Comment # 33
on bug 70388
from Linas Vepstas
FYI,
New kernel, new CPU, same old graphics cards: if vram_pushbuf=1 not specified
on boot line, then kernel hangs about 4 seconds in (near first modesetting); I
can't capture the dmesg ...
This is with this-weeks kernel on mint-qiana-17 Annotated dmesg below; I can
provide the full dmesg if you really want it ...
[ 0.000000] Command line: BOOT_IMAGE=/boot/vmlinuz-3.18.0-rc2-linas
root=UUID=18a92961-a17b-4205-9038-a6eb91849706 ro nouveau.vram_pushbuf=1
nomdmonddf nomdmonisw
...
[ 0.000000] AGP: Checking aperture...
[ 0.000000] AGP: No AGP bridge found
[ 0.000000] AGP: Node 0: aperture [bus addr 0x00000000-0x01ffffff] (32MB)
[ 0.000000] AGP: Your BIOS doesn't leave a aperture memory hole
[ 0.000000] AGP: Please enable the IOMMU option in the BIOS setup
[ 0.000000] AGP: This costs you 64MB of RAM
[ 0.000000] AGP: Mapping aperture over RAM [mem 0x94000000-0x97ffffff]
(65536KB)
strange to see above, as I'm pretty sure I have the IOMMU turned on ...
I'm pretty sure the machine does not have any agp slots, instead, there are two
PCIe x16 slots ...
...
[ 0.000000] DMI: ASUSTeK Computer INC. KGP(M)E-D16/KGP(M)E-D16, BIOS 3201
02/19/2013
...
[ 0.648524] pci 0000:00:14.4: can't claim BAR 8 [mem 0xf5800000-0xf8ffffff]:
address conflict with amd_iommu [mem 0xf6000000-0xf6003fff]
Note that one of the two graphics cards is a PCI card NOT a PCIe card!!! I
mention this because of the above BAR road-bump
00:14.4 PCI bridge: Advanced Micro Devices, Inc. [AMD/ATI] SBx00 PCI to PCI
Bridge
[ 1.947594] AMD-Vi: Found IOMMU at 0000:00:00.2 cap 0x40
[ 1.947599] AMD-Vi: Interrupt remapping enabled
[ 1.947617] pci 0000:00:00.2: irq 25 for MSI/MSI-X
[ 1.958894] AMD-Vi: Lazy IO/TLB flushing enabled
[ 2.042066] PCI-DMA: Using software bounce buffering for IO (SWIOTLB)
[ 4.009084] fbcon: nouveaufb (fb1) is primary device
[ 4.009086] fbcon: Remapping primary device, fb1, to tty 1-63
[ 4.044782] Switched to clocksource tsc
[ 4.050013] nouveau [ DRM] 0xC633: Parsing digital output script table
[ 4.161782] nouveau E[ DRM] GPU lockup - switching to software fbcon
I have two DVI cables, one VGA cable. The edid to the VGA fails ... it didn't
use to ...
[ 16.869739] [drm:drm_edid_block_valid] *ERROR* EDID checksum is invalid,
remainder is 178
[ 16.869747] Raw EDID:
[ 16.869749] 00 ff ff ff ff ff ff 00 04 72 a6 01 12 12 12 12
[ 16.869751] 15 15 01 03 68 27 ff ff ff ff ff ff ff ff ff ff
etc.
it tries multiple times, gets different results ...
Xorg.0.log shows no startup errors, but the framebuffer shows mostly
corruption.
kill -9 of X sends X into 100% cpu usage. While it does that, this gets
logged:
Oct 31 15:36:35 fanny kernel: [ 2449.191822] nouveau E[Xorg[2290]] failed to
idle channel 0xcccc0002 [Xorg[2290]]
Oct 31 15:36:50 fanny kernel: [ 2464.194418] nouveau E[Xorg[2290]] failed to
idle channel 0xcccc0002 [Xorg[2290]]
Oct 31 15:37:05 fanny kernel: [ 2479.197017] nouveau E[Xorg[2290]] failed to
idle channel 0xcccc0001 [Xorg[2290]]
Oct 31 15:37:20 fanny kernel: [ 2494.199611] nouveau E[Xorg[2290]] failed to
idle channel 0xcccc0001 [Xorg[2290]]
Oct 31 15:37:35 fanny kernel: [ 2509.202213] nouveau E[Xorg[2290]] failed to
idle channel 0xcccc0000 [Xorg[2290]]
Oct 31 15:37:50 fanny kernel: [ 2524.204809] nouveau E[Xorg[2290]] failed to
idle channel 0xcccc0000 [Xorg[2290]]
After this, starting X a second time does not generate any video-out, so
monitors stay blanked. Nothing at all printed in /var/log/syslog. However
kern.log shows:
Oct 31 15:41:47 fanny kernel: [ 2761.045902] INFO: task kworker/10:1:135
blocked for more than 120 seconds.
Oct 31 15:41:47 fanny kernel: [ 2761.045912] Not tainted 3.18.0-rc2-linas
#5
Oct 31 15:41:47 fanny kernel: [ 2761.045915] "echo 0 >
/proc/sys/kernel/hung_task_timeout_secs" disables this message.
Oct 31 15:41:47 fanny kernel: [ 2761.045920] kworker/10:1 D ffff88085dd53040
0 135 2 0x00000000
Oct 31 15:41:47 fanny kernel: [ 2761.045959] Workqueue: events
output_poll_execute [drm_kms_helper]
Oct 31 15:41:47 fanny kernel: [ 2761.045964] ffff88082ea0fc28 0000000000000002
ffff88082e9e1910 ffff88082ea0ffd8
Oct 31 15:41:47 fanny kernel: [ 2761.045971] 0000000000013040 0000000000013040
ffff880831f88000 ffff88082e9e1910
Oct 31 15:41:47 fanny kernel: [ 2761.045978] ffff88082ea0fc38 ffff8808315a7898
ffff8808315a7946 ffff8808315a79c8
Oct 31 15:41:47 fanny kernel: [ 2761.045985] Call Trace:
Oct 31 15:41:47 fanny kernel: [ 2761.046000] [<ffffffff81592cf9>]
schedule+0x29/0x70
Oct 31 15:41:47 fanny kernel: [ 2761.046010] [<ffffffff813cd64d>]
rpm_resume+0x22d/0x620
Oct 31 15:41:47 fanny kernel: [ 2761.046020] [<ffffffff81080770>] ?
prepare_to_wait_event+0x110/0x110
Oct 31 15:41:47 fanny kernel: [ 2761.046028] [<ffffffff813cda8d>]
__pm_runtime_resume+0x4d/0x70
Oct 31 15:41:47 fanny kernel: [ 2761.046035] [<ffffffff8105a4f6>] ?
__queue_delayed_work+0x166/0x1a0
Oct 31 15:41:47 fanny kernel: [ 2761.046106] [<ffffffffa040c2b6>]
nouveau_connector_detect+0x66/0x3d0 [nouveau]
Oct 31 15:41:47 fanny kernel: [ 2761.046124] [<ffffffffa011c608>]
output_poll_execute+0xb8/0x180 [drm_kms_helper]
Oct 31 15:41:47 fanny kernel: [ 2761.046132] [<ffffffff8105ab06>]
process_one_work+0x146/0x400
Oct 31 15:41:47 fanny kernel: [ 2761.046139] [<ffffffff8105aee0>]
worker_thread+0x120/0x490
Oct 31 15:41:47 fanny kernel: [ 2761.046146] [<ffffffff8105adc0>] ?
process_one_work+0x400/0x400
Oct 31 15:41:47 fanny kernel: [ 2761.046153] [<ffffffff8105fd82>]
kthread+0xd2/0xf0
Oct 31 15:41:47 fanny kernel: [ 2761.046159] [<ffffffff8105fcb0>] ?
kthread_create_on_node+0x1a0/0x1a0
Oct 31 15:41:47 fanny kernel: [ 2761.046166] [<ffffffff8159722c>]
ret_from_fork+0x7c/0xb0
Oct 31 15:41:47 fanny kernel: [ 2761.046172] [<ffffffff8105fcb0>] ?
kthread_create_on_node+0x1a0/0x1a0
Oct 31 15:41:47 fanny kernel: [ 2761.046179] INFO: task kworker/3:1:175 blocked
for more than 120 seconds.
Oct 31 15:41:47 fanny kernel: [ 2761.046183] Not tainted 3.18.0-rc2-linas
#5
Oct 31 15:41:47 fanny kernel: [ 2761.046186] "echo 0 >
/proc/sys/kernel/hung_task_timeout_secs" disables this message.
Oct 31 15:41:47 fanny kernel: [ 2761.046189] kworker/3:1 D ffff88085dc73040
0 175 2 0x00000000
Oct 31 15:41:47 fanny kernel: [ 2761.046197] Workqueue: pm pm_runtime_work
Oct 31 15:41:47 fanny kernel: [ 2761.046201] ffff88082e593bb8 0000000000000002
ffff88082ebbcb30 ffff88082e593fd8
Oct 31 15:41:47 fanny kernel: [ 2761.046207] 0000000000013040 0000000000013040
ffff880831f4cb30 ffff88082ebbcb30
Oct 31 15:41:47 fanny kernel: [ 2761.046214] ffff88082e593bc8 ffff8808315a7898
ffff8808315a7946 ffff8808315a79c8
Oct 31 15:41:47 fanny kernel: [ 2761.046221] Call Trace:
Oct 31 15:41:47 fanny kernel: [ 2761.046229] [<ffffffff81592cf9>]
schedule+0x29/0x70
Oct 31 15:41:47 fanny kernel: [ 2761.046236] [<ffffffff813cd64d>]
rpm_resume+0x22d/0x620
Oct 31 15:41:47 fanny kernel: [ 2761.046243] [<ffffffff81080770>] ?
prepare_to_wait_event+0x110/0x110
Oct 31 15:41:47 fanny kernel: [ 2761.046251] [<ffffffff813cdd5c>]
pm_runtime_forbid+0x4c/0x60
Oct 31 15:41:47 fanny kernel: [ 2761.046310] [<ffffffffa03f6ba8>]
nouveau_pmops_runtime_suspend+0xc8/0xf0 [nouveau]
Oct 31 15:41:47 fanny kernel: [ 2761.046320] [<ffffffff8131a8ba>]
pci_pm_runtime_suspend+0x6a/0x120
Oct 31 15:41:47 fanny kernel: [ 2761.046326] [<ffffffff8131a850>] ?
pci_pm_runtime_resume+0x100/0x100
Oct 31 15:41:47 fanny kernel: [ 2761.046332] [<ffffffff813cc4bd>]
__rpm_callback+0x3d/0x90
Oct 31 15:41:47 fanny kernel: [ 2761.046338] [<ffffffff81059fd1>] ?
try_to_grab_pending+0xb1/0x150
Oct 31 15:41:47 fanny kernel: [ 2761.046345] [<ffffffff813cc53f>]
rpm_callback+0x2f/0x90
Oct 31 15:41:47 fanny kernel: [ 2761.046352] [<ffffffff813cca61>]
rpm_suspend+0x111/0x640
Oct 31 15:41:47 fanny kernel: [ 2761.046359] [<ffffffff8106b251>] ?
get_parent_ip+0x11/0x50
Oct 31 15:41:47 fanny kernel: [ 2761.046365] [<ffffffff8106b2e5>] ?
preempt_count_add+0x55/0xb0
Oct 31 15:41:47 fanny kernel: [ 2761.046370] [<ffffffff813ce3f2>]
pm_runtime_work+0x82/0xb0
Oct 31 15:41:47 fanny kernel: [ 2761.046377] [<ffffffff8105ab06>]
process_one_work+0x146/0x400
Oct 31 15:41:47 fanny kernel: [ 2761.046384] [<ffffffff8105aee0>]
worker_thread+0x120/0x490
Oct 31 15:41:47 fanny kernel: [ 2761.046391] [<ffffffff8105adc0>] ?
process_one_work+0x400/0x400
Oct 31 15:41:47 fanny kernel: [ 2761.046396] [<ffffffff8105fd82>]
kthread+0xd2/0xf0
Oct 31 15:41:47 fanny kernel: [ 2761.046403] [<ffffffff8105fcb0>] ?
kthread_create_on_node+0x1a0/0x1a0
Oct 31 15:41:47 fanny kernel: [ 2761.046408] [<ffffffff8159722c>]
ret_from_fork+0x7c/0xb0
Oct 31 15:41:47 fanny kernel: [ 2761.046414] [<ffffffff8105fcb0>] ?
kthread_create_on_node+0x1a0/0x1a0
FWIW: my kernel has CONFIG_PM_CORE enabled, but much/most of the rest of ACPI
is not enabled (its not a laptop, and never gets turned off or rebooted).
Don't know if any of my rambling helped, but I conclude:
-- on latest rc kernel, with three-head, dual old graphics cards, one of which
is PCI not PCIe, that nouveau has multiple possibly unrelated issues ...