm/i915/intel_uc.c +10 -0
+
+
+	/* Invalidate GuC TLB to let GuC take the latest updates to GTT. */
+	I915_WRITE(GEN8_GTCR, GEN8_GTCR_INVALIDATE);
+
+	/* Specify auth action and where public signature is. */
+	data[0] = INTEL_GUC_ACTION_AUTHENTICATE_HUC;
+	data[1] = i915_ggtt_offset(vma) + huc->huc_fw.rsa_offset;
+
+	ret = intel_guc_send(guc, data, ARRAY_SIZE(data));

Hmm, maybe this function shall be split into two parts:

intel_huc_auth() in intel_huc_loader.c that contains most of the logic from current function, but calls intel_guc_auth_huc() from this file that just sends action to the guc (similar to other simple functions in this file.

Quoted 34 lines
drivers/gpu/drm/i915/intel_uc.c +15 -0
+	if (ret) {
+		DRM_ERROR("HuC: GuC did not ack Auth request\n");
+		goto out;
+	}
+
+	/* Check authentication status, it should be done by now */
+	ret = wait_for((I915_READ(HUC_STATUS2) & HUC_FW_VERIFIED) > 0, 50);
+	if (ret) {
+		DRM_ERROR("HuC: Authentication failed\n");
+		goto out;
+	}
+
+out:
+	i915_vma_unpin(vma);
+}
drivers/gpu/drm/i915/intel_uc.h +1 -0
index ac92946..1db8bc2 100644
@@ -196,6 +196,7 @@ int intel_guc_sample_forcewake(struct intel_guc *guc);
 int intel_guc_log_flush_complete(struct intel_guc *guc);
 int intel_guc_log_flush(struct intel_guc *guc);
 int intel_guc_log_control(struct intel_guc *guc, u32 control_val);
+void intel_guc_auth_huc(struct drm_i915_private *dev_priv);
 
 /* intel_guc_loader.c */
 extern void intel_guc_init(struct drm_i915_private *dev_priv);
Signature · 6 lines
2.7.4

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

#30 lore raw

Re: [PATCH 7/8] drm/i915/huc: Support HuC authentication

Quoted 88 lines

-----Original Message-----
From: Michal Wajdeczko [mailto:[email protected]]
Sent: Friday, December 9, 2016 4:37 AM
To: Srivatsa, Anusha <[email protected]>
Cc: [email protected]; Alex Dai <[email protected]>; Peter Antoine <[email protected]>
Subject: Re: [Intel-gfx] [PATCH 7/8] drm/i915/huc: Support HuC authentication

Quoted 22 lines · anushasr

From: Peter Antoine <[email protected]>

The HuC authentication is done by host2guc call. The HuC RSA keys are sent to GuC for authentication.

v2: rebased on top of drm-intel-nightly.
changed name format and upped version 1.7.
v3: rebased on top of drm-intel-nightly.
v4: changed wait_for_automic to wait_for
v5: rebased.
v7: rebased.
v8: rebased.
v9: rebased. Rename intel_huc_auh() to intel_guc_auth_huc() and place the prototype in intel_guc.h,correct the comments.
v10: rebased.
v11: rebased.
v12: rebased on top of drm-tip
v13: rebased. Moved intel_guc_auth_huc from i915_guc_submission.c to intel_uc.c.Update dev to dev_priv in intel_guc_auth_huc(). Renamed HOST2GUC_ACTION_AUTHENTICATE_HUC TO INTEL_GUC_ACTION_ AUTHENTICATE_HUC

Quoted 8 lines
---
 drivers/gpu/drm/i915/intel_guc_fwif.h   |  1 +
 drivers/gpu/drm/i915/intel_guc_loader.c |  2 ++
 drivers/gpu/drm/i915/intel_uc.c         | 61

+++++++++++++++++++++++++++++++++

Quoted 35 lines
 drivers/gpu/drm/i915/intel_uc.h         |  1 +
 4 files changed, 65 insertions(+)
drivers/gpu/drm/i915/intel_guc_fwif.h +1 -0
index c1e7faf..94a974d 100644
@@ -504,6 +504,7 @@ enum intel_guc_action {
 	INTEL_GUC_ACTION_EXIT_S_STATE = 0x502,
 	INTEL_GUC_ACTION_SLPC_REQUEST = 0x3003,
 	INTEL_GUC_ACTION_UK_LOG_ENABLE_LOGGING = 0x0E000,
+	INTEL_GUC_ACTION_AUTHENTICATE_HUC = 0x4000,
 	INTEL_GUC_ACTION_LIMIT
 };
 
drivers/gpu/drm/i915/intel_guc_loader.c +2 -0
index b971351..89d092b 100644
@@ -529,6 +529,8 @@ int intel_guc_setup(struct drm_i915_private *dev_priv)
 		intel_uc_fw_status_repr(guc_fw->fetch_status),
 		intel_uc_fw_status_repr(guc_fw->load_status));
 
+	intel_guc_auth_huc(dev_priv);
+
 	if (i915.enable_guc_submission) {
 		if (i915.guc_log_level >= 0)
 			gen9_enable_guc_interrupts(dev_priv);
(quoted diff) +0 -0

b/drivers/gpu/drm/i915/intel_uc.c index 8ae6795..445b9ad 100644

drivers/gpu/drm/i915/intel_uc.c int intel_guc_log_control(struct intel_guc *guc, · line 144

Quoted 7 lines
drivers/gpu/drm/i915/intel_uc.c +4 -0
@@ -138,3 +138,64 @@ int intel_guc_log_control(struct intel_guc *guc,
 u32 control_val)
 
 	return intel_guc_send(guc, action, ARRAY_SIZE(action));  }
+
+/**
+ * intel_guc_auth_huc() - authenticate ucode
+ * @dev: the drm device

Mismatched param name.

Quoted 8 lines

drivers/gpu/drm/i915/intel_uc.c int intel_guc_log_control(struct intel_guc *guc, u32 control_val) · line 149

Quoted 5 lines
drivers/gpu/drm/i915/intel_uc.c +5 -0
+ *
+ * Triggers a HuC fw authentication request to the GuC via host-2-guc
+ * interface.
+ */
+void intel_guc_auth_huc(struct drm_i915_private *dev_priv)

Can we use *guc as the first param to match other intel_guc functions?

