lkml.org 
[lkml]   [2014]   [Oct]   [10]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: RCU bug with v3.17-rc3 ?
On Fri, Oct 10, 2014 at 05:18:35PM +0100, Russell King - ARM Linux wrote:
> On Fri, Oct 10, 2014 at 12:47:06AM +0300, Aaro Koskinen wrote:
> > On Thu, Oct 09, 2014 at 10:41:01PM +0200, Rabin Vincent wrote:
> > > What GCC version are you using?
> > >
> > > 4.8.1 and 4.8.2 are known to miscompile the ARM kernel and these
> > > find_get_entry() crashes with 0xffffffff involved smell a lot like the
> > > earlier reports from kernels build with those compilers:
> > >
> > > https://lkml.org/lkml/2014/6/25/456
> > > https://lkml.org/lkml/2014/6/30/375
> > > https://lkml.org/lkml/2014/6/30/660
> > > https://gcc.gnu.org/bugzilla/show_bug.cgi?id=58854
> > > https://lkml.org/lkml/2014/5/9/330
> >
> > Is it possible to blacklist those GCC versions on ARM somehow as it
> > seems people are still using them?
> >
> > This bug also ruined a file system on one of my boxes last year
> > (see e.g. http://marc.info/?l=linux-arm-kernel&m=139033442527244&w=2).
>
> Given that, why the fsck (pun intended) did you not shout a little louder
> about getting it blacklisted. Looking at your marc.info URL, there's
> very little information there which hints at filesystem corruption, and
> it's a thread of only *one* message according to marc.info.
>
> Even _if_ I did read the message you point to above, that on its own did
> not hint at filesystem corruption.
>
> So, would you please mind passing on further details about this,
> specifically which function in the ext4 code is affected, so it can
> be properly written up.

I have not done any proper deeper analysis. After I first mailed about
the issue I just downgraded GCC and pretty much forgot about it until
an engineer from some commercial Linux vendor replied privately months
later and kindly pointed me the needed GCC fix (which I then shared
in the reply). Then I just moved on using a newer GCC with no issues.
Obviously this was not a widespread problem since no one else
reported the same.

Today I again booted a kernel compiled with GCC 4.8.2 and still was able
reproduce the issue, and I think below shows that at least ext3 can
easily end up in inconsistent state using these compiler versions:

0) Run the bad kernel:

~ # dmesg|grep GCC
[ 0.000000] Linux version 3.17.0-mvebu-los_9755+ (aaro@cooljazz) (gcc version 4.8.2 (GCC) ) #1 Fri Oct 10 21:05:20 EEST 2014

1) Start with small ext3 (writeback) fs with gcc tarball:

/mnt/test # ls -l
total 84092
-rw-r--r-- 1 root root 85999682 Apr 24 21:52 gcc-4.8.2.tar.bz2
drwx------ 2 root root 16384 Oct 10 10:33 lost+found
/mnt/test # df -h .
Filesystem Size Used Available Use% Mounted on
/dev/sdc1 3.8G 90.2M 3.5G 2% /mnt/test

2) Extract, delete & crash:

/mnt/test # tar xjf gcc-4.8.2.tar.bz2
/mnt/test # rm -rf gcc-4.8.2
rm: can't remove 'gcc-4.8.2/libgfortran/generated': Directory not empty
rm: can't remove 'gcc-4.8.2/libgfortran': Directory not empty
rm: can't remove 'gcc-4.8.2/gcc/testsuite/gcc.dg/compat/struct-by-value-18a_y.c': No such file or directory
rm: can't remove 'gcc-4.8.2/gcc/testsuite/gcc.dg/compat': Directory not empty
rm: can't remove 'gcc-4.8.2/gcc/testsuite/gcc.dg/tree-ssa': Directory not empty
rm: can't remove 'gcc-4.8.2/gcc/testsuite/gcc.dg': Directory not empty
rm: can't remove 'gcc-4.8.2/gcc/testsuite/gfortran.dg/result_default_init_1.f90': No such file or directory
rm: can't remove 'gcc-4.8.2/gcc/testsuite/gfortran.dg': Directory not empty
[ 960.864433] Unable to handle kernel paging request at virtual address ffffffff
[ 960.930597] pgd = df6e0000
[ 960.990849] [ffffffff] *pgd=1fffd831, *pte=00000000, *ppte=00000000
[ 961.056512] Internal error: Oops: 1 [#1] ARM
[ 961.120063] Modules linked in:
[ 961.180974] CPU: 0 PID: 684 Comm: rm Not tainted 3.17.0-mvebu-los_9755+ #1
[ 961.247146] task: df447b00 ti: df4de000 task.ti: df4de000
[ 961.311524] PC is at find_get_entry+0x28/0x84
[ 961.375037] LR is at radix_tree_lookup_slot+0x1c/0x2c
[ 961.439061] pc : [<c006e418>] lr : [<c018392c>] psr: a0000013
[ 961.439061] sp : df4dfc68 ip : 00000000 fp : df4dfc7c
[ 961.570018] r10: 00000001 r9 : c04e3253 r8 : df020b60
[ 961.634596] r7 : 0009001a r6 : 00000000 r5 : 0009001a r4 : df020c90
[ 961.700070] r3 : ffffffff r2 : 00000000 r1 : 0009001a r0 : ffffffff
[ 961.764437] Flags: NzCv IRQs on FIQs on Mode SVC_32 ISA ARM Segment user
[ 961.830518] Control: 0005317f Table: 1f6e0000 DAC: 00000015
[ 961.895866] Process rm (pid: 684, stack limit = 0xdf4de1c0)
[ 961.960597] Stack: (0xdf4dfc68 to 0xdf4e0000)
[ 962.022968] fc60: 00000001 df020c8c df4dfcb4 df4dfc80 c006eef68 c006e400
[ 962.091214] fc80: c00d4e80 c00d4764 00001000 0009001a 00000000 00000000 df0200b60 df020b60
[ 962.159490] fca0: df020bd8 df04e4d8 df4dfd04 df4dfcb8 c00d34c0 c006ef44 000000000 df4dfcc8
[ 962.226940] fcc0: c00d4e80 c00d4764 00001000 00000001 df4dfd84 dd1c73f0 000900306 00000000
[ 962.295558] fce0: 00090068 00000000 00000000 df020b60 df04e4d8 00000181 df4dffd4c df4dfd08
[ 962.364710] fd00: c00d4828 c00d347c 00000000 00000001 df4dfdc4 dd1c73f0 000000000 00000000
[ 962.433394] fd20: 00000000 00000000 df4dfd84 00090002 00001000 dbaa2200 df0200b60 df04e4d8
[ 962.501810] fd40: df4dfdbc df4dfd50 c00d4e80 c00d4764 00001000 df4dfd60 c01411284 c0148708
[ 962.569685] fd60: 0009001a 00000000 c0ebc7c0 df041180 00000002 00000000 df4dffd9c df4dfd88
[ 962.639143] fd80: c003813c c0038084 df041180 df0b7320 df4dfdac 00090002 000000000 dbaa2200
[ 962.708562] fda0: df4dfe4c df04e4d8 00000181 df04e4d8 df4dfe24 df4dfdc0 c010887c0 c00d4e6c
[ 962.778108] fdc0: 00001000 c038caf8 0000128f 00000000 00000000 00011000 000000001 c9c59740
[ 962.846670] fde0: 0009001a 00000000 00000a26 c824f240 00000010 00000000 df4dffe1c df04e4d8
[ 962.913956] fe00: df04e4d8 df4dfe4c de53cf40 de53cf40 00000000 df04e4d8 df4dffe44 df4dfe28
[ 962.980679] fe20: c010c5a8 c01086c4 df04e4d8 dee12000 dbaa2200 df04e4b4 df4dffe84 df4dfe48
[ 963.046696] fe40: c0115dc4 c010c584 dd1c73f0 00000000 00000100 00000012 000000000 c0fbfe00
[ 963.112648] fe60: df04e4d8 dd1c73f0 de53cf40 00000000 df4dff04 df04e4d8 df4dffecc df4dfe88
[ 963.178402] fe80: c0116b24 c0115ce0 00000000 c00b3b24 df4dfeac c067b174 5437dd0a4 22921900
[ 963.244947] fea0: df4dfecc df4dfeb0 c00b7a50 c19ca440 df04e4d8 df04e534 dd1c773f0 000b6650
[ 963.311517] fec0: df4dfefc df4dfed0 c00b7e4c c01168d8 df4dfefc df4dfee0 c19caa440 00000000
[ 963.377319] fee0: df4e6000 00000000 000b6650 ffffff9c df4dff94 df4dff00 c00b880b0 c00b7d94
[ 963.443083] ff00: 5437d035 00000000 dba4a8d0 d899f6e8 78ae7ba4 0000000d df4e6603c 0000000c
[ 963.509416] ff20: 00000000 c0009624 dd1c73f0 00000000 00000004 00000038 000000000 00000000
[ 963.575556] ff40: 00024182 00000000 00800021 c04c81b4 00000001 000003e8 0000003e8 00000000
[ 963.641281] ff60: 0000024d 00000000 4bfad53f 000b6650 00000008 0000000c 00000000a c0009624
[ 963.707194] ff80: df4de000 00000000 df4dffa4 df4dff98 c00b8e20 c00b7ed0 000000000 df4dffa8
[ 963.773584] ffa0: c00094c0 c00b8e18 000b6650 00000008 000b6650 bed03990 bed033990 00008000
[ 963.841022] ffc0: 000b6650 00000008 0000000c 0000000a 000b6650 00000000 b6fccc000 00000000
[ 963.907530] ffe0: 00093224 bed0398c 00071284 b6efa39c 60000010 000b6650 0000fffff 0000ffff
[ 963.973653] Backtrace: [ 964.032680] [<c006e3f0>] (find_get_entry) from [<c006ef68>] (pagecache_get_page+0x34/0x1fc)
[ 964.100751] r5:df020c8c r4:00000001
[ 964.162591] [<c006ef34>] (pagecache_get_page) from [<c00d34c0>] (__find_get_b
block_slow+0x54/0x16c)
[ 964.291505] r10:df04e4d8 r9:df020bd8 r8:df020b60 r7:df020b60 r6:00000000 r5:
:00000000
[ 964.361857] r4:0009001a
[ 964.425342] [<c00d346c>] (__find_get_block_slow) from [<c00d4828>] (__find_ge
et_block+0xd4/0x1e4)
[ 964.498345] r9:00000181 r8:df04e4d8 r7:df020b60 r6:00000000 r5:00000000 r4:0
00090068
[ 964.570979] [<c00d4754>] (__find_get_block) from [<c00d4e80>] (__getblk+0x24/
/0x358)
[ 964.643833] r8:df04e4d8 r7:df020b60 r6:dbaa2200 r5:00001000 r4:00090002
[ 964.716031] [<c00d4e5c>] (__getblk) from [<c01087c0>] (__ext4_get_inode_loc+0
0x10c/0x454)
[ 964.790734] r10:df04e4d8 r9:00000181 r8:df04e4d8 r7:df4dfe4c r6:dbaa2200 r5:
:00000000
[ 964.865945] r4:00090002
[ 964.934187] [<c01086b4>] (__ext4_get_inode_loc) from [<c010c5a8>] (ext4_reser
rve_inode_write+0x34/0x9c)
[ 965.080216] r10:df04e4d8 r9:00000000 r8:de53cf40 r7:de53cf40 r6:df4dfe4c r5:
:df04e4d8
[ 965.159656] r4:df04e4d8
[ 965.232230] [<c010c574>] (ext4_reserve_inode_write) from [<c0115dc4>] (ext4_o
orphan_add+0xf4/0x218)
[ 965.385687] r7:df04e4b4 r6:dbaa2200 r5:dee12000 r4:df04e4d8
[ 965.464523] [<c0115cd0>] (ext4_orphan_add) from [<c0116b24>] (ext4_unlink+0x2
25c/0x26c)
[ 965.547430] r10:df04e4d8 r9:df4dff04 r8:00000000 r7:de53cf40 r6:dd1c73f0 r5:
:df04e4d8
[ 965.631429] r4:c0fbfe00
[ 965.708445] [<c01168c8>] (ext4_unlink) from [<c00b7e4c>] (vfs_unlink+0xc8/0x1
13c)
[ 965.792677] r8:000b6650 r7:dd1c73f0 r6:df04e534 r5:df04e4d8 r4:c19ca440
[ 965.877297] [<c00b7d84>] (vfs_unlink) from [<c00b80b0>] (do_unlinkat+0x1f0/0x
x210)
[ 965.963851] r9:ffffff9c r8:000b6650 r7:00000000 r6:df4e6000 r5:00000000 r4:c
c19ca440
[ 966.051666] [<c00b7ec0>] (do_unlinkat) from [<c00b8e20>] (SyS_unlink+0x18/0x1
1c)
[ 966.139262] r10:00000000 r9:df4de000 r8:c0009624 r7:0000000a r6:0000000c r5:
:00000008
[ 966.228970] r4:000b6650
[ 966.311776] [<c00b8e08>] (SyS_unlink) from [<c00094c0>] (ret_fast_syscall+0x0
0/0x2c)
[ 966.401452] Code: e1a01005 eb04553f e2503000 0a00000f (e5930000)
[ 966.608250] ---[ end trace a1b54af48fda09ed ]---
[ 966.693854] Kernel panic - not syncing: Fatal exception
[ 966.781707] ---[ end Kernel panic - not syncing: Fatal exception

