dri-devel April 2019

dri-devel@lists.freedesktop.org

307 participants
2114 discussions

[PATCH 1/2] drm/edid: Add a EDID edp panel quirk for forcing max lane count
by Manasi Navare 03 Apr '19

03 Apr '19

For certain eDP 1.4 panels, we need to use max lane count for the link training to succeed. This patch adds a EDID quirk for such eDP panels using their vendor ID and product ID to force using max lane count in the driver. Cc: Clint Taylor <Clinton.A.Taylor(a)intel.com> Cc: Ville Syrjälä <ville.syrjala(a)linux.intel.com> Tested-by: Albert Astals Cid <aacid(a)kde.org> Tested-by: Emanuele Panigati <ilpanich(a)gmail.com> Tested-by: Ralgor <ralgorfdb(a)compuspex.org> Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=109959 Signed-off-by: Manasi Navare <manasi.d.navare(a)intel.com> --- drivers/gpu/drm/drm_edid.c | 10 ++++++++++ include/drm/drm_connector.h | 5 +++++ 2 files changed, 15 insertions(+) diff --git a/drivers/gpu/drm/drm_edid.c b/drivers/gpu/drm/drm_edid.c index 2c22ea446075..fbc661806484 100644 --- a/drivers/gpu/drm/drm_edid.c +++ b/drivers/gpu/drm/drm_edid.c @@ -82,6 +82,8 @@ #define EDID_QUIRK_FORCE_10BPC (1 << 11) /* Non desktop display (i.e. HMD) */ #define EDID_QUIRK_NON_DESKTOP (1 << 12) +/* Force max lane count */ +#define EDID_QUIRK_FORCE_MAX_LANE_COUNT (1 << 13) struct detailed_mode_closure { struct drm_connector *connector; @@ -189,6 +191,10 @@ static const struct edid_quirk { /* OSVR HDK and HDK2 VR Headsets */ { "SVR", 0x1019, EDID_QUIRK_NON_DESKTOP }, + + /* SHP eDP 1.4 panel only works with max lane count */ + { "SHP", 0x149a, EDID_QUIRK_FORCE_MAX_LANE_COUNT }, + { "SHP", 0x148e, EDID_QUIRK_FORCE_MAX_LANE_COUNT }, }; /* @@ -4463,6 +4469,7 @@ drm_reset_display_info(struct drm_connector *connector) memset(&info->hdmi, 0, sizeof(info->hdmi)); info->non_desktop = 0; + info->force_max_lane_count = 0; } u32 drm_add_display_info(struct drm_connector *connector, const struct edid *edid) @@ -4744,6 +4751,9 @@ int drm_add_edid_modes(struct drm_connector *connector, struct edid *edid) if (quirks & EDID_QUIRK_FORCE_12BPC) connector->display_info.bpc = 12; + if (quirks & EDID_QUIRK_FORCE_MAX_LANE_COUNT) + connector->display_info.force_max_lane_count = true; + return num_modes; } EXPORT_SYMBOL(drm_add_edid_modes); diff --git a/include/drm/drm_connector.h b/include/drm/drm_connector.h index 02a131202add..45436d40ffe3 100644 --- a/include/drm/drm_connector.h +++ b/include/drm/drm_connector.h @@ -457,6 +457,11 @@ struct drm_display_info { * @non_desktop: Non desktop display (HMD). */ bool non_desktop; + + /** + * @force_max_lane_count: Link training requires max lane count to pass + */ + bool force_max_lane_count; }; int drm_display_info_set_bus_formats(struct drm_display_info *info, -- 2.19.1

4 10

[PATCH v2 1/2] drm/vc4: Use common helpers for debugfs setup by the driver components.
by Eric Anholt 03 Apr '19

03 Apr '19

The global list of all debugfs entries for the driver was painful: the list couldn't see into the components' structs, so each component had its own debugs show function to find the component, then find the regset and dump it. The components also had to be careful to check that they were actually registered in vc4 before dereferencing themselves, in case they weren't probed on a particular platform. They routinely failed at that. Instead, we can have the components add their debugfs callbacks to a little list in vc4 to be registered at drm_dev_register() time, which gets vc4_debugfs.c out of the business of knowing the whole list of components. Thanks to this change, dsi0 (if it existed) would register its node. v2: Rebase on hvs_underrun addition. Signed-off-by: Eric Anholt <eric(a)anholt.net> --- drivers/gpu/drm/vc4/vc4_bo.c | 6 +-- drivers/gpu/drm/vc4/vc4_crtc.c | 33 +++---------- drivers/gpu/drm/vc4/vc4_debugfs.c | 82 ++++++++++++++++++++++++------- drivers/gpu/drm/vc4/vc4_dpi.c | 20 +------- drivers/gpu/drm/vc4/vc4_drv.c | 1 + drivers/gpu/drm/vc4/vc4_drv.h | 41 +++++++++++----- drivers/gpu/drm/vc4/vc4_dsi.c | 24 ++------- drivers/gpu/drm/vc4/vc4_hdmi.c | 6 +-- drivers/gpu/drm/vc4/vc4_hvs.c | 20 ++------ drivers/gpu/drm/vc4/vc4_txp.c | 20 +------- drivers/gpu/drm/vc4/vc4_v3d.c | 19 ++----- drivers/gpu/drm/vc4/vc4_vec.c | 20 +------- 12 files changed, 126 insertions(+), 166 deletions(-) diff --git a/drivers/gpu/drm/vc4/vc4_bo.c b/drivers/gpu/drm/vc4/vc4_bo.c index 276ea9c550c0..88ebd681d7eb 100644 --- a/drivers/gpu/drm/vc4/vc4_bo.c +++ b/drivers/gpu/drm/vc4/vc4_bo.c @@ -66,8 +66,7 @@ static void vc4_bo_stats_print(struct drm_printer *p, struct vc4_dev *vc4) mutex_unlock(&vc4->purgeable.lock); } -#ifdef CONFIG_DEBUG_FS -int vc4_bo_stats_debugfs(struct seq_file *m, void *unused) +static int vc4_bo_stats_debugfs(struct seq_file *m, void *unused) { struct drm_info_node *node = (struct drm_info_node *)m->private; struct drm_device *dev = node->minor->dev; @@ -78,7 +77,6 @@ int vc4_bo_stats_debugfs(struct seq_file *m, void *unused) return 0; } -#endif /* Takes ownership of *name and returns the appropriate slot for it in * the bo_labels[] array, extending it as necessary. @@ -1005,6 +1003,8 @@ int vc4_bo_cache_init(struct drm_device *dev) mutex_init(&vc4->bo_lock); + vc4_debugfs_add_file(dev, "bo_stats", vc4_bo_stats_debugfs, NULL); + INIT_LIST_HEAD(&vc4->bo_cache.time_list); INIT_WORK(&vc4->bo_cache.time_work, vc4_bo_cache_time_work); diff --git a/drivers/gpu/drm/vc4/vc4_crtc.c b/drivers/gpu/drm/vc4/vc4_crtc.c index 6e564783ba94..b2891ca0e7f4 100644 --- a/drivers/gpu/drm/vc4/vc4_crtc.c +++ b/drivers/gpu/drm/vc4/vc4_crtc.c @@ -84,33 +84,6 @@ static const struct debugfs_reg32 crtc_regs[] = { VC4_REG32(PV_HACT_ACT), }; -#ifdef CONFIG_DEBUG_FS -int vc4_crtc_debugfs_regs(struct seq_file *m, void *unused) -{ - struct drm_info_node *node = (struct drm_info_node *)m->private; - struct drm_device *dev = node->minor->dev; - int crtc_index = (uintptr_t)node->info_ent->data; - struct drm_printer p = drm_seq_file_printer(m); - struct drm_crtc *crtc; - struct vc4_crtc *vc4_crtc; - int i; - - i = 0; - list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) { - if (i == crtc_index) - break; - i++; - } - if (!crtc) - return 0; - vc4_crtc = to_vc4_crtc(crtc); - - drm_print_regset32(&p, &vc4_crtc->regset); - - return 0; -} -#endif - bool vc4_crtc_get_scanoutpos(struct drm_device *dev, unsigned int crtc_id, bool in_vblank_irq, int *vpos, int *hpos, ktime_t *stime, ktime_t *etime, @@ -1070,6 +1043,7 @@ static const struct drm_crtc_helper_funcs vc4_crtc_helper_funcs = { static const struct vc4_crtc_data pv0_data = { .hvs_channel = 0, + .debugfs_name = "crtc0_regs", .encoder_types = { [PV_CONTROL_CLK_SELECT_DSI] = VC4_ENCODER_TYPE_DSI0, [PV_CONTROL_CLK_SELECT_DPI_SMI_HDMI] = VC4_ENCODER_TYPE_DPI, @@ -1078,6 +1052,7 @@ static const struct vc4_crtc_data pv0_data = { static const struct vc4_crtc_data pv1_data = { .hvs_channel = 2, + .debugfs_name = "crtc1_regs", .encoder_types = { [PV_CONTROL_CLK_SELECT_DSI] = VC4_ENCODER_TYPE_DSI1, [PV_CONTROL_CLK_SELECT_DPI_SMI_HDMI] = VC4_ENCODER_TYPE_SMI, @@ -1086,6 +1061,7 @@ static const struct vc4_crtc_data pv1_data = { static const struct vc4_crtc_data pv2_data = { .hvs_channel = 1, + .debugfs_name = "crtc2_regs", .encoder_types = { [PV_CONTROL_CLK_SELECT_DPI_SMI_HDMI] = VC4_ENCODER_TYPE_HDMI, [PV_CONTROL_CLK_SELECT_VEC] = VC4_ENCODER_TYPE_VEC, @@ -1247,6 +1223,9 @@ static int vc4_crtc_bind(struct device *dev, struct device *master, void *data) platform_set_drvdata(pdev, vc4_crtc); + vc4_debugfs_add_regset32(drm, vc4_crtc->data->debugfs_name, + &vc4_crtc->regset); + return 0; err_destroy_planes: diff --git a/drivers/gpu/drm/vc4/vc4_debugfs.c b/drivers/gpu/drm/vc4/vc4_debugfs.c index 59cdad89f844..69df84bdf904 100644 --- a/drivers/gpu/drm/vc4/vc4_debugfs.c +++ b/drivers/gpu/drm/vc4/vc4_debugfs.c @@ -15,28 +15,20 @@ #include "vc4_drv.h" #include "vc4_regs.h" -static const struct drm_info_list vc4_debugfs_list[] = { - {"bo_stats", vc4_bo_stats_debugfs, 0}, - {"dpi_regs", vc4_dpi_debugfs_regs, 0}, - {"dsi1_regs", vc4_dsi_debugfs_regs, 0, (void *)(uintptr_t)1}, - {"hdmi_regs", vc4_hdmi_debugfs_regs, 0}, - {"vec_regs", vc4_vec_debugfs_regs, 0}, - {"txp_regs", vc4_txp_debugfs_regs, 0}, - {"hvs_regs", vc4_hvs_debugfs_regs, 0}, - {"hvs_underrun", vc4_hvs_debugfs_underrun, 0}, - {"crtc0_regs", vc4_crtc_debugfs_regs, 0, (void *)(uintptr_t)0}, - {"crtc1_regs", vc4_crtc_debugfs_regs, 0, (void *)(uintptr_t)1}, - {"crtc2_regs", vc4_crtc_debugfs_regs, 0, (void *)(uintptr_t)2}, - {"v3d_ident", vc4_v3d_debugfs_ident, 0}, - {"v3d_regs", vc4_v3d_debugfs_regs, 0}, +struct vc4_debugfs_info_entry { + struct list_head link; + struct drm_info_list info; }; -#define VC4_DEBUGFS_ENTRIES ARRAY_SIZE(vc4_debugfs_list) - +/** + * Called at drm_dev_register() time on each of the minors registered + * by the DRM device, to attach the debugfs files. + */ int vc4_debugfs_init(struct drm_minor *minor) { struct vc4_dev *vc4 = to_vc4_dev(minor->dev); + struct vc4_debugfs_info_entry *entry; struct dentry *dentry; dentry = debugfs_create_bool("hvs_load_tracker", S_IRUGO | S_IWUSR, @@ -45,6 +37,60 @@ vc4_debugfs_init(struct drm_minor *minor) if (!dentry) return -ENOMEM; - return drm_debugfs_create_files(vc4_debugfs_list, VC4_DEBUGFS_ENTRIES, - minor->debugfs_root, minor); + list_for_each_entry(entry, &vc4->debugfs_list, link) { + int ret = drm_debugfs_create_files(&entry->info, 1, + minor->debugfs_root, minor); + + if (ret) + return ret; + } + + return 0; +} + +int vc4_debugfs_regset32(struct seq_file *m, void *unused) +{ + struct drm_info_node *node = (struct drm_info_node *)m->private; + struct debugfs_regset32 *regset = node->info_ent->data; + struct drm_printer p = drm_seq_file_printer(m); + + drm_print_regset32(&p, regset); + + return 0; +} + +/** + * Registers a debugfs file with a callback function for a vc4 component. + * + * This is like drm_debugfs_create_files(), but that can only be + * called a given DRM minor, while the various VC4 components want to + * register their debugfs files during the component bind process. We + * track the request and delay it to be called on each minor during + * vc4_debugfs_init(). + */ +void vc4_debugfs_add_file(struct drm_device *dev, + const char *name, + int (*show)(struct seq_file*, void*), + void *data) +{ + struct vc4_dev *vc4 = to_vc4_dev(dev); + + struct vc4_debugfs_info_entry *entry = + devm_kzalloc(dev->dev, sizeof(*entry), GFP_KERNEL); + + if (!entry) + return; + + entry->info.name = name; + entry->info.show = show; + entry->info.data = data; + + list_add(&entry->link, &vc4->debugfs_list); +} + +void vc4_debugfs_add_regset32(struct drm_device *drm, + const char *name, + struct debugfs_regset32 *regset) +{ + vc4_debugfs_add_file(drm, name, vc4_debugfs_regset32, regset); } diff --git a/drivers/gpu/drm/vc4/vc4_dpi.c b/drivers/gpu/drm/vc4/vc4_dpi.c index 8be2264f496a..34f90ca8f479 100644 --- a/drivers/gpu/drm/vc4/vc4_dpi.c +++ b/drivers/gpu/drm/vc4/vc4_dpi.c @@ -125,24 +125,6 @@ static const struct debugfs_reg32 dpi_regs[] = { VC4_REG32(DPI_ID), }; -#ifdef CONFIG_DEBUG_FS -int vc4_dpi_debugfs_regs(struct seq_file *m, void *unused) -{ - struct drm_info_node *node = (struct drm_info_node *)m->private; - struct drm_device *dev = node->minor->dev; - struct vc4_dev *vc4 = to_vc4_dev(dev); - struct vc4_dpi *dpi = vc4->dpi; - struct drm_printer p = drm_seq_file_printer(m); - - if (!dpi) - return 0; - - drm_print_regset32(&p, &dpi->regset); - - return 0; -} -#endif - static const struct drm_encoder_funcs vc4_dpi_encoder_funcs = { .destroy = drm_encoder_cleanup, }; @@ -349,6 +331,8 @@ static int vc4_dpi_bind(struct device *dev, struct device *master, void *data) vc4->dpi = dpi; + vc4_debugfs_add_regset32(drm, "dpi_regs", &dpi->regset); + return 0; err_destroy_encoder: diff --git a/drivers/gpu/drm/vc4/vc4_drv.c b/drivers/gpu/drm/vc4/vc4_drv.c index f2d0fb3aa729..3d6fccea5080 100644 --- a/drivers/gpu/drm/vc4/vc4_drv.c +++ b/drivers/gpu/drm/vc4/vc4_drv.c @@ -262,6 +262,7 @@ static int vc4_drm_bind(struct device *dev) platform_set_drvdata(pdev, drm); vc4->dev = drm; drm->dev_private = vc4; + INIT_LIST_HEAD(&vc4->debugfs_list); ret = vc4_bo_cache_init(drm); if (ret) diff --git a/drivers/gpu/drm/vc4/vc4_drv.h b/drivers/gpu/drm/vc4/vc4_drv.h index f6c9de75d465..be85656024fa 100644 --- a/drivers/gpu/drm/vc4/vc4_drv.h +++ b/drivers/gpu/drm/vc4/vc4_drv.h @@ -210,7 +210,12 @@ struct vc4_dev { struct drm_modeset_lock ctm_state_lock; struct drm_private_obj ctm_manager; - struct drm_private_obj load_tracker; + struct drm_private_obj load_tracker; + + /* List of vc4_debugfs_info_entry for adding to debugfs once + * the minor is available (after drm_dev_register()). + */ + struct list_head debugfs_list; }; static inline struct vc4_dev * @@ -429,6 +434,7 @@ struct vc4_crtc_data { int hvs_channel; enum vc4_encoder_type encoder_types[4]; + const char *debugfs_name; }; struct vc4_crtc { @@ -715,7 +721,6 @@ struct drm_gem_object *vc4_prime_import_sg_table(struct drm_device *dev, void *vc4_prime_vmap(struct drm_gem_object *obj); int vc4_bo_cache_init(struct drm_device *dev); void vc4_bo_cache_destroy(struct drm_device *dev); -int vc4_bo_stats_debugfs(struct seq_file *m, void *arg); int vc4_bo_inc_usecnt(struct vc4_bo *bo); void vc4_bo_dec_usecnt(struct vc4_bo *bo); void vc4_bo_add_to_purgeable_pool(struct vc4_bo *bo); @@ -723,7 +728,6 @@ void vc4_bo_remove_from_purgeable_pool(struct vc4_bo *bo); /* vc4_crtc.c */ extern struct platform_driver vc4_crtc_driver; -int vc4_crtc_debugfs_regs(struct seq_file *m, void *arg); bool vc4_crtc_get_scanoutpos(struct drm_device *dev, unsigned int crtc_id, bool in_vblank_irq, int *vpos, int *hpos, ktime_t *stime, ktime_t *etime, @@ -736,17 +740,37 @@ void vc4_crtc_get_margins(struct drm_crtc_state *state, /* vc4_debugfs.c */ int vc4_debugfs_init(struct drm_minor *minor); +#ifdef CONFIG_DEBUG_FS +void vc4_debugfs_add_file(struct drm_device *drm, + const char *filename, + int (*show)(struct seq_file*, void*), + void *data); +void vc4_debugfs_add_regset32(struct drm_device *drm, + const char *filename, + struct debugfs_regset32 *regset); +#else +static inline void vc4_debugfs_add_file(struct drm_device *drm, + const char *filename, + int (*show)(struct seq_file*, void*), + void *data) +{ +} + +static inline void vc4_debugfs_add_regset32(struct drm_device *drm, + const char *filename, + struct debugfs_regset32 *regset) +{ +} +#endif /* vc4_drv.c */ void __iomem *vc4_ioremap_regs(struct platform_device *dev, int index); /* vc4_dpi.c */ extern struct platform_driver vc4_dpi_driver; -int vc4_dpi_debugfs_regs(struct seq_file *m, void *unused); /* vc4_dsi.c */ extern struct platform_driver vc4_dsi_driver; -int vc4_dsi_debugfs_regs(struct seq_file *m, void *unused); /* vc4_fence.c */ extern const struct dma_fence_ops vc4_fence_ops; @@ -774,15 +798,12 @@ int vc4_gem_madvise_ioctl(struct drm_device *dev, void *data, /* vc4_hdmi.c */ extern struct platform_driver vc4_hdmi_driver; -int vc4_hdmi_debugfs_regs(struct seq_file *m, void *unused); /* vc4_vec.c */ extern struct platform_driver vc4_vec_driver; -int vc4_vec_debugfs_regs(struct seq_file *m, void *unused); /* vc4_txp.c */ extern struct platform_driver vc4_txp_driver; -int vc4_txp_debugfs_regs(struct seq_file *m, void *unused); /* vc4_irq.c */ irqreturn_t vc4_irq(int irq, void *arg); @@ -794,8 +815,6 @@ void vc4_irq_reset(struct drm_device *dev); /* vc4_hvs.c */ extern struct platform_driver vc4_hvs_driver; void vc4_hvs_dump_state(struct drm_device *dev); -int vc4_hvs_debugfs_regs(struct seq_file *m, void *unused); -int vc4_hvs_debugfs_underrun(struct seq_file *m, void *unused); void vc4_hvs_unmask_underrun(struct drm_device *dev, int channel); void vc4_hvs_mask_underrun(struct drm_device *dev, int channel); @@ -812,8 +831,6 @@ void vc4_plane_async_set_fb(struct drm_plane *plane, /* vc4_v3d.c */ extern struct platform_driver vc4_v3d_driver; -int vc4_v3d_debugfs_ident(struct seq_file *m, void *unused); -int vc4_v3d_debugfs_regs(struct seq_file *m, void *unused); int vc4_v3d_get_bin_slot(struct vc4_dev *vc4); int vc4_v3d_pm_get(struct vc4_dev *vc4); void vc4_v3d_pm_put(struct vc4_dev *vc4); diff --git a/drivers/gpu/drm/vc4/vc4_dsi.c b/drivers/gpu/drm/vc4/vc4_dsi.c index 806cfaa2a6a7..9412709067f5 100644 --- a/drivers/gpu/drm/vc4/vc4_dsi.c +++ b/drivers/gpu/drm/vc4/vc4_dsi.c @@ -657,25 +657,6 @@ static const struct debugfs_reg32 dsi1_regs[] = { VC4_REG32(DSI1_ID), }; -#ifdef CONFIG_DEBUG_FS -int vc4_dsi_debugfs_regs(struct seq_file *m, void *unused) -{ - struct drm_info_node *node = (struct drm_info_node *)m->private; - struct drm_device *drm = node->minor->dev; - struct vc4_dev *vc4 = to_vc4_dev(drm); - int dsi_index = (uintptr_t)node->info_ent->data; - struct vc4_dsi *dsi = (dsi_index == 1 ? vc4->dsi1 : NULL); - struct drm_printer p = drm_seq_file_printer(m); - - if (!dsi) - return 0; - - drm_print_regset32(&p, &dsi->regset); - - return 0; -} -#endif - static void vc4_dsi_encoder_destroy(struct drm_encoder *encoder) { drm_encoder_cleanup(encoder); @@ -1637,6 +1618,11 @@ static int vc4_dsi_bind(struct device *dev, struct device *master, void *data) */ dsi->encoder->bridge = NULL; + if (dsi->port == 0) + vc4_debugfs_add_regset32(drm, "dsi0_regs", &dsi->regset); + else + vc4_debugfs_add_regset32(drm, "dsi1_regs", &dsi->regset); + pm_runtime_enable(dev); return 0; diff --git a/drivers/gpu/drm/vc4/vc4_hdmi.c b/drivers/gpu/drm/vc4/vc4_hdmi.c index 38c9172cfe52..99fc8569e0f5 100644 --- a/drivers/gpu/drm/vc4/vc4_hdmi.c +++ b/drivers/gpu/drm/vc4/vc4_hdmi.c @@ -187,8 +187,7 @@ static const struct debugfs_reg32 hd_regs[] = { VC4_REG32(VC4_HD_FRAME_COUNT), }; -#ifdef CONFIG_DEBUG_FS -int vc4_hdmi_debugfs_regs(struct seq_file *m, void *unused) +static int vc4_hdmi_debugfs_regs(struct seq_file *m, void *unused) { struct drm_info_node *node = (struct drm_info_node *)m->private; struct drm_device *dev = node->minor->dev; @@ -201,7 +200,6 @@ int vc4_hdmi_debugfs_regs(struct seq_file *m, void *unused) return 0; } -#endif /* CONFIG_DEBUG_FS */ static enum drm_connector_status vc4_hdmi_connector_detect(struct drm_connector *connector, bool force) @@ -1432,6 +1430,8 @@ static int vc4_hdmi_bind(struct device *dev, struct device *master, void *data) if (ret) goto err_destroy_encoder; + vc4_debugfs_add_file(drm, "hdmi_regs", vc4_hdmi_debugfs_regs, hdmi); + return 0; #ifdef CONFIG_DRM_VC4_HDMI_CEC diff --git a/drivers/gpu/drm/vc4/vc4_hvs.c b/drivers/gpu/drm/vc4/vc4_hvs.c index c463453f941b..f746e9a7a88c 100644 --- a/drivers/gpu/drm/vc4/vc4_hvs.c +++ b/drivers/gpu/drm/vc4/vc4_hvs.c @@ -80,20 +80,7 @@ void vc4_hvs_dump_state(struct drm_device *dev) } } -#ifdef CONFIG_DEBUG_FS -int vc4_hvs_debugfs_regs(struct seq_file *m, void *unused) -{ - struct drm_info_node *node = (struct drm_info_node *)m->private; - struct drm_device *dev = node->minor->dev; - struct vc4_dev *vc4 = to_vc4_dev(dev); - struct drm_printer p = drm_seq_file_printer(m); - - drm_print_regset32(&p, &vc4->hvs->regset); - - return 0; -} - -int vc4_hvs_debugfs_underrun(struct seq_file *m, void *data) +static int vc4_hvs_debugfs_underrun(struct seq_file *m, void *data) { struct drm_info_node *node = m->private; struct drm_device *dev = node->minor->dev; @@ -104,7 +91,6 @@ int vc4_hvs_debugfs_underrun(struct seq_file *m, void *data) return 0; } -#endif /* The filter kernel is composed of dwords each containing 3 9-bit * signed integers packed next to each other. @@ -316,6 +302,10 @@ static int vc4_hvs_bind(struct device *dev, struct device *master, void *data) if (ret) return ret; + vc4_debugfs_add_regset32(drm, "hvs_regs", &hvs->regset); + vc4_debugfs_add_file(drm, "hvs_underrun", vc4_hvs_debugfs_underrun, + NULL); + return 0; } diff --git a/drivers/gpu/drm/vc4/vc4_txp.c b/drivers/gpu/drm/vc4/vc4_txp.c index c8f80064c179..1220f4a5aac4 100644 --- a/drivers/gpu/drm/vc4/vc4_txp.c +++ b/drivers/gpu/drm/vc4/vc4_txp.c @@ -169,24 +169,6 @@ static const struct debugfs_reg32 txp_regs[] = { VC4_REG32(TXP_PROGRESS), }; -#ifdef CONFIG_DEBUG_FS -int vc4_txp_debugfs_regs(struct seq_file *m, void *unused) -{ - struct drm_info_node *node = (struct drm_info_node *)m->private; - struct drm_device *dev = node->minor->dev; - struct vc4_dev *vc4 = to_vc4_dev(dev); - struct vc4_txp *txp = vc4->txp; - struct drm_printer p = drm_seq_file_printer(m); - - if (!txp) - return 0; - - drm_print_regset32(&p, &txp->regset); - - return 0; -} -#endif - static int vc4_txp_connector_get_modes(struct drm_connector *connector) { struct drm_device *dev = connector->dev; @@ -424,6 +406,8 @@ static int vc4_txp_bind(struct device *dev, struct device *master, void *data) dev_set_drvdata(dev, txp); vc4->txp = txp; + vc4_debugfs_add_regset32(drm, "txp_regs", &txp->regset); + return 0; } diff --git a/drivers/gpu/drm/vc4/vc4_v3d.c b/drivers/gpu/drm/vc4/vc4_v3d.c index 27c70eb52405..36e6c7086ecf 100644 --- a/drivers/gpu/drm/vc4/vc4_v3d.c +++ b/drivers/gpu/drm/vc4/vc4_v3d.c @@ -22,7 +22,6 @@ #include "vc4_drv.h" #include "vc4_regs.h" -#ifdef CONFIG_DEBUG_FS static const struct debugfs_reg32 v3d_regs[] = { VC4_REG32(V3D_IDENT0), VC4_REG32(V3D_IDENT1), @@ -104,19 +103,7 @@ static const struct debugfs_reg32 v3d_regs[] = { VC4_REG32(V3D_ERRSTAT), }; -int vc4_v3d_debugfs_regs(struct seq_file *m, void *unused) -{ - struct drm_info_node *node = (struct drm_info_node *)m->private; - struct drm_device *dev = node->minor->dev; - struct vc4_dev *vc4 = to_vc4_dev(dev); - struct drm_printer p = drm_seq_file_printer(m); - - drm_print_regset32(&p, &vc4->v3d->regset); - - return 0; -} - -int vc4_v3d_debugfs_ident(struct seq_file *m, void *unused) +static int vc4_v3d_debugfs_ident(struct seq_file *m, void *unused) { struct drm_info_node *node = (struct drm_info_node *)m->private; struct drm_device *dev = node->minor->dev; @@ -141,7 +128,6 @@ int vc4_v3d_debugfs_ident(struct seq_file *m, void *unused) return 0; } -#endif /* CONFIG_DEBUG_FS */ /** * Wraps pm_runtime_get_sync() in a refcount, so that we can reliably @@ -442,6 +428,9 @@ static int vc4_v3d_bind(struct device *dev, struct device *master, void *data) pm_runtime_set_autosuspend_delay(dev, 40); /* a little over 2 frames. */ pm_runtime_enable(dev); + vc4_debugfs_add_file(drm, "v3d_ident", vc4_v3d_debugfs_ident, NULL); + vc4_debugfs_add_regset32(drm, "v3d_regs", &v3d->regset); + return 0; } diff --git a/drivers/gpu/drm/vc4/vc4_vec.c b/drivers/gpu/drm/vc4/vc4_vec.c index 83227d2440aa..0a27e48fab31 100644 --- a/drivers/gpu/drm/vc4/vc4_vec.c +++ b/drivers/gpu/drm/vc4/vc4_vec.c @@ -252,24 +252,6 @@ static const struct debugfs_reg32 vec_regs[] = { VC4_REG32(VEC_DAC_MISC), }; -#ifdef CONFIG_DEBUG_FS -int vc4_vec_debugfs_regs(struct seq_file *m, void *unused) -{ - struct drm_info_node *node = (struct drm_info_node *)m->private; - struct drm_device *dev = node->minor->dev; - struct vc4_dev *vc4 = to_vc4_dev(dev); - struct vc4_vec *vec = vc4->vec; - struct drm_printer p = drm_seq_file_printer(m); - - if (!vec) - return 0; - - drm_print_regset32(&p, &vec->regset); - - return 0; -} -#endif - static void vc4_vec_ntsc_mode_set(struct vc4_vec *vec) { VEC_WRITE(VEC_CONFIG0, VEC_CONFIG0_NTSC_STD | VEC_CONFIG0_PDEN); @@ -609,6 +591,8 @@ static int vc4_vec_bind(struct device *dev, struct device *master, void *data) vc4->vec = vec; + vc4_debugfs_add_regset32(drm, "vec_regs", &vec->regset); + return 0; err_destroy_encoder: -- 2.20.1

