From e4fa8457b2197118538a1400b75c898f9faaf164 Mon Sep 17 00:00:00 2001 From: Daniel Vetter Date: Fri, 14 Jun 2019 22:35:25 +0200 Subject: drm/prime: Align gem_prime_export with obj_funcs.export MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The idea is that gem_prime_export is deprecated in favor of obj_funcs.export. That's much easier to do if both have matching function signatures. Reviewed-by: Eric Anholt Reviewed-by: Emil Velikov Acked-by: Christian König Acked-by: Thierry Reding Signed-off-by: Daniel Vetter Cc: Russell King Cc: Maarten Lankhorst Cc: Maxime Ripard Cc: Sean Paul Cc: David Airlie Cc: Daniel Vetter Cc: Zhenyu Wang Cc: Zhi Wang Cc: Jani Nikula Cc: Joonas Lahtinen Cc: Rodrigo Vivi Cc: Tomi Valkeinen Cc: Alex Deucher Cc: "Christian König" Cc: "David (ChunMing) Zhou" Cc: Thierry Reding Cc: Jonathan Hunter Cc: Dave Airlie Cc: Eric Anholt Cc: "Michel Dänzer" Cc: Chris Wilson Cc: Huang Rui Cc: Felix Kuehling Cc: Hawking Zhang Cc: Feifei Xu Cc: Jim Qu Cc: Evan Quan Cc: Matthew Auld Cc: Mika Kuoppala Cc: Thomas Zimmermann Cc: Kate Stewart Cc: Sumit Semwal Cc: Jilayne Lovejoy Cc: Thomas Gleixner Cc: Mikulas Patocka Cc: Greg Kroah-Hartman Cc: Junwei Zhang Cc: intel-gvt-dev@lists.freedesktop.org Cc: intel-gfx@lists.freedesktop.org Cc: amd-gfx@lists.freedesktop.org Cc: linux-tegra@vger.kernel.org Link: https://patchwork.freedesktop.org/patch/msgid/20190614203615.12639-10-daniel.vetter@ffwll.ch --- drivers/gpu/drm/i915/gvt/dmabuf.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers/gpu/drm/i915/gvt') diff --git a/drivers/gpu/drm/i915/gvt/dmabuf.c b/drivers/gpu/drm/i915/gvt/dmabuf.c index 41c8ebc60c63..13044c027f27 100644 --- a/drivers/gpu/drm/i915/gvt/dmabuf.c +++ b/drivers/gpu/drm/i915/gvt/dmabuf.c @@ -491,7 +491,7 @@ int intel_vgpu_get_dmabuf(struct intel_vgpu *vgpu, unsigned int dmabuf_id) obj->gvt_info = dmabuf_obj->info; - dmabuf = i915_gem_prime_export(dev, &obj->base, DRM_CLOEXEC | DRM_RDWR); + dmabuf = i915_gem_prime_export(&obj->base, DRM_CLOEXEC | DRM_RDWR); if (IS_ERR(dmabuf)) { gvt_vgpu_err("export dma-buf failed\n"); ret = PTR_ERR(dmabuf); -- cgit v1.2.3 From b5893ffc274be966f95aa35f35916fa8725af154 Mon Sep 17 00:00:00 2001 From: Janusz Krzysztofik Date: Fri, 12 Jul 2019 13:24:25 +0200 Subject: drm/i915: Drop extern qualifiers from header function prototypes Follow dim checkpatch recommendation so it doesn't complain on that now and again on header file modifications. v2: drop testing leftover (Chris) Signed-off-by: Janusz Krzysztofik Reviewed-by: Chris Wilson Signed-off-by: Chris Wilson Link: https://patchwork.freedesktop.org/patch/msgid/20190712112429.740-2-janusz.krzysztofik@linux.intel.com --- drivers/gpu/drm/i915/gem/i915_gem_object.h | 2 +- drivers/gpu/drm/i915/gvt/gtt.h | 13 +++++---- drivers/gpu/drm/i915/i915_drv.h | 47 ++++++++++++++---------------- drivers/gpu/drm/i915/i915_irq.h | 4 +-- drivers/gpu/drm/i915/oa/i915_oa_bdw.h | 2 +- drivers/gpu/drm/i915/oa/i915_oa_bxt.h | 2 +- drivers/gpu/drm/i915/oa/i915_oa_cflgt2.h | 2 +- drivers/gpu/drm/i915/oa/i915_oa_cflgt3.h | 2 +- drivers/gpu/drm/i915/oa/i915_oa_chv.h | 2 +- drivers/gpu/drm/i915/oa/i915_oa_cnl.h | 2 +- drivers/gpu/drm/i915/oa/i915_oa_glk.h | 2 +- drivers/gpu/drm/i915/oa/i915_oa_hsw.h | 2 +- drivers/gpu/drm/i915/oa/i915_oa_icl.h | 2 +- drivers/gpu/drm/i915/oa/i915_oa_kblgt2.h | 2 +- drivers/gpu/drm/i915/oa/i915_oa_kblgt3.h | 2 +- drivers/gpu/drm/i915/oa/i915_oa_sklgt2.h | 2 +- drivers/gpu/drm/i915/oa/i915_oa_sklgt3.h | 2 +- drivers/gpu/drm/i915/oa/i915_oa_sklgt4.h | 2 +- include/drm/i915_drm.h | 10 +++---- 19 files changed, 51 insertions(+), 53 deletions(-) (limited to 'drivers/gpu/drm/i915/gvt') diff --git a/drivers/gpu/drm/i915/gem/i915_gem_object.h b/drivers/gpu/drm/i915/gem/i915_gem_object.h index 20754c15412a..67aea07ea019 100644 --- a/drivers/gpu/drm/i915/gem/i915_gem_object.h +++ b/drivers/gpu/drm/i915/gem/i915_gem_object.h @@ -81,7 +81,7 @@ i915_gem_object_lookup(struct drm_file *file, u32 handle) } __deprecated -extern struct drm_gem_object * +struct drm_gem_object * drm_gem_object_lookup(struct drm_file *file, u32 handle); __attribute__((nonnull)) diff --git a/drivers/gpu/drm/i915/gvt/gtt.h b/drivers/gpu/drm/i915/gvt/gtt.h index 42d0394f0de2..88789316807d 100644 --- a/drivers/gpu/drm/i915/gvt/gtt.h +++ b/drivers/gpu/drm/i915/gvt/gtt.h @@ -205,17 +205,18 @@ struct intel_vgpu_gtt { struct intel_vgpu_scratch_pt scratch_pt[GTT_TYPE_MAX]; }; -extern int intel_vgpu_init_gtt(struct intel_vgpu *vgpu); -extern void intel_vgpu_clean_gtt(struct intel_vgpu *vgpu); +int intel_vgpu_init_gtt(struct intel_vgpu *vgpu); +void intel_vgpu_clean_gtt(struct intel_vgpu *vgpu); void intel_vgpu_reset_ggtt(struct intel_vgpu *vgpu, bool invalidate_old); void intel_vgpu_invalidate_ppgtt(struct intel_vgpu *vgpu); -extern int intel_gvt_init_gtt(struct intel_gvt *gvt); +int intel_gvt_init_gtt(struct intel_gvt *gvt); void intel_vgpu_reset_gtt(struct intel_vgpu *vgpu); -extern void intel_gvt_clean_gtt(struct intel_gvt *gvt); +void intel_gvt_clean_gtt(struct intel_gvt *gvt); -extern struct intel_vgpu_mm *intel_gvt_find_ppgtt_mm(struct intel_vgpu *vgpu, - int page_table_level, void *root_entry); +struct intel_vgpu_mm *intel_gvt_find_ppgtt_mm(struct intel_vgpu *vgpu, + int page_table_level, + void *root_entry); struct intel_vgpu_oos_page { struct intel_vgpu_ppgtt_spt *spt; diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h index 56527a7a1666..1a6b4e14a405 100644 --- a/drivers/gpu/drm/i915/i915_drv.h +++ b/drivers/gpu/drm/i915/i915_drv.h @@ -2390,19 +2390,17 @@ __i915_printk(struct drm_i915_private *dev_priv, const char *level, __i915_printk(dev_priv, KERN_ERR, fmt, ##__VA_ARGS__) #ifdef CONFIG_COMPAT -extern long i915_compat_ioctl(struct file *filp, unsigned int cmd, - unsigned long arg); +long i915_compat_ioctl(struct file *filp, unsigned int cmd, unsigned long arg); #else #define i915_compat_ioctl NULL #endif extern const struct dev_pm_ops i915_pm_ops; -extern int i915_driver_load(struct pci_dev *pdev, - const struct pci_device_id *ent); -extern void i915_driver_unload(struct drm_device *dev); +int i915_driver_load(struct pci_dev *pdev, const struct pci_device_id *ent); +void i915_driver_unload(struct drm_device *dev); -extern void intel_engine_init_hangcheck(struct intel_engine_cs *engine); -extern void intel_hangcheck_init(struct drm_i915_private *dev_priv); +void intel_engine_init_hangcheck(struct intel_engine_cs *engine); +void intel_hangcheck_init(struct drm_i915_private *dev_priv); int vlv_force_gfx_clock(struct drm_i915_private *dev_priv, bool on); u32 intel_calculate_mcr_s_ss_select(struct drm_i915_private *dev_priv); @@ -2672,14 +2670,14 @@ int intel_engine_cmd_parser(struct intel_engine_cs *engine, bool is_master); /* i915_perf.c */ -extern void i915_perf_init(struct drm_i915_private *dev_priv); -extern void i915_perf_fini(struct drm_i915_private *dev_priv); -extern void i915_perf_register(struct drm_i915_private *dev_priv); -extern void i915_perf_unregister(struct drm_i915_private *dev_priv); +void i915_perf_init(struct drm_i915_private *dev_priv); +void i915_perf_fini(struct drm_i915_private *dev_priv); +void i915_perf_register(struct drm_i915_private *dev_priv); +void i915_perf_unregister(struct drm_i915_private *dev_priv); /* i915_suspend.c */ -extern int i915_save_state(struct drm_i915_private *dev_priv); -extern int i915_restore_state(struct drm_i915_private *dev_priv); +int i915_save_state(struct drm_i915_private *dev_priv); +int i915_restore_state(struct drm_i915_private *dev_priv); /* i915_sysfs.c */ void i915_setup_sysfs(struct drm_i915_private *dev_priv); @@ -2693,23 +2691,22 @@ mkwrite_device_info(struct drm_i915_private *dev_priv) } /* modesetting */ -extern void intel_modeset_init_hw(struct drm_device *dev); -extern int intel_modeset_init(struct drm_device *dev); -extern void intel_modeset_cleanup(struct drm_device *dev); -extern int intel_modeset_vga_set_state(struct drm_i915_private *dev_priv, - bool state); -extern void intel_display_resume(struct drm_device *dev); -extern void i915_redisable_vga(struct drm_i915_private *dev_priv); -extern void i915_redisable_vga_power_on(struct drm_i915_private *dev_priv); -extern void intel_init_pch_refclk(struct drm_i915_private *dev_priv); +void intel_modeset_init_hw(struct drm_device *dev); +int intel_modeset_init(struct drm_device *dev); +void intel_modeset_cleanup(struct drm_device *dev); +int intel_modeset_vga_set_state(struct drm_i915_private *dev_priv, bool state); +void intel_display_resume(struct drm_device *dev); +void i915_redisable_vga(struct drm_i915_private *dev_priv); +void i915_redisable_vga_power_on(struct drm_i915_private *dev_priv); +void intel_init_pch_refclk(struct drm_i915_private *dev_priv); int i915_reg_read_ioctl(struct drm_device *dev, void *data, struct drm_file *file); -extern struct intel_display_error_state * +struct intel_display_error_state * intel_display_capture_error_state(struct drm_i915_private *dev_priv); -extern void intel_display_print_error_state(struct drm_i915_error_state_buf *e, - struct intel_display_error_state *error); +void intel_display_print_error_state(struct drm_i915_error_state_buf *e, + struct intel_display_error_state *error); #define __I915_REG_OP(op__, dev_priv__, ...) \ intel_uncore_##op__(&(dev_priv__)->uncore, __VA_ARGS__) diff --git a/drivers/gpu/drm/i915/i915_irq.h b/drivers/gpu/drm/i915/i915_irq.h index d93fa4e75442..4f803f910177 100644 --- a/drivers/gpu/drm/i915/i915_irq.h +++ b/drivers/gpu/drm/i915/i915_irq.h @@ -13,8 +13,8 @@ struct drm_i915_private; struct intel_crtc; -extern void intel_irq_init(struct drm_i915_private *dev_priv); -extern void intel_irq_fini(struct drm_i915_private *dev_priv); +void intel_irq_init(struct drm_i915_private *dev_priv); +void intel_irq_fini(struct drm_i915_private *dev_priv); int intel_irq_install(struct drm_i915_private *dev_priv); void intel_irq_uninstall(struct drm_i915_private *dev_priv); diff --git a/drivers/gpu/drm/i915/oa/i915_oa_bdw.h b/drivers/gpu/drm/i915/oa/i915_oa_bdw.h index 0e667f1a8aa1..b5ed68882588 100644 --- a/drivers/gpu/drm/i915/oa/i915_oa_bdw.h +++ b/drivers/gpu/drm/i915/oa/i915_oa_bdw.h @@ -10,6 +10,6 @@ #ifndef __I915_OA_BDW_H__ #define __I915_OA_BDW_H__ -extern void i915_perf_load_test_config_bdw(struct drm_i915_private *dev_priv); +void i915_perf_load_test_config_bdw(struct drm_i915_private *dev_priv); #endif diff --git a/drivers/gpu/drm/i915/oa/i915_oa_bxt.h b/drivers/gpu/drm/i915/oa/i915_oa_bxt.h index 679e92cf4f1d..43c3e4ab030a 100644 --- a/drivers/gpu/drm/i915/oa/i915_oa_bxt.h +++ b/drivers/gpu/drm/i915/oa/i915_oa_bxt.h @@ -10,6 +10,6 @@ #ifndef __I915_OA_BXT_H__ #define __I915_OA_BXT_H__ -extern void i915_perf_load_test_config_bxt(struct drm_i915_private *dev_priv); +void i915_perf_load_test_config_bxt(struct drm_i915_private *dev_priv); #endif diff --git a/drivers/gpu/drm/i915/oa/i915_oa_cflgt2.h b/drivers/gpu/drm/i915/oa/i915_oa_cflgt2.h index 4d6025559bbe..1b4b563bc585 100644 --- a/drivers/gpu/drm/i915/oa/i915_oa_cflgt2.h +++ b/drivers/gpu/drm/i915/oa/i915_oa_cflgt2.h @@ -10,6 +10,6 @@ #ifndef __I915_OA_CFLGT2_H__ #define __I915_OA_CFLGT2_H__ -extern void i915_perf_load_test_config_cflgt2(struct drm_i915_private *dev_priv); +void i915_perf_load_test_config_cflgt2(struct drm_i915_private *dev_priv); #endif diff --git a/drivers/gpu/drm/i915/oa/i915_oa_cflgt3.h b/drivers/gpu/drm/i915/oa/i915_oa_cflgt3.h index 0697f4077402..500565e055cd 100644 --- a/drivers/gpu/drm/i915/oa/i915_oa_cflgt3.h +++ b/drivers/gpu/drm/i915/oa/i915_oa_cflgt3.h @@ -10,6 +10,6 @@ #ifndef __I915_OA_CFLGT3_H__ #define __I915_OA_CFLGT3_H__ -extern void i915_perf_load_test_config_cflgt3(struct drm_i915_private *dev_priv); +void i915_perf_load_test_config_cflgt3(struct drm_i915_private *dev_priv); #endif diff --git a/drivers/gpu/drm/i915/oa/i915_oa_chv.h b/drivers/gpu/drm/i915/oa/i915_oa_chv.h index 0986eae3135f..ad85d6a6a573 100644 --- a/drivers/gpu/drm/i915/oa/i915_oa_chv.h +++ b/drivers/gpu/drm/i915/oa/i915_oa_chv.h @@ -10,6 +10,6 @@ #ifndef __I915_OA_CHV_H__ #define __I915_OA_CHV_H__ -extern void i915_perf_load_test_config_chv(struct drm_i915_private *dev_priv); +void i915_perf_load_test_config_chv(struct drm_i915_private *dev_priv); #endif diff --git a/drivers/gpu/drm/i915/oa/i915_oa_cnl.h b/drivers/gpu/drm/i915/oa/i915_oa_cnl.h index e830a406aff2..9faaca38b587 100644 --- a/drivers/gpu/drm/i915/oa/i915_oa_cnl.h +++ b/drivers/gpu/drm/i915/oa/i915_oa_cnl.h @@ -10,6 +10,6 @@ #ifndef __I915_OA_CNL_H__ #define __I915_OA_CNL_H__ -extern void i915_perf_load_test_config_cnl(struct drm_i915_private *dev_priv); +void i915_perf_load_test_config_cnl(struct drm_i915_private *dev_priv); #endif diff --git a/drivers/gpu/drm/i915/oa/i915_oa_glk.h b/drivers/gpu/drm/i915/oa/i915_oa_glk.h index 06dedf991edb..cc13a1e9fd3e 100644 --- a/drivers/gpu/drm/i915/oa/i915_oa_glk.h +++ b/drivers/gpu/drm/i915/oa/i915_oa_glk.h @@ -10,6 +10,6 @@ #ifndef __I915_OA_GLK_H__ #define __I915_OA_GLK_H__ -extern void i915_perf_load_test_config_glk(struct drm_i915_private *dev_priv); +void i915_perf_load_test_config_glk(struct drm_i915_private *dev_priv); #endif diff --git a/drivers/gpu/drm/i915/oa/i915_oa_hsw.h b/drivers/gpu/drm/i915/oa/i915_oa_hsw.h index 3d0c870cd0bd..f0ddcc79c761 100644 --- a/drivers/gpu/drm/i915/oa/i915_oa_hsw.h +++ b/drivers/gpu/drm/i915/oa/i915_oa_hsw.h @@ -10,6 +10,6 @@ #ifndef __I915_OA_HSW_H__ #define __I915_OA_HSW_H__ -extern void i915_perf_load_test_config_hsw(struct drm_i915_private *dev_priv); +void i915_perf_load_test_config_hsw(struct drm_i915_private *dev_priv); #endif diff --git a/drivers/gpu/drm/i915/oa/i915_oa_icl.h b/drivers/gpu/drm/i915/oa/i915_oa_icl.h index 24eaa97d61ba..e501651d385b 100644 --- a/drivers/gpu/drm/i915/oa/i915_oa_icl.h +++ b/drivers/gpu/drm/i915/oa/i915_oa_icl.h @@ -10,6 +10,6 @@ #ifndef __I915_OA_ICL_H__ #define __I915_OA_ICL_H__ -extern void i915_perf_load_test_config_icl(struct drm_i915_private *dev_priv); +void i915_perf_load_test_config_icl(struct drm_i915_private *dev_priv); #endif diff --git a/drivers/gpu/drm/i915/oa/i915_oa_kblgt2.h b/drivers/gpu/drm/i915/oa/i915_oa_kblgt2.h index a55398a904de..dc460e6e0fae 100644 --- a/drivers/gpu/drm/i915/oa/i915_oa_kblgt2.h +++ b/drivers/gpu/drm/i915/oa/i915_oa_kblgt2.h @@ -10,6 +10,6 @@ #ifndef __I915_OA_KBLGT2_H__ #define __I915_OA_KBLGT2_H__ -extern void i915_perf_load_test_config_kblgt2(struct drm_i915_private *dev_priv); +void i915_perf_load_test_config_kblgt2(struct drm_i915_private *dev_priv); #endif diff --git a/drivers/gpu/drm/i915/oa/i915_oa_kblgt3.h b/drivers/gpu/drm/i915/oa/i915_oa_kblgt3.h index 3ddd3483b7cc..5926992b735a 100644 --- a/drivers/gpu/drm/i915/oa/i915_oa_kblgt3.h +++ b/drivers/gpu/drm/i915/oa/i915_oa_kblgt3.h @@ -10,6 +10,6 @@ #ifndef __I915_OA_KBLGT3_H__ #define __I915_OA_KBLGT3_H__ -extern void i915_perf_load_test_config_kblgt3(struct drm_i915_private *dev_priv); +void i915_perf_load_test_config_kblgt3(struct drm_i915_private *dev_priv); #endif diff --git a/drivers/gpu/drm/i915/oa/i915_oa_sklgt2.h b/drivers/gpu/drm/i915/oa/i915_oa_sklgt2.h index be6256037239..353db35b36c1 100644 --- a/drivers/gpu/drm/i915/oa/i915_oa_sklgt2.h +++ b/drivers/gpu/drm/i915/oa/i915_oa_sklgt2.h @@ -10,6 +10,6 @@ #ifndef __I915_OA_SKLGT2_H__ #define __I915_OA_SKLGT2_H__ -extern void i915_perf_load_test_config_sklgt2(struct drm_i915_private *dev_priv); +void i915_perf_load_test_config_sklgt2(struct drm_i915_private *dev_priv); #endif diff --git a/drivers/gpu/drm/i915/oa/i915_oa_sklgt3.h b/drivers/gpu/drm/i915/oa/i915_oa_sklgt3.h index 650beb068e56..52f94c674b62 100644 --- a/drivers/gpu/drm/i915/oa/i915_oa_sklgt3.h +++ b/drivers/gpu/drm/i915/oa/i915_oa_sklgt3.h @@ -10,6 +10,6 @@ #ifndef __I915_OA_SKLGT3_H__ #define __I915_OA_SKLGT3_H__ -extern void i915_perf_load_test_config_sklgt3(struct drm_i915_private *dev_priv); +void i915_perf_load_test_config_sklgt3(struct drm_i915_private *dev_priv); #endif diff --git a/drivers/gpu/drm/i915/oa/i915_oa_sklgt4.h b/drivers/gpu/drm/i915/oa/i915_oa_sklgt4.h index 8dcf849d131e..8e364820cc63 100644 --- a/drivers/gpu/drm/i915/oa/i915_oa_sklgt4.h +++ b/drivers/gpu/drm/i915/oa/i915_oa_sklgt4.h @@ -10,6 +10,6 @@ #ifndef __I915_OA_SKLGT4_H__ #define __I915_OA_SKLGT4_H__ -extern void i915_perf_load_test_config_sklgt4(struct drm_i915_private *dev_priv); +void i915_perf_load_test_config_sklgt4(struct drm_i915_private *dev_priv); #endif diff --git a/include/drm/i915_drm.h b/include/drm/i915_drm.h index eb30062359d1..23274cf92712 100644 --- a/include/drm/i915_drm.h +++ b/include/drm/i915_drm.h @@ -30,11 +30,11 @@ #include /* For use by IPS driver */ -extern unsigned long i915_read_mch_val(void); -extern bool i915_gpu_raise(void); -extern bool i915_gpu_lower(void); -extern bool i915_gpu_busy(void); -extern bool i915_gpu_turbo_disable(void); +unsigned long i915_read_mch_val(void); +bool i915_gpu_raise(void); +bool i915_gpu_lower(void); +bool i915_gpu_busy(void); +bool i915_gpu_turbo_disable(void); /* Exported from arch/x86/kernel/early-quirks.c */ extern struct resource intel_graphics_stolen_res; -- cgit v1.2.3 From f5d974f9d2a811ef08c044b6fce95c94a6a6e19b Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Tue, 30 Jul 2019 15:32:09 +0100 Subject: drm/i915/gt: Provide a local intel_context.vm Track the currently bound address space used by the HW context. Minor conversions to use the local intel_context.vm are made, leaving behind some more surgery required to make intel_context the primary through the selftests. Signed-off-by: Chris Wilson Reviewed-by: Tvrtko Ursulin Link: https://patchwork.freedesktop.org/patch/msgid/20190730143209.4549-2-chris@chris-wilson.co.uk --- drivers/gpu/drm/i915/gem/i915_gem_client_blt.c | 4 +--- drivers/gpu/drm/i915/gem/i915_gem_context.c | 15 +++++++++++---- drivers/gpu/drm/i915/gem/i915_gem_execbuffer.c | 11 +++-------- drivers/gpu/drm/i915/gem/i915_gem_object_blt.c | 6 +----- drivers/gpu/drm/i915/gem/selftests/i915_gem_context.c | 2 +- drivers/gpu/drm/i915/gt/intel_context.c | 4 ++++ drivers/gpu/drm/i915/gt/intel_context_types.h | 4 +++- drivers/gpu/drm/i915/gt/intel_lrc.c | 9 +++------ drivers/gpu/drm/i915/gt/intel_ringbuffer.c | 6 +++--- drivers/gpu/drm/i915/gvt/scheduler.c | 2 +- 10 files changed, 31 insertions(+), 32 deletions(-) (limited to 'drivers/gpu/drm/i915/gvt') diff --git a/drivers/gpu/drm/i915/gem/i915_gem_client_blt.c b/drivers/gpu/drm/i915/gem/i915_gem_client_blt.c index 6f537e8e4dea..2312a0c6af89 100644 --- a/drivers/gpu/drm/i915/gem/i915_gem_client_blt.c +++ b/drivers/gpu/drm/i915/gem/i915_gem_client_blt.c @@ -250,13 +250,11 @@ int i915_gem_schedule_fill_pages_blt(struct drm_i915_gem_object *obj, u32 value) { struct drm_i915_private *i915 = to_i915(obj->base.dev); - struct i915_gem_context *ctx = ce->gem_context; - struct i915_address_space *vm = ctx->vm ?: &i915->ggtt.vm; struct clear_pages_work *work; struct i915_sleeve *sleeve; int err; - sleeve = create_sleeve(vm, obj, pages, page_sizes); + sleeve = create_sleeve(ce->vm, obj, pages, page_sizes); if (IS_ERR(sleeve)) return PTR_ERR(sleeve); diff --git a/drivers/gpu/drm/i915/gem/i915_gem_context.c b/drivers/gpu/drm/i915/gem/i915_gem_context.c index 0f6b0678f548..b28c7ca681a8 100644 --- a/drivers/gpu/drm/i915/gem/i915_gem_context.c +++ b/drivers/gpu/drm/i915/gem/i915_gem_context.c @@ -475,10 +475,18 @@ static struct i915_address_space * __set_ppgtt(struct i915_gem_context *ctx, struct i915_address_space *vm) { struct i915_address_space *old = ctx->vm; + struct i915_gem_engines_iter it; + struct intel_context *ce; ctx->vm = i915_vm_get(vm); ctx->desc_template = default_desc_template(ctx->i915, vm); + for_each_gem_engine(ce, i915_gem_context_lock_engines(ctx), it) { + i915_vm_put(ce->vm); + ce->vm = i915_vm_get(vm); + } + i915_gem_context_unlock_engines(ctx); + return old; } @@ -1004,7 +1012,7 @@ static void set_ppgtt_barrier(void *data) static int emit_ppgtt_update(struct i915_request *rq, void *data) { - struct i915_address_space *vm = rq->gem_context->vm; + struct i915_address_space *vm = rq->hw_context->vm; struct intel_engine_cs *engine = rq->engine; u32 base = engine->mmio_base; u32 *cs; @@ -1113,9 +1121,8 @@ static int set_ppgtt(struct drm_i915_file_private *file_priv, set_ppgtt_barrier, old); if (err) { - ctx->vm = old; - ctx->desc_template = default_desc_template(ctx->i915, old); - i915_vm_put(vm); + i915_vm_put(__set_ppgtt(ctx, old)); + i915_vm_put(old); } unlock: diff --git a/drivers/gpu/drm/i915/gem/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/gem/i915_gem_execbuffer.c index 8a2047c4e7c3..cbd7c6e3a1f8 100644 --- a/drivers/gpu/drm/i915/gem/i915_gem_execbuffer.c +++ b/drivers/gpu/drm/i915/gem/i915_gem_execbuffer.c @@ -223,7 +223,6 @@ struct i915_execbuffer { struct intel_engine_cs *engine; /** engine to queue the request to */ struct intel_context *context; /* logical state for the request */ struct i915_gem_context *gem_context; /** caller's context */ - struct i915_address_space *vm; /** GTT and vma for the request */ struct i915_request *request; /** our request to build */ struct i915_vma *batch; /** identity of the batch obj/vma */ @@ -697,7 +696,7 @@ static int eb_reserve(struct i915_execbuffer *eb) case 1: /* Too fragmented, unbind everything and retry */ - err = i915_gem_evict_vm(eb->vm); + err = i915_gem_evict_vm(eb->context->vm); if (err) return err; break; @@ -725,12 +724,8 @@ static int eb_select_context(struct i915_execbuffer *eb) return -ENOENT; eb->gem_context = ctx; - if (ctx->vm) { - eb->vm = ctx->vm; + if (ctx->vm) eb->invalid_flags |= EXEC_OBJECT_NEEDS_GTT; - } else { - eb->vm = &eb->i915->ggtt.vm; - } eb->context_flags = 0; if (test_bit(UCONTEXT_NO_ZEROMAP, &ctx->user_flags)) @@ -832,7 +827,7 @@ static int eb_lookup_vmas(struct i915_execbuffer *eb) goto err_vma; } - vma = i915_vma_instance(obj, eb->vm, NULL); + vma = i915_vma_instance(obj, eb->context->vm, NULL); if (IS_ERR(vma)) { err = PTR_ERR(vma); goto err_obj; diff --git a/drivers/gpu/drm/i915/gem/i915_gem_object_blt.c b/drivers/gpu/drm/i915/gem/i915_gem_object_blt.c index cb42e3a312e2..685064af32d1 100644 --- a/drivers/gpu/drm/i915/gem/i915_gem_object_blt.c +++ b/drivers/gpu/drm/i915/gem/i915_gem_object_blt.c @@ -47,15 +47,11 @@ int i915_gem_object_fill_blt(struct drm_i915_gem_object *obj, struct intel_context *ce, u32 value) { - struct drm_i915_private *i915 = to_i915(obj->base.dev); - struct i915_gem_context *ctx = ce->gem_context; - struct i915_address_space *vm = ctx->vm ?: &i915->ggtt.vm; struct i915_request *rq; struct i915_vma *vma; int err; - /* XXX: ce->vm please */ - vma = i915_vma_instance(obj, vm, NULL); + vma = i915_vma_instance(obj, ce->vm, NULL); if (IS_ERR(vma)) return PTR_ERR(vma); diff --git a/drivers/gpu/drm/i915/gem/selftests/i915_gem_context.c b/drivers/gpu/drm/i915/gem/selftests/i915_gem_context.c index bbd17d4b8ffd..7f9f6701b32c 100644 --- a/drivers/gpu/drm/i915/gem/selftests/i915_gem_context.c +++ b/drivers/gpu/drm/i915/gem/selftests/i915_gem_context.c @@ -747,7 +747,7 @@ emit_rpcs_query(struct drm_i915_gem_object *obj, GEM_BUG_ON(!intel_engine_can_store_dword(ce->engine)); - vma = i915_vma_instance(obj, ce->gem_context->vm, NULL); + vma = i915_vma_instance(obj, ce->vm, NULL); if (IS_ERR(vma)) return PTR_ERR(vma); diff --git a/drivers/gpu/drm/i915/gt/intel_context.c b/drivers/gpu/drm/i915/gt/intel_context.c index d64b45f7ec6d..f30441a140f8 100644 --- a/drivers/gpu/drm/i915/gt/intel_context.c +++ b/drivers/gpu/drm/i915/gt/intel_context.c @@ -191,6 +191,8 @@ intel_context_init(struct intel_context *ce, kref_init(&ce->ref); ce->gem_context = ctx; + ce->vm = i915_vm_get(ctx->vm ?: &engine->gt->ggtt->vm); + ce->engine = engine; ce->ops = engine->cops; ce->sseu = engine->sseu; @@ -206,6 +208,8 @@ intel_context_init(struct intel_context *ce, void intel_context_fini(struct intel_context *ce) { + i915_vm_put(ce->vm); + mutex_destroy(&ce->pin_mutex); i915_active_fini(&ce->active); } diff --git a/drivers/gpu/drm/i915/gt/intel_context_types.h b/drivers/gpu/drm/i915/gt/intel_context_types.h index 