Resolution problem solved

Yesterday I posted about trying to get my old CRT to do 1600×1200 again using Fedora 11. Yesterday I had an incorrect xorg.conf, here’s the correct one that is working for me. The key to my problem was adding the Modeline to the Monitor section NOT to the Display section.

After putting this xorg.conf into action I got 1600×1200 from the gdm
login screen onwards. I don’t think there is any hope for the bootup screens they still show up with some awful resolution.

# Xorg configuration created by system-config-display

Section "ServerLayout"
	Identifier     "single head configuration"
	Screen      0  "Screen0" 0 0
	InputDevice    "Keyboard0" "CoreKeyboard"
EndSection

Section "InputDevice"
	Identifier  "Keyboard0"
	Driver      "kbd"
	Option	    "XkbModel" "pc105"
	Option	    "XkbLayout" "us+inet"
EndSection

Section "Monitor"
	Identifier   "Monitor0"
	ModelName    "NEC MultiSync E900+"
	HorizSync    31.0 - 96.0
	VertRefresh  55.0 - 160.0
	Option	    "dpms"
	Modeline "1600x1200"  190.50  1600 1720 1888 2176  1200 1203 1207 1252 -hsync +vsync
EndSection

Section "Device"
	Identifier  "Videocard0"
	Driver      "radeon"
EndSection

Section "Screen"
	Identifier "Screen0"
	Device     "Videocard0"
	Monitor    "Monitor0"
	DefaultDepth     24
    
	SubSection "Display"
		Viewport   0 0
		Depth     24
		Modes    "1600x1200"
        #"1600x1024" "1440x900" "1400x1050" "1360x768" "1280x1024" "1280x960" "1152x864" "1024x768" "832x624" "800x600" "720x400" "640x480" "640x400" "640x350"
	EndSubSection
EndSection

Where is my 1600×1200 Leonidas?

After upgrading both my Thinkpad T43 laptop and Dell workstation to Fedora 11, I decided to upgrade my home machine from Fedora 9 to 11. Actually it wasn’t a traditional upgrade, it was a reinstall. The hardest part about the installation has been trying to figure out all of the things I needed to backup.

I rsync’d /etc/, $HOME, and some scripts I had in /opt over to my server and proceeded to wipe the box clean. This time I chose to partition the drive so that /home is in a separate partition, hopefully this will make it easier to upgrade/install in the future.  The installer seemed more streamlined than previous versions, less questions, either that or I’m just used to clicking through it without paying much attention 🙂 Once I selected the packages to install I left it copying data overnight.

The following morning I rebooted and logged in via the command line to start the rsync of my data back to the machine. After several hours of data copying, and package installation I was up and running. All is well except for my screen resolution, the new ‘enhancements’ won’t detect that my monitor can do 1600×1200, so it limits me to a paltry 1280×1024. Using xrandr I was able to add a new modeline, then changed the resolution to 1600×1200.

$ gtf 1600 1200 70
# 1600x1200 @ 70.00 Hz (GTF) hsync: 87.43 kHz; pclk: 190.25 MHz
Modeline "1600x1200_70.00" 190.25 1600 1712 1888 2176 1200 1201 1204 1249 -HSync +Vsync
xrandr --newmode "1600x1200_70.00" 190.25 1600 1712 1888 2176 1200 1201 1204 1249 -HSync +Vsync
xrandr --addmode VGA-0 "1600x1200_70.00"

This is great except I have to do this every time I login. Before I figured out xrandr, I tried using system-config-display which created an xorg.conf (which in Fedora 11 is gone now).  After rebooting with the ‘generated’ config, the machine was stuck showing me a blue screen with the fedora logo in it. I had to boot up in rescue mode to be able to see what was going on. Xorg.0.log showed the following:

(WW) xf86CloseConsole: KDSETMODE failed: Bad file descriptor
(WW) xf86CloseConsole: VT_GETMODE failed: Bad file descriptor

Hrm something is wrong with my ‘generated’ xorg.conf, maybe that’s the reason system-config-display isn’t installed by default 🙂

# Xorg configuration created by system-config-display

Section "ServerLayout"
    Identifier     "single head configuration"
    Screen      0  "Screen0" 0 0
    InputDevice    "Keyboard0" "CoreKeyboard"
EndSection

Section "InputDevice"
# keyboard added by rhpxl
    Identifier  "Keyboard0"
    Driver      "kbd"
    Option      "XkbModel" "pc105+inet"
    Option      "XkbLayout" "us"
EndSection

Section "Monitor"
    Identifier   "Monitor0"
    ModelName    "NEC MultiSync E900+"
    HorizSync    31.0 - 96.0
    VertRefresh  55.0 - 160.0
    Option      "dpms"
EndSection

