BUG: Bad page state in process swapper pfn:94d23

Bug #605739 reported by Bryan Wu
16
This bug affects 3 people
Affects Status Importance Assigned to Milestone
linux-ti-omap4 (Ubuntu)
Fix Released
High
Bryan Wu
Maverick
Fix Released
High
Bryan Wu

Bug Description

This oops always show up during kernel booting

---
BUG: Bad page state in process swapper pfn:94d23
page:c0a11460 count:0 mapcount:-897 mapping:(null) index:0x0
page flags: 0x0()
[<c00445b4>] (unwind_backtrace+0x0/0xf0) from [<c00ccde8>] (bad_page+0xc4/0xf0)
[<c00ccde8>] (bad_page+0xc4/0xf0) from [<c00cdf80>] (__free_pages_ok+0xf0/0x1ec)
[<c00cdf80>] (__free_pages_ok+0xf0/0x1ec) from [<c0016420>] (free_all_bootmem_core+0xf8/0x214)
[<c0016420>] (free_all_bootmem_core+0xf8/0x214) from [<c000cbcc>] (mem_init+0xec/0x3cc)
[<c000cbcc>] (mem_init+0xec/0x3cc) from [<c0008a0c>] (start_kernel+0x148/0x300)
[<c0008a0c>] (start_kernel+0x148/0x300) from [<80008080>] (0x80008080)
---

-Bryan

Tim Gardner (timg-tpi)
Changed in linux-ti-omap4 (Ubuntu Maverick):
assignee: nobody → Tim Gardner (timg-tpi)
importance: Undecided → High
milestone: none → maverick-alpha-3
status: New → In Progress
Revision history for this message
Bryan Wu (cooloney) wrote :
Download full text (5.6 KiB)

I added bootmem_debug=1 to kernel cmdline, got following messages:

---
bootmem::init_bootmem_core nid=0 start=80000 map=80773 end=9cf00 mapsize=39e0
bootmem::mark_bootmem_node nid=0 start=80000 end=9cf00 reserve=0 flags=0
bootmem::__free nid=0 start=80000 end=9cf00
bootmem::mark_bootmem_node nid=0 start=80773 end=80777 reserve=1 flags=0
bootmem::__reserve nid=0 start=80773 end=80777 flags=0
bootmem::mark_bootmem_node nid=0 start=80008 end=80773 reserve=1 flags=0
bootmem::__reserve nid=0 start=80008 end=80773 flags=0
bootmem::mark_bootmem_node nid=0 start=80004 end=80008 reserve=1 flags=0
bootmem::__reserve nid=0 start=80004 end=80008 flags=0
bootmem::alloc_bootmem_core nid=0 size=3a0000 [928 pages] align=20 goal=bfffffff limit=0
bootmem::__reserve nid=0 start=80777 end=80b17 flags=1
bootmem::alloc_bootmem_core nid=0 size=2c [1 pages] align=20 goal=bfffffff limit=0
bootmem::__reserve nid=0 start=80b17 end=80b18 flags=1
bootmem::alloc_bootmem_core nid=0 size=2000 [2 pages] align=20 goal=bfffffff limit=0
bootmem::__reserve nid=0 start=80b18 end=80b1a flags=1
bootmem::alloc_bootmem_core nid=0 size=1000 [1 pages] align=1000 goal=0 limit=ffffffff
bootmem::__reserve nid=0 start=80b1a end=80b1b flags=1
bootmem::alloc_bootmem_core nid=0 size=1000 [1 pages] align=1000 goal=0 limit=ffffffff
bootmem::__reserve nid=0 start=80b1b end=80b1c flags=1
OMAP4430 ES1.0
SRAM: Mapped pa 0x40300000 to va 0xfe400000 size: 0x100000
bootmem::alloc_bootmem_core nid=0 size=2000000 [8192 pages] align=1000 goal=0 limit=0
bootmem::__reserve nid=0 start=80b1c end=82b1c flags=1
Reserving 33554432 bytes SDRAM for VRAM
bootmem::alloc_bootmem_core nid=0 size=1000 [1 pages] align=1000 goal=0 limit=ffffffff
bootmem::__reserve nid=0 start=82b1c end=82b1d flags=1
bootmem::alloc_bootmem_core nid=0 size=1c [1 pages] align=20 goal=0 limit=ffffffff
bootmem::__reserve nid=0 start=82b1d end=82b1e flags=1
bootmem::alloc_bootmem_core nid=0 size=72 [1 pages] align=20 goal=0 limit=0
bootmem::__reserve nid=0 start=82b1e end=82b1e flags=1
bootmem::alloc_bootmem_core nid=0 size=72 [1 pages] align=20 goal=0 limit=0
bootmem::__reserve nid=0 start=82b1e end=82b1e flags=1
bootmem::alloc_bootmem_core nid=0 size=1000 [1 pages] align=20 goal=0 limit=0
bootmem::__reserve nid=0 start=82b1e end=82b1f flags=1
bootmem::alloc_bootmem_core nid=0 size=1000 [1 pages] align=20 goal=0 limit=0
bootmem::__reserve nid=0 start=82b1f end=82b20 flags=1
bootmem::alloc_bootmem_core nid=0 size=20000 [32 pages] align=1000 goal=0 limit=0
bootmem::__reserve nid=0 start=82b20 end=82b40 flags=1
bootmem::mark_bootmem_node nid=0 start=82b28 end=82b30 reserve=0 flags=0
bootmem::__free nid=0 start=82b28 end=82b30
bootmem::mark_bootmem_node nid=0 start=82b38 end=82b40 reserve=0 flags=0
bootmem::__free nid=0 start=82b38 end=82b40
PERCPU: Embedded 8 pages/cpu @c2b20000 s9088 r8192 d15488 u65536
bootmem::alloc_bootmem_core nid=0 size=4 [1 pages] align=20 goal=0 limit=0
bootmem::__reserve nid=0 start=82b28 end=82b29 flags=1
bootmem::alloc_bootmem_core nid=0 size=4 [1 pages] align=20 goal=0 limit=0
bootmem::__reserve nid=0 start=82b29 end=82b29 flags=1
bootmem::alloc_bootmem_core nid=0 size=8 [1 pages] align=20 goal=0 limit=0
bootmem::...