In function intel_guc_init() and intel_guc_load() we are using dev_priv as the first parameter.....

Quoted 9 lines
Quoted 9 lines
drivers/gpu/drm/i915/intel_uc.c +9 -0
+{
+	struct intel_guc *guc = &dev_priv->guc;
+	struct intel_huc *huc = &dev_priv->huc;
+	struct i915_vma *vma;
+	int ret;
+	u32 data[2];
+
+	/* Bypass the case where there is no HuC firmware */
+	if (huc->huc_fw.fetch_status == UC_FIRMWARE_NONE ||
Quoted 8 lines

drivers/gpu/drm/i915/intel_uc.c int intel_guc_log_control(struct intel_guc *guc, u32 control_val) · line 163

Quoted 5 lines
drivers/gpu/drm/i915/intel_uc.c +5 -0
+		huc->huc_fw.load_status == UC_FIRMWARE_NONE)
+		return;
+
+	if (guc->guc_fw.load_status != UC_FIRMWARE_SUCCESS) {
+		DRM_ERROR("HuC: GuC fw wasn't loaded. Can't authenticate");

Hmm, this looks like late handling of earlier error.
Note that other functions in this file assume that Guc is working fine.

Michal, after intel_uc_init_early() which initializes a mutex lock, intel_guc_auth_huc() is the first function that the control goes to and hence all checking happens here. If anything is wrong a suitable error message is displayed and we return out immediately. If on the other hand things are proper, as expected post error checking, the other functions are called....

Hope this clears your concern.

-Anusha

Quoted 28 lines

drivers/gpu/drm/i915/intel_uc.c int intel_guc_log_control(struct intel_guc *guc, u32 control_val)

Quoted 10 lines
drivers/gpu/drm/i915/intel_uc.c +9 -0
+		return;
+	}
+
+	if (huc->huc_fw.load_status != UC_FIRMWARE_SUCCESS) {
+		DRM_ERROR("HuC: fw wasn't loaded. Nothing to authenticate");
+		return;
+	}
+
+	vma = i915_gem_object_ggtt_pin(huc->huc_fw.uc_fw_obj, NULL, 0, 0,

0);

Quoted 4 lines
drivers/gpu/drm/i915/intel_uc.c +4 -0
+	if (IS_ERR(vma)) {
+		DRM_DEBUG_DRIVER("pin failed %d\n", (int)PTR_ERR(vma));
+		return;
+	}

drivers/gpu/drm/i915/intel_uc.c int intel_guc_log_control(struct intel_guc *guc, u32 control_val) · line 186

Quoted 10 lines
drivers/gpu/drm/i915/intel_uc.c +10 -0
+
+
+	/* Invalidate GuC TLB to let GuC take the latest updates to GTT. */
+	I915_WRITE(GEN8_GTCR, GEN8_GTCR_INVALIDATE);
+
+	/* Specify auth action and where public signature is. */
+	data[0] = INTEL_GUC_ACTION_AUTHENTICATE_HUC;
+	data[1] = i915_ggtt_offset(vma) + huc->huc_fw.rsa_offset;
+
+	ret = intel_guc_send(guc, data, ARRAY_SIZE(data));

Hmm, maybe this function shall be split into two parts:

Quoted 4 lines

intel_huc_auth() in intel_huc_loader.c that contains most of the logic from current function, but calls intel_guc_auth_huc() from this file that just sends action to the guc (similar to other simple functions in this file.

The intention is to make intel_guc_auth_huc() as simple as other functions in the file?
Two points on my mind-
1. Wont splitting into intel_huc_auth() and intel_guc_auth_huc() be more confusing than just intel_guc_auth_huc?
2.Even on splitting we will have only the preliminary check - check if guc is loaded, check if huc is loaded in intel_huc_Auth. The actual specification of authentication action, sending it to guc, checking if that is a success or not has to happen in intel_huc_auth_guc(). Am I correct?

Quoted 26 lines

drivers/gpu/drm/i915/intel_uc.c int intel_guc_log_control(struct intel_guc *guc, u32 control_val)

Quoted 7 lines
drivers/gpu/drm/i915/intel_uc.c +7 -0
+	if (ret) {
+		DRM_ERROR("HuC: GuC did not ack Auth request\n");
+		goto out;
+	}
+
+	/* Check authentication status, it should be done by now */
+	ret = wait_for((I915_READ(HUC_STATUS2) & HUC_FW_VERIFIED) > 0,

50);

Quoted 18 lines
drivers/gpu/drm/i915/intel_uc.c +8 -0
+	if (ret) {
+		DRM_ERROR("HuC: Authentication failed\n");
+		goto out;
+	}
+
+out:
+	i915_vma_unpin(vma);
+}
drivers/gpu/drm/i915/intel_uc.h +1 -0
index ac92946..1db8bc2 100644
@@ -196,6 +196,7 @@ int intel_guc_sample_forcewake(struct intel_guc *guc);
 int intel_guc_log_flush_complete(struct intel_guc *guc);
 int intel_guc_log_flush(struct intel_guc *guc);
 int intel_guc_log_control(struct intel_guc *guc, u32 control_val);
+void intel_guc_auth_huc(struct drm_i915_private *dev_priv);
 
Quoted 5 lines
Quoted 5 lines
drivers/gpu/drm/i915/intel_uc.h +0 -1
 /* intel_guc_loader.c */
 extern void intel_guc_init(struct drm_i915_private *dev_priv);
--

2.7.4

Cheers,
Anusha

Quoted 4 lines
Quoted 4 lines

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

author
#31 lore raw

PATCH 8/8 drm/i915/get_params: Add HuC status to getparams

In reply to anushasr

From: Peter Antoine <[email protected]>

This patch will allow for getparams to return the status of the HuC. As the HuC has to be validated by the GuC this patch uses the validated status to show when the HuC is loaded and ready for use. You cannot use the loaded status as with the GuC as the HuC is verified after it is loaded and is not usable until it is verified.

v2: removed the forewakes as the registers are already force-woken.