2 6

Re: remove NULL struct device support in the DMA API
by Russell King - ARM Linux admin 03 Apr '19

03 Apr '19

On Thu, Mar 21, 2019 at 03:52:28PM -0700, Christoph Hellwig wrote: > We still have a few drivers which pass a NULL struct device pointer > to DMA API functions, which generally is a bad idea as the API > implementations rely on the device not only for ops selection, but > also the dma mask and various other attributes, and many implementations > have been broken for NULL device support for a while. I think I must be missing something, but... My understanding is that ISA DMA is normally limited to 24 bits of address - indeed, the x86 version only programs 24 bits of DMA address. Looking through this series, it appears that the conversions mean that the DMA mask for ISA becomes the full all-ones DMA mask, which would of course lead to memory corruption if only 24 bits of the address end up being programmed into the hardware. Maybe you could say why you think this series is safe in regard to ISA DMA? > > This series removes the few remaning users that weren't picked up in > the last merge window and then removes core support for this "feature". > > A git tree is also available at: > > git://git.infradead.org/users/hch/misc.git dma-remove-NULL-dev-support > > Gitweb: > > http://git.infradead.org/users/hch/misc.git/shortlog/refs/heads/dma-remove-… > -- RMK's Patch system: https://www.armlinux.org.uk/developer/patches/ FTTC broadband for 0.8mile line in suburbia: sync at 12.1Mbps down 622kbps up According to speedtest.net: 11.9Mbps down 500kbps up

1 0

[radeon-alex:drm-next-5.2-wip 21/42] drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c:798:7: error: implicit declaration of function 'hmm_vma_range_done'; did you mean '__range_ok'?
by kbuild test robot 03 Apr '19

03 Apr '19

tree: git://people.freedesktop.org/~agd5f/linux.git drm-next-5.2-wip head: 5666aea3ea494d4dd96df8f092cab32dbeeac321 commit: 95db8d6001df8966e3370a66c9f358925fbcc890 [21/42] drm/amdgpu: replace get_user_pages with HMM mirror helpers config: s390-allyesconfig (attached as .config) compiler: s390x-linux-gnu-gcc (Debian 7.2.0-11) 7.2.0 reproduce: wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross chmod +x ~/bin/make.cross git checkout 95db8d6001df8966e3370a66c9f358925fbcc890 # save the attached .config to linux build tree GCC_VERSION=7.2.0 make.cross ARCH=s390 Note: the radeon-alex/drm-next-5.2-wip HEAD 5666aea3ea494d4dd96df8f092cab32dbeeac321 builds fine. It only hurts bisectibility. All errors (new ones prefixed by >>): drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c:713:19: error: field 'range' has incomplete type struct hmm_range range; ^~~~~ drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c: In function 'amdgpu_ttm_tt_get_user_pages': drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c:738:7: error: dereferencing pointer to incomplete type 'struct hmm_range' range->vma = find_vma(mm, gtt->userptr); ^~ drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c:756:32: error: 'HMM_PFN_VALID' undeclared (first use in this function); did you mean 'DP_SU_VALID'? range->pfns[0] = range->flags[HMM_PFN_VALID]; ^~~~~~~~~~~~~ DP_SU_VALID drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c:756:32: note: each undeclared identifier is reported only once for each function it appears in drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c:758:22: error: 'HMM_PFN_WRITE' undeclared (first use in this function); did you mean 'HMM_PFN_VALID'? 0 : range->flags[HMM_PFN_WRITE]; ^~~~~~~~~~~~~ HMM_PFN_VALID drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c:763:6: error: implicit declaration of function 'hmm_vma_fault'; did you mean 'hmm_mm_init'? [-Werror=implicit-function-declaration] r = hmm_vma_fault(range, true); ^~~~~~~~~~~~~ hmm_mm_init drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c:770:14: error: implicit declaration of function 'hmm_pfn_to_page'; did you mean '__pfn_to_page'? [-Werror=implicit-function-declaration] pages[i] = hmm_pfn_to_page(range, range->pfns[i]); ^~~~~~~~~~~~~~~ __pfn_to_page drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c: In function 'amdgpu_ttm_tt_get_user_pages_done': >> drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c:798:7: error: implicit declaration of function 'hmm_vma_range_done'; did you mean '__range_ok'? [-Werror=implicit-function-declaration] r = hmm_vma_range_done(&gtt->range); ^~~~~~~~~~~~~~~~~~ __range_ok cc1: some warnings being treated as errors vim +798 drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c 715 716 /** 717 * amdgpu_ttm_tt_get_user_pages - get device accessible pages that back user 718 * memory and start HMM tracking CPU page table update 719 * 720 * Calling function must call amdgpu_ttm_tt_userptr_range_done() once and only 721 * once afterwards to stop HMM tracking 722 */ 723 int amdgpu_ttm_tt_get_user_pages(struct ttm_tt *ttm, struct page **pages) 724 { 725 struct amdgpu_ttm_tt *gtt = (void *)ttm; 726 struct mm_struct *mm = gtt->usertask->mm; 727 unsigned long end = gtt->userptr + ttm->num_pages * PAGE_SIZE; 728 struct hmm_range *range = &gtt->range; 729 int r = 0, i; 730 731 if (!mm) /* Happens during process shutdown */ 732 return -ESRCH; 733 734 amdgpu_hmm_init_range(range); 735 736 down_read(&mm->mmap_sem); 737 738 range->vma = find_vma(mm, gtt->userptr); 739 if (!range_in_vma(range->vma, gtt->userptr, end)) 740 r = -EFAULT; 741 else if ((gtt->userflags & AMDGPU_GEM_USERPTR_ANONONLY) && 742 range->vma->vm_file) 743 r = -EPERM; 744 if (r) 745 goto out; 746 747 range->pfns = kvmalloc_array(ttm->num_pages, sizeof(uint64_t), 748 GFP_KERNEL); 749 if (range->pfns == NULL) { 750 r = -ENOMEM; 751 goto out; 752 } 753 range->start = gtt->userptr; 754 range->end = end; 755 756 range->pfns[0] = range->flags[HMM_PFN_VALID]; 757 range->pfns[0] |= amdgpu_ttm_tt_is_readonly(ttm) ? 758 0 : range->flags[HMM_PFN_WRITE]; 759 for (i = 1; i < ttm->num_pages; i++) 760 range->pfns[i] = range->pfns[0]; 761 762 /* This may trigger page table update */ > 763 r = hmm_vma_fault(range, true); 764 if (r) 765 goto out_free_pfns; 766 767 up_read(&mm->mmap_sem); 768 769 for (i = 0; i < ttm->num_pages; i++) 770 pages[i] = hmm_pfn_to_page(range, range->pfns[i]); 771 772 return 0; 773 774 out_free_pfns: 775 kvfree(range->pfns); 776 range->pfns = NULL; 777 out: 778 up_read(&mm->mmap_sem); 779 return r; 780 } 781 782 /** 783 * amdgpu_ttm_tt_userptr_range_done - stop HMM track the CPU page table change 784 * Check if the pages backing this ttm range have been invalidated 785 * 786 * Returns: true if pages are still valid 787 */ 788 bool amdgpu_ttm_tt_get_user_pages_done(struct ttm_tt *ttm) 789 { 790 struct amdgpu_ttm_tt *gtt = (void *)ttm; 791 bool r = false; 792 793 if (!gtt || !gtt->userptr) 794 return false; 795 796 WARN_ONCE(!gtt->range.pfns, "No user pages to check\n"); 797 if (gtt->range.pfns) { > 798 r = hmm_vma_range_done(&gtt->range); 799 kvfree(gtt->range.pfns); 800 gtt->range.pfns = NULL; 801 } 802 803 return r; 804 } 805 --- 0-DAY kernel test infrastructure Open Source Technology Center https://lists.01.org/pipermail/kbuild-all Intel Corporation

