* gma500: invalid frequency for monitor
@ 2012-02-04 3:42 Michael D Labriola
2012-02-04 11:41 ` Alan Cox
0 siblings, 1 reply; 3+ messages in thread
From: Michael D Labriola @ 2012-02-04 3:42 UTC (permalink / raw)
To: Alan Cox, David Airlie; +Cc: Michael D Labriola, dri-devel, linux-kernel
I've got a VersaLogic Ocelot (Atom Z530, Menlow) and I've been test
driving the new GMA500 driver w/out much luck. I tried the old
staging driver a few times, and just compiled and tested a checkout
from Linus' tree (d12566). I've got the gma500 driver compiled as a
module, with the optional parts disabled. Here's what I'm seeing.
My system boots up, loads drm, loads gma500_gfx, and then my monitor's
OSD says "Out of Frequency". Pretty much the same behavior with the
old psb_gfx module and the new gma500_gfx module.
My system is configured with a LVDS-VGA adapter board, which is
plugged into the VGA port on a 1680x1050 LCD. The BIOS for the Ocelot
is configured for a 1280x1024 LVDS panel, which seems like a
resolution my monitor should be able to handle. I've attached
relevant kernel log with drm.debug set to 999.
I think my setup is correct, because the vesa frambuffer seems to work
correctly. All is well unless I load the gma500_gfx driver and let it
set the mode.
I'd love to help nail this down, let me know if you want me to try
anything.
[ 102.969863] [drm] Initialized drm 1.1.0 20060810
[ 110.192778] [drm:drm_pci_init],
[ 110.192863] [drm:drm_get_pci_dev],
[ 110.192930] gma500 0000:00:02.0: setting latency timer to 64
[ 110.193545] [drm:drm_get_minor],
[ 110.193863] [drm:drm_get_minor], new minor assigned 64
[ 110.193887] [drm:drm_get_minor],
[ 110.194753] [drm:drm_get_minor], new minor assigned 0
[ 110.197942] ioremap error for 0x7f7f8000-0x7f7fb000, requested 0x10, got 0x0
[ 110.206418] [drm:drm_mode_debug_printmodeline], Modeline 0:"1280x1024" 0 108000 1280 1328 1440 1688 1024 1025 1028 1072 0x8 0x0
[ 110.207738] Stolen memory information
[ 110.207757] base in RAM: 0x7f800000
[ 110.207771] size: 8060K, calculated by (GTT RAM base) - (Stolen base), seems wrong
[ 110.207789] the correct size should be: 8M(dvmt mode=3)
[ 110.216955] Set up 2015 stolen pages starting at 0x7f800000, GTT offset 0K
[ 110.217662] [drm] SGX core id = 0x01130000
[ 110.217682] [drm] SGX core rev major = 0x01, minor = 0x02
[ 110.217701] [drm] SGX core rev maintenance = 0x01, designer = 0x00
[ 110.223579] input: Video Bus as /devices/LNXSYSTM:00/device:00/PNP0A03:00/LNXVIDEO:00/input/input6
[ 110.223800] ACPI: Video Device [GFX0] (multi-head: yes rom: no post: no)
[ 110.223940] [drm] Supports vblank timestamp caching Rev 1 (10.10.2010).
[ 110.223958] [drm] No driver support for vblank timestamp query.
[ 110.223980] [drm:drm_irq_install], irq=16
[ 110.257714] [drm:drm_sysfs_connector_add], adding "LVDS-1" to sysfs
[ 110.257932] [drm:drm_sysfs_hotplug_event], generating hotplug event
[ 110.269604] [drm:psb_intel_sdvo_read_byte], i2c transfer returned -6
[ 110.269633] [drm:psb_intel_sdvo_init], No SDVO device found on SDVOB
[ 110.561304] [drm:drm_helper_probe_single_connector_modes], [CONNECTOR:7:LVDS-1]
[ 110.586242] [drm:drm_helper_probe_single_connector_modes], [CONNECTOR:7:LVDS-1] probed modes :
[ 110.586274] [drm:drm_mode_debug_printmodeline], Modeline 10:"1280x1024" 40 71500 1280 1328 1440 1688 1024 1025 1028 1072 0x8 0x0
[ 110.586314] [drm:drm_setup_crtcs],
[ 110.586335] [drm:drm_enable_connectors], connector 7 enabled? yes
[ 110.586357] [drm:drm_target_preferred], looking for cmdline mode on connector 7
[ 110.586380] [drm:drm_target_preferred], looking for preferred mode on connector 7
[ 110.586402] [drm:drm_target_preferred], found mode 1280x1024
[ 110.586423] [drm:drm_setup_crtcs], picking CRTCs for 2048x2048 config
[ 110.586450] [drm:drm_setup_crtcs], desired mode 1280x1024 set on crtc 4
[ 110.586513] gma500 0000:00:02.0: allocated 1280x1024 fb
[ 110.586819] fbcon: psbfb (fb0) is primary device
[ 110.596288] [drm:drm_crtc_helper_set_config],
[ 110.596306] [drm:drm_crtc_helper_set_config], [CRTC:3] [NOFB]
[ 110.637408] [drm:drm_crtc_helper_set_config],
[ 110.637427] [drm:drm_crtc_helper_set_config], [CRTC:4] [FB:12] #connectors=1 (x y) (0 0)
[ 110.637467] [drm:drm_crtc_helper_set_config], crtc has no fb, full mode set
