dri-devel February 2021

dri-devel@lists.freedesktop.org

314 participants
684 discussions

[PATCH 01/15] drm/arc: Switch to devm_drm_dev_alloc
by Daniel Vetter 25 Feb '21

25 Feb '21

- Need to embedded the drm_device, but for now we keep the usual pointer chasing. - No more devm_kzalloc, which fixes a lifetime issues on driver remove. - No more drm_dev_put, that's done by devm_ now. Acked-by: Sam Ravnborg <sam(a)ravnborg.org> Cc: Eugeniy Paltsev <Eugeniy.Paltsev(a)synopsys.com> Signed-off-by: Daniel Vetter <daniel.vetter(a)intel.com> Cc: Alexey Brodkin <abrodkin(a)synopsys.com> --- drivers/gpu/drm/arc/arcpgu.h | 1 + drivers/gpu/drm/arc/arcpgu_drv.c | 33 +++++++++++++------------------- 2 files changed, 14 insertions(+), 20 deletions(-) diff --git a/drivers/gpu/drm/arc/arcpgu.h b/drivers/gpu/drm/arc/arcpgu.h index 6aac44b953ad..cd9e932f501e 100644 --- a/drivers/gpu/drm/arc/arcpgu.h +++ b/drivers/gpu/drm/arc/arcpgu.h @@ -9,6 +9,7 @@ #define _ARCPGU_H_ struct arcpgu_drm_private { + struct drm_device drm; void __iomem *regs; struct clk *clk; struct drm_framebuffer *fb; diff --git a/drivers/gpu/drm/arc/arcpgu_drv.c b/drivers/gpu/drm/arc/arcpgu_drv.c index f164818ec477..68eb4a31c54b 100644 --- a/drivers/gpu/drm/arc/arcpgu_drv.c +++ b/drivers/gpu/drm/arc/arcpgu_drv.c @@ -42,18 +42,14 @@ static void arcpgu_setup_mode_config(struct drm_device *drm) DEFINE_DRM_GEM_CMA_FOPS(arcpgu_drm_ops); -static int arcpgu_load(struct drm_device *drm) +static int arcpgu_load(struct arcpgu_drm_private *arcpgu) { - struct platform_device *pdev = to_platform_device(drm->dev); - struct arcpgu_drm_private *arcpgu; + struct platform_device *pdev = to_platform_device(arcpgu->drm.dev); struct device_node *encoder_node = NULL, *endpoint_node = NULL; + struct drm_device *drm = &arcpgu->drm; struct resource *res; int ret; - arcpgu = devm_kzalloc(&pdev->dev, sizeof(*arcpgu), GFP_KERNEL); - if (arcpgu == NULL) - return -ENOMEM; - drm->dev_private = arcpgu; arcpgu->clk = devm_clk_get(drm->dev, "pxlclk"); @@ -162,30 +158,28 @@ static struct drm_driver arcpgu_drm_driver = { static int arcpgu_probe(struct platform_device *pdev) { - struct drm_device *drm; + struct arcpgu_drm_private *arcpgu; int ret; - drm = drm_dev_alloc(&arcpgu_drm_driver, &pdev->dev); - if (IS_ERR(drm)) - return PTR_ERR(drm); + arcpgu = devm_drm_dev_alloc(&pdev->dev, &arcpgu_drm_driver, + struct arcpgu_drm_private, drm); + if (IS_ERR(arcpgu)) + return PTR_ERR(arcpgu); - ret = arcpgu_load(drm); + ret = arcpgu_load(arcpgu); if (ret) - goto err_unref; + return ret; - ret = drm_dev_register(drm, 0); + ret = drm_dev_register(&arcpgu->drm, 0); if (ret) goto err_unload; - drm_fbdev_generic_setup(drm, 16); + drm_fbdev_generic_setup(&arcpgu->drm, 16); return 0; err_unload: - arcpgu_unload(drm); - -err_unref: - drm_dev_put(drm); + arcpgu_unload(&arcpgu->drm); return ret; } @@ -196,7 +190,6 @@ static int arcpgu_remove(struct platform_device *pdev) drm_dev_unregister(drm); arcpgu_unload(drm); - drm_dev_put(drm); return 0; } -- 2.29.2