3) Boot a good kernel:

~ # dmesg | grep GCC
[ 0.000000] Linux version 3.17.0-mvebu-los_1b42 (aaro@cooljazz) (gcc version 4.9.1 (GCC) ) #1 Thu Oct 9 06:46:07 EEST 2014

4) Use the beforementioned file system and try to clean the mess:

/mnt/test # df -h .
Filesystem Size Used Available Use% Mounted on
/dev/sdc1 3.8G 796.2M 2.8G 22% /mnt/test
/mnt/test # rm -rf gcc-4.8.2
rm: can't remove 'gcc-4.8.2/gcc/testsuite/gcc.dg/tree-ssa': Directory not empty
rm: can't remove 'gcc-4.8.2/gcc/testsuite/gcc.dg': Directory not empty
rm: can't remove 'gcc-4.8.2/gcc/testsuite/gfortran.dg': Directory not empty
rm: can't remove 'gcc-4.8.2/gcc/testsuite': Directory not empty
rm: can't remove 'gcc-4.8.2/gcc': Directory not empty
rm: can't remove 'gcc-4.8.2': Directory not empty
/mnt/test # rm -rf gcc-4.8.2
rm: can't remove 'gcc-4.8.2/gcc/testsuite/gcc.dg/tree-ssa': Directory not empty
rm: can't remove 'gcc-4.8.2/gcc/testsuite/gcc.dg': Directory not empty
rm: can't remove 'gcc-4.8.2/gcc/testsuite/gfortran.dg': Directory not empty
rm: can't remove 'gcc-4.8.2/gcc/testsuite': Directory not empty
rm: can't remove 'gcc-4.8.2/gcc': Directory not empty
rm: can't remove 'gcc-4.8.2': Directory not empty
/mnt/test # df -h .
Filesystem Size Used Available Use% Mounted on
/dev/sdc1 3.8G 90.5M 3.5G 2% /mnt/test
/mnt/test # find gcc-4.8.2
gcc-4.8.2
gcc-4.8.2/gcc
gcc-4.8.2/gcc/testsuite
gcc-4.8.2/gcc/testsuite/gcc.dg
gcc-4.8.2/gcc/testsuite/gcc.dg/tree-ssa
find: gcc-4.8.2/gcc/testsuite/gcc.dg/tree-ssa/forwprop-8.c: No such file or directory
gcc-4.8.2/gcc/testsuite/gfortran.dg
find: gcc-4.8.2/gcc/testsuite/gfortran.dg/result_default_init_1.f90: No such file or directory

5) fsck to rescue:

/mnt/test # cd /
~ # umount /mnt/test
~ # fsck /dev/sdc1
fsck 1.42.9 (28-Dec-2013)
e2fsck 1.42.9 (28-Dec-2013)
/dev/sdc1: clean, 21/262144 files, 72408/1048576 blocks
~ # fsck -f /dev/sdc1
fsck 1.42.9 (28-Dec-2013)
e2fsck 1.42.9 (28-Dec-2013)
Pass 1: Checking inodes, blocks, and sizes
Pass 2: Checking directory structure
Problem in HTREE directory inode 118267: block #4 has bad min hash
Problem in HTREE directory inode 118267: block #26 has bad max hash
Invalid HTREE directory inode 118267 (/gcc-4.8.2/gcc/testsuite/gfortran.dg). Clear HTree index<y>? yes
Problem in HTREE directory inode 174218: block #8 has bad min hash
Invalid HTREE directory inode 174218 (/gcc-4.8.2/gcc/testsuite/gcc.dg/tree-ssa). Clear HTree index<y>? yes
Pass 3: Checking directory connectivity
Pass 3A: Optimizing directories
Pass 4: Checking reference counts
Pass 5: Checking group summary information

/dev/sdc1: ***** FILE SYSTEM WAS MODIFIED *****
/dev/sdc1: 21/262144 files (19.0% non-contiguous), 72368/1048576 blocks
~ # mount /dev/sdc1 /mnt/
~ # rm -rf /mnt/gcc-4.8.2
~ #

So in this case fsck was able to fix it.

A.


\
 
 \ /
  Last update: 2014-10-10 23:21    [W:0.103 / U:0.044 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site