Tags

drm fixes for 5.4-rc4

dma-resv:
- shared fences for lima/panfrost

ttm:
- prefault regression fix
- lifetime fix

panfrost:
- stopped job timeout fix
- missing register values

amdgpu:
- smu7 powerplay fix
- bail earlier for cik/si detection
- navi SDMA fix

radeon:
- revert a ppc64 shutdown fix that broke x86

i915:
- VBT information handling fix
- Circular locking fix
- preemption vs resubmission virtual requests fix

-dma-resv: Change shared_count to post-increment to fix lima crash (Qiang)
-ttm: A couple fixes related to lifetime and restore prefault behavior
 (Christian & Thomas)
-panfrost: Fill in missing feature reg values and fix stoppedjob timeouts
 (Steven)

Cc: Qiang Yu <yuq825@gmail.com>
Cc: Thomas Hellstrom <thellstrom@vmware.com>
Cc: Christian König <christian.koenig@amd.com>
Cc: Steven Price <steven.price@arm.com>

drm-misc-next for 5.5:

UAPI Changes:
-omap:
    -Add OMAP_BO_MEM_* flags to specify how to allocate BO (Tomi)
    -Reorder OMAP_BO_* #defines, no functional change (Tomi)
-Change unsupported error code from EINVAL to EOPNOTSUPP for: (Rodrigo)
    -drm_wait_vblank_ioctl
    -drm_crtc_get_sequence_ioctl
    -drm_crtc_queue_sequence_ioctl

Cross-subsystem Changes:
-None

Core Changes:
-Delete drmP.h \o/ (Sam)
-kerneldoc clarifications on zpos collisions and plane rects (Simon & Maarten)
-dp_helpers: Add link training repeater definitions added in DP 1.4 (Rodrigo)
-TODO: Add item to convert fbdev drivers to drm (Thomas)
-prime: Add mmap to drm_gem_object_funcs giving more control than vm_ops (Gerd)
-shmem/ttm/vram: Use new mmap gem_object callback (Gerd)

Driver Changes:
-malidp: Add display QoS configuration via devicetree (Wen)
-vkms: Add prime import support (Oleg)
-panfrost: Properly handle job timeouts when cancelling them (Steven)
trockchip/meson/sun4i(via dw-hdmi): Add Dynamic Range and Mastering infoframe					    support (Jonas)
-mxsfb: Add bridge support to accommodate dsi outputs (Robert)
-vboxvideo: Drop hand-rolled implementations and use fbdev emulation,
	    dirtyfb and drm_framebuffer struct from core/core helpers (Thomas)
-komeda: Add D71-specific line sizes and respect connector color fmt (Lowry)
-lima: Use shmem and reservation lock helpers from gem (Qiang)
-rockchip: Add gamma LUT support on vop crtcs (Ezequiel)
-omap:
  -Use refcount_t instead of rolling custom refcounting (Jean-Jacques)

Cc: Wen He <wen.he_1@nxp.com>
Cc: Sam Ravnborg <sam@ravnborg.org>
Cc: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Cc: Oleg Vasilev <omrigann@gmail.com>
Cc: Steven Price <steven.price@arm.com>
Cc: Jonas Karlman <jonas@kwiboo.se>
Cc: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Cc: Simon Ser <contact@emersion.fr>
Cc: Robert Chiras <robert.chiras@nxp.com>
Cc: Thomas Zimmermann <tzimmermann@suse.de>
Cc: Lowry Li <Lowry.Li@arm.com>
Cc: Gerd Hoffmann <kraxel@redhat.com>
Cc: Qiang Yu <yuq825@gmail.com>
Cc: Tomi Valkeinen <tomi.valkeinen@ti.com>
Cc: Ezequiel Garcia <ezequiel@collabora.com>
Cc: Jean-Jacques Hiblot <jjhiblot@ti.com>

- Display fix on handling VBT information.
- Important circular locking fix
- Fix for preemption vs resubmission on virtual requests
  - and a prep patch to make this last one to apply cleanly

Linux 5.4-rc3

drm fixes for 5.4-rc3

i915:
- execlist access fixes
- list deletion fix
- CML display fix
- HSW workaround extension to GT2
- chicken bit whitelist
- GGTT resume issue
- SKL GPU hangs for Vulkan compute

amdgpu:
- memory leak fix

panel:
- spi aliases

tc358767:
- bridge artifacts fix.

