summaryrefslogtreecommitdiff
AgeCommit message (Collapse)AuthorFilesLines
2019-01-29mm/hmm: add helpers for driver to safely take the mmap_semhmm-for-5.1Jérôme Glisse1-3/+47
The device driver context which holds reference to mirror and thus to core hmm struct might outlive the mm against which it was created. To avoid every driver to check for that case provide an helper that check if mm is still alive and take the mmap_sem in read mode if so. If the mm have been destroy (mmu_notifier release call back did happen) then we return -EINVAL so that calling code knows that it is trying to do something against a mm that is no longer valid. Signed-off-by: Jérôme Glisse <jglisse@redhat.com> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Ralph Campbell <rcampbell@nvidia.com> Cc: John Hubbard <jhubbard@nvidia.com>
2019-01-29mm/hmm: allow to mirror vma of a file on a DAX backed filesystemJérôme Glisse1-21/+112
This add support to mirror vma which is an mmap of a file which is on a filesystem that using a DAX block device. There is no reason not to support that case. Note that unlike GUP code we do not take page reference hence when we back-off we have nothing to undo. Signed-off-by: Jérôme Glisse <jglisse@redhat.com> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Dan Williams <dan.j.williams@intel.com> Cc: Ralph Campbell <rcampbell@nvidia.com> Cc: John Hubbard <jhubbard@nvidia.com>
2019-01-29mm/hmm: support hugetlbfs (snap shoting, faulting and DMA mapping)Jérôme Glisse2-23/+147
This adds support for hugetlbfs so that HMM user can map mirror range of virtual address back by hugetlbfs. Note that now the range allows user to optimize DMA mapping of such page so that we can map a huge page as one chunk. Signed-off-by: Jérôme Glisse <jglisse@redhat.com> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Ralph Campbell <rcampbell@nvidia.com> Cc: John Hubbard <jhubbard@nvidia.com>
2019-01-29mm/hmm: add an helper function that fault pages and map them to a deviceJérôme Glisse2-0/+161
This is a all in one helper that fault pages in a range and map them to a device so that every single device driver do not have to re-implement this common pattern. Signed-off-by: Jérôme Glisse <jglisse@redhat.com> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Ralph Campbell <rcampbell@nvidia.com> Cc: John Hubbard <jhubbard@nvidia.com>
2019-01-29mm/hmm: add default fault flags to avoid the need to pre-fill pfns arrays.Jérôme Glisse2-0/+19
The HMM mirror API can be use in two fashions. The first one where the HMM user coalesce multiple page faults into one request and set flags per pfns for of those faults. The second one where the HMM user want to pre-fault a range with specific flags. For the latter one it is a waste to have the user pre-fill the pfn arrays with a default flags value. This patch adds a default flags value allowing user to set them for a range without having to pre-fill the pfn array. Signed-off-by: Jérôme Glisse <jglisse@redhat.com> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Ralph Campbell <rcampbell@nvidia.com> Cc: John Hubbard <jhubbard@nvidia.com>
2019-01-29mm/hmm: improve driver API to work and wait over a rangeJérôme Glisse2-304/+430
A common use case for HMM mirror is user trying to mirror a range and before they could program the hardware it get invalidated by some core mm event. Instead of having user re-try right away to mirror the range provide a completion mechanism for them to wait for any active invalidation affecting the range. This also changes how hmm_range_snapshot() and hmm_range_fault() works by not relying on vma so that we can drop the mmap_sem when waiting and lookup the vma again on retry. Signed-off-by: Jérôme Glisse <jglisse@redhat.com> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Ralph Campbell <rcampbell@nvidia.com> Cc: John Hubbard <jhubbard@nvidia.com>
2019-01-29mm/hmm: improve and rename hmm_vma_fault() to hmm_range_fault()Jérôme Glisse2-53/+53
Rename for consistency between code, comments and documentation. Also improves the comments on all the possible returns values. Improve the function by returning the number of populated entries in pfns array. Signed-off-by: Jérôme Glisse <jglisse@redhat.com> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Ralph Campbell <rcampbell@nvidia.com> Cc: John Hubbard <jhubbard@nvidia.com>
2019-01-29mm/hmm: improve and rename hmm_vma_get_pfns() to hmm_range_snapshot()Jérôme Glisse2-15/+12
Rename for consistency between code, comments and documentation. Also improves the comments on all the possible returns values. Improve the function by returning the number of populated entries in pfns array. Signed-off-by: Jérôme Glisse <jglisse@redhat.com> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Ralph Campbell <rcampbell@nvidia.com> Cc: John Hubbard <jhubbard@nvidia.com>
2019-01-29mm/hmm: do not erase snapshot when a range is invalidatedJérôme Glisse1-6/+0
Users of HMM might be using the snapshot information to do preparatory step like dma mapping pages to a device before checking for invalidation through hmm_vma_range_done() so do not erase that information and assume users will do the right thing. Signed-off-by: Jérôme Glisse <jglisse@redhat.com> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Ralph Campbell <rcampbell@nvidia.com> Cc: John Hubbard <jhubbard@nvidia.com>
2019-01-29mm/hmm: use reference counting for HMM structJérôme Glisse2-60/+120
Every time i read the code to check that the HMM structure does not vanish before it should thanks to the many lock protecting its removal i get a headache. Switch to reference counting instead it is much easier to follow and harder to break. This also remove some code that is no longer needed with refcounting. Signed-off-by: Jérôme Glisse <jglisse@redhat.com> Cc: Ralph Campbell <rcampbell@nvidia.com> Cc: John Hubbard <jhubbard@nvidia.com> Cc: Andrew Morton <akpm@linux-foundation.org>
2019-01-29HMM FOR 5.1 COVER LETTERJérôme Glisse1-0/+27
Signed-off-by: Jérôme Glisse <jglisse@redhat.com>
2019-01-29gpu/nouveau: new ioctl to migrate process memory to GPU memoryJérôme Glisse5-0/+57
This add an ioctl to migrate a range of process address space to the device memory. On platform without cache coherent bus (x86, ARM, ...) this means that CPU can not access that range directly, instead CPU will fault which will migrate the memory back to system memory. This is behind a staging flag so that we can evolve the API. Signed-off-by: Jérôme Glisse <jglisse@redhat.com>
2019-01-29gpu/nouveau: device memory helpers for SVMJérôme Glisse6-0/+1082
Device memory can be use in SVM, in which case we do not have any of the existing buffer object. This commit add infrastructure to allow use of device memory without nouveau_bo. Again this is a temporary solution until a rework of GPU memory management. Signed-off-by: Jérôme Glisse <jglisse@redhat.com>
2019-01-29gpu/nouveau: allow to create nouveau_mem without ttm objectJérôme Glisse2-0/+31
New helper to create nouveau_mem object without ttm object, directly using scatterlist object instead. This is a temporary solution to allocate device memory. Bigger re-work of GPU memory management is needed. Signed-off-by: Jérôme Glisse <jglisse@redhat.com>
2019-01-29gpu/nouveau/svm: have reserved vma comes as a special mmap of device fileJérôme Glisse4-1/+61
Enforce the reserved area to be the result of mmap on the device file. Signed-off-by: Jérôme Glisse <jglisse@redhat.com>
2019-01-29gpu/nouveau/svm: initial support for share virtual memory (SVM/SVA)Ben Skeggs9-1/+818
Add support for share virtual memory (SVM) or share virtual address (SVA) which means that the GPU use the same address space as a process which register itself with the device. Once the process at registered itself then it can run threads on the GPU and those threads will be able to access the same address space the CPU threads of the process. This use HMM to mirror the CPU page table into the GPU page table and to keep both of them synchronize at all times so that both GPU and CPU access the same physical memory for the same virtual address. Signed-off-by: Ben Skeggs <bskeggs@redhat.com> Tested-by: Jérôme Glisse <jglisse@redhat.com>