     (T.Ursulin)

v4: rebased.
v5: rebased on top of drm-tip.
v6: rebased. Removed any reference to intel_huc.h

---
 drivers/gpu/drm/i915/i915_drv.c         |  4 ++++
 drivers/gpu/drm/i915/intel_huc_loader.c | 12 ++++++++++++
 drivers/gpu/drm/i915/intel_uc.h         |  1 +
 include/uapi/drm/i915_drm.h             |  1 +
 4 files changed, 18 insertions(+)
drivers/gpu/drm/i915/i915_drv.c +4 -0
index 85a47c2..6be06a27 100644
@@ -49,6 +49,7 @@
 #include "i915_trace.h"
 #include "i915_vgpu.h"
 #include "intel_drv.h"
+#include "intel_uc.h"
 
 static struct drm_driver driver;
 
@@ -349,6 +350,9 @@ static int i915_getparam(struct drm_device *dev, void *data,
 		 */
 		value = 1;
 		break;
+	case I915_PARAM_HAS_HUC:
+		value = intel_is_huc_valid(dev_priv);
+		break;
 	default:
 		DRM_DEBUG("Unknown parameter %d\n", param->param);
 		return -EINVAL;
drivers/gpu/drm/i915/intel_huc_loader.c +12 -0
index 96fc727..6704cc8 100644
@@ -289,3 +289,15 @@ void intel_huc_fini(struct drm_device *dev)
 	huc_fw->fetch_status = UC_FIRMWARE_NONE;
 }
 
+/**
+ * intel_is_huc_valid() - Check to see if the HuC is fully loaded.
+ * @dev_priv:	drm device to check.
+ *
+ * This function will return true if the guc has been loaded and
+ * has valid firmware. The simplest way of doing this is to check
+ * if the HuC has been validated, if so it must have been loaded.
+ */
+int intel_is_huc_valid(struct drm_i915_private *dev_priv)
+{
+	return ((I915_READ(HUC_STATUS2) & HUC_FW_VERIFIED) != 0);
+}
drivers/gpu/drm/i915/intel_uc.h +1 -0
index 1db8bc2..ccd3f69 100644
@@ -226,6 +226,7 @@ int i915_guc_log_control(struct drm_i915_private *dev_priv, u64 control_val);
 void intel_huc_init(struct drm_i915_private *dev_priv);
 void intel_huc_fini(struct drm_device *dev);
 int intel_huc_load(struct drm_i915_private *dev_priv);
+int intel_is_huc_valid(struct drm_i915_private *dev_priv);
 
 #endif
 #endif
include/uapi/drm/i915_drm.h +1 -0
index da32c2f..3e1964c 100644
@@ -395,6 +395,7 @@ typedef struct drm_i915_irq_wait {
  * priorities and the driver will attempt to execute batches in priority order.
  */
 #define I915_PARAM_HAS_SCHEDULER	 41
+#define I915_PARAM_HAS_HUC		 42
 
 typedef struct drm_i915_getparam {
 	__s32 param;
Signature · 6 lines
2.7.4

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
#32 lore raw

Re: [PATCH 8/8] drm/i915/get_params: Add HuC status to getparams

Quoted 36 lines · anushasr

From: Peter Antoine <[email protected]>

This patch will allow for getparams to return the status of the HuC. As the HuC has to be validated by the GuC this patch uses the validated status to show when the HuC is loaded and ready for use. You cannot use the loaded status as with the GuC as the HuC is verified after it is loaded and is not usable until it is verified.

v2: removed the forewakes as the registers are already force-woken.

     (T.Ursulin)

v4: rebased.
v5: rebased on top of drm-tip.
v6: rebased. Removed any reference to intel_huc.h

---
 drivers/gpu/drm/i915/i915_drv.c         |  4 ++++
 drivers/gpu/drm/i915/intel_huc_loader.c | 12 ++++++++++++
 drivers/gpu/drm/i915/intel_uc.h         |  1 +
 include/uapi/drm/i915_drm.h             |  1 +
 4 files changed, 18 insertions(+)
drivers/gpu/drm/i915/i915_drv.c +1 -0
index 85a47c2..6be06a27 100644
@@ -49,6 +49,7 @@
 #include "i915_trace.h"
 #include "i915_vgpu.h"
 #include "intel_drv.h"
+#include "intel_uc.h"
 
 static struct drm_driver driver;
 

drivers/gpu/drm/i915/i915_drv.c static int i915_getparam(struct drm_device *dev, void *data, · line 355

Quoted 6 lines
drivers/gpu/drm/i915/i915_drv.c +3 -0
@@ -349,6 +350,9 @@ static int i915_getparam(struct drm_device *dev, void *data,
 		 */
 		value = 1;
 		break;
+	case I915_PARAM_HAS_HUC:
+		value = intel_is_huc_valid(dev_priv);
+		break;

Why did you put it here? It breaks the pattern of case statements.

Quoted 10 lines
drivers/gpu/drm/i915/i915_drv.c +0 -0
 	default:
 		DRM_DEBUG("Unknown parameter %d\n", param->param);
 		return -EINVAL;
drivers/gpu/drm/i915/intel_huc_loader.c +0 -0
index 96fc727..6704cc8 100644
@@ -289,3 +289,15 @@ void intel_huc_fini(struct drm_device *dev)
 	huc_fw->fetch_status = UC_FIRMWARE_NONE;
 }

drivers/gpu/drm/i915/intel_huc_loader.c void intel_huc_fini(struct drm_device *dev) · line 300

Quoted 10 lines
drivers/gpu/drm/i915/intel_huc_loader.c +9 -0
 
+/**
+ * intel_is_huc_valid() - Check to see if the HuC is fully loaded.
+ * @dev_priv:	drm device to check.
+ *
+ * This function will return true if the guc has been loaded and
+ * has valid firmware. The simplest way of doing this is to check
+ * if the HuC has been validated, if so it must have been loaded.
+ */
+int intel_is_huc_valid(struct drm_i915_private *dev_priv)

bool

drivers/gpu/drm/i915/intel_huc_loader.c void intel_huc_fini(struct drm_device *dev) · line 302

Quoted 2 lines
drivers/gpu/drm/i915/intel_huc_loader.c +2 -0
+{
+	return ((I915_READ(HUC_STATUS2) & HUC_FW_VERIFIED) != 0);

(brackets (because (brackets)))

But what I really wanted to ask... Does this register access require the device to be awake and powered?
-Chris

Signature · 5 lines
Chris Wilson, Intel Open Source Technology Centre
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
#33 lore raw

Re: [PATCH 8/8] drm/i915/get_params: Add HuC status to getparams

Quoted 69 lines · Chris Wilson
Quoted 36 lines · anushasr

From: Peter Antoine <[email protected]>

This patch will allow for getparams to return the status of the HuC. As the HuC has to be validated by the GuC this patch uses the validated status to show when the HuC is loaded and ready for use. You cannot use the loaded status as with the GuC as the HuC is verified after it is loaded and is not usable until it is verified.

v2: removed the forewakes as the registers are already force-woken.

