[ANNOUNCE] xf86-video-intel 2.4.0 - Error underrun on pipe A

Barry Scott barry.scott at onelan.co.uk
Wed Jul 23 08:25:21 PDT 2008


This version of the driver also gives me a black screen
or a screen slowly changing between full screen colours.

This happens in 1 in 4 cold boots of a 965GME/GL
the logs are full of:

(EE) intel(0): underrun on pipe A!

Full logs and conf are available on:

http://www.onelanftp.co.uk/bscott/xorg-2.4.0-pipe-underrun/

ntb10117 screen has sync but is black
ntb10457 screen is flashing between full screen colours

Barry

Zhenyu Wang wrote:
> Hi, here's xf86-video-intel 2.4.0 release.
>
> Main new features included within this release are
> - New Intel 4 series chipset support
> - Improved 965 exa render performance
> - Integrated HDMI support
> - SDVO-HDMI support (e.g ASUS P5E-VM board)
>
> With other kinds of cleanups and bug fixes. Credit goes to
> people listed in following Changelog. Thanks!
>
> Please try and report regression back to us.
>
> ---
>
> Full Changelog from 2.3.2:
>
> Alan Coopersmith (1):
>       Define DEFFILEMODE for OS'es that don't have it
>
> Bryce Harrington (4):
>       Quirks from Ubuntu/Dell
>       More quirks from ubuntu/dell
>       More Pipe A force quirks
>       Two more Dell quirks
>
> Carl Worth (7):
>       Enumerate all possible cc_state objects
>       Enumerate all possible src,mask sampler state pairs
>       Enumerate all possible wm_state objects
>       Associate one sf_state object with each sf_kernel
>       Move VERTEX_BUFFERS setup from prepare_composite to composite
>       Allow for multiple vertex buffers (though only use one for now)
>       Use up to 256 separate vertex buffers
>
> Dave Airlie (1):
>       intel: fix drm check.
>
> David Schleef (1):
>       Bug #15219: Fix default overlay CSC factors, and use BT709 if available.
>
> Eric Anholt (55):
>       Disable a bunch of clock gating disables on IGD_GM, which doesn't need them.
>       Revert "Use mprotect on unbound AGP memory to attempt to catch use while unbound."
>       Fix 965+ rendering issues with DRI disabled.
>       Add a "total" field to intel_idle output (set when any unit is busy).
>       Add SDVO TV-out registers.
>       Add WIP SDVO TV-out support.
>       SDVO: Handle RGB outputs that are really TV outputs, and select a TV format.
>       SDVO: Fix mixups with input and output channels.
>       SDVO: Use the TV reference clock for SDVO TV outputs.
>       SDVO: Fix lols TV mode list handling.
>       SDVO: Set adjusted_mode in mode_fixup from the preferred input timings for TV.
>       SDVO: Rename output_dtd variable to input_dtd, which is its new job.
>       Add dolt to speed up compile time and reduce insanity.
>       Keep static copies of the 965 render programs in video memory.
>       Add copyright information for recent editors of this file.
>       Statically allocate the sampler default color, which we never change.
>       Fix compiler warning from 24-bit lvds change.
>       Fix the offset to sampler default color, and a compiler warning.
>       Rename some variables in i965_render.c for clarity.
>       Make the VS unit state static for 965 render.
>       Move the 965 static variables local to the function they're used in.
>       Move the binding table for 965 to the state structure.
>       Move the vertex buffer to the state structure as well.
>       Abstract surface setup into a separate function.
>       Update dolt from git, fixing fallback to libtool.
>       Make the binding table and surface state be arrays to reduce syncing.
>       Bug #15807: Fix use of the ring while VT-switched, hit by fast user switching.
>       Add i915 support to intel_idle.
>       Formatting fixes for intel_idle.
>       Just remove the mprotect kludge.
>       Replace a couple of wait-for-ring-idles with a single function to do so.
>       Move DSPARB setup into a separate function, and save/restore it at VT switch.
>       Remove duplicated i830_stop_ring()/SetHWOperatingState() in EnterVT().
>       Move the remaining two lines of SetHWOperatingState to the caller.
>       Remove gratuitous wait_ring_idle after I830Sync.  Syncing implies that.
>       Move BIOS scratch register setup to EnterVT instead of PreInit.
>       Initialize clock gating from EnterVT and save/restore it appropriately.
>       Remove SVG_WORK_CONTROL init.
>       Move debug clock printout from ErrorF to X_INFO.
>       Set up/restore PWRCTXA from enter/leavevt not server start/exit.