- Fix CML display by adding a missing ID.
- Drop redundant list_del_init
- Only enqueue already completed requests to avoid races
- Fixup preempt-to-busy vs reset of a virtual request
- Protect peeking at execlists->active
- execlists->active is serialised by the tasklet

drm-intel-next-fixes-2019-09-19:
- Extend old HSW workaround to fix some GPU hangs on Haswell GT2
- Fix return error code on GEM mmap.
- White list a chicken bit register for push constants legacy mode on Mesa
- Fix resume issue related to GGTT restore
- Remove incorrect BUG_ON on execlist's schedule-out
- Fix unrecoverable GPU hangs with Vulkan compute workloads on SKL

drm-intel-next-fixes-2019-09-26:
- Fix concurrence on cases where requests where getting retired at same time as resubmitted to HW
- Fix gen9 display resolutions by setting the right max plane width
- Fix GPU hang on preemption
- Mark contents as dirty on a write fault. This was breaking cursor sprite with dumb buffers.

Short summary of fixes pull (less than what git shortlog provides):
- SPI Aliases fixes for panels
- One fix for the tc358767 bridge dealing with visual artifacts

drm-misc-next for 5.5:

UAPI Changes:
-Colorspace: Expose different prop values for DP vs. HDMI (Gwan-gyeong Mun)
-fourcc: Add DRM_FORMAT_MOD_ARM_16X16_BLOCK_U_INTERLEAVED (Raymond)
-not_actually: s/ENOTSUPP/EOPNOTSUPP/ in drm_edid and drm_mipi_dbi. This should
    not reach userspace, but adding here to specifically call that out (Daniel)
-i810: Prevent underflow in dispatch ioctls (Dan)
-komeda: Add ACLK sysfs attribute (Mihail)
-v3d: Allow userspace to clean up after render jobs (Iago)

Cross-subsystem Changes:
-MAINTAINERS:
 -Add Alyssa & Steven as panfrost reviewers (Rob)
 -Add Jernej as DE2 reviewer (Maxime)
 -Add Chen-Yu as Allwinner maintainer (Maxime)
-staging: Make some stack arrays static const (Colin)

Core Changes:
-ttm: Allow drivers to specify their vma manager (to use gem mgr) (Gerd)
-docs: Various fixes in connector/encoder/bridge docs (Daniel, Lyude, Laurent)
-connector: Allow more than 3 possible encoders for a connector (José)
-dp_cec: Allow a connector to be associated with a cec device (Dariusz)
-various: Fix some compile/sparse warnings (Ville)
-mm: Ensure mm node removals are properly serialised (Chris)
-panel: Specify the type of panel for drm_panels for later use (Laurent)
-panel: Use drm_panel_init to init device and funcs (Laurent)
-mst: Refactors and cleanups in anticipation of suspend/resume support (Lyude)
-vram:
 -Add lazy unmapping for gem bo's (Thomas)
 -Unify and rationalize vram mm and gem vram (Thomas)
 -Expose vmap and vunmap for gem vram objects (Thomas)
 -Allow objects to be pinned at the top of vram to avoid fragmentation (Thomas)

Driver Changes:
-various: Include drm_bridge.h instead of relying on drm_crtc.h (Boris)
-ast/mgag200: Refactor show_cursor(), move cursor to top of video mem (Thomas)
-komeda:
 -Add error event printing (behind CONFIG) and reg dump support (Lowry)
 -Add suspend/resume support (Lowry)
 -Workaround D71 shadow registers not flushing on disable (Lowry)
-meson: Add suspend/resume support (Neil)
-omap: Miscellaneous refactors and improvements (Tomi/Jyri)
-panfrost/shmem: Silence lockdep by using mutex_trylock (Rob)
-panfrost: Miscellaneous small fixes (Rob/Steven)
-sti: Fix warnings (Benjamin/Linus)
-sun4i:
 -Add vcc-dsi regulator to sun6i_mipi_dsi (Jagan)
 -A few patches to figure out the DRQ/start delay calc on dsi (Jagan/Icenowy)
-virtio:
 -Add module param to switch resource reuse workaround on/off (Gerd)
 -Avoid calling vmexit while holding spinlock (Gerd)
 -Use gem shmem helpers instead of ttm (Gerd)
 -Accommodate command buffer allocations too big for cma (David)

Cc: Rob Herring <robh@kernel.org>
Cc: Maxime Ripard <mripard@kernel.org>
Cc: Gwan-gyeong Mun <gwan-gyeong.mun@intel.com>
Cc: Gerd Hoffmann <kraxel@redhat.com>
Cc: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Cc: Lyude Paul <lyude@redhat.com>
Cc: José Roberto de Souza <jose.souza@intel.com>
Cc: Dariusz Marcinkiewicz <darekm@google.com>
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Cc: Raymond Smith <raymond.smith@arm.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Colin Ian King <colin.king@canonical.com>
Cc: Thomas Zimmermann <tzimmermann@suse.de>
Cc: Dan Carpenter <dan.carpenter@oracle.com>
Cc: Mihail Atanassov <Mihail.Atanassov@arm.com>
Cc: Lowry Li <Lowry.Li@arm.com>
Cc: Neil Armstrong <narmstrong@baylibre.com>
Cc: Jyri Sarha <jsarha@ti.com>
Cc: Tomi Valkeinen <tomi.valkeinen@ti.com>
Cc: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Cc: Steven Price <steven.price@arm.com>
Cc: Benjamin Gaignard <benjamin.gaignard@st.com>
Cc: Linus Walleij <linus.walleij@linaro.org>
Cc: Jagan Teki <jagan@amarulasolutions.com>
Cc: Icenowy Zheng <icenowy@aosc.io>
Cc: Iago Toral Quiroga <itoral@igalia.com>
Cc: David Riley <davidriley@chromium.org>

UAPI Changes:
- Never allow userptr into the mappable GGTT (Chris)
  No existing users. Avoid anyone from even trying to
  spare a deadlock scenario.

Cross-subsystem Changes:

Core Changes:

Driver Changes:

- Eliminate struct_mutex use as BKL! (Chris)
  Only used for execbuf serialisation.

- Initialize DDI TC and TBT ports (D-I) on Tigerlake (Lucas)
- Fix DKL link training for 2.7GHz and 1.62GHz (Jose)
- Add Tigerlake DKL PHY programming sequences (Clinton)
- Add Tigerlake Thunderbolt PLL divider values (Imre)

- drm/i915: Use helpers for drm_mm_node booleans (Chris)
- Restrict L3 remapping sysfs interface to dwords (Chris)
- Fix audio power up sequence for gen10+ display (Kai)
- Skip redundant execlist resubmission (Chris)
- Only unwedge if we can reset GPU first (Chris)
- Initialise breadcrumb lists on the virtual engine (Chris)
- Don't rely on kernel context existing during early errors (Matt A)
- Update Icelake+ MG_DP_MODE programming table (Clinton)
- Update DMC firmware for Icelake (Anusha)
- Downgrade DP MST error after unplugging TypeC cable (Srinivasan)
- Limit MST modes based on plane size too (Ville)
- Polish intel_tv_mode_valid() (Ville)
- Fix g4x sprite scaling stride check with GTT remapping (Ville)
- Don't advertize non-exisiting crtcs (Ville)
- Clean up encoder->crtc_mask setup (Ville)
- Use tc_port instead of port parameter to MG registers (Jose)
- Remove static variable for aux last status (Jani)
- Implement a better i945gm vblank irq vs. C-states workaround (Ville)

- Make the object creation interface consistent (CQ)
- Rename intel_vga_msr_write() to intel_vga_reset_io_mem() (Jani, Ville)
- Eliminate previous drm_dbg/drm_err usage (Jani)
- Move gmbus setup down to intel_modeset_init() (Jani)
- Abstract all vgaarb access to intel_vga.[ch] (Jani)
- Split out i915_switcheroo.[ch] from i915_drv.c (Jani)
- Use intel_gt in has_reset* (Chris)
- Eliminate return value for i915_gem_init_early (Matt A)
- Selftest improvements (Chris)
- Update HuC firmware header version number format (Daniele)

Linux 5.4-rc2

drm fixes for 5.4-rc2

core:
- writeback fixes

i915:
- Fix DP-MST crtc_mask
- Fix dsc dpp calculations
- Fix g4x sprite scaling stride check with GTT remapping
- Fix concurrence on cases where requests where getting retired at same time as resubmitted to HW
- Fix gen9 display resolutions by setting the right max plane width
- Fix GPU hang on preemption
- Mark contents as dirty on a write fault. This was breaking cursor sprite with dumb buffers.

komeda:
- memory leak fix

tilcdc:
- include fix

amdgpu:
- Enable bulk moves
- Power metrics fixes for Navi
- Fix S4 regression
- Add query for tcc disabled mask
- Fix several leaks in error paths
- randconfig fixes
- clang fixes

- Fix DP-MST crtc_mask
- Fix dsc dpp calculations
- Fix g4x sprite scaling stride check with GTT remapping

Short summary of fixes pull (less than what git shortlog provides):
- explain anything non-fixes (e.g. cleanups) and why it's appropriate
- highlight regressions
- summarize pull requests contained
This shouldn't be more than a few lines (or it indicates your fixes pull is a
bit too big).

- Fix dsc dpp calculations
- Fix g4x sprite scaling stride check with GTT remapping

 - One include fix for tilcdc
 - A clock fix for OMAP
 - A memory leak fix for Komeda
 - Some fixes for resources cleanups with writeback

 - One include fix for tilcdc
 - A memory leak fix for Komeda
 - Some fixes for resources cleanups with writeback

- One include fix for tilcdc

Linux 5.4-rc1

UAPI Changes:
- Revert "drm/i915: Fix DP-MST crtc_mask" to avoid MST regressions (Ville)
- Disable set/get_tiling ioctl on Gen12+ as hardware is gone (Daniel)
- Add immutable zpos plane properties (Ville)
- Report dual-subslice count as subslices for Tigerlake (Daniele)

Driver Changes:

<TIGERLAKE ENABLING>
- Enable HDCP 1.4 and 2.2 on Gen12+ (Ramalingam)
- Enable display state buffer (DSB) batch-programming (Animesh)
- Add 12 BPC support for Tigerlake (Anusha)
- Add maximum resolution supported by PSR2 HW for Tigerlake (Jose)
- Only allow PSR2 on supporting transcoders (Jose)
- Disable pipes in reverse order to comply with MST for Tigerlake+ (Jose)
- Implement Tigerlake DisplayPort training sequence (Jose)
- Do not apply WaIncreaseDefaultTLBEntries from Gen12 onwards (Michel)
- Reuse Icelake OA context logic for Tigerlake (Michel)
- Enable VD HCP/MFX sub-pipe power gating (Michel)
- Use separate context for relocations to deal with Tigerlake pre-parser (Daniele)
- Enabling DSC on Pipe A for Tigerlake (Madhumitha)
- Remove Yf tiling and legacy CCS support starting Tigerlake (Dhinakaran)
- Remove PSR link standby support starting Tigerlake (Jose)
- Access the right register when handling PSR interruptions (Jose)
- Move DP_TP_* registers from port to transcoder for Tigerlake (Lucas)
- Disable SAGV for Tigerlake (Lucas)
- Reuse Gen11 stolen initialization for Gen12 (Lucas)
- Apply FBC WA for Tigerlake too (Jose)
- Use engine relative LRIs on context setup for Tigerlake (Mika, Daniele)
- Register state context definition for Gen12 (Michel)
- Extend MI_SEMAPHORE_WAIT instruction for Tigerlake (Chris)
- Disable various Tigerlake features in attempt to have stable CI results (Chris)
- Add Tigerlake W/A to disable CPS aware color pipe by setting chicken bit (Radhakrishna)
- Add Tigerlake W/A to Enable Small PL for power benefit (Michel)
- Add missing DDI clock select during DP init sequence for Tigerlake (Clinton)
- Add missing update_active_dpll callback on Tigerlake (Clinton)
- Finish modular FIA support on registers for Tigerlake (Jose)
- Unify disable and enable phy clock gating functions on Tigerlake (Jose)
- Check the UC health of TC controllers after power on (Jose)
- Add TigerLake bandwidth checking (Stanislav)
- Add Pipe D cursor ctrl register for Gen12 (Ankit)
- Add DKL PHY PLL calculations (Lucas, Vandita, Jose)
- Add memory type decoding for bandwidth checking (James)
</TIGERLAKE ENABLING>

- Downgrade Gen7 and Cherryview back to aliasing-ppGTT (Chris)
- Limit MST to <= 8bpc once again (Ville)
- Restrict the aliasing-ppgtt to the size of the ggtt (Chris)
- Restore relaxed padding (OCL_OOB_SUPPRES_ENABLE) for SKL+ (Chris, Jason)
- Whitelist COMMON_SLICE_CHICKEN2 (Kenneth)
- Include GTT page-size info in error state (Matt A)
- Clear STOP_RING bit on reset (Chris)
- Ignore lost CSB completion events (Chris)
- Use a high priority wq for nonblocking plane updates (Ville)
- Bump up Skylake/Icelake+ display/plane/fb size restrictions (Manasi, Ville)
- Update Gen11/Gen12 forcewake ranges from BSpec (Mika, Daniele, Michel)
- Allow downscale factor of <3.0 on GLK+ for all formats (Ville)
- Add missing Comet Lake PCH PCI ID (Matt)
- Fix Gen11 SFC reset flow (Daniele)
- Fix YCbCr programming for ILK-IVB,HSW+ (Ville)
- Save audio frequency programming state at audio domain suspend (Kai)
- Fix DisplayPort DSC BPP calculations (Maarten)
- Add hardware readout for FEC (Maarten)
- Do not add all planes when checking scalers on GLK+ (Maarten)
- Make small joiner RAM buffer size platform-specific (Matt R)
- Use per-process HWSP as scratch (Michal Wi)