3 18

[PATCH] drm/shmem-helpers: vunmap: Don't put pages for dma-buf
by Noralf Trønnes 25 Feb '21

25 Feb '21

dma-buf importing was reworked in commit 7d2cd72a9aa3 ("drm/shmem-helpers: Simplify dma-buf importing"). Before that commit drm_gem_shmem_prime_import_sg_table() did set ->pages_use_count=1 and drm_gem_shmem_vunmap_locked() could call drm_gem_shmem_put_pages() unconditionally. Now without the use count set, put pages is called also on dma-bufs. Fix this by only putting pages if it's not imported. Fixes: 7d2cd72a9aa3 ("drm/shmem-helpers: Simplify dma-buf importing") Cc: Daniel Vetter <daniel.vetter(a)ffwll.ch> Cc: Thomas Zimmermann <tzimmermann(a)suse.de> Signed-off-by: Noralf Trønnes <noralf(a)tronnes.org> --- drivers/gpu/drm/drm_gem_shmem_helper.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/drivers/gpu/drm/drm_gem_shmem_helper.c b/drivers/gpu/drm/drm_gem_shmem_helper.c index 9825c378dfa6..c8a6547a1757 100644 --- a/drivers/gpu/drm/drm_gem_shmem_helper.c +++ b/drivers/gpu/drm/drm_gem_shmem_helper.c @@ -357,13 +357,14 @@ static void drm_gem_shmem_vunmap_locked(struct drm_gem_shmem_object *shmem, if (--shmem->vmap_use_count > 0) return; - if (obj->import_attach) + if (obj->import_attach) { dma_buf_vunmap(obj->import_attach->dmabuf, map); - else + } else { vunmap(shmem->vaddr); + drm_gem_shmem_put_pages(shmem); + } shmem->vaddr = NULL; - drm_gem_shmem_put_pages(shmem); } /* -- 2.23.0

4 8

[PULL] drm-misc-next-fixes
by Maarten Lankhorst 25 Feb '21

25 Feb '21

drm-misc-next-fixes-2021-02-25: drm-misc-next tasty fixes for v5.12: - Cherry pick of drm-misc-fixes pull: "here's this week's PR for drm-misc-fixes. One of the patches is a memory leak; the rest is for hardware issues." - Fix dt bindings for dp connector. - Fix build error in atyfb. - Improve error handling for dma-buf heaps. - Make vblank timestamp more correct, by recording timestamp to be set when signaling. The following changes since commit e2183fb135a7f62d317aa1c61eb3d1919080edba: Revert "drm/scheduler: Job timeout handler returns status (v3)" (2021-02-10 15:26:00 +0100) are available in the Git repository at: git://anongit.freedesktop.org/drm/drm-misc tags/drm-misc-next-fixes-2021-02-25 for you to fetch changes up to d922d58fedcd98ba625e89b625a98e222b090b10: drm/panel: kd35t133: allow using non-continuous dsi clock (2021-02-25 10:18:45 +0100) ---------------------------------------------------------------- drm-misc-next tasty fixes for v5.12: - Cherry pick of drm-misc-fixes pull: "here's this week's PR for drm-misc-fixes. One of the patches is a memory leak; the rest is for hardware issues." - Fix dt bindings for dp connector. - Fix build error in atyfb. - Improve error handling for dma-buf heaps. - Make vblank timestamp more correct, by recording timestamp to be set when signaling. ---------------------------------------------------------------- Alyssa Rosenzweig (1): drm/rockchip: Require the YTR modifier for AFBC Bjorn Andersson (1): dt-bindings: dp-connector: Drop maxItems from -supply Heiko Stuebner (1): drm/panel: kd35t133: allow using non-continuous dsi clock John Stultz (2): dma-buf: system_heap: Make sure to return an error if we abort dma-buf: heaps: Rework heap allocation hooks to return struct dma_buf instead of fd Randy Dunlap (1): fbdev: atyfb: add stubs for aty_{ld,st}_lcd() Veera Sundaram Sankaran (2): dma-fence: allow signaling drivers to set fence timestamp drm/drm_vblank: set the dma-fence timestamp during send_vblank_event xinhui pan (1): drm/ttm: Fix a memory leak .../bindings/display/connector/dp-connector.yaml | 1 - drivers/dma-buf/dma-fence.c | 70 +++++++++++++++++++--- drivers/dma-buf/dma-heap.c | 14 ++++- drivers/dma-buf/heaps/cma_heap.c | 22 +++---- drivers/dma-buf/heaps/system_heap.c | 25 ++++---- drivers/gpu/drm/drm_file.c | 68 +++++++++++++++++---- drivers/gpu/drm/drm_vblank.c | 9 ++- drivers/gpu/drm/panel/panel-elida-kd35t133.c | 3 +- drivers/gpu/drm/rockchip/rockchip_drm_vop.h | 11 ++++ drivers/gpu/drm/ttm/ttm_bo.c | 9 ++- drivers/video/fbdev/aty/atyfb_base.c | 9 +++ include/drm/drm_file.h | 3 + include/linux/dma-fence.h | 3 + include/linux/dma-heap.h | 12 ++-- 14 files changed, 197 insertions(+), 62 deletions(-)