[ 110.637484] [drm:drm_crtc_helper_set_config], modes are different, full mode set
[ 110.637500] [drm:drm_mode_debug_printmodeline], Modeline 0:"" 0 0 0 0 0 0 0 0 0 0 0x0 0x0
[ 110.637529] [drm:drm_mode_debug_printmodeline], Modeline 11:"1280x1024" 40 71500 1280 1328 1440 1688 1024 1025 1028 1072 0x8 0x0
[ 110.637564] [drm:drm_crtc_helper_set_config], encoder changed, full mode switch
[ 110.637580] [drm:drm_crtc_helper_set_config], crtc changed, full mode switch
[ 110.637599] [drm:drm_crtc_helper_set_config], [CONNECTOR:7:LVDS-1] to [CRTC:4]
[ 110.637618] [drm:drm_crtc_helper_set_config], attempting to set mode from userspace
[ 110.637634] [drm:drm_mode_debug_printmodeline], Modeline 11:"1280x1024" 40 71500 1280 1328 1440 1688 1024 1025 1028 1072 0x8 0x0
[ 110.637678] [drm:drm_crtc_helper_set_mode], [CRTC:4]
[ 110.658501] [drm:drm_mode_debug_printmodeline], Modeline 11:"1280x1024" 40 71500 1280 1328 1440 1688 1024 1025 1028 1072 0x8 0x0
[ 110.699109] [drm:drm_crtc_helper_set_mode], [ENCODER:8:LVDS-8] set [MODE:11:1280x1024]
[ 110.977395] [drm:drm_calc_timestamping_constants], crtc 4: hwmode: htotal 1688, vtotal 1072, vdisplay 1024
[ 110.977415] [drm:drm_calc_timestamping_constants], crtc 4: clock 71500 kHz framedur 25307776 linedur 23608, pixeldur 13
[ 110.977439] [drm:drm_crtc_helper_set_config], Setting connector DPMS state to on
[ 110.977453] [drm:drm_crtc_helper_set_config], [CONNECTOR:7:LVDS-1] set DPMS on
[ 111.032366] Console: switching to colour frame buffer device 160x64
[ 111.064932] fb0: psbfb frame buffer device
[ 111.064944] drm: registered panic notifier
[ 111.066827] gma500 0000:00:02.0: Backlight lvds set brightness 7a120000
[ 111.066860] [drm] Initialized gma500 1.0.0 2011-06-06 for 0000:00:02.0 on minor 0
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: gma500: invalid frequency for monitor
2012-02-04 3:42 gma500: invalid frequency for monitor Michael D Labriola
@ 2012-02-04 11:41 ` Alan Cox
2012-02-05 3:05 ` Michael Labriola
0 siblings, 1 reply; 3+ messages in thread
From: Alan Cox @ 2012-02-04 11:41 UTC (permalink / raw)
To: Michael D Labriola
Cc: Alan Cox, David Airlie, Michael D Labriola, dri-devel,
linux-kernel
> My system boots up, loads drm, loads gma500_gfx, and then my monitor's
> OSD says "Out of Frequency". Pretty much the same behavior with the
> old psb_gfx module and the new gma500_gfx module.
They are all basically the same logic until the most recent release which
at least queries the monitor for the actual system ports it knows how to.
>
> My system is configured with a LVDS-VGA adapter board, which is
> plugged into the VGA port on a 1680x1050 LCD. The BIOS for the Ocelot
> is configured for a 1280x1024 LVDS panel, which seems like a
> resolution my monitor should be able to handle. I've attached
> relevant kernel log with drm.debug set to 999.
I don't have a setup with an LVDS-VGA adapter board to test, but my guess
would be that even if we pick a mode the panel can do we'll pick a clock
rate that it cannot.
> I think my setup is correct, because the vesa frambuffer seems to work
> correctly. All is well unless I load the gma500_gfx driver and let it
> set the mode.
What mode is the vesa driver using and if you set the same mode what
happens ?
> [ 110.197942] ioremap error for 0x7f7f8000-0x7f7fb000, requested 0x10, got 0x0
That is expected
> [ 110.206418] [drm:drm_mode_debug_printmodeline], Modeline 0:"1280x1024" 0 108000 1280 1328 1440 1688 1024 1025 1028 1072 0x8 0x0
So we picked 1280x1024
> [ 110.257714] [drm:drm_sysfs_connector_add], adding "LVDS-1" to sysfs
> [ 110.257932] [drm:drm_sysfs_hotplug_event], generating hotplug event
> [ 110.269604] [drm:psb_intel_sdvo_read_byte], i2c transfer returned -6
> [ 110.269633] [drm:psb_intel_sdvo_init], No SDVO device found on SDVOB
We found the LVDS, we found nothing on the SDVO
> [ 110.561304] [drm:drm_helper_probe_single_connector_modes], [CONNECTOR:7:LVDS-1]
> [ 110.586242] [drm:drm_helper_probe_single_connector_modes], [CONNECTOR:7:LVDS-1] probed modes :
> [ 110.586274] [drm:drm_mode_debug_printmodeline], Modeline 10:"1280x1024" 40 71500 1280 1328 1440 1688 1024 1025 1028 1072 0x8 0x0
And we probed a 1280x1024 mode with a clock of 71.5 (which might be a bit
high for the panel ?)