- Match allowed Gen11+ CDCLK values to BSpec (Matt R)
- Rework CDCLK code for clarity and table format (Matt R)
- Unify CDCLK code to reuse functions (Ville)
- Enhance CDCLK sanitization (Matt R)
- Preallocate Braswell top-level page directory (Chris)
- Make vgpu ppgtt notificaiton as atomic operation (Xiaolin)
- Use NOEVICT for first pass on attemping to pin a GGTT mmap (Chris)
- Disable PSR if more than one eDP panel is present (Jose)
- Make breadcrumb flushes more robust (Chris)
- Extend non readable MCR range (Mika)
- Protect our local workers against I915_FENCE_TIMEOUT (Chris)
- Allow stolen memory (and future local memory) addresses in sg_table (Matt A)
- Better organize the disable sequence in atomic_commit_tail() (Manasi)
- Fix regression with crtc disable ordering (Maarten)
- Add HW Gamma LUT readout (Swati)
- Hook up power management code to use intel_gt (Andi)
- Rework codebase towards use of intel_gt (Tvrtko)
- Remove incorrect BUG_ON for schedule-out (Chris, Vinay)
- Cleanup cache coloring code (Matt A)
- Flush writes before RING_TAIL update on SNB (Chris)
- Perform GGTT restore much earlier during resume (Chris)
- Make shrink pinning atomic (Chris)
- Make i915_vma.flags atomic for mutex reduction (Chris)
- Make sure the gen6 ppgtt is bound before first use without struct mutex (Chris)
- Report IOMMU status in debugfs (Chris)
- Disable FBC if BIOS reserved memory (stolen) is unavailable (Chris)
- Add a paranoid flushes and context reload around GPU reset (Chris)
- Skip engine busyness sampling when and where not needed (Tvrtko)
- Use GT parked time for estimating RC6 while asleep (Chris)
- Get the correct wakeref for reading hotplug registers from debugfs (Arkadiusz)
- Only apply a rmw mmio update if the value changes (Chris, Daniele)
- Extend Haswell GT1 PSMI workaround to all HSW (Chris)
- Only enqueue already completed requests (Chris)
- Fix preempt-to-busy interactions of virtual requests (Chris)
- Prevent bonded requests from overtaking each other on preemption (Chris)
- Mark contents as dirty on a write fault (Chris)
- Adjust length of MI_LOAD_REGISTER_REG (Michal Wi)
- Don't disable interrupts for intel_engine_breadcrumbs_irq() (Sebastian)
- Extract GT render sleep (rc6) management (Andi)

- Rework SSEU reporting code (Stuart)
- Use correct DSC registers in intel_configure_pps_for_dsc_encoder (Manasi)
- Use enum pipe instead of crtc index to track active pipes (Ville)
- Enforce irq-off lockdep check for for timeline locks (Chris)
- Flush the existing fence before GGTT read/write (Chris)
- Keep drm_i915_file_private around under RCU (Chris)
- Call dma_set_max_seg_size() to silence spurious warnings (Lyude)
- Make engine's batch pool safe for use with virtual engines (Chris)
- Align power domain names with port names (Imre)
- Parameterize and unify HPD code (Lucas)
- Use RCU for unlocked vm_idr lookup (Chris)
- Replace obj->pin_global with obj->frontbuffer (Chris)
- Rework code to use INTEL_NUM_PIPES() (Jani)
- Convert device info num_pipes to pipe_mask (Jani)
- Introduce INTEL_DISPLAY_ENABLED() (Jani)
- Stop conflating HAS_DISPLAY() and disabled display (Jani)
- Modularize i915 modesetting probing/init code (Jani)
- Use drm_format_info_is_yuv_semiplanar() instead of rolling own (Ville)
- Other display codebase cleanups (Ville)
- Other GEM codebase cleanup, lockdep and selftest improvements (Chris)
- Future-proof DDC pin mapping to reuse ICP variant (Matt R)
- Rewrite timeline handling to be RCU based (Chris)
- Define explicit wedged on init reset state (Michal Wi)
- Add GuC firmware for Elkhartlake (Daniele)
- Update HuC firmware naming convention and bump versions (Anusha)
- Extract common code from GuC stop/disable comm (Fernando)
- Fix perf kernel-doc formatting for struct members (Anna)
- Documentation fixes (Joonas)

drm fixes for 5.4-rc1

core:
- Some cleanups and fixes in the self-refresh helpers
- Some cleanups and fixes in the atomic helpers

amdgpu:
- Fix a 64 bit divide
- Prevent a memory leak in a failure case in dc
- Load proper gfx firmware on navi14 variants
- Add more navi12 and navi14 PCI ids
- Misc fixes for renoir
- Fix bandwidth issues with multiple displays on vega20
- Support for Dali
- Fix a possible oops with KFD on hawaii
- Fix for backlight level after resume on some APUs
- Other misc fixes

panfrost:
- Multiple panfrost fixes for regulator support and page fault handling