4c0e211c715d..68a7e979b1a9 100644 --- a/drivers/gpu/drm/i915/gt/intel_context_types.h +++ b/drivers/gpu/drm/i915/gt/intel_context_types.h @@ -36,7 +36,6 @@ struct intel_context_ops { struct intel_context { struct kref ref; - struct i915_gem_context *gem_context; struct intel_engine_cs *engine; struct intel_engine_cs *inflight; #define intel_context_inflight(ce) ptr_mask_bits((ce)->inflight, 2) @@ -44,6 +43,9 @@ struct intel_context { #define intel_context_inflight_inc(ce) ptr_count_inc(&(ce)->inflight) #define intel_context_inflight_dec(ce) ptr_count_dec(&(ce)->inflight) + struct i915_address_space *vm; + struct i915_gem_context *gem_context; + struct list_head signal_link; struct list_head signals; diff --git a/drivers/gpu/drm/i915/gt/intel_lrc.c b/drivers/gpu/drm/i915/gt/intel_lrc.c index 4d7c4d0dbf75..d9061d9348cb 100644 --- a/drivers/gpu/drm/i915/gt/intel_lrc.c +++ b/drivers/gpu/drm/i915/gt/intel_lrc.c @@ -1605,8 +1605,6 @@ __execlists_context_pin(struct intel_context *ce, void *vaddr; int ret; - GEM_BUG_ON(!ce->gem_context->vm); - ret = execlists_context_deferred_alloc(ce, engine); if (ret) goto err; @@ -1716,8 +1714,7 @@ static int gen8_emit_init_breadcrumb(struct i915_request *rq) static int emit_pdps(struct i915_request *rq) { const struct intel_engine_cs * const engine = rq->engine; - struct i915_ppgtt * const ppgtt = - i915_vm_to_ppgtt(rq->gem_context->vm); + struct i915_ppgtt * const ppgtt = i915_vm_to_ppgtt(rq->hw_context->vm); int err, i; u32 *cs; @@ -1790,7 +1787,7 @@ static int execlists_request_alloc(struct i915_request *request) */ /* Unconditionally invalidate GPU caches and TLBs. */ - if (i915_vm_is_4lvl(request->gem_context->vm)) + if (i915_vm_is_4lvl(request->hw_context->vm)) ret = request->engine->emit_flush(request, EMIT_INVALIDATE); else ret = emit_pdps(request); @@ -2920,7 +2917,7 @@ static void execlists_init_reg_state(u32 *regs, struct intel_engine_cs *engine, struct intel_ring *ring) { - struct i915_ppgtt *ppgtt = i915_vm_to_ppgtt(ce->gem_context->vm); + struct i915_ppgtt *ppgtt = i915_vm_to_ppgtt(ce->vm); bool rcs = engine->class == RENDER_CLASS; u32 base = engine->mmio_base; diff --git a/drivers/gpu/drm/i915/gt/intel_ringbuffer.c b/drivers/gpu/drm/i915/gt/intel_ringbuffer.c index 11afe64cc256..8d24a49e5139 100644 --- a/drivers/gpu/drm/i915/gt/intel_ringbuffer.c +++ b/drivers/gpu/drm/i915/gt/intel_ringbuffer.c @@ -1380,9 +1380,9 @@ static struct i915_address_space *vm_alias(struct intel_context *ce) { struct i915_address_space *vm; - vm = ce->gem_context->vm; - if (!vm) - vm = &ce->engine->gt->ggtt->alias->vm; + vm = ce->vm; + if (i915_is_ggtt(vm)) + vm = &i915_vm_to_ggtt(vm)->alias->vm; return vm; } diff --git a/drivers/gpu/drm/i915/gvt/scheduler.c b/drivers/gpu/drm/i915/gvt/scheduler.c index 2144fb46d0e1..f68798ab1e7c 100644 --- a/drivers/gpu/drm/i915/gvt/scheduler.c +++ b/drivers/gpu/drm/i915/gvt/scheduler.c @@ -1156,7 +1156,7 @@ void intel_vgpu_clean_submission(struct intel_vgpu *vgpu) intel_vgpu_select_submission_ops(vgpu, ALL_ENGINES, 0); - i915_context_ppgtt_root_restore(s, i915_vm_to_ppgtt(s->shadow[0]->gem_context->vm)); + i915_context_ppgtt_root_restore(s, i915_vm_to_ppgtt(s->shadow[0]->vm)); for_each_engine(engine, vgpu->gvt->dev_priv, id) intel_context_unpin(s->shadow[id]); -- cgit v1.2.3 From a1c9ca223c3df1b8993abedde777f5462165387c Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Tue, 30 Jul 2019 14:30:26 +0100 Subject: drm/i915: Remove lrc default desc from GEM context We only compute the lrc_descriptor() on pinning the context, i.e. infrequently, so we do not benefit from storing the template as the addressing mode is also fixed for the lifetime of the intel_context. Signed-off-by: Chris Wilson Reviewed-by: Prathap Kumar Valsan Acked-by: Tvrtko Ursulin Link: https://patchwork.freedesktop.org/patch/msgid/20190730133035.1977-9-chris@chris-wilson.co.uk --- drivers/gpu/drm/i915/gem/i915_gem_context.c | 28 ++--------------------- drivers/gpu/drm/i915/gem/i915_gem_context_types.h | 2 -- drivers/gpu/drm/i915/gt/intel_lrc.c | 12 ++++++---- drivers/gpu/drm/i915/gvt/scheduler.c | 3 --- 4 files changed, 10 insertions(+), 35 deletions(-) (limited to 'drivers/gpu/drm/i915/gvt') diff --git a/drivers/gpu/drm/i915/gem/i915_gem_context.c b/drivers/gpu/drm/i915/gem/i915_gem_context.c index eb2d28a37b58..64f7a533e886 100644 --- a/drivers/gpu/drm/i915/gem/i915_gem_context.c +++ b/drivers/gpu/drm/i915/gem/i915_gem_context.c @@ -397,30 +397,6 @@ static void context_close(struct i915_gem_context *ctx) i915_gem_context_put(ctx); } -static u32 default_desc_template(const struct drm_i915_private *i915, - const struct i915_address_space *vm) -{ - u32 address_mode; - u32 desc; - - desc = GEN8_CTX_VALID | GEN8_CTX_PRIVILEGE; - - address_mode = INTEL_LEGACY_32B_CONTEXT; - if (vm && i915_vm_is_4lvl(vm)) - address_mode = INTEL_LEGACY_64B_CONTEXT; - desc |= address_mode << GEN8_CTX_ADDRESSING_MODE_SHIFT; - - if (IS_GEN(i915, 8)) - desc |= GEN8_CTX_L3LLC_COHERENT; - - /* TODO: WaDisableLiteRestore when we start using semaphore - * signalling between Command Streamers - * ring->ctx_desc_template |= GEN8_CTX_FORCE_RESTORE; - */ - - return desc; -} - static struct i915_gem_context * __create_context(struct drm_i915_private *i915) { @@ -459,7 +435,6 @@ __create_context(struct drm_i915_private *i915) i915_gem_context_set_recoverable(ctx); ctx->ring_size = 4 * PAGE_SIZE; - ctx->desc_template = default_desc_template(i915, NULL); for (i = 0; i < ARRAY_SIZE(ctx->hang_timestamp); i++) ctx->hang_timestamp[i] = jiffies - CONTEXT_FAST_HANG_JIFFIES; @@ -478,8 +453,9 @@ __set_ppgtt(struct i915_gem_context *ctx, struct i915_address_space *vm) struct i915_gem_engines_iter it; struct intel_context *ce; + GEM_BUG_ON(old && i915_vm_is_4lvl(vm) != i915_vm_is_4lvl(old)); + ctx->vm = i915_vm_get(vm); - ctx->desc_template = default_desc_template(ctx->i915, vm); for_each_gem_engine(ce, i915_gem_context_lock_engines(ctx), it) { i915_vm_put(ce->vm); diff --git a/drivers/gpu/drm/i915/gem/i915_gem_context_types.h b/drivers/gpu/drm/i915/gem/i915_gem_context_types.h index 0ee61482ef94..a02d98494078 100644 --- a/drivers/gpu/drm/i915/gem/i915_gem_context_types.h +++ b/drivers/gpu/drm/i915/gem/i915_gem_context_types.h @@ -171,8 +171,6 @@ struct i915_gem_context { /** ring_size: size for allocating the per-engine ring buffer */ u32 ring_size; - /** desc_template: invariant fields for the HW context descriptor */ - u32 desc_template; /** guilty_count: How many times this context has caused a GPU hang. */ atomic_t guilty_count; diff --git a/drivers/gpu/drm/i915/gt/intel_lrc.c b/drivers/gpu/drm/i915/gt/intel_lrc.c index 5181d29d272e..232f40fcb490 100644 --- a/drivers/gpu/drm/i915/gt/intel_lrc.c +++ b/drivers/gpu/drm/i915/gt/intel_lrc.c @@ -417,13 +417,17 @@ lrc_descriptor(struct intel_context *ce, struct intel_engine_cs *engine) BUILD_BUG_ON(MAX_CONTEXT_HW_ID > (BIT(GEN8_CTX_ID_WIDTH))); BUILD_BUG_ON(GEN11_MAX_CONTEXT_HW_ID > (BIT(GEN11_SW_CTX_ID_WIDTH))); - desc = ctx->desc_template; /* bits 0-11 */ - GEM_BUG_ON(desc & GENMASK_ULL(63, 12)); + desc = INTEL_LEGACY_32B_CONTEXT; + if (i915_vm_is_4lvl(ce->vm)) + desc = INTEL_LEGACY_64B_CONTEXT; + desc <<= GEN8_CTX_ADDRESSING_MODE_SHIFT; + + desc |= GEN8_CTX_VALID | GEN8_CTX_PRIVILEGE; + if (IS_GEN(engine->i915, 8)) + desc |= GEN8_CTX_L3LLC_COHERENT; desc |= i915_ggtt_offset(ce->state) + LRC_HEADER_PAGES * PAGE_SIZE; /* bits 12-31 */ - GEM_BUG_ON(desc & GENMASK_ULL(63, 32)); - /* * The following 32bits are copied into the OA reports (dword 2). * Consider updating oa_get_render_ctx_id in i915_perf.c when changing diff --git a/drivers/gpu/drm/i915/gvt/scheduler.c b/drivers/gpu/drm/i915/gvt/scheduler.c index f68798ab1e7c..4c018fb1359c 100644 --- a/drivers/gpu/drm/i915/gvt/scheduler.c +++ b/drivers/gpu/drm/i915/gvt/scheduler.c @@ -291,9 +291,6 @@ shadow_context_descriptor_update(struct intel_context *ce, * Update bits 0-11 of the context descriptor which includes flags * like GEN8_CTX_* cached in desc_template */ - desc &= U64_MAX << 12; - desc |= ce->gem_context->desc_template & ((1ULL << 12) - 1); - desc &= ~(0x3 << GEN8_CTX_ADDRESSING_MODE_SHIFT); desc |= workload->ctx_desc.addressing_mode << GEN8_CTX_ADDRESSING_MODE_SHIFT; -- cgit v1.2.3 From a37f08a882b01a6e86a07512a082b14d20ee0773 Mon Sep 17 00:00:00 2001 From: Umesh Nerlige Ramappa Date: Tue, 6 Aug 2019 16:30:02 -0700 Subject: drm/i915/perf: Refactor oa object to better manage resources The oa object manages the oa buffer and must be allocated when the user intends to read performance counter snapshots. This can be achieved by making the oa object part of the stream object which is allocated when a stream is opened by the user. Attributes in the oa object that are gen-specific are moved to the perf object so that they can be initialized on driver load. The split provides a better separation of the objects used in perf implementation of i915 driver so that resources are allocated and initialized only when needed. v2: Fix checkpatch warnings v3: Addressed Lionel's review comment v4: Rebase v5: Fix rebase/merge issue with ratelimit_state_init Signed-off-by: Umesh Nerlige Ramappa Reviewed-by: Lionel Landwerlin Signed-off-by: Chris Wilson Link: https://patchwork.freedesktop.org/patch/msgid/20190806233002.984-1-umesh.nerlige.ramappa@intel.com --- drivers/gpu/drm/i915/gt/intel_sseu.c | 2 +- drivers/gpu/drm/i915/gvt/scheduler.c | 4 +- drivers/gpu/drm/i915/i915_drv.h | 219 ++++++------- drivers/gpu/drm/i915/i915_perf.c | 540 ++++++++++++++++--------------- drivers/gpu/drm/i915/oa/i915_oa_bdw.c | 30 +- drivers/gpu/drm/i915/oa/i915_oa_bxt.c | 30 +- drivers/gpu/drm/i915/oa/i915_oa_cflgt2.c | 30 +- drivers/gpu/drm/i915/oa/i915_oa_cflgt3.c | 30 +- drivers/gpu/drm/i915/oa/i915_oa_chv.c | 30 +- drivers/gpu/drm/i915/oa/i915_oa_cnl.c | 30 +- drivers/gpu/drm/i915/oa/i915_oa_glk.c | 30 +- drivers/gpu/drm/i915/oa/i915_oa_hsw.c | 30 +- drivers/gpu/drm/i915/oa/i915_oa_icl.c | 30 +- drivers/gpu/drm/i915/oa/i915_oa_kblgt2.c | 30 +- drivers/gpu/drm/i915/oa/i915_oa_kblgt3.c | 30 +- drivers/gpu/drm/i915/oa/i915_oa_sklgt2.c | 30 +- drivers/gpu/drm/i915/oa/i915_oa_sklgt3.c | 30 +- drivers/gpu/drm/i915/oa/i915_oa_sklgt4.c | 30 +- 18 files changed, 596 insertions(+), 589 deletions(-) (limited to 'drivers/gpu/drm/i915/gvt') diff --git a/drivers/gpu/drm/i915/gt/intel_sseu.c b/drivers/gpu/drm/i915/gt/intel_sseu.c index a0756f006f5f..6bf2d87da109 100644 --- a/drivers/gpu/drm/i915/gt/intel_sseu.c +++ b/drivers/gpu/drm/i915/gt/intel_sseu.c @@ -49,7 +49,7 @@ u32 intel_sseu_make_rpcs(struct drm_i915_private *i915, * cases which disable slices for functional, apart for performance * reasons. So in this case we select a known stable subset. */ - if (!i915->perf.oa.exclusive_stream) { + if (!i915->perf.exclusive_stream) { ctx_sseu = *req_sseu; } else { ctx_sseu = intel_sseu_from_device_info(sseu); diff --git a/drivers/gpu/drm/i915/gvt/scheduler.c b/drivers/gpu/drm/i915/gvt/scheduler.c index 4c018fb1359c..431fd427c49c 100644 --- a/drivers/gpu/drm/i915/gvt/scheduler.c +++ b/drivers/gpu/drm/i915/gvt/scheduler.c @@ -84,8 +84,8 @@ static void sr_oa_regs(struct intel_vgpu_workload *workload, u32 *reg_state, bool save) { struct drm_i915_private *dev_priv = workload->vgpu->gvt->dev_priv; - u32 ctx_oactxctrl = dev_priv->perf.oa.ctx_oactxctrl_offset; - u32 ctx_flexeu0 = dev_priv->perf.oa.ctx_flexeu0_offset; + u32 ctx_oactxctrl = dev_priv->perf.ctx_oactxctrl_offset; + u32 ctx_flexeu0 = dev_priv->perf.ctx_flexeu0_offset; int i = 0; u32 flex_mmio[] = { i915_mmio_reg_offset(EU_PERF_CNTL0), diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h index 1f70d9e2fc35..7232e203aab3 100644 --- a/drivers/gpu/drm/i915/i915_drv.h +++ b/drivers/gpu/drm/i915/i915_drv.h @@ -1241,6 +1241,86 @@ struct i915_perf_stream { * @oa_config: The OA configuration used by the stream. */ struct i915_oa_config *oa_config; + + /** + * The OA context specific information. + */ + struct intel_context *pinned_ctx; + u32 specific_ctx_id; + u32 specific_ctx_id_mask; + + struct hrtimer poll_check_timer; + wait_queue_head_t poll_wq; + bool pollin; + + bool periodic; + int period_exponent; + + /** + * State of the OA buffer. + */ + struct { + struct i915_vma *vma; + u8 *vaddr; + u32 last_ctx_id; + int format; + int format_size; + int size_exponent; + + /** + * Locks reads and writes to all head/tail state + * + * Consider: the head and tail pointer state needs to be read + * consistently from a hrtimer callback (atomic context) and + * read() fop (user context) with tail pointer updates happening + * in atomic context and head updates in user context and the + * (unlikely) possibility of read() errors needing to reset all + * head/tail state. + * + * Note: Contention/performance aren't currently a significant + * concern here considering the relatively low frequency of + * hrtimer callbacks (5ms period) and that reads typically only + * happen in response to a hrtimer event and likely complete + * before the next callback. + * + * Note: This lock is not held *while* reading and copying data + * to userspace so the value of head observed in htrimer + * callbacks won't represent any partial consumption of data. + */ + spinlock_t ptr_lock; + + /** + * One 'aging' tail pointer and one 'aged' tail pointer ready to + * used for reading. + * + * Initial values of 0xffffffff are invalid and imply that an + * update is required (and should be ignored by an attempted + * read) + */ + struct { + u32 offset; + } tails[2]; + + /** + * Index for the aged tail ready to read() data up to. + */ + unsigned int aged_tail_idx; + + /** + * A monotonic timestamp for when the current aging tail pointer + * was read; used to determine when it is old enough to trust. + */ + u64 aging_timestamp; + + /** + * Although we can always read back the head pointer register, + * we prefer to avoid trusting the HW state, just to avoid any + * risk that some hardware condition could * somehow bump the + * head pointer unpredictably and cause us to forward the wrong + * OA buffer data to userspace. + */ + u32 head; + } oa_buffer; }; /** @@ -1278,7 +1358,7 @@ struct i915_oa_ops { * @disable_metric_set: Remove system constraints associated with using * the OA unit. */ - void (*disable_metric_set)(struct drm_i915_private *dev_priv); + void (*disable_metric_set)(struct i915_perf_stream *stream); /** * @oa_enable: Enable periodic sampling @@ -1306,7 +1386,7 @@ struct i915_oa_ops { * handling the OA unit tail pointer race that affects multiple * generations. */ - u32 (*oa_hw_tail_read)(struct drm_i915_private *dev_priv); + u32 (*oa_hw_tail_read)(struct i915_perf_stream *stream); }; struct intel_cdclk_state { @@ -1705,120 +1785,35 @@ struct drm_i915_private { struct mutex lock; struct list_head streams; - struct { - /* - * The stream currently using the OA unit. If accessed - * outside a syscall associated to its file - * descriptor, you need to hold - * dev_priv->drm.struct_mutex. - */ - struct i915_perf_stream *exclusive_stream; + /* + * The stream currently using the OA unit. If accessed + * outside a syscall associated to its file + * descriptor, you need to hold + * dev_priv->drm.struct_mutex. + */ + struct i915_perf_stream *exclusive_stream; + + /** + * For rate limiting any notifications of spurious + * invalid OA reports + */ + struct ratelimit_state spurious_report_rs; - struct intel_context *pinned_ctx; - u32 specific_ctx_id; - u32 specific_ctx_id_mask; + struct i915_oa_config test_config; - struct hrtimer poll_check_timer; - wait_queue_head_t poll_wq; - bool pollin; + u32 gen7_latched_oastatus1; + u32 ctx_oactxctrl_offset; + u32 ctx_flexeu0_offset; - /** - * For rate limiting any notifications of spurious - * invalid OA reports - */ - struct ratelimit_state spurious_report_rs; - - bool periodic; - int period_exponent; - - struct i915_oa_config test_config; - - struct { - struct i915_vma *vma; - u8 *vaddr; - u32 last_ctx_id; - int format; - int format_size; - - /** - * Locks reads and writes to all head/tail state - * - * Consider: the head and tail pointer state - * needs to be read consistently from a hrtimer - * callback (atomic context) and read() fop - * (user context) with tail pointer updates - * happening in atomic context and head updates - * in user context and the (unlikely) - * possibility of read() errors needing to - * reset all head/tail state. - * - * Note: Contention or performance aren't - * currently a significant concern here - * considering the relatively low frequency of - * hrtimer callbacks (5ms period) and that - * reads typically only happen in response to a - * hrtimer event and likely complete before the - * next callback. - * - * Note: This lock is not held *while* reading - * and copying data to userspace so the value - * of head observed in htrimer callbacks won't - * represent any partial consumption of data. - */ - spinlock_t ptr_lock; - - /** - * One 'aging' tail pointer and one 'aged' - * tail pointer ready to used for reading. - * - * Initial values of 0xffffffff are invalid - * and imply that an update is required - * (and should be ignored by an attempted - * read) - */ - struct { - u32 offset; - } tails[2]; - - /** - * Index for the aged tail ready to read() - * data up to. - */ - unsigned int aged_tail_idx; - - /** - * A monotonic timestamp for when the current - * aging tail pointer was read; used to - * determine when it is old enough to trust. - */ - u64 aging_timestamp; - - /** - * Although we can always read back the head - * pointer register, we prefer to avoid - * trusting the HW state, just to avoid any - * risk that some hardware condition could - * somehow bump the head pointer unpredictably - * and cause us to forward the wrong OA buffer - * data to userspace. - */ - u32 head; - } oa_buffer; - - u32 gen7_latched_oastatus1; - u32 ctx_oactxctrl_offset; - u32 ctx_flexeu0_offset; - - /** - * The RPT_ID/reason field for Gen8+ includes a bit - * to determine if the CTX ID in the report is valid - * but the specific bit differs between Gen 8 and 9 - */ - u32 gen8_valid_ctx_bit; + /** + * The RPT_ID/reason field for Gen8+ includes a bit + * to determine if the CTX ID in the report is valid + * but the specific bit differs between Gen 8 and 9 + */ + u32 gen8_valid_ctx_bit; - struct i915_oa_ops ops; - const struct i915_oa_format *oa_formats; - } oa; + struct i915_oa_ops ops; + const struct i915_oa_format *oa_formats; } perf; /* Abstract the submission mechanism (legacy ringbuffer or execlists) away */ diff --git a/drivers/gpu/drm/i915/i915_perf.c b/drivers/gpu/drm/i915/i915_perf.c index 2bcf0d6e2ee2..7b59a7038556 100644 --- a/drivers/gpu/drm/i915/i915_perf.c +++ b/drivers/gpu/drm/i915/i915_perf.c @@ -364,6 +364,8 @@ struct perf_open_properties { int oa_period_exponent; }; +static enum hrtimer_restart oa_poll_check_timer_cb(struct hrtimer *hrtimer); + static void free_oa_config(struct drm_i915_private *dev_priv, struct i915_oa_config *oa_config) { @@ -392,8 +394,8 @@ static int get_oa_config(struct drm_i915_private *dev_priv, int ret; if (metrics_set == 1) { - *out_config = &dev_priv->perf.oa.test_config; - atomic_inc(&dev_priv->perf.oa.test_config.ref_count); + *out_config = &dev_priv->perf.test_config; + atomic_inc(&dev_priv->perf.test_config.ref_count); return 0; } @@ -412,13 +414,16 @@ static int get_oa_config(struct drm_i915_private *dev_priv, return ret; } -static u32 gen8_oa_hw_tail_read(struct drm_i915_private *dev_priv) +static u32 gen8_oa_hw_tail_read(struct i915_perf_stream *stream) { + struct drm_i915_private *dev_priv = stream->dev_priv; + return I915_READ(GEN8_OATAILPTR) & GEN8_OATAILPTR_MASK; } -static u32 gen7_oa_hw_tail_read(struct drm_i915_private *dev_priv) +static u32 gen7_oa_hw_tail_read(struct i915_perf_stream *stream) { + struct drm_i915_private *dev_priv = stream->dev_priv; u32 oastatus1 = I915_READ(GEN7_OASTATUS1); return oastatus1 & GEN7_OASTATUS1_TAIL_MASK; @@ -426,7 +431,7 @@ static u32 gen7_oa_hw_tail_read(struct drm_i915_private *dev_priv) /** * oa_buffer_check_unlocked - check for data and update tail ptr state - * @dev_priv: i915 device instance + * @stream: i915 stream instance * * This is either called via fops (for blocking reads in user ctx) or the poll * check hrtimer (atomic ctx) to check the OA buffer tail pointer and check @@ -448,9 +453,10 @@ static u32 gen7_oa_hw_tail_read(struct drm_i915_private *dev_priv) * * Returns: %true if the OA buffer contains data, else %false */ -static bool oa_buffer_check_unlocked(struct drm_i915_private *dev_priv) +static bool oa_buffer_check_unlocked(struct i915_perf_stream *stream) { - int report_size = dev_priv->perf.oa.oa_buffer.format_size; + struct drm_i915_private *dev_priv = stream->dev_priv; + int report_size = stream->oa_buffer.format_size; unsigned long flags; unsigned int aged_idx; u32 head, hw_tail, aged_tail, aging_tail; @@ -460,19 +466,19 @@ static bool oa_buffer_check_unlocked(struct drm_i915_private *dev_priv) * could result in an OA buffer reset which might reset the head, * tails[] and aged_tail state. */ - spin_lock_irqsave(&dev_priv->perf.oa.oa_buffer.ptr_lock, flags); + spin_lock_irqsave(&stream->oa_buffer.ptr_lock, flags); /* NB: The head we observe here might effectively be a little out of * date (between head and tails[aged_idx].offset if there is currently * a read() in progress. */ - head = dev_priv->perf.oa.oa_buffer.head; + head = stream->oa_buffer.head; - aged_idx = dev_priv->perf.oa.oa_buffer.aged_tail_idx; - aged_tail = dev_priv->perf.oa.oa_buffer.tails[aged_idx].offset; - aging_tail = dev_priv->perf.oa.oa_buffer.tails[!aged_idx].offset; + aged_idx = stream->oa_buffer.aged_tail_idx; + aged_tail = stream->oa_buffer.tails[aged_idx].offset; + aging_tail = stream->oa_buffer.tails[!aged_idx].offset; - hw_tail = dev_priv->perf.oa.ops.oa_hw_tail_read(dev_priv); + hw_tail = dev_priv->perf.ops.oa_hw_tail_read(stream); /* The tail pointer increases in 64 byte increments, * not in report_size steps... @@ -492,16 +498,16 @@ static bool oa_buffer_check_unlocked(struct drm_i915_private *dev_priv) * available) without needing to wait for a later hrtimer callback. */ if (aging_tail != INVALID_TAIL_PTR && - ((now - dev_priv->perf.oa.oa_buffer.aging_timestamp) > + ((now - stream->oa_buffer.aging_timestamp) > OA_TAIL_MARGIN_NSEC)) { aged_idx ^= 1; - dev_priv->perf.oa.oa_buffer.aged_tail_idx = aged_idx; + stream->oa_buffer.aged_tail_idx = aged_idx; aged_tail = aging_tail; /* Mark that we need a new pointer to start aging... */ - dev_priv->perf.oa.oa_buffer.tails[!aged_idx].offset = INVALID_TAIL_PTR; + stream->oa_buffer.tails[!aged_idx].offset = INVALID_TAIL_PTR; aging_tail = INVALID_TAIL_PTR; } @@ -516,7 +522,7 @@ static bool oa_buffer_check_unlocked(struct drm_i915_private *dev_priv) if (aging_tail == INVALID_TAIL_PTR && (aged_tail == INVALID_TAIL_PTR || OA_TAKEN(hw_tail, aged_tail) >= report_size)) { - struct i915_vma *vma = dev_priv->perf.oa.oa_buffer.vma; + struct i915_vma *vma = stream->oa_buffer.vma; u32 gtt_offset = i915_ggtt_offset(vma); /* Be paranoid and do a bounds check on the pointer read back @@ -525,16 +531,16 @@ static bool oa_buffer_check_unlocked(struct drm_i915_private *dev_priv) */ if (hw_tail >= gtt_offset && hw_tail < (gtt_offset + OA_BUFFER_SIZE)) { - dev_priv->perf.oa.oa_buffer.tails[!aged_idx].offset = + stream->oa_buffer.tails[!aged_idx].offset = aging_tail = hw_tail; - dev_priv->perf.oa.oa_buffer.aging_timestamp = now; + stream->oa_buffer.aging_timestamp = now; } else { DRM_ERROR("Ignoring spurious out of range OA buffer tail pointer = %u\n", hw_tail); } } - spin_unlock_irqrestore(&dev_priv->perf.oa.oa_buffer.ptr_lock, flags); + spin_unlock_irqrestore(&stream->oa_buffer.ptr_lock, flags); return aged_tail == INVALID_TAIL_PTR ? false : OA_TAKEN(aged_tail, head) >= report_size; @@ -597,8 +603,7 @@ static int append_oa_sample(struct i915_perf_stream *stream, size_t *offset, const u8 *report) { - struct drm_i915_private *dev_priv = stream->dev_priv; - int report_size = dev_priv->perf.oa.oa_buffer.format_size; + int report_size = stream->oa_buffer.format_size; struct drm_i915_perf_record_header header; u32 sample_flags = stream->sample_flags; @@ -650,9 +655,9 @@ static int gen8_append_oa_reports(struct i915_perf_stream *stream, size_t *offset) { struct drm_i915_private *dev_priv = stream->dev_priv; - int report_size = dev_priv->perf.oa.oa_buffer.format_size; - u8 *oa_buf_base = dev_priv->perf.oa.oa_buffer.vaddr; - u32 gtt_offset = i915_ggtt_offset(dev_priv->perf.oa.oa_buffer.vma); + int report_size = stream->oa_buffer.format_size; + u8 *oa_buf_base = stream->oa_buffer.vaddr; + u32 gtt_offset = i915_ggtt_offset(stream->oa_buffer.vma); u32 mask = (OA_BUFFER_SIZE - 1); size_t start_offset = *offset; unsigned long flags; @@ -664,13 +669,13 @@ static int gen8_append_oa_reports(struct i915_perf_stream *stream, if (WARN_ON(!stream->enabled)) return -EIO; - spin_lock_irqsave(&dev_priv->perf.oa.oa_buffer.ptr_lock, flags); + spin_lock_irqsave(&stream->oa_buffer.ptr_lock, flags); - head = dev_priv->perf.oa.oa_buffer.head; - aged_tail_idx = dev_priv->perf.oa.oa_buffer.aged_tail_idx; - tail = dev_priv->perf.oa.oa_buffer.tails[aged_tail_idx].offset; + head = stream->oa_buffer.head; + aged_tail_idx = stream->oa_buffer.aged_tail_idx; + tail = stream->oa_buffer.tails[aged_tail_idx].offset; - spin_unlock_irqrestore(&dev_priv->perf.oa.oa_buffer.ptr_lock, flags); + spin_unlock_irqrestore(&stream->oa_buffer.ptr_lock, flags); /* * An invalid tail pointer here means we're still waiting for the poll @@ -734,12 +739,12 @@ static int gen8_append_oa_reports(struct i915_perf_stream *stream, reason = ((report32[0] >> OAREPORT_REASON_SHIFT) & OAREPORT_REASON_MASK); if (reason == 0) { - if (__ratelimit(&dev_priv->perf.oa.spurious_report_rs)) + if (__ratelimit(&dev_priv->perf.spurious_report_rs)) DRM_NOTE("Skipping spurious, invalid OA report\n"); continue; } - ctx_id = report32[2] & dev_priv->perf.oa.specific_ctx_id_mask; + ctx_id = report32[2] & stream->specific_ctx_id_mask; /* * Squash whatever is in the CTX_ID field if it's marked as @@ -749,7 +754,7 @@ static int gen8_append_oa_reports(struct i915_perf_stream *stream, * Note: that we don't clear the valid_ctx_bit so userspace can * understand that the ID has been squashed by the kernel. */ - if (!