Section "Device"
    Identifier  "Videocard0"
    Driver      "radeon"
EndSection

Section "Screen"
    Identifier "Screen0"
    Device     "Videocard0"
    Monitor    "Monitor0"
    DefaultDepth     24
    Modeline "1600x1200"  190.50  1600 1720 1888 2176  1200 1203 1207 1252 -hsync +vsync
    Option "PreferredMode" "1600x1200"
    SubSection "Display"
        Viewport   0 0
        Depth     24
    EndSubSection
EndSection

I wasn’t in the mood to figure out the problem right then, so I did the obvious: mv xorg.conf dontusethisxorg_conf and rebooted. This time it booted up to the gdm login screen, sadly at the crappy 1152×768 resolution. Right now I’m just staying logged in 🙂 until I can figure out what the heck is wrong and how to specify the information properly in xorg.conf.

Hardware

  • ATI Technologies Inc RV370 5B60 [Radeon X300 (PCIE)]
  • NEC Multisync E900+ 19″ CRT (yes it’s still a CRT, I really want a Dell Ultrasharp 2408WFP LCD!)
  • monitor is connected to the ATI via VGA connection

I’ll keep digging to see how to get this working after reboots, if anyone has any suggestions feel free to leave me a comment.

banshee and older iPods

With the release of banshee 1.0, podsleuth was re-written to work with later models of iPods but ended up breaking compatibility with 3rd gen iPods (like mine). I tried creating a patch but C# is not my forte and time was lacking as well.

But today I found that a patch was committed that fixes bug #486661: commit 8c17bbb6e3fa9bd39c4c43c0e28a30a6795a189c in podsleuth.

I downloaded the F9 src.rpm from koji and created a patch from the above mentioned commit. I rebuilt the rpm and installed it on my Fedora 9 x86_64 machine. I’m happy to say I’m back in business!

[jmrodri@firebird podsleuth]$ rpm -qa 'podsleuth*'
podsleuth-0.6.4-2.zeus.fc9.x86_64
podsleuth-devel-0.6.4-2.zeus.fc9.x86_64

Thanks to Nicolas Cortot for submitting the patch, and to Gabriel Burt for applying the patch.

banshee-ipod

Leonidas broke suspend

Fedora 11 has broken my suspend 😦 It worked find in Fedora 8/9/10. Sigh, time to figure out what the heck is going on.

WARNING: at kernel/hrtimer.c:625 hres_timers_resume+0x34/0x4a() (Not tainted)
Hardware name: 2687D7U
hres_timers_resume() called with IRQs enabled!Modules linked in: arc4 ecb lib80211_crypt_wep fuse sco bridge stp llc bnep l2cap bluetooth sunrpc ip6t_REJECT nf_conntrack_ipv6 ip6table_filter ip6_tables ipv6 cpufreq_ondemand acpi_cpufreq dm_multipath uinput ppdev snd_intel8x0m snd_intel8x0 snd_ac97_codec ac97_bus snd_pcm thinkpad_acpi ipw2200 snd_timer nsc_ircc parport_pc snd iTCO_wdt libipw irda iTCO_vendor_support hwmon yenta_socket parport i2c_i801 video soundcore lib80211 pcspkr tg3 snd_page_alloc output rsrc_nonstatic joydev crc_ccitt ata_generic pata_acpi xts gf128mul aes_i586 aes_generic dm_crypt radeon drm i2c_algo_bit i2c_core [last unloaded: scsi_wait_scan]
Pid: 2724, comm: pm-suspend Not tainted 2.6.29.5-191.fc11.i686.PAE #1
Call Trace:
[] warn_slowpath+0x7c/0xa4
[] ? ktime_get_ts+0x4f/0x53
[] ? lapic_next_event+0x18/0x1c
[] ? clockevents_program_event+0xe1/0xf0
[] ? tick_dev_program_event+0x47/0xb5
[] ? tick_program_event+0x26/0x2e
[] ? tick_notify+0x2e8/0x2f7
[] ? hpet_resume_counter+0x0/0x51
[] ? notifier_call_chain+0x26/0x48
[] hres_timers_resume+0x34/0x4a
[] timekeeping_resume+0x130/0x138
[] __sysdev_resume+0x19/0x3d
[] sysdev_resume+0x26/0x59
[] suspend_devices_and_enter+0x112/0x186
[] enter_state+0x142/0x19d
[] state_store+0x98/0xac
[] ? state_store+0x0/0xac
[] kobj_attr_store+0x16/0x22
[] sysfs_write_file+0xc9/0xf4
[] ? sysfs_write_file+0x0/0xf4
[] vfs_write+0x95/0xf4
[] sys_write+0x4c/0x70
[] sysenter_do_call+0x12/0x34
---[ end trace 98ea38b8c74f376f ]---