1 0

[PULL] drm-misc-fixes
by Thomas Zimmermann 25 Feb '21

25 Feb '21

Hi Dave and Daniel, here's this week's PR for drm-misc-fixes. One of the patches is a memory leak; the rest is for hardware issues. Best regards Thomas drm-misc-fixes-2021-02-24: * drm/panel: kd35t133: Work with non-continuous DSI clock * drm/rockchip: Require YTR modifier for AFBC * drm/ttm: Fix a memory leak in error handling The following changes since commit f40ddce88593482919761f74910f42f4b84c004b: Linux 5.11 (2021-02-14 14:32:24 -0800) are available in the Git repository at: git://anongit.freedesktop.org/drm/drm-misc tags/drm-misc-fixes-2021-02-24 for you to fetch changes up to 54dab3a718f7094532daf7d25cd14121a0e00e34: drm/panel: kd35t133: allow using non-continuous dsi clock (2021-02-23 22:44:58 +0100) ---------------------------------------------------------------- * drm/panel: kd35t133: Work with non-continuous DSI clock * drm/rockchip: Require YTR modifier for AFBC * drm/ttm: Fix a memory leak in error handling ---------------------------------------------------------------- Alyssa Rosenzweig (1): drm/rockchip: Require the YTR modifier for AFBC Heiko Stuebner (1): drm/panel: kd35t133: allow using non-continuous dsi clock xinhui pan (1): drm/ttm: Fix a memory leak drivers/gpu/drm/panel/panel-elida-kd35t133.c | 3 ++- drivers/gpu/drm/rockchip/rockchip_drm_vop.h | 11 +++++++++++ drivers/gpu/drm/ttm/ttm_bo.c | 9 ++++++--- 3 files changed, 19 insertions(+), 4 deletions(-) -- Thomas Zimmermann Graphics Driver Developer SUSE Software Solutions Germany GmbH Maxfeldstr. 5, 90409 Nürnberg, Germany (HRB 36809, AG Nürnberg) Geschäftsführer: Felix Imendörffer

4 4

[PATCH v1] drm/nouveau: delete unused variables "dev" in nouveau_ttm_tt_populate()
by Luo Jiaxing 25 Feb '21

25 Feb '21

After switch to new allocator, dev is no longer used here and should be deleted. Fixes: 461619f5c324 ("drm/nouveau: switch to new allocator") Signed-off-by: Luo Jiaxing <luojiaxing(a)huawei.com> --- drivers/gpu/drm/nouveau/nouveau_bo.c | 4 ---- 1 file changed, 4 deletions(-) diff --git a/drivers/gpu/drm/nouveau/nouveau_bo.c b/drivers/gpu/drm/nouveau/nouveau_bo.c index 2375711..0b94f8a 100644 --- a/drivers/gpu/drm/nouveau/nouveau_bo.c +++ b/drivers/gpu/drm/nouveau/nouveau_bo.c @@ -1251,7 +1251,6 @@ nouveau_ttm_tt_populate(struct ttm_bo_device *bdev, { struct ttm_tt *ttm_dma = (void *)ttm; struct nouveau_drm *drm; - struct device *dev; bool slave = !!(ttm->page_flags & TTM_PAGE_FLAG_SG); if (ttm_tt_is_populated(ttm)) @@ -1264,7 +1263,6 @@ nouveau_ttm_tt_populate(struct ttm_bo_device *bdev, } drm = nouveau_bdev(bdev); - dev = drm->dev->dev; return ttm_pool_alloc(&drm->ttm.bdev.pool, ttm, ctx); } @@ -1274,14 +1272,12 @@ nouveau_ttm_tt_unpopulate(struct ttm_bo_device *bdev, struct ttm_tt *ttm) { struct nouveau_drm *drm; - struct device *dev; bool slave = !!(ttm->page_flags & TTM_PAGE_FLAG_SG); if (slave) return; drm = nouveau_bdev(bdev); - dev = drm->dev->dev; return ttm_pool_free(&drm->ttm.bdev.pool, ttm); } -- 2.7.4

1 0

[PATCH v1] drm/nouveau/kms: delete an useless function call in nouveau_framebuffer_new()
by Luo Jiaxing 25 Feb '21

25 Feb '21

nouveau_framebuffer_new() call drm_format_info_plane_width() to get a width of plane, but width is not used then, so it's a useless function call here. Signed-off-by: Luo Jiaxing <luojiaxing(a)huawei.com> --- drivers/gpu/drm/nouveau/nouveau_display.c | 5 +---- 1 file changed, 1 insertion(+), 4 deletions(-) diff --git a/drivers/gpu/drm/nouveau/nouveau_display.c b/drivers/gpu/drm/nouveau/nouveau_display.c index 17831ee..7d3b1f8 100644 --- a/drivers/gpu/drm/nouveau/nouveau_display.c +++ b/drivers/gpu/drm/nouveau/nouveau_display.c @@ -306,7 +306,7 @@ nouveau_framebuffer_new(struct drm_device *dev, struct nouveau_bo *nvbo = nouveau_gem_object(gem); struct drm_framebuffer *fb; const struct drm_format_info *info; - unsigned int width, height, i; + unsigned int height, i; uint32_t tile_mode; uint8_t kind; int ret; @@ -346,9 +346,6 @@ nouveau_framebuffer_new(struct drm_device *dev, info = drm_get_format_info(dev, mode_cmd); for (i = 0; i < info->num_planes; i++) { - width = drm_format_info_plane_width(info, - mode_cmd->width, - i); height = drm_format_info_plane_height(info, mode_cmd->height, i); -- 2.7.4

1 0

[PATCH] drm/msm/dsi_pll_7nm: Solve TODO for multiplier frac_bits assignment
by Dmitry Baryshkov 25 Feb '21

25 Feb '21

The number of fractional registers bits is known and already set in the frac_bits variable of the dsi_pll_config struct here in 7nm: remove the TODO by simply using that variable. This is a copy of 196145eb1af1 ("drm/msm/dsi_pll_10nm: Solve TODO for multiplier frac_bits assignment"). Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov(a)linaro.org> --- drivers/gpu/drm/msm/dsi/pll/dsi_pll_7nm.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/msm/dsi/pll/dsi_pll_7nm.c b/drivers/gpu/drm/msm/dsi/pll/dsi_pll_7nm.c index c1f6708367ae..0458eda15114 100644 --- a/drivers/gpu/drm/msm/dsi/pll/dsi_pll_7nm.c +++ b/drivers/gpu/drm/msm/dsi/pll/dsi_pll_7nm.c @@ -509,6 +509,7 @@ static unsigned long dsi_pll_7nm_vco_recalc_rate(struct clk_hw *hw, { struct msm_dsi_pll *pll = hw_clk_to_pll(hw); struct dsi_pll_7nm *pll_7nm = to_pll_7nm(pll); + struct dsi_pll_config *config = &pll_7nm->pll_configuration; void __iomem *base = pll_7nm->mmio; u64 ref_clk = pll_7nm->vco_ref_clk_rate; u64 vco_rate = 0x0; @@ -529,9 +530,8 @@ static unsigned long dsi_pll_7nm_vco_recalc_rate(struct clk_hw *hw, /* * TODO: * 1. Assumes prescaler is disabled - * 2. Multiplier is 2^18. it should be 2^(num_of_frac_bits) */ - multiplier = 1 << 18; + multiplier = 1 << config->frac_bits; pll_freq = dec * (ref_clk * 2); tmp64 = (ref_clk * 2 * frac); pll_freq += div_u64(tmp64, multiplier); -- 2.30.0

2 1

[PATCH v1 1/2] drm/nouveau/kms/nv50-: Remove several set but not used variables "ret" in disp.c
by Luo Jiaxing 25 Feb '21

25 Feb '21

Fixes the following warning when using W=1 to build kernel: drivers/gpu/drm/nouveau/dispnv50/disp.c: In function ‘nv50_mstm_cleanup’: drivers/gpu/drm/nouveau/dispnv50/disp.c:1389:6: warning: variable ‘ret’ set but not used [-Wunused-but-set-variable] 1389 | int ret; | ^~~ drivers/gpu/drm/nouveau/dispnv50/disp.c: In function ‘nv50_mstm_prepare’: drivers/gpu/drm/nouveau/dispnv50/disp.c:1413:6: warning: variable ‘ret’ set but not used [-Wunused-but-set-variable] 1413 | int ret; | ^~~ Signed-off-by: Luo Jiaxing <luojiaxing(a)huawei.com> --- drivers/gpu/drm/nouveau/dispnv50/disp.c | 8 +++----- 1 file changed, 3 insertions(+), 5 deletions(-) diff --git a/drivers/gpu/drm/nouveau/dispnv50/disp.c b/drivers/gpu/drm/nouveau/dispnv50/disp.c index 196612a..828f48d 100644 --- a/drivers/gpu/drm/nouveau/dispnv50/disp.c +++ b/drivers/gpu/drm/nouveau/dispnv50/disp.c @@ -1386,12 +1386,11 @@ nv50_mstm_cleanup(struct nv50_mstm *mstm) { struct nouveau_drm *drm = nouveau_drm(mstm->outp->base.base.dev); struct drm_encoder *encoder; - int ret; NV_ATOMIC(drm, "%s: mstm cleanup\n", mstm->outp->base.base.name); - ret = drm_dp_check_act_status(&mstm->mgr); + drm_dp_check_act_status(&mstm->mgr); - ret = drm_dp_update_payload_part2(&mstm->mgr); + drm_dp_update_payload_part2(&mstm->mgr); drm_for_each_encoder(encoder, mstm->outp->base.base.dev) { if (encoder->encoder_type == DRM_MODE_ENCODER_DPMST) { @@ -1410,10 +1409,9 @@ nv50_mstm_prepare(struct nv50_mstm *mstm) { struct nouveau_drm *drm = nouveau_drm(mstm->outp->base.base.dev); struct drm_encoder *encoder; - int ret; NV_ATOMIC(drm, "%s: mstm prepare\n", mstm->outp->base.base.name); - ret = drm_dp_update_payload_part1(&mstm->mgr); + drm_dp_update_payload_part1(&mstm->mgr); drm_for_each_encoder(encoder, mstm->outp->base.base.dev) { if (encoder->encoder_type == DRM_MODE_ENCODER_DPMST) { -- 2.7.4

