dri-devel February 2021

dri-devel@lists.freedesktop.org

314 participants
684 discussions

[PATCH] Revert "drm/atomic: document and enforce rules around "spurious" EBUSY"
by Manasi Navare 11 Feb '21

11 Feb '21

This reverts commit fb6473a48b635c55d04eb94e579eede52ef39550. These additional checks added to avoid EBUSY give unnecessary WARN_ON in case of big joiner used in i915 in which case even if the modeset is requested on a single pipe, internally another consecutive pipe is stolen and used to drive half of the transcoder timings. So in this case it is expected that requested crtc and affected crtcs do not match. Hence the added WARN ON becomes irrelevant. But there is no easy solution to get the bigjoiner information here at drm level. So for now revert this until we work out a better solution. Cc: Ville Syrjala <ville.syrjala(a)linux.intel.com> Cc: Daniel Vetter <daniel.vetter(a)intel.com> Signed-off-by: Manasi Navare <manasi.d.navare(a)intel.com> --- drivers/gpu/drm/drm_atomic.c | 29 ----------------------------- 1 file changed, 29 deletions(-) diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c index b1efa9322be2..48b2262d69f6 100644 --- a/drivers/gpu/drm/drm_atomic.c +++ b/drivers/gpu/drm/drm_atomic.c @@ -320,10 +320,6 @@ EXPORT_SYMBOL(__drm_atomic_state_free); * needed. It will also grab the relevant CRTC lock to make sure that the state * is consistent. * - * WARNING: Drivers may only add new CRTC states to a @state if - * drm_atomic_state.allow_modeset is set, or if it's a driver-internal commit - * not created by userspace through an IOCTL call. - * * Returns: * * Either the allocated state or the error code encoded into the pointer. When @@ -1306,15 +1302,10 @@ int drm_atomic_check_only(struct drm_atomic_state *state) struct drm_crtc_state *new_crtc_state; struct drm_connector *conn; struct drm_connector_state *conn_state; - unsigned requested_crtc = 0; - unsigned affected_crtc = 0; int i, ret = 0; DRM_DEBUG_ATOMIC("checking %p\n", state); - for_each_new_crtc_in_state(state, crtc, new_crtc_state, i) - requested_crtc |= drm_crtc_mask(crtc); - for_each_oldnew_plane_in_state(state, plane, old_plane_state, new_plane_state, i) { ret = drm_atomic_plane_check(old_plane_state, new_plane_state); if (ret) { @@ -1362,26 +1353,6 @@ int drm_atomic_check_only(struct drm_atomic_state *state) } } - for_each_new_crtc_in_state(state, crtc, new_crtc_state, i) - affected_crtc |= drm_crtc_mask(crtc); - - /* - * For commits that allow modesets drivers can add other CRTCs to the - * atomic commit, e.g. when they need to reallocate global resources. - * This can cause spurious EBUSY, which robs compositors of a very - * effective sanity check for their drawing loop. Therefor only allow - * drivers to add unrelated CRTC states for modeset commits. - * - * FIXME: Should add affected_crtc mask to the ATOMIC IOCTL as an output - * so compositors know what's going on. - */ - if (affected_crtc != requested_crtc) { - DRM_DEBUG_ATOMIC("driver added CRTC to commit: requested 0x%x, affected 0x%0x\n", - requested_crtc, affected_crtc); - WARN(!state->allow_modeset, "adding CRTC not allowed without modesets: requested 0x%x, affected 0x%0x\n", - requested_crtc, affected_crtc); - } - return 0; } EXPORT_SYMBOL(drm_atomic_check_only); -- 2.19.1

6 7

[PATCH] drm/ttm: make sure pool pages are cleared
by Christian König 11 Feb '21

11 Feb '21

The old implementation wasn't consistend on this. But it looks like we depend on this so better bring it back. Signed-off-by: Christian König <christian.koenig(a)amd.com> Reported-and-tested-by: Mike Galbraith <efault(a)gmx.de> Fixes: d099fc8f540a ("drm/ttm: new TT backend allocation pool v3") --- drivers/gpu/drm/ttm/ttm_pool.c | 10 ++++++++++ 1 file changed, 10 insertions(+) diff --git a/drivers/gpu/drm/ttm/ttm_pool.c b/drivers/gpu/drm/ttm/ttm_pool.c index 74bf1c84b637..6e27cb1bf48b 100644 --- a/drivers/gpu/drm/ttm/ttm_pool.c +++ b/drivers/gpu/drm/ttm/ttm_pool.c @@ -33,6 +33,7 @@ #include <linux/module.h> #include <linux/dma-mapping.h> +#include <linux/highmem.h> #ifdef CONFIG_X86 #include <asm/set_memory.h> @@ -218,6 +219,15 @@ static void ttm_pool_unmap(struct ttm_pool *pool, dma_addr_t dma_addr, /* Give pages into a specific pool_type */ static void ttm_pool_type_give(struct ttm_pool_type *pt, struct page *p) { + unsigned int i, num_pages = 1 << pt->order; + + for (i = 0; i < num_pages; ++i) { + if (PageHighMem(p)) + clear_highpage(p + i); + else + clear_page(page_address(p + i)); + } + spin_lock(&pt->lock); list_add(&p->lru, &pt->pages); spin_unlock(&pt->lock); -- 2.25.1