     (T.Ursulin)

v4: rebased.
v5: rebased on top of drm-tip.
v6: rebased. Removed any reference to intel_huc.h

---
 drivers/gpu/drm/i915/i915_drv.c         |  4 ++++
 drivers/gpu/drm/i915/intel_huc_loader.c | 12 ++++++++++++
 drivers/gpu/drm/i915/intel_uc.h         |  1 +
 include/uapi/drm/i915_drm.h             |  1 +
 4 files changed, 18 insertions(+)
drivers/gpu/drm/i915/i915_drv.c +1 -0
index 85a47c2..6be06a27 100644
@@ -49,6 +49,7 @@
 #include "i915_trace.h"
 #include "i915_vgpu.h"
 #include "intel_drv.h"
+#include "intel_uc.h"
 
 static struct drm_driver driver;
 

drivers/gpu/drm/i915/i915_drv.c static int i915_getparam(struct drm_device *dev, void *data, · line 355

Quoted 6 lines
drivers/gpu/drm/i915/i915_drv.c +3 -0
@@ -349,6 +350,9 @@ static int i915_getparam(struct drm_device *dev, void *data,
 		 */
 		value = 1;
 		break;
+	case I915_PARAM_HAS_HUC:
+		value = intel_is_huc_valid(dev_priv);
+		break;

Why did you put it here? It breaks the pattern of case statements.

Quoted 10 lines
drivers/gpu/drm/i915/i915_drv.c +0 -0
 	default:
 		DRM_DEBUG("Unknown parameter %d\n", param->param);
 		return -EINVAL;
drivers/gpu/drm/i915/intel_huc_loader.c +0 -0
index 96fc727..6704cc8 100644
@@ -289,3 +289,15 @@ void intel_huc_fini(struct drm_device *dev)
 	huc_fw->fetch_status = UC_FIRMWARE_NONE;
 }

drivers/gpu/drm/i915/intel_huc_loader.c void intel_huc_fini(struct drm_device *dev) · line 300

Quoted 10 lines
drivers/gpu/drm/i915/intel_huc_loader.c +9 -0
 
+/**
+ * intel_is_huc_valid() - Check to see if the HuC is fully loaded.
+ * @dev_priv:	drm device to check.
+ *
+ * This function will return true if the guc has been loaded and
+ * has valid firmware. The simplest way of doing this is to check
+ * if the HuC has been validated, if so it must have been loaded.
+ */
+int intel_is_huc_valid(struct drm_i915_private *dev_priv)

bool

Quoted 7 lines

drivers/gpu/drm/i915/intel_huc_loader.c void intel_huc_fini(struct drm_device *dev) · line 302

Quoted 2 lines
drivers/gpu/drm/i915/intel_huc_loader.c +2 -0
+{
+	return ((I915_READ(HUC_STATUS2) & HUC_FW_VERIFIED) != 0);

(brackets (because (brackets)))

But what I really wanted to ask... Does this register access require the device to be awake and powered?

Just confirmed that this register indeed requires MEDIA FW.

Quoted 1 line

-Chris

Signature · 6 lines
Cheers,
Arek
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
#34 lore raw

Re: [PATCH 8/8] drm/i915/get_params: Add HuC status to getparams

Quoted 87 lines

-----Original Message-----
From: Hiler, Arkadiusz
Sent: Tuesday, December 13, 2016 1:41 AM
To: Chris Wilson <[email protected]>; Srivatsa, Anusha <[email protected]>; [email protected]; Peter Antoine <[email protected]>
Subject: Re: [Intel-gfx] [PATCH 8/8] drm/i915/get_params: Add HuC status to getparams

Quoted 34 lines · Chris Wilson
Quoted 33 lines · anushasr

From: Peter Antoine <[email protected]>

This patch will allow for getparams to return the status of the HuC. As the HuC has to be validated by the GuC this patch uses the validated status to show when the HuC is loaded and ready for use. You cannot use the loaded status as with the GuC as the HuC is verified after it is loaded and is not usable until it is verified.

v2: removed the forewakes as the registers are already force-woken.

     (T.Ursulin)

v4: rebased.
v5: rebased on top of drm-tip.
v6: rebased. Removed any reference to intel_huc.h

---
 drivers/gpu/drm/i915/i915_drv.c         |  4 ++++
 drivers/gpu/drm/i915/intel_huc_loader.c | 12 ++++++++++++
 drivers/gpu/drm/i915/intel_uc.h         |  1 +
 include/uapi/drm/i915_drm.h             |  1 +
 4 files changed, 18 insertions(+)
(quoted diff) +0 -0

b/drivers/gpu/drm/i915/i915_drv.c index 85a47c2..6be06a27 100644

drivers/gpu/drm/i915/i915_drv.c +1 -0
@@ -49,6 +49,7 @@
 #include "i915_trace.h"
 #include "i915_vgpu.h"
 #include "intel_drv.h"
+#include "intel_uc.h"
 
Quoted 3 lines

drivers/gpu/drm/i915/i915_drv.c static int i915_getparam(struct drm_device *dev, void

Quoted 3 lines
drivers/gpu/drm/i915/i915_drv.c +0 -0
 static struct drm_driver driver;
 
@@ -349,6 +350,9 @@ static int i915_getparam(struct drm_device *dev, void

*data,

Quoted 37 lines

drivers/gpu/drm/i915/i915_drv.c static int i915_getparam(struct drm_device *dev, void *data, · line 355

Quoted 6 lines
drivers/gpu/drm/i915/i915_drv.c +3 -0
 		 */
 		value = 1;
 		break;
+	case I915_PARAM_HAS_HUC:
+		value = intel_is_huc_valid(dev_priv);
+		break;

Why did you put it here? It breaks the pattern of case statements.

Quoted 10 lines
drivers/gpu/drm/i915/i915_drv.c +0 -0
 	default:
 		DRM_DEBUG("Unknown parameter %d\n", param->param);
 		return -EINVAL;
drivers/gpu/drm/i915/intel_huc_loader.c +0 -0
index 96fc727..6704cc8 100644
@@ -289,3 +289,15 @@ void intel_huc_fini(struct drm_device *dev)
 	huc_fw->fetch_status = UC_FIRMWARE_NONE;  }

drivers/gpu/drm/i915/intel_huc_loader.c void intel_huc_fini(struct drm_device *dev) · line 299

Quoted 10 lines
drivers/gpu/drm/i915/intel_huc_loader.c +9 -0
 
+/**
+ * intel_is_huc_valid() - Check to see if the HuC is fully loaded.
+ * @dev_priv:	drm device to check.
+ *
+ * This function will return true if the guc has been loaded and
+ * has valid firmware. The simplest way of doing this is to check
+ * if the HuC has been validated, if so it must have been loaded.
+ */
+int intel_is_huc_valid(struct drm_i915_private *dev_priv)

bool

drivers/gpu/drm/i915/intel_huc_loader.c void intel_huc_fini(struct drm_device *dev) · line 302

Quoted 2 lines
drivers/gpu/drm/i915/intel_huc_loader.c +2 -0
+{
+	return ((I915_READ(HUC_STATUS2) & HUC_FW_VERIFIED) != 0);

(brackets (because (brackets)))

Quoted 4 lines
Quoted 2 lines

But what I really wanted to ask... Does this register access require the device to be awake and powered?

Just confirmed that this register indeed requires MEDIA FW.

Thanks Arek.

Quoted 5 lines
Quoted 1 line

-Chris

--
Cheers,
Arek

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

#35 lore raw

Re: [PATCH 8/8] drm/i915/get_params: Add HuC status to getparams

In reply to anushasr
Quoted 36 lines · anushasr

From: Peter Antoine <[email protected]>

This patch will allow for getparams to return the status of the HuC. As the HuC has to be validated by the GuC this patch uses the validated status to show when the HuC is loaded and ready for use. You cannot use the loaded status as with the GuC as the HuC is verified after it is loaded and is not usable until it is verified.

v2: removed the forewakes as the registers are already force-woken.

     (T.Ursulin)

v4: rebased.
v5: rebased on top of drm-tip.
v6: rebased. Removed any reference to intel_huc.h

---
 drivers/gpu/drm/i915/i915_drv.c         |  4 ++++
 drivers/gpu/drm/i915/intel_huc_loader.c | 12 ++++++++++++
 drivers/gpu/drm/i915/intel_uc.h         |  1 +
 include/uapi/drm/i915_drm.h             |  1 +
 4 files changed, 18 insertions(+)
drivers/gpu/drm/i915/i915_drv.c +1 -0
index 85a47c2..6be06a27 100644
@@ -49,6 +49,7 @@
 #include "i915_trace.h"
 #include "i915_vgpu.h"
 #include "intel_drv.h"
+#include "intel_uc.h"
 
 static struct drm_driver driver;
 

drivers/gpu/drm/i915/i915_drv.c static int i915_getparam(struct drm_device *dev, void *data, · line 353

Quoted 4 lines
drivers/gpu/drm/i915/i915_drv.c +1 -0
@@ -349,6 +350,9 @@ static int i915_getparam(struct drm_device *dev, void *data,
 		 */
 		value = 1;
 		break;
+	case I915_PARAM_HAS_HUC:

For me this param name does not match returned result which maybe misleading. Note that other HAS params return static driver/hw capability, not runtime.

I guess PARAM_HUC_STATUS would be better (0=no huc, 1=pending, 2=ok, -1=failed) And you can cache huc status in intel_huc struct and make final modification in intel_huc_auth() function to avoid registry read (unless we want to detect later crash of the huc using this reg read)

Quoted 10 lines
drivers/gpu/drm/i915/i915_drv.c +2 -0
+		value = intel_is_huc_valid(dev_priv);
+		break;
 	default:
 		DRM_DEBUG("Unknown parameter %d\n", param->param);
 		return -EINVAL;
drivers/gpu/drm/i915/intel_huc_loader.c +0 -0
index 96fc727..6704cc8 100644

drivers/gpu/drm/i915/intel_huc_loader.c void intel_huc_fini(struct drm_device *dev) · line 296

Quoted 8 lines
drivers/gpu/drm/i915/intel_huc_loader.c +5 -0
@@ -289,3 +289,15 @@ void intel_huc_fini(struct drm_device *dev)
 	huc_fw->fetch_status = UC_FIRMWARE_NONE;
 }
 
+/**
+ * intel_is_huc_valid() - Check to see if the HuC is fully loaded.
+ * @dev_priv:	drm device to check.
+ *
+ * This function will return true if the guc has been loaded and

Please change function return type to bool to match this description.

Quoted 38 lines
drivers/gpu/drm/i915/intel_huc_loader.c +7 -0
+ * has valid firmware. The simplest way of doing this is to check
+ * if the HuC has been validated, if so it must have been loaded.
+ */
+int intel_is_huc_valid(struct drm_i915_private *dev_priv)
+{
+	return ((I915_READ(HUC_STATUS2) & HUC_FW_VERIFIED) != 0);
+}
drivers/gpu/drm/i915/intel_uc.h +1 -0
index 1db8bc2..ccd3f69 100644
@@ -226,6 +226,7 @@ int i915_guc_log_control(struct drm_i915_private *dev_priv, u64 control_val);
 void intel_huc_init(struct drm_i915_private *dev_priv);
 void intel_huc_fini(struct drm_device *dev);
 int intel_huc_load(struct drm_i915_private *dev_priv);
+int intel_is_huc_valid(struct drm_i915_private *dev_priv);
 
 #endif
 #endif
include/uapi/drm/i915_drm.h +1 -0
index da32c2f..3e1964c 100644
@@ -395,6 +395,7 @@ typedef struct drm_i915_irq_wait {
  * priorities and the driver will attempt to execute batches in priority order.
  */
 #define I915_PARAM_HAS_SCHEDULER	 41
+#define I915_PARAM_HAS_HUC		 42
 
 typedef struct drm_i915_getparam {
 	__s32 param;
Signature · 6 lines
2.7.4

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

#36 lore raw

Re: [PATCH 8/8] drm/i915/get_params: Add HuC status to getparams

Quoted 45 lines · Michal Wajdeczko
Quoted 36 lines · anushasr

From: Peter Antoine <[email protected]>

This patch will allow for getparams to return the status of the HuC. As the HuC has to be validated by the GuC this patch uses the validated status to show when the HuC is loaded and ready for use. You cannot use the loaded status as with the GuC as the HuC is verified after it is loaded and is not usable until it is verified.

v2: removed the forewakes as the registers are already force-woken.