> [ 110.637467] [drm:drm_crtc_helper_set_config], crtc has no fb, full mode set
> [ 110.637484] [drm:drm_crtc_helper_set_config], modes are different, full mode set
We didn't have a mode before, we have one now.
> [ 110.637500] [drm:drm_mode_debug_printmodeline], Modeline 0:"" 0 0 0 0 0 0 0 0 0 0 0x0 0x0
> [ 110.637529] [drm:drm_mode_debug_printmodeline], Modeline 11:"1280x1024" 40 71500 1280 1328 1440 1688 1024 1025 1028 1072 0x8 0x0
> [ 110.637564] [drm:drm_crtc_helper_set_config], encoder changed, full mode switch
> [ 110.637580] [drm:drm_crtc_helper_set_config], crtc changed, full mode switch
> [ 110.637599] [drm:drm_crtc_helper_set_config], [CONNECTOR:7:LVDS-1] to [CRTC:4]
Do a full load
> [ 110.637618] [drm:drm_crtc_helper_set_config], attempting to set mode from userspace
> [ 110.637634] [drm:drm_mode_debug_printmodeline], Modeline 11:"1280x1024" 40 71500 1280 1328 1440 1688 1024 1025 1028 1072 0x8 0x0
> [ 110.637678] [drm:drm_crtc_helper_set_mode], [CRTC:4]
> [ 110.658501] [drm:drm_mode_debug_printmodeline], Modeline 11:"1280x1024" 40 71500 1280 1328 1440 1688 1024 1025 1028 1072 0x8 0x0
> [ 110.699109] [drm:drm_crtc_helper_set_mode], [ENCODER:8:LVDS-8] set [MODE:11:1280x1024]
> [ 110.977395] [drm:drm_calc_timestamping_constants], crtc 4: hwmode: htotal 1688, vtotal 1072, vdisplay 1024
> [ 110.977415] [drm:drm_calc_timestamping_constants], crtc 4: clock 71500 kHz framedur 25307776 linedur 23608, pixeldur 13
We set it all up and seem happy.
> [ 110.977439] [drm:drm_crtc_helper_set_config], Setting connector DPMS state to on
> [ 110.977453] [drm:drm_crtc_helper_set_config], [CONNECTOR:7:LVDS-1] set DPMS on
> [ 111.032366] Console: switching to colour frame buffer device 160x64
> [ 111.064932] fb0: psbfb frame buffer device
> [ 111.064944] drm: registered panic notifier
> [ 111.066827] gma500 0000:00:02.0: Backlight lvds set brightness 7a120000
> [ 111.066860] [drm] Initialized gma500 1.0.0 2011-06-06 for 0000:00:02.0 on minor 0
Do you have a different monitor around to try and also hand setting lower
modes ?
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: gma500: invalid frequency for monitor
2012-02-04 11:41 ` Alan Cox
@ 2012-02-05 3:05 ` Michael Labriola
0 siblings, 0 replies; 3+ messages in thread
From: Michael Labriola @ 2012-02-05 3:05 UTC (permalink / raw)
To: Alan Cox
Cc: Alan Cox, David Airlie, Michael D Labriola, dri-devel,
linux-kernel
On 2/4/12, Alan Cox <alan@lxorguk.ukuu.org.uk> wrote:
*snip*
>> I think my setup is correct, because the vesa frambuffer seems to work
>> correctly. All is well unless I load the gma500_gfx driver and let it
>> set the mode.
>
> What mode is the vesa driver using and if you set the same mode what
> happens ?
The vesa driver was using a resolution of 1280x1024x32... is there a
way to get a whole modeline out of that?
*snip*
>> [ 110.586242] [drm:drm_helper_probe_single_connector_modes],
>> [CONNECTOR:7:LVDS-1] probed modes :
>> [ 110.586274] [drm:drm_mode_debug_printmodeline], Modeline 10:"1280x1024"
>> 40 71500 1280 1328 1440 1688 1024 1025 1028 1072 0x8 0x0
>
> And we probed a 1280x1024 mode with a clock of 71.5 (which might be a bit
> high for the panel ?)
I'm assuming the clock is at fault, yes.
*snip*
>
> Do you have a different monitor around to try and also hand setting lower
> modes ?
Compiling on my home system overnight (these atom boards are pretty
slick, but slow as dirt at compiling the kernel). I've got a
different monitor at home to try, so we'll see what happens. How do I
hand set the mode that gma500 uses?
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2012-02-05 3:06 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2012-02-04 3:42 gma500: invalid frequency for monitor Michael D Labriola
2012-02-04 11:41 ` Alan Cox
2012-02-05 3:05 ` Michael Labriola
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox