summaryrefslogtreecommitdiff
AgeCommit message (Collapse)AuthorFilesLines
2009-03-02Bump version to 2.6.3.2.6.3Eric Anholt1-1/+1
2009-03-02Only allocate pixmaps aligned for tiling when requested by DRI2 GetBuffers.Eric Anholt3-27/+44
This saves massive quantities of memory on pre-965 since the DRI2 tiling enable caused the minimum size of any pixmap to be 1MB. (cherry picked from commit 5bfd73cd31ba197a62f549cdbad1a1270b571027)
2009-03-02Disable fb resizing for DRI1-only server so that DRI1 can initialize.Eric Anholt1-0/+4
(cherry picked from commit 70e0261208654c6c875ad462da2734c6aa9eeb96)
2009-02-24Bump version to 2.6.2 for release.2.6.2Eric Anholt1-1/+1
2009-02-24Fix distcheck from drmmode_display.h deletion.Eric Anholt1-2/+1
(cherry picked from commit 9d8e5c21a1688b915bf39261d4c3b0bf2906daef)
2009-02-24Update kms to work with drmModeModeInfo API update.Kristian Høgsberg1-3/+3
(cherry picked from commit a6b31f38ebf470c61de0e10b0ce2af0d7ee1684b)
2009-02-24Don't do AdjustFrame in KMS mode.Eric Anholt1-0/+3
This was hit by xv86vm's SwitchMode path, and for that the CRTC offsets get set at mode setting time anyway. (cherry picked from commit 53108994616d9751ac3a29fd61eb269cfaeab967)
2009-02-24KMS: Fix bug that prevented EDID data from getting propagated.Kristian Høgsberg1-1/+2
(cherry picked from commit 73bc7f113969834d00cd92be8374dbadc62f96a9)
2009-02-24Move disable_render_standby to EnterVT instead of startup.Helge Bahmann1-3/+3
Otherwise, with a pre-2.6.28 older kernel the disable would be lost at resume time and cause hangs. Bug #20214 (cherry picked from commit 81c652e9a666a7459bcc5217c8a5ec518b6e00da)
2009-02-24uxa: Ask for BOs ready for rendering for pixmaps.Eric Anholt2-2/+2
The assumption is that we're almost always accelerating our drawing to new pixmaps (fill, copy, etc.). (cherry picked from commit 0621ba12a3b694720e67a49b25ca52f0e09b3802)
2009-02-24KMS: Hook up rotated shadow buffers.Kristian Høgsberg1-24/+26
(cherry picked from commit 5018d0f16cb8b44c743b5b37d194fe806d955568)
2009-02-24Access the Xv buffer through the GTT for the non-KMS case.Kristian Høgsberg1-21/+42
(cherry picked from commit beca598bc2848093b710bd47828d622205d273df)
2009-02-24Fix i915 textured video to work with the i830_memory -> bo change.Kristian Høgsberg1-4/+5
Forgot to update i915_video.c in 872aadc7102bd5131e1582ede081e22672911ba2. (cherry picked from commit e97e2571703e3d6188bf18f211b793fc50383f9c)
2009-02-24Dont allocate overlay registers in KMS mode.Kristian Høgsberg3-8/+7
(cherry picked from commit 96da26b6813a8c1da8a43036c375aa0d2bb70f16)
2009-02-24Make Xv used a buffer object instead of i830_memory.Kristian Høgsberg4-60/+53
We still pin the buffer object in case of overlay, but for textured video we're now no longer using i830_memory for Xv anymore. (cherry picked from commit 872aadc7102bd5131e1582ede081e22672911ba2)
2009-02-24KMS: Hook up crtc::gamma_set.Kristian Høgsberg1-1/+12
(cherry picked from commit 723b6065093adb56a2d7204bd990ceae41bfafc9)
2009-02-24Don't disable vga centering bit.Ma Ling1-4/+0
commit id b9f5915ce812144ffd9d2aa42e8ba856129c35e, which resolved bug #17235, but generate new regression-bug #19715. This patch intends to resolve bug #17235, and avoid regression as well. We have successfully re-tested it for bug #17235 and #19715 respectively. (cherry picked from commit 3aa8591abfbe8db0f13912910c850fdd748808df)
2009-02-24TV quirk for HP Compaq nx6310Zhenyu Wang1-0/+2
(cherry picked from commit 9fe5fca3fe761a4f11857d9766138a60f471a9e6)
2009-02-24Fix front buffer memset() for non-KMS case.Kristian Høgsberg1-1/+1
Missed the pI830->FbBase condition when removing the KMS hook. (cherry picked from commit 5c370091620b38447172ebeffbc6ed3256e86c9d)
2009-02-24Un-revert the I915_SETPARAM_NUM_USED_FENCES commit reverted by accident.Kristian Høgsberg6-27/+84
Oops, my bad. Reverted 8d4bc36fae50b09a73ba2cfab920adb32141a358 since my kernel doesn't yet have the new param, committed the revert by accident. (cherry picked from commit 2013799b20599a58de48cb21a5a389e898a58af1)
2009-02-24Fix last-minute "cleanup" that broke the patch.Kristian Høgsberg7-101/+59
(cherry picked from commit 127330bfd53ac7571bdd12a551142528b972893f)
