dri-devel September 2019

dri-devel@lists.freedesktop.org

315 participants
2278 discussions

[PATCH v7 0/3] CRTC background color
by Matt Roper 11 Oct '19

11 Oct '19

The previous version of this series was posted in February here: https://lists.freedesktop.org/archives/dri-devel/2019-February/208068.html Right before we merged this in February Maarten noticed that we should be setting up the initial property state in a CRTC reset function (which didn't exist yet) instead of when the property was attached. Maarten landed the CRTC reset functionality a week or two later, but I was busy with travel and other work at the time, so revisiting and rebasing this background color series fell through the cracks and I'm just getting back to it now. Userspace consumer is chromeos; these are the links the ChromeOS folks gave me back in February: https://chromium-review.googlesource.com/c/chromium/src/+/1278858 https://chromium-review.googlesource.com/c/chromiumos/platform/drm-tests/+/… IGT is still the same as posted in February: https://lists.freedesktop.org/archives/igt-dev/2019-February/009637.html Cc: Maarten Lankhorst <maarten.lankhorst(a)linux.intel.com> Matt Roper (3): drm: Add CRTC background color property drm/i915/gen9+: Add support for pipe background color drm/i915: Add background color hardware readout and state check drivers/gpu/drm/drm_atomic_state_helper.c | 4 +- drivers/gpu/drm/drm_atomic_uapi.c | 4 ++ drivers/gpu/drm/drm_blend.c | 35 +++++++++++++-- drivers/gpu/drm/drm_mode_config.c | 6 +++ drivers/gpu/drm/i915/display/intel_color.c | 11 +++-- drivers/gpu/drm/i915/display/intel_display.c | 45 ++++++++++++++++++++ drivers/gpu/drm/i915/i915_debugfs.c | 9 ++++ include/drm/drm_blend.h | 1 + include/drm/drm_crtc.h | 12 ++++++ include/drm/drm_mode_config.h | 5 +++ include/uapi/drm/drm_mode.h | 28 ++++++++++++ 11 files changed, 153 insertions(+), 7 deletions(-) -- 2.21.0

6 10

[PATCH v2 1/8] dt-bindings: rng: exynos4-rng: Convert Exynos PRNG bindings to json-schema
by Krzysztof Kozlowski 10 Oct '19

10 Oct '19

Convert Samsung Exynos Pseudo Random Number Generator bindings to DT schema format using json-schema. Signed-off-by: Krzysztof Kozlowski <krzk(a)kernel.org> --- Changes since v1: 1. Indent example with four spaces (more readable). --- .../bindings/rng/samsung,exynos4-rng.txt | 19 --------- .../bindings/rng/samsung,exynos4-rng.yaml | 41 +++++++++++++++++++ MAINTAINERS | 2 +- 3 files changed, 42 insertions(+), 20 deletions(-) delete mode 100644 Documentation/devicetree/bindings/rng/samsung,exynos4-rng.txt create mode 100644 Documentation/devicetree/bindings/rng/samsung,exynos4-rng.yaml diff --git a/Documentation/devicetree/bindings/rng/samsung,exynos4-rng.txt b/Documentation/devicetree/bindings/rng/samsung,exynos4-rng.txt deleted file mode 100644 index a13fbdb4bd88..000000000000 --- a/Documentation/devicetree/bindings/rng/samsung,exynos4-rng.txt +++ /dev/null @@ -1,19 +0,0 @@ -Exynos Pseudo Random Number Generator - -Required properties: - -- compatible : One of: - - "samsung,exynos4-rng" for Exynos4210 and Exynos4412 - - "samsung,exynos5250-prng" for Exynos5250+ -- reg : Specifies base physical address and size of the registers map. -- clocks : Phandle to clock-controller plus clock-specifier pair. -- clock-names : "secss" as a clock name. - -Example: - - rng@10830400 { - compatible = "samsung,exynos4-rng"; - reg = <0x10830400 0x200>; - clocks = <&clock CLK_SSS>; - clock-names = "secss"; - }; diff --git a/Documentation/devicetree/bindings/rng/samsung,exynos4-rng.yaml b/Documentation/devicetree/bindings/rng/samsung,exynos4-rng.yaml new file mode 100644 index 000000000000..2d075d6c87b6 --- /dev/null +++ b/Documentation/devicetree/bindings/rng/samsung,exynos4-rng.yaml @@ -0,0 +1,41 @@ +# SPDX-License-Identifier: GPL-2.0 +%YAML 1.2 +--- +$id: http://devicetree.org/schemas/rng/samsung,exynos4-rng.yaml# +$schema: http://devicetree.org/meta-schemas/core.yaml# + +title: Samsung Exynos SoC Pseudo Random Number Generator + +maintainers: + - Krzysztof Kozlowski <krzk(a)kernel.org> + +properties: + compatible: + enum: + - samsung,exynos4-rng # for Exynos4210 and Exynos4412 + - samsung,exynos5250-prng # for Exynos5250+ + + reg: + maxItems: 1 + + clocks: + maxItems: 1 + + clock-names: + items: + - const: secss + +required: + - compatible + - reg + - clock-names + - clocks + +examples: + - | + rng@10830400 { + compatible = "samsung,exynos4-rng"; + reg = <0x10830400 0x200>; + clocks = <&clock 255>; // CLK_SSS + clock-names = "secss"; + }; diff --git a/MAINTAINERS b/MAINTAINERS index 00969a90f94c..9cec4494b9a8 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -14181,7 +14181,7 @@ L: linux-crypto(a)vger.kernel.org L: linux-samsung-soc(a)vger.kernel.org S: Maintained F: drivers/crypto/exynos-rng.c -F: Documentation/devicetree/bindings/rng/samsung,exynos4-rng.txt +F: Documentation/devicetree/bindings/rng/samsung,exynos4-rng.yaml SAMSUNG EXYNOS TRUE RANDOM NUMBER GENERATOR (TRNG) DRIVER M: Łukasz Stelmach <l.stelmach(a)samsung.com> -- 2.17.1

