linux_dsm_epyc7002/drivers/gpu/drm/i915/i915_drv.c

2785 lines
75 KiB
C
Raw Normal View History

/* i915_drv.c -- i830,i845,i855,i865,i915 driver -*- linux-c -*-
*/
/*
*
* Copyright 2003 Tungsten Graphics, Inc., Cedar Park, Texas.
* All Rights Reserved.
*
* 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, sub license, 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 NON-INFRINGEMENT.
* IN NO EVENT SHALL TUNGSTEN GRAPHICS AND/OR ITS SUPPLIERS BE LIABLE FOR
* ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
* TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
* SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
*
*/
#include <linux/acpi.h>
#include <linux/device.h>
#include <linux/oom.h>
#include <linux/module.h>
#include <linux/pci.h>
#include <linux/pm.h>
#include <linux/pm_runtime.h>
#include <linux/pnp.h>
#include <linux/slab.h>
#include <linux/vga_switcheroo.h>
#include <linux/vt.h>
#include <acpi/video.h>
#include <drm/drm_atomic_helper.h>
#include <drm/drm_ioctl.h>
#include <drm/drm_irq.h>
#include <drm/drm_probe_helper.h>
#include <drm/i915_drm.h>
#include "display/intel_acpi.h"
#include "display/intel_audio.h"
#include "display/intel_bw.h"
#include "display/intel_cdclk.h"
#include "display/intel_display_types.h"
#include "display/intel_dp.h"
#include "display/intel_fbdev.h"
#include "display/intel_hotplug.h"
#include "display/intel_overlay.h"
#include "display/intel_pipe_crc.h"
#include "display/intel_sprite.h"
#include "display/intel_vga.h"
#include "gem/i915_gem_context.h"
#include "gem/i915_gem_ioctls.h"
#include "gt/intel_gt.h"
drm/i915: Invert the GEM wakeref hierarchy In the current scheme, on submitting a request we take a single global GEM wakeref, which trickles down to wake up all GT power domains. This is undesirable as we would like to be able to localise our power management to the available power domains and to remove the global GEM operations from the heart of the driver. (The intent there is to push global GEM decisions to the boundary as used by the GEM user interface.) Now during request construction, each request is responsible via its logical context to acquire a wakeref on each power domain it intends to utilize. Currently, each request takes a wakeref on the engine(s) and the engines themselves take a chipset wakeref. This gives us a transition on each engine which we can extend if we want to insert more powermangement control (such as soft rc6). The global GEM operations that currently require a struct_mutex are reduced to listening to pm events from the chipset GT wakeref. As we reduce the struct_mutex requirement, these listeners should evaporate. Perhaps the biggest immediate change is that this removes the struct_mutex requirement around GT power management, allowing us greater flexibility in request construction. Another important knock-on effect, is that by tracking engine usage, we can insert a switch back to the kernel context on that engine immediately, avoiding any extra delay or inserting global synchronisation barriers. This makes tracking when an engine and its associated contexts are idle much easier -- important for when we forgo our assumed execution ordering and need idle barriers to unpin used contexts. In the process, it means we remove a large chunk of code whose only purpose was to switch back to the kernel context. Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Cc: Imre Deak <imre.deak@intel.com> Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190424200717.1686-5-chris@chris-wilson.co.uk
2019-04-25 03:07:17 +07:00
#include "gt/intel_gt_pm.h"
#include "gt/intel_rc6.h"
#include "i915_debugfs.h"
#include "i915_drv.h"
#include "i915_irq.h"
#include "i915_memcpy.h"
#include "i915_perf.h"
drm/i915: add query uAPI There are a number of information that are readable from hardware registers and that we would like to make accessible to userspace. One particular example is the topology of the execution units (how are execution units grouped in subslices and slices and also which ones have been fused off for die recovery). At the moment the GET_PARAM ioctl covers some basic needs, but generally is only able to return a single value for each defined parameter. This is a bit problematic with topology descriptions which are array/maps of available units. This change introduces a new ioctl that can deal with requests to fill structures of potentially variable lengths. The user is expected fill a query with length fields set at 0 on the first call, the kernel then sets the length fields to the their expected values. A second call to the kernel with length fields at their expected values will trigger a copy of the data to the pointed memory locations. The scope of this uAPI is only to provide information to userspace, not to allow configuration of the device. v2: Simplify dispatcher code iteration (Tvrtko) Tweak uapi drm_i915_query_item structure (Tvrtko) v3: Rename pad fields into flags (Chris) Return error on flags field != 0 (Chris) Only copy length back to userspace in drm_i915_query_item (Chris) v4: Use array of functions instead of switch (Chris) v5: More comments in uapi (Tvrtko) Return query item errors in length field (All) v6: Tweak uapi comments style to match the coding style (Lionel) v7: Add i915_query.h (Joonas) v8: (Lionel) Change the behavior of the item iterator to report invalid queries into the query item rather than stopping the iteration. This enables userspace applications to query newer items on older kernels and only have failure on the items that are not supported. v9: Edit copyright headers (Joonas) v10: Typos & comments in uapi (Joonas) Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com> Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Acked-by: Chris Wilson <chris@chris-wilson.co.uk> Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20180306122857.27317-6-lionel.g.landwerlin@intel.com
2018-03-06 19:28:56 +07:00
#include "i915_query.h"
#include "i915_suspend.h"
#include "i915_switcheroo.h"
#include "i915_sysfs.h"
#include "i915_trace.h"
#include "i915_vgpu.h"
#include "intel_csr.h"
#include "intel_memory_region.h"
#include "intel_pm.h"
static struct drm_driver driver;
struct vlv_s0ix_state {
/* GAM */
u32 wr_watermark;
u32 gfx_prio_ctrl;
u32 arb_mode;
u32 gfx_pend_tlb0;
u32 gfx_pend_tlb1;
u32 lra_limits[GEN7_LRA_LIMITS_REG_NUM];
u32 media_max_req_count;
u32 gfx_max_req_count;
u32 render_hwsp;
u32 ecochk;
u32 bsd_hwsp;
u32 blt_hwsp;
u32 tlb_rd_addr;
/* MBC */
u32 g3dctl;
u32 gsckgctl;
u32 mbctl;
/* GCP */
u32 ucgctl1;
u32 ucgctl3;
u32 rcgctl1;
u32 rcgctl2;
u32 rstctl;
u32 misccpctl;
/* GPM */
u32 gfxpause;
u32 rpdeuhwtc;
u32 rpdeuc;
u32 ecobus;
u32 pwrdwnupctl;
u32 rp_down_timeout;
u32 rp_deucsw;
u32 rcubmabdtmr;
u32 rcedata;
u32 spare2gh;
/* Display 1 CZ domain */
u32 gt_imr;
u32 gt_ier;
u32 pm_imr;
u32 pm_ier;
u32 gt_scratch[GEN7_GT_SCRATCH_REG_NUM];
/* GT SA CZ domain */
u32 tilectl;
u32 gt_fifoctl;
u32 gtlc_wake_ctrl;
u32 gtlc_survive;
u32 pmwgicz;
/* Display 2 CZ domain */
u32 gu_ctl0;
u32 gu_ctl1;
u32 pcbr;
u32 clock_gate_dis2;
};
static int i915_get_bridge_dev(struct drm_i915_private *dev_priv)
{
int domain = pci_domain_nr(dev_priv->drm.pdev->bus);
dev_priv->bridge_dev =
pci_get_domain_bus_and_slot(domain, 0, PCI_DEVFN(0, 0));
if (!dev_priv->bridge_dev) {
DRM_ERROR("bridge device not found\n");
return -1;
}
return 0;
}
/* Allocate space for the MCH regs if needed, return nonzero on error */
static int
intel_alloc_mchbar_resource(struct drm_i915_private *dev_priv)
{
int reg = INTEL_GEN(dev_priv) >= 4 ? MCHBAR_I965 : MCHBAR_I915;
u32 temp_lo, temp_hi = 0;
u64 mchbar_addr;
int ret;
if (INTEL_GEN(dev_priv) >= 4)
pci_read_config_dword(dev_priv->bridge_dev, reg + 4, &temp_hi);
pci_read_config_dword(dev_priv->bridge_dev, reg, &temp_lo);
mchbar_addr = ((u64)temp_hi << 32) | temp_lo;
/* If ACPI doesn't have it, assume we need to allocate it ourselves */
#ifdef CONFIG_PNP
if (mchbar_addr &&
pnp_range_reserved(mchbar_addr, mchbar_addr + MCHBAR_SIZE))
return 0;
#endif
/* Get some space for it */
dev_priv->mch_res.name = "i915 MCHBAR";
dev_priv->mch_res.flags = IORESOURCE_MEM;
ret = pci_bus_alloc_resource(dev_priv->bridge_dev->bus,
&dev_priv->mch_res,
MCHBAR_SIZE, MCHBAR_SIZE,
PCIBIOS_MIN_MEM,
0, pcibios_align_resource,
dev_priv->bridge_dev);
if (ret) {
DRM_DEBUG_DRIVER("failed bus alloc: %d\n", ret);
dev_priv->mch_res.start = 0;
return ret;
}
if (INTEL_GEN(dev_priv) >= 4)
pci_write_config_dword(dev_priv->bridge_dev, reg + 4,
upper_32_bits(dev_priv->mch_res.start));
pci_write_config_dword(dev_priv->bridge_dev, reg,
lower_32_bits(dev_priv->mch_res.start));
return 0;
}
/* Setup MCHBAR if possible, return true if we should disable it again */
static void
intel_setup_mchbar(struct drm_i915_private *dev_priv)
{
int mchbar_reg = INTEL_GEN(dev_priv) >= 4 ? MCHBAR_I965 : MCHBAR_I915;
u32 temp;
bool enabled;
if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv))
return;
dev_priv->mchbar_need_disable = false;
if (IS_I915G(dev_priv) || IS_I915GM(dev_priv)) {
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);
enabled = temp & 1;
}
/* If it's already enabled, don't have to do anything */
if (enabled)
return;
if (intel_alloc_mchbar_resource(dev_priv))
return;
dev_priv->mchbar_need_disable = true;
/* Space is allocated or reserved, so enable it. */
if (IS_I915G(dev_priv) || IS_I915GM(dev_priv)) {
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);
pci_write_config_dword(dev_priv->bridge_dev, mchbar_reg, temp | 1);
}
}
static void
intel_teardown_mchbar(struct drm_i915_private *dev_priv)
{
int mchbar_reg = INTEL_GEN(dev_priv) >= 4 ? MCHBAR_I965 : MCHBAR_I915;
if (dev_priv->mchbar_need_disable) {
if (IS_I915G(dev_priv) || IS_I915GM(dev_priv)) {
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 {
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);
}
}
if (dev_priv->mch_res.start)
release_resource(&dev_priv->mch_res);
}
static int i915_driver_modeset_probe(struct drm_i915_private *i915)
{
int ret;
if (i915_inject_probe_failure(i915))
return -ENODEV;
if (HAS_DISPLAY(i915) && INTEL_DISPLAY_ENABLED(i915)) {
ret = drm_vblank_init(&i915->drm,
INTEL_NUM_PIPES(i915));
if (ret)
goto out;
}
intel_bios_init(i915);
ret = intel_vga_register(i915);
if (ret)
goto out;
intel_register_dsm_handler();
ret = i915_switcheroo_register(i915);
if (ret)
goto cleanup_vga_client;
intel_power_domains_init_hw(i915, false);
intel_csr_ucode_init(i915);
ret = intel_irq_install(i915);
if (ret)
goto cleanup_csr;
/* Important: The output setup functions called by modeset_init need
* working irqs for e.g. gmbus and dp aux transfers. */
ret = intel_modeset_init(i915);
if (ret)
goto cleanup_irq;
ret = i915_gem_init(i915);
if (ret)
goto cleanup_modeset;
intel_overlay_setup(i915);
if (!HAS_DISPLAY(i915) || !INTEL_DISPLAY_ENABLED(i915))
return 0;
ret = intel_fbdev_init(&i915->drm);
if (ret)
goto cleanup_gem;
/* Only enable hotplug handling once the fbdev is fully set up. */
intel_hpd_init(i915);
intel_init_ipc(i915);
return 0;
cleanup_gem:
i915_gem_suspend(i915);
i915_gem_driver_remove(i915);
i915_gem_driver_release(i915);
cleanup_modeset:
intel_modeset_driver_remove(i915);
cleanup_irq:
intel_irq_uninstall(i915);
cleanup_csr:
intel_csr_ucode_fini(i915);
intel_power_domains_driver_remove(i915);
i915_switcheroo_unregister(i915);
cleanup_vga_client:
intel_vga_unregister(i915);
out:
return ret;
}
static void i915_driver_modeset_remove(struct drm_i915_private *i915)
{
intel_modeset_driver_remove(i915);
intel_irq_uninstall(i915);
intel_bios_driver_remove(i915);
i915_switcheroo_unregister(i915);
intel_vga_unregister(i915);
intel_csr_ucode_fini(i915);
}
static void intel_init_dpio(struct drm_i915_private *dev_priv)
{
/*
* IOSF_PORT_DPIO is used for VLV x2 PHY (DP/HDMI B and C),
* CHV x1 PHY (DP/HDMI D)
* IOSF_PORT_DPIO_2 is used for CHV x2 PHY (DP/HDMI B and C)
*/
if (IS_CHERRYVIEW(dev_priv)) {
DPIO_PHY_IOSF_PORT(DPIO_PHY0) = IOSF_PORT_DPIO_2;
DPIO_PHY_IOSF_PORT(DPIO_PHY1) = IOSF_PORT_DPIO;
} else if (IS_VALLEYVIEW(dev_priv)) {
DPIO_PHY_IOSF_PORT(DPIO_PHY0) = IOSF_PORT_DPIO;
}
}
static int i915_workqueues_init(struct drm_i915_private *dev_priv)
{
/*
* The i915 workqueue is primarily used for batched retirement of
* requests (and thus managing bo) once the task has been completed
* by the GPU. i915_retire_requests() is called directly when we
* need high-priority retirement, such as waiting for an explicit
* bo.
*
* It is also used for periodic low-priority events, such as
* idle-timers and recording error state.
*
* All tasks on the workqueue are expected to acquire the dev mutex
* so there is no point in running more than one instance of the
* workqueue at any time. Use an ordered one.
*/
dev_priv->wq = alloc_ordered_workqueue("i915", 0);
if (dev_priv->wq == NULL)
goto out_err;
dev_priv->hotplug.dp_wq = alloc_ordered_workqueue("i915-dp", 0);
if (dev_priv->hotplug.dp_wq == NULL)
goto out_free_wq;
return 0;
out_free_wq:
destroy_workqueue(dev_priv->wq);
out_err:
DRM_ERROR("Failed to allocate workqueues.\n");
return -ENOMEM;
}
static void i915_workqueues_cleanup(struct drm_i915_private *dev_priv)
{
destroy_workqueue(dev_priv->hotplug.dp_wq);
destroy_workqueue(dev_priv->wq);
}
/*
* We don't keep the workarounds for pre-production hardware, so we expect our
* driver to fail on these machines in one way or another. A little warning on
* dmesg may help both the user and the bug triagers.
*
* Our policy for removing pre-production workarounds is to keep the
* current gen workarounds as a guide to the bring-up of the next gen
* (workarounds have a habit of persisting!). Anything older than that
* should be removed along with the complications they introduce.
*/
static void intel_detect_preproduction_hw(struct drm_i915_private *dev_priv)
{
bool pre = false;
pre |= IS_HSW_EARLY_SDV(dev_priv);
pre |= IS_SKL_REVID(dev_priv, 0, SKL_REVID_F0);
pre |= IS_BXT_REVID(dev_priv, 0, BXT_REVID_B_LAST);
pre |= IS_KBL_REVID(dev_priv, 0, KBL_REVID_A0);
if (pre) {
DRM_ERROR("This is a pre-production stepping. "
"It may not be fully functional.\n");
add_taint(TAINT_MACHINE_CHECK, LOCKDEP_STILL_OK);
}
}
static int vlv_alloc_s0ix_state(struct drm_i915_private *i915)
{
if (!IS_VALLEYVIEW(i915))
return 0;
/* we write all the values in the struct, so no need to zero it out */
i915->vlv_s0ix_state = kmalloc(sizeof(*i915->vlv_s0ix_state),
GFP_KERNEL);
if (!i915->vlv_s0ix_state)
return -ENOMEM;
return 0;
}
static void vlv_free_s0ix_state(struct drm_i915_private *i915)
{
if (!i915->vlv_s0ix_state)
return;
kfree(i915->vlv_s0ix_state);
i915->vlv_s0ix_state = NULL;
}
/**
* i915_driver_early_probe - setup state not requiring device access
* @dev_priv: device private
*
* Initialize everything that is a "SW-only" state, that is state not
* requiring accessing the device or exposing the driver via kernel internal
* or userspace interfaces. Example steps belonging here: lock initialization,
* system memory allocation, setting up device specific attributes and
* function hooks not requiring accessing the device.
*/
static int i915_driver_early_probe(struct drm_i915_private *dev_priv)
{
int ret = 0;
if (i915_inject_probe_failure(dev_priv))
return -ENODEV;
drm/i915: Introduce concept of a sub-platform Concept of a sub-platform already exist in our code (like ULX and ULT platform variants and similar),implemented via the macros which check a list of device ids to determine a match. With this patch we consolidate device ids checking into a single function called during early driver load. A few low bits in the platform mask are reserved for sub-platform identification and defined as a per-platform namespace. At the same time it future proofs the platform_mask handling by preparing the code for easy extending, and tidies the very verbose WARN strings generated when IS_PLATFORM macros are embedded into a WARN type statements. v2: Fixed IS_SUBPLATFORM. Updated commit msg. v3: Chris was right, there is an ordering problem. v4: * Catch-up with new sub-platforms. * Rebase for RUNTIME_INFO. * Drop subplatform mask union tricks and convert platform_mask to an array for extensibility. v5: * Fix subplatform check. * Protect against forgetting to expand subplatform bits. * Remove platform enum tallying. * Add subplatform to error state. (Chris) * Drop macros and just use static inlines. * Remove redundant IRONLAKE_M. (Ville) v6: * Split out Ironlake change. * Optimize subplatform check. * Use __always_inline. (Lucas) * Add platform_mask comment. (Paulo) * Pass stored runtime info in error capture. (Chris) v7: * Rebased for new AML ULX device id. * Bump platform mask array size for EHL. * Stop mentioning device ids in intel_device_subplatform_init by using the trick of splitting macros i915_pciids.h. (Jani) * AML seems to be either a subplatform of KBL or CFL so express it like that. v8: * Use one device id table per subplatform. (Jani) Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Suggested-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Chris Wilson <chris@chris-wilson.co.uk> Cc: Jani Nikula <jani.nikula@intel.com> Cc: Lucas De Marchi <lucas.demarchi@intel.com> Cc: Jose Souza <jose.souza@intel.com> Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Cc: Paulo Zanoni <paulo.r.zanoni@intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Acked-by: Jani Nikula <jani.nikula@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190327142328.31780-1-tvrtko.ursulin@linux.intel.com
2019-03-27 21:23:28 +07:00
intel_device_info_subplatform_init(dev_priv);
intel_uncore_mmio_debug_init_early(&dev_priv->mmio_debug);
intel_uncore_init_early(&dev_priv->uncore, dev_priv);
spin_lock_init(&dev_priv->irq_lock);
spin_lock_init(&dev_priv->gpu_error.lock);
mutex_init(&dev_priv->backlight_lock);
mutex_init(&dev_priv->sb_lock);
drm/i915: Disable preemption and sleeping while using the punit sideband While we talk to the punit over its sideband, we need to prevent the cpu from sleeping in order to prevent a potential machine hang. Note that by itself, it appears that pm_qos_update_request (via intel_idle) doesn't provide a sufficient barrier to ensure that all core are indeed awake (out of Cstate) and that the package is awake. To do so, we need to supplement the pm_qos with a manual ping on_each_cpu. v2: Restrict the heavy-weight wakeup to just the ISOF_PORT_PUNIT, there is insufficient evidence to implicate a wider problem atm. Similarly, restrict the w/a to Valleyview, as Cherryview doesn't have an angry cadre of users. The working theory, courtesy of Ville and Hans, is the issue lies within the power delivery and so is likely to be unit and board specific and occurs when both the unit/fw require extra power at the same time as the cpu package is changing its own power state. References: https://bugzilla.kernel.org/show_bug.cgi?id=109051 References: https://bugs.freedesktop.org/show_bug.cgi?id=102657 References: https://bugzilla.kernel.org/show_bug.cgi?id=195255 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com> Cc: Hans de Goede <hdegoede@redhat.com> Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190426081725.31217-1-chris@chris-wilson.co.uk
2019-04-26 15:17:18 +07:00
pm_qos_add_request(&dev_priv->sb_qos,
PM_QOS_CPU_DMA_LATENCY, PM_QOS_DEFAULT_VALUE);
mutex_init(&dev_priv->av_mutex);
mutex_init(&dev_priv->wm.wm_mutex);
mutex_init(&dev_priv->pps_mutex);
mutex_init(&dev_priv->hdcp_comp_mutex);
drm/i915: Use SSE4.1 movntdqa to accelerate reads from WC memory This patch provides the infrastructure for performing a 16-byte aligned read from WC memory using non-temporal instructions introduced with sse4.1. Using movntdqa we can bypass the CPU caches and read directly from memory and ignoring the page attributes set on the CPU PTE i.e. negating the impact of an otherwise UC access. Copying using movntdqa from WC is almost as fast as reading from WB memory, modulo the possibility of both hitting the CPU cache or leaving the data in the CPU cache for the next consumer. (The CPU cache itself my be flushed for the region of the movntdqa and on later access the movntdqa reads from a separate internal buffer for the cacheline.) The write back to the memory is however cached. This will be used in later patches to accelerate accessing WC memory. v2: Report whether the accelerated copy is successful/possible. v3: Function alignment override was only necessary when using the function target("sse4.1") - which is not necessary for emitting movntdqa from __asm__. v4: Improve notes on CPU cache behaviour vs non-temporal stores. v5: Fix byte offsets for unrolled moves. v6: Find all remaining typos of "movntqda", use kernel_fpu_begin. Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Akash Goel <akash.goel@intel.com> Cc: Damien Lespiau <damien.lespiau@intel.com> Cc: Mika Kuoppala <mika.kuoppala@intel.com> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Link: http://patchwork.freedesktop.org/patch/msgid/1471001999-17787-2-git-send-email-chris@chris-wilson.co.uk
2016-08-12 18:39:59 +07:00
i915_memcpy_init_early(dev_priv);
intel_runtime_pm_init_early(&dev_priv->runtime_pm);
drm/i915: Use SSE4.1 movntdqa to accelerate reads from WC memory This patch provides the infrastructure for performing a 16-byte aligned read from WC memory using non-temporal instructions introduced with sse4.1. Using movntdqa we can bypass the CPU caches and read directly from memory and ignoring the page attributes set on the CPU PTE i.e. negating the impact of an otherwise UC access. Copying using movntdqa from WC is almost as fast as reading from WB memory, modulo the possibility of both hitting the CPU cache or leaving the data in the CPU cache for the next consumer. (The CPU cache itself my be flushed for the region of the movntdqa and on later access the movntdqa reads from a separate internal buffer for the cacheline.) The write back to the memory is however cached. This will be used in later patches to accelerate accessing WC memory. v2: Report whether the accelerated copy is successful/possible. v3: Function alignment override was only necessary when using the function target("sse4.1") - which is not necessary for emitting movntdqa from __asm__. v4: Improve notes on CPU cache behaviour vs non-temporal stores. v5: Fix byte offsets for unrolled moves. v6: Find all remaining typos of "movntqda", use kernel_fpu_begin. Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Akash Goel <akash.goel@intel.com> Cc: Damien Lespiau <damien.lespiau@intel.com> Cc: Mika Kuoppala <mika.kuoppala@intel.com> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Link: http://patchwork.freedesktop.org/patch/msgid/1471001999-17787-2-git-send-email-chris@chris-wilson.co.uk
2016-08-12 18:39:59 +07:00
ret = i915_workqueues_init(dev_priv);
if (ret < 0)
return ret;
ret = vlv_alloc_s0ix_state(dev_priv);
if (ret < 0)
goto err_workqueues;
intel_wopcm_init_early(&dev_priv->wopcm);
intel_gt_init_early(&dev_priv->gt, dev_priv);
i915_gem_init_early(dev_priv);
/* This must be called before any calls to HAS_PCH_* */
intel_detect_pch(dev_priv);
intel_pm_setup(dev_priv);
intel_init_dpio(dev_priv);
ret = intel_power_domains_init(dev_priv);
if (ret < 0)
goto err_gem;
intel_irq_init(dev_priv);
intel_init_display_hooks(dev_priv);
intel_init_clock_gating_hooks(dev_priv);
intel_init_audio_hooks(dev_priv);
intel_display_crc_init(dev_priv);
intel_detect_preproduction_hw(dev_priv);
return 0;
err_gem:
i915_gem_cleanup_early(dev_priv);
intel_gt_driver_late_release(&dev_priv->gt);
vlv_free_s0ix_state(dev_priv);
err_workqueues:
i915_workqueues_cleanup(dev_priv);
return ret;
}
/**
* i915_driver_late_release - cleanup the setup done in
* i915_driver_early_probe()
* @dev_priv: device private
*/
static void i915_driver_late_release(struct drm_i915_private *dev_priv)
{
intel_irq_fini(dev_priv);
intel_power_domains_cleanup(dev_priv);
i915_gem_cleanup_early(dev_priv);
intel_gt_driver_late_release(&dev_priv->gt);
vlv_free_s0ix_state(dev_priv);
i915_workqueues_cleanup(dev_priv);
drm/i915: Disable preemption and sleeping while using the punit sideband While we talk to the punit over its sideband, we need to prevent the cpu from sleeping in order to prevent a potential machine hang. Note that by itself, it appears that pm_qos_update_request (via intel_idle) doesn't provide a sufficient barrier to ensure that all core are indeed awake (out of Cstate) and that the package is awake. To do so, we need to supplement the pm_qos with a manual ping on_each_cpu. v2: Restrict the heavy-weight wakeup to just the ISOF_PORT_PUNIT, there is insufficient evidence to implicate a wider problem atm. Similarly, restrict the w/a to Valleyview, as Cherryview doesn't have an angry cadre of users. The working theory, courtesy of Ville and Hans, is the issue lies within the power delivery and so is likely to be unit and board specific and occurs when both the unit/fw require extra power at the same time as the cpu package is changing its own power state. References: https://bugzilla.kernel.org/show_bug.cgi?id=109051 References: https://bugs.freedesktop.org/show_bug.cgi?id=102657 References: https://bugzilla.kernel.org/show_bug.cgi?id=195255 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com> Cc: Hans de Goede <hdegoede@redhat.com> Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190426081725.31217-1-chris@chris-wilson.co.uk
2019-04-26 15:17:18 +07:00
pm_qos_remove_request(&dev_priv->sb_qos);
mutex_destroy(&dev_priv->sb_lock);
}
/**
* i915_driver_mmio_probe - setup device MMIO
* @dev_priv: device private
*
* Setup minimal device state necessary for MMIO accesses later in the
* initialization sequence. The setup here should avoid any other device-wide
* side effects or exposing the driver via kernel internal or user space
* interfaces.
*/
static int i915_driver_mmio_probe(struct drm_i915_private *dev_priv)
{
int ret;
if (i915_inject_probe_failure(dev_priv))
return -ENODEV;
if (i915_get_bridge_dev(dev_priv))
return -EIO;
ret = intel_uncore_init_mmio(&dev_priv->uncore);
if (ret < 0)
goto err_bridge;
/* Try to make sure MCHBAR is enabled before poking at it */
intel_setup_mchbar(dev_priv);
drm/i915/icl: Check for fused-off VDBOX and VEBOX instances In Gen11, the Video Decode engines (aka VDBOX, aka VCS, aka BSD) and the Video Enhancement engines (aka VEBOX, aka VECS) could be fused off. Also, each VDBOX and VEBOX has its own power well, which only exist if the related engine exists in the HW. Unfortunately, we have a Catch-22 situation going on: we need the blitter forcewake to read the register with the fuse info, but we cannot initialize the forcewake domains without knowin about the engines present in the HW. We workaround this problem by allowing the initialization of all forcewake domains and then pruning the fused off ones, as per the fuse information. Bspec: 20680 v2: We were shifting incorrectly for vebox disable (Vinay) v3: Assert mmio is ready and warn if we have attempted to initialize forcewake for fused-off engines (Paulo) v4: - Use INTEL_GEN in new code (Tvrtko) - Shorter local variable (Tvrtko, Michal) - Keep "if (!...) continue" style (Tvrtko) - No unnecessary BUG_ON (Tvrtko) - WARN_ON and cleanup if wrong mask (Tvrtko, Michal) - Use I915_READ_FW (Michal) - Use I915_MAX_VCS/VECS macros (Michal) v5: Rebased by Rodrigo fixing conflicts on top of: "drm/i915: Simplify intel_engines_init" v6: Fix v5. Remove info->num_rings. (by Oscar) v7: Rebase (Rodrigo). v8: - s/intel_device_info_fused_off_engines/ intel_device_info_init_mmio (Chris) - Make vdbox_disable & vebox_disable local variables (Chris) v9: - Move function declaration to intel_device_info.h (Michal) - Missing indent in bit fields definitions (Michal) - When RC6 is enabled by BIOS, the fuse register cannot be read until the blitter powerwell is awake. Shuffle where the fuse is read, prune the forcewake domains after the fact and change the commit message accordingly (Vinay, Sagar, Chris). v10: - Improved commit message (Sagar) - New line in header file (Sagar) - Specify the message in fw_domain_reset applies to ICL+ (Sagar) Cc: Paulo Zanoni <paulo.r.zanoni@intel.com> Cc: Vinay Belgaumkar <vinay.belgaumkar@intel.com> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Cc: Michal Wajdeczko <michal.wajdeczko@intel.com> Cc: Chris Wilson <chris@chris-wilson.co.uk> Cc: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com> Cc: Sagar Arun Kamble <sagar.a.kamble@intel.com> Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Signed-off-by: Oscar Mateo <oscar.mateo@intel.com> Reviewed-by: Sagar Arun Kamble <sagar.a.kamble@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20180316121456.11577-1-mika.kuoppala@linux.intel.com [Mika: soothe checkpatch on commit msg] Signed-off-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
2018-03-16 19:14:49 +07:00
intel_device_info_init_mmio(dev_priv);
intel_uncore_prune_mmio_domains(&dev_priv->uncore);
drm/i915/icl: Check for fused-off VDBOX and VEBOX instances In Gen11, the Video Decode engines (aka VDBOX, aka VCS, aka BSD) and the Video Enhancement engines (aka VEBOX, aka VECS) could be fused off. Also, each VDBOX and VEBOX has its own power well, which only exist if the related engine exists in the HW. Unfortunately, we have a Catch-22 situation going on: we need the blitter forcewake to read the register with the fuse info, but we cannot initialize the forcewake domains without knowin about the engines present in the HW. We workaround this problem by allowing the initialization of all forcewake domains and then pruning the fused off ones, as per the fuse information. Bspec: 20680 v2: We were shifting incorrectly for vebox disable (Vinay) v3: Assert mmio is ready and warn if we have attempted to initialize forcewake for fused-off engines (Paulo) v4: - Use INTEL_GEN in new code (Tvrtko) - Shorter local variable (Tvrtko, Michal) - Keep "if (!...) continue" style (Tvrtko) - No unnecessary BUG_ON (Tvrtko) - WARN_ON and cleanup if wrong mask (Tvrtko, Michal) - Use I915_READ_FW (Michal) - Use I915_MAX_VCS/VECS macros (Michal) v5: Rebased by Rodrigo fixing conflicts on top of: "drm/i915: Simplify intel_engines_init" v6: Fix v5. Remove info->num_rings. (by Oscar) v7: Rebase (Rodrigo). v8: - s/intel_device_info_fused_off_engines/ intel_device_info_init_mmio (Chris) - Make vdbox_disable & vebox_disable local variables (Chris) v9: - Move function declaration to intel_device_info.h (Michal) - Missing indent in bit fields definitions (Michal) - When RC6 is enabled by BIOS, the fuse register cannot be read until the blitter powerwell is awake. Shuffle where the fuse is read, prune the forcewake domains after the fact and change the commit message accordingly (Vinay, Sagar, Chris). v10: - Improved commit message (Sagar) - New line in header file (Sagar) - Specify the message in fw_domain_reset applies to ICL+ (Sagar) Cc: Paulo Zanoni <paulo.r.zanoni@intel.com> Cc: Vinay Belgaumkar <vinay.belgaumkar@intel.com> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Cc: Michal Wajdeczko <michal.wajdeczko@intel.com> Cc: Chris Wilson <chris@chris-wilson.co.uk> Cc: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com> Cc: Sagar Arun Kamble <sagar.a.kamble@intel.com> Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Signed-off-by: Oscar Mateo <oscar.mateo@intel.com> Reviewed-by: Sagar Arun Kamble <sagar.a.kamble@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20180316121456.11577-1-mika.kuoppala@linux.intel.com [Mika: soothe checkpatch on commit msg] Signed-off-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
2018-03-16 19:14:49 +07:00
intel_uc_init_mmio(&dev_priv->gt.uc);
ret = intel_engines_init_mmio(&dev_priv->gt);
if (ret)
goto err_uncore;
return 0;
err_uncore:
intel_teardown_mchbar(dev_priv);
intel_uncore_fini_mmio(&dev_priv->uncore);
err_bridge:
pci_dev_put(dev_priv->bridge_dev);
return ret;
}
/**
* i915_driver_mmio_release - cleanup the setup done in i915_driver_mmio_probe()
* @dev_priv: device private
*/
static void i915_driver_mmio_release(struct drm_i915_private *dev_priv)
{
intel_engines_cleanup(&dev_priv->gt);
intel_teardown_mchbar(dev_priv);
intel_uncore_fini_mmio(&dev_priv->uncore);
pci_dev_put(dev_priv->bridge_dev);
}
static void intel_sanitize_options(struct drm_i915_private *dev_priv)
{
intel_gvt_sanitize_options(dev_priv);
}
#define DRAM_TYPE_STR(type) [INTEL_DRAM_ ## type] = #type
static const char *intel_dram_type_str(enum intel_dram_type type)
{
static const char * const str[] = {
DRAM_TYPE_STR(UNKNOWN),
DRAM_TYPE_STR(DDR3),
DRAM_TYPE_STR(DDR4),
DRAM_TYPE_STR(LPDDR3),
DRAM_TYPE_STR(LPDDR4),
};
if (type >= ARRAY_SIZE(str))
type = INTEL_DRAM_UNKNOWN;
return str[type];
}
#undef DRAM_TYPE_STR
static int intel_dimm_num_devices(const struct dram_dimm_info *dimm)
{
return dimm->ranks * 64 / (dimm->width ?: 1);
}
/* Returns total GB for the whole DIMM */
static int skl_get_dimm_size(u16 val)
{
return val & SKL_DRAM_SIZE_MASK;
}
static int skl_get_dimm_width(u16 val)
{
if (skl_get_dimm_size(val) == 0)
return 0;
switch (val & SKL_DRAM_WIDTH_MASK) {
case SKL_DRAM_WIDTH_X8:
case SKL_DRAM_WIDTH_X16:
case SKL_DRAM_WIDTH_X32:
val = (val & SKL_DRAM_WIDTH_MASK) >> SKL_DRAM_WIDTH_SHIFT;
return 8 << val;
default:
MISSING_CASE(val);
return 0;
}
}
static int skl_get_dimm_ranks(u16 val)
{
if (skl_get_dimm_size(val) == 0)
return 0;
val = (val & SKL_DRAM_RANK_MASK) >> SKL_DRAM_RANK_SHIFT;
return val + 1;
}
/* Returns total GB for the whole DIMM */
static int cnl_get_dimm_size(u16 val)
{
return (val & CNL_DRAM_SIZE_MASK) / 2;
}
static int cnl_get_dimm_width(u16 val)
{
if (cnl_get_dimm_size(val) == 0)
return 0;
switch (val & CNL_DRAM_WIDTH_MASK) {
case CNL_DRAM_WIDTH_X8:
case CNL_DRAM_WIDTH_X16:
case CNL_DRAM_WIDTH_X32:
val = (val & CNL_DRAM_WIDTH_MASK) >> CNL_DRAM_WIDTH_SHIFT;
return 8 << val;
default:
MISSING_CASE(val);
return 0;
}
}
static int cnl_get_dimm_ranks(u16 val)
{
if (cnl_get_dimm_size(val) == 0)
return 0;
val = (val & CNL_DRAM_RANK_MASK) >> CNL_DRAM_RANK_SHIFT;
return val + 1;
}
static bool
skl_is_16gb_dimm(const struct dram_dimm_info *dimm)
{
/* Convert total GB to Gb per DRAM device */
return 8 * dimm->size / (intel_dimm_num_devices(dimm) ?: 1) == 16;
}
static void
skl_dram_get_dimm_info(struct drm_i915_private *dev_priv,
struct dram_dimm_info *dimm,
int channel, char dimm_name, u16 val)
{
if (INTEL_GEN(dev_priv) >= 10) {
dimm->size = cnl_get_dimm_size(val);
dimm->width = cnl_get_dimm_width(val);
dimm->ranks = cnl_get_dimm_ranks(val);
} else {
dimm->size = skl_get_dimm_size(val);
dimm->width = skl_get_dimm_width(val);
dimm->ranks = skl_get_dimm_ranks(val);
}
DRM_DEBUG_KMS("CH%u DIMM %c size: %u GB, width: X%u, ranks: %u, 16Gb DIMMs: %s\n",
channel, dimm_name, dimm->size, dimm->width, dimm->ranks,
yesno(skl_is_16gb_dimm(dimm)));
}
static int
skl_dram_get_channel_info(struct drm_i915_private *dev_priv,
struct dram_channel_info *ch,
int channel, u32 val)
{
skl_dram_get_dimm_info(dev_priv, &ch->dimm_l,
channel, 'L', val & 0xffff);
skl_dram_get_dimm_info(dev_priv, &ch->dimm_s,
channel, 'S', val >> 16);
if (ch->dimm_l.size == 0 && ch->dimm_s.size == 0) {
DRM_DEBUG_KMS("CH%u not populated\n", channel);
return -EINVAL;
}
if (ch->dimm_l.ranks == 2 || ch->dimm_s.ranks == 2)
ch->ranks = 2;
else if (ch->dimm_l.ranks == 1 && ch->dimm_s.ranks == 1)
ch->ranks = 2;
else
ch->ranks = 1;
ch->is_16gb_dimm =
skl_is_16gb_dimm(&ch->dimm_l) ||
skl_is_16gb_dimm(&ch->dimm_s);
DRM_DEBUG_KMS("CH%u ranks: %u, 16Gb DIMMs: %s\n",
channel, ch->ranks, yesno(ch->is_16gb_dimm));
return 0;
}
static bool
intel_is_dram_symmetric(const struct dram_channel_info *ch0,
const struct dram_channel_info *ch1)
{
return !memcmp(ch0, ch1, sizeof(*ch0)) &&
(ch0->dimm_s.size == 0 ||
!memcmp(&ch0->dimm_l, &ch0->dimm_s, sizeof(ch0->dimm_l)));
}
static int
skl_dram_get_channels_info(struct drm_i915_private *dev_priv)
{
struct dram_info *dram_info = &dev_priv->dram_info;
struct dram_channel_info ch0 = {}, ch1 = {};
u32 val;
int ret;
val = I915_READ(SKL_MAD_DIMM_CH0_0_0_0_MCHBAR_MCMAIN);
ret = skl_dram_get_channel_info(dev_priv, &ch0, 0, val);
if (ret == 0)
dram_info->num_channels++;
val = I915_READ(SKL_MAD_DIMM_CH1_0_0_0_MCHBAR_MCMAIN);
ret = skl_dram_get_channel_info(dev_priv, &ch1, 1, val);
if (ret == 0)
dram_info->num_channels++;
if (dram_info->num_channels == 0) {
DRM_INFO("Number of memory channels is zero\n");
return -EINVAL;
}
/*
* If any of the channel is single rank channel, worst case output
* will be same as if single rank memory, so consider single rank
* memory.
*/
if (ch0.ranks == 1 || ch1.ranks == 1)
dram_info->ranks = 1;
else
dram_info->ranks = max(ch0.ranks, ch1.ranks);
if (dram_info->ranks == 0) {
DRM_INFO("couldn't get memory rank information\n");
return -EINVAL;
}
dram_info->is_16gb_dimm = ch0.is_16gb_dimm || ch1.is_16gb_dimm;
dram_info->symmetric_memory = intel_is_dram_symmetric(&ch0, &ch1);
DRM_DEBUG_KMS("Memory configuration is symmetric? %s\n",
yesno(dram_info->symmetric_memory));
return 0;
}
static enum intel_dram_type
skl_get_dram_type(struct drm_i915_private *dev_priv)
{
u32 val;
val = I915_READ(SKL_MAD_INTER_CHANNEL_0_0_0_MCHBAR_MCMAIN);
switch (val & SKL_DRAM_DDR_TYPE_MASK) {
case SKL_DRAM_DDR_TYPE_DDR3:
return INTEL_DRAM_DDR3;
case SKL_DRAM_DDR_TYPE_DDR4:
return INTEL_DRAM_DDR4;
case SKL_DRAM_DDR_TYPE_LPDDR3:
return INTEL_DRAM_LPDDR3;
case SKL_DRAM_DDR_TYPE_LPDDR4:
return INTEL_DRAM_LPDDR4;
default:
MISSING_CASE(val);
return INTEL_DRAM_UNKNOWN;
}
}
static int
skl_get_dram_info(struct drm_i915_private *dev_priv)
{
struct dram_info *dram_info = &dev_priv->dram_info;
u32 mem_freq_khz, val;
int ret;
dram_info->type = skl_get_dram_type(dev_priv);
DRM_DEBUG_KMS("DRAM type: %s\n", intel_dram_type_str(dram_info->type));
ret = skl_dram_get_channels_info(dev_priv);
if (ret)
return ret;
val = I915_READ(SKL_MC_BIOS_DATA_0_0_0_MCHBAR_PCU);
mem_freq_khz = DIV_ROUND_UP((val & SKL_REQ_DATA_MASK) *
SKL_MEMORY_FREQ_MULTIPLIER_HZ, 1000);
dram_info->bandwidth_kbps = dram_info->num_channels *
mem_freq_khz * 8;
if (dram_info->bandwidth_kbps == 0) {
DRM_INFO("Couldn't get system memory bandwidth\n");
return -EINVAL;
}
dram_info->valid = true;
return 0;
}
/* Returns Gb per DRAM device */
static int bxt_get_dimm_size(u32 val)
{
switch (val & BXT_DRAM_SIZE_MASK) {
case BXT_DRAM_SIZE_4GBIT:
return 4;
case BXT_DRAM_SIZE_6GBIT:
return 6;
case BXT_DRAM_SIZE_8GBIT:
return 8;
case BXT_DRAM_SIZE_12GBIT:
return 12;
case BXT_DRAM_SIZE_16GBIT:
return 16;
default:
MISSING_CASE(val);
return 0;
}
}
static int bxt_get_dimm_width(u32 val)
{
if (!bxt_get_dimm_size(val))
return 0;
val = (val & BXT_DRAM_WIDTH_MASK) >> BXT_DRAM_WIDTH_SHIFT;
return 8 << val;
}
static int bxt_get_dimm_ranks(u32 val)
{
if (!bxt_get_dimm_size(val))
return 0;
switch (val & BXT_DRAM_RANK_MASK) {
case BXT_DRAM_RANK_SINGLE:
return 1;
case BXT_DRAM_RANK_DUAL:
return 2;
default:
MISSING_CASE(val);
return 0;
}
}
static enum intel_dram_type bxt_get_dimm_type(u32 val)
{
if (!bxt_get_dimm_size(val))
return INTEL_DRAM_UNKNOWN;
switch (val & BXT_DRAM_TYPE_MASK) {
case BXT_DRAM_TYPE_DDR3:
return INTEL_DRAM_DDR3;
case BXT_DRAM_TYPE_LPDDR3:
return INTEL_DRAM_LPDDR3;
case BXT_DRAM_TYPE_DDR4:
return INTEL_DRAM_DDR4;
case BXT_DRAM_TYPE_LPDDR4:
return INTEL_DRAM_LPDDR4;
default:
MISSING_CASE(val);
return INTEL_DRAM_UNKNOWN;
}
}
static void bxt_get_dimm_info(struct dram_dimm_info *dimm,
u32 val)
{
dimm->width = bxt_get_dimm_width(val);
dimm->ranks = bxt_get_dimm_ranks(val);
/*
* Size in register is Gb per DRAM device. Convert to total
* GB to match the way we report this for non-LP platforms.
*/
dimm->size = bxt_get_dimm_size(val) * intel_dimm_num_devices(dimm) / 8;
}
static int
bxt_get_dram_info(struct drm_i915_private *dev_priv)
{
struct dram_info *dram_info = &dev_priv->dram_info;
u32 dram_channels;
u32 mem_freq_khz, val;
u8 num_active_channels;
int i;
val = I915_READ(BXT_P_CR_MC_BIOS_REQ_0_0_0);
mem_freq_khz = DIV_ROUND_UP((val & BXT_REQ_DATA_MASK) *
BXT_MEMORY_FREQ_MULTIPLIER_HZ, 1000);
dram_channels = val & BXT_DRAM_CHANNEL_ACTIVE_MASK;
num_active_channels = hweight32(dram_channels);
/* Each active bit represents 4-byte channel */
dram_info->bandwidth_kbps = (mem_freq_khz * num_active_channels * 4);
if (dram_info->bandwidth_kbps == 0) {
DRM_INFO("Couldn't get system memory bandwidth\n");
return -EINVAL;
}
/*
* Now read each DUNIT8/9/10/11 to check the rank of each dimms.
*/
for (i = BXT_D_CR_DRP0_DUNIT_START; i <= BXT_D_CR_DRP0_DUNIT_END; i++) {
struct dram_dimm_info dimm;
enum intel_dram_type type;
val = I915_READ(BXT_D_CR_DRP0_DUNIT(i));
if (val == 0xFFFFFFFF)
continue;
dram_info->num_channels++;
bxt_get_dimm_info(&dimm, val);
type = bxt_get_dimm_type(val);
WARN_ON(type != INTEL_DRAM_UNKNOWN &&
dram_info->type != INTEL_DRAM_UNKNOWN &&
dram_info->type != type);
DRM_DEBUG_KMS("CH%u DIMM size: %u GB, width: X%u, ranks: %u, type: %s\n",
i - BXT_D_CR_DRP0_DUNIT_START,
dimm.size, dimm.width, dimm.ranks,
intel_dram_type_str(type));
/*
* If any of the channel is single rank channel,
* worst case output will be same as if single rank
* memory, so consider single rank memory.
*/
if (dram_info->ranks == 0)
dram_info->ranks = dimm.ranks;
else if (dimm.ranks == 1)
dram_info->ranks = 1;
if (type != INTEL_DRAM_UNKNOWN)
dram_info->type = type;
}
if (dram_info->type == INTEL_DRAM_UNKNOWN ||
dram_info->ranks == 0) {
DRM_INFO("couldn't get memory information\n");
return -EINVAL;
}
dram_info->valid = true;
return 0;
}
static void
intel_get_dram_info(struct drm_i915_private *dev_priv)
{
struct dram_info *dram_info = &dev_priv->dram_info;
int ret;
/*
* Assume 16Gb DIMMs are present until proven otherwise.
* This is only used for the level 0 watermark latency
* w/a which does not apply to bxt/glk.
*/
dram_info->is_16gb_dimm = !IS_GEN9_LP(dev_priv);
if (INTEL_GEN(dev_priv) < 9 || !HAS_DISPLAY(dev_priv))
return;
if (IS_GEN9_LP(dev_priv))
ret = bxt_get_dram_info(dev_priv);
else
ret = skl_get_dram_info(dev_priv);
if (ret)
return;
DRM_DEBUG_KMS("DRAM bandwidth: %u kBps, channels: %u\n",
dram_info->bandwidth_kbps,
dram_info->num_channels);
DRM_DEBUG_KMS("DRAM ranks: %u, 16Gb DIMMs: %s\n",
dram_info->ranks, yesno(dram_info->is_16gb_dimm));
}
static u32 gen9_edram_size_mb(struct drm_i915_private *dev_priv, u32 cap)
{
static const u8 ways[8] = { 4, 8, 12, 16, 16, 16, 16, 16 };
static const u8 sets[4] = { 1, 1, 2, 2 };
return EDRAM_NUM_BANKS(cap) *
ways[EDRAM_WAYS_IDX(cap)] *
sets[EDRAM_SETS_IDX(cap)];
}
static void edram_detect(struct drm_i915_private *dev_priv)
{
u32 edram_cap = 0;
if (!(IS_HASWELL(dev_priv) ||
IS_BROADWELL(dev_priv) ||
INTEL_GEN(dev_priv) >= 9))
return;
edram_cap = __raw_uncore_read32(&dev_priv->uncore, HSW_EDRAM_CAP);
/* NB: We can't write IDICR yet because we don't have gt funcs set up */
if (!(edram_cap & EDRAM_ENABLED))
return;
/*
* The needed capability bits for size calculation are not there with
* pre gen9 so return 128MB always.
*/
if (INTEL_GEN(dev_priv) < 9)
dev_priv->edram_size_mb = 128;
else
dev_priv->edram_size_mb =
gen9_edram_size_mb(dev_priv, edram_cap);
dev_info(dev_priv->drm.dev,
"Found %uMB of eDRAM\n", dev_priv->edram_size_mb);
}
/**
* i915_driver_hw_probe - setup state requiring device access
* @dev_priv: device private
*
* Setup state that requires accessing the device, but doesn't require
* exposing the driver via kernel internal or userspace interfaces.
*/
static int i915_driver_hw_probe(struct drm_i915_private *dev_priv)
{
struct pci_dev *pdev = dev_priv->drm.pdev;
int ret;
if (i915_inject_probe_failure(dev_priv))
return -ENODEV;
intel_device_info_runtime_init(dev_priv);
if (HAS_PPGTT(dev_priv)) {
if (intel_vgpu_active(dev_priv) &&
!intel_vgpu_has_full_ppgtt(dev_priv)) {
i915_report_error(dev_priv,
"incompatible vGPU found, support for isolated ppGTT required\n");
return -ENXIO;
}
}
if (HAS_EXECLISTS(dev_priv)) {
/*
* Older GVT emulation depends upon intercepting CSB mmio,
* which we no longer use, preferring to use the HWSP cache
* instead.
*/
if (intel_vgpu_active(dev_priv) &&
!intel_vgpu_has_hwsp_emulation(dev_priv)) {
i915_report_error(dev_priv,
"old vGPU host found, support for HWSP emulation required\n");
return -ENXIO;
}
}
intel_sanitize_options(dev_priv);
/* needs to be done before ggtt probe */
edram_detect(dev_priv);
i915_perf_init(dev_priv);
ret = i915_ggtt_probe_hw(dev_priv);
if (ret)
goto err_perf;
ret = drm_fb_helper_remove_conflicting_pci_framebuffers(pdev, "inteldrmfb");
if (ret)
goto err_ggtt;
ret = i915_ggtt_init_hw(dev_priv);
if (ret)
goto err_ggtt;
ret = intel_memory_regions_hw_probe(dev_priv);
if (ret)
goto err_ggtt;
intel_gt_init_hw_early(&dev_priv->gt, &dev_priv->ggtt);
ret = i915_ggtt_enable_hw(dev_priv);
if (ret) {
DRM_ERROR("failed to enable GGTT\n");
goto err_mem_regions;
}
pci_set_master(pdev);
drm/i915: Call dma_set_max_seg_size() in i915_driver_hw_probe() Currently, we don't call dma_set_max_seg_size() for i915 because we intentionally do not limit the segment length that the device supports. However, this results in a warning being emitted if we try to map anything larger than SZ_64K on a kernel with CONFIG_DMA_API_DEBUG_SG enabled: [ 7.751926] DMA-API: i915 0000:00:02.0: mapping sg segment longer than device claims to support [len=98304] [max=65536] [ 7.751934] WARNING: CPU: 5 PID: 474 at kernel/dma/debug.c:1220 debug_dma_map_sg+0x20f/0x340 This was originally brought up on https://bugs.freedesktop.org/show_bug.cgi?id=108517 , and the consensus there was it wasn't really useful to set a limit (and that dma-debug isn't really all that useful for i915 in the first place). Unfortunately though, CONFIG_DMA_API_DEBUG_SG is enabled in the debug configs for various distro kernels. Since a WARN_ON() will disable automatic problem reporting (and cause any CI with said option enabled to start complaining), we really should just fix the problem. Note that as me and Chris Wilson discussed, the other solution for this would be to make DMA-API not make such assumptions when a driver hasn't explicitly set a maximum segment size. But, taking a look at the commit which originally introduced this behavior, commit 78c47830a5cb ("dma-debug: check scatterlist segments"), there is an explicit mention of this assumption and how it applies to devices with no segment size: Conversely, devices which are less limited than the rather conservative defaults, or indeed have no limitations at all (e.g. GPUs with their own internal MMU), should be encouraged to set appropriate dma_parms, as they may get more efficient DMA mapping performance out of it. So unless there's any concerns (I'm open to discussion!), let's just follow suite and call dma_set_max_seg_size() with UINT_MAX as our limit to silence any warnings. Changes since v3: * Drop patch for enabling CONFIG_DMA_API_DEBUG_SG in CI. It looks like just turning it on causes the kernel to spit out bogus WARN_ONs() during some igt tests which would otherwise require teaching igt to disable the various DMA-API debugging options causing this. This is too much work to be worth it, since DMA-API debugging is useless for us. So, we'll just settle with this single patch to squelch WARN_ONs() during driver load for users that have CONFIG_DMA_API_DEBUG_SG turned on for some reason. * Move dma_set_max_seg_size() call into i915_driver_hw_probe() - Chris Wilson Signed-off-by: Lyude Paul <lyude@redhat.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: <stable@vger.kernel.org> # v4.18+ Link: https://patchwork.freedesktop.org/patch/msgid/20190823205251.14298-1-lyude@redhat.com
2019-08-24 03:52:51 +07:00
/*
* We don't have a max segment size, so set it to the max so sg's
* debugging layer doesn't complain
*/
dma_set_max_seg_size(&pdev->dev, UINT_MAX);
/* overlay on gen2 is broken and can't address above 1G */
if (IS_GEN(dev_priv, 2)) {
ret = dma_set_coherent_mask(&pdev->dev, DMA_BIT_MASK(30));
if (ret) {
DRM_ERROR("failed to set DMA mask\n");
goto err_mem_regions;
}
}
/* 965GM sometimes incorrectly writes to hardware status page (HWS)
* using 32bit addressing, overwriting memory if HWS is located
* above 4GB.
*
* The documentation also mentions an issue with undefined
* behaviour if any general state is accessed within a page above 4GB,
* which also needs to be handled carefully.
*/
if (IS_I965G(dev_priv) || IS_I965GM(dev_priv)) {
ret = dma_set_coherent_mask(&pdev->dev, DMA_BIT_MASK(32));
if (ret) {
DRM_ERROR("failed to set DMA mask\n");
goto err_mem_regions;
}
}
pm_qos_add_request(&dev_priv->pm_qos, PM_QOS_CPU_DMA_LATENCY,
PM_QOS_DEFAULT_VALUE);
2018-12-03 20:33:19 +07:00
intel_gt_init_workarounds(dev_priv);
/* On the 945G/GM, the chipset reports the MSI capability on the
* integrated graphics even though the support isn't actually there
* according to the published specs. It doesn't appear to function
* correctly in testing on 945G.
* This may be a side effect of MSI having been made available for PEG
* and the registers being closely associated.
*
* According to chipset errata, on the 965GM, MSI interrupts may
drm/i915: Disable MSI for all pre-gen5 We have pretty clear evidence that MSIs are getting lost on g4x and somehow the interrupt logic doesn't seem to recover from that state even if we try hard to clear the IIR. Disabling IER around the normal IIR clearing in the irq handler isn't sufficient to avoid this, so the problem really seems to be further up the interrupt chain. This should guarantee that there's always an edge if any IIR bits are set after the interrupt handler is done, which should normally guarantee that the CPU interrupt is generated. That approach seems to work perfectly on VLV/CHV, but apparently not on g4x. MSI is documented to be broken on 965gm at least. The chipset spec says MSI is defeatured because interrupts can be delayed or lost, which fits well with what we're seeing on g4x. Previously we've already disabled GMBUS interrupts on g4x because somehow GMBUS manages to raise legacy interrupts even when MSI is enabled. Since there's such widespread MSI breakahge all over in the pre-gen5 land let's just give up on MSI on these platforms. Seqno reporting might be negatively affected by this since the legcy interrupts aren't guaranteed to be ordered with the seqno writes, whereas MSI interrupts may be? But an occasioanlly missed seqno seems like a small price to pay for generally working interrupts. Cc: stable@vger.kernel.org Cc: Diego Viola <diego.viola@gmail.com> Tested-by: Diego Viola <diego.viola@gmail.com> Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=101261 Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: http://patchwork.freedesktop.org/patch/msgid/20170626203051.28480-1-ville.syrjala@linux.intel.com Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2017-06-27 03:30:51 +07:00
* be lost or delayed, and was defeatured. MSI interrupts seem to
* get lost on g4x as well, and interrupt delivery seems to stay
* properly dead afterwards. So we'll just disable them for all
* pre-gen5 chipsets.
*
* 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 legacy irq no. is shared with another
* device. The kernel then disables that interrupt source and so
* prevents the other device from working properly.
*/
drm/i915: Disable MSI for all pre-gen5 We have pretty clear evidence that MSIs are getting lost on g4x and somehow the interrupt logic doesn't seem to recover from that state even if we try hard to clear the IIR. Disabling IER around the normal IIR clearing in the irq handler isn't sufficient to avoid this, so the problem really seems to be further up the interrupt chain. This should guarantee that there's always an edge if any IIR bits are set after the interrupt handler is done, which should normally guarantee that the CPU interrupt is generated. That approach seems to work perfectly on VLV/CHV, but apparently not on g4x. MSI is documented to be broken on 965gm at least. The chipset spec says MSI is defeatured because interrupts can be delayed or lost, which fits well with what we're seeing on g4x. Previously we've already disabled GMBUS interrupts on g4x because somehow GMBUS manages to raise legacy interrupts even when MSI is enabled. Since there's such widespread MSI breakahge all over in the pre-gen5 land let's just give up on MSI on these platforms. Seqno reporting might be negatively affected by this since the legcy interrupts aren't guaranteed to be ordered with the seqno writes, whereas MSI interrupts may be? But an occasioanlly missed seqno seems like a small price to pay for generally working interrupts. Cc: stable@vger.kernel.org Cc: Diego Viola <diego.viola@gmail.com> Tested-by: Diego Viola <diego.viola@gmail.com> Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=101261 Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: http://patchwork.freedesktop.org/patch/msgid/20170626203051.28480-1-ville.syrjala@linux.intel.com Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2017-06-27 03:30:51 +07:00
if (INTEL_GEN(dev_priv) >= 5) {
if (pci_enable_msi(pdev) < 0)
DRM_DEBUG_DRIVER("can't enable MSI");
}
ret = intel_gvt_init(dev_priv);
if (ret)
drm/i915: Unwind HW init after GVT setup failure Following intel_gvt_init() failure, we missed unwinding our setup leaving pointers dangling past the module unload. For our example, the pm_qos: [ 441.057615] top: 000000006b3baf1c, n: 0000000054d8ef33, p: 0000000097cdf1a2 prev: 0000000054d8ef33, n: 0000000097cdf1a2, p: 000000006b3baf1c next: 0000000097cdf1a2, n: 000000006de8fc8b, p: 0000000081087253 [ 441.057627] WARNING: CPU: 4 PID: 9277 at lib/plist.c:42 plist_check_prev_next+0x2d/0x40 [ 441.057628] Modules linked in: i915(+) vgem snd_hda_codec_hdmi snd_hda_codec_realtek snd_hda_codec_generic x86_pkg_temp_thermal intel_powerclamp coretemp crct10dif_pclmul crc32_pclmul ghash_clmulni_intel snd_hda_codec snd_hwdep snd_hda_core e1000e snd_pcm mei_me mei prime_numbers [last unloaded: i915] [ 441.057652] CPU: 4 PID: 9277 Comm: drv_selftest Tainted: G U 4.18.0-rc4-CI-CI_DRM_4464+ #1 [ 441.057653] Hardware name: System manufacturer System Product Name/Z170 PRO GAMING, BIOS 3402 04/26/2017 [ 441.057656] RIP: 0010:plist_check_prev_next+0x2d/0x40 [ 441.057657] Code: 08 48 39 f0 74 2b 49 89 f0 48 8b 4f 08 50 ff 32 52 48 89 fe 41 ff 70 08 48 8b 17 48 c7 c7 d8 ae 14 82 4d 8b 08 e8 63 0e 76 ff <0f> 0b 48 83 c4 20 c3 48 39 10 75 d0 f3 c3 0f 1f 44 00 00 41 54 55 [ 441.057717] RSP: 0018:ffffc900003a3a68 EFLAGS: 00010082 [ 441.057720] RAX: 0000000000000000 RBX: ffff8802193978c0 RCX: 0000000000000002 [ 441.057721] RDX: 0000000080000002 RSI: ffffffff820c65a4 RDI: 00000000ffffffff [ 441.057722] RBP: ffff8802193978c0 R08: 0000000000000000 R09: 0000000000000001 [ 441.057724] R10: ffffc900003a3a70 R11: 0000000000000000 R12: ffffffff82243de0 [ 441.057725] R13: ffffffff82243de0 R14: ffff88021a6c78c0 R15: 0000000077359400 [ 441.057726] FS: 00007fc23a4a9980(0000) GS:ffff880236d00000(0000) knlGS:0000000000000000 [ 441.057728] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 441.057729] CR2: 0000563e4503d038 CR3: 0000000138f86005 CR4: 00000000003606e0 [ 441.057730] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 441.057731] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 [ 441.057732] Call Trace: [ 441.057736] plist_check_list+0x2e/0x40 [ 441.057738] plist_add+0x23/0x130 [ 441.057743] pm_qos_update_target+0x1bd/0x2f0 [ 441.057771] i915_driver_load+0xec4/0x1060 [i915] [ 441.057775] ? trace_hardirqs_on_caller+0xe0/0x1b0 [ 441.057800] i915_pci_probe+0x29/0x90 [i915] [ 441.057804] pci_device_probe+0xa1/0x130 [ 441.057807] driver_probe_device+0x306/0x480 [ 441.057810] __driver_attach+0xdb/0x100 [ 441.057812] ? driver_probe_device+0x480/0x480 [ 441.057813] ? driver_probe_device+0x480/0x480 [ 441.057816] bus_for_each_dev+0x74/0xc0 [ 441.057819] bus_add_driver+0x15f/0x250 [ 441.057821] ? 0xffffffffa0696000 [ 441.057823] driver_register+0x56/0xe0 [ 441.057825] ? 0xffffffffa0696000 [ 441.057827] do_one_initcall+0x58/0x370 [ 441.057830] ? do_init_module+0x1d/0x1ea [ 441.057832] ? rcu_read_lock_sched_held+0x6f/0x80 [ 441.057834] ? kmem_cache_alloc_trace+0x282/0x2e0 [ 441.057838] do_init_module+0x56/0x1ea [ 441.057841] load_module+0x2435/0x2b20 [ 441.057852] ? __se_sys_finit_module+0xd3/0xf0 [ 441.057854] __se_sys_finit_module+0xd3/0xf0 [ 441.057861] do_syscall_64+0x55/0x190 [ 441.057863] entry_SYSCALL_64_after_hwframe+0x49/0xbe [ 441.057865] RIP: 0033:0x7fc239d75839 [ 441.057866] Code: 00 f3 c3 66 2e 0f 1f 84 00 00 00 00 00 0f 1f 40 00 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 8b 0d 1f f6 2c 00 f7 d8 64 89 01 48 [ 441.057927] RSP: 002b:00007fffb7825d38 EFLAGS: 00000246 ORIG_RAX: 0000000000000139 [ 441.057930] RAX: ffffffffffffffda RBX: 0000563e45035dd0 RCX: 00007fc239d75839 [ 441.057931] RDX: 0000000000000000 RSI: 0000563e4502f8a0 RDI: 0000000000000004 [ 441.057932] RBP: 0000563e4502f8a0 R08: 0000000000000004 R09: 0000000000000000 [ 441.057933] R10: 00007fffb7825ea0 R11: 0000000000000246 R12: 0000000000000000 [ 441.057934] R13: 0000563e4502f690 R14: 0000000000000000 R15: 000000000000003f [ 441.057940] irq event stamp: 231338 [ 441.057943] hardirqs last enabled at (231337): [<ffffffff8193e3fc>] _raw_spin_unlock_irqrestore+0x4c/0x60 [ 441.057944] hardirqs last disabled at (231338): [<ffffffff8193e26d>] _raw_spin_lock_irqsave+0xd/0x50 [ 441.057947] softirqs last enabled at (231024): [<ffffffff81c0034f>] __do_softirq+0x34f/0x505 [ 441.057949] softirqs last disabled at (231005): [<ffffffff8108c7b9>] irq_exit+0xa9/0xc0 [ 441.057951] WARNING: CPU: 4 PID: 9277 at lib/plist.c:42 plist_check_prev_next+0x2d/0x40 v2: Add a load failure point to intel_gvt_init() so that we always exercise this path in future. Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=107129 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Matthew Auld <matthew.auld@intel.com> Cc: Michał Winiarski <michal.winiarski@intel.com> Reviewed-by: Michał Winiarski <michal.winiarski@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20180710143821.1889-1-chris@chris-wilson.co.uk
2018-07-10 21:38:21 +07:00
goto err_msi;
intel_opregion_setup(dev_priv);
/*
* Fill the dram structure to get the system raw bandwidth and
* dram info. This will be used for memory latency calculation.
*/
intel_get_dram_info(dev_priv);
drm/i915: Make sure we have enough memory bandwidth on ICL ICL has so many planes that it can easily exceed the maximum effective memory bandwidth of the system. We must therefore check that we don't exceed that limit. The algorithm is very magic number heavy and lacks sufficient explanation for now. We also have no sane way to query the memory clock and timings, so we must rely on a combination of raw readout from the memory controller and hardcoded assumptions. The memory controller values obviously change as the system jumps between the different SAGV points, so we try to stabilize it first by disabling SAGV for the duration of the readout. The utilized bandwidth is tracked via a device wide atomic private object. That is actually not robust because we can't afford to enforce strict global ordering between the pipes. Thus I think I'll need to change this to simply chop up the available bandwidth between all the active pipes. Each pipe can then do whatever it wants as long as it doesn't exceed its budget. That scheme will also require that we assume that any number of planes could be active at any time. TODO: make it robust and deal with all the open questions v2: Sleep longer after disabling SAGV v3: Poll for the dclk to get raised (seen it take 250ms!) If the system has 2133MT/s memory then we pointlessly wait one full second :( v4: Use the new pcode interface to get the qgv points rather that using hardcoded numbers v5: Move the pcode stuff into intel_bw.c (Matt) s/intel_sagv_info/intel_qgv_info/ Do the NV12/P010 as per spec for now (Matt) s/IS_ICELAKE/IS_GEN11/ v6: Ignore bandwidth limits if the pcode query fails Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Matt Roper <matthew.d.roper@intel.com> Acked-by: Clint Taylor <Clinton.A.Taylor@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190524153614.32410-1-ville.syrjala@linux.intel.com
2019-05-24 22:36:14 +07:00
intel_bw_init_hw(dev_priv);
return 0;
drm/i915: Unwind HW init after GVT setup failure Following intel_gvt_init() failure, we missed unwinding our setup leaving pointers dangling past the module unload. For our example, the pm_qos: [ 441.057615] top: 000000006b3baf1c, n: 0000000054d8ef33, p: 0000000097cdf1a2 prev: 0000000054d8ef33, n: 0000000097cdf1a2, p: 000000006b3baf1c next: 0000000097cdf1a2, n: 000000006de8fc8b, p: 0000000081087253 [ 441.057627] WARNING: CPU: 4 PID: 9277 at lib/plist.c:42 plist_check_prev_next+0x2d/0x40 [ 441.057628] Modules linked in: i915(+) vgem snd_hda_codec_hdmi snd_hda_codec_realtek snd_hda_codec_generic x86_pkg_temp_thermal intel_powerclamp coretemp crct10dif_pclmul crc32_pclmul ghash_clmulni_intel snd_hda_codec snd_hwdep snd_hda_core e1000e snd_pcm mei_me mei prime_numbers [last unloaded: i915] [ 441.057652] CPU: 4 PID: 9277 Comm: drv_selftest Tainted: G U 4.18.0-rc4-CI-CI_DRM_4464+ #1 [ 441.057653] Hardware name: System manufacturer System Product Name/Z170 PRO GAMING, BIOS 3402 04/26/2017 [ 441.057656] RIP: 0010:plist_check_prev_next+0x2d/0x40 [ 441.057657] Code: 08 48 39 f0 74 2b 49 89 f0 48 8b 4f 08 50 ff 32 52 48 89 fe 41 ff 70 08 48 8b 17 48 c7 c7 d8 ae 14 82 4d 8b 08 e8 63 0e 76 ff <0f> 0b 48 83 c4 20 c3 48 39 10 75 d0 f3 c3 0f 1f 44 00 00 41 54 55 [ 441.057717] RSP: 0018:ffffc900003a3a68 EFLAGS: 00010082 [ 441.057720] RAX: 0000000000000000 RBX: ffff8802193978c0 RCX: 0000000000000002 [ 441.057721] RDX: 0000000080000002 RSI: ffffffff820c65a4 RDI: 00000000ffffffff [ 441.057722] RBP: ffff8802193978c0 R08: 0000000000000000 R09: 0000000000000001 [ 441.057724] R10: ffffc900003a3a70 R11: 0000000000000000 R12: ffffffff82243de0 [ 441.057725] R13: ffffffff82243de0 R14: ffff88021a6c78c0 R15: 0000000077359400 [ 441.057726] FS: 00007fc23a4a9980(0000) GS:ffff880236d00000(0000) knlGS:0000000000000000 [ 441.057728] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 441.057729] CR2: 0000563e4503d038 CR3: 0000000138f86005 CR4: 00000000003606e0 [ 441.057730] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 441.057731] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 [ 441.057732] Call Trace: [ 441.057736] plist_check_list+0x2e/0x40 [ 441.057738] plist_add+0x23/0x130 [ 441.057743] pm_qos_update_target+0x1bd/0x2f0 [ 441.057771] i915_driver_load+0xec4/0x1060 [i915] [ 441.057775] ? trace_hardirqs_on_caller+0xe0/0x1b0 [ 441.057800] i915_pci_probe+0x29/0x90 [i915] [ 441.057804] pci_device_probe+0xa1/0x130 [ 441.057807] driver_probe_device+0x306/0x480 [ 441.057810] __driver_attach+0xdb/0x100 [ 441.057812] ? driver_probe_device+0x480/0x480 [ 441.057813] ? driver_probe_device+0x480/0x480 [ 441.057816] bus_for_each_dev+0x74/0xc0 [ 441.057819] bus_add_driver+0x15f/0x250 [ 441.057821] ? 0xffffffffa0696000 [ 441.057823] driver_register+0x56/0xe0 [ 441.057825] ? 0xffffffffa0696000 [ 441.057827] do_one_initcall+0x58/0x370 [ 441.057830] ? do_init_module+0x1d/0x1ea [ 441.057832] ? rcu_read_lock_sched_held+0x6f/0x80 [ 441.057834] ? kmem_cache_alloc_trace+0x282/0x2e0 [ 441.057838] do_init_module+0x56/0x1ea [ 441.057841] load_module+0x2435/0x2b20 [ 441.057852] ? __se_sys_finit_module+0xd3/0xf0 [ 441.057854] __se_sys_finit_module+0xd3/0xf0 [ 441.057861] do_syscall_64+0x55/0x190 [ 441.057863] entry_SYSCALL_64_after_hwframe+0x49/0xbe [ 441.057865] RIP: 0033:0x7fc239d75839 [ 441.057866] Code: 00 f3 c3 66 2e 0f 1f 84 00 00 00 00 00 0f 1f 40 00 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 8b 0d 1f f6 2c 00 f7 d8 64 89 01 48 [ 441.057927] RSP: 002b:00007fffb7825d38 EFLAGS: 00000246 ORIG_RAX: 0000000000000139 [ 441.057930] RAX: ffffffffffffffda RBX: 0000563e45035dd0 RCX: 00007fc239d75839 [ 441.057931] RDX: 0000000000000000 RSI: 0000563e4502f8a0 RDI: 0000000000000004 [ 441.057932] RBP: 0000563e4502f8a0 R08: 0000000000000004 R09: 0000000000000000 [ 441.057933] R10: 00007fffb7825ea0 R11: 0000000000000246 R12: 0000000000000000 [ 441.057934] R13: 0000563e4502f690 R14: 0000000000000000 R15: 000000000000003f [ 441.057940] irq event stamp: 231338 [ 441.057943] hardirqs last enabled at (231337): [<ffffffff8193e3fc>] _raw_spin_unlock_irqrestore+0x4c/0x60 [ 441.057944] hardirqs last disabled at (231338): [<ffffffff8193e26d>] _raw_spin_lock_irqsave+0xd/0x50 [ 441.057947] softirqs last enabled at (231024): [<ffffffff81c0034f>] __do_softirq+0x34f/0x505 [ 441.057949] softirqs last disabled at (231005): [<ffffffff8108c7b9>] irq_exit+0xa9/0xc0 [ 441.057951] WARNING: CPU: 4 PID: 9277 at lib/plist.c:42 plist_check_prev_next+0x2d/0x40 v2: Add a load failure point to intel_gvt_init() so that we always exercise this path in future. Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=107129 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Matthew Auld <matthew.auld@intel.com> Cc: Michał Winiarski <michal.winiarski@intel.com> Reviewed-by: Michał Winiarski <michal.winiarski@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20180710143821.1889-1-chris@chris-wilson.co.uk
2018-07-10 21:38:21 +07:00
err_msi:
if (pdev->msi_enabled)
pci_disable_msi(pdev);
pm_qos_remove_request(&dev_priv->pm_qos);
err_mem_regions:
intel_memory_regions_driver_release(dev_priv);
err_ggtt:
i915_ggtt_driver_release(dev_priv);
err_perf:
i915_perf_fini(dev_priv);
return ret;
}
/**
* i915_driver_hw_remove - cleanup the setup done in i915_driver_hw_probe()
* @dev_priv: device private
*/
static void i915_driver_hw_remove(struct drm_i915_private *dev_priv)
{
struct pci_dev *pdev = dev_priv->drm.pdev;
i915_perf_fini(dev_priv);
if (pdev->msi_enabled)
pci_disable_msi(pdev);
pm_qos_remove_request(&dev_priv->pm_qos);
}
/**
* i915_driver_register - register the driver with the rest of the system
* @dev_priv: device private
*
* Perform any steps necessary to make the driver available via kernel
* internal or userspace interfaces.
*/
static void i915_driver_register(struct drm_i915_private *dev_priv)
{
struct drm_device *dev = &dev_priv->drm;
i915_gem_driver_register(dev_priv);
drm/i915/pmu: Expose a PMU interface for perf queries From: Chris Wilson <chris@chris-wilson.co.uk> From: Tvrtko Ursulin <tvrtko.ursulin@intel.com> From: Dmitry Rogozhkin <dmitry.v.rogozhkin@intel.com> The first goal is to be able to measure GPU (and invidual ring) busyness without having to poll registers from userspace. (Which not only incurs holding the forcewake lock indefinitely, perturbing the system, but also runs the risk of hanging the machine.) As an alternative we can use the perf event counter interface to sample the ring registers periodically and send those results to userspace. Functionality we are exporting to userspace is via the existing perf PMU API and can be exercised via the existing tools. For example: perf stat -a -e i915/rcs0-busy/ -I 1000 Will print the render engine busynnes once per second. All the performance counters can be enumerated (perf list) and have their unit of measure correctly reported in sysfs. v1-v2 (Chris Wilson): v2: Use a common timer for the ring sampling. v3: (Tvrtko Ursulin) * Decouple uAPI from i915 engine ids. * Complete uAPI defines. * Refactor some code to helpers for clarity. * Skip sampling disabled engines. * Expose counters in sysfs. * Pass in fake regs to avoid null ptr deref in perf core. * Convert to class/instance uAPI. * Use shared driver code for rc6 residency, power and frequency. v4: (Dmitry Rogozhkin) * Register PMU with .task_ctx_nr=perf_invalid_context * Expose cpumask for the PMU with the single CPU in the mask * Properly support pmu->stop(): it should call pmu->read() * Properly support pmu->del(): it should call stop(event, PERF_EF_UPDATE) * Introduce refcounting of event subscriptions. * Make pmu.busy_stats a refcounter to avoid busy stats going away with some deleted event. * Expose cpumask for i915 PMU to avoid multiple events creation of the same type followed by counter aggregation by perf-stat. * Track CPUs getting online/offline to migrate perf context. If (likely) cpumask will initially set CPU0, CONFIG_BOOTPARAM_HOTPLUG_CPU0 will be needed to see effect of CPU status tracking. * End result is that only global events are supported and perf stat works correctly. * Deny perf driver level sampling - it is prohibited for uncore PMU. v5: (Tvrtko Ursulin) * Don't hardcode number of engine samplers. * Rewrite event ref-counting for correctness and simplicity. * Store initial counter value when starting already enabled events to correctly report values to all listeners. * Fix RC6 residency readout. * Comments, GPL header. v6: * Add missing entry to v4 changelog. * Fix accounting in CPU hotplug case by copying the approach from arch/x86/events/intel/cstate.c. (Dmitry Rogozhkin) v7: * Log failure message only on failure. * Remove CPU hotplug notification state on unregister. v8: * Fix error unwind on failed registration. * Checkpatch cleanup. v9: * Drop the energy metric, it is available via intel_rapl_perf. (Ville Syrjälä) * Use HAS_RC6(p). (Chris Wilson) * Handle unsupported non-engine events. (Dmitry Rogozhkin) * Rebase for intel_rc6_residency_ns needing caller managed runtime pm. * Drop HAS_RC6 checks from the read callback since creating those events will be rejected at init time already. * Add counter units to sysfs so perf stat output is nicer. * Cleanup the attribute tables for brevity and readability. v10: * Fixed queued accounting. v11: * Move intel_engine_lookup_user to intel_engine_cs.c * Commit update. (Joonas Lahtinen) v12: * More accurate sampling. (Chris Wilson) * Store and report frequency in MHz for better usability from perf stat. * Removed metrics: queued, interrupts, rc6 counters. * Sample engine busyness based on seqno difference only for less MMIO (and forcewake) on all platforms. (Chris Wilson) v13: * Comment spelling, use mul_u32_u32 to work around potential GCC issue and somne code alignment changes. (Chris Wilson) v14: * Rebase. v15: * Rebase for RPS refactoring. v16: * Use the dynamic slot in the CPU hotplug state machine so that we are free to setup our state as multi-instance. Previously we were re-using the CPUHP_AP_PERF_X86_UNCORE_ONLINE slot which is neither used as multi-instance, nor owned by our driver to start with. * Register the CPU hotplug handlers after the PMU, otherwise the callback will get called before the PMU is initialized which can end up in perf_pmu_migrate_context with an un-initialized base. * Added workaround for a probable bug in cpuhp core. v17: * Remove workaround for the cpuhp bug. v18: * Rebase for drm_i915_gem_engine_class getting upstream before us. v19: * Rebase. (trivial) Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Signed-off-by: Dmitry Rogozhkin <dmitry.v.rogozhkin@intel.com> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Cc: Chris Wilson <chris@chris-wilson.co.uk> Cc: Dmitry Rogozhkin <dmitry.v.rogozhkin@intel.com> Cc: Peter Zijlstra <peterz@infradead.org> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20171121181852.16128-2-tvrtko.ursulin@linux.intel.com
2017-11-22 01:18:45 +07:00
i915_pmu_register(dev_priv);
/*
* Notify a valid surface after modesetting,
* when running inside a VM.
*/
if (intel_vgpu_active(dev_priv))
I915_WRITE(vgtif_reg(display_ready), VGT_DRV_DISPLAY_READY);
/* Reveal our presence to userspace */
if (drm_dev_register(dev, 0) == 0) {
i915_debugfs_register(dev_priv);
i915_setup_sysfs(dev_priv);
/* Depends on sysfs having been initialized */
i915_perf_register(dev_priv);
} else
DRM_ERROR("Failed to register driver for userspace access!\n");
if (HAS_DISPLAY(dev_priv) && INTEL_DISPLAY_ENABLED(dev_priv)) {
/* Must be done after probing outputs */
intel_opregion_register(dev_priv);
acpi_video_register();
}
intel_gt_driver_register(&dev_priv->gt);
intel_audio_init(dev_priv);
/*
* Some ports require correctly set-up hpd registers for detection to
* work properly (leading to ghost connected connector status), e.g. VGA
* on gm45. Hence we can only set up the initial fbdev config after hpd
* irqs are fully enabled. We do it last so that the async config
* cannot run before the connectors are registered.
*/
intel_fbdev_initial_config_async(dev);
/*
* We need to coordinate the hotplugs with the asynchronous fbdev
* configuration, for which we use the fbdev->async_cookie.
*/
if (HAS_DISPLAY(dev_priv) && INTEL_DISPLAY_ENABLED(dev_priv))
drm_kms_helper_poll_init(dev);
intel_power_domains_enable(dev_priv);
intel_runtime_pm_enable(&dev_priv->runtime_pm);
}
/**
* i915_driver_unregister - cleanup the registration done in i915_driver_regiser()
* @dev_priv: device private
*/
static void i915_driver_unregister(struct drm_i915_private *dev_priv)
{
intel_runtime_pm_disable(&dev_priv->runtime_pm);
intel_power_domains_disable(dev_priv);
intel_fbdev_unregister(dev_priv);
intel_audio_deinit(dev_priv);
/*
* After flushing the fbdev (incl. a late async config which will
* have delayed queuing of a hotplug event), then flush the hotplug
* events.
*/
drm_kms_helper_poll_fini(&dev_priv->drm);
intel_gt_driver_unregister(&dev_priv->gt);
acpi_video_unregister();
intel_opregion_unregister(dev_priv);
i915_perf_unregister(dev_priv);
drm/i915/pmu: Expose a PMU interface for perf queries From: Chris Wilson <chris@chris-wilson.co.uk> From: Tvrtko Ursulin <tvrtko.ursulin@intel.com> From: Dmitry Rogozhkin <dmitry.v.rogozhkin@intel.com> The first goal is to be able to measure GPU (and invidual ring) busyness without having to poll registers from userspace. (Which not only incurs holding the forcewake lock indefinitely, perturbing the system, but also runs the risk of hanging the machine.) As an alternative we can use the perf event counter interface to sample the ring registers periodically and send those results to userspace. Functionality we are exporting to userspace is via the existing perf PMU API and can be exercised via the existing tools. For example: perf stat -a -e i915/rcs0-busy/ -I 1000 Will print the render engine busynnes once per second. All the performance counters can be enumerated (perf list) and have their unit of measure correctly reported in sysfs. v1-v2 (Chris Wilson): v2: Use a common timer for the ring sampling. v3: (Tvrtko Ursulin) * Decouple uAPI from i915 engine ids. * Complete uAPI defines. * Refactor some code to helpers for clarity. * Skip sampling disabled engines. * Expose counters in sysfs. * Pass in fake regs to avoid null ptr deref in perf core. * Convert to class/instance uAPI. * Use shared driver code for rc6 residency, power and frequency. v4: (Dmitry Rogozhkin) * Register PMU with .task_ctx_nr=perf_invalid_context * Expose cpumask for the PMU with the single CPU in the mask * Properly support pmu->stop(): it should call pmu->read() * Properly support pmu->del(): it should call stop(event, PERF_EF_UPDATE) * Introduce refcounting of event subscriptions. * Make pmu.busy_stats a refcounter to avoid busy stats going away with some deleted event. * Expose cpumask for i915 PMU to avoid multiple events creation of the same type followed by counter aggregation by perf-stat. * Track CPUs getting online/offline to migrate perf context. If (likely) cpumask will initially set CPU0, CONFIG_BOOTPARAM_HOTPLUG_CPU0 will be needed to see effect of CPU status tracking. * End result is that only global events are supported and perf stat works correctly. * Deny perf driver level sampling - it is prohibited for uncore PMU. v5: (Tvrtko Ursulin) * Don't hardcode number of engine samplers. * Rewrite event ref-counting for correctness and simplicity. * Store initial counter value when starting already enabled events to correctly report values to all listeners. * Fix RC6 residency readout. * Comments, GPL header. v6: * Add missing entry to v4 changelog. * Fix accounting in CPU hotplug case by copying the approach from arch/x86/events/intel/cstate.c. (Dmitry Rogozhkin) v7: * Log failure message only on failure. * Remove CPU hotplug notification state on unregister. v8: * Fix error unwind on failed registration. * Checkpatch cleanup. v9: * Drop the energy metric, it is available via intel_rapl_perf. (Ville Syrjälä) * Use HAS_RC6(p). (Chris Wilson) * Handle unsupported non-engine events. (Dmitry Rogozhkin) * Rebase for intel_rc6_residency_ns needing caller managed runtime pm. * Drop HAS_RC6 checks from the read callback since creating those events will be rejected at init time already. * Add counter units to sysfs so perf stat output is nicer. * Cleanup the attribute tables for brevity and readability. v10: * Fixed queued accounting. v11: * Move intel_engine_lookup_user to intel_engine_cs.c * Commit update. (Joonas Lahtinen) v12: * More accurate sampling. (Chris Wilson) * Store and report frequency in MHz for better usability from perf stat. * Removed metrics: queued, interrupts, rc6 counters. * Sample engine busyness based on seqno difference only for less MMIO (and forcewake) on all platforms. (Chris Wilson) v13: * Comment spelling, use mul_u32_u32 to work around potential GCC issue and somne code alignment changes. (Chris Wilson) v14: * Rebase. v15: * Rebase for RPS refactoring. v16: * Use the dynamic slot in the CPU hotplug state machine so that we are free to setup our state as multi-instance. Previously we were re-using the CPUHP_AP_PERF_X86_UNCORE_ONLINE slot which is neither used as multi-instance, nor owned by our driver to start with. * Register the CPU hotplug handlers after the PMU, otherwise the callback will get called before the PMU is initialized which can end up in perf_pmu_migrate_context with an un-initialized base. * Added workaround for a probable bug in cpuhp core. v17: * Remove workaround for the cpuhp bug. v18: * Rebase for drm_i915_gem_engine_class getting upstream before us. v19: * Rebase. (trivial) Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Signed-off-by: Dmitry Rogozhkin <dmitry.v.rogozhkin@intel.com> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Cc: Chris Wilson <chris@chris-wilson.co.uk> Cc: Dmitry Rogozhkin <dmitry.v.rogozhkin@intel.com> Cc: Peter Zijlstra <peterz@infradead.org> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20171121181852.16128-2-tvrtko.ursulin@linux.intel.com
2017-11-22 01:18:45 +07:00
i915_pmu_unregister(dev_priv);
i915_teardown_sysfs(dev_priv);
drm_dev_unplug(&dev_priv->drm);
i915_gem_driver_unregister(dev_priv);
}
static void i915_welcome_messages(struct drm_i915_private *dev_priv)
{
if (drm_debug & DRM_UT_DRIVER) {
struct drm_printer p = drm_debug_printer("i915 device info:");
drm/i915: Introduce concept of a sub-platform Concept of a sub-platform already exist in our code (like ULX and ULT platform variants and similar),implemented via the macros which check a list of device ids to determine a match. With this patch we consolidate device ids checking into a single function called during early driver load. A few low bits in the platform mask are reserved for sub-platform identification and defined as a per-platform namespace. At the same time it future proofs the platform_mask handling by preparing the code for easy extending, and tidies the very verbose WARN strings generated when IS_PLATFORM macros are embedded into a WARN type statements. v2: Fixed IS_SUBPLATFORM. Updated commit msg. v3: Chris was right, there is an ordering problem. v4: * Catch-up with new sub-platforms. * Rebase for RUNTIME_INFO. * Drop subplatform mask union tricks and convert platform_mask to an array for extensibility. v5: * Fix subplatform check. * Protect against forgetting to expand subplatform bits. * Remove platform enum tallying. * Add subplatform to error state. (Chris) * Drop macros and just use static inlines. * Remove redundant IRONLAKE_M. (Ville) v6: * Split out Ironlake change. * Optimize subplatform check. * Use __always_inline. (Lucas) * Add platform_mask comment. (Paulo) * Pass stored runtime info in error capture. (Chris) v7: * Rebased for new AML ULX device id. * Bump platform mask array size for EHL. * Stop mentioning device ids in intel_device_subplatform_init by using the trick of splitting macros i915_pciids.h. (Jani) * AML seems to be either a subplatform of KBL or CFL so express it like that. v8: * Use one device id table per subplatform. (Jani) Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Suggested-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Chris Wilson <chris@chris-wilson.co.uk> Cc: Jani Nikula <jani.nikula@intel.com> Cc: Lucas De Marchi <lucas.demarchi@intel.com> Cc: Jose Souza <jose.souza@intel.com> Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Cc: Paulo Zanoni <paulo.r.zanoni@intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Acked-by: Jani Nikula <jani.nikula@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190327142328.31780-1-tvrtko.ursulin@linux.intel.com
2019-03-27 21:23:28 +07:00
drm_printf(&p, "pciid=0x%04x rev=0x%02x platform=%s (subplatform=0x%x) gen=%i\n",
INTEL_DEVID(dev_priv),
INTEL_REVID(dev_priv),
intel_platform_name(INTEL_INFO(dev_priv)->platform),
drm/i915: Introduce concept of a sub-platform Concept of a sub-platform already exist in our code (like ULX and ULT platform variants and similar),implemented via the macros which check a list of device ids to determine a match. With this patch we consolidate device ids checking into a single function called during early driver load. A few low bits in the platform mask are reserved for sub-platform identification and defined as a per-platform namespace. At the same time it future proofs the platform_mask handling by preparing the code for easy extending, and tidies the very verbose WARN strings generated when IS_PLATFORM macros are embedded into a WARN type statements. v2: Fixed IS_SUBPLATFORM. Updated commit msg. v3: Chris was right, there is an ordering problem. v4: * Catch-up with new sub-platforms. * Rebase for RUNTIME_INFO. * Drop subplatform mask union tricks and convert platform_mask to an array for extensibility. v5: * Fix subplatform check. * Protect against forgetting to expand subplatform bits. * Remove platform enum tallying. * Add subplatform to error state. (Chris) * Drop macros and just use static inlines. * Remove redundant IRONLAKE_M. (Ville) v6: * Split out Ironlake change. * Optimize subplatform check. * Use __always_inline. (Lucas) * Add platform_mask comment. (Paulo) * Pass stored runtime info in error capture. (Chris) v7: * Rebased for new AML ULX device id. * Bump platform mask array size for EHL. * Stop mentioning device ids in intel_device_subplatform_init by using the trick of splitting macros i915_pciids.h. (Jani) * AML seems to be either a subplatform of KBL or CFL so express it like that. v8: * Use one device id table per subplatform. (Jani) Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Suggested-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Chris Wilson <chris@chris-wilson.co.uk> Cc: Jani Nikula <jani.nikula@intel.com> Cc: Lucas De Marchi <lucas.demarchi@intel.com> Cc: Jose Souza <jose.souza@intel.com> Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Cc: Paulo Zanoni <paulo.r.zanoni@intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Acked-by: Jani Nikula <jani.nikula@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190327142328.31780-1-tvrtko.ursulin@linux.intel.com
2019-03-27 21:23:28 +07:00
intel_subplatform(RUNTIME_INFO(dev_priv),
INTEL_INFO(dev_priv)->platform),
INTEL_GEN(dev_priv));
intel_device_info_dump_flags(INTEL_INFO(dev_priv), &p);
intel_device_info_dump_runtime(RUNTIME_INFO(dev_priv), &p);
}
if (IS_ENABLED(CONFIG_DRM_I915_DEBUG))
DRM_INFO("DRM_I915_DEBUG enabled\n");
if (IS_ENABLED(CONFIG_DRM_I915_DEBUG_GEM))
DRM_INFO("DRM_I915_DEBUG_GEM enabled\n");
if (IS_ENABLED(CONFIG_DRM_I915_DEBUG_RUNTIME_PM))
DRM_INFO("DRM_I915_DEBUG_RUNTIME_PM enabled\n");
}
static struct drm_i915_private *
i915_driver_create(struct pci_dev *pdev, const struct pci_device_id *ent)
{
const struct intel_device_info *match_info =
(struct intel_device_info *)ent->driver_data;
struct intel_device_info *device_info;
struct drm_i915_private *i915;
int err;
i915 = kzalloc(sizeof(*i915), GFP_KERNEL);
if (!i915)
return ERR_PTR(-ENOMEM);
err = drm_dev_init(&i915->drm, &driver, &pdev->dev);
if (err) {
kfree(i915);
return ERR_PTR(err);
}
i915->drm.dev_private = i915;
i915->drm.pdev = pdev;
pci_set_drvdata(pdev, i915);
/* Setup the write-once "constant" device info */
device_info = mkwrite_device_info(i915);
memcpy(device_info, match_info, sizeof(*device_info));
RUNTIME_INFO(i915)->device_id = pdev->device;
BUG_ON(device_info->gen > BITS_PER_TYPE(device_info->gen_mask));
return i915;
}
static void i915_driver_destroy(struct drm_i915_private *i915)
{
struct pci_dev *pdev = i915->drm.pdev;
drm_dev_fini(&i915->drm);
kfree(i915);
/* And make sure we never chase our dangling pointer from pci_dev */
pci_set_drvdata(pdev, NULL);
}
/**
* i915_driver_probe - setup chip and create an initial config
* @pdev: PCI device
* @ent: matching PCI ID entry
*
* The driver probe routine has to do several things:
* - drive output discovery via intel_modeset_init()
* - initialize the memory manager
* - allocate initial config memory
* - setup the DRM framebuffer with the allocated memory
*/
int i915_driver_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
{
const struct intel_device_info *match_info =
(struct intel_device_info *)ent->driver_data;
struct drm_i915_private *dev_priv;
int ret;
dev_priv = i915_driver_create(pdev, ent);
if (IS_ERR(dev_priv))
return PTR_ERR(dev_priv);
/* Disable nuclear pageflip by default on pre-ILK */
if (!i915_modparams.nuclear_pageflip && match_info->gen < 5)
dev_priv->drm.driver_features &= ~DRIVER_ATOMIC;
/*
* Check if we support fake LMEM -- for now we only unleash this for
* the live selftests(test-and-exit).
*/
#if IS_ENABLED(CONFIG_DRM_I915_SELFTEST)
if (IS_ENABLED(CONFIG_DRM_I915_UNSTABLE_FAKE_LMEM)) {
if (INTEL_GEN(dev_priv) >= 9 && i915_selftest.live < 0 &&
i915_modparams.fake_lmem_start) {
mkwrite_device_info(dev_priv)->memory_regions =
REGION_SMEM | REGION_LMEM | REGION_STOLEN;
mkwrite_device_info(dev_priv)->is_dgfx = true;
GEM_BUG_ON(!HAS_LMEM(dev_priv));
GEM_BUG_ON(!IS_DGFX(dev_priv));
}
}
#endif
ret = pci_enable_device(pdev);
if (ret)
goto out_fini;
ret = i915_driver_early_probe(dev_priv);
if (ret < 0)
goto out_pci_disable;
disable_rpm_wakeref_asserts(&dev_priv->runtime_pm);
i915_detect_vgpu(dev_priv);
ret = i915_driver_mmio_probe(dev_priv);
if (ret < 0)
goto out_runtime_pm_put;
ret = i915_driver_hw_probe(dev_priv);
if (ret < 0)
goto out_cleanup_mmio;
ret = i915_driver_modeset_probe(dev_priv);
if (ret < 0)
goto out_cleanup_hw;
i915_driver_register(dev_priv);
enable_rpm_wakeref_asserts(&dev_priv->runtime_pm);
i915_welcome_messages(dev_priv);
return 0;
out_cleanup_hw:
i915_driver_hw_remove(dev_priv);
intel_memory_regions_driver_release(dev_priv);
i915_ggtt_driver_release(dev_priv);
out_cleanup_mmio:
i915_driver_mmio_release(dev_priv);
out_runtime_pm_put:
enable_rpm_wakeref_asserts(&dev_priv->runtime_pm);
i915_driver_late_release(dev_priv);
out_pci_disable:
pci_disable_device(pdev);
out_fini:
i915_probe_error(dev_priv, "Device initialization failed (%d)\n", ret);
i915_driver_destroy(dev_priv);
return ret;
}
void i915_driver_remove(struct drm_i915_private *i915)
{
disable_rpm_wakeref_asserts(&i915->runtime_pm);
i915_driver_unregister(i915);
/*
* After unregistering the device to prevent any new users, cancel
* all in-flight requests so that we can quickly unbind the active
* resources.
*/
intel_gt_set_wedged(&i915->gt);
/* Flush any external code that still may be under the RCU lock */
synchronize_rcu();
i915_gem_suspend(i915);
drm_atomic_helper_shutdown(&i915->drm);
intel_gvt_driver_remove(i915);
i915_driver_modeset_remove(i915);
drm/i915: fix pch pci device enumeration pci_get_class(class, from) drops the refcount for 'from', so the extra pci_dev_put we do on it will result in a use after free bug starting with the WARN below. Regression introduced in commit 6a9c4b35e6696a63805b6da5e4889c6986e9ee1b Author: Rui Guo <firemeteor@users.sourceforge.net> Date: Wed Jun 19 21:10:23 2013 +0800 drm/i915: Fix PCH detect with multiple ISA bridges in VM [ 164.338460] WARNING: CPU: 1 PID: 2094 at include/linux/kref.h:47 klist_next+0xae/0x110() [ 164.347731] CPU: 1 PID: 2094 Comm: modprobe Tainted: G O 3.13.0-imre+ #354 [ 164.356468] Hardware name: Intel Corp. VALLEYVIEW B0 PLATFORM/NOTEBOOK, BIOS BYTICRB1.X64.0062.R70.1310112051 10/11/2013 [ 164.368796] Call Trace: [ 164.371609] [<ffffffff816a32a6>] dump_stack+0x4e/0x7a [ 164.377447] [<ffffffff8104f75d>] warn_slowpath_common+0x7d/0xa0 [ 164.384238] [<ffffffff8104f83a>] warn_slowpath_null+0x1a/0x20 [ 164.390851] [<ffffffff8169aeae>] klist_next+0xae/0x110 [ 164.396777] [<ffffffff8130a110>] ? pci_do_find_bus+0x70/0x70 [ 164.403286] [<ffffffff813cb4a9>] bus_find_device+0x89/0xc0 [ 164.409719] [<ffffffff8130a373>] pci_get_dev_by_id+0x63/0xa0 [ 164.416238] [<ffffffff8130a4e4>] pci_get_class+0x44/0x50 [ 164.422433] [<ffffffffa034821f>] intel_dsm_detect+0x16f/0x1f0 [i915] [ 164.429801] [<ffffffffa03482ae>] intel_register_dsm_handler+0xe/0x10 [i915] [ 164.437831] [<ffffffffa02d30fe>] i915_driver_load+0xafe/0xf30 [i915] [ 164.445126] [<ffffffff8158a150>] ? intel_alloc_coherent+0x110/0x110 [ 164.452340] [<ffffffffa0148c07>] drm_dev_register+0xc7/0x150 [drm] [ 164.459462] [<ffffffffa014b23f>] drm_get_pci_dev+0x11f/0x1f0 [drm] [ 164.466554] [<ffffffff816abb81>] ? _raw_spin_unlock_irqrestore+0x51/0x70 [ 164.474287] [<ffffffffa02cf7a6>] i915_pci_probe+0x56/0x60 [i915] [ 164.481185] [<ffffffff8130a028>] pci_device_probe+0x78/0xf0 [ 164.487603] [<ffffffff813cd495>] driver_probe_device+0x155/0x350 [ 164.494505] [<ffffffff813cd74e>] __driver_attach+0x6e/0xa0 [ 164.500826] [<ffffffff813cd6e0>] ? __device_attach+0x50/0x50 [ 164.507333] [<ffffffff813cb2be>] bus_for_each_dev+0x6e/0xc0 [ 164.513752] [<ffffffff813ccefe>] driver_attach+0x1e/0x20 [ 164.519870] [<ffffffff813cc958>] bus_add_driver+0x138/0x260 [ 164.526289] [<ffffffffa0188000>] ? 0xffffffffa0187fff [ 164.532116] [<ffffffff813cde78>] driver_register+0x98/0xe0 [ 164.538558] [<ffffffffa0188000>] ? 0xffffffffa0187fff [ 164.544389] [<ffffffff813087b0>] __pci_register_driver+0x60/0x70 [ 164.551336] [<ffffffffa014b37d>] drm_pci_init+0x6d/0x120 [drm] [ 164.558040] [<ffffffffa0188000>] ? 0xffffffffa0187fff [ 164.563928] [<ffffffffa018806a>] i915_init+0x6a/0x6c [i915] [ 164.570363] [<ffffffff810002da>] do_one_initcall+0xaa/0x160 [ 164.576783] [<ffffffff8103b140>] ? set_memory_nx+0x40/0x50 [ 164.583100] [<ffffffff810ce7f5>] load_module+0x1fb5/0x2550 [ 164.589410] [<ffffffff810caab0>] ? store_uevent+0x40/0x40 [ 164.595628] [<ffffffff810cee7d>] SyS_init_module+0xed/0x100 [ 164.602048] [<ffffffff816b3c52>] system_call_fastpath+0x16/0x1b v2: simplify the loop further (Chris) Signed-off-by: Imre Deak <imre.deak@intel.com> Cc: Jesse Barnes <jbarnes@virtuousgeek.org> Reported-by: Jesse Barnes <jbarnes@virtuousgeek.org> Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=65652 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=74161 Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: stable@vger.kernel.org Signed-off-by: Jani Nikula <jani.nikula@intel.com>
2014-02-15 01:23:54 +07:00
i915_reset_error_state(i915);
i915_gem_driver_remove(i915);
intel_power_domains_driver_remove(i915);
i915_driver_hw_remove(i915);
enable_rpm_wakeref_asserts(&i915->runtime_pm);
}
static void i915_driver_release(struct drm_device *dev)
{
struct drm_i915_private *dev_priv = to_i915(dev);
struct intel_runtime_pm *rpm = &dev_priv->runtime_pm;
disable_rpm_wakeref_asserts(rpm);
i915_gem_driver_release(dev_priv);
intel_memory_regions_driver_release(dev_priv);
i915_ggtt_driver_release(dev_priv);
i915_driver_mmio_release(dev_priv);
enable_rpm_wakeref_asserts(rpm);
intel_runtime_pm_driver_release(rpm);
i915_driver_late_release(dev_priv);
i915_driver_destroy(dev_priv);
}
static int i915_driver_open(struct drm_device *dev, struct drm_file *file)
{
struct drm_i915_private *i915 = to_i915(dev);
int ret;
ret = i915_gem_open(i915, file);
if (ret)
return ret;
return 0;
}
/**
* i915_driver_lastclose - clean up after all DRM clients have exited
* @dev: DRM device
*
* Take care of cleaning up after all DRM clients have exited. In the
* mode setting case, we want to restore the kernel's initial mode (just
* in case the last client left us in a bad state).
*
* Additionally, in the non-mode setting case, we'll tear down the GTT
* and DMA structures, since the kernel won't be using them, and clea
* up any GEM state.
*/
static void i915_driver_lastclose(struct drm_device *dev)
{
intel_fbdev_restore_mode(dev);
vga_switcheroo_process_delayed_switch();
}
static void i915_driver_postclose(struct drm_device *dev, struct drm_file *file)
{
struct drm_i915_file_private *file_priv = file->driver_priv;
i915_gem_context_close(file);
i915_gem_release(dev, file);
drm/i915: Keep drm_i915_file_private around under RCU Ensure that the drm_i915_file_private continues to exist as we attempt to remove a request from its list, which may race with the destruction of the file. <6> [38.380714] [IGT] gem_ctx_create: starting subtest basic-files <0> [42.201329] BUG: spinlock bad magic on CPU#0, kworker/u16:0/7 <4> [42.201356] general protection fault: 0000 [#1] PREEMPT SMP PTI <4> [42.201371] CPU: 0 PID: 7 Comm: kworker/u16:0 Tainted: G U 5.3.0-rc5-CI-Patchwork_14169+ #1 <4> [42.201391] Hardware name: Dell Inc. OptiPlex 745 /0GW726, BIOS 2.3.1 05/21/2007 <4> [42.201594] Workqueue: i915 retire_work_handler [i915] <4> [42.201614] RIP: 0010:spin_dump+0x5a/0x90 <4> [42.201625] Code: 00 48 8d 88 c0 06 00 00 48 c7 c7 00 71 09 82 e8 35 ef 00 00 48 85 db 44 8b 4d 08 41 b8 ff ff ff ff 48 c7 c1 0b cd 0f 82 74 0e <44> 8b 83 e0 04 00 00 48 8d 8b c0 06 00 00 8b 55 04 48 89 ee 48 c7 <4> [42.201660] RSP: 0018:ffffc9000004bd80 EFLAGS: 00010202 <4> [42.201673] RAX: 0000000000000031 RBX: 6b6b6b6b6b6b6b6b RCX: ffffffff820fcd0b <4> [42.201688] RDX: 0000000000000000 RSI: ffff88803de266f8 RDI: 00000000ffffffff <4> [42.201703] RBP: ffff888038381ff8 R08: 00000000ffffffff R09: 000000006b6b6b6b <4> [42.201718] R10: 0000000041cb0b89 R11: 646162206b636f6c R12: ffff88802a618500 <4> [42.201733] R13: ffff88802b32c288 R14: ffff888038381ff8 R15: ffff88802b32c250 <4> [42.201748] FS: 0000000000000000(0000) GS:ffff88803de00000(0000) knlGS:0000000000000000 <4> [42.201765] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 <4> [42.201778] CR2: 00007f2cefc6d180 CR3: 00000000381ee000 CR4: 00000000000006f0 <4> [42.201793] Call Trace: <4> [42.201805] do_raw_spin_lock+0x66/0xb0 <4> [42.201898] i915_request_retire+0x548/0x7c0 [i915] <4> [42.201989] retire_requests+0x4d/0x60 [i915] <4> [42.202078] i915_retire_requests+0x144/0x2e0 [i915] <4> [42.202169] retire_work_handler+0x10/0x40 [i915] Recently, in commit 44c22f3f1a0a ("drm/i915: Serialize insertion into the file->mm.request_list"), we fixed a race on insertion. Now, it appears we also have a race with destruction! Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Matthew Auld <matthew.auld@intel.com> Reviewed-by: Matthew Auld <matthew.auld@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190823181455.31910-1-chris@chris-wilson.co.uk
2019-08-24 01:14:55 +07:00
kfree_rcu(file_priv, rcu);
/* Catch up with all the deferred frees from "this" client */
i915_gem_flush_free_objects(to_i915(dev));
}
static void intel_suspend_encoders(struct drm_i915_private *dev_priv)
{
struct drm_device *dev = &dev_priv->drm;
struct intel_encoder *encoder;
drm_modeset_lock_all(dev);
for_each_intel_encoder(dev, encoder)
if (encoder->suspend)
encoder->suspend(encoder);
drm_modeset_unlock_all(dev);
}
static int vlv_resume_prepare(struct drm_i915_private *dev_priv,
bool rpm_resume);
static int vlv_suspend_complete(struct drm_i915_private *dev_priv);
drm/i915/skl: Add DC6 Trigger sequence. Add triggers for DC6 as per details provided in skl_enable_dc6 and skl_disable_dc6 implementations. Also Call POSTING_READ for every write to a register to ensure it is written to immediately v1: Remove POSTING_READ and intel_prepare_ddi calls as they've been added in previous patches. v2: 1] Remove check for backlight disabled as it should be the case by that time. 2] Mark DC5 as disabled when enabling DC6. 3] Return from DC5-disabling function early if DC5 is already be disabled which can happen due to DC6-enabling earlier. 3] Ensure CSR firmware is loaded after resume from DC6 as corresponding memory contents won't be retained after runtime-suspend. 4] Ensure that CSR isn't identified as loaded before CSR-loading program is called during runtime-resume. v3: Rebase to latest Modified as per review comments from Imre and after discussion with Art: 1] DC6 should be preferably enabled when PG2 is disabled by SW as the check for PG1 being disabled is taken of by HW to enter DC6, and disabled when PG2 is enabled respectively. This helps save more power, especially in the case when display is disabled but GT is enabled. Accordingly, replacing DC5 trigger sequence with DC6 for SKL. 2] DC6 could be enabled from intel_runtime_suspend() function, if DC5 is already enabled. 3] Move CSR-load-status setting code from intel_runtime_suspend function to a new function. v4: 1] Enable/disable DC6 only when toggling the power-well using a newly defined macro ENABLE_DC6. v5: 1] Load CSR on system resume too as firmware may be lost on system suspend preventing enabling DC5, DC6. 2] DDI buffers shouldn't be programmed during driver-load/resume as it's already done during modeset initialization then and also that the encoder list is still uninitialized by then. Therefore, call intel_prepare_ddi function right after disabling DC6 but outside skl_disable_dc6 function and not during driver-load/resume. v6: 1] Rebase to latest. 2] Move SKL_ENABLE_DC6 macro definition from intel_display.c to intel_runtime_pm.c. v7: 1) Refactored the code for removing the warning got from checkpatch. 2) After adding dmc ver 1.0 support rebased on top of nightly. (Animesh) v8: - Reverted the changes done in v7. - Removed the condition check in skl_prepare_resune(). (Animesh) Issue: VIZ-2819 Signed-off-by: A.Sunil Kamath <sunil.kamath@intel.com> Signed-off-by: Suketu Shah <suketu.j.shah@intel.com> Signed-off-by: Damien Lespiau <damien.lespiau@intel.com> Signed-off-by: Animesh Manna <animesh.manna@intel.com> Reviewed-by: Imre Deak <imre.deak@intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2015-04-16 15:52:11 +07:00
static bool suspend_to_idle(struct drm_i915_private *dev_priv)
{
#if IS_ENABLED(CONFIG_ACPI_SLEEP)
if (acpi_target_system_state() < ACPI_STATE_S3)
return true;
#endif
return false;
}
static int i915_drm_prepare(struct drm_device *dev)
{
struct drm_i915_private *i915 = to_i915(dev);
/*
* NB intel_display_suspend() may issue new requests after we've
* ostensibly marked the GPU as ready-to-sleep here. We need to
* split out that work and pull it forward so that after point,
* the GPU is not woken again.
*/
i915_gem_suspend(i915);
return 0;
}
static int i915_drm_suspend(struct drm_device *dev)
{
struct drm_i915_private *dev_priv = to_i915(dev);
struct pci_dev *pdev = dev_priv->drm.pdev;
pci_power_t opregion_target_state;
disable_rpm_wakeref_asserts(&dev_priv->runtime_pm);
drm/i915: add support for checking if we hold an RPM reference Atm, we assert that the device is not suspended until the point when the device is truly put to a suspended state. This is fine, but we can catch more problems if we check that RPM refcount is non-zero. After that one drops to zero we shouldn't access the device any more, even if the actual device suspend may be delayed. Change assert_rpm_wakelock_held() accordingly to check for a non-zero RPM refcount in addition to the current device-not-suspended check. For the new asserts to work we need to annotate every place explicitly in the code where we expect that the device is powered. The places where we only assume this, but may not hold an RPM reference: - driver load We assume the device to be powered until we enable RPM. Make this explicit by taking an RPM reference around the load function. - system and runtime sudpend/resume handlers These handlers are called when the RPM reference becomes 0 and know the exact point after which the device can get powered off. Disable the RPM-reference-held check for their duration. - the IRQ, hangcheck and RPS work handlers These handlers are flushed in the system/runtime suspend handler before the device is powered off, so it's guaranteed that they won't run while the device is powered off even though they don't hold any RPM reference. Disable the RPM-reference-held check for their duration. In all these cases we still check that the device is not suspended. These explicit annotations also have the positive side effect of documenting our assumptions better. This caught additional WARNs from the atomic modeset path, those should be fixed separately. v2: - remove the redundant HAS_RUNTIME_PM check (moved to patch 1) (Ville) v3: - use a new dedicated RPM wakelock refcount to also catch cases where our own RPM get/put functions were not called (Chris) - assert also that the new RPM wakelock refcount is 0 in the RPM suspend handler (Chris) - change the assert error message to be more meaningful (Chris) - prevent false assert errors and check that the RPM wakelock is 0 in the RPM resume handler too - prevent false assert errors in the hangcheck work too - add a device not suspended assert check to the hangcheck work v4: - rename disable/enable_rpm_asserts to disable/enable_rpm_wakeref_asserts and wakelock_count to wakeref_count - disable the wakeref asserts in the IRQ handlers and RPS work too - update/clarify commit message v5: - mark places we plan to change to use proper RPM refcounting with separate DISABLE/ENABLE_RPM_WAKEREF_ASSERTS aliases (Chris) Signed-off-by: Imre Deak <imre.deak@intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Link: http://patchwork.freedesktop.org/patch/msgid/1450227139-13471-1-git-send-email-imre.deak@intel.com
2015-12-16 07:52:19 +07:00
drm/i915: allow package C8+ states on Haswell (disabled) This patch allows PC8+ states on Haswell. These states can only be reached when all the display outputs are disabled, and they allow some more power savings. The fact that the graphics device is allowing PC8+ doesn't mean that the machine will actually enter PC8+: all the other devices also need to allow PC8+. For now this option is disabled by default. You need i915.allow_pc8=1 if you want it. This patch adds a big comment inside i915_drv.h explaining how it works and how it tracks things. Read it. v2: (this is not really v2, many previous versions were already sent, but they had different names) - Use the new functions to enable/disable GTIMR and GEN6_PMIMR - Rename almost all variables and functions to names suggested by Chris - More WARNs on the IRQ handling code - Also disable PC8 when there's GPU work to do (thanks to Ben for the help on this), so apps can run caster - Enable PC8 on a delayed work function that is delayed for 5 seconds. This makes sure we only enable PC8+ if we're really idle - Make sure we're not in PC8+ when suspending v3: - WARN if IRQs are disabled on __wait_seqno - Replace some DRM_ERRORs with WARNs - Fix calls to restore GT and PM interrupts - Use intel_mark_busy instead of intel_ring_advance to disable PC8 v4: - Use the force_wake, Luke! v5: - Remove the "IIR is not zero" WARNs - Move the force_wake chunk to its own patch - Only restore what's missing from RC6, not everything Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2013-08-19 23:18:09 +07:00
/* We do a lot of poking in a lot of registers, make sure they work
* properly. */
intel_power_domains_disable(dev_priv);
drm_kms_helper_poll_disable(dev);
pci_save_state(pdev);
intel_display_suspend(dev);
intel_dp_mst_suspend(dev_priv);
intel_runtime_pm_disable_interrupts(dev_priv);
intel_hpd_cancel_work(dev_priv);
intel_suspend_encoders(dev_priv);
2014-05-02 11:02:48 +07:00
intel_suspend_hw(dev_priv);
i915_gem_suspend_gtt_mappings(dev_priv);
drm/i915: Disable GGTT PTEs on GEN6+ suspend Once the machine gets to a certain point in the suspend process, we expect the GPU to be idle. If it is not, we might corrupt memory. Empirically (with an early version of this patch) we have seen this is not the case. We cannot currently explain why the latent GPU writes occur. In the technical sense, this patch is a workaround in that we have an issue we can't explain, and the patch indirectly solves the issue. However, it's really better than a workaround because we understand why it works, and it really should be a safe thing to do in all cases. The noticeable effect other than the debug messages would be an increase in the suspend time. I have not measure how expensive it actually is. I think it would be good to spend further time to root cause why we're seeing these latent writes, but it shouldn't preclude preventing the fallout. NOTE: It should be safe (and makes some sense IMO) to also keep the VALID bit unset on resume when we clear_range(). I've opted not to do this as properly clearing those bits at some later point would be extra work. v2: Fix bugzilla link Bugzilla: http://bugs.freedesktop.org/show_bug.cgi?id=65496 Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=59321 Tested-by: Takashi Iwai <tiwai@suse.de> Tested-by: Paulo Zanoni <paulo.r.zanoni@intel.com> Signed-off-by: Ben Widawsky <ben@bwidawsk.net> Tested-By: Todd Previte <tprevite@gmail.com> Cc: stable@vger.kernel.org Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2013-10-16 23:21:30 +07:00
i915_save_state(dev_priv);
opregion_target_state = suspend_to_idle(dev_priv) ? PCI_D1 : PCI_D3cold;
intel_opregion_suspend(dev_priv, opregion_target_state);
intel_fbdev_set_suspend(dev, FBINFO_STATE_SUSPENDED, true);
dev_priv->suspend_count++;
intel_csr_ucode_suspend(dev_priv);
enable_rpm_wakeref_asserts(&dev_priv->runtime_pm);
drm/i915: add support for checking if we hold an RPM reference Atm, we assert that the device is not suspended until the point when the device is truly put to a suspended state. This is fine, but we can catch more problems if we check that RPM refcount is non-zero. After that one drops to zero we shouldn't access the device any more, even if the actual device suspend may be delayed. Change assert_rpm_wakelock_held() accordingly to check for a non-zero RPM refcount in addition to the current device-not-suspended check. For the new asserts to work we need to annotate every place explicitly in the code where we expect that the device is powered. The places where we only assume this, but may not hold an RPM reference: - driver load We assume the device to be powered until we enable RPM. Make this explicit by taking an RPM reference around the load function. - system and runtime sudpend/resume handlers These handlers are called when the RPM reference becomes 0 and know the exact point after which the device can get powered off. Disable the RPM-reference-held check for their duration. - the IRQ, hangcheck and RPS work handlers These handlers are flushed in the system/runtime suspend handler before the device is powered off, so it's guaranteed that they won't run while the device is powered off even though they don't hold any RPM reference. Disable the RPM-reference-held check for their duration. In all these cases we still check that the device is not suspended. These explicit annotations also have the positive side effect of documenting our assumptions better. This caught additional WARNs from the atomic modeset path, those should be fixed separately. v2: - remove the redundant HAS_RUNTIME_PM check (moved to patch 1) (Ville) v3: - use a new dedicated RPM wakelock refcount to also catch cases where our own RPM get/put functions were not called (Chris) - assert also that the new RPM wakelock refcount is 0 in the RPM suspend handler (Chris) - change the assert error message to be more meaningful (Chris) - prevent false assert errors and check that the RPM wakelock is 0 in the RPM resume handler too - prevent false assert errors in the hangcheck work too - add a device not suspended assert check to the hangcheck work v4: - rename disable/enable_rpm_asserts to disable/enable_rpm_wakeref_asserts and wakelock_count to wakeref_count - disable the wakeref asserts in the IRQ handlers and RPS work too - update/clarify commit message v5: - mark places we plan to change to use proper RPM refcounting with separate DISABLE/ENABLE_RPM_WAKEREF_ASSERTS aliases (Chris) Signed-off-by: Imre Deak <imre.deak@intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Link: http://patchwork.freedesktop.org/patch/msgid/1450227139-13471-1-git-send-email-imre.deak@intel.com
2015-12-16 07:52:19 +07:00
return 0;
}
static enum i915_drm_suspend_mode
get_suspend_mode(struct drm_i915_private *dev_priv, bool hibernate)
{
if (hibernate)
return I915_DRM_SUSPEND_HIBERNATE;
if (suspend_to_idle(dev_priv))
return I915_DRM_SUSPEND_IDLE;
return I915_DRM_SUSPEND_MEM;
}
static int i915_drm_suspend_late(struct drm_device *dev, bool hibernation)
{
struct drm_i915_private *dev_priv = to_i915(dev);
struct pci_dev *pdev = dev_priv->drm.pdev;
struct intel_runtime_pm *rpm = &dev_priv->runtime_pm;
int ret = 0;
disable_rpm_wakeref_asserts(rpm);
drm/i915: add support for checking if we hold an RPM reference Atm, we assert that the device is not suspended until the point when the device is truly put to a suspended state. This is fine, but we can catch more problems if we check that RPM refcount is non-zero. After that one drops to zero we shouldn't access the device any more, even if the actual device suspend may be delayed. Change assert_rpm_wakelock_held() accordingly to check for a non-zero RPM refcount in addition to the current device-not-suspended check. For the new asserts to work we need to annotate every place explicitly in the code where we expect that the device is powered. The places where we only assume this, but may not hold an RPM reference: - driver load We assume the device to be powered until we enable RPM. Make this explicit by taking an RPM reference around the load function. - system and runtime sudpend/resume handlers These handlers are called when the RPM reference becomes 0 and know the exact point after which the device can get powered off. Disable the RPM-reference-held check for their duration. - the IRQ, hangcheck and RPS work handlers These handlers are flushed in the system/runtime suspend handler before the device is powered off, so it's guaranteed that they won't run while the device is powered off even though they don't hold any RPM reference. Disable the RPM-reference-held check for their duration. In all these cases we still check that the device is not suspended. These explicit annotations also have the positive side effect of documenting our assumptions better. This caught additional WARNs from the atomic modeset path, those should be fixed separately. v2: - remove the redundant HAS_RUNTIME_PM check (moved to patch 1) (Ville) v3: - use a new dedicated RPM wakelock refcount to also catch cases where our own RPM get/put functions were not called (Chris) - assert also that the new RPM wakelock refcount is 0 in the RPM suspend handler (Chris) - change the assert error message to be more meaningful (Chris) - prevent false assert errors and check that the RPM wakelock is 0 in the RPM resume handler too - prevent false assert errors in the hangcheck work too - add a device not suspended assert check to the hangcheck work v4: - rename disable/enable_rpm_asserts to disable/enable_rpm_wakeref_asserts and wakelock_count to wakeref_count - disable the wakeref asserts in the IRQ handlers and RPS work too - update/clarify commit message v5: - mark places we plan to change to use proper RPM refcounting with separate DISABLE/ENABLE_RPM_WAKEREF_ASSERTS aliases (Chris) Signed-off-by: Imre Deak <imre.deak@intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Link: http://patchwork.freedesktop.org/patch/msgid/1450227139-13471-1-git-send-email-imre.deak@intel.com
2015-12-16 07:52:19 +07:00
i915_gem_suspend_late(dev_priv);
intel_uncore_suspend(&dev_priv->uncore);
intel_power_domains_suspend(dev_priv,
get_suspend_mode(dev_priv, hibernation));
drm/i915/skl: init/uninit display core as part of the HW power domain state We need to initialize the display core part early, before initializing the rest of the display power state. This is also described in the bspec termed "Display initialization sequence". Atm we run this sequence during driver loading after power domain HW state initialization which is too late and during runtime suspend/resume which is unneeded and can interere with DMC functionality which handles HW resources toggled by this init/uninit sequence automatically. The init sequence must be run as the first step of HW power state initialization and during system resume. The uninit sequence must be run during system suspend. To address the above move the init sequence to the initial HW power state setup and the uninit sequence to a new power domains suspend function called during system suspend. As part of the init sequence we also have to reprogram the DMC firmware as it's lost across a system suspend/resume cycle. After this change CD clock initialization during driver loading will happen only later after other dependent HW/SW parts are initialized, while during system resume it will get initialized as the last step of the init sequence. This distinction can be removed by some refactoring of platform independent parts. I left this refactoring out from this series since I didn't want to change non-SKL parts. This is a TODO for later. v2: - fix error path in i915_drm_suspend_late() - don't try to re-program the DMC firmware if it failed to load Signed-off-by: Imre Deak <imre.deak@intel.com> Reviewed-by: Patrik Jakobsson <patrik.jakobsson@linux.intel.com> Link: http://patchwork.freedesktop.org/patch/msgid/1447774433-20834-1-git-send-email-imre.deak@intel.com
2015-11-17 22:33:53 +07:00
intel_display_power_suspend_late(dev_priv);
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);
intel_power_domains_resume(dev_priv);
drm/i915: add support for checking if we hold an RPM reference Atm, we assert that the device is not suspended until the point when the device is truly put to a suspended state. This is fine, but we can catch more problems if we check that RPM refcount is non-zero. After that one drops to zero we shouldn't access the device any more, even if the actual device suspend may be delayed. Change assert_rpm_wakelock_held() accordingly to check for a non-zero RPM refcount in addition to the current device-not-suspended check. For the new asserts to work we need to annotate every place explicitly in the code where we expect that the device is powered. The places where we only assume this, but may not hold an RPM reference: - driver load We assume the device to be powered until we enable RPM. Make this explicit by taking an RPM reference around the load function. - system and runtime sudpend/resume handlers These handlers are called when the RPM reference becomes 0 and know the exact point after which the device can get powered off. Disable the RPM-reference-held check for their duration. - the IRQ, hangcheck and RPS work handlers These handlers are flushed in the system/runtime suspend handler before the device is powered off, so it's guaranteed that they won't run while the device is powered off even though they don't hold any RPM reference. Disable the RPM-reference-held check for their duration. In all these cases we still check that the device is not suspended. These explicit annotations also have the positive side effect of documenting our assumptions better. This caught additional WARNs from the atomic modeset path, those should be fixed separately. v2: - remove the redundant HAS_RUNTIME_PM check (moved to patch 1) (Ville) v3: - use a new dedicated RPM wakelock refcount to also catch cases where our own RPM get/put functions were not called (Chris) - assert also that the new RPM wakelock refcount is 0 in the RPM suspend handler (Chris) - change the assert error message to be more meaningful (Chris) - prevent false assert errors and check that the RPM wakelock is 0 in the RPM resume handler too - prevent false assert errors in the hangcheck work too - add a device not suspended assert check to the hangcheck work v4: - rename disable/enable_rpm_asserts to disable/enable_rpm_wakeref_asserts and wakelock_count to wakeref_count - disable the wakeref asserts in the IRQ handlers and RPS work too - update/clarify commit message v5: - mark places we plan to change to use proper RPM refcounting with separate DISABLE/ENABLE_RPM_WAKEREF_ASSERTS aliases (Chris) Signed-off-by: Imre Deak <imre.deak@intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Link: http://patchwork.freedesktop.org/patch/msgid/1450227139-13471-1-git-send-email-imre.deak@intel.com
2015-12-16 07:52:19 +07:00
goto out;
}
pci_disable_device(pdev);
2015-03-02 18:04:41 +07:00
/*
drm/i915: apply the PCI_D0/D3 hibernation workaround everywhere on pre GEN6 commit da2bc1b9db3351addd293e5b82757efe1f77ed1d Author: Imre Deak <imre.deak@intel.com> Date: Thu Oct 23 19:23:26 2014 +0300 drm/i915: add poweroff_late handler introduced a regression on old platforms during hibernation. A workaround was added in commit ab3be73fa7b43f4c3648ce29b5fd649ea54d3adb Author: Imre Deak <imre.deak@intel.com> Date: Mon Mar 2 13:04:41 2015 +0200 drm/i915: gen4: work around hang during hibernation using an explicit blacklist for the GENs/BIOS vendors where the issue was reported. Later there we had reports of the same failure on platforms not on this list. To my best knowledge the correct thing to do is still to put the device to PCI D3 state during hibernation, see [1] and [2] for the reasons. This also aligns with our future plans to unify more the runtime and system suspend/resume paths. Since an exact blacklist seems to be impractical (multiple GENs and BIOS vendors are affected) apply the workaround on everything pre GEN6. [1] http://lists.freedesktop.org/archives/intel-gfx/2015-February/060710.html [2] https://lkml.org/lkml/2015/6/22/274 Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=95061 Reported-by: Ilya Tumaykin <itumaykin@gmail.com> Reported-by: Dirk Griesbach <spamthis@freenet.de> Reported-by: Pavel Machek <pavel@ucw.cz> Reported-by: Mikko Rapeli <mikko.rapeli@iki.fi> Tested-by: Mikko Rapeli <mikko.rapeli@iki.fi> Reported-by: Paul Bolle <pebolle@tiscali.nl> CC: stable@vger.kernel.org Signed-off-by: Imre Deak <imre.deak@intel.com> Signed-off-by: Jani Nikula <jani.nikula@intel.com>
2015-06-30 21:06:47 +07:00
* During hibernation on some platforms the BIOS may try to access
2015-03-02 18:04:41 +07:00
* the device even though it's already in D3 and hang the machine. So
* leave the device in D0 on those platforms and hope the BIOS will
drm/i915: apply the PCI_D0/D3 hibernation workaround everywhere on pre GEN6 commit da2bc1b9db3351addd293e5b82757efe1f77ed1d Author: Imre Deak <imre.deak@intel.com> Date: Thu Oct 23 19:23:26 2014 +0300 drm/i915: add poweroff_late handler introduced a regression on old platforms during hibernation. A workaround was added in commit ab3be73fa7b43f4c3648ce29b5fd649ea54d3adb Author: Imre Deak <imre.deak@intel.com> Date: Mon Mar 2 13:04:41 2015 +0200 drm/i915: gen4: work around hang during hibernation using an explicit blacklist for the GENs/BIOS vendors where the issue was reported. Later there we had reports of the same failure on platforms not on this list. To my best knowledge the correct thing to do is still to put the device to PCI D3 state during hibernation, see [1] and [2] for the reasons. This also aligns with our future plans to unify more the runtime and system suspend/resume paths. Since an exact blacklist seems to be impractical (multiple GENs and BIOS vendors are affected) apply the workaround on everything pre GEN6. [1] http://lists.freedesktop.org/archives/intel-gfx/2015-February/060710.html [2] https://lkml.org/lkml/2015/6/22/274 Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=95061 Reported-by: Ilya Tumaykin <itumaykin@gmail.com> Reported-by: Dirk Griesbach <spamthis@freenet.de> Reported-by: Pavel Machek <pavel@ucw.cz> Reported-by: Mikko Rapeli <mikko.rapeli@iki.fi> Tested-by: Mikko Rapeli <mikko.rapeli@iki.fi> Reported-by: Paul Bolle <pebolle@tiscali.nl> CC: stable@vger.kernel.org Signed-off-by: Imre Deak <imre.deak@intel.com> Signed-off-by: Jani Nikula <jani.nikula@intel.com>
2015-06-30 21:06:47 +07:00
* power down the device properly. The issue was seen on multiple old
* GENs with different BIOS vendors, so having an explicit blacklist
* is inpractical; apply the workaround on everything pre GEN6. The
* platforms where the issue was seen:
* Lenovo Thinkpad X301, X61s, X60, T60, X41
* Fujitsu FSC S7110
* Acer Aspire 1830T
2015-03-02 18:04:41 +07:00
*/
if (!(hibernation && INTEL_GEN(dev_priv) < 6))
pci_set_power_state(pdev, PCI_D3hot);
drm/i915: add support for checking if we hold an RPM reference Atm, we assert that the device is not suspended until the point when the device is truly put to a suspended state. This is fine, but we can catch more problems if we check that RPM refcount is non-zero. After that one drops to zero we shouldn't access the device any more, even if the actual device suspend may be delayed. Change assert_rpm_wakelock_held() accordingly to check for a non-zero RPM refcount in addition to the current device-not-suspended check. For the new asserts to work we need to annotate every place explicitly in the code where we expect that the device is powered. The places where we only assume this, but may not hold an RPM reference: - driver load We assume the device to be powered until we enable RPM. Make this explicit by taking an RPM reference around the load function. - system and runtime sudpend/resume handlers These handlers are called when the RPM reference becomes 0 and know the exact point after which the device can get powered off. Disable the RPM-reference-held check for their duration. - the IRQ, hangcheck and RPS work handlers These handlers are flushed in the system/runtime suspend handler before the device is powered off, so it's guaranteed that they won't run while the device is powered off even though they don't hold any RPM reference. Disable the RPM-reference-held check for their duration. In all these cases we still check that the device is not suspended. These explicit annotations also have the positive side effect of documenting our assumptions better. This caught additional WARNs from the atomic modeset path, those should be fixed separately. v2: - remove the redundant HAS_RUNTIME_PM check (moved to patch 1) (Ville) v3: - use a new dedicated RPM wakelock refcount to also catch cases where our own RPM get/put functions were not called (Chris) - assert also that the new RPM wakelock refcount is 0 in the RPM suspend handler (Chris) - change the assert error message to be more meaningful (Chris) - prevent false assert errors and check that the RPM wakelock is 0 in the RPM resume handler too - prevent false assert errors in the hangcheck work too - add a device not suspended assert check to the hangcheck work v4: - rename disable/enable_rpm_asserts to disable/enable_rpm_wakeref_asserts and wakelock_count to wakeref_count - disable the wakeref asserts in the IRQ handlers and RPS work too - update/clarify commit message v5: - mark places we plan to change to use proper RPM refcounting with separate DISABLE/ENABLE_RPM_WAKEREF_ASSERTS aliases (Chris) Signed-off-by: Imre Deak <imre.deak@intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Link: http://patchwork.freedesktop.org/patch/msgid/1450227139-13471-1-git-send-email-imre.deak@intel.com
2015-12-16 07:52:19 +07:00
out:
enable_rpm_wakeref_asserts(rpm);
if (!dev_priv->uncore.user_forcewake_count)
intel_runtime_pm_driver_release(rpm);
drm/i915: add support for checking if we hold an RPM reference Atm, we assert that the device is not suspended until the point when the device is truly put to a suspended state. This is fine, but we can catch more problems if we check that RPM refcount is non-zero. After that one drops to zero we shouldn't access the device any more, even if the actual device suspend may be delayed. Change assert_rpm_wakelock_held() accordingly to check for a non-zero RPM refcount in addition to the current device-not-suspended check. For the new asserts to work we need to annotate every place explicitly in the code where we expect that the device is powered. The places where we only assume this, but may not hold an RPM reference: - driver load We assume the device to be powered until we enable RPM. Make this explicit by taking an RPM reference around the load function. - system and runtime sudpend/resume handlers These handlers are called when the RPM reference becomes 0 and know the exact point after which the device can get powered off. Disable the RPM-reference-held check for their duration. - the IRQ, hangcheck and RPS work handlers These handlers are flushed in the system/runtime suspend handler before the device is powered off, so it's guaranteed that they won't run while the device is powered off even though they don't hold any RPM reference. Disable the RPM-reference-held check for their duration. In all these cases we still check that the device is not suspended. These explicit annotations also have the positive side effect of documenting our assumptions better. This caught additional WARNs from the atomic modeset path, those should be fixed separately. v2: - remove the redundant HAS_RUNTIME_PM check (moved to patch 1) (Ville) v3: - use a new dedicated RPM wakelock refcount to also catch cases where our own RPM get/put functions were not called (Chris) - assert also that the new RPM wakelock refcount is 0 in the RPM suspend handler (Chris) - change the assert error message to be more meaningful (Chris) - prevent false assert errors and check that the RPM wakelock is 0 in the RPM resume handler too - prevent false assert errors in the hangcheck work too - add a device not suspended assert check to the hangcheck work v4: - rename disable/enable_rpm_asserts to disable/enable_rpm_wakeref_asserts and wakelock_count to wakeref_count - disable the wakeref asserts in the IRQ handlers and RPS work too - update/clarify commit message v5: - mark places we plan to change to use proper RPM refcounting with separate DISABLE/ENABLE_RPM_WAKEREF_ASSERTS aliases (Chris) Signed-off-by: Imre Deak <imre.deak@intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Link: http://patchwork.freedesktop.org/patch/msgid/1450227139-13471-1-git-send-email-imre.deak@intel.com
2015-12-16 07:52:19 +07:00
return ret;
}
int i915_suspend_switcheroo(struct drm_i915_private *i915, pm_message_t state)
{
int error;
if (WARN_ON_ONCE(state.event != PM_EVENT_SUSPEND &&
state.event != PM_EVENT_FREEZE))
return -EINVAL;
if (i915->drm.switch_power_state == DRM_SWITCH_POWER_OFF)
return 0;
error = i915_drm_suspend(&i915->drm);
if (error)
return error;
return i915_drm_suspend_late(&i915->drm, false);
}
static int i915_drm_resume(struct drm_device *dev)
{
struct drm_i915_private *dev_priv = to_i915(dev);
2016-05-07 01:35:55 +07:00
int ret;
disable_rpm_wakeref_asserts(&dev_priv->runtime_pm);
drm/i915: add support for checking if we hold an RPM reference Atm, we assert that the device is not suspended until the point when the device is truly put to a suspended state. This is fine, but we can catch more problems if we check that RPM refcount is non-zero. After that one drops to zero we shouldn't access the device any more, even if the actual device suspend may be delayed. Change assert_rpm_wakelock_held() accordingly to check for a non-zero RPM refcount in addition to the current device-not-suspended check. For the new asserts to work we need to annotate every place explicitly in the code where we expect that the device is powered. The places where we only assume this, but may not hold an RPM reference: - driver load We assume the device to be powered until we enable RPM. Make this explicit by taking an RPM reference around the load function. - system and runtime sudpend/resume handlers These handlers are called when the RPM reference becomes 0 and know the exact point after which the device can get powered off. Disable the RPM-reference-held check for their duration. - the IRQ, hangcheck and RPS work handlers These handlers are flushed in the system/runtime suspend handler before the device is powered off, so it's guaranteed that they won't run while the device is powered off even though they don't hold any RPM reference. Disable the RPM-reference-held check for their duration. In all these cases we still check that the device is not suspended. These explicit annotations also have the positive side effect of documenting our assumptions better. This caught additional WARNs from the atomic modeset path, those should be fixed separately. v2: - remove the redundant HAS_RUNTIME_PM check (moved to patch 1) (Ville) v3: - use a new dedicated RPM wakelock refcount to also catch cases where our own RPM get/put functions were not called (Chris) - assert also that the new RPM wakelock refcount is 0 in the RPM suspend handler (Chris) - change the assert error message to be more meaningful (Chris) - prevent false assert errors and check that the RPM wakelock is 0 in the RPM resume handler too - prevent false assert errors in the hangcheck work too - add a device not suspended assert check to the hangcheck work v4: - rename disable/enable_rpm_asserts to disable/enable_rpm_wakeref_asserts and wakelock_count to wakeref_count - disable the wakeref asserts in the IRQ handlers and RPS work too - update/clarify commit message v5: - mark places we plan to change to use proper RPM refcounting with separate DISABLE/ENABLE_RPM_WAKEREF_ASSERTS aliases (Chris) Signed-off-by: Imre Deak <imre.deak@intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Link: http://patchwork.freedesktop.org/patch/msgid/1450227139-13471-1-git-send-email-imre.deak@intel.com
2015-12-16 07:52:19 +07:00
intel_gt_sanitize(&dev_priv->gt, true);
ret = i915_ggtt_enable_hw(dev_priv);
2016-05-07 01:35:55 +07:00
if (ret)
DRM_ERROR("failed to re-enable GGTT\n");
i915_gem_restore_gtt_mappings(dev_priv);
i915_gem_restore_fences(&dev_priv->ggtt);
intel_csr_ucode_resume(dev_priv);
i915_restore_state(dev_priv);
intel_pps_unlock_regs_wa(dev_priv);
intel_init_pch_refclk(dev_priv);
/*
* Interrupts have to be enabled before any batches are run. If not the
* GPU will hang. i915_gem_init_hw() will initiate batches to
* update/restore the context.
*
* drm_mode_config_reset() needs AUX interrupts.
*
* Modeset enabling in intel_modeset_init_hw() also needs working
* interrupts.
*/
intel_runtime_pm_enable_interrupts(dev_priv);
drm_mode_config_reset(dev);
i915_gem_resume(dev_priv);
intel_modeset_init_hw(dev_priv);
intel_init_clock_gating(dev_priv);
spin_lock_irq(&dev_priv->irq_lock);
if (dev_priv->display.hpd_irq_setup)
drm/i915: Small display interrupt handlers tidy I have noticed some of our interrupt handlers use both dev and dev_priv while they could get away with only dev_priv in the huge majority of cases. Tidying that up had a cascading effect on changing functions prototypes, so relatively big churn factor, but I think it is for the better. For example even where changes cascade out of i915_irq.c, for functions prefixed with intel_, genX_ or <plat>_, it makes more sense to take dev_priv directly anyway. This allows us to eliminate local variables and intermixed usage of dev and dev_priv where only one is good enough. End result is shrinkage of both source and the resulting binary. i915.ko: - .text 000b0899 + .text 000b0619 Or if we look at the Gen8 display irq chain: -00000000000006ad t gen8_irq_handler +0000000000000663 t gen8_irq_handler -0000000000000028 T intel_opregion_asle_intr +0000000000000024 T intel_opregion_asle_intr -000000000000008c t ilk_hpd_irq_handler +000000000000007f t ilk_hpd_irq_handler -0000000000000116 T intel_check_page_flip +0000000000000112 T intel_check_page_flip -000000000000011a T intel_prepare_page_flip +0000000000000119 T intel_prepare_page_flip -0000000000000014 T intel_finish_page_flip_plane +0000000000000013 T intel_finish_page_flip_plane -0000000000000053 t hsw_pipe_crc_irq_handler +000000000000004c t hsw_pipe_crc_irq_handler -000000000000022e t cpt_irq_handler +0000000000000213 t cpt_irq_handler So small shrinkage but it is all fast paths so doesn't harm. Situation is similar in other interrupt handlers as well. v2: Tidy intel_queue_rps_boost_for_request as well. (Chris Wilson) Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Cc: Chris Wilson <chris@chris-wilson.co.uk> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
2016-05-06 20:48:28 +07:00
dev_priv->display.hpd_irq_setup(dev_priv);
spin_unlock_irq(&dev_priv->irq_lock);
2014-05-02 11:02:48 +07:00
intel_dp_mst_resume(dev_priv);
intel_display_resume(dev);
drm_kms_helper_poll_enable(dev);
/*
* ... but also need to make sure that hotplug processing
* doesn't cause havoc. Like in the driver load code we don't
* bother with the tiny race here where we might lose hotplug
* notifications.
* */
intel_hpd_init(dev_priv);
intel_opregion_resume(dev_priv);
intel_fbdev_set_suspend(dev, FBINFO_STATE_RUNNING, false);
intel_power_domains_enable(dev_priv);
enable_rpm_wakeref_asserts(&dev_priv->runtime_pm);
drm/i915: add support for checking if we hold an RPM reference Atm, we assert that the device is not suspended until the point when the device is truly put to a suspended state. This is fine, but we can catch more problems if we check that RPM refcount is non-zero. After that one drops to zero we shouldn't access the device any more, even if the actual device suspend may be delayed. Change assert_rpm_wakelock_held() accordingly to check for a non-zero RPM refcount in addition to the current device-not-suspended check. For the new asserts to work we need to annotate every place explicitly in the code where we expect that the device is powered. The places where we only assume this, but may not hold an RPM reference: - driver load We assume the device to be powered until we enable RPM. Make this explicit by taking an RPM reference around the load function. - system and runtime sudpend/resume handlers These handlers are called when the RPM reference becomes 0 and know the exact point after which the device can get powered off. Disable the RPM-reference-held check for their duration. - the IRQ, hangcheck and RPS work handlers These handlers are flushed in the system/runtime suspend handler before the device is powered off, so it's guaranteed that they won't run while the device is powered off even though they don't hold any RPM reference. Disable the RPM-reference-held check for their duration. In all these cases we still check that the device is not suspended. These explicit annotations also have the positive side effect of documenting our assumptions better. This caught additional WARNs from the atomic modeset path, those should be fixed separately. v2: - remove the redundant HAS_RUNTIME_PM check (moved to patch 1) (Ville) v3: - use a new dedicated RPM wakelock refcount to also catch cases where our own RPM get/put functions were not called (Chris) - assert also that the new RPM wakelock refcount is 0 in the RPM suspend handler (Chris) - change the assert error message to be more meaningful (Chris) - prevent false assert errors and check that the RPM wakelock is 0 in the RPM resume handler too - prevent false assert errors in the hangcheck work too - add a device not suspended assert check to the hangcheck work v4: - rename disable/enable_rpm_asserts to disable/enable_rpm_wakeref_asserts and wakelock_count to wakeref_count - disable the wakeref asserts in the IRQ handlers and RPS work too - update/clarify commit message v5: - mark places we plan to change to use proper RPM refcounting with separate DISABLE/ENABLE_RPM_WAKEREF_ASSERTS aliases (Chris) Signed-off-by: Imre Deak <imre.deak@intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Link: http://patchwork.freedesktop.org/patch/msgid/1450227139-13471-1-git-send-email-imre.deak@intel.com
2015-12-16 07:52:19 +07:00
return 0;
}
static int i915_drm_resume_early(struct drm_device *dev)
{
struct drm_i915_private *dev_priv = to_i915(dev);
struct pci_dev *pdev = dev_priv->drm.pdev;
drm/i915: Fix system resume if PCI device remained enabled During system resume we depended on pci_enable_device() also putting the device into PCI D0 state. This won't work if the PCI device was already enabled but still in D3 state. This is because pci_enable_device() is refcounted and will not change the HW state if called with a non-zero refcount. Leaving the device in D3 will make all subsequent device accesses fail. This didn't cause a problem most of the time, since we resumed with an enable refcount of 0. But it fails at least after module reload because after that we also happen to leak a PCI device enable reference: During probing we call drm_get_pci_dev() which will enable the PCI device, but during device removal drm_put_dev() won't disable it. This is a bug of its own in DRM core, but without much harm as it only leaves the PCI device enabled. Fixing it is also a bit more involved, due to DRM mid-layering and because it affects non-i915 drivers too. The fix in this patch is valid regardless of the problem in DRM core. v2: - Add a code comment about the relation of this fix to the freeze/thaw vs. the suspend/resume phases. (Ville) - Add a code comment about the inconsistent ordering of set power state and device enable calls. (Chris) CC: Ville Syrjälä <ville.syrjala@linux.intel.com> CC: Chris Wilson <chris@chris-wilson.co.uk> CC: stable@vger.kernel.org Signed-off-by: Imre Deak <imre.deak@intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: http://patchwork.freedesktop.org/patch/msgid/1460979954-14503-1-git-send-email-imre.deak@intel.com
2016-04-18 18:45:54 +07:00
int ret;
/*
* We have a resume ordering issue with the snd-hda driver also
* requiring our device to be power up. Due to the lack of a
* parent/child relationship we currently solve this with an early
* resume hook.
*
* FIXME: This should be solved with a special hdmi sink device or
* similar so that power domains can be employed.
*/
drm/i915: Fix system resume if PCI device remained enabled During system resume we depended on pci_enable_device() also putting the device into PCI D0 state. This won't work if the PCI device was already enabled but still in D3 state. This is because pci_enable_device() is refcounted and will not change the HW state if called with a non-zero refcount. Leaving the device in D3 will make all subsequent device accesses fail. This didn't cause a problem most of the time, since we resumed with an enable refcount of 0. But it fails at least after module reload because after that we also happen to leak a PCI device enable reference: During probing we call drm_get_pci_dev() which will enable the PCI device, but during device removal drm_put_dev() won't disable it. This is a bug of its own in DRM core, but without much harm as it only leaves the PCI device enabled. Fixing it is also a bit more involved, due to DRM mid-layering and because it affects non-i915 drivers too. The fix in this patch is valid regardless of the problem in DRM core. v2: - Add a code comment about the relation of this fix to the freeze/thaw vs. the suspend/resume phases. (Ville) - Add a code comment about the inconsistent ordering of set power state and device enable calls. (Chris) CC: Ville Syrjälä <ville.syrjala@linux.intel.com> CC: Chris Wilson <chris@chris-wilson.co.uk> CC: stable@vger.kernel.org Signed-off-by: Imre Deak <imre.deak@intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: http://patchwork.freedesktop.org/patch/msgid/1460979954-14503-1-git-send-email-imre.deak@intel.com
2016-04-18 18:45:54 +07:00
/*
* 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(pdev, PCI_D0);
drm/i915: Fix system resume if PCI device remained enabled During system resume we depended on pci_enable_device() also putting the device into PCI D0 state. This won't work if the PCI device was already enabled but still in D3 state. This is because pci_enable_device() is refcounted and will not change the HW state if called with a non-zero refcount. Leaving the device in D3 will make all subsequent device accesses fail. This didn't cause a problem most of the time, since we resumed with an enable refcount of 0. But it fails at least after module reload because after that we also happen to leak a PCI device enable reference: During probing we call drm_get_pci_dev() which will enable the PCI device, but during device removal drm_put_dev() won't disable it. This is a bug of its own in DRM core, but without much harm as it only leaves the PCI device enabled. Fixing it is also a bit more involved, due to DRM mid-layering and because it affects non-i915 drivers too. The fix in this patch is valid regardless of the problem in DRM core. v2: - Add a code comment about the relation of this fix to the freeze/thaw vs. the suspend/resume phases. (Ville) - Add a code comment about the inconsistent ordering of set power state and device enable calls. (Chris) CC: Ville Syrjälä <ville.syrjala@linux.intel.com> CC: Chris Wilson <chris@chris-wilson.co.uk> CC: stable@vger.kernel.org Signed-off-by: Imre Deak <imre.deak@intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: http://patchwork.freedesktop.org/patch/msgid/1460979954-14503-1-git-send-email-imre.deak@intel.com
2016-04-18 18:45:54 +07:00
if (ret) {
DRM_ERROR("failed to set PCI D0 power state (%d)\n", ret);
return ret;
drm/i915: Fix system resume if PCI device remained enabled During system resume we depended on pci_enable_device() also putting the device into PCI D0 state. This won't work if the PCI device was already enabled but still in D3 state. This is because pci_enable_device() is refcounted and will not change the HW state if called with a non-zero refcount. Leaving the device in D3 will make all subsequent device accesses fail. This didn't cause a problem most of the time, since we resumed with an enable refcount of 0. But it fails at least after module reload because after that we also happen to leak a PCI device enable reference: During probing we call drm_get_pci_dev() which will enable the PCI device, but during device removal drm_put_dev() won't disable it. This is a bug of its own in DRM core, but without much harm as it only leaves the PCI device enabled. Fixing it is also a bit more involved, due to DRM mid-layering and because it affects non-i915 drivers too. The fix in this patch is valid regardless of the problem in DRM core. v2: - Add a code comment about the relation of this fix to the freeze/thaw vs. the suspend/resume phases. (Ville) - Add a code comment about the inconsistent ordering of set power state and device enable calls. (Chris) CC: Ville Syrjälä <ville.syrjala@linux.intel.com> CC: Chris Wilson <chris@chris-wilson.co.uk> CC: stable@vger.kernel.org Signed-off-by: Imre Deak <imre.deak@intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: http://patchwork.freedesktop.org/patch/msgid/1460979954-14503-1-git-send-email-imre.deak@intel.com
2016-04-18 18:45:54 +07:00
}
/*
* 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(pdev))
return -EIO;
pci_set_master(pdev);
disable_rpm_wakeref_asserts(&dev_priv->runtime_pm);
drm/i915: add support for checking if we hold an RPM reference Atm, we assert that the device is not suspended until the point when the device is truly put to a suspended state. This is fine, but we can catch more problems if we check that RPM refcount is non-zero. After that one drops to zero we shouldn't access the device any more, even if the actual device suspend may be delayed. Change assert_rpm_wakelock_held() accordingly to check for a non-zero RPM refcount in addition to the current device-not-suspended check. For the new asserts to work we need to annotate every place explicitly in the code where we expect that the device is powered. The places where we only assume this, but may not hold an RPM reference: - driver load We assume the device to be powered until we enable RPM. Make this explicit by taking an RPM reference around the load function. - system and runtime sudpend/resume handlers These handlers are called when the RPM reference becomes 0 and know the exact point after which the device can get powered off. Disable the RPM-reference-held check for their duration. - the IRQ, hangcheck and RPS work handlers These handlers are flushed in the system/runtime suspend handler before the device is powered off, so it's guaranteed that they won't run while the device is powered off even though they don't hold any RPM reference. Disable the RPM-reference-held check for their duration. In all these cases we still check that the device is not suspended. These explicit annotations also have the positive side effect of documenting our assumptions better. This caught additional WARNs from the atomic modeset path, those should be fixed separately. v2: - remove the redundant HAS_RUNTIME_PM check (moved to patch 1) (Ville) v3: - use a new dedicated RPM wakelock refcount to also catch cases where our own RPM get/put functions were not called (Chris) - assert also that the new RPM wakelock refcount is 0 in the RPM suspend handler (Chris) - change the assert error message to be more meaningful (Chris) - prevent false assert errors and check that the RPM wakelock is 0 in the RPM resume handler too - prevent false assert errors in the hangcheck work too - add a device not suspended assert check to the hangcheck work v4: - rename disable/enable_rpm_asserts to disable/enable_rpm_wakeref_asserts and wakelock_count to wakeref_count - disable the wakeref asserts in the IRQ handlers and RPS work too - update/clarify commit message v5: - mark places we plan to change to use proper RPM refcounting with separate DISABLE/ENABLE_RPM_WAKEREF_ASSERTS aliases (Chris) Signed-off-by: Imre Deak <imre.deak@intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Link: http://patchwork.freedesktop.org/patch/msgid/1450227139-13471-1-git-send-email-imre.deak@intel.com
2015-12-16 07:52:19 +07:00
if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv))
ret = vlv_resume_prepare(dev_priv, false);
if (ret)
DRM_ERROR("Resume prepare failed: %d, continuing anyway\n",
ret);
intel_uncore_resume_early(&dev_priv->uncore);
intel_gt_check_and_clear_faults(&dev_priv->gt);
intel_display_power_resume_early(dev_priv);
intel_power_domains_resume(dev_priv);
enable_rpm_wakeref_asserts(&dev_priv->runtime_pm);
return ret;
}
int i915_resume_switcheroo(struct drm_i915_private *i915)
{
int ret;
if (i915->drm.switch_power_state == DRM_SWITCH_POWER_OFF)
return 0;
ret = i915_drm_resume_early(&i915->drm);
if (ret)
return ret;
return i915_drm_resume(&i915->drm);
}
static int i915_pm_prepare(struct device *kdev)
{
struct drm_i915_private *i915 = kdev_to_i915(kdev);
if (!i915) {
dev_err(kdev, "DRM not initialized, aborting suspend.\n");
return -ENODEV;
}
if (i915->drm.switch_power_state == DRM_SWITCH_POWER_OFF)
return 0;
return i915_drm_prepare(&i915->drm);
}
static int i915_pm_suspend(struct device *kdev)
{
struct drm_i915_private *i915 = kdev_to_i915(kdev);
if (!i915) {
dev_err(kdev, "DRM not initialized, aborting suspend.\n");
return -ENODEV;
}
if (i915->drm.switch_power_state == DRM_SWITCH_POWER_OFF)
return 0;
return i915_drm_suspend(&i915->drm);
}
static int i915_pm_suspend_late(struct device *kdev)
{
struct drm_i915_private *i915 = kdev_to_i915(kdev);
/*
* We have a suspend ordering issue with the snd-hda driver also
* requiring our device to be power up. Due to the lack of a
* parent/child relationship we currently solve this with an late
* suspend hook.
*
* FIXME: This should be solved with a special hdmi sink device or
* similar so that power domains can be employed.
*/
if (i915->drm.switch_power_state == DRM_SWITCH_POWER_OFF)
return 0;
return i915_drm_suspend_late(&i915->drm, false);
2015-03-02 18:04:41 +07:00
}
static int i915_pm_poweroff_late(struct device *kdev)
2015-03-02 18:04:41 +07:00
{
struct drm_i915_private *i915 = kdev_to_i915(kdev);
2015-03-02 18:04:41 +07:00
if (i915->drm.switch_power_state == DRM_SWITCH_POWER_OFF)
2015-03-02 18:04:41 +07:00
return 0;
return i915_drm_suspend_late(&i915->drm, true);
}
static int i915_pm_resume_early(struct device *kdev)
{
struct drm_i915_private *i915 = kdev_to_i915(kdev);
if (i915->drm.switch_power_state == DRM_SWITCH_POWER_OFF)
return 0;
return i915_drm_resume_early(&i915->drm);
}
static int i915_pm_resume(struct device *kdev)
{
struct drm_i915_private *i915 = kdev_to_i915(kdev);
if (i915->drm.switch_power_state == DRM_SWITCH_POWER_OFF)
return 0;
return i915_drm_resume(&i915->drm);
}
/* freeze: before creating the hibernation_image */
static int i915_pm_freeze(struct device *kdev)
{
struct drm_i915_private *i915 = kdev_to_i915(kdev);
drm/i915: Only shrink the unbound objects during freeze At the point of creating the hibernation image, the runtime power manage core is disabled - and using the rpm functions triggers a warn. i915_gem_shrink_all() tries to unbind objects, which requires device access and so tries to how an rpm reference triggering a warning: [ 44.235420] ------------[ cut here ]------------ [ 44.235424] WARNING: CPU: 2 PID: 2199 at drivers/gpu/drm/i915/intel_runtime_pm.c:2688 intel_runtime_pm_get_if_in_use+0xe6/0xf0 [ 44.235426] WARN_ON_ONCE(ret < 0) [ 44.235445] Modules linked in: ctr ccm arc4 rt2800usb rt2x00usb rt2800lib rt2x00lib crc_ccitt mac80211 cmac cfg80211 btusb rfcomm bnep btrtl btbcm btintel bluetooth dcdbas x86_pkg_temp_thermal intel_powerclamp coretemp snd_hda_codec_realtek crct10dif_pclmul crc32_pclmul ghash_clmulni_intel snd_hda_codec_generic aesni_intel snd_hda_codec_hdmi aes_x86_64 lrw gf128mul snd_hda_intel glue_helper ablk_helper cryptd snd_hda_codec hid_multitouch joydev snd_hda_core binfmt_misc i2c_hid serio_raw snd_pcm acpi_pad snd_timer snd i2c_designware_platform 8250_dw nls_iso8859_1 i2c_designware_core lpc_ich mfd_core soundcore usbhid hid psmouse ahci libahci [ 44.235447] CPU: 2 PID: 2199 Comm: kworker/u8:8 Not tainted 4.8.0-rc5+ #130 [ 44.235447] Hardware name: Dell Inc. XPS 13 9343/0310JH, BIOS A07 11/11/2015 [ 44.235450] Workqueue: events_unbound async_run_entry_fn [ 44.235453] 0000000000000000 ffff8801b2f7fb98 ffffffff81306c2f ffff8801b2f7fbe8 [ 44.235454] 0000000000000000 ffff8801b2f7fbd8 ffffffff81056c01 00000a801f50ecc0 [ 44.235456] ffff88020ce50000 ffff88020ce59b60 ffffffff81a60b5c ffffffff81414840 [ 44.235456] Call Trace: [ 44.235459] [<ffffffff81306c2f>] dump_stack+0x4d/0x6e [ 44.235461] [<ffffffff81056c01>] __warn+0xd1/0xf0 [ 44.235464] [<ffffffff81414840>] ? i915_pm_suspend_late+0x30/0x30 [ 44.235465] [<ffffffff81056c6f>] warn_slowpath_fmt+0x4f/0x60 [ 44.235468] [<ffffffff814e73ce>] ? pm_runtime_get_if_in_use+0x6e/0xa0 [ 44.235469] [<ffffffff81433526>] intel_runtime_pm_get_if_in_use+0xe6/0xf0 [ 44.235471] [<ffffffff81458a26>] i915_gem_shrink+0x306/0x360 [ 44.235473] [<ffffffff81343fd4>] ? pci_platform_power_transition+0x24/0x90 [ 44.235475] [<ffffffff81414840>] ? i915_pm_suspend_late+0x30/0x30 [ 44.235476] [<ffffffff81458dfb>] i915_gem_shrink_all+0x1b/0x30 [ 44.235478] [<ffffffff814560b3>] i915_gem_freeze_late+0x33/0x90 [ 44.235479] [<ffffffff81414877>] i915_pm_freeze_late+0x37/0x40 [ 44.235481] [<ffffffff814e9b8e>] dpm_run_callback+0x4e/0x130 [ 44.235483] [<ffffffff814ea5db>] __device_suspend_late+0xdb/0x1f0 [ 44.235484] [<ffffffff814ea70f>] async_suspend_late+0x1f/0xa0 [ 44.235486] [<ffffffff81077557>] async_run_entry_fn+0x37/0x150 [ 44.235488] [<ffffffff8106f518>] process_one_work+0x148/0x3f0 [ 44.235490] [<ffffffff8106f8eb>] worker_thread+0x12b/0x490 [ 44.235491] [<ffffffff8106f7c0>] ? process_one_work+0x3f0/0x3f0 [ 44.235492] [<ffffffff81074d09>] kthread+0xc9/0xe0 [ 44.235495] [<ffffffff816e257f>] ret_from_fork+0x1f/0x40 [ 44.235496] [<ffffffff81074c40>] ? kthread_park+0x60/0x60 [ 44.235497] ---[ end trace e438706b97c7f132 ]--- Alternatively, to actually shrink everything we have to do so slightly earlier in the hibernation process. To keep lockdep silent, we need to take struct_mutex for the shrinker even though we know that we are the only user during the freeze. Fixes: 7aab2d534e35 ("drm/i915: Shrink objects prior to hibernation") Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com> Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com> Link: http://patchwork.freedesktop.org/patch/msgid/20160921135108.29574-2-chris@chris-wilson.co.uk
2016-09-21 20:51:07 +07:00
int ret;
if (i915->drm.switch_power_state != DRM_SWITCH_POWER_OFF) {
ret = i915_drm_suspend(&i915->drm);
if (ret)
return ret;
}
drm/i915: Only shrink the unbound objects during freeze At the point of creating the hibernation image, the runtime power manage core is disabled - and using the rpm functions triggers a warn. i915_gem_shrink_all() tries to unbind objects, which requires device access and so tries to how an rpm reference triggering a warning: [ 44.235420] ------------[ cut here ]------------ [ 44.235424] WARNING: CPU: 2 PID: 2199 at drivers/gpu/drm/i915/intel_runtime_pm.c:2688 intel_runtime_pm_get_if_in_use+0xe6/0xf0 [ 44.235426] WARN_ON_ONCE(ret < 0) [ 44.235445] Modules linked in: ctr ccm arc4 rt2800usb rt2x00usb rt2800lib rt2x00lib crc_ccitt mac80211 cmac cfg80211 btusb rfcomm bnep btrtl btbcm btintel bluetooth dcdbas x86_pkg_temp_thermal intel_powerclamp coretemp snd_hda_codec_realtek crct10dif_pclmul crc32_pclmul ghash_clmulni_intel snd_hda_codec_generic aesni_intel snd_hda_codec_hdmi aes_x86_64 lrw gf128mul snd_hda_intel glue_helper ablk_helper cryptd snd_hda_codec hid_multitouch joydev snd_hda_core binfmt_misc i2c_hid serio_raw snd_pcm acpi_pad snd_timer snd i2c_designware_platform 8250_dw nls_iso8859_1 i2c_designware_core lpc_ich mfd_core soundcore usbhid hid psmouse ahci libahci [ 44.235447] CPU: 2 PID: 2199 Comm: kworker/u8:8 Not tainted 4.8.0-rc5+ #130 [ 44.235447] Hardware name: Dell Inc. XPS 13 9343/0310JH, BIOS A07 11/11/2015 [ 44.235450] Workqueue: events_unbound async_run_entry_fn [ 44.235453] 0000000000000000 ffff8801b2f7fb98 ffffffff81306c2f ffff8801b2f7fbe8 [ 44.235454] 0000000000000000 ffff8801b2f7fbd8 ffffffff81056c01 00000a801f50ecc0 [ 44.235456] ffff88020ce50000 ffff88020ce59b60 ffffffff81a60b5c ffffffff81414840 [ 44.235456] Call Trace: [ 44.235459] [<ffffffff81306c2f>] dump_stack+0x4d/0x6e [ 44.235461] [<ffffffff81056c01>] __warn+0xd1/0xf0 [ 44.235464] [<ffffffff81414840>] ? i915_pm_suspend_late+0x30/0x30 [ 44.235465] [<ffffffff81056c6f>] warn_slowpath_fmt+0x4f/0x60 [ 44.235468] [<ffffffff814e73ce>] ? pm_runtime_get_if_in_use+0x6e/0xa0 [ 44.235469] [<ffffffff81433526>] intel_runtime_pm_get_if_in_use+0xe6/0xf0 [ 44.235471] [<ffffffff81458a26>] i915_gem_shrink+0x306/0x360 [ 44.235473] [<ffffffff81343fd4>] ? pci_platform_power_transition+0x24/0x90 [ 44.235475] [<ffffffff81414840>] ? i915_pm_suspend_late+0x30/0x30 [ 44.235476] [<ffffffff81458dfb>] i915_gem_shrink_all+0x1b/0x30 [ 44.235478] [<ffffffff814560b3>] i915_gem_freeze_late+0x33/0x90 [ 44.235479] [<ffffffff81414877>] i915_pm_freeze_late+0x37/0x40 [ 44.235481] [<ffffffff814e9b8e>] dpm_run_callback+0x4e/0x130 [ 44.235483] [<ffffffff814ea5db>] __device_suspend_late+0xdb/0x1f0 [ 44.235484] [<ffffffff814ea70f>] async_suspend_late+0x1f/0xa0 [ 44.235486] [<ffffffff81077557>] async_run_entry_fn+0x37/0x150 [ 44.235488] [<ffffffff8106f518>] process_one_work+0x148/0x3f0 [ 44.235490] [<ffffffff8106f8eb>] worker_thread+0x12b/0x490 [ 44.235491] [<ffffffff8106f7c0>] ? process_one_work+0x3f0/0x3f0 [ 44.235492] [<ffffffff81074d09>] kthread+0xc9/0xe0 [ 44.235495] [<ffffffff816e257f>] ret_from_fork+0x1f/0x40 [ 44.235496] [<ffffffff81074c40>] ? kthread_park+0x60/0x60 [ 44.235497] ---[ end trace e438706b97c7f132 ]--- Alternatively, to actually shrink everything we have to do so slightly earlier in the hibernation process. To keep lockdep silent, we need to take struct_mutex for the shrinker even though we know that we are the only user during the freeze. Fixes: 7aab2d534e35 ("drm/i915: Shrink objects prior to hibernation") Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com> Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com> Link: http://patchwork.freedesktop.org/patch/msgid/20160921135108.29574-2-chris@chris-wilson.co.uk
2016-09-21 20:51:07 +07:00
ret = i915_gem_freeze(i915);
drm/i915: Only shrink the unbound objects during freeze At the point of creating the hibernation image, the runtime power manage core is disabled - and using the rpm functions triggers a warn. i915_gem_shrink_all() tries to unbind objects, which requires device access and so tries to how an rpm reference triggering a warning: [ 44.235420] ------------[ cut here ]------------ [ 44.235424] WARNING: CPU: 2 PID: 2199 at drivers/gpu/drm/i915/intel_runtime_pm.c:2688 intel_runtime_pm_get_if_in_use+0xe6/0xf0 [ 44.235426] WARN_ON_ONCE(ret < 0) [ 44.235445] Modules linked in: ctr ccm arc4 rt2800usb rt2x00usb rt2800lib rt2x00lib crc_ccitt mac80211 cmac cfg80211 btusb rfcomm bnep btrtl btbcm btintel bluetooth dcdbas x86_pkg_temp_thermal intel_powerclamp coretemp snd_hda_codec_realtek crct10dif_pclmul crc32_pclmul ghash_clmulni_intel snd_hda_codec_generic aesni_intel snd_hda_codec_hdmi aes_x86_64 lrw gf128mul snd_hda_intel glue_helper ablk_helper cryptd snd_hda_codec hid_multitouch joydev snd_hda_core binfmt_misc i2c_hid serio_raw snd_pcm acpi_pad snd_timer snd i2c_designware_platform 8250_dw nls_iso8859_1 i2c_designware_core lpc_ich mfd_core soundcore usbhid hid psmouse ahci libahci [ 44.235447] CPU: 2 PID: 2199 Comm: kworker/u8:8 Not tainted 4.8.0-rc5+ #130 [ 44.235447] Hardware name: Dell Inc. XPS 13 9343/0310JH, BIOS A07 11/11/2015 [ 44.235450] Workqueue: events_unbound async_run_entry_fn [ 44.235453] 0000000000000000 ffff8801b2f7fb98 ffffffff81306c2f ffff8801b2f7fbe8 [ 44.235454] 0000000000000000 ffff8801b2f7fbd8 ffffffff81056c01 00000a801f50ecc0 [ 44.235456] ffff88020ce50000 ffff88020ce59b60 ffffffff81a60b5c ffffffff81414840 [ 44.235456] Call Trace: [ 44.235459] [<ffffffff81306c2f>] dump_stack+0x4d/0x6e [ 44.235461] [<ffffffff81056c01>] __warn+0xd1/0xf0 [ 44.235464] [<ffffffff81414840>] ? i915_pm_suspend_late+0x30/0x30 [ 44.235465] [<ffffffff81056c6f>] warn_slowpath_fmt+0x4f/0x60 [ 44.235468] [<ffffffff814e73ce>] ? pm_runtime_get_if_in_use+0x6e/0xa0 [ 44.235469] [<ffffffff81433526>] intel_runtime_pm_get_if_in_use+0xe6/0xf0 [ 44.235471] [<ffffffff81458a26>] i915_gem_shrink+0x306/0x360 [ 44.235473] [<ffffffff81343fd4>] ? pci_platform_power_transition+0x24/0x90 [ 44.235475] [<ffffffff81414840>] ? i915_pm_suspend_late+0x30/0x30 [ 44.235476] [<ffffffff81458dfb>] i915_gem_shrink_all+0x1b/0x30 [ 44.235478] [<ffffffff814560b3>] i915_gem_freeze_late+0x33/0x90 [ 44.235479] [<ffffffff81414877>] i915_pm_freeze_late+0x37/0x40 [ 44.235481] [<ffffffff814e9b8e>] dpm_run_callback+0x4e/0x130 [ 44.235483] [<ffffffff814ea5db>] __device_suspend_late+0xdb/0x1f0 [ 44.235484] [<ffffffff814ea70f>] async_suspend_late+0x1f/0xa0 [ 44.235486] [<ffffffff81077557>] async_run_entry_fn+0x37/0x150 [ 44.235488] [<ffffffff8106f518>] process_one_work+0x148/0x3f0 [ 44.235490] [<ffffffff8106f8eb>] worker_thread+0x12b/0x490 [ 44.235491] [<ffffffff8106f7c0>] ? process_one_work+0x3f0/0x3f0 [ 44.235492] [<ffffffff81074d09>] kthread+0xc9/0xe0 [ 44.235495] [<ffffffff816e257f>] ret_from_fork+0x1f/0x40 [ 44.235496] [<ffffffff81074c40>] ? kthread_park+0x60/0x60 [ 44.235497] ---[ end trace e438706b97c7f132 ]--- Alternatively, to actually shrink everything we have to do so slightly earlier in the hibernation process. To keep lockdep silent, we need to take struct_mutex for the shrinker even though we know that we are the only user during the freeze. Fixes: 7aab2d534e35 ("drm/i915: Shrink objects prior to hibernation") Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com> Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com> Link: http://patchwork.freedesktop.org/patch/msgid/20160921135108.29574-2-chris@chris-wilson.co.uk
2016-09-21 20:51:07 +07:00
if (ret)
return ret;
return 0;
}
static int i915_pm_freeze_late(struct device *kdev)
{
struct drm_i915_private *i915 = kdev_to_i915(kdev);
int ret;
if (i915->drm.switch_power_state != DRM_SWITCH_POWER_OFF) {
ret = i915_drm_suspend_late(&i915->drm, true);
if (ret)
return ret;
}
ret = i915_gem_freeze_late(i915);
if (ret)
return ret;
return 0;
}
/* thaw: called after creating the hibernation image, but before turning off. */
static int i915_pm_thaw_early(struct device *kdev)
{
return i915_pm_resume_early(kdev);
}
static int i915_pm_thaw(struct device *kdev)
{
return i915_pm_resume(kdev);
}
/* restore: called after loading the hibernation image. */
static int i915_pm_restore_early(struct device *kdev)
{
return i915_pm_resume_early(kdev);
}
static int i915_pm_restore(struct device *kdev)
{
return i915_pm_resume(kdev);
}
/*
* 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
* defined in the VLV2_S0IXRegs document. This documents marks all Gunit
* registers in the following way:
* - Driver: saved/restored by the driver
* - Punit : saved/restored by the Punit firmware
* - No, w/o marking: no need to save/restore, since the register is R/O or
* used internally by the HW in a way that doesn't depend
* keeping the content across a suspend/resume.
* - Debug : used for debugging
*
* We save/restore all registers marked with 'Driver', with the following
* exceptions:
* - Registers out of use, including also registers marked with 'Debug'.
* These have no effect on the driver's operation, so we don't save/restore
* them to reduce the overhead.
* - Registers that are fully setup by an initialization function called from
* the resume path. For example many clock gating and RPS/RC6 registers.
* - Registers that provide the right functionality with their reset defaults.
*
* TODO: Except for registers that based on the above 3 criteria can be safely
* ignored, we save/restore all others, practically treating the HW context as
* a black-box for the driver. Further investigation is needed to reduce the
* saved/restored registers even further, by following the same 3 criteria.
*/
static void vlv_save_gunit_s0ix_state(struct drm_i915_private *dev_priv)
{
struct vlv_s0ix_state *s = dev_priv->vlv_s0ix_state;
int i;
if (!s)
return;
/* GAM 0x4000-0x4770 */
s->wr_watermark = I915_READ(GEN7_WR_WATERMARK);
s->gfx_prio_ctrl = I915_READ(GEN7_GFX_PRIO_CTRL);
s->arb_mode = I915_READ(ARB_MODE);
s->gfx_pend_tlb0 = I915_READ(GEN7_GFX_PEND_TLB0);
s->gfx_pend_tlb1 = I915_READ(GEN7_GFX_PEND_TLB1);
for (i = 0; i < ARRAY_SIZE(s->lra_limits); i++)
s->lra_limits[i] = I915_READ(GEN7_LRA_LIMITS(i));
s->media_max_req_count = I915_READ(GEN7_MEDIA_MAX_REQ_COUNT);
s->gfx_max_req_count = I915_READ(GEN7_GFX_MAX_REQ_COUNT);
s->render_hwsp = I915_READ(RENDER_HWS_PGA_GEN7);
s->ecochk = I915_READ(GAM_ECOCHK);
s->bsd_hwsp = I915_READ(BSD_HWS_PGA_GEN7);
s->blt_hwsp = I915_READ(BLT_HWS_PGA_GEN7);
s->tlb_rd_addr = I915_READ(GEN7_TLB_RD_ADDR);
/* MBC 0x9024-0x91D0, 0x8500 */
s->g3dctl = I915_READ(VLV_G3DCTL);
s->gsckgctl = I915_READ(VLV_GSCKGCTL);
s->mbctl = I915_READ(GEN6_MBCTL);
/* GCP 0x9400-0x9424, 0x8100-0x810C */
s->ucgctl1 = I915_READ(GEN6_UCGCTL1);
s->ucgctl3 = I915_READ(GEN6_UCGCTL3);
s->rcgctl1 = I915_READ(GEN6_RCGCTL1);
s->rcgctl2 = I915_READ(GEN6_RCGCTL2);
s->rstctl = I915_READ(GEN6_RSTCTL);
s->misccpctl = I915_READ(GEN7_MISCCPCTL);
/* GPM 0xA000-0xAA84, 0x8000-0x80FC */
s->gfxpause = I915_READ(GEN6_GFXPAUSE);
s->rpdeuhwtc = I915_READ(GEN6_RPDEUHWTC);
s->rpdeuc = I915_READ(GEN6_RPDEUC);
s->ecobus = I915_READ(ECOBUS);
s->pwrdwnupctl = I915_READ(VLV_PWRDWNUPCTL);
s->rp_down_timeout = I915_READ(GEN6_RP_DOWN_TIMEOUT);
s->rp_deucsw = I915_READ(GEN6_RPDEUCSW);
s->rcubmabdtmr = I915_READ(GEN6_RCUBMABDTMR);
s->rcedata = I915_READ(VLV_RCEDATA);
s->spare2gh = I915_READ(VLV_SPAREG2H);
/* Display CZ domain, 0x4400C-0x4402C, 0x4F000-0x4F11F */
s->gt_imr = I915_READ(GTIMR);
s->gt_ier = I915_READ(GTIER);
s->pm_imr = I915_READ(GEN6_PMIMR);
s->pm_ier = I915_READ(GEN6_PMIER);
for (i = 0; i < ARRAY_SIZE(s->gt_scratch); i++)
s->gt_scratch[i] = I915_READ(GEN7_GT_SCRATCH(i));
/* GT SA CZ domain, 0x100000-0x138124 */
s->tilectl = I915_READ(TILECTL);
s->gt_fifoctl = I915_READ(GTFIFOCTL);
s->gtlc_wake_ctrl = I915_READ(VLV_GTLC_WAKE_CTRL);
s->gtlc_survive = I915_READ(VLV_GTLC_SURVIVABILITY_REG);
s->pmwgicz = I915_READ(VLV_PMWGICZ);
/* Gunit-Display CZ domain, 0x182028-0x1821CF */
s->gu_ctl0 = I915_READ(VLV_GU_CTL0);
s->gu_ctl1 = I915_READ(VLV_GU_CTL1);
s->pcbr = I915_READ(VLV_PCBR);
s->clock_gate_dis2 = I915_READ(VLV_GUNIT_CLOCK_GATE2);
/*
* Not saving any of:
* DFT, 0x9800-0x9EC0
* SARB, 0xB000-0xB1FC
* GAC, 0x5208-0x524C, 0x14000-0x14C000
* PCI CFG
*/
}
static void vlv_restore_gunit_s0ix_state(struct drm_i915_private *dev_priv)
{
struct vlv_s0ix_state *s = dev_priv->vlv_s0ix_state;
u32 val;
int i;
if (!s)
return;
/* GAM 0x4000-0x4770 */
I915_WRITE(GEN7_WR_WATERMARK, s->wr_watermark);
I915_WRITE(GEN7_GFX_PRIO_CTRL, s->gfx_prio_ctrl);
I915_WRITE(ARB_MODE, s->arb_mode | (0xffff << 16));
I915_WRITE(GEN7_GFX_PEND_TLB0, s->gfx_pend_tlb0);
I915_WRITE(GEN7_GFX_PEND_TLB1, s->gfx_pend_tlb1);
for (i = 0; i < ARRAY_SIZE(s->lra_limits); i++)
I915_WRITE(GEN7_LRA_LIMITS(i), s->lra_limits[i]);
I915_WRITE(GEN7_MEDIA_MAX_REQ_COUNT, s->media_max_req_count);
I915_WRITE(GEN7_GFX_MAX_REQ_COUNT, s->gfx_max_req_count);
I915_WRITE(RENDER_HWS_PGA_GEN7, s->render_hwsp);
I915_WRITE(GAM_ECOCHK, s->ecochk);
I915_WRITE(BSD_HWS_PGA_GEN7, s->bsd_hwsp);
I915_WRITE(BLT_HWS_PGA_GEN7, s->blt_hwsp);
I915_WRITE(GEN7_TLB_RD_ADDR, s->tlb_rd_addr);
/* MBC 0x9024-0x91D0, 0x8500 */
I915_WRITE(VLV_G3DCTL, s->g3dctl);
I915_WRITE(VLV_GSCKGCTL, s->gsckgctl);
I915_WRITE(GEN6_MBCTL, s->mbctl);
/* GCP 0x9400-0x9424, 0x8100-0x810C */
I915_WRITE(GEN6_UCGCTL1, s->ucgctl1);
I915_WRITE(GEN6_UCGCTL3, s->ucgctl3);
I915_WRITE(GEN6_RCGCTL1, s->rcgctl1);
I915_WRITE(GEN6_RCGCTL2, s->rcgctl2);
I915_WRITE(GEN6_RSTCTL, s->rstctl);
I915_WRITE(GEN7_MISCCPCTL, s->misccpctl);
/* GPM 0xA000-0xAA84, 0x8000-0x80FC */
I915_WRITE(GEN6_GFXPAUSE, s->gfxpause);
I915_WRITE(GEN6_RPDEUHWTC, s->rpdeuhwtc);
I915_WRITE(GEN6_RPDEUC, s->rpdeuc);
I915_WRITE(ECOBUS, s->ecobus);
I915_WRITE(VLV_PWRDWNUPCTL, s->pwrdwnupctl);
I915_WRITE(GEN6_RP_DOWN_TIMEOUT,s->rp_down_timeout);
I915_WRITE(GEN6_RPDEUCSW, s->rp_deucsw);
I915_WRITE(GEN6_RCUBMABDTMR, s->rcubmabdtmr);
I915_WRITE(VLV_RCEDATA, s->rcedata);
I915_WRITE(VLV_SPAREG2H, s->spare2gh);
/* Display CZ domain, 0x4400C-0x4402C, 0x4F000-0x4F11F */
I915_WRITE(GTIMR, s->gt_imr);
I915_WRITE(GTIER, s->gt_ier);
I915_WRITE(GEN6_PMIMR, s->pm_imr);
I915_WRITE(GEN6_PMIER, s->pm_ier);
for (i = 0; i < ARRAY_SIZE(s->gt_scratch); i++)
I915_WRITE(GEN7_GT_SCRATCH(i), s->gt_scratch[i]);
/* GT SA CZ domain, 0x100000-0x138124 */
I915_WRITE(TILECTL, s->tilectl);
I915_WRITE(GTFIFOCTL, s->gt_fifoctl);
/*
* Preserve the GT allow wake and GFX force clock bit, they are not
* be restored, as they are used to control the s0ix suspend/resume
* sequence by the caller.
*/
val = I915_READ(VLV_GTLC_WAKE_CTRL);
val &= VLV_GTLC_ALLOWWAKEREQ;
val |= s->gtlc_wake_ctrl & ~VLV_GTLC_ALLOWWAKEREQ;
I915_WRITE(VLV_GTLC_WAKE_CTRL, val);
val = I915_READ(VLV_GTLC_SURVIVABILITY_REG);
val &= VLV_GFX_CLK_FORCE_ON_BIT;
val |= s->gtlc_survive & ~VLV_GFX_CLK_FORCE_ON_BIT;
I915_WRITE(VLV_GTLC_SURVIVABILITY_REG, val);
I915_WRITE(VLV_PMWGICZ, s->pmwgicz);
/* Gunit-Display CZ domain, 0x182028-0x1821CF */
I915_WRITE(VLV_GU_CTL0, s->gu_ctl0);
I915_WRITE(VLV_GU_CTL1, s->gu_ctl1);
I915_WRITE(VLV_PCBR, s->pcbr);
I915_WRITE(VLV_GUNIT_CLOCK_GATE2, s->clock_gate_dis2);
}
static int vlv_wait_for_pw_status(struct drm_i915_private *i915,
drm/i915: Avoid busy-spinning on VLV_GLTC_PW_STATUS mmio The busy-spin, as the first stage of intel_wait_for_register(), is currently under suspicion for causing: [ 62.034926] NMI watchdog: Watchdog detected hard LOCKUP on cpu 1 [ 62.034928] Modules linked in: i2c_dev i915 intel_gtt drm_kms_helper prime_numbers [ 62.034932] CPU: 1 PID: 183 Comm: kworker/1:2 Not tainted 4.11.0-rc7+ #471 [ 62.034933] Hardware name: / , BIOS PYBSWCEL.86A.0027.2015.0507.1758 05/07/2015 [ 62.034934] Workqueue: pm pm_runtime_work [ 62.034936] task: ffff880275a04ec0 task.stack: ffffc900002d8000 [ 62.034936] RIP: 0010:__intel_wait_for_register_fw+0x77/0x1a0 [i915] [ 62.034937] RSP: 0018:ffffc900002dbc38 EFLAGS: 00000082 [ 62.034939] RAX: ffffc90003530094 RBX: 0000000000130094 RCX: 0000000000000001 [ 62.034940] RDX: 00000000000000a1 RSI: ffff88027fd15e58 RDI: 0000000000000000 [ 62.034941] RBP: ffffc900002dbc78 R08: 0000000000000002 R09: 0000000000000000 [ 62.034942] R10: ffffc900002dbc18 R11: ffff880276429dd0 R12: ffff8802707c0000 [ 62.034943] R13: 00000000000000a0 R14: 0000000000000000 R15: 00000000fffefc10 [ 62.034945] FS: 0000000000000000(0000) GS:ffff88027fd00000(0000) knlGS:0000000000000000 [ 62.034945] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 62.034947] CR2: 00007ffd3cd98ff8 CR3: 0000000274c19000 CR4: 00000000001006e0 [ 62.034947] Call Trace: [ 62.034948] intel_wait_for_register+0x77/0x140 [i915] [ 62.034949] vlv_suspend_complete+0x23/0x5b0 [i915] [ 62.034950] intel_runtime_suspend+0x16c/0x2a0 [i915] [ 62.034950] pci_pm_runtime_suspend+0x50/0x180 [ 62.034951] ? pci_pm_runtime_resume+0xa0/0xa0 [ 62.034952] __rpm_callback+0xc5/0x210 [ 62.034953] rpm_callback+0x1f/0x80 [ 62.034953] ? pci_pm_runtime_resume+0xa0/0xa0 [ 62.034954] rpm_suspend+0x118/0x580 [ 62.034955] pm_runtime_work+0x64/0x90 [ 62.034956] process_one_work+0x1bb/0x3e0 [ 62.034956] worker_thread+0x46/0x4f0 [ 62.034957] ? __schedule+0x18b/0x610 [ 62.034958] kthread+0xff/0x140 [ 62.034958] ? process_one_work+0x3e0/0x3e0 [ 62.034959] ? kthread_create_on_node+ and related hard lockups in CI for byt and bsw. Note this effectively reverts commits 41ce405e6894 and b27366958869 ("drm/i915: Convert wait_for(I915_READ(reg)) to intel_wait_for_register()") v2: Convert bool allow into a u32 mask for clarity and repeat the comment on vlv rc6 timing to justify the 3ms timeout used for the wait (Ville) Fixes: 41ce405e6894 ("drm/i915: Convert wait_for(I915_READ(reg)) to intel_wait_for_register()") Fixes: b27366958869 ("drm/i915: Convert wait_for(I915_READ(reg)) to intel_wait_for_register()") Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=100718 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Cc: Tomi Sarvela <tomi.p.sarvela@intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: http://patchwork.freedesktop.org/patch/msgid/20170421135815.11897-1-chris@chris-wilson.co.uk Tested-by: Tomi Sarvela <tomi.p.sarvela@intel.com>
2017-04-21 20:58:15 +07:00
u32 mask, u32 val)
{
i915_reg_t reg = VLV_GTLC_PW_STATUS;
u32 reg_value;
int ret;
drm/i915: Avoid busy-spinning on VLV_GLTC_PW_STATUS mmio The busy-spin, as the first stage of intel_wait_for_register(), is currently under suspicion for causing: [ 62.034926] NMI watchdog: Watchdog detected hard LOCKUP on cpu 1 [ 62.034928] Modules linked in: i2c_dev i915 intel_gtt drm_kms_helper prime_numbers [ 62.034932] CPU: 1 PID: 183 Comm: kworker/1:2 Not tainted 4.11.0-rc7+ #471 [ 62.034933] Hardware name: / , BIOS PYBSWCEL.86A.0027.2015.0507.1758 05/07/2015 [ 62.034934] Workqueue: pm pm_runtime_work [ 62.034936] task: ffff880275a04ec0 task.stack: ffffc900002d8000 [ 62.034936] RIP: 0010:__intel_wait_for_register_fw+0x77/0x1a0 [i915] [ 62.034937] RSP: 0018:ffffc900002dbc38 EFLAGS: 00000082 [ 62.034939] RAX: ffffc90003530094 RBX: 0000000000130094 RCX: 0000000000000001 [ 62.034940] RDX: 00000000000000a1 RSI: ffff88027fd15e58 RDI: 0000000000000000 [ 62.034941] RBP: ffffc900002dbc78 R08: 0000000000000002 R09: 0000000000000000 [ 62.034942] R10: ffffc900002dbc18 R11: ffff880276429dd0 R12: ffff8802707c0000 [ 62.034943] R13: 00000000000000a0 R14: 0000000000000000 R15: 00000000fffefc10 [ 62.034945] FS: 0000000000000000(0000) GS:ffff88027fd00000(0000) knlGS:0000000000000000 [ 62.034945] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 62.034947] CR2: 00007ffd3cd98ff8 CR3: 0000000274c19000 CR4: 00000000001006e0 [ 62.034947] Call Trace: [ 62.034948] intel_wait_for_register+0x77/0x140 [i915] [ 62.034949] vlv_suspend_complete+0x23/0x5b0 [i915] [ 62.034950] intel_runtime_suspend+0x16c/0x2a0 [i915] [ 62.034950] pci_pm_runtime_suspend+0x50/0x180 [ 62.034951] ? pci_pm_runtime_resume+0xa0/0xa0 [ 62.034952] __rpm_callback+0xc5/0x210 [ 62.034953] rpm_callback+0x1f/0x80 [ 62.034953] ? pci_pm_runtime_resume+0xa0/0xa0 [ 62.034954] rpm_suspend+0x118/0x580 [ 62.034955] pm_runtime_work+0x64/0x90 [ 62.034956] process_one_work+0x1bb/0x3e0 [ 62.034956] worker_thread+0x46/0x4f0 [ 62.034957] ? __schedule+0x18b/0x610 [ 62.034958] kthread+0xff/0x140 [ 62.034958] ? process_one_work+0x3e0/0x3e0 [ 62.034959] ? kthread_create_on_node+ and related hard lockups in CI for byt and bsw. Note this effectively reverts commits 41ce405e6894 and b27366958869 ("drm/i915: Convert wait_for(I915_READ(reg)) to intel_wait_for_register()") v2: Convert bool allow into a u32 mask for clarity and repeat the comment on vlv rc6 timing to justify the 3ms timeout used for the wait (Ville) Fixes: 41ce405e6894 ("drm/i915: Convert wait_for(I915_READ(reg)) to intel_wait_for_register()") Fixes: b27366958869 ("drm/i915: Convert wait_for(I915_READ(reg)) to intel_wait_for_register()") Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=100718 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Cc: Tomi Sarvela <tomi.p.sarvela@intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: http://patchwork.freedesktop.org/patch/msgid/20170421135815.11897-1-chris@chris-wilson.co.uk Tested-by: Tomi Sarvela <tomi.p.sarvela@intel.com>
2017-04-21 20:58:15 +07:00
/* The HW does not like us polling for PW_STATUS frequently, so
* use the sleeping loop rather than risk the busy spin within
* intel_wait_for_register().
*
* Transitioning between RC6 states should be at most 2ms (see
* valleyview_enable_rps) so use a 3ms timeout.
*/
ret = wait_for(((reg_value =
intel_uncore_read_notrace(&i915->uncore, reg)) & mask)
== val, 3);
/* just trace the final value */
trace_i915_reg_rw(false, reg, reg_value, sizeof(reg_value), true);
return ret;
drm/i915: Avoid busy-spinning on VLV_GLTC_PW_STATUS mmio The busy-spin, as the first stage of intel_wait_for_register(), is currently under suspicion for causing: [ 62.034926] NMI watchdog: Watchdog detected hard LOCKUP on cpu 1 [ 62.034928] Modules linked in: i2c_dev i915 intel_gtt drm_kms_helper prime_numbers [ 62.034932] CPU: 1 PID: 183 Comm: kworker/1:2 Not tainted 4.11.0-rc7+ #471 [ 62.034933] Hardware name: / , BIOS PYBSWCEL.86A.0027.2015.0507.1758 05/07/2015 [ 62.034934] Workqueue: pm pm_runtime_work [ 62.034936] task: ffff880275a04ec0 task.stack: ffffc900002d8000 [ 62.034936] RIP: 0010:__intel_wait_for_register_fw+0x77/0x1a0 [i915] [ 62.034937] RSP: 0018:ffffc900002dbc38 EFLAGS: 00000082 [ 62.034939] RAX: ffffc90003530094 RBX: 0000000000130094 RCX: 0000000000000001 [ 62.034940] RDX: 00000000000000a1 RSI: ffff88027fd15e58 RDI: 0000000000000000 [ 62.034941] RBP: ffffc900002dbc78 R08: 0000000000000002 R09: 0000000000000000 [ 62.034942] R10: ffffc900002dbc18 R11: ffff880276429dd0 R12: ffff8802707c0000 [ 62.034943] R13: 00000000000000a0 R14: 0000000000000000 R15: 00000000fffefc10 [ 62.034945] FS: 0000000000000000(0000) GS:ffff88027fd00000(0000) knlGS:0000000000000000 [ 62.034945] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 62.034947] CR2: 00007ffd3cd98ff8 CR3: 0000000274c19000 CR4: 00000000001006e0 [ 62.034947] Call Trace: [ 62.034948] intel_wait_for_register+0x77/0x140 [i915] [ 62.034949] vlv_suspend_complete+0x23/0x5b0 [i915] [ 62.034950] intel_runtime_suspend+0x16c/0x2a0 [i915] [ 62.034950] pci_pm_runtime_suspend+0x50/0x180 [ 62.034951] ? pci_pm_runtime_resume+0xa0/0xa0 [ 62.034952] __rpm_callback+0xc5/0x210 [ 62.034953] rpm_callback+0x1f/0x80 [ 62.034953] ? pci_pm_runtime_resume+0xa0/0xa0 [ 62.034954] rpm_suspend+0x118/0x580 [ 62.034955] pm_runtime_work+0x64/0x90 [ 62.034956] process_one_work+0x1bb/0x3e0 [ 62.034956] worker_thread+0x46/0x4f0 [ 62.034957] ? __schedule+0x18b/0x610 [ 62.034958] kthread+0xff/0x140 [ 62.034958] ? process_one_work+0x3e0/0x3e0 [ 62.034959] ? kthread_create_on_node+ and related hard lockups in CI for byt and bsw. Note this effectively reverts commits 41ce405e6894 and b27366958869 ("drm/i915: Convert wait_for(I915_READ(reg)) to intel_wait_for_register()") v2: Convert bool allow into a u32 mask for clarity and repeat the comment on vlv rc6 timing to justify the 3ms timeout used for the wait (Ville) Fixes: 41ce405e6894 ("drm/i915: Convert wait_for(I915_READ(reg)) to intel_wait_for_register()") Fixes: b27366958869 ("drm/i915: Convert wait_for(I915_READ(reg)) to intel_wait_for_register()") Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=100718 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Cc: Tomi Sarvela <tomi.p.sarvela@intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: http://patchwork.freedesktop.org/patch/msgid/20170421135815.11897-1-chris@chris-wilson.co.uk Tested-by: Tomi Sarvela <tomi.p.sarvela@intel.com>
2017-04-21 20:58:15 +07:00
}
int vlv_force_gfx_clock(struct drm_i915_private *dev_priv, bool force_on)
{
u32 val;
int err;
val = I915_READ(VLV_GTLC_SURVIVABILITY_REG);
val &= ~VLV_GFX_CLK_FORCE_ON_BIT;
if (force_on)
val |= VLV_GFX_CLK_FORCE_ON_BIT;
I915_WRITE(VLV_GTLC_SURVIVABILITY_REG, val);
if (!force_on)
return 0;
err = intel_wait_for_register(&dev_priv->uncore,
VLV_GTLC_SURVIVABILITY_REG,
VLV_GFX_CLK_STATUS_BIT,
VLV_GFX_CLK_STATUS_BIT,
20);
if (err)
DRM_ERROR("timeout waiting for GFX clock force-on (%08x)\n",
I915_READ(VLV_GTLC_SURVIVABILITY_REG));
return err;
}
static int vlv_allow_gt_wake(struct drm_i915_private *dev_priv, bool allow)
{
drm/i915: Avoid busy-spinning on VLV_GLTC_PW_STATUS mmio The busy-spin, as the first stage of intel_wait_for_register(), is currently under suspicion for causing: [ 62.034926] NMI watchdog: Watchdog detected hard LOCKUP on cpu 1 [ 62.034928] Modules linked in: i2c_dev i915 intel_gtt drm_kms_helper prime_numbers [ 62.034932] CPU: 1 PID: 183 Comm: kworker/1:2 Not tainted 4.11.0-rc7+ #471 [ 62.034933] Hardware name: / , BIOS PYBSWCEL.86A.0027.2015.0507.1758 05/07/2015 [ 62.034934] Workqueue: pm pm_runtime_work [ 62.034936] task: ffff880275a04ec0 task.stack: ffffc900002d8000 [ 62.034936] RIP: 0010:__intel_wait_for_register_fw+0x77/0x1a0 [i915] [ 62.034937] RSP: 0018:ffffc900002dbc38 EFLAGS: 00000082 [ 62.034939] RAX: ffffc90003530094 RBX: 0000000000130094 RCX: 0000000000000001 [ 62.034940] RDX: 00000000000000a1 RSI: ffff88027fd15e58 RDI: 0000000000000000 [ 62.034941] RBP: ffffc900002dbc78 R08: 0000000000000002 R09: 0000000000000000 [ 62.034942] R10: ffffc900002dbc18 R11: ffff880276429dd0 R12: ffff8802707c0000 [ 62.034943] R13: 00000000000000a0 R14: 0000000000000000 R15: 00000000fffefc10 [ 62.034945] FS: 0000000000000000(0000) GS:ffff88027fd00000(0000) knlGS:0000000000000000 [ 62.034945] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 62.034947] CR2: 00007ffd3cd98ff8 CR3: 0000000274c19000 CR4: 00000000001006e0 [ 62.034947] Call Trace: [ 62.034948] intel_wait_for_register+0x77/0x140 [i915] [ 62.034949] vlv_suspend_complete+0x23/0x5b0 [i915] [ 62.034950] intel_runtime_suspend+0x16c/0x2a0 [i915] [ 62.034950] pci_pm_runtime_suspend+0x50/0x180 [ 62.034951] ? pci_pm_runtime_resume+0xa0/0xa0 [ 62.034952] __rpm_callback+0xc5/0x210 [ 62.034953] rpm_callback+0x1f/0x80 [ 62.034953] ? pci_pm_runtime_resume+0xa0/0xa0 [ 62.034954] rpm_suspend+0x118/0x580 [ 62.034955] pm_runtime_work+0x64/0x90 [ 62.034956] process_one_work+0x1bb/0x3e0 [ 62.034956] worker_thread+0x46/0x4f0 [ 62.034957] ? __schedule+0x18b/0x610 [ 62.034958] kthread+0xff/0x140 [ 62.034958] ? process_one_work+0x3e0/0x3e0 [ 62.034959] ? kthread_create_on_node+ and related hard lockups in CI for byt and bsw. Note this effectively reverts commits 41ce405e6894 and b27366958869 ("drm/i915: Convert wait_for(I915_READ(reg)) to intel_wait_for_register()") v2: Convert bool allow into a u32 mask for clarity and repeat the comment on vlv rc6 timing to justify the 3ms timeout used for the wait (Ville) Fixes: 41ce405e6894 ("drm/i915: Convert wait_for(I915_READ(reg)) to intel_wait_for_register()") Fixes: b27366958869 ("drm/i915: Convert wait_for(I915_READ(reg)) to intel_wait_for_register()") Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=100718 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Cc: Tomi Sarvela <tomi.p.sarvela@intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: http://patchwork.freedesktop.org/patch/msgid/20170421135815.11897-1-chris@chris-wilson.co.uk Tested-by: Tomi Sarvela <tomi.p.sarvela@intel.com>
2017-04-21 20:58:15 +07:00
u32 mask;
u32 val;
drm/i915: Avoid busy-spinning on VLV_GLTC_PW_STATUS mmio The busy-spin, as the first stage of intel_wait_for_register(), is currently under suspicion for causing: [ 62.034926] NMI watchdog: Watchdog detected hard LOCKUP on cpu 1 [ 62.034928] Modules linked in: i2c_dev i915 intel_gtt drm_kms_helper prime_numbers [ 62.034932] CPU: 1 PID: 183 Comm: kworker/1:2 Not tainted 4.11.0-rc7+ #471 [ 62.034933] Hardware name: / , BIOS PYBSWCEL.86A.0027.2015.0507.1758 05/07/2015 [ 62.034934] Workqueue: pm pm_runtime_work [ 62.034936] task: ffff880275a04ec0 task.stack: ffffc900002d8000 [ 62.034936] RIP: 0010:__intel_wait_for_register_fw+0x77/0x1a0 [i915] [ 62.034937] RSP: 0018:ffffc900002dbc38 EFLAGS: 00000082 [ 62.034939] RAX: ffffc90003530094 RBX: 0000000000130094 RCX: 0000000000000001 [ 62.034940] RDX: 00000000000000a1 RSI: ffff88027fd15e58 RDI: 0000000000000000 [ 62.034941] RBP: ffffc900002dbc78 R08: 0000000000000002 R09: 0000000000000000 [ 62.034942] R10: ffffc900002dbc18 R11: ffff880276429dd0 R12: ffff8802707c0000 [ 62.034943] R13: 00000000000000a0 R14: 0000000000000000 R15: 00000000fffefc10 [ 62.034945] FS: 0000000000000000(0000) GS:ffff88027fd00000(0000) knlGS:0000000000000000 [ 62.034945] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 62.034947] CR2: 00007ffd3cd98ff8 CR3: 0000000274c19000 CR4: 00000000001006e0 [ 62.034947] Call Trace: [ 62.034948] intel_wait_for_register+0x77/0x140 [i915] [ 62.034949] vlv_suspend_complete+0x23/0x5b0 [i915] [ 62.034950] intel_runtime_suspend+0x16c/0x2a0 [i915] [ 62.034950] pci_pm_runtime_suspend+0x50/0x180 [ 62.034951] ? pci_pm_runtime_resume+0xa0/0xa0 [ 62.034952] __rpm_callback+0xc5/0x210 [ 62.034953] rpm_callback+0x1f/0x80 [ 62.034953] ? pci_pm_runtime_resume+0xa0/0xa0 [ 62.034954] rpm_suspend+0x118/0x580 [ 62.034955] pm_runtime_work+0x64/0x90 [ 62.034956] process_one_work+0x1bb/0x3e0 [ 62.034956] worker_thread+0x46/0x4f0 [ 62.034957] ? __schedule+0x18b/0x610 [ 62.034958] kthread+0xff/0x140 [ 62.034958] ? process_one_work+0x3e0/0x3e0 [ 62.034959] ? kthread_create_on_node+ and related hard lockups in CI for byt and bsw. Note this effectively reverts commits 41ce405e6894 and b27366958869 ("drm/i915: Convert wait_for(I915_READ(reg)) to intel_wait_for_register()") v2: Convert bool allow into a u32 mask for clarity and repeat the comment on vlv rc6 timing to justify the 3ms timeout used for the wait (Ville) Fixes: 41ce405e6894 ("drm/i915: Convert wait_for(I915_READ(reg)) to intel_wait_for_register()") Fixes: b27366958869 ("drm/i915: Convert wait_for(I915_READ(reg)) to intel_wait_for_register()") Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=100718 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Cc: Tomi Sarvela <tomi.p.sarvela@intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: http://patchwork.freedesktop.org/patch/msgid/20170421135815.11897-1-chris@chris-wilson.co.uk Tested-by: Tomi Sarvela <tomi.p.sarvela@intel.com>
2017-04-21 20:58:15 +07:00
int err;
val = I915_READ(VLV_GTLC_WAKE_CTRL);
val &= ~VLV_GTLC_ALLOWWAKEREQ;
if (allow)
val |= VLV_GTLC_ALLOWWAKEREQ;
I915_WRITE(VLV_GTLC_WAKE_CTRL, val);
POSTING_READ(VLV_GTLC_WAKE_CTRL);
drm/i915: Avoid busy-spinning on VLV_GLTC_PW_STATUS mmio The busy-spin, as the first stage of intel_wait_for_register(), is currently under suspicion for causing: [ 62.034926] NMI watchdog: Watchdog detected hard LOCKUP on cpu 1 [ 62.034928] Modules linked in: i2c_dev i915 intel_gtt drm_kms_helper prime_numbers [ 62.034932] CPU: 1 PID: 183 Comm: kworker/1:2 Not tainted 4.11.0-rc7+ #471 [ 62.034933] Hardware name: / , BIOS PYBSWCEL.86A.0027.2015.0507.1758 05/07/2015 [ 62.034934] Workqueue: pm pm_runtime_work [ 62.034936] task: ffff880275a04ec0 task.stack: ffffc900002d8000 [ 62.034936] RIP: 0010:__intel_wait_for_register_fw+0x77/0x1a0 [i915] [ 62.034937] RSP: 0018:ffffc900002dbc38 EFLAGS: 00000082 [ 62.034939] RAX: ffffc90003530094 RBX: 0000000000130094 RCX: 0000000000000001 [ 62.034940] RDX: 00000000000000a1 RSI: ffff88027fd15e58 RDI: 0000000000000000 [ 62.034941] RBP: ffffc900002dbc78 R08: 0000000000000002 R09: 0000000000000000 [ 62.034942] R10: ffffc900002dbc18 R11: ffff880276429dd0 R12: ffff8802707c0000 [ 62.034943] R13: 00000000000000a0 R14: 0000000000000000 R15: 00000000fffefc10 [ 62.034945] FS: 0000000000000000(0000) GS:ffff88027fd00000(0000) knlGS:0000000000000000 [ 62.034945] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 62.034947] CR2: 00007ffd3cd98ff8 CR3: 0000000274c19000 CR4: 00000000001006e0 [ 62.034947] Call Trace: [ 62.034948] intel_wait_for_register+0x77/0x140 [i915] [ 62.034949] vlv_suspend_complete+0x23/0x5b0 [i915] [ 62.034950] intel_runtime_suspend+0x16c/0x2a0 [i915] [ 62.034950] pci_pm_runtime_suspend+0x50/0x180 [ 62.034951] ? pci_pm_runtime_resume+0xa0/0xa0 [ 62.034952] __rpm_callback+0xc5/0x210 [ 62.034953] rpm_callback+0x1f/0x80 [ 62.034953] ? pci_pm_runtime_resume+0xa0/0xa0 [ 62.034954] rpm_suspend+0x118/0x580 [ 62.034955] pm_runtime_work+0x64/0x90 [ 62.034956] process_one_work+0x1bb/0x3e0 [ 62.034956] worker_thread+0x46/0x4f0 [ 62.034957] ? __schedule+0x18b/0x610 [ 62.034958] kthread+0xff/0x140 [ 62.034958] ? process_one_work+0x3e0/0x3e0 [ 62.034959] ? kthread_create_on_node+ and related hard lockups in CI for byt and bsw. Note this effectively reverts commits 41ce405e6894 and b27366958869 ("drm/i915: Convert wait_for(I915_READ(reg)) to intel_wait_for_register()") v2: Convert bool allow into a u32 mask for clarity and repeat the comment on vlv rc6 timing to justify the 3ms timeout used for the wait (Ville) Fixes: 41ce405e6894 ("drm/i915: Convert wait_for(I915_READ(reg)) to intel_wait_for_register()") Fixes: b27366958869 ("drm/i915: Convert wait_for(I915_READ(reg)) to intel_wait_for_register()") Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=100718 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Cc: Tomi Sarvela <tomi.p.sarvela@intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: http://patchwork.freedesktop.org/patch/msgid/20170421135815.11897-1-chris@chris-wilson.co.uk Tested-by: Tomi Sarvela <tomi.p.sarvela@intel.com>
2017-04-21 20:58:15 +07:00
mask = VLV_GTLC_ALLOWWAKEACK;
val = allow ? mask : 0;
err = vlv_wait_for_pw_status(dev_priv, mask, val);
if (err)
DRM_ERROR("timeout disabling GT waking\n");
return err;
}
drm/i915: Avoid busy-spinning on VLV_GLTC_PW_STATUS mmio The busy-spin, as the first stage of intel_wait_for_register(), is currently under suspicion for causing: [ 62.034926] NMI watchdog: Watchdog detected hard LOCKUP on cpu 1 [ 62.034928] Modules linked in: i2c_dev i915 intel_gtt drm_kms_helper prime_numbers [ 62.034932] CPU: 1 PID: 183 Comm: kworker/1:2 Not tainted 4.11.0-rc7+ #471 [ 62.034933] Hardware name: / , BIOS PYBSWCEL.86A.0027.2015.0507.1758 05/07/2015 [ 62.034934] Workqueue: pm pm_runtime_work [ 62.034936] task: ffff880275a04ec0 task.stack: ffffc900002d8000 [ 62.034936] RIP: 0010:__intel_wait_for_register_fw+0x77/0x1a0 [i915] [ 62.034937] RSP: 0018:ffffc900002dbc38 EFLAGS: 00000082 [ 62.034939] RAX: ffffc90003530094 RBX: 0000000000130094 RCX: 0000000000000001 [ 62.034940] RDX: 00000000000000a1 RSI: ffff88027fd15e58 RDI: 0000000000000000 [ 62.034941] RBP: ffffc900002dbc78 R08: 0000000000000002 R09: 0000000000000000 [ 62.034942] R10: ffffc900002dbc18 R11: ffff880276429dd0 R12: ffff8802707c0000 [ 62.034943] R13: 00000000000000a0 R14: 0000000000000000 R15: 00000000fffefc10 [ 62.034945] FS: 0000000000000000(0000) GS:ffff88027fd00000(0000) knlGS:0000000000000000 [ 62.034945] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 62.034947] CR2: 00007ffd3cd98ff8 CR3: 0000000274c19000 CR4: 00000000001006e0 [ 62.034947] Call Trace: [ 62.034948] intel_wait_for_register+0x77/0x140 [i915] [ 62.034949] vlv_suspend_complete+0x23/0x5b0 [i915] [ 62.034950] intel_runtime_suspend+0x16c/0x2a0 [i915] [ 62.034950] pci_pm_runtime_suspend+0x50/0x180 [ 62.034951] ? pci_pm_runtime_resume+0xa0/0xa0 [ 62.034952] __rpm_callback+0xc5/0x210 [ 62.034953] rpm_callback+0x1f/0x80 [ 62.034953] ? pci_pm_runtime_resume+0xa0/0xa0 [ 62.034954] rpm_suspend+0x118/0x580 [ 62.034955] pm_runtime_work+0x64/0x90 [ 62.034956] process_one_work+0x1bb/0x3e0 [ 62.034956] worker_thread+0x46/0x4f0 [ 62.034957] ? __schedule+0x18b/0x610 [ 62.034958] kthread+0xff/0x140 [ 62.034958] ? process_one_work+0x3e0/0x3e0 [ 62.034959] ? kthread_create_on_node+ and related hard lockups in CI for byt and bsw. Note this effectively reverts commits 41ce405e6894 and b27366958869 ("drm/i915: Convert wait_for(I915_READ(reg)) to intel_wait_for_register()") v2: Convert bool allow into a u32 mask for clarity and repeat the comment on vlv rc6 timing to justify the 3ms timeout used for the wait (Ville) Fixes: 41ce405e6894 ("drm/i915: Convert wait_for(I915_READ(reg)) to intel_wait_for_register()") Fixes: b27366958869 ("drm/i915: Convert wait_for(I915_READ(reg)) to intel_wait_for_register()") Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=100718 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Cc: Tomi Sarvela <tomi.p.sarvela@intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: http://patchwork.freedesktop.org/patch/msgid/20170421135815.11897-1-chris@chris-wilson.co.uk Tested-by: Tomi Sarvela <tomi.p.sarvela@intel.com>
2017-04-21 20:58:15 +07:00
static void vlv_wait_for_gt_wells(struct drm_i915_private *dev_priv,
bool wait_for_on)
{
u32 mask;
u32 val;
mask = VLV_GTLC_PW_MEDIA_STATUS_MASK | VLV_GTLC_PW_RENDER_STATUS_MASK;
val = wait_for_on ? mask : 0;
/*
* RC6 transitioning can be delayed up to 2 msec (see
* valleyview_enable_rps), use 3 msec for safety.
*
* This can fail to turn off the rc6 if the GPU is stuck after a failed
* reset and we are trying to force the machine to sleep.
*/
drm/i915: Avoid busy-spinning on VLV_GLTC_PW_STATUS mmio The busy-spin, as the first stage of intel_wait_for_register(), is currently under suspicion for causing: [ 62.034926] NMI watchdog: Watchdog detected hard LOCKUP on cpu 1 [ 62.034928] Modules linked in: i2c_dev i915 intel_gtt drm_kms_helper prime_numbers [ 62.034932] CPU: 1 PID: 183 Comm: kworker/1:2 Not tainted 4.11.0-rc7+ #471 [ 62.034933] Hardware name: / , BIOS PYBSWCEL.86A.0027.2015.0507.1758 05/07/2015 [ 62.034934] Workqueue: pm pm_runtime_work [ 62.034936] task: ffff880275a04ec0 task.stack: ffffc900002d8000 [ 62.034936] RIP: 0010:__intel_wait_for_register_fw+0x77/0x1a0 [i915] [ 62.034937] RSP: 0018:ffffc900002dbc38 EFLAGS: 00000082 [ 62.034939] RAX: ffffc90003530094 RBX: 0000000000130094 RCX: 0000000000000001 [ 62.034940] RDX: 00000000000000a1 RSI: ffff88027fd15e58 RDI: 0000000000000000 [ 62.034941] RBP: ffffc900002dbc78 R08: 0000000000000002 R09: 0000000000000000 [ 62.034942] R10: ffffc900002dbc18 R11: ffff880276429dd0 R12: ffff8802707c0000 [ 62.034943] R13: 00000000000000a0 R14: 0000000000000000 R15: 00000000fffefc10 [ 62.034945] FS: 0000000000000000(0000) GS:ffff88027fd00000(0000) knlGS:0000000000000000 [ 62.034945] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 62.034947] CR2: 00007ffd3cd98ff8 CR3: 0000000274c19000 CR4: 00000000001006e0 [ 62.034947] Call Trace: [ 62.034948] intel_wait_for_register+0x77/0x140 [i915] [ 62.034949] vlv_suspend_complete+0x23/0x5b0 [i915] [ 62.034950] intel_runtime_suspend+0x16c/0x2a0 [i915] [ 62.034950] pci_pm_runtime_suspend+0x50/0x180 [ 62.034951] ? pci_pm_runtime_resume+0xa0/0xa0 [ 62.034952] __rpm_callback+0xc5/0x210 [ 62.034953] rpm_callback+0x1f/0x80 [ 62.034953] ? pci_pm_runtime_resume+0xa0/0xa0 [ 62.034954] rpm_suspend+0x118/0x580 [ 62.034955] pm_runtime_work+0x64/0x90 [ 62.034956] process_one_work+0x1bb/0x3e0 [ 62.034956] worker_thread+0x46/0x4f0 [ 62.034957] ? __schedule+0x18b/0x610 [ 62.034958] kthread+0xff/0x140 [ 62.034958] ? process_one_work+0x3e0/0x3e0 [ 62.034959] ? kthread_create_on_node+ and related hard lockups in CI for byt and bsw. Note this effectively reverts commits 41ce405e6894 and b27366958869 ("drm/i915: Convert wait_for(I915_READ(reg)) to intel_wait_for_register()") v2: Convert bool allow into a u32 mask for clarity and repeat the comment on vlv rc6 timing to justify the 3ms timeout used for the wait (Ville) Fixes: 41ce405e6894 ("drm/i915: Convert wait_for(I915_READ(reg)) to intel_wait_for_register()") Fixes: b27366958869 ("drm/i915: Convert wait_for(I915_READ(reg)) to intel_wait_for_register()") Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=100718 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Cc: Tomi Sarvela <tomi.p.sarvela@intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: http://patchwork.freedesktop.org/patch/msgid/20170421135815.11897-1-chris@chris-wilson.co.uk Tested-by: Tomi Sarvela <tomi.p.sarvela@intel.com>
2017-04-21 20:58:15 +07:00
if (vlv_wait_for_pw_status(dev_priv, mask, val))
DRM_DEBUG_DRIVER("timeout waiting for GT wells to go %s\n",
onoff(wait_for_on));
}
static void vlv_check_no_gt_access(struct drm_i915_private *dev_priv)
{
if (!(I915_READ(VLV_GTLC_PW_STATUS) & VLV_GTLC_ALLOWWAKEERR))
return;
DRM_DEBUG_DRIVER("GT register access while GT waking disabled\n");
I915_WRITE(VLV_GTLC_PW_STATUS, VLV_GTLC_ALLOWWAKEERR);
}
static int vlv_suspend_complete(struct drm_i915_private *dev_priv)
{
u32 mask;
int err;
/*
* Bspec defines the following GT well on flags as debug only, so
* don't treat them as hard failures.
*/
drm/i915: Avoid busy-spinning on VLV_GLTC_PW_STATUS mmio The busy-spin, as the first stage of intel_wait_for_register(), is currently under suspicion for causing: [ 62.034926] NMI watchdog: Watchdog detected hard LOCKUP on cpu 1 [ 62.034928] Modules linked in: i2c_dev i915 intel_gtt drm_kms_helper prime_numbers [ 62.034932] CPU: 1 PID: 183 Comm: kworker/1:2 Not tainted 4.11.0-rc7+ #471 [ 62.034933] Hardware name: / , BIOS PYBSWCEL.86A.0027.2015.0507.1758 05/07/2015 [ 62.034934] Workqueue: pm pm_runtime_work [ 62.034936] task: ffff880275a04ec0 task.stack: ffffc900002d8000 [ 62.034936] RIP: 0010:__intel_wait_for_register_fw+0x77/0x1a0 [i915] [ 62.034937] RSP: 0018:ffffc900002dbc38 EFLAGS: 00000082 [ 62.034939] RAX: ffffc90003530094 RBX: 0000000000130094 RCX: 0000000000000001 [ 62.034940] RDX: 00000000000000a1 RSI: ffff88027fd15e58 RDI: 0000000000000000 [ 62.034941] RBP: ffffc900002dbc78 R08: 0000000000000002 R09: 0000000000000000 [ 62.034942] R10: ffffc900002dbc18 R11: ffff880276429dd0 R12: ffff8802707c0000 [ 62.034943] R13: 00000000000000a0 R14: 0000000000000000 R15: 00000000fffefc10 [ 62.034945] FS: 0000000000000000(0000) GS:ffff88027fd00000(0000) knlGS:0000000000000000 [ 62.034945] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 62.034947] CR2: 00007ffd3cd98ff8 CR3: 0000000274c19000 CR4: 00000000001006e0 [ 62.034947] Call Trace: [ 62.034948] intel_wait_for_register+0x77/0x140 [i915] [ 62.034949] vlv_suspend_complete+0x23/0x5b0 [i915] [ 62.034950] intel_runtime_suspend+0x16c/0x2a0 [i915] [ 62.034950] pci_pm_runtime_suspend+0x50/0x180 [ 62.034951] ? pci_pm_runtime_resume+0xa0/0xa0 [ 62.034952] __rpm_callback+0xc5/0x210 [ 62.034953] rpm_callback+0x1f/0x80 [ 62.034953] ? pci_pm_runtime_resume+0xa0/0xa0 [ 62.034954] rpm_suspend+0x118/0x580 [ 62.034955] pm_runtime_work+0x64/0x90 [ 62.034956] process_one_work+0x1bb/0x3e0 [ 62.034956] worker_thread+0x46/0x4f0 [ 62.034957] ? __schedule+0x18b/0x610 [ 62.034958] kthread+0xff/0x140 [ 62.034958] ? process_one_work+0x3e0/0x3e0 [ 62.034959] ? kthread_create_on_node+ and related hard lockups in CI for byt and bsw. Note this effectively reverts commits 41ce405e6894 and b27366958869 ("drm/i915: Convert wait_for(I915_READ(reg)) to intel_wait_for_register()") v2: Convert bool allow into a u32 mask for clarity and repeat the comment on vlv rc6 timing to justify the 3ms timeout used for the wait (Ville) Fixes: 41ce405e6894 ("drm/i915: Convert wait_for(I915_READ(reg)) to intel_wait_for_register()") Fixes: b27366958869 ("drm/i915: Convert wait_for(I915_READ(reg)) to intel_wait_for_register()") Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=100718 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com> Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Cc: Tomi Sarvela <tomi.p.sarvela@intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: http://patchwork.freedesktop.org/patch/msgid/20170421135815.11897-1-chris@chris-wilson.co.uk Tested-by: Tomi Sarvela <tomi.p.sarvela@intel.com>
2017-04-21 20:58:15 +07:00
vlv_wait_for_gt_wells(dev_priv, false);
mask = VLV_GTLC_RENDER_CTX_EXISTS | VLV_GTLC_MEDIA_CTX_EXISTS;
WARN_ON((I915_READ(VLV_GTLC_WAKE_CTRL) & mask) != mask);
vlv_check_no_gt_access(dev_priv);
err = vlv_force_gfx_clock(dev_priv, true);
if (err)
goto err1;
err = vlv_allow_gt_wake(dev_priv, false);
if (err)
goto err2;
vlv_save_gunit_s0ix_state(dev_priv);
err = vlv_force_gfx_clock(dev_priv, false);
if (err)
goto err2;
return 0;
err2:
/* For safety always re-enable waking and disable gfx clock forcing */
vlv_allow_gt_wake(dev_priv, true);
err1:
vlv_force_gfx_clock(dev_priv, false);
return err;
}
drm/i915: Sharing platform specific sequence between runtime and system suspend/ resume paths On VLV, post S0i3 during i915_drm_thaw following issue is observed during ring initialization. [ 335.604039] [drm:stop_ring] ERROR render ring :timed out trying to stop ring [ 336.607340] [drm:stop_ring] ERROR render ring :timed out trying to stop ring [ 336.607345] [drm:init_ring_common] ERROR failed to set render ring head to zero ctl 00000000 head 00000000 tail 00000000 start 00000000 [ 337.610645] [drm:stop_ring] ERROR bsd ring :timed out trying to stop ring [ 338.613952] [drm:stop_ring] ERROR bsd ring :timed out trying to stop ring [ 338.613956] [drm:init_ring_common] ERROR failed to set bsd ring head to zero ctl 00000000 head 00000000 tail 00000000 start 00000000 [ 339.617256] [drm:stop_ring] ERROR render ring :timed out trying to stop ring [ 339.617258] -----------[ cut here ]----------- [ 339.617267] WARNING: CPU: 0 PID: 6 at drivers/gpu/drm/i915/intel_ringbuffer.c:1666 intel_cleanup_ring+0xe6/0xf0() [ 339.617396] --[ end trace 5ef5ed1a3c92e2a6 ]-- [ 339.617428] [drm:__i915_drm_thaw] ERROR failed to re-initialize GPU, declaring wedged! This is happening since wake is not enabled and Gunit registers are not restored. For this system suspend/resume paths need to follow save/restore and additional platform specific setup in suspend_complete and resume_prepare. suspend_complete is shared unconditionaly for VLV, HSW, BDW. resume_prepare for HSW and BDW has pc8 disabling which is needed during thaw_early so sharing uncondtionally. For VLV and SNB runtime resume specific sequence exists. Cc: Imre Deak <imre.deak@intel.com> Cc: Paulo Zanoni <paulo.r.zanoni@intel.com> Cc: Daniel Vetter <daniel.vetter@ffwll.ch> Cc: Jani Nikula <jani.nikula@linux.intel.com> Cc: Goel, Akash <akash.goel@intel.com> Signed-off-by: Sagar Kamble <sagar.a.kamble@intel.com> Reviewed-by: Imre Deak <imre.deak@intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-08-14 00:37:06 +07:00
static int vlv_resume_prepare(struct drm_i915_private *dev_priv,
bool rpm_resume)
{
int err;
int ret;
/*
* If any of the steps fail just try to continue, that's the best we
* can do at this point. Return the first error code (which will also
* leave RPM permanently disabled).
*/
ret = vlv_force_gfx_clock(dev_priv, true);
vlv_restore_gunit_s0ix_state(dev_priv);
err = vlv_allow_gt_wake(dev_priv, true);
if (!ret)
ret = err;
err = vlv_force_gfx_clock(dev_priv, false);
if (!ret)
ret = err;
vlv_check_no_gt_access(dev_priv);
if (rpm_resume)
intel_init_clock_gating(dev_priv);
return ret;
}
static int intel_runtime_suspend(struct device *kdev)
{
struct drm_i915_private *dev_priv = kdev_to_i915(kdev);
struct intel_runtime_pm *rpm = &dev_priv->runtime_pm;
int ret = 0;
if (WARN_ON_ONCE(!HAS_RUNTIME_PM(dev_priv)))
return -ENODEV;
DRM_DEBUG_KMS("Suspending device\n");
disable_rpm_wakeref_asserts(rpm);
drm/i915: add support for checking if we hold an RPM reference Atm, we assert that the device is not suspended until the point when the device is truly put to a suspended state. This is fine, but we can catch more problems if we check that RPM refcount is non-zero. After that one drops to zero we shouldn't access the device any more, even if the actual device suspend may be delayed. Change assert_rpm_wakelock_held() accordingly to check for a non-zero RPM refcount in addition to the current device-not-suspended check. For the new asserts to work we need to annotate every place explicitly in the code where we expect that the device is powered. The places where we only assume this, but may not hold an RPM reference: - driver load We assume the device to be powered until we enable RPM. Make this explicit by taking an RPM reference around the load function. - system and runtime sudpend/resume handlers These handlers are called when the RPM reference becomes 0 and know the exact point after which the device can get powered off. Disable the RPM-reference-held check for their duration. - the IRQ, hangcheck and RPS work handlers These handlers are flushed in the system/runtime suspend handler before the device is powered off, so it's guaranteed that they won't run while the device is powered off even though they don't hold any RPM reference. Disable the RPM-reference-held check for their duration. In all these cases we still check that the device is not suspended. These explicit annotations also have the positive side effect of documenting our assumptions better. This caught additional WARNs from the atomic modeset path, those should be fixed separately. v2: - remove the redundant HAS_RUNTIME_PM check (moved to patch 1) (Ville) v3: - use a new dedicated RPM wakelock refcount to also catch cases where our own RPM get/put functions were not called (Chris) - assert also that the new RPM wakelock refcount is 0 in the RPM suspend handler (Chris) - change the assert error message to be more meaningful (Chris) - prevent false assert errors and check that the RPM wakelock is 0 in the RPM resume handler too - prevent false assert errors in the hangcheck work too - add a device not suspended assert check to the hangcheck work v4: - rename disable/enable_rpm_asserts to disable/enable_rpm_wakeref_asserts and wakelock_count to wakeref_count - disable the wakeref asserts in the IRQ handlers and RPS work too - update/clarify commit message v5: - mark places we plan to change to use proper RPM refcounting with separate DISABLE/ENABLE_RPM_WAKEREF_ASSERTS aliases (Chris) Signed-off-by: Imre Deak <imre.deak@intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Link: http://patchwork.freedesktop.org/patch/msgid/1450227139-13471-1-git-send-email-imre.deak@intel.com
2015-12-16 07:52:19 +07:00
/*
* We are safe here against re-faults, since the fault handler takes
* an RPM reference.
*/
i915_gem_runtime_suspend(dev_priv);
intel_gt_runtime_suspend(&dev_priv->gt);
intel_runtime_pm_disable_interrupts(dev_priv);
intel_uncore_suspend(&dev_priv->uncore);
intel_display_power_suspend(dev_priv);
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_uncore_runtime_resume(&dev_priv->uncore);
intel_runtime_pm_enable_interrupts(dev_priv);
intel_gt_runtime_resume(&dev_priv->gt);
i915_gem_restore_fences(&dev_priv->ggtt);
enable_rpm_wakeref_asserts(rpm);
drm/i915: add support for checking if we hold an RPM reference Atm, we assert that the device is not suspended until the point when the device is truly put to a suspended state. This is fine, but we can catch more problems if we check that RPM refcount is non-zero. After that one drops to zero we shouldn't access the device any more, even if the actual device suspend may be delayed. Change assert_rpm_wakelock_held() accordingly to check for a non-zero RPM refcount in addition to the current device-not-suspended check. For the new asserts to work we need to annotate every place explicitly in the code where we expect that the device is powered. The places where we only assume this, but may not hold an RPM reference: - driver load We assume the device to be powered until we enable RPM. Make this explicit by taking an RPM reference around the load function. - system and runtime sudpend/resume handlers These handlers are called when the RPM reference becomes 0 and know the exact point after which the device can get powered off. Disable the RPM-reference-held check for their duration. - the IRQ, hangcheck and RPS work handlers These handlers are flushed in the system/runtime suspend handler before the device is powered off, so it's guaranteed that they won't run while the device is powered off even though they don't hold any RPM reference. Disable the RPM-reference-held check for their duration. In all these cases we still check that the device is not suspended. These explicit annotations also have the positive side effect of documenting our assumptions better. This caught additional WARNs from the atomic modeset path, those should be fixed separately. v2: - remove the redundant HAS_RUNTIME_PM check (moved to patch 1) (Ville) v3: - use a new dedicated RPM wakelock refcount to also catch cases where our own RPM get/put functions were not called (Chris) - assert also that the new RPM wakelock refcount is 0 in the RPM suspend handler (Chris) - change the assert error message to be more meaningful (Chris) - prevent false assert errors and check that the RPM wakelock is 0 in the RPM resume handler too - prevent false assert errors in the hangcheck work too - add a device not suspended assert check to the hangcheck work v4: - rename disable/enable_rpm_asserts to disable/enable_rpm_wakeref_asserts and wakelock_count to wakeref_count - disable the wakeref asserts in the IRQ handlers and RPS work too - update/clarify commit message v5: - mark places we plan to change to use proper RPM refcounting with separate DISABLE/ENABLE_RPM_WAKEREF_ASSERTS aliases (Chris) Signed-off-by: Imre Deak <imre.deak@intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Link: http://patchwork.freedesktop.org/patch/msgid/1450227139-13471-1-git-send-email-imre.deak@intel.com
2015-12-16 07:52:19 +07:00
return ret;
}
drm/i915: make PC8 be part of runtime PM suspend/resume Currently, when our driver becomes idle for i915.pc8_timeout (default: 5s) we enable PC8, so we save some power, but not everything we can. Then, while PC8 is enabled, if we stay idle for more autosuspend_delay_ms (default: 10s) we'll enter runtime PM and put the graphics device in D3 state, saving even more power. The two features are separate things with increasing levels of power savings, but if we disable PC8 we'll never get into D3. While from the modularity point of view it would be nice to keep these features as separate, we have reasons to merge them: - We are not aware of anybody wanting a "PC8 without D3" environment. - If we keep both features as separate, we'll have to to test both PC8 and PC8+D3 code paths. We're already having a major pain to make QA do automated testing of just one thing, testing both paths will cost even more. - Only Haswell+ supports PC8, so if we want to add runtime PM support to, for example, IVB, we'll have to copy some code from the PC8 feature to runtime PM, so merging both features as a single thing will make it easier for enabling runtime PM on other platforms. This patch only does the very basic steps required to have PC8 and runtime PM merged on a single feature: the next patches will take care of cleaning up everything. v2: - Rebase. v3: - Rebase. - Fully remove the deprecated i915 params since Daniel doesn't consider them as part of the ABI. v4: - Rebase. - Fix typo in the commit message. v5: - Rebase, again. - Add a huge comment explaining the different forcewake usage (Chris, Daniel). - Use open-coded forcewake functions (Daniel). Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com> Reviewed-by: Imre Deak <imre.deak@intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-03-08 06:08:05 +07:00
enable_rpm_wakeref_asserts(rpm);
intel_runtime_pm_driver_release(rpm);
if (intel_uncore_arm_unclaimed_mmio_detection(&dev_priv->uncore))
DRM_ERROR("Unclaimed access detected prior to suspending\n");
rpm->suspended = true;
/*
* FIXME: We really should find a document that references the arguments
* used below!
*/
if (IS_BROADWELL(dev_priv)) {
/*
* On Broadwell, if we use PCI_D1 the PCH DDI ports will stop
* being detected, and the call we do at intel_runtime_resume()
* won't be able to restore them. Since PCI_D3hot matches the
* actual specification and appears to be working, use it.
*/
intel_opregion_notify_adapter(dev_priv, PCI_D3hot);
} else {
/*
* current versions of firmware which depend on this opregion
* notification have repurposed the D1 definition to mean
* "runtime suspended" vs. what you would normally expect (D3)
* to distinguish it from notifications that might be sent via
* the suspend path.
*/
intel_opregion_notify_adapter(dev_priv, PCI_D1);
}
assert_forcewakes_inactive(&dev_priv->uncore);
if (!IS_VALLEYVIEW(dev_priv) && !IS_CHERRYVIEW(dev_priv))
drm/i915: Enable polling when we don't have hpd Unfortunately, there's two situations where we lose hpd right now: - Runtime suspend - When we've shut off all of the power wells on Valleyview/Cherryview While it would be nice if this didn't cause issues, this has the ability to get us in some awkward states where a user won't be able to get their display to turn on. For instance; if we boot a Valleyview system without any monitors connected, it won't need any of it's power wells and thus shut them off. Since this causes us to lose HPD, this means that unless the user knows how to ssh into their machine and do a manual reprobe for monitors, none of the monitors they connect after booting will actually work. Eventually we should come up with a better fix then having to enable polling for this, since this makes rpm a lot less useful, but for now the infrastructure in i915 just isn't there yet to get hpd in these situations. Changes since v1: - Add comment explaining the addition of the if (!mode_config->poll_running) in intel_hpd_init() - Remove unneeded if (!dev->mode_config.poll_enabled) in i915_hpd_poll_init_work() - Call to drm_helper_hpd_irq_event() after we disable polling - Add cancel_work_sync() call to intel_hpd_cancel_work() Changes since v2: - Apparently dev->mode_config.poll_running doesn't actually reflect whether or not a poll is currently in progress, and is actually used for dynamic module paramter enabling/disabling. So now we instead keep track of our own poll_running variable in dev_priv->hotplug - Clean i915_hpd_poll_init_work() a little bit Changes since v3: - Remove the now-redundant connector loop in intel_hpd_init(), just rely on intel_hpd_poll_enable() for setting connector->polled correctly on each connector - Get rid of poll_running - Don't assign enabled in i915_hpd_poll_init_work before we actually lock dev->mode_config.mutex - Wrap enabled assignment in i915_hpd_poll_init_work() in READ_ONCE() for doc purposes - Do the same for dev_priv->hotplug.poll_enabled with WRITE_ONCE in intel_hpd_poll_enable() - Add some comments about racing not mattering in intel_hpd_poll_enable Changes since v4: - Rename intel_hpd_poll_enable() to intel_hpd_poll_init() - Drop the bool argument from intel_hpd_poll_init() - Remove redundant calls to intel_hpd_poll_init() - Rename poll_enable_work to poll_init_work - Add some kerneldoc for intel_hpd_poll_init() - Cross-reference intel_hpd_poll_init() in intel_hpd_init() - Just copy the loop from intel_hpd_init() in intel_hpd_poll_init() Changes since v5: - Minor kerneldoc nitpicks Cc: stable@vger.kernel.org Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch> Signed-off-by: Lyude <cpaul@redhat.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2016-06-22 04:03:44 +07:00
intel_hpd_poll_init(dev_priv);
drm/i915: make PC8 be part of runtime PM suspend/resume Currently, when our driver becomes idle for i915.pc8_timeout (default: 5s) we enable PC8, so we save some power, but not everything we can. Then, while PC8 is enabled, if we stay idle for more autosuspend_delay_ms (default: 10s) we'll enter runtime PM and put the graphics device in D3 state, saving even more power. The two features are separate things with increasing levels of power savings, but if we disable PC8 we'll never get into D3. While from the modularity point of view it would be nice to keep these features as separate, we have reasons to merge them: - We are not aware of anybody wanting a "PC8 without D3" environment. - If we keep both features as separate, we'll have to to test both PC8 and PC8+D3 code paths. We're already having a major pain to make QA do automated testing of just one thing, testing both paths will cost even more. - Only Haswell+ supports PC8, so if we want to add runtime PM support to, for example, IVB, we'll have to copy some code from the PC8 feature to runtime PM, so merging both features as a single thing will make it easier for enabling runtime PM on other platforms. This patch only does the very basic steps required to have PC8 and runtime PM merged on a single feature: the next patches will take care of cleaning up everything. v2: - Rebase. v3: - Rebase. - Fully remove the deprecated i915 params since Daniel doesn't consider them as part of the ABI. v4: - Rebase. - Fix typo in the commit message. v5: - Rebase, again. - Add a huge comment explaining the different forcewake usage (Chris, Daniel). - Use open-coded forcewake functions (Daniel). Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com> Reviewed-by: Imre Deak <imre.deak@intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2014-03-08 06:08:05 +07:00
DRM_DEBUG_KMS("Device suspended\n");
return 0;
}
static int intel_runtime_resume(struct device *kdev)
{
struct drm_i915_private *dev_priv = kdev_to_i915(kdev);
struct intel_runtime_pm *rpm = &dev_priv->runtime_pm;
int ret = 0;
if (WARN_ON_ONCE(!HAS_RUNTIME_PM(dev_priv)))
return -ENODEV;
DRM_DEBUG_KMS("Resuming device\n");
WARN_ON_ONCE(atomic_read(&rpm->wakeref_count));
disable_rpm_wakeref_asserts(rpm);
drm/i915: add support for checking if we hold an RPM reference Atm, we assert that the device is not suspended until the point when the device is truly put to a suspended state. This is fine, but we can catch more problems if we check that RPM refcount is non-zero. After that one drops to zero we shouldn't access the device any more, even if the actual device suspend may be delayed. Change assert_rpm_wakelock_held() accordingly to check for a non-zero RPM refcount in addition to the current device-not-suspended check. For the new asserts to work we need to annotate every place explicitly in the code where we expect that the device is powered. The places where we only assume this, but may not hold an RPM reference: - driver load We assume the device to be powered until we enable RPM. Make this explicit by taking an RPM reference around the load function. - system and runtime sudpend/resume handlers These handlers are called when the RPM reference becomes 0 and know the exact point after which the device can get powered off. Disable the RPM-reference-held check for their duration. - the IRQ, hangcheck and RPS work handlers These handlers are flushed in the system/runtime suspend handler before the device is powered off, so it's guaranteed that they won't run while the device is powered off even though they don't hold any RPM reference. Disable the RPM-reference-held check for their duration. In all these cases we still check that the device is not suspended. These explicit annotations also have the positive side effect of documenting our assumptions better. This caught additional WARNs from the atomic modeset path, those should be fixed separately. v2: - remove the redundant HAS_RUNTIME_PM check (moved to patch 1) (Ville) v3: - use a new dedicated RPM wakelock refcount to also catch cases where our own RPM get/put functions were not called (Chris) - assert also that the new RPM wakelock refcount is 0 in the RPM suspend handler (Chris) - change the assert error message to be more meaningful (Chris) - prevent false assert errors and check that the RPM wakelock is 0 in the RPM resume handler too - prevent false assert errors in the hangcheck work too - add a device not suspended assert check to the hangcheck work v4: - rename disable/enable_rpm_asserts to disable/enable_rpm_wakeref_asserts and wakelock_count to wakeref_count - disable the wakeref asserts in the IRQ handlers and RPS work too - update/clarify commit message v5: - mark places we plan to change to use proper RPM refcounting with separate DISABLE/ENABLE_RPM_WAKEREF_ASSERTS aliases (Chris) Signed-off-by: Imre Deak <imre.deak@intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Link: http://patchwork.freedesktop.org/patch/msgid/1450227139-13471-1-git-send-email-imre.deak@intel.com
2015-12-16 07:52:19 +07:00
intel_opregion_notify_adapter(dev_priv, PCI_D0);
rpm->suspended = false;
if (intel_uncore_unclaimed_mmio(&dev_priv->uncore))
DRM_DEBUG_DRIVER("Unclaimed access during suspend, bios?\n");
intel_display_power_resume(dev_priv);
if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv))
ret = vlv_resume_prepare(dev_priv, true);
intel_uncore_runtime_resume(&dev_priv->uncore);
intel_runtime_pm_enable_interrupts(dev_priv);
/*
* No point of rolling back things in case of an error, as the best
* we can do is to hope that things will still work (and disable RPM).
*/
intel_gt_runtime_resume(&dev_priv->gt);
i915_gem_restore_fences(&dev_priv->ggtt);
/*
* On VLV/CHV display interrupts are part of the display
* power well, so hpd is reinitialized from there. For
* everyone else do it here.
*/
if (!IS_VALLEYVIEW(dev_priv) && !IS_CHERRYVIEW(dev_priv))
intel_hpd_init(dev_priv);
2017-08-17 20:45:28 +07:00
intel_enable_ipc(dev_priv);
enable_rpm_wakeref_asserts(rpm);
drm/i915: add support for checking if we hold an RPM reference Atm, we assert that the device is not suspended until the point when the device is truly put to a suspended state. This is fine, but we can catch more problems if we check that RPM refcount is non-zero. After that one drops to zero we shouldn't access the device any more, even if the actual device suspend may be delayed. Change assert_rpm_wakelock_held() accordingly to check for a non-zero RPM refcount in addition to the current device-not-suspended check. For the new asserts to work we need to annotate every place explicitly in the code where we expect that the device is powered. The places where we only assume this, but may not hold an RPM reference: - driver load We assume the device to be powered until we enable RPM. Make this explicit by taking an RPM reference around the load function. - system and runtime sudpend/resume handlers These handlers are called when the RPM reference becomes 0 and know the exact point after which the device can get powered off. Disable the RPM-reference-held check for their duration. - the IRQ, hangcheck and RPS work handlers These handlers are flushed in the system/runtime suspend handler before the device is powered off, so it's guaranteed that they won't run while the device is powered off even though they don't hold any RPM reference. Disable the RPM-reference-held check for their duration. In all these cases we still check that the device is not suspended. These explicit annotations also have the positive side effect of documenting our assumptions better. This caught additional WARNs from the atomic modeset path, those should be fixed separately. v2: - remove the redundant HAS_RUNTIME_PM check (moved to patch 1) (Ville) v3: - use a new dedicated RPM wakelock refcount to also catch cases where our own RPM get/put functions were not called (Chris) - assert also that the new RPM wakelock refcount is 0 in the RPM suspend handler (Chris) - change the assert error message to be more meaningful (Chris) - prevent false assert errors and check that the RPM wakelock is 0 in the RPM resume handler too - prevent false assert errors in the hangcheck work too - add a device not suspended assert check to the hangcheck work v4: - rename disable/enable_rpm_asserts to disable/enable_rpm_wakeref_asserts and wakelock_count to wakeref_count - disable the wakeref asserts in the IRQ handlers and RPS work too - update/clarify commit message v5: - mark places we plan to change to use proper RPM refcounting with separate DISABLE/ENABLE_RPM_WAKEREF_ASSERTS aliases (Chris) Signed-off-by: Imre Deak <imre.deak@intel.com> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Link: http://patchwork.freedesktop.org/patch/msgid/1450227139-13471-1-git-send-email-imre.deak@intel.com
2015-12-16 07:52:19 +07:00
if (ret)
DRM_ERROR("Runtime resume failed, disabling it (%d)\n", ret);
else
DRM_DEBUG_KMS("Device resumed\n");
return ret;
}
const struct dev_pm_ops i915_pm_ops = {
/*
* S0ix (via system suspend) and S3 event handlers [PMSG_SUSPEND,
* PMSG_RESUME]
*/
.prepare = i915_pm_prepare,
.suspend = i915_pm_suspend,
.suspend_late = i915_pm_suspend_late,
.resume_early = i915_pm_resume_early,
.resume = i915_pm_resume,
/*
* S4 event handlers
* @freeze, @freeze_late : called (1) before creating the
* hibernation image [PMSG_FREEZE] and
* (2) after rebooting, before restoring
* the image [PMSG_QUIESCE]
* @thaw, @thaw_early : called (1) after creating the hibernation
* image, before writing it [PMSG_THAW]
* and (2) after failing to create or
* restore the image [PMSG_RECOVER]
* @poweroff, @poweroff_late: called after writing the hibernation
* image, before rebooting [PMSG_HIBERNATE]
* @restore, @restore_early : called after rebooting and restoring the
* hibernation image [PMSG_RESTORE]
*/
.freeze = i915_pm_freeze,
.freeze_late = i915_pm_freeze_late,
.thaw_early = i915_pm_thaw_early,
.thaw = i915_pm_thaw,
.poweroff = i915_pm_suspend,
2015-03-02 18:04:41 +07:00
.poweroff_late = i915_pm_poweroff_late,
.restore_early = i915_pm_restore_early,
.restore = i915_pm_restore,
/* S0ix (via runtime suspend) event handlers */
.runtime_suspend = intel_runtime_suspend,
.runtime_resume = intel_runtime_resume,
};
static const struct vm_operations_struct i915_gem_vm_ops = {
.fault = i915_gem_fault,
.open = drm_gem_vm_open,
.close = drm_gem_vm_close,
};
static const struct file_operations i915_driver_fops = {
.owner = THIS_MODULE,
.open = drm_open,
.release = drm_release,
.unlocked_ioctl = drm_ioctl,
.mmap = drm_gem_mmap,
.poll = drm_poll,
.read = drm_read,
.compat_ioctl = i915_compat_ioctl,
.llseek = noop_llseek,
};
static int
i915_gem_reject_pin_ioctl(struct drm_device *dev, void *data,
struct drm_file *file)
{
return -ENODEV;
}
static const struct drm_ioctl_desc i915_ioctls[] = {
DRM_IOCTL_DEF_DRV(I915_INIT, drm_noop, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
DRM_IOCTL_DEF_DRV(I915_FLUSH, drm_noop, DRM_AUTH),
DRM_IOCTL_DEF_DRV(I915_FLIP, drm_noop, DRM_AUTH),
DRM_IOCTL_DEF_DRV(I915_BATCHBUFFER, drm_noop, DRM_AUTH),
DRM_IOCTL_DEF_DRV(I915_IRQ_EMIT, drm_noop, DRM_AUTH),
DRM_IOCTL_DEF_DRV(I915_IRQ_WAIT, drm_noop, DRM_AUTH),
DRM_IOCTL_DEF_DRV(I915_GETPARAM, i915_getparam_ioctl, DRM_RENDER_ALLOW),
DRM_IOCTL_DEF_DRV(I915_SETPARAM, drm_noop, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
DRM_IOCTL_DEF_DRV(I915_ALLOC, drm_noop, DRM_AUTH),
DRM_IOCTL_DEF_DRV(I915_FREE, drm_noop, DRM_AUTH),
DRM_IOCTL_DEF_DRV(I915_INIT_HEAP, drm_noop, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
DRM_IOCTL_DEF_DRV(I915_CMDBUFFER, drm_noop, DRM_AUTH),
DRM_IOCTL_DEF_DRV(I915_DESTROY_HEAP, drm_noop, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
DRM_IOCTL_DEF_DRV(I915_SET_VBLANK_PIPE, drm_noop, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
DRM_IOCTL_DEF_DRV(I915_GET_VBLANK_PIPE, drm_noop, DRM_AUTH),
DRM_IOCTL_DEF_DRV(I915_VBLANK_SWAP, drm_noop, DRM_AUTH),
DRM_IOCTL_DEF_DRV(I915_HWS_ADDR, drm_noop, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
DRM_IOCTL_DEF_DRV(I915_GEM_INIT, drm_noop, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
DRM_IOCTL_DEF_DRV(I915_GEM_EXECBUFFER, i915_gem_execbuffer_ioctl, DRM_AUTH),
DRM_IOCTL_DEF_DRV(I915_GEM_EXECBUFFER2_WR, i915_gem_execbuffer2_ioctl, DRM_RENDER_ALLOW),
DRM_IOCTL_DEF_DRV(I915_GEM_PIN, i915_gem_reject_pin_ioctl, DRM_AUTH|DRM_ROOT_ONLY),
DRM_IOCTL_DEF_DRV(I915_GEM_UNPIN, i915_gem_reject_pin_ioctl, DRM_AUTH|DRM_ROOT_ONLY),
DRM_IOCTL_DEF_DRV(I915_GEM_BUSY, i915_gem_busy_ioctl, DRM_RENDER_ALLOW),
DRM_IOCTL_DEF_DRV(I915_GEM_SET_CACHING, i915_gem_set_caching_ioctl, DRM_RENDER_ALLOW),
DRM_IOCTL_DEF_DRV(I915_GEM_GET_CACHING, i915_gem_get_caching_ioctl, DRM_RENDER_ALLOW),
DRM_IOCTL_DEF_DRV(I915_GEM_THROTTLE, i915_gem_throttle_ioctl, DRM_RENDER_ALLOW),
DRM_IOCTL_DEF_DRV(I915_GEM_ENTERVT, drm_noop, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
DRM_IOCTL_DEF_DRV(I915_GEM_LEAVEVT, drm_noop, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
DRM_IOCTL_DEF_DRV(I915_GEM_CREATE, i915_gem_create_ioctl, DRM_RENDER_ALLOW),
DRM_IOCTL_DEF_DRV(I915_GEM_PREAD, i915_gem_pread_ioctl, DRM_RENDER_ALLOW),
DRM_IOCTL_DEF_DRV(I915_GEM_PWRITE, i915_gem_pwrite_ioctl, DRM_RENDER_ALLOW),
DRM_IOCTL_DEF_DRV(I915_GEM_MMAP, i915_gem_mmap_ioctl, DRM_RENDER_ALLOW),
DRM_IOCTL_DEF_DRV(I915_GEM_MMAP_GTT, i915_gem_mmap_gtt_ioctl, DRM_RENDER_ALLOW),
DRM_IOCTL_DEF_DRV(I915_GEM_SET_DOMAIN, i915_gem_set_domain_ioctl, DRM_RENDER_ALLOW),
DRM_IOCTL_DEF_DRV(I915_GEM_SW_FINISH, i915_gem_sw_finish_ioctl, DRM_RENDER_ALLOW),
DRM_IOCTL_DEF_DRV(I915_GEM_SET_TILING, i915_gem_set_tiling_ioctl, DRM_RENDER_ALLOW),
DRM_IOCTL_DEF_DRV(I915_GEM_GET_TILING, i915_gem_get_tiling_ioctl, DRM_RENDER_ALLOW),
DRM_IOCTL_DEF_DRV(I915_GEM_GET_APERTURE, i915_gem_get_aperture_ioctl, DRM_RENDER_ALLOW),
DRM_IOCTL_DEF_DRV(I915_GET_PIPE_FROM_CRTC_ID, intel_get_pipe_from_crtc_id_ioctl, 0),
DRM_IOCTL_DEF_DRV(I915_GEM_MADVISE, i915_gem_madvise_ioctl, DRM_RENDER_ALLOW),
DRM_IOCTL_DEF_DRV(I915_OVERLAY_PUT_IMAGE, intel_overlay_put_image_ioctl, DRM_MASTER),
DRM_IOCTL_DEF_DRV(I915_OVERLAY_ATTRS, intel_overlay_attrs_ioctl, DRM_MASTER),
DRM_IOCTL_DEF_DRV(I915_SET_SPRITE_COLORKEY, intel_sprite_set_colorkey_ioctl, DRM_MASTER),
DRM_IOCTL_DEF_DRV(I915_GET_SPRITE_COLORKEY, drm_noop, DRM_MASTER),
DRM_IOCTL_DEF_DRV(I915_GEM_WAIT, i915_gem_wait_ioctl, DRM_RENDER_ALLOW),
DRM_IOCTL_DEF_DRV(I915_GEM_CONTEXT_CREATE_EXT, i915_gem_context_create_ioctl, DRM_RENDER_ALLOW),
DRM_IOCTL_DEF_DRV(I915_GEM_CONTEXT_DESTROY, i915_gem_context_destroy_ioctl, DRM_RENDER_ALLOW),
DRM_IOCTL_DEF_DRV(I915_REG_READ, i915_reg_read_ioctl, DRM_RENDER_ALLOW),
DRM_IOCTL_DEF_DRV(I915_GET_RESET_STATS, i915_gem_context_reset_stats_ioctl, DRM_RENDER_ALLOW),
DRM_IOCTL_DEF_DRV(I915_GEM_USERPTR, i915_gem_userptr_ioctl, DRM_RENDER_ALLOW),
DRM_IOCTL_DEF_DRV(I915_GEM_CONTEXT_GETPARAM, i915_gem_context_getparam_ioctl, DRM_RENDER_ALLOW),
DRM_IOCTL_DEF_DRV(I915_GEM_CONTEXT_SETPARAM, i915_gem_context_setparam_ioctl, DRM_RENDER_ALLOW),
drm/i915: Add i915 perf infrastructure Adds base i915 perf infrastructure for Gen performance metrics. This adds a DRM_IOCTL_I915_PERF_OPEN ioctl that takes an array of uint64 properties to configure a stream of metrics and returns a new fd usable with standard VFS system calls including read() to read typed and sized records; ioctl() to enable or disable capture and poll() to wait for data. A stream is opened something like: uint64_t properties[] = { /* Single context sampling */ DRM_I915_PERF_PROP_CTX_HANDLE, ctx_handle, /* Include OA reports in samples */ DRM_I915_PERF_PROP_SAMPLE_OA, true, /* OA unit configuration */ DRM_I915_PERF_PROP_OA_METRICS_SET, metrics_set_id, DRM_I915_PERF_PROP_OA_FORMAT, report_format, DRM_I915_PERF_PROP_OA_EXPONENT, period_exponent, }; struct drm_i915_perf_open_param parm = { .flags = I915_PERF_FLAG_FD_CLOEXEC | I915_PERF_FLAG_FD_NONBLOCK | I915_PERF_FLAG_DISABLED, .properties_ptr = (uint64_t)properties, .num_properties = sizeof(properties) / 16, }; int fd = drmIoctl(drm_fd, DRM_IOCTL_I915_PERF_OPEN, &param); Records read all start with a common { type, size } header with DRM_I915_PERF_RECORD_SAMPLE being of most interest. Sample records contain an extensible number of fields and it's the DRM_I915_PERF_PROP_SAMPLE_xyz properties given when opening that determine what's included in every sample. No specific streams are supported yet so any attempt to open a stream will return an error. v2: use i915_gem_context_get() - Chris Wilson v3: update read() interface to avoid passing state struct - Chris Wilson fix some rebase fallout, with i915-perf init/deinit v4: s/DRM_IORW/DRM_IOW/ - Emil Velikov Signed-off-by: Robert Bragg <robert@sixbynine.org> Reviewed-by: Matthew Auld <matthew.auld@intel.com> Reviewed-by: Sourab Gupta <sourab.gupta@intel.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Link: http://patchwork.freedesktop.org/patch/msgid/20161107194957.3385-2-robert@sixbynine.org
2016-11-08 02:49:47 +07:00
DRM_IOCTL_DEF_DRV(I915_PERF_OPEN, i915_perf_open_ioctl, DRM_RENDER_ALLOW),
DRM_IOCTL_DEF_DRV(I915_PERF_ADD_CONFIG, i915_perf_add_config_ioctl, DRM_RENDER_ALLOW),
DRM_IOCTL_DEF_DRV(I915_PERF_REMOVE_CONFIG, i915_perf_remove_config_ioctl, DRM_RENDER_ALLOW),
DRM_IOCTL_DEF_DRV(I915_QUERY, i915_query_ioctl, DRM_RENDER_ALLOW),
DRM_IOCTL_DEF_DRV(I915_GEM_VM_CREATE, i915_gem_vm_create_ioctl, DRM_RENDER_ALLOW),
DRM_IOCTL_DEF_DRV(I915_GEM_VM_DESTROY, i915_gem_vm_destroy_ioctl, DRM_RENDER_ALLOW),
};
static struct drm_driver driver = {
/* Don't use MTRRs here; the Xserver or userspace app should
* deal with them for Intel hardware.
*/
.driver_features =
drm/prime: Actually remove DRIVER_PRIME everywhere Split out to make the functional changes stick out more. All places where DRIVER_PRIME was used have been removed in previous patches already. v2: amdgpu gained DRIVER_SYNCOBJ_TIMELINE. v3: amdgpu lost DRIVER_SYNCOBJ_TIMELINE. v4: Don't add a space in i915_drv.c (Sam) v5: Add note that previous patches removed all the DRIVER_PRIME users already (Emil). v6: Fixupe ingenic (new driver) while applying. Cc: Sam Ravnborg <sam@ravnborg.org> Reviewed-by: Emil Velikov <emil.velikov@collabora.com> Reviewed-by: Eric Anholt <eric@anholt.net> Signed-off-by: Daniel Vetter <daniel.vetter@intel.com> Cc: amd-gfx@lists.freedesktop.org Cc: etnaviv@lists.freedesktop.org Cc: freedreno@lists.freedesktop.org Cc: intel-gfx@lists.freedesktop.org Cc: lima@lists.freedesktop.org Cc: linux-amlogic@lists.infradead.org Cc: linux-arm-kernel@lists.infradead.org Cc: linux-arm-msm@vger.kernel.org Cc: linux-aspeed@lists.ozlabs.org Cc: linux-renesas-soc@vger.kernel.org Cc: linux-rockchip@lists.infradead.org Cc: linux-samsung-soc@vger.kernel.org Cc: linux-stm32@st-md-mailman.stormreply.com Cc: linux-tegra@vger.kernel.org Cc: nouveau@lists.freedesktop.org Cc: NXP Linux Team <linux-imx@nxp.com> Cc: spice-devel@lists.freedesktop.org Cc: virtualization@lists.linux-foundation.org Cc: VMware Graphics <linux-graphics-maintainer@vmware.com> Cc: xen-devel@lists.xenproject.org Link: https://patchwork.freedesktop.org/patch/msgid/20190617153924.414-1-daniel.vetter@ffwll.ch
2019-06-17 22:39:24 +07:00
DRIVER_GEM |
DRIVER_RENDER | DRIVER_MODESET | DRIVER_ATOMIC | DRIVER_SYNCOBJ,
.release = i915_driver_release,
.open = i915_driver_open,
.lastclose = i915_driver_lastclose,
.postclose = i915_driver_postclose,
.gem_close_object = i915_gem_close_object,
.gem_free_object_unlocked = i915_gem_free_object,
.gem_vm_ops = &i915_gem_vm_ops,
i915: add dmabuf/prime buffer sharing support. This adds handle->fd and fd->handle support to i915, this is to allow for offloading of rendering in one direction and outputs in the other. v2 from Daniel Vetter: - fixup conflicts with the prepare/finish gtt prep work. - implement ppgtt binding support. Note that we have squat i-g-t testcoverage for any of the lifetime and access rules dma_buf/prime support brings along. And there are quite a few intricate situations here. Also note that the integration with the existing code is a bit hackish, especially around get_gtt_pages and put_gtt_pages. It imo would be easier with the prep code from Chris Wilson's unbound series, but that is for 3.6. Also note that I didn't bother to put the new prepare/finish gtt hooks to good use by moving the dma_buf_map/unmap_attachment calls in there (like we've originally planned for). Last but not least this patch is only compile-tested, but I've changed very little compared to Dave Airlie's version. So there's a decent chance v2 on drm-next works as well as v1 on 3.4-rc. v3: Right when I've hit sent I've noticed that I've screwed up one obj->sg_list (for dmar support) and obj->sg_table (for prime support) disdinction. We should be able to merge these 2 paths, but that's material for another patch. v4: fix the error reporting bugs pointed out by ickle. v5: fix another error, and stop non-gtt mmaps on shared objects stop pread/pwrite on imported objects, add fake kmap Signed-off-by: Dave Airlie <airlied@redhat.com> Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
2012-05-10 20:25:09 +07:00
.prime_handle_to_fd = drm_gem_prime_handle_to_fd,
.prime_fd_to_handle = drm_gem_prime_fd_to_handle,
.gem_prime_export = i915_gem_prime_export,
.gem_prime_import = i915_gem_prime_import,
.get_vblank_timestamp = drm_calc_vbltimestamp_from_scanoutpos,
.get_scanout_position = i915_get_crtc_scanoutpos,
.dumb_create = i915_gem_dumb_create,
.dumb_map_offset = i915_gem_mmap_gtt,
.ioctls = i915_ioctls,
.num_ioctls = ARRAY_SIZE(i915_ioctls),
.fops = &i915_driver_fops,
.name = DRIVER_NAME,
.desc = DRIVER_DESC,
.date = DRIVER_DATE,
.major = DRIVER_MAJOR,
.minor = DRIVER_MINOR,
.patchlevel = DRIVER_PATCHLEVEL,
};