     (T.Ursulin)

v4: rebased.
v5: rebased on top of drm-tip.
v6: rebased. Removed any reference to intel_huc.h

---
 drivers/gpu/drm/i915/i915_drv.c         |  4 ++++
 drivers/gpu/drm/i915/intel_huc_loader.c | 12 ++++++++++++
 drivers/gpu/drm/i915/intel_uc.h         |  1 +
 include/uapi/drm/i915_drm.h             |  1 +
 4 files changed, 18 insertions(+)
drivers/gpu/drm/i915/i915_drv.c +1 -0
index 85a47c2..6be06a27 100644
@@ -49,6 +49,7 @@
 #include "i915_trace.h"
 #include "i915_vgpu.h"
 #include "intel_drv.h"
+#include "intel_uc.h"
 
 static struct drm_driver driver;
 

drivers/gpu/drm/i915/i915_drv.c static int i915_getparam(struct drm_device *dev, void *data, · line 353

Quoted 4 lines
drivers/gpu/drm/i915/i915_drv.c +1 -0
@@ -349,6 +350,9 @@ static int i915_getparam(struct drm_device *dev, void *data,
 		 */
 		value = 1;
 		break;
+	case I915_PARAM_HAS_HUC:

For me this param name does not match returned result which maybe misleading. Note that other HAS params return static driver/hw capability, not runtime.

Quoted 4 lines

I guess PARAM_HUC_STATUS would be better (0=no huc, 1=pending, 2=ok, -1=failed) And you can cache huc status in intel_huc struct and make final modification in intel_huc_auth() function to avoid registry read (unless we want to detect later crash of the huc using this reg read)

Why should userspace care for those intermediary states? From what I know (docs and patches from the cover letter) userspace is interested only in being able to use HuC. If something is not working you have DebugFS for that exact purpose.

As of cacheing - seems like a good idea to limit reg reads.

Quoted 63 lines
Quoted 10 lines
drivers/gpu/drm/i915/i915_drv.c +2 -0
+		value = intel_is_huc_valid(dev_priv);
+		break;
 	default:
 		DRM_DEBUG("Unknown parameter %d\n", param->param);
 		return -EINVAL;
drivers/gpu/drm/i915/intel_huc_loader.c +0 -0
index 96fc727..6704cc8 100644

drivers/gpu/drm/i915/intel_huc_loader.c void intel_huc_fini(struct drm_device *dev) · line 296

Quoted 8 lines
drivers/gpu/drm/i915/intel_huc_loader.c +5 -0
@@ -289,3 +289,15 @@ void intel_huc_fini(struct drm_device *dev)
 	huc_fw->fetch_status = UC_FIRMWARE_NONE;
 }
 
+/**
+ * intel_is_huc_valid() - Check to see if the HuC is fully loaded.
+ * @dev_priv:	drm device to check.
+ *
+ * This function will return true if the guc has been loaded and

Please change function return type to bool to match this description.

Quoted 38 lines
drivers/gpu/drm/i915/intel_huc_loader.c +7 -0
+ * has valid firmware. The simplest way of doing this is to check
+ * if the HuC has been validated, if so it must have been loaded.
+ */
+int intel_is_huc_valid(struct drm_i915_private *dev_priv)
+{
+	return ((I915_READ(HUC_STATUS2) & HUC_FW_VERIFIED) != 0);
+}
drivers/gpu/drm/i915/intel_uc.h +1 -0
index 1db8bc2..ccd3f69 100644
@@ -226,6 +226,7 @@ int i915_guc_log_control(struct drm_i915_private *dev_priv, u64 control_val);
 void intel_huc_init(struct drm_i915_private *dev_priv);
 void intel_huc_fini(struct drm_device *dev);
 int intel_huc_load(struct drm_i915_private *dev_priv);
+int intel_is_huc_valid(struct drm_i915_private *dev_priv);
 
 #endif
 #endif
include/uapi/drm/i915_drm.h +1 -0
index da32c2f..3e1964c 100644
@@ -395,6 +395,7 @@ typedef struct drm_i915_irq_wait {
  * priorities and the driver will attempt to execute batches in priority order.
  */
 #define I915_PARAM_HAS_SCHEDULER	 41
+#define I915_PARAM_HAS_HUC		 42
 
 typedef struct drm_i915_getparam {
 	__s32 param;
Signature · 6 lines
2.7.4

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Signature · 6 lines
Cheers,
Arek
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
#37 lore raw

Re: [PATCH 8/8] drm/i915/get_params: Add HuC status to getparams

Quoted 51 lines · Arkadiusz Hiler
Quoted 45 lines · Michal Wajdeczko
Quoted 36 lines · anushasr

From: Peter Antoine <[email protected]>

This patch will allow for getparams to return the status of the HuC. As the HuC has to be validated by the GuC this patch uses the validated status to show when the HuC is loaded and ready for use. You cannot use the loaded status as with the GuC as the HuC is verified after it is loaded and is not usable until it is verified.

v2: removed the forewakes as the registers are already force-woken.

