summaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2008-11-16gem: protect idr_pre with the spinlockJerome Glisse
There seems to be an issue if we don't protect idr_pre with the spinlock we use for idr operations.
2008-11-16radeon: protect cs ioctl atomic part with a mutexJerome Glisse
A small subset of CS need to be atomic (relocation+IB commit to ring) right now, because of the way relocation are handled, we need to protect the whole ioctl.
2008-11-16libdrm-radeon: add print callback to cs & small fixesJerome Glisse
2008-11-15libdrm-radeon: unref return current BO ptr to reflect BO destructionJerome Glisse
2008-11-15drm/bo: fix stupid lock imbalanceDave Airlie
2008-11-14libdrm-radeon: unreference buffer once cs stream is submited or on cs cleanJerome Glisse
BO are referenced once by reloc to make sure that they not destroyed before we get a chance to flush the cmd stream, so we need to unreference them once in cs submit or cs erase if cs i never submitted so bo can be destructed.
2008-11-14libdrm-radeon: new tracker toolsJerome Glisse
To keep record of bo activities and print them when necessary, should help in tracking unbalanced ref/unref calls.
2008-11-12mode: Minor reodering and renamingJakob Bornecrantz
2008-11-12mode: Reorder the ioctls and numberingJakob Bornecrantz
This is to fill in the gaps left by the removal of the hotplug ioctls. And they also look better :)
2008-11-12mode: Unify userspace API to uint_tJakob Bornecrantz
2008-11-12mode: Remove hotplug support from ioctl interfaceJakob Bornecrantz
2008-11-12mode: Unify types for ids and stringsJakob Bornecrantz
2008-11-12radeon+libdrm-radeon: change relocation informationsJerome Glisse
Relocation now consist of the following informations (in this order) : handle buffer object handle identifier start_offset start offset of first data of the buffer object used by the cs end_offset end offset of last data of the buffer object used by the cs read_domain read domain (either VRAM, or GTT as GPU is invalid for CS) write_domain write domain (either VRAM, or GTT as GPU is invalid for CS) flags flags used for further optimization (like discard previous buffer content or forget buffer content after cs which can help in avoiding moving content in or out)
2008-11-10Merge branch 'modesetting-gem' of ssh://git.freedesktop.org/git/mesa/drm ↵Jerome Glisse
into modesetting-gem
2008-11-10libdrm-radeon: be verbose on bo failure and cleanup cs a bitJerome Glisse
2008-11-10radeon: add gart useable size to report to userspaceDave Airlie
2008-11-10radeon: upgrade atom headersDave Airlie
2008-11-10radeon: fix fence race condition hopefullyDave Airlie
For some reason reading the SCRATCH reg from RAM causes some race to occur. Hopefully fix this.
2008-11-10drm/radeon: add dpms connector functionsDave Airlie
2008-11-10radeon: avivo cursors are across the full surface.Dave Airlie
fixes cursor on second head
2008-11-10modesetting: set the crtc x,y after the mode base changeDave Airlie
2008-11-10drm/radeon: add uncached allocator to drm ttm code.Dave Airlie
2008-11-10radeon: fix ring tail overflow issue since alignmentDave Airlie
2008-11-10radeon: disable HDP read cache for nowDave Airlie
2008-11-10radeon: force all ring writes to 16-dword alignment.Dave Airlie
2008-11-10radeon: fixup vram visible calculation to take a/c pinned objects for nowDave Airlie
2008-11-09radeon: add more packet3 relocations handlingJerome Glisse
2008-11-09libdrm-radeon: update libdrm-radeon to match current CS relocation structuresJerome Glisse
2008-11-06radeon: lib radeon add bo & cs gem backendJerome Glisse
2008-11-05radeon: libdrm_radeon add handle to debug stringJerome Glisse
2008-11-05Merge branch 'modesetting-gem' of ssh://git.freedesktop.org/git/mesa/drm ↵Jerome Glisse
into modesetting-gem
2008-11-03radeon: make build againDave Airlie
2008-11-03radeon: fix ROP values for the paint ROPDave Airlie
2008-11-03radeon: add mtrr support for VRAM aperture.Dave Airlie
2008-11-03radeon: disable AGP for certain chips if not specified until we figure it outDave Airlie
2008-11-03radeon: disable debugging messageDave Airlie
2008-11-03radeon: commit ring after emitting the buffer discardsDave Airlie
2008-11-03radeon: add more HDMI bitsDave Airlie
2008-11-03radeon: set dma bufs bo type to a kernel typeDave Airlie
2008-11-03radeon: enable DVI-D + HDMI connectors.Dave Airlie
This allows the rs690 to work on DVI
2008-11-03radeon: setup isync cntl properlyDave Airlie
2008-11-03radeon: add more debuggingDave Airlie
2008-11-03radeon: overhaul ring interactionsDave Airlie
emit in 16-dword blocks, emit irqs at same time as everything else
2008-11-03radeon: fix race in sysfsDave Airlie
2008-11-03radeon: add proc debugging for interrupts/ringDave Airlie
2008-11-03radeon: only enable dynclks if asked forDave Airlie
2008-11-03radeon: add wait rendering APIDave Airlie
2008-11-03radeon: rs480 fixes for bus masteringDave Airlie
2008-11-03radeon: remove unused gem indirect ioctlDave Airlie
2008-11-03radeon: fix some warningsDave Airlie