[#7106] run testing on 4.5 toolchain would make kernel panic and not restart again

Document created by Aaronwu Employee on Oct 17, 2013
Version 1Show Document
  • View in full screen mode

[#7106] run testing on 4.5 toolchain would make kernel panic and not restart again

Submitted By: Mingquan Pan

Open Date

2012-05-16 02:08:05    

Priority:

Medium     Assignee:

Mingquan Pan

Stuart Henderson

Board:

N/A     Silicon Revision:

Resolution:

Not reproducible     Fixed In Release:

N/A

Processor:

BF561     

Host Operating System:

toolchain rev.:

    kernel rev.:

State:

Open     Found In Release:

N/A

Is this bug repeatable?:

N/A     

Summary: run testing on 4.5 toolchain would make kernel panic and not restart again

Details:

 

run testing on 4.5 toolchain would make kernel panic frequently and it can't get restart over again.

 

Targets:                    uclinux

Components:                 binutils gas ld sim gdb newlib gcc g++ gfortran objc libstdc++ libmudflap libgomp

Host:                       i686-pc-linux-gnu

Build:                      i686-pc-linux-gnu

Removing from PATH: /home/test/work/cruise/temp/bfin-uclinux/bin

Removing from PATH: /home/test/work/cruise/temp/bfin-elf/bin

Removing from PATH: /home/test/work/cruise/temp/bfin-linux-uclibc/bin

Found multiple versions of runtest, using the one at /usr/local/bin/runtest

Toolchain source:           /home/test/work/cruise/checkouts/toolchain

Path to logs output dir:    /home/test/work/cruise/temp/regtest_build4.5/logs

Path to tests output dir:   /home/test/work/cruise/temp/regtest_build4.5/tests

Path to uclinux output dir: /home/test/work/cruise/temp/regtest_build4.5/tests/uclinux

Previous uclinux test logs: /home/test/work/cruise/checkouts/toolchain/buildscript/test_results/uclinux

Creating log file:          /home/test/work/cruise/temp/regtest_build4.5/logs/log

Checking host sysctl:       /proc/sys/net/ipv4/tcp_tw_recycle

Checking target sysctl:     /proc/sys/net/ipv4/tcp_tw_recycle

Detected board cpu:         bf561-0.5

    Board at '10.100.4.50' being set so syslog doesn't fill /var/log/messages

Path to boards output dir:  /home/test/work/cruise/temp/regtest_build4.5/boards

*** Testing bfin-uclinux via hardware (10.100.4.50)         15 May 2012 15:08:15

binutils version:           2.21

binutils path:              /home/test/work/cruise/temp/bfin-uclinux/bin

*** Running tests on binutils                               15 May 2012 15:08:16

*** Running tests on gas                                    15 May 2012 15:08:21

*** Running tests on ld                                     15 May 2012 15:08:37

gdb version:                6.6

gdb path:                   /home/test/work/cruise/temp/bfin-uclinux/bin

*** Running tests on gdb                                    15 May 2012 15:09:18

gcc version:                4.5.3

gcc path:                   /home/test/work/cruise/temp/bfin-uclinux/bin

*** Running tests on gcc                                    15 May 2012 17:18:38

*** I think the board or network died during testing - sorry

*** Kill the current testing

/home/test/work/cruise/checkouts/toolchain/buildscript/lib/funcs.sh: line 164: 17303 Terminated              run_cmd_nodie $RUNTEST --tool ${pkg} $RUNTEST_FLAGS

*** Found old processes (Xvfb rsh); killing them to avoid   15 May 2012 18:22:03

    confusing the testsuites                               

*** Running tests on g++                                    15 May 2012 18:22:12

*** I think the board or network died during testing - sorry

*** Kill the current testing

/home/test/work/cruise/checkouts/toolchain/buildscript/lib/funcs.sh: line 164:  2078 Terminated              run_cmd_nodie $RUNTEST --tool ${pkg} $RUNTEST_FLAGS

*** Found old processes (Xvfb rsh); killing them to avoid   15 May 2012 18:22:34

    confusing the testsuites                               

*** Running tests on gfortran                               15 May 2012 18:22:43

*** I think the board or network died during testing - sorry

*** Kill the current testing

/home/test/work/cruise/checkouts/toolchain/buildscript/lib/funcs.sh: line 164:  2423 Terminated              run_cmd_nodie $RUNTEST --tool ${pkg} $RUNTEST_FLAGS

*** Found old processes (Xvfb rsh); killing them to avoid   15 May 2012 18:23:05

    confusing the testsuites                               

objc support not found in bfin-uclinux toolchain

*** Found old processes (Xvfb rsh); killing them to avoid   15 May 2012 18:23:14

    confusing the testsuites                               

*** Running tests on libstdc++                              15 May 2012 18:23:23

*** I think the board or network died during testing - sorry

*** Kill the current testing

/home/test/work/cruise/checkouts/toolchain/buildscript/lib/funcs.sh: line 164:  2917 Terminated              run_cmd_nodie $RUNTEST --tool ${pkg} $RUNTEST_FLAGS

*** Found old processes (Xvfb rsh); killing them to avoid   15 May 2012 18:23:45

    confusing the testsuites                               

*** Running tests on libmudflap                             15 May 2012 18:23:54

*** I think the board or network died during testing - sorry

*** Kill the current testing

/home/test/work/cruise/checkouts/toolchain/buildscript/lib/funcs.sh: line 143:  3114 Terminated              run_cmd_nodie $RUNTEST --tool ${pkg} $RUNTEST_FLAGS

*** Found old processes (Xvfb rsh); killing them to avoid   15 May 2012 18:24:15

    confusing the testsuites                               

*** Running tests on libgomp                                15 May 2012 18:24:24

*** I think the board or network died during testing - sorry

*** Kill the current testing

/home/test/work/cruise/checkouts/toolchain/buildscript/lib/funcs.sh: line 143:  3382 Terminated              run_cmd_nodie $RUNTEST --tool ${pkg} $RUNTEST_FLAGS

*** Legacify result names                                   15 May 2012 18:24:54

*** Done testing - results                                  15 May 2012 18:24:54

 

### New: binutils-2.21.sum in /home/test/work/cruise/temp/regtest_build4.5/tests/uclinux

        === binutils Summary ===

# of expected passes        71

# of unsupported tests        5

 

### New: g++-4.5.sum in /home/test/work/cruise/temp/regtest_build4.5/tests/uclinux

 

### New: gas-2.21.sum in /home/test/work/cruise/temp/regtest_build4.5/tests/uclinux

        === gas Summary ===

# of expected passes        125

# of expected failures        1

 

### New: gcc-4.5.sum in /home/test/work/cruise/temp/regtest_build4.5/tests/uclinux

 

### New: gdb-6.6.sum in /home/test/work/cruise/temp/regtest_build4.5/tests/uclinux

        === gdb Summary ===

# of expected passes        9402

# of unexpected failures    821

# of unexpected successes    2

# of expected failures        39

# of known failures        58

# of unresolved testcases    15

# of untested testcases        10

# of unsupported tests        18

 

### New: gfortran-4.5.sum in /home/test/work/cruise/temp/regtest_build4.5/tests/uclinux

 

 

On the console log:

 

testing uclinux gcc^M

NULL pointer access^M

Kernel OOPS in progress^M

Deferred Exception context^M

CURRENT PROCESS:^M

COMM=run-id-1.exe.17 PID=17376  CPU=0^M

TEXT = 0x02860040-0x028609e0        DATA = 0x02860a00-0x02860a78^M

BSS = 0x02860a78-0x02860ac0  USER-STACK = 0x02874f40^M

^M

return address: [0x00000470]; contents of:^M

0x00000450:  9101  5601  9300  0010  0000  0000  0000  0000 ^M

0x00000460:  9101  5401  9300  0010  0000  0000  0000  0000 ^M

0x00000470: [9101] 5801  9300  0010  0000  0000  0000  0000 ^M

0x00000480:  0000  0000  0000  0000  00a4  0000  0000  0000 ^M

^M

ADSP-BF561-0.5 600(MHz CCLK) 100(MHz SCLK) (mpu off)^M

Linux version 3.3.0-ADI-2012R1-pre-02011-g68bbda2 (test@43-bf561-toolchain) (gcc version 4.5.3 (ADI-trunk/svn-5845) ) #22 Sun Apr 29 18:39:22 GMT 2012^M

