atom D525 trouble

atom D525 trouble

[Patrick Welche]

In ancient pre-KMS times, I was able to get a particular pineview-based
(Atom D525) desktop computer & monitor to run at decent resolutions.
Recently, I have tried the 2.99.917 snapshot, and the best I can
manage is 1024x768.

Bad parts of the system are:
- the monitor's EDID is invalid
- it presents a
  - low resolution LVDS connected to a DVI socket(!)
  - a high resolution VGA connector

It used to be sufficient simply to

Section "Monitor"
        Identifier      "VGA"
        Option "Ignore" "false"
EndSection

Section "Monitor"
        Identifier      "LVDS"
        Option "Ignore" "true"
EndSection

Section "Monitor"
        Identifier      "LVDS-1"
        Option "Ignore" "true"
EndSection

in xorg.conf.


How is the connector / resolution meant to be selected (what are the rules)?

Any hints on where to look / how to improve the situation?


Cheers,

Patrick

debugging highlights:

DRM debug in intel_parse_bios: Using VBT from OpRegion: $VBT PINEVIEW       d
DRM debug in parse_general_features: BDB_GENERAL_FEATURES int_tv_support 0 int_crt_support 1 lvds_use_ssc 0 lvds_ssc_freq 100000 display
_clock_mode 0 fdi_rx_polarity_inverted 0
DRM debug in parse_general_definitions: crt_ddc_bus_pin: 2
DRM debug in parse_lfp_panel_data: Found panel mode in BIOS VBT tables:
DRM debug in drm_mode_debug_printmodeline: Modeline 0:"1024x768" 0 65000 1024 1048 1184 1344 768 771 777 806 0x8 0xa
DRM debug in parse_lfp_panel_data: VBT initial LVDS value 40000300
DRM debug in parse_lfp_backlight: VBT backlight PWM modulation frequency 200 Hz, active high, min brightness 0, level 0
DRM debug in parse_sdvo_panel_data: Found SDVO panel mode in BIOS VBT tables:
DRM debug in drm_mode_debug_printmodeline: Modeline 0:"1600x1200" 0 162000 1600 1664 1856 2160 1200 1201 1204 1250 0x8 0xa
DRM debug in parse_sdvo_device_mapping: No SDVO device info is found in VBT
DRM debug in parse_mipi: No MIPI BDB found
DRM debug in intel_dsm_pci_probe: no _DSM method for intel device
...
DRM debug in intel_connector_check_state: [CONNECTOR:8:LVDS-1]
DRM debug in intel_connector_check_state: [CONNECTOR:12:VGA-1]
DRM debug in check_encoder_state: [ENCODER:9:LVDS-9]
DRM debug in check_encoder_state: [ENCODER:13:DAC-13]
DRM debug in check_crtc_state: [CRTC:5]
DRM debug in check_crtc_state: [CRTC:7]
...
DRM debug in drm_helper_probe_single_connector_modes: [CONNECTOR:8:LVDS-1]
DRM debug in intel_lvds_detect: [CONNECTOR:8:LVDS-1]
DRM debug in drm_helper_probe_single_connector_modes: [CONNECTOR:8:LVDS-1] probed modes :
DRM debug in drm_mode_debug_printmodeline: Modeline 14:"1024x768" 60 65000 1024 1048 1184 1344 768 771 777 806 0x8 0xa
DRM debug in drm_helper_probe_single_connector_modes: [CONNECTOR:12:VGA-1]
DRM debug in intel_crt_detect: [CONNECTOR:12:VGA-1] force=1
DRM debug in intel_crt_detect: CRT detected via hotplug
DRM error in drm_edid_block_valid: EDID checksum is invalid, remainder is 194
drm kern error: Raw EDID:
00 ff ff ff ff ff ff 00 3f ff ff ff ff ff ff ff
...
DRM debug in intel_crt_get_edid: CRT GMBUS EDID read failed, retry using GPIO bit-banging
DRM debug in intel_gmbus_force_bit: enabling bit-banging on i915 gmbus vga. force bit now 1
DRM debug in intel_gmbus_force_bit: disabling bit-banging on i915 gmbus vga. force bit now 0
DRM debug in drm_helper_probe_single_connector_modes: [CONNECTOR:12:VGA-1] probed modes :
DRM debug in drm_mode_debug_printmodeline: Modeline 19:"1024x768" 60 65000 1024 1048 1184 1344 768 771 777 806 0x40 0xa
DRM debug in drm_mode_debug_printmodeline: Modeline 17:"800x600" 60 40000 800 840 968 1056 600 601 605 628 0x40 0x5
DRM debug in drm_mode_debug_printmodeline: Modeline 16:"800x600" 56 36000 800 824 896 1024 600 601 603 625 0x40 0x5
DRM debug in drm_mode_debug_printmodeline: Modeline 18:"848x480" 60 33750 848 864 976 1088 480 486 494 517 0x40 0x5
DRM debug in drm_mode_debug_printmodeline: Modeline 15:"640x480" 60 25175 640 656 752 800 480 489 492 525 0x40 0xa
DRM debug in drm_setup_crtcs: 
DRM debug in drm_enable_connectors: connector 8 enabled? yes
DRM debug in drm_enable_connectors: connector 12 enabled? yes
DRM debug in intel_fb_initial_config: looking for cmdline mode on connector 8
DRM debug in intel_fb_initial_config: looking for preferred mode on connector 8
DRM debug in intel_fb_initial_config: connector LVDS-1 on crtc 7: 1024x768
DRM debug in intel_fb_initial_config: fallback: cloned configuration
DRM debug in intel_fb_initial_config: Not using firmware configuration
DRM debug in drm_target_preferred: looking for cmdline mode on connector 8
DRM debug in drm_target_preferred: looking for preferred mode on connector 8
DRM debug in drm_target_preferred: found mode 1024x768
DRM debug in drm_target_preferred: looking for cmdline mode on connector 12
DRM debug in drm_target_preferred: looking for preferred mode on connector 12
DRM debug in drm_target_preferred: found mode 1024x768
DRM debug in drm_setup_crtcs: picking CRTCs for 4096x4096 config
DRM debug in drm_setup_crtcs: desired mode 1024x768 set on crtc 7
DRM debug in drm_setup_crtcs: desired mode 1024x768 set on crtc 5
DRM debug in intelfb_create: no BIOS fb, allocating a new one
DRM debug in drm_framebuffer_reference: FB ID: 22
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: atom D525 trouble

[Chris Wilson]

On Tue, Apr 28, 2015 at 11:46:59AM +0100, Patrick Welche wrote:
> In ancient pre-KMS times, I was able to get a particular pineview-based
> (Atom D525) desktop computer & monitor to run at decent resolutions.
> Recently, I have tried the 2.99.917 snapshot, and the best I can
> manage is 1024x768.
> 
> Bad parts of the system are:
> - the monitor's EDID is invalid
> - it presents a
>   - low resolution LVDS connected to a DVI socket(!)
>   - a high resolution VGA connector
> 
> It used to be sufficient simply to
> 
> Section "Monitor"
>         Identifier      "VGA"
>         Option "Ignore" "false"
> EndSection
> 
> Section "Monitor"
>         Identifier      "LVDS"
>         Option "Ignore" "true"
> EndSection
> 
> Section "Monitor"
>         Identifier      "LVDS-1"
>         Option "Ignore" "true"
> EndSection
> 
> in xorg.conf.
> 
> 
> How is the connector / resolution meant to be selected (what are the rules)?

If we cannot inherit a config, one is chosen such that a similar pixel
size is presented across all outputs - a "cloned" config.
 
> Any hints on where to look / how to improve the situation?

First try kernel cmdline video=LVDS-1:d to disable the spurious LVDS-DVI
connection (I am sorry it will never work with the current code without
some hacks as we stongly expect a LVDS connection to be a laptop panel).

To fixup the missing EDID, either find a genuine one for your panel (or
fake one) and use drm_kms_helper.edid_firmware=VGA-1:my-edid.bin or
provide sufficent Modelines in either your xorg.conf or through xrandr.
-Chris

-- 
Chris Wilson, Intel Open Source Technology Centre
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx