| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
| |
set Interlace flag for ModeLines created for interlaced Detailed
Timings
|
|
|
|
|
|
| |
add support for Short Video Descriptor formats 1-34 (as defined by
EIA/CEA-861-B) in Video Data Blocks of CEA EDID Timing Extension
|
|
|
|
|
|
| |
add support for Detailed Timing Descriptors in CEA EDID Timing
Extension
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
|
| |
o remove all x86emu and vga softbootloader code, and always use the
LRMI interface which was previously the backup one on 32-bit x86;
libx86 is now used instead of liblrmi, allowing the use of LRMI
interface on non-x86 hosts as well (fixes Mandriva bug #53866, which
was caused by a bug in the removed code)
|
| |
|
|
|
|
|
|
| |
use -sharevts for X server when plymouth is active, as VT switching
does not work at that point (fixes Mandriva bug #53736)
|
|
|
|
|
|
|
| |
disable glx module to speed up X startup, especially if a proprietary
glx module is in use (fixes a timeout I observed when booting One
through a slow USB bus)
|
| |
|
|
|
|
|
|
|
|
|
| |
ignore the Manufacturer Specified Range Timing descriptor if the
first detailed timing descriptor appears to violate it (this probably
means that the descriptor is actually something else or in a
different format; this fixes Lenovo W500 detection, reported by Udo
Rader)
|
| |
|
|
|
|
|
|
|
| |
probe ports 0-2 by default instead of just 0-1 (many NVIDIA cards
seem to be using port 2 and the changes in monitor-get-edid-using-vbe
should make it safe for old problematic NVIDIA cards as well)
|
|
|
|
|
|
| |
check if the port supports DDC before trying to read EDID data (this
should prevent problems when probing nonexistent ports on old cards)
|
|
|
|
|
|
| |
update x86emu from xserver git (fixes issues at least with newish NVIDIA
cards)
|
| |
|
|
|
|
|
| |
reports that it is exact, making it work with nouveau driver
|
|
|
|
|
| |
enabled
|
| |
|
|
|
|
|
|
| |
o instead of checking vt >= 7 to know if X is running,
check wether $DISPLAY is set
|
| |
|
|
|
|
|
| |
o handle parsing of EDIDs found in "xrandr --prop" or Xorg.log
|
|
|
|
|
| |
o compute and display the "dpi" of the preferred modelines
|
| |
|
| |
|
| |
|
|
|
|
|
| |
(Remi Collet and Ville Skytta)
|
|
|
|
|
| |
- prefix mandriva bug numbers with mdvbz
|
| |
|
|
|
|
|
| |
available (Remi Collet)
|
| |
|
|
|
|
|
|
|
|
| |
o probe "using DMI" before "using X"
- monitor-probe-using-X:
o in last resort, get Intel BIOS mode when "BIOS panel mode is bigger than
probed programmed mode"
|
| |
|
|
|
|
|
|
| |
o call monitor-get-edid-using-vbe with a range of ports, it stops on first
success (by default it tries port 0 then port 1)
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
o minimal support for getting EDID from different DDC port
(experimental, need testing before using it in monitor-edid)
this is inspired from SuSE's hwinfo. note:
- we may need to not redo vbe_check_vbe_info() for each vbe_get_edid_info() if
we want to get many ports. It would need a little re-architecturing
(since returning list in C is not great fun)
- hwinfo is probing ports 0,1 by default,
(hwinfo says: some BIOSes crash when you try more)
and 0,1,2 on Dell and nvidia-based laptops
|
|
|
|
|
|
|
|
| |
(#34417)
(drop the monitor-parse-edid from NEWS since it's not that useful anymore
since monitor-edid will not give it EDID containing "<not supported>")
|
|
parse "<not supported>" /proc/acpi/video/**/EDID files will give a more
understandable error message
|