[meta-freescale] Wandboard Quad experience (good!)

Tue Jul 16 08:58:25 PDT 2013

I do get the 
[ 1685.143891] Physical memory allocation error!

errors as well! (same back-trace so I will not repeat)

I'm using a nitrogen6x right now.
I've already set CONFIG_SWAP=n, which didn't help for me.

This morning I used to somehow(! read a bit below why it's not a solution for 
me) work around this issue with this patch:
http://pastebin.com/hRgpitvP

inspired by a patch from:
https://community.freescale.com/message/316472#316472

However this morning I switched to master for all my layers (meta-fsl-arm-
extra is master-next), now if I apply this patch I get kernel-panic at start-
up:
< I don't have the log/back-trace right now, if requested I can provide it >

Anyway .. I'm not sure it's the right solution!... because although I got rid 
of the kernel errors I had a different problem then:

Somehow the kernel and/or the application started to misbehave at some point.
The framebuffer went black (and even /dev/urandom couldn't do anything about 
it), and if I tried to kill the application (SIGTERM or SIGKILL) it became a 
zombie and consumed ~25-30% CPU continuously. Only a reboot helped.

Sound like 'it ended up in a kernel endless loop' .. doesn't it? 

Greets
Thomas

On Monday, 15 July, 2013 21:53:03 John Weber wrote:
> Hi Chris,
> 
> You might want to consider emailing the wandboard-dev mailing list for
> Wandboard kernel questions.
> 
> wandboard-dev at lists.wandboard.org
> 
> You'll need to sign up of course:
> http://wandboard.org/cgi-bin/mailman/listinfo/wandboard-dev
> 
> Or the main user group at
> wandboard at googlegroups.com
> 
> I haven't seen that message in using the video encoder functionality of the
> VPU from Gstreamer.  Just greping the source, that message comes from the
> vpu driver (as expected).  During vpu_open(), it seems to be enabling the
> clock to the VPU, then checking to see if the clock is enabled by checking
> a program counter of what is I'm guessing the integrated bitstream
> processor in the VPU.  If that is greater than 0x0, then it pops that debug
> message.  Then it disables the vpu clock, then moves on.
> 
> John
> 
> On 7/15/13 3:27 PM, Chris Tapp wrote:
> > Hi John,
> > 
> > This (2nd crash) still happens when CONFIG_SWAP is not set, but it's just
> > possible that the Vivante one has gone! I've had a board running my app.
> > for over 6 hours hours and the problem hasn't shown yet. I'll keep it
> > running...
> > 
> > However, I now see lots of:
> > 
> > [ 1481.696860] Not power off before vpu open!
> > 
> > Any idea what these are about?
> > 
> > I've also noticed that the Wandboard is very sensitive to power quality.
> > The specs say a 5v, 2A PSU is suitable - but it's not if it's got a very
> > fast current limit :-) I get reboots (uboot reports 'POR' reset cause) if
> > I use a lab PSU set to 2A.> 
> > On 14 Jul 2013, at 20:38, John Weber wrote:
> >> Hi Chris,
> >> 
> >> Thanks.  You've probably noticed, but this is a different error from the
> >> first one you sent.  The first one was in the Vivante driver.
> >> 
> >> This one seems related to a memory limitation and perhaps related to the
> >> CONFIG_SWAP being on by default in the kernel build.  Since there is no
> >> swap partition, having CONFIG_SWAP seems a little useless.  Oddly
> >> enough, all of the default i.MX6 defconfigs set CONFIG_SWAP.  I'm seeing
> >> if we can remove that option.
> >> 
> >> John
> >> 
> >> On 7/13/13 4:26 PM, Chris Tapp wrote:
> >>> Hi John,
> >>> 
> >>> On 13 Jul 2013, at 01:07, John Weber wrote:
> >>>> Chris,
> >>>> 
> >>>> This looks like it is coming from the Vivante GPU driver in:
> >>>> drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel.c, line 1315.
> >>>> 
> >>>> How would I replicate this problem?
> >>> 
> >>> Good question! Simply running (with GST_DEBUG="*:2")
> >>> 
> >>> gst-launch playbin2 uri=http://media.w3.org/2010/05/sintel/trailer.webm
> >>> video-sink="queue2 ! mfw_v4lsink"
> >>> 
> >>> sometimes gives this:
> >>> 
> >>> [ 2445.396718] source:src: page allocation failure: order:11, mode:0xd1
> >>> [ 2445.403170] Backtrace:
> >>> [ 2445.405710] [<c003e358>] (dump_backtrace+0x0/0x104) from [<c0421760>]
> >>> (dump_stack+0x18/0x1c) [ 2445.414222]  r6:e9b80000 r5:000000d1
> >>> r4:00000001 r3:00000000
> >>> [ 2445.420029] [<c0421748>] (dump_stack+0x0/0x1c) from [<c00b7be8>]
> >>> (warn_alloc_failed+0xe4/0x104) [ 2445.428825] [<c00b7b04>]
> >>> (warn_alloc_failed+0x0/0x104) from [<c00ba12c>]
> >>> (__alloc_pages_nodemask+0x5b8/0x634) [ 2445.438806]  r3:00000000
> >>> r2:00000000
> >>> [ 2445.442482]  r8:00000000 r7:00000000 r6:e9b80000 r5:0000000b
> >>> r4:000000d1
> >>> [ 2445.449342] [<c00b9b74>] (__alloc_pages_nodemask+0x0/0x634) from
> >>> [<c004400c>] (__dma_alloc+0xc4/0x2b0) [ 2445.458728] [<c0043f48>]
> >>> (__dma_alloc+0x0/0x2b0) from [<c0044550>]
> >>> (dma_alloc_coherent+0x5c/0x68) [ 2445.467695] [<c00444f4>]
> >>> (dma_alloc_coherent+0x0/0x68) from [<c02f693c>]
> >>> (vpu_alloc_dma_buffer+0x34/0x5c) [ 2445.477323]  r7:e9b80000
> >>> r6:e9cb6f28 r5:e9cb6f20 r4:e9cb6f28
> >>> [ 2445.483422] [<c02f6908>] (vpu_alloc_dma_buffer+0x0/0x5c) from
> >>> [<c02f6a38>] (vpu_ioctl+0xd4/0x7ac) [ 2445.492366]  r4:41efe940
> >>> r3:00000000
> >>> [ 2445.496042] [<c02f6964>] (vpu_ioctl+0x0/0x7ac) from [<c00f144c>]
> >>> (vfs_ioctl+0x28/0x44) [ 2445.504025]  r8:e9b468d0 r7:00000007
> >>> r6:00005600 r5:e9b667a0 r4:e9b667a0 [ 2445.510888] [<c00f1424>]
> >>> (vfs_ioctl+0x0/0x44) from [<c00f1e78>] (do_vfs_ioctl+0x414/0x508) [
> >>> 2445.519167] [<c00f1a64>] (do_vfs_ioctl+0x0/0x508) from [<c00f1fa8>]
> >>> (sys_ioctl+0x3c/0x68) [ 2445.527381] [<c00f1f6c>] (sys_ioctl+0x0/0x68)
> >>> from [<c003ac00>] (ret_fast_syscall+0x0/0x30) [ 2445.535753] 
> >>> r7:00000036 r6:415c11ac r5:420762a8 r4:41efe940
> >>> [ 2445.541491] Mem-info:
> >>> [ 2445.543770] DMA per-cpu:
> >>> [ 2445.546309] CPU    0: hi:   90, btch:  15 usd:  84
> >>> [ 2445.551119] CPU    1: hi:   90, btch:  15 usd:  84
> >>> [ 2445.555918] CPU    2: hi:   90, btch:  15 usd:   0
> >>> [ 2445.560733] CPU    3: hi:   90, btch:  15 usd:  98
> >>> [ 2445.565529] Normal per-cpu:
> >>> [ 2445.568330] CPU    0: hi:   90, btch:  15 usd:  80
> >>> [ 2445.573146] CPU    1: hi:   90, btch:  15 usd:  66
> >>> [ 2445.577944] CPU    2: hi:   90, btch:  15 usd:  77
> >>> [ 2445.582757] CPU    3: hi:   90, btch:  15 usd:  19
> >>> [ 2445.587553] HighMem per-cpu:
> >>> [ 2445.590455] CPU    0: hi:  186, btch:  31 usd:  16
> >>> [ 2445.595254] CPU    1: hi:  186, btch:  31 usd:  37
> >>> [ 2445.600052] CPU    2: hi:  186, btch:  31 usd: 172
> >>> [ 2445.604863] CPU    3: hi:  186, btch:  31 usd: 181
> >>> [ 2445.609675] active_anon:1321 inactive_anon:19 isolated_anon:0