2019-01-19gpu/nouveau/fault: GV100 expose VoltaFaultBufferABen Skeggs2-0/+11
"non-replayable" fault buffer.. need to figure out exact semantics here, it appears both UVM and RM process this one, not sure exactly why/what parts need to be done by each. Signed-off-by: Ben Skeggs <bskeggs@redhat.com> Tested-by: Jérôme Glisse <jglisse@redhat.com>
2019-01-19gpu/nouveau/fault: GP100 expose MaxwellFaultBufferABen Skeggs9-8/+141
Expose page fault on volta GPU. Signed-off-by: Ben Skeggs <bskeggs@redhat.com> Tested-by: Jérôme Glisse <jglisse@redhat.com>
2019-01-19gpu/nouveau/fault: add explicit control over fault buffer interruptsBen Skeggs1-2/+0
The GPU will continually fire interrupts while a fault buffer GET != PUT, and to stop the spurious interrupts while the handler does its thing, we were disabling the fault buffer temporarily. This is not actually a great idea to begin with, and made worse by Volta resetting GET/PUT when it's reactivated. So, let's not do that. Signed-off-by: Ben Skeggs <bskeggs@redhat.com> Tested-by: Jérôme Glisse <jglisse@redhat.com>
2019-01-19gpu/nouveau/mmu: GP100- support vmms with gcc/tex replayable faults enabledBen Skeggs9-10/+54
Add methods to replay or cancel fault. The event flow is: 1- interrupt handler trigger page fault handler 2- page fault handler try to fault memory (using HMM for SVM/SVA) 3a- if page fault is successful update GPU page table 4a- use method introduced in this patch to resume the GPU threads 3b- if page fault fails use method introduced in this patch to cancel/kill GPU threads Signed-off-by: Ben Skeggs <bskeggs@redhat.com> Tested-by: Jérôme Glisse <jglisse@redhat.com>
2019-01-19gpu/nouveau/mmu: GP100- add privileged methods for fault replay/cancelBen Skeggs7-0/+107
Host methods exist to do at least some of what we need, but we are not currently pushing replay/cancels through a channel like UVM does as it's not clear whether it's necessary in our case (UVM also updates PTEs with the GPU). UVM pushes a software method for fault cancels on Pascal, which could indicate that we potentially need some special SW handling anyway. Keep it simple for now, until we figure out exactly what we need here. TBD: not sure targeted fault cancel is sufficient Signed-off-by: Ben Skeggs <bskeggs@redhat.com> Tested-by: Jérôme Glisse <jglisse@redhat.com>
2019-01-19gpu/nouveau/mmu: add a privileged method to directly manage PTEsBen Skeggs6-34/+445
Privileged method to allow direct manipulation of the GPU page table, this is for share virtual memory (SVM) or share virtual address (SVA). They are still things to improve on top of this: - avoid PT readback, generally *way* too slow - mutliple page sizes (huge pages) - systems without a 4KiB page size Signed-off-by: Ben Skeggs <bskeggs@redhat.com> Tested-by: Jérôme Glisse <jglisse@redhat.com>
2019-01-19gpu/nouveau/mmu: store mapped flag separately from memory pointerBen Skeggs2-5/+11
This will be used to support a privileged client providing PTEs directly. Signed-off-by: Ben Skeggs <bskeggs@redhat.com> Tested-by: Jérôme Glisse <jglisse@redhat.com>
2019-01-19gpu/nouveau/mmu: support initialisation of client-managed address-spacesBen Skeggs22-115/+172
A client managed address space means that only a small range of virtual address space is manage by core nouveau virtual address space management. This is to allow nouveau to restrict itself to a small range of virtual address space so that a process address space can be use directly for the GPU virtual address space. Signed-off-by: Ben Skeggs <bskeggs@redhat.com> Tested-by: Jérôme Glisse <jglisse@redhat.com>
2019-01-19RDMA/uverbs: Mark ioctl responses with UVERBS_ATTR_F_VALID_OUTPUTJason Gunthorpe4-13/+59
