Update to 2.4.63:

Libdrm 2.4.63 has been released. It contains the new libdrm_amdgpu library.

Alan Coopersmith (1):
      include <sys/types.h> & <sys/mkdev.h> directly for major() and minor()

Alex Deucher (10):
      drm: consolidate common list implementations (v2)
      drm: add util_math.h
      drm: add libdrm_amdgpu (v7)
      drm: add tests/amdgpu (v3)
      amdgpu: update to the latest kernel header
      fix configuration when amdgpu is disabled
      fix amdgpu cunit configure test harder
      move up cunit workaround for ubuntu/debian
      add a note about which version of cunit is fixed for debian/ubuntu
      radeon: add new OLAND pci id

Alexandr Akulich (1):
      libdrm/amdgpu: Fixed drm.h include.

Anuj Phogat (2):
      i965/gen9: Pass alignment as function parameter in drm_intel_gem_bo_alloc_internal()
      Set alignment value in drm_intel_add_validate_buffer()

Chris Wilson (1):
      drm: Detect no-op drmModeAtomicRequest and return early

Christian K旦nig (16):
      amdgpu: cleanup public interface v2
      amdgpu: add public bo list interface v3
      amdgpu: compare the primary device names instead
      amdgpu: remove bo_vas hash table v2
      amdgpu: add helper for VM mapping v2
      amdgpu: stop checking flag masks
      amdgpu: explicitly unmap GPU mapping on BO destruction
      amdgpu: remove flink export workaround v2
      amdgpu: cleanup VA IOCTL handling
      amdgpu: remove pointer arithmetic from command submission
      amdgpu: add CS dependencies v2
      gitignore: add some generated amdgpu files
      amdgpu: cleanup public interface style
      amdgpu: remove reference to AMD specific error codes
      amdgpu: use common fence structure for dependencies as well.
      amdgpu: fix bs buffer size for vce test

Emil Velikov (15):
      configure: default --enable-valgrind to auto
      freedreno: zero is a valid fd number, treat it as such
      omap: zero is a valid fd number, treat it as such
      xf86drm: fix incorrect fd comparison in drmOpenOnce{,WithType}
      Consistently check the fd value
      man: remove .man_fixup workaround
      Force enable amdgpu for the dist build/check.
      amdgpu/util_hash: hide private symbols from global namespace
      amdgpu/util_hash_table: hide private symbols from global namespace
      amdgpu: add a bunch of missing config.h includes
      amdgpu: cosmetic chances in license boilerplate
      amdgpu: squash trivial documentation typo
      amdgpu/amdgpu_vamgr: hide private symbols from global namespace
      amdgpu: hide the final internal functions from global namespace
      amdgpu: add symbols check test

Jack Xiao (2):
      amdgpu: fix round down/up page size error
      amdgpu: add zero timeout check in amdgpu_cs_query_fence_status

Jammy Zhou (24):
      amdgpu: remove active_rb_pipes from amdgpu_gpu_info
      amdgpu: remove AMDGPU_GEM_CREATE_CPU_GTT_UC
      amdgpu: fix 32-on-64 support (v2)
      amdgpu: add ctx_id for wait_cs
      amdgpu: reuse the kernel IB flags v2
      amdgpu: validate the upper limit of virtual address v2
      amdgpu: fix the number of IB size enums
      amdgpu: remove unused AMDGPU_IB_RESOURCE_PRIORITY
      amdgpu: replace alloca with calloc v2
      amdgpu: add amdgpu_bo_list_update interface v2
      amdgpu: add IB sharing support v2
      tests/amdgpu: add shared IB submission test v2
      amdgpu: get rid of IB pool management v3
      tests/amdgpu: manage IB in client side
      amdgpu: add amdgpu_query_gds_info
      amdgpu: cleanup gds specific alloc/free functions
      amdgpu: merge amdgpu_drm.h from kernel
      amdgpu: do NULL check for bo handle in amdgpu_bo_query_info
      amdgpu: improve the amdgpu_cs_query_fence_status interface
      drm: fix the ALIGN macro to avoid value clamp
      tests/amdgpu: remove the duplicate IB allocation for VCE test
      amdgpu: add flags parameter for amdgpu_va_range_alloc
      amdgpu: add amdgpu_bo_va_op for va map/unmap support v3
      amdgpu: expose the PCI revision ID

Jonathan Gray (3):
      xf86drmMode: Implement drmCheckModesettingSupported() for OpenBSD
      xf86drm: correct the OpenBSD DRM_MAJOR define
      xf86drm: use the correct device minor names on OpenBSD

Joonyoung Shim (1):
      Build vbltest irrespective of the presence of libudev.

Julien Cristau (1):
      Fix headers inclusion in xf86drmMode.c

Ken Wang (8):
      amdgpu: make vamgr global
      amdgpu: add max_memory_clock for interface query
      amdgpu: add vram_type and vram_bit_width for interface query
      amdgpu: add ce_ram_size for interface query
      amdgpu: add ib_start_alignment and ib_size_alignment for interface query
      amdgpu: add base_preferred parameter to amdgpu_vamgr_find_va
      amdgpu : move management of user fence from libdrm to UMD
      test/amdgpu : fix a bug in VCE/UVD test introduced by previous change

Kristian H淡gsberg Kristensen (1):
      intel: Drop aub dumping functionality

Leo Liu (2):
      tests/amdgpu: implement VCE unit tests
      amdgpu: add VCE harvesting instance query

Marek Ol邸叩k (15):
      amdgpu: implement amdgpu_cs_query_reset_state
      amdgpu: fix a valgrind warning
      amdgpu: add new AMDGPU_TILING flags
      amdgpu: rename GEM_OP_SET_INITIAL_DOMAIN -> GEM_OP_SET_PLACEMENT
      amdgpu: don't use amdgpu_cs_create_ib for allocation of the fence BO
      amdgpu: remove amdgpu_ib
      amdgpu: remove amdgpu_ib helpers
      amdgpu: remove bo_handle from amdgpu_cs_ib_info, IBs should be in buffer list
      amdgpu: allow exporting KMS handles with render nodes
      amdgpu: use alloca and malloc in critical codepaths (v2)
      amdgpu: fix valgrind warnings
      amdgpu: fix double mutex_unlock in amdgpu_bo_import
      drm/amdgpu: allow passing absolute timeouts to amdgpu_cs_query_fence_status
      amdgpu: check the user fence only if the IP supports user fences
      Bump version for release

Mario Kleiner (1):
      libdrm/amdgpu: Use private fd for amdgpu_device and winsys hash table to fix ZaphodHeads. (v2)

Michel D辰nzer (4):
      amdgpu: Use drmIoctl in amdgpu_ioctl_wait_cs
      amdgpu/tests: Use buf_handle in amdgpu_bo_alloc_and_map() error paths
      amdgpu/tests: Cast CPU map argument of amdgpu_bo_alloc_and_map to void**
      amdgpu: Remove unused local variables bo_size/offset from amdgpu_cs_submit

Sabre Shao (2):
      amdgpu: add va allocation intefaces
      amdgpu: add va range query interface

Tapani P辰lli (3):
      remove usage of 'c_plusplus' preprocessor macro
      intel: wrap intel_bufmgr.h C code for C++ compilation/linking
      modeprint: cleanup, remove compile warnings

monk.liu (7):
      amdgpu: support non-page-aligned userptr
      amdgpu: add missing mutex lock/unlock pair
      amdgpu: fix an error of bo_list handler
      amdgpu: do not call BO_LIST ioctl when no resource referred
      amdgpu: fix segfault when resources are NULL
      amdgpu: fix code alignment
      amdgpu: fix vamgr_free_va logic


(wiz)