diff options
| author | Dave Airlie <airlied@redhat.com> | 2016-05-04 07:25:30 (GMT) |
|---|---|---|
| committer | Dave Airlie <airlied@redhat.com> | 2016-05-04 07:25:30 (GMT) |
| commit | fffb675106aef326bbd837612ad74c31ab060b93 (patch) | |
| tree | d48fe2eaee9346b49b8bc3beec4d235124745f9e | |
| parent | b89359bdf0f1e95a4c5f92300594ba9dde323fc4 (diff) | |
| parent | 5b4fd5b1111b1230cd037df3b314e7b36d45d483 (diff) | |
| download | linux-fffb675106aef326bbd837612ad74c31ab060b93.tar.xz | |
Merge tag 'drm-intel-next-2016-04-25' of git://anongit.freedesktop.org/drm-intel into drm-next
- more userptr cornercase fixes from Chris
- clean up and tune forcewake handling (Tvrtko)
- more underrun fixes from Ville, mostly for ilk to appeas CI
- fix unclaimed register warnings on vlv/chv and enable the debug code to catch
them by default (Ville)
- skl gpu hang fixes for gt3/4 (Mika Kuoppala)
- edram improvements for gen9+ (Mika again)
- clean up gpu reset corner cases (Chris)
- fix ctx/ring machine deaths on snb/ilk (Chris)
- MOCS programming for all engines (Peter Antoine)
- robustify/clean up vlv/chv irq handler (Ville)
- split gen8+ irq handlers into ack/handle phase (Ville)
- tons of bxt rpm fixes (mostly around firmware interactions), from Imre
- hook up panel fitting for dsi panels (Ville)
- more runtime PM fixes all over from Imre
- shrinker polish (Chris)
- more guc fixes from Alex Dai and Dave Gordon
- tons of bugfixes and small polish all over (but with a big focus on bxt)
* tag 'drm-intel-next-2016-04-25' of git://anongit.freedesktop.org/drm-intel: (142 commits)
drm/i915: Update DRIVER_DATE to 20160425
drm/i915/bxt: Explicitly clear the Turbo control register
drm/i915: Correct the i915_frequency_info debugfs output
drm/i915: Macros to convert PM time interval values to microseconds
drm/i915: Make RPS EI/thresholds multiple of 25 on SNB-BDW
drm/i915: Fake HDMI live status
drm/i915/bxt: Force reprogramming a PHY with invalid HW state
drm/i915/bxt: Wait for PHY1 GRC done if PHY0 was already enabled
drm/i915/bxt: Use PHY0 GRC value for HW state verification
drm/i915: use dev_priv directly in gen8_ppgtt_notify_vgt
drm/i915/bxt: Enable DC5 during runtime resume
drm/i915/bxt: Sanitize DC state tracking during system resume
drm/i915/bxt: Don't uninit/init display core twice during system suspend/resume
drm/i915: Inline intel_suspend_complete
drm/i915/kbl: Don't WARN for expected secondary MISC IO power well request
drm/i915: Fix eDP low vswing for Broadwell
drm/i915: check for ERR_PTR from i915_gem_object_pin_map()
drm/i915/guc: local optimisations and updating comments
drm/i915/guc: drop cached copy of 'wq_head'
drm/i915/guc: keep GuC doorbell & process descriptor mapped in kernel
...
42 files changed, 2689 insertions, 1653 deletions
diff --git a/drivers/gpu/drm/i915/Kconfig.debug b/drivers/gpu/drm/i915/Kconfig.debug index 649a562..8f40410 100644 --- a/drivers/gpu/drm/i915/Kconfig.debug +++ b/drivers/gpu/drm/i915/Kconfig.debug @@ -1,3 +1,20 @@ +config DRM_I915_WERROR + bool "Force GCC to throw an error instead of a warning when compiling" + # As this may inadvertently break the build, only allow the user + # to shoot oneself in the foot iff they aim really hard + depends on EXPERT + # We use the dependency on !COMPILE_TEST to not be enabled in + # allmodconfig or allyesconfig configurations + depends on !COMPILE_TEST + default n + help + Add -Werror to the build flags for (and only for) i915.ko. + Do not enable this unless you are writing code for the i915.ko module. + + Recommended for driver developers only. + + If in doubt, say "N". + config DRM_I915_DEBUG bool "Enable additional driver debugging" depends on DRM_I915 @@ -10,3 +27,15 @@ config DRM_I915_DEBUG If in doubt, say "N". +config DRM_I915_DEBUG_GEM + bool "Insert extra checks into the GEM internals" + default n + depends on DRM_I915_WERROR + help + Enable extra sanity checks (including BUGs) along the GEM driver + paths that may slow the system down and if hit hang the machine. + + Recommended for driver developers only. + + If in doubt, say "N". + diff --git a/drivers/gpu/drm/i915/Makefile b/drivers/gpu/drm/i915/Makefile index 7ffb51b..0b88ba0 100644 --- a/drivers/gpu/drm/i915/Makefile +++ b/drivers/gpu/drm/i915/Makefile @@ -2,6 +2,8 @@ # Makefile for the drm device driver. This driver provides support for the # Direct Rendering Infrastructure (DRI) in XFree86 4.1.0 and higher. +subdir-ccflags-$(CONFIG_DRM_I915_WERROR) := -Werror + # Please keep these build lists sorted! # core driver code diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c index 8593394..4950d05 100644 --- a/drivers/gpu/drm/i915/i915_debugfs.c +++ b/drivers/gpu/drm/i915/i915_debugfs.c @@ -89,27 +89,34 @@ static int i915_capabilities(struct seq_file *m, void *data) return 0; } -static const char *get_pin_flag(struct drm_i915_gem_object *obj) +static const char get_active_flag(struct drm_i915_gem_object *obj) { - if (obj->pin_display) - return "p"; - else - return " "; + return obj->active ? '*' : ' '; } -static const char *get_tiling_flag(struct drm_i915_gem_object *obj) +static const char get_pin_flag(struct drm_i915_gem_object *obj) +{ + return obj->pin_display ? 'p' : ' '; +} + +static const char get_tiling_flag(struct drm_i915_gem_object *obj) { switch (obj->tiling_mode) { default: - case I915_TILING_NONE: return " "; - case I915_TILING_X: return "X"; - case I915_TILING_Y: return "Y"; + case I915_TILING_NONE: return ' '; + case I915_TILING_X: return 'X'; + case I915_TILING_Y: return 'Y'; } } -static inline const char *get_global_flag(struct drm_i915_gem_object *obj) +static inline const char get_global_flag(struct drm_i915_gem_object *obj) { - return i915_gem_obj_to_ggtt(obj) ? "g" : " "; + return i915_gem_obj_to_ggtt(obj) ? 'g' : ' '; +} + +static inline const char get_pin_mapped_flag(struct drm_i915_gem_object *obj) +{ + return obj->mapping ? 'M' : ' '; } static u64 i915_gem_obj_total_ggtt_size(struct drm_i915_gem_object *obj) @@ -136,12 +143,13 @@ describe_obj(struct seq_file *m, struct drm_i915_gem_object *obj) lockdep_assert_held(&obj->base.dev->struct_mutex); - seq_printf(m, "%pK: %s%s%s%s %8zdKiB %02x %02x [ ", + seq_printf(m, "%pK: %c%c%c%c%c %8zdKiB %02x %02x [ ", &obj->base, - obj->active ? "*" : " ", + get_active_flag(obj), get_pin_flag(obj), get_tiling_flag(obj), get_global_flag(obj), + get_pin_mapped_flag(obj), obj->base.size / 1024, obj->base.read_domains, obj->base.write_domain); @@ -435,6 +443,8 @@ static int i915_gem_object_info(struct seq_file *m, void* data) struct i915_ggtt *ggtt = &dev_priv->ggtt; u32 count, mappable_count, purgeable_count; u64 size, mappable_size, purgeable_size; + unsigned long pin_mapped_count = 0, pin_mapped_purgeable_count = 0; + u64 pin_mapped_size = 0, pin_mapped_purgeable_size = 0; struct drm_i915_gem_object *obj; struct drm_file *file; struct i915_vma *vma; @@ -468,6 +478,14 @@ static int i915_gem_object_info(struct seq_file *m, void* data) size += obj->base.size, ++count; if (obj->madv == I915_MADV_DONTNEED) purgeable_size += obj->base.size, ++purgeable_count; + if (obj->mapping) { + pin_mapped_count++; + pin_mapped_size += obj->base.size; + if (obj->pages_pin_count == 0) { + pin_mapped_purgeable_count++; + pin_mapped_purgeable_size += obj->base.size; + } + } } seq_printf(m, "%u unbound objects, %llu bytes\n", count, size); @@ -485,6 +503,14 @@ static int i915_gem_object_info(struct seq_file *m, void* data) purgeable_size += obj->base.size; ++purgeable_count; } + if (obj->mapping) { + pin_mapped_count++; + pin_mapped_size += obj->base.size; + if (obj->pages_pin_count == 0) { + pin_mapped_purgeable_count++; + pin_mapped_purgeable_size += obj->base.size; + } + } } seq_printf(m, "%u purgeable objects, %llu bytes\n", purgeable_count, purgeable_size); @@ -492,6 +518,10 @@ static int i915_gem_object_info(struct seq_file *m, void* data) mappable_count, mappable_size); seq_printf(m, "%u fault mappable objects, %llu bytes\n", count, size); + seq_printf(m, + "%lu [%lu] pin mapped objects, %llu [%llu] bytes [purgeable]\n", + pin_mapped_count, pin_mapped_purgeable_count, + pin_mapped_size, pin_mapped_purgeable_size); seq_printf(m, "%llu [%llu] gtt total\n", ggtt->base.total, ggtt->mappable_end - ggtt->base.start); @@ -1216,12 +1246,12 @@ static int i915_frequency_info(struct seq_file *m, void *unused) rpdeclimit = I915_READ(GEN6_RP_DOWN_THRESHOLD); rpstat = I915_READ(GEN6_RPSTAT1); - rpupei = I915_READ(GEN6_RP_CUR_UP_EI); - rpcurup = I915_READ(GEN6_RP_CUR_UP); - rpprevup = I915_READ(GEN6_RP_PREV_UP); - rpdownei = I915_READ(GEN6_RP_CUR_DOWN_EI); - rpcurdown = I915_READ(GEN6_RP_CUR_DOWN); - rpprevdown = I915_READ(GEN6_RP_PREV_DOWN); + rpupei = I915_READ(GEN6_RP_CUR_UP_EI) & GEN6_CURICONT_MASK; + rpcurup = I915_READ(GEN6_RP_CUR_UP) & GEN6_CURBSYTAVG_MASK; + rpprevup = I915_READ(GEN6_RP_PREV_UP) & GEN6_CURBSYTAVG_MASK; + rpdownei = I915_READ(GEN6_RP_CUR_DOWN_EI) & GEN6_CURIAVG_MASK; + rpcurdown = I915_READ(GEN6_RP_CUR_DOWN) & GEN6_CURBSYTAVG_MASK; + rpprevdown = I915_READ(GEN6_RP_PREV_DOWN) & GEN6_CURBSYTAVG_MASK; if (IS_GEN9(dev)) cagf = (rpstat & GEN9_CAGF_MASK) >> GEN9_CAGF_SHIFT; else if (IS_HASWELL(dev) || IS_BROADWELL(dev)) @@ -1261,21 +1291,21 @@ static int i915_frequency_info(struct seq_file *m, void *unused) seq_printf(m, "RPDECLIMIT: 0x%08x\n", rpdeclimit); seq_printf(m, "RPNSWREQ: %dMHz\n", reqf); seq_printf(m, "CAGF: %dMHz\n", cagf); - seq_printf(m, "RP CUR UP EI: %dus\n", rpupei & - GEN6_CURICONT_MASK); - seq_printf(m, "RP CUR UP: %dus\n", rpcurup & - GEN6_CURBSYTAVG_MASK); - seq_printf(m, "RP PREV UP: %dus\n", rpprevup & - GEN6_CURBSYTAVG_MASK); + seq_printf(m, "RP CUR UP EI: %d (%dus)\n", + rpupei, GT_PM_INTERVAL_TO_US(dev_priv, rpupei)); + seq_printf(m, "RP CUR UP: %d (%dus)\n", + rpcurup, GT_PM_INTERVAL_TO_US(dev_priv, rpcurup)); + seq_printf(m, "RP PREV UP: %d (%dus)\n", + rpprevup, GT_PM_INTERVAL_TO_US(dev_priv, rpprevup)); seq_printf(m, "Up threshold: %d%%\n", dev_priv->rps.up_threshold); - seq_printf(m, "RP CUR DOWN EI: %dus\n", rpdownei & - GEN6_CURIAVG_MASK); - seq_printf(m, "RP CUR DOWN: %dus\n", rpcurdown & - GEN6_CURBSYTAVG_MASK); - seq_printf(m, "RP PREV DOWN: %dus\n", rpprevdown & - GEN6_CURBSYTAVG_MASK); + seq_printf(m, "RP CUR DOWN EI: %d (%dus)\n", + rpdownei, GT_PM_INTERVAL_TO_US(dev_priv, rpdownei)); + seq_printf(m, "RP CUR DOWN: %d (%dus)\n", + rpcurdown, GT_PM_INTERVAL_TO_US(dev_priv, rpcurdown)); + seq_printf(m, "RP PREV DOWN: %d (%dus)\n", + rpprevdown, GT_PM_INTERVAL_TO_US(dev_priv, rpprevdown)); seq_printf(m, "Down threshold: %d%%\n", dev_priv->rps.down_threshold); @@ -1469,12 +1499,11 @@ static int i915_forcewake_domains(struct seq_file *m, void *data) struct drm_device *dev = node->minor->dev; struct drm_i915_private *dev_priv = dev->dev_private; struct intel_uncore_forcewake_domain *fw_domain; - int i; spin_lock_irq(&dev_priv->uncore.lock); - for_each_fw_domain(fw_domain, dev_priv, i) { + for_each_fw_domain(fw_domain, dev_priv) { seq_printf(m, "%s.wake_count = %u\n", - intel_uncore_forcewake_domain_to_str(i), + intel_uncore_forcewake_domain_to_str(fw_domain->id), fw_domain->wake_count); } spin_unlock_irq(&dev_priv->uncore.lock); @@ -2405,10 +2434,11 @@ static int i915_llc(struct seq_file *m, void *data) struct drm_info_node *node = m->private; struct drm_device *dev = node->minor->dev; struct drm_i915_private *dev_priv = dev->dev_private; + const bool edram = INTEL_GEN(dev_priv) > 8; - /* Size calculation for LLC is a bit of a pain. Ignore for now. */ seq_printf(m, "LLC: %s\n", yesno(HAS_LLC(dev))); - seq_printf(m, "eLLC: %zuMB\n", dev_priv->ellc_size); + seq_printf(m, "%s: %lluMB\n", edram ? "eDRAM" : "eLLC", + intel_uncore_edram_size(dev_priv)/1024/1024); return 0; } @@ -4723,7 +4753,7 @@ i915_wedged_get(void *data, u64 *val) struct drm_device *dev = data; struct drm_i915_private *dev_priv = dev->dev_private; - *val = atomic_read(&dev_priv->gpu_error.reset_counter); + *val = i915_terminally_wedged(&dev_priv->gpu_error); return 0; } diff --git a/drivers/gpu/drm/i915/i915_dma.c b/drivers/gpu/drm/i915/i915_dma.c index b377753..5c76150 100644 --- a/drivers/gpu/drm/i915/i915_dma.c +++ b/drivers/gpu/drm/i915/i915_dma.c @@ -257,13 +257,6 @@ static int i915_get_bridge_dev(struct drm_device *dev) return 0; } -#define MCHBAR_I915 0x44 -#define MCHBAR_I965 0x48 -#define MCHBAR_SIZE (4*4096) - -#define DEVEN_REG 0x54 -#define DEVEN_MCHBAR_EN (1 << 28) - /* Allocate space for the MCH regs if needed, return nonzero on error */ static int intel_alloc_mchbar_resource(struct drm_device *dev) @@ -325,7 +318,7 @@ intel_setup_mchbar(struct drm_device *dev) dev_priv->mchbar_need_disable = false; if (IS_I915G(dev) || IS_I915GM(dev)) { - pci_read_config_dword(dev_priv->bridge_dev, DEVEN_REG, &temp); + pci_read_config_dword(dev_priv->bridge_dev, DEVEN, &temp); enabled = !!(temp & DEVEN_MCHBAR_EN); } else { pci_read_config_dword(dev_priv->bridge_dev, mchbar_reg, &temp); @@ -343,7 +336,7 @@ intel_setup_mchbar(struct drm_device *dev) /* Space is allocated or reserved, so enable it. */ if (IS_I915G(dev) || IS_I915GM(dev)) { - pci_write_config_dword(dev_priv->bridge_dev, DEVEN_REG, + pci_write_config_dword(dev_priv->bridge_dev, DEVEN, temp | DEVEN_MCHBAR_EN); } else { pci_read_config_dword(dev_priv->bridge_dev, mchbar_reg, &temp); @@ -356,17 +349,24 @@ intel_teardown_mchbar(struct drm_device *dev) { struct drm_i915_private *dev_priv = dev->dev_private; int mchbar_reg = INTEL_INFO(dev)->gen >= 4 ? MCHBAR_I965 : MCHBAR_I915; - u32 temp; if (dev_priv->mchbar_need_disable) { if (IS_I915G(dev) || IS_I915GM(dev)) { - pci_read_config_dword(dev_priv->bridge_dev, DEVEN_REG, &temp); - temp &= ~DEVEN_MCHBAR_EN; - pci_write_config_dword(dev_priv->bridge_dev, DEVEN_REG, temp); + u32 deven_val; + + pci_read_config_dword(dev_priv->bridge_dev, DEVEN, + &deven_val); + deven_val &= ~DEVEN_MCHBAR_EN; + pci_write_config_dword(dev_priv->bridge_dev, DEVEN, + deven_val); } else { - pci_read_config_dword(dev_priv->bridge_dev, mchbar_reg, &temp); - temp &= ~1; - pci_write_config_dword(dev_priv->bridge_dev, mchbar_reg, temp); + u32 mchbar_val; + + pci_read_config_dword(dev_priv->bridge_dev, mchbar_reg, + &mchbar_val); + mchbar_val &= ~1; + pci_write_config_dword(dev_priv->bridge_dev, mchbar_reg, + mchbar_val); } } diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c index 29b4e79..d37c0a6 100644 --- a/drivers/gpu/drm/i915/i915_drv.c +++ b/drivers/gpu/drm/i915/i915_drv.c @@ -567,10 +567,9 @@ static void intel_suspend_encoders(struct drm_i915_private *dev_priv) drm_modeset_unlock_all(dev); } -static int intel_suspend_complete(struct drm_i915_private *dev_priv); static int vlv_resume_prepare(struct drm_i915_private *dev_priv, bool rpm_resume); -static int bxt_resume_prepare(struct drm_i915_private *dev_priv); +static int vlv_suspend_complete(struct drm_i915_private *dev_priv); static bool suspend_to_idle(struct drm_i915_private *dev_priv) { @@ -640,8 +639,7 @@ static int i915_drm_suspend(struct drm_device *dev) intel_display_set_init_power(dev_priv, false); - if (HAS_CSR(dev_priv)) - flush_work(&dev_priv->csr.work); + intel_csr_ucode_suspend(dev_priv); out: enable_rpm_wakeref_asserts(dev_priv); @@ -657,7 +655,8 @@ static int i915_drm_suspend_late(struct drm_device *drm_dev, bool hibernation) disable_rpm_wakeref_asserts(dev_priv); - fw_csr = suspend_to_idle(dev_priv) && dev_priv->csr.dmc_payload; + fw_csr = !IS_BROXTON(dev_priv) && + suspend_to_idle(dev_priv) && dev_priv->csr.dmc_payload; /* * In case of firmware assisted context save/restore don't manually * deinit the power domains. This also means the CSR/DMC firmware will @@ -668,7 +667,13 @@ static int i915_drm_suspend_late(struct drm_device *drm_dev, bool hibernation) if (!fw_csr) intel_power_domains_suspend(dev_priv); - ret = intel_suspend_complete(dev_priv); + ret = 0; + if (IS_BROXTON(dev_priv)) + bxt_enable_dc9(dev_priv); + else if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv)) + hsw_enable_pc8(dev_priv); + else if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv)) + ret = vlv_suspend_complete(dev_priv); if (ret) { DRM_ERROR("Suspend complete failed: %d\n", ret); @@ -732,6 +737,8 @@ static int i915_drm_resume(struct drm_device *dev) disable_rpm_wakeref_asserts(dev_priv); + intel_csr_ucode_resume(dev_priv); + mutex_lock(&dev->struct_mutex); i915_gem_restore_gtt_mappings(dev); mutex_unlock(&dev->struct_mutex); @@ -802,7 +809,7 @@ static int i915_drm_resume(struct drm_device *dev) static int i915_drm_resume_early(struct drm_device *dev) { struct drm_i915_private *dev_priv = dev->dev_private; - int ret = 0; + int ret; /* * We have a resume ordering issue with the snd-hda driver also @@ -813,6 +820,36 @@ static int i915_drm_resume_early(struct drm_device *dev) * FIXME: This should be solved with a special hdmi sink device or * similar so that power domains can be employed. */ + + /* + * Note that we need to set the power state explicitly, since we + * powered off the device during freeze and the PCI core won't power + * it back up for us during thaw. Powering off the device during + * freeze is not a hard requirement though, and during the + * suspend/resume phases the PCI core makes sure we get here with the + * device powered on. So in case we change our freeze logic and keep + * the device powered we can also remove the following set power state + * call. + */ + ret = pci_set_power_state(dev->pdev, PCI_D0); + if (ret) { + DRM_ERROR("failed to set PCI D0 power state (%d)\n", ret); + goto out; + } + + /* + * Note that pci_enable_device() first enables any parent bridge + * device and only then sets the power state for this device. The + * bridge enabling is a nop though, since bridge devices are resumed + * first. The order of enabling power and enabling the device is + * imposed by the PCI core as described above, so here we preserve the + * same order for the freeze/thaw phases. + * + * TODO: eventually we should remove pci_disable_device() / + * pci_enable_enable_device() from suspend/resume. Due to how they + * depend on the device enable refcount we can't anyway depend on them + * disabling/enabling the device. + */ if (pci_enable_device(dev->pdev)) { ret = -EIO; goto out; @@ -830,21 +867,25 @@ static int i915_drm_resume_early(struct drm_device *dev) intel_uncore_early_sanitize(dev, true); - if (IS_BROXTON(dev)) - ret = bxt_resume_prepare(dev_priv); - else if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv)) + if (IS_BROXTON(dev)) { + if (!dev_priv->suspended_to_idle) + gen9_sanitize_dc_state(dev_priv); + bxt_disable_dc9(dev_priv); + } else if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv)) { hsw_disable_pc8(dev_priv); + } intel_uncore_sanitize(dev); - if (!(dev_priv->suspended_to_idle && dev_priv->csr.dmc_payload)) + if (IS_BROXTON(dev_priv) || + !(dev_priv->suspended_to_idle && dev_priv->csr.dmc_payload)) intel_power_domains_init_hw(dev_priv, true); + enable_rpm_wakeref_asserts(dev_priv); + out: dev_priv->suspended_to_idle = false; - enable_rpm_wakeref_asserts(dev_priv); - return ret; } @@ -880,23 +921,32 @@ int i915_resume_switcheroo(struct drm_device *dev) int i915_reset(struct drm_device *dev) { struct drm_i915_private *dev_priv = dev->dev_private; - bool simulated; + struct i915_gpu_error *error = &dev_priv->gpu_error; + unsigned reset_counter; int ret; intel_reset_gt_powersave(dev); mutex_lock(&dev->struct_mutex); - i915_gem_reset(dev); + /* Clear any previous failed attempts at recovery. Time to try again. */ + atomic_andnot(I915_WEDGED, &error->reset_counter); - simulated = dev_priv->gpu_error.stop_rings != 0; + /* Clear the reset-in-progress flag and increment the reset epoch. */ + reset_counter = atomic_inc_return(&error->reset_counter); + if (WARN_ON(__i915_reset_in_progress(reset_counter))) { + ret = -EIO; + goto error; + } + + i915_gem_reset(dev); ret = intel_gpu_reset(dev, ALL_ENGINES); /* Also reset the gpu hangman. */ - if (simulated) { + if (error->stop_rings != 0) { DRM_INFO("Simulated gpu hang, resetting stop_rings\n"); - dev_priv->gpu_error.stop_rings = 0; + error->stop_rings = 0; if (ret == -ENODEV) { DRM_INFO("Reset not implemented, but ignoring " "error for simulated gpu hangs\n"); @@ -908,9 +958,11 @@ int i915_reset(struct drm_device *dev) pr_notice("drm/i915: Resetting chip after gpu hang\n"); if (ret) { - DRM_ERROR("Failed to reset chip: %i\n", ret); - mutex_unlock(&dev->struct_mutex); - return ret; + if (ret != -ENODEV) + DRM_ERROR("Failed to reset chip: %i\n", ret); + else + DRM_DEBUG_DRIVER("GPU reset disabled\n"); + goto error; } intel_overlay_reset(dev_priv); @@ -929,20 +981,14 @@ int i915_reset(struct drm_device *dev) * was running at the time of the reset (i.e. we weren't VT * switched away). */ - - /* Used to prevent gem_check_wedged returning -EAGAIN during gpu reset */ - dev_priv->gpu_error.reload_in_reset = true; - ret = i915_gem_init_hw(dev); - - dev_priv->gpu_error.reload_in_reset = false; - - mutex_unlock(&dev->struct_mutex); if (ret) { DRM_ERROR("Failed hw init on reset %d\n", ret); - return ret; + goto error; } + mutex_unlock(&dev->struct_mutex); + /* * rps/rc6 re-init is necessary to restore state lost after the * reset and the re-install of gt irqs. Skip for ironlake per @@ -953,6 +999,11 @@ int i915_reset(struct drm_device *dev) intel_enable_gt_powersave(dev); return 0; + +error: + atomic_or(I915_WEDGED, &error->reset_counter); + mutex_unlock(&dev->struct_mutex); + return ret; } static int i915_pci_probe(struct pci_dev *pdev, const struct pci_device_id *ent) @@ -1059,44 +1110,6 @@ static int i915_pm_resume(struct device *dev) return i915_drm_resume(drm_dev); } -static int hsw_suspend_complete(struct drm_i915_private *dev_priv) -{ - hsw_enable_pc8(dev_priv); - - return 0; -} - -static int bxt_suspend_complete(struct drm_i915_private *dev_priv) -{ - struct drm_device *dev = dev_priv->dev; - - /* TODO: when DC5 support is added disable DC5 here. */ - - broxton_ddi_phy_uninit(dev); - broxton_uninit_cdclk(dev); - bxt_enable_dc9(dev_priv); - - return 0; -} - -static int bxt_resume_prepare(struct drm_i915_private *dev_priv) -{ - struct drm_device *dev = dev_priv->dev; - - /* TODO: when CSR FW support is added make sure the FW is loaded */ - - bxt_disable_dc9(dev_priv); - - /* - * TODO: when DC5 support is added enable DC5 here if the CSR FW - * is available. - */ - broxton_init_cdclk(dev); - broxton_ddi_phy_init(dev); - - return 0; -} - /* * Save all Gunit registers that may be lost after a D3 and a subsequent * S0i[R123] transition. The list of registers needing a save/restore is @@ -1502,7 +1515,16 @@ static int intel_runtime_suspend(struct device *device) intel_suspend_gt_powersave(dev); intel_runtime_pm_disable_interrupts(dev_priv); - ret = intel_suspend_complete(dev_priv); + ret = 0; + if (IS_BROXTON(dev_priv)) { + bxt_display_core_uninit(dev_priv); + bxt_enable_dc9(dev_priv); + } else if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv)) { + hsw_enable_pc8(dev_priv); + } else if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv)) { + ret = vlv_suspend_complete(dev_priv); + } + if (ret) { DRM_ERROR("Runtime suspend failed, disabling it (%d)\n", ret); intel_runtime_pm_enable_interrupts(dev_priv); @@ -1576,12 +1598,17 @@ static int intel_runtime_resume(struct device *device) if (IS_GEN6(dev_priv)) intel_init_pch_refclk(dev); - if (IS_BROXTON(dev)) - ret = bxt_resume_prepare(dev_priv); - else if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv)) + if (IS_BROXTON(dev)) { + bxt_disable_dc9(dev_priv); + bxt_display_core_init(dev_priv, true); + if (dev_priv->csr.dmc_payload && + (dev_priv->csr.allowed_dc_mask & DC_STATE_EN_UPTO_DC5)) + gen9_enable_dc5(dev_priv); + } else if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv)) { hsw_disable_pc8(dev_priv); - else if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv)) + } else if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv)) { ret = vlv_resume_prepare(dev_priv, true); + } /* * No point of rolling back things in case of an error, as the best @@ -1612,26 +1639,6 @@ static int intel_runtime_resume(struct device *device) return ret; } -/* - * This function implements common functionality of runtime and system - * suspend sequence. - */ -static int intel_suspend_complete(struct drm_i915_private *dev_priv) -{ - int ret; - - if (IS_BROXTON(dev_priv)) - ret = bxt_suspend_complete(dev_priv); - else if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv)) - ret = hsw_suspend_complete(dev_priv); - else if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv)) - ret = vlv_suspend_complete(dev_priv); - else - ret = 0; - - return ret; -} - static const struct dev_pm_ops i915_pm_ops = { /* * S0ix (via system suspend) and S3 event handlers [PMSG_SUSPEND, diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h index a9c8211..9d7b54e 100644 --- a/drivers/gpu/drm/i915/i915_drv.h +++ b/drivers/gpu/drm/i915/i915_drv.h @@ -33,34 +33,40 @@ #include <uapi/drm/i915_drm.h> #include <uapi/drm/drm_fourcc.h> -#include <drm/drmP.h> -#include "i915_params.h" -#include "i915_reg.h" -#include "intel_bios.h" -#include "intel_ringbuffer.h" -#include "intel_lrc.h" -#include "i915_gem_gtt.h" -#include "i915_gem_render_state.h" #include <linux/io-mapping.h> #include <linux/i2c.h> #include <linux/i2c-algo-bit.h> -#include <drm/intel-gtt.h> -#include <drm/drm_legacy.h> /* for struct drm_dma_handle */ -#include <drm/drm_gem.h> #include <linux/backlight.h> #include <linux/hashtable.h> #include <linux/intel-iommu.h> #include <linux/kref.h> #include <linux/pm_qos.h> -#include "intel_guc.h" +#include <linux/shmem_fs.h> + +#include <drm/drmP.h> +#include <drm/intel-gtt.h> +#include <drm/drm_legacy.h> /* for struct drm_dma_handle */ +#include <drm/drm_gem.h> + +#include "i915_params.h" +#include "i915_reg.h" + +#include "intel_bios.h" #include "intel_dpll_mgr.h" +#include "intel_guc.h" +#include "intel_lrc.h" +#include "intel_ringbuffer.h" + +#include "i915_gem.h" +#include "i915_gem_gtt.h" +#include "i915_gem_render_state.h" /* General customization: */ #define DRIVER_NAME "i915" #define DRIVER_DESC "Intel Graphics" -#define DRIVER_DATE "20160411" +#define DRIVER_DATE "20160425" #undef WARN_ON /* Many gcc seem to no see through this and fall over :( */ @@ -634,6 +640,13 @@ enum forcewake_domains { FORCEWAKE_MEDIA) }; +#define FW_REG_READ (1) +#define FW_REG_WRITE (2) + +enum forcewake_domains +intel_uncore_forcewake_for_reg(struct drm_i915_private *dev_priv, + i915_reg_t reg, unsigned int op); + struct intel_uncore_funcs { void (*force_wake_get)(struct drm_i915_private *dev_priv, enum forcewake_domains domains); @@ -666,8 +679,9 @@ struct intel_uncore { struct intel_uncore_forcewake_domain { struct drm_i915_private *i915; enum forcewake_domain_id id; + enum forcewake_domains mask; unsigned wake_count; - struct timer_list timer; + struct hrtimer timer; i915_reg_t reg_set; u32 val_set; u32 val_clear; @@ -680,14 +694,14 @@ struct intel_uncore { }; /* Iterate over initialised fw domains */ -#define for_each_fw_domain_mask(domain__, mask__, dev_priv__, i__) \ - for ((i__) = 0, (domain__) = &(dev_priv__)->uncore.fw_domain[0]; \ - (i__) < FW_DOMAIN_ID_COUNT; \ - (i__)++, (domain__) = &(dev_priv__)->uncore.fw_domain[i__]) \ - for_each_if (((mask__) & (dev_priv__)->uncore.fw_domains) & (1 << (i__))) +#define for_each_fw_domain_masked(domain__, mask__, dev_priv__) \ + for ((domain__) = &(dev_priv__)->uncore.fw_domain[0]; \ + (domain__) < &(dev_priv__)->uncore.fw_domain[FW_DOMAIN_ID_COUNT]; \ + (domain__)++) \ + for_each_if ((mask__) & (domain__)->mask) -#define for_each_fw_domain(domain__, dev_priv__, i__) \ - for_each_fw_domain_mask(domain__, FORCEWAKE_ALL, dev_priv__, i__) +#define for_each_fw_domain(domain__, dev_priv__) \ + for_each_fw_domain_masked(domain__, FORCEWAKE_ALL, dev_priv__) #define CSR_VERSION(major, minor) ((major) << 16 | (minor)) #define CSR_VERSION_MAJOR(version) ((version) >> 16) @@ -996,6 +1010,7 @@ struct intel_fbc_work; struct intel_gmbus { struct i2c_adapter adapter; +#define GMBUS_FORCE_BIT_RETRY (1U << 31) u32 force_bit; u32 reg0; i915_reg_t gpio_reg; @@ -1385,9 +1400,6 @@ struct i915_gpu_error { /* For missed irq/seqno simulation. */ unsigned int test_irq_rings; - - /* Used to prevent gem_check_wedged returning -EAGAIN during gpu reset */ - bool reload_in_reset; }; enum modeset_restore { @@ -1444,6 +1456,7 @@ struct intel_vbt_data { unsigned int lvds_use_ssc:1; unsigned int display_clock_mode:1; unsigned int fdi_rx_polarity_inverted:1; + unsigned int panel_type:4; int lvds_ssc_freq; unsigned int bios_lvds_val; /* initial [PCH_]LVDS reg val in VBIOS */ @@ -1863,7 +1876,7 @@ struct drm_i915_private { struct intel_l3_parity l3_parity; /* Cannot be determined by PCIID. You must always read a register. */ - size_t ellc_size; + u32 edram_cap; /* gen6+ rps state */ struct intel_gen6_power_mgmt rps; @@ -1911,6 +1924,7 @@ struct drm_i915_private { * crappiness (can't read out DPLL_MD for pipes B & C). */ u32 chv_dpll_md[I915_MAX_PIPES]; + u32 bxt_phy_grc; u32 suspend_count; bool suspended_to_idle; @@ -2237,6 +2251,7 @@ struct drm_i915_gem_request { /** On Which ring this request was generated */ struct drm_i915_private *i915; struct intel_engine_cs *engine; + unsigned reset_counter; /** GEM sequence number associated with the previous request, * when the HWS breadcrumb is equal to this the GPU is processing @@ -2317,7 +2332,6 @@ struct drm_i915_gem_request { struct drm_i915_gem_request * __must_check i915_gem_request_alloc(struct intel_engine_cs *engine, struct intel_context *ctx); -void i915_gem_request_cancel(struct drm_i915_gem_request *req); void i915_gem_request_free(struct kref *req_ref); int i915_gem_request_add_to_client(struct drm_i915_gem_request *req, struct drm_file *file); @@ -2487,6 +2501,7 @@ struct drm_i915_cmd_table { __p; \ }) #define INTEL_INFO(p) (&__I915__(p)->info) +#define INTEL_GEN(p) (INTEL_INFO(p)->gen) #define INTEL_DEVID(p) (INTEL_INFO(p)->device_id) #define INTEL_REVID(p) (__I915__(p)->dev->pdev->revision) @@ -2613,8 +2628,9 @@ struct drm_i915_cmd_table { #define HAS_VEBOX(dev) (INTEL_INFO(dev)->ring_mask & VEBOX_RING) #define HAS_LLC(dev) (INTEL_INFO(dev)->has_llc) #define HAS_SNOOP(dev) (INTEL_INFO(dev)->has_snoop) +#define HAS_EDRAM(dev) (__I915__(dev)->edram_cap & EDRAM_ENABLED) #define HAS_WT(dev) ((IS_HASWELL(dev) || IS_BROADWELL(dev)) && \ - __I915__(dev)->ellc_size) + HAS_EDRAM(dev)) #define I915_NEED_GFX_HWS(dev) (INTEL_INFO(dev)->need_gfx_hws) #define HAS_HW_CONTEXTS(dev) (INTEL_INFO(dev)->gen >= 6) @@ -2631,8 +2647,9 @@ struct drm_i915_cmd_table { /* WaRsDisableCoarsePowerGating:skl,bxt */ #define NEEDS_WaRsDisableCoarsePowerGating(dev) (IS_BXT_REVID(dev, 0, BXT_REVID_A1) || \ - ((IS_SKL_GT3(dev) || IS_SKL_GT4(dev)) && \ - IS_SKL_REVID(dev, 0, SKL_REVID_F0))) + IS_SKL_GT3(dev) || \ + IS_SKL_GT4(dev)) + /* * dp aux and gmbus irq on gen4 seems to be able to generate legacy interrupts * even when in MSI mode. This results in spurious interrupt warnings if the @@ -2667,7 +2684,7 @@ struct drm_i915_cmd_table { #define HAS_RUNTIME_PM(dev) (IS_GEN6(dev) || IS_HASWELL(dev) || \ IS_BROADWELL(dev) || IS_VALLEYVIEW(dev) || \ IS_CHERRYVIEW(dev) || IS_SKYLAKE(dev) || \ - IS_KABYLAKE(dev)) + IS_KABYLAKE(dev) || IS_BROXTON(dev)) #define HAS_RC6(dev) (INTEL_INFO(dev)->gen >= 6) #define HAS_RC6p(dev) (INTEL_INFO(dev)->gen == 6 || IS_IVYBRIDGE(dev)) @@ -2791,6 +2808,8 @@ void intel_uncore_forcewake_get__locked(struct drm_i915_private *dev_priv, enum forcewake_domains domains); void intel_uncore_forcewake_put__locked(struct drm_i915_private *dev_priv, enum forcewake_domains domains); +u64 intel_uncore_edram_size(struct drm_i915_private *dev_priv); + void assert_forcewakes_inactive(struct drm_i915_private *dev_priv); static inline bool intel_vgpu_active(struct drm_device *dev) { @@ -2869,7 +2888,6 @@ int i915_gem_sw_finish_ioctl(struct drm_device *dev, void *data, struct drm_file *file_priv); void i915_gem_execbuffer_move_to_active(struct list_head *vmas, struct drm_i915_gem_request *req); -void i915_gem_execbuffer_retire_commands(struct i915_execbuffer_params *params); int i915_gem_ringbuffer_submission(struct i915_execbuffer_params *params, struct drm_i915_gem_execbuffer2 *args, struct list_head *vmas); @@ -3000,9 +3018,11 @@ static inline void i915_gem_object_unpin_pages(struct drm_i915_gem_object *obj) * pages and then returns a contiguous mapping of the backing storage into * the kernel address space. * - * The caller must hold the struct_mutex. + * The caller must hold the struct_mutex, and is responsible for calling + * i915_gem_object_unpin_map() when the mapping is no longer required. * - * Returns the pointer through which to access the backing storage. + * Returns the pointer through which to access the mapped object, or an + * ERR_PTR() on error. */ void *__must_check i915_gem_object_pin_map(struct drm_i915_gem_object *obj); @@ -3069,23 +3089,45 @@ i915_gem_find_active_request(struct intel_engine_cs *engine); bool i915_gem_retire_requests(struct drm_device *dev); void i915_gem_retire_requests_ring(struct intel_engine_cs *engine); -int __must_check i915_gem_check_wedge(struct i915_gpu_error *error, - bool interruptible); + +static inline u32 i915_reset_counter(struct i915_gpu_error *error) +{ + return atomic_read(&error->reset_counter); +} + +static inline bool __i915_reset_in_progress(u32 reset) +{ + return unlikely(reset & I915_RESET_IN_PROGRESS_FLAG); +} + +static inline bool __i915_reset_in_progress_or_wedged(u32 reset) +{ + return unlikely(reset & (I915_RESET_IN_PROGRESS_FLAG | I915_WEDGED)); +} + +static inline bool __i915_terminally_wedged(u32 reset) +{ + return unlikely(reset & I915_WEDGED); +} static inline