3 12

[PATCH v2 0/5] DRM/MSM: Add support for MSM8956 and Adreno 510
by kholk11＠gmail.com 10 Oct '19

10 Oct '19

From: AngeloGioacchino Del Regno <kholk11(a)gmail.com> This patch series enables support for MSM8956/76 and its Adreno 510 GPU on the current DRM driver. The personal aim is to upstream MSM8956 as much as possible. This code has been tested on two Sony phones featuring the Qualcomm MSM8956 SoC. Changes in v2: - MDP5: Documented tbu and tbu_rt clocks (Jeffrey) - Adreno510: - Lower case hex where required (Jordan) - Direct register writes (Jordan) - Used gpu_rmw() where required (Jordan) - No mentioning of unsupported A5xx (Jordan) - ZAP firmware exclusions not per-model (Rob) Angelo G. Del Regno (5): drm/msm/mdp5: Add optional TBU and TBU_RT clocks drm/msm/mdp5: Add configuration for msm8x56 drm/msm/dsi: Add configuration for 28nm PLL on family B drm/msm/dsi: Add configuration for 8x56 drm/msm/adreno: Add support for Adreno 510 GPU .../devicetree/bindings/display/msm/dsi.txt | 1 + .../devicetree/bindings/display/msm/mdp5.txt | 2 + drivers/gpu/drm/msm/adreno/a5xx_gpu.c | 73 +++++++++++--- drivers/gpu/drm/msm/adreno/a5xx_power.c | 7 ++ drivers/gpu/drm/msm/adreno/adreno_device.c | 15 +++ drivers/gpu/drm/msm/adreno/adreno_gpu.h | 5 + drivers/gpu/drm/msm/disp/mdp5/mdp5_cfg.c | 99 +++++++++++++++++++ drivers/gpu/drm/msm/disp/mdp5/mdp5_kms.c | 10 ++ drivers/gpu/drm/msm/disp/mdp5/mdp5_kms.h | 2 + drivers/gpu/drm/msm/dsi/dsi_cfg.c | 22 +++++ drivers/gpu/drm/msm/dsi/dsi_cfg.h | 1 + drivers/gpu/drm/msm/dsi/phy/dsi_phy.c | 2 + drivers/gpu/drm/msm/dsi/phy/dsi_phy.h | 1 + drivers/gpu/drm/msm/dsi/phy/dsi_phy_28nm.c | 18 ++++ 14 files changed, 244 insertions(+), 14 deletions(-) -- 2.21.0

2 7

[PATCH] dma-buf/resv: fix exclusive fence get
by Qiang Yu 10 Oct '19

10 Oct '19

This causes kernel crash when testing lima driver. Cc: Christian König <christian.koenig(a)amd.com> Fixes: b8c036dfc66f ("dma-buf: simplify reservation_object_get_fences_rcu a bit") Signed-off-by: Qiang Yu <yuq825(a)gmail.com> --- drivers/dma-buf/dma-resv.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/dma-buf/dma-resv.c b/drivers/dma-buf/dma-resv.c index 42a8f3f11681..709002515550 100644 --- a/drivers/dma-buf/dma-resv.c +++ b/drivers/dma-buf/dma-resv.c @@ -471,7 +471,7 @@ int dma_resv_get_fences_rcu(struct dma_resv *obj, if (pfence_excl) *pfence_excl = fence_excl; else if (fence_excl) - shared[++shared_count] = fence_excl; + shared[shared_count++] = fence_excl; if (!shared_count) { kfree(shared); -- 2.17.1

4 7

[PATCH] drm/edid: Add 6 bpc quirk for SDC panel in Lenovo G50
by Kai-Heng Feng 10 Oct '19

10 Oct '19

Another panel that needs 6BPC quirk. BugLink: https://bugs.launchpad.net/bugs/1819968 Cc: <stable(a)vger.kernel.org> # v4.8+ Signed-off-by: Kai-Heng Feng <kai.heng.feng(a)canonical.com> --- drivers/gpu/drm/drm_edid.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/gpu/drm/drm_edid.c b/drivers/gpu/drm/drm_edid.c index 990b1909f9d7..1cb4d0052efe 100644 --- a/drivers/gpu/drm/drm_edid.c +++ b/drivers/gpu/drm/drm_edid.c @@ -166,6 +166,9 @@ static const struct edid_quirk { /* Medion MD 30217 PG */ { "MED", 0x7b8, EDID_QUIRK_PREFER_LARGE_75 }, + /* Lenovo G50 */ + { "SDC", 18514, EDID_QUIRK_FORCE_6BPC }, + /* Panel in Samsung NP700G7A-S01PL notebook reports 6bpc */ { "SEC", 0xd033, EDID_QUIRK_FORCE_8BPC }, -- 2.17.1

2 3

[PATCH] drm/tiny: Kconfig: Remove always-y THERMAL dep. from TINYDRM_REPAPER
by Ulf Magnusson 10 Oct '19

10 Oct '19

Commit 554b3529fe01 ("thermal/drivers/core: Remove the module Kconfig's option") changed the type of THERMAL from tristate to bool, so THERMAL || !THERMAL is now always y. Remove the redundant dependency. Discovered through Kconfiglib detecting a dependency loop. The C tools simplify the expression to y before running dependency loop detection, and so don't see it. Changing the type of THERMAL back to tristate makes the C tools detect the same loop. Not sure if running dep. loop detection after simplification can be called a bug. Fixing this nit unbreaks Kconfiglib on the kernel at least. Signed-off-by: Ulf Magnusson <ulfalizer(a)gmail.com> --- drivers/gpu/drm/tiny/Kconfig | 1 - 1 file changed, 1 deletion(-) diff --git a/drivers/gpu/drm/tiny/Kconfig b/drivers/gpu/drm/tiny/Kconfig index 504763423d46..a46ac284dd5e 100644 --- a/drivers/gpu/drm/tiny/Kconfig +++ b/drivers/gpu/drm/tiny/Kconfig @@ -63,7 +63,6 @@ config TINYDRM_REPAPER depends on DRM && SPI select DRM_KMS_HELPER select DRM_KMS_CMA_HELPER - depends on THERMAL || !THERMAL help DRM driver for the following Pervasive Displays panels: 1.44" TFT EPD Panel (E1144CS021) -- 2.20.1

2 2

[PATCH 0/6] drm/lima: simplify driver by using more drm helpers
by Qiang Yu 10 Oct '19

