Comment 7 for bug 1507255

This also impacts me. Screen becomes garbled quicker if Firefox is running. Screen takes a bit longer to become garbled if only Thunderbird is running.

I tried Cory's fix above, and it hasn't garbled the screen in about 45 minutes minutes with Firefox, Thunderbird, and other graphical apps running simultaneously. On my system the file /usr/share/X11/xorg.conf.d/20-intel.conf didn't exist, so I created it with the contents that Cory provided.

From uname -a:
Linux gandalf 4.2.0-16-generic #19-Ubuntu SMP Thu Oct 8 14:46:51 UTC 2015 i686 i686 i686 GNU/Linux

From lspci:
00:02.0 VGA compatible controller: Intel Corporation 82945G/GZ Integrated Graphics Controller (rev 02) (prog-if 00 [VGA controller])
        Subsystem: ASUSTeK Computer Inc. Device 817a
        Flags: bus master, fast devsel, latency 0, IRQ 16
        Memory at dfd00000 (32-bit, non-prefetchable) [size=512K]
        I/O ports at b800 [size=8]
        Memory at c0000000 (32-bit, prefetchable) [size=256M]
        Memory at dfd80000 (32-bit, non-prefetchable) [size=256K]
        Expansion ROM at <unassigned> [disabled]
        Capabilities: [90] MSI: Enable- Count=1/1 Maskable- 64bit-
        Capabilities: [d0] Power Management version 2
        Kernel driver in use: i915

I see these messages in the syslog that seem to correspond to each time that the screen becomes garbled:
Oct 31 18:25:16 gandalf kernel: [22669.004026] [drm] stuck on render ring
Oct 31 18:25:16 gandalf kernel: [22669.006035] [drm] GPU HANG: ecode 3:0:0x677fff41, in Xorg [960], reason: Ring hung, action: reset
Oct 31 18:25:16 gandalf kernel: [22669.006040] [drm] GPU hangs can indicate a bug anywhere in the entire gfx stack, including userspace.
Oct 31 18:25:16 gandalf kernel: [22669.006042] [drm] Please file a _new_ bug report on bugs.freedesktop.org against DRI -> DRM/Intel
Oct 31 18:25:16 gandalf kernel: [22669.006045] [drm] drm/i915 developers can then reassign to the right component if it's not a kernel issue.
Oct 31 18:25:16 gandalf kernel: [22669.006047] [drm] The gpu crash dump is required to analyze gpu hangs, so please always attach it.
Oct 31 18:25:16 gandalf kernel: [22669.006050] [drm] GPU crash dump saved to /sys/class/drm/card0/error
Oct 31 18:25:16 gandalf kernel: [22669.006151] ------------[ cut here ]------------
Oct 31 18:25:16 gandalf kernel: [22669.006225] WARNING: CPU: 0 PID: 357 at /build/linux-hXawIp/linux-4.2.0/drivers/gpu/drm/i915/intel_display.c:3294 intel_crtc_wait_for_pending_flips+0x17c/0x220 [i915]()
Oct 31 18:25:16 gandalf kernel: [22669.006234] WARN_ON(ret)
Oct 31 18:25:16 gandalf kernel: [22669.006241] Modules linked in: ftdi_sio usbserial snd_hda_codec_analog snd_hda_codec_generic gpio_ich snd_hda_intel snd_hda_codec snd_hda_core snd_hwdep joydev snd_pcm snd_seq_midi snd_seq_midi_event snd_rawmidi input_leds snd_seq serio_raw snd_seq_device snd_timer snd 8250_fintek soundcore asus_atk0110 lpc_ich shpchp mac_hid parport_pc ppdev lp parport autofs4 jitterentropy_rng drbg ansi_cprng xts gf128mul algif_skcipher af_alg dm_crypt hid_generic usbhid hid i915 video firewire_ohci i2c_algo_bit drm_kms_helper firewire_core r8169 crc_itu_t pata_acpi drm mii
Oct 31 18:25:16 gandalf kernel: [22669.006379] CPU: 0 PID: 357 Comm: kworker/u8:3 Not tainted 4.2.0-16-generic #19-Ubuntu
Oct 31 18:25:16 gandalf kernel: [22669.006385] Hardware name: System manufacturer System Product Name/P5L8L-SE, BIOS 0201 10/09/2007
Oct 31 18:25:16 gandalf kernel: [22669.006439] Workqueue: i915-hangcheck i915_hangcheck_elapsed [i915]
Oct 31 18:25:16 gandalf kernel: [22669.006446] c1a629a7 fc2bbb5e 00000000 f5701d48 c175dae1 f5701d88 f5701d78 c106a687
Oct 31 18:25:16 gandalf kernel: [22669.006462] f8d6a8e9 f5701da8 00000165 f8d5ca44 00000cde f8cff22c f8cff22c f5460000
Oct 31 18:25:16 gandalf kernel: [22669.006478] f5a1e434 00000001 f5701d94 c106a6fe 00000009 f5701d88 f8d6a8e9 f5701da8
Oct 31 18:25:16 gandalf kernel: [22669.006493] Call Trace:
Oct 31 18:25:16 gandalf kernel: [22669.006510] [<c175dae1>] dump_stack+0x41/0x52
Oct 31 18:25:16 gandalf kernel: [22669.006523] [<c106a687>] warn_slowpath_common+0x87/0xc0
Oct 31 18:25:16 gandalf kernel: [22669.006581] [<f8cff22c>] ? intel_crtc_wait_for_pending_flips+0x17c/0x220 [i915]
Oct 31 18:25:16 gandalf kernel: [22669.006636] [<f8cff22c>] ? intel_crtc_wait_for_pending_flips+0x17c/0x220 [i915]
Oct 31 18:25:16 gandalf kernel: [22669.006649] [<c106a6fe>] warn_slowpath_fmt+0x3e/0x60
Oct 31 18:25:16 gandalf kernel: [22669.006707] [<f8cff22c>] intel_crtc_wait_for_pending_flips+0x17c/0x220 [i915]
Oct 31 18:25:16 gandalf kernel: [22669.006753] [<f8b1ffd4>] ? drm_modeset_lock_all_crtcs+0x84/0x90 [drm]
Oct 31 18:25:16 gandalf kernel: [22669.006809] [<f8d005f4>] intel_crtc_disable_planes+0x34/0xf0 [i915]
Oct 31 18:25:16 gandalf kernel: [22669.006863] [<f8d0075a>] intel_prepare_reset+0x6a/0x80 [i915]
Oct 31 18:25:16 gandalf kernel: [22669.006912] [<f8cd27b7>] i915_handle_error+0x147/0x6e0 [i915]
Oct 31 18:25:16 gandalf kernel: [22669.006924] [<c10be1bd>] ? log_store+0x1cd/0x210
Oct 31 18:25:16 gandalf kernel: [22669.006937] [<c10bfa91>] ? vprintk_default+0x41/0x60
Oct 31 18:25:16 gandalf kernel: [22669.006986] [<f8cd2fd3>] i915_hangcheck_elapsed+0x233/0x410 [i915]
Oct 31 18:25:16 gandalf kernel: [22669.006999] [<c10810b5>] process_one_work+0x155/0x3e0
Oct 31 18:25:16 gandalf kernel: [22669.007010] [<c1081377>] worker_thread+0x37/0x490
Oct 31 18:25:16 gandalf kernel: [22669.007019] [<c1081340>] ? process_one_work+0x3e0/0x3e0
Oct 31 18:25:16 gandalf kernel: [22669.007030] [<c1086a56>] kthread+0xa6/0xc0
Oct 31 18:25:16 gandalf kernel: [22669.007019] [<c1081340>] ? process_one_work+0x3e0/0x3e0
Oct 31 18:25:16 gandalf kernel: [22669.007030] [<c1086a56>] kthread+0xa6/0xc0
Oct 31 18:25:16 gandalf kernel: [22669.007042] [<c1763741>] ret_from_kernel_thread+0x21/0x30
Oct 31 18:25:16 gandalf kernel: [22669.007050] [<c10869b0>] ? kthread_create_on_node+0x170/0x170
Oct 31 18:25:16 gandalf kernel: [22669.007056] ---[ end trace b71e971205ef4bc7 ]---
Oct 31 18:25:16 gandalf kernel: [22669.042261] drm/i915: Resetting chip after gpu hang