bool i915_reset_in_progress(struct i915_gpu_error *error) { - return unlikely(atomic_read(&error->reset_counter) - & (I915_RESET_IN_PROGRESS_FLAG | I915_WEDGED)); + return __i915_reset_in_progress(i915_reset_counter(error)); +} + +static inline bool i915_reset_in_progress_or_wedged(struct i915_gpu_error *error) +{ + return __i915_reset_in_progress_or_wedged(i915_reset_counter(error)); } static inline bool i915_terminally_wedged(struct i915_gpu_error *error) { - return atomic_read(&error->reset_counter) & I915_WEDGED; + return __i915_terminally_wedged(i915_reset_counter(error)); } static inline u32 i915_reset_count(struct i915_gpu_error *error) { - return ((atomic_read(&error->reset_counter) & ~I915_WEDGED) + 1) / 2; + return ((i915_reset_counter(error) & ~I915_WEDGED) + 1) / 2; } static inline bool i915_stop_ring_allow_ban(struct drm_i915_private *dev_priv) @@ -3118,7 +3160,6 @@ void __i915_add_request(struct drm_i915_gem_request *req, #define i915_add_request_no_flush(req) \ __i915_add_request(req, NULL, false) int __i915_wait_request(struct drm_i915_gem_request *req, - unsigned reset_counter, bool interruptible, s64 *timeout, struct intel_rps_client *rps); @@ -3455,6 +3496,7 @@ extern int intel_opregion_notify_encoder(struct intel_encoder *intel_encoder, bool enable); extern int intel_opregion_notify_adapter(struct drm_device *dev, pci_power_t state); +extern int intel_opregion_get_panel_type(struct drm_device *dev); #else static inline int intel_opregion_setup(struct drm_device *dev) { return 0; } static inline void intel_opregion_init(struct drm_device *dev) { return; } @@ -3470,6 +3512,10 @@ intel_opregion_notify_adapter(struct drm_device *dev, pci_power_t state) { return 0; } +static inline int intel_opregion_get_panel_type(struct drm_device *dev) +{ + return -ENODEV; +} #endif /* intel_acpi.c */ diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c index f4abf3a..261a3ef 100644 --- a/drivers/gpu/drm/i915/i915_gem.c +++ b/drivers/gpu/drm/i915/i915_gem.c @@ -32,14 +32,13 @@ #include "i915_vgpu.h" #include "i915_trace.h" #include "intel_drv.h" +#include "intel_mocs.h" #include <linux/shmem_fs.h> #include <linux/slab.h> #include <linux/swap.h> #include <linux/pci.h> #include <linux/dma-buf.h> -#define RQ_BUG_ON(expr) - static void i915_gem_object_flush_gtt_write_domain(struct drm_i915_gem_object *obj); static void i915_gem_object_flush_cpu_write_domain(struct drm_i915_gem_object *obj); static void @@ -85,9 +84,7 @@ i915_gem_wait_for_error(struct i915_gpu_error *error) { int ret; -#define EXIT_COND (!i915_reset_in_progress(error) || \ - i915_terminally_wedged(error)) - if (EXIT_COND) + if (!i915_reset_in_progress(error)) return 0; /* @@ -96,17 +93,16 @@ i915_gem_wait_for_error(struct i915_gpu_error *error) * we should simply try to bail out and fail as gracefully as possible. */ ret = wait_event_interruptible_timeout(error->reset_queue, - EXIT_COND, + !i915_reset_in_progress(error), 10*HZ); if (ret == 0) { DRM_ERROR("Timed out waiting for the gpu reset to complete\n"); return -EIO; } else if (ret < 0) { return ret; + } else { + return 0; } -#undef EXIT_COND - - return 0; } int i915_mutex_lock_interruptible(struct drm_device *dev) @@ -211,11 +207,10 @@ i915_gem_object_put_pages_phys(struct drm_i915_gem_object *obj) BUG_ON(obj->madv == __I915_MADV_PURGED); ret = i915_gem_object_set_to_cpu_domain(obj, true); - if (ret) { + if (WARN_ON(ret)) { /* In the event of a disaster, abandon all caches and * hope for the best. */ - WARN_ON(ret != -EIO); obj->base.read_domains = obj->base.write_domain = I915_GEM_DOMAIN_CPU; } @@ -1110,27 +1105,19 @@ put_rpm: return ret; } -int -i915_gem_check_wedge(struct i915_gpu_error *error, - bool interruptible) +static int +i915_gem_check_wedge(unsigned reset_counter, bool interruptible) { - if (i915_reset_in_progress(error)) { + if (__i915_terminally_wedged(reset_counter)) + return -EIO; + + if (__i915_reset_in_progress(reset_counter)) { /* Non-interruptible callers can't handle -EAGAIN, hence return * -EIO unconditionally for these. */ if (!interruptible) return -EIO; - /* Recovery complete, but the reset failed ... */ - if (i915_terminally_wedged(error)) - return -EIO; - - /* - * Check if GPU Reset is in progress - we need intel_ring_begin - * to work properly to reinit the hw state while the gpu is - * still marked as reset-in-progress. Handle this with a flag. - */ - if (!error->reload_in_reset) - return -EAGAIN; + return -EAGAIN; } return 0; @@ -1224,7 +1211,6 @@ static int __i915_spin_request(struct drm_i915_gem_request *req, int state) /** * __i915_wait_request - wait until execution of request has finished * @req: duh! - * @reset_counter: reset sequence associated with the given request * @interruptible: do an interruptible wait (normally yes) * @timeout: in - how long to wait (NULL forever); out - how much time remaining * @@ -1239,7 +1225,6 @@ static int __i915_spin_request(struct drm_i915_gem_request *req, int state) * errno with remaining time filled in timeout argument. */ int __i915_wait_request(struct drm_i915_gem_request *req, - unsigned reset_counter, bool interruptible, s64 *timeout, struct intel_rps_client *rps) @@ -1300,13 +1285,14 @@ int __i915_wait_request(struct drm_i915_gem_request *req, prepare_to_wait(&engine->irq_queue, &wait, state); /* We need to check whether any gpu reset happened in between - * the caller grabbing the seqno and now ... */ - if (reset_counter != atomic_read(&dev_priv->gpu_error.reset_counter)) { - /* ... but upgrade the -EAGAIN to an -EIO if the gpu - * is truely gone. */ - ret = i915_gem_check_wedge(&dev_priv->gpu_error, interruptible); - if (ret == 0) - ret = -EAGAIN; + * the request being submitted and now. If a reset has occurred, + * the request is effectively complete (we either are in the + * process of or have discarded the rendering and completely + * reset the GPU. The results of the request are lost and we + * are free to continue on with the original operation. + */ + if (req->reset_counter != i915_reset_counter(&dev_priv->gpu_error)) { + ret = 0; break; } @@ -1458,26 +1444,15 @@ __i915_gem_request_retire__upto(struct drm_i915_gem_request *req) int i915_wait_request(struct drm_i915_gem_request *req) { - struct drm_device *dev; - struct drm_i915_private *dev_priv; + struct drm_i915_private *dev_priv = req->i915; bool interruptible; int ret; - BUG_ON(req == NULL); - - dev = req->engine->dev; - dev_priv = dev->dev_private; interruptible = dev_priv->mm.interruptible; - BUG_ON(!mutex_is_locked(&dev->struct_mutex)); - - ret = i915_gem_check_wedge(&dev_priv->gpu_error, interruptible); - if (ret) - return ret; + BUG_ON(!mutex_is_locked(&dev_priv->dev->struct_mutex)); - ret = __i915_wait_request(req, - atomic_read(&dev_priv->gpu_error.reset_counter), - interruptible, NULL, NULL); + ret = __i915_wait_request(req, interruptible, NULL, NULL); if (ret) return ret; @@ -1521,7 +1496,7 @@ i915_gem_object_wait_rendering(struct drm_i915_gem_object *obj, i915_gem_object_retire__read(obj, i); } - RQ_BUG_ON(obj->active); + GEM_BUG_ON(obj->active); } return 0; @@ -1552,7 +1527,6 @@ i915_gem_object_wait_rendering__nonblocking(struct drm_i915_gem_object *obj, struct drm_device *dev = obj->base.dev; struct drm_i915_private *dev_priv = dev->dev_private; struct drm_i915_gem_request *requests[I915_NUM_ENGINES]; - unsigned reset_counter; int ret, i, n = 0; BUG_ON(!mutex_is_locked(&dev->struct_mutex)); @@ -1561,12 +1535,6 @@ i915_gem_object_wait_rendering__nonblocking(struct drm_i915_gem_object *obj, if (!obj->active) return 0; - ret = i915_gem_check_wedge(&dev_priv->gpu_error, true); - if (ret) - return ret; - - reset_counter = atomic_read(&dev_priv->gpu_error.reset_counter); - if (readonly) { struct drm_i915_gem_request *req; @@ -1588,9 +1556,9 @@ i915_gem_object_wait_rendering__nonblocking(struct drm_i915_gem_object *obj, } mutex_unlock(&dev->struct_mutex); + ret = 0; for (i = 0; ret == 0 && i < n; i++) - ret = __i915_wait_request(requests[i], reset_counter, true, - NULL, rps); + ret = __i915_wait_request(requests[i], true, NULL, rps); mutex_lock(&dev->struct_mutex); for (i = 0; i < n; i++) { @@ -1964,11 +1932,27 @@ out: void i915_gem_release_mmap(struct drm_i915_gem_object *obj) { + /* Serialisation between user GTT access and our code depends upon + * revoking the CPU's PTE whilst the mutex is held. The next user + * pagefault then has to wait until we release the mutex. + */ + lockdep_assert_held(&obj->base.dev->struct_mutex); + if (!obj->fault_mappable) return; drm_vma_node_unmap(&obj->base.vma_node, obj->base.dev->anon_inode->i_mapping); + + /* Ensure that the CPU's PTE are revoked and there are not outstanding + * memory transactions from userspace before we return. The TLB + * flushing implied above by changing the PTE above *should* be + * sufficient, an extra barrier here just provides us with a bit + * of paranoid documentation about our requirement to serialise + * memory writes before touching registers / GSM. + */ + wmb(); + obj->fault_mappable = false; } @@ -2177,11 +2161,10 @@ i915_gem_object_put_pages_gtt(struct drm_i915_gem_object *obj) BUG_ON(obj->madv == __I915_MADV_PURGED); ret = i915_gem_object_set_to_cpu_domain(obj, true); - if (ret) { + if (WARN_ON(ret)) { /* In the event of a disaster, abandon all caches and * hope for the best. */ - WARN_ON(ret != -EIO); i915_gem_clflush_object(obj, true); obj->base.read_domains = obj->base.write_domain = I915_GEM_DOMAIN_CPU; } @@ -2470,8 +2453,8 @@ void i915_vma_move_to_active(struct i915_vma *vma, static void i915_gem_object_retire__write(struct drm_i915_gem_object *obj) { - RQ_BUG_ON(obj->last_write_req == NULL); - RQ_BUG_ON(!(obj->active & intel_engine_flag(obj->last_write_req->engine))); + GEM_BUG_ON(obj->last_write_req == NULL); + GEM_BUG_ON(!(obj->active & intel_engine_flag(obj->last_write_req->engine))); i915_gem_request_assign(&obj->last_write_req, NULL); intel_fb_obj_flush(obj, true, ORIGIN_CS); @@ -2482,8 +2465,8 @@ i915_gem_object_retire__read(struct drm_i915_gem_object *obj, int ring) { struct i915_vma *vma; - RQ_BUG_ON(obj->last_read_req[ring] == NULL); - RQ_BUG_ON(!(obj->active & (1 << ring))); + GEM_BUG_ON(obj->last_read_req[ring] == NULL); + GEM_BUG_ON(!(obj->active & (1 << ring))); list_del_init(&obj->engine_list[ring]); i915_gem_request_assign(&obj->last_read_req[ring], NULL); @@ -2743,6 +2726,7 @@ __i915_gem_request_alloc(struct intel_engine_cs *engine, struct drm_i915_gem_request **req_out) { struct drm_i915_private *dev_priv = to_i915(engine->dev); + unsigned reset_counter = i915_reset_counter(&dev_priv->gpu_error); struct drm_i915_gem_request *req; int ret; @@ -2751,6 +2735,14 @@ __i915_gem_request_alloc(struct intel_engine_cs *engine, *req_out = NULL; + /* ABI: Before userspace accesses the GPU (e.g. execbuffer), report + * EIO if the GPU is already wedged, or EAGAIN to drop the struct_mutex + * and restart. + */ + ret = i915_gem_check_wedge(reset_counter, dev_priv->mm.interruptible); + if (ret) + return ret; + req = kmem_cache_zalloc(dev_priv->requests, GFP_KERNEL); if (req == NULL) return -ENOMEM; @@ -2762,6 +2754,7 @@ __i915_gem_request_alloc(struct intel_engine_cs *engine, kref_init(&req->ref); req->i915 = dev_priv; req->engine = engine; + req->reset_counter = reset_counter; req->ctx = ctx; i915_gem_context_reference(req->ctx); @@ -2791,7 +2784,8 @@ __i915_gem_request_alloc(struct intel_engine_cs *engine, * fully prepared. Thus it can be cleaned up using the proper * free code. */ - i915_gem_request_cancel(req); + intel_ring_reserved_space_cancel(req->ringbuf); + i915_gem_request_unreference(req); return ret; } @@ -2828,13 +2822,6 @@ i915_gem_request_alloc(struct intel_engine_cs *engine, return err ? ERR_PTR(err) : req; } -void i915_gem_request_cancel(struct drm_i915_gem_request *req) -{ - intel_ring_reserved_space_cancel(req->ringbuf); - - i915_gem_request_unreference(req); -} - struct drm_i915_gem_request * i915_gem_find_active_request(struct intel_engine_cs *engine) { @@ -3140,11 +3127,9 @@ retire: int i915_gem_wait_ioctl(struct drm_device *dev, void *data, struct drm_file *file) { - struct drm_i915_private *dev_priv = dev->dev_private; struct drm_i915_gem_wait *args = data; struct drm_i915_gem_object *obj; struct drm_i915_gem_request *req[I915_NUM_ENGINES]; - unsigned reset_counter; int i, n = 0; int ret; @@ -3178,7 +3163,6 @@ i915_gem_wait_ioctl(struct drm_device *dev, void *data, struct drm_file *file) } drm_gem_object_unreference(&obj->base); - reset_counter = atomic_read(&dev_priv->gpu_error.reset_counter); for (i = 0; i < I915_NUM_ENGINES; i++) { if (obj->last_read_req[i] == NULL) @@ -3191,7 +3175,7 @@ i915_gem_wait_ioctl(struct drm_device *dev, void *data, struct drm_file *file) for (i = 0; i < n; i++) { if (ret == 0) - ret = __i915_wait_request(req[i], reset_counter, true, + ret = __i915_wait_request(req[i], true, args->timeout_ns > 0 ? &args->timeout_ns : NULL, to_rps_client(file)); i915_gem_request_unreference__unlocked(req[i]); @@ -3223,7 +3207,6 @@ __i915_gem_object_sync(struct drm_i915_gem_object *obj, if (!i915_semaphore_is_enabled(obj->base.dev)) { struct drm_i915_private *i915 = to_i915(obj->base.dev); ret = __i915_wait_request(from_req, - atomic_read(&i915->gpu_error.reset_counter), i915->mm.interruptible, NULL, &i915->rps.semaphores); @@ -3344,9 +3327,6 @@ static void i915_gem_object_finish_gtt(struct drm_i915_gem_object *obj) if ((obj->base.read_domains & I915_GEM_DOMAIN_GTT) == 0) return; - /* Wait for any direct GTT access to complete */ - mb(); - old_read_domains = obj->base.read_domains; old_write_domain = obj->base.write_domain; @@ -3451,12 +3431,9 @@ int i915_gpu_idle(struct drm_device *dev) return PTR_ERR(req); ret = i915_switch_context(req); - if (ret) { - i915_gem_request_cancel(req); - return ret; - } - i915_add_request_no_flush(req); + if (ret) + return ret; } ret = intel_engine_idle(engine); @@ -4179,16 +4156,15 @@ i915_gem_ring_throttle(struct drm_device *dev, struct drm_file *file) struct drm_i915_file_private *file_priv = file->driver_priv; unsigned long recent_enough = jiffies - DRM_I915_THROTTLE_JIFFIES; struct drm_i915_gem_request *request, *target = NULL; - unsigned reset_counter; int ret; ret = i915_gem_wait_for_error(&dev_priv->gpu_error); if (ret) return ret; - ret = i915_gem_check_wedge(&dev_priv->gpu_error, false); - if (ret) - return ret; + /* ABI: return -EIO if already wedged */ + if (i915_terminally_wedged(&dev_priv->gpu_error)) + return -EIO; spin_lock(&file_priv->mm.lock); list_for_each_entry(request, &file_priv->mm.request_list, client_list) { @@ -4204,7 +4180,6 @@ i915_gem_ring_throttle(struct drm_device *dev, struct drm_file *file) target = request; } - reset_counter = atomic_read(&dev_priv->gpu_error.reset_counter); if (target) i915_gem_request_reference(target); spin_unlock(&file_priv->mm.lock); @@ -4212,7 +4187,7 @@ i915_gem_ring_throttle(struct drm_device *dev, struct drm_file *file) if (target == NULL) return 0; - ret = __i915_wait_request(target, reset_counter, true, NULL, NULL); + ret = __i915_wait_request(target, true, NULL, NULL); if (ret == 0) queue_delayed_work(dev_priv->wq, &dev_priv->mm.retire_work, 0); @@ -4372,7 +4347,6 @@ i915_gem_object_ggtt_unpin_view(struct drm_i915_gem_object *obj, { struct i915_vma *vma = i915_gem_obj_to_ggtt_view(obj, view); - BUG_ON(!vma); WARN_ON(vma->pin_count == 0); WARN_ON(!i915_gem_obj_ggtt_bound_view(obj, view)); @@ -4889,7 +4863,7 @@ i915_gem_init_hw(struct drm_device *dev) /* Double layer security blanket, see i915_gem_init() */ intel_uncore_forcewake_get(dev_priv, FORCEWAKE_ALL); - if (dev_priv->ellc_size) + if (HAS_EDRAM(dev) && INTEL_GEN(dev_priv) < 9) I915_WRITE(HSW_IDICR, I915_READ(HSW_IDICR) | IDIHASHMSK(0xf)); if (IS_HASWELL(dev)) @@ -4933,6 +4907,8 @@ i915_gem_init_hw(struct drm_device *dev) goto out; } + intel_mocs_init_l3cc_table(dev); + /* We can't enable contexts until all firmware is loaded */ if (HAS_GUC_UCODE(dev)) { ret = intel_guc_ucode_load(dev); @@ -4958,34 +4934,33 @@ i915_gem_init_hw(struct drm_device *dev) req = i915_gem_request_alloc(engine, NULL); if (IS_ERR(req)) { ret = PTR_ERR(req); - i915_gem_cleanup_engines(dev); - goto out; + break; } if (engine->id == RCS) { - for (j = 0; j < NUM_L3_SLICES(dev); j++) - i915_gem_l3_remap(req, j); + for (j = 0; j < NUM_L3_SLICES(dev); j++) { + ret = i915_gem_l3_remap(req, j); + if (ret) + goto err_request; + } } ret = i915_ppgtt_init_ring(req); - if (ret && ret != -EIO) { - DRM_ERROR("PPGTT enable %s failed %d\n", - engine->name, ret); - i915_gem_request_cancel(req); - i915_gem_cleanup_engines(dev); - goto out; - } + if (ret) + goto err_request; ret = i915_gem_context_enable(req); - if (ret && ret != -EIO) { - DRM_ERROR("Context enable %s failed %d\n", + if (ret) + goto err_request; + +err_request: + i915_add_request_no_flush(req); + if (ret) { + DRM_ERROR("Failed to enable %s, error=%d\n", engine->name, ret); - i915_gem_request_cancel(req); i915_gem_cleanup_engines(dev); - goto out; + break; } - - i915_add_request_no_flush(req); } out: diff --git a/drivers/gpu/drm/i915/i915_gem.h b/drivers/gpu/drm/i915/i915_gem.h new file mode 100644 index 0000000..8292e79 --- /dev/null +++ b/drivers/gpu/drm/i915/i915_gem.h @@ -0,0 +1,34 @@ +/* + * Copyright © 2016 Intel Corporation + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice (including the next + * paragraph) shall be included in all copies or substantial portions of the + * Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + * + */ + +#ifndef __I915_GEM_H__ +#define __I915_GEM_H__ + +#ifdef CONFIG_DRM_I915_DEBUG_GEM +#define GEM_BUG_ON(expr) BUG_ON(expr) +#else +#define GEM_BUG_ON(expr) +#endif + +#endif /* __I915_GEM_H__ */ diff --git a/drivers/gpu/drm/i915/i915_gem_context.c b/drivers/gpu/drm/i915/i915_gem_context.c index fe580cb..e5acc39 100644 --- a/drivers/gpu/drm/i915/i915_gem_context.c +++ b/drivers/gpu/drm/i915/i915_gem_context.c @@ -342,7 +342,7 @@ void i915_gem_context_reset(struct drm_device *dev) struct intel_context *ctx; list_for_each_entry(ctx, &dev_priv->context_list, link) - intel_lr_context_reset(dev, ctx); + intel_lr_context_reset(dev_priv, ctx); } for (i = 0; i < I915_NUM_ENGINES; i++) { @@ -539,7 +539,7 @@ mi_set_context(struct drm_i915_gem_request *req, u32 hw_flags) len = 4; if (INTEL_INFO(engine->dev)->gen >= 7) - len += 2 + (num_rings ? 4*num_rings + 2 : 0); + len += 2 + (num_rings ? 4*num_rings + 6 : 0); ret = intel_ring_begin(req, len); if (ret) @@ -579,6 +579,7 @@ mi_set_context(struct drm_i915_gem_request *req, u32 hw_flags) if (INTEL_INFO(engine->dev)->gen >= 7) { if (num_rings) { struct intel_engine_cs *signaller; + i915_reg_t last_reg = {}; /* keep gcc quiet */ intel_ring_emit(engine, MI_LOAD_REGISTER_IMM(num_rings)); @@ -586,11 +587,19 @@ mi_set_context(struct drm_i915_gem_request *req, u32 hw_flags) if (signaller == engine) continue; - intel_ring_emit_reg(engine, - RING_PSMI_CTL(signaller->mmio_base)); + last_reg = RING_PSMI_CTL(signaller->mmio_base); + intel_ring_emit_reg(engine, last_reg); intel_ring_emit(engine, _MASKED_BIT_DISABLE(GEN6_PSMI_SLEEP_MSG_DISABLE)); } + + /* Insert a delay before the next switch! */ + intel_ring_emit(engine, + MI_STORE_REGISTER_MEM | + MI_SRM_LRM_GLOBAL_GTT); + intel_ring_emit_reg(engine, last_reg); + intel_ring_emit(engine, engine->scratch.gtt_offset); + intel_ring_emit(engine, MI_NOOP); } intel_ring_emit(engine, MI_ARB_ON_OFF | MI_ARB_ENABLE); } @@ -600,50 +609,48 @@ mi_set_context(struct drm_i915_gem_request *req, u32 hw_flags) return ret; } -static inline bool should_skip_switch(struct intel_engine_cs *engine, - struct intel_context *from, - struct intel_context *to) +static inline bool skip_rcs_switch(struct intel_engine_cs *engine, + struct intel_context *to) { if (to->remap_slice) return false; - if (to->ppgtt && from == to && + if (!to->legacy_hw_ctx.initialized) + return false; + + if (to->ppgtt && !(intel_engine_flag(engine) & to->ppgtt->pd_dirty_rings)) - return true; + return false; - return false; + return to == engine->last_context; } static bool needs_pd_load_pre(struct intel_engine_cs *engine, struct intel_context *to) { - struct drm_i915_private *dev_priv = engine->dev->dev_private; - if (!to->ppgtt) return false; - if (INTEL_INFO(engine->dev)->gen < 8) + if (engine->last_context == to && + !(intel_engine_flag(engine) & to->ppgtt->pd_dirty_rings)) + return false; + + if (engine->id != RCS) return true; - if (engine != &dev_priv->engine[RCS]) + if (INTEL_INFO(engine->dev)->gen < 8) return true; return false; } static bool -needs_pd_load_post(struct intel_engine_cs *engine, struct intel_context *to, - u32 hw_flags) +needs_pd_load_post(struct intel_context *to, u32 hw_flags) { - struct drm_i915_private *dev_priv = engine->dev->dev_private; - if (!to->ppgtt) return false; - if (!IS_GEN8(engine->dev)) - return false; - - if (engine != &dev_priv->engine[RCS]) + if (!IS_GEN8(to->i915)) return false; if (hw_flags & MI_RESTORE_INHIBIT) @@ -652,60 +659,33 @@ needs_pd_load_post(struct intel_engine_cs *engine, struct intel_context *to, return false; } -static int do_switch(struct drm_i915_gem_request *req) +static int do_rcs_switch(struct drm_i915_gem_request *req) { struct intel_context *to = req->ctx; struct intel_engine_cs *engine = req->engine; - struct drm_i915_private *dev_priv = req->i915; - struct intel_context *from = engine->last_context; - u32 hw_flags = 0; - bool uninitialized = false; + struct intel_context *from; + u32 hw_flags; int ret, i; - if (from != NULL && engine == &dev_priv->engine[RCS]) { - BUG_ON(from->legacy_hw_ctx.rcs_state == NULL); - BUG_ON(!i915_gem_obj_is_pinned(from->legacy_hw_ctx.rcs_state)); - } - - if (should_skip_switch(engine, from, to)) + if (skip_rcs_switch(engine, to)) return 0; /* Trying to pin first makes error handling easier. */ - if (engine == &dev_priv->engine[RCS]) { - ret = i915_gem_obj_ggtt_pin(to->legacy_hw_ctx.rcs_state, - get_context_alignment(engine->dev), - 0); - if (ret) - return ret; - } + ret = i915_gem_obj_ggtt_pin(to->legacy_hw_ctx.rcs_state, + get_context_alignment(engine->dev), + 0); + if (ret) + return ret; /* * Pin can switch back to the default context if we end up calling into * evict_everything - as a last ditch gtt defrag effort that also * switches to the default context. Hence we need to reload from here. + * + * XXX: Doing so is painfully broken! */ from = engine->last_context; - if (needs_pd_load_pre(engine, to)) { - /* Older GENs and non render rings still want the load first, - * "PP_DCLV followed by PP_DIR_BASE register through Load - * Register Immediate commands in Ring Buffer before submitting - * a context."*/ - trace_switch_mm(engine, to); - ret = to->ppgtt->switch_mm(to->ppgtt, req); - if (ret) - goto unpin_out; - - /* Doing a PD load always reloads the page dirs */ - to->ppgtt->pd_dirty_rings &= ~intel_engine_flag(engine); - } - - if (engine != &dev_priv->engine[RCS]) { - if (from) - i915_gem_context_unreference(from); - goto done; - } - /* * Clear this page out of any CPU caches for coherent swap-in/out. Note * that thanks to write = false in this call and us not setting any gpu @@ -718,53 +698,37 @@ static int do_switch(struct drm_i915_gem_request *req) if (ret) goto unpin_out; - if (!to->legacy_hw_ctx.initialized || i915_gem_context_is_default(to)) { - hw_flags |= MI_RESTORE_INHIBIT; + if (needs_pd_load_pre(engine, to)) { + /* Older GENs and non render rings still want the load first, + * "PP_DCLV followed by PP_DIR_BASE register through Load + * Register Immediate commands in Ring Buffer before submitting + * a context."*/ + trace_switch_mm(engine, to); + ret = to->ppgtt->switch_mm(to->ppgtt, req); + if (ret) + goto unpin_out; + } + + if (!to->legacy_hw_ctx.initialized || i915_gem_context_is_default(to)) /* NB: If we inhibit the restore, the context is not allowed to * die because future work may end up depending on valid address * space. This means we must enforce that a page table load * occur when this occurs. */ - } else if (to->ppgtt && - (intel_engine_flag(engine) & to->ppgtt->pd_dirty_rings)) { - hw_flags |= MI_FORCE_RESTORE; - to->ppgtt->pd_dirty_rings &= ~intel_engine_flag(engine); - } + hw_flags = MI_RESTORE_INHIBIT; + else if (to->ppgtt && + intel_engine_flag(engine) & to->ppgtt->pd_dirty_rings) + hw_flags = MI_FORCE_RESTORE; + else + hw_flags = 0; /* We should never emit switch_mm more than once */ WARN_ON(needs_pd_load_pre(engine, to) && - needs_pd_load_post(engine, to, hw_flags)); - - ret = mi_set_context(req, hw_flags); - if (ret) - goto unpin_out; + needs_pd_load_post(to, hw_flags)); - /* GEN8 does *not* require an explicit reload if the PDPs have been - * setup, and we do not wish to move them. - */ - if (needs_pd_load_post(engine, to, hw_flags)) { - trace_switch_mm(engine, to); - ret = to->ppgtt->switch_mm(to->ppgtt, req); - /* The hardware context switch is emitted, but we haven't - * actually changed the state - so it's probably safe to bail - * here. Still, let the user know something dangerous has - * happened. - */ - if (ret) { - DRM_ERROR("Failed to change address space on context switch\n"); - goto unpin_out; - } - } - - for (i = 0; i < MAX_L3_SLICES; i++) { - if (!(to->remap_slice & (1<<i))) - continue; - - ret = i915_gem_l3_remap(req, i); - /* If it failed, try again next round */ + if (to != from || (hw_flags & MI_FORCE_RESTORE)) { + ret = mi_set_context(req, hw_flags); if (ret) - DRM_DEBUG_DRIVER("L3 remapping failed\n"); - else - to->remap_slice &= ~(1<<i); + goto unpin_out; } /* The backing object for the context is done after switching to the @@ -789,27 +753,51 @@ static int do_switch(struct drm_i915_gem_request *req) i915_gem_object_ggtt_unpin(from->legacy_hw_ctx.rcs_state); i915_gem_context_unreference(from); } - - uninitialized = !to->legacy_hw_ctx.initialized; - to->legacy_hw_ctx.initialized = true; - -done: i915_gem_context_reference(to); engine->last_context = to; - if (uninitialized) { + /* GEN8 does *not* require an explicit reload if the PDPs have been + * setup, and we do not wish to move them. + */ + if (needs_pd_load_post(to, hw_flags)) { + trace_switch_mm(engine, to); + ret = to->ppgtt->switch_mm(to->ppgtt, req); + /* The hardware context switch is emitted, but we haven't + * actually changed the state - so it's probably safe to bail + * here. Still, let the user know something dangerous has + * happened. + */ + if (ret) + return ret; + } + + if (to->ppgtt) + to->ppgtt->pd_dirty_rings &= ~intel_engine_flag(engine); + + for (i = 0; i < MAX_L3_SLICES; i++) { + if (!(to->remap_slice & (1<<i))) + continue; + + ret = i915_gem_l3_remap(req, i); + if (ret) + return ret; + + to->remap_slice &= ~(1<<i); + } + + if (!to->legacy_hw_ctx.initialized) { if (engine->init_context) { ret = engine->init_context(req); if (ret) - DRM_ERROR("ring init context: %d\n", ret); + return ret; } + to->legacy_hw_ctx.initialized = true; } return 0; unpin_out: - if (engine->id == RCS) - i915_gem_object_ggtt_unpin(to->legacy_hw_ctx.rcs_state); + i915_gem_object_ggtt_unpin(to->legacy_hw_ctx.rcs_state); return ret; } @@ -834,17 +822,33 @@ int i915_switch_context(struct drm_i915_gem_request *req) WARN_ON(i915.enable_execlists); WARN_ON(!mutex_is_locked(&dev_priv->dev->struct_mutex)); - if (req->ctx->legacy_hw_ctx.rcs_state == NULL) { /* We have the fake context */ - if (req->ctx != engine->last_context) { - i915_gem_context_reference(req->ctx); + if (engine->id != RCS || + req->ctx->legacy_hw_ctx.rcs_state == NULL) { + struct intel_context *to = req->ctx; + + if (needs_pd_load_pre(engine, to)) { + int ret; + + trace_switch_mm(engine, to); + ret = to->ppgtt->switch_mm(to->ppgtt, req); + if (ret) + return ret; + + /* Doing a PD load always reloads the page dirs */ + to->ppgtt->pd_dirty_rings &= ~intel_engine_flag(engine); + } + + if (to != engine->last_context) { + i915_gem_context_reference(to); if (engine->last_context) i915_gem_context_unreference(engine->last_context); - engine->last_context = req->ctx; + engine->last_context = to; } + return 0; } - return do_switch(req); + return do_rcs_switch(req); } static bool contexts_enabled(struct drm_device *dev) diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c index 6ee4f00..6f4f2a6 100644 --- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c +++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c @@ -1137,7 +1137,7 @@ i915_gem_execbuffer_move_to_active(struct list_head *vmas, } } -void +static void i915_gem_execbuffer_retire_commands(struct i915_execbuffer_params *params) { /* Unconditionally force add_request to emit a full flush. */ @@ -1322,7 +1322,6 @@ i915_gem_ringbuffer_submission(struct i915_execbuffer_params *params, trace_i915_gem_ring_dispatch(params->request, params->dispatch_flags); i915_gem_execbuffer_move_to_active(vmas, params->request); - i915_gem_execbuffer_retire_commands(params); return 0; } @@ -1624,7 +1623,7 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data, ret = i915_gem_request_add_to_client(req, file); if (ret) - goto err_batch_unpin; + goto err_request; /* * Save assorted stuff away to pass through to *_submission(). @@ -1641,6 +1640,8 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data, params->request = req; ret = dev_priv->gt.execbuf_submit(params, args, &eb->vmas); +err_request: + i915_gem_execbuffer_retire_commands(params); err_batch_unpin: /* @@ -1657,14 +1658,6 @@ err: i915_gem_context_unreference(ctx); eb_destroy(eb); - /* - * If the request was created but not successfully submitted then it - * must be freed again. If it was submitted then it is being tracked - * on the active request list and no clean up is required here. - */ - if (ret && !IS_ERR_OR_NULL(req)) - i915_gem_request_cancel(req); - mutex_unlock(&dev->struct_mutex); pre_mutex_err: diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c index c5cb049..0d666b3 100644 --- a/drivers/gpu/drm/i915/i915_gem_gtt.c +++ b/drivers/gpu/drm/i915/i915_gem_gtt.c @@ -745,7 +745,7 @@ static void gen8_ppgtt_clear_pte_range(struct i915_address_space *vm, num_entries--; } - kunmap_px(ppgtt, pt); + kunmap_px(ppgtt, pt_vaddr); pte = 0; if (++pde == I915_PDES) { @@ -905,11 +905,10 @@ static int gen8_init_scratch(struct i915_address_space *vm) static int gen8_ppgtt_notify_vgt(struct i915_hw_ppgtt *ppgtt, bool create) { enum vgt_g2v_type msg; - struct drm_device *dev = ppgtt->base.dev; - struct drm_i915_private *dev_priv = dev->dev_private; + struct drm_i915_private *dev_priv = to_i915(ppgtt->base.dev); int i; - if (USES_FULL_48BIT_PPGTT(dev)) { + if (USES_FULL_48BIT_PPGTT(dev_priv)) { u64 daddr = px_dma(&ppgtt->pml4); I915_WRITE(vgtif_reg(pdp[0].lo), lower_32_bits(daddr)); @@ -3172,7 +3171,8 @@ int i915_ggtt_init_hw(struct drm_device *dev) } else if (INTEL_INFO(dev)->gen < 8) { ggtt->probe = gen6_gmch_probe; ggtt->base.cleanup = gen6_gmch_remove; - if (IS_HASWELL(dev) && dev_priv->ellc_size) + + if (HAS_EDRAM(dev)) ggtt->base.pte_encode = iris_pte_encode; else if (IS_HASWELL(dev)) ggtt->base.pte_encode = hsw_pte_encode; diff --git a/drivers/gpu/drm/i915/i915_gem_shrinker.c b/drivers/gpu/drm/i915/i915_gem_shrinker.c index d46388f..425e721 100644 --- a/drivers/gpu/drm/i915/i915_gem_shrinker.c +++ b/drivers/gpu/drm/i915/i915_gem_shrinker.c @@ -70,6 +70,10 @@ static bool swap_available(void) static bool can_release_pages(struct drm_i915_gem_object *obj) { + /* Only shmemfs objects are backed by swap */ + if (!obj->base.filp) + return false; + /* Only report true if by unbinding the object and putting its pages * we can actually make forward progress towards freeing physical * pages. @@ -336,7 +340,7 @@ i915_gem_shrinker_oom(struct notifier_block *nb, unsigned long event, void *ptr) container_of(nb, struct drm_i915_private, mm.oom_notifier); struct shrinker_lock_uninterruptible slu; struct drm_i915_gem_object *obj; - unsigned long pinned, bound, unbound, freed_pages; + unsigned long unevictable, bound, unbound, freed_pages; if (!i915_gem_shrinker_lock_uninterruptible(dev_priv, &slu, 5000)) return NOTIFY_DONE; @@ -347,33 +351,28 @@ i915_gem_shrinker_oom(struct notifier_block *nb, unsigned long event, void *ptr) * assert that there are no objects with pinned pages that are not * being pointed to by hardware. */ - unbound = bound = pinned = 0; + unbound = bound = unevictable = 0; list_for_each_entry(obj, &dev_priv->mm.unbound_list, global_list) { - if (!obj->base.filp) /* not backed by a freeable object */ - continue; - - if (obj->pages_pin_count) - pinned += obj->base.size; + if (!can_release_pages(obj)) + unevictable += obj->base.size >> PAGE_SHIFT; else - unbound += obj->base.size; + unbound += obj->base.size >> PAGE_SHIFT; } list_for_each_entry(obj, &dev_priv->mm.bound_list, global_list) { - if (!obj->base.filp) - continue; - - if (obj->pages_pin_count) - pinned += obj->base.size; + if (!can_release_pages(obj)) + unevictable += obj->base.size >> PAGE_SHIFT; else - bound += obj->base.size; + bound += obj->base.size >> PAGE_SHIFT; } i915_gem_shrinker_unlock_uninterruptible(dev_priv, &slu); if (freed_pages || unbound || bound) - pr_info("Purging GPU memory, %lu bytes freed, %lu bytes still pinned.\n", - freed_pages << PAGE_SHIFT, pinned); + pr_info("Purging GPU memory, %lu pages freed, " + "%lu pages still pinned.\n", + freed_pages, unevictable); if (unbound || bound) - pr_err("%lu and %lu bytes still available in the " + pr_err("%lu and %lu pages still available in the " "bound and unbound GPU page lists.\n", bound, unbound); diff --git a/drivers/gpu/drm/i915/i915_gem_stolen.c b/drivers/gpu/drm/i915/i915_gem_stolen.c index ea06da0..b7ce963 100644 --- a/drivers/gpu/drm/i915/i915_gem_stolen.c +++ b/drivers/gpu/drm/i915/i915_gem_stolen.c @@ -95,9 +95,9 @@ static unsigned long i915_stolen_to_physical(struct drm_device *dev) u32 base; /* Almost universally we can find the Graphics Base of Stolen Memory - * at offset 0x5c in the igfx configuration space. On a few (desktop) - * machines this is also mirrored in the bridge device at different - * locations, or in the MCHBAR. + * at register BSM (0x5c) in the igfx configuration space. On a few + * (desktop) machines this is also mirrored in the bridge device at + * different locations, or in the MCHBAR. * * On 865 we just check the TOUD register. * @@ -107,9 +107,11 @@ static unsigned long i915_stolen_to_physical(struct drm_device *dev) */ base = 0; if (INTEL_INFO(dev)->gen >= 3) { - /* Read Graphics Base of Stolen Memory directly */ - pci_read_config_dword(dev->pdev, 0x5c, &base); - base &= ~((1<<20) - 1); + u32 bsm; + + pci_read_config_dword(dev->pdev, BSM, &bsm); + + base = bsm & BSM_MASK; } else if (IS_I865G(dev)) { u16 toud = 0; diff --git a/drivers/gpu/drm/i915/i915_gem_userptr.c b/drivers/gpu/drm/i915/i915_gem_userptr.c index 0f94b6c..32d9726 100644 --- a/drivers/gpu/drm/i915/i915_gem_userptr.c +++ b/drivers/gpu/drm/i915/i915_gem_userptr.c @@ -34,7 +34,7 @@ struct i915_mm_struct { struct mm_struct *mm; - struct drm_device *dev; + struct drm_i915_private *i915; struct i915_mmu_notifier *mn; struct hlist_node node; struct kref kref; @@ -49,6 +49,7 @@ struct i915_mmu_notifier { struct hlist_node node; struct mmu_notifier mn; struct rb_root objects; + struct workqueue_struct *wq; }; struct i915_mmu_object { @@ -60,6 +61,37 @@ struct i915_mmu_object { bool attached; }; +static void wait_rendering(struct drm_i915_gem_object *obj) +{ + struct drm_device *dev = obj->base.dev; + struct drm_i915_gem_request *requests[I915_NUM_ENGINES]; + int i, n; + + if (!obj->active) + return; + + n = 0; + for (i = 0; i < I915_NUM_ENGINES; i++) { + struct drm_i915_gem_request *req; + + req = obj->last_read_req[i]; + if (req == NULL) + continue; + + requests[n++] = i915_gem_request_reference(req); + } + + mutex_unlock(&dev->struct_mutex); + + for (i = 0; i < n; i++) + __i915_wait_request(requests[i], false, NULL, NULL); + + mutex_lock(&dev->struct_mutex); + + for (i = 0; i < n; i++) + i915_gem_request_unreference(requests[i]); +} + static void cancel_userptr(struct work_struct *work) { struct i915_mmu_object *mo = container_of(work, typeof(*mo), work); @@ -75,13 +107,13 @@ static void cancel_userptr(struct work_struct *work) struct i915_vma *vma, *tmp; bool was_interruptible; + wait_rendering(obj); + was_interruptible = dev_priv->mm.interruptible; dev_priv->mm.interruptible = false; - list_for_each_entry_safe(vma, tmp, &obj->vma_list, obj_link) { - int ret = i915_vma_unbind(vma); - WARN_ON(ret && ret != -EIO); - } + list_for_each_entry_safe(vma, tmp, &obj->vma_list, obj_link) + WARN_ON(i915_vma_unbind(vma)); WARN_ON(i915_gem_object_put_pages(obj)); dev_priv->mm.interruptible = was_interruptible; @@ -140,7 +172,7 @@ static void i915_gem_userptr_mn_invalidate_range_start(struct mmu_notifier *_mn, */ mo = container_of(it, struct i915_mmu_object, it); if (kref_get_unless_zero(&mo->obj->base.refcount)) - schedule_work(&mo->work); + queue_work(mn->wq, &mo->work); list_add(&mo->link, &cancelled); it = interval_tree_iter_next(it, start, end); @@ -148,6 +180,8 @@ static void i915_gem_userptr_mn_invalidate_range_start(struct mmu_notifier *_mn, list_for_each_entry(mo, &cancelled, link) del_object(mo); spin_unlock(&mn->lock); + + flush_workqueue(mn->wq); } static const struct mmu_notifier_ops i915_gem_userptr_notifier = { @@ -167,10 +201,16 @@ i915_mmu_notifier_create(struct mm_struct *mm) spin_lock_init(&mn->lock); mn->mn.ops = &i915_gem_userptr_notifier; mn->objects = RB_ROOT; + mn->wq = alloc_workqueue("i915-userptr-release", WQ_UNBOUND, 0); + if (mn->wq == NULL) { + kfree(mn); + return ERR_PTR(-ENOMEM); + } /* Protected by mmap_sem (write-lock) */ ret = __mmu_notifier_register(&mn->mn, mm); if (ret) { + destroy_workqueue(mn->wq); kfree(mn); return ERR_PTR(ret); } @@ -205,13 +245,13 @@ i915_mmu_notifier_find(struct i915_mm_struct *mm) return mn; down_write(&mm->mm->mmap_sem); - mutex_lock(&to_i915(mm->dev)->mm_lock); + mutex_lock(&mm->i915->mm_lock); if ((mn = mm->mn) == NULL) { mn = i915_mmu_notifier_create(mm->mm); if (!IS_ERR(mn)) mm->mn = mn; } - mutex_unlock(&to_i915(mm->dev)->mm_lock); + mutex_unlock(&mm->i915->mm_lock); up_write(&mm->mm->mmap_sem); return mn; @@ -256,6 +296,7 @@ i915_mmu_notifier_free(struct i915_mmu_notifier *mn, return; mmu_notifier_unregister(&mn->mn, mm); + destroy_workqueue(mn->wq); kfree(mn); } @@ -327,7 +368,7 @@ i915_gem_userptr_init__mm_struct(struct drm_i915_gem_object *obj) } kref_init(&mm->kref); - mm->dev = obj->base.dev; + mm->i915 = to_i915(obj->base.dev); mm->mm = current->mm; atomic_inc(¤t->mm->mm_count); @@ -362,7 +403,7 @@ __i915_mm_struct_free(struct kref *kref) /* Protected by dev_priv->mm_lock */ hash_del(&mm->node); - mutex_unlock(&to_i915(mm->dev)->mm_lock); + mutex_unlock(&mm->i915->mm_lock); INIT_WORK(&mm->work, __i915_mm_struct_free__worker); schedule_work(&mm->work); @@ -498,19 +539,24 @@ __i915_gem_userptr_get_pages_worker(struct work_struct *_work) if (pvec != NULL) { struct mm_struct *mm = obj->userptr.mm->mm; - down_read(&mm->mmap_sem); - while (pinned < npages) { - ret = get_user_pages_remote(work->task, mm, - obj->userptr.ptr + pinned * PAGE_SIZE, - npages - pinned, - !obj->userptr.read_only, 0, - pvec + pinned, NULL); - if (ret < 0) - break; - - pinned += ret; + ret = -EFAULT; + if (atomic_inc_not_zero(&mm->mm_users)) { + down_read(&mm->mmap_sem); + while (pinned < npages) { + ret = get_user_pages_remote + (work->task, mm, + obj->userptr.ptr + pinned * PAGE_SIZE, + npages - pinned, + !obj->userptr.read_only, 0, + pvec + pinned, NULL); + if (ret < 0) + break; + + pinned += ret; + } + up_read(&mm->mmap_sem); + mmput(mm); } - up_read(&mm->mmap_sem); } mutex_lock(&dev->struct_mutex); diff --git a/drivers/gpu/drm/i915/i915_guc_submission.c b/drivers/gpu/drm/i915/i915_guc_submission.c index da86bdb..d40c13f 100644 --- a/drivers/gpu/drm/i915/i915_guc_submission.c +++ b/drivers/gpu/drm/i915/i915_guc_submission.c @@ -179,15 +179,11 @@ static void guc_init_doorbell(struct intel_guc *guc, struct i915_guc_client *client) { struct guc_doorbell_info *doorbell; - void *base; - base = kmap_atomic(i915_gem_object_get_page(client->client_obj, 0)); - doorbell = base + client->doorbell_offset; + doorbell = client->client_base + client->doorbell_offset; - doorbell->db_status = 1; + doorbell->db_status = GUC_DOORBELL_ENABLED; doorbell->cookie = 0; - - kunmap_atomic(base); } static int guc_ring_doorbell(struct i915_guc_client *gc) @@ -195,11 +191,9 @@ static int guc_ring_doorbell(struct i915_guc_client *gc) struct guc_process_desc *desc; union guc_doorbell_qw db_cmp, db_exc, db_ret; union guc_doorbell_qw *db; - void *base; int attempt = 2, ret = -EAGAIN; - base = kmap_atomic(i915_gem_object_get_page(gc->client_obj, 0)); - desc = base + gc->proc_desc_offset; + desc = gc->client_base + gc->proc_desc_offset; /* Update the tail so it is visible to GuC */ desc->tail = gc->wq_tail; @@ -215,7 +209,7 @@ static int guc_ring_doorbell(struct i915_guc_client *gc) db_exc.cookie = 1; /* pointer of current doorbell cacheline */ - db = base + gc->doorbell_offset; + db = gc->client_base + gc->doorbell_offset; while (attempt--) { /* lets ring the doorbell */ @@ -244,10 +238,6 @@ static int guc_ring_doorbell(struct i915_guc_client *gc) db_exc.cookie = 1; } - /* Finally, update the cached copy of the GuC's WQ head */ - gc->wq_head = desc->head; - - kunmap_atomic(base); return ret; } @@ -256,16 +246,12 @@ static void guc_disable_doorbell(struct intel_guc *guc, { struct drm_i915_private *dev_priv = guc_to_i915(guc); struct guc_doorbell_info *doorbell; - void *base; i915_reg_t drbreg = GEN8_DRBREGL(client->doorbell_id); int value; - base = kmap_atomic(i915_gem_object_get_page(client->client_obj, 0)); - doorbell = base + client->doorbell_offset; - - doorbell->db_status = 0; + doorbell = client->client_base + client->doorbell_offset; - kunmap_atomic(base); + doorbell->db_status = GUC_DOORBELL_DISABLED; I915_WRITE(drbreg, I915_READ(drbreg) & ~GEN8_DRB_VALID); @@ -341,10 +327,8 @@ static void guc_init_proc_desc(struct intel_guc *guc, struct i915_guc_client *client) { struct guc_process_desc *desc; - void *base; - base = kmap_atomic(i915_gem_object_get_page(client->client_obj, 0)); - desc = base + client->proc_desc_offset; + desc = client->client_base + client->proc_desc_offset; memset(desc, 0, sizeof(*desc)); @@ -361,8 +345,6 @@ static void guc_init_proc_desc(struct intel_guc *guc, desc->wq_size_bytes = client->wq_size; desc->wq_status = WQ_STATUS_ACTIVE; desc->priority = client->priority; - - kunmap_atomic(base); } /* @@ -376,12 +358,14 @@ static void guc_init_proc_desc(struct intel_guc *guc, static void guc_init_ctx_desc(struct intel_guc *guc, struct i915_guc_client *client) { + struct drm_i915_gem_object *client_obj = client->client_obj; struct drm_i915_private *dev_priv = guc_to_i915(guc); struct intel_engine_cs *engine; struct intel_context *ctx = client->owner; struct guc_context_desc desc; struct sg_table *sg; enum intel_engine_id id; + u32 gfx_addr; memset(&desc, 0, sizeof(desc)); @@ -410,16 +394,17 @@ static void guc_init_ctx_desc(struct intel_guc *guc, lrc->context_desc = (u32)ctx_desc; /* The state page is after PPHWSP */ - lrc->ring_lcra = i915_gem_obj_ggtt_offset(obj) + - LRC_STATE_PN * PAGE_SIZE; + gfx_addr = i915_gem_obj_ggtt_offset(obj); + lrc->ring_lcra = gfx_addr + LRC_STATE_PN * PAGE_SIZE; lrc->context_id = (client->ctx_index << GUC_ELC_CTXID_OFFSET) | (engine->guc_id << GUC_ELC_ENGINE_OFFSET); obj = ctx->engine[id].ringbuf->obj; + gfx_addr = i915_gem_obj_ggtt_offset(obj); - lrc->ring_begin = i915_gem_obj_ggtt_offset(obj); - lrc->ring_end = lrc->ring_begin + obj->base.size - 1; - lrc->ring_next_free_location = lrc->ring_begin; + lrc->ring_begin = gfx_addr; + lrc->ring_end = gfx_addr + obj->base.size - 1; + lrc->ring_next_free_location = gfx_addr; lrc->ring_current_tail_pointer_value = 0; desc.engines_used |= (1 << engine->guc_id); @@ -428,22 +413,17 @@ static void guc_init_ctx_desc(struct intel_guc *guc, WARN_ON(desc.engines_used == 0); /* - * The CPU address is only needed at certain points, so kmap_atomic on - * demand instead of storing it in the ctx descriptor. - * XXX: May make debug easier to have it mapped + * The doorbell, process descriptor, and workqueue are all parts + * of the client object, which the GuC will reference via the GGTT */ - desc.db_trigger_cpu = 0; - desc.db_trigger_uk = client->doorbell_offset + - i915_gem_obj_ggtt_offset(client->client_obj); - desc.db_trigger_phy = client->doorbell_offset + - sg_dma_address(client->client_obj->pages->sgl); - - desc.process_desc = client->proc_desc_offset + - i915_gem_obj_ggtt_offset(client->client_obj); - - desc.wq_addr = client->wq_offset + - i915_gem_obj_ggtt_offset(client->client_obj); - + gfx_addr = i915_gem_obj_ggtt_offset(client_obj); + desc.db_trigger_phy = sg_dma_address(client_obj->pages->sgl) + + client->doorbell_offset; + desc.db_trigger_cpu = (uintptr_t)client->client_base + + client->doorbell_offset; + desc.db_trigger_uk = gfx_addr + client->doorbell_offset; + desc.process_desc = gfx_addr + client->proc_desc_offset; + desc.wq_addr = gfx_addr + client->wq_offset; desc.wq_size = client->wq_size; /* @@ -474,25 +454,16 @@ static void guc_fini_ctx_desc(struct intel_guc *guc, int i915_guc_wq_check_space(struct i915_guc_client *gc) { struct guc_process_desc *desc; - void *base; u32 size = sizeof(struct guc_wq_item); int ret = -ETIMEDOUT, timeout_counter = 200; if (!gc) return 0; - /* Quickly return if wq space is available since last time we cache the - * head position. */ - if (CIRC_SPACE(gc->wq_tail, gc->wq_head, gc->wq_size) >= size) - return 0; - - base = kmap_atomic(i915_gem_object_get_page(gc->client_obj, 0)); - desc = base + gc->proc_desc_offset; + desc = gc->client_base + gc->proc_desc_offset; while (timeout_counter-- > 0) { - gc->wq_head = desc->head; - - if (CIRC_SPACE(gc->wq_tail, gc->wq_head, gc->wq_size) >= size) { + if (CIRC_SPACE(gc->wq_tail, desc->head, gc->wq_size) >= size) { ret = 0; break; } @@ -501,19 +472,19 @@ int i915_guc_wq_check_space(struct i915_guc_client *gc) usleep_range(1000, 2000); }; - kunmap_atomic(base); - return ret; } static int guc_add_workqueue_item(struct i915_guc_client *gc, struct drm_i915_gem_request *rq) { + struct guc_process_desc *desc; struct guc_wq_item *wqi; void *base; u32 tail, wq_len, wq_off, space; - space = CIRC_SPACE(gc->wq_tail, gc->wq_head, gc->wq_size); + desc = gc->client_base + gc->proc_desc_offset; + space = CIRC_SPACE(gc->wq_tail, desc->head, gc->wq_size); if (WARN_ON(space < sizeof(struct guc_wq_item))) return -ENOSPC; /* shouldn't happen */ @@ -661,21 +632,28 @@ static void guc_client_free(struct drm_device *dev, if (!client) return; - if (client->doorbell_id != GUC_INVALID_DOORBELL_ID) { - /* - * First disable the doorbell, then tell the GuC we've - * finished with it, finally deallocate it in our bitmap - */ - guc_disable_doorbell(guc, client); - host2guc_release_doorbell(guc, client); - release_doorbell(guc, client->doorbell_id); - } - /* * XXX: wait for any outstanding submissions before freeing memory. * Be sure to drop any locks */ + if (client->client_base) { + /* + * If we got as far as setting up a doorbell, make sure + * we shut it down before unmapping & deallocating the + * memory. So first disable the doorbell, then tell the + * GuC that we've finished with it, finally deallocate + * it in our bitmap + */ + if (client->doorbell_id != GUC_INVALID_DOORBELL_ID) { + guc_disable_doorbell(guc, client); + host2guc_release_doorbell(guc, client); + release_doorbell(guc, client->doorbell_id); + } + + kunmap(kmap_to_page(client->client_base)); + } + gem_release_guc_obj(client->client_obj); if (client->ctx_index != GUC_INVALID_CTX_ID) { @@ -696,7 +674,7 @@ static void guc_client_free(struct drm_device *dev, * @ctx: the context that owns the client (we use the default render * context) * - * Return: An i915_guc_client object if success. + * Return: An i915_guc_client object if success, else NULL. */ static struct i915_guc_client *guc_client_alloc(struct drm_device *dev, uint32_t priority, @@ -728,7 +706,9 @@ static struct i915_guc_client *guc_client_alloc(struct drm_device *dev, if (!obj) goto err; + /* We'll keep just the first (doorbell/proc) page permanently kmap'd. */ client->client_obj = obj; + client->client_base = kmap(i915_gem_object_get_page(obj, 0)); client->wq_offset = GUC_DB_SIZE; client->wq_size = GUC_WQ_SIZE; diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c index 679f08c..2f6fd33 100644 --- a/drivers/gpu/drm/i915/i915_irq.c +++ b/drivers/gpu/drm/i915/i915_irq.c @@ -1264,18 +1264,17 @@ out: mutex_unlock(&dev_priv->dev->struct_mutex); } -static void ivybridge_parity_error_irq_handler(struct drm_device *dev, u32 iir) +static void ivybridge_parity_error_irq_handler(struct drm_i915_private *dev_priv, + u32 iir) { - struct drm_i915_private *dev_priv = dev->dev_private; - - if (!HAS_L3_DPF(dev)) + if (!HAS_L3_DPF(dev_priv)) return; spin_lock(&dev_priv->irq_lock); - gen5_disable_gt_irq(dev_priv, GT_PARITY_ERROR(dev)); + gen5_disable_gt_irq(dev_priv, GT_PARITY_ERROR(dev_priv)); spin_unlock(&dev_priv->irq_lock); - iir &= GT_PARITY_ERROR(dev); + iir &= GT_PARITY_ERROR(dev_priv); if (iir & GT_RENDER_L3_PARITY_ERROR_INTERRUPT_S1) dev_priv->l3_parity.which_slice |= 1 << 1; @@ -1285,8 +1284,7 @@ static void ivybridge_parity_error_irq_handler(struct drm_device *dev, u32 iir) queue_work(dev_priv->wq, &dev_priv->l3_parity.error_work); } -static void ilk_gt_irq_handler(struct drm_device *dev, - struct drm_i915_private *dev_priv, +static void ilk_gt_irq_handler(struct drm_i915_private *dev_priv, u32 gt_iir) { if (gt_iir & @@ -1296,8 +1294,7 @@ static void ilk_gt_irq_handler(struct drm_device *dev, notify_ring(&dev_priv->engine[VCS]); } -static void snb_gt_irq_handler(struct drm_device *dev, - struct drm_i915_private *dev_priv, +static void snb_gt_irq_handler(struct drm_i915_private *dev_priv, u32 gt_iir) { @@ -1314,8 +1311,8 @@ static void snb_gt_irq_handler(struct drm_device *dev, GT_RENDER_CS_MASTER_ERROR_INTERRUPT)) DRM_DEBUG("Command parser error, gt_iir 0x%08x\n", gt_iir); - if (gt_iir & GT_PARITY_ERROR(dev)) - ivybridge_parity_error_irq_handler(dev, gt_iir); + if (gt_iir & GT_PARITY_ERROR(dev_priv)) + ivybridge_parity_error_irq_handler(dev_priv, gt_iir); } static __always_inline void @@ -1327,60 +1324,45 @@ gen8_cs_irq_handler(struct intel_engine_cs *engine, u32 iir, int test_shift) tasklet_schedule(&engine->irq_tasklet); } -static irqreturn_t gen8_gt_irq_handler(struct drm_i915_private *dev_priv, - u32 master_ctl) +static irqreturn_t gen8_gt_irq_ack(struct drm_i915_private *dev_priv, + u32 master_ctl, + u32 gt_iir[4]) { irqreturn_t ret = IRQ_NONE; if (master_ctl & (GEN8_GT_RCS_IRQ | GEN8_GT_BCS_IRQ)) { - u32 iir = I915_READ_FW(GEN8_GT_IIR(0)); - if (iir) { - I915_WRITE_FW(GEN8_GT_IIR(0), iir); + gt_iir[0] = I915_READ_FW(GEN8_GT_IIR(0)); + if (gt_iir[0]) { + I915_WRITE_FW(GEN8_GT_IIR(0), gt_iir[0]); ret = IRQ_HANDLED; - - gen8_cs_irq_handler(&dev_priv->engine[RCS], - iir, GEN8_RCS_IRQ_SHIFT); - - gen8_cs_irq_handler(&dev_priv->engine[BCS], - iir, GEN8_BCS_IRQ_SHIFT); } else DRM_ERROR("The master control interrupt lied (GT0)!\n"); } if (master_ctl & (GEN8_GT_VCS1_IRQ | GEN8_GT_VCS2_IRQ)) { - u32 iir = I915_READ_FW(GEN8_GT_IIR(1)); - if (iir) { - I915_WRITE_FW(GEN8_GT_IIR(1), iir); + gt_iir[1] = I915_READ_FW(GEN8_GT_IIR(1)); + if (gt_iir[1]) { + I915_WRITE_FW(GEN8_GT_IIR(1), gt_iir[1]); ret = IRQ_HANDLED; - - gen8_cs_irq_handler(&dev_priv->engine[VCS], - iir, GEN8_VCS1_IRQ_SHIFT); - - gen8_cs_irq_handler(&dev_priv->engine[VCS2], - iir, GEN8_VCS2_IRQ_SHIFT); } else DRM_ERROR("The master control interrupt lied (GT1)!\n"); } if (master_ctl & GEN8_GT_VECS_IRQ) { - u32 iir = I915_READ_FW(GEN8_GT_IIR(3)); - if (iir) { - I915_WRITE_FW(GEN8_GT_IIR(3), iir); + gt_iir[3] = I915_READ_FW(GEN8_GT_IIR(3)); + if (gt_iir[3]) { + I915_WRITE_FW(GEN8_GT_IIR(3), gt_iir[3]); ret = IRQ_HANDLED; - - gen8_cs_irq_handler(&dev_priv->engine[VECS], - iir, GEN8_VECS_IRQ_SHIFT); } else DRM_ERROR("The master control interrupt lied (GT3)!\n"); } if (master_ctl & GEN8_GT_PM_IRQ) { - u32 iir = I915_READ_FW(GEN8_GT_IIR(2)); - if (iir & dev_priv->pm_rps_events) { + gt_iir[2] = I915_READ_FW(GEN8_GT_IIR(2)); + if (gt_iir[2] & dev_priv->pm_rps_events) { I915_WRITE_FW(GEN8_GT_IIR(2), - iir & dev_priv->pm_rps_events); + gt_iir[2] & dev_priv->pm_rps_events); ret = IRQ_HANDLED; - gen6_rps_irq_handler(dev_priv, iir); } else DRM_ERROR("The master control interrupt lied (PM)!\n"); } @@ -1388,6 +1370,31 @@ static irqreturn_t gen8_gt_irq_handler(struct drm_i915_private *dev_priv, return ret; } +static void gen8_gt_irq_handler(struct drm_i915_private *dev_priv, + u32 gt_iir[4]) +{ + if (gt_iir[0]) { + gen8_cs_irq_handler(&dev_priv->engine[RCS], + gt_iir[0], GEN8_RCS_IRQ_SHIFT); + gen8_cs_irq_handler(&dev_priv->engine[BCS], + gt_iir[0], GEN8_BCS_IRQ_SHIFT); + } + + if (gt_iir[1]) { + gen8_cs_irq_handler(&dev_priv->engine[VCS], + gt_iir[1], GEN8_VCS1_IRQ_SHIFT); + gen8_cs_irq_handler(&dev_priv->engine[VCS2], + gt_iir[1], GEN8_VCS2_IRQ_SHIFT); + } + + if (gt_iir[3]) + gen8_cs_irq_handler(&dev_priv->engine[VECS], + gt_iir[3], GEN8_VECS_IRQ_SHIFT); + + if (gt_iir[2] & dev_priv->pm_rps_events) + gen6_rps_irq_handler(dev_priv, gt_iir[2]); +} + static bool bxt_port_hotplug_long_detect(enum port port, u32 val) { switch (port) { @@ -1644,10 +1651,10 @@ static bool intel_pipe_handle_vblank(struct drm_device *dev, enum pipe pipe) return true; } -static void valleyview_pipestat_irq_handler(struct drm_device *dev, u32 iir) +static void valleyview_pipestat_irq_ack(struct drm_device *dev, u32 iir, + u32 pipe_stats[I915_MAX_PIPES]) { struct drm_i915_private *dev_priv = dev->dev_private; - u32 pipe_stats[I915_MAX_PIPES] = { }; int pipe; spin_lock(&dev_priv->irq_lock); @@ -1701,6 +1708,13 @@ static void valleyview_pipestat_irq_handler(struct drm_device *dev, u32 iir) I915_WRITE(reg, pipe_stats[pipe]); } spin_unlock(&dev_priv->irq_lock); +} + +static void valleyview_pipestat_irq_handler(struct drm_device *dev, + u32 pipe_stats[I915_MAX_PIPES]) +{ + struct drm_i915_private *dev_priv = to_i915(dev); + enum pipe pipe; for_each_pipe(dev_priv, pipe) { if (pipe_stats[pipe] & PIPE_START_VBLANK_INTERRUPT_STATUS && @@ -1723,21 +1737,20 @@ static void valleyview_pipestat_irq_handler(struct drm_device *dev, u32 iir) gmbus_irq_handler(dev); } -static void i9xx_hpd_irq_handler(struct drm_device *dev) +static u32 i9xx_hpd_irq_ack(struct drm_i915_private *dev_priv) { - struct drm_i915_private *dev_priv = dev->dev_private; u32 hotplug_status = I915_READ(PORT_HOTPLUG_STAT); - u32 pin_mask = 0, long_mask = 0; - if (!hotplug_status) - return; + if (hotplug_status) + I915_WRITE(PORT_HOTPLUG_STAT, hotplug_status); - I915_WRITE(PORT_HOTPLUG_STAT, hotplug_status); - /* - * Make sure hotplug status is cleared before we clear IIR, or else we - * may miss hotplug events. - */ - POSTING_READ(PORT_HOTPLUG_STAT); + return hotplug_status; +} + +static void i9xx_hpd_irq_handler(struct drm_device *dev, + u32 hotplug_status) +{ + u32 pin_mask = 0, long_mask = 0; if (IS_G4X(dev) || IS_VALLEYVIEW(dev) || IS_CHERRYVIEW(dev)) { u32 hotplug_trigger = hotplug_status & HOTPLUG_INT_STATUS_G4X; @@ -1768,7 +1781,6 @@ static irqreturn_t valleyview_irq_handler(int irq, void *arg) { struct drm_device *dev = arg; struct drm_i915_private *dev_priv = dev->dev_private; - u32 iir, gt_iir, pm_iir; irqreturn_t ret = IRQ_NONE; if (!intel_irqs_enabled(dev_priv)) @@ -1777,40 +1789,72 @@ static irqreturn_t valleyview_irq_handler(int irq, void *arg) /* IRQs are synced during runtime_suspend, we don't require a wakeref */ disable_rpm_wakeref_asserts(dev_priv); - while (true) { - /* Find, clear, then process each source of interrupt */ + do { + u32 iir, gt_iir, pm_iir; + u32 pipe_stats[I915_MAX_PIPES] = {}; + u32 hotplug_status = 0; + u32 ier = 0; gt_iir = I915_READ(GTIIR); - if (gt_iir) - I915_WRITE(GTIIR, gt_iir); - pm_iir = I915_READ(GEN6_PMIIR); - if (pm_iir) - I915_WRITE(GEN6_PMIIR, pm_iir); - iir = I915_READ(VLV_IIR); - if (iir) { - /* Consume port before clearing IIR or we'll miss events */ - if (iir & I915_DISPLAY_PORT_INTERRUPT) - i9xx_hpd_irq_handler(dev); - I915_WRITE(VLV_IIR, iir); - } if (gt_iir == 0 && pm_iir == 0 && iir == 0) - goto out; + break; ret = IRQ_HANDLED; + /* + * Theory on interrupt generation, based on empirical evidence: + * + * x = ((VLV_IIR & VLV_IER) || + * (((GT_IIR & GT_IER) || (GEN6_PMIIR & GEN6_PMIER)) && + * (VLV_MASTER_IER & MASTER_INTERRUPT_ENABLE))); + * + * A CPU interrupt will only be raised when 'x' has a 0->1 edge. + * Hence we clear MASTER_INTERRUPT_ENABLE and VLV_IER to + * guarantee the CPU interrupt will be raised again even if we + * don't end up clearing all the VLV_IIR, GT_IIR, GEN6_PMIIR + * bits this time around. + */ + I915_WRITE(VLV_MASTER_IER, 0); + ier = I915_READ(VLV_IER); + I915_WRITE(VLV_IER, 0); + if (gt_iir) - snb_gt_irq_handler(dev, dev_priv, gt_iir); + I915_WRITE(GTIIR, gt_iir); if (pm_iir) - gen6_rps_irq_handler(dev_priv, pm_iir); + I915_WRITE(GEN6_PMIIR, pm_iir); + + if (iir & I915_DISPLAY_PORT_INTERRUPT) + hotplug_status = i9xx_hpd_irq_ack(dev_priv); + /* Call regardless, as some status bits might not be * signalled in iir */ - valleyview_pipestat_irq_handler(dev, iir); - } + valleyview_pipestat_irq_ack(dev, iir, pipe_stats); + + /* + * VLV_IIR is single buffered, and reflects the level + * from PIPESTAT/PORT_HOTPLUG_STAT, hence clear it last. + */ + if (iir) + I915_WRITE(VLV_IIR, iir); + + I915_WRITE(VLV_IER, ier); + I915_WRITE(VLV_MASTER_IER, MASTER_INTERRUPT_ENABLE); + POSTING_READ(VLV_MASTER_IER); + + if (gt_iir) + snb_gt_irq_handler(dev_priv, gt_iir); + if (pm_iir) + gen6_rps_irq_handler(dev_priv, pm_iir); + + if (hotplug_status) + i9xx_hpd_irq_handler(dev, hotplug_status); + + valleyview_pipestat_irq_handler(dev, pipe_stats); + } while (0); -out: enable_rpm_wakeref_asserts(dev_priv); return ret; @@ -1820,7 +1864,6 @@ static irqreturn_t cherryview_irq_handler(int irq, void *arg) { struct drm_device *dev = arg; struct drm_i915_private *dev_priv = dev->dev_private; - u32 master_ctl, iir; irqreturn_t ret = IRQ_NONE; if (!intel_irqs_enabled(dev_priv)) @@ -1830,6 +1873,12 @@ static irqreturn_t cherryview_irq_handler(int irq, void *arg) disable_rpm_wakeref_asserts(dev_priv); do { + u32 master_ctl, iir; + u32 gt_iir[4] = {}; + u32 pipe_stats[I915_MAX_PIPES] = {}; + u32 hotplug_status = 0; + u32 ier = 0; + master_ctl = I915_READ(GEN8_MASTER_IRQ) & ~GEN8_MASTER_IRQ_CONTROL; iir = I915_READ(VLV_IIR); @@ -1838,25 +1887,49 @@ static irqreturn_t cherryview_irq_handler(int irq, void *arg) ret = IRQ_HANDLED; + /* + * Theory on interrupt generation, based on empirical evidence: + * + * x = ((VLV_IIR & VLV_IER) || + * ((GEN8_MASTER_IRQ & ~GEN8_MASTER_IRQ_CONTROL) && + * (GEN8_MASTER_IRQ & GEN8_MASTER_IRQ_CONTROL))); + * + * A CPU interrupt will only be raised when 'x' has a 0->1 edge. + * Hence we clear GEN8_MASTER_IRQ_CONTROL and VLV_IER to + * guarantee the CPU interrupt will be raised again even if we + * don't end up clearing all the VLV_IIR and GEN8_MASTER_IRQ_CONTROL + * bits this time around. + */ I915_WRITE(GEN8_MASTER_IRQ, 0); + ier = I915_READ(VLV_IER); + I915_WRITE(VLV_IER, 0); - /* Find, clear, then process each source of interrupt */ + gen8_gt_irq_ack(dev_priv, master_ctl, gt_iir); - if (iir) { - /* Consume port before clearing IIR or we'll miss events */ - if (iir & I915_DISPLAY_PORT_INTERRUPT) - i9xx_hpd_irq_handler(dev); - I915_WRITE(VLV_IIR, iir); - } - - gen8_gt_irq_handler(dev_priv, master_ctl); + if (iir & I915_DISPLAY_PORT_INTERRUPT) + hotplug_status = i9xx_hpd_irq_ack(dev_priv); /* Call regardless, as some status bits might not be * signalled in iir */ - valleyview_pipestat_irq_handler(dev, iir); + valleyview_pipestat_irq_ack(dev, iir, pipe_stats); - I915_WRITE(GEN8_MASTER_IRQ, DE_MASTER_IRQ_CONTROL); + /* + * VLV_IIR is single buffered, and reflects the level + * from PIPESTAT/PORT_HOTPLUG_STAT, hence clear it last. + */ + if (iir) + I915_WRITE(VLV_IIR, iir); + + I915_WRITE(VLV_IER, ier); + I915_WRITE(GEN8_MASTER_IRQ, GEN8_MASTER_IRQ_CONTROL); POSTING_READ(GEN8_MASTER_IRQ); + + gen8_gt_irq_handler(dev_priv, gt_iir); + + if (hotplug_status) + i9xx_hpd_irq_handler(dev, hotplug_status); + + valleyview_pipestat_irq_handler(dev, pipe_stats); } while (0); enable_rpm_wakeref_asserts(dev_priv); @@ -2217,9 +2290,9 @@ static irqreturn_t ironlake_irq_handler(int irq, void *arg) I915_WRITE(GTIIR, gt_iir); ret = IRQ_HANDLED; if (INTEL_INFO(dev)->gen >= 6) - snb_gt_irq_handler(dev, dev_priv, gt_iir); + snb_gt_irq_handler(dev_priv, gt_iir); else - ilk_gt_irq_handler(dev, dev_priv, gt_iir); + ilk_gt_irq_handler(dev_priv, gt_iir); } de_iir = I915_READ(DEIIR); @@ -2419,6 +2492,7 @@ static irqreturn_t gen8_irq_handler(int irq, void *arg) struct drm_device *dev = arg; struct drm_i915_private *dev_priv = dev->dev_private; u32 master_ctl; + u32 gt_iir[4] = {}; irqreturn_t ret; if (!intel_irqs_enabled(dev_priv)) @@ -2435,7 +2509,8 @@ static irqreturn_t gen8_irq_handler(int irq, void *arg) disable_rpm_wakeref_asserts(dev_priv); /* Find, clear, then process each source of interrupt */ - ret = gen8_gt_irq_handler(dev_priv, master_ctl); + ret = gen8_gt_irq_ack(dev_priv, master_ctl, gt_iir); + gen8_gt_irq_handler(dev_priv, gt_iir); ret |= gen8_de_irq_handler(dev_priv, master_ctl); I915_WRITE_FW(GEN8_MASTER_IRQ, GEN8_MASTER_IRQ_CONTROL); @@ -2483,7 +2558,6 @@ static void i915_error_wake_up(struct drm_i915_private *dev_priv, static void i915_reset_and_wakeup(struct drm_device *dev) { struct drm_i915_private *dev_priv = to_i915(dev); - struct i915_gpu_error *error = &dev_priv->gpu_error; char *error_event[] = { I915_ERROR_UEVENT "=1", NULL }; char *reset_event[] = { I915_RESET_UEVENT "=1", NULL }; char *reset_done_event[] = { I915_ERROR_UEVENT "=0", NULL }; @@ -2501,7 +2575,7 @@ static void i915_reset_and_wakeup(struct drm_device *dev) * the reset in-progress bit is only ever set by code outside of this * work we don't need to worry about any other races. */ - if (i915_reset_in_progress(error) && !i915_terminally_wedged(error)) { + if (i915_reset_in_progress(&dev_priv->gpu_error)) { DRM_DEBUG_DRIVER("resetting chip\n"); kobject_uevent_env(&dev->primary->kdev->kobj, KOBJ_CHANGE, reset_event); @@ -2529,25 +2603,9 @@ static void i915_reset_and_wakeup(struct drm_device *dev) intel_runtime_pm_put(dev_priv); - if (ret == 0) { - /* - * After all the gem state is reset, increment the reset - * counter and wake up everyone waiting for the reset to - * complete. - * - * Since unlock operations are a one-sided barrier only, - * we need to insert a barrier here to order any seqno - * updates before - * the counter increment. - */ - smp_mb__before_atomic(); - atomic_inc(&dev_priv->gpu_error.reset_counter); - + if (ret == 0) kobject_uevent_env(&dev->primary->kdev->kobj, KOBJ_CHANGE, reset_done_event); - } else { - atomic_or(I915_WEDGED, &error->reset_counter); - } /* * Note: The wake_up also serves as a memory barrier so that @@ -3285,6 +3343,55 @@ static void gen5_gt_irq_reset(struct drm_device *dev) GEN5_IRQ_RESET(GEN6_PM); } +static void vlv_display_irq_reset(struct drm_i915_private *dev_priv) +{ + enum pipe pipe; + + if (IS_CHERRYVIEW(dev_priv)) + I915_WRITE(DPINVGTT, DPINVGTT_STATUS_MASK_CHV); + else + I915_WRITE(DPINVGTT, DPINVGTT_STATUS_MASK); + + i915_hotplug_interrupt_update_locked(dev_priv, 0xffffffff, 0); + I915_WRITE(PORT_HOTPLUG_STAT, I915_READ(PORT_HOTPLUG_STAT)); + + for_each_pipe(dev_priv, pipe) { + I915_WRITE(PIPESTAT(pipe), + PIPE_FIFO_UNDERRUN_STATUS | + PIPESTAT_INT_STATUS_MASK); + dev_priv->pipestat_irq_mask[pipe] = 0; + } + + GEN5_IRQ_RESET(VLV_); + dev_priv->irq_mask = ~0; +} + +static void vlv_display_irq_postinstall(struct drm_i915_private *dev_priv) +{ + u32 pipestat_mask; + u32 enable_mask; + enum pipe pipe; + + pipestat_mask = PLANE_FLIP_DONE_INT_STATUS_VLV | + PIPE_CRC_DONE_INTERRUPT_STATUS; + + i915_enable_pipestat(dev_priv, PIPE_A, PIPE_GMBUS_INTERRUPT_STATUS); + for_each_pipe(dev_priv, pipe) + i915_enable_pipestat(dev_priv, pipe, pipestat_mask); + + enable_mask = I915_DISPLAY_PORT_INTERRUPT | + I915_DISPLAY_PIPE_A_EVENT_INTERRUPT | + I915_DISPLAY_PIPE_B_EVENT_INTERRUPT; + if (IS_CHERRYVIEW(dev_priv)) + enable_mask |= I915_DISPLAY_PIPE_C_EVENT_INTERRUPT; + + WARN_ON(dev_priv->irq_mask != ~0); + + dev_priv->irq_mask = ~enable_mask; + + GEN5_IRQ_INIT(VLV_, dev_priv->irq_mask, enable_mask); +} + /* drm_dma.h hooks */ static void ironlake_irq_reset(struct drm_device *dev) @@ -3302,34 +3409,19 @@ static void ironlake_irq_reset(struct drm_device *dev) ibx_irq_reset(dev); } -static void vlv_display_irq_reset(struct drm_i915_private *dev_priv) -{ - enum pipe pipe; - - i915_hotplug_interrupt_update(dev_priv, 0xFFFFFFFF, 0); - I915_WRITE(PORT_HOTPLUG_STAT, I915_READ(PORT_HOTPLUG_STAT)); - - for_each_pipe(dev_priv, pipe) - I915_WRITE(PIPESTAT(pipe), 0xffff); - - GEN5_IRQ_RESET(VLV_); -} - static void valleyview_irq_preinstall(struct drm_device *dev) { struct drm_i915_private *dev_priv = dev->dev_private; - /* VLV magic */ - I915_WRITE(VLV_IMR, 0); - I915_WRITE(RING_IMR(RENDER_RING_BASE), 0); - I915_WRITE(RING_IMR(GEN6_BSD_RING_BASE), 0); - I915_WRITE(RING_IMR(BLT_RING_BASE), 0); + I915_WRITE(VLV_MASTER_IER, 0); + POSTING_READ(VLV_MASTER_IER); gen5_gt_irq_reset(dev); - I915_WRITE(DPINVGTT, DPINVGTT_STATUS_MASK); - - vlv_display_irq_reset(dev_priv); + spin_lock_irq(&dev_priv->irq_lock); + if (dev_priv->display_irqs_enabled) + vlv_display_irq_reset(dev_priv); + spin_unlock_irq(&dev_priv->irq_lock); } static void gen8_gt_irq_reset(struct drm_i915_private *dev_priv) @@ -3402,9 +3494,10 @@ static void cherryview_irq_preinstall(struct drm_device *dev) GEN5_IRQ_RESET(GEN8_PCU_); - I915_WRITE(DPINVGTT, DPINVGTT_STATUS_MASK_CHV); - - vlv_display_irq_reset(dev_priv); + spin_lock_irq(&dev_priv->irq_lock); + if (dev_priv->display_irqs_enabled) + vlv_display_irq_reset(dev_priv); + spin_unlock_irq(&dev_priv->irq_lock); } static u32 intel_hpd_enabled_irqs(struct drm_device *dev, @@ -3651,74 +3744,6 @@ static int ironlake_irq_postinstall(struct drm_device *dev) return 0; } -static void valleyview_display_irqs_install(struct drm_i915_private *dev_priv) -{ - u32 pipestat_mask; - u32 iir_mask; - enum pipe pipe; - - pipestat_mask = PIPESTAT_INT_STATUS_MASK | - PIPE_FIFO_UNDERRUN_STATUS; - - for_each_pipe(dev_priv, pipe) - I915_WRITE(PIPESTAT(pipe), pipestat_mask); - POSTING_READ(PIPESTAT(PIPE_A)); - - pipestat_mask = PLANE_FLIP_DONE_INT_STATUS_VLV | - PIPE_CRC_DONE_INTERRUPT_STATUS; - - i915_enable_pipestat(dev_priv, PIPE_A, PIPE_GMBUS_INTERRUPT_STATUS); - for_each_pipe(dev_priv, pipe) - i915_enable_pipestat(dev_priv, pipe, pipestat_mask); - - iir_mask = I915_DISPLAY_PORT_INTERRUPT | - I915_DISPLAY_PIPE_A_EVENT_INTERRUPT | - I915_DISPLAY_PIPE_B_EVENT_INTERRUPT; - if (IS_CHERRYVIEW(dev_priv)) - iir_mask |= I915_DISPLAY_PIPE_C_EVENT_INTERRUPT; - dev_priv->irq_mask &= ~iir_mask; - - I915_WRITE(VLV_IIR, iir_mask); - I915_WRITE(VLV_IIR, iir_mask); - I915_WRITE(VLV_IER, ~dev_priv->irq_mask); - I915_WRITE(VLV_IMR, dev_priv->irq_mask); - POSTING_READ(VLV_IMR); -} - -static void valleyview_display_irqs_uninstall(struct drm_i915_private *dev_priv) -{ - u32 pipestat_mask; - u32 iir_mask; - enum pipe pipe; - - iir_mask = I915_DISPLAY_PORT_INTERRUPT | - I915_DISPLAY_PIPE_A_EVENT_INTERRUPT | - I915_DISPLAY_PIPE_B_EVENT_INTERRUPT; - if (IS_CHERRYVIEW(dev_priv)) - iir_mask |= I915_DISPLAY_PIPE_C_EVENT_INTERRUPT; - - dev_priv->irq_mask |= iir_mask; - I915_WRITE(VLV_IMR, dev_priv->irq_mask); - I915_WRITE(VLV_IER, ~dev_priv->irq_mask); - I915_WRITE(VLV_IIR, iir_mask); - I915_WRITE(VLV_IIR, iir_mask); - POSTING_READ(VLV_IIR); - - pipestat_mask = PLANE_FLIP_DONE_INT_STATUS_VLV | - PIPE_CRC_DONE_INTERRUPT_STATUS; - - i915_disable_pipestat(dev_priv, PIPE_A, PIPE_GMBUS_INTERRUPT_STATUS); - for_each_pipe(dev_priv, pipe) - i915_disable_pipestat(dev_priv, pipe, pipestat_mask); - - pipestat_mask = PIPESTAT_INT_STATUS_MASK | - PIPE_FIFO_UNDERRUN_STATUS; - - for_each_pipe(dev_priv, pipe) - I915_WRITE(PIPESTAT(pipe), pipestat_mask); - POSTING_READ(PIPESTAT(PIPE_A)); -} - void valleyview_enable_display_irqs(struct drm_i915_private *dev_priv) { assert_spin_locked(&dev_priv->irq_lock); @@ -3728,8 +3753,10 @@ void valleyview_enable_display_irqs(struct drm_i915_private *dev_priv) dev_priv->display_irqs_enabled = true; - if (intel_irqs_enabled(dev_priv)) - valleyview_display_irqs_install(dev_priv); + if (intel_irqs_enabled(dev_priv)) { + vlv_display_irq_reset(dev_priv); + vlv_display_irq_postinstall(dev_priv); + } } void valleyview_disable_display_irqs(struct drm_i915_private *dev_priv) @@ -3742,45 +3769,23 @@ void valleyview_disable_display_irqs(struct drm_i915_private *dev_priv) dev_priv->display_irqs_enabled = false; if (intel_irqs_enabled(dev_priv)) - valleyview_display_irqs_uninstall(dev_priv); + vlv_display_irq_reset(dev_priv); } -static void vlv_display_irq_postinstall(struct drm_i915_private *dev_priv) -{ - dev_priv->irq_mask = ~0; - - i915_hotplug_interrupt_update(dev_priv, 0xffffffff, 0); - POSTING_READ(PORT_HOTPLUG_EN); - - I915_WRITE(VLV_IIR, 0xffffffff); - I915_WRITE(VLV_IIR, 0xffffffff); - I915_WRITE(VLV_IER, ~dev_priv->irq_mask); - I915_WRITE(VLV_IMR, dev_priv->irq_mask); - POSTING_READ(VLV_IMR); - - /* Interrupt setup is already guaranteed to be single-threaded, this is - * just to make the assert_spin_locked check happy. */ - spin_lock_irq(&dev_priv->irq_lock); - if (dev_priv->display_irqs_enabled) - valleyview_display_irqs_install(dev_priv); - spin_unlock_irq(&dev_priv->irq_lock); -} static int valleyview_irq_postinstall(struct drm_device *dev) { struct drm_i915_private *dev_priv = dev->dev_private; - vlv_display_irq_postinstall(dev_priv); - gen5_gt_irq_postinstall(dev); - /* ack & enable invalid PTE error interrupts */ -#if 0 /* FIXME: add support to irq handler for checking these bits */ - I915_WRITE(DPINVGTT, DPINVGTT_STATUS_MASK); - I915_WRITE(DPINVGTT, DPINVGTT_EN_MASK); -#endif + spin_lock_irq(&dev_priv->irq_lock); + if (dev_priv->display_irqs_enabled) + vlv_display_irq_postinstall(dev_priv); + spin_unlock_irq(&dev_priv->irq_lock); I915_WRITE(VLV_MASTER_IER, MASTER_INTERRUPT_ENABLE); + POSTING_READ(VLV_MASTER_IER); return 0; } @@ -3791,7 +3796,6 @@ static void gen8_gt_irq_postinstall(struct drm_i915_private *dev_priv) uint32_t gt_interrupts[] = { GT_RENDER_USER_INTERRUPT << GEN8_RCS_IRQ_SHIFT | GT_CONTEXT_SWITCH_INTERRUPT << GEN8_RCS_IRQ_SHIFT | - GT_RENDER_L3_PARITY_ERROR_INTERRUPT | GT_RENDER_USER_INTERRUPT << GEN8_BCS_IRQ_SHIFT | GT_CONTEXT_SWITCH_INTERRUPT << GEN8_BCS_IRQ_SHIFT, GT_RENDER_USER_INTERRUPT << GEN8_VCS1_IRQ_SHIFT | @@ -3803,6 +3807,9 @@ static void gen8_gt_irq_postinstall(struct drm_i915_private *dev_priv) GT_CONTEXT_SWITCH_INTERRUPT << GEN8_VECS_IRQ_SHIFT }; + if (HAS_L3_DPF(dev_priv)) + gt_interrupts[0] |= GT_RENDER_L3_PARITY_ERROR_INTERRUPT; + dev_priv->pm_irq_mask = 0xffffffff; GEN8_IRQ_INIT_NDX(GT, 0, ~gt_interrupts[0], gt_interrupts[0]); GEN8_IRQ_INIT_NDX(GT, 1, ~gt_interrupts[1], gt_interrupts[1]); @@ -3870,7 +3877,7 @@ static int gen8_irq_postinstall(struct drm_device *dev) if (HAS_PCH_SPLIT(dev)) ibx_irq_postinstall(dev); - I915_WRITE(GEN8_MASTER_IRQ, DE_MASTER_IRQ_CONTROL); + I915_WRITE(GEN8_MASTER_IRQ, GEN8_MASTER_IRQ_CONTROL); POSTING_READ(GEN8_MASTER_IRQ); return 0; @@ -3880,11 +3887,14 @@ static int cherryview_irq_postinstall(struct drm_device *dev) { struct drm_i915_private *dev_priv = dev->dev_private; - vlv_display_irq_postinstall(dev_priv); - gen8_gt_irq_postinstall(dev_priv); - I915_WRITE(GEN8_MASTER_IRQ, MASTER_INTERRUPT_ENABLE); + spin_lock_irq(&dev_priv->irq_lock); + if (dev_priv->display_irqs_enabled) + vlv_display_irq_postinstall(dev_priv); + spin_unlock_irq(&dev_priv->irq_lock); + + I915_WRITE(GEN8_MASTER_IRQ, GEN8_MASTER_IRQ_CONTROL); POSTING_READ(GEN8_MASTER_IRQ); return 0; @@ -3900,20 +3910,6 @@ static void gen8_irq_uninstall(struct drm_device *dev) gen8_irq_reset(dev); } -static void vlv_display_irq_uninstall(struct drm_i915_private *dev_priv) -{ - /* Interrupt setup is already guaranteed to be single-threaded, this is - * just to make the assert_spin_locked check happy. */ - spin_lock_irq(&dev_priv->irq_lock); - if (dev_priv->display_irqs_enabled) - valleyview_display_irqs_uninstall(dev_priv); - spin_unlock_irq(&dev_priv->irq_lock); - - vlv_display_irq_reset(dev_priv); - - dev_priv->irq_mask = ~0; -} - static void valleyview_irq_uninstall(struct drm_device *dev) { struct drm_i915_private *dev_priv = dev->dev_private; @@ -3922,12 +3918,16 @@ static void valleyview_irq_uninstall(struct drm_device *dev) return; I915_WRITE(VLV_MASTER_IER, 0); + POSTING_READ(VLV_MASTER_IER); gen5_gt_irq_reset(dev); I915_WRITE(HWSTAM, 0xffffffff); - vlv_display_irq_uninstall(dev_priv); + spin_lock_irq(&dev_priv->irq_lock); + if (dev_priv->display_irqs_enabled) + vlv_display_irq_reset(dev_priv); + spin_unlock_irq(&dev_priv->irq_lock); } static void cherryview_irq_uninstall(struct drm_device *dev) @@ -3944,7 +3944,10 @@ static void cherryview_irq_uninstall(struct drm_device *dev) GEN5_IRQ_RESET(GEN8_PCU_); - vlv_display_irq_uninstall(dev_priv); + spin_lock_irq(&dev_priv->irq_lock); + if (dev_priv->display_irqs_enabled) + vlv_display_irq_reset(dev_priv); + spin_unlock_irq(&dev_priv->irq_lock); } static void ironlake_irq_uninstall(struct drm_device *dev) @@ -4271,8 +4274,11 @@ static irqreturn_t i915_irq_handler(int irq, void *arg) /* Consume port. Then clear IIR or we'll miss events */ if (I915_HAS_HOTPLUG(dev) && - iir & I915_DISPLAY_PORT_INTERRUPT) - i9xx_hpd_irq_handler(dev); + iir & I915_DISPLAY_PORT_INTERRUPT) { + u32 hotplug_status = i9xx_hpd_irq_ack(dev_priv); + if (hotplug_status) + i9xx_hpd_irq_handler(dev, hotplug_status); + } I915_WRITE(IIR, iir & ~flip_mask); new_iir = I915_READ(IIR); /* Flush posted writes */ @@ -4501,8 +4507,11 @@ static irqreturn_t i965_irq_handler(int irq, void *arg) ret = IRQ_HANDLED; /* Consume port. Then clear IIR or we'll miss events */ - if (iir & I915_DISPLAY_PORT_INTERRUPT) - i9xx_hpd_irq_handler(dev); + if (iir & I915_DISPLAY_PORT_INTERRUPT) { + u32 hotplug_status = i9xx_hpd_irq_ack(dev_priv); + if (hotplug_status) + i9xx_hpd_irq_handler(dev, hotplug_status); + } I915_WRITE(IIR, iir & ~flip_mask); new_iir = I915_READ(IIR); /* Flush posted writes */ diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h index cea5a39..58ac6c7 100644 --- a/drivers/gpu/drm/i915/i915_reg.h +++ b/drivers/gpu/drm/i915/i915_reg.h @@ -79,6 +79,16 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg) /* PCI config space */ +#define MCHBAR_I915 0x44 +#define MCHBAR_I965 0x48 +#define MCHBAR_SIZE (4 * 4096) + +#define DEVEN 0x54 +#define DEVEN_MCHBAR_EN (1 << 28) + +#define BSM 0x5c +#define BSM_MASK (0xFFFF << 20) + #define HPLLCC 0xc0 /* 85x only */ #define GC_CLOCK_CONTROL_MASK (0x7 << 0) #define GC_CLOCK_133_200 (0 << 0) @@ -90,6 +100,16 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg) #define GC_CLOCK_166_266 (6 << 0) #define GC_CLOCK_166_250 (7 << 0) +#define I915_GDRST 0xc0 /* PCI config register */ +#define GRDOM_FULL (0 << 2) +#define GRDOM_RENDER (1 << 2) +#define GRDOM_MEDIA (3 << 2) +#define GRDOM_MASK (3 << 2) +#define GRDOM_RESET_STATUS (1 << 1) +#define GRDOM_RESET_ENABLE (1 << 0) + +#define GCDGMBUS 0xcc + #define GCFGC2 0xda #define GCFGC 0xf0 /* 915+ only */ #define GC_LOW_FREQUENCY_ENABLE (1 << 7) @@ -121,18 +141,16 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg) #define I915_GC_RENDER_CLOCK_166_MHZ (0 << 0) #define I915_GC_RENDER_CLOCK_200_MHZ (1 << 0) #define I915_GC_RENDER_CLOCK_333_MHZ (4 << 0) -#define GCDGMBUS 0xcc -#define PCI_LBPC 0xf4 /* legacy/combination backlight modes, also called LBB */ +#define ASLE 0xe4 +#define ASLS 0xfc + +#define SWSCI 0xe8 +#define SWSCI_SCISEL (1 << 15) +#define SWSCI_GSSCIE (1 << 0) + +#define LBPC 0xf4 /* legacy/combination backlight modes, also called LBB */ -/* Graphics reset regs */ -#define I915_GDRST 0xc0 /* PCI config register */ -#define GRDOM_FULL (0<<2) -#define GRDOM_RENDER (1<<2) -#define GRDOM_MEDIA (3<<2) -#define GRDOM_MASK (3<<2) -#define GRDOM_RESET_STATUS (1<<1) -#define GRDOM_RESET_ENABLE (1<<0) #define ILK_GDSR _MMIO(MCHBAR_MIRROR_BASE + 0x2ca4) #define ILK_GRDOM_FULL (0<<1) @@ -1375,14 +1393,10 @@ enum skl_disp_power_wells { #define _PORT_REF_DW6_A 0x162198 #define _PORT_REF_DW6_BC 0x6C198 -/* - * FIXME: BSpec/CHV ConfigDB disagrees on the following two fields, fix them - * after testing. - */ -#define GRC_CODE_SHIFT 23 -#define GRC_CODE_MASK (0x1FF << GRC_CODE_SHIFT) +#define GRC_CODE_SHIFT 24 +#define GRC_CODE_MASK (0xFF << GRC_CODE_SHIFT) #define GRC_CODE_FAST_SHIFT 16 -#define GRC_CODE_FAST_MASK (0x7F << GRC_CODE_FAST_SHIFT) +#define GRC_CODE_FAST_MASK (0xFF << GRC_CODE_FAST_SHIFT) #define GRC_CODE_SLOW_SHIFT 8 #define GRC_CODE_SLOW_MASK (0xFF << GRC_CODE_SLOW_SHIFT) #define GRC_CODE_NOM_MASK 0xFF @@ -2934,7 +2948,14 @@ enum skl_disp_power_wells { #define GEN6_RP_STATE_CAP _MMIO(MCHBAR_MIRROR_BASE_SNB + 0x5998) #define BXT_RP_STATE_CAP _MMIO(0x138170) -#define INTERVAL_1_28_US(us) (((us) * 100) >> 7) +/* + * Make these a multiple of magic 25 to avoid SNB (eg. Dell XPS + * 8300) freezing up around GPU hangs. Looks as if even + * scheduling/timer interrupts start misbehaving if the RPS + * EI/thresholds are "bad", leading to a very sluggish or even + * frozen machine. + */ +#define INTERVAL_1_28_US(us) roundup(((us) * 100) >> 7, 25) #define INTERVAL_1_33_US(us) (((us) * 3) >> 2) #define INTERVAL_0_833_US(us) (((us) * 6) / 5) #define GT_INTERVAL_FROM_US(dev_priv, us) (IS_GEN9(dev_priv) ? \ @@ -2943,6 +2964,15 @@ enum skl_disp_power_wells { INTERVAL_1_33_US(us)) : \ INTERVAL_1_28_US(us)) +#define INTERVAL_1_28_TO_US(interval) (((interval) << 7) / 100) +#define INTERVAL_1_33_TO_US(interval) (((interval) << 2) / 3) +#define INTERVAL_0_833_TO_US(interval) (((interval) * 5) / 6) +#define GT_PM_INTERVAL_TO_US(dev_priv, interval) (IS_GEN9(dev_priv) ? \ + (IS_BROXTON(dev_priv) ? \ + INTERVAL_0_833_TO_US(interval) : \ + INTERVAL_1_33_TO_US(interval)) : \ + INTERVAL_1_28_TO_US(interval)) + /* * Logical Context regs */ @@ -6866,6 +6896,8 @@ enum skl_disp_power_wells { #define VLV_SPAREG2H _MMIO(0xA194) #define GTFIFODBG _MMIO(0x120000) +#define GT_FIFO_SBDEDICATE_FREE_ENTRY_CHV (0x1f << 20) +#define GT_FIFO_FREE_ENTRIES_CHV (0x7f << 13) #define GT_FIFO_SBDROPERR (1<<6) #define GT_FIFO_BLOBDROPERR (1<<5) #define GT_FIFO_SB_READ_ABORTERR (1<<4) @@ -6882,8 +6914,11 @@ enum skl_disp_power_wells { #define HSW_IDICR _MMIO(0x9008) #define IDIHASHMSK(x) (((x) & 0x3f) << 16) -#define HSW_EDRAM_PRESENT _MMIO(0x120010) +#define HSW_EDRAM_CAP _MMIO(0x120010) #define EDRAM_ENABLED 0x1 +#define EDRAM_NUM_BANKS(cap) (((cap) >> 1) & 0xf) +#define EDRAM_WAYS_IDX(cap) (((cap) >> 5) & 0x7) +#define EDRAM_SETS_IDX(cap) (((cap) >> 8) & 0x3) #define GEN6_UCGCTL1 _MMIO(0x9400) # define GEN6_EU_TCUNIT_CLOCK_GATE_DISABLE (1 << 16) @@ -7161,6 +7196,7 @@ enum skl_disp_power_wells { #define GEN9_HALF_SLICE_CHICKEN7 _MMIO(0xe194) #define GEN9_ENABLE_YV12_BUGFIX (1<<4) +#define GEN9_ENABLE_GPGPU_PREEMPTION (1<<2) /* Audio */ #define G4X_AUD_VID_DID _MMIO(dev_priv->info.display_mmio_offset + 0x62020) diff --git a/drivers/gpu/drm/i915/intel_bios.c b/drivers/gpu/drm/i915/intel_bios.c index eb756c4..e72dd9a 100644 --- a/drivers/gpu/drm/i915/intel_bios.c +++ b/drivers/gpu/drm/i915/intel_bios.c @@ -58,8 +58,6 @@ #define SLAVE_ADDR1 0x70 #define SLAVE_ADDR2 0x72 -static int panel_type; - /* Get BDB block size given a pointer to Block ID. */ static u32 _get_blocksize(const u8 *block_base) { @@ -205,17 +203,32 @@ parse_lfp_panel_data(struct drm_i915_private *dev_priv, const struct lvds_dvo_timing *panel_dvo_timing; const struct lvds_fp_timing *fp_timing; struct drm_display_mode *panel_fixed_mode; + int panel_type; int drrs_mode; + int ret; lvds_options = find_section(bdb, BDB_LVDS_OPTIONS); if (!lvds_options) return; dev_priv->vbt.lvds_dither = lvds_options->pixel_dither; - if (lvds_options->panel_type == 0xff) - return; - panel_type = lvds_options->panel_type; + ret = intel_opregion_get_panel_type(dev_priv->dev); + if (ret >= 0) { + WARN_ON(ret > 0xf); + panel_type = ret; + DRM_DEBUG_KMS("Panel type: %d (OpRegion)\n", panel_type); + } else { + if (lvds_options->panel_type > 0xf) { + DRM_DEBUG_KMS("Invalid VBT panel type 0x%x\n", + lvds_options->panel_type); + return; + } + panel_type = lvds_options->panel_type; + DRM_DEBUG_KMS("Panel type: %d (VBT)\n", panel_type); + } + + dev_priv->vbt.panel_type = panel_type; drrs_mode = (lvds_options->dps_panel_type_bits >> (panel_type * 2)) & MODE_MASK; @@ -251,7 +264,7 @@ parse_lfp_panel_data(struct drm_i915_private *dev_priv, panel_dvo_timing = get_lvds_dvo_timing(lvds_lfp_data, lvds_lfp_data_ptrs, - lvds_options->panel_type); + panel_type); panel_fixed_mode = kzalloc(sizeof(*panel_fixed_mode), GFP_KERNEL); if (!panel_fixed_mode) @@ -266,7 +279,7 @@ parse_lfp_panel_data(struct drm_i915_private *dev_priv, fp_timing = get_lvds_fp_timing(bdb, lvds_lfp_data, lvds_lfp_data_ptrs, - lvds_options->panel_type); + panel_type); if (fp_timing) { /* check the resolution, just to be sure */ if (fp_timing->x_res == panel_fixed_mode->hdisplay && @@ -284,6 +297,7 @@ parse_lfp_backlight(struct drm_i915_private *dev_priv, { const struct bdb_lfp_backlight_data *backlight_data; const struct bdb_lfp_backlight_data_entry *entry; + int panel_type = dev_priv->vbt.panel_type; backlight_data = find_section(bdb, BDB_LVDS_BACKLIGHT); if (!backlight_data) @@ -546,6 +560,7 @@ parse_edp(struct drm_i915_private *dev_priv, const struct bdb_header *bdb) const struct bdb_edp *edp; const struct edp_power_seq *edp_pps; const struct edp_link_params *edp_link_params; + int panel_type = dev_priv->vbt.panel_type; edp = find_section(bdb, BDB_EDP); if (!edp) { @@ -657,6 +672,7 @@ parse_psr(struct drm_i915_private *dev_priv, const struct bdb_header *bdb) { const struct bdb_psr *psr; const struct psr_table *psr_table; + int panel_type = dev_priv->vbt.panel_type; psr = find_section(bdb, BDB_PSR); if (!psr) { @@ -703,6 +719,7 @@ parse_mipi_config(struct drm_i915_private *dev_priv, const struct bdb_mipi_config *start; const struct mipi_config *config; const struct mipi_pps_data *pps; + int panel_type = dev_priv->vbt.panel_type; /* parse MIPI blocks only if LFP type is MIPI */ if (!intel_bios_is_dsi_present(dev_priv, NULL)) @@ -910,6 +927,7 @@ static void parse_mipi_sequence(struct drm_i915_private *dev_priv, const struct bdb_header *bdb) { + int panel_type = dev_priv->vbt.panel_type; const struct bdb_mipi_sequence *sequence; const u8 *seq_data; u32 seq_size; diff --git a/drivers/gpu/drm/i915/intel_csr.c b/drivers/gpu/drm/i915/intel_csr.c index 3f57cb9..a34c23e 100644 --- a/drivers/gpu/drm/i915/intel_csr.c +++ b/drivers/gpu/drm/i915/intel_csr.c @@ -50,6 +50,7 @@ MODULE_FIRMWARE(I915_CSR_SKL); MODULE_FIRMWARE(I915_CSR_BXT); #define SKL_CSR_VERSION_REQUIRED CSR_VERSION(1, 23) +#define BXT_CSR_VERSION_REQUIRED CSR_VERSION(1, 7) #define CSR_MAX_FW_SIZE 0x2FFF #define CSR_DEFAULT_FW_OFFSET 0xFFFFFFFF @@ -281,6 +282,7 @@ static uint32_t *parse_csr_fw(struct drm_i915_private *dev_priv, uint32_t dmc_offset = CSR_DEFAULT_FW_OFFSET, readcount = 0, nbytes; uint32_t i; uint32_t *dmc_payload; + uint32_t required_min_version; if (!fw) return NULL; @@ -296,15 +298,23 @@ static uint32_t *parse_csr_fw(struct drm_i915_private *dev_priv, csr->version = css_header->version; - if ((IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv)) && - csr->version < SKL_CSR_VERSION_REQUIRED) { - DRM_INFO("Refusing to load old Skylake DMC firmware v%u.%u," + if (IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv)) { + required_min_version = SKL_CSR_VERSION_REQUIRED; + } else if (IS_BROXTON(dev_priv)) { + required_min_version = BXT_CSR_VERSION_REQUIRED; + } else { + MISSING_CASE(INTEL_REVID(dev_priv)); + required_min_version = 0; + } + + if (csr->version < required_min_version) { + DRM_INFO("Refusing to load old DMC firmware v%u.%u," " please upgrade to v%u.%u or later" " [" FIRMWARE_URL "].\n", CSR_VERSION_MAJOR(csr->version), CSR_VERSION_MINOR(csr->version), - CSR_VERSION_MAJOR(SKL_CSR_VERSION_REQUIRED), - CSR_VERSION_MINOR(SKL_CSR_VERSION_REQUIRED)); + CSR_VERSION_MAJOR(required_min_version), + CSR_VERSION_MINOR(required_min_version)); return NULL; } @@ -457,10 +467,50 @@ void intel_csr_ucode_init(struct drm_i915_private *dev_priv) } /** + * intel_csr_ucode_suspend() - prepare CSR firmware before system suspend + * @dev_priv: i915 drm device + * + * Prepare the DMC firmware before entering system suspend. This includes + * flushing pending work items and releasing any resources acquired during + * init. + */ +void intel_csr_ucode_suspend(struct drm_i915_private *dev_priv) +{ + if (!HAS_CSR(dev_priv)) + return; + + flush_work(&dev_priv->csr.work); + + /* Drop the reference held in case DMC isn't loaded. */ + if (!dev_priv->csr.dmc_payload) + intel_display_power_put(dev_priv, POWER_DOMAIN_INIT); +} + +/** + * intel_csr_ucode_resume() - init CSR firmware during system resume + * @dev_priv: i915 drm device + * + * Reinitialize the DMC firmware during system resume, reacquiring any + * resources released in intel_csr_ucode_suspend(). + */ +void intel_csr_ucode_resume(struct drm_i915_private *dev_priv) +{ + if (!HAS_CSR(dev_priv)) + return; + + /* + * Reacquire the reference to keep RPM disabled in case DMC isn't + * loaded. + */ + if (!dev_priv->csr.dmc_payload) + intel_display_power_get(dev_priv, POWER_DOMAIN_INIT); +} + +/** * intel_csr_ucode_fini() - unload the CSR firmware. * @dev_priv: i915 drm device. * - * Firmmware unloading includes freeing the internal momory and reset the + * Firmmware unloading includes freeing the internal memory and reset the * firmware loading status. */ void intel_csr_ucode_fini(struct drm_i915_private *dev_priv) @@ -468,7 +518,7 @@ void intel_csr_ucode_fini(struct drm_i915_private *dev_priv) if (!HAS_CSR(dev_priv)) return; - flush_work(&dev_priv->csr.work); + intel_csr_ucode_suspend(dev_priv); kfree(dev_priv->csr.dmc_payload); } diff --git a/drivers/gpu/drm/i915/intel_ddi.c b/drivers/gpu/drm/i915/intel_ddi.c index 921edf1..e30e178 100644 --- a/drivers/gpu/drm/i915/intel_ddi.c +++ b/drivers/gpu/drm/i915/intel_ddi.c @@ -443,9 +443,17 @@ void intel_prepare_ddi_buffer(struct intel_encoder *encoder) } else if (IS_BROADWELL(dev_priv)) { ddi_translations_fdi = bdw_ddi_translations_fdi; ddi_translations_dp = bdw_ddi_translations_dp; - ddi_translations_edp = bdw_ddi_translations_edp; + + if (dev_priv->vbt.edp.low_vswing) { + ddi_translations_edp = bdw_ddi_translations_edp; + n_edp_entries = ARRAY_SIZE(bdw_ddi_translations_edp); + } else { + ddi_translations_edp = bdw_ddi_translations_dp; + n_edp_entries = ARRAY_SIZE(bdw_ddi_translations_dp); + } + ddi_translations_hdmi = bdw_ddi_translations_hdmi; - n_edp_entries = ARRAY_SIZE(bdw_ddi_translations_edp); + n_dp_entries = ARRAY_SIZE(bdw_ddi_translations_dp); n_hdmi_entries = ARRAY_SIZE(bdw_ddi_translations_hdmi); hdmi_default_entry = 7; @@ -1722,12 +1730,78 @@ static void intel_disable_ddi(struct intel_encoder *intel_encoder) } } +static bool broxton_phy_is_enabled(struct drm_i915_private *dev_priv, + enum dpio_phy phy) +{ + if (!(I915_READ(BXT_P_CR_GT_DISP_PWRON) & GT_DISPLAY_POWER_ON(phy))) + return false; + + if ((I915_READ(BXT_PORT_CL1CM_DW0(phy)) & + (PHY_POWER_GOOD | PHY_RESERVED)) != PHY_POWER_GOOD) { + DRM_DEBUG_DRIVER("DDI PHY %d powered, but power hasn't settled\n", + phy); + + return false; + } + + if (phy == DPIO_PHY1 && + !(I915_READ(BXT_PORT_REF_DW3(DPIO_PHY1)) & GRC_DONE)) { + DRM_DEBUG_DRIVER("DDI PHY 1 powered, but GRC isn't done\n"); + + return false; + } + + if (!(I915_READ(BXT_PHY_CTL_FAMILY(phy)) & COMMON_RESET_DIS)) { + DRM_DEBUG_DRIVER("DDI PHY %d powered, but still in reset\n", + phy); + + return false; + } + + return true; +} + +static u32 broxton_get_grc(struct drm_i915_private *dev_priv, enum dpio_phy phy) +{ + u32 val = I915_READ(BXT_PORT_REF_DW6(phy)); + + return (val & GRC_CODE_MASK) >> GRC_CODE_SHIFT; +} + +static void broxton_phy_wait_grc_done(struct drm_i915_private *dev_priv, + enum dpio_phy phy) +{ + if (wait_for(I915_READ(BXT_PORT_REF_DW3(phy)) & GRC_DONE, 10)) + DRM_ERROR("timeout waiting for PHY%d GRC\n", phy); +} + +static bool broxton_phy_verify_state(struct drm_i915_private *dev_priv, + enum dpio_phy phy); + static void broxton_phy_init(struct drm_i915_private *dev_priv, enum dpio_phy phy) { enum port port; u32 ports, val; + if (broxton_phy_is_enabled(dev_priv, phy)) { + /* Still read out the GRC value for state verification */ + if (phy == DPIO_PHY0) + dev_priv->bxt_phy_grc = broxton_get_grc(dev_priv, phy); + + if (broxton_phy_verify_state(dev_priv, phy)) { + DRM_DEBUG_DRIVER("DDI PHY %d already enabled, " + "won't reprogram it\n", phy); + + return; + } + + DRM_DEBUG_DRIVER("DDI PHY %d enabled with invalid state, " + "force reprogramming it\n", phy); + } else { + DRM_DEBUG_DRIVER("DDI PHY %d not enabled, enabling it\n", phy); + } + val = I915_READ(BXT_P_CR_GT_DISP_PWRON); val |= GT_DISPLAY_POWER_ON(phy); I915_WRITE(BXT_P_CR_GT_DISP_PWRON, val); @@ -1798,6 +1872,9 @@ static void broxton_phy_init(struct drm_i915_private *dev_priv, * enabled. * TODO: port C is only connected on BXT-P, so on BXT0/1 we should * power down the second channel on PHY0 as well. + * + * FIXME: Clarify programming of the following, the register is + * read-only with bit 6 fixed at 0 at least in stepping A. */ if (phy == DPIO_PHY1) val |= OCL2_LDOFUSE_PWR_DIS; @@ -1810,12 +1887,10 @@ static void broxton_phy_init(struct drm_i915_private *dev_priv, * the corresponding calibrated value from PHY1, and disable * the automatic calibration on PHY0. */ - if (wait_for(I915_READ(BXT_PORT_REF_DW3(DPIO_PHY1)) & GRC_DONE, - 10)) - DRM_ERROR("timeout waiting for PHY1 GRC\n"); + broxton_phy_wait_grc_done(dev_priv, DPIO_PHY1); - val = I915_READ(BXT_PORT_REF_DW6(DPIO_PHY1)); - val = (val & GRC_CODE_MASK) >> GRC_CODE_SHIFT; + val = dev_priv->bxt_phy_grc = broxton_get_grc(dev_priv, + DPIO_PHY1); grc_code = val << GRC_CODE_FAST_SHIFT | val << GRC_CODE_SLOW_SHIFT | val; @@ -1825,17 +1900,27 @@ static void broxton_phy_init(struct drm_i915_private *dev_priv, val |= GRC_DIS | GRC_RDY_OVRD; I915_WRITE(BXT_PORT_REF_DW8(DPIO_PHY0), val); } + /* + * During PHY1 init delay waiting for GRC calibration to finish, since + * it can happen in parallel with the subsequent PHY0 init. + */ val = I915_READ(BXT_PHY_CTL_FAMILY(phy)); val |= COMMON_RESET_DIS; I915_WRITE(BXT_PHY_CTL_FAMILY(phy), val); } -void broxton_ddi_phy_init(struct drm_device *dev) +void broxton_ddi_phy_init(struct drm_i915_private *dev_priv) { /* Enable PHY1 first since it provides Rcomp for PHY0 */ - broxton_phy_init(dev->dev_private, DPIO_PHY1); - broxton_phy_init(dev->dev_private, DPIO_PHY0); + broxton_phy_init(dev_priv, DPIO_PHY1); + broxton_phy_init(dev_priv, DPIO_PHY0); + + /* + * If BIOS enabled only PHY0 and not PHY1, we skipped waiting for the + * PHY1 GRC calibration to finish, so wait for it here. + */ + broxton_phy_wait_grc_done(dev_priv, DPIO_PHY1); } static void broxton_phy_uninit(struct drm_i915_private *dev_priv, @@ -1846,17 +1931,126 @@ static void broxton_phy_uninit(struct drm_i915_private *dev_priv, val = I915_READ(BXT_PHY_CTL_FAMILY(phy)); val &= ~COMMON_RESET_DIS; I915_WRITE(BXT_PHY_CTL_FAMILY(phy), val); + + val = I915_READ(BXT_P_CR_GT_DISP_PWRON); + val &= ~GT_DISPLAY_POWER_ON(phy); + I915_WRITE(BXT_P_CR_GT_DISP_PWRON, val); } -void broxton_ddi_phy_uninit(struct drm_device *dev) +void broxton_ddi_phy_uninit(struct drm_i915_private *dev_priv) { - struct drm_i915_private *dev_priv = dev->dev_private; - broxton_phy_uninit(dev_priv, DPIO_PHY1); broxton_phy_uninit(dev_priv, DPIO_PHY0); +} + +static bool __printf(6, 7) +__phy_reg_verify_state(struct drm_i915_private *dev_priv, enum dpio_phy phy, + i915_reg_t reg, u32 mask, u32 expected, + const char *reg_fmt, ...) +{ + struct va_format vaf; + va_list args; + u32 val; + + val = I915_READ(reg); + if ((val & mask) == expected) + return true; + + va_start(args, reg_fmt); + vaf.fmt = reg_fmt; + vaf.va = &args; - /* FIXME: do this in broxton_phy_uninit per phy */ - I915_WRITE(BXT_P_CR_GT_DISP_PWRON, 0); + DRM_DEBUG_DRIVER("DDI PHY %d reg %pV [%08x] state mismatch: " + "current %08x, expected %08x (mask %08x)\n", + phy, &vaf, reg.reg, val, (val & ~mask) | expected, + mask); + + va_end(args); + + return false; +} + +static bool broxton_phy_verify_state(struct drm_i915_private *dev_priv, + enum dpio_phy phy) +{ + enum port port; + u32 ports; + uint32_t mask; + bool ok; + +#define _CHK(reg, mask, exp, fmt, ...) \ + __phy_reg_verify_state(dev_priv, phy, reg, mask, exp, fmt, \ + ## __VA_ARGS__) + + /* We expect the PHY to be always enabled */ + if (!broxton_phy_is_enabled(dev_priv, phy)) + return false; + + ok = true; + + if (phy == DPIO_PHY0) + ports = BIT(PORT_B) | BIT(PORT_C); + else + ports = BIT(PORT_A); + + for_each_port_masked(port, ports) { + int lane; + + for (lane = 0; lane < 4; lane++) + ok &= _CHK(BXT_PORT_TX_DW14_LN(port, lane), + LATENCY_OPTIM, + lane != 1 ? LATENCY_OPTIM : 0, + "BXT_PORT_TX_DW14_LN(%d, %d)", port, lane); + } + + /* PLL Rcomp code offset */ + ok &= _CHK(BXT_PORT_CL1CM_DW9(phy), + IREF0RC_OFFSET_MASK, 0xe4 << IREF0RC_OFFSET_SHIFT, + "BXT_PORT_CL1CM_DW9(%d)", phy); + ok &= _CHK(BXT_PORT_CL1CM_DW10(phy), + IREF1RC_OFFSET_MASK, 0xe4 << IREF1RC_OFFSET_SHIFT, + "BXT_PORT_CL1CM_DW10(%d)", phy); + + /* Power gating */ + mask = OCL1_POWER_DOWN_EN | DW28_OLDO_DYN_PWR_DOWN_EN | SUS_CLK_CONFIG; + ok &= _CHK(BXT_PORT_CL1CM_DW28(phy), mask, mask, + "BXT_PORT_CL1CM_DW28(%d)", phy); + + if (phy == DPIO_PHY0) + ok &= _CHK(BXT_PORT_CL2CM_DW6_BC, + DW6_OLDO_DYN_PWR_DOWN_EN, DW6_OLDO_DYN_PWR_DOWN_EN, + "BXT_PORT_CL2CM_DW6_BC"); + + /* + * TODO: Verify BXT_PORT_CL1CM_DW30 bit OCL2_LDOFUSE_PWR_DIS, + * at least on stepping A this bit is read-only and fixed at 0. + */ + + if (phy == DPIO_PHY0) { + u32 grc_code = dev_priv->bxt_phy_grc; + + grc_code = grc_code << GRC_CODE_FAST_SHIFT | + grc_code << GRC_CODE_SLOW_SHIFT | + grc_code; + mask = GRC_CODE_FAST_MASK | GRC_CODE_SLOW_MASK | + GRC_CODE_NOM_MASK; + ok &= _CHK(BXT_PORT_REF_DW6(DPIO_PHY0), mask, grc_code, + "BXT_PORT_REF_DW6(%d)", DPIO_PHY0); + + mask = GRC_DIS | GRC_RDY_OVRD; + ok &= _CHK(BXT_PORT_REF_DW8(DPIO_PHY0), mask, mask, + "BXT_PORT_REF_DW8(%d)", DPIO_PHY0); + } + + return ok; +#undef _CHK +} + +void broxton_ddi_phy_verify_state(struct drm_i915_private *dev_priv) +{ + if (!broxton_phy_verify_state(dev_priv, DPIO_PHY0) || + !broxton_phy_verify_state(dev_priv, DPIO_PHY1)) + i915_report_error(dev_priv, "DDI PHY state mismatch\n"); } void intel_ddi_prepare_link_retrain(struct intel_dp *intel_dp) @@ -2044,12 +2238,6 @@ void intel_ddi_get_config(struct intel_encoder *encoder, intel_ddi_clock_get(encoder, pipe_config); } -static void intel_ddi_destroy(struct drm_encoder *encoder) -{ - /* HDMI has nothing special to destroy, so we can go with this. */ - intel_dp_encoder_destroy(encoder); -} - static bool intel_ddi_compute_config(struct intel_encoder *encoder, struct intel_crtc_state *pipe_config) { @@ -2068,7 +2256,8 @@ static bool intel_ddi_compute_config(struct intel_encoder *encoder, } static const struct drm_encoder_funcs intel_ddi_funcs = { - .destroy = intel_ddi_destroy, + .reset = intel_dp_encoder_reset, + .destroy = intel_dp_encoder_destroy, }; static struct intel_connector * @@ -2167,6 +2356,7 @@ void intel_ddi_init(struct drm_device *dev, enum port port) intel_encoder->post_disable = intel_ddi_post_disable; intel_encoder->get_hw_state = intel_ddi_get_hw_state; intel_encoder->get_config = intel_ddi_get_config; + intel_encoder->suspend = intel_dp_encoder_suspend; intel_dig_port->port = port; intel_dig_port->saved_port_bits = I915_READ(DDI_BUF_CTL(port)) & diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 551541b303..ff60241 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -1530,45 +1530,47 @@ static void assert_pch_ports_disabled(struct drm_i915_private *dev_priv, assert_pch_hdmi_disabled(dev_priv, pipe, PCH_HDMID); } +static void _vlv_enable_pll(struct intel_crtc *crtc, + const struct intel_crtc_state *pipe_config) +{ + struct drm_i915_private *dev_priv = to_i915(crtc->base.dev); + enum pipe pipe = crtc->pipe; + + I915_WRITE(DPLL(pipe), pipe_config->dpll_hw_state.dpll); + POSTING_READ(DPLL(pipe)); + udelay(150); + + if (wait_for(((I915_READ(DPLL(pipe)) & DPLL_LOCK_VLV) == DPLL_LOCK_VLV), 1)) + DRM_ERROR("DPLL %d failed to lock\n", pipe); +} + static void vlv_enable_pll(struct intel_crtc *crtc, const struct intel_crtc_state *pipe_config) { - struct drm_device *dev = crtc->base.dev; - struct drm_i915_private *dev_priv = dev->dev_private; + struct drm_i915_private *dev_priv = to_i915(crtc->base.dev); enum pipe pipe = crtc->pipe; - i915_reg_t reg = DPLL(pipe); - u32 dpll = pipe_config->dpll_hw_state.dpll; assert_pipe_disabled(dev_priv, pipe); /* PLL is protected by panel, make sure we can write it */ assert_panel_unlocked(dev_priv, pipe); - I915_WRITE(reg, dpll); - POSTING_READ(reg); - udelay(150); - - if (wait_for(((I915_READ(reg) & DPLL_LOCK_VLV) == DPLL_LOCK_VLV), 1)) - DRM_ERROR("DPLL %d failed to lock\n", pipe); + if (pipe_config->dpll_hw_state.dpll & DPLL_VCO_ENABLE) + _vlv_enable_pll(crtc, pipe_config); I915_WRITE(DPLL_MD(pipe), pipe_config->dpll_hw_state.dpll_md); POSTING_READ(DPLL_MD(pipe)); } -static void chv_enable_pll(struct intel_crtc *crtc, - const struct intel_crtc_state *pipe_config) + +static void _chv_enable_pll(struct intel_crtc *crtc, + const struct intel_crtc_state *pipe_config) { - struct drm_device *dev = crtc->base.dev; - struct drm_i915_private *dev_priv = dev->dev_private; + struct drm_i915_private *dev_priv = to_i915(crtc->base.dev); enum pipe pipe = crtc->pipe; enum dpio_channel port = vlv_pipe_to_channel(pipe); u32 tmp; - assert_pipe_disabled(dev_priv, pipe); - - /* PLL is protected by panel, make sure we can write it */ - assert_panel_unlocked(dev_priv, pipe); - mutex_lock(&dev_priv->sb_lock); /* Enable back the 10bit clock to display controller */ @@ -1589,6 +1591,21 @@ static void chv_enable_pll(struct intel_crtc *crtc, /* Check PLL is locked */ if (wait_for(((I915_READ(DPLL(pipe)) & DPLL_LOCK_VLV) == DPLL_LOCK_VLV), 1)) DRM_ERROR("PLL %d failed to lock\n", pipe); +} + +static void chv_enable_pll(struct intel_crtc *crtc, + const struct intel_crtc_state *pipe_config) +{ + struct drm_i915_private *dev_priv = to_i915(crtc->base.dev); + enum pipe pipe = crtc->pipe; + + assert_pipe_disabled(dev_priv, pipe); + + /* PLL is protected by panel, make sure we can write it */ + assert_panel_unlocked(dev_priv, pipe); + + if (pipe_config->dpll_hw_state.dpll & DPLL_VCO_ENABLE) + _chv_enable_pll(crtc, pipe_config); if (pipe != PIPE_A) { /* @@ -3198,12 +3215,12 @@ void intel_finish_reset(struct drm_device *dev) static bool intel_crtc_has_pending_flip(struct drm_crtc *crtc) { struct drm_device *dev = crtc->dev; - struct drm_i915_private *dev_priv = dev->dev_private; struct intel_crtc *intel_crtc = to_intel_crtc(crtc); + unsigned reset_counter; bool pending; - if (i915_reset_in_progress(&dev_priv->gpu_error) || - intel_crtc->reset_counter != atomic_read(&dev_priv->gpu_error.reset_counter)) + reset_counter = i915_reset_counter(&to_i915(dev)->gpu_error); + if (intel_crtc->reset_counter != reset_counter) return false; spin_lock_irq(&dev->event_lock); @@ -3805,9 +3822,7 @@ static void page_flip_completed(struct intel_crtc *intel_crtc) intel_crtc->unpin_work = NULL; if (work->event) - drm_send_vblank_event(intel_crtc->base.dev, - intel_crtc->pipe, - work->event); + drm_crtc_send_vblank_event(&intel_crtc->base, work->event); drm_crtc_vblank_put(&intel_crtc->base); @@ -4088,12 +4103,6 @@ static void ironlake_pch_enable(struct drm_crtc *crtc) I915_WRITE(FDI_RX_TUSIZE1(pipe), I915_READ(PIPE_DATA_M1(pipe)) & TU_SIZE_MASK); - /* - * Sometimes spurious CPU pipe underruns happen during FDI - * training, at least with VGA+HDMI cloning. Suppress them. - */ - intel_set_cpu_fifo_underrun_reporting(dev_priv, pipe, false); - /* For PCH output, training FDI link */ dev_priv->display.fdi_link_train(crtc); @@ -4128,8 +4137,6 @@ static void ironlake_pch_enable(struct drm_crtc *crtc) intel_fdi_normal_train(crtc); - intel_set_cpu_fifo_underrun_reporting(dev_priv, pipe, true); - /* For PCH DP, enable TRANS_DP_CTL */ if (HAS_PCH_CPT(dev) && intel_crtc->config->has_dp_encoder) { const struct drm_display_mode *adjusted_mode = @@ -4732,6 +4739,18 @@ static void ironlake_crtc_enable(struct drm_crtc *crtc) if (WARN_ON(intel_crtc->active)) return; + /* + * Sometimes spurious CPU pipe underruns happen during FDI + * training, at least with VGA+HDMI cloning. Suppress them. + * + * On ILK we get an occasional spurious CPU pipe underruns + * between eDP port A enable and vdd enable. Also PCH port + * enable seems to result in the occasional CPU pipe underrun. + * + * Spurious PCH underruns also occur during PCH enabling. + */ + if (intel_crtc->config->has_pch_encoder || IS_GEN5(dev_priv)) + intel_set_cpu_fifo_underrun_reporting(dev_priv, pipe, false); if (intel_crtc->config->has_pch_encoder) intel_set_pch_fifo_underrun_reporting(dev_priv, pipe, false); @@ -4753,8 +4772,6 @@ static void ironlake_crtc_enable(struct drm_crtc *crtc) intel_crtc->active = true; - intel_set_cpu_fifo_underrun_reporting(dev_priv, pipe, true); - for_each_encoder_on_crtc(dev, crtc, encoder) if (encoder->pre_enable) encoder->pre_enable(encoder); @@ -4796,6 +4813,7 @@ static void ironlake_crtc_enable(struct drm_crtc *crtc) /* Must wait for vblank to avoid spurious PCH FIFO underruns */ if (intel_crtc->config->has_pch_encoder) intel_wait_for_vblank(dev, pipe); + intel_set_cpu_fifo_underrun_reporting(dev_priv, pipe, true); intel_set_pch_fifo_underrun_reporting(dev_priv, pipe, true); } @@ -4948,8 +4966,15 @@ static void ironlake_crtc_disable(struct drm_crtc *crtc) struct intel_encoder *encoder; int pipe = intel_crtc->pipe; - if (intel_crtc->config->has_pch_encoder) + /* + * Sometimes spurious CPU pipe underruns happen when the + * pipe is already disabled, but FDI RX/TX is still enabled. + * Happens at least with VGA+HDMI cloning. Suppress them. + */ + if (intel_crtc->config->has_pch_encoder) { + intel_set_cpu_fifo_underrun_reporting(dev_priv, pipe, false); intel_set_pch_fifo_underrun_reporting(dev_priv, pipe, false); + } for_each_encoder_on_crtc(dev, crtc, encoder) encoder->disable(encoder); @@ -4957,22 +4982,12 @@ static void ironlake_crtc_disable(struct drm_crtc *crtc) drm_crtc_vblank_off(crtc); assert_vblank_disabled(crtc); - /* - * Sometimes spurious CPU pipe underruns happen when the - * pipe is already disabled, but FDI RX/TX is still enabled. - * Happens at least with VGA+HDMI cloning. Suppress them. - */ - if (intel_crtc->config->has_pch_encoder) - intel_set_cpu_fifo_underrun_reporting(dev_priv, pipe, false); - intel_disable_pipe(intel_crtc); ironlake_pfit_disable(intel_crtc, false); - if (intel_crtc->config->has_pch_encoder) { + if (intel_crtc->config->has_pch_encoder) ironlake_fdi_disable(crtc); - intel_set_cpu_fifo_underrun_reporting(dev_priv, pipe, true); - } for_each_encoder_on_crtc(dev, crtc, encoder) if (encoder->post_disable) @@ -5002,6 +5017,7 @@ static void ironlake_crtc_disable(struct drm_crtc *crtc) ironlake_fdi_pll_disable(intel_crtc); } + intel_set_cpu_fifo_underrun_reporting(dev_priv, pipe, true); intel_set_pch_fifo_underrun_reporting(dev_priv, pipe, true); } @@ -5329,9 +5345,8 @@ static void intel_update_cdclk(struct drm_device *dev) intel_update_max_cdclk(dev); } -static void broxton_set_cdclk(struct drm_device *dev, int frequency) +static void broxton_set_cdclk(struct drm_i915_private *dev_priv, int frequency) { - struct drm_i915_private *dev_priv = dev->dev_private; uint32_t divider; uint32_t ratio; uint32_t current_freq; @@ -5445,33 +5460,46 @@ static void broxton_set_cdclk(struct drm_device *dev, int frequency) return; } - intel_update_cdclk(dev); + intel_update_cdclk(dev_priv->dev); } -void broxton_init_cdclk(struct drm_device *dev) +static bool broxton_cdclk_is_enabled(struct drm_i915_private *dev_priv) { - struct drm_i915_private *dev_priv = dev->dev_private; - uint32_t val; + if (!(I915_READ(BXT_DE_PLL_ENABLE) & BXT_DE_PLL_PLL_ENABLE)) + return false; - /* - * NDE_RSTWRN_OPT RST PCH Handshake En must always be 0b on BXT - * or else the reset will hang because there is no PCH to respond. - * Move the handshake programming to initialization sequence. - * Previously was left up to BIOS. - */ - val = I915_READ(HSW_NDE_RSTWRN_OPT); - val &= ~RESET_PCH_HANDSHAKE_ENABLE; - I915_WRITE(HSW_NDE_RSTWRN_OPT, val); + /* TODO: Check for a valid CDCLK rate */ + + if (!(I915_READ(DBUF_CTL) & DBUF_POWER_REQUEST)) { + DRM_DEBUG_DRIVER("CDCLK enabled, but DBUF power not requested\n"); + + return false; + } + + if (!(I915_READ(DBUF_CTL) & DBUF_POWER_STATE)) { + DRM_DEBUG_DRIVER("CDCLK enabled, but DBUF power hasn't settled\n"); + + return false; + } + + return true; +} - /* Enable PG1 for cdclk */ - intel_display_power_get(dev_priv, POWER_DOMAIN_PLLS); +bool broxton_cdclk_verify_state(struct drm_i915_private *dev_priv) +{ + return broxton_cdclk_is_enabled(dev_priv); +} +void broxton_init_cdclk(struct drm_i915_private *dev_priv) +{ /* check if cd clock is enabled */ - if (I915_READ(BXT_DE_PLL_ENABLE) & BXT_DE_PLL_PLL_ENABLE) { - DRM_DEBUG_KMS("Display already initialized\n"); + if (broxton_cdclk_is_enabled(dev_priv)) { + DRM_DEBUG_KMS("CDCLK already enabled, won't reprogram it\n"); return; } + DRM_DEBUG_KMS("CDCLK not enabled, enabling it\n"); + /* * FIXME: * - The initial CDCLK needs to be read from VBT. @@ -5479,7 +5507,7 @@ void broxton_init_cdclk(struct drm_device *dev) * - check if setting the max (or any) cdclk freq is really necessary * here, it belongs to modeset time */ - broxton_set_cdclk(dev, 624000); + broxton_set_cdclk(dev_priv, 624000); I915_WRITE(DBUF_CTL, I915_READ(DBUF_CTL) | DBUF_POWER_REQUEST); POSTING_READ(DBUF_CTL); @@ -5490,10 +5518,8 @@ void broxton_init_cdclk(struct drm_device *dev) DRM_ERROR("DBuf power enable timeout!\n"); } -void broxton_uninit_cdclk(struct drm_device *dev) +void broxton_uninit_cdclk(struct drm_i915_private *dev_priv) { - struct drm_i915_private *dev_priv = dev->dev_private; - I915_WRITE(DBUF_CTL, I915_READ(DBUF_CTL) & ~DBUF_POWER_REQUEST); POSTING_READ(DBUF_CTL); @@ -5503,9 +5529,7 @@ void broxton_uninit_cdclk(struct drm_device *dev) DRM_ERROR("DBuf power disable timeout!\n"); /* Set minimum (bypass) frequency, in effect turning off the DE PLL */ - broxton_set_cdclk(dev, 19200); - - intel_display_power_put(dev_priv, POWER_DOMAIN_PLLS); + broxton_set_cdclk(dev_priv, 19200); } static const struct skl_cdclk_entry { @@ -6072,14 +6096,12 @@ static void valleyview_crtc_enable(struct drm_crtc *crtc) if (encoder->pre_pll_enable) encoder->pre_pll_enable(encoder); - if (!intel_crtc->config->has_dsi_encoder) { - if (IS_CHERRYVIEW(dev)) { - chv_prepare_pll(intel_crtc, intel_crtc->config); - chv_enable_pll(intel_crtc, intel_crtc->config); - } else { - vlv_prepare_pll(intel_crtc, intel_crtc->config); - vlv_enable_pll(intel_crtc, intel_crtc->config); - } + if (IS_CHERRYVIEW(dev)) { + chv_prepare_pll(intel_crtc, intel_crtc->config); + chv_enable_pll(intel_crtc, intel_crtc->config); + } else { + vlv_prepare_pll(intel_crtc, intel_crtc->config); + vlv_enable_pll(intel_crtc, intel_crtc->config); } for_each_encoder_on_crtc(dev, crtc, encoder) @@ -6117,7 +6139,7 @@ static void i9xx_crtc_enable(struct drm_crtc *crtc) struct intel_encoder *encoder; struct intel_crtc_state *pipe_config = to_intel_crtc_state(crtc->state); - int pipe = intel_crtc->pipe; + enum pipe pipe = intel_crtc->pipe; if (WARN_ON(intel_crtc->active)) return; @@ -7173,11 +7195,15 @@ static void vlv_compute_dpll(struct intel_crtc *crtc, struct intel_crtc_state *pipe_config) { pipe_config->dpll_hw_state.dpll = DPLL_INTEGRATED_REF_CLK_VLV | - DPLL_REF_CLK_ENABLE_VLV | DPLL_VGA_MODE_DIS | - DPLL_VCO_ENABLE | DPLL_EXT_BUFFER_ENABLE_VLV; + DPLL_REF_CLK_ENABLE_VLV | DPLL_VGA_MODE_DIS; if (crtc->pipe != PIPE_A) pipe_config->dpll_hw_state.dpll |= DPLL_INTEGRATED_CRI_CLK_VLV; + /* DPLL not used with DSI, but still need the rest set up */ + if (!pipe_config->has_dsi_encoder) + pipe_config->dpll_hw_state.dpll |= DPLL_VCO_ENABLE | + DPLL_EXT_BUFFER_ENABLE_VLV; + pipe_config->dpll_hw_state.dpll_md = (pipe_config->pixel_multiplier - 1) << DPLL_MD_UDI_MULTIPLIER_SHIFT; } @@ -7186,11 +7212,14 @@ static void chv_compute_dpll(struct intel_crtc *crtc, struct intel_crtc_state *pipe_config) { pipe_config->dpll_hw_state.dpll = DPLL_SSC_REF_CLK_CHV | - DPLL_REF_CLK_ENABLE_VLV | DPLL_VGA_MODE_DIS | - DPLL_VCO_ENABLE; + DPLL_REF_CLK_ENABLE_VLV | DPLL_VGA_MODE_DIS; if (crtc->pipe != PIPE_A) pipe_config->dpll_hw_state.dpll |= DPLL_INTEGRATED_CRI_CLK_VLV; + /* DPLL not used with DSI, but still need the rest set up */ + if (!pipe_config->has_dsi_encoder) + pipe_config->dpll_hw_state.dpll |= DPLL_VCO_ENABLE; + pipe_config->dpll_hw_state.dpll_md = (pipe_config->pixel_multiplier - 1) << DPLL_MD_UDI_MULTIPLIER_SHIFT; } @@ -7200,11 +7229,20 @@ static void vlv_prepare_pll(struct intel_crtc *crtc, { struct drm_device *dev = crtc->base.dev; struct drm_i915_private *dev_priv = dev->dev_private; - int pipe = crtc->pipe; + enum pipe pipe = crtc->pipe; u32 mdiv; u32 bestn, bestm1, bestm2, bestp1, bestp2; u32 coreclk, reg_val; + /* Enable Refclk */ + I915_WRITE(DPLL(pipe), + pipe_config->dpll_hw_state.dpll & + ~(DPLL_VCO_ENABLE | DPLL_EXT_BUFFER_ENABLE_VLV)); + + /* No need to actually set up the DPLL with DSI */ + if ((pipe_config->dpll_hw_state.dpll & DPLL_VCO_ENABLE) == 0) + return; + mutex_lock(&dev_priv->sb_lock); bestn = pipe_config->dpll.n; @@ -7291,14 +7329,21 @@ static void chv_prepare_pll(struct intel_crtc *crtc, { struct drm_device *dev = crtc->base.dev; struct drm_i915_private *dev_priv = dev->dev_private; - int pipe = crtc->pipe; - i915_reg_t dpll_reg = DPLL(crtc->pipe); + enum pipe pipe = crtc->pipe; enum dpio_channel port = vlv_pipe_to_channel(pipe); u32 loopfilter, tribuf_calcntr; u32 bestn, bestm1, bestm2, bestp1, bestp2, bestm2_frac; u32 dpio_val; int vco; + /* Enable Refclk and SSC */ + I915_WRITE(DPLL(pipe), + pipe_config->dpll_hw_state.dpll & ~DPLL_VCO_ENABLE); + + /* No need to actually set up the DPLL with DSI */ + if ((pipe_config->dpll_hw_state.dpll & DPLL_VCO_ENABLE) == 0) + return; + bestn = pipe_config->dpll.n; bestm2_frac = pipe_config->dpll.m2 & 0x3fffff; bestm1 = pipe_config->dpll.m1; @@ -7309,12 +7354,6 @@ static void chv_prepare_pll(struct intel_crtc *crtc, dpio_val = 0; loopfilter = 0; - /* - * Enable Refclk and SSC - */ - I915_WRITE(dpll_reg, - pipe_config->dpll_hw_state.dpll & ~DPLL_VCO_ENABLE); - mutex_lock(&dev_priv->sb_lock); /* p1 and p2 divider */ @@ -7929,9 +7968,6 @@ static int chv_crtc_compute_clock(struct intel_crtc *crtc, memset(&crtc_state->dpll_hw_state, 0, sizeof(crtc_state->dpll_hw_state)); - if (crtc_state->has_dsi_encoder) - return 0; - if (!crtc_state->clock_set && !chv_find_best_dpll(limit, crtc_state, crtc_state->port_clock, refclk, NULL, &crtc_state->dpll)) { @@ -7953,9 +7989,6 @@ static int vlv_crtc_compute_clock(struct intel_crtc *crtc, memset(&crtc_state->dpll_hw_state, 0, sizeof(crtc_state->dpll_hw_state)); - if (crtc_state->has_dsi_encoder) - return 0; - if (!crtc_state->clock_set && !vlv_find_best_dpll(limit, crtc_state, crtc_state->port_clock, refclk, NULL, &crtc_state->dpll)) { @@ -8008,8 +8041,8 @@ static void vlv_crtc_clock_get(struct intel_crtc *crtc, u32 mdiv; int refclk = 100000; - /* In case of MIPI DPLL will not even be used */ - if (!(pipe_config->dpll_hw_state.dpll & DPLL_VCO_ENABLE)) + /* In case of DSI, DPLL will not be used */ + if ((pipe_config->dpll_hw_state.dpll & DPLL_VCO_ENABLE) == 0) return; mutex_lock(&dev_priv->sb_lock); @@ -8105,6 +8138,10 @@ static void chv_crtc_clock_get(struct intel_crtc *crtc, u32 cmn_dw13, pll_dw0, pll_dw1, pll_dw2, pll_dw3; int refclk = 100000; + /* In case of DSI, DPLL will not be used */ + if ((pipe_config->dpll_hw_state.dpll & DPLL_VCO_ENABLE) == 0) + return; + mutex_lock(&dev_priv->sb_lock); cmn_dw13 = vlv_dpio_read(dev_priv, pipe, CHV_CMN_DW13(port)); pll_dw0 = vlv_dpio_read(dev_priv, pipe, CHV_PLL_DW0(port)); @@ -9533,7 +9570,7 @@ static void broxton_modeset_commit_cdclk(struct drm_atomic_state *old_state) to_intel_atomic_state(old_state); unsigned int req_cdclk = old_intel_state->dev_cdclk; - broxton_set_cdclk(dev, req_cdclk); + broxton_set_cdclk(to_i915(dev), req_cdclk); } /* compute the max rate for new configuration */ @@ -10903,9 +10940,10 @@ static bool page_flip_finished(struct intel_crtc *crtc) { struct drm_device *dev = crtc->base.dev; struct drm_i915_private *dev_priv = dev->dev_private; + unsigned reset_counter; - if (i915_reset_in_progress(&dev_priv->gpu_error) || - crtc->reset_counter != atomic_read(&dev_priv->gpu_error.reset_counter)) + reset_counter = i915_reset_counter(&dev_priv->gpu_error); + if (crtc->reset_counter != reset_counter) return true; /* @@ -11359,7 +11397,6 @@ static void intel_mmio_flip_work_func(struct work_struct *work) if (mmio_flip->req) { WARN_ON(__i915_wait_request(mmio_flip->req, - mmio_flip->crtc->reset_counter, false, NULL, &mmio_flip->i915->rps.mmioflips)); i915_gem_request_unreference__unlocked(mmio_flip->req); @@ -11567,8 +11604,13 @@ static int intel_crtc_page_flip(struct drm_crtc *crtc, if (ret) goto cleanup; + intel_crtc->reset_counter = i915_reset_counter(&dev_priv->gpu_error); + if (__i915_reset_in_progress_or_wedged(intel_crtc->reset_counter)) { + ret = -EIO; + goto cleanup; + } + atomic_inc(&intel_crtc->unpin_work_count); - intel_crtc->reset_counter = atomic_read(&dev_priv->gpu_error.reset_counter); if (INTEL_INFO(dev)->gen >= 5 || IS_G4X(dev)) work->flip_count = I915_READ(PIPE_FLIPCOUNT_G4X(pipe)) + 1; @@ -11654,7 +11696,7 @@ cleanup_unpin: intel_unpin_fb_obj(fb, crtc->primary->state->rotation); cleanup_pending: if (!IS_ERR_OR_NULL(request)) - i915_gem_request_cancel(request); + i915_add_request_no_flush(request); atomic_dec(&intel_crtc->unpin_work_count); mutex_unlock(&dev->struct_mutex); cleanup: @@ -11704,7 +11746,7 @@ retry: if (ret == 0 && event) { spin_lock_irq(&dev->event_lock); - drm_send_vblank_event(dev, pipe, event); + drm_crtc_send_vblank_event(crtc, event); spin_unlock_irq(&dev->event_lock); } } @@ -12686,7 +12728,7 @@ intel_pipe_config_compare(struct drm_device *dev, PIPE_CONF_CHECK_X(gmch_pfit.control); /* pfit ratios are autocomputed by the hw on gen4+ */ if (INTEL_INFO(dev)->gen < 4) - PIPE_CONF_CHECK_I(gmch_pfit.pgm_ratios); + PIPE_CONF_CHECK_X(gmch_pfit.pgm_ratios); PIPE_CONF_CHECK_X(gmch_pfit.lvds_border_bits); if (!adjust) { @@ -12721,6 +12763,9 @@ intel_pipe_config_compare(struct drm_device *dev, PIPE_CONF_CHECK_X(dpll_hw_state.cfgcr1); PIPE_CONF_CHECK_X(dpll_hw_state.cfgcr2); + PIPE_CONF_CHECK_X(dsi_pll.ctrl); + PIPE_CONF_CHECK_X(dsi_pll.div); + if (IS_G4X(dev) || INTEL_INFO(dev)->gen >= 5) PIPE_CONF_CHECK_I(pipe_bpp); @@ -13401,6 +13446,9 @@ static int intel_atomic_prepare_commit(struct drm_device *dev, } for_each_crtc_in_state(state, crtc, crtc_state, i) { + if (state->legacy_cursor_update) + continue; + ret = intel_crtc_wait_for_pending_flips(crtc); if (ret) return ret; @@ -13414,12 +13462,9 @@ static int intel_atomic_prepare_commit(struct drm_device *dev, return ret; ret = drm_atomic_helper_prepare_planes(dev, state); - if (!ret && !async && !i915_reset_in_progress(&dev_priv->gpu_error)) { - u32 reset_counter; - - reset_counter = atomic_read(&dev_priv->gpu_error.reset_counter); - mutex_unlock(&dev->struct_mutex); + mutex_unlock(&dev->struct_mutex); + if (!ret && !async) { for_each_plane_in_state(state, plane, plane_state, i) { struct intel_plane_state *intel_plane_state = to_intel_plane_state(plane_state); @@ -13428,25 +13473,18 @@ static int intel_atomic_prepare_commit(struct drm_device *dev, continue; ret = __i915_wait_request(intel_plane_state->wait_req, - reset_counter, true, - NULL, NULL); - - /* Swallow -EIO errors to allow updates during hw lockup. */ - if (ret == -EIO) - ret = 0; - - if (ret) + true, NULL, NULL); + if (ret) { + /* Any hang should be swallowed by the wait */ + WARN_ON(ret == -EIO); + mutex_lock(&dev->struct_mutex); + drm_atomic_helper_cleanup_planes(dev, state); + mutex_unlock(&dev->struct_mutex); break; + } } - - if (!ret) - return 0; - - mutex_lock(&dev->struct_mutex); - drm_atomic_helper_cleanup_planes(dev, state); } - mutex_unlock(&dev->struct_mutex); return ret; } @@ -13488,7 +13526,7 @@ static void intel_atomic_wait_for_vblanks(struct drm_device *dev, drm_crtc_vblank_count(crtc), msecs_to_jiffies(50)); - WARN_ON(!lret); + WARN(!lret, "pipe %c vblank wait timed out\n", pipe_name(pipe)); drm_crtc_vblank_put(crtc); } @@ -13790,10 +13828,11 @@ intel_prepare_plane_fb(struct drm_plane *plane, */ if (needs_modeset(crtc_state)) ret = i915_gem_object_wait_rendering(old_obj, true); - - /* Swallow -EIO errors to allow updates during hw lockup. */ - if (ret && ret != -EIO) + if (ret) { + /* GPU hangs should have been swallowed by the wait */ + WARN_ON(ret == -EIO); return ret; + } } /* For framebuffer backed by dmabuf, wait for fence */ diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c index da0c3d2..a3fc494 100644 --- a/drivers/gpu/drm/i915/intel_dp.c +++ b/drivers/gpu/drm/i915/intel_dp.c @@ -2215,6 +2215,15 @@ static void ironlake_edp_pll_on(struct intel_dp *intel_dp) POSTING_READ(DP_A); udelay(500); + /* + * [DevILK] Work around required when enabling DP PLL + * while a pipe is enabled going to FDI: + * 1. Wait for the start of vertical blank on the enabled pipe going to FDI + * 2. Program DP PLL enable + */ + if (IS_GEN5(dev_priv)) + intel_wait_for_vblank_if_active(dev_priv->dev, !crtc->pipe); + intel_dp->DP |= DP_PLL_ENABLE; I915_WRITE(DP_A, intel_dp->DP); @@ -2630,7 +2639,6 @@ static void intel_enable_dp(struct intel_encoder *encoder) struct drm_i915_private *dev_priv = dev->dev_private; struct intel_crtc *crtc = to_intel_crtc(encoder->base.crtc); uint32_t dp_reg = I915_READ(intel_dp->output_reg); - enum port port = dp_to_dig_port(intel_dp)->port; enum pipe pipe = crtc->pipe; if (WARN_ON(dp_reg & DP_PORT_EN)) @@ -2641,35 +2649,12 @@ static void intel_enable_dp(struct intel_encoder *encoder) if (IS_VALLEYVIEW(dev) || IS_CHERRYVIEW(dev)) vlv_init_panel_power_sequencer(intel_dp); - /* - * We get an occasional spurious underrun between the port - * enable and vdd enable, when enabling port A eDP. - * - * FIXME: Not sure if this applies to (PCH) port D eDP as well - */ - if (port == PORT_A) - intel_set_cpu_fifo_underrun_reporting(dev_priv, pipe, false); - intel_dp_enable_port(intel_dp); - if (port == PORT_A && IS_GEN5(dev_priv)) { - /* - * Underrun reporting for the other pipe was disabled in - * g4x_pre_enable_dp(). The eDP PLL and port have now been - * enabled, so it's now safe to re-enable underrun reporting. - */ - intel_wait_for_vblank_if_active(dev_priv->dev, !pipe); - intel_set_cpu_fifo_underrun_reporting(dev_priv, !pipe, true); - intel_set_pch_fifo_underrun_reporting(dev_priv, !pipe, true); - } - edp_panel_vdd_on(intel_dp); edp_panel_on(intel_dp); edp_panel_vdd_off(intel_dp, true); - if (port == PORT_A) - intel_set_cpu_fifo_underrun_reporting(dev_priv, pipe, true); - pps_unlock(intel_dp); if (IS_VALLEYVIEW(dev) || IS_CHERRYVIEW(dev)) { @@ -2711,26 +2696,11 @@ static void vlv_enable_dp(struct intel_encoder *encoder) static void g4x_pre_enable_dp(struct intel_encoder *encoder) { - struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base); enum port port = dp_to_dig_port(intel_dp)->port; - enum pipe pipe = to_intel_crtc(encoder->base.crtc)->pipe; intel_dp_prepare(encoder); - if (port == PORT_A && IS_GEN5(dev_priv)) { - /* - * We get FIFO underruns on the other pipe when - * enabling the CPU eDP PLL, and when enabling CPU - * eDP port. We could potentially avoid the PLL - * underrun with a vblank wait just prior to enabling - * the PLL, but that doesn't appear to help the port - * enable case. Just sweep it all under the rug. - */ - intel_set_cpu_fifo_underrun_reporting(dev_priv, !pipe, false); - intel_set_pch_fifo_underrun_reporting(dev_priv, !pipe, false); - } - /* Only ilk+ has port A */ if (port == PORT_A) ironlake_edp_pll_on(intel_dp); @@ -3806,7 +3776,7 @@ intel_dp_get_dpcd(struct intel_dp *intel_dp) * downstream port information. So, an early return here saves * time from performing other operations which are not required. */ - if (!intel_dp->sink_count) + if (!is_edp(intel_dp) && !intel_dp->sink_count) return false; /* Check if the panel supports PSR */ @@ -4339,6 +4309,9 @@ intel_dp_detect_dpcd(struct intel_dp *intel_dp) if (!intel_dp_get_dpcd(intel_dp)) return connector_status_disconnected; + if (is_edp(intel_dp)) + return connector_status_connected; + /* if there's no downstream port, we're done */ if (!(dpcd[DP_DOWNSTREAMPORT_PRESENT] & DP_DWN_STRM_PORT_PRESENT)) return connector_status_connected; @@ -4608,6 +4581,15 @@ intel_dp_long_pulse(struct intel_connector *intel_connector) intel_dp->compliance_test_type = 0; intel_dp->compliance_test_data = 0; + if (intel_dp->is_mst) { + DRM_DEBUG_KMS("MST device may have disappeared %d vs %d\n", + intel_dp->is_mst, + intel_dp->mst_mgr.mst_state); + intel_dp->is_mst = false; + drm_dp_mst_topology_mgr_set_mst(&intel_dp->mst_mgr, + intel_dp->is_mst); + } + goto out; } @@ -4665,20 +4647,9 @@ intel_dp_long_pulse(struct intel_connector *intel_connector) } out: - if (status != connector_status_connected) { + if ((status != connector_status_connected) && + (intel_dp->is_mst == false)) intel_dp_unset_edid(intel_dp); - /* - * If we were in MST mode, and device is not there, - * get out of MST mode - */ - if (intel_dp->is_mst) { - DRM_DEBUG_KMS("MST device may have disappeared %d vs %d\n", - intel_dp->is_mst, intel_dp->mst_mgr.mst_state); - intel_dp->is_mst = false; - drm_dp_mst_topology_mgr_set_mst(&intel_dp->mst_mgr, - intel_dp->is_mst); - } - } intel_display_power_put(to_i915(dev), power_domain); return; @@ -4851,6 +4822,11 @@ intel_dp_set_property(struct drm_connector *connector, DRM_DEBUG_KMS("no scaling not supported\n"); return -EINVAL; } + if (HAS_GMCH_DISPLAY(dev_priv) && + val == DRM_MODE_SCALE_CENTER) { + DRM_DEBUG_KMS("centering not supported\n"); + return -EINVAL; + } if (intel_connector->panel.fitting_mode == val) { /* the eDP scaling property is not changed */ @@ -4914,7 +4890,7 @@ void intel_dp_encoder_destroy(struct drm_encoder *encoder) kfree(intel_dig_port); } -static void intel_dp_encoder_suspend(struct intel_encoder *intel_encoder) +void intel_dp_encoder_suspend(struct intel_encoder *intel_encoder) { struct intel_dp *intel_dp = enc_to_intel_dp(&intel_encoder->base); @@ -4956,7 +4932,7 @@ static void intel_edp_panel_vdd_sanitize(struct intel_dp *intel_dp) edp_panel_vdd_schedule_off(intel_dp); } -static void intel_dp_encoder_reset(struct drm_encoder *encoder) +void intel_dp_encoder_reset(struct drm_encoder *encoder) { struct intel_dp *intel_dp; diff --git a/drivers/gpu/drm/i915/intel_dpll_mgr.c b/drivers/gpu/drm/i915/intel_dpll_mgr.c index 0bde6a4..639bf02 100644 --- a/drivers/gpu/drm/i915/intel_dpll_mgr.c +++ b/drivers/gpu/drm/i915/intel_dpll_mgr.c @@ -1295,17 +1295,9 @@ static void bxt_ddi_pll_enable(struct drm_i915_private *dev_priv, uint32_t temp; enum port port = (enum port)pll->id; /* 1:1 port->PLL mapping */ - temp = I915_READ(BXT_PORT_PLL_ENABLE(port)); - /* - * Definition of each bit polarity has been changed - * after A1 stepping - */ - if (IS_BXT_REVID(dev_priv, 0, BXT_REVID_A1)) - temp &= ~PORT_PLL_REF_SEL; - else - temp |= PORT_PLL_REF_SEL; - /* Non-SSC reference */ + temp = I915_READ(BXT_PORT_PLL_ENABLE(port)); + temp |= PORT_PLL_REF_SEL; I915_WRITE(BXT_PORT_PLL_ENABLE(port), temp); /* Disable 10 bit clock */ @@ -1652,10 +1644,7 @@ static void intel_ddi_pll_init(struct drm_device *dev) DRM_DEBUG_KMS("Sanitized cdclk programmed by pre-os\n"); if (!(I915_READ(LCPLL1_CTL) & LCPLL_PLL_ENABLE)) DRM_ERROR("LCPLL1 is disabled\n"); - } else if (IS_BROXTON(dev)) { - broxton_init_cdclk(dev); - broxton_ddi_phy_init(dev); - } else { + } else if (!IS_BROXTON(dev_priv)) { /* * The LCPLL register should be turned on by the BIOS. For now * let's just check its state and print errors in case diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h index e0fcfa1..b9f1304 100644 --- a/drivers/gpu/drm/i915/intel_drv.h +++ b/drivers/gpu/drm/i915/intel_drv.h @@ -497,6 +497,11 @@ struct intel_crtc_state { /* Actual register state of the dpll, for shared dpll cross-checking. */ struct intel_dpll_hw_state dpll_hw_state; + /* DSI PLL registers */ + struct { + u32 ctrl, div; + } dsi_pll; + int pipe_bpp; struct intel_link_m_n dp_m_n; @@ -1224,12 +1229,16 @@ void intel_prepare_reset(struct drm_device *dev); void intel_finish_reset(struct drm_device *dev); void hsw_enable_pc8(struct drm_i915_private *dev_priv); void hsw_disable_pc8(struct drm_i915_private *dev_priv); -void broxton_init_cdclk(struct drm_device *dev); -void broxton_uninit_cdclk(struct drm_device *dev); -void broxton_ddi_phy_init(struct drm_device *dev); -void broxton_ddi_phy_uninit(struct drm_device *dev); +void broxton_init_cdclk(struct drm_i915_private *dev_priv); +void broxton_uninit_cdclk(struct drm_i915_private *dev_priv); +bool broxton_cdclk_verify_state(struct drm_i915_private *dev_priv); +void broxton_ddi_phy_init(struct drm_i915_private *dev_priv); +void broxton_ddi_phy_uninit(struct drm_i915_private *dev_priv); +void broxton_ddi_phy_verify_state(struct drm_i915_private *dev_priv); +void gen9_sanitize_dc_state(struct drm_i915_private *dev_priv); void bxt_enable_dc9(struct drm_i915_private *dev_priv); void bxt_disable_dc9(struct drm_i915_private *dev_priv); +void gen9_enable_dc5(struct drm_i915_private *dev_priv); void skl_init_cdclk(struct drm_i915_private *dev_priv); int skl_sanitize_cdclk(struct drm_i915_private *dev_priv); void skl_uninit_cdclk(struct drm_i915_private *dev_priv); @@ -1268,6 +1277,8 @@ u32 skl_plane_ctl_rotation(unsigned int rotation); void intel_csr_ucode_init(struct drm_i915_private *); void intel_csr_load_program(struct drm_i915_private *); void intel_csr_ucode_fini(struct drm_i915_private *); +void intel_csr_ucode_suspend(struct drm_i915_private *); +void intel_csr_ucode_resume(struct drm_i915_private *); /* intel_dp.c */ void intel_dp_init(struct drm_device *dev, i915_reg_t output_reg, enum port port); @@ -1278,6 +1289,8 @@ void intel_dp_set_link_params(struct intel_dp *intel_dp, void intel_dp_start_link_train(struct intel_dp *intel_dp); void intel_dp_stop_link_train(struct intel_dp *intel_dp); void intel_dp_sink_dpms(struct intel_dp *intel_dp, int mode); +void intel_dp_encoder_reset(struct drm_encoder *encoder); +void intel_dp_encoder_suspend(struct intel_encoder *intel_encoder); void intel_dp_encoder_destroy(struct drm_encoder *encoder); int intel_dp_sink_crc(struct intel_dp *intel_dp, u8 *crc); bool intel_dp_compute_config(struct intel_encoder *encoder, @@ -1462,8 +1475,8 @@ int intel_power_domains_init(struct drm_i915_private *); void intel_power_domains_fini(struct drm_i915_private *); void intel_power_domains_init_hw(struct drm_i915_private *dev_priv, bool resume); void intel_power_domains_suspend(struct drm_i915_private *dev_priv); -void skl_pw1_misc_io_init(struct drm_i915_private *dev_priv); -void skl_pw1_misc_io_fini(struct drm_i915_private *dev_priv); +void bxt_display_core_init(struct drm_i915_private *dev_priv, bool resume); +void bxt_display_core_uninit(struct drm_i915_private *dev_priv); void intel_runtime_pm_enable(struct drm_i915_private *dev_priv); const char * intel_display_power_domain_str(enum intel_display_power_domain domain); diff --git a/drivers/gpu/drm/i915/intel_dsi.c b/drivers/gpu/drm/i915/intel_dsi.c index a1e0547..2b22bb9 100644 --- a/drivers/gpu/drm/i915/intel_dsi.c +++ b/drivers/gpu/drm/i915/intel_dsi.c @@ -290,16 +290,26 @@ static bool intel_dsi_compute_config(struct intel_encoder *encoder, struct intel_dsi *intel_dsi = container_of(encoder, struct intel_dsi, base); struct intel_connector *intel_connector = intel_dsi->attached_connector; - struct drm_display_mode *fixed_mode = intel_connector->panel.fixed_mode; + struct intel_crtc *crtc = to_intel_crtc(pipe_config->base.crtc); + const struct drm_display_mode *fixed_mode = intel_connector->panel.fixed_mode; struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode; + int ret; DRM_DEBUG_KMS("\n"); pipe_config->has_dsi_encoder = true; - if (fixed_mode) + if (fixed_mode) { intel_fixed_panel_mode(fixed_mode, adjusted_mode); + if (HAS_GMCH_DISPLAY(dev_priv)) + intel_gmch_panel_fitting(crtc, pipe_config, + intel_connector->panel.fitting_mode); + else + intel_pch_panel_fitting(crtc, pipe_config, + intel_connector->panel.fitting_mode); + } + /* DSI uses short packets for sync events, so clear mode flags for DSI */ adjusted_mode->flags = 0; @@ -311,6 +321,12 @@ static bool intel_dsi_compute_config(struct intel_encoder *encoder, pipe_config->cpu_transcoder = TRANSCODER_DSI_A; } + ret = intel_compute_dsi_pll(encoder, pipe_config); + if (ret) + return false; + + pipe_config->clock_set = true; + return true; } @@ -498,14 +514,19 @@ static void intel_dsi_pre_enable(struct intel_encoder *encoder) struct drm_device *dev = encoder->base.dev; struct drm_i915_private *dev_priv = dev->dev_private; struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base); - struct intel_crtc *intel_crtc = to_intel_crtc(encoder->base.crtc); - enum pipe pipe = intel_crtc->pipe; + struct intel_crtc *crtc = to_intel_crtc(encoder->base.crtc); enum port port; u32 tmp; DRM_DEBUG_KMS("\n"); - intel_enable_dsi_pll(encoder); + /* + * The BIOS may leave the PLL in a wonky state where it doesn't + * lock. It needs to be fully powered down to fix it. + */ + intel_disable_dsi_pll(encoder); + intel_enable_dsi_pll(encoder, crtc->config); + intel_dsi_prepare(encoder); /* Panel Enable over CRC PMIC */ @@ -515,19 +536,7 @@ static void intel_dsi_pre_enable(struct intel_encoder *encoder) msleep(intel_dsi->panel_on_delay); if (IS_VALLEYVIEW(dev) || IS_CHERRYVIEW(dev)) { - /* - * Disable DPOunit clock gating, can stall pipe - * and we need DPLL REFA always enabled - */ - tmp = I915_READ(DPLL(pipe)); - tmp |= DPLL_REF_CLK_ENABLE_VLV; - I915_WRITE(DPLL(pipe), tmp); - - /* update the hw state for DPLL */ - intel_crtc->config->dpll_hw_state.dpll = - DPLL_INTEGRATED_REF_CLK_VLV | - DPLL_REF_CLK_ENABLE_VLV | DPLL_VGA_MODE_DIS; - + /* Disable DPOunit clock gating, can stall pipe */ tmp = I915_READ(DSPCLK_GATE_D); tmp |= DPOUNIT_CLOCK_GATE_DISABLE; I915_WRITE(DSPCLK_GATE_D, tmp); @@ -679,11 +688,16 @@ static void intel_dsi_post_disable(struct intel_encoder *encoder) drm_panel_unprepare(intel_dsi->panel); msleep(intel_dsi->panel_off_delay); - msleep(intel_dsi->panel_pwr_cycle_delay); /* Panel Disable over CRC PMIC */ if (intel_dsi->gpio_panel) gpiod_set_value_cansleep(intel_dsi->gpio_panel, 0); + + /* + * FIXME As we do with eDP, just make a note of the time here + * and perform the wait before the next panel power on. + */ + msleep(intel_dsi->panel_pwr_cycle_delay); } static bool intel_dsi_get_hw_state(struct intel_encoder *encoder, @@ -716,11 +730,12 @@ static bool intel_dsi_get_hw_state(struct intel_encoder *encoder, BXT_MIPI_PORT_CTRL(port) : MIPI_PORT_CTRL(port); bool enabled = I915_READ(ctrl_reg) & DPI_ENABLE; - /* Due to some hardware limitations on BYT, MIPI Port C DPI - * Enable bit does not get set. To check whether DSI Port C - * was enabled in BIOS, check the Pipe B enable bit + /* + * Due to some hardware limitations on VLV/CHV, the DPI enable + * bit in port C control register does not get set. As a + * workaround, check pipe B conf instead. */ - if (IS_VALLEYVIEW(dev) && port == PORT_C) + if ((IS_VALLEYVIEW(dev) || IS_CHERRYVIEW(dev)) && port == PORT_C) enabled = I915_READ(PIPECONF(PIPE_B)) & PIPECONF_ENABLE; /* Try command mode if video mode not enabled */ @@ -826,13 +841,8 @@ static void intel_dsi_get_config(struct intel_encoder *encoder, if (IS_BROXTON(dev)) bxt_dsi_get_pipe_config(encoder, pipe_config); - /* - * DPLL_MD is not used in case of DSI, reading will get some default value - * set dpll_md = 0 - */ - pipe_config->dpll_hw_state.dpll_md = 0; - - pclk = intel_dsi_get_pclk(encoder, pipe_config->pipe_bpp); + pclk = intel_dsi_get_pclk(encoder, pipe_config->pipe_bpp, + pipe_config); if (!pclk) return; @@ -845,7 +855,7 @@ intel_dsi_mode_valid(struct drm_connector *connector, struct drm_display_mode *mode) { struct intel_connector *intel_connector = to_intel_connector(connector); - struct drm_display_mode *fixed_mode = intel_connector->panel.fixed_mode; + const struct drm_display_mode *fixed_mode = intel_connector->panel.fixed_mode; int max_dotclk = to_i915(connector->dev)->max_dotclk_freq; DRM_DEBUG_KMS("\n"); @@ -1183,6 +1193,48 @@ static int intel_dsi_get_modes(struct drm_connector *connector) return 1; } +static int intel_dsi_set_property(struct drm_connector *connector, + struct drm_property *property, + uint64_t val) +{ + struct drm_device *dev = connector->dev; + struct intel_connector *intel_connector = to_intel_connector(connector); + struct drm_crtc *crtc; + int ret; + + ret = drm_object_property_set_value(&connector->base, property, val); + if (ret) + return ret; + + if (property == dev->mode_config.scaling_mode_property) { + if (val == DRM_MODE_SCALE_NONE) { + DRM_DEBUG_KMS("no scaling not supported\n"); + return -EINVAL; + } + if (HAS_GMCH_DISPLAY(dev) && + val == DRM_MODE_SCALE_CENTER) { + DRM_DEBUG_KMS("centering not supported\n"); + return -EINVAL; + } + + if (intel_connector->panel.fitting_mode == val) + return 0; + + intel_connector->panel.fitting_mode = val; + } + + crtc = intel_attached_encoder(connector)->base.crtc; + if (crtc && crtc->state->enable) { + /* + * If the CRTC is enabled, the display will be changed + * according to the new panel fitting mode. + */ + intel_crtc_restore_mode(crtc); + } + + return 0; +} + static void intel_dsi_connector_destroy(struct drm_connector *connector) { struct intel_connector *intel_connector = to_intel_connector(connector); @@ -1225,11 +1277,25 @@ static const struct drm_connector_funcs intel_dsi_connector_funcs = { .detect = intel_dsi_detect, .destroy = intel_dsi_connector_destroy, .fill_modes = drm_helper_probe_single_connector_modes, + .set_property = intel_dsi_set_property, .atomic_get_property = intel_connector_atomic_get_property, .atomic_destroy_state = drm_atomic_helper_connector_destroy_state, .atomic_duplicate_state = drm_atomic_helper_connector_duplicate_state, }; +static void intel_dsi_add_properties(struct intel_connector *connector) +{ + struct drm_device *dev = connector->base.dev; + + if (connector->panel.fixed_mode) { + drm_mode_create_scaling_mode_property(dev); + drm_object_attach_property(&connector->base.base, + dev->mode_config.scaling_mode_property, + DRM_MODE_SCALE_ASPECT); + connector->panel.fitting_mode = DRM_MODE_SCALE_ASPECT; + } +} + void intel_dsi_init(struct drm_device *dev) { struct intel_dsi *intel_dsi; @@ -1353,8 +1419,6 @@ void intel_dsi_init(struct drm_device *dev) intel_connector_attach_encoder(intel_connector, intel_encoder); - drm_connector_register(connector); - drm_panel_attach(intel_dsi->panel, connector); mutex_lock(&dev->mode_config.mutex); @@ -1373,6 +1437,11 @@ void intel_dsi_init(struct drm_device *dev) } intel_panel_init(&intel_connector->panel, fixed_mode, NULL); + + intel_dsi_add_properties(intel_connector); + + drm_connector_register(connector); + intel_panel_setup_backlight(connector, INVALID_PIPE); return; diff --git a/drivers/gpu/drm/i915/intel_dsi.h b/drivers/gpu/drm/i915/intel_dsi.h index dabde19..61a6957 100644 --- a/drivers/gpu/drm/i915/intel_dsi.h +++ b/drivers/gpu/drm/i915/intel_dsi.h @@ -127,11 +127,15 @@ static inline struct intel_dsi *enc_to_intel_dsi(struct drm_encoder *encoder) } bool intel_dsi_pll_is_enabled(struct drm_i915_private *dev_priv); -extern void intel_enable_dsi_pll(struct intel_encoder *encoder); -extern void intel_disable_dsi_pll(struct intel_encoder *encoder); -extern u32 intel_dsi_get_pclk(struct intel_encoder *encoder, int pipe_bpp); -extern void intel_dsi_reset_clocks(struct intel_encoder *encoder, - enum port port); +int intel_compute_dsi_pll(struct intel_encoder *encoder, + struct intel_crtc_state *config); +void intel_enable_dsi_pll(struct intel_encoder *encoder, + const struct intel_crtc_state *config); +void intel_disable_dsi_pll(struct intel_encoder *encoder); +u32 intel_dsi_get_pclk(struct intel_encoder *encoder, int pipe_bpp, + struct intel_crtc_state *config); +void intel_dsi_reset_clocks(struct intel_encoder *encoder, + enum port port); struct drm_panel *vbt_panel_init(struct intel_dsi *intel_dsi, u16 panel_id); enum mipi_dsi_pixel_format pixel_format_from_register_bits(u32 fmt); diff --git a/drivers/gpu/drm/i915/intel_dsi_pll.c b/drivers/gpu/drm/i915/intel_dsi_pll.c index 4e53fcf..1765e6e 100644 --- a/drivers/gpu/drm/i915/intel_dsi_pll.c +++ b/drivers/gpu/drm/i915/intel_dsi_pll.c @@ -30,12 +30,7 @@ #include "i915_drv.h" #include "intel_dsi.h" -struct dsi_mnp { - u32 dsi_pll_ctrl; - u32 dsi_pll_div; -}; - -static const u32 lfsr_converts[] = { +static const u16 lfsr_converts[] = { 426, 469, 234, 373, 442, 221, 110, 311, 411, /* 62 - 70 */ 461, 486, 243, 377, 188, 350, 175, 343, 427, 213, /* 71 - 80 */ 106, 53, 282, 397, 454, 227, 113, 56, 284, 142, /* 81 - 90 */ @@ -57,7 +52,8 @@ static u32 dsi_clk_from_pclk(u32 pclk, enum mipi_dsi_pixel_format fmt, } static int dsi_calc_mnp(struct drm_i915_private *dev_priv, - struct dsi_mnp *dsi_mnp, int target_dsi_clk) + struct intel_crtc_state *config, + int target_dsi_clk) { unsigned int calc_m = 0, calc_p = 0; unsigned int m_min, m_max, p_min = 2, p_max = 6; @@ -103,8 +99,8 @@ static int dsi_calc_mnp(struct drm_i915_private *dev_priv, /* register has log2(N1), this works fine for powers of two */ n = ffs(n) - 1; m_seed = lfsr_converts[calc_m - 62]; - dsi_mnp->dsi_pll_ctrl = 1 << (DSI_PLL_P1_POST_DIV_SHIFT + calc_p - 2); - dsi_mnp->dsi_pll_div = n << DSI_PLL_N1_DIV_SHIFT | + config->dsi_pll.ctrl = 1 << (DSI_PLL_P1_POST_DIV_SHIFT + calc_p - 2); + config->dsi_pll.div = n << DSI_PLL_N1_DIV_SHIFT | m_seed << DSI_PLL_M1_DIV_SHIFT; return 0; @@ -114,54 +110,55 @@ static int dsi_calc_mnp(struct drm_i915_private *dev_priv, * XXX: The muxing and gating is hard coded for now. Need to add support for * sharing PLLs with two DSI outputs. */ -static void vlv_configure_dsi_pll(struct intel_encoder *encoder) +static int vlv_compute_dsi_pll(struct intel_encoder *encoder, + struct intel_crtc_state *config) { struct drm_i915_private *dev_priv = encoder->base.dev->dev_private; struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base); int ret; - struct dsi_mnp dsi_mnp; u32 dsi_clk; dsi_clk = dsi_clk_from_pclk(intel_dsi->pclk, intel_dsi->pixel_format, intel_dsi->lane_count); - ret = dsi_calc_mnp(dev_priv, &dsi_mnp, dsi_clk); + ret = dsi_calc_mnp(dev_priv, config, dsi_clk); if (ret) { DRM_DEBUG_KMS("dsi_calc_mnp failed\n"); - return; + return ret; } if (intel_dsi->ports & (1 << PORT_A)) - dsi_mnp.dsi_pll_ctrl |= DSI_PLL_CLK_GATE_DSI0_DSIPLL; + config->dsi_pll.ctrl |= DSI_PLL_CLK_GATE_DSI0_DSIPLL; if (intel_dsi->ports & (1 << PORT_C)) - dsi_mnp.dsi_pll_ctrl |= DSI_PLL_CLK_GATE_DSI1_DSIPLL; + config->dsi_pll.ctrl |= DSI_PLL_CLK_GATE_DSI1_DSIPLL; + + config->dsi_pll.ctrl |= DSI_PLL_VCO_EN; DRM_DEBUG_KMS("dsi pll div %08x, ctrl %08x\n", - dsi_mnp.dsi_pll_div, dsi_mnp.dsi_pll_ctrl); + config->dsi_pll.div, config->dsi_pll.ctrl); - vlv_cck_write(dev_priv, CCK_REG_DSI_PLL_CONTROL, 0); - vlv_cck_write(dev_priv, CCK_REG_DSI_PLL_DIVIDER, dsi_mnp.dsi_pll_div); - vlv_cck_write(dev_priv, CCK_REG_DSI_PLL_CONTROL, dsi_mnp.dsi_pll_ctrl); + return 0; } -static void vlv_enable_dsi_pll(struct intel_encoder *encoder) +static void vlv_enable_dsi_pll(struct intel_encoder *encoder, + const struct intel_crtc_state *config) { - struct drm_i915_private *dev_priv = encoder->base.dev->dev_private; - u32 tmp; + struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); DRM_DEBUG_KMS("\n"); mutex_lock(&dev_priv->sb_lock); - vlv_configure_dsi_pll(encoder); + vlv_cck_write(dev_priv, CCK_REG_DSI_PLL_CONTROL, 0); + vlv_cck_write(dev_priv, CCK_REG_DSI_PLL_DIVIDER, config->dsi_pll.div); + vlv_cck_write(dev_priv, CCK_REG_DSI_PLL_CONTROL, + config->dsi_pll.ctrl & ~DSI_PLL_VCO_EN); /* wait at least 0.5 us after ungating before enabling VCO */ usleep_range(1, 10); - tmp = vlv_cck_read(dev_priv, CCK_REG_DSI_PLL_CONTROL); - tmp |= DSI_PLL_VCO_EN; - vlv_cck_write(dev_priv, CCK_REG_DSI_PLL_CONTROL, tmp); + vlv_cck_write(dev_priv, CCK_REG_DSI_PLL_CONTROL, config->dsi_pll.ctrl); if (wait_for(vlv_cck_read(dev_priv, CCK_REG_DSI_PLL_CONTROL) & DSI_PLL_LOCK, 20)) { @@ -177,7 +174,7 @@ static void vlv_enable_dsi_pll(struct intel_encoder *encoder) static void vlv_disable_dsi_pll(struct intel_encoder *encoder) { - struct drm_i915_private *dev_priv = encoder->base.dev->dev_private; + struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); u32 tmp; DRM_DEBUG_KMS("\n"); @@ -224,7 +221,7 @@ static bool bxt_dsi_pll_is_enabled(struct drm_i915_private *dev_priv) static void bxt_disable_dsi_pll(struct intel_encoder *encoder) { - struct drm_i915_private *dev_priv = encoder->base.dev->dev_private; + struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); u32 val; DRM_DEBUG_KMS("\n"); @@ -251,14 +248,15 @@ static void assert_bpp_mismatch(enum mipi_dsi_pixel_format fmt, int pipe_bpp) bpp, pipe_bpp); } -static u32 vlv_dsi_get_pclk(struct intel_encoder *encoder, int pipe_bpp) +static u32 vlv_dsi_get_pclk(struct intel_encoder *encoder, int pipe_bpp, + struct intel_crtc_state *config) { - struct drm_i915_private *dev_priv = encoder->base.dev->dev_private; + struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base); u32 dsi_clock, pclk; u32 pll_ctl, pll_div; u32 m = 0, p = 0, n; - int refclk = 25000; + int refclk = IS_CHERRYVIEW(dev_priv) ? 100000 : 25000; int i; DRM_DEBUG_KMS("\n"); @@ -268,6 +266,9 @@ static u32 vlv_dsi_get_pclk(struct intel_encoder *encoder, int pipe_bpp) pll_div = vlv_cck_read(dev_priv, CCK_REG_DSI_PLL_DIVIDER); mutex_unlock(&dev_priv->sb_lock); + config->dsi_pll.ctrl = pll_ctl & ~DSI_PLL_LOCK; + config->dsi_pll.div = pll_div; + /* mask out other bits and extract the P1 divisor */ pll_ctl &= DSI_PLL_P1_POST_DIV_MASK; pll_ctl = pll_ctl >> (DSI_PLL_P1_POST_DIV_SHIFT - 2); @@ -313,7 +314,8 @@ static u32 vlv_dsi_get_pclk(struct intel_encoder *encoder, int pipe_bpp) return pclk; } -static u32 bxt_dsi_get_pclk(struct intel_encoder *encoder, int pipe_bpp) +static u32 bxt_dsi_get_pclk(struct intel_encoder *encoder, int pipe_bpp, + struct intel_crtc_state *config) { u32 pclk; u32 dsi_clk; @@ -327,15 +329,9 @@ static u32 bxt_dsi_get_pclk(struct intel_encoder *encoder, int pipe_bpp) return 0; } - dsi_ratio = I915_READ(BXT_DSI_PLL_CTL) & - BXT_DSI_PLL_RATIO_MASK; + config->dsi_pll.ctrl = I915_READ(BXT_DSI_PLL_CTL); - /* Invalid DSI ratio ? */ - if (dsi_ratio < BXT_DSI_PLL_RATIO_MIN || - dsi_ratio > BXT_DSI_PLL_RATIO_MAX) { - DRM_ERROR("Invalid DSI pll ratio(%u) programmed\n", dsi_ratio); - return 0; - } + dsi_ratio = config->dsi_pll.ctrl & BXT_DSI_PLL_RATIO_MASK; dsi_clk = (dsi_ratio * BXT_REF_CLOCK_KHZ) / 2; @@ -348,12 +344,13 @@ static u32 bxt_dsi_get_pclk(struct intel_encoder *encoder, int pipe_bpp) return pclk; } -u32 intel_dsi_get_pclk(struct intel_encoder *encoder, int pipe_bpp) +u32 intel_dsi_get_pclk(struct intel_encoder *encoder, int pipe_bpp, + struct intel_crtc_state *config) { if (IS_BROXTON(encoder->base.dev)) - return bxt_dsi_get_pclk(encoder, pipe_bpp); + return bxt_dsi_get_pclk(encoder, pipe_bpp, config); else - return vlv_dsi_get_pclk(encoder, pipe_bpp); + return vlv_dsi_get_pclk(encoder, pipe_bpp, config); } static void vlv_dsi_reset_clocks(struct intel_encoder *encoder, enum port port) @@ -370,7 +367,8 @@ static void vlv_dsi_reset_clocks(struct intel_encoder *encoder, enum port port) } /* Program BXT Mipi clocks and dividers */ -static void bxt_dsi_program_clocks(struct drm_device *dev, enum port port) +static void bxt_dsi_program_clocks(struct drm_device *dev, enum port port, + const struct intel_crtc_state *config) { struct drm_i915_private *dev_priv = dev->dev_private; u32 tmp; @@ -390,8 +388,7 @@ static void bxt_dsi_program_clocks(struct drm_device *dev, enum port port) tmp &= ~(BXT_MIPI_RX_ESCLK_LOWER_FIXDIV_MASK(port)); /* Get the current DSI rate(actual) */ - pll_ratio = I915_READ(BXT_DSI_PLL_CTL) & - BXT_DSI_PLL_RATIO_MASK; + pll_ratio = config->dsi_pll.ctrl & BXT_DSI_PLL_RATIO_MASK; dsi_rate = (BXT_REF_CLOCK_KHZ * pll_ratio) / 2; /* @@ -427,16 +424,15 @@ static void bxt_dsi_program_clocks(struct drm_device *dev, enum port port) I915_WRITE(BXT_MIPI_CLOCK_CTL, tmp); } -static bool bxt_configure_dsi_pll(struct intel_encoder *encoder) +static int bxt_compute_dsi_pll(struct intel_encoder *encoder, + struct intel_crtc_state *config) { - struct drm_i915_private *dev_priv = encoder->base.dev->dev_private; struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base); u8 dsi_ratio; u32 dsi_clk; - u32 val; dsi_clk = dsi_clk_from_pclk(intel_dsi->pclk, intel_dsi->pixel_format, - intel_dsi->lane_count); + intel_dsi->lane_count); /* * From clock diagram, to get PLL ratio divider, divide double of DSI @@ -445,9 +441,9 @@ static bool bxt_configure_dsi_pll(struct intel_encoder *encoder) */ dsi_ratio = DIV_ROUND_UP(dsi_clk * 2, BXT_REF_CLOCK_KHZ); if (dsi_ratio < BXT_DSI_PLL_RATIO_MIN || - dsi_ratio > BXT_DSI_PLL_RATIO_MAX) { + dsi_ratio > BXT_DSI_PLL_RATIO_MAX) { DRM_ERROR("Cant get a suitable ratio from DSI PLL ratios\n"); - return false; + return -ECHRNG; } /* @@ -455,27 +451,19 @@ static bool bxt_configure_dsi_pll(struct intel_encoder *encoder) * Spec says both have to be programmed, even if one is not getting * used. Configure MIPI_CLOCK_CTL dividers in modeset */ - val = I915_READ(BXT_DSI_PLL_CTL); - val &= ~BXT_DSI_PLL_PVD_RATIO_MASK; - val &= ~BXT_DSI_FREQ_SEL_MASK; - val &= ~BXT_DSI_PLL_RATIO_MASK; - val |= (dsi_ratio | BXT_DSIA_16X_BY2 | BXT_DSIC_16X_BY2); + config->dsi_pll.ctrl = dsi_ratio | BXT_DSIA_16X_BY2 | BXT_DSIC_16X_BY2; /* As per recommendation from hardware team, * Prog PVD ratio =1 if dsi ratio <= 50 */ - if (dsi_ratio <= 50) { - val &= ~BXT_DSI_PLL_PVD_RATIO_MASK; - val |= BXT_DSI_PLL_PVD_RATIO_1; - } + if (dsi_ratio <= 50) + config->dsi_pll.ctrl |= BXT_DSI_PLL_PVD_RATIO_1; - I915_WRITE(BXT_DSI_PLL_CTL, val); - POSTING_READ(BXT_DSI_PLL_CTL); - - return true; + return 0; } -static void bxt_enable_dsi_pll(struct intel_encoder *encoder) +static void bxt_enable_dsi_pll(struct intel_encoder *encoder, + const struct intel_crtc_state *config) { struct drm_i915_private *dev_priv = encoder->base.dev->dev_private; struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base); @@ -484,23 +472,13 @@ static void bxt_enable_dsi_pll(struct intel_encoder *encoder) DRM_DEBUG_KMS("\n"); - val = I915_READ(BXT_DSI_PLL_ENABLE); - - if (val & BXT_DSI_PLL_DO_ENABLE) { - WARN(1, "DSI PLL already enabled. Disabling it.\n"); - val &= ~BXT_DSI_PLL_DO_ENABLE; - I915_WRITE(BXT_DSI_PLL_ENABLE, val); - } - /* Configure PLL vales */ - if (!bxt_configure_dsi_pll(encoder)) { - DRM_ERROR("Configure DSI PLL failed, abort PLL enable\n"); - return; - } + I915_WRITE(BXT_DSI_PLL_CTL, config->dsi_pll.ctrl); + POSTING_READ(BXT_DSI_PLL_CTL); /* Program TX, RX, Dphy clocks */ for_each_dsi_port(port, intel_dsi->ports) - bxt_dsi_program_clocks(encoder->base.dev, port); + bxt_dsi_program_clocks(encoder->base.dev, port, config); /* Enable DSI PLL */ val = I915_READ(BXT_DSI_PLL_ENABLE); @@ -526,14 +504,28 @@ bool intel_dsi_pll_is_enabled(struct drm_i915_private *dev_priv) return false; } -void intel_enable_dsi_pll(struct intel_encoder *encoder) +int intel_compute_dsi_pll(struct intel_encoder *encoder, + struct intel_crtc_state *config) +{ + struct drm_device *dev = encoder->base.dev; + + if (IS_VALLEYVIEW(dev) || IS_CHERRYVIEW(dev)) + return vlv_compute_dsi_pll(encoder, config); + else if (IS_BROXTON(dev)) + return bxt_compute_dsi_pll(encoder, config); + + return -ENODEV; +} + +void intel_enable_dsi_pll(struct intel_encoder *encoder, + const struct intel_crtc_state *config) { struct drm_device *dev = encoder->base.dev; if (IS_VALLEYVIEW(dev) || IS_CHERRYVIEW(dev)) - vlv_enable_dsi_pll(encoder); + vlv_enable_dsi_pll(encoder, config); else if (IS_BROXTON(dev)) - bxt_enable_dsi_pll(encoder); + bxt_enable_dsi_pll(encoder, config); } void intel_disable_dsi_pll(struct intel_encoder *encoder) diff --git a/drivers/gpu/drm/i915/intel_guc.h b/drivers/gpu/drm/i915/intel_guc.h index 73002e9..9d79c4c 100644 --- a/drivers/gpu/drm/i915/intel_guc.h +++ b/drivers/gpu/drm/i915/intel_guc.h @@ -27,8 +27,34 @@ #include "intel_guc_fwif.h" #include "i915_guc_reg.h" +struct drm_i915_gem_request; + +/* + * This structure primarily describes the GEM object shared with the GuC. + * The GEM object is held for the entire lifetime of our interaction with + * the GuC, being allocated before the GuC is loaded with its firmware. + * Because there's no way to update the address used by the GuC after + * initialisation, the shared object must stay pinned into the GGTT as + * long as the GuC is in use. We also keep the first page (only) mapped + * into kernel address space, as it includes shared data that must be + * updated on every request submission. + * + * The single GEM object described here is actually made up of several + * separate areas, as far as the GuC is concerned. The first page (kept + * kmap'd) includes the "process decriptor" which holds sequence data for + * the doorbell, and one cacheline which actually *is* the doorbell; a + * write to this will "ring the doorbell" (i.e. send an interrupt to the + * GuC). The subsequent pages of the client object constitute the work + * queue (a circular array of work items), again described in the process + * descriptor. Work queue pages are mapped momentarily as required. + * + * Finally, we also keep a few statistics here, including the number of + * submissions to each engine, and a record of the last submission failure + * (if any). + */ struct i915_guc_client { struct drm_i915_gem_object *client_obj; + void *client_base; /* first page (only) of above */ struct intel_context *owner; struct intel_guc *guc; uint32_t priority; @@ -43,13 +69,14 @@ struct i915_guc_client { uint32_t wq_offset; uint32_t wq_size; uint32_t wq_tail; - uint32_t wq_head; + uint32_t unused; /* Was 'wq_head' */ /* GuC submission statistics & status */ uint64_t submissions[GUC_MAX_ENGINES_NUM]; uint32_t q_fail; uint32_t b_fail; int retcode; + int spare; /* pad to 32 DWords */ }; enum intel_guc_fw_status { diff --git a/drivers/gpu/drm/i915/intel_hdmi.c b/drivers/gpu/drm/i915/intel_hdmi.c index b199ede..2cdab73 100644 --- a/drivers/gpu/drm/i915/intel_hdmi.c +++ b/drivers/gpu/drm/i915/intel_hdmi.c @@ -1412,8 +1412,16 @@ intel_hdmi_detect(struct drm_connector *connector, bool force) hdmi_to_dig_port(intel_hdmi)); } - if (!live_status) - DRM_DEBUG_KMS("Live status not up!"); + if (!live_status) { + DRM_DEBUG_KMS("HDMI live status down\n"); + /* + * Live status register is not reliable on all intel platforms. + * So consider live_status only for certain platforms, for + * others, read EDID to determine presence of sink. + */ + if (INTEL_INFO(dev_priv)->gen < 7 || IS_IVYBRIDGE(dev_priv)) + live_status = true; + } intel_hdmi_unset_edid(connector); diff --git a/drivers/gpu/drm/i915/intel_i2c.c b/drivers/gpu/drm/i915/intel_i2c.c index 6dbe73e..81de230 100644 --- a/drivers/gpu/drm/i915/intel_i2c.c +++ b/drivers/gpu/drm/i915/intel_i2c.c @@ -571,15 +571,14 @@ clear_err: goto out; timeout: - DRM_INFO("GMBUS [%s] timed out, falling back to bit banging on pin %d\n", - bus->adapter.name, bus->reg0 & 0xff); + DRM_DEBUG_KMS("GMBUS [%s] timed out, falling back to bit banging on pin %d\n", + bus->adapter.name, bus->reg0 & 0xff); I915_WRITE(GMBUS0, 0); /* * Hardware may not support GMBUS over these pins? Try GPIO bitbanging * instead. Use EAGAIN to have i2c core retry. */ - bus->force_bit = 1; ret = -EAGAIN; out: @@ -597,10 +596,15 @@ gmbus_xfer(struct i2c_adapter *adapter, struct i2c_msg *msgs, int num) intel_display_power_get(dev_priv, POWER_DOMAIN_GMBUS); mutex_lock(&dev_priv->gmbus_mutex); - if (bus->force_bit) + if (bus->force_bit) { ret = i2c_bit_algo.master_xfer(adapter, msgs, num); - else + if (ret < 0) + bus->force_bit &= ~GMBUS_FORCE_BIT_RETRY; + } else { ret = do_gmbus_xfer(adapter, msgs, num); + if (ret == -EAGAIN) + bus->force_bit |= GMBUS_FORCE_BIT_RETRY; + } mutex_unlock(&dev_priv->gmbus_mutex); intel_display_power_put(dev_priv, POWER_DOMAIN_GMBUS); @@ -718,11 +722,16 @@ void intel_gmbus_set_speed(struct i2c_adapter *adapter, int speed) void intel_gmbus_force_bit(struct i2c_adapter *adapter, bool force_bit) { struct intel_gmbus *bus = to_intel_gmbus(adapter); + struct drm_i915_private *dev_priv = bus->dev_priv; + + mutex_lock(&dev_priv->gmbus_mutex); bus->force_bit += force_bit ? 1 : -1; DRM_DEBUG_KMS("%sabling bit-banging on %s. force bit now %d\n", force_bit ? "en" : "dis", adapter->name, bus->force_bit); + + mutex_unlock(&dev_priv->gmbus_mutex); } void intel_teardown_gmbus(struct drm_device *dev) diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c index 0d6dc5e..6179b59 100644 --- a/drivers/gpu/drm/i915/intel_lrc.c +++ b/drivers/gpu/drm/i915/intel_lrc.c @@ -229,9 +229,6 @@ enum { static int intel_lr_context_pin(struct intel_context *ctx, struct intel_engine_cs *engine); -static void lrc_setup_hardware_status_page(struct intel_engine_cs *engine, - struct drm_i915_gem_object *default_ctx_obj); - /** * intel_sanitize_enable_execlists() - sanitize i915.enable_execlists @@ -418,6 +415,7 @@ static void execlists_submit_requests(struct drm_i915_gem_request *rq0, struct drm_i915_gem_request *rq1) { struct drm_i915_private *dev_priv = rq0->i915; + unsigned int fw_domains = rq0->engine->fw_domains; execlists_update_context(rq0); @@ -425,11 +423,11 @@ static void execlists_submit_requests(struct drm_i915_gem_request *rq0, execlists_update_context(rq1); spin_lock_irq(&dev_priv->uncore.lock); - intel_uncore_forcewake_get__locked(dev_priv, FORCEWAKE_ALL); + intel_uncore_forcewake_get__locked(dev_priv, fw_domains); execlists_elsp_write(rq0, rq1); - intel_uncore_forcewake_put__locked(dev_priv, FORCEWAKE_ALL); + intel_uncore_forcewake_put__locked(dev_priv, fw_domains); spin_unlock_irq(&dev_priv->uncore.lock); } @@ -552,7 +550,7 @@ static void intel_lrc_irq_handler(unsigned long data) unsigned int csb_read = 0, i; unsigned int submit_contexts = 0; - intel_uncore_forcewake_get(dev_priv, FORCEWAKE_ALL); + intel_uncore_forcewake_get(dev_priv, engine->fw_domains); status_pointer = I915_READ_FW(RING_CONTEXT_STATUS_PTR(engine)); @@ -577,7 +575,7 @@ static void intel_lrc_irq_handler(unsigned long data) _MASKED_FIELD(GEN8_CSB_READ_PTR_MASK, engine->next_context_status_buffer << 8)); - intel_uncore_forcewake_put(dev_priv, FORCEWAKE_ALL); + intel_uncore_forcewake_put(dev_priv, engine->fw_domains); spin_lock(&engine->execlist_lock); @@ -892,17 +890,8 @@ static int logical_ring_prepare(struct drm_i915_gem_request *req, int bytes) */ int intel_logical_ring_begin(struct drm_i915_gem_request *req, int num_dwords) { - struct drm_i915_private *dev_priv; int ret; - WARN_ON(req == NULL); - dev_priv = req->i915; - - ret = i915_gem_check_wedge(&dev_priv->gpu_error, - dev_priv->mm.interruptible); - if (ret) - return ret; - ret = logical_ring_prepare(req, num_dwords * sizeof(uint32_t)); if (ret) return ret; @@ -1016,7 +1005,6 @@ int intel_execlists_submission(struct i915_execbuffer_params *params, trace_i915_gem_ring_dispatch(params->request, params->dispatch_flags); i915_gem_execbuffer_move_to_active(vmas, params->request); - i915_gem_execbuffer_retire_commands(params); return 0; } @@ -1057,7 +1045,7 @@ void intel_logical_ring_stop(struct intel_engine_cs *engine) return; ret = intel_engine_idle(engine); - if (ret && !i915_reset_in_progress(&to_i915(engine->dev)->gpu_error)) + if (ret) DRM_ERROR("failed to quiesce %s whilst cleaning up: %d\n", engine->name, ret); @@ -1093,8 +1081,8 @@ static int intel_lr_context_do_pin(struct intel_context *ctx, struct drm_i915_private *dev_priv = dev->dev_private; struct drm_i915_gem_object *ctx_obj = ctx->engine[engine->id].state; struct intel_ringbuffer *ringbuf = ctx->engine[engine->id].ringbuf; - struct page *lrc_state_page; - uint32_t *lrc_reg_state; + void *vaddr; + u32 *lrc_reg_state; int ret; WARN_ON(!mutex_is_locked(&engine->dev->struct_mutex)); @@ -1104,19 +1092,20 @@ static int intel_lr_context_do_pin(struct intel_context *ctx, if (ret) return ret; - lrc_state_page = i915_gem_object_get_dirty_page(ctx_obj, LRC_STATE_PN); - if (WARN_ON(!lrc_state_page)) { - ret = -ENODEV; + vaddr = i915_gem_object_pin_map(ctx_obj); + if (IS_ERR(vaddr)) { + ret = PTR_ERR(vaddr); goto unpin_ctx_obj; } + lrc_reg_state = vaddr + LRC_STATE_PN * PAGE_SIZE; + ret = intel_pin_and_map_ringbuffer_obj(engine->dev, ringbuf); if (ret) - goto unpin_ctx_obj; + goto unpin_map; ctx->engine[engine->id].lrc_vma = i915_gem_obj_to_ggtt(ctx_obj); intel_lr_context_descriptor_update(ctx, engine); - lrc_reg_state = kmap(lrc_state_page); lrc_reg_state[CTX_RING_BUFFER_START+1] = ringbuf->vma->node.start; ctx->engine[engine->id].lrc_reg_state = lrc_reg_state; ctx_obj->dirty = true; @@ -1127,6 +1116,8 @@ static int intel_lr_context_do_pin(struct intel_context *ctx, return ret; +unpin_map: + i915_gem_object_unpin_map(ctx_obj); unpin_ctx_obj: i915_gem_object_ggtt_unpin(ctx_obj); @@ -1159,7 +1150,7 @@ void intel_lr_context_unpin(struct intel_context *ctx, WARN_ON(!mutex_is_locked(&ctx->i915->dev->struct_mutex)); if (--ctx->engine[engine->id].pin_count == 0) { - kunmap(kmap_to_page(ctx->engine[engine->id].lrc_reg_state)); + i915_gem_object_unpin_map(ctx_obj); intel_unpin_ringbuffer_obj(ctx->engine[engine->id].ringbuf); i915_gem_object_ggtt_unpin(ctx_obj); ctx->engine[engine->id].lrc_vma = NULL; @@ -1579,14 +1570,22 @@ out: return ret; } +static void lrc_init_hws(struct intel_engine_cs *engine) +{ + struct drm_i915_private *dev_priv = engine->dev->dev_private; + + I915_WRITE(RING_HWS_PGA(engine->mmio_base), + (u32)engine->status_page.gfx_addr); + POSTING_READ(RING_HWS_PGA(engine->mmio_base)); +} + static int gen8_init_common_ring(struct intel_engine_cs *engine) { struct drm_device *dev = engine->dev; struct drm_i915_private *dev_priv = dev->dev_private; unsigned int next_context_status_buffer_hw; - lrc_setup_hardware_status_page(engine, - dev_priv->kernel_context->engine[engine->id].state); + lrc_init_hws(engine); I915_WRITE_IMR(engine, ~(engine->irq_enable_mask | engine->irq_keep_mask)); @@ -1625,7 +1624,7 @@ static int gen8_init_common_ring(struct intel_engine_cs *engine) intel_engine_init_hangcheck(engine); - return 0; + return intel_mocs_init_engine(engine); } static int gen8_init_render_ring(struct intel_engine_cs *engine) @@ -1945,15 +1944,18 @@ static int gen8_emit_request_render(struct drm_i915_gem_request *request) struct intel_ringbuffer *ringbuf = request->ringbuf; int ret; - ret = intel_logical_ring_begin(request, 6 + WA_TAIL_DWORDS); + ret = intel_logical_ring_begin(request, 8 + WA_TAIL_DWORDS); if (ret) return ret; + /* We're using qword write, seqno should be aligned to 8 bytes. */ + BUILD_BUG_ON(I915_GEM_HWS_INDEX & 1); + /* w/a for post sync ops following a GPGPU operation we * need a prior CS_STALL, which is emitted by the flush * following the batch. */ - intel_logical_ring_emit(ringbuf, GFX_OP_PIPE_CONTROL(5)); + intel_logical_ring_emit(ringbuf, GFX_OP_PIPE_CONTROL(6)); intel_logical_ring_emit(ringbuf, (PIPE_CONTROL_GLOBAL_GTT_IVB | PIPE_CONTROL_CS_STALL | @@ -1961,7 +1963,10 @@ static int gen8_emit_request_render(struct drm_i915_gem_request *request) intel_logical_ring_emit(ringbuf, hws_seqno_address(request->engine)); intel_logical_ring_emit(ringbuf, 0); intel_logical_ring_emit(ringbuf, i915_gem_request_get_seqno(request)); + /* We're thrashing one dword of HWS. */ + intel_logical_ring_emit(ringbuf, 0); intel_logical_ring_emit(ringbuf, MI_USER_INTERRUPT); + intel_logical_ring_emit(ringbuf, MI_NOOP); return intel_logical_ring_advance_and_submit(request); } @@ -2048,7 +2053,7 @@ void intel_logical_ring_cleanup(struct intel_engine_cs *engine) i915_gem_batch_pool_fini(&engine->batch_pool); if (engine->status_page.obj) { - kunmap(sg_page(engine->status_page.obj->pages->sgl)); + i915_gem_object_unpin_map(engine->status_page.obj); engine->status_page.obj = NULL; } @@ -2087,9 +2092,29 @@ logical_ring_default_irqs(struct intel_engine_cs *engine, unsigned shift) } static int +lrc_setup_hws(struct intel_engine_cs *engine, + struct drm_i915_gem_object *dctx_obj) +{ + void *hws; + + /* The HWSP is part of the default context object in LRC mode. */ + engine->status_page.gfx_addr = i915_gem_obj_ggtt_offset(dctx_obj) + + LRC_PPHWSP_PN * PAGE_SIZE; + hws = i915_gem_object_pin_map(dctx_obj); + if (IS_ERR(hws)) + return PTR_ERR(hws); + engine->status_page.page_addr = hws + LRC_PPHWSP_PN * PAGE_SIZE; + engine->status_page.obj = dctx_obj; + + return 0; +} + +static int logical_ring_init(struct drm_device *dev, struct intel_engine_cs *engine) { - struct intel_context *dctx = to_i915(dev)->kernel_context; + struct drm_i915_private *dev_priv = to_i915(dev); + struct intel_context *dctx = dev_priv->kernel_context; + enum forcewake_domains fw_domains; int ret; /* Intentionally left blank. */ @@ -2111,6 +2136,20 @@ logical_ring_init(struct drm_device *dev, struct intel_engine_cs *engine) logical_ring_init_platform_invariants(engine); + fw_domains = intel_uncore_forcewake_for_reg(dev_priv, + RING_ELSP(engine), + FW_REG_WRITE); + + fw_domains |= intel_uncore_forcewake_for_reg(dev_priv, + RING_CONTEXT_STATUS_PTR(engine), + FW_REG_READ | FW_REG_WRITE); + + fw_domains |= intel_uncore_forcewake_for_reg(dev_priv, + RING_CONTEXT_STATUS_BUF_BASE(engine), + FW_REG_READ); + + engine->fw_domains = fw_domains; + ret = i915_cmd_parser_init_ring(engine); if (ret) goto error; @@ -2128,6 +2167,13 @@ logical_ring_init(struct drm_device *dev, struct intel_engine_cs *engine) goto error; } + /* And setup the hardware status page. */ + ret = lrc_setup_hws(engine, dctx->engine[engine->id].state); + if (ret) { + DRM_ERROR("Failed to set up hws %s: %d\n", engine->name, ret); + goto error; + } + return 0; error: @@ -2378,15 +2424,16 @@ static u32 intel_lr_indirect_ctx_offset(struct intel_engine_cs *engine) } static int -populate_lr_context(struct intel_context *ctx, struct drm_i915_gem_object *ctx_obj, +populate_lr_context(struct intel_context *ctx, + struct drm_i915_gem_object *ctx_obj, struct intel_engine_cs *engine, struct intel_ringbuffer *ringbuf) { struct drm_device *dev = engine->dev; struct drm_i915_private *dev_priv = dev->dev_private; struct i915_hw_ppgtt *ppgtt = ctx->ppgtt; - struct page *page; - uint32_t *reg_state; + void *vaddr; + u32 *reg_state; int ret; if (!ppgtt) @@ -2398,18 +2445,17 @@ populate_lr_context(struct intel_context *ctx, struct drm_i915_gem_object *ctx_o return ret; } - ret = i915_gem_object_get_pages(ctx_obj); - if (ret) { - DRM_DEBUG_DRIVER("Could not get object pages\n"); + vaddr = i915_gem_object_pin_map(ctx_obj); + if (IS_ERR(vaddr)) { + ret = PTR_ERR(vaddr); + DRM_DEBUG_DRIVER("Could not map object pages! (%d)\n", ret); return ret; } - - i915_gem_object_pin_pages(ctx_obj); + ctx_obj->dirty = true; /* The second page of the context object contains some fields which must * be set up prior to the first execution. */ - page = i915_gem_object_get_dirty_page(ctx_obj, LRC_STATE_PN); - reg_state = kmap_atomic(page); + reg_state = vaddr + LRC_STATE_PN * PAGE_SIZE; /* A context is actually a big batch buffer with several MI_LOAD_REGISTER_IMM * commands followed by (reg, value) pairs. The values we are setting here are @@ -2514,8 +2560,7 @@ populate_lr_context(struct intel_context *ctx, struct drm_i915_gem_object *ctx_o make_rpcs(dev)); } - kunmap_atomic(reg_state); - i915_gem_object_unpin_pages(ctx_obj); + i915_gem_object_unpin_map(ctx_obj); return 0; } @@ -2542,6 +2587,7 @@ void intel_lr_context_free(struct intel_context *ctx) if (ctx == ctx->i915->kernel_context) { intel_unpin_ringbuffer_obj(ringbuf); i915_gem_object_ggtt_unpin(ctx_obj); + i915_gem_object_unpin_map(ctx_obj); } WARN_ON(ctx->engine[i].pin_count); @@ -2588,24 +2634,6 @@ uint32_t intel_lr_context_size(struct intel_engine_cs *engine) return ret; } -static void lrc_setup_hardware_status_page(struct intel_engine_cs *engine, - struct drm_i915_gem_object *default_ctx_obj) -{ - struct drm_i915_private *dev_priv = engine->dev->dev_private; - struct page *page; - - /* The HWSP is part of the default context object in LRC mode. */ - engine->status_page.gfx_addr = i915_gem_obj_ggtt_offset(default_ctx_obj) - + LRC_PPHWSP_PN * PAGE_SIZE; - page = i915_gem_object_get_page(default_ctx_obj, LRC_PPHWSP_PN); - engine->status_page.page_addr = kmap(page); - engine->status_page.obj = default_ctx_obj; - - I915_WRITE(RING_HWS_PGA(engine->mmio_base), - (u32)engine->status_page.gfx_addr); - POSTING_READ(RING_HWS_PGA(engine->mmio_base)); -} - /** * intel_lr_context_deferred_alloc() - create the LRC specific bits of a context * @ctx: LR context to create. @@ -2669,13 +2697,12 @@ int intel_lr_context_deferred_alloc(struct intel_context *ctx, } ret = engine->init_context(req); + i915_add_request_no_flush(req); if (ret) { DRM_ERROR("ring init context: %d\n", ret); - i915_gem_request_cancel(req); goto error_ringbuf; } - i915_add_request_no_flush(req); } return 0; @@ -2688,10 +2715,9 @@ error_deref_obj: return ret; } -void intel_lr_context_reset(struct drm_device *dev, - struct intel_context *ctx) +void intel_lr_context_reset(struct drm_i915_private *dev_priv, + struct intel_context *ctx) { - struct drm_i915_private *dev_priv = dev->dev_private; struct intel_engine_cs *engine; for_each_engine(engine, dev_priv) { @@ -2699,23 +2725,23 @@ void intel_lr_context_reset(struct drm_device *dev, ctx->engine[engine->id].state; struct intel_ringbuffer *ringbuf = ctx->engine[engine->id].ringbuf; + void *vaddr; uint32_t *reg_state; - struct page *page; if (!ctx_obj) continue; - if (i915_gem_object_get_pages(ctx_obj)) { - WARN(1, "Failed get_pages for context obj\n"); + vaddr = i915_gem_object_pin_map(ctx_obj); + if (WARN_ON(IS_ERR(vaddr))) continue; - } - page = i915_gem_object_get_dirty_page(ctx_obj, LRC_STATE_PN); - reg_state = kmap_atomic(page); + + reg_state = vaddr + LRC_STATE_PN * PAGE_SIZE; + ctx_obj->dirty = true; reg_state[CTX_RING_HEAD+1] = 0; reg_state[CTX_RING_TAIL+1] = 0; - kunmap_atomic(reg_state); + i915_gem_object_unpin_map(ctx_obj); ringbuf->head = 0; ringbuf->tail = 0; diff --git a/drivers/gpu/drm/i915/intel_lrc.h b/drivers/gpu/drm/i915/intel_lrc.h index 0b0853e..461f1ef 100644 --- a/drivers/gpu/drm/i915/intel_lrc.h +++ b/drivers/gpu/drm/i915/intel_lrc.h @@ -24,6 +24,8 @@ #ifndef _INTEL_LRC_H_ #define _INTEL_LRC_H_ +#include "intel_ringbuffer.h" + #define GEN8_LR_CONTEXT_ALIGN 4096 /* Execlists regs */ @@ -34,6 +36,7 @@ #define CTX_CTRL_INHIBIT_SYN_CTX_SWITCH (1 << 3) #define CTX_CTRL_ENGINE_CTX_RESTORE_INHIBIT (1 << 0) #define CTX_CTRL_RS_CTX_ENABLE (1 << 1) +#define RING_CONTEXT_STATUS_BUF_BASE(ring) _MMIO((ring)->mmio_base + 0x370) #define RING_CONTEXT_STATUS_BUF_LO(ring, i) _MMIO((ring)->mmio_base + 0x370 + (i) * 8) #define RING_CONTEXT_STATUS_BUF_HI(ring, i) _MMIO((ring)->mmio_base + 0x370 + (i) * 8 + 4) #define RING_CONTEXT_STATUS_PTR(ring) _MMIO((ring)->mmio_base + 0x3a0) @@ -103,8 +106,11 @@ int intel_lr_context_deferred_alloc(struct intel_context *ctx, struct intel_engine_cs *engine); void intel_lr_context_unpin(struct intel_context *ctx, struct intel_engine_cs *engine); -void intel_lr_context_reset(struct drm_device *dev, - struct intel_context *ctx); + +struct drm_i915_private; + +void intel_lr_context_reset(struct drm_i915_private *dev_priv, + struct intel_context *ctx); uint64_t intel_lr_context_descriptor(struct intel_context *ctx, struct intel_engine_cs *engine); diff --git a/drivers/gpu/drm/i915/intel_mocs.c b/drivers/gpu/drm/i915/intel_mocs.c index 7c7ac0a..23b8545 100644 --- a/drivers/gpu/drm/i915/intel_mocs.c +++ b/drivers/gpu/drm/i915/intel_mocs.c @@ -128,9 +128,9 @@ static const struct drm_i915_mocs_entry broxton_mocs_table[] = { /** * get_mocs_settings() - * @dev: DRM device. + * @dev_priv: i915 device. * @table: Output table that will be made to point at appropriate - * MOCS values for the device. + * MOCS values for the device. * * This function will return the values of the MOCS table that needs to * be programmed for the platform. It will return the values that need @@ -138,21 +138,21 @@ static const struct drm_i915_mocs_entry broxton_mocs_table[] = { * * Return: true if there are applicable MOCS settings for the device. */ -static bool get_mocs_settings(struct drm_device *dev, +static bool get_mocs_settings(struct drm_i915_private *dev_priv, struct drm_i915_mocs_table *table) { bool result = false; - if (IS_SKYLAKE(dev) || IS_KABYLAKE(dev)) { + if (IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv)) { table->size = ARRAY_SIZE(skylake_mocs_table); table->table = skylake_mocs_table; result = true; - } else if (IS_BROXTON(dev)) { + } else if (IS_BROXTON(dev_priv)) { table->size = ARRAY_SIZE(broxton_mocs_table); table->table = broxton_mocs_table; result = true; } else { - WARN_ONCE(INTEL_INFO(dev)->gen >= 9, + WARN_ONCE(INTEL_INFO(dev_priv)->gen >= 9, "Platform that should have a MOCS table does not.\n"); } @@ -179,10 +179,49 @@ static i915_reg_t mocs_register(enum intel_engine_id ring, int index) } /** + * intel_mocs_init_engine() - emit the mocs control table + * @engine: The engine for whom to emit the registers. + * + * This function simply emits a MI_LOAD_REGISTER_IMM command for the + * given table starting at the given address. + * + * Return: 0 on success, otherwise the error status. + */ +int intel_mocs_init_engine(struct intel_engine_cs *engine) +{ + struct drm_i915_private *dev_priv = to_i915(engine->dev); + struct drm_i915_mocs_table table; + unsigned int index; + + if (!get_mocs_settings(dev_priv, &table)) + return 0; + + if (WARN_ON(table.size > GEN9_NUM_MOCS_ENTRIES)) + return -ENODEV; + + for (index = 0; index < table.size; index++) + I915_WRITE(mocs_register(engine->id, index), + table.table[index].control_value); + + /* + * Ok, now set the unused entries to uncached. These entries + * are officially undefined and no contract for the contents + * and settings is given for these entries. + * + * Entry 0 in the table is uncached - so we are just writing + * that value to all the used entries. + */ + for (; index < GEN9_NUM_MOCS_ENTRIES; index++) + I915_WRITE(mocs_register(engine->id, index), + table.table[0].control_value); + + return 0; +} + +/** * emit_mocs_control_table() - emit the mocs control table * @req: Request to set up the MOCS table for. * @table: The values to program into the control regs. - * @ring: The engine for whom to emit the registers. * * This function simply emits a MI_LOAD_REGISTER_IMM command for the * given table starting at the given address. @@ -190,10 +229,10 @@ static i915_reg_t mocs_register(enum intel_engine_id ring, int index) * Return: 0 on success, otherwise the error status. */ static int emit_mocs_control_table(struct drm_i915_gem_request *req, - const struct drm_i915_mocs_table *table, - enum intel_engine_id ring) + const struct drm_i915_mocs_table *table) { struct intel_ringbuffer *ringbuf = req->ringbuf; + enum intel_engine_id engine = req->engine->id; unsigned int index; int ret; @@ -210,7 +249,8 @@ static int emit_mocs_control_table(struct drm_i915_gem_request *req, MI_LOAD_REGISTER_IMM(GEN9_NUM_MOCS_ENTRIES)); for (index = 0; index < table->size; index++) { - intel_logical_ring_emit_reg(ringbuf, mocs_register(ring, index)); + intel_logical_ring_emit_reg(ringbuf, + mocs_register(engine, index)); intel_logical_ring_emit(ringbuf, table->table[index].control_value); } @@ -224,8 +264,10 @@ static int emit_mocs_control_table(struct drm_i915_gem_request *req, * that value to all the used entries. */ for (; index < GEN9_NUM_MOCS_ENTRIES; index++) { - intel_logical_ring_emit_reg(ringbuf, mocs_register(ring, index)); - intel_logical_ring_emit(ringbuf, table->table[0].control_value); + intel_logical_ring_emit_reg(ringbuf, + mocs_register(engine, index)); + intel_logical_ring_emit(ringbuf, + table->table[0].control_value); } intel_logical_ring_emit(ringbuf, MI_NOOP); @@ -234,6 +276,14 @@ static int emit_mocs_control_table(struct drm_i915_gem_request *req, return 0; } +static inline u32 l3cc_combine(const struct drm_i915_mocs_table *table, + u16 low, + u16 high) +{ + return table->table[low].l3cc_value | + table->table[high].l3cc_value << 16; +} + /** * emit_mocs_l3cc_table() - emit the mocs control table * @req: Request to set up the MOCS table for. @@ -249,11 +299,7 @@ static int emit_mocs_l3cc_table(struct drm_i915_gem_request *req, const struct drm_i915_mocs_table *table) { struct intel_ringbuffer *ringbuf = req->ringbuf; - unsigned int count; unsigned int i; - u32 value; - u32 filler = (table->table[0].l3cc_value & 0xffff) | - ((table->table[0].l3cc_value & 0xffff) << 16); int ret; if (WARN_ON(table->size > GEN9_NUM_MOCS_ENTRIES)) @@ -268,20 +314,18 @@ static int emit_mocs_l3cc_table(struct drm_i915_gem_request *req, intel_logical_ring_emit(ringbuf, MI_LOAD_REGISTER_IMM(GEN9_NUM_MOCS_ENTRIES / 2)); - for (i = 0, count = 0; i < table->size / 2; i++, count += 2) { - value = (table->table[count].l3cc_value & 0xffff) | - ((table->table[count + 1].l3cc_value & 0xffff) << 16); - + for (i = 0; i < table->size/2; i++) { intel_logical_ring_emit_reg(ringbuf, GEN9_LNCFCMOCS(i)); - intel_logical_ring_emit(ringbuf, value); + intel_logical_ring_emit(ringbuf, + l3cc_combine(table, 2*i, 2*i+1)); } if (table->size & 0x01) { /* Odd table size - 1 left over */ - value = (table->table[count].l3cc_value & 0xffff) | - ((table->table[0].l3cc_value & 0xffff) << 16); - } else - value = filler; + intel_logical_ring_emit_reg(ringbuf, GEN9_LNCFCMOCS(i)); + intel_logical_ring_emit(ringbuf, l3cc_combine(table, 2*i, 0)); + i++; + } /* * Now set the rest of the table to uncached - use entry 0 as @@ -290,9 +334,7 @@ static int emit_mocs_l3cc_table(struct drm_i915_gem_request *req, */ for (; i < GEN9_NUM_MOCS_ENTRIES / 2; i++) { intel_logical_ring_emit_reg(ringbuf, GEN9_LNCFCMOCS(i)); - intel_logical_ring_emit(ringbuf, value); - - value = filler; + intel_logical_ring_emit(ringbuf, l3cc_combine(table, 0, 0)); } intel_logical_ring_emit(ringbuf, MI_NOOP); @@ -302,6 +344,47 @@ static int emit_mocs_l3cc_table(struct drm_i915_gem_request *req, } /** + * intel_mocs_init_l3cc_table() - program the mocs control table + * @dev: The the device to be programmed. + * + * This function simply programs the mocs registers for the given table + * starting at the given address. This register set is programmed in pairs. + * + * These registers may get programmed more than once, it is simpler to + * re-program 32 registers than maintain the state of when they were programmed. + * We are always reprogramming with the same values and this only on context + * start. + * + * Return: Nothing. + */ +void intel_mocs_init_l3cc_table(struct drm_device *dev) +{ + struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_mocs_table table; + unsigned int i; + + if (!get_mocs_settings(dev_priv, &table)) + return; + + for (i = 0; i < table.size/2; i++) + I915_WRITE(GEN9_LNCFCMOCS(i), l3cc_combine(&table, 2*i, 2*i+1)); + + /* Odd table size - 1 left over */ + if (table.size & 0x01) { + I915_WRITE(GEN9_LNCFCMOCS(i), l3cc_combine(&table, 2*i, 0)); + i++; + } + + /* + * Now set the rest of the table to uncached - use entry 0 as + * this will be uncached. Leave the last pair as initialised as + * they are reserved by the hardware. + */ + for (; i < (GEN9_NUM_MOCS_ENTRIES / 2); i++) + I915_WRITE(GEN9_LNCFCMOCS(i), l3cc_combine(&table, 0, 0)); +} + +/** * intel_rcs_context_init_mocs() - program the MOCS register. * @req: Request to set up the MOCS tables for. * @@ -322,17 +405,11 @@ int intel_rcs_context_init_mocs(struct drm_i915_gem_request *req) struct drm_i915_mocs_table t; int ret; - if (get_mocs_settings(req->engine->dev, &t)) { - struct drm_i915_private *dev_priv = req->i915; - struct intel_engine_cs *engine; - enum intel_engine_id id; - - /* Program the control registers */ - for_each_engine_id(engine, dev_priv, id) { - ret = emit_mocs_control_table(req, &t, id); - if (ret) - return ret; - } + if (get_mocs_settings(req->i915, &t)) { + /* Program the RCS control registers */ + ret = emit_mocs_control_table(req, &t); + if (ret) + return ret; /* Now program the l3cc registers */ ret = emit_mocs_l3cc_table(req, &t); diff --git a/drivers/gpu/drm/i915/intel_mocs.h b/drivers/gpu/drm/i915/intel_mocs.h index 76e45b1..4640299 100644 --- a/drivers/gpu/drm/i915/intel_mocs.h +++ b/drivers/gpu/drm/i915/intel_mocs.h @@ -53,5 +53,7 @@ #include "i915_drv.h" int intel_rcs_context_init_mocs(struct drm_i915_gem_request *req); +void intel_mocs_init_l3cc_table(struct drm_device *dev); +int intel_mocs_init_engine(struct intel_engine_cs *ring); #endif diff --git a/drivers/gpu/drm/i915/intel_opregion.c b/drivers/gpu/drm/i915/intel_opregion.c index c15718b..99e2603 100644 --- a/drivers/gpu/drm/i915/intel_opregion.c +++ b/drivers/gpu/drm/i915/intel_opregion.c @@ -34,12 +34,6 @@ #include "i915_drv.h" #include "intel_drv.h" -#define PCI_ASLE 0xe4 -#define PCI_ASLS 0xfc -#define PCI_SWSCI 0xe8 -#define PCI_SWSCI_SCISEL (1 << 15) -#define PCI_SWSCI_GSSCIE (1 << 0) - #define OPREGION_HEADER_OFFSET 0 #define OPREGION_ACPI_OFFSET 0x100 #define ACPI_CLID 0x01ac /* current lid state indicator */ @@ -246,13 +240,12 @@ struct opregion_asle_ext { #define MAX_DSLP 1500 -#ifdef CONFIG_ACPI static int swsci(struct drm_device *dev, u32 function, u32 parm, u32 *parm_out) { struct drm_i915_private *dev_priv = dev->dev_private; struct opregion_swsci *swsci = dev_priv->opregion.swsci; u32 main_function, sub_function, scic; - u16 pci_swsci; + u16 swsci_val; u32 dslp; if (!swsci) @@ -300,16 +293,16 @@ static int swsci(struct drm_device *dev, u32 function, u32 parm, u32 *parm_out) swsci->scic = scic; /* Ensure SCI event is selected and event trigger is cleared. */ - pci_read_config_word(dev->pdev, PCI_SWSCI, &pci_swsci); - if (!(pci_swsci & PCI_SWSCI_SCISEL) || (pci_swsci & PCI_SWSCI_GSSCIE)) { - pci_swsci |= PCI_SWSCI_SCISEL; - pci_swsci &= ~PCI_SWSCI_GSSCIE; - pci_write_config_word(dev->pdev, PCI_SWSCI, pci_swsci); + pci_read_config_word(dev->pdev, SWSCI, &swsci_val); + if (!(swsci_val & SWSCI_SCISEL) || (swsci_val & SWSCI_GSSCIE)) { + swsci_val |= SWSCI_SCISEL; + swsci_val &= ~SWSCI_GSSCIE; + pci_write_config_word(dev->pdev, SWSCI, swsci_val); } /* Use event trigger to tell bios to check the mail. */ - pci_swsci |= PCI_SWSCI_GSSCIE; - pci_write_config_word(dev->pdev, PCI_SWSCI, pci_swsci); + swsci_val |= SWSCI_GSSCIE; + pci_write_config_word(dev->pdev, SWSCI, swsci_val); /* Poll for the result. */ #define C (((scic = swsci->scic) & SWSCI_SCIC_INDICATOR) == 0) @@ -905,9 +898,6 @@ static void swsci_setup(struct drm_device *dev) opregion->swsci_gbda_sub_functions, opregion->swsci_sbcb_sub_functions); } -#else /* CONFIG_ACPI */ -static inline void swsci_setup(struct drm_device *dev) {} -#endif /* CONFIG_ACPI */ static int intel_no_opregion_vbt_callback(const struct dmi_system_id *id) { @@ -943,16 +933,14 @@ int intel_opregion_setup(struct drm_device *dev) BUILD_BUG_ON(sizeof(struct opregion_asle) != 0x100); BUILD_BUG_ON(sizeof(struct opregion_asle_ext) != 0x400); - pci_read_config_dword(dev->pdev, PCI_ASLS, &asls); + pci_read_config_dword(dev->pdev, ASLS, &asls); DRM_DEBUG_DRIVER("graphic opregion physical addr: 0x%x\n", asls); if (asls == 0) { DRM_DEBUG_DRIVER("ACPI OpRegion not supported!\n"); return -ENOTSUPP; } -#ifdef CONFIG_ACPI INIT_WORK(&opregion->asle_work, asle_work); -#endif base = memremap(asls, OPREGION_SIZE, MEMREMAP_WB); if (!base) @@ -1024,3 +1012,31 @@ err_out: memunmap(base); return err; } + +int +intel_opregion_get_panel_type(struct drm_device *dev) +{ + u32 panel_details; + int ret; + + ret = swsci(dev, SWSCI_GBDA_PANEL_DETAILS, 0x0, &panel_details); + if (ret) { + DRM_DEBUG_KMS("Failed to get panel details from OpRegion (%d)\n", + ret); + return ret; + } + + ret = (panel_details >> 8) & 0xff; + if (ret > 0x10) { + DRM_DEBUG_KMS("Invalid OpRegion panel type 0x%x\n", ret); + return -EINVAL; + } + + /* fall back to VBT panel type? */ + if (ret == 0x0) { + DRM_DEBUG_KMS("No panel type in OpRegion\n"); + return -ENODEV; + } + + return ret - 1; +} diff --git a/drivers/gpu/drm/i915/intel_overlay.c b/drivers/gpu/drm/i915/intel_overlay.c index 6694e92..bcc3b6a 100644 --- a/drivers/gpu/drm/i915/intel_overlay.c +++ b/drivers/gpu/drm/i915/intel_overlay.c @@ -247,7 +247,7 @@ static int intel_overlay_on(struct intel_overlay *overlay) ret = intel_ring_begin(req, 4); if (ret) { - i915_gem_request_cancel(req); + i915_add_request_no_flush(req); return ret; } @@ -290,7 +290,7 @@ static int intel_overlay_continue(struct intel_overlay *overlay, ret = intel_ring_begin(req, 2); if (ret) { - i915_gem_request_cancel(req); + i915_add_request_no_flush(req); return ret; } @@ -356,7 +356,7 @@ static int intel_overlay_off(struct intel_overlay *overlay) ret = intel_ring_begin(req, 6); if (ret) { - i915_gem_request_cancel(req); + i915_add_request_no_flush(req); return ret; } @@ -431,7 +431,7 @@ static int intel_overlay_release_old_vid(struct intel_overlay *overlay) ret = intel_ring_begin(req, 2); if (ret) { - i915_gem_request_cancel(req); + i915_add_request_no_flush(req); return ret; } diff --git a/drivers/gpu/drm/i915/intel_panel.c b/drivers/gpu/drm/i915/intel_panel.c index 8c8996f..a078876 100644 --- a/drivers/gpu/drm/i915/intel_panel.c +++ b/drivers/gpu/drm/i915/intel_panel.c @@ -504,7 +504,7 @@ static u32 i9xx_get_backlight(struct intel_connector *connector) if (panel->backlight.combination_mode) { u8 lbpc; - pci_read_config_byte(dev_priv->dev->pdev, PCI_LBPC, &lbpc); + pci_read_config_byte(dev_priv->dev->pdev, LBPC, &lbpc); val *= lbpc; } @@ -592,7 +592,7 @@ static void i9xx_set_backlight(struct intel_connector *connector, u32 level) lbpc = level * 0xfe / panel->backlight.max + 1; level /= lbpc; - pci_write_config_byte(dev_priv->dev->pdev, PCI_LBPC, lbpc); + pci_write_config_byte(dev_priv->dev->pdev, LBPC, lbpc); } if (IS_GEN4(dev_priv)) { diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 43b24a1..695a464 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -2483,7 +2483,7 @@ static void ilk_wm_merge(struct drm_device *dev, /* ILK/SNB/IVB: LP1+ watermarks only w/ single pipe */ if ((INTEL_INFO(dev)->gen <= 6 || IS_IVYBRIDGE(dev)) && config->num_pipes_active > 1) - return; + last_enabled_level = 0; /* ILK: FBC WM must be disabled always */ merged->fbc_wm_enabled = INTEL_INFO(dev)->gen >= 6; @@ -4587,7 +4587,7 @@ void intel_set_rps(struct drm_device *dev, u8 val) gen6_set_rps(dev, val); } -static void gen9_disable_rps(struct drm_device *dev) +static void gen9_disable_rc6(struct drm_device *dev) { struct drm_i915_private *dev_priv = dev->dev_private; @@ -4595,12 +4595,20 @@ static void gen9_disable_rps(struct drm_device *dev) I915_WRITE(GEN9_PG_ENABLE, 0); } +static void gen9_disable_rps(struct drm_device *dev) +{ + struct drm_i915_private *dev_priv = dev->dev_private; + + I915_WRITE(GEN6_RP_CONTROL, 0); +} + static void gen6_disable_rps(struct drm_device *dev) { struct drm_i915_private *dev_priv = dev->dev_private; I915_WRITE(GEN6_RC_CONTROL, 0); I915_WRITE(GEN6_RPNSWREQ, 1 << 31); + I915_WRITE(GEN6_RP_CONTROL, 0); } static void cherryview_disable_rps(struct drm_device *dev) @@ -4804,6 +4812,16 @@ static void gen9_enable_rps(struct drm_device *dev) /* WaGsvDisableTurbo: Workaround to disable turbo on BXT A* */ if (IS_BXT_REVID(dev, 0, BXT_REVID_A1)) { + /* + * BIOS could leave the Hw Turbo enabled, so need to explicitly + * clear out the Control register just to avoid inconsitency + * with debugfs interface, which will show Turbo as enabled + * only and that is not expected by the User after adding the + * WaGsvDisableTurbo. Apart from this there is no problem even + * if the Turbo is left enabled in the Control register, as the + * Up/Down interrupts would remain masked. + */ + gen9_disable_rps(dev); intel_uncore_forcewake_put(dev_priv, FORCEWAKE_ALL); return; } @@ -4997,7 +5015,8 @@ static void gen6_enable_rps(struct drm_device *dev) I915_WRITE(GEN6_RC_STATE, 0); /* Clear the DBG now so we don't confuse earlier errors */ - if ((gtfifodbg = I915_READ(GTFIFODBG))) { + gtfifodbg = I915_READ(GTFIFODBG); + if (gtfifodbg) { DRM_ERROR("GT fifo had a previous error %x\n", gtfifodbg); I915_WRITE(GTFIFODBG, gtfifodbg); } @@ -5528,7 +5547,8 @@ static void cherryview_enable_rps(struct drm_device *dev) WARN_ON(!mutex_is_locked(&dev_priv->rps.hw_lock)); - gtfifodbg = I915_READ(GTFIFODBG); + gtfifodbg = I915_READ(GTFIFODBG) & ~(GT_FIFO_SBDEDICATE_FREE_ENTRY_CHV | + GT_FIFO_FREE_ENTRIES_CHV); if (gtfifodbg) { DRM_DEBUG_DRIVER("GT fifo had a previous error %x\n", gtfifodbg); @@ -5627,7 +5647,8 @@ static void valleyview_enable_rps(struct drm_device *dev) valleyview_check_pctx(dev_priv); - if ((gtfifodbg = I915_READ(GTFIFODBG))) { + gtfifodbg = I915_READ(GTFIFODBG); + if (gtfifodbg) { DRM_DEBUG_DRIVER("GT fifo had a previous error %x\n", gtfifodbg); I915_WRITE(GTFIFODBG, gtfifodbg); @@ -6265,9 +6286,10 @@ void intel_disable_gt_powersave(struct drm_device *dev) intel_suspend_gt_powersave(dev); mutex_lock(&dev_priv->rps.hw_lock); - if (INTEL_INFO(dev)->gen >= 9) + if (INTEL_INFO(dev)->gen >= 9) { + gen9_disable_rc6(dev); gen9_disable_rps(dev); - else if (IS_CHERRYVIEW(dev)) + } else if (IS_CHERRYVIEW(dev)) cherryview_disable_rps(dev); else if (IS_VALLEYVIEW(dev)) valleyview_disable_rps(dev); @@ -6882,23 +6904,10 @@ static void ivybridge_init_clock_gating(struct drm_device *dev) gen6_check_mch_setup(dev); } -static void vlv_init_display_clock_gating(struct drm_i915_private *dev_priv) -{ - I915_WRITE(DSPCLK_GATE_D, VRHUNIT_CLOCK_GATE_DISABLE); - - /* - * Disable trickle feed and enable pnd deadline calculation - */ - I915_WRITE(MI_ARB_VLV, MI_ARB_DISPLAY_TRICKLE_FEED_DISABLE); - I915_WRITE(CBR1_VLV, 0); -} - static void valleyview_init_clock_gating(struct drm_device *dev) { struct drm_i915_private *dev_priv = dev->dev_private; - vlv_init_display_clock_gating(dev_priv); - /* WaDisableEarlyCull:vlv */ I915_WRITE(_3D_CHICKEN3, _MASKED_BIT_ENABLE(_3D_CHICKEN_SF_DISABLE_OBJEND_CULL)); @@ -6981,8 +6990,6 @@ static void cherryview_init_clock_gating(struct drm_device *dev) { struct drm_i915_private *dev_priv = dev->dev_private; - vlv_init_display_clock_gating(dev_priv); - /* WaVSRefCountFullforceMissDisable:chv */ /* WaDSRefCountFullforceMissDisable:chv */ I915_WRITE(GEN7_FF_THREAD_MODE, diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c index 41b604e..245386e 100644 --- a/drivers/gpu/drm/i915/intel_ringbuffer.c +++ b/drivers/gpu/drm/i915/intel_ringbuffer.c @@ -959,9 +959,10 @@ static int gen9_init_workarounds(struct intel_engine_cs *engine) } /* WaEnableYV12BugFixInHalfSliceChicken7:skl,bxt */ - if (IS_SKL_REVID(dev, SKL_REVID_C0, REVID_FOREVER) || IS_BROXTON(dev)) - WA_SET_BIT_MASKED(GEN9_HALF_SLICE_CHICKEN7, - GEN9_ENABLE_YV12_BUGFIX); + /* WaEnableSamplerGPGPUPreemptionSupport:skl,bxt */ + WA_SET_BIT_MASKED(GEN9_HALF_SLICE_CHICKEN7, + GEN9_ENABLE_YV12_BUGFIX | + GEN9_ENABLE_GPGPU_PREEMPTION); /* Wa4x4STCOptimizationDisable:skl,bxt */ /* WaDisablePartialResolveInVc:skl,bxt */ @@ -980,7 +981,7 @@ static int gen9_init_workarounds(struct intel_engine_cs *engine) /* WaForceContextSaveRestoreNonCoherent:skl,bxt */ tmp = HDC_FORCE_CONTEXT_SAVE_RESTORE_NON_COHERENT; - if (IS_SKL_REVID(dev, SKL_REVID_F0, SKL_REVID_F0) || + if (IS_SKL_REVID(dev, SKL_REVID_F0, REVID_FOREVER) || IS_BXT_REVID(dev, BXT_REVID_B0, REVID_FOREVER)) tmp |= HDC_FORCE_CSR_NON_COHERENT_OVR_DISABLE; WA_SET_BIT_MASKED(HDC_CHICKEN0, tmp); @@ -1097,7 +1098,8 @@ static int skl_init_workarounds(struct intel_engine_cs *engine) WA_SET_BIT_MASKED(HIZ_CHICKEN, BDW_HIZ_POWER_COMPILER_CLOCK_GATING_DISABLE); - if (IS_SKL_REVID(dev, 0, SKL_REVID_F0)) { + /* This is tied to WaForceContextSaveRestoreNonCoherent */ + if (IS_SKL_REVID(dev, 0, REVID_FOREVER)) { /* *Use Force Non-Coherent whenever executing a 3D context. This * is a workaround for a possible hang in the unlikely event @@ -2086,6 +2088,7 @@ void intel_unpin_ringbuffer_obj(struct intel_ringbuffer *ringbuf) i915_gem_object_unpin_map(ringbuf->obj); else iounmap(ringbuf->virtual_start); + ringbuf->virtual_start = NULL; ringbuf->vma = NULL; i915_gem_object_ggtt_unpin(ringbuf->obj); } @@ -2096,10 +2099,13 @@ int intel_pin_and_map_ringbuffer_obj(struct drm_device *dev, struct drm_i915_private *dev_priv = to_i915(dev); struct i915_ggtt *ggtt = &dev_priv->ggtt; struct drm_i915_gem_object *obj = ringbuf->obj; + /* Ring wraparound at offset 0 sometimes hangs. No idea why. */ + unsigned flags = PIN_OFFSET_BIAS | 4096; + void *addr; int ret; if (HAS_LLC(dev_priv) && !obj->stolen) { - ret = i915_gem_obj_ggtt_pin(obj, PAGE_SIZE, 0); + ret = i915_gem_obj_ggtt_pin(obj, PAGE_SIZE, flags); if (ret) return ret; @@ -2107,13 +2113,14 @@ int intel_pin_and_map_ringbuffer_obj(struct drm_device *dev, if (ret) goto err_unpin; - ringbuf->virtual_start = i915_gem_object_pin_map(obj); - if (ringbuf->virtual_start == NULL) { - ret = -ENOMEM; + addr = i915_gem_object_pin_map(obj); + if (IS_ERR(addr)) { + ret = PTR_ERR(addr); goto err_unpin; } } else { - ret = i915_gem_obj_ggtt_pin(obj, PAGE_SIZE, PIN_MAPPABLE); + ret = i915_gem_obj_ggtt_pin(obj, PAGE_SIZE, + flags | PIN_MAPPABLE); if (ret) return ret; @@ -2124,14 +2131,15 @@ int intel_pin_and_map_ringbuffer_obj(struct drm_device *dev, /* Access through the GTT requires the device to be awake. */ assert_rpm_wakelock_held(dev_priv); - ringbuf->virtual_start = ioremap_wc(ggtt->mappable_base + - i915_gem_obj_ggtt_offset(obj), ringbuf->size); - if (ringbuf->virtual_start == NULL) { + addr = ioremap_wc(ggtt->mappable_base + + i915_gem_obj_ggtt_offset(obj), ringbuf->size); + if (addr == NULL) { ret = -ENOMEM; goto err_unpin; } } + ringbuf->virtual_start = addr; ringbuf->vma = i915_gem_obj_to_ggtt(obj); return 0; @@ -2363,8 +2371,7 @@ int intel_engine_idle(struct intel_engine_cs *engine) /* Make sure we do not trigger any retires */ return __i915_wait_request(req, - atomic_read(&to_i915(engine->dev)->gpu_error.reset_counter), - to_i915(engine->dev)->mm.interruptible, + req->i915->mm.interruptible, NULL, NULL); } @@ -2486,19 +2493,9 @@ static int __intel_ring_prepare(struct intel_engine_cs *engine, int bytes) int intel_ring_begin(struct drm_i915_gem_request *req, int num_dwords) { - struct intel_engine_cs *engine; - struct drm_i915_private *dev_priv; + struct intel_engine_cs *engine = req->engine; int ret; - WARN_ON(req == NULL); - engine = req->engine; - dev_priv = req->i915; - - ret = i915_gem_check_wedge(&dev_priv->gpu_error, - dev_priv->mm.interruptible); - if (ret) - return ret; - ret = __intel_ring_prepare(engine, num_dwords * sizeof(uint32_t)); if (ret) return ret; @@ -3189,7 +3186,7 @@ intel_stop_engine(struct intel_engine_cs *engine) return; ret = intel_engine_idle(engine); - if (ret && !i915_reset_in_progress(&to_i915(engine->dev)->gpu_error)) + if (ret) DRM_ERROR("failed to quiesce %s whilst cleaning up: %d\n", engine->name, ret); diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h index 78dc468..2ade194 100644 --- a/drivers/gpu/drm/i915/intel_ringbuffer.h +++ b/drivers/gpu/drm/i915/intel_ringbuffer.h @@ -270,6 +270,7 @@ struct intel_engine_cs { spinlock_t execlist_lock; /* used inside tasklet, use spin_lock_bh */ struct list_head execlist_queue; struct list_head execlist_retired_req_list; + unsigned int fw_domains; unsigned int next_context_status_buffer; unsigned int idle_lite_restore_wa; bool disable_lite_restore_wa; diff --git a/drivers/gpu/drm/i915/intel_runtime_pm.c b/drivers/gpu/drm/i915/intel_runtime_pm.c index 80e8bd4..7fb1da4 100644 --- a/drivers/gpu/drm/i915/intel_runtime_pm.c +++ b/drivers/gpu/drm/i915/intel_runtime_pm.c @@ -397,11 +397,6 @@ static void hsw_set_power_well(struct drm_i915_private *dev_priv, BIT(POWER_DOMAIN_MODESET) | \ BIT(POWER_DOMAIN_AUX_A) | \ BIT(POWER_DOMAIN_INIT)) -#define SKL_DISPLAY_ALWAYS_ON_POWER_DOMAINS ( \ - (POWER_DOMAIN_MASK & ~( \ - SKL_DISPLAY_POWERWELL_2_POWER_DOMAINS | \ - SKL_DISPLAY_DC_OFF_POWER_DOMAINS)) | \ - BIT(POWER_DOMAIN_INIT)) #define BXT_DISPLAY_POWERWELL_2_POWER_DOMAINS ( \ BIT(POWER_DOMAIN_TRANSCODER_A) | \ @@ -419,39 +414,21 @@ static void hsw_set_power_well(struct drm_i915_private *dev_priv, BIT(POWER_DOMAIN_VGA) | \ BIT(POWER_DOMAIN_GMBUS) | \ BIT(POWER_DOMAIN_INIT)) -#define BXT_DISPLAY_POWERWELL_1_POWER_DOMAINS ( \ - BXT_DISPLAY_POWERWELL_2_POWER_DOMAINS | \ - BIT(POWER_DOMAIN_PIPE_A) | \ - BIT(POWER_DOMAIN_TRANSCODER_EDP) | \ - BIT(POWER_DOMAIN_TRANSCODER_DSI_A) | \ - BIT(POWER_DOMAIN_TRANSCODER_DSI_C) | \ - BIT(POWER_DOMAIN_PIPE_A_PANEL_FITTER) | \ - BIT(POWER_DOMAIN_PORT_DDI_A_LANES) | \ - BIT(POWER_DOMAIN_PORT_DSI) | \ - BIT(POWER_DOMAIN_AUX_A) | \ - BIT(POWER_DOMAIN_PLLS) | \ - BIT(POWER_DOMAIN_INIT)) #define BXT_DISPLAY_DC_OFF_POWER_DOMAINS ( \ BXT_DISPLAY_POWERWELL_2_POWER_DOMAINS | \ BIT(POWER_DOMAIN_MODESET) | \ BIT(POWER_DOMAIN_AUX_A) | \ BIT(POWER_DOMAIN_INIT)) -#define BXT_DISPLAY_ALWAYS_ON_POWER_DOMAINS ( \ - (POWER_DOMAIN_MASK & ~(BXT_DISPLAY_POWERWELL_1_POWER_DOMAINS | \ - BXT_DISPLAY_POWERWELL_2_POWER_DOMAINS)) | \ - BIT(POWER_DOMAIN_INIT)) static void assert_can_enable_dc9(struct drm_i915_private *dev_priv) { - struct drm_device *dev = dev_priv->dev; - - WARN(!IS_BROXTON(dev), "Platform doesn't support DC9.\n"); - WARN((I915_READ(DC_STATE_EN) & DC_STATE_EN_DC9), - "DC9 already programmed to be enabled.\n"); - WARN(I915_READ(DC_STATE_EN) & DC_STATE_EN_UPTO_DC5, - "DC5 still not disabled to enable DC9.\n"); - WARN(I915_READ(HSW_PWR_WELL_DRIVER), "Power well on.\n"); - WARN(intel_irqs_enabled(dev_priv), "Interrupts not disabled yet.\n"); + WARN_ONCE((I915_READ(DC_STATE_EN) & DC_STATE_EN_DC9), + "DC9 already programmed to be enabled.\n"); + WARN_ONCE(I915_READ(DC_STATE_EN) & DC_STATE_EN_UPTO_DC5, + "DC5 still not disabled to enable DC9.\n"); + WARN_ONCE(I915_READ(HSW_PWR_WELL_DRIVER), "Power well on.\n"); + WARN_ONCE(intel_irqs_enabled(dev_priv), + "Interrupts not disabled yet.\n"); /* * TODO: check for the following to verify the conditions to enter DC9 @@ -464,9 +441,10 @@ static void assert_can_enable_dc9(struct drm_i915_private *dev_priv) static void assert_can_disable_dc9(struct drm_i915_private *dev_priv) { - WARN(intel_irqs_enabled(dev_priv), "Interrupts not disabled yet.\n"); - WARN(I915_READ(DC_STATE_EN) & DC_STATE_EN_UPTO_DC5, - "DC5 still not disabled.\n"); + WARN_ONCE(intel_irqs_enabled(dev_priv), + "Interrupts not disabled yet.\n"); + WARN_ONCE(I915_READ(DC_STATE_EN) & DC_STATE_EN_UPTO_DC5, + "DC5 still not disabled.\n"); /* * TODO: check for the following to verify DC9 state was indeed @@ -514,10 +492,9 @@ static void gen9_write_dc_state(struct drm_i915_private *dev_priv, state, rewrites); } -static void gen9_set_dc_state(struct drm_i915_private *dev_priv, uint32_t state) +static u32 gen9_dc_mask(struct drm_i915_private *dev_priv) { - uint32_t val; - uint32_t mask; + u32 mask; mask = DC_STATE_EN_UPTO_DC5; if (IS_BROXTON(dev_priv)) @@ -525,10 +502,30 @@ static void gen9_set_dc_state(struct drm_i915_private *dev_priv, uint32_t state) else mask |= DC_STATE_EN_UPTO_DC6; + return mask; +} + +void gen9_sanitize_dc_state(struct drm_i915_private *dev_priv) +{ + u32 val; + + val = I915_READ(DC_STATE_EN) & gen9_dc_mask(dev_priv); + + DRM_DEBUG_KMS("Resetting DC state tracking from %02x to %02x\n", + dev_priv->csr.dc_state, val); + dev_priv->csr.dc_state = val; +} + +static void gen9_set_dc_state(struct drm_i915_private *dev_priv, uint32_t state) +{ + uint32_t val; + uint32_t mask; + if (WARN_ON_ONCE(state & ~dev_priv->csr.allowed_dc_mask)) state &= dev_priv->csr.allowed_dc_mask; val = I915_READ(DC_STATE_EN); + mask = gen9_dc_mask(dev_priv); DRM_DEBUG_KMS("Setting DC state from %02x to %02x\n", val & mask, state); @@ -573,13 +570,9 @@ static void assert_csr_loaded(struct drm_i915_private *dev_priv) static void assert_can_enable_dc5(struct drm_i915_private *dev_priv) { - struct drm_device *dev = dev_priv->dev; bool pg2_enabled = intel_display_power_well_is_enabled(dev_priv, SKL_DISP_PW_2); - WARN_ONCE(!IS_SKYLAKE(dev) && !IS_KABYLAKE(dev), - "Platform doesn't support DC5.\n"); - WARN_ONCE(!HAS_RUNTIME_PM(dev), "Runtime PM not enabled.\n"); WARN_ONCE(pg2_enabled, "PG2 not disabled to enable DC5.\n"); WARN_ONCE((I915_READ(DC_STATE_EN) & DC_STATE_EN_UPTO_DC5), @@ -589,7 +582,7 @@ static void assert_can_enable_dc5(struct drm_i915_private *dev_priv) assert_csr_loaded(dev_priv); } -static void gen9_enable_dc5(struct drm_i915_private *dev_priv) +void gen9_enable_dc5(struct drm_i915_private *dev_priv) { assert_can_enable_dc5(dev_priv); @@ -600,11 +593,6 @@ static void gen9_enable_dc5(struct drm_i915_private *dev_priv) static void assert_can_enable_dc6(struct drm_i915_private *dev_priv) { - struct drm_device *dev = dev_priv->dev; - - WARN_ONCE(!IS_SKYLAKE(dev) && !IS_KABYLAKE(dev), - "Platform doesn't support DC6.\n"); - WARN_ONCE(!HAS_RUNTIME_PM(dev), "Runtime PM not enabled.\n"); WARN_ONCE(I915_READ(UTIL_PIN_CTL) & UTIL_PIN_ENABLE, "Backlight is not disabled.\n"); WARN_ONCE((I915_READ(DC_STATE_EN) & DC_STATE_EN_UPTO_DC6), @@ -630,6 +618,45 @@ void skl_disable_dc6(struct drm_i915_private *dev_priv) gen9_set_dc_state(dev_priv, DC_STATE_DISABLE); } +static void +gen9_sanitize_power_well_requests(struct drm_i915_private *dev_priv, + struct i915_power_well *power_well) +{ + enum skl_disp_power_wells power_well_id = power_well->data; + u32 val; + u32 mask; + + mask = SKL_POWER_WELL_REQ(power_well_id); + + val = I915_READ(HSW_PWR_WELL_KVMR); + if (WARN_ONCE(val & mask, "Clearing unexpected KVMR request for %s\n", + power_well->name)) + I915_WRITE(HSW_PWR_WELL_KVMR, val & ~mask); + + val = I915_READ(HSW_PWR_WELL_BIOS); + val |= I915_READ(HSW_PWR_WELL_DEBUG); + + if (!(val & mask)) + return; + + /* + * DMC is known to force on the request bits for power well 1 on SKL + * and BXT and the misc IO power well on SKL but we don't expect any + * other request bits to be set, so WARN for those. + */ + if (power_well_id == SKL_DISP_PW_1 || + ((IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv)) && + power_well_id == SKL_DISP_PW_MISC_IO)) + DRM_DEBUG_DRIVER("Clearing auxiliary requests for %s forced on " + "by DMC\n", power_well->name); + else + WARN_ONCE(1, "Clearing unexpected auxiliary requests for %s\n", + power_well->name); + + I915_WRITE(HSW_PWR_WELL_BIOS, val & ~mask); + I915_WRITE(HSW_PWR_WELL_DEBUG, val & ~mask); +} + static void skl_set_power_well(struct drm_i915_private *dev_priv, struct i915_power_well *power_well, bool enable) { @@ -684,10 +711,6 @@ static void skl_set_power_well(struct drm_i915_private *dev_priv, if (!is_enabled) { DRM_DEBUG_KMS("Enabling %s\n", power_well->name); - if (wait_for((I915_READ(HSW_PWR_WELL_DRIVER) & - state_mask), 1)) - DRM_ERROR("%s enable timeout\n", - power_well->name); check_fuse_status = true; } } else { @@ -696,8 +719,16 @@ static void skl_set_power_well(struct drm_i915_private *dev_priv, POSTING_READ(HSW_PWR_WELL_DRIVER); DRM_DEBUG_KMS("Disabling %s\n", power_well->name); } + + if (IS_GEN9(dev_priv)) + gen9_sanitize_power_well_requests(dev_priv, power_well); } + if (wait_for(!!