10 Oct '19

By using shared drm helpers: 1. drm_gem_objects_lookup 2. drm_gem_(un)lock_reservations 3. drm_gem_shmem_helpers we can simplify lima driver a lot and benifit from updates to these functions. drm_gem_objects_lookup need a refine in order to be used by lima. Note: 1. changes to panfrost and v3d are just compile tested. 2. patch series is based on drm-misc-next branch Qiang Yu (6): drm/gem: refine drm_gem_objects_lookup drm/v3d: use drm_gem_objects_lookup drm/lima: use drm_gem_objects_lookup drm/lima: use drm_gem_shmem_helpers drm/lima: use drm_gem_(un)lock_reservations drm/lima: add __GFP_NOWARN flag to all dma_alloc_wc drivers/gpu/drm/drm_gem.c | 28 +-- drivers/gpu/drm/lima/Kconfig | 1 + drivers/gpu/drm/lima/Makefile | 4 +- drivers/gpu/drm/lima/lima_device.c | 2 +- drivers/gpu/drm/lima/lima_drv.c | 27 +-- drivers/gpu/drm/lima/lima_gem.c | 254 ++++++++++-------------- drivers/gpu/drm/lima/lima_gem.h | 32 ++- drivers/gpu/drm/lima/lima_gem_prime.c | 46 ----- drivers/gpu/drm/lima/lima_gem_prime.h | 13 -- drivers/gpu/drm/lima/lima_mmu.c | 1 - drivers/gpu/drm/lima/lima_object.c | 119 ----------- drivers/gpu/drm/lima/lima_object.h | 35 ---- drivers/gpu/drm/lima/lima_sched.c | 6 +- drivers/gpu/drm/lima/lima_vm.c | 87 ++++---- drivers/gpu/drm/panfrost/panfrost_drv.c | 23 ++- drivers/gpu/drm/v3d/v3d_gem.c | 35 +--- include/drm/drm_gem.h | 2 +- 17 files changed, 222 insertions(+), 493 deletions(-) delete mode 100644 drivers/gpu/drm/lima/lima_gem_prime.c delete mode 100644 drivers/gpu/drm/lima/lima_gem_prime.h delete mode 100644 drivers/gpu/drm/lima/lima_object.c delete mode 100644 drivers/gpu/drm/lima/lima_object.h -- 2.17.1

4 12

[PATCH 0/3] drm/tinydrm: Rename to drm/tiny
by Noralf Trønnes 10 Oct '19

10 Oct '19

This is the final polish on tinydrm turning it into _the_ place for tiny DRM drivers. Noralf. Noralf Trønnes (3): drm/tinydrm/Kconfig: Remove menuconfig DRM_TINYDRM drm/tinydrm: Rename folder to tiny drm/gm12u320: Move driver to drm/tiny MAINTAINERS | 14 +++---- drivers/gpu/drm/Kconfig | 4 +- drivers/gpu/drm/Makefile | 3 +- drivers/gpu/drm/gm12u320/Kconfig | 9 ---- drivers/gpu/drm/gm12u320/Makefile | 2 - drivers/gpu/drm/{tinydrm => tiny}/Kconfig | 41 +++++++++++++------ drivers/gpu/drm/{tinydrm => tiny}/Makefile | 1 + drivers/gpu/drm/{gm12u320 => tiny}/gm12u320.c | 0 drivers/gpu/drm/{tinydrm => tiny}/hx8357d.c | 0 drivers/gpu/drm/{tinydrm => tiny}/ili9225.c | 0 drivers/gpu/drm/{tinydrm => tiny}/ili9341.c | 0 drivers/gpu/drm/{tinydrm => tiny}/mi0283qt.c | 0 drivers/gpu/drm/{tinydrm => tiny}/repaper.c | 0 drivers/gpu/drm/{tinydrm => tiny}/st7586.c | 0 drivers/gpu/drm/{tinydrm => tiny}/st7735r.c | 0 15 files changed, 38 insertions(+), 36 deletions(-) delete mode 100644 drivers/gpu/drm/gm12u320/Kconfig delete mode 100644 drivers/gpu/drm/gm12u320/Makefile rename drivers/gpu/drm/{tinydrm => tiny}/Kconfig (73%) rename drivers/gpu/drm/{tinydrm => tiny}/Makefile (89%) rename drivers/gpu/drm/{gm12u320 => tiny}/gm12u320.c (100%) rename drivers/gpu/drm/{tinydrm => tiny}/hx8357d.c (100%) rename drivers/gpu/drm/{tinydrm => tiny}/ili9225.c (100%) rename drivers/gpu/drm/{tinydrm => tiny}/ili9341.c (100%) rename drivers/gpu/drm/{tinydrm => tiny}/mi0283qt.c (100%) rename drivers/gpu/drm/{tinydrm => tiny}/repaper.c (100%) rename drivers/gpu/drm/{tinydrm => tiny}/st7586.c (100%) rename drivers/gpu/drm/{tinydrm => tiny}/st7735r.c (100%) -- 2.20.1

7 21

[PATCH v2 00/24] drm/msm: Various dpu locking and legacy cleanups
by Sean Paul 10 Oct '19

10 Oct '19

From: Sean Paul <seanpaul(a)chromium.org> This was originally 3 patchsets, but none have gotten full review, so I figured I'd package the v2's all up into one set so it's easier to track. Set 1- https://lists.freedesktop.org/archives/dri-devel/2018-November/196170.html Set 2- https://lists.freedesktop.org/archives/dri-devel/2018-November/196184.html Set 3- https://lists.freedesktop.org/archives/dri-devel/2018-November/196276.html Thanks, Sean Sean Paul (24): drm/msm: dpu: Remove dpu_power_handle_get_dbus_name() drm/msm: dpu: Remove unused trace_dpu_perf_update_bus() drm/msm: dpu: Remove dpu_power_client drm/msm: dpu: Don't use power_event for vbif_init_memtypes drm/msm: dpu: Handle crtc pm_runtime_resume() directly drm/msm: dpu: Remove power_handle from core_perf drm/msm: dpu: Include dpu_io_util.h directly in dpu_kms.h drm/msm: dpu: Move DPU_POWER_HANDLE_DBUS_ID to core_perf drm/msm: dpu: Remove dpu_power_handle drm/msm: dpu: Fix typo in dpu_encoder drm/msm: dpu: Add ->enabled to dpu_encoder_virt drm/msm: dpu: Move crtc runtime resume to encoder drm/msm: dpu: Don't drop locks in crtc_vblank_enable drm/msm: dpu: Grab the modeset locks in frame_event drm/msm: dpu: Stop using encoder->crtc pointer drm/msm: dpu: Add modeset lock checks where applicable drm/msm: dpu: Move pm_runtime_(get|put) from vblank_enable drm/msm: dpu: Remove crtc_lock from setup_mixers drm/msm: dpu: Remove vblank_callback from encoder drm/msm: dpu: Use atomic_disable for dpu_crtc_disable drm/msm: dpu: Don't bother checking ->enabled in dpu_crtc_vblank drm/msm: dpu: Separate crtc assignment from vblank enable drm/msm: dpu: Remove vblank_requested flag from dpu_crtc drm/msm: dpu: Remove crtc_lock drivers/gpu/drm/msm/Makefile | 1 - drivers/gpu/drm/msm/disp/dpu1/dpu_core_perf.c | 37 ++- drivers/gpu/drm/msm/disp/dpu1/dpu_core_perf.h | 22 +- drivers/gpu/drm/msm/disp/dpu1/dpu_crtc.c | 216 +++++----------- drivers/gpu/drm/msm/disp/dpu1/dpu_crtc.h | 15 +- drivers/gpu/drm/msm/disp/dpu1/dpu_encoder.c | 97 ++++--- drivers/gpu/drm/msm/disp/dpu1/dpu_encoder.h | 24 +- drivers/gpu/drm/msm/disp/dpu1/dpu_kms.c | 76 ++---- drivers/gpu/drm/msm/disp/dpu1/dpu_kms.h | 6 +- .../gpu/drm/msm/disp/dpu1/dpu_power_handle.c | 240 ------------------ .../gpu/drm/msm/disp/dpu1/dpu_power_handle.h | 217 ---------------- drivers/gpu/drm/msm/disp/dpu1/dpu_trace.h | 43 +--- 12 files changed, 204 insertions(+), 790 deletions(-) delete mode 100644 drivers/gpu/drm/msm/disp/dpu1/dpu_power_handle.c delete mode 100644 drivers/gpu/drm/msm/disp/dpu1/dpu_power_handle.h -- Sean Paul, Software Engineer, Google / Chromium OS

3 51

[PATCH 1/2] drm/nouveau: move io_reserve_lru handling into the driver
by Christian König 10 Oct '19

10 Oct '19

While working on TTM cleanups I've found that the io_reserve_lru used by Nouveau is actually not working at all. In general we should remove driver specific handling from the memory management, so this patch moves the io_reserve_lru handling into Nouveau instead. The patch should be functional correct, but is only compile tested! Signed-off-by: Christian König <christian.koenig(a)amd.com> --- drivers/gpu/drm/nouveau/nouveau_bo.c | 106 ++++++++++++++++++++++++++-------- drivers/gpu/drm/nouveau/nouveau_bo.h | 3 + drivers/gpu/drm/nouveau/nouveau_drv.h | 2 + drivers/gpu/drm/nouveau/nouveau_ttm.c | 43 +++++++++++++- 4 files changed, 130 insertions(+), 24 deletions(-) diff --git a/drivers/gpu/drm/nouveau/nouveau_bo.c b/drivers/gpu/drm/nouveau/nouveau_bo.c index e918b437af17..9532ad1bd474 100644 --- a/drivers/gpu/drm/nouveau/nouveau_bo.c +++ b/drivers/gpu/drm/nouveau/nouveau_bo.c @@ -139,6 +139,7 @@ nouveau_bo_del_ttm(struct ttm_buffer_object *bo) if (unlikely(nvbo->bo.base.filp)) DRM_ERROR("bo %p still attached to GEM object\n", bo); WARN_ON(nvbo->pin_refcnt > 0); + nouveau_bo_del_io_reserve_lru(bo); nv10_bo_put_tile_region(dev, nvbo->tile, NULL); kfree(nvbo); } @@ -297,6 +298,7 @@ nouveau_bo_init(struct nouveau_bo *nvbo, u64 size, int align, u32 flags, nouveau_bo_fixup_align(nvbo, flags, &align, &size); nvbo->bo.mem.num_pages = size >> PAGE_SHIFT; nouveau_bo_placement_set(nvbo, flags, 0); + INIT_LIST_HEAD(&nvbo->io_reserve_lru); ret = ttm_bo_init(nvbo->bo.bdev, &nvbo->bo, size, type, &nvbo->placement, align >> PAGE_SHIFT, false, @@ -566,6 +568,26 @@ nouveau_bo_sync_for_cpu(struct nouveau_bo *nvbo) PAGE_SIZE, DMA_FROM_DEVICE); } +void nouveau_bo_add_io_reserve_lru(struct ttm_buffer_object *bo) +{ + struct nouveau_drm *drm = nouveau_bdev(bo->bdev); + struct nouveau_bo *nvbo = nouveau_bo(bo); + + mutex_lock(&drm->ttm.io_reserve_mutex); + list_move_tail(&nvbo->io_reserve_lru, &drm->ttm.io_reserve_lru); + mutex_unlock(&drm->ttm.io_reserve_mutex); +} + +void nouveau_bo_del_io_reserve_lru(struct ttm_buffer_object *bo) +{ + struct nouveau_drm *drm = nouveau_bdev(bo->bdev); + struct nouveau_bo *nvbo = nouveau_bo(bo); + + mutex_lock(&drm->ttm.io_reserve_mutex); + list_del_init(&nvbo->io_reserve_lru); + mutex_unlock(&drm->ttm.io_reserve_mutex); +} + int nouveau_bo_validate(struct nouveau_bo *nvbo, bool interruptible, bool no_wait_gpu) @@ -674,8 +696,6 @@ nouveau_bo_init_mem_type(struct ttm_bo_device *bdev, uint32_t type, } man->func = &nouveau_vram_manager; - man->io_reserve_fastpath = false; - man->use_io_reserve_lru = true; } else { man->func = &ttm_bo_manager_func; } @@ -1304,6 +1324,8 @@ nouveau_bo_move_ntfy(struct ttm_buffer_object *bo, bool evict, if (bo->destroy != nouveau_bo_del_ttm) return; + nouveau_bo_del_io_reserve_lru(bo); + if (mem && new_reg->mem_type != TTM_PL_SYSTEM && mem->mem.page == nvbo->page) { list_for_each_entry(vma, &nvbo->vma_list, head) { @@ -1426,6 +1448,30 @@ nouveau_bo_verify_access(struct ttm_buffer_object *bo, struct file *filp) filp->private_data); } +static void +nouveau_ttm_io_mem_free_locked(struct nouveau_drm *drm, struct ttm_mem_reg *reg) +{ + struct nouveau_mem *mem = nouveau_mem(reg); + + if (!reg->bus.base) + return; /* already freed */ + + if (drm->client.mem->oclass >= NVIF_CLASS_MEM_NV50) { + switch (reg->mem_type) { + case TTM_PL_TT: + if (mem->kind) + nvif_object_unmap_handle(&mem->mem.object); + break; + case TTM_PL_VRAM: + nvif_object_unmap_handle(&mem->mem.object); + break; + default: + break; + } + } + reg->bus.base = 0; +} + static int nouveau_ttm_io_mem_reserve(struct ttm_bo_device *bdev, struct ttm_mem_reg *reg) { @@ -1433,18 +1479,26 @@ nouveau_ttm_io_mem_reserve(struct ttm_bo_device *bdev, struct ttm_mem_reg *reg) struct nouveau_drm *drm = nouveau_bdev(bdev); struct nvkm_device *device = nvxx_device(&drm->client.device); struct nouveau_mem *mem = nouveau_mem(reg); + struct nouveau_bo *nvbo; + int ret; + + if (reg->bus.base) + return 0; /* already mapped */ reg->bus.addr = NULL; reg->bus.offset = 0; reg->bus.size = reg->num_pages << PAGE_SHIFT; - reg->bus.base = 0; reg->bus.is_iomem = false; if (!(man->flags & TTM_MEMTYPE_FLAG_MAPPABLE)) return -EINVAL; + + mutex_lock(&drm->ttm.io_reserve_mutex); +retry: switch (reg->mem_type) { case TTM_PL_SYSTEM: /* System memory */ - return 0; + ret = 0; + goto out; case TTM_PL_TT: #if IS_ENABLED(CONFIG_AGP) if (drm->agp.bridge) { @@ -1468,7 +1522,6 @@ nouveau_ttm_io_mem_reserve(struct ttm_bo_device *bdev, struct ttm_mem_reg *reg) } args; u64 handle, length; u32 argc = 0; - int ret; switch (mem->mem.object.oclass) { case NVIF_CLASS_MEM_NV50: @@ -1492,38 +1545,45 @@ nouveau_ttm_io_mem_reserve(struct ttm_bo_device *bdev, struct ttm_mem_reg *reg) ret = nvif_object_map_handle(&mem->mem.object, &args, argc, &handle, &length); - if (ret != 1) - return ret ? ret : -EINVAL; + if (ret != 1) { + ret = ret ? ret : -EINVAL; + goto out; + } + ret = 0; reg->bus.base = 0; reg->bus.offset = handle; } break; default: - return -EINVAL; + ret = -EINVAL; } - return 0; + +out: + if (ret == -EAGAIN) { + nvbo = list_first_entry_or_null(&drm->ttm.io_reserve_lru, + typeof(*nvbo), + io_reserve_lru); + if (nvbo) { + list_del_init(&nvbo->io_reserve_lru); + ttm_bo_unmap_virtual(&nvbo->bo); + nouveau_ttm_io_mem_free_locked(drm, &nvbo->bo.mem); + goto retry; + } + + } + mutex_unlock(&drm->ttm.io_reserve_mutex); + return ret; } static void nouveau_ttm_io_mem_free(struct ttm_bo_device *bdev, struct ttm_mem_reg *reg) { struct nouveau_drm *drm = nouveau_bdev(bdev); - struct nouveau_mem *mem = nouveau_mem(reg); - if (drm->client.mem->oclass >= NVIF_CLASS_MEM_NV50) { - switch (reg->mem_type) { - case TTM_PL_TT: - if (mem->kind) - nvif_object_unmap_handle(&mem->mem.object); - break; - case TTM_PL_VRAM: - nvif_object_unmap_handle(&mem->mem.object); - break; - default: - break; - } - } + mutex_lock(&drm->ttm.io_reserve_mutex); + nouveau_ttm_io_mem_free_locked(drm, reg); + mutex_unlock(&drm->ttm.io_reserve_mutex); } static int diff --git a/drivers/gpu/drm/nouveau/nouveau_bo.h b/drivers/gpu/drm/nouveau/nouveau_bo.h index 62930d834fba..b8a17bc19d7b 100644 --- a/drivers/gpu/drm/nouveau/nouveau_bo.h +++ b/drivers/gpu/drm/nouveau/nouveau_bo.h @@ -17,6 +17,7 @@ struct nouveau_bo { bool force_coherent; struct ttm_bo_kmap_obj kmap; struct list_head head; + struct list_head io_reserve_lru; /* protected by ttm_bo_reserve() */ struct drm_file *reserved_by; @@ -92,6 +93,8 @@ int nouveau_bo_validate(struct nouveau_bo *, bool interruptible, bool no_wait_gpu); void nouveau_bo_sync_for_device(struct nouveau_bo *nvbo); void nouveau_bo_sync_for_cpu(struct nouveau_bo *nvbo); +void nouveau_bo_add_io_reserve_lru(struct ttm_buffer_object *bo); +void nouveau_bo_del_io_reserve_lru(struct ttm_buffer_object *bo); /* TODO: submit equivalent to TTM generic API upstream? */ static inline void __iomem * diff --git a/drivers/gpu/drm/nouveau/nouveau_drv.h b/drivers/gpu/drm/nouveau/nouveau_drv.h index 70f34cacc552..060e2ab1fd95 100644 --- a/drivers/gpu/drm/nouveau/nouveau_drv.h +++ b/drivers/gpu/drm/nouveau/nouveau_drv.h @@ -158,6 +158,8 @@ struct nouveau_drm { int type_vram; int type_host[2]; int type_ncoh[2]; + struct mutex io_reserve_mutex; + struct list_head io_reserve_lru; } ttm; /* GEM interface support */ diff --git a/drivers/gpu/drm/nouveau/nouveau_ttm.c b/drivers/gpu/drm/nouveau/nouveau_ttm.c index f0daf958e03a..7f8495066e8b 100644 --- a/drivers/gpu/drm/nouveau/nouveau_ttm.c +++ b/drivers/gpu/drm/nouveau/nouveau_ttm.c @@ -162,13 +162,51 @@ const struct ttm_mem_type_manager_func nv04_gart_manager = { .debug = nouveau_manager_debug }; +static vm_fault_t nouveau_ttm_fault(struct vm_fault *vmf) +{ + struct vm_area_struct *vma = vmf->vma; + struct ttm_buffer_object *bo = vma->vm_private_data; + pgprot_t prot; + vm_fault_t ret; + + ret = ttm_bo_vm_reserve(bo, vmf); + if (ret) + return ret; + + nouveau_bo_del_io_reserve_lru(bo); + + prot = vm_get_page_prot(vma->vm_flags); + ret = ttm_bo_vm_fault_reserved(vmf, prot, TTM_BO_VM_NUM_PREFAULT); + if (ret == VM_FAULT_RETRY && !(vmf->flags & FAULT_FLAG_RETRY_NOWAIT)) + return ret; + + nouveau_bo_add_io_reserve_lru(bo); + + dma_resv_unlock(bo->base.resv); + + return ret; +} + +static struct vm_operations_struct nouveau_ttm_vm_ops = { + .fault = nouveau_ttm_fault, + .open = ttm_bo_vm_open, + .close = ttm_bo_vm_close, + .access = ttm_bo_vm_access +}; + int nouveau_ttm_mmap(struct file *filp, struct vm_area_struct *vma) { struct drm_file *file_priv = filp->private_data; struct nouveau_drm *drm = nouveau_drm(file_priv->minor->dev); + int ret; - return ttm_bo_mmap(filp, vma, &drm->ttm.bdev); + ret = ttm_bo_mmap(filp, vma, &drm->ttm.bdev); + if (ret) + return ret; + + vma->vm_ops = &nouveau_ttm_vm_ops; + return 0; } static int @@ -272,6 +310,9 @@ nouveau_ttm_init(struct nouveau_drm *drm) return ret; } + mutex_init(&drm->ttm.io_reserve_mutex); + INIT_LIST_HEAD(&drm->ttm.io_reserve_lru); + NV_INFO(drm, "VRAM: %d MiB\n", (u32)(drm->gem.vram_available >> 20)); NV_INFO(drm, "GART: %d MiB\n", (u32)(drm->gem.gart_available >> 20)); return 0; -- 2.14.1

2 4

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

dri-devel September 2019