(report32[0] & dev_priv->perf.oa.gen8_valid_ctx_bit)) + if (!(report32[0] & dev_priv->perf.gen8_valid_ctx_bit)) ctx_id = report32[2] = INVALID_CTX_ID; /* @@ -783,18 +788,17 @@ static int gen8_append_oa_reports(struct i915_perf_stream *stream, * switches since it's not-uncommon for periodic samples to * identify a switch before any 'context switch' report. */ - if (!dev_priv->perf.oa.exclusive_stream->ctx || - dev_priv->perf.oa.specific_ctx_id == ctx_id || - (dev_priv->perf.oa.oa_buffer.last_ctx_id == - dev_priv->perf.oa.specific_ctx_id) || + if (!dev_priv->perf.exclusive_stream->ctx || + stream->specific_ctx_id == ctx_id || + stream->oa_buffer.last_ctx_id == stream->specific_ctx_id || reason & OAREPORT_REASON_CTX_SWITCH) { /* * While filtering for a single context we avoid * leaking the IDs of other contexts. */ - if (dev_priv->perf.oa.exclusive_stream->ctx && - dev_priv->perf.oa.specific_ctx_id != ctx_id) { + if (dev_priv->perf.exclusive_stream->ctx && + stream->specific_ctx_id != ctx_id) { report32[2] = INVALID_CTX_ID; } @@ -803,7 +807,7 @@ static int gen8_append_oa_reports(struct i915_perf_stream *stream, if (ret) break; - dev_priv->perf.oa.oa_buffer.last_ctx_id = ctx_id; + stream->oa_buffer.last_ctx_id = ctx_id; } /* @@ -817,7 +821,7 @@ static int gen8_append_oa_reports(struct i915_perf_stream *stream, } if (start_offset != *offset) { - spin_lock_irqsave(&dev_priv->perf.oa.oa_buffer.ptr_lock, flags); + spin_lock_irqsave(&stream->oa_buffer.ptr_lock, flags); /* * We removed the gtt_offset for the copy loop above, indexing @@ -826,9 +830,9 @@ static int gen8_append_oa_reports(struct i915_perf_stream *stream, head += gtt_offset; I915_WRITE(GEN8_OAHEADPTR, head & GEN8_OAHEADPTR_MASK); - dev_priv->perf.oa.oa_buffer.head = head; + stream->oa_buffer.head = head; - spin_unlock_irqrestore(&dev_priv->perf.oa.oa_buffer.ptr_lock, flags); + spin_unlock_irqrestore(&stream->oa_buffer.ptr_lock, flags); } return ret; @@ -863,7 +867,7 @@ static int gen8_oa_read(struct i915_perf_stream *stream, u32 oastatus; int ret; - if (WARN_ON(!dev_priv->perf.oa.oa_buffer.vaddr)) + if (WARN_ON(!stream->oa_buffer.vaddr)) return -EIO; oastatus = I915_READ(GEN8_OASTATUS); @@ -889,10 +893,10 @@ static int gen8_oa_read(struct i915_perf_stream *stream, return ret; DRM_DEBUG("OA buffer overflow (exponent = %d): force restart\n", - dev_priv->perf.oa.period_exponent); + stream->period_exponent); - dev_priv->perf.oa.ops.oa_disable(stream); - dev_priv->perf.oa.ops.oa_enable(stream); + dev_priv->perf.ops.oa_disable(stream); + dev_priv->perf.ops.oa_enable(stream); /* * Note: .oa_enable() is expected to re-init the oabuffer and @@ -939,9 +943,9 @@ static int gen7_append_oa_reports(struct i915_perf_stream *stream, size_t *offset) { struct drm_i915_private *dev_priv = stream->dev_priv; - int report_size = dev_priv->perf.oa.oa_buffer.format_size; - u8 *oa_buf_base = dev_priv->perf.oa.oa_buffer.vaddr; - u32 gtt_offset = i915_ggtt_offset(dev_priv->perf.oa.oa_buffer.vma); + int report_size = stream->oa_buffer.format_size; + u8 *oa_buf_base = stream->oa_buffer.vaddr; + u32 gtt_offset = i915_ggtt_offset(stream->oa_buffer.vma); u32 mask = (OA_BUFFER_SIZE - 1); size_t start_offset = *offset; unsigned long flags; @@ -953,13 +957,13 @@ static int gen7_append_oa_reports(struct i915_perf_stream *stream, if (WARN_ON(!stream->enabled)) return -EIO; - spin_lock_irqsave(&dev_priv->perf.oa.oa_buffer.ptr_lock, flags); + spin_lock_irqsave(&stream->oa_buffer.ptr_lock, flags); - head = dev_priv->perf.oa.oa_buffer.head; - aged_tail_idx = dev_priv->perf.oa.oa_buffer.aged_tail_idx; - tail = dev_priv->perf.oa.oa_buffer.tails[aged_tail_idx].offset; + head = stream->oa_buffer.head; + aged_tail_idx = stream->oa_buffer.aged_tail_idx; + tail = stream->oa_buffer.tails[aged_tail_idx].offset; - spin_unlock_irqrestore(&dev_priv->perf.oa.oa_buffer.ptr_lock, flags); + spin_unlock_irqrestore(&stream->oa_buffer.ptr_lock, flags); /* An invalid tail pointer here means we're still waiting for the poll * hrtimer callback to give us a pointer @@ -1012,7 +1016,7 @@ static int gen7_append_oa_reports(struct i915_perf_stream *stream, * copying it to userspace... */ if (report32[0] == 0) { - if (__ratelimit(&dev_priv->perf.oa.spurious_report_rs)) + if (__ratelimit(&dev_priv->perf.spurious_report_rs)) DRM_NOTE("Skipping spurious, invalid OA report\n"); continue; } @@ -1031,7 +1035,7 @@ static int gen7_append_oa_reports(struct i915_perf_stream *stream, } if (start_offset != *offset) { - spin_lock_irqsave(&dev_priv->perf.oa.oa_buffer.ptr_lock, flags); + spin_lock_irqsave(&stream->oa_buffer.ptr_lock, flags); /* We removed the gtt_offset for the copy loop above, indexing * relative to oa_buf_base so put back here... @@ -1041,9 +1045,9 @@ static int gen7_append_oa_reports(struct i915_perf_stream *stream, I915_WRITE(GEN7_OASTATUS2, ((head & GEN7_OASTATUS2_HEAD_MASK) | GEN7_OASTATUS2_MEM_SELECT_GGTT)); - dev_priv->perf.oa.oa_buffer.head = head; + stream->oa_buffer.head = head; - spin_unlock_irqrestore(&dev_priv->perf.oa.oa_buffer.ptr_lock, flags); + spin_unlock_irqrestore(&stream->oa_buffer.ptr_lock, flags); } return ret; @@ -1074,7 +1078,7 @@ static int gen7_oa_read(struct i915_perf_stream *stream, u32 oastatus1; int ret; - if (WARN_ON(!dev_priv->perf.oa.oa_buffer.vaddr)) + if (WARN_ON(!stream->oa_buffer.vaddr)) return -EIO; oastatus1 = I915_READ(GEN7_OASTATUS1); @@ -1084,7 +1088,7 @@ static int gen7_oa_read(struct i915_perf_stream *stream, * may be updated asynchronously) so we ignore status bits * that have already been reported to userspace. */ - oastatus1 &= ~dev_priv->perf.oa.gen7_latched_oastatus1; + oastatus1 &= ~dev_priv->perf.gen7_latched_oastatus1; /* We treat OABUFFER_OVERFLOW as a significant error: * @@ -1113,10 +1117,10 @@ static int gen7_oa_read(struct i915_perf_stream *stream, return ret; DRM_DEBUG("OA buffer overflow (exponent = %d): force restart\n", - dev_priv->perf.oa.period_exponent); + stream->period_exponent); - dev_priv->perf.oa.ops.oa_disable(stream); - dev_priv->perf.oa.ops.oa_enable(stream); + dev_priv->perf.ops.oa_disable(stream); + dev_priv->perf.ops.oa_enable(stream); oastatus1 = I915_READ(GEN7_OASTATUS1); } @@ -1126,7 +1130,7 @@ static int gen7_oa_read(struct i915_perf_stream *stream, DRM_I915_PERF_RECORD_OA_REPORT_LOST); if (ret) return ret; - dev_priv->perf.oa.gen7_latched_oastatus1 |= + dev_priv->perf.gen7_latched_oastatus1 |= GEN7_OASTATUS1_REPORT_LOST; } @@ -1149,14 +1153,12 @@ static int gen7_oa_read(struct i915_perf_stream *stream, */ static int i915_oa_wait_unlocked(struct i915_perf_stream *stream) { - struct drm_i915_private *dev_priv = stream->dev_priv; - /* We would wait indefinitely if periodic sampling is not enabled */ - if (!dev_priv->perf.oa.periodic) + if (!stream->periodic) return -EIO; - return wait_event_interruptible(dev_priv->perf.oa.poll_wq, - oa_buffer_check_unlocked(dev_priv)); + return wait_event_interruptible(stream->poll_wq, + oa_buffer_check_unlocked(stream)); } /** @@ -1173,9 +1175,7 @@ static void i915_oa_poll_wait(struct i915_perf_stream *stream, struct file *file, poll_table *wait) { - struct drm_i915_private *dev_priv = stream->dev_priv; - - poll_wait(file, &dev_priv->perf.oa.poll_wq, wait); + poll_wait(file, &stream->poll_wq, wait); } /** @@ -1197,13 +1197,14 @@ static int i915_oa_read(struct i915_perf_stream *stream, { struct drm_i915_private *dev_priv = stream->dev_priv; - return dev_priv->perf.oa.ops.read(stream, buf, count, offset); + return dev_priv->perf.ops.read(stream, buf, count, offset); } -static struct intel_context *oa_pin_context(struct drm_i915_private *i915, - struct i915_gem_context *ctx) +static struct intel_context *oa_pin_context(struct i915_perf_stream *stream) { struct i915_gem_engines_iter it; + struct drm_i915_private *i915 = stream->dev_priv; + struct i915_gem_context *ctx = stream->ctx; struct intel_context *ce; int err; @@ -1221,7 +1222,7 @@ static struct intel_context *oa_pin_context(struct drm_i915_private *i915, */ err = intel_context_pin(ce); if (err == 0) { - i915->perf.oa.pinned_ctx = ce; + stream->pinned_ctx = ce; break; } } @@ -1231,7 +1232,7 @@ static struct intel_context *oa_pin_context(struct drm_i915_private *i915, if (err) return ERR_PTR(err); - return i915->perf.oa.pinned_ctx; + return stream->pinned_ctx; } /** @@ -1249,7 +1250,7 @@ static int oa_get_render_ctx_id(struct i915_perf_stream *stream) struct drm_i915_private *i915 = stream->dev_priv; struct intel_context *ce; - ce = oa_pin_context(i915, stream->ctx); + ce = oa_pin_context(stream); if (IS_ERR(ce)) return PTR_ERR(ce); @@ -1259,8 +1260,8 @@ static int oa_get_render_ctx_id(struct i915_perf_stream *stream) * On Haswell we don't do any post processing of the reports * and don't need to use the mask. */ - i915->perf.oa.specific_ctx_id = i915_ggtt_offset(ce->state); - i915->perf.oa.specific_ctx_id_mask = 0; + stream->specific_ctx_id = i915_ggtt_offset(ce->state); + stream->specific_ctx_id_mask = 0; break; } @@ -1278,33 +1279,33 @@ static int oa_get_render_ctx_id(struct i915_perf_stream *stream) * dropped by GuC. They won't be part of the context * ID in the OA reports, so squash those lower bits. */ - i915->perf.oa.specific_ctx_id = + stream->specific_ctx_id = lower_32_bits(ce->lrc_desc) >> 12; /* * GuC uses the top bit to signal proxy submission, so * ignore that bit. */ - i915->perf.oa.specific_ctx_id_mask = + stream->specific_ctx_id_mask = (1U << (GEN8_CTX_ID_WIDTH - 1)) - 1; } else { - i915->perf.oa.specific_ctx_id_mask = + stream->specific_ctx_id_mask = (1U << GEN8_CTX_ID_WIDTH) - 1; - i915->perf.oa.specific_ctx_id = + stream->specific_ctx_id = upper_32_bits(ce->lrc_desc); - i915->perf.oa.specific_ctx_id &= - i915->perf.oa.specific_ctx_id_mask; + stream->specific_ctx_id &= + stream->specific_ctx_id_mask; } break; case 11: { - i915->perf.oa.specific_ctx_id_mask = + stream->specific_ctx_id_mask = ((1U << GEN11_SW_CTX_ID_WIDTH) - 1) << (GEN11_SW_CTX_ID_SHIFT - 32) | ((1U << GEN11_ENGINE_INSTANCE_WIDTH) - 1) << (GEN11_ENGINE_INSTANCE_SHIFT - 32) | ((1 << GEN11_ENGINE_CLASS_WIDTH) - 1) << (GEN11_ENGINE_CLASS_SHIFT - 32); - i915->perf.oa.specific_ctx_id = upper_32_bits(ce->lrc_desc); - i915->perf.oa.specific_ctx_id &= - i915->perf.oa.specific_ctx_id_mask; + stream->specific_ctx_id = upper_32_bits(ce->lrc_desc); + stream->specific_ctx_id &= + stream->specific_ctx_id_mask; break; } @@ -1313,8 +1314,8 @@ static int oa_get_render_ctx_id(struct i915_perf_stream *stream) } DRM_DEBUG_DRIVER("filtering on ctx_id=0x%x ctx_id_mask=0x%x\n", - i915->perf.oa.specific_ctx_id, - i915->perf.oa.specific_ctx_id_mask); + stream->specific_ctx_id, + stream->specific_ctx_id_mask); return 0; } @@ -1331,10 +1332,10 @@ static void oa_put_render_ctx_id(struct i915_perf_stream *stream) struct drm_i915_private *dev_priv = stream->dev_priv; struct intel_context *ce; - dev_priv->perf.oa.specific_ctx_id = INVALID_CTX_ID; - dev_priv->perf.oa.specific_ctx_id_mask = 0; + stream->specific_ctx_id = INVALID_CTX_ID; + stream->specific_ctx_id_mask = 0; - ce = fetch_and_zero(&dev_priv->perf.oa.pinned_ctx); + ce = fetch_and_zero(&stream->pinned_ctx); if (ce) { mutex_lock(&dev_priv->drm.struct_mutex); intel_context_unpin(ce); @@ -1343,34 +1344,36 @@ static void oa_put_render_ctx_id(struct i915_perf_stream *stream) } static void -free_oa_buffer(struct drm_i915_private *i915) +free_oa_buffer(struct i915_perf_stream *stream) { + struct drm_i915_private *i915 = stream->dev_priv; + mutex_lock(&i915->drm.struct_mutex); - i915_vma_unpin_and_release(&i915->perf.oa.oa_buffer.vma, + i915_vma_unpin_and_release(&stream->oa_buffer.vma, I915_VMA_RELEASE_MAP); mutex_unlock(&i915->drm.struct_mutex); - i915->perf.oa.oa_buffer.vaddr = NULL; + stream->oa_buffer.vaddr = NULL; } static void i915_oa_stream_destroy(struct i915_perf_stream *stream) { struct drm_i915_private *dev_priv = stream->dev_priv; - BUG_ON(stream != dev_priv->perf.oa.exclusive_stream); + BUG_ON(stream != dev_priv->perf.exclusive_stream); /* * Unset exclusive_stream first, it will be checked while disabling * the metric set on gen8+. */ mutex_lock(&dev_priv->drm.struct_mutex); - dev_priv->perf.oa.exclusive_stream = NULL; - dev_priv->perf.oa.ops.disable_metric_set(dev_priv); + dev_priv->perf.exclusive_stream = NULL; + dev_priv->perf.ops.disable_metric_set(stream); mutex_unlock(&dev_priv->drm.struct_mutex); - free_oa_buffer(dev_priv); + free_oa_buffer(stream); intel_uncore_forcewake_put(&dev_priv->uncore, FORCEWAKE_ALL); intel_runtime_pm_put(&dev_priv->runtime_pm, stream->wakeref); @@ -1380,41 +1383,42 @@ static void i915_oa_stream_destroy(struct i915_perf_stream *stream) put_oa_config(dev_priv, stream->oa_config); - if (dev_priv->perf.oa.spurious_report_rs.missed) { + if (dev_priv->perf.spurious_report_rs.missed) { DRM_NOTE("%d spurious OA report notices suppressed due to ratelimiting\n", - dev_priv->perf.oa.spurious_report_rs.missed); + dev_priv->perf.spurious_report_rs.missed); } } -static void gen7_init_oa_buffer(struct drm_i915_private *dev_priv) +static void gen7_init_oa_buffer(struct i915_perf_stream *stream) { - u32 gtt_offset = i915_ggtt_offset(dev_priv->perf.oa.oa_buffer.vma); + struct drm_i915_private *dev_priv = stream->dev_priv; + u32 gtt_offset = i915_ggtt_offset(stream->oa_buffer.vma); unsigned long flags; - spin_lock_irqsave(&dev_priv->perf.oa.oa_buffer.ptr_lock, flags); + spin_lock_irqsave(&stream->oa_buffer.ptr_lock, flags); /* Pre-DevBDW: OABUFFER must be set with counters off, * before OASTATUS1, but after OASTATUS2 */ I915_WRITE(GEN7_OASTATUS2, gtt_offset | GEN7_OASTATUS2_MEM_SELECT_GGTT); /* head */ - dev_priv->perf.oa.oa_buffer.head = gtt_offset; + stream->oa_buffer.head = gtt_offset; I915_WRITE(GEN7_OABUFFER, gtt_offset); I915_WRITE(GEN7_OASTATUS1, gtt_offset | OABUFFER_SIZE_16M); /* tail */ /* Mark that we need updated tail pointers to read from... */ - dev_priv->perf.oa.oa_buffer.tails[0].offset = INVALID_TAIL_PTR; - dev_priv->perf.oa.oa_buffer.tails[1].offset = INVALID_TAIL_PTR; + stream->oa_buffer.tails[0].offset = INVALID_TAIL_PTR; + stream->oa_buffer.tails[1].offset = INVALID_TAIL_PTR; - spin_unlock_irqrestore(&dev_priv->perf.oa.oa_buffer.ptr_lock, flags); + spin_unlock_irqrestore(&stream->oa_buffer.ptr_lock, flags); /* On Haswell we have to track which OASTATUS1 flags we've * already seen since they can't be cleared while periodic * sampling is enabled. */ - dev_priv->perf.oa.gen7_latched_oastatus1 = 0; + dev_priv->perf.gen7_latched_oastatus1 = 0; /* NB: although the OA buffer will initially be allocated * zeroed via shmfs (and so this memset is redundant when @@ -1427,24 +1431,25 @@ static void gen7_init_oa_buffer(struct drm_i915_private *dev_priv) * the assumption that new reports are being written to zeroed * memory... */ - memset(dev_priv->perf.oa.oa_buffer.vaddr, 0, OA_BUFFER_SIZE); + memset(stream->oa_buffer.vaddr, 0, OA_BUFFER_SIZE); /* Maybe make ->pollin per-stream state if we support multiple * concurrent streams in the future. */ - dev_priv->perf.oa.pollin = false; + stream->pollin = false; } -static void gen8_init_oa_buffer(struct drm_i915_private *dev_priv) +static void gen8_init_oa_buffer(struct i915_perf_stream *stream) { - u32 gtt_offset = i915_ggtt_offset(dev_priv->perf.oa.oa_buffer.vma); + struct drm_i915_private *dev_priv = stream->dev_priv; + u32 gtt_offset = i915_ggtt_offset(stream->oa_buffer.vma); unsigned long flags; - spin_lock_irqsave(&dev_priv->perf.oa.oa_buffer.ptr_lock, flags); + spin_lock_irqsave(&stream->oa_buffer.ptr_lock, flags); I915_WRITE(GEN8_OASTATUS, 0); I915_WRITE(GEN8_OAHEADPTR, gtt_offset); - dev_priv->perf.oa.oa_buffer.head = gtt_offset; + stream->oa_buffer.head = gtt_offset; I915_WRITE(GEN8_OABUFFER_UDW, 0); @@ -1461,17 +1466,17 @@ static void gen8_init_oa_buffer(struct drm_i915_private *dev_priv) I915_WRITE(GEN8_OATAILPTR, gtt_offset & GEN8_OATAILPTR_MASK); /* Mark that we need updated tail pointers to read from... */ - dev_priv->perf.oa.oa_buffer.tails[0].offset = INVALID_TAIL_PTR; - dev_priv->perf.oa.oa_buffer.tails[1].offset = INVALID_TAIL_PTR; + stream->oa_buffer.tails[0].offset = INVALID_TAIL_PTR; + stream->oa_buffer.tails[1].offset = INVALID_TAIL_PTR; /* * Reset state used to recognise context switches, affecting which * reports we will forward to userspace while filtering for a single * context. */ - dev_priv->perf.oa.oa_buffer.last_ctx_id = INVALID_CTX_ID; + stream->oa_buffer.last_ctx_id = INVALID_CTX_ID; - spin_unlock_irqrestore(&dev_priv->perf.oa.oa_buffer.ptr_lock, flags); + spin_unlock_irqrestore(&stream->oa_buffer.ptr_lock, flags); /* * NB: although the OA buffer will initially be allocated @@ -1485,22 +1490,23 @@ static void gen8_init_oa_buffer(struct drm_i915_private *dev_priv) * the assumption that new reports are being written to zeroed * memory... */ - memset(dev_priv->perf.oa.oa_buffer.vaddr, 0, OA_BUFFER_SIZE); + memset(stream->oa_buffer.vaddr, 0, OA_BUFFER_SIZE); /* * Maybe make ->pollin per-stream state if we support multiple * concurrent streams in the future. */ - dev_priv->perf.oa.pollin = false; + stream->pollin = false; } -static int alloc_oa_buffer(struct drm_i915_private *dev_priv) +static int alloc_oa_buffer(struct i915_perf_stream *stream) { struct drm_i915_gem_object *bo; + struct drm_i915_private *dev_priv = stream->dev_priv; struct i915_vma *vma; int ret; - if (WARN_ON(dev_priv->perf.oa.oa_buffer.vma)) + if (WARN_ON(stream->oa_buffer.vma)) return -ENODEV; ret = i915_mutex_lock_interruptible(&dev_priv->drm); @@ -1525,18 +1531,18 @@ static int alloc_oa_buffer(struct drm_i915_private *dev_priv) ret = PTR_ERR(vma); goto err_unref; } - dev_priv->perf.oa.oa_buffer.vma = vma; + stream->oa_buffer.vma = vma; - dev_priv->perf.oa.oa_buffer.vaddr = + stream->oa_buffer.vaddr = i915_gem_object_pin_map(bo, I915_MAP_WB); - if (IS_ERR(dev_priv->perf.oa.oa_buffer.vaddr)) { - ret = PTR_ERR(dev_priv->perf.oa.oa_buffer.vaddr); + if (IS_ERR(stream->oa_buffer.vaddr)) { + ret = PTR_ERR(stream->oa_buffer.vaddr); goto err_unpin; } DRM_DEBUG_DRIVER("OA Buffer initialized, gtt offset = 0x%x, vaddr = %p\n", - i915_ggtt_offset(dev_priv->perf.oa.oa_buffer.vma), - dev_priv->perf.oa.oa_buffer.vaddr); + i915_ggtt_offset(stream->oa_buffer.vma), + stream->oa_buffer.vaddr); goto unlock; @@ -1546,8 +1552,8 @@ err_unpin: err_unref: i915_gem_object_put(bo); - dev_priv->perf.oa.oa_buffer.vaddr = NULL; - dev_priv->perf.oa.oa_buffer.vma = NULL; + stream->oa_buffer.vaddr = NULL; + stream->oa_buffer.vma = NULL; unlock: mutex_unlock(&dev_priv->drm.struct_mutex); @@ -1623,8 +1629,10 @@ static int hsw_enable_metric_set(struct i915_perf_stream *stream) return 0; } -static void hsw_disable_metric_set(struct drm_i915_private *dev_priv) +static void hsw_disable_metric_set(struct i915_perf_stream *stream) { + struct drm_i915_private *dev_priv = stream->dev_priv; + I915_WRITE(GEN6_UCGCTL1, (I915_READ(GEN6_UCGCTL1) & ~GEN6_CSUNIT_CLOCK_GATE_DISABLE)); I915_WRITE(GEN7_MISCCPCTL, (I915_READ(GEN7_MISCCPCTL) | @@ -1663,13 +1671,14 @@ static u32 oa_config_flex_reg(const struct i915_oa_config *oa_config, * in the case that the OA unit has been disabled. */ static void -gen8_update_reg_state_unlocked(struct intel_context *ce, +gen8_update_reg_state_unlocked(struct i915_perf_stream *stream, + struct intel_context *ce, u32 *reg_state, const struct i915_oa_config *oa_config) { struct drm_i915_private *i915 = ce->engine->i915; - u32 ctx_oactxctrl = i915->perf.oa.ctx_oactxctrl_offset; - u32 ctx_flexeu0 = i915->perf.oa.ctx_flexeu0_offset; + u32 ctx_oactxctrl = i915->perf.ctx_oactxctrl_offset; + u32 ctx_flexeu0 = i915->perf.ctx_flexeu0_offset; /* The MMIO offsets for Flex EU registers aren't contiguous */ i915_reg_t flex_regs[] = { EU_PERF_CNTL0, @@ -1683,8 +1692,8 @@ gen8_update_reg_state_unlocked(struct intel_context *ce, int i; CTX_REG(reg_state, ctx_oactxctrl, GEN8_OACTXCONTROL, - (i915->perf.oa.period_exponent << GEN8_OA_TIMER_PERIOD_SHIFT) | - (i915->perf.oa.periodic ? GEN8_OA_TIMER_ENABLE : 0) | + (stream->period_exponent << GEN8_OA_TIMER_PERIOD_SHIFT) | + (stream->periodic ? GEN8_OA_TIMER_ENABLE : 0) | GEN8_OA_COUNTER_RESUME); for (i = 0; i < ARRAY_SIZE(flex_regs); i++) { @@ -1846,11 +1855,12 @@ static int gen8_configure_context(struct i915_gem_context *ctx, * * Note: it's only the RCS/Render context that has any OA state. */ -static int gen8_configure_all_contexts(struct drm_i915_private *i915, +static int gen8_configure_all_contexts(struct i915_perf_stream *stream, const struct i915_oa_config *oa_config) { + struct drm_i915_private *i915 = stream->dev_priv; /* The MMIO offsets for Flex EU registers aren't contiguous */ - const u32 ctx_flexeu0 = i915->perf.oa.ctx_flexeu0_offset; + const u32 ctx_flexeu0 = i915->perf.ctx_flexeu0_offset; #define ctx_flexeuN(N) (ctx_flexeu0 + 2 * (N)) struct flex regs[] = { { @@ -1859,9 +1869,9 @@ static int gen8_configure_all_contexts(struct drm_i915_private *i915, }, { GEN8_OACTXCONTROL, - i915->perf.oa.ctx_oactxctrl_offset, - ((i915->perf.oa.period_exponent << GEN8_OA_TIMER_PERIOD_SHIFT) | - (i915->perf.oa.periodic ? GEN8_OA_TIMER_ENABLE : 0) | + i915->perf.ctx_oactxctrl_offset, + ((stream->period_exponent << GEN8_OA_TIMER_PERIOD_SHIFT) | + (stream->periodic ? GEN8_OA_TIMER_ENABLE : 0) | GEN8_OA_COUNTER_RESUME) }, { EU_PERF_CNTL0, ctx_flexeuN(0) }, @@ -1971,7 +1981,7 @@ static int gen8_enable_metric_set(struct i915_perf_stream *stream) * to make sure all slices/subslices are ON before writing to NOA * registers. */ - ret = gen8_configure_all_contexts(dev_priv, oa_config); + ret = gen8_configure_all_contexts(stream, oa_config); if (ret) return ret; @@ -1984,19 +1994,23 @@ static int gen8_enable_metric_set(struct i915_perf_stream *stream) return 0; } -static void gen8_disable_metric_set(struct drm_i915_private *dev_priv) +static void gen8_disable_metric_set(struct i915_perf_stream *stream) { + struct drm_i915_private *dev_priv = stream->dev_priv; + /* Reset all contexts' slices/subslices configurations. */ - gen8_configure_all_contexts(dev_priv, NULL); + gen8_configure_all_contexts(stream, NULL); I915_WRITE(GDT_CHICKEN_BITS, (I915_READ(GDT_CHICKEN_BITS) & ~GT_NOA_ENABLE)); } -static void gen10_disable_metric_set(struct drm_i915_private *dev_priv) +static void gen10_disable_metric_set(struct i915_perf_stream *stream) { + struct drm_i915_private *dev_priv = stream->dev_priv; + /* Reset all contexts' slices/subslices configurations. */ - gen8_configure_all_contexts(dev_priv, NULL); + gen8_configure_all_contexts(stream, NULL); /* Make sure we disable noa to save power. */ I915_WRITE(RPM_CONFIG1, @@ -2007,10 +2021,10 @@ static void gen7_oa_enable(struct i915_perf_stream *stream) { struct drm_i915_private *dev_priv = stream->dev_priv; struct i915_gem_context *ctx = stream->ctx; - u32 ctx_id = dev_priv->perf.oa.specific_ctx_id; - bool periodic = dev_priv->perf.oa.periodic; - u32 period_exponent = dev_priv->perf.oa.period_exponent; - u32 report_format = dev_priv->perf.oa.oa_buffer.format; + u32 ctx_id = stream->specific_ctx_id; + bool periodic = stream->periodic; + u32 period_exponent = stream->period_exponent; + u32 report_format = stream->oa_buffer.format; /* * Reset buf pointers so we don't forward reports from before now. @@ -2021,7 +2035,7 @@ static void gen7_oa_enable(struct i915_perf_stream *stream) * on the assumption that certain fields are written to zeroed * memory which this helps maintains. */ - gen7_init_oa_buffer(dev_priv); + gen7_init_oa_buffer(stream); I915_WRITE(GEN7_OACONTROL, (ctx_id & GEN7_OACONTROL_CTX_MASK) | @@ -2036,7 +2050,7 @@ static void gen7_oa_enable(struct i915_perf_stream *stream) static void gen8_oa_enable(struct i915_perf_stream *stream) { struct drm_i915_private *dev_priv = stream->dev_priv; - u32 report_format = dev_priv->perf.oa.oa_buffer.format; + u32 report_format = stream->oa_buffer.format; /* * Reset buf pointers so we don't forward reports from before now. @@ -2047,7 +2061,7 @@ static void gen8_oa_enable(struct i915_perf_stream *stream) * on the assumption that certain fields are written to zeroed * memory which this helps maintains. */ - gen8_init_oa_buffer(dev_priv); + gen8_init_oa_buffer(stream); /* * Note: we don't rely on the hardware to perform single context @@ -2072,10 +2086,10 @@ static void i915_oa_stream_enable(struct i915_perf_stream *stream) { struct drm_i915_private *dev_priv = stream->dev_priv; - dev_priv->perf.oa.ops.oa_enable(stream); + dev_priv->perf.ops.oa_enable(stream); - if (dev_priv->perf.oa.periodic) - hrtimer_start(&dev_priv->perf.oa.poll_check_timer, + if (stream->periodic) + hrtimer_start(&stream->poll_check_timer, ns_to_ktime(POLL_PERIOD), HRTIMER_MODE_REL_PINNED); } @@ -2114,10 +2128,10 @@ static void i915_oa_stream_disable(struct i915_perf_stream *stream) { struct drm_i915_private *dev_priv = stream->dev_priv; - dev_priv->perf.oa.ops.oa_disable(stream); + dev_priv->perf.ops.oa_disable(stream); - if (dev_priv->perf.oa.periodic) - hrtimer_cancel(&dev_priv->perf.oa.poll_check_timer); + if (stream->periodic) + hrtimer_cancel(&stream->poll_check_timer); } static const struct i915_perf_stream_ops i915_oa_stream_ops = { @@ -2169,7 +2183,7 @@ static int i915_oa_stream_init(struct i915_perf_stream *stream, return -EINVAL; } - if (!dev_priv->perf.oa.ops.enable_metric_set) { + if (!dev_priv->perf.ops.enable_metric_set) { DRM_DEBUG("OA unit not supported\n"); return -ENODEV; } @@ -2178,7 +2192,7 @@ static int i915_oa_stream_init(struct i915_perf_stream *stream, * counter reports and marshal to the appropriate client * we currently only allow exclusive access */ - if (dev_priv->perf.oa.exclusive_stream) { + if (dev_priv->perf.exclusive_stream) { DRM_DEBUG("OA unit already in use\n"); return -EBUSY; } @@ -2188,43 +2202,23 @@ static int i915_oa_stream_init(struct i915_perf_stream *stream, return -EINVAL; } - /* We set up some ratelimit state to potentially throttle any _NOTES - * about spurious, invalid OA reports which we don't forward to - * userspace. - * - * The initialization is associated with opening the stream (not driver - * init) considering we print a _NOTE about any throttling when closing - * the stream instead of waiting until driver _fini which no one would - * ever see. - * - * Using the same limiting factors as printk_ratelimit() - */ - ratelimit_state_init(&dev_priv->perf.oa.spurious_report_rs, - 5 * HZ, 10); - /* Since we use a DRM_NOTE for spurious reports it would be - * inconsistent to let __ratelimit() automatically print a warning for - * throttling. - */ - ratelimit_set_flags(&dev_priv->perf.oa.spurious_report_rs, - RATELIMIT_MSG_ON_RELEASE); - stream->sample_size = sizeof(struct drm_i915_perf_record_header); - format_size = dev_priv->perf.oa.oa_formats[props->oa_format].size; + format_size = dev_priv->perf.oa_formats[props->oa_format].size; stream->sample_flags |= SAMPLE_OA_REPORT; stream->sample_size += format_size; - dev_priv->perf.oa.oa_buffer.format_size = format_size; - if (WARN_ON(dev_priv->perf.oa.oa_buffer.format_size == 0)) + stream->oa_buffer.format_size = format_size; + if (WARN_ON(stream->oa_buffer.format_size == 0)) return -EINVAL; - dev_priv->perf.oa.oa_buffer.format = - dev_priv->perf.oa.oa_formats[props->oa_format].format; + stream->oa_buffer.format = + dev_priv->perf.oa_formats[props->oa_format].format; - dev_priv->perf.oa.periodic = props->oa_periodic; - if (dev_priv->perf.oa.periodic) - dev_priv->perf.oa.period_exponent = props->oa_period_exponent; + stream->periodic = props->oa_periodic; + if (stream->periodic) + stream->period_exponent = props->oa_period_exponent; if (stream->ctx) { ret = oa_get_render_ctx_id(stream); @@ -2255,7 +2249,7 @@ static int i915_oa_stream_init(struct i915_perf_stream *stream, stream->wakeref = intel_runtime_pm_get(&dev_priv->runtime_pm); intel_uncore_forcewake_get(&dev_priv->uncore, FORCEWAKE_ALL); - ret = alloc_oa_buffer(dev_priv); + ret = alloc_oa_buffer(stream); if (ret) goto err_oa_buf_alloc; @@ -2264,9 +2258,9 @@ static int i915_oa_stream_init(struct i915_perf_stream *stream, goto err_lock; stream->ops = &i915_oa_stream_ops; - dev_priv->perf.oa.exclusive_stream = stream; + dev_priv->perf.exclusive_stream = stream; - ret = dev_priv->perf.oa.ops.enable_metric_set(stream); + ret = dev_priv->perf.ops.enable_metric_set(stream); if (ret) { DRM_DEBUG("Unable to enable metric set\n"); goto err_enable; @@ -2274,15 +2268,21 @@ static int i915_oa_stream_init(struct i915_perf_stream *stream, mutex_unlock(&dev_priv->drm.struct_mutex); + hrtimer_init(&stream->poll_check_timer, + CLOCK_MONOTONIC, HRTIMER_MODE_REL); + stream->poll_check_timer.function = oa_poll_check_timer_cb; + init_waitqueue_head(&stream->poll_wq); + spin_lock_init(&stream->oa_buffer.ptr_lock); + return 0; err_enable: - dev_priv->perf.oa.exclusive_stream = NULL; - dev_priv->perf.oa.ops.disable_metric_set(dev_priv); + dev_priv->perf.exclusive_stream = NULL; + dev_priv->perf.ops.disable_metric_set(stream); mutex_unlock(&dev_priv->drm.struct_mutex); err_lock: - free_oa_buffer(dev_priv); + free_oa_buffer(stream); err_oa_buf_alloc: put_oa_config(dev_priv, stream->oa_config); @@ -2306,9 +2306,9 @@ void i915_oa_init_reg_state(struct intel_engine_cs *engine, if (engine->class != RENDER_CLASS) return; - stream = engine->i915->perf.oa.exclusive_stream; + stream = engine->i915->perf.exclusive_stream; if (stream) - gen8_update_reg_state_unlocked(ce, regs, stream->oa_config); + gen8_update_reg_state_unlocked(stream, ce, regs, stream->oa_config); } /** @@ -2424,7 +2424,7 @@ static ssize_t i915_perf_read(struct file *file, /* Maybe make ->pollin per-stream state if we support multiple * concurrent streams in the future. */ - dev_priv->perf.oa.pollin = false; + stream->pollin = false; } return ret; @@ -2432,13 +2432,12 @@ static ssize_t i915_perf_read(struct file *file, static enum hrtimer_restart oa_poll_check_timer_cb(struct hrtimer *hrtimer) { - struct drm_i915_private *dev_priv = - container_of(hrtimer, typeof(*dev_priv), - perf.oa.poll_check_timer); + struct i915_perf_stream *stream = + container_of(hrtimer, typeof(*stream), poll_check_timer); - if (oa_buffer_check_unlocked(dev_priv)) { - dev_priv->perf.oa.pollin = true; - wake_up(&dev_priv->perf.oa.poll_wq); + if (oa_buffer_check_unlocked(stream)) { + stream->pollin = true; + wake_up(&stream->poll_wq); } hrtimer_forward_now(hrtimer, ns_to_ktime(POLL_PERIOD)); @@ -2477,7 +2476,7 @@ static __poll_t i915_perf_poll_locked(struct drm_i915_private *dev_priv, * the hrtimer/oa_poll_check_timer_cb to notify us when there are * samples to read. */ - if (dev_priv->perf.oa.pollin) + if (stream->pollin) events |= EPOLLIN; return events; @@ -2903,7 +2902,7 @@ static int read_properties_unlocked(struct drm_i915_private *dev_priv, value); return -EINVAL; } - if (!dev_priv->perf.oa.oa_formats[value].size) { + if (!dev_priv->perf.oa_formats[value].size) { DRM_DEBUG("Unsupported OA report format %llu\n", value); return -EINVAL; @@ -3047,7 +3046,7 @@ void i915_perf_register(struct drm_i915_private *dev_priv) if (!dev_priv->perf.metrics_kobj) goto exit; - sysfs_attr_init(&dev_priv->perf.oa.test_config.sysfs_metric_id.attr); + sysfs_attr_init(&dev_priv->perf.test_config.sysfs_metric_id.attr); if (INTEL_GEN(dev_priv) >= 11) { i915_perf_load_test_config_icl(dev_priv); @@ -3082,15 +3081,15 @@ void i915_perf_register(struct drm_i915_private *dev_priv) i915_perf_load_test_config_hsw(dev_priv); } - if (dev_priv->perf.oa.test_config.id == 0) + if (dev_priv->perf.test_config.id == 0) goto sysfs_error; ret = sysfs_create_group(dev_priv->perf.metrics_kobj, - &dev_priv->perf.oa.test_config.sysfs_metric); + &dev_priv->perf.test_config.sysfs_metric); if (ret) goto sysfs_error; - atomic_set(&dev_priv->perf.oa.test_config.ref_count, 1); + atomic_set(&dev_priv->perf.test_config.ref_count, 1); goto exit; @@ -3117,7 +3116,7 @@ void i915_perf_unregister(struct drm_i915_private *dev_priv) return; sysfs_remove_group(dev_priv->perf.metrics_kobj, - &dev_priv->perf.oa.test_config.sysfs_metric); + &dev_priv->perf.test_config.sysfs_metric); kobject_put(dev_priv->perf.metrics_kobj); dev_priv->perf.metrics_kobj = NULL; @@ -3362,7 +3361,7 @@ int i915_perf_add_config_ioctl(struct drm_device *dev, void *data, oa_config->mux_regs_len = args->n_mux_regs; oa_config->mux_regs = alloc_oa_regs(dev_priv, - dev_priv->perf.oa.ops.is_valid_mux_reg, + dev_priv->perf.ops.is_valid_mux_reg, u64_to_user_ptr(args->mux_regs_ptr), args->n_mux_regs); @@ -3375,7 +3374,7 @@ int i915_perf_add_config_ioctl(struct drm_device *dev, void *data, oa_config->b_counter_regs_len = args->n_boolean_regs; oa_config->b_counter_regs = alloc_oa_regs(dev_priv, - dev_priv->perf.oa.ops.is_valid_b_counter_reg, + dev_priv->perf.ops.is_valid_b_counter_reg, u64_to_user_ptr(args->boolean_regs_ptr), args->n_boolean_regs); @@ -3394,7 +3393,7 @@ int i915_perf_add_config_ioctl(struct drm_device *dev, void *data, oa_config->flex_regs_len = args->n_flex_regs; oa_config->flex_regs = alloc_oa_regs(dev_priv, - dev_priv->perf.oa.ops.is_valid_flex_reg, + dev_priv->perf.ops.is_valid_flex_reg, u64_to_user_ptr(args->flex_regs_ptr), args->n_flex_regs); @@ -3561,20 +3560,20 @@ static struct ctl_table dev_root[] = { void i915_perf_init(struct drm_i915_private *dev_priv) { if (IS_HASWELL(dev_priv)) { - dev_priv->perf.oa.ops.is_valid_b_counter_reg = + dev_priv->perf.ops.is_valid_b_counter_reg = gen7_is_valid_b_counter_addr; - dev_priv->perf.oa.ops.is_valid_mux_reg = + dev_priv->perf.ops.is_valid_mux_reg = hsw_is_valid_mux_addr; - dev_priv->perf.oa.ops.is_valid_flex_reg = NULL; - dev_priv->perf.oa.ops.enable_metric_set = hsw_enable_metric_set; - dev_priv->perf.oa.ops.disable_metric_set = hsw_disable_metric_set; - dev_priv->perf.oa.ops.oa_enable = gen7_oa_enable; - dev_priv->perf.oa.ops.oa_disable = gen7_oa_disable; - dev_priv->perf.oa.ops.read = gen7_oa_read; - dev_priv->perf.oa.ops.oa_hw_tail_read = + dev_priv->perf.ops.is_valid_flex_reg = NULL; + dev_priv->perf.ops.enable_metric_set = hsw_enable_metric_set; + dev_priv->perf.ops.disable_metric_set = hsw_disable_metric_set; + dev_priv->perf.ops.oa_enable = gen7_oa_enable; + dev_priv->perf.ops.oa_disable = gen7_oa_disable; + dev_priv->perf.ops.read = gen7_oa_read; + dev_priv->perf.ops.oa_hw_tail_read = gen7_oa_hw_tail_read; - dev_priv->perf.oa.oa_formats = hsw_oa_formats; + dev_priv->perf.oa_formats = hsw_oa_formats; } else if (HAS_LOGICAL_RING_CONTEXTS(dev_priv)) { /* Note: that although we could theoretically also support the * legacy ringbuffer mode on BDW (and earlier iterations of @@ -3582,71 +3581,65 @@ void i915_perf_init(struct drm_i915_private *dev_priv) * worth the complexity to maintain now that BDW+ enable * execlist mode by default. */ - dev_priv->perf.oa.oa_formats = gen8_plus_oa_formats; + dev_priv->perf.oa_formats = gen8_plus_oa_formats; - dev_priv->perf.oa.ops.oa_enable = gen8_oa_enable; - dev_priv->perf.oa.ops.oa_disable = gen8_oa_disable; - dev_priv->perf.oa.ops.read = gen8_oa_read; - dev_priv->perf.oa.ops.oa_hw_tail_read = gen8_oa_hw_tail_read; + dev_priv->perf.ops.oa_enable = gen8_oa_enable; + dev_priv->perf.ops.oa_disable = gen8_oa_disable; + dev_priv->perf.ops.read = gen8_oa_read; + dev_priv->perf.ops.oa_hw_tail_read = gen8_oa_hw_tail_read; if (IS_GEN_RANGE(dev_priv, 8, 9)) { - dev_priv->perf.oa.ops.is_valid_b_counter_reg = + dev_priv->perf.ops.is_valid_b_counter_reg = gen7_is_valid_b_counter_addr; - dev_priv->perf.oa.ops.is_valid_mux_reg = + dev_priv->perf.ops.is_valid_mux_reg = gen8_is_valid_mux_addr; - dev_priv->perf.oa.ops.is_valid_flex_reg = + dev_priv->perf.ops.is_valid_flex_reg = gen8_is_valid_flex_addr; if (IS_CHERRYVIEW(dev_priv)) { - dev_priv->perf.oa.ops.is_valid_mux_reg = + dev_priv->perf.ops.is_valid_mux_reg = chv_is_valid_mux_addr; } - dev_priv->perf.oa.ops.enable_metric_set = gen8_enable_metric_set; - dev_priv->perf.oa.ops.disable_metric_set = gen8_disable_metric_set; + dev_priv->perf.ops.enable_metric_set = gen8_enable_metric_set; + dev_priv->perf.ops.disable_metric_set = gen8_disable_metric_set; if (IS_GEN(dev_priv, 8)) { - dev_priv->perf.oa.ctx_oactxctrl_offset = 0x120; - dev_priv->perf.oa.ctx_flexeu0_offset = 0x2ce; + dev_priv->perf.ctx_oactxctrl_offset = 0x120; + dev_priv->perf.ctx_flexeu0_offset = 0x2ce; - dev_priv->perf.oa.gen8_valid_ctx_bit = (1<<25); + dev_priv->perf.gen8_valid_ctx_bit = BIT(25); } else { - dev_priv->perf.oa.ctx_oactxctrl_offset = 0x128; - dev_priv->perf.oa.ctx_flexeu0_offset = 0x3de; + dev_priv->perf.ctx_oactxctrl_offset = 0x128; + dev_priv->perf.ctx_flexeu0_offset = 0x3de; - dev_priv->perf.oa.gen8_valid_ctx_bit = (1<<16); + dev_priv->perf.gen8_valid_ctx_bit = BIT(16); } } else if (IS_GEN_RANGE(dev_priv, 10, 11)) { - dev_priv->perf.oa.ops.is_valid_b_counter_reg = + dev_priv->perf.ops.is_valid_b_counter_reg = gen7_is_valid_b_counter_addr; - dev_priv->perf.oa.ops.is_valid_mux_reg = + dev_priv->perf.ops.is_valid_mux_reg = gen10_is_valid_mux_addr; - dev_priv->perf.oa.ops.is_valid_flex_reg = + dev_priv->perf.ops.is_valid_flex_reg = gen8_is_valid_flex_addr; - dev_priv->perf.oa.ops.enable_metric_set = gen8_enable_metric_set; - dev_priv->perf.oa.ops.disable_metric_set = gen10_disable_metric_set; + dev_priv->perf.ops.enable_metric_set = gen8_enable_metric_set; + dev_priv->perf.ops.disable_metric_set = gen10_disable_metric_set; if (IS_GEN(dev_priv, 10)) { - dev_priv->perf.oa.ctx_oactxctrl_offset = 0x128; - dev_priv->perf.oa.ctx_flexeu0_offset = 0x3de; + dev_priv->perf.ctx_oactxctrl_offset = 0x128; + dev_priv->perf.ctx_flexeu0_offset = 0x3de; } else { - dev_priv->perf.oa.ctx_oactxctrl_offset = 0x124; - dev_priv->perf.oa.ctx_flexeu0_offset = 0x78e; + dev_priv->perf.ctx_oactxctrl_offset = 0x124; + dev_priv->perf.ctx_flexeu0_offset = 0x78e; } - dev_priv->perf.oa.gen8_valid_ctx_bit = (1<<16); + dev_priv->perf.gen8_valid_ctx_bit = BIT(16); } } - if (dev_priv->perf.oa.ops.enable_metric_set) { - hrtimer_init(&dev_priv->perf.oa.poll_check_timer, - CLOCK_MONOTONIC, HRTIMER_MODE_REL); - dev_priv->perf.oa.poll_check_timer.function = oa_poll_check_timer_cb; - init_waitqueue_head(&dev_priv->perf.oa.poll_wq); - + if (dev_priv->perf.ops.enable_metric_set) { INIT_LIST_HEAD(&dev_priv->perf.streams); mutex_init(&dev_priv->perf.lock); - spin_lock_init(&dev_priv->perf.oa.oa_buffer.ptr_lock); oa_sample_rate_hard_limit = 1000 * (RUNTIME_INFO(dev_priv)->cs_timestamp_frequency_khz / 2); @@ -3655,6 +3648,25 @@ void i915_perf_init(struct drm_i915_private *dev_priv) mutex_init(&dev_priv->perf.metrics_lock); idr_init(&dev_priv->perf.metrics_idr); + /* We set up some ratelimit state to potentially throttle any + * _NOTES about spurious, invalid OA reports which we don't + * forward to userspace. + * + * We print a _NOTE about any throttling when closing the + * stream instead of waiting until driver _fini which no one + * would ever see. + * + * Using the same limiting factors as printk_ratelimit() + */ + ratelimit_state_init(&dev_priv->perf.spurious_report_rs, + 5 * HZ, 10); + /* Since we use a DRM_NOTE for spurious reports it would be + * inconsistent to let __ratelimit() automatically print a + * warning for throttling. + */ + ratelimit_set_flags(&dev_priv->perf.spurious_report_rs, + RATELIMIT_MSG_ON_RELEASE); + dev_priv->perf.initialized = true; } } @@ -3683,7 +3695,7 @@ void i915_perf_fini(struct drm_i915_private *dev_priv) unregister_sysctl_table(dev_priv->perf.sysctl_header); - memset(&dev_priv->perf.oa.ops, 0, sizeof(dev_priv->perf.oa.ops)); + memset(&dev_priv->perf.ops, 0, sizeof(dev_priv->perf.ops)); dev_priv->perf.initialized = false; } diff --git a/drivers/gpu/drm/i915/oa/i915_oa_bdw.c b/drivers/gpu/drm/i915/oa/i915_oa_bdw.c index 440b6b9058da..14da5c3b569d 100644 --- a/drivers/gpu/drm/i915/oa/i915_oa_bdw.c +++ b/drivers/gpu/drm/i915/oa/i915_oa_bdw.c @@ -65,26 +65,26 @@ show_test_oa_id(struct device *kdev, struct device_attribute *attr, char *buf) void i915_perf_load_test_config_bdw(struct drm_i915_private *dev_priv) { - strlcpy(dev_priv->perf.oa.test_config.uuid, + strlcpy(dev_priv->perf.test_config.uuid, "d6de6f55-e526-4f79-a6a6-d7315c09044e", - sizeof(dev_priv->perf.oa.test_config.uuid)); - dev_priv->perf.oa.test_config.id = 1; + sizeof(dev_priv->perf.test_config.uuid)); + dev_priv->perf.test_config.id = 1; - dev_priv->perf.oa.test_config.mux_regs = mux_config_test_oa; - dev_priv->perf.oa.test_config.mux_regs_len = ARRAY_SIZE(mux_config_test_oa); + dev_priv->perf.test_config.mux_regs = mux_config_test_oa; + dev_priv->perf.test_config.mux_regs_len = ARRAY_SIZE(mux_config_test_oa); - dev_priv->perf.oa.test_config.b_counter_regs = b_counter_config_test_oa; - dev_priv->perf.oa.test_config.b_counter_regs_len = ARRAY_SIZE(b_counter_config_test_oa); + dev_priv->perf.test_config.b_counter_regs = b_counter_config_test_oa; + dev_priv->perf.test_config.b_counter_regs_len = ARRAY_SIZE(b_counter_config_test_oa); - dev_priv->perf.oa.test_config.flex_regs = flex_eu_config_test_oa; - dev_priv->perf.oa.test_config.flex_regs_len = ARRAY_SIZE(flex_eu_config_test_oa); + dev_priv->perf.test_config.flex_regs = flex_eu_config_test_oa; + dev_priv->perf.test_config.flex_regs_len = ARRAY_SIZE(flex_eu_config_test_oa); - dev_priv->perf.oa.test_config.sysfs_metric.name = "d6de6f55-e526-4f79-a6a6-d7315c09044e"; - dev_priv->perf.oa.test_config.sysfs_metric.attrs = dev_priv->perf.oa.test_config.attrs; + dev_priv->perf.test_config.sysfs_metric.name = "d6de6f55-e526-4f79-a6a6-d7315c09044e"; + dev_priv->perf.test_config.sysfs_metric.attrs = dev_priv->perf.test_config.attrs; - dev_priv->perf.oa.test_config.attrs[0] = &dev_priv->perf.oa.test_config.sysfs_metric_id.attr; + dev_priv->perf.test_config.attrs[0] = &dev_priv->perf.test_config.sysfs_metric_id.attr; - dev_priv->perf.oa.test_config.sysfs_metric_id.attr.name = "id"; - dev_priv->perf.oa.test_config.sysfs_metric_id.attr.mode = 0444; - dev_priv->perf.oa.test_config.sysfs_metric_id.show = show_test_oa_id; + dev_priv->perf.test_config.sysfs_metric_id.attr.name = "id"; + dev_priv->perf.test_config.sysfs_metric_id.attr.mode = 0444; + dev_priv->perf.test_config.sysfs_metric_id.show = show_test_oa_id; } diff --git a/drivers/gpu/drm/i915/oa/i915_oa_bxt.c b/drivers/gpu/drm/i915/oa/i915_oa_bxt.c index 0b268e3fe2ab..3e785bafcf99 100644 --- a/drivers/gpu/drm/i915/oa/i915_oa_bxt.c +++ b/drivers/gpu/drm/i915/oa/i915_oa_bxt.c @@ -63,26 +63,26 @@ show_test_oa_id(struct device *kdev, struct device_attribute *attr, char *buf) void i915_perf_load_test_config_bxt(struct drm_i915_private *dev_priv) { - strlcpy(dev_priv->perf.oa.test_config.uuid, + strlcpy(dev_priv->perf.test_config.uuid, "5ee72f5c-092f-421e-8b70-225f7c3e9612", - sizeof(dev_priv->perf.oa.test_config.uuid)); - dev_priv->perf.oa.test_config.id = 1; + sizeof(dev_priv->perf.test_config.uuid)); + dev_priv->perf.test_config.id = 1; - dev_priv->perf.oa.test_config.mux_regs = mux_config_test_oa; - dev_priv->perf.oa.test_config.mux_regs_len = ARRAY_SIZE(mux_config_test_oa); + dev_priv->perf.test_config.mux_regs = mux_config_test_oa; + dev_priv->perf.test_config.mux_regs_len = ARRAY_SIZE(mux_config_test_oa); - dev_priv->perf.oa.test_config.b_counter_regs = b_counter_config_test_oa; - dev_priv->perf.oa.test_config.b_counter_regs_len = ARRAY_SIZE(b_counter_config_test_oa); + dev_priv->perf.test_config.b_counter_regs = b_counter_config_test_oa; + dev_priv->perf.test_config.b_counter_regs_len = ARRAY_SIZE(b_counter_config_test_oa); - dev_priv->perf.oa.test_config.flex_regs = flex_eu_config_test_oa; - dev_priv->perf.oa.test_config.flex_regs_len = ARRAY_SIZE(flex_eu_config_test_oa); + dev_priv->perf.test_config.flex_regs = flex_eu_config_test_oa; + dev_priv->perf.test_config.flex_regs_len = ARRAY_SIZE(flex_eu_config_test_oa); - dev_priv->perf.oa.test_config.sysfs_metric.name = "5ee72f5c-092f-421e-8b70-225f7c3e9612"; - dev_priv->perf.oa.test_config.sysfs_metric.attrs = dev_priv->perf.oa.test_config.attrs; + dev_priv->perf.test_config.sysfs_metric.name = "5ee72f5c-092f-421e-8b70-225f7c3e9612"; + dev_priv->perf.test_config.sysfs_metric.attrs = dev_priv->perf.test_config.attrs; - dev_priv->perf.oa.test_config.attrs[0] = &dev_priv->perf.oa.test_config.sysfs_metric_id.attr; + dev_priv->perf.test_config.attrs[0] = &dev_priv->perf.test_config.sysfs_metric_id.attr; - dev_priv->perf.oa.test_config.sysfs_metric_id.attr.name = "id"; - dev_priv->perf.oa.test_config.sysfs_metric_id.attr.mode = 0444; - dev_priv->perf.oa.test_config.sysfs_metric_id.show = show_test_oa_id; + dev_priv->perf.test_config.sysfs_metric_id.attr.name = "id"; + dev_priv->perf.test_config.sysfs_metric_id.attr.mode = 0444; + dev_priv->perf.test_config.sysfs_metric_id.show = show_test_oa_id; } diff --git a/drivers/gpu/drm/i915/oa/i915_oa_cflgt2.c b/drivers/gpu/drm/i915/oa/i915_oa_cflgt2.c index 89542d30c55c..0ea86f70a06c 100644 --- a/drivers/gpu/drm/i915/oa/i915_oa_cflgt2.c +++ b/drivers/gpu/drm/i915/oa/i915_oa_cflgt2.c @@ -64,26 +64,26 @@ show_test_oa_id(struct device *kdev, struct device_attribute *attr, char *buf) void i915_perf_load_test_config_cflgt2(struct drm_i915_private *dev_priv) { - strlcpy(dev_priv->perf.oa.test_config.uuid, + strlcpy(dev_priv->perf.test_config.uuid, "74fb4902-d3d3-4237-9e90-cbdc68d0a446", - sizeof(dev_priv->perf.oa.test_config.uuid)); - dev_priv->perf.oa.test_config.id = 1; + sizeof(dev_priv->perf.test_config.uuid)); + dev_priv->perf.test_config.id = 1; - dev_priv->perf.oa.test_config.mux_regs = mux_config_test_oa; - dev_priv->perf.oa.test_config.mux_regs_len = ARRAY_SIZE(mux_config_test_oa); + dev_priv->perf.test_config.mux_regs = mux_config_test_oa; + dev_priv->perf.test_config.mux_regs_len = ARRAY_SIZE(mux_config_test_oa); - dev_priv->perf.oa.test_config.b_counter_regs = b_counter_config_test_oa; - dev_priv->perf.oa.test_config.b_counter_regs_len = ARRAY_SIZE(b_counter_config_test_oa); + dev_priv->perf.test_config.b_counter_regs = b_counter_config_test_oa; + dev_priv->perf.test_config.b_counter_regs_len = ARRAY_SIZE(b_counter_config_test_oa); - dev_priv->perf.oa.test_config.flex_regs = flex_eu_config_test_oa; - dev_priv->perf.oa.test_config.flex_regs_len = ARRAY_SIZE(flex_eu_config_test_oa); + dev_priv->perf.test_config.flex_regs = flex_eu_config_test_oa; + dev_priv->perf.test_config.flex_regs_len = ARRAY_SIZE(flex_eu_config_test_oa); - dev_priv->perf.oa.test_config.sysfs_metric.name = "74fb4902-d3d3-4237-9e90-cbdc68d0a446"; - dev_priv->perf.oa.test_config.sysfs_metric.attrs = dev_priv->perf.oa.test_config.attrs; + dev_priv->perf.test_config.sysfs_metric.name = "74fb4902-d3d3-4237-9e90-cbdc68d0a446"; + dev_priv->perf.test_config.sysfs_metric.attrs = dev_priv->perf.test_config.attrs; - dev_priv->perf.oa.test_config.attrs[0] = &dev_priv->perf.oa.test_config.sysfs_metric_id.attr; + dev_priv->perf.test_config.attrs[0] = &dev_priv->perf.test_config.sysfs_metric_id.attr; - dev_priv->perf.oa.test_config.sysfs_metric_id.attr.name = "id"; - dev_priv->perf.oa.test_config.sysfs_metric_id.attr.mode = 0444; - dev_priv->perf.oa.test_config.sysfs_metric_id.show = show_test_oa_id; + dev_priv->perf.test_config.sysfs_metric_id.attr.name = "id"; + dev_priv->perf.test_config.sysfs_metric_id.attr.mode = 0444; + dev_priv->perf.test_config.sysfs_metric_id.show = show_test_oa_id; } diff --git a/drivers/gpu/drm/i915/oa/i915_oa_cflgt3.c b/drivers/gpu/drm/i915/oa/i915_oa_cflgt3.c index 51b118b03716..fc632dd890bf 100644 --- a/drivers/gpu/drm/i915/oa/i915_oa_cflgt3.c +++ b/drivers/gpu/drm/i915/oa/i915_oa_cflgt3.c @@ -64,26 +64,26 @@ show_test_oa_id(struct device *kdev, struct device_attribute *attr, char *buf) void i915_perf_load_test_config_cflgt3(struct drm_i915_private *dev_priv) { - strlcpy(dev_priv->perf.oa.test_config.uuid, + strlcpy(dev_priv->perf.test_config.uuid, "577e8e2c-3fa0-4875-8743-3538d585e3b0", - sizeof(dev_priv->perf.oa.test_config.uuid)); - dev_priv->perf.oa.test_config.id = 1; + sizeof(dev_priv->perf.test_config.uuid)); + dev_priv->perf.test_config.id = 1; - dev_priv->perf.oa.test_config.mux_regs = mux_config_test_oa; - dev_priv->perf.oa.test_config.mux_regs_len = ARRAY_SIZE(mux_config_test_oa); + dev_priv->perf.test_config.mux_regs = mux_config_test_oa; + dev_priv->perf.test_config.mux_regs_len = ARRAY_SIZE(mux_config_test_oa); - dev_priv->perf.oa.test_config.b_counter_regs = b_counter_config_test_oa; - dev_priv->perf.oa.test_config.b_counter_regs_len = ARRAY_SIZE(b_counter_config_test_oa); + dev_priv->perf.test_config.b_counter_regs = b_counter_config_test_oa; + dev_priv->perf.test_config.b_counter_regs_len = ARRAY_SIZE(b_counter_config_test_oa); - dev_priv->perf.oa.test_config.flex_regs = flex_eu_config_test_oa; - dev_priv->perf.oa.test_config.flex_regs_len = ARRAY_SIZE(flex_eu_config_test_oa); + dev_priv->perf.test_config.flex_regs = flex_eu_config_test_oa; + dev_priv->perf.test_config.flex_regs_len = ARRAY_SIZE(flex_eu_config_test_oa); - dev_priv->perf.oa.test_config.sysfs_metric.name = "577e8e2c-3fa0-4875-8743-3538d585e3b0"; - dev_priv->perf.oa.test_config.sysfs_metric.attrs = dev_priv->perf.oa.test_config.attrs; + dev_priv->perf.test_config.sysfs_metric.name = "577e8e2c-3fa0-4875-8743-3538d585e3b0"; + dev_priv->perf.test_config.sysfs_metric.attrs = dev_priv->perf.test_config.attrs; - dev_priv->perf.oa.test_config.attrs[0] = &dev_priv->perf.oa.test_config.sysfs_metric_id.attr; + dev_priv->perf.test_config.attrs[0] = &dev_priv->perf.test_config.sysfs_metric_id.attr; - dev_priv->perf.oa.test_config.sysfs_metric_id.attr.name = "id"; - dev_priv->perf.oa.test_config.sysfs_metric_id.attr.mode = 0444; - dev_priv->perf.oa.test_config.sysfs_metric_id.show = show_test_oa_id; + dev_priv->perf.test_config.sysfs_metric_id.attr.name = "id"; + dev_priv->perf.test_config.sysfs_metric_id.attr.mode = 0444; + dev_priv->perf.test_config.sysfs_metric_id.show = show_test_oa_id; } diff --git a/drivers/gpu/drm/i915/oa/i915_oa_chv.c b/drivers/gpu/drm/i915/oa/i915_oa_chv.c index c70c5af8a765..6cd4e9921a8a 100644 --- a/drivers/gpu/drm/i915/oa/i915_oa_chv.c +++ b/drivers/gpu/drm/i915/oa/i915_oa_chv.c @@ -64,26 +64,26 @@ show_test_oa_id(struct device *kdev, struct device_attribute *attr, char *buf) void i915_perf_load_test_config_chv(struct drm_i915_private *dev_priv) { - strlcpy(dev_priv->perf.oa.test_config.uuid, + strlcpy(dev_priv->perf.test_config.uuid, "4a534b07-cba3-414d-8d60-874830e883aa", - sizeof(dev_priv->perf.oa.test_config.uuid)); - dev_priv->perf.oa.test_config.id = 1; + sizeof(dev_priv->perf.test_config.uuid)); + dev_priv->perf.test_config.id = 1; - dev_priv->perf.oa.test_config.mux_regs = mux_config_test_oa; - dev_priv->perf.oa.test_config.mux_regs_len = ARRAY_SIZE(mux_config_test_oa); + dev_priv->perf.test_config.mux_regs = mux_config_test_oa; + dev_priv->perf.test_config.mux_regs_len = ARRAY_SIZE(mux_config_test_oa); - dev_priv->perf.oa.test_config.b_counter_regs = b_counter_config_test_oa; - dev_priv->perf.oa.test_config.b_counter_regs_len = ARRAY_SIZE(b_counter_config_test_oa); + dev_priv->perf.test_config.b_counter_regs = b_counter_config_test_oa; + dev_priv->perf.test_config.b_counter_regs_len = ARRAY_SIZE(b_counter_config_test_oa); - dev_priv->perf.oa.test_config.flex_regs = flex_eu_config_test_oa; - dev_priv->perf.oa.test_config.flex_regs_len = ARRAY_SIZE(flex_eu_config_test_oa); + dev_priv->perf.test_config.flex_regs = flex_eu_config_test_oa; + dev_priv->perf.test_config.flex_regs_len = ARRAY_SIZE(flex_eu_config_test_oa); - dev_priv->perf.oa.test_config.sysfs_metric.name = "4a534b07-cba3-414d-8d60-874830e883aa"; - dev_priv->perf.oa.test_config.sysfs_metric.attrs = dev_priv->perf.oa.test_config.attrs; + dev_priv->perf.test_config.sysfs_metric.name = "4a534b07-cba3-414d-8d60-874830e883aa"; + dev_priv->perf.test_config.sysfs_metric.attrs = dev_priv->perf.test_config.attrs; - dev_priv->perf.oa.test_config.attrs[0] = &dev_priv->perf.oa.test_config.sysfs_metric_id.attr; + dev_priv->perf.test_config.attrs[0] = &dev_priv->perf.test_config.sysfs_metric_id.attr; - dev_priv->perf.oa.test_config.sysfs_metric_id.attr.name = "id"; - dev_priv->perf.oa.test_config.sysfs_metric_id.attr.mode = 0444; - dev_priv->perf.oa.test_config.sysfs_metric_id.show = show_test_oa_id; + dev_priv->perf.test_config.sysfs_metric_id.attr.name = "id"; + dev_priv->perf.test_config.sysfs_metric_id.attr.mode = 0444; + dev_priv->perf.test_config.sysfs_metric_id.show = show_test_oa_id; } diff --git a/drivers/gpu/drm/i915/oa/i915_oa_cnl.c b/drivers/gpu/drm/i915/oa/i915_oa_cnl.c index 6d13983600e7..1041e8914993 100644 --- a/drivers/gpu/drm/i915/oa/i915_oa_cnl.c +++ b/drivers/gpu/drm/i915/oa/i915_oa_cnl.c @@ -76,26 +76,26 @@ show_test_oa_id(struct device *kdev, struct device_attribute *attr, char *buf) void i915_perf_load_test_config_cnl(struct drm_i915_private *dev_priv) { - strlcpy(dev_priv->perf.oa.test_config.uuid, + strlcpy(dev_priv->perf.test_config.uuid, "db41edd4-d8e7-4730-ad11-b9a2d6833503", - sizeof(dev_priv->perf.oa.test_config.uuid)); - dev_priv->perf.oa.test_config.id = 1; + sizeof(dev_priv->perf.test_config.uuid)); + dev_priv->perf.test_config.id = 1; - dev_priv->perf.oa.test_config.mux_regs = mux_config_test_oa; - dev_priv->perf.oa.test_config.mux_regs_len = ARRAY_SIZE(mux_config_test_oa); + dev_priv->perf.test_config.mux_regs = mux_config_test_oa; + dev_priv->perf.test_config.mux_regs_len = ARRAY_SIZE(mux_config_test_oa); - dev_priv->perf.oa.test_config.b_counter_regs = b_counter_config_test_oa; - dev_priv->perf.oa.test_config.b_counter_regs_len = ARRAY_SIZE(b_counter_config_test_oa); + dev_priv->perf.test_config.b_counter_regs = b_counter_config_test_oa; + dev_priv->perf.test_config.b_counter_regs_len = ARRAY_SIZE(b_counter_config_test_oa); - dev_priv->perf.oa.test_config.flex_regs = flex_eu_config_test_oa; - dev_priv->perf.oa.test_config.flex_regs_len = ARRAY_SIZE(flex_eu_config_test_oa); + dev_priv->perf.test_config.flex_regs = flex_eu_config_test_oa; + dev_priv->perf.test_config.flex_regs_len = ARRAY_SIZE(flex_eu_config_test_oa); - dev_priv->perf.oa.test_config.sysfs_metric.name = "db41edd4-d8e7-4730-ad11-b9a2d6833503"; - dev_priv->perf.oa.test_config.sysfs_metric.attrs = dev_priv->perf.oa.test_config.attrs; + dev_priv->perf.test_config.sysfs_metric.name = "db41edd4-d8e7-4730-ad11-b9a2d6833503"; + dev_priv->perf.test_config.sysfs_metric.attrs = dev_priv->perf.test_config.attrs; - dev_priv->perf.oa.test_config.attrs[0] = &dev_priv->perf.oa.test_config.sysfs_metric_id.attr; + dev_priv->perf.test_config.attrs[0] = &dev_priv->perf.test_config.sysfs_metric_id.attr; - dev_priv->perf.oa.test_config.sysfs_metric_id.attr.name = "id"; - dev_priv->perf.oa.test_config.sysfs_metric_id.attr.mode = 0444; - dev_priv->perf.oa.test_config.sysfs_metric_id.show = show_test_oa_id; + dev_priv->perf.test_config.sysfs_metric_id.attr.name = "id"; + dev_priv->perf.test_config.sysfs_metric_id.attr.mode = 0444; + dev_priv->perf.test_config.sysfs_metric_id.show = show_test_oa_id; } diff --git a/drivers/gpu/drm/i915/oa/i915_oa_glk.c b/drivers/gpu/drm/i915/oa/i915_oa_glk.c index 668c9aa15bc5..bd15ebe9aeeb 100644 --- a/drivers/gpu/drm/i915/oa/i915_oa_glk.c +++ b/drivers/gpu/drm/i915/oa/i915_oa_glk.c @@ -63,26 +63,26 @@ show_test_oa_id(struct device *kdev, struct device_attribute *attr, char *buf) void i915_perf_load_test_config_glk(struct drm_i915_private *dev_priv) { - strlcpy(dev_priv->perf.oa.test_config.uuid, + strlcpy(dev_priv->perf.test_config.uuid, "dd3fd789-e783-4204-8cd0-b671bbccb0cf", - sizeof(dev_priv->perf.oa.test_config.uuid)); - dev_priv->perf.oa.test_config.id = 1; + sizeof(dev_priv->perf.test_config.uuid)); + dev_priv->perf.test_config.id = 1; - dev_priv->perf.oa.test_config.mux_regs = mux_config_test_oa; - dev_priv->perf.oa.test_config.mux_regs_len = ARRAY_SIZE(mux_config_test_oa); + dev_priv->perf.test_config.mux_regs = mux_config_test_oa; + dev_priv->perf.test_config.mux_regs_len = ARRAY_SIZE(mux_config_test_oa); - dev_priv->perf.oa.test_config.b_counter_regs = b_counter_config_test_oa; - dev_priv->perf.oa.test_config.b_counter_regs_len = ARRAY_SIZE(b_counter_config_test_oa); + dev_priv->perf.test_config.b_counter_regs = b_counter_config_test_oa; + dev_priv->perf.test_config.b_counter_regs_len = ARRAY_SIZE(b_counter_config_test_oa); - dev_priv->perf.oa.test_config.flex_regs = flex_eu_config_test_oa; - dev_priv->perf.oa.test_config.flex_regs_len = ARRAY_SIZE(flex_eu_config_test_oa); + dev_priv->perf.test_config.flex_regs = flex_eu_config_test_oa; + dev_priv->perf.test_config.flex_regs_len = ARRAY_SIZE(flex_eu_config_test_oa); - dev_priv->perf.oa.test_config.sysfs_metric.name = "dd3fd789-e783-4204-8cd0-b671bbccb0cf"; - dev_priv->perf.oa.test_config.sysfs_metric.attrs = dev_priv->perf.oa.test_config.attrs; + dev_priv->perf.test_config.sysfs_metric.name = "dd3fd789-e783-4204-8cd0-b671bbccb0cf"; + dev_priv->perf.test_config.sysfs_metric.attrs = dev_priv->perf.test_config.attrs; - dev_priv->perf.oa.test_config.attrs[0] = &dev_priv->perf.oa.test_config.sysfs_metric_id.attr; + dev_priv->perf.test_config.attrs[0] = &dev_priv->perf.test_config.sysfs_metric_id.attr; - dev_priv->perf.oa.test_config.sysfs_metric_id.attr.name = "id"; - dev_priv->perf.oa.test_config.sysfs_metric_id.attr.mode = 0444; - dev_priv->perf.oa.test_config.sysfs_metric_id.show = show_test_oa_id; + dev_priv->perf.test_config.sysfs_metric_id.attr.name = "id"; + dev_priv->perf.test_config.sysfs_metric_id.attr.mode = 0444; + dev_priv->perf.test_config.sysfs_metric_id.show = show_test_oa_id; } diff --git a/drivers/gpu/drm/i915/oa/i915_oa_hsw.c b/drivers/gpu/drm/i915/oa/i915_oa_hsw.c index 2e10a73127a2..133721a8619f 100644 --- a/drivers/gpu/drm/i915/oa/i915_oa_hsw.c +++ b/drivers/gpu/drm/i915/oa/i915_oa_hsw.c @@ -93,26 +93,26 @@ show_render_basic_id(struct device *kdev, struct device_attribute *attr, char *b void i915_perf_load_test_config_hsw(struct drm_i915_private *dev_priv) { - strlcpy(dev_priv->perf.oa.test_config.uuid, + strlcpy(dev_priv->perf.test_config.uuid, "403d8832-1a27-4aa6-a64e-f5389ce7b212", - sizeof(dev_priv->perf.oa.test_config.uuid)); - dev_priv->perf.oa.test_config.id = 1; + sizeof(dev_priv->perf.test_config.uuid)); + dev_priv->perf.test_config.id = 1; - dev_priv->perf.oa.test_config.mux_regs = mux_config_render_basic; - dev_priv->perf.oa.test_config.mux_regs_len = ARRAY_SIZE(mux_config_render_basic); + dev_priv->perf.test_config.mux_regs = mux_config_render_basic; + dev_priv->perf.test_config.mux_regs_len = ARRAY_SIZE(mux_config_render_basic); - dev_priv->perf.oa.test_config.b_counter_regs = b_counter_config_render_basic; - dev_priv->perf.oa.test_config.b_counter_regs_len = ARRAY_SIZE(b_counter_config_render_basic); + dev_priv->perf.test_config.b_counter_regs = b_counter_config_render_basic; + dev_priv->perf.test_config.b_counter_regs_len = ARRAY_SIZE(b_counter_config_render_basic); - dev_priv->perf.oa.test_config.flex_regs = flex_eu_config_render_basic; - dev_priv->perf.oa.test_config.flex_regs_len = ARRAY_SIZE(flex_eu_config_render_basic); + dev_priv->perf.test_config.flex_regs = flex_eu_config_render_basic; + dev_priv->perf.test_config.flex_regs_len = ARRAY_SIZE(flex_eu_config_render_basic); - dev_priv->perf.oa.test_config.sysfs_metric.name = "403d8832-1a27-4aa6-a64e-f5389ce7b212"; - dev_priv->perf.oa.test_config.sysfs_metric.attrs = dev_priv->perf.oa.test_config.attrs; + dev_priv->perf.test_config.sysfs_metric.name = "403d8832-1a27-4aa6-a64e-f5389ce7b212"; + dev_priv->perf.test_config.sysfs_metric.attrs = dev_priv->perf.test_config.attrs; - dev_priv->perf.oa.test_config.attrs[0] = &dev_priv->perf.oa.test_config.sysfs_metric_id.attr; + dev_priv->perf.test_config.attrs[0] = &dev_priv->perf.test_config.sysfs_metric_id.attr; - dev_priv->perf.oa.test_config.sysfs_metric_id.attr.name = "id"; - dev_priv->perf.oa.test_config.sysfs_metric_id.attr.mode = 0444; - dev_priv->perf.oa.test_config.sysfs_metric_id.show = show_render_basic_id; + dev_priv->perf.test_config.sysfs_metric_id.attr.name = "id"; + dev_priv->perf.test_config.sysfs_metric_id.attr.mode = 0444; + dev_priv->perf.test_config.sysfs_metric_id.show = show_render_basic_id; } diff --git a/drivers/gpu/drm/i915/oa/i915_oa_icl.c b/drivers/gpu/drm/i915/oa/i915_oa_icl.c index 6dc388de4518..2d92041b754f 100644 --- a/drivers/gpu/drm/i915/oa/i915_oa_icl.c +++ b/drivers/gpu/drm/i915/oa/i915_oa_icl.c @@ -73,26 +73,26 @@ show_test_oa_id(struct device *kdev, struct device_attribute *attr, char *buf) void i915_perf_load_test_config_icl(struct drm_i915_private *dev_priv) { - strlcpy(dev_priv->perf.oa.test_config.uuid, + strlcpy(dev_priv->perf.test_config.uuid, "a291665e-244b-4b76-9b9a-01de9d3c8068", - sizeof(dev_priv->perf.oa.test_config.uuid)); - dev_priv->perf.oa.test_config.id = 1; + sizeof(dev_priv->perf.test_config.uuid)); + dev_priv->perf.test_config.id = 1; - dev_priv->perf.oa.test_config.mux_regs = mux_config_test_oa; - dev_priv->perf.oa.test_config.mux_regs_len = ARRAY_SIZE(mux_config_test_oa); + dev_priv->perf.test_config.mux_regs = mux_config_test_oa; + dev_priv->perf.test_config.mux_regs_len = ARRAY_SIZE(mux_config_test_oa); - dev_priv->perf.oa.test_config.b_counter_regs = b_counter_config_test_oa; - dev_priv->perf.oa.test_config.b_counter_regs_len = ARRAY_SIZE(b_counter_config_test_oa); + dev_priv->perf.test_config.b_counter_regs = b_counter_config_test_oa; + dev_priv->perf.test_config.b_counter_regs_len = ARRAY_SIZE(b_counter_config_test_oa); - dev_priv->perf.oa.test_config.flex_regs = flex_eu_config_test_oa; - dev_priv->perf.oa.test_config.flex_regs_len = ARRAY_SIZE(flex_eu_config_test_oa); + dev_priv->perf.test_config.flex_regs = flex_eu_config_test_oa; + dev_priv->perf.test_config.flex_regs_len = ARRAY_SIZE(flex_eu_config_test_oa); - dev_priv->perf.oa.test_config.sysfs_metric.name = "a291665e-244b-4b76-9b9a-01de9d3c8068"; - dev_priv->perf.oa.test_config.sysfs_metric.attrs = dev_priv->perf.oa.test_config.attrs; + dev_priv->perf.test_config.sysfs_metric.name = "a291665e-244b-4b76-9b9a-01de9d3c8068"; + dev_priv->perf.test_config.sysfs_metric.attrs = dev_priv->perf.test_config.attrs; - dev_priv->perf.oa.test_config.attrs[0] = &dev_priv->perf.oa.test_config.sysfs_metric_id.attr; + dev_priv->perf.test_config.attrs[0] = &dev_priv->perf.test_config.sysfs_metric_id.attr; - dev_priv->perf.oa.test_config.sysfs_metric_id.attr.name = "id"; - dev_priv->perf.oa.test_config.sysfs_metric_id.attr.mode = 0444; - dev_priv->perf.oa.test_config.sysfs_metric_id.show = show_test_oa_id; + dev_priv->perf.test_config.sysfs_metric_id.attr.name = "id"; + dev_priv->perf.test_config.sysfs_metric_id.attr.mode = 0444; + dev_priv->perf.test_config.sysfs_metric_id.show = show_test_oa_id; } diff --git a/drivers/gpu/drm/i915/oa/i915_oa_kblgt2.c b/drivers/gpu/drm/i915/oa/i915_oa_kblgt2.c index 45b76366c4f3..1c3a67c9cfe0 100644 --- a/drivers/gpu/drm/i915/oa/i915_oa_kblgt2.c +++ b/drivers/gpu/drm/i915/oa/i915_oa_kblgt2.c @@ -64,26 +64,26 @@ show_test_oa_id(struct device *kdev, struct device_attribute *attr, char *buf) void i915_perf_load_test_config_kblgt2(struct drm_i915_private *dev_priv) { - strlcpy(dev_priv->perf.oa.test_config.uuid, + strlcpy(dev_priv->perf.test_config.uuid, "baa3c7e4-52b6-4b85-801e-465a94b746dd", - sizeof(dev_priv->perf.oa.test_config.uuid)); - dev_priv->perf.oa.test_config.id = 1; + sizeof(dev_priv->perf.test_config.uuid)); + dev_priv->perf.test_config.id = 1; - dev_priv->perf.oa.test_config.mux_regs = mux_config_test_oa; - dev_priv->perf.oa.test_config.mux_regs_len = ARRAY_SIZE(mux_config_test_oa); + dev_priv->perf.test_config.mux_regs = mux_config_test_oa; + dev_priv->perf.test_config.mux_regs_len = ARRAY_SIZE(mux_config_test_oa); - dev_priv->perf.oa.test_config.b_counter_regs = b_counter_config_test_oa; - dev_priv->perf.oa.test_config.b_counter_regs_len = ARRAY_SIZE(b_counter_config_test_oa); + dev_priv->perf.test_config.b_counter_regs = b_counter_config_test_oa; + dev_priv->perf.test_config.b_counter_regs_len = ARRAY_SIZE(b_counter_config_test_oa); - dev_priv->perf.oa.test_config.flex_regs = flex_eu_config_test_oa; - dev_priv->perf.oa.test_config.flex_regs_len = ARRAY_SIZE(flex_eu_config_test_oa); + dev_priv->perf.test_config.flex_regs = flex_eu_config_test_oa; + dev_priv->perf.test_config.flex_regs_len = ARRAY_SIZE(flex_eu_config_test_oa); - dev_priv->perf.oa.test_config.sysfs_metric.name = "baa3c7e4-52b6-4b85-801e-465a94b746dd"; - dev_priv->perf.oa.test_config.sysfs_metric.attrs = dev_priv->perf.oa.test_config.attrs; + dev_priv->perf.test_config.sysfs_metric.name = "baa3c7e4-52b6-4b85-801e-465a94b746dd"; + dev_priv->perf.test_config.sysfs_metric.attrs = dev_priv->perf.test_config.attrs; - dev_priv->perf.oa.test_config.attrs[0] = &dev_priv->perf.oa.test_config.sysfs_metric_id.attr; + dev_priv->perf.test_config.attrs[0] = &dev_priv->perf.test_config.sysfs_metric_id.attr; - dev_priv->perf.oa.test_config.sysfs_metric_id.attr.name = "id"; - dev_priv->perf.oa.test_config.sysfs_metric_id.attr.mode = 0444; - dev_priv->perf.oa.test_config.sysfs_metric_id.show = show_test_oa_id; + dev_priv->perf.test_config.sysfs_metric_id.attr.name = "id"; + dev_priv->perf.test_config.sysfs_metric_id.attr.mode = 0444; + dev_priv->perf.test_config.sysfs_metric_id.show = show_test_oa_id; } diff --git a/drivers/gpu/drm/i915/oa/i915_oa_kblgt3.c b/drivers/gpu/drm/i915/oa/i915_oa_kblgt3.c index 35a340f48353..ebbe5a9c9fdc 100644 --- a/drivers/gpu/drm/i915/oa/i915_oa_kblgt3.c +++ b/drivers/gpu/drm/i915/oa/i915_oa_kblgt3.c @@ -64,26 +64,26 @@ show_test_oa_id(struct device *kdev, struct device_attribute *attr, char *buf) void i915_perf_load_test_config_kblgt3(struct drm_i915_private *dev_priv) { - strlcpy(dev_priv->perf.oa.test_config.uuid, + strlcpy(dev_priv->perf.test_config.uuid, "f1792f32-6db2-4b50-b4b2-557128f1688d", - sizeof(dev_priv->perf.oa.test_config.uuid)); - dev_priv->perf.oa.test_config.id = 1; + sizeof(dev_priv->perf.test_config.uuid)); + dev_priv->perf.test_config.id = 1; - dev_priv->perf.oa.test_config.mux_regs = mux_config_test_oa; - dev_priv->perf.oa.test_config.mux_regs_len = ARRAY_SIZE(mux_config_test_oa); + dev_priv->perf.test_config.mux_regs = mux_config_test_oa; + dev_priv->perf.test_config.mux_regs_len = ARRAY_SIZE(mux_config_test_oa); - dev_priv->perf.oa.test_config.b_counter_regs = b_counter_config_test_oa; - dev_priv->perf.oa.test_config.b_counter_regs_len = ARRAY_SIZE(b_counter_config_test_oa); + dev_priv->perf.test_config.b_counter_regs = b_counter_config_test_oa; + dev_priv->perf.test_config.b_counter_regs_len = ARRAY_SIZE(b_counter_config_test_oa); - dev_priv->perf.oa.test_config.flex_regs = flex_eu_config_test_oa; - dev_priv->perf.oa.test_config.flex_regs_len = ARRAY_SIZE(flex_eu_config_test_oa); + dev_priv->perf.test_config.flex_regs = flex_eu_config_test_oa; + dev_priv->perf.test_config.flex_regs_len = ARRAY_SIZE(flex_eu_config_test_oa); - dev_priv->perf.oa.test_config.sysfs_metric.name = "f1792f32-6db2-4b50-b4b2-557128f1688d"; - dev_priv->perf.oa.test_config.sysfs_metric.attrs = dev_priv->perf.oa.test_config.attrs; + dev_priv->perf.test_config.sysfs_metric.name = "f1792f32-6db2-4b50-b4b2-557128f1688d"; + dev_priv->perf.test_config.sysfs_metric.attrs = dev_priv->perf.test_config.attrs; - dev_priv->perf.oa.test_config.attrs[0] = &dev_priv->perf.oa.test_config.sysfs_metric_id.attr; + dev_priv->perf.test_config.attrs[0] = &dev_priv->perf.test_config.sysfs_metric_id.attr; - dev_priv->perf.oa.test_config.sysfs_metric_id.attr.name = "id"; - dev_priv->perf.oa.test_config.sysfs_metric_id.attr.mode = 0444; - dev_priv->perf.oa.test_config.sysfs_metric_id.show = show_test_oa_id; + dev_priv->perf.test_config.sysfs_metric_id.attr.name = "id"; + dev_priv->perf.test_config.sysfs_metric_id.attr.mode = 0444; + dev_priv->perf.test_config.sysfs_metric_id.show = show_test_oa_id; } diff --git a/drivers/gpu/drm/i915/oa/i915_oa_sklgt2.c b/drivers/gpu/drm/i915/oa/i915_oa_sklgt2.c index 309bb49d234c..1bc359ed34e8 100644 --- a/drivers/gpu/drm/i915/oa/i915_oa_sklgt2.c +++ b/drivers/gpu/drm/i915/oa/i915_oa_sklgt2.c @@ -63,26 +63,26 @@ show_test_oa_id(struct device *kdev, struct device_attribute *attr, char *buf) void i915_perf_load_test_config_sklgt2(struct drm_i915_private *dev_priv) { - strlcpy(dev_priv->perf.oa.test_config.uuid, + strlcpy(dev_priv->perf.test_config.uuid, "1651949f-0ac0-4cb1-a06f-dafd74a407d1", - sizeof(dev_priv->perf.oa.test_config.uuid)); - dev_priv->perf.oa.test_config.id = 1; + sizeof(dev_priv->perf.test_config.uuid)); + dev_priv->perf.test_config.id = 1; - dev_priv->perf.oa.test_config.mux_regs = mux_config_test_oa; - dev_priv->perf.oa.test_config.mux_regs_len = ARRAY_SIZE(mux_config_test_oa); + dev_priv->perf.test_config.mux_regs = mux_config_test_oa; + dev_priv->perf.test_config.mux_regs_len = ARRAY_SIZE(mux_config_test_oa); - dev_priv->perf.oa.test_config.b_counter_regs = b_counter_config_test_oa; - dev_priv->perf.oa.test_config.b_counter_regs_len = ARRAY_SIZE(b_counter_config_test_oa); + dev_priv->perf.test_config.b_counter_regs = b_counter_config_test_oa; + dev_priv->perf.test_config.b_counter_regs_len = ARRAY_SIZE(b_counter_config_test_oa); - dev_priv->perf.oa.test_config.flex_regs = flex_eu_config_test_oa; - dev_priv->perf.oa.test_config.flex_regs_len = ARRAY_SIZE(flex_eu_config_test_oa); + dev_priv->perf.test_config.flex_regs = flex_eu_config_test_oa; + dev_priv->perf.test_config.flex_regs_len = ARRAY_SIZE(flex_eu_config_test_oa); - dev_priv->perf.oa.test_config.sysfs_metric.name = "1651949f-0ac0-4cb1-a06f-dafd74a407d1"; - dev_priv->perf.oa.test_config.sysfs_metric.attrs = dev_priv->perf.oa.test_config.attrs; + dev_priv->perf.test_config.sysfs_metric.name = "1651949f-0ac0-4cb1-a06f-dafd74a407d1"; + dev_priv->perf.test_config.sysfs_metric.attrs = dev_priv->perf.test_config.attrs; - dev_priv->perf.oa.test_config.attrs[0] = &dev_priv->perf.oa.test_config.sysfs_metric_id.attr; + dev_priv->perf.test_config.attrs[0] = &dev_priv->perf.test_config.sysfs_metric_id.attr; - dev_priv->perf.oa.test_config.sysfs_metric_id.attr.name = "id"; - dev_priv->perf.oa.test_config.sysfs_metric_id.attr.mode = 0444; - dev_priv->perf.oa.test_config.sysfs_metric_id.show = show_test_oa_id; + dev_priv->perf.test_config.sysfs_metric_id.attr.name = "id"; + dev_priv->perf.test_config.sysfs_metric_id.attr.mode = 0444; + dev_priv->perf.test_config.sysfs_metric_id.show = show_test_oa_id; } diff --git a/drivers/gpu/drm/i915/oa/i915_oa_sklgt3.c b/drivers/gpu/drm/i915/oa/i915_oa_sklgt3.c index 4096e3ba5bb5..6e352f881310 100644 --- a/drivers/gpu/drm/i915/oa/i915_oa_sklgt3.c +++ b/drivers/gpu/drm/i915/oa/i915_oa_sklgt3.c @@ -64,26 +64,26 @@ show_test_oa_id(struct device *kdev, struct device_attribute *attr, char *buf) void i915_perf_load_test_config_sklgt3(struct drm_i915_private *dev_priv) { - strlcpy(dev_priv->perf.oa.test_config.uuid, + strlcpy(dev_priv->perf.test_config.uuid, "2b985803-d3c9-4629-8a4f-634bfecba0e8", - sizeof(dev_priv->perf.oa.test_config.uuid)); - dev_priv->perf.oa.test_config.id = 1; + sizeof(dev_priv->perf.test_config.uuid)); + dev_priv->perf.test_config.id = 1; - dev_priv->perf.oa.test_config.mux_regs = mux_config_test_oa; - dev_priv->perf.oa.test_config.mux_regs_len = ARRAY_SIZE(mux_config_test_oa); + dev_priv->perf.test_config.mux_regs = mux_config_test_oa; + dev_priv->perf.test_config.mux_regs_len = ARRAY_SIZE(mux_config_test_oa); - dev_priv->perf.oa.test_config.b_counter_regs = b_counter_config_test_oa; - dev_priv->perf.oa.test_config.b_counter_regs_len = ARRAY_SIZE(b_counter_config_test_oa); + dev_priv->perf.test_config.b_counter_regs = b_counter_config_test_oa; + dev_priv->perf.test_config.b_counter_regs_len = ARRAY_SIZE(b_counter_config_test_oa); - dev_priv->perf.oa.test_config.flex_regs = flex_eu_config_test_oa; - dev_priv->perf.oa.test_config.flex_regs_len = ARRAY_SIZE(flex_eu_config_test_oa); + dev_priv->perf.test_config.flex_regs = flex_eu_config_test_oa; + dev_priv->perf.test_config.flex_regs_len = ARRAY_SIZE(flex_eu_config_test_oa); - dev_priv->perf.oa.test_config.sysfs_metric.name = "2b985803-d3c9-4629-8a4f-634bfecba0e8"; - dev_priv->perf.oa.test_config.sysfs_metric.attrs = dev_priv->perf.oa.test_config.attrs; + dev_priv->perf.test_config.sysfs_metric.name = "2b985803-d3c9-4629-8a4f-634bfecba0e8"; + dev_priv->perf.test_config.sysfs_metric.attrs = dev_priv->perf.test_config.attrs; - dev_priv->perf.oa.test_config.attrs[0] = &dev_priv->perf.oa.test_config.sysfs_metric_id.attr; + dev_priv->perf.test_config.attrs[0] = &dev_priv->perf.test_config.sysfs_metric_id.attr; - dev_priv->perf.oa.test_config.sysfs_metric_id.attr.name = "id"; - dev_priv->perf.oa.test_config.sysfs_metric_id.attr.mode = 0444; - dev_priv->perf.oa.test_config.sysfs_metric_id.show = show_test_oa_id; + dev_priv->perf.test_config.sysfs_metric_id.attr.name = "id"; + dev_priv->perf.test_config.sysfs_metric_id.attr.mode = 0444; + dev_priv->perf.test_config.sysfs_metric_id.show = show_test_oa_id; } diff --git a/drivers/gpu/drm/i915/oa/i915_oa_sklgt4.c b/drivers/gpu/drm/i915/oa/i915_oa_sklgt4.c index b9575b04f752..8f345115a306 100644 --- a/drivers/gpu/drm/i915/oa/i915_oa_sklgt4.c +++ b/drivers/gpu/drm/i915/oa/i915_oa_sklgt4.c @@ -64,26 +64,26 @@ show_test_oa_id(struct device *kdev, struct device_attribute *attr, char *buf) void i915_perf_load_test_config_sklgt4(struct drm_i915_private *dev_priv) { - strlcpy(dev_priv->perf.oa.test_config.uuid, + strlcpy(dev_priv->perf.test_config.uuid, "882fa433-1f4a-4a67-a962-c741888fe5f5", - sizeof(dev_priv->perf.oa.test_config.uuid)); - dev_priv->perf.oa.test_config.id = 1; + sizeof(dev_priv->perf.test_config.uuid)); + dev_priv->perf.test_config.id = 1; - dev_priv->perf.oa.test_config.mux_regs = mux_config_test_oa; - dev_priv->perf.oa.test_config.mux_regs_len = ARRAY_SIZE(mux_config_test_oa); + dev_priv->perf.test_config.mux_regs = mux_config_test_oa; + dev_priv->perf.test_config.mux_regs_len = ARRAY_SIZE(mux_config_test_oa); - dev_priv->perf.oa.test_config.b_counter_regs = b_counter_config_test_oa; - dev_priv->perf.oa.test_config.b_counter_regs_len = ARRAY_SIZE(b_counter_config_test_oa); + dev_priv->perf.test_config.b_counter_regs = b_counter_config_test_oa; + dev_priv->perf.test_config.b_counter_regs_len = ARRAY_SIZE(b_counter_config_test_oa); - dev_priv->perf.oa.test_config.flex_regs = flex_eu_config_test_oa; - dev_priv->perf.oa.test_config.flex_regs_len = ARRAY_SIZE(flex_eu_config_test_oa); + dev_priv->perf.test_config.flex_regs = flex_eu_config_test_oa; + dev_priv->perf.test_config.flex_regs_len = ARRAY_SIZE(flex_eu_config_test_oa); - dev_priv->perf.oa.test_config.sysfs_metric.name = "882fa433-1f4a-4a67-a962-c741888fe5f5"; - dev_priv->perf.oa.test_config.sysfs_metric.attrs = dev_priv->perf.oa.test_config.attrs; + dev_priv->perf.test_config.sysfs_metric.name = "882fa433-1f4a-4a67-a962-c741888fe5f5"; + dev_priv->perf.test_config.sysfs_metric.attrs = dev_priv->perf.test_config.attrs; - dev_priv->perf.oa.test_config.attrs[0] = &dev_priv->perf.oa.test_config.sysfs_metric_id.attr; + dev_priv->perf.test_config.attrs[0] = &dev_priv->perf.test_config.sysfs_metric_id.attr; - dev_priv->perf.oa.test_config.sysfs_metric_id.attr.name = "id"; - dev_priv->perf.oa.test_config.sysfs_metric_id.attr.mode = 0444; - dev_priv->perf.oa.test_config.sysfs_metric_id.show = show_test_oa_id; + dev_priv->perf.test_config.sysfs_metric_id.attr.name = "id"; + dev_priv->perf.test_config.sysfs_metric_id.attr.mode = 0444; + dev_priv->perf.test_config.sysfs_metric_id.show = show_test_oa_id; } -- cgit v1.2.3 From 387758298bfdec9fbb74406207eb648bb6391670 Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Thu, 8 Aug 2019 12:06:11 +0100 Subject: drm/i915: Allocate kernel_contexts directly Ignore the central i915->kernel_context for allocating an engine, as that GEM context is being phased out. For internal clients, we just need the per-engine logical state, so allocate it at the point of use. Signed-off-by: Chris Wilson Reviewed-by: Mika Kuoppala Link: https://patchwork.freedesktop.org/patch/msgid/20190808110612.23539-1-chris@chris-wilson.co.uk --- drivers/gpu/drm/i915/Makefile | 2 - drivers/gpu/drm/i915/gt/intel_engine_cs.c | 58 +++++++++++---------- drivers/gpu/drm/i915/gt/mock_engine.c | 15 ++---- drivers/gpu/drm/i915/gvt/scheduler.c | 3 +- drivers/gpu/drm/i915/i915_gem.c | 87 ++++++++++++++++++------------- 5 files changed, 90 insertions(+), 75 deletions(-) (limited to 'drivers/gpu/drm/i915/gvt') diff --git a/drivers/gpu/drm/i915/Makefile b/drivers/gpu/drm/i915/Makefile index 3403d79d595e..033121791f90 100644 --- a/drivers/gpu/drm/i915/Makefile +++ b/drivers/gpu/drm/i915/Makefile @@ -95,8 +95,6 @@ gt-y += \ gt/gen7_renderstate.o \ gt/gen8_renderstate.o \ gt/gen9_renderstate.o -gt-$(CONFIG_DRM_I915_SELFTEST) += \ - gt/mock_engine.o i915-y += $(gt-y) # GEM (Graphics Execution Management) code diff --git a/drivers/gpu/drm/i915/gt/intel_engine_cs.c b/drivers/gpu/drm/i915/gt/intel_engine_cs.c index 86247eeb6f2b..8d44d0d8a758 100644 --- a/drivers/gpu/drm/i915/gt/intel_engine_cs.c +++ b/drivers/gpu/drm/i915/gt/intel_engine_cs.c @@ -707,26 +707,6 @@ out_frame: return dw; } -static int pin_context(struct i915_gem_context *ctx, - struct intel_engine_cs *engine, - struct intel_context **out) -{ - struct intel_context *ce; - int err; - - ce = i915_gem_context_get_engine(ctx, engine->id); - if (IS_ERR(ce)) - return PTR_ERR(ce); - - err = intel_context_pin(ce); - intel_context_put(ce); - if (err) - return err; - - *out = ce; - return 0; -} - void intel_engine_init_active(struct intel_engine_cs *engine, unsigned int subclass) { @@ -748,6 +728,25 @@ intel_engine_init_active(struct intel_engine_cs *engine, unsigned int subclass) #endif } +static struct intel_context * +create_kernel_context(struct intel_engine_cs *engine) +{ + struct intel_context *ce; + int err; + + ce = intel_context_create(engine->i915->kernel_context, engine); + if (IS_ERR(ce)) + return ce; + + err = intel_context_pin(ce); + if (err) { + intel_context_put(ce); + return ERR_PTR(err); + } + + return ce; +} + /** * intel_engines_init_common - initialize cengine state which might require hw access * @engine: Engine to initialize. @@ -761,22 +760,24 @@ intel_engine_init_active(struct intel_engine_cs *engine, unsigned int subclass) */ int intel_engine_init_common(struct intel_engine_cs *engine) { - struct drm_i915_private *i915 = engine->i915; + struct intel_context *ce; int ret; engine->set_default_submission(engine); - /* We may need to do things with the shrinker which + /* + * We may need to do things with the shrinker which * require us to immediately switch back to the default * context. This can cause a problem as pinning the * default context also requires GTT space which may not * be available. To avoid this we always pin the default * context. */ - ret = pin_context(i915->kernel_context, engine, - &engine->kernel_context); - if (ret) - return ret; + ce = create_kernel_context(engine); + if (IS_ERR(ce)) + return PTR_ERR(ce); + + engine->kernel_context = ce; ret = measure_breadcrumb_dw(engine); if (ret < 0) @@ -787,7 +788,8 @@ int intel_engine_init_common(struct intel_engine_cs *engine) return 0; err_unpin: - intel_context_unpin(engine->kernel_context); + intel_context_unpin(ce); + intel_context_put(ce); return ret; } @@ -812,6 +814,7 @@ void intel_engine_cleanup_common(struct intel_engine_cs *engine) i915_gem_object_put(engine->default_state); intel_context_unpin(engine->kernel_context); + intel_context_put(engine->kernel_context); GEM_BUG_ON(!llist_empty(&engine->barrier_tasks)); intel_wa_list_free(&engine->ctx_wa_list); @@ -1573,5 +1576,6 @@ intel_engine_find_active_request(struct intel_engine_cs *engine) } #if IS_ENABLED(CONFIG_DRM_I915_SELFTEST) +#include "mock_engine.c" #include "selftest_engine_cs.c" #endif diff --git a/drivers/gpu/drm/i915/gt/mock_engine.c b/drivers/gpu/drm/i915/gt/mock_engine.c index 8a5f07935b84..c64790864795 100644 --- a/drivers/gpu/drm/i915/gt/mock_engine.c +++ b/drivers/gpu/drm/i915/gt/mock_engine.c @@ -286,8 +286,7 @@ struct intel_engine_cs *mock_engine(struct drm_i915_private *i915, int mock_engine_init(struct intel_engine_cs *engine) { - struct drm_i915_private *i915 = engine->i915; - int err; + struct intel_context *ce; intel_engine_init_active(engine, ENGINE_MOCK); intel_engine_init_breadcrumbs(engine); @@ -295,16 +294,11 @@ int mock_engine_init(struct intel_engine_cs *engine) intel_engine_init__pm(engine); intel_engine_pool_init(&engine->pool); - engine->kernel_context = - i915_gem_context_get_engine(i915->kernel_context, engine->id); - if (IS_ERR(engine->kernel_context)) - goto err_breadcrumbs; - - err = intel_context_pin(engine->kernel_context); - intel_context_put(engine->kernel_context); - if (err) + ce = create_kernel_context(engine); + if (IS_ERR(ce)) goto err_breadcrumbs; + engine->kernel_context = ce; return 0; err_breadcrumbs: @@ -338,6 +332,7 @@ void mock_engine_free(struct intel_engine_cs *engine) GEM_BUG_ON(timer_pending(&mock->hw_delay)); intel_context_unpin(engine->kernel_context); + intel_context_put(engine->kernel_context); intel_engine_fini_breadcrumbs(engine); diff --git a/drivers/gpu/drm/i915/gvt/scheduler.c b/drivers/gpu/drm/i915/gvt/scheduler.c index 431fd427c49c..0a0a17bea540 100644 --- a/drivers/gpu/drm/i915/gvt/scheduler.c +++ b/drivers/gpu/drm/i915/gvt/scheduler.c @@ -1229,7 +1229,7 @@ int intel_vgpu_setup_submission(struct intel_vgpu *vgpu) INIT_LIST_HEAD(&s->workload_q_head[i]); s->shadow[i] = ERR_PTR(-EINVAL); - ce = i915_gem_context_get_engine(ctx, i); + ce = intel_context_create(ctx, engine); if (IS_ERR(ce)) { ret = PTR_ERR(ce); goto out_shadow_ctx; @@ -1270,6 +1270,7 @@ out_shadow_ctx: break; intel_context_unpin(s->shadow[i]); + intel_context_put(s->shadow[i]); } i915_gem_context_put(ctx); return ret; diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c index 3f888d6d6a77..6bae073fb570 100644 --- a/drivers/gpu/drm/i915/i915_gem.c +++ b/drivers/gpu/drm/i915/i915_gem.c @@ -1256,9 +1256,8 @@ out: static int __intel_engines_record_defaults(struct drm_i915_private *i915) { + struct i915_request *requests[I915_NUM_ENGINES] = {}; struct intel_engine_cs *engine; - struct i915_gem_context *ctx; - struct i915_gem_engines *e; enum intel_engine_id id; int err = 0; @@ -1271,20 +1270,25 @@ static int __intel_engines_record_defaults(struct drm_i915_private *i915) * from the same default HW values. */ - ctx = i915_gem_context_create_kernel(i915, 0); - if (IS_ERR(ctx)) - return PTR_ERR(ctx); - - e = i915_gem_context_lock_engines(ctx); - for_each_engine(engine, i915, id) { - struct intel_context *ce = e->engines[id]; + struct intel_context *ce; struct i915_request *rq; + /* We must be able to switch to something! */ + GEM_BUG_ON(!engine->kernel_context); + engine->serial++; /* force the kernel context switch */ + + ce = intel_context_create(i915->kernel_context, engine); + if (IS_ERR(ce)) { + err = PTR_ERR(ce); + goto out; + } + rq = intel_context_create_request(ce); if (IS_ERR(rq)) { err = PTR_ERR(rq); - goto err_active; + intel_context_put(ce); + goto out; } err = intel_engine_emit_ctx_wa(rq); @@ -1305,26 +1309,33 @@ static int __intel_engines_record_defaults(struct drm_i915_private *i915) goto err_rq; err_rq: + requests[id] = i915_request_get(rq); i915_request_add(rq); if (err) - goto err_active; + goto out; } /* Flush the default context image to memory, and enable powersaving. */ if (!i915_gem_load_power_context(i915)) { err = -EIO; - goto err_active; + goto out; } - for_each_engine(engine, i915, id) { - struct intel_context *ce = e->engines[id]; - struct i915_vma *state = ce->state; + for (id = 0; id < ARRAY_SIZE(requests); id++) { + struct i915_request *rq; + struct i915_vma *state; void *vaddr; - if (!state) + rq = requests[id]; + if (!rq) continue; - GEM_BUG_ON(intel_context_is_pinned(ce)); + /* We want to be able to unbind the state from the GGTT */ + GEM_BUG_ON(intel_context_is_pinned(rq->hw_context)); + + state = rq->hw_context->state; + if (!state) + continue; /* * As we will hold a reference to the logical state, it will @@ -1336,43 +1347,49 @@ err_rq: */ err = i915_vma_unbind(state); if (err) - goto err_active; + goto out; i915_gem_object_lock(state->obj); err = i915_gem_object_set_to_cpu_domain(state->obj, false); i915_gem_object_unlock(state->obj); if (err) - goto err_active; + goto out; - engine->default_state = i915_gem_object_get(state->obj); - i915_gem_object_set_cache_coherency(engine->default_state, - I915_CACHE_LLC); + i915_gem_object_set_cache_coherency(state->obj, I915_CACHE_LLC); /* Check we can acquire the image of the context state */ - vaddr = i915_gem_object_pin_map(engine->default_state, - I915_MAP_FORCE_WB); + vaddr = i915_gem_object_pin_map(state->obj, I915_MAP_FORCE_WB); if (IS_ERR(vaddr)) { err = PTR_ERR(vaddr); - goto err_active; + goto out; } - i915_gem_object_unpin_map(engine->default_state); + rq->engine->default_state = i915_gem_object_get(state->obj); + i915_gem_object_unpin_map(state->obj); } -out_ctx: - i915_gem_context_unlock_engines(ctx); - i915_gem_context_set_closed(ctx); - i915_gem_context_put(ctx); - return err; - -err_active: +out: /* * If we have to abandon now, we expect the engines to be idle * and ready to be torn-down. The quickest way we can accomplish * this is by declaring ourselves wedged. */ - intel_gt_set_wedged(&i915->gt); - goto out_ctx; + if (err) + intel_gt_set_wedged(&i915->gt); + + for (id = 0; id < ARRAY_SIZE(requests); id++) { + struct intel_context *ce; + struct i915_request *rq; + + rq = requests[id]; + if (!rq) + continue; + + ce = rq->hw_context; + i915_request_put(rq); + intel_context_put(ce); + } + return err; } static int -- cgit v1.2.3 From cd3e05836fc66539f2b387a0697ce1262ad121ce Mon Sep 17 00:00:00 2001 From: "Gustavo A. R. Silva" Date: Fri, 24 May 2019 12:14:21 -0500 Subject: drm/i915/kvmgt: Use struct_size() helper Make use of the struct_size() helper instead of an open-coded version in order to avoid any potential type mistakes, in particular in the context in which this code is being used. So, replace the following form: sizeof(*sparse) + (nr_areas * sizeof(*sparse->areas) with: struct_size(sparse, areas, sparse->nr_areas) and so on... Also, notice that variable size is unnecessary, hence it is removed. This code was detected with the help of Coccinelle. Signed-off-by: Gustavo A. R. Silva Signed-off-by: Zhenyu Wang --- drivers/gpu/drm/i915/gvt/kvmgt.c | 12 +++++------- 1 file changed, 5 insertions(+), 7 deletions(-) (limited to 'drivers/gpu/drm/i915/gvt') diff --git a/drivers/gpu/drm/i915/gvt/kvmgt.c b/drivers/gpu/drm/i915/gvt/kvmgt.c index 144301b778df..9674738b89df 100644 --- a/drivers/gpu/drm/i915/gvt/kvmgt.c +++ b/drivers/gpu/drm/i915/gvt/kvmgt.c @@ -1306,7 +1306,6 @@ static long intel_vgpu_ioctl(struct mdev_device *mdev, unsigned int cmd, unsigned int i; int ret; struct vfio_region_info_cap_sparse_mmap *sparse = NULL; - size_t size; int nr_areas = 1; int cap_type_id; @@ -1349,9 +1348,8 @@ static long intel_vgpu_ioctl(struct mdev_device *mdev, unsigned int cmd, VFIO_REGION_INFO_FLAG_WRITE; info.size = gvt_aperture_sz(vgpu->gvt); - size = sizeof(*sparse) + - (nr_areas * sizeof(*sparse->areas)); - sparse = kzalloc(size, GFP_KERNEL); + sparse = kzalloc(struct_size(sparse, areas, nr_areas), + GFP_KERNEL); if (!sparse) return -ENOMEM; @@ -1416,9 +1414,9 @@ static long intel_vgpu_ioctl(struct mdev_device *mdev, unsigned int cmd, switch (cap_type_id) { case VFIO_REGION_INFO_CAP_SPARSE_MMAP: ret = vfio_info_add_capability(&caps, - &sparse->header, sizeof(*sparse) + - (sparse->nr_areas * - sizeof(*sparse->areas))); + &sparse->header, + struct_size(sparse, areas, + sparse->nr_areas)); if (ret) { kfree(sparse); return ret; -- cgit v1.2.3 From 72e277759340dacdf3d68b0b2b4eae267e601d55 Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Fri, 9 Aug 2019 19:25:15 +0100 Subject: drm/i915: Remove i915_gem_context_create_gvt() As we are phasing out using the GEM context for internal clients that need to manipulate logical context state directly, remove the constructor for the GVT context. We are not using it for anything other than default setup and allocation of an i915_ppgtt. Signed-off-by: Chris Wilson Reviewed-by: Mika Kuoppala Link: https://patchwork.freedesktop.org/patch/msgid/20190809182518.20486-1-chris@chris-wilson.co.uk --- drivers/gpu/drm/i915/gem/i915_gem_context.c | 47 ----------------------------- drivers/gpu/drm/i915/gem/i915_gem_context.h | 2 -- drivers/gpu/drm/i915/gvt/scheduler.c | 22 +++++++++++--- 3 files changed, 17 insertions(+), 54 deletions(-) (limited to 'drivers/gpu/drm/i915/gvt') diff --git a/drivers/gpu/drm/i915/gem/i915_gem_context.c b/drivers/gpu/drm/i915/gem/i915_gem_context.c index b407baaf0014..afd994391ad7 100644 --- a/drivers/gpu/drm/i915/gem/i915_gem_context.c +++ b/drivers/gpu/drm/i915/gem/i915_gem_context.c @@ -529,53 +529,6 @@ i915_gem_create_context(struct drm_i915_private *dev_priv, unsigned int flags) return ctx; } -/** - * i915_gem_context_create_gvt - create a GVT GEM context - * @dev: drm device * - * - * This function is used to create a GVT specific GEM context. - * - * Returns: - * pointer to i915_gem_context on success, error pointer if failed - * - */ -struct i915_gem_context * -i915_gem_context_create_gvt(struct drm_device *dev) -{ - struct i915_gem_context *ctx; - int ret; - - if (!IS_ENABLED(CONFIG_DRM_I915_GVT)) - return ERR_PTR(-ENODEV); - - ret = i915_mutex_lock_interruptible(dev); - if (ret) - return ERR_PTR(ret); - - ctx = i915_gem_create_context(to_i915(dev), 0); - if (IS_ERR(ctx)) - goto out; - - ret = i915_gem_context_pin_hw_id(ctx); - if (ret) { - context_close(ctx); - ctx = ERR_PTR(ret); - goto out; - } - - ctx->file_priv = ERR_PTR(-EBADF); - i915_gem_context_set_closed(ctx); /* not user accessible */ - i915_gem_context_clear_bannable(ctx); - i915_gem_context_set_force_single_submission(ctx); - if (!USES_GUC_SUBMISSION(to_i915(dev))) - ctx->ring_size = 512 * PAGE_SIZE; /* Max ring buffer size */ - - GEM_BUG_ON(i915_gem_context_is_kernel(ctx)); -out: - mutex_unlock(&dev->struct_mutex); - return ctx; -} - static void destroy_kernel_context(struct i915_gem_context **ctxp) { diff --git a/drivers/gpu/drm/i915/gem/i915_gem_context.h b/drivers/gpu/drm/i915/gem/i915_gem_context.h index 106e2ccf7a4c..176978608b6f 100644 --- a/drivers/gpu/drm/i915/gem/i915_gem_context.h +++ b/drivers/gpu/drm/i915/gem/i915_gem_context.h @@ -141,8 +141,6 @@ int i915_gem_context_open(struct drm_i915_private *i915, void i915_gem_context_close(struct drm_file *file); void i915_gem_context_release(struct kref *ctx_ref); -struct i915_gem_context * -i915_gem_context_create_gvt(struct drm_device *dev); int i915_gem_vm_create_ioctl(struct drm_device *dev, void *data, struct drm_file *file); diff --git a/drivers/gpu/drm/i915/gvt/scheduler.c b/drivers/gpu/drm/i915/gvt/scheduler.c index 0a0a17bea540..4cb3c71327fe 100644 --- a/drivers/gpu/drm/i915/gvt/scheduler.c +++ b/drivers/gpu/drm/i915/gvt/scheduler.c @@ -1211,19 +1211,28 @@ i915_context_ppgtt_root_save(struct intel_vgpu_submission *s, */ int intel_vgpu_setup_submission(struct intel_vgpu *vgpu) { + struct drm_i915_private *i915 = vgpu->gvt->dev_priv; struct intel_vgpu_submission *s = &vgpu->submission; struct intel_engine_cs *engine; struct i915_gem_context *ctx; enum intel_engine_id i; int ret; - ctx = i915_gem_context_create_gvt(&vgpu->gvt->dev_priv->drm); - if (IS_ERR(ctx)) - return PTR_ERR(ctx); + mutex_lock(&i915->drm.struct_mutex); + + ctx = i915_gem_context_create_kernel(i915, I915_PRIORITY_MAX); + if (IS_ERR(ctx)) { + ret = PTR_ERR(ctx); + goto out_unlock; + } + + i915_gem_context_set_force_single_submission(ctx); + if (!USES_GUC_SUBMISSION(i915)) + ctx->ring_size = 512 * PAGE_SIZE; /* Max ring buffer size */ i915_context_ppgtt_root_save(s, i915_vm_to_ppgtt(ctx->vm)); - for_each_engine(engine, vgpu->gvt->dev_priv, i) { + for_each_engine(engine, i915, i) { struct intel_context *ce; INIT_LIST_HEAD(&s->workload_q_head[i]); @@ -1261,11 +1270,12 @@ int intel_vgpu_setup_submission(struct intel_vgpu *vgpu) bitmap_zero(s->tlb_handle_pending, I915_NUM_ENGINES); i915_gem_context_put(ctx); + mutex_unlock(&i915->drm.struct_mutex); return 0; out_shadow_ctx: i915_context_ppgtt_root_restore(s, i915_vm_to_ppgtt(ctx->vm)); - for_each_engine(engine, vgpu->gvt->dev_priv, i) { + for_each_engine(engine, i915, i) { if (IS_ERR(s->shadow[i])) break; @@ -1273,6 +1283,8 @@ out_shadow_ctx: intel_context_put(s->shadow[i]); } i915_gem_context_put(ctx); +out_unlock: + mutex_unlock(&i915->drm.struct_mutex); return ret; } -- cgit v1.2.3 From 48ae397b6b935c6733f15476c338df27eac9293c Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Fri, 9 Aug 2019 19:25:17 +0100 Subject: drm/i915: Push the ring creation flags to the backend Push the ring creation flags from the outer GEM context to the inner intel_context to avoid an unsightly back-reference from inside the backend. Signed-off-by: Chris Wilson Reviewed-by: Andi Shyti Link: https://patchwork.freedesktop.org/patch/msgid/20190809182518.20486-3-chris@chris-wilson.co.uk --- drivers/gpu/drm/i915/gem/i915_gem_context.c | 40 +++++++++++++++-------- drivers/gpu/drm/i915/gem/i915_gem_context_types.h | 3 -- drivers/gpu/drm/i915/gt/intel_context.c | 1 + drivers/gpu/drm/i915/gt/intel_context.h | 5 +++ drivers/gpu/drm/i915/gt/intel_engine_cs.c | 2 ++ drivers/gpu/drm/i915/gt/intel_lrc.c | 5 ++- drivers/gpu/drm/i915/gt/intel_ringbuffer.c | 2 +- drivers/gpu/drm/i915/gt/mock_engine.c | 2 +- drivers/gpu/drm/i915/gvt/scheduler.c | 8 +++-- drivers/gpu/drm/i915/i915_debugfs.c | 23 ++++++++----- 10 files changed, 59 insertions(+), 32 deletions(-) (limited to 'drivers/gpu/drm/i915/gvt') diff --git a/drivers/gpu/drm/i915/gem/i915_gem_context.c b/drivers/gpu/drm/i915/gem/i915_gem_context.c index afd994391ad7..d1952637cd91 100644 --- a/drivers/gpu/drm/i915/gem/i915_gem_context.c +++ b/drivers/gpu/drm/i915/gem/i915_gem_context.c @@ -436,8 +436,6 @@ __create_context(struct drm_i915_private *i915) i915_gem_context_set_bannable(ctx); i915_gem_context_set_recoverable(ctx); - ctx->ring_size = 4 * PAGE_SIZE; - for (i = 0; i < ARRAY_SIZE(ctx->hang_timestamp); i++) ctx->hang_timestamp[i] = jiffies - CONTEXT_FAST_HANG_JIFFIES; @@ -448,22 +446,34 @@ err_free: return ERR_PTR(err); } +static void +context_apply_all(struct i915_gem_context *ctx, + void (*fn)(struct intel_context *ce, void *data), + void *data) +{ + struct i915_gem_engines_iter it; + struct intel_context *ce; + + for_each_gem_engine(ce, i915_gem_context_lock_engines(ctx), it) + fn(ce, data); + i915_gem_context_unlock_engines(ctx); +} + +static void __apply_ppgtt(struct intel_context *ce, void *vm) +{ + i915_vm_put(ce->vm); + ce->vm = i915_vm_get(vm); +} + static struct i915_address_space * __set_ppgtt(struct i915_gem_context *ctx, struct i915_address_space *vm) { struct i915_address_space *old = ctx->vm; - struct i915_gem_engines_iter it; - struct intel_context *ce; GEM_BUG_ON(old && i915_vm_is_4lvl(vm) != i915_vm_is_4lvl(old)); ctx->vm = i915_vm_get(vm); - - for_each_gem_engine(ce, i915_gem_context_lock_engines(ctx), it) { - i915_vm_put(ce->vm); - ce->vm = i915_vm_get(vm); - } - i915_gem_context_unlock_engines(ctx); + context_apply_all(ctx, __apply_ppgtt, vm); return old; } @@ -560,7 +570,6 @@ i915_gem_context_create_kernel(struct drm_i915_private *i915, int prio) i915_gem_context_clear_bannable(ctx); ctx->sched.priority = I915_USER_PRIORITY(prio); - ctx->ring_size = PAGE_SIZE; GEM_BUG_ON(!i915_gem_context_is_kernel(ctx)); @@ -1544,6 +1553,7 @@ set_engines(struct i915_gem_context *ctx, for (n = 0; n < num_engines; n++) { struct i915_engine_class_instance ci; struct intel_engine_cs *engine; + struct intel_context *ce; if (copy_from_user(&ci, &user->engines[n], sizeof(ci))) { __free_engines(set.engines, n); @@ -1566,11 +1576,13 @@ set_engines(struct i915_gem_context *ctx, return -ENOENT; } - set.engines->engines[n] = intel_context_create(ctx, engine); - if (!set.engines->engines[n]) { + ce = intel_context_create(ctx, engine); + if (IS_ERR(ce)) { __free_engines(set.engines, n); - return -ENOMEM; + return PTR_ERR(ce); } + + set.engines->engines[n] = ce; } set.engines->num_engines = num_engines; diff --git a/drivers/gpu/drm/i915/gem/i915_gem_context_types.h b/drivers/gpu/drm/i915/gem/i915_gem_context_types.h index a02d98494078..260d59cc3de8 100644 --- a/drivers/gpu/drm/i915/gem/i915_gem_context_types.h +++ b/drivers/gpu/drm/i915/gem/i915_gem_context_types.h @@ -169,9 +169,6 @@ struct i915_gem_context { struct i915_sched_attr sched; - /** ring_size: size for allocating the per-engine ring buffer */ - u32 ring_size; - /** guilty_count: How many times this context has caused a GPU hang. */ atomic_t guilty_count; /** diff --git a/drivers/gpu/drm/i915/gt/intel_context.c b/drivers/gpu/drm/i915/gt/intel_context.c index 41d38e661de7..6d1d4e8dbfc9 100644 --- a/drivers/gpu/drm/i915/gt/intel_context.c +++ b/drivers/gpu/drm/i915/gt/intel_context.c @@ -222,6 +222,7 @@ intel_context_init(struct intel_context *ce, ce->engine = engine; ce->ops = engine->cops; ce->sseu = engine->sseu; + ce->ring = __intel_context_ring_size(SZ_16K); INIT_LIST_HEAD(&ce->signal_link); INIT_LIST_HEAD(&ce->signals); diff --git a/drivers/gpu/drm/i915/gt/intel_context.h b/drivers/gpu/drm/i915/gt/intel_context.h index 07f9924de48f..13f28dd316bc 100644 --- a/drivers/gpu/drm/i915/gt/intel_context.h +++ b/drivers/gpu/drm/i915/gt/intel_context.h @@ -136,4 +136,9 @@ int intel_context_prepare_remote_request(struct intel_context *ce, struct i915_request *intel_context_create_request(struct intel_context *ce); +static inline struct intel_ring *__intel_context_ring_size(u64 sz) +{ + return u64_to_ptr(struct intel_ring, sz); +} + #endif /* __INTEL_CONTEXT_H__ */ diff --git a/drivers/gpu/drm/i915/gt/intel_engine_cs.c b/drivers/gpu/drm/i915/gt/intel_engine_cs.c index 634ef45b77da..ece731ab7b3a 100644 --- a/drivers/gpu/drm/i915/gt/intel_engine_cs.c +++ b/drivers/gpu/drm/i915/gt/intel_engine_cs.c @@ -738,6 +738,8 @@ create_kernel_context(struct intel_engine_cs *engine) if (IS_ERR(ce)) return ce; + ce->ring = __intel_context_ring_size(SZ_4K); + err = intel_context_pin(ce); if (err) { intel_context_put(ce); diff --git a/drivers/gpu/drm/i915/gt/intel_lrc.c b/drivers/gpu/drm/i915/gt/intel_lrc.c index a986ea87fbd9..fc5fcc2d03ac 100644 --- a/drivers/gpu/drm/i915/gt/intel_lrc.c +++ b/drivers/gpu/drm/i915/gt/intel_lrc.c @@ -3115,9 +3115,8 @@ static int __execlists_context_alloc(struct intel_context *ce, goto error_deref_obj; } - ring = intel_engine_create_ring(engine, - timeline, - ce->gem_context->ring_size); + ring = intel_engine_create_ring(engine, timeline, + (unsigned long)ce->ring); intel_timeline_put(timeline); if (IS_ERR(ring)) { ret = PTR_ERR(ring); diff --git a/drivers/gpu/drm/i915/gt/intel_ringbuffer.c b/drivers/gpu/drm/i915/gt/intel_ringbuffer.c index 6a2892033457..d5b7e4dde762 100644 --- a/drivers/gpu/drm/i915/gt/intel_ringbuffer.c +++ b/drivers/gpu/drm/i915/gt/intel_ringbuffer.c @@ -2350,7 +2350,7 @@ int intel_ring_submission_init(struct intel_engine_cs *engine) } GEM_BUG_ON(timeline->has_initial_breadcrumb); - ring = intel_engine_create_ring(engine, timeline, 32 * PAGE_SIZE); + ring = intel_engine_create_ring(engine, timeline, SZ_16K); intel_timeline_put(timeline); if (IS_ERR(ring)) { err = PTR_ERR(ring); diff --git a/drivers/gpu/drm/i915/gt/mock_engine.c b/drivers/gpu/drm/i915/gt/mock_engine.c index 848a83a38b08..79baed1aaf38 100644 --- a/drivers/gpu/drm/i915/gt/mock_engine.c +++ b/drivers/gpu/drm/i915/gt/mock_engine.c @@ -140,7 +140,7 @@ static void mock_context_destroy(struct kref *ref) GEM_BUG_ON(intel_context_is_pinned(ce)); - if (ce->ring) + if (test_bit(CONTEXT_ALLOC_BIT, &ce->flags)) mock_ring_free(ce->ring); intel_context_fini(ce); diff --git a/drivers/gpu/drm/i915/gvt/scheduler.c b/drivers/gpu/drm/i915/gvt/scheduler.c index 4cb3c71327fe..dfb64a6ec838 100644 --- a/drivers/gpu/drm/i915/gvt/scheduler.c +++ b/drivers/gpu/drm/i915/gvt/scheduler.c @@ -1227,8 +1227,6 @@ int intel_vgpu_setup_submission(struct intel_vgpu *vgpu) } i915_gem_context_set_force_single_submission(ctx); - if (!USES_GUC_SUBMISSION(i915)) - ctx->ring_size = 512 * PAGE_SIZE; /* Max ring buffer size */ i915_context_ppgtt_root_save(s, i915_vm_to_ppgtt(ctx->vm)); @@ -1244,6 +1242,12 @@ int intel_vgpu_setup_submission(struct intel_vgpu *vgpu) goto out_shadow_ctx; } + if (!USES_GUC_SUBMISSION(i915)) { /* Max ring buffer size */ + const unsigned int ring_size = 512 * SZ_4K; + + ce->ring = __intel_context_ring_size(ring_size); + } + ret = intel_context_pin(ce); intel_context_put(ce); if (ret) diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c index bddbbd959d1b..e4f173f2f3c3 100644 --- a/drivers/gpu/drm/i915/i915_debugfs.c +++ b/drivers/gpu/drm/i915/i915_debugfs.c @@ -324,10 +324,14 @@ static void print_context_stats(struct seq_file *m, for_each_gem_engine(ce, i915_gem_context_lock_engines(ctx), it) { - if (ce->state) - per_file_stats(0, ce->state->obj, &kstats); - if (ce->ring) + intel_context_lock_pinned(ce); + if (intel_context_is_pinned(ce)) { + if (ce->state) + per_file_stats(0, + ce->state->obj, &kstats); per_file_stats(0, ce->ring->vma->obj, &kstats); + } + intel_context_unlock_pinned(ce); } i915_gem_context_unlock_engines(ctx); @@ -1627,12 +1631,15 @@ static int i915_context_status(struct seq_file *m, void *unused) for_each_gem_engine(ce, i915_gem_context_lock_engines(ctx), it) { - seq_printf(m, "%s: ", ce->engine->name); - if (ce->state) - describe_obj(m, ce->state->obj); - if (ce->ring) + intel_context_lock_pinned(ce); + if (intel_context_is_pinned(ce)) { + seq_printf(m, "%s: ", ce->engine->name); + if (ce->state) + describe_obj(m, ce->state->obj); describe_ctx_ring(m, ce->ring); - seq_putc(m, '\n'); + seq_putc(m, '\n'); + } + intel_context_unlock_pinned(ce); } i915_gem_context_unlock_engines(ctx); -- cgit v1.2.3 From e14006d939fea6d89ddcbb40a72f5c0089a28c85 Mon Sep 17 00:00:00 2001 From: Zhenyu Wang Date: Mon, 10 Jun 2019 17:27:50 +0800 Subject: drm/i915/gvt: Fix typo of VBLANK_TIMER_PERIOD This fixes typo for VBLANK_TIMER_PERIOD. Reviewed-by: Xiong Zhang Signed-off-by: Zhenyu Wang --- drivers/gpu/drm/i915/gvt/interrupt.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'drivers/gpu/drm/i915/gvt') diff --git a/drivers/gpu/drm/i915/gvt/interrupt.c b/drivers/gpu/drm/i915/gvt/interrupt.c index 951681813230..11accd3e1023 100644 --- a/drivers/gpu/drm/i915/gvt/interrupt.c +++ b/drivers/gpu/drm/i915/gvt/interrupt.c @@ -672,7 +672,7 @@ void intel_gvt_clean_irq(struct intel_gvt *gvt) hrtimer_cancel(&irq->vblank_timer.timer); } -#define VBLNAK_TIMER_PERIOD 16000000 +#define VBLANK_TIMER_PERIOD 16000000 /** * intel_gvt_init_irq - initialize GVT-g IRQ emulation subsystem @@ -704,7 +704,7 @@ int intel_gvt_init_irq(struct intel_gvt *gvt) hrtimer_init(&vblank_timer->timer, CLOCK_MONOTONIC, HRTIMER_MODE_ABS); vblank_timer->timer.function = vblank_timer_fn; - vblank_timer->period = VBLNAK_TIMER_PERIOD; + vblank_timer->period = VBLANK_TIMER_PERIOD; return 0; } -- cgit v1.2.3 From f8871ec8fc73f57295703a8d61c8c33d7ab4805b Mon Sep 17 00:00:00 2001 From: Greg Kroah-Hartman Date: Thu, 13 Jun 2019 15:34:19 +0200 Subject: drm/i915/gvt: no need to check return value of debugfs_create functions When calling debugfs functions, there is no need to ever check the return value. The function can work or not, but the code logic should never do something different based on this. Because there is no need to check these functions, a number of local functions can be made to return void to simplify things as nothing can fail. Cc: Zhenyu Wang Cc: Zhi Wang Cc: Jani Nikula Cc: Joonas Lahtinen Cc: Rodrigo Vivi Cc: David Airlie Cc: Daniel Vetter Cc: intel-gvt-dev@lists.freedesktop.org Cc: intel-gfx@lists.freedesktop.org Cc: dri-devel@lists.freedesktop.org Reviewed-by: Zhenyu Wang Signed-off-by: Greg Kroah-Hartman Signed-off-by: Zhenyu Wang --- drivers/gpu/drm/i915/gvt/debugfs.c | 47 ++++++++------------------------------ drivers/gpu/drm/i915/gvt/gvt.c | 4 +--- drivers/gpu/drm/i915/gvt/gvt.h | 4 ++-- drivers/gpu/drm/i915/gvt/kvmgt.c | 3 --- drivers/gpu/drm/i915/gvt/vgpu.c | 4 +--- 5 files changed, 13 insertions(+), 49 deletions(-) (limited to 'drivers/gpu/drm/i915/gvt') diff --git a/drivers/gpu/drm/i915/gvt/debugfs.c b/drivers/gpu/drm/i915/gvt/debugfs.c index 2fb7b73b260d..285f6011a537 100644 --- a/drivers/gpu/drm/i915/gvt/debugfs.c +++ b/drivers/gpu/drm/i915/gvt/debugfs.c @@ -189,36 +189,19 @@ DEFINE_SIMPLE_ATTRIBUTE(vgpu_scan_nonprivbb_fops, /** * intel_gvt_debugfs_add_vgpu - register debugfs entries for a vGPU * @vgpu: a vGPU - * - * Returns: - * Zero on success, negative error code if failed. */ -int intel_gvt_debugfs_add_vgpu(struct intel_vgpu *vgpu) +void intel_gvt_debugfs_add_vgpu(struct intel_vgpu *vgpu) { - struct dentry *ent; char name[16] = ""; snprintf(name, 16, "vgpu%d", vgpu->id); vgpu->debugfs = debugfs_create_dir(name, vgpu->gvt->debugfs_root); - if (!vgpu->debugfs) - return -ENOMEM; - - ent = debugfs_create_bool("active", 0444, vgpu->debugfs, - &vgpu->active); - if (!ent) - return -ENOMEM; - - ent = debugfs_create_file("mmio_diff", 0444, vgpu->debugfs, - vgpu, &vgpu_mmio_diff_fops); - if (!ent) - return -ENOMEM; - ent = debugfs_create_file("scan_nonprivbb", 0644, vgpu->debugfs, - vgpu, &vgpu_scan_nonprivbb_fops); - if (!ent) - return -ENOMEM; - - return 0; + debugfs_create_bool("active", 0444, vgpu->debugfs, &vgpu->active); + debugfs_create_file("mmio_diff", 0444, vgpu->debugfs, vgpu, + &vgpu_mmio_diff_fops); + debugfs_create_file("scan_nonprivbb", 0644, vgpu->debugfs, vgpu, + &vgpu_scan_nonprivbb_fops); } /** @@ -234,27 +217,15 @@ void intel_gvt_debugfs_remove_vgpu(struct intel_vgpu *vgpu) /** * intel_gvt_debugfs_init - register gvt debugfs root entry * @gvt: GVT device - * - * Returns: - * zero on success, negative if failed. */ -int intel_gvt_debugfs_init(struct intel_gvt *gvt) +void intel_gvt_debugfs_init(struct intel_gvt *gvt) { struct drm_minor *minor = gvt->dev_priv->drm.primary; - struct dentry *ent; gvt->debugfs_root = debugfs_create_dir("gvt", minor->debugfs_root); - if (!gvt->debugfs_root) { - gvt_err("Cannot create debugfs dir\n"); - return -ENOMEM; - } - ent = debugfs_create_ulong("num_tracked_mmio", 0444, gvt->debugfs_root, - &gvt->mmio.num_tracked_mmio); - if (!ent) - return -ENOMEM; - - return 0; + debugfs_create_ulong("num_tracked_mmio", 0444, gvt->debugfs_root, + &gvt->mmio.num_tracked_mmio); } /** diff --git a/drivers/gpu/drm/i915/gvt/gvt.c b/drivers/gpu/drm/i915/gvt/gvt.c index 43f4242062dd..8f37eefa0a02 100644 --- a/drivers/gpu/drm/i915/gvt/gvt.c +++ b/drivers/gpu/drm/i915/gvt/gvt.c @@ -375,9 +375,7 @@ int intel_gvt_init_device(struct drm_i915_private *dev_priv) } gvt->idle_vgpu = vgpu; - ret = intel_gvt_debugfs_init(gvt); - if (ret) - gvt_err("debugfs registration failed, go on.\n"); + intel_gvt_debugfs_init(gvt); gvt_dbg_core("gvt device initialization is done\n"); dev_priv->gvt = gvt; diff --git a/drivers/gpu/drm/i915/gvt/gvt.h b/drivers/gpu/drm/i915/gvt/gvt.h index 7a1fe44d45af..3ff59dcf6977 100644 --- a/drivers/gpu/drm/i915/gvt/gvt.h +++ b/drivers/gpu/drm/i915/gvt/gvt.h @@ -682,9 +682,9 @@ static inline void intel_gvt_mmio_set_in_ctx( gvt->mmio.mmio_attribute[offset >> 2] |= F_IN_CTX; } -int intel_gvt_debugfs_add_vgpu(struct intel_vgpu *vgpu); +void intel_gvt_debugfs_add_vgpu(struct intel_vgpu *vgpu); void intel_gvt_debugfs_remove_vgpu(struct intel_vgpu *vgpu); -int intel_gvt_debugfs_init(struct intel_gvt *gvt); +void intel_gvt_debugfs_init(struct intel_gvt *gvt); void intel_gvt_debugfs_clean(struct intel_gvt *gvt); diff --git a/drivers/gpu/drm/i915/gvt/kvmgt.c b/drivers/gpu/drm/i915/gvt/kvmgt.c index 9674738b89df..d91136b39dad 100644 --- a/drivers/gpu/drm/i915/gvt/kvmgt.c +++ b/drivers/gpu/drm/i915/gvt/kvmgt.c @@ -1796,9 +1796,6 @@ static int kvmgt_guest_init(struct mdev_device *mdev) "kvmgt_nr_cache_entries", 0444, vgpu->debugfs, &vgpu->vdev.nr_cache_entries); - if (!info->debugfs_cache_entries) - gvt_vgpu_err("Cannot create kvmgt debugfs entry\n"); - return 0; } diff --git a/drivers/gpu/drm/i915/gvt/vgpu.c b/drivers/gpu/drm/i915/gvt/vgpu.c index 44ce3c2b9ac1..d5a6e4e3d0fd 100644 --- a/drivers/gpu/drm/i915/gvt/vgpu.c +++ b/drivers/gpu/drm/i915/gvt/vgpu.c @@ -420,9 +420,7 @@ static struct intel_vgpu *__intel_gvt_create_vgpu(struct intel_gvt *gvt, if (ret) goto out_clean_submission; - ret = intel_gvt_debugfs_add_vgpu(vgpu); - if (ret) - goto out_clean_sched_policy; + intel_gvt_debugfs_add_vgpu(vgpu); ret = intel_gvt_hypervisor_set_opregion(vgpu); if (ret) -- cgit v1.2.3 From 8cfbca7848ffe3f5d49155748814cd68a774e449 Mon Sep 17 00:00:00 2001 From: Zhi Wang Date: Mon, 22 Jul 2019 14:07:07 +0300 Subject: drm/i915/gvt: factor out tlb and mocs register offset table Factor out tlb and mocs register offset table to fix the issues reported by klocwork, #512 and #550. Mostly, the reason why the klocwork reports these problems is because there can be possbilities for platforms, which have more rings than the ring offset table, to take the dirty data from the stack as the register offset. It results to a random HW register offset writting in this scenairo when doing context switch between vGPUs. After the factoring, the ring offset table of TLB and MOCS should be per platform. v2: - Enable TLB register switch for GEN8. (Zhenyu) Reviewed-by: Zhenyu Wang Signed-off-by: Zhi Wang Signed-off-by: Zhenyu Wang --- drivers/gpu/drm/i915/gvt/gvt.h | 4 +++ drivers/gpu/drm/i915/gvt/mmio_context.c | 57 ++++++++++++++++++++++----------- 2 files changed, 43 insertions(+), 18 deletions(-) (limited to 'drivers/gpu/drm/i915/gvt') diff --git a/drivers/gpu/drm/i915/gvt/gvt.h b/drivers/gpu/drm/i915/gvt/gvt.h index 3ff59dcf6977..b47c6acaf9c0 100644 --- a/drivers/gpu/drm/i915/gvt/gvt.h +++ b/drivers/gpu/drm/i915/gvt/gvt.h @@ -334,6 +334,10 @@ struct intel_gvt { struct { struct engine_mmio *mmio; int ctx_mmio_count[I915_NUM_ENGINES]; + u32 *tlb_mmio_offset_list; + u32 tlb_mmio_offset_list_cnt; + u32 *mocs_mmio_offset_list; + u32 mocs_mmio_offset_list_cnt; } engine_mmio_list; struct dentry *debugfs_root; diff --git a/drivers/gpu/drm/i915/gvt/mmio_context.c b/drivers/gpu/drm/i915/gvt/mmio_context.c index 2998999e8568..4208e40445b1 100644 --- a/drivers/gpu/drm/i915/gvt/mmio_context.c +++ b/drivers/gpu/drm/i915/gvt/mmio_context.c @@ -148,19 +148,27 @@ static struct { u32 l3cc_table[GEN9_MOCS_SIZE / 2]; } gen9_render_mocs; +static u32 gen9_mocs_mmio_offset_list[] = { + [RCS0] = 0xc800, + [VCS0] = 0xc900, + [VCS1] = 0xca00, + [BCS0] = 0xcc00, + [VECS0] = 0xcb00, +}; + static void load_render_mocs(struct drm_i915_private *dev_priv) { + struct intel_gvt *gvt = dev_priv->gvt; i915_reg_t offset; - u32 regs[] = { - [RCS0] = 0xc800, - [VCS0] = 0xc900, - [VCS1] = 0xca00, - [BCS0] = 0xcc00, - [VECS0] = 0xcb00, - }; + u32 cnt = gvt->engine_mmio_list.mocs_mmio_offset_list_cnt; + u32 *regs = gvt->engine_mmio_list.mocs_mmio_offset_list; int ring_id, i; - for (ring_id = 0; ring_id < ARRAY_SIZE(regs); ring_id++) { + /* Platform doesn't have mocs mmios. */ + if (!regs) + return; + + for (ring_id = 0; ring_id < cnt; ring_id++) { if (!HAS_ENGINE(dev_priv, ring_id)) continue; offset.reg = regs[ring_id]; @@ -327,22 +335,28 @@ out: return ret; } +static u32 gen8_tlb_mmio_offset_list[] = { + [RCS0] = 0x4260, + [VCS0] = 0x4264, + [VCS1] = 0x4268, + [BCS0] = 0x426c, + [VECS0] = 0x4270, +}; + static void handle_tlb_pending_event(struct intel_vgpu *vgpu, int ring_id) { struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv; struct intel_uncore *uncore = &dev_priv->uncore; struct intel_vgpu_submission *s = &vgpu->submission; + u32 *regs = vgpu->gvt->engine_mmio_list.tlb_mmio_offset_list; + u32 cnt = vgpu->gvt->engine_mmio_list.tlb_mmio_offset_list_cnt; enum forcewake_domains fw; i915_reg_t reg; - u32 regs[] = { - [RCS0] = 0x4260, - [VCS0] = 0x4264, - [VCS1] = 0x4268, - [BCS0] = 0x426c, - [VECS0] = 0x4270, - }; - if (WARN_ON(ring_id >= ARRAY_SIZE(regs))) + if (!regs) + return; + + if (WARN_ON(ring_id >= cnt)) return; if (!test_and_clear_bit(ring_id, (void *)s->tlb_handle_pending)) @@ -565,10 +579,17 @@ void intel_gvt_init_engine_mmio_context(struct intel_gvt *gvt) { struct engine_mmio *mmio; - if (INTEL_GEN(gvt->dev_priv) >= 9) + if (INTEL_GEN(gvt->dev_priv) >= 9) { gvt->engine_mmio_list.mmio = gen9_engine_mmio_list; - else + gvt->engine_mmio_list.tlb_mmio_offset_list = gen8_tlb_mmio_offset_list; + gvt->engine_mmio_list.tlb_mmio_offset_list_cnt = ARRAY_SIZE(gen8_tlb_mmio_offset_list); + gvt->engine_mmio_list.mocs_mmio_offset_list = gen9_mocs_mmio_offset_list; + gvt->engine_mmio_list.mocs_mmio_offset_list_cnt = ARRAY_SIZE(gen9_mocs_mmio_offset_list); + } else { gvt->engine_mmio_list.mmio = gen8_engine_mmio_list; + gvt->engine_mmio_list.tlb_mmio_offset_list = gen8_tlb_mmio_offset_list; + gvt->engine_mmio_list.tlb_mmio_offset_list_cnt = ARRAY_SIZE(gen8_tlb_mmio_offset_list); + } for (mmio = gvt->engine_mmio_list.mmio; i915_mmio_reg_valid(mmio->reg); mmio++) { -- cgit v1.2.3 From 4f870f1fba3443705389d9a2405c10c1cf73f2e0 Mon Sep 17 00:00:00 2001 From: "Gao, Fred" Date: Thu, 18 Jul 2019 09:39:01 +0800 Subject: drm/i915/gvt: Utility for valid command length check Add utility for valid command length check. v2: Add F_VAL_CONST flag to identify the value is const although LEN maybe variable. (Zhenyu) v3: unused code removal, flag rename/conflict. (Zhenyu) v4: redefine F_IP_ADVANCE_CUSTOM and move the check function to next patch. (Zhenyu) Reviewed-by: Zhenyu Wang Signed-off-by: Gao, Fred Signed-off-by: Zhenyu Wang --- drivers/gpu/drm/i915/gvt/cmd_parser.c | 15 +++++++++------ 1 file changed, 9 insertions(+), 6 deletions(-) (limited to 'drivers/gpu/drm/i915/gvt') diff --git a/drivers/gpu/drm/i915/gvt/cmd_parser.c b/drivers/gpu/drm/i915/gvt/cmd_parser.c index 6ea88270c818..5f00e5d04929 100644 --- a/drivers/gpu/drm/i915/gvt/cmd_parser.c +++ b/drivers/gpu/drm/i915/gvt/cmd_parser.c @@ -378,17 +378,17 @@ struct cmd_info { const char *name; u32 opcode; -#define F_LEN_MASK (1U<<0) +#define F_LEN_MASK 3U #define F_LEN_CONST 1U #define F_LEN_VAR 0U +/* value is const although LEN maybe variable */ +#define F_LEN_VAR_FIXED (1<<1) /* * command has its own ip advance logic * e.g. MI_BATCH_START, MI_BATCH_END */ -#define F_IP_ADVANCE_CUSTOM (1<<1) - -#define F_POST_HANDLE (1<<2) +#define F_IP_ADVANCE_CUSTOM (1<<2) u32 flag; #define R_RCS BIT(RCS0) @@ -418,9 +418,12 @@ struct cmd_info { * flag == F_LEN_VAR : length bias bits * Note: length is in DWord */ - u8 len; + u32 len; parser_cmd_handler handler; + + /* valid length in DWord */ + u32 valid_len; }; struct cmd_entry { @@ -1912,7 +1915,7 @@ static const struct cmd_info cmd_info[] = { {"MI_RS_CONTEXT", OP_MI_RS_CONTEXT, F_LEN_CONST, R_RCS, D_ALL, 0, 1, NULL}, - {"MI_DISPLAY_FLIP", OP_MI_DISPLAY_FLIP, F_LEN_VAR | F_POST_HANDLE, + {"MI_DISPLAY_FLIP", OP_MI_DISPLAY_FLIP, F_LEN_VAR, R_RCS | R_BCS, D_ALL, 0, 8, cmd_handler_mi_display_flip}, {"MI_SEMAPHORE_MBOX", OP_MI_SEMAPHORE_MBOX, F_LEN_VAR, R_ALL, D_ALL, -- cgit v1.2.3 From 1e2adc0d037076fb1de6492dd152b368adea0d2d Mon Sep 17 00:00:00 2001 From: "Gao, Fred" Date: Thu, 18 Jul 2019 09:39:09 +0800 Subject: drm/i915/gvt: Add MI command valid length check Add the constant valid length of MI command. v2: Add F_VAL_CONST flag. (Zhenyu Wang) Reviewed-by: Zhenyu Wang Signed-off-by: Gao, Fred Signed-off-by: Zhenyu Wang --- drivers/gpu/drm/i915/gvt/cmd_parser.c | 74 ++++++++++++++++++++++++----------- 1 file changed, 51 insertions(+), 23 deletions(-) (limited to 'drivers/gpu/drm/i915/gvt') diff --git a/drivers/gpu/drm/i915/gvt/cmd_parser.c b/drivers/gpu/drm/i915/gvt/cmd_parser.c index 5f00e5d04929..a0b12defea9b 100644 --- a/drivers/gpu/drm/i915/gvt/cmd_parser.c +++ b/drivers/gpu/drm/i915/gvt/cmd_parser.c @@ -374,6 +374,19 @@ typedef int (*parser_cmd_handler)(struct parser_exec_state *s); #define ADDR_FIX_4(x1, x2, x3, x4) (ADDR_FIX_1(x1) | ADDR_FIX_3(x2, x3, x4)) #define ADDR_FIX_5(x1, x2, x3, x4, x5) (ADDR_FIX_1(x1) | ADDR_FIX_4(x2, x3, x4, x5)) +#define OP_LENGTH_BIAS 2 +#define CMD_LEN(value) (value + OP_LENGTH_BIAS) + +static int gvt_check_valid_cmd_length(int len, int valid_len) +{ + if (valid_len != len) { + gvt_err("len is not valid: len=%u valid_len=%u\n", + len, valid_len); + return -EFAULT; + } + return 0; +} + struct cmd_info { const char *name; u32 opcode; @@ -1918,18 +1931,21 @@ static const struct cmd_info cmd_info[] = { {"MI_DISPLAY_FLIP", OP_MI_DISPLAY_FLIP, F_LEN_VAR, R_RCS | R_BCS, D_ALL, 0, 8, cmd_handler_mi_display_flip}, - {"MI_SEMAPHORE_MBOX", OP_MI_SEMAPHORE_MBOX, F_LEN_VAR, R_ALL, D_ALL, - 0, 8, NULL}, + {"MI_SEMAPHORE_MBOX", OP_MI_SEMAPHORE_MBOX, F_LEN_VAR | F_LEN_VAR_FIXED, + R_ALL, D_ALL, 0, 8, NULL, CMD_LEN(1)}, {"MI_MATH", OP_MI_MATH, F_LEN_VAR, R_ALL, D_ALL, 0, 8, NULL}, - {"MI_URB_CLEAR", OP_MI_URB_CLEAR, F_LEN_VAR, R_RCS, D_ALL, 0, 8, NULL}, + {"MI_URB_CLEAR", OP_MI_URB_CLEAR, F_LEN_VAR | F_LEN_VAR_FIXED, R_RCS, + D_ALL, 0, 8, NULL, CMD_LEN(0)}, - {"MI_SEMAPHORE_SIGNAL", OP_MI_SEMAPHORE_SIGNAL, F_LEN_VAR, R_ALL, - D_BDW_PLUS, 0, 8, NULL}, + {"MI_SEMAPHORE_SIGNAL", OP_MI_SEMAPHORE_SIGNAL, + F_LEN_VAR | F_LEN_VAR_FIXED, R_ALL, D_BDW_PLUS, 0, 8, + NULL, CMD_LEN(0)}, - {"MI_SEMAPHORE_WAIT", OP_MI_SEMAPHORE_WAIT, F_LEN_VAR, R_ALL, - D_BDW_PLUS, ADDR_FIX_1(2), 8, cmd_handler_mi_semaphore_wait}, + {"MI_SEMAPHORE_WAIT", OP_MI_SEMAPHORE_WAIT, + F_LEN_VAR | F_LEN_VAR_FIXED, R_ALL, D_BDW_PLUS, ADDR_FIX_1(2), + 8, cmd_handler_mi_semaphore_wait, CMD_LEN(2)}, {"MI_STORE_DATA_IMM", OP_MI_STORE_DATA_IMM, F_LEN_VAR, R_ALL, D_BDW_PLUS, ADDR_FIX_1(1), 10, cmd_handler_mi_store_data_imm}, @@ -1943,8 +1959,9 @@ static const struct cmd_info cmd_info[] = { {"MI_UPDATE_GTT", OP_MI_UPDATE_GTT, F_LEN_VAR, R_ALL, D_BDW_PLUS, 0, 10, cmd_handler_mi_update_gtt}, - {"MI_STORE_REGISTER_MEM", OP_MI_STORE_REGISTER_MEM, F_LEN_VAR, R_ALL, - D_ALL, ADDR_FIX_1(2), 8, cmd_handler_srm}, + {"MI_STORE_REGISTER_MEM", OP_MI_STORE_REGISTER_MEM, + F_LEN_VAR | F_LEN_VAR_FIXED, R_ALL, D_ALL, ADDR_FIX_1(2), 8, + cmd_handler_srm, CMD_LEN(2)}, {"MI_FLUSH_DW", OP_MI_FLUSH_DW, F_LEN_VAR, R_ALL, D_ALL, 0, 6, cmd_handler_mi_flush_dw}, @@ -1952,26 +1969,30 @@ static const struct cmd_info cmd_info[] = { {"MI_CLFLUSH", OP_MI_CLFLUSH, F_LEN_VAR, R_ALL, D_ALL, ADDR_FIX_1(1), 10, cmd_handler_mi_clflush}, - {"MI_REPORT_PERF_COUNT", OP_MI_REPORT_PERF_COUNT, F_LEN_VAR, R_ALL, - D_ALL, ADDR_FIX_1(1), 