^M

SEQUENCER STATUS:               Not tainted^M

SEQSTAT: 00060027  IPEND: 0048  IMASK: 003f  SYSCFG: 0006^M

  Peripheral interrupts masked off^M

  Kernel interrupts masked off^M

  EXCAUSE   : 0x27^M

  physical IVG3 asserted : <0xffa00828> { _trap + 0x0 }^M

  physical IVG6 asserted : <0xffa00eb0> { _evt_timer + 0x0 }^M

  logical irq   6 mapped  : <0xffa00468> { _bfin_coretmr_interrupt + 0x0 }^M

  logical irq  35 mapped  : <0x00000397> /* Maybe null pointer? */^M

Double Fault^M

Kernel OOPS in progress^M

Deferred Exception context^M

CURRENT PROCESS:^M

COMM=run-id-1.exe.17 PID=17376  CPU=0^M

TEXT = 0x02860040-0x028609e0        DATA = 0x02860a00-0x02860a78^M

BSS = 0x02860a78-0x02860ac0  USER-STACK = 0x02874f40^M

^M

return address: [0x00008462]; contents of:^M

0x00008440:  acfb  2edc  60f9  4f71  5408  c682  0390  e140 ^M

0x00008450:  0014  e100  d400  e308  acef  a0a1  544f  2e7d ^M

0x00008460:  3046 [9129] 64c0  e3ff  f9c9  e140  0014  304e ^M

0x00008470:  e100  d494  64c1  e308  acdf  ad2d  2fdb  e140 ^M

^M

ADSP-BF561-0.5 600(MHz CCLK) 100(MHz SCLK) (mpu off)^M

Linux version 3.3.0-ADI-2012R1-pre-02011-g68bbda2 (test@43-bf561-toolchain) (gcc version 4.5.3 (ADI-trunk/svn-5845) ) #22 Sun Apr 29 18:39:22 GMT 2012^M

^MSEQUENCER STATUS:               Not tainted^M

SEQSTAT: 00060027  IPEND: 0048  IMASK: 003f  SYSCFG: 0006^M

  Peripheral interrupts masked off^M

  Kernel interrupts masked off^M

  EXCAUSE   : 0x27^M

  physical IVG3 asserted : <0xffa00828> { _trap + 0x0 }^M

  physical IVG6 asserted : <0xffa00eb0> { _evt_timer + 0x0 }^M

  logical irq   6 mapped  : <0xffa00468> { _bfin_coretmr_interrupt + 0x0 }^M

  logical irq  35 mapped  : <0x00000397> /* Maybe null pointer? */^M

Double Fault^M

Kernel OOPS in progress^M

Deferred Exception context^M

CURRENT PROCESS:^M

COMM=run-id-1.exe.17 PID=17376  CPU=0^M

TEXT = 0x02860040-0x028609e0        DATA = 0x02860a00-0x02860a78^M

BSS = 0x02860a78-0x02860ac0  USER-STACK = 0x02874f40^M

^M

return address: [0x00008462]; contents of:^M

0x00008440:  acfb  2edc  60f9  4f71  5408  c682  0390  e140 ^M

0x00008450:  0014  e100  d400  e308  acef  a0a1  544f  2e7d ^M

0x00008460:  3046 [9129] 64c0  e3ff  f9c9  e140  0014  304e ^M

0x00008470:  e100  d494  64c1  e308  acdf  ad2d  2fdb  e140 ^M

^M

ADSP-BF561-0.5 600(MHz CCLK) 100(MHz SCLK) (mpu off)^M

Linux version 3.3.0-ADI-2012R1-pre-02011-g68bbda2 (test@43-bf561-toolchain) (gcc version 4.5.3 (ADI-trunk/svn-5845) ) #22 Sun Apr 29 18:39:22 GMT 2012^M

^M

SEQUENCER STATUS:               Not tainted^M

SEQSTAT: 00060027  IPEND: 0068  IMASK: 003f  SYSCFG: 0006^M

  Peripheral interrupts masked off^M

  Kernel interrupts masked off^M

  EXCAUSE   : 0x27^M

  physical IVG3 asserted : <0xffa00828> { _trap + 0x0 }^M

  physical IVG5 asserted : <0xffa00d94> { _evt_ivhw + 0x0 }^M

  physical IVG6 asserted : <0xffa00eb0> { _evt_timer + 0x0 }^M

  logical irq   6 mapped  : <0xffa00468> { _bfin_coretmr_interrupt + 0x0 }^M

@                                                                                                                                                                                                             @                                logical irq  35 mapped  : <0x00000397> /* Maybe null pointer? */, <0x0000003f> /* Maybe null pointer? */, <0x0000003f> /* Maybe null pointer? */, <0x0000003f> /* Maybe null pointer? */, <0x0000003f> /* Maybe null pointer? */, <0x0000003f> /* Maybe null pointer? */, <0x0000003f> /* Maybe null pointer? */, <0x0000003f> /* Maybe null pointer? */, <0x0000003f> /* Maybe null pointer? */, <0x0000003f> /* Maybe null pointer? */, <0x0000003f> /* Maybe null pointer? */, <0x0000003f> /* Maybe null pointer? */, <0x0000003f> /* Maybe null pointer? */, <0x0000003f> /* Maybe null pointer? */, <0x0000003f> /* Maybe null pointer? */, <0x0000003f> /* Maybe null pointer? */, <0x0000003f> /* Maybe null pointer? */, <0x0000003f> /* Maybe null pointer? */, <0x0000003f> /* Maybe null pointer? */, <0x0000003f> /* Maybe null pointer? */, <0x0000003f> /* Maybe null pointer? */, <0x0000003f> /* Maybe null pointer? */, <0x0000003f> /* Maybe null pointer? */, <0x0000003f> /* Maybe null pointer? */, <0x0000003f> /* Maybe null pointer? */, <0x0000003f> /* Maybe null pointer? */, <0x0000003f> /* Maybe null pointer? */, <0x0000003f> /* Maybe null pointer? */, <0x0000003f> /* Maybe null pointer? */, <0x0000003f> /* Maybe null pointer? */, <0x0000003f> /* Maybe null pointer? */, <0x0000003f> /* Maybe null pointer? */, <0x0000003f> /* Maybe null pointer? */, <0x0000003f> /* Maybe null pointer? */, <0x0000003f> /* Maybe null pointer? */, <0x0000003f> /* Maybe null pointer? */, <0x0000003f> /* Maybe null pointer? */, <0x0000003f> /* Maybe null pointer? */, <0x0000003f> /* Maybe null pointer? */, <0x0000003f> /* Maybe null pointer? */, <0x0000003f> /* Maybe null pointer? */, <0x00000                                                             

 

 

Follow-ups

 

--- Stuart Henderson                                         2012-06-06 11:03:30

I'm not seeing this.  Can you confirm if it still occurs, and if so provide me

with exactly how to reproduce?

 

--- Mingquan Pan                                             2012-06-07 04:09:35

Running 4.5 toolchain testing is always seeing stuffs like above here. I execute

the testing running on trunk head kernel:

 

************* Tested uclinux-dist svn info *************

U    testsuites/ltp/ltp_auto_test.exp

U    user/blkfin-apps/icc_utils/include/debug.h

U    user/blkfin-apps/icc_utils/icc_loader/icc_loader.c

U    user/blkfin-apps/icc_utils/icc_core/corebboot.c

U    user/blkfin-apps/icc_utils/example/test_app/Makefile

A    user/blkfin-apps/icc_utils/example/test_app/icc_latency.c

U    user/blkfin-apps/icc_utils/example/task/Makefile

A    user/blkfin-apps/icc_utils/example/task/icc_latency.c

Updated to revision 10701.

Path: .

URL: svn://10.99.29.20/uclinux-dist/trunk

Repository Root: svn://10.99.29.20/uclinux-dist

Repository UUID: 3e2a5524-39e8-4cb9-84a9-c49da69cba3a

Revision: 10701

Node Kind: directory

Schedule: normal

Last Changed Author: steven.miao

Last Changed Rev: 10701

Last Changed Date: 2012-06-06 04:26:03 +0000 (Wed, 06 Jun 2012)

************* Tested linux svn info ********************

Updating from ed5abf215c05106d5dc19e24385619307ca1ed1e to

2fa2531e3460ad5e4a7bc1e951d4d12fe40f40a4

Fast forward

arch/blackfin/include/asm/bfin_simple_timer.h |    6 +

arch/blackfin/mach-bf561/boards/ezkit.c       |   28 +++

arch/blackfin/mach-bf561/icc.c                |   14 -

arch/blackfin/mach-bf609/boards/ezkit.c       |   28 +++

arch/blackfin/mach-bf609/icc.c                |   16 --

arch/blackfin/mach-common/Makefile            |    1 -

arch/blackfin/mach-common/icc.c               |   51 ----

drivers/char/bfin_simple_timer.c              |   23 ++-

drivers/staging/icc/core/protocol.c           |  323

+++++++++++++++++--------

9 files changed, 306 insertions(+), 184 deletions(-)

delete mode 100644 arch/blackfin/mach-common/icc.c

commit 2fa2531e3460ad5e4a7bc1e951d4d12fe40f40a4

Author: Sonic Zhang <sonic.zhang@analog.com>

Date:   Wed Jun 6 13:40:36 2012 +0800

 

Testing is running on target with following versions:

kernel:    Linux release 3.3.0-ADI-2012R1-pre-02011-g2fa2531, build #14 Thu Jun

7 02:49:40 GMT 2012

toolchain: bfin-uclinux-gcc release gcc version 4.5.3 (ADI-trunk/svn-5845)

user-dist: release svn-