     (T.Ursulin)

v4: rebased.
v5: rebased on top of drm-tip.
v6: rebased. Removed any reference to intel_huc.h

---
 drivers/gpu/drm/i915/i915_drv.c         |  4 ++++
 drivers/gpu/drm/i915/intel_huc_loader.c | 12 ++++++++++++
 drivers/gpu/drm/i915/intel_uc.h         |  1 +
 include/uapi/drm/i915_drm.h             |  1 +
 4 files changed, 18 insertions(+)
drivers/gpu/drm/i915/i915_drv.c +1 -0
index 85a47c2..6be06a27 100644
@@ -49,6 +49,7 @@
 #include "i915_trace.h"
 #include "i915_vgpu.h"
 #include "intel_drv.h"
+#include "intel_uc.h"
 
 static struct drm_driver driver;
 

drivers/gpu/drm/i915/i915_drv.c static int i915_getparam(struct drm_device *dev, void *data, · line 353

Quoted 4 lines
drivers/gpu/drm/i915/i915_drv.c +1 -0
@@ -349,6 +350,9 @@ static int i915_getparam(struct drm_device *dev, void *data,
 		 */
 		value = 1;
 		break;
+	case I915_PARAM_HAS_HUC:

For me this param name does not match returned result which maybe misleading. Note that other HAS params return static driver/hw capability, not runtime.

Quoted 4 lines

I guess PARAM_HUC_STATUS would be better (0=no huc, 1=pending, 2=ok, -1=failed) And you can cache huc status in intel_huc struct and make final modification in intel_huc_auth() function to avoid registry read (unless we want to detect later crash of the huc using this reg read)

Quoted 6 lines

Why should userspace care for those intermediary states? From what I know (docs and patches from the cover letter) userspace is interested only in being able to use HuC. If something is not working you have DebugFS for that exact purpose.

As of cacheing - seems like a good idea to limit reg reads.

Is GETPARAM(HAS_HUC) a hot path? Should we even encourage it?

Are there any other users of intel_huc_is_valid()?

As for userspace simply asking where huc is enabled, we already have that in the ABI via the module parameter, so you need to justify why this is preferred (in addition to the available information). -Chris

Signature · 5 lines
Chris Wilson, Intel Open Source Technology Centre
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
#38 lore raw

Re: [PATCH 8/8] drm/i915/get_params: Add HuC status to getparams

Quoted 52 lines · Chris Wilson
Quoted 51 lines · Arkadiusz Hiler
Quoted 49 lines · Michal Wajdeczko
Quoted 36 lines · anushasr

From: Peter Antoine <[email protected]>

This patch will allow for getparams to return the status of the HuC. As the HuC has to be validated by the GuC this patch uses the validated status to show when the HuC is loaded and ready for use. You cannot use the loaded status as with the GuC as the HuC is verified after it is loaded and is not usable until it is verified.

v2: removed the forewakes as the registers are already force-woken.

     (T.Ursulin)

v4: rebased.
v5: rebased on top of drm-tip.
v6: rebased. Removed any reference to intel_huc.h

---
 drivers/gpu/drm/i915/i915_drv.c         |  4 ++++
 drivers/gpu/drm/i915/intel_huc_loader.c | 12 ++++++++++++
 drivers/gpu/drm/i915/intel_uc.h         |  1 +
 include/uapi/drm/i915_drm.h             |  1 +
 4 files changed, 18 insertions(+)
drivers/gpu/drm/i915/i915_drv.c +1 -0
index 85a47c2..6be06a27 100644
@@ -49,6 +49,7 @@
 #include "i915_trace.h"
 #include "i915_vgpu.h"
 #include "intel_drv.h"
+#include "intel_uc.h"
 
 static struct drm_driver driver;
 

drivers/gpu/drm/i915/i915_drv.c static int i915_getparam(struct drm_device *dev, void *data, · line 353

Quoted 4 lines
drivers/gpu/drm/i915/i915_drv.c +1 -0
@@ -349,6 +350,9 @@ static int i915_getparam(struct drm_device *dev, void *data,
 		 */
 		value = 1;
 		break;
+	case I915_PARAM_HAS_HUC:

For me this param name does not match returned result which maybe misleading. Note that other HAS params return static driver/hw capability, not runtime.

I guess PARAM_HUC_STATUS would be better (0=no huc, 1=pending, 2=ok, -1=failed) And you can cache huc status in intel_huc struct and make final modification in intel_huc_auth() function to avoid registry read (unless we want to detect later crash of the huc using this reg read)

Quoted 8 lines
Quoted 6 lines

Why should userspace care for those intermediary states? From what I know (docs and patches from the cover letter) userspace is interested only in being able to use HuC. If something is not working you have DebugFS for that exact purpose.

As of cacheing - seems like a good idea to limit reg reads.

Is GETPARAM(HAS_HUC) a hot path? Should we even encourage it?

Actually... Good point. HAS_HUC is used once each time you initialise libva client.

Quoted 1 line

Are there any other users of intel_huc_is_valid()?

intel_guc_auth_huc() uses
ret = wait_for((I915_READ(HUC_STATUS2) & HUC_FW_VERIFIED) > 0, 50);

So it can be reused there.

Quoted 3 lines

As for userspace simply asking where huc is enabled, we already have that in the ABI via the module parameter, so you need to justify why this is preferred (in addition to the available information).

Yeah, we do change the values of module parameters. But a lot of them are uid 0 only and we have PARAMS for those.

Do anything userspace use those actually? Do we plan to use them instead of the getparams since now on?

Quoted 4 lines

-Chris

Signature · 1 line
Chris Wilson, Intel Open Source Technology Centre
Signature · 6 lines
Cheers,
Arek
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
#39 lore raw

Re: [PATCH 8/8] drm/i915/get_params: Add HuC status to getparams

Quoted 10 lines · Arkadiusz Hiler
Quoted 3 lines · Chris Wilson

As for userspace simply asking where huc is enabled, we already have that in the ABI via the module parameter, so you need to justify why this is preferred (in addition to the available information).

Yeah, we do change the values of module parameters. But a lot of them are uid 0 only and we have PARAMS for those.

Do anything userspace use those actually? Do we plan to use them instead of the getparams since now on?

I've done both from userspace... I don't really have a preference, once you have an fd, an ioctl/GETPARAM is trivial. But for quick querying and reporting of driver state, reading the module parameter is easier. So I have a tendency to use an ioctl in production code and module parameter in igt. (And I would say that for one-off validation purposes, i.e. did hte module actually enable huc?, just check the module parameter. If userspace is expected to interact and respond to the setting during its early probe, make it an ioctl so it fits in with the similar code also being run at that time.)

It is just worth explaining the intended usecase in the cover note, so that the use can be sanity checked and reflected upon later if need be. -Chris

Signature · 5 lines
Chris Wilson, Intel Open Source Technology Centre
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
#40 lore raw

Re: [PATCH 8/8] drm/i915/get_params: Add HuC status to getparams

Quoted 12 lines · Chris Wilson
Quoted 10 lines · Arkadiusz Hiler
Quoted 3 lines · Chris Wilson

As for userspace simply asking where huc is enabled, we already have that in the ABI via the module parameter, so you need to justify why this is preferred (in addition to the available information).

Yeah, we do change the values of module parameters. But a lot of them are uid 0 only and we have PARAMS for those.

Do anything userspace use those actually? Do we plan to use them instead of the getparams since now on?

Quoted 6 lines

I've done both from userspace... I don't really have a preference, once you have an fd, an ioctl/GETPARAM is trivial. But for quick querying and reporting of driver state, reading the module parameter is easier. So I have a tendency to use an ioctl in production code and module parameter in igt. (And I would say that for one-off validation purposes, i.e. did hte module actually enable huc?, just check the module parameter.

Thank you for explaining it :)

Quoted 3 lines

If userspace is expected to interact and respond to the setting during its early probe, make it an ioctl so it fits in with the similar code also being run at that time.)

Okay. Seems like the scenario Anusha's got.

Quoted 3 lines

It is just worth explaining the intended usecase in the cover note, so that the use can be sanity checked and reflected upon later if need be. -Chris

The cover letter links patches from the libva which is pretty much a template usecase, but I agree, it wold be nice if it used words/pseudocode instead of portions of code from some other project.

Quoted 2 lines
Signature · 1 line
Chris Wilson, Intel Open Source Technology Centre
Signature · 6 lines
Cheers,
Arek
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
#41 lore raw

Re: [PATCH 8/8] drm/i915/get_params: Add HuC status to getparams

In reply to Chris Wilson
Quoted 33 lines

-----Original Message-----
From: Intel-gfx [mailto:[email protected]] On Behalf Of Chris Wilson
Sent: Monday, December 12, 2016 7:17 AM
To: Hiler, Arkadiusz <[email protected]>
Cc: [email protected]; Michal Wajdeczko <[email protected]>
Subject: Re: [Intel-gfx] [PATCH 8/8] drm/i915/get_params: Add HuC status to getparams

Quoted 10 lines · Arkadiusz Hiler
Quoted 3 lines · Chris Wilson

As for userspace simply asking where huc is enabled, we already have that in the ABI via the module parameter, so you need to justify why this is preferred (in addition to the available information).

Yeah, we do change the values of module parameters. But a lot of them are uid 0 only and we have PARAMS for those.

Do anything userspace use those actually? Do we plan to use them instead of the getparams since now on?

I've done both from userspace... I don't really have a preference, once you have an fd, an ioctl/GETPARAM is trivial. But for quick querying and reporting of driver state, reading the module parameter is easier. So I have a tendency to use an ioctl in production code and module parameter in igt. (And I would say that for one-off validation purposes, i.e. did hte module actually enable huc?, just check the module parameter. If userspace is expected to interact and respond to the setting during its early probe, make it an ioctl so it fits in with the similar code also being run at that time.)

Quoted 4 lines

It is just worth explaining the intended usecase in the cover note, so that the use can be sanity checked and reflected upon later if need be.
-Chris

Thanks for explaining Chris....

Anusha

Quoted 5 lines

Chris Wilson, Intel Open Source Technology Centre
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

#42 lore raw

✓ Fi.CI.BAT: success for HuC Loading Patches

In reply to anushasr
== Series Details ==

== Series Details ==

Series: HuC Loading Patches
URL : https://patchwork.freedesktop.org/series/16584/
State : success

== Summary ==

Series 16584v1 HuC Loading Patches
https://patchwork.freedesktop.org/api/1.0/series/16584/revisions/1/mbox/

fi-bdw-5557u     total:247  pass:233  dwarn:0   dfail:0   fail:0   skip:14 
fi-bsw-n3050     total:247  pass:208  dwarn:0   dfail:0   fail:0   skip:39 
fi-bxt-t5700     total:247  pass:220  dwarn:0   dfail:0   fail:0   skip:27 
fi-byt-j1900     total:247  pass:220  dwarn:0   dfail:0   fail:0   skip:27 
fi-byt-n2820     total:247  pass:216  dwarn:0   dfail:0   fail:0   skip:31 
fi-hsw-4770      total:247  pass:228  dwarn:0   dfail:0   fail:0   skip:19 
fi-hsw-4770r     total:247  pass:228  dwarn:0   dfail:0   fail:0   skip:19 
fi-ilk-650       total:247  pass:195  dwarn:0   dfail:0   fail:0   skip:52 
fi-ivb-3520m     total:247  pass:226  dwarn:0   dfail:0   fail:0   skip:21 
fi-ivb-3770      total:247  pass:226  dwarn:0   dfail:0   fail:0   skip:21 
fi-kbl-7500u     total:247  pass:226  dwarn:0   dfail:0   fail:0   skip:21 
fi-skl-6260u     total:247  pass:234  dwarn:0   dfail:0   fail:0   skip:13 
fi-skl-6700hq    total:247  pass:227  dwarn:0   dfail:0   fail:0   skip:20 
fi-skl-6700k     total:247  pass:224  dwarn:3   dfail:0   fail:0   skip:20 
fi-skl-6770hq    total:247  pass:234  dwarn:0   dfail:0   fail:0   skip:13 
fi-snb-2600      total:247  pass:215  dwarn:0   dfail:0   fail:0   skip:32 

a967a5db05519675cd3a0f419fba30814ed457eb drm-tip: 2016y-12m-08d-20h-50m-10s UTC integration manifest b50355c drm/i915/get_params: Add HuC status to getparams
0c33af1 drm/i915/huc: Support HuC authentication
8c9354e drm/i915/huc: Add debugfs for HuC loading status check
73b8768 drm/i915/HuC: Add KBL huC loading Support
0ef1bd0 drm/i915/huc: Add BXT HuC Loading Support
70b8dd7 drm/i915/huc: Add HuC fw loading support
1dec5f7 drm/i915/huc: Unified css_header struct for GuC and HuC
5941062 drm/i915/guc: Make the GuC fw loading helper functions general

== Logs ==

For more details see: https://intel-gfx-ci.01.org/CI/Patchwork_3245/
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx