linux/drivers/gpu/drm/i915
Daniel Vetter 9cca306880 drm/i915: Handle inaccurate time conversion issues
So apparently jiffies<->nsec<->ktime isn't accurate or something. At
elast if we timeout there's occasionally still a few hundred us left
(in a 2 second timeout).

Stuff I've tried and thrown out again:
- Sampling the before timestamp before jiffies. Doesn't improve test
  path rate at all.
- Using jiffies. Way to inaccurate, which means way too much drift
  with signals plus automatic ioctl restarting in userspace. In
  hindsight we should have used an absolute timeout, but hey we need
  something for v3 of the i915 gem wait interfaces ;-)
- Trying to figure out where accuracy gets lost. gl testcase really
  don't care all that much about this (as long as isn't not massively
  off), it's just that the testcase gets a bit upset if it receives an
  EITME with timeout > 0.

So as long as we're in the ballbark it's good enough. So patch
everything up if we're at most one jiffies off. I get's me a solid
test again.

This regression is probably introduced in

commit 5ed0bdf21a
Author: Thomas Gleixner <tglx@linutronix.de>
Date:   Wed Jul 16 21:05:06 2014 +0000

    drm: i915: Use nsec based interfaces

    Use ktime_get_raw_ns() and get rid of the back and forth timespec
    conversions.

    Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
    Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
    Signed-off-by: John Stultz <john.stultz@linaro.org>

Probably because I'm too lazy to confirm myself and still waiting for
QA ;-)

Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: John Stultz <john.stultz@linaro.org>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=82749
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
2014-12-05 15:24:45 +02:00
..
dvo_ch7xxx.c
dvo_ch7017.c
dvo_ivch.c
dvo_ns2501.c drm/i915: Check pixel clock in ns2501 mode_valid hook 2014-09-03 11:05:21 +02:00
dvo_sil164.c
dvo_tfp410.c
dvo.h
i915_cmd_parser.c drm/i915: Add the predicate source registers to the register whitelist 2014-11-14 10:29:24 +01:00
i915_debugfs.c drm/i915: Handle runtime pm in the CRC setup code 2014-12-03 09:29:40 +01:00
i915_dma.c Linux 3.18-rc7 2014-12-02 10:58:33 +10:00
i915_drv.c drm/i915: Grab modeset locks for GPU rest on pre-ctg 2014-12-03 09:29:38 +01:00
i915_drv.h drm/i915: compute wait_ioctl timeout correctly 2014-12-05 15:20:24 +02:00
i915_gem_context.c drm/i915: Initialize workarounds in logical ring mode too 2014-11-14 10:29:25 +01:00
i915_gem_debug.c
i915_gem_dmabuf.c
i915_gem_evict.c drm/i915: fix another use-after-free in i915_gem_evict_everything 2014-09-19 14:41:16 +02:00
i915_gem_execbuffer.c Merge tag 'drm-intel-next-2014-11-21-fixed' of git://anongit.freedesktop.org/drm-intel into drm-next 2014-12-03 08:25:59 +10:00
i915_gem_gtt.c Linux 3.18-rc7 2014-12-02 10:58:33 +10:00
i915_gem_gtt.h drm/i915: Can i915_gem_init_ioctl 2014-11-20 13:03:31 +01:00
i915_gem_render_state.c drm/i915 Add golden context support for Gen9 2014-11-04 14:04:55 +01:00
i915_gem_render_state.h drm/i915/bdw: Render state init for Execlists 2014-09-03 11:04:52 +02:00
i915_gem_stolen.c drm/i915: Move flags describing VMA mappings into the VMA 2014-11-04 14:04:51 +01:00
i915_gem_tiling.c Linux 3.18-rc7 2014-12-02 10:58:33 +10:00
i915_gem_userptr.c drm/i915: Do not leak pages when freeing userptr objects 2014-09-29 15:31:01 +02:00
i915_gem.c drm/i915: Handle inaccurate time conversion issues 2014-12-05 15:24:45 +02:00
i915_gpu_error.c drm/i915: Use ggtt error obj capture helper for gen8 semaphores 2014-11-20 16:59:49 +01:00
i915_ioc32.c drm/i915: remove redundant #ifdef CONFIG_COMPAT 2014-10-24 16:34:07 +02:00
i915_irq.c drm/i915: mask RPS IRQs properly when disabling RPS 2014-12-03 09:31:53 +01:00
i915_params.c Merge branch 'drm-next' of git://people.freedesktop.org/~airlied/linux 2014-10-14 09:39:08 +02:00
i915_reg.h drm/i915: Implement GPU reset for 915/945 2014-12-03 09:29:37 +01:00
i915_suspend.c drm/i915: unify remaining register save/restore code a bit 2014-11-14 10:29:28 +01:00
i915_sysfs.c drm/i915: Do not export RC6p and RC6pp if they don't exist 2014-10-24 16:34:00 +02:00
i915_trace_points.c
i915_trace.h drm/i915: Add tracepoints to track a vm during its lifetime 2014-11-14 10:29:13 +01:00
i915_ums.c drm/i915: don't save/restore backlight hist ctl registers 2014-11-14 10:29:28 +01:00
intel_acpi.c
intel_audio.c drm/i915/audio: fix monitor presence indication after disable 2014-11-18 11:17:37 +01:00
intel_bios.c Merge branch 'drm-next' of git://people.freedesktop.org/~airlied/linux 2014-10-14 09:39:08 +02:00
intel_bios.h drm/i915/bios: add missing __packed to structs used for reading vbt 2014-09-19 14:43:14 +02:00
intel_crt.c drm/i915: Don't claim that we're resetting PCH ADPA register 2014-10-24 16:34:10 +02:00
intel_ddi.c drm/i915: Don't rely upon encoder->type for infoframe hw state readout 2014-11-20 22:37:04 +01:00
intel_display.c drm/i915: don't always do full mode sets when infoframes are enabled 2014-12-05 15:03:46 +02:00
intel_dp_mst.c drm: Miscellaneous fixes for v3.19-rc1 2014-11-15 09:37:20 +10:00
intel_dp.c drm/i915/eDP: When enabling panel VDD cancel pending disable worker 2014-12-03 09:29:40 +01:00
intel_drv.h drm/i915: Grab modeset locks for GPU rest on pre-ctg 2014-12-03 09:29:38 +01:00
intel_dsi_cmd.c drm/i915: Align intel_dsi*.c files a bit 2014-08-08 17:43:45 +02:00
intel_dsi_cmd.h drm/i915: wait for all DSI FIFOs to be empty 2014-08-07 11:07:15 +02:00
intel_dsi_panel_vbt.c drm/i915: Add support for Video Burst Mode for MIPI DSI 2014-08-08 17:43:45 +02:00
intel_dsi_pll.c drm/i915: Align intel_dsi*.c files a bit 2014-08-08 17:43:45 +02:00
intel_dsi.c drm/i915: Bikeshed rpm functions name a bit. 2014-10-01 10:52:59 +02:00
intel_dsi.h drm/i915: Add support for Video Burst Mode for MIPI DSI 2014-08-08 17:43:45 +02:00
intel_dvo.c drm/i915: Don't call DVO mode_set hook on DPMS changes 2014-09-03 11:05:14 +02:00
intel_fbdev.c drm/i915: Make intel_pin_and_fence_fb_obj take plane and framebuffer 2014-11-07 18:41:51 +01:00
intel_fifo_underrun.c drm/i915: kerneldoc for intel_fifo_underrun.c 2014-10-24 16:33:55 +02:00
intel_frontbuffer.c drm/i915: Introduce intel_psr.c 2014-11-17 19:12:28 +01:00
intel_hdmi.c drm/i915/chv: Enable AVI, SPD and HDMI infoframes for CHV. 2014-12-03 09:29:34 +01:00
intel_i2c.c
intel_lrc.c drm/i915/bdw: Pin the ringbuffer backing object to GGTT on-demand 2014-11-19 19:56:44 +01:00
intel_lrc.h drm/i915/bdw: Pin the context backing objects to GGTT on-demand 2014-11-19 19:32:58 +01:00
intel_lvds.c drm/i915: Pass the current pipe from eDP init to backlight setup 2014-11-14 10:29:20 +01:00
intel_modes.c
intel_opregion.c ACPI / i915: Update the condition to ignore firmware backlight change request 2014-09-30 01:11:18 +02:00
intel_overlay.c
intel_panel.c Linux 3.18-rc7 2014-12-02 10:58:33 +10:00
intel_pm.c Linux 3.18-rc7 2014-12-02 10:58:33 +10:00
intel_psr.c drm/i915: Add PSR docbook 2014-11-17 19:13:24 +01:00
intel_renderstate_gen6.c
intel_renderstate_gen7.c
intel_renderstate_gen8.c drm/i915 Update Gen8 golden context batch buffer 2014-11-04 14:04:54 +01:00
intel_renderstate_gen9.c drm/i915 Add golden context support for Gen9 2014-11-04 14:04:55 +01:00
intel_renderstate.h drm/i915 Add golden context support for Gen9 2014-11-04 14:04:55 +01:00
intel_ringbuffer.c drm/i915: Remove DRI1 ring accessors and API 2014-11-19 21:17:11 +01:00
intel_ringbuffer.h drm/i915: Remove DRI1 ring accessors and API 2014-11-19 21:17:11 +01:00
intel_runtime_pm.c drm/i915: Reinit display irqs and hpd from chv pipe-a power well 2014-11-17 09:16:53 +01:00
intel_sdvo_regs.h
intel_sdvo.c drm/i915: Don't destroy DRM properties in the driver 2014-10-31 18:38:17 +01:00
intel_sideband.c
intel_sprite.c drm/i915: use the correct obj when preparing the sprite plane 2014-11-14 10:29:13 +01:00
intel_tv.c drm/i915: Clarify irq_lock locking, intel_tv_detect 2014-09-19 14:43:19 +02:00
intel_uncore.c drm/i915/skl: Update in Gen9 multi-engine forcewake range 2014-12-03 09:29:41 +01:00
Kconfig
Makefile drm/i915: Introduce intel_psr.c 2014-11-17 19:12:28 +01:00