Read more...

Revision history for this message
Bryan Wu (cooloney) wrote :

I suspect it is an issue about swap. So I turned off CONFIG_SWAP=y, tried the new kernel. This oops never happens

But swap is required by our Ubuntu, need more investigation then.

Revision history for this message
Oliver Grawert (ogra) wrote :

we only have 512M, generally ubuntu doesnt need swap if there is enough ram.
but with 512M on the omap4 boards you will hit OOM very quickly if you use no swap and some app like firefox or openoffice. support for swap partitions and files should be there and function the right way in our kernel.

tags: added: armel
Revision history for this message
Bryan Wu (cooloney) wrote :

After upgraded into TI latest release, this bug was gone.

Kernel git tree is here:
http://kernel.ubuntu.com/git?p=roc/ubuntu-maverick.git;a=shortlog;h=refs/heads/ti-omap4

Please help me to test the kernel here:
http://people.canonical.com/~roc/kernel/omap4/

Thanks,
-Bryan

Lee Jones (lag)
Changed in linux-ti-omap4 (Ubuntu Maverick):
assignee: Tim Gardner (timg-tpi) → Bryan Wu (cooloney)
Revision history for this message
Martin Pitt (pitti) wrote :

Workaround is in place by dropping swap, so not an alpha-3 blocker.

Changed in linux-ti-omap4 (Ubuntu Maverick):
milestone: maverick-alpha-3 → ubuntu-10.10-beta
Tobin Davis (gruemaster)
tags: added: verification-done
Revision history for this message
Tobin Davis (gruemaster) wrote :

Sorry for not testing this earlier. Not getting the proper email for some reason. Please add "verification-needed" to the tags when testing is needed.

I have not seen this issue now in the last 24 hours running this kernel on the 20100802 image.

tags: added: iso-testing
Revision history for this message
Bryan Wu (cooloney) wrote :

In the latest 903.7 omap4 kernel, I didn't see this issue again. I believe it was fixed by TI's new kernel release.

Tobin,

Could you please help to test the new kernel package?
http://people.canonical.com/~roc/kernel/omap4/

Thanks a lot,
-Bryan

tags: added: verification-needede
removed: verification-done
tags: added: verification-needed
removed: verification-needede
Revision history for this message
Ricardo Salveti (rsalveti) wrote :

Tested with linux-image-2.6.34-903-omap4 and I can't reproduce the error anymore (using a swap file as before).

Changed in linux-ti-omap4 (Ubuntu Maverick):
status: In Progress → Fix Released
Oliver Grawert (ogra)
Changed in linux-ti-omap4 (Ubuntu Maverick):
milestone: ubuntu-10.10-beta → ubuntu-10.10
Tobin Davis (gruemaster)
tags: added: verification-done
removed: verification-needed
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.