2 4

[PULL] drm-misc-next-fixes
by Maarten Lankhorst 11 Feb '21

11 Feb '21

hi Dave, Here a pull request for drm-misc-next-fixes, I'm not 100% sure about all the CEC fixes, but seems like it wouldn't hurt. We could revert the flag that enables CEC if needed. I just picked everything that looked like fixes from drm-misc-next. drm-misc-next-fixes-2021-02-11: drm-misc-next-fixes cherry picked from drm-misc-next for v5.12: - Assorted small fixes. - Disable and remove gma3600 support. - Fix CEC for vc4/hdmi. The following changes since commit 4c3a3292730c56591472717d8c5c0faf74f6c6bb: drm/amd/display: fix unused variable warning (2021-02-05 09:49:44 +1000) are available in the Git repository at: git://anongit.freedesktop.org/drm/drm-misc tags/drm-misc-next-fixes-2021-02-11 for you to fetch changes up to e2183fb135a7f62d317aa1c61eb3d1919080edba: Revert "drm/scheduler: Job timeout handler returns status (v3)" (2021-02-10 15:26:00 +0100) ---------------------------------------------------------------- drm-misc-next-fixes cherry picked from drm-misc-next for v5.12: - Assorted small fixes. - Disable and remove gma3600 support. - Fix CEC for vc4/hdmi. ---------------------------------------------------------------- Bernard Zhao (1): drm/vc4: remove unneeded variable: "ret" Dan Carpenter (2): drm/vmwgfx/vmwgfx_drv: Fix an error path in vmw_setup_pci_resources() drm/virtio: fix an error code in virtio_gpu_init() Daniel Vetter (1): drm/todo: Add entry for moving to dma_resv_lock Dom Cobley (5): drm/vc4: hdmi: Move hdmi reset to bind drm/vc4: hdmi: Fix register offset with longer CEC messages drm/vc4: hdmi: Fix up CEC registers drm/vc4: hdmi: Restore cec physical address on reconnect drm/vc4: hdmi: Remove cec_available flag Imre Deak (1): drm/dp_mst: Don't cache EDIDs for physical ports Joe Perches (1): dma-buf: Avoid comma separated statements Joseph Schulte (1): drm: replace drm_modeset_lock_all() in drm_client_modeset_dpms_legacy() Luben Tuikov (1): drm/scheduler: Job timeout handler returns status (v3) Maarten Lankhorst (1): Revert "drm/scheduler: Job timeout handler returns status (v3)" Maxime Ripard (7): drm/vc4: hdmi: Compute the CEC clock divider from the clock rate drm/vc4: hdmi: Update the CEC clock divider on HSM rate change drm/vc4: hdmi: Introduce a CEC clock drm/vc4: hdmi: Split the interrupt handlers drm/vc4: hdmi: Support BCM2711 CEC interrupt setup drm/vc4: hdmi: Don't register the CEC adapter if there's no interrupts dt-binding: display: bcm2711-hdmi: Add CEC and hotplug interrupts Qinglang Miao (1): drm/lima: fix reference leak in lima_pm_busy Thomas Zimmermann (4): drm/gma500: Remove Medfield support drm/gma500: Drop DRM_GMA3600 config option drm/gma500: Remove CONFIG_X86 conditionals from source files drm/gma500: Remove dependency on TTM Ye Bin (1): drm/nouveau: remove set but not used variable ‘pdev’ in nouveau_bios_init Zack Rusin (1): drm/vmwgfx: Fix some memory leaks on errors .../bindings/display/brcm,bcm2711-hdmi.yaml | 20 +- Documentation/gpu/todo.rst | 19 + drivers/dma-buf/st-dma-fence.c | 7 +- drivers/gpu/drm/drm_client_modeset.c | 7 +- drivers/gpu/drm/drm_dp_mst_topology.c | 3 +- drivers/gpu/drm/gma500/Kconfig | 17 +- drivers/gpu/drm/gma500/Makefile | 37 +- drivers/gpu/drm/gma500/cdv_intel_hdmi.c | 4 - drivers/gpu/drm/gma500/mdfld_device.c | 564 ----------- drivers/gpu/drm/gma500/mdfld_dsi_dpi.c | 1017 -------------------- drivers/gpu/drm/gma500/mdfld_dsi_dpi.h | 79 -- drivers/gpu/drm/gma500/mdfld_dsi_output.c | 603 ------------ drivers/gpu/drm/gma500/mdfld_dsi_output.h | 377 -------- drivers/gpu/drm/gma500/mdfld_dsi_pkg_sender.c | 679 ------------- drivers/gpu/drm/gma500/mdfld_dsi_pkg_sender.h | 80 -- drivers/gpu/drm/gma500/mdfld_intel_display.c | 966 ------------------- drivers/gpu/drm/gma500/mdfld_output.c | 74 -- drivers/gpu/drm/gma500/mdfld_output.h | 76 -- drivers/gpu/drm/gma500/mdfld_tmd_vid.c | 197 ---- drivers/gpu/drm/gma500/mdfld_tpo_vid.c | 83 -- drivers/gpu/drm/gma500/mmu.c | 21 - drivers/gpu/drm/gma500/psb_drv.c | 16 +- drivers/gpu/drm/gma500/psb_drv.h | 66 -- drivers/gpu/drm/gma500/psb_intel_reg.h | 12 +- drivers/gpu/drm/gma500/psb_irq.c | 72 +- drivers/gpu/drm/gma500/psb_irq.h | 2 - drivers/gpu/drm/gma500/psb_reg.h | 14 - drivers/gpu/drm/gma500/tc35876x-dsi-lvds.c | 805 ---------------- drivers/gpu/drm/gma500/tc35876x-dsi-lvds.h | 38 - drivers/gpu/drm/lima/lima_sched.c | 2 +- drivers/gpu/drm/nouveau/nouveau_bios.c | 2 - drivers/gpu/drm/vc4/vc4_gem.c | 3 +- drivers/gpu/drm/vc4/vc4_hdmi.c | 223 ++++- drivers/gpu/drm/vc4/vc4_hdmi.h | 11 +- drivers/gpu/drm/vc4/vc4_hdmi_regs.h | 4 +- drivers/gpu/drm/virtio/virtgpu_kms.c | 1 + drivers/gpu/drm/vmwgfx/vmwgfx_drv.c | 14 +- 37 files changed, 261 insertions(+), 5954 deletions(-) delete mode 100644 drivers/gpu/drm/gma500/mdfld_device.c delete mode 100644 drivers/gpu/drm/gma500/mdfld_dsi_dpi.c delete mode 100644 drivers/gpu/drm/gma500/mdfld_dsi_dpi.h delete mode 100644 drivers/gpu/drm/gma500/mdfld_dsi_output.c delete mode 100644 drivers/gpu/drm/gma500/mdfld_dsi_output.h delete mode 100644 drivers/gpu/drm/gma500/mdfld_dsi_pkg_sender.c delete mode 100644 drivers/gpu/drm/gma500/mdfld_dsi_pkg_sender.h delete mode 100644 drivers/gpu/drm/gma500/mdfld_intel_display.c delete mode 100644 drivers/gpu/drm/gma500/mdfld_output.c delete mode 100644 drivers/gpu/drm/gma500/mdfld_output.h delete mode 100644 drivers/gpu/drm/gma500/mdfld_tmd_vid.c delete mode 100644 drivers/gpu/drm/gma500/mdfld_tpo_vid.c delete mode 100644 drivers/gpu/drm/gma500/tc35876x-dsi-lvds.c delete mode 100644 drivers/gpu/drm/gma500/tc35876x-dsi-lvds.h

1 0

[PATCH] PCI: Also set up legacy files only after sysfs init
by Daniel Vetter 11 Feb '21

11 Feb '21

We are already doing this for all the regular sysfs files on PCI devices, but not yet on the legacy io files on the PCI buses. Thus far no problem, but in the next patch I want to wire up iomem revoke support. That needs the vfs up and running already to make sure that iomem_get_mapping() works. Wire it up exactly like the existing code in pci_create_sysfs_dev_files(). Note that pci_remove_legacy_files() doesn't need a check since the one for pci_bus->legacy_io is sufficient. An alternative solution would be to implement a callback in sysfs to set up the address space from iomem_get_mapping() when userspace calls mmap(). This also works, but Greg didn't really like that just to work around an ordering issue when the kernel loads initially. v2: Improve commit message (Bjorn) Signed-off-by: Daniel Vetter <daniel.vetter(a)intel.com> Cc: Stephen Rothwell <sfr(a)canb.auug.org.au> Cc: Jason Gunthorpe <jgg(a)ziepe.ca> Cc: Kees Cook <keescook(a)chromium.org> Cc: Dan Williams <dan.j.williams(a)intel.com> Cc: Andrew Morton <akpm(a)linux-foundation.org> Cc: John Hubbard <jhubbard(a)nvidia.com> Cc: Jérôme Glisse <jglisse(a)redhat.com> Cc: Jan Kara <jack(a)suse.cz> Cc: Dan Williams <dan.j.williams(a)intel.com> Cc: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Cc: linux-mm(a)kvack.org Cc: linux-arm-kernel(a)lists.infradead.org Cc: linux-samsung-soc(a)vger.kernel.org Cc: linux-media(a)vger.kernel.org Cc: Bjorn Helgaas <bhelgaas(a)google.com> Cc: linux-pci(a)vger.kernel.org --- drivers/pci/pci-sysfs.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/drivers/pci/pci-sysfs.c b/drivers/pci/pci-sysfs.c index fb072f4b3176..0c45b4f7b214 100644 --- a/drivers/pci/pci-sysfs.c +++ b/drivers/pci/pci-sysfs.c @@ -927,6 +927,9 @@ void pci_create_legacy_files(struct pci_bus *b) { int error; + if (!sysfs_initialized) + return; + b->legacy_io = kcalloc(2, sizeof(struct bin_attribute), GFP_ATOMIC); if (!b->legacy_io) @@ -1448,6 +1451,7 @@ void pci_remove_sysfs_dev_files(struct pci_dev *pdev) static int __init pci_sysfs_init(void) { struct pci_dev *pdev = NULL; + struct pci_bus *pbus = NULL; int retval; sysfs_initialized = 1; @@ -1459,6 +1463,9 @@ static int __init pci_sysfs_init(void) } } + while ((pbus = pci_find_next_bus(pbus))) + pci_create_legacy_files(pbus); + return 0; } late_initcall(pci_sysfs_init); -- 2.30.0