6, cmd_handler_mi_report_perf_count}, + {"MI_REPORT_PERF_COUNT", OP_MI_REPORT_PERF_COUNT, + F_LEN_VAR | F_LEN_VAR_FIXED, R_ALL, D_ALL, ADDR_FIX_1(1), 6, + cmd_handler_mi_report_perf_count, CMD_LEN(2)}, - {"MI_LOAD_REGISTER_MEM", OP_MI_LOAD_REGISTER_MEM, F_LEN_VAR, R_ALL, - D_ALL, ADDR_FIX_1(2), 8, cmd_handler_lrm}, + {"MI_LOAD_REGISTER_MEM", OP_MI_LOAD_REGISTER_MEM, + F_LEN_VAR | F_LEN_VAR_FIXED, R_ALL, D_ALL, ADDR_FIX_1(2), 8, + cmd_handler_lrm, CMD_LEN(2)}, - {"MI_LOAD_REGISTER_REG", OP_MI_LOAD_REGISTER_REG, F_LEN_VAR, R_ALL, - D_ALL, 0, 8, cmd_handler_lrr}, + {"MI_LOAD_REGISTER_REG", OP_MI_LOAD_REGISTER_REG, + F_LEN_VAR | F_LEN_VAR_FIXED, R_ALL, D_ALL, 0, 8, + cmd_handler_lrr, CMD_LEN(1)}, - {"MI_RS_STORE_DATA_IMM", OP_MI_RS_STORE_DATA_IMM, F_LEN_VAR, R_RCS, - D_ALL, 0, 8, NULL}, + {"MI_RS_STORE_DATA_IMM", OP_MI_RS_STORE_DATA_IMM, + F_LEN_VAR | F_LEN_VAR_FIXED, R_RCS, D_ALL, 0, + 8, NULL, CMD_LEN(2)}, - {"MI_LOAD_URB_MEM", OP_MI_LOAD_URB_MEM, F_LEN_VAR, R_RCS, D_ALL, - ADDR_FIX_1(2), 8, NULL}, + {"MI_LOAD_URB_MEM", OP_MI_LOAD_URB_MEM, F_LEN_VAR | F_LEN_VAR_FIXED, + R_RCS, D_ALL, ADDR_FIX_1(2), 8, NULL, CMD_LEN(2)}, {"MI_STORE_URM_MEM", OP_MI_STORE_URM_MEM, F_LEN_VAR, R_RCS, D_ALL, ADDR_FIX_1(2), 8, NULL}, - {"MI_OP_2E", OP_MI_2E, F_LEN_VAR, R_ALL, D_BDW_PLUS, ADDR_FIX_2(1, 2), - 8, cmd_handler_mi_op_2e}, + {"MI_OP_2E", OP_MI_2E, F_LEN_VAR | F_LEN_VAR_FIXED, R_ALL, D_BDW_PLUS, + ADDR_FIX_2(1, 2), 8, cmd_handler_mi_op_2e, CMD_LEN(3)}, {"MI_OP_2F", OP_MI_2F, F_LEN_VAR, R_ALL, D_BDW_PLUS, ADDR_FIX_1(1), 8, cmd_handler_mi_op_2f}, @@ -1981,8 +2002,8 @@ static const struct cmd_info cmd_info[] = { cmd_handler_mi_batch_buffer_start}, {"MI_CONDITIONAL_BATCH_BUFFER_END", OP_MI_CONDITIONAL_BATCH_BUFFER_END, - F_LEN_VAR, R_ALL, D_ALL, ADDR_FIX_1(2), 8, - cmd_handler_mi_conditional_batch_buffer_end}, + F_LEN_VAR | F_LEN_VAR_FIXED, R_ALL, D_ALL, ADDR_FIX_1(2), 8, + cmd_handler_mi_conditional_batch_buffer_end, CMD_LEN(2)}, {"MI_LOAD_SCAN_LINES_INCL", OP_MI_LOAD_SCAN_LINES_INCL, F_LEN_CONST, R_RCS | R_BCS, D_ALL, 0, 2, NULL}, @@ -2572,6 +2593,13 @@ static int cmd_parser_exec(struct parser_exec_state *s) cmd_length(s), s->buf_type, s->buf_addr_type, s->workload, info->name); + if ((info->flag & F_LEN_MASK) == F_LEN_VAR_FIXED) { + ret = gvt_check_valid_cmd_length(cmd_length(s), + info->valid_len); + if (ret) + return ret; + } + if (info->handler) { ret = info->handler(s); if (ret < 0) { -- cgit v1.2.3 From 00a33be406349ae04a676e05c535987750f838ca Mon Sep 17 00:00:00 2001 From: "Gao, Fred" Date: Thu, 18 Jul 2019 09:39:18 +0800 Subject: drm/i915/gvt: Add valid length check for MI variable commands Add valid length check for the commands with variable length. v2: remove the macro definition. (Zhenyu) v3: refine the LRI command. (Zhenyu) Reviewed-by: Zhenyu Wang Signed-off-by: Gao, Fred Signed-off-by: Zhenyu Wang --- drivers/gpu/drm/i915/gvt/cmd_parser.c | 52 +++++++++++++++++++++++++++++++++++ 1 file changed, 52 insertions(+) (limited to 'drivers/gpu/drm/i915/gvt') diff --git a/drivers/gpu/drm/i915/gvt/cmd_parser.c b/drivers/gpu/drm/i915/gvt/cmd_parser.c index a0b12defea9b..b5a89631a59e 100644 --- a/drivers/gpu/drm/i915/gvt/cmd_parser.c +++ b/drivers/gpu/drm/i915/gvt/cmd_parser.c @@ -374,6 +374,9 @@ typedef int (*parser_cmd_handler)(struct parser_exec_state *s); #define ADDR_FIX_4(x1, x2, x3, x4) (ADDR_FIX_1(x1) | ADDR_FIX_3(x2, x3, x4)) #define ADDR_FIX_5(x1, x2, x3, x4, x5) (ADDR_FIX_1(x1) | ADDR_FIX_4(x2, x3, x4, x5)) +#define DWORD_FIELD(dword, end, start) \ + FIELD_GET(GENMASK(end, start), cmd_val(s, dword)) + #define OP_LENGTH_BIAS 2 #define CMD_LEN(value) (value + OP_LENGTH_BIAS) @@ -960,6 +963,18 @@ static int cmd_handler_lri(struct parser_exec_state *s) int i, ret = 0; int cmd_len = cmd_length(s); struct intel_gvt *gvt = s->vgpu->gvt; + u32 valid_len = CMD_LEN(1); + + /* + * Official intel docs are somewhat sloppy , check the definition of + * MI_LOAD_REGISTER_IMM. + */ + #define MAX_VALID_LEN 127 + if ((cmd_len < valid_len) || (cmd_len > MAX_VALID_LEN)) { + gvt_err("len is not valid: len=%u valid_len=%u\n", + cmd_len, valid_len); + return -EFAULT; + } for (i = 1; i < cmd_len; i += 2) { if (IS_BROADWELL(gvt->dev_priv) && s->ring_id != RCS0) { @@ -1391,6 +1406,15 @@ static int cmd_handler_mi_display_flip(struct parser_exec_state *s) int ret; int i; int len = cmd_length(s); + u32 valid_len = CMD_LEN(1); + + /* Flip Type == Stereo 3D Flip */ + if (DWORD_FIELD(2, 1, 0) == 2) + valid_len++; + ret = gvt_check_valid_cmd_length(cmd_length(s), + valid_len); + if (ret) + return ret; ret = decode_mi_display_flip(s, &info); if (ret) { @@ -1510,12 +1534,21 @@ static int cmd_handler_mi_store_data_imm(struct parser_exec_state *s) int op_size = (cmd_length(s) - 3) * sizeof(u32); int core_id = (cmd_val(s, 2) & (1 << 0)) ? 1 : 0; unsigned long gma, gma_low, gma_high; + u32 valid_len = CMD_LEN(2); int ret = 0; /* check ppggt */ if (!(cmd_val(s, 0) & (1 << 22))) return 0; + /* check if QWORD */ + if (DWORD_FIELD(0, 21, 21)) + valid_len++; + ret = gvt_check_valid_cmd_length(cmd_length(s), + valid_len); + if (ret) + return ret; + gma = cmd_val(s, 2) & GENMASK(31, 2); if (gmadr_bytes == 8) { @@ -1558,11 +1591,20 @@ static int cmd_handler_mi_op_2f(struct parser_exec_state *s) int op_size = (1 << ((cmd_val(s, 0) & GENMASK(20, 19)) >> 19)) * sizeof(u32); unsigned long gma, gma_high; + u32 valid_len = CMD_LEN(1); int ret = 0; if (!(cmd_val(s, 0) & (1 << 22))) return ret; + /* check if QWORD */ + if (DWORD_FIELD(0, 20, 19) == 1) + valid_len += 8; + ret = gvt_check_valid_cmd_length(cmd_length(s), + valid_len); + if (ret) + return ret; + gma = cmd_val(s, 1) & GENMASK(31, 2); if (gmadr_bytes == 8) { gma_high = cmd_val(s, 2) & GENMASK(15, 0); @@ -1600,6 +1642,16 @@ static int cmd_handler_mi_flush_dw(struct parser_exec_state *s) bool index_mode = false; int ret = 0; u32 hws_pga, val; + u32 valid_len = CMD_LEN(2); + + ret = gvt_check_valid_cmd_length(cmd_length(s), + valid_len); + if (ret) { + /* Check again for Qword */ + ret = gvt_check_valid_cmd_length(cmd_length(s), + ++valid_len); + return ret; + } /* Check post-sync and ppgtt bit */ if (((cmd_val(s, 0) >> 14) & 0x3) && (cmd_val(s, 1) & (1 << 2))) { -- cgit v1.2.3 From 220b65d8009dff397a112ee560c297299ba5e423 Mon Sep 17 00:00:00 2001 From: Tina Zhang Date: Mon, 12 Aug 2019 14:47:51 +0800 Subject: drm/i915/gvt: Double check batch buffer size after copy Double check the end of the privilege buffer to make sure the size of the privilege buffer remains unchanged after copy. v4: - Refine the commit message. (Zhenyu) v3: - To get the right offset of the batch buffer end cmd. (Yan) v2: - Use lightweight way to audit batch buffer end. (Yan) Reviewed-by: Yan Zhao Signed-off-by: Tina Zhang Signed-off-by: Zhenyu Wang --- drivers/gpu/drm/i915/gvt/cmd_parser.c | 39 +++++++++++++++++++++++++++++++++-- 1 file changed, 37 insertions(+), 2 deletions(-) (limited to 'drivers/gpu/drm/i915/gvt') diff --git a/drivers/gpu/drm/i915/gvt/cmd_parser.c b/drivers/gpu/drm/i915/gvt/cmd_parser.c index b5a89631a59e..58afd36617db 100644 --- a/drivers/gpu/drm/i915/gvt/cmd_parser.c +++ b/drivers/gpu/drm/i915/gvt/cmd_parser.c @@ -1729,7 +1729,9 @@ static int batch_buffer_needs_scan(struct parser_exec_state *s) return 1; } -static int find_bb_size(struct parser_exec_state *s, unsigned long *bb_size) +static int find_bb_size(struct parser_exec_state *s, + unsigned long *bb_size, + unsigned long *bb_end_cmd_offset) { unsigned long gma = 0; const struct cmd_info *info; @@ -1741,6 +1743,7 @@ static int find_bb_size(struct parser_exec_state *s, unsigned long *bb_size) s->vgpu->gtt.ggtt_mm : s->workload->shadow_mm; *bb_size = 0; + *bb_end_cmd_offset = 0; /* get the start gm address of the batch buffer */ gma = get_gma_bb_from_cmd(s, 1); @@ -1776,6 +1779,10 @@ static int find_bb_size(struct parser_exec_state *s, unsigned long *bb_size) /* chained batch buffer */ bb_end = true; } + + if (bb_end) + *bb_end_cmd_offset = *bb_size; + cmd_len = get_cmd_length(info, cmd) << 2; *bb_size += cmd_len; gma += cmd_len; @@ -1784,12 +1791,36 @@ static int find_bb_size(struct parser_exec_state *s, unsigned long *bb_size) return 0; } +static int audit_bb_end(struct parser_exec_state *s, void *va) +{ + struct intel_vgpu *vgpu = s->vgpu; + u32 cmd = *(u32 *)va; + const struct cmd_info *info; + + info = get_cmd_info(s->vgpu->gvt, cmd, s->ring_id); + if (info == NULL) { + gvt_vgpu_err("unknown cmd 0x%x, opcode=0x%x, addr_type=%s, ring %d, workload=%p\n", + cmd, get_opcode(cmd, s->ring_id), + (s->buf_addr_type == PPGTT_BUFFER) ? + "ppgtt" : "ggtt", s->ring_id, s->workload); + return -EBADRQC; + } + + if ((info->opcode == OP_MI_BATCH_BUFFER_END) || + ((info->opcode == OP_MI_BATCH_BUFFER_START) && + (BATCH_BUFFER_2ND_LEVEL_BIT(cmd) == 0))) + return 0; + + return -EBADRQC; +} + static int perform_bb_shadow(struct parser_exec_state *s) { struct intel_vgpu *vgpu = s->vgpu; struct intel_vgpu_shadow_bb *bb; unsigned long gma = 0; unsigned long bb_size; + unsigned long bb_end_cmd_offset; int ret = 0; struct intel_vgpu_mm *mm = (s->buf_addr_type == GTT_BUFFER) ? s->vgpu->gtt.ggtt_mm : s->workload->shadow_mm; @@ -1800,7 +1831,7 @@ static int perform_bb_shadow(struct parser_exec_state *s) if (gma == INTEL_GVT_INVALID_ADDR) return -EFAULT; - ret = find_bb_size(s, &bb_size); + ret = find_bb_size(s, &bb_size, &bb_end_cmd_offset); if (ret) return ret; @@ -1856,6 +1887,10 @@ static int perform_bb_shadow(struct parser_exec_state *s) goto err_unmap; } + ret = audit_bb_end(s, bb->va + start_offset + bb_end_cmd_offset); + if (ret) + goto err_unmap; + INIT_LIST_HEAD(&bb->list); list_add(&bb->list, &s->workload->shadow_bb); -- cgit v1.2.3 From e2ccc50a3a6c90cace6a9aef40072e558dd13863 Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Thu, 22 Aug 2019 07:09:12 +0100 Subject: drm/i915: Track ggtt fence reservations under its own mutex We can reduce the locking for fence registers from the dev->struct_mutex to a local mutex. We could introduce a mutex for the sole purpose of tracking the fence acquisition, except there is a little bit of overlap with the fault tracking, so use the i915_ggtt.mutex as it covers both. Signed-off-by: Chris Wilson Cc: Mika Kuoppala Reviewed-by: Matthew Auld Link: https://patchwork.freedesktop.org/patch/msgid/20190822060914.2671-1-chris@chris-wilson.co.uk --- drivers/gpu/drm/i915/gt/selftest_hangcheck.c | 7 ++ drivers/gpu/drm/i915/gvt/aperture_gm.c | 10 +-- drivers/gpu/drm/i915/i915_debugfs.c | 5 +- drivers/gpu/drm/i915/i915_gem_fence_reg.c | 108 +++++++++++++++++---------- drivers/gpu/drm/i915/i915_gem_fence_reg.h | 2 +- drivers/gpu/drm/i915/i915_vma.h | 4 +- 6 files changed, 87 insertions(+), 49 deletions(-) (limited to 'drivers/gpu/drm/i915/gvt') diff --git a/drivers/gpu/drm/i915/gt/selftest_hangcheck.c b/drivers/gpu/drm/i915/gt/selftest_hangcheck.c index 298c4d191439..a0098fc35921 100644 --- a/drivers/gpu/drm/i915/gt/selftest_hangcheck.c +++ b/drivers/gpu/drm/i915/gt/selftest_hangcheck.c @@ -1157,7 +1157,14 @@ static int evict_fence(void *data) goto out_unlock; } + err = i915_vma_pin(arg->vma, 0, 0, PIN_GLOBAL | PIN_MAPPABLE); + if (err) { + pr_err("Unable to pin vma for Y-tiled fence; err:%d\n", err); + goto out_unlock; + } + err = i915_vma_pin_fence(arg->vma); + i915_vma_unpin(arg->vma); if (err) { pr_err("Unable to pin Y-tiled fence; err:%d\n", err); goto out_unlock; diff --git a/drivers/gpu/drm/i915/gvt/aperture_gm.c b/drivers/gpu/drm/i915/gvt/aperture_gm.c index c3d19d88da40..5ff2437b2998 100644 --- a/drivers/gpu/drm/i915/gvt/aperture_gm.c +++ b/drivers/gpu/drm/i915/gvt/aperture_gm.c @@ -172,14 +172,14 @@ static void free_vgpu_fence(struct intel_vgpu *vgpu) intel_runtime_pm_get(&dev_priv->runtime_pm); - mutex_lock(&dev_priv->drm.struct_mutex); + mutex_lock(&dev_priv->ggtt.vm.mutex); _clear_vgpu_fence(vgpu); for (i = 0; i < vgpu_fence_sz(vgpu); i++) { reg = vgpu->fence.regs[i]; i915_unreserve_fence(reg); vgpu->fence.regs[i] = NULL; } - mutex_unlock(&dev_priv->drm.struct_mutex); + mutex_unlock(&dev_priv->ggtt.vm.mutex); intel_runtime_pm_put_unchecked(&dev_priv->runtime_pm); } @@ -195,7 +195,7 @@ static int alloc_vgpu_fence(struct intel_vgpu *vgpu) intel_runtime_pm_get(rpm); /* Request fences from host */ - mutex_lock(&dev_priv->drm.struct_mutex); + mutex_lock(&dev_priv->ggtt.vm.mutex); for (i = 0; i < vgpu_fence_sz(vgpu); i++) { reg = i915_reserve_fence(dev_priv); @@ -207,7 +207,7 @@ static int alloc_vgpu_fence(struct intel_vgpu *vgpu) _clear_vgpu_fence(vgpu); - mutex_unlock(&dev_priv->drm.struct_mutex); + mutex_unlock(&dev_priv->ggtt.vm.mutex); intel_runtime_pm_put_unchecked(rpm); return 0; out_free_fence: @@ -220,7 +220,7 @@ out_free_fence: i915_unreserve_fence(reg); vgpu->fence.regs[i] = NULL; } - mutex_unlock(&dev_priv->drm.struct_mutex); + mutex_unlock(&dev_priv->ggtt.vm.mutex); intel_runtime_pm_put_unchecked(rpm); return -ENOSPC; } diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c index b39226d7f8d2..f5d6702ec7df 100644 --- a/drivers/gpu/drm/i915/i915_debugfs.c +++ b/drivers/gpu/drm/i915/i915_debugfs.c @@ -652,10 +652,11 @@ static int i915_gem_fence_regs_info(struct seq_file *m, void *data) rcu_read_lock(); for (i = 0; i < i915->ggtt.num_fences; i++) { - struct i915_vma *vma = i915->ggtt.fence_regs[i].vma; + struct i915_fence_reg *reg = &i915->ggtt.fence_regs[i]; + struct i915_vma *vma = reg->vma; seq_printf(m, "Fence %d, pin count = %d, object = ", - i, i915->ggtt.fence_regs[i].pin_count); + i, atomic_read(®->pin_count)); if (!vma) seq_puts(m, "unused"); else diff --git a/drivers/gpu/drm/i915/i915_gem_fence_reg.c b/drivers/gpu/drm/i915/i915_gem_fence_reg.c index c9654f1a468f..6a33a0bb97a9 100644 --- a/drivers/gpu/drm/i915/i915_gem_fence_reg.c +++ b/drivers/gpu/drm/i915/i915_gem_fence_reg.c @@ -299,15 +299,24 @@ static int fence_update(struct i915_fence_reg *fence, */ int i915_vma_put_fence(struct i915_vma *vma) { + struct i915_ggtt *ggtt = i915_vm_to_ggtt(vma->vm); struct i915_fence_reg *fence = vma->fence; + int err; if (!fence) return 0; - if (fence->pin_count) + if (atomic_read(&fence->pin_count)) return -EBUSY; - return fence_update(fence, NULL); + err = mutex_lock_interruptible(&ggtt->vm.mutex); + if (err) + return err; + + err = fence_update(fence, NULL); + mutex_unlock(&ggtt->vm.mutex); + + return err; } static struct i915_fence_reg *fence_find(struct drm_i915_private *i915) @@ -317,7 +326,7 @@ static struct i915_fence_reg *fence_find(struct drm_i915_private *i915) list_for_each_entry(fence, &i915->ggtt.fence_list, link) { GEM_BUG_ON(fence->vma && fence->vma->fence != fence); - if (fence->pin_count) + if (atomic_read(&fence->pin_count)) continue; return fence; @@ -330,6 +339,48 @@ static struct i915_fence_reg *fence_find(struct drm_i915_private *i915) return ERR_PTR(-EDEADLK); } +static int __i915_vma_pin_fence(struct i915_vma *vma) +{ + struct i915_ggtt *ggtt = i915_vm_to_ggtt(vma->vm); + struct i915_fence_reg *fence; + struct i915_vma *set = i915_gem_object_is_tiled(vma->obj) ? vma : NULL; + int err; + + /* Just update our place in the LRU if our fence is getting reused. */ + if (vma->fence) { + fence = vma->fence; + GEM_BUG_ON(fence->vma != vma); + atomic_inc(&fence->pin_count); + if (!fence->dirty) { + list_move_tail(&fence->link, &ggtt->fence_list); + return 0; + } + } else if (set) { + fence = fence_find(vma->vm->i915); + if (IS_ERR(fence)) + return PTR_ERR(fence); + + GEM_BUG_ON(atomic_read(&fence->pin_count)); + atomic_inc(&fence->pin_count); + } else { + return 0; + } + + err = fence_update(fence, set); + if (err) + goto out_unpin; + + GEM_BUG_ON(fence->vma != set); + GEM_BUG_ON(vma->fence != (set ? fence : NULL)); + + if (set) + return 0; + +out_unpin: + atomic_dec(&fence->pin_count); + return err; +} + /** * i915_vma_pin_fence - set up fencing for a vma * @vma: vma to map through a fence reg @@ -350,8 +401,6 @@ static struct i915_fence_reg *fence_find(struct drm_i915_private *i915) */ int i915_vma_pin_fence(struct i915_vma *vma) { - struct i915_fence_reg *fence; - struct i915_vma *set = i915_gem_object_is_tiled(vma->obj) ? vma : NULL; int err; /* @@ -359,39 +408,16 @@ int i915_vma_pin_fence(struct i915_vma *vma) * must keep the device awake whilst using the fence. */ assert_rpm_wakelock_held(&vma->vm->i915->runtime_pm); + GEM_BUG_ON(!i915_vma_is_pinned(vma)); + GEM_BUG_ON(!i915_vma_is_ggtt(vma)); - /* Just update our place in the LRU if our fence is getting reused. */ - if (vma->fence) { - fence = vma->fence; - GEM_BUG_ON(fence->vma != vma); - fence->pin_count++; - if (!fence->dirty) { - list_move_tail(&fence->link, - &fence->i915->ggtt.fence_list); - return 0; - } - } else if (set) { - fence = fence_find(vma->vm->i915); - if (IS_ERR(fence)) - return PTR_ERR(fence); - - GEM_BUG_ON(fence->pin_count); - fence->pin_count++; - } else - return 0; - - err = fence_update(fence, set); + err = mutex_lock_interruptible(&vma->vm->mutex); if (err) - goto out_unpin; + return err; - GEM_BUG_ON(fence->vma != set); - GEM_BUG_ON(vma->fence != (set ? fence : NULL)); - - if (set) - return 0; + err = __i915_vma_pin_fence(vma); + mutex_unlock(&vma->vm->mutex); -out_unpin: - fence->pin_count--; return err; } @@ -404,16 +430,17 @@ out_unpin: */ struct i915_fence_reg *i915_reserve_fence(struct drm_i915_private *i915) { + struct i915_ggtt *ggtt = &i915->ggtt; struct i915_fence_reg *fence; int count; int ret; - lockdep_assert_held(&i915->drm.struct_mutex); + lockdep_assert_held(&ggtt->vm.mutex); /* Keep at least one fence available for the display engine. */ count = 0; - list_for_each_entry(fence, &i915->ggtt.fence_list, link) - count += !fence->pin_count; + list_for_each_entry(fence, &ggtt->fence_list, link) + count += !atomic_read(&fence->pin_count); if (count <= 1) return ERR_PTR(-ENOSPC); @@ -429,6 +456,7 @@ struct i915_fence_reg *i915_reserve_fence(struct drm_i915_private *i915) } list_del(&fence->link); + return fence; } @@ -440,9 +468,11 @@ struct i915_fence_reg *i915_reserve_fence(struct drm_i915_private *i915) */ void i915_unreserve_fence(struct i915_fence_reg *fence) { - lockdep_assert_held(&fence->i915->drm.struct_mutex); + struct i915_ggtt *ggtt = &fence->i915->ggtt; + + lockdep_assert_held(&ggtt->vm.mutex); - list_add(&fence->link, &fence->i915->ggtt.fence_list); + list_add(&fence->link, &ggtt->fence_list); } /** diff --git a/drivers/gpu/drm/i915/i915_gem_fence_reg.h b/drivers/gpu/drm/i915/i915_gem_fence_reg.h index 37e4f104f7c0..99866fb9d94f 100644 --- a/drivers/gpu/drm/i915/i915_gem_fence_reg.h +++ b/drivers/gpu/drm/i915/i915_gem_fence_reg.h @@ -41,7 +41,7 @@ struct i915_fence_reg { struct list_head link; struct drm_i915_private *i915; struct i915_vma *vma; - int pin_count; + atomic_t pin_count; int id; /** * Whether the tiling parameters for the currently diff --git a/drivers/gpu/drm/i915/i915_vma.h b/drivers/gpu/drm/i915/i915_vma.h index 9e7d8f4154b2..cf6c0437091d 100644 --- a/drivers/gpu/drm/i915/i915_vma.h +++ b/drivers/gpu/drm/i915/i915_vma.h @@ -426,8 +426,8 @@ int __must_check i915_vma_put_fence(struct i915_vma *vma); static inline void __i915_vma_unpin_fence(struct i915_vma *vma) { - GEM_BUG_ON(vma->fence->pin_count <= 0); - vma->fence->pin_count--; + GEM_BUG_ON(atomic_read(&vma->fence->pin_count) <= 0); + atomic_dec(&vma->fence->pin_count); } /** -- cgit v1.2.3 From 0a3242bdb47713e09cb004a0ba4947d3edf82d8a Mon Sep 17 00:00:00 2001 From: Xiaolin Zhang Date: Tue, 27 Aug 2019 16:39:23 +0800 Subject: drm/i915/gvt: update vgpu workload head pointer correctly when creating a vGPU workload, the guest context head pointer should be updated correctly by comparing with the exsiting workload in the guest worklod queue including the current running context. in some situation, there is a running context A and then received 2 new vGPU workload context B and A. in the new workload context A, it's head pointer should be updated with the running context A's tail. v2: walk through guest workload list in backward way. Cc: stable@vger.kernel.org Signed-off-by: Xiaolin Zhang Reviewed-by: Zhenyu Wang Signed-off-by: Zhenyu Wang --- drivers/gpu/drm/i915/gvt/scheduler.c | 28 +++++++++++++++------------- 1 file changed, 15 insertions(+), 13 deletions(-) (limited to 'drivers/gpu/drm/i915/gvt') diff --git a/drivers/gpu/drm/i915/gvt/scheduler.c b/drivers/gpu/drm/i915/gvt/scheduler.c index 7c99bbc3e2b8..affd38238725 100644 --- a/drivers/gpu/drm/i915/gvt/scheduler.c +++ b/drivers/gpu/drm/i915/gvt/scheduler.c @@ -1371,9 +1371,6 @@ static int prepare_mm(struct intel_vgpu_workload *workload) #define same_context(a, b) (((a)->context_id == (b)->context_id) && \ ((a)->lrca == (b)->lrca)) -#define get_last_workload(q) \ - (list_empty(q) ? NULL : container_of(q->prev, \ - struct intel_vgpu_workload, list)) /** * intel_vgpu_create_workload - create a vGPU workload * @vgpu: a vGPU @@ -1393,7 +1390,7 @@ intel_vgpu_create_workload(struct intel_vgpu *vgpu, int ring_id, { struct intel_vgpu_submission *s = &vgpu->submission; struct list_head *q = workload_q_head(vgpu, ring_id); - struct intel_vgpu_workload *last_workload = get_last_workload(q); + struct intel_vgpu_workload *last_workload = NULL; struct intel_vgpu_workload *workload = NULL; struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv; u64 ring_context_gpa; @@ -1416,15 +1413,20 @@ intel_vgpu_create_workload(struct intel_vgpu *vgpu, int ring_id, head &= RB_HEAD_OFF_MASK; tail &= RB_TAIL_OFF_MASK; - if (last_workload && same_context(&last_workload->ctx_desc, desc)) { - gvt_dbg_el("ring id %d cur workload == last\n", ring_id); - gvt_dbg_el("ctx head %x real head %lx\n", head, - last_workload->rb_tail); - /* - * cannot use guest context head pointer here, - * as it might not be updated at this time - */ - head = last_workload->rb_tail; + list_for_each_entry_reverse(last_workload, q, list) { + + if (same_context(&last_workload->ctx_desc, desc)) { + gvt_dbg_el("ring id %d cur workload == last\n", + ring_id); + gvt_dbg_el("ctx head %x real head %lx\n", head, + last_workload->rb_tail); + /* + * cannot use guest context head pointer here, + * as it might not be updated at this time + */ + head = last_workload->rb_tail; + break; + } } gvt_dbg_el("ring id %d begin a new workload\n", ring_id); -- cgit v1.2.3 From 4a5322560aa235efa84c0aa34c00e5749a0792fd Mon Sep 17 00:00:00 2001 From: Weinan Li Date: Mon, 2 Sep 2019 13:57:59 +0800 Subject: drm/i915/gvt: update RING_START reg of vGPU when the context is submitted to i915 The guest may use this register to identify the running state of one context. Emulate it as the value in context image as if the context runs on the GPU hardware. Signed-off-by: Weinan Li Reviewed-by: Zhenyu Wang Signed-off-by: Zhenyu Wang --- drivers/gpu/drm/i915/gvt/scheduler.c | 17 +++++++++++++++++ 1 file changed, 17 insertions(+) (limited to 'drivers/gpu/drm/i915/gvt') diff --git a/drivers/gpu/drm/i915/gvt/scheduler.c b/drivers/gpu/drm/i915/gvt/scheduler.c index affd38238725..30807a458050 100644 --- a/drivers/gpu/drm/i915/gvt/scheduler.c +++ b/drivers/gpu/drm/i915/gvt/scheduler.c @@ -573,6 +573,16 @@ static int prepare_shadow_wa_ctx(struct intel_shadow_wa_ctx *wa_ctx) return 0; } +static void update_vreg_in_ctx(struct intel_vgpu_workload *workload) +{ + struct intel_vgpu *vgpu = workload->vgpu; + struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv; + u32 ring_base; + + ring_base = dev_priv->engine[workload->ring_id]->mmio_base; + vgpu_vreg_t(vgpu, RING_START(ring_base)) = workload->rb_start; +} + static void release_shadow_batch_buffer(struct intel_vgpu_workload *workload) { struct intel_vgpu *vgpu = workload->vgpu; @@ -1002,6 +1012,13 @@ static int workload_thread(void *priv) if (need_force_wake) intel_uncore_forcewake_get(&gvt->dev_priv->uncore, FORCEWAKE_ALL); + /* + * Update the vReg of the vGPU which submitted this + * workload. The vGPU may use these registers for checking + * the context state. The value comes from GPU commands + * in this workload. + */ + update_vreg_in_ctx(workload); ret = dispatch_workload(workload); -- cgit v1.2.3