> >>> [ 2445.609680]  active_file:2311 inactive_file:2276 isolated_file:0
> >>> [ 2445.609686]  unevictable:0 dirty:1 writeback:0 unstable:0
> >>> [ 2445.609690]  free:456985 slab_reclaimable:305 slab_unreclaimable:1705
> >>> [ 2445.609696]  mapped:1837 shmem:43 pagetables:153 bounce:0
> >>> [ 2445.638709] DMA free:48392kB min:1052kB low:1312kB high:1576kB
> >>> active_anon:0kB inactive_anon:0kB active_file:0kB inactive_file:0kB
> >>> unevictable:0kB isolated(anon):0kB isolated(file):0kB present:186944kB
> >>> mlocked:0kB dirty:0kB writeback:0kB mapped:0kB shmem:0kB
> >>> slab_reclaimable:0kB slab_unreclaimable:0kB kernel_stack:0kB
> >>> pagetables:0kB unstable:0kB bounce:0kB writeback_tmp:0kB
> >>> pages_scanned:0 all_unreclaimable? yes [ 2445.675225] lowmem_reserve[]:
> >>> 0 308 1673 1673
> >>> [ 2445.679676] Normal free:392244kB min:1776kB low:2220kB high:2664kB
> >>> active_anon:0kB inactive_anon:0kB active_file:2036kB
> >>> inactive_file:1676kB unevictable:0kB isolated(anon):0kB
> >>> isolated(file):0kB present:315584kB mlocked:0kB dirty:4kB writeback:0kB
> >>> mapped:0kB shmem:0kB slab_reclaimable:1220kB slab_unreclaimable:6820kB
> >>> kernel_stack:656kB pagetables:612kB unstable:0kB bounce:0kB
> >>> writeback_tmp:0kB pages_scanned:0 all_unreclaimable? no [ 2445.717837]
> >>> lowmem_reserve[]: 0 0 10922 10922
> >>> [ 2445.722299] HighMem free:1387304kB min:512kB low:2480kB high:4448kB
> >>> active_anon:5284kB inactive_anon:76kB active_file:7208kB
> >>> inactive_file:7428kB unevictable:0kB isolated(anon):0kB
> >>> isolated(file):0kB present:1398016kB mlocked:0kB dirty:0kB
> >>> writeback:0kB mapped:7348kB shmem:172kB slab_reclaimable:0kB
> >>> slab_unreclaimable:0kB kernel_stack:0kB pagetables:0kB unstable:0kB
> >>> bounce:0kB writeback_tmp:0kB pages_scanned:0 all_unreclaimable? no [
> >>> 2445.760550] lowmem_reserve[]: 0 0 0 0
> >>> [ 2445.764287] DMA: 10*4kB 12*8kB 14*16kB 13*32kB 12*64kB 12*128kB
> >>> 7*256kB 3*512kB 5*1024kB 6*2048kB 6*4096kB 0*8192kB 0*16384kB 0*32768kB
> >>> = 48392kB [ 2445.777566] Normal: 99*4kB 71*8kB 41*16kB 9*32kB 5*64kB
> >>> 3*128kB 2*256kB 2*512kB 1*1024kB 1*2048kB 2*4096kB 4*8192kB 3*16384kB
> >>> 9*32768kB = 392244kB [ 2445.790945] HighMem: 400*4kB 205*8kB 114*16kB
> >>> 131*32kB 62*64kB 29*128kB 13*256kB 2*512kB 4*1024kB 1*2048kB 2*4096kB
> >>> 1*8192kB 2*16384kB 40*32768kB = 1387304kB [ 2445.805269] 4614 total
> >>> pagecache pages
> >>> [ 2445.809023] 0 pages in swap cache
> >>> [ 2445.812355] Swap cache stats: add 0, delete 0, find 0/0
> >>> [ 2445.817586] Free swap  = 0kB
> >>> [ 2445.820483] Total swap = 0kB
> >>> [ 2445.870417] 524288 pages of RAM
> >>> [ 2445.873566] 458040 free pages
> >>> [ 2445.876537] 50744 reserved pages
> >>> [ 2445.879767] 2012 slab pages
> >>> [ 2445.882584] 5157 pages shared
> >>> [ 2445.885557] 0 pages swap cached
> >>> [ 2445.888705] Physical memory allocation error!
> >>> [ 2445.893083] Physical memory allocation error!
> >>> 
> >>> I'll have a go at getting the other one to show...
> >>> 
> >>> Chris Tapp
> >>> 
> >>> opensource at keylevel.com
> >>> www.keylevel.com
> > 
> > Chris Tapp
> > 
> > opensource at keylevel.com
> > www.keylevel.com
> 
> _______________________________________________
> meta-freescale mailing list
> meta-freescale at yoctoproject.org
> https://lists.yoctoproject.org/listinfo/meta-freescale