When the ioctl interface for the write commands was introduced it did not mark the core response with UVERBS_ATTR_F_VALID_OUTPUT. This causes rdma-core in userspace to not mark the buffers as written for valgrind. Along the same lines it turns out we have always missed marking the driver data. Fixing both of these makes valgrind work properly with rdma-core and ioctl. Fixes: 4785860e04bc ("RDMA/uverbs: Implement an ioctl that can call write and write_ex handlers") Signed-off-by: Jason Gunthorpe <jgg@mellanox.com> Reviewed-by: Artemy Kovalyov <artemyko@mellanox.com> Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
2019-01-19RDMA/mthca: Clear QP objects during their allocationLeon Romanovsky1-2/+2
As part of audit process to update drivers to use rdma_restrack_add() ensure that QP objects is cleared before access. Such change fixes the crash observed with uninitialized non zero sgid attr accessed by ib_destroy_qp(). CPU: 3 PID: 74 Comm: kworker/u16:1 Not tainted 4.19.10-300.fc29.x86_64 Workqueue: ipoib_wq ipoib_cm_tx_reap [ib_ipoib] RIP: 0010:rdma_put_gid_attr+0x9/0x30 [ib_core] RSP: 0018:ffffb7ad819dbde8 EFLAGS: 00010202 RAX: 0000000000000000 RBX: ffff8d1bdf5a2e00 RCX: 0000000000002699 RDX: 206c656e72656af8 RSI: ffff8d1bf7ae6160 RDI: 206c656e72656b20 RBP: 0000000000000000 R08: 0000000000026160 R09: ffffffffc06b45bf R10: ffffe849887da000 R11: 0000000000000002 R12: ffff8d1be30cb400 R13: ffff8d1bdf681800 R14: ffff8d1be2272400 R15: ffff8d1be30ca000 FS: 0000000000000000(0000) GS:ffff8d1bf7ac0000(0000) knlGS:0000000000000000 Trace: ib_destroy_qp+0xc9/0x240 [ib_core] ipoib_cm_tx_reap+0x1f9/0x4e0 [ib_ipoib] process_one_work+0x1a1/0x3a0 worker_thread+0x30/0x380 ? pwq_unbound_release_workfn+0xd0/0xd0 kthread+0x112/0x130 ? kthread_create_worker_on_cpu+0x70/0x70 ret_from_fork+0x22/0x40 Reported-by: Alexander Murashkin <AlexanderMurashkin@msn.com> Tested-by: Alexander Murashkin <AlexanderMurashkin@msn.com> Fixes: 1a1f460ff151 ("RDMA: Hold the sgid_attr inside the struct ib_ah/qp") Signed-off-by: Parav Pandit <parav@mellanox.com> Signed-off-by: Leon Romanovsky <leonro@mellanox.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
2019-01-19RDMA/vmw_pvrdma: Return the correct opcode when creating WRAdit Ranadive3-1/+41
Since the IB_WR_REG_MR opcode value changed, let's set the PVRDMA device opcodes explicitly. Reported-by: Ruishuang Wang <ruishuangw@vmware.com> Fixes: 9a59739bd01f ("IB/rxe: Revise the ib_wr_opcode enum") Cc: stable@vger.kernel.org Reviewed-by: Bryan Tan <bryantan@vmware.com> Reviewed-by: Ruishuang Wang <ruishuangw@vmware.com> Reviewed-by: Vishnu Dasa <vdasa@vmware.com> Signed-off-by: Adit Ranadive <aditr@vmware.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
2019-01-19RDMA/cma: Add cm_id restrack resource based on kernel or user cm_id typeSteve Wise1-1/+4
A recent regression causes a null ptr crash when dumping cm_id resources. The cma is incorrectly adding all cm_id restrack resources as kernel mode. Fixes: af8d70375d56 ("RDMA/restrack: Resource-tracker should not use uobject pointers") Signed-off-by: Steve Wise <swise@opengridcomputing.com> Reviewed-by: Leon Romanovsky <leonro@mellanox.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
2019-01-19RDMA/nldev: Don't expose unsafe global rkey to regular userLeon Romanovsky1-4/+0
Unsafe global rkey is considered dangerous because it exposes memory registered for all memory in the system. Only users with a QP on the same PD can use the rkey, and generally those QPs will already know the value. However, out of caution, do not expose the value to unprivleged users on the local system. Require CAP_NET_ADMIN instead. Cc: <stable@vger.kernel.org> # 4.16 Fixes: 29cf1351d450 ("RDMA/nldev: provide detailed PD information") Signed-off-by: Leon Romanovsky <leonro@mellanox.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
2019-01-19RDMA/uverbs: Fix post send success return value in case of errorGal Pressman1-1/+3