1 0

[radeon-alex:drm-next-5.2-wip 19/42] drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c:87:20: error: field 'mirror' has incomplete type
by kbuild test robot 03 Apr '19

03 Apr '19

tree: git://people.freedesktop.org/~agd5f/linux.git drm-next-5.2-wip head: 5666aea3ea494d4dd96df8f092cab32dbeeac321 commit: ba5896bd6a1a175b21609c6c81dea9813cbf296c [19/42] drm/amdgpu: use HMM callback to replace mmu notifier config: xtensa-allyesconfig (attached as .config) compiler: xtensa-linux-gcc (GCC) 8.1.0 reproduce: wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross chmod +x ~/bin/make.cross git checkout ba5896bd6a1a175b21609c6c81dea9813cbf296c # save the attached .config to linux build tree GCC_VERSION=8.1.0 make.cross ARCH=xtensa Note: the radeon-alex/drm-next-5.2-wip HEAD 5666aea3ea494d4dd96df8f092cab32dbeeac321 builds fine. It only hurts bisectibility. All error/warnings (new ones prefixed by >>): >> drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c:87:20: error: field 'mirror' has incomplete type struct hmm_mirror mirror; ^~~~~~ drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c: In function 'amdgpu_mn_destroy': >> drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c:131:2: error: implicit declaration of function 'hmm_mirror_unregister'; did you mean 'drm_dp_aux_unregister'? [-Werror=implicit-function-declaration] hmm_mirror_unregister(&amn->mirror); ^~~~~~~~~~~~~~~~~~~~~ drm_dp_aux_unregister In file included from include/linux/firmware.h:6, from drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c:46: drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c: In function 'amdgpu_hmm_mirror_release': >> include/linux/kernel.h:979:32: error: dereferencing pointer to incomplete type 'struct hmm_mirror' BUILD_BUG_ON_MSG(!__same_type(*(ptr), ((type *)0)->member) && \ ^~~~~~ include/linux/compiler.h:324:9: note: in definition of macro '__compiletime_assert' if (!(condition)) \ ^~~~~~~~~ include/linux/compiler.h:344:2: note: in expansion of macro '_compiletime_assert' _compiletime_assert(condition, msg, __compiletime_assert_, __LINE__) ^~~~~~~~~~~~~~~~~~~ include/linux/build_bug.h:39:37: note: in expansion of macro 'compiletime_assert' #define BUILD_BUG_ON_MSG(cond, msg) compiletime_assert(!(cond), msg) ^~~~~~~~~~~~~~~~~~ include/linux/kernel.h:979:2: note: in expansion of macro 'BUILD_BUG_ON_MSG' BUILD_BUG_ON_MSG(!__same_type(*(ptr), ((type *)0)->member) && \ ^~~~~~~~~~~~~~~~ include/linux/kernel.h:979:20: note: in expansion of macro '__same_type' BUILD_BUG_ON_MSG(!__same_type(*(ptr), ((type *)0)->member) && \ ^~~~~~~~~~~ drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c:144:26: note: in expansion of macro 'container_of' struct amdgpu_mn *amn = container_of(mirror, struct amdgpu_mn, mirror); ^~~~~~~~~~~~ drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c: At top level: >> drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c:238:17: warning: 'struct hmm_update' declared inside parameter list will not be visible outside of this definition or declaration const struct hmm_update *update) ^~~~~~~~~~ drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c: In function 'amdgpu_mn_sync_pagetables_gfx': >> drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c:241:30: error: dereferencing pointer to incomplete type 'const struct hmm_update' unsigned long start = update->start; ^~ drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c: At top level: drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c:286:17: warning: 'struct hmm_update' declared inside parameter list will not be visible outside of this definition or declaration const struct hmm_update *update) ^~~~~~~~~~ drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c: In function 'amdgpu_mn_sync_pagetables_hsa': drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c:289:30: error: dereferencing pointer to incomplete type 'const struct hmm_update' unsigned long start = update->start; ^~ drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c: At top level: >> drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c:333:30: error: array type has incomplete element type 'struct hmm_mirror_ops' static struct hmm_mirror_ops amdgpu_hmm_mirror_ops[] = { ^~~~~~~~~~~~~~~~~~~~~ >> drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c:334:3: error: array index in non-array initializer [AMDGPU_MN_TYPE_GFX] = { ^~~~~~~~~~~~~~~~~~ drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c:334:3: note: (near initialization for 'amdgpu_hmm_mirror_ops') >> drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c:335:3: error: field name not in record or union initializer .sync_cpu_device_pagetables = amdgpu_mn_sync_pagetables_gfx, ^ drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c:335:3: note: (near initialization for 'amdgpu_hmm_mirror_ops') drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c:336:3: error: field name not in record or union initializer .release = amdgpu_hmm_mirror_release ^ drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c:336:3: note: (near initialization for 'amdgpu_hmm_mirror_ops') drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c:338:3: error: array index in non-array initializer [AMDGPU_MN_TYPE_HSA] = { ^~~~~~~~~~~~~~~~~~ drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c:338:3: note: (near initialization for 'amdgpu_hmm_mirror_ops') drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c:339:3: error: field name not in record or union initializer .sync_cpu_device_pagetables = amdgpu_mn_sync_pagetables_hsa, ^ drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c:339:3: note: (near initialization for 'amdgpu_hmm_mirror_ops') drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c:340:3: error: field name not in record or union initializer .release = amdgpu_hmm_mirror_release ^ drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c:340:3: note: (near initialization for 'amdgpu_hmm_mirror_ops') drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c: In function 'amdgpu_mn_get': >> drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c:383:6: error: implicit declaration of function 'hmm_mirror_register'; did you mean 'drm_dp_aux_register'? [-Werror=implicit-function-declaration] r = hmm_mirror_register(&amn->mirror, mm); ^~~~~~~~~~~~~~~~~~~ drm_dp_aux_register At top level: drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c:333:30: warning: 'amdgpu_hmm_mirror_ops' defined but not used [-Wunused-variable] static struct hmm_mirror_ops amdgpu_hmm_mirror_ops[] = { ^~~~~~~~~~~~~~~~~~~~~ cc1: some warnings being treated as errors vim +/mirror +87 drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c 55 56 /** 57 * struct amdgpu_mn 58 * 59 * @adev: amdgpu device pointer 60 * @mm: process address space 61 * @type: type of MMU notifier 62 * @work: destruction work item 63 * @node: hash table node to find structure by adev and mn 64 * @lock: rw semaphore protecting the notifier nodes 65 * @objects: interval tree containing amdgpu_mn_nodes 66 * @mirror: HMM mirror function support 67 * 68 * Data for each amdgpu device and process address space. 69 */ 70 struct amdgpu_mn { 71 /* constant after initialisation */ 72 struct amdgpu_device *adev; 73 struct mm_struct *mm; 74 enum amdgpu_mn_type type; 75 76 /* only used on destruction */ 77 struct work_struct work; 78 79 /* protected by adev->mn_lock */ 80 struct hlist_node node; 81 82 /* objects protected by lock */ 83 struct rw_semaphore lock; 84 struct rb_root_cached objects; 85 86 /* HMM mirror */ > 87 struct hmm_mirror mirror; 88 }; 89 90 /** 91 * struct amdgpu_mn_node 92 * 93 * @it: interval node defining start-last of the affected address range 94 * @bos: list of all BOs in the affected address range 95 * 96 * Manages all BOs which are affected of a certain range of address space. 97 */ 98 struct amdgpu_mn_node { 99 struct interval_tree_node it; 100 struct list_head bos; 101 }; 102 103 /** 104 * amdgpu_mn_destroy - destroy the HMM mirror 105 * 106 * @work: previously sheduled work item 107 * 108 * Lazy destroys the notifier from a work item 109 */ 110 static void amdgpu_mn_destroy(struct work_struct *work) 111 { 112 struct amdgpu_mn *amn = container_of(work, struct amdgpu_mn, work); 113 struct amdgpu_device *adev = amn->adev; 114 struct amdgpu_mn_node *node, *next_node; 115 struct amdgpu_bo *bo, *next_bo; 116 117 mutex_lock(&adev->mn_lock); 118 down_write(&amn->lock); 119 hash_del(&amn->node); 120 rbtree_postorder_for_each_entry_safe(node, next_node, 121 &amn->objects.rb_root, it.rb) { 122 list_for_each_entry_safe(bo, next_bo, &node->bos, mn_list) { 123 bo->mn = NULL; 124 list_del_init(&bo->mn_list); 125 } 126 kfree(node); 127 } 128 up_write(&amn->lock); 129 mutex_unlock(&adev->mn_lock); 130 > 131 hmm_mirror_unregister(&amn->mirror); 132 kfree(amn); 133 } 134 135 /** 136 * amdgpu_hmm_mirror_release - callback to notify about mm destruction 137 * 138 * @mirror: the HMM mirror (mm) this callback is about 139 * 140 * Shedule a work item to lazy destroy HMM mirror. 141 */ 142 static void amdgpu_hmm_mirror_release(struct hmm_mirror *mirror) 143 { > 144 struct amdgpu_mn *amn = container_of(mirror, struct amdgpu_mn, mirror); 145 146 INIT_WORK(&amn->work, amdgpu_mn_destroy); 147 schedule_work(&amn->work); 148 } 149 150 /** 151 * amdgpu_mn_lock - take the write side lock for this notifier 152 * 153 * @mn: our notifier 154 */ 155 void amdgpu_mn_lock(struct amdgpu_mn *mn) 156 { 157 if (mn) 158 down_write(&mn->lock); 159 } 160 161 /** 162 * amdgpu_mn_unlock - drop the write side lock for this notifier 163 * 164 * @mn: our notifier 165 */ 166 void amdgpu_mn_unlock(struct amdgpu_mn *mn) 167 { 168 if (mn) 169 up_write(&mn->lock); 170 } 171 172 /** 173 * amdgpu_mn_read_lock - take the read side lock for this notifier 174 * 175 * @amn: our notifier 176 */ 177 static int amdgpu_mn_read_lock(struct amdgpu_mn *amn, bool blockable) 178 { 179 if (blockable) 180 down_read(&amn->lock); 181 else if (!down_read_trylock(&amn->lock)) 182 return -EAGAIN; 183 184 return 0; 185 } 186 187 /** 188 * amdgpu_mn_read_unlock - drop the read side lock for this notifier 189 * 190 * @amn: our notifier 191 */ 192 static void amdgpu_mn_read_unlock(struct amdgpu_mn *amn) 193 { 194 up_read(&amn->lock); 195 } 196 197 /** 198 * amdgpu_mn_invalidate_node - unmap all BOs of a node 199 * 200 * @node: the node with the BOs to unmap 201 * @start: start of address range affected 202 * @end: end of address range affected 203 * 204 * Block for operations on BOs to finish and mark pages as accessed and 205 * potentially dirty. 206 */ 207 static void amdgpu_mn_invalidate_node(struct amdgpu_mn_node *node, 208 unsigned long start, 209 unsigned long end) 210 { 211 struct amdgpu_bo *bo; 212 long r; 213 214 list_for_each_entry(bo, &node->bos, mn_list) { 215 216 if (!amdgpu_ttm_tt_affect_userptr(bo->tbo.ttm, start, end)) 217 continue; 218 219 r = reservation_object_wait_timeout_rcu(bo->tbo.resv, 220 true, false, MAX_SCHEDULE_TIMEOUT); 221 if (r <= 0) 222 DRM_ERROR("(%ld) failed to wait for user bo\n", r); 223 224 amdgpu_ttm_tt_mark_user_pages(bo->tbo.ttm); 225 } 226 } 227 228 /** 229 * amdgpu_mn_sync_pagetables_gfx - callback to notify about mm change 230 * 231 * @mirror: the hmm_mirror (mm) is about to update 232 * @update: the update start, end address 233 * 234 * Block for operations on BOs to finish and mark pages as accessed and 235 * potentially dirty. 236 */ 237 static int amdgpu_mn_sync_pagetables_gfx(struct hmm_mirror *mirror, > 238 const struct hmm_update *update) 239 { 240 struct amdgpu_mn *amn = container_of(mirror, struct amdgpu_mn, mirror); > 241 unsigned long start = update->start; 242 unsigned long end = update->end; 243 bool blockable = update->blockable; 244 struct interval_tree_node *it; 245 246 /* notification is exclusive, but interval is inclusive */ 247 end -= 1; 248 249 /* TODO we should be able to split locking for interval tree and 250 * amdgpu_mn_invalidate_node 251 */ 252 if (amdgpu_mn_read_lock(amn, blockable)) 253 return -EAGAIN; 254 255 it = interval_tree_iter_first(&amn->objects, start, end); 256 while (it) { 257 struct amdgpu_mn_node *node; 258 259 if (!blockable) { 260 amdgpu_mn_read_unlock(amn); 261 return -EAGAIN; 262 } 263 264 node = container_of(it, struct amdgpu_mn_node, it); 265 it = interval_tree_iter_next(it, start, end); 266 267 amdgpu_mn_invalidate_node(node, start, end); 268 } 269 270 amdgpu_mn_read_unlock(amn); 271 272 return 0; 273 } 274 275 /** 276 * amdgpu_mn_sync_pagetables_hsa - callback to notify about mm change 277 * 278 * @mirror: the hmm_mirror (mm) is about to update 279 * @update: the update start, end address 280 * 281 * We temporarily evict all BOs between start and end. This 282 * necessitates evicting all user-mode queues of the process. The BOs 283 * are restorted in amdgpu_mn_invalidate_range_end_hsa. 284 */ 285 static int amdgpu_mn_sync_pagetables_hsa(struct hmm_mirror *mirror, > 286 const struct hmm_update *update) 287 { 288 struct amdgpu_mn *amn = container_of(mirror, struct amdgpu_mn, mirror); 289 unsigned long start = update->start; 290 unsigned long end = update->end; 291 bool blockable = update->blockable; 292 struct interval_tree_node *it; 293 294 /* notification is exclusive, but interval is inclusive */ 295 end -= 1; 296 297 if (amdgpu_mn_read_lock(amn, blockable)) 298 return -EAGAIN; 299 300 it = interval_tree_iter_first(&amn->objects, start, end); 301 while (it) { 302 struct amdgpu_mn_node *node; 303 struct amdgpu_bo *bo; 304 305 if (!blockable) { 306 amdgpu_mn_read_unlock(amn); 307 return -EAGAIN; 308 } 309 310 node = container_of(it, struct amdgpu_mn_node, it); 311 it = interval_tree_iter_next(it, start, end); 312 313 list_for_each_entry(bo, &node->bos, mn_list) { 314 struct kgd_mem *mem = bo->kfd_bo; 315 316 if (amdgpu_ttm_tt_affect_userptr(bo->tbo.ttm, 317 start, end)) 318 amdgpu_amdkfd_evict_userptr(mem, amn->mm); 319 } 320 } 321 322 amdgpu_mn_read_unlock(amn); 323 324 return 0; 325 } 326 --- 0-DAY kernel test infrastructure Open Source Technology Center https://lists.01.org/pipermail/kbuild-all Intel Corporation

1 0

[PATCH] drm/panel: otm8009a: set clock to 29.70 Mhz
by Yannick Fertré 03 Apr '19

03 Apr '19

The panel does not support clock frequency over 30.74 Mhz. The clock rate has been reduced to 29.70 Mhz & new timings have been computed to get a framerate of 50fps. Signed-off-by: Yannick Fertré <yannick.fertre(a)st.com> --- drivers/gpu/drm/panel/panel-orisetech-otm8009a.c | 14 +++++++------- 1 file changed, 7 insertions(+), 7 deletions(-) diff --git a/drivers/gpu/drm/panel/panel-orisetech-otm8009a.c b/drivers/gpu/drm/panel/panel-orisetech-otm8009a.c index 87fa316..f715bbe 100644 --- a/drivers/gpu/drm/panel/panel-orisetech-otm8009a.c +++ b/drivers/gpu/drm/panel/panel-orisetech-otm8009a.c @@ -67,15 +67,15 @@ struct otm8009a { }; static const struct drm_display_mode default_mode = { - .clock = 32729, + .clock = 29700, .hdisplay = 480, - .hsync_start = 480 + 120, - .hsync_end = 480 + 120 + 63, - .htotal = 480 + 120 + 63 + 120, + .hsync_start = 480 + 98, + .hsync_end = 480 + 98 + 32, + .htotal = 480 + 98 + 32 + 98, .vdisplay = 800, - .vsync_start = 800 + 12, - .vsync_end = 800 + 12 + 12, - .vtotal = 800 + 12 + 12 + 12, + .vsync_start = 800 + 15, + .vsync_end = 800 + 15 + 10, + .vtotal = 800 + 15 + 10 + 14, .vrefresh = 50, .flags = 0, .width_mm = 52, -- 2.7.4

3 2

[PATCH] drm/panel: rm68200: no error msg if probe deferred
by Yannick Fertré 03 Apr '19

03 Apr '19

Do not print an error message if the regulator framework returns EPROBE_DEFER. Signed-off-by: Yannick Fertré <yannick.fertre(a)st.com> --- drivers/gpu/drm/panel/panel-raydium-rm68200.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/panel/panel-raydium-rm68200.c b/drivers/gpu/drm/panel/panel-raydium-rm68200.c index 7759353..1418682 100644 --- a/drivers/gpu/drm/panel/panel-raydium-rm68200.c +++ b/drivers/gpu/drm/panel/panel-raydium-rm68200.c @@ -383,7 +383,8 @@ static int rm68200_probe(struct mipi_dsi_device *dsi) ctx->supply = devm_regulator_get(dev, "power"); if (IS_ERR(ctx->supply)) { ret = PTR_ERR(ctx->supply); - dev_err(dev, "cannot get regulator: %d\n", ret); + if (ret != -EPROBE_DEFER) + dev_err(dev, "cannot get regulator: %d\n", ret); return ret; } -- 2.7.4

3 2

[PATCH] drm/panel: otm8009a: no error msg if probe deferred
by Yannick Fertré 03 Apr '19

03 Apr '19

Do not print an error message if the regulator framework returns EPROBE_DEFER. Signed-off-by: Yannick Fertré <yannick.fertre(a)st.com> --- drivers/gpu/drm/panel/panel-orisetech-otm8009a.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/panel/panel-orisetech-otm8009a.c b/drivers/gpu/drm/panel/panel-orisetech-otm8009a.c index 87fa316..aa3db38 100644 --- a/drivers/gpu/drm/panel/panel-orisetech-otm8009a.c +++ b/drivers/gpu/drm/panel/panel-orisetech-otm8009a.c @@ -433,7 +433,8 @@ static int otm8009a_probe(struct mipi_dsi_device *dsi) ctx->supply = devm_regulator_get(dev, "power"); if (IS_ERR(ctx->supply)) { ret = PTR_ERR(ctx->supply); - dev_err(dev, "failed to request regulator: %d\n", ret); + if (ret != -EPROBE_DEFER) + dev_err(dev, "failed to request regulator: %d\n", ret); return ret; } -- 2.7.4

3 2

[PATCH] drm/panel: otm8009a: add delay at the end of initialization
by Yannick Fertré 03 Apr '19

03 Apr '19

At the end of initialization, a delay is required by the panel. Without this delay, the panel could received a frame early & generate a crash of panel (black screen). Signed-off-by: Yannick Fertré <yannick.fertre(a)st.com> --- drivers/gpu/drm/panel/panel-orisetech-otm8009a.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/gpu/drm/panel/panel-orisetech-otm8009a.c b/drivers/gpu/drm/panel/panel-orisetech-otm8009a.c index 87fa316..58ccf64 100644 --- a/drivers/gpu/drm/panel/panel-orisetech-otm8009a.c +++ b/drivers/gpu/drm/panel/panel-orisetech-otm8009a.c @@ -248,6 +248,9 @@ static int otm8009a_init_sequence(struct otm8009a *ctx) /* Send Command GRAM memory write (no parameters) */ dcs_write_seq(ctx, MIPI_DCS_WRITE_MEMORY_START); + /* Wait a short while to let the panel be ready before the 1st frame */ + mdelay(10); + return 0; } -- 2.7.4

3 2

[PATCH] dt-bindings: display/panel: Add missing unit names
by Guido Günther 03 Apr '19

03 Apr '19

Some examples were missing the unit names triggering Warning (unit_address_vs_reg): .../panel: node has a reg or ranges property, but no unit name warnings when used verbatim in DTs and running dtc with W=1. Signed-off-by: Guido Günther <agx(a)sigxcpu.org> --- .../devicetree/bindings/display/panel/innolux,p079zca.txt | 2 +- .../devicetree/bindings/display/panel/innolux,p097pfg.txt | 2 +- .../devicetree/bindings/display/panel/kingdisplay,kd097d04.txt | 2 +- 3 files changed, 3 insertions(+), 3 deletions(-) diff --git a/Documentation/devicetree/bindings/display/panel/innolux,p079zca.txt b/Documentation/devicetree/bindings/display/panel/innolux,p079zca.txt index d0f55161579a..3ab8c7412cf6 100644 --- a/Documentation/devicetree/bindings/display/panel/innolux,p079zca.txt +++ b/Documentation/devicetree/bindings/display/panel/innolux,p079zca.txt @@ -12,7 +12,7 @@ Optional properties: Example: &mipi_dsi { - panel { + panel@0 { compatible = "innolux,p079zca"; reg = <0>; power-supply = <...>; diff --git a/Documentation/devicetree/bindings/display/panel/innolux,p097pfg.txt b/Documentation/devicetree/bindings/display/panel/innolux,p097pfg.txt index 595d9dfeffd3..d1cab3a8f0fb 100644 --- a/Documentation/devicetree/bindings/display/panel/innolux,p097pfg.txt +++ b/Documentation/devicetree/bindings/display/panel/innolux,p097pfg.txt @@ -13,7 +13,7 @@ Optional properties: Example: &mipi_dsi { - panel { + panel@0 { compatible = "innolux,p079zca"; reg = <0>; avdd-supply = <...>; diff --git a/Documentation/devicetree/bindings/display/panel/kingdisplay,kd097d04.txt b/Documentation/devicetree/bindings/display/panel/kingdisplay,kd097d04.txt index 164a5fa236da..cfefff688614 100644 --- a/Documentation/devicetree/bindings/display/panel/kingdisplay,kd097d04.txt +++ b/Documentation/devicetree/bindings/display/panel/kingdisplay,kd097d04.txt @@ -12,7 +12,7 @@ Optional properties: Example: &mipi_dsi { - panel { + panel@0 { compatible = "kingdisplay,kd097d04"; reg = <0>; power-supply = <...>; -- 2.20.1

3 2

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

dri-devel April 2019