>       Add DisplayPort registers.
>       Initial HDMI work.  Not currently hooked up at startup.
>       The phase shift its are now reserved, and add HDMI clock limits.
>       Add pixel multiplier support for HDMI
>       Set the sync active bits like we're supposed to, matching the BIOS.
>       Get HDMI output working.
>       Fix hdmi POSTING_READ to use the register number instead of the register value.
>       Automatically detect the presence of HDMI.
>       Fix a crash in i830_sdvo_init error paths by setting up dev_priv earlier.
>       Fix DSPARB setting on 845/865, which have only the AEND field and 96 entries.
>       Get prototype for i830_bios_get_tv().
>       Fix uninitialized-use warning in i830_debug.c ring dumping.
>       Fix distcheck.
>       Remove VGA regs from debug output.
>       Add little hotplug detector app.
>
> Hong Liu (6):
>       SDVO: Fix DDC bus selection based on output priority list.
>       Bug #14935: Fix i9xx reference clock for spread spectrum.
>       fix possible segfault in I830FreeScreen
>       Fix DSPARB change on 965G
>       Set SDVO sync polarity to default on 965
>       Fix SDVO HDMI output.
>
> Jesse Barnes (24):
>       Initial panel fitting changes
>       Fixup backlight control and panel fitting property names
>       Use combo backlight control rather than pure legacy
>       Fix LVDS regression
>       Add a new quirk for BIOSes that reprogram regs at lid close/open time
>       Add FIFO watermark regs to register dumper
>       Save/restore cursor registers
>       Panel fitting:  fix letterbox modes
>       Revert "Disable FBC by default on 965GM"
>       Fixup power saving registers
>       Handle display FIFOs better
>       Fixup FIFO comment
>       Fixup DSPARB for 855 & 945
>       Fix TV programming:  add vblank wait after TV_CTL writes
>       Add support for keeping vblank counters sane across mode setting
>       Remove ErrorF debugging from modeset ioctl
>       Add pipe a force enable quirk for Lenovo T60
>       Add pipea force enable quirk for HP Pavilion ze4944ea
>       Improve FBC size checking
>       Improve VBIOS feature detection, add SSC support
>       Add VBIOS based TV connector detection
>       Don't disable pipe A on 855 chips
>       Choose a split for DSPARB based on the configured modes for both planes.
>       Add no LVDS quirk for Transtec Senyo 610 mini PC
>
> Joerg Sonnenberger (1):
>       Fix DRI config for NetBSD
>
> Julien Cristau (3):
>       Don't try to build XvMC if DRI is disabled
>       Fix gen4asm rule to work with a build dir
>       Link the driver with -lpciaccess and -ldrm if needed
>
> Keith Packard (41):
>       Transformed coord computed using floats. Don't move bilinear composite dst.
>       Disable cursors while switching modes
>       Compute du/dv/dw in no-mask SF prog
>       Clarify comment in exa_sf program
>       Reimplement wm program for nomask case to handle affine transforms
>       Fix the sf_mask program to compute and pass corrected uvw cooefficients
>       Remove rotation sf and wm progs
>       Use m4 to clean up gen4 asm progs. Start adding projective transform support.
>       Dont set the compr bit on 8-unit sends
>       Back to new shaders, fix urb read length
>       remove old monolithic shaders
>       remove old monolithic shaders
>       remove old shader source files
>       Fix composite with mask using new compositing thread code
>       Shrink WM thread to 32 registers and 1024 scratch space.
>       Add projective versions of the PS kernels
>       Remove extra flushing
>       Forgot to build exa_wm_src_sample_argb.g4b
>       Rename src/mask/data registers to indicate channel
>       Use symbolic names for channels in YUV code
>       Compute pixel values directly into data port
>       Use shared exa_wm code for packed yuv decode
>       Add planer video decode kernel
>       Remove sync after 965 video put.
>       Revert "Compute pixel values directly into data port"
>       Update yuv->rgb conversion programs to write to src_sample regs
>       Remove .g4b files on clean
>       Use available symbolic register name
>       Single memcpy when pitches align on planar image transfer
>       Revert "remove '#line NUM ...' in macro process"
>       Use compressed instructions for yuv conversion
>       Set some mask bits to try and get yuv conversion working
>       Skip copying on FOURCC_XVMC surfaces
>       Kludge to make planar video work - duplicate sampler state
>       Move wm_scratch first to meet alignment requirements
>       Only use FOURCC_XVMC when INTEL_XVMC is defined
>       Add a kludge-around to fix cd/wt bits in fb ptes on linux.
>       Add a bunch of 965 ring stuff to the debug dump
>       Dump batch buffers found from main ring
>       Overlay video doesn't require that the target pixmap be in video memory.
>       Use new xf86RotateFreeShadow function to clean up shadow buffers.
>
> Lukas Hejtmanek (1):
>       Fix maximum backlight issue
>
> Matthias Hopf (3):
>       Fix obviously stray '(' in non-I830_USE_EXA case.
>       Fix obviously wrong verification of DRI_SUPPORTS_CLIP_NOTIFY.
>       Include missing stdlib.h for free()
>
> Maxim Levitsky (1):
>       Fix crash triggered by dpms low power mode with hardware overlay running
>
> Mike Isely (2):
>       Implement support for 24 bit pixel format
>       Implement option to ignore external fixed mode settings
>
> Paulo Cesar Pereira de Andrade (1):
>       More compile warning fixes.
>
> Robert Lowery (1):
>       Fix TV out connection type detection
>
> Roland Test-tools Bär (1):
>       i810: Remove an effectively unused variable (only used in an incorrect free())
>
> Rémi Cardona (1):
>       Add glproto to DRI dependencies
>
> Shuang He (1):
>       Fix typo in xvmc block destroy
>
> Stefan Dirsch (1):
>       Add pipea force enable quirk for another 855GM machine.
>
> Tomas Carnecky (1):
>       Fix compile error in LeaveVT
>
> Zhenyu Wang (44):
>       xvmc: fix distcheck
>       Disable overlay on new mobile chipset
>       Disable DRI earlier if fb width > 2048
>       Include <stdint.h> in i810.h for integer types definition
>       remove '#line NUM ...' in macro process
>       Remove old sf/wm program from Makefile
>       Change default panel fitting mode to origin
>       Only check xvmc lib when xvmc is enabled.
>       Check pitch for EXA operation
>       Use I830FALLBACK when check pitch for debug
>       Fix i965 render getting dest format error
>       Revert last HP 965GM pipe A quirk
>       Revert "Add FIFO watermark regs to register dumper"
>       xvmc: remove unused dri drawable
>       xvmc: a little cleanup
>       xvmc: move block size lookup helper to generic
>       xvmc: add render dump function
>       Fix compiler warning when disable xvmc config
>       Fix compiling with server master in LVDS backlight patch
>       Add support for Intel 4 series chipsets.
>       xvmc: formalize surface type definition
>       xvmc: environ debug option support
>       xvmc: init one-time mc context once
>       xvmc: init dest and reference buffer once
>       xvmc: init mc render load indirect command once
>       xvmc: only allocate memory requiring physical address on 915G
>       xvmc: enlarge batch buffer size
>       xvmc: Don't copy on xvmc surface in PutImage
>       xvmc: fix motion_type dump for frame/field picture
>       Fix SDVOC typo
>       Give asus and eeepc backlight method higher priority
>       Fix official name for GM45 chipset
>       move FBC register dump out of display registers
>       xvmc: use vector based structure
>       Only initialize integrated TV encoder for mobile chips
>       Thinkpad R60e TV quirk via DMI info
>       Check underrun on enabled pipe
>       SSC is LVDS only
>       only check pipe underrun if vtSema is true.
>       Bug #16801: fix X crash when NoAccel on 965
>       Fix HDMI output number
>       Fix distcheck for clean doltcompile
>       shut up gcc warning
>       Bump version 2.4.0
>
> lipeng (1):
>       Fix GPIO pin usage for DDC on second HDMI port.
>
> git tag: xf86-video-intel-2.4.0
>
> http://xorg.freedesktop.org/archive/individual/driver/xf86-video-intel-2.4.0.tar.bz2
> MD5: 81a84e891c138b713947ed06556ec53c  xf86-video-intel-2.4.0.tar.bz2
> SHA1: f58223cfbd10bdf1f66ffc4a31ab9e716f13952a  xf86-video-intel-2.4.0.tar.bz2
>
> http://xorg.freedesktop.org/archive/individual/driver/xf86-video-intel-2.4.0.tar.gz
> MD5: 49e4321055d40435b0cf62c57de555e5  xf86-video-intel-2.4.0.tar.gz
> SHA1: 6ae0ab75885da1eb60612939dee93796e69115d7  xf86-video-intel-2.4.0.tar.gz
>
>   
> ------------------------------------------------------------------------
>
> _______________________________________________
> xorg mailing list
> xorg at lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/xorg




More information about the xorg mailing list