If get QP object fails 'ret' must be assigned with a proper error code. Fixes: 9a0738575f26 ("RDMA/uverbs: Use uverbs_response() for remaining response copying") Signed-off-by: Gal Pressman <galpress@amazon.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
2019-01-19drm/sun4i: backend: add missing of_node_putsJulia Lawall1-2/+3
The device node iterators perform an of_node_get on each iteration, so a jump out of the loop requires an of_node_put. Remote and port also have augmented reference counts, so drop them on each iteration and at the end of the function, respectively. Remote is only used for the address it contains, not for the contents of that address, so the reference count can be dropped immediately. The semantic patch that fixes the first part of this problem is as follows (http://coccinelle.lip6.fr): // <smpl> @@ expression root,e; local idexpression child; iterator name for_each_child_of_node; @@ for_each_available_child_of_node(root, child) { ... when != of_node_put(child) when != e = child + of_node_put(child); ? break; ... } ... when != child // </smpl> Signed-off-by: Julia Lawall <Julia.Lawall@lip6.fr> Signed-off-by: Maxime Ripard <maxime.ripard@bootlin.com> Link: https://patchwork.freedesktop.org/patch/msgid/1547369264-24831-5-git-send-email-Julia.Lawall@lip6.fr
2019-01-19drm/meson: Fix atomic mode switching regressionNeil Armstrong2-21/+7
Since commit 2bcd3ecab773 when switching mode from X11 (ubuntu mate for example) the display gets blurry, looking like an invalid framebuffer width. This commit fixed atomic crtc modesetting in a totally wrong way and introduced a local unnecessary ->enabled crtc state. This commit reverts the crctc _begin() and _enable() changes and simply adds drm_atomic_helper_commit_tail_rpm as helper. Reported-by: Tony McKahan <tonymckahan@gmail.com> Suggested-by: Daniel Vetter <daniel@ffwll.ch> Fixes: 2bcd3ecab773 ("drm/meson: Fixes for drm_crtc_vblank_on/off support") Signed-off-by: Neil Armstrong <narmstrong@baylibre.com> Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch> [narmstrong: fixed blank line issue from checkpatch] Link: https://patchwork.freedesktop.org/patch/msgid/20190114153118.8024-1-narmstrong@baylibre.com
2019-01-19drm/meson: add missing of_node_putJulia Lawall1-2/+7
Add an of_node_put when the result of of_graph_get_remote_port_parent is not available. An of_node_put is also needed when meson_probe_remote completes. This was present at the recursive call, but not in the call from meson_drv_probe. The semantic match that finds this problem is as follows (http://coccinelle.lip6.fr): // <smpl> @r exists@ local idexpression e; expression x; @@ e = of_graph_get_remote_port_parent(...); ... when != x = e when != true e == NULL when != of_node_put(e) when != of_fwnode_handle(e) ( return e; | *return ...; ) // </smpl> Signed-off-by: Julia Lawall <Julia.Lawall@lip6.fr> Acked-by: Neil Armstrong <narmstrong@baylibre.com> Signed-off-by: Neil Armstrong <narmstrong@baylibre.com> Link: https://patchwork.freedesktop.org/patch/msgid/1547372691-28324-4-git-send-email-Julia.Lawall@lip6.fr
2019-01-19drm/virtio: drop prime import/export callbacksGerd Hoffmann3-22/+0
Also set prime_handle_to_fd and prime_fd_to_handle to NULL, so drm will not advertive DRM_PRIME_CAP_{IMPORT,EXPORT} to userspace. Signed-off-by: Gerd Hoffmann <kraxel@redhat.com> Reviewed-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com> Reviewed-by: Dave Airlie <airlied@redhat.com> Link: http://patchwork.freedesktop.org/patch/msgid/20190110111545.26768-1-kraxel@redhat.com
2019-01-19drm/qxl: drop prime import/export callbacksGerd Hoffmann2-18/+0
Also set prime_handle_to_fd and prime_fd_to_handle to NULL, so drm will not advertive DRM_PRIME_CAP_{IMPORT,EXPORT} to userspace. Signed-off-by: Gerd Hoffmann <kraxel@redhat.com> Reviewed-by: Dave Airlie <airlied@redhat.com> Link: http://patchwork.freedesktop.org/patch/msgid/20190110081750.11358-1-kraxel@redhat.com
2019-01-19drm/rockchip: add missing of_node_putJulia Lawall1-1/+3
The device node iterators perform an of_node_get on each iteration, so a jump out of the loop requires an of_node_put. The semantic patch that fixes this problem is as follows (http://coccinelle.lip6.fr): // <smpl> @@ expression root,e; local idexpression child; iterator name for_each_child_of_node; @@ for_each_child_of_node(root, child) { ... when != of_node_put(child) when != e = child + of_node_put(child); ? break; ... } ... when != child // </smpl> Fixes: 1f0f01515172 ("drm/rockchip: Add support for Rockchip Soc RGB output interface") Cc: stable@vger.kernel.org Signed-off-by: Julia Lawall <Julia.Lawall@lip6.fr> [added fixes and cc-stable] Signed-off-by: Heiko Stuebner <heiko@sntech.de> Link: https://patchwork.freedesktop.org/patch/msgid/1547369264-24831-4-git-send-email-Julia.Lawall@lip6.fr
2019-01-19drm/i915/gvt: Fix mmap range checkZhenyu Wang1-2/+12
This is to fix missed mmap range check on vGPU bar2 region and only allow to map vGPU allocated GMADDR range, which means user space should support sparse mmap to get proper offset for mmap vGPU aperture. And this takes care of actual pgoff in mmap request as original code always does from beginning of vGPU aperture. Fixes: 659643f7d814 ("drm/i915/gvt/kvmgt: add vfio/mdev support to KVMGT") Cc: "Monroy, Rodrigo Axel" <rodrigo.axel.monroy@intel.com> Cc: "Orrala Contreras, Alfredo" <alfredo.orrala.contreras@intel.com> Cc: stable@vger.kernel.org # v4.10+ Reviewed-by: Hang Yuan <hang.yuan@intel.com> Signed-off-by: Zhenyu Wang <zhenyuw@linux.intel.com>
2019-01-19drm/i915/gvt: free VFIO region space in vgpu detachHang Yuan3-4/+16
VFIO region space is allocated when one region is registered for one vgpu. So free the space when destroy the vgpu. Also change the parameter of detach_vgpu callback to use vgpu directly. Fixes: b851adeac0858c7d257b3 ("drm/i915/gvt: Add opregion support") Reviewed-by: Zhenyu Wang <zhenyuw@linux.intel.com> Signed-off-by: Hang Yuan <hang.yuan@linux.intel.com> Signed-off-by: Zhenyu Wang <zhenyuw@linux.intel.com>
2019-01-19drm/i915/gvt: Allow F_CMD_ACCESS on mmio 0x21f0Colin Xu1-0/+1
The newly updated guest driver could program 0x21f0 via lri and srm, without F_CMD_ACCESS flag cmd parser will stop parser reset cmd. 0x21f0 applies to BDW, SKL, KBL, BXT and CFL. Reviewed-by: Yan Zhao <yan.y.zhao@intel.com> Signed-off-by: Colin Xu <colin.xu@intel.com> Signed-off-by: Zhenyu Wang <zhenyuw@linux.intel.com>
2019-01-19drm/amd/display: Detach backlight from streamDavid Francis4-15/+3
[Why] Backlight is conceptually a property of links, not streams. All backlight programming is done on links, but there is a stream property bl_pwm_level that is used to restore backlight on dpms on and s3 resume. This is unnecessary, as backlight is already restored by hardware with no driver intervention. [How] Remove bl_pwm_level, and the stream argument to set_backlight Bug: https://bugs.freedesktop.org/show_bug.cgi?id=109375 Signed-off-by: David Francis <David.Francis@amd.com> Reviewed-by: Harry Wentland <Harry.Wentland@amd.com> Acked-by: Anthony Koo <Anthony.Koo@amd.com> Acked-by: Leo Li <sunpeng.li@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com> (cherry picked from commit 923fe4951282cbdfce05186c10380bbc45b5e03b)
2019-01-19Revert "drm/amdgpu: validate user pitch alignment"Michel Dänzer1-10/+0
The check turned out to be too strict in some cases. Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2019-01-19Revert "drm/amdgpu: validate user GEM object size"Alex Deucher1-8/+0
It was at the same time too strict (for linear tiling modes, where no height alignment is required) and too lenient (for 2D tiling modes, where height may need to be aligned to values > 8). Acked-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2019-01-19drm/amd/display: Fix disabled cursor on top screen edgeNicholas Kazlauskas2-2/+2
[Why] The cursor vanishes when touching the top of edge of the screen for Raven on Linux. This occurs because the cursor height is not taken into account when deciding to disable the cursor. [How] Factor in the cursor height into the cursor calculations - and mimic the existing x position calculations. Fixes: 94a4ffd1d40b ("drm/amd/display: fix PIP bugs on Dal3") Signed-off-by: Nicholas Kazlauskas <nicholas.kazlauskas@amd.com> Reviewed-by: Harry Wentland <Harry.Wentland@amd.com> Acked-by: Leo Li <sunpeng.li@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2019-01-19drm/amd/display: fix warning on raven hotplugRoman Li1-13/+6
[Why] Hotplug on raven results in REG_WAIT_TIMEOUT warning due to failing attempt to lock disabled otg for the hubp interdependent pipes programming. [How] Don't setup pipe interdependencies for disabled otg. Also removed the unnecessary duplicate logic checks. Signed-off-by: Roman Li <Roman.Li@amd.com> Reviewed-by: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com> Acked-by: Leo Li <sunpeng.li@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2019-01-19drm/amd/display: fix PME notification not working in RV desktopCharlene Liu1-1/+6
[Why] PPLIB not receive the PME when unplug. Signed-off-by: Charlene Liu <charlene.liu@amd.com> Reviewed-by: Chris Park <Chris.Park@amd.com> Acked-by: Leo Li <sunpeng.li@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2019-01-19drm/amd/display: Only get the connector state for VRR when toggledNicholas Kazlauskas1-1/+1
[Why] This fixes a stuttering issue that occurs when moving a hardware cursor when VRR is enabled. Previously when VRR is enabled atomic check will grab the connector state for every atomic update. This has to lock the connector in order to do so. The locking is bad enough by itself for performance, but it gets worse with what we do just below that - add all the planes for the CRTC to the commit. This prevents the cursor fast path from working - there's more than one plane now. With state->allow_modeset = true on top of this, it also adds and removes all the planes from the DC context triggering a full (very slow) update in DC. [How] We need the connector state to get the VRR min/max capbilities, but we only need them when there's a CRTC mode change or when VRR is toggled. The condition has been updated accordingly. Fixes: 3cc22f281318 ("drm/amdgpu: Set FreeSync state using drm VRR properties") Signed-off-by: Nicholas Kazlauskas <nicholas.kazlauskas@amd.com> Reviewed-by: Leo Li <sunpeng.li@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2019-01-19drm/amd/display: Pack DMCU iRAM alignmentJosip Pavic1-0/+2
[Why] When the DMCU's iRAM definition was moved to the newly created power_helpers, a #pragma pack was lost, causing the iRAM to be misaligned [How] Restore the #pragma pack Signed-off-by: Josip Pavic <Josip.Pavic@amd.com> Reviewed-by: Anthony Koo <Anthony.Koo@amd.com> Acked-by: Leo Li <sunpeng.li@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2019-01-19drm/amd/powerplay: run acg btc for Vega12Kenneth Feng1-0/+21
acg btc was added to Vega12 Signed-off-by: Kenneth Feng <kenneth.feng@amd.com> Reviewed-by: Evan Quan <evan.quan@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2019-01-19drm/amdkfd: Don't assign dGPUs to APU topology devicesFelix Kuehling1-3/+9
dGPUs need their own topology devices. Don't assign them to APU topology devices with CPU cores. Bug: https://github.com/RadeonOpenCompute/ROCK-Kernel-Driver/issues/66 Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com> Tested-by: Elias Konstantinidis <ekondis@gmail.com> Acked-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2019-01-19drm/amdkfd: Allow building KFD on ARM64 (v2)Felix Kuehling3-6/+15
ifdef x86_64 specific code. Allow enabling CONFIG_HSA_AMD on ARM64. v2: Fixed a compiler warning due to an unused variable CC: Mark Nutter <Mark.Nutter@arm.com> Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com> Tested-by: Mark Nutter <Mark.Nutter@arm.com> Acked-by: Alex Deucher <alexander.deucher@amd.com> Acked-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
generated by cgit v1.2.3 (git 2.39.1) at 2024-11-24 09:15:32 +0000