1 0

[PATCH] drm/msm/dsi_pll_7nm: Fix variable usage for pll_lockdet_rate
by Dmitry Baryshkov 25 Feb '21

25 Feb '21

The PLL_LOCKDET_RATE_1 was being programmed with a hardcoded value directly, but the same value was also being specified in the dsi_pll_regs struct pll_lockdet_rate variable: let's use it! Based on 362cadf34b9f ("drm/msm/dsi_pll_10nm: Fix variable usage for pll_lockdet_rate") Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov(a)linaro.org> --- drivers/gpu/drm/msm/dsi/pll/dsi_pll_7nm.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/msm/dsi/pll/dsi_pll_7nm.c b/drivers/gpu/drm/msm/dsi/pll/dsi_pll_7nm.c index 0458eda15114..e29b3bfd63d1 100644 --- a/drivers/gpu/drm/msm/dsi/pll/dsi_pll_7nm.c +++ b/drivers/gpu/drm/msm/dsi/pll/dsi_pll_7nm.c @@ -325,7 +325,7 @@ static void dsi_pll_commit(struct dsi_pll_7nm *pll) pll_write(base + REG_DSI_7nm_PHY_PLL_FRAC_DIV_START_LOW_1, reg->frac_div_start_low); pll_write(base + REG_DSI_7nm_PHY_PLL_FRAC_DIV_START_MID_1, reg->frac_div_start_mid); pll_write(base + REG_DSI_7nm_PHY_PLL_FRAC_DIV_START_HIGH_1, reg->frac_div_start_high); - pll_write(base + REG_DSI_7nm_PHY_PLL_PLL_LOCKDET_RATE_1, 0x40); + pll_write(base + REG_DSI_7nm_PHY_PLL_PLL_LOCKDET_RATE_1, reg->pll_lockdet_rate); pll_write(base + REG_DSI_7nm_PHY_PLL_PLL_LOCK_DELAY, 0x06); pll_write(base + REG_DSI_7nm_PHY_PLL_CMODE_1, 0x10); /* TODO: 0x00 for CPHY */ pll_write(base + REG_DSI_7nm_PHY_PLL_CLOCK_INVERTERS, reg->pll_clock_inverters); -- 2.30.0

2 1

[PATCH v2] drm/nouveau/pmu: fix timeout on GP108
by Diego Viola 25 Feb '21

25 Feb '21

This code times out on GP108, probably because the BIOS puts it into a bad state. Since we reset the PMU on driver load anyway, we are at no risk from missing a response from it since we are not waiting for one to begin with. Signed-off-by: Diego Viola <diego.viola(a)gmail.com> --- drivers/gpu/drm/nouveau/nvkm/subdev/pmu/base.c | 6 +----- 1 file changed, 1 insertion(+), 5 deletions(-) diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/pmu/base.c b/drivers/gpu/drm/nouveau/nvkm/subdev/pmu/base.c index a0fe607c9c07..5c802f2d00cb 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/pmu/base.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/pmu/base.c @@ -102,12 +102,8 @@ nvkm_pmu_reset(struct nvkm_pmu *pmu) if (!pmu->func->enabled(pmu)) return 0; - /* Inhibit interrupts, and wait for idle. */ + /* Inhibit interrupts. */ nvkm_wr32(device, 0x10a014, 0x0000ffff); - nvkm_msec(device, 2000, - if (!nvkm_rd32(device, 0x10a04c)) - break; - ); /* Reset. */ if (pmu->func->reset) -- 2.30.1

3 5

← Newer
1
...
21
22
23
24
25
26
27
...
69
Older →

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

dri-devel February 2021