2 3

[PATCH 1/3] drm/ttm: move swapout logic around
by Christian König 11 Feb '21

11 Feb '21

Move the iteration of the global lru into the new function ttm_global_swapout() and use that instead in drivers. Signed-off-by: Christian König <christian.koenig(a)amd.com> --- drivers/gpu/drm/ttm/ttm_bo.c | 57 ++++++++--------------------- drivers/gpu/drm/ttm/ttm_device.c | 29 +++++++++++++++ drivers/gpu/drm/ttm/ttm_tt.c | 2 +- drivers/gpu/drm/vmwgfx/ttm_memory.c | 3 +- drivers/gpu/drm/vmwgfx/vmwgfx_drv.c | 2 +- include/drm/ttm/ttm_bo_api.h | 3 +- include/drm/ttm/ttm_device.h | 2 + 7 files changed, 53 insertions(+), 45 deletions(-) diff --git a/drivers/gpu/drm/ttm/ttm_bo.c b/drivers/gpu/drm/ttm/ttm_bo.c index e38102282fd5..d33578a112b4 100644 --- a/drivers/gpu/drm/ttm/ttm_bo.c +++ b/drivers/gpu/drm/ttm/ttm_bo.c @@ -1184,56 +1184,35 @@ int ttm_bo_wait(struct ttm_buffer_object *bo, } EXPORT_SYMBOL(ttm_bo_wait); -/* - * A buffer object shrink method that tries to swap out the first - * buffer object on the bo_global::swap_lru list. - */ -int ttm_bo_swapout(struct ttm_operation_ctx *ctx, gfp_t gfp_flags) +int ttm_bo_swapout(struct ttm_buffer_object *bo, struct ttm_operation_ctx *ctx, + gfp_t gfp_flags) { struct ttm_global *glob = &ttm_glob; - struct ttm_buffer_object *bo; - int ret = -EBUSY; bool locked; - unsigned i; - - spin_lock(&glob->lru_lock); - for (i = 0; i < TTM_MAX_BO_PRIORITY; ++i) { - list_for_each_entry(bo, &glob->swap_lru[i], swap) { - if (!ttm_bo_evict_swapout_allowable(bo, ctx, &locked, - NULL)) - continue; - - if (!ttm_bo_get_unless_zero(bo)) { - if (locked) - dma_resv_unlock(bo->base.resv); - continue; - } + int ret; - ret = 0; - break; - } - if (!ret) - break; - } + if (!ttm_bo_evict_swapout_allowable(bo, ctx, &locked, NULL)) + return -EBUSY; - if (ret) { - spin_unlock(&glob->lru_lock); - return ret; + if (!ttm_bo_get_unless_zero(bo)) { + if (locked) + dma_resv_unlock(bo->base.resv); + return -EBUSY; } if (bo->deleted) { - ret = ttm_bo_cleanup_refs(bo, false, false, locked); + ttm_bo_cleanup_refs(bo, false, false, locked); ttm_bo_put(bo); - return ret; + return 0; } ttm_bo_del_from_lru(bo); + /* TODO: Cleanup the locking */ spin_unlock(&glob->lru_lock); - /** + /* * Move to system cached */ - if (bo->mem.mem_type != TTM_PL_SYSTEM) { struct ttm_operation_ctx ctx = { false, false }; struct ttm_resource evict_mem; @@ -1253,29 +1232,26 @@ int ttm_bo_swapout(struct ttm_operation_ctx *ctx, gfp_t gfp_flags) } } - /** + /* * Make sure BO is idle. */ - ret = ttm_bo_wait(bo, false, false); if (unlikely(ret != 0)) goto out; ttm_bo_unmap_virtual(bo); - /** + /* * Swap out. Buffer will be swapped in again as soon as * anyone tries to access a ttm page. */ - if (bo->bdev->funcs->swap_notify) bo->bdev->funcs->swap_notify(bo); ret = ttm_tt_swapout(bo->bdev, bo->ttm, gfp_flags); out: - /** - * + /* * Unreserve without putting on LRU to avoid swapping out an * already swapped buffer. */ @@ -1284,7 +1260,6 @@ int ttm_bo_swapout(struct ttm_operation_ctx *ctx, gfp_t gfp_flags) ttm_bo_put(bo); return ret; } -EXPORT_SYMBOL(ttm_bo_swapout); void ttm_bo_tt_destroy(struct ttm_buffer_object *bo) { diff --git a/drivers/gpu/drm/ttm/ttm_device.c b/drivers/gpu/drm/ttm/ttm_device.c index 95e1b7b1f2e6..dfc2a7e4e490 100644 --- a/drivers/gpu/drm/ttm/ttm_device.c +++ b/drivers/gpu/drm/ttm/ttm_device.c @@ -102,6 +102,35 @@ static int ttm_global_init(void) return ret; } +/** + * A buffer object