2009-02-24Implement front buffer resize for KMS.Kristian Høgsberg5-205/+154
This adds back the resize hook so we can resize the front buffer under kernel mode setting as well. The patch also pulls the drmmode_* structs from drmmode_display.h into drmmode_display.c and eliminates the header file. (cherry picked from commit 0cb87ccfe97b0e016e47dcf236fd5ce78dddfc4b)
2009-02-24dri2: Use modesetting's master fd instead of opening our own non-master.Eric Anholt1-1/+11
This fixes failure to auth DRI2 clients under KMS. (cherry picked from commit 66bc44e8f9a0505c0b11b8042243ca74079da85f)
2009-02-24Support tiled back/depth on 915-class hardware with DRI2.Jesse Barnes6-27/+84
Set alignments, tile settings and flags correctly in the 2D driver to support tiled rendering. UXA's create pixmap function currently assumes the worst about the alignment constraints; that should probably be fixed. Some of the 1M alignment fixes could probably be done more cleanly as well. (cherry picked from commit 8d4bc36fae50b09a73ba2cfab920adb32141a358)
2009-02-24Don't forget the new state bos in check_aperture.Eric Anholt1-35/+47
They're tiny so it shouldn't have been a problem, but play it safe. This is another <5% loss on top of the previously reported value, bringing the whole series to about 8%. (cherry picked from commit 6c0ca1676bf60529dd331cc739abdf68fa9e918d)
2009-02-24Fixup bogus VBT modes when detectedJesse Barnes1-0/+6
Several VBT modes out in the wild have H or VSyncEnd values greater than the H or VTotal value. This clearly ends up creating a bad mode, causing some panels to either ignore the timing or display some sort of corrupt image. Check for these cases and fix them up by default, making things work for several Dell and Sony machines. Fixes FDO bug #17292. (cherry picked from commit 57a02b50c60c10a25ff0f3cd93af9f37fa0d1b11)
2009-02-24Quirk MSI IM-945GSE-A LVDS, TV outputs.Bill Nottingham1-0/+17
The IM-945GSE-A claims to have a TV output, and always claims a connected LVDS output. It has neither. (cherry picked from commit e20e1cf76fb00ba4f933a1ed6d1a4896be346c91)
2009-02-24quirk for AOpen MP45Vincent Mussard1-0/+2
(cherry picked from commit 05ff561234cc2b93fe1ea2a35041fa2e119a7e38)
2009-02-24Move i965 render sampler state to BOs.Eric Anholt3-122/+85
This eliminates the pinned memory allocation for 965 render state. (cherry picked from commit fbf003ef2767a1a9f5e4064f04a17992030d8f5c)
2009-02-24Move i965 render kernels to BOs.Eric Anholt1-45/+30
(cherry picked from commit befd4ad8beae39377f804e06c7cbd926ec4251db)
2009-02-24Move 965 render unit state to BOs.Eric Anholt2-138/+269
This is a first step in a series of changes to avoid requiring a pinned object, which gets in the way of doing non-root KMS. This change appears to result in about a 2-6% loss in x11perf -aa10text, which better algorithms in libdrm could make up for (it hasn't really had to deal with code this bad before). (cherry picked from commit 5d705de5d11297f4d6bd5237fb67619e703745bc)
2009-02-24Remove 965 render wm scratch space, which was just unused.Eric Anholt1-10/+3
(cherry picked from commit 64b08ed5ad46ccd76964972e39bfed8721a920aa)
2009-02-24Fix build with server 1.4.Eric Anholt2-1/+6
Debian "unstable" is still stuck with this ancient version. (cherry picked from commit f126aabdf8952177bb15f392041da7a7094eb31b)
2009-02-24Use drm_intel_bo_subdata to put render vb data in.Eric Anholt1-11/+4
This improves performance by avoiding repeated map/unmap cycles, which are a bit expensive on my machine with lock debugging on in the kernel. It could do much better if we did more than 18 or so floats at a time. (cherry picked from commit b6f3ce32e295929f461a7bc37e61f126fb51e4aa)
2009-02-24Move i965 render vb setup to use time, and decouple state emit from it.Eric Anholt1-34/+55
The require_space had failed since it only checked for the space required by the batch emits in the function itself, but not in the i965_emit_composite_state() that it called (the state we were concerned about having set up for that 12 * 4 dwords to follow!). This is replaced by intel_batch_start_atomic(), which will catch such mistakes in the future. (cherry picked from commit 9a8bbb1951ad0ca0a9407a97348fc7fa03127900)
2009-02-24Move i965 render transform setup from emit_composite_state to prepare_composite.Eric Anholt1-28/+22
(cherry picked from commit 3d739597c4f5817079efd9067ad5db2f4105f765)
2009-02-24i965: Pull check_aperture out to a separate function and make it dtrt.Eric Anholt1-62/+32
Previously it wouldn't count the pixmaps that were about to be used, which is pretty much the only purpose of having the pain around. This also eliminates the check_twice confusion with emit_batch_header_for_composite(). (cherry picked from commit e20f7278f3abb44a3a151ac91f83c45cf1a2745a)
2009-02-24Move filter computation from emit_batch_header to prepare_composite.Eric Anholt1-21/+31
(cherry picked from commit 013e2adfbf955cb21450b610091542ebd54392c2)
2009-02-24Use intel_emit_reloc from video to prettify 965 render bind_bo setup.Eric Anholt3-33/+33
(cherry picked from commit a340fe5e4227ebea5493e658eb6289624b07ab0b)
2009-02-24Move i965 render state bo setup back to prepare_composite.Eric Anholt1-66/+60
We want the objects to be created once per prepare/done both for efficiency and so we can handle aperture checking better. (cherry picked from commit aefe198ca427a5ad69717f49948eb3ede713bb28)
2009-02-24Do check_aperture_space and batch_start_atomic for i965 video.Eric Anholt1-2/+27
This increases the overhead for video in the presence of cliprects, but we were already doing nasty things in that case and don't seem to care. This could fix potential bad rendering or hangs with video, particularly with DRI2. (cherry picked from commit 946c7ef8170e74ac178c83b1465242d57fa86f2e)
2009-02-24Move 965 video setup to a separate function so we can move it around.Eric Anholt1-172/+174
(cherry picked from commit 7be668179a12918918cad863f6936ced4ab78dbf)
2009-02-24Assign rotation memory dri_bo to rotation pixmap.Keith Packard1-0/+3
As the rotation memory and rotation pixmap are allocated separately (to make rotation at startup work), the allocate dri_bo needs to be set in the pixmap for acceleration to work. This restores the performance in rotated modes. Signed-off-by: Keith Packard <keithp@keithp.com> (cherry picked from commit a4b023c17b9c3bd65fb9466ddb8a953f60244402)
2009-02-24Resize framebuffer on screen size change (requires UXA and DRI2)Keith Packard6-153/+184
Signed-off-by: Keith Packard <keithp@keithp.com> (cherry picked from commit 8237faf8f3ca73ecdf0ef009a7d361b318726f6f)
2009-02-24Make i830_allocate_memory take tiling parameters.Keith Packard5-190/+189
This eliminates the separate i830_allocate_memory_tiled function which means that all memory objects will have tiling parameters set correctly. Signed-off-by: Keith Packard <keithp@keithp.com> (cherry picked from commit 21bd4e8974e4c0e83f5f95adb0fc17290444caf5)
2009-02-24warning fix.Eric Anholt1-1/+0
(cherry picked from commit d96f774d1bb39640486c72338fe8b19ee1ceaa23)
2009-02-24Fix compile failure after 45f45c73469f1bd46a1b6fb206f2e9e5e4fd66b3Eric Anholt1-1/+0
(cherry picked from commit 750d8e105831718d4a44a145fdb87571fa9f9d8e)
2009-02-24Remove logical context setup.Eric Anholt3-22/+0
This should be a noop. If it wasn't a noop, it means that on pre-g33 chipsets we were spamming some data into a page of system memory because we used a virtual instead of a physical address. It was also supposed to not work when we submit it from a batchbuffer, as we have been doing for some time now. This code has existed since about the beginning of the driver's existence, with no justification. (cherry picked from commit 45f45c73469f1bd46a1b6fb206f2e9e5e4fd66b3)
2009-01-21Bump version to 2.6.1.xf86-video-intel-2.6.1Eric Anholt1-1/+1
2009-01-21Don't run in KD_TEXT mode even with KMSJesse Barnes1-8/+0
Leaving the VT in KD_TEXT mode keeps the kernel's blanking code active, so when a DPMS event happens, the fb console is restored rather than X's configuration. On the downside it means the kernel won't print messages in the background, which would be visible if a panic or emergency switch occurred. The proper fix here is a new kernel mode, which we can move to when ready. (cherry picked from commit 253b8db298f38676e47dc902534465054f7b58b8)
generated by cgit v1.2.3 (git 2.25.1) at 2024-12-01 09:11:12 +0000