(I915_READ(HSW_PWR_WELL_DRIVER) & state_mask) == enable, + 1)) + DRM_ERROR("%s %s timeout\n", + power_well->name, enable ? "enable" : "disable"); + if (check_fuse_status) { if (power_well->data == SKL_DISP_PW_1) { if (wait_for((I915_READ(SKL_FUSE_STATUS) & @@ -779,11 +810,19 @@ static void gen9_dc_off_power_well_enable(struct drm_i915_private *dev_priv, struct i915_power_well *power_well) { gen9_set_dc_state(dev_priv, DC_STATE_DISABLE); + + if (IS_BROXTON(dev_priv)) { + broxton_cdclk_verify_state(dev_priv); + broxton_ddi_phy_verify_state(dev_priv); + } } static void gen9_dc_off_power_well_disable(struct drm_i915_private *dev_priv, struct i915_power_well *power_well) { + if (!dev_priv->csr.dmc_payload) + return; + if (dev_priv->csr.allowed_dc_mask & DC_STATE_EN_UPTO_DC6) skl_enable_dc6(dev_priv); else if (dev_priv->csr.allowed_dc_mask & DC_STATE_EN_UPTO_DC5) @@ -900,6 +939,17 @@ static bool vlv_power_well_enabled(struct drm_i915_private *dev_priv, return enabled; } +static void vlv_init_display_clock_gating(struct drm_i915_private *dev_priv) +{ + I915_WRITE(DSPCLK_GATE_D, VRHUNIT_CLOCK_GATE_DISABLE); + + /* + * Disable trickle feed and enable pnd deadline calculation + */ + I915_WRITE(MI_ARB_VLV, MI_ARB_DISPLAY_TRICKLE_FEED_DISABLE); + I915_WRITE(CBR1_VLV, 0); +} + static void vlv_display_power_well_init(struct drm_i915_private *dev_priv) { enum pipe pipe; @@ -922,6 +972,8 @@ static void vlv_display_power_well_init(struct drm_i915_private *dev_priv) I915_WRITE(DPLL(pipe), val); } + vlv_init_display_clock_gating(dev_priv); + spin_lock_irq(&dev_priv->irq_lock); valleyview_enable_display_irqs(dev_priv); spin_unlock_irq(&dev_priv->irq_lock); @@ -1560,34 +1612,56 @@ void intel_display_power_put(struct drm_i915_private *dev_priv, intel_runtime_pm_put(dev_priv); } -#define HSW_ALWAYS_ON_POWER_DOMAINS ( \ - BIT(POWER_DOMAIN_PIPE_A) | \ - BIT(POWER_DOMAIN_TRANSCODER_EDP) | \ - BIT(POWER_DOMAIN_PORT_DDI_A_LANES) | \ +#define HSW_DISPLAY_POWER_DOMAINS ( \ + BIT(POWER_DOMAIN_PIPE_B) | \ + BIT(POWER_DOMAIN_PIPE_C) | \ + BIT(POWER_DOMAIN_PIPE_A_PANEL_FITTER) | \ + BIT(POWER_DOMAIN_PIPE_B_PANEL_FITTER) | \ + BIT(POWER_DOMAIN_PIPE_C_PANEL_FITTER) | \ + BIT(POWER_DOMAIN_TRANSCODER_A) | \ + BIT(POWER_DOMAIN_TRANSCODER_B) | \ + BIT(POWER_DOMAIN_TRANSCODER_C) | \ BIT(POWER_DOMAIN_PORT_DDI_B_LANES) | \ BIT(POWER_DOMAIN_PORT_DDI_C_LANES) | \ BIT(POWER_DOMAIN_PORT_DDI_D_LANES) | \ - BIT(POWER_DOMAIN_PORT_CRT) | \ - BIT(POWER_DOMAIN_PLLS) | \ - BIT(POWER_DOMAIN_AUX_A) | \ - BIT(POWER_DOMAIN_AUX_B) | \ - BIT(POWER_DOMAIN_AUX_C) | \ - BIT(POWER_DOMAIN_AUX_D) | \ - BIT(POWER_DOMAIN_GMBUS) | \ - BIT(POWER_DOMAIN_INIT)) -#define HSW_DISPLAY_POWER_DOMAINS ( \ - (POWER_DOMAIN_MASK & ~HSW_ALWAYS_ON_POWER_DOMAINS) | \ + BIT(POWER_DOMAIN_PORT_CRT) | /* DDI E */ \ + BIT(POWER_DOMAIN_VGA) | \ + BIT(POWER_DOMAIN_AUDIO) | \ BIT(POWER_DOMAIN_INIT)) -#define BDW_ALWAYS_ON_POWER_DOMAINS ( \ - HSW_ALWAYS_ON_POWER_DOMAINS | \ - BIT(POWER_DOMAIN_PIPE_A_PANEL_FITTER)) -#define BDW_DISPLAY_POWER_DOMAINS ( \ - (POWER_DOMAIN_MASK & ~BDW_ALWAYS_ON_POWER_DOMAINS) | \ +#define BDW_DISPLAY_POWER_DOMAINS ( \ + BIT(POWER_DOMAIN_PIPE_B) | \ + BIT(POWER_DOMAIN_PIPE_C) | \ + BIT(POWER_DOMAIN_PIPE_B_PANEL_FITTER) | \ + BIT(POWER_DOMAIN_PIPE_C_PANEL_FITTER) | \ + BIT(POWER_DOMAIN_TRANSCODER_A) | \ + BIT(POWER_DOMAIN_TRANSCODER_B) | \ + BIT(POWER_DOMAIN_TRANSCODER_C) | \ + BIT(POWER_DOMAIN_PORT_DDI_B_LANES) | \ + BIT(POWER_DOMAIN_PORT_DDI_C_LANES) | \ + BIT(POWER_DOMAIN_PORT_DDI_D_LANES) | \ + BIT(POWER_DOMAIN_PORT_CRT) | /* DDI E */ \ + BIT(POWER_DOMAIN_VGA) | \ + BIT(POWER_DOMAIN_AUDIO) | \ BIT(POWER_DOMAIN_INIT)) -#define VLV_ALWAYS_ON_POWER_DOMAINS BIT(POWER_DOMAIN_INIT) -#define VLV_DISPLAY_POWER_DOMAINS POWER_DOMAIN_MASK +#define VLV_DISPLAY_POWER_DOMAINS ( \ + BIT(POWER_DOMAIN_PIPE_A) | \ + BIT(POWER_DOMAIN_PIPE_B) | \ + BIT(POWER_DOMAIN_PIPE_A_PANEL_FITTER) | \ + BIT(POWER_DOMAIN_PIPE_B_PANEL_FITTER) | \ + BIT(POWER_DOMAIN_TRANSCODER_A) | \ + BIT(POWER_DOMAIN_TRANSCODER_B) | \ + BIT(POWER_DOMAIN_PORT_DDI_B_LANES) | \ + BIT(POWER_DOMAIN_PORT_DDI_C_LANES) | \ + BIT(POWER_DOMAIN_PORT_DSI) | \ + BIT(POWER_DOMAIN_PORT_CRT) | \ + BIT(POWER_DOMAIN_VGA) | \ + BIT(POWER_DOMAIN_AUDIO) | \ + BIT(POWER_DOMAIN_AUX_B) | \ + BIT(POWER_DOMAIN_AUX_C) | \ + BIT(POWER_DOMAIN_GMBUS) | \ + BIT(POWER_DOMAIN_INIT)) #define VLV_DPIO_CMN_BC_POWER_DOMAINS ( \ BIT(POWER_DOMAIN_PORT_DDI_B_LANES) | \ @@ -1617,6 +1691,28 @@ void intel_display_power_put(struct drm_i915_private *dev_priv, BIT(POWER_DOMAIN_AUX_C) | \ BIT(POWER_DOMAIN_INIT)) +#define CHV_DISPLAY_POWER_DOMAINS ( \ + BIT(POWER_DOMAIN_PIPE_A) | \ + BIT(POWER_DOMAIN_PIPE_B) | \ + BIT(POWER_DOMAIN_PIPE_C) | \ + BIT(POWER_DOMAIN_PIPE_A_PANEL_FITTER) | \ + BIT(POWER_DOMAIN_PIPE_B_PANEL_FITTER) | \ + BIT(POWER_DOMAIN_PIPE_C_PANEL_FITTER) | \ + BIT(POWER_DOMAIN_TRANSCODER_A) | \ + BIT(POWER_DOMAIN_TRANSCODER_B) | \ + BIT(POWER_DOMAIN_TRANSCODER_C) | \ + BIT(POWER_DOMAIN_PORT_DDI_B_LANES) | \ + BIT(POWER_DOMAIN_PORT_DDI_C_LANES) | \ + BIT(POWER_DOMAIN_PORT_DDI_D_LANES) | \ + BIT(POWER_DOMAIN_PORT_DSI) | \ + BIT(POWER_DOMAIN_VGA) | \ + BIT(POWER_DOMAIN_AUDIO) | \ + BIT(POWER_DOMAIN_AUX_B) | \ + BIT(POWER_DOMAIN_AUX_C) | \ + BIT(POWER_DOMAIN_AUX_D) | \ + BIT(POWER_DOMAIN_GMBUS) | \ + BIT(POWER_DOMAIN_INIT)) + #define CHV_DPIO_CMN_BC_POWER_DOMAINS ( \ BIT(POWER_DOMAIN_PORT_DDI_B_LANES) | \ BIT(POWER_DOMAIN_PORT_DDI_C_LANES) | \ @@ -1684,7 +1780,7 @@ static struct i915_power_well hsw_power_wells[] = { { .name = "always-on", .always_on = 1, - .domains = HSW_ALWAYS_ON_POWER_DOMAINS, + .domains = POWER_DOMAIN_MASK, .ops = &i9xx_always_on_power_well_ops, }, { @@ -1698,7 +1794,7 @@ static struct i915_power_well bdw_power_wells[] = { { .name = "always-on", .always_on = 1, - .domains = BDW_ALWAYS_ON_POWER_DOMAINS, + .domains = POWER_DOMAIN_MASK, .ops = &i9xx_always_on_power_well_ops, }, { @@ -1733,7 +1829,7 @@ static struct i915_power_well vlv_power_wells[] = { { .name = "always-on", .always_on = 1, - .domains = VLV_ALWAYS_ON_POWER_DOMAINS, + .domains = POWER_DOMAIN_MASK, .ops = &i9xx_always_on_power_well_ops, .data = PUNIT_POWER_WELL_ALWAYS_ON, }, @@ -1791,7 +1887,7 @@ static struct i915_power_well chv_power_wells[] = { { .name = "always-on", .always_on = 1, - .domains = VLV_ALWAYS_ON_POWER_DOMAINS, + .domains = POWER_DOMAIN_MASK, .ops = &i9xx_always_on_power_well_ops, }, { @@ -1801,7 +1897,7 @@ static struct i915_power_well chv_power_wells[] = { * power wells don't actually exist. Pipe A power well is * required for any pipe to work. */ - .domains = VLV_DISPLAY_POWER_DOMAINS, + .domains = CHV_DISPLAY_POWER_DOMAINS, .data = PIPE_A, .ops = &chv_pipe_power_well_ops, }, @@ -1835,7 +1931,7 @@ static struct i915_power_well skl_power_wells[] = { { .name = "always-on", .always_on = 1, - .domains = SKL_DISPLAY_ALWAYS_ON_POWER_DOMAINS, + .domains = POWER_DOMAIN_MASK, .ops = &i9xx_always_on_power_well_ops, .data = SKL_DISP_PW_ALWAYS_ON, }, @@ -1891,44 +1987,16 @@ static struct i915_power_well skl_power_wells[] = { }, }; -void skl_pw1_misc_io_init(struct drm_i915_private *dev_priv) -{ - struct i915_power_well *well; - - if (!(IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv))) - return; - - well = lookup_power_well(dev_priv, SKL_DISP_PW_1); - intel_power_well_enable(dev_priv, well); - - well = lookup_power_well(dev_priv, SKL_DISP_PW_MISC_IO); - intel_power_well_enable(dev_priv, well); -} - -void skl_pw1_misc_io_fini(struct drm_i915_private *dev_priv) -{ - struct i915_power_well *well; - - if (!(IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv))) - return; - - well = lookup_power_well(dev_priv, SKL_DISP_PW_1); - intel_power_well_disable(dev_priv, well); - - well = lookup_power_well(dev_priv, SKL_DISP_PW_MISC_IO); - intel_power_well_disable(dev_priv, well); -} - static struct i915_power_well bxt_power_wells[] = { { .name = "always-on", .always_on = 1, - .domains = BXT_DISPLAY_ALWAYS_ON_POWER_DOMAINS, + .domains = POWER_DOMAIN_MASK, .ops = &i9xx_always_on_power_well_ops, }, { .name = "power well 1", - .domains = BXT_DISPLAY_POWERWELL_1_POWER_DOMAINS, + .domains = 0, .ops = &skl_power_well_ops, .data = SKL_DISP_PW_1, }, @@ -1953,11 +2021,6 @@ sanitize_disable_power_well_option(const struct drm_i915_private *dev_priv, if (disable_power_well >= 0) return !!disable_power_well; - if (IS_BROXTON(dev_priv)) { - DRM_DEBUG_KMS("Disabling display power well support\n"); - return 0; - } - return 1; } @@ -2109,9 +2172,10 @@ static void intel_power_domains_sync_hw(struct drm_i915_private *dev_priv) } static void skl_display_core_init(struct drm_i915_private *dev_priv, - bool resume) + bool resume) { struct i915_power_domains *power_domains = &dev_priv->power_domains; + struct i915_power_well *well; uint32_t val; gen9_set_dc_state(dev_priv, DC_STATE_DISABLE); @@ -2122,7 +2186,13 @@ static void skl_display_core_init(struct drm_i915_private *dev_priv, /* enable PG1 and Misc I/O */ mutex_lock(&power_domains->lock); - skl_pw1_misc_io_init(dev_priv); + + well = lookup_power_well(dev_priv, SKL_DISP_PW_1); + intel_power_well_enable(dev_priv, well); + + well = lookup_power_well(dev_priv, SKL_DISP_PW_MISC_IO); + intel_power_well_enable(dev_priv, well); + mutex_unlock(&power_domains->lock); if (!resume) @@ -2137,6 +2207,7 @@ static void skl_display_core_init(struct drm_i915_private *dev_priv, static void skl_display_core_uninit(struct drm_i915_private *dev_priv) { struct i915_power_domains *power_domains = &dev_priv->power_domains; + struct i915_power_well *well; gen9_set_dc_state(dev_priv, DC_STATE_DISABLE); @@ -2144,8 +2215,73 @@ static void skl_display_core_uninit(struct drm_i915_private *dev_priv) /* The spec doesn't call for removing the reset handshake flag */ /* disable PG1 and Misc I/O */ + mutex_lock(&power_domains->lock); - skl_pw1_misc_io_fini(dev_priv); + + well = lookup_power_well(dev_priv, SKL_DISP_PW_MISC_IO); + intel_power_well_disable(dev_priv, well); + + well = lookup_power_well(dev_priv, SKL_DISP_PW_1); + intel_power_well_disable(dev_priv, well); + + mutex_unlock(&power_domains->lock); +} + +void bxt_display_core_init(struct drm_i915_private *dev_priv, + bool resume) +{ + struct i915_power_domains *power_domains = &dev_priv->power_domains; + struct i915_power_well *well; + uint32_t val; + + gen9_set_dc_state(dev_priv, DC_STATE_DISABLE); + + /* + * NDE_RSTWRN_OPT RST PCH Handshake En must always be 0b on BXT + * or else the reset will hang because there is no PCH to respond. + * Move the handshake programming to initialization sequence. + * Previously was left up to BIOS. + */ + val = I915_READ(HSW_NDE_RSTWRN_OPT); + val &= ~RESET_PCH_HANDSHAKE_ENABLE; + I915_WRITE(HSW_NDE_RSTWRN_OPT, val); + + /* Enable PG1 */ + mutex_lock(&power_domains->lock); + + well = lookup_power_well(dev_priv, SKL_DISP_PW_1); + intel_power_well_enable(dev_priv, well); + + mutex_unlock(&power_domains->lock); + + broxton_init_cdclk(dev_priv); + broxton_ddi_phy_init(dev_priv); + + broxton_cdclk_verify_state(dev_priv); + broxton_ddi_phy_verify_state(dev_priv); + + if (resume && dev_priv->csr.dmc_payload) + intel_csr_load_program(dev_priv); +} + +void bxt_display_core_uninit(struct drm_i915_private *dev_priv) +{ + struct i915_power_domains *power_domains = &dev_priv->power_domains; + struct i915_power_well *well; + + gen9_set_dc_state(dev_priv, DC_STATE_DISABLE); + + broxton_ddi_phy_uninit(dev_priv); + broxton_uninit_cdclk(dev_priv); + + /* The spec doesn't call for removing the reset handshake flag */ + + /* Disable PG1 */ + mutex_lock(&power_domains->lock); + + well = lookup_power_well(dev_priv, SKL_DISP_PW_1); + intel_power_well_disable(dev_priv, well); + mutex_unlock(&power_domains->lock); } @@ -2280,6 +2416,8 @@ void intel_power_domains_init_hw(struct drm_i915_private *dev_priv, bool resume) if (IS_SKYLAKE(dev) || IS_KABYLAKE(dev)) { skl_display_core_init(dev_priv, resume); + } else if (IS_BROXTON(dev)) { + bxt_display_core_init(dev_priv, resume); } else if (IS_CHERRYVIEW(dev)) { mutex_lock(&power_domains->lock); chv_phy_control_init(dev_priv); @@ -2317,6 +2455,8 @@ void intel_power_domains_suspend(struct drm_i915_private *dev_priv) if (IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv)) skl_display_core_uninit(dev_priv); + else if (IS_BROXTON(dev_priv)) + bxt_display_core_uninit(dev_priv); } /** diff --git a/drivers/gpu/drm/i915/intel_uncore.c b/drivers/gpu/drm/i915/intel_uncore.c index ac2ac07..4f1dfe6 100644 --- a/drivers/gpu/drm/i915/intel_uncore.c +++ b/drivers/gpu/drm/i915/intel_uncore.c @@ -60,7 +60,11 @@ fw_domain_reset(const struct intel_uncore_forcewake_domain *d) static inline void fw_domain_arm_timer(struct intel_uncore_forcewake_domain *d) { - mod_timer_pinned(&d->timer, jiffies + 1); + d->wake_count++; + hrtimer_start_range_ns(&d->timer, + ktime_set(0, NSEC_PER_MSEC), + NSEC_PER_MSEC, + HRTIMER_MODE_REL); } static inline void @@ -107,22 +111,22 @@ static void fw_domains_get(struct drm_i915_private *dev_priv, enum forcewake_domains fw_domains) { struct intel_uncore_forcewake_domain *d; - enum forcewake_domain_id id; - for_each_fw_domain_mask(d, fw_domains, dev_priv, id) { + for_each_fw_domain_masked(d, fw_domains, dev_priv) { fw_domain_wait_ack_clear(d); fw_domain_get(d); - fw_domain_wait_ack(d); } + + for_each_fw_domain_masked(d, fw_domains, dev_priv) + fw_domain_wait_ack(d); } static void fw_domains_put(struct drm_i915_private *dev_priv, enum forcewake_domains fw_domains) { struct intel_uncore_forcewake_domain *d; - enum forcewake_domain_id id; - for_each_fw_domain_mask(d, fw_domains, dev_priv, id) { + for_each_fw_domain_masked(d, fw_domains, dev_priv) { fw_domain_put(d); fw_domain_posting_read(d); } @@ -132,10 +136,9 @@ static void fw_domains_posting_read(struct drm_i915_private *dev_priv) { struct intel_uncore_forcewake_domain *d; - enum forcewake_domain_id id; /* No need to do for all, just do for first found */ - for_each_fw_domain(d, dev_priv, id) { + for_each_fw_domain(d, dev_priv) { fw_domain_posting_read(d); break; } @@ -145,12 +148,11 @@ static void fw_domains_reset(struct drm_i915_private *dev_priv, enum forcewake_domains fw_domains) { struct intel_uncore_forcewake_domain *d; - enum forcewake_domain_id id; if (dev_priv->uncore.fw_domains == 0) return; - for_each_fw_domain_mask(d, fw_domains, dev_priv, id) + for_each_fw_domain_masked(d, fw_domains, dev_priv) fw_domain_reset(d); fw_domains_posting_read(dev_priv); @@ -224,9 +226,11 @@ static int __gen6_gt_wait_for_fifo(struct drm_i915_private *dev_priv) return ret; } -static void intel_uncore_fw_release_timer(unsigned long arg) +static enum hrtimer_restart +intel_uncore_fw_release_timer(struct hrtimer *timer) { - struct intel_uncore_forcewake_domain *domain = (void *)arg; + struct intel_uncore_forcewake_domain *domain = + container_of(timer, struct intel_uncore_forcewake_domain, timer); unsigned long irqflags; assert_rpm_device_not_suspended(domain->i915); @@ -240,6 +244,8 @@ static void intel_uncore_fw_release_timer(unsigned long arg) 1 << domain->id); spin_unlock_irqrestore(&domain->i915->uncore.lock, irqflags); + + return HRTIMER_NORESTART; } void intel_uncore_forcewake_reset(struct drm_device *dev, bool restore) @@ -248,7 +254,6 @@ void intel_uncore_forcewake_reset(struct drm_device *dev, bool restore) unsigned long irqflags; struct intel_uncore_forcewake_domain *domain; int retry_count = 100; - enum forcewake_domain_id id; enum forcewake_domains fw = 0, active_domains; /* Hold uncore.lock across reset to prevent any register access @@ -258,18 +263,18 @@ void intel_uncore_forcewake_reset(struct drm_device *dev, bool restore) while (1) { active_domains = 0; - for_each_fw_domain(domain, dev_priv, id) { - if (del_timer_sync(&domain->timer) == 0) + for_each_fw_domain(domain, dev_priv) { + if (hrtimer_cancel(&domain->timer) == 0) continue; - intel_uncore_fw_release_timer((unsigned long)domain); + intel_uncore_fw_release_timer(&domain->timer); } spin_lock_irqsave(&dev_priv->uncore.lock, irqflags); - for_each_fw_domain(domain, dev_priv, id) { - if (timer_pending(&domain->timer)) - active_domains |= (1 << id); + for_each_fw_domain(domain, dev_priv) { + if (hrtimer_active(&domain->timer)) + active_domains |= domain->mask; } if (active_domains == 0) @@ -286,9 +291,9 @@ void intel_uncore_forcewake_reset(struct drm_device *dev, bool restore) WARN_ON(active_domains); - for_each_fw_domain(domain, dev_priv, id) + for_each_fw_domain(domain, dev_priv) if (domain->wake_count) - fw |= 1 << id; + fw |= domain->mask; if (fw) dev_priv->uncore.funcs.force_wake_put(dev_priv, fw); @@ -310,21 +315,49 @@ void intel_uncore_forcewake_reset(struct drm_device *dev, bool restore) spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags); } -static void intel_uncore_ellc_detect(struct drm_device *dev) +static u64 gen9_edram_size(struct drm_i915_private *dev_priv) { - struct drm_i915_private *dev_priv = dev->dev_private; + const unsigned int ways[8] = { 4, 8, 12, 16, 16, 16, 16, 16 }; + const unsigned int sets[4] = { 1, 1, 2, 2 }; + const u32 cap = dev_priv->edram_cap; + + return EDRAM_NUM_BANKS(cap) * + ways[EDRAM_WAYS_IDX(cap)] * + sets[EDRAM_SETS_IDX(cap)] * + 1024 * 1024; +} + +u64 intel_uncore_edram_size(struct drm_i915_private *dev_priv) +{ + if (!HAS_EDRAM(dev_priv)) + return 0; + + /* The needed capability bits for size calculation + * are not there with pre gen9 so return 128MB always. + */ + if (INTEL_GEN(dev_priv) < 9) + return 128 * 1024 * 1024; + + return gen9_edram_size(dev_priv); +} + +static void intel_uncore_edram_detect(struct drm_i915_private *dev_priv) +{ + if (IS_HASWELL(dev_priv) || + IS_BROADWELL(dev_priv) || + INTEL_GEN(dev_priv) >= 9) { + dev_priv->edram_cap = __raw_i915_read32(dev_priv, + HSW_EDRAM_CAP); - if ((IS_HASWELL(dev) || IS_BROADWELL(dev) || - INTEL_INFO(dev)->gen >= 9) && - (__raw_i915_read32(dev_priv, HSW_EDRAM_PRESENT) & EDRAM_ENABLED)) { - /* The docs do not explain exactly how the calculation can be - * made. It is somewhat guessable, but for now, it's always - * 128MB. - * NB: We can't write IDICR yet because we do not have gt funcs + /* NB: We can't write IDICR yet because we do not have gt funcs * set up */ - dev_priv->ellc_size = 128; - DRM_INFO("Found %zuMB of eLLC\n", dev_priv->ellc_size); + } else { + dev_priv->edram_cap = 0; } + + if (HAS_EDRAM(dev_priv)) + DRM_INFO("Found %lluMB of eDRAM\n", + intel_uncore_edram_size(dev_priv) / (1024 * 1024)); } static bool @@ -410,16 +443,15 @@ static void __intel_uncore_forcewake_get(struct drm_i915_private *dev_priv, enum forcewake_domains fw_domains) { struct intel_uncore_forcewake_domain *domain; - enum forcewake_domain_id id; if (!dev_priv->uncore.funcs.force_wake_get) return; fw_domains &= dev_priv->uncore.fw_domains; - for_each_fw_domain_mask(domain, fw_domains, dev_priv, id) { + for_each_fw_domain_masked(domain, fw_domains, dev_priv) { if (domain->wake_count++) - fw_domains &= ~(1 << id); + fw_domains &= ~domain->mask; } if (fw_domains) @@ -477,21 +509,19 @@ static void __intel_uncore_forcewake_put(struct drm_i915_private *dev_priv, enum forcewake_domains fw_domains) { struct intel_uncore_forcewake_domain *domain; - enum forcewake_domain_id id; if (!dev_priv->uncore.funcs.force_wake_put) return; fw_domains &= dev_priv->uncore.fw_domains; - for_each_fw_domain_mask(domain, fw_domains, dev_priv, id) { + for_each_fw_domain_masked(domain, fw_domains, dev_priv) { if (WARN_ON(domain->wake_count == 0)) continue; if (--domain->wake_count) continue; - domain->wake_count++; fw_domain_arm_timer(domain); } } @@ -539,18 +569,27 @@ void intel_uncore_forcewake_put__locked(struct drm_i915_private *dev_priv, void assert_forcewakes_inactive(struct drm_i915_private *dev_priv) { struct intel_uncore_forcewake_domain *domain; - enum forcewake_domain_id id; if (!dev_priv->uncore.funcs.force_wake_get) return; - for_each_fw_domain(domain, dev_priv, id) + for_each_fw_domain(domain, dev_priv) WARN_ON(domain->wake_count); } /* We give fast paths for the really cool registers */ #define NEEDS_FORCE_WAKE(reg) ((reg) < 0x40000) +#define __gen6_reg_read_fw_domains(offset) \ +({ \ + enum forcewake_domains __fwd; \ + if (NEEDS_FORCE_WAKE(offset)) \ + __fwd = FORCEWAKE_RENDER; \ + else \ + __fwd = 0; \ + __fwd; \ +}) + #define REG_RANGE(reg, start, end) ((reg) >= (start) && (reg) < (end)) #define FORCEWAKE_VLV_RENDER_RANGE_OFFSET(reg) \ @@ -564,6 +603,48 @@ void assert_forcewakes_inactive(struct drm_i915_private *dev_priv) REG_RANGE((reg), 0x22000, 0x24000) || \ REG_RANGE((reg), 0x30000, 0x40000)) +#define __vlv_reg_read_fw_domains(offset) \ +({ \ + enum forcewake_domains __fwd = 0; \ + if (!NEEDS_FORCE_WAKE(offset)) \ + __fwd = 0; \ + else if (FORCEWAKE_VLV_RENDER_RANGE_OFFSET(offset)) \ + __fwd = FORCEWAKE_RENDER; \ + else if (FORCEWAKE_VLV_MEDIA_RANGE_OFFSET(offset)) \ + __fwd = FORCEWAKE_MEDIA; \ + __fwd; \ +}) + +static const i915_reg_t gen8_shadowed_regs[] = { + GEN6_RPNSWREQ, + GEN6_RC_VIDEO_FREQ, + RING_TAIL(RENDER_RING_BASE), + RING_TAIL(GEN6_BSD_RING_BASE), + RING_TAIL(VEBOX_RING_BASE), + RING_TAIL(BLT_RING_BASE), + /* TODO: Other registers are not yet used */ +}; + +static bool is_gen8_shadowed(u32 offset) +{ + int i; + for (i = 0; i < ARRAY_SIZE(gen8_shadowed_regs); i++) + if (offset == gen8_shadowed_regs[i].reg) + return true; + + return false; +} + +#define __gen8_reg_write_fw_domains(offset) \ +({ \ + enum forcewake_domains __fwd; \ + if (NEEDS_FORCE_WAKE(offset) && !is_gen8_shadowed(offset)) \ + __fwd = FORCEWAKE_RENDER; \ + else \ + __fwd = 0; \ + __fwd; \ +}) + #define FORCEWAKE_CHV_RENDER_RANGE_OFFSET(reg) \ (REG_RANGE((reg), 0x2000, 0x4000) || \ REG_RANGE((reg), 0x5200, 0x8000) || \ @@ -586,6 +667,34 @@ void assert_forcewakes_inactive(struct drm_i915_private *dev_priv) REG_RANGE((reg), 0x9000, 0xB000) || \ REG_RANGE((reg), 0xF000, 0x10000)) +#define __chv_reg_read_fw_domains(offset) \ +({ \ + enum forcewake_domains __fwd = 0; \ + if (!NEEDS_FORCE_WAKE(offset)) \ + __fwd = 0; \ + else if (FORCEWAKE_CHV_RENDER_RANGE_OFFSET(offset)) \ + __fwd = FORCEWAKE_RENDER; \ + else if (FORCEWAKE_CHV_MEDIA_RANGE_OFFSET(offset)) \ + __fwd = FORCEWAKE_MEDIA; \ + else if (FORCEWAKE_CHV_COMMON_RANGE_OFFSET(offset)) \ + __fwd = FORCEWAKE_RENDER | FORCEWAKE_MEDIA; \ + __fwd; \ +}) + +#define __chv_reg_write_fw_domains(offset) \ +({ \ + enum forcewake_domains __fwd = 0; \ + if (!NEEDS_FORCE_WAKE(offset) || is_gen8_shadowed(offset)) \ + __fwd = 0; \ + else if (FORCEWAKE_CHV_RENDER_RANGE_OFFSET(offset)) \ + __fwd = FORCEWAKE_RENDER; \ + else if (FORCEWAKE_CHV_MEDIA_RANGE_OFFSET(offset)) \ + __fwd = FORCEWAKE_MEDIA; \ + else if (FORCEWAKE_CHV_COMMON_RANGE_OFFSET(offset)) \ + __fwd = FORCEWAKE_RENDER | FORCEWAKE_MEDIA; \ + __fwd; \ +}) + #define FORCEWAKE_GEN9_UNCORE_RANGE_OFFSET(reg) \ REG_RANGE((reg), 0xB00, 0x2000) @@ -618,6 +727,61 @@ void assert_forcewakes_inactive(struct drm_i915_private *dev_priv) !FORCEWAKE_GEN9_MEDIA_RANGE_OFFSET(reg) && \ !FORCEWAKE_GEN9_COMMON_RANGE_OFFSET(reg)) +#define SKL_NEEDS_FORCE_WAKE(reg) \ + ((reg) < 0x40000 && !FORCEWAKE_GEN9_UNCORE_RANGE_OFFSET(reg)) + +#define __gen9_reg_read_fw_domains(offset) \ +({ \ + enum forcewake_domains __fwd; \ + if (!SKL_NEEDS_FORCE_WAKE(offset)) \ + __fwd = 0; \ + else if (FORCEWAKE_GEN9_RENDER_RANGE_OFFSET(offset)) \ + __fwd = FORCEWAKE_RENDER; \ + else if (FORCEWAKE_GEN9_MEDIA_RANGE_OFFSET(offset)) \ + __fwd = FORCEWAKE_MEDIA; \ + else if (FORCEWAKE_GEN9_COMMON_RANGE_OFFSET(offset)) \ + __fwd = FORCEWAKE_RENDER | FORCEWAKE_MEDIA; \ + else \ + __fwd = FORCEWAKE_BLITTER; \ + __fwd; \ +}) + +static const i915_reg_t gen9_shadowed_regs[] = { + RING_TAIL(RENDER_RING_BASE), + RING_TAIL(GEN6_BSD_RING_BASE), + RING_TAIL(VEBOX_RING_BASE), + RING_TAIL(BLT_RING_BASE), + GEN6_RPNSWREQ, + GEN6_RC_VIDEO_FREQ, + /* TODO: Other registers are not yet used */ +}; + +static bool is_gen9_shadowed(u32 offset) +{ + int i; + for (i = 0; i < ARRAY_SIZE(gen9_shadowed_regs); i++) + if (offset == gen9_shadowed_regs[i].reg) + return true; + + return false; +} + +#define __gen9_reg_write_fw_domains(offset) \ +({ \ + enum forcewake_domains __fwd; \ + if (!SKL_NEEDS_FORCE_WAKE(offset) || is_gen9_shadowed(offset)) \ + __fwd = 0; \ + else if (FORCEWAKE_GEN9_RENDER_RANGE_OFFSET(offset)) \ + __fwd = FORCEWAKE_RENDER; \ + else if (FORCEWAKE_GEN9_MEDIA_RANGE_OFFSET(offset)) \ + __fwd = FORCEWAKE_MEDIA; \ + else if (FORCEWAKE_GEN9_COMMON_RANGE_OFFSET(offset)) \ + __fwd = FORCEWAKE_RENDER | FORCEWAKE_MEDIA; \ + else \ + __fwd = FORCEWAKE_BLITTER; \ + __fwd; \ +}) + static void ilk_dummy_write(struct drm_i915_private *dev_priv) { @@ -633,15 +797,6 @@ __unclaimed_reg_debug(struct drm_i915_private *dev_priv, const bool read, const bool before) { - /* XXX. We limit the auto arming traces for mmio - * debugs on these platforms. There are just too many - * revealed by these and CI/Bat suffers from the noise. - * Please fix and then re-enable the automatic traces. - */ - if (i915.mmio_debug < 2 && - (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv))) - return; - if (WARN(check_for_unclaimed_mmio(dev_priv), "Unclaimed register detected %s %s register 0x%x\n", before ? "before" : "after", @@ -720,19 +875,17 @@ static inline void __force_wake_auto(struct drm_i915_private *dev_priv, enum forcewake_domains fw_domains) { struct intel_uncore_forcewake_domain *domain; - enum forcewake_domain_id id; if (WARN_ON(!fw_domains)) return; /* Ideally GCC would be constant-fold and eliminate this loop */ - for_each_fw_domain_mask(domain, fw_domains, dev_priv, id) { + for_each_fw_domain_masked(domain, fw_domains, dev_priv) { if (domain->wake_count) { - fw_domains &= ~(1 << id); + fw_domains &= ~domain->mask; continue; } - domain->wake_count++; fw_domain_arm_timer(domain); } @@ -743,9 +896,11 @@ static inline void __force_wake_auto(struct drm_i915_private *dev_priv, #define __gen6_read(x) \ static u##x \ gen6_read##x(struct drm_i915_private *dev_priv, i915_reg_t reg, bool trace) { \ + enum forcewake_domains fw_engine; \ GEN6_READ_HEADER(x); \ - if (NEEDS_FORCE_WAKE(offset)) \ - __force_wake_auto(dev_priv, FORCEWAKE_RENDER); \ + fw_engine = __gen6_reg_read_fw_domains(offset); \ + if (fw_engine) \ + __force_wake_auto(dev_priv, fw_engine); \ val = __raw_i915_read##x(dev_priv, reg); \ GEN6_READ_FOOTER; \ } @@ -753,14 +908,9 @@ gen6_read##x(struct drm_i915_private *dev_priv, i915_reg_t reg, bool trace) { \ #define __vlv_read(x) \ static u##x \ vlv_read##x(struct drm_i915_private *dev_priv, i915_reg_t reg, bool trace) { \ - enum forcewake_domains fw_engine = 0; \ + enum forcewake_domains fw_engine; \ GEN6_READ_HEADER(x); \ - if (!NEEDS_FORCE_WAKE(offset)) \ - fw_engine = 0; \ - else if (FORCEWAKE_VLV_RENDER_RANGE_OFFSET(offset)) \ - fw_engine = FORCEWAKE_RENDER; \ - else if (FORCEWAKE_VLV_MEDIA_RANGE_OFFSET(offset)) \ - fw_engine = FORCEWAKE_MEDIA; \ + fw_engine = __vlv_reg_read_fw_domains(offset); \ if (fw_engine) \ __force_wake_auto(dev_priv, fw_engine); \ val = __raw_i915_read##x(dev_priv, reg); \ @@ -770,40 +920,21 @@ vlv_read##x(struct drm_i915_private *dev_priv, i915_reg_t reg, bool trace) { \ #define __chv_read(x) \ static u##x \ chv_read##x(struct drm_i915_private *dev_priv, i915_reg_t reg, bool trace) { \ - enum forcewake_domains fw_engine = 0; \ + enum forcewake_domains fw_engine; \ GEN6_READ_HEADER(x); \ - if (!NEEDS_FORCE_WAKE(offset)) \ - fw_engine = 0; \ - else if (FORCEWAKE_CHV_RENDER_RANGE_OFFSET(offset)) \ - fw_engine = FORCEWAKE_RENDER; \ - else if (FORCEWAKE_CHV_MEDIA_RANGE_OFFSET(offset)) \ - fw_engine = FORCEWAKE_MEDIA; \ - else if (FORCEWAKE_CHV_COMMON_RANGE_OFFSET(offset)) \ - fw_engine = FORCEWAKE_RENDER | FORCEWAKE_MEDIA; \ + fw_engine = __chv_reg_read_fw_domains(offset); \ if (fw_engine) \ __force_wake_auto(dev_priv, fw_engine); \ val = __raw_i915_read##x(dev_priv, reg); \ GEN6_READ_FOOTER; \ } -#define SKL_NEEDS_FORCE_WAKE(reg) \ - ((reg) < 0x40000 && !FORCEWAKE_GEN9_UNCORE_RANGE_OFFSET(reg)) - #define __gen9_read(x) \ static u##x \ gen9_read##x(struct drm_i915_private *dev_priv, i915_reg_t reg, bool trace) { \ enum forcewake_domains fw_engine; \ GEN6_READ_HEADER(x); \ - if (!SKL_NEEDS_FORCE_WAKE(offset)) \ - fw_engine = 0; \ - else if (FORCEWAKE_GEN9_RENDER_RANGE_OFFSET(offset)) \ - fw_engine = FORCEWAKE_RENDER; \ - else if (FORCEWAKE_GEN9_MEDIA_RANGE_OFFSET(offset)) \ - fw_engine = FORCEWAKE_MEDIA; \ - else if (FORCEWAKE_GEN9_COMMON_RANGE_OFFSET(offset)) \ - fw_engine = FORCEWAKE_RENDER | FORCEWAKE_MEDIA; \ - else \ - fw_engine = FORCEWAKE_BLITTER; \ + fw_engine = __gen9_reg_read_fw_domains(offset); \ if (fw_engine) \ __force_wake_auto(dev_priv, fw_engine); \ val = __raw_i915_read##x(dev_priv, reg); \ @@ -942,34 +1073,14 @@ hsw_write##x(struct drm_i915_private *dev_priv, i915_reg_t reg, u##x val, bool t GEN6_WRITE_FOOTER; \ } -static const i915_reg_t gen8_shadowed_regs[] = { - FORCEWAKE_MT, - GEN6_RPNSWREQ, - GEN6_RC_VIDEO_FREQ, - RING_TAIL(RENDER_RING_BASE), - RING_TAIL(GEN6_BSD_RING_BASE), - RING_TAIL(VEBOX_RING_BASE), - RING_TAIL(BLT_RING_BASE), - /* TODO: Other registers are not yet used */ -}; - -static bool is_gen8_shadowed(struct drm_i915_private *dev_priv, - i915_reg_t reg) -{ - int i; - for (i = 0; i < ARRAY_SIZE(gen8_shadowed_regs); i++) - if (i915_mmio_reg_equal(reg, gen8_shadowed_regs[i])) - return true; - - return false; -} - #define __gen8_write(x) \ static void \ gen8_write##x(struct drm_i915_private *dev_priv, i915_reg_t reg, u##x val, bool trace) { \ + enum forcewake_domains fw_engine; \ GEN6_WRITE_HEADER; \ - if (NEEDS_FORCE_WAKE(offset) && !is_gen8_shadowed(dev_priv, reg)) \ - __force_wake_auto(dev_priv, FORCEWAKE_RENDER); \ + fw_engine = __gen8_reg_write_fw_domains(offset); \ + if (fw_engine) \ + __force_wake_auto(dev_priv, fw_engine); \ __raw_i915_write##x(dev_priv, reg, val); \ GEN6_WRITE_FOOTER; \ } @@ -977,64 +1088,22 @@ gen8_write##x(struct drm_i915_private *dev_priv, i915_reg_t reg, u##x val, bool #define __chv_write(x) \ static void \ chv_write##x(struct drm_i915_private *dev_priv, i915_reg_t reg, u##x val, bool trace) { \ - enum forcewake_domains fw_engine = 0; \ + enum forcewake_domains fw_engine; \ GEN6_WRITE_HEADER; \ - if (!NEEDS_FORCE_WAKE(offset) || \ - is_gen8_shadowed(dev_priv, reg)) \ - fw_engine = 0; \ - else if (FORCEWAKE_CHV_RENDER_RANGE_OFFSET(offset)) \ - fw_engine = FORCEWAKE_RENDER; \ - else if (FORCEWAKE_CHV_MEDIA_RANGE_OFFSET(offset)) \ - fw_engine = FORCEWAKE_MEDIA; \ - else if (FORCEWAKE_CHV_COMMON_RANGE_OFFSET(offset)) \ - fw_engine = FORCEWAKE_RENDER | FORCEWAKE_MEDIA; \ + fw_engine = __chv_reg_write_fw_domains(offset); \ if (fw_engine) \ __force_wake_auto(dev_priv, fw_engine); \ __raw_i915_write##x(dev_priv, reg, val); \ GEN6_WRITE_FOOTER; \ } -static const i915_reg_t gen9_shadowed_regs[] = { - RING_TAIL(RENDER_RING_BASE), - RING_TAIL(GEN6_BSD_RING_BASE), - RING_TAIL(VEBOX_RING_BASE), - RING_TAIL(BLT_RING_BASE), - FORCEWAKE_BLITTER_GEN9, - FORCEWAKE_RENDER_GEN9, - FORCEWAKE_MEDIA_GEN9, - GEN6_RPNSWREQ, - GEN6_RC_VIDEO_FREQ, - /* TODO: Other registers are not yet used */ -}; - -static bool is_gen9_shadowed(struct drm_i915_private *dev_priv, - i915_reg_t reg) -{ - int i; - for (i = 0; i < ARRAY_SIZE(gen9_shadowed_regs); i++) - if (i915_mmio_reg_equal(reg, gen9_shadowed_regs[i])) - return true; - - return false; -} - #define __gen9_write(x) \ static void \ gen9_write##x(struct drm_i915_private *dev_priv, i915_reg_t reg, u##x val, \ bool trace) { \ enum forcewake_domains fw_engine; \ GEN6_WRITE_HEADER; \ - if (!SKL_NEEDS_FORCE_WAKE(offset) || \ - is_gen9_shadowed(dev_priv, reg)) \ - fw_engine = 0; \ - else if (FORCEWAKE_GEN9_RENDER_RANGE_OFFSET(offset)) \ - fw_engine = FORCEWAKE_RENDER; \ - else if (FORCEWAKE_GEN9_MEDIA_RANGE_OFFSET(offset)) \ - fw_engine = FORCEWAKE_MEDIA; \ - else if (FORCEWAKE_GEN9_COMMON_RANGE_OFFSET(offset)) \ - fw_engine = FORCEWAKE_RENDER | FORCEWAKE_MEDIA; \ - else \ - fw_engine = FORCEWAKE_BLITTER; \ + fw_engine = __gen9_reg_write_fw_domains(offset); \ if (fw_engine) \ __force_wake_auto(dev_priv, fw_engine); \ __raw_i915_write##x(dev_priv, reg, val); \ @@ -1150,7 +1219,14 @@ static void fw_domain_init(struct drm_i915_private *dev_priv, d->i915 = dev_priv; d->id = domain_id; - setup_timer(&d->timer, intel_uncore_fw_release_timer, (unsigned long)d); + BUILD_BUG_ON(FORCEWAKE_RENDER != (1 << FW_DOMAIN_ID_RENDER)); + BUILD_BUG_ON(FORCEWAKE_BLITTER != (1 << FW_DOMAIN_ID_BLITTER)); + BUILD_BUG_ON(FORCEWAKE_MEDIA != (1 << FW_DOMAIN_ID_MEDIA)); + + d->mask = 1 << domain_id; + + hrtimer_init(&d->timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); + d->timer.function = intel_uncore_fw_release_timer; dev_priv->uncore.fw_domains |= (1 << domain_id); @@ -1189,7 +1265,11 @@ static void intel_uncore_fw_domains_init(struct drm_device *dev) } else if (IS_HASWELL(dev) || IS_BROADWELL(dev)) { dev_priv->uncore.funcs.force_wake_get = fw_domains_get_with_thread_status; - dev_priv->uncore.funcs.force_wake_put = fw_domains_put; + if (IS_HASWELL(dev)) + dev_priv->uncore.funcs.force_wake_put = + fw_domains_put_with_fifo; + else + dev_priv->uncore.funcs.force_wake_put = fw_domains_put; fw_domain_init(dev_priv, FW_DOMAIN_ID_RENDER, FORCEWAKE_MT, FORCEWAKE_ACK_HSW); } else if (IS_IVYBRIDGE(dev)) { @@ -1253,7 +1333,7 @@ void intel_uncore_init(struct drm_device *dev) i915_check_vgpu(dev); - intel_uncore_ellc_detect(dev); + intel_uncore_edram_detect(dev_priv); intel_uncore_fw_domains_init(dev); __intel_uncore_early_sanitize(dev, false); @@ -1715,3 +1795,111 @@ intel_uncore_arm_unclaimed_mmio_detection(struct drm_i915_private *dev_priv) return false; } + +static enum forcewake_domains +intel_uncore_forcewake_for_read(struct drm_i915_private *dev_priv, + i915_reg_t reg) +{ + enum forcewake_domains fw_domains; + + if (intel_vgpu_active(dev_priv->dev)) + return 0; + + switch (INTEL_INFO(dev_priv)->gen) { + case 9: + fw_domains = __gen9_reg_read_fw_domains(i915_mmio_reg_offset(reg)); + break; + case 8: + if (IS_CHERRYVIEW(dev_priv)) + fw_domains = __chv_reg_read_fw_domains(i915_mmio_reg_offset(reg)); + else + fw_domains = __gen6_reg_read_fw_domains(i915_mmio_reg_offset(reg)); + break; + case 7: + case 6: + if (IS_VALLEYVIEW(dev_priv)) + fw_domains = __vlv_reg_read_fw_domains(i915_mmio_reg_offset(reg)); + else + fw_domains = __gen6_reg_read_fw_domains(i915_mmio_reg_offset(reg)); + break; + default: + MISSING_CASE(INTEL_INFO(dev_priv)->gen); + case 5: /* forcewake was introduced with gen6 */ + case 4: + case 3: + case 2: + return 0; + } + + WARN_ON(fw_domains & ~dev_priv->uncore.fw_domains); + + return fw_domains; +} + +static enum forcewake_domains +intel_uncore_forcewake_for_write(struct drm_i915_private *dev_priv, + i915_reg_t reg) +{ + enum forcewake_domains fw_domains; + + if (intel_vgpu_active(dev_priv->dev)) + return 0; + + switch (INTEL_INFO(dev_priv)->gen) { + case 9: + fw_domains = __gen9_reg_write_fw_domains(i915_mmio_reg_offset(reg)); + break; + case 8: + if (IS_CHERRYVIEW(dev_priv)) + fw_domains = __chv_reg_write_fw_domains(i915_mmio_reg_offset(reg)); + else + fw_domains = __gen8_reg_write_fw_domains(i915_mmio_reg_offset(reg)); + break; + case 7: + case 6: + fw_domains = FORCEWAKE_RENDER; + break; + default: + MISSING_CASE(INTEL_INFO(dev_priv)->gen); + case 5: + case 4: + case 3: + case 2: + return 0; + } + + WARN_ON(fw_domains & ~dev_priv->uncore.fw_domains); + + return fw_domains; +} + +/** + * intel_uncore_forcewake_for_reg - which forcewake domains are needed to access + * a register + * @dev_priv: pointer to struct drm_i915_private + * @reg: register in question + * @op: operation bitmask of FW_REG_READ and/or FW_REG_WRITE + * + * Returns a set of forcewake domains required to be taken with for example + * intel_uncore_forcewake_get for the specified register to be accessible in the + * specified mode (read, write or read/write) with raw mmio accessors. + * + * NOTE: On Gen6 and Gen7 write forcewake domain (FORCEWAKE_RENDER) requires the + * callers to do FIFO management on their own or risk losing writes. + */ +enum forcewake_domains +intel_uncore_forcewake_for_reg(struct drm_i915_private *dev_priv, + i915_reg_t reg, unsigned int op) +{ + enum forcewake_domains fw_domains = 0; + + WARN_ON(!op); + + if (op & FW_REG_READ) + fw_domains = intel_uncore_forcewake_for_read(dev_priv, reg); + + if (op & FW_REG_WRITE) + fw_domains |= intel_uncore_forcewake_for_write(dev_priv, reg); + + return fw_domains; +} |