shrink method that tries to swap out the first + * buffer object on the global::swap_lru list. + */ +long ttm_global_swapout(struct ttm_operation_ctx *ctx, gfp_t gfp_flags) +{ + struct ttm_global *glob = &ttm_glob; + struct ttm_buffer_object *bo; + unsigned i; + int ret; + + spin_lock(&glob->lru_lock); + for (i = 0; i < TTM_MAX_BO_PRIORITY; ++i) { + list_for_each_entry(bo, &glob->swap_lru[i], swap) { + uint32_t num_pages = bo->ttm->num_pages; + + ret = ttm_bo_swapout(bo, ctx, gfp_flags); + /* ttm_bo_swapout has dropped the lru_lock */ + if (!ret) + return num_pages; + if (ret != -EBUSY) + return ret; + } + } + spin_unlock(&glob->lru_lock); + return 0; +} +EXPORT_SYMBOL(ttm_global_swapout); + static void ttm_init_sysman(struct ttm_device *bdev) { struct ttm_resource_manager *man = &bdev->sysman; diff --git a/drivers/gpu/drm/ttm/ttm_tt.c b/drivers/gpu/drm/ttm/ttm_tt.c index 2f0833c98d2c..95b5cff25f4c 100644 --- a/drivers/gpu/drm/ttm/ttm_tt.c +++ b/drivers/gpu/drm/ttm/ttm_tt.c @@ -369,7 +369,7 @@ static unsigned long ttm_tt_shrinker_scan(struct shrinker *shrink, }; int ret; - ret = ttm_bo_swapout(&ctx, GFP_NOFS); + ret = ttm_global_swapout(&ctx, GFP_NOFS); return ret < 0 ? SHRINK_EMPTY : ret; } diff --git a/drivers/gpu/drm/vmwgfx/ttm_memory.c b/drivers/gpu/drm/vmwgfx/ttm_memory.c index e972af07d029..104b95a8c7a2 100644 --- a/drivers/gpu/drm/vmwgfx/ttm_memory.c +++ b/drivers/gpu/drm/vmwgfx/ttm_memory.c @@ -38,6 +38,7 @@ #include <drm/drm_device.h> #include <drm/drm_file.h> +#include <drm/ttm/ttm_device.h> #include "ttm_memory.h" @@ -277,7 +278,7 @@ static void ttm_shrink(struct ttm_mem_global *glob, bool from_wq, while (ttm_zones_above_swap_target(glob, from_wq, extra)) { spin_unlock(&glob->lock); - ret = ttm_bo_swapout(ctx, GFP_KERNEL); + ret = ttm_global_swapout(ctx, GFP_KERNEL); spin_lock(&glob->lock); if (unlikely(ret < 0)) break; diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c index 4efed3bf0ef9..01da355b86f3 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c @@ -1384,7 +1384,7 @@ static int vmw_pm_freeze(struct device *kdev) vmw_execbuf_release_pinned_bo(dev_priv); vmw_resource_evict_all(dev_priv); vmw_release_device_early(dev_priv); - while (ttm_bo_swapout(&ctx, GFP_KERNEL) > 0); + while (ttm_global_swapout(&ctx, GFP_KERNEL) > 0); if (dev_priv->enable_fb) vmw_fifo_resource_dec(dev_priv); if (atomic_read(&dev_priv->num_fifo_resources) != 0) { diff --git a/include/drm/ttm/ttm_bo_api.h b/include/drm/ttm/ttm_bo_api.h index 4fb523dfab32..5044ac330858 100644 --- a/include/drm/ttm/ttm_bo_api.h +++ b/include/drm/ttm/ttm_bo_api.h @@ -560,7 +560,8 @@ ssize_t ttm_bo_io(struct ttm_device *bdev, struct file *filp, const char __user *wbuf, char __user *rbuf, size_t count, loff_t *f_pos, bool write); -int ttm_bo_swapout(struct ttm_operation_ctx *ctx, gfp_t gfp_flags); +int ttm_bo_swapout(struct ttm_buffer_object *bo, struct ttm_operation_ctx *ctx, + gfp_t gfp_flags); /** * ttm_bo_uses_embedded_gem_object - check if the given bo uses the diff --git a/include/drm/ttm/ttm_device.h b/include/drm/ttm/ttm_device.h index 035bbc044a3b..6a0b267d4fe6 100644 --- a/include/drm/ttm/ttm_device.h +++ b/include/drm/ttm/ttm_device.h @@ -297,6 +297,8 @@ struct ttm_device { struct delayed_work wq; }; +long ttm_global_swapout(struct ttm_operation_ctx *ctx, gfp_t gfp_flags); + static inline struct ttm_resource_manager * ttm_manager_type(struct ttm_device *bdev, int mem_type) { -- 2.25.1

1 2

[Bug 211651] New: Dual screen not working with Nvidia GTX 1050ti in a notebook
by bugzilla-daemon＠bugzilla.kernel.org 11 Feb '21

11 Feb '21

https://bugzilla.kernel.org/show_bug.cgi?id=211651 Bug ID: 211651 Summary: Dual screen not working with Nvidia GTX 1050ti in a notebook Product: Drivers Version: 2.5 Kernel Version: 5.8.0.43 Hardware: Intel OS: Linux Tree: Mainline Status: NEW Severity: normal Priority: P1 Component: Video(DRI - non Intel) Assignee: drivers_video-dri(a)kernel-bugs.osdl.org Reporter: kernel(a)fily.com.de Regression: No Hello, I can't connect my second monitor with kernel 5.8.0.43 and driver nvidia-driver-450 and nvidia-driver-460. With 5.3.18-050318-generic this is possible. So, I think the problem is the kernel. Maybe I can do a Kernel bisect, but I don't know how to perform it... Can you help me? -- You may reply to this email to add a comment. You are receiving this mail because: You are watching the assignee of the bug.

1 4

[PULL] drm-intel-fixes
by Jani Nikula 11 Feb '21

11 Feb '21

Hi Dave & Daniel - Calming down, just a couple of Cc: stable fixes now. drm-intel-fixes-2021-02-11: drm/i915 fixes for v5.11 final: - Ensure Type-C FIA is powered when initializing - Fix overlay frontbuffer tracking BR, Jani. The following changes since commit 92bf22614b21a2706f4993b278017e437f7785b3: Linux 5.11-rc7 (2021-02-07 13:57:38 -0800) are available in the Git repository at: git://anongit.freedesktop.org/drm/drm-intel tags/drm-intel-fixes-2021-02-11 for you to fetch changes up to 5feba0e905c495a217aea9db4ea91093d8fe5dde: drm/i915: Fix overlay frontbuffer tracking (2021-02-10 11:03:56 +0200) ---------------------------------------------------------------- drm/i915 fixes for v5.11 final: - Ensure Type-C FIA is powered when initializing - Fix overlay frontbuffer tracking ---------------------------------------------------------------- Imre Deak (1): drm/i915/tgl+: Make sure TypeC FIA is powered up when initializing it Ville Syrjälä (1): drm/i915: Fix overlay frontbuffer tracking drivers/gpu/drm/i915/display/intel_overlay.c | 17 ++++--- drivers/gpu/drm/i915/display/intel_tc.c | 67 +++++++++++++++------------- 2 files changed, 45 insertions(+), 39 deletions(-) -- Jani Nikula, Intel Open Source Graphics Center

1 0

[PATCH 0/3] drm/v3d: CL/CSD job timeout fixes
by Yukimasa Sugizaki 11 Feb '21

11 Feb '21

From: Yukimasa Sugizaki <ysugi(a)idein.jp> Hi, The current V3D scheduler has two issues where CSD jobs are resubmitted regardless of the previous timed-out flag, and where the timer is not restarted for timed-out CL/CSD jobs (which we wish to continue running). The second one is due to the DRM scheduler API change and fixed in a similar way to [1]. A kernel command-line option to set the default timeout value is also added. I tested this patchset with Piglit and our CSD programs in [2]. Because it is hard to get the current upstream kernel to work on BCM2711, I used the kernel from rpi-5.8.y tree [3]. There still are problems where some Piglit tests get longer time to finish running (3610 minutes to 3650 minutes in total), and some ones result in the invalid memory read errors with unknown reasons: [17086.230959] v3d fec00000.v3d: MMU error from client CLE (4) at 0xac1000, pte invalid [17086.238722] v3d fec00000.v3d: MMU error from client CLE (4) at 0x1b61000, pte invalid [18643.303188] v3d fec00000.v3d: MMU error from client L2T (0) at 0x15bff00, pte invalid [18655.933748] v3d fec00000.v3d: MMU error from client L2T (0) at 0x15bff00, pte invalid However, most of the CL/CSD programs are now working happily without kernel warnings and errors. Regards, Sugizaki [1] https://patchwork.kernel.org/patch/11732895/ [2] https://github.com/Idein/py-videocore6 [3] https://github.com/raspberrypi/linux/tree/rpi-5.8.y Yukimasa Sugizaki (3): drm/v3d: Don't resubmit guilty CSD jobs drm/v3d: Correctly restart the timer when progress is made drm/v3d: Add job timeout module param drivers/gpu/drm/v3d/v3d_sched.c | 62 +++++++++++++++++++++++++++++++++-------- 1 file changed, 51 insertions(+), 11 deletions(-) -- 2.7.4

3 9

[PATCH] drm/i915/adl_s: Add gmbus pin mapping
by Anand Moon 11 Feb '21

11 Feb '21

Add table to map the GMBUS pin pairs to GPIO registers and port to DDC mapping for ADL_S as per below Bspec. Bspec:20124, 53597. Cc: Aditya Swarup <aditya.swarup(a)intel.com> Cc: Matt Roper <matthew.d.roper(a)intel.com> Cc: Lucas De Marchi <lucas.demarchi(a)intel.com> Signed-off-by: Anand Moon <anandx.ram.moon(a)intel.com> --- drivers/gpu/drm/i915/display/intel_gmbus.c | 16 ++++++++++++++-- 1 file changed, 14 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/i915/display/intel_gmbus.c b/drivers/gpu/drm/i915/display/intel_gmbus.c index 0c952e1d720e..58b8e42d4f90 100644 --- a/drivers/gpu/drm/i915/display/intel_gmbus.c +++ b/drivers/gpu/drm/i915/display/intel_gmbus.c @@ -52,6 +52,14 @@ static const struct gmbus_pin gmbus_pins[] = { [GMBUS_PIN_DPD] = { "dpd", GPIOF }, }; +static const struct gmbus_pin gmbus_pins_adls[] = { + [GMBUS_PIN_1_BXT] = { "edp", GPIOA }, + [GMBUS_PIN_9_TC1_ICP] = { "tc1", GPIOD }, + [GMBUS_PIN_10_TC2_ICP] = { "tc2", GPIOE }, + [GMBUS_PIN_11_TC3_ICP] = { "tc3", GPIOF }, + [GMBUS_PIN_12_TC4_ICP] = { "tc4", GPIOG }, +}; + static const struct gmbus_pin gmbus_pins_bdw[] = { [GMBUS_PIN_VGADDC] = { "vga", GPIOA }, [GMBUS_PIN_DPC] = { "dpc", GPIOD }, @@ -101,7 +109,9 @@ static const struct gmbus_pin gmbus_pins_dg1[] = { static const struct gmbus_pin *get_gmbus_pin(struct drm_i915_private *dev_priv, unsigned int pin) { - if (INTEL_PCH_TYPE(dev_priv) >= PCH_DG1) + if (INTEL_PCH_TYPE(dev_priv) == PCH_ADP) + return &gmbus_pins_adls[pin]; + else if (INTEL_PCH_TYPE(dev_priv) >= PCH_DG1) return &gmbus_pins_dg1[pin]; else if (INTEL_PCH_TYPE(dev_priv) >= PCH_ICP) return &gmbus_pins_icp[pin]; @@ -122,7 +132,9 @@ bool intel_gmbus_is_valid_pin(struct drm_i915_private *dev_priv, { unsigned int size; - if (INTEL_PCH_TYPE(dev_priv) >= PCH_DG1) + if (INTEL_PCH_TYPE(dev_priv) == PCH_ADP) + size = ARRAY_SIZE(gmbus_pins_adls); + else if (INTEL_PCH_TYPE(dev_priv) >= PCH_DG1) size = ARRAY_SIZE(gmbus_pins_dg1); else if (INTEL_PCH_TYPE(dev_priv) >= PCH_ICP) size = ARRAY_SIZE(gmbus_pins_icp); -- 2.30.0

3 2

Re: [Mesa-dev] should we do GSoC 2021?
by Trevor Woerner 11 Feb '21

11 Feb '21

Hi Omar, On Wed 2021-02-10 @ 02:31:11 PM, Omar Emara wrote: > If applying to GSoC will not take much time from you, I think you should apply > regardless and leave the acceptance decision on a per-project basis. GSoC works, literally, the other way around. Unfortunately GSoC doesn't start with students expressing their interest, then organizations applying based on the student interest they receive. GSoC starts with mentors willing to volunteer to mentor, then organizations deciding to apply based on the level of interest expressed by potential volunteers, then google selecting which organizations can participate, then students expressing an interest. Without a proper task list XOrg will not be selected by Google to participate in GSoC. Google has always been clear that the task list is one of the most important criteria for deciding which organizations they pick to participate. Mentors need to update the task list with ideas they're willing to mentor. So far nobody has volunteered to mentor, and by extension, nobody has volunteered to update the tasks to fit into the new constraints Google is applying to GSoC 2021. Organization applications close Feb 19, if there are no mentors we can't possibly consider participating.

1 0

← Newer
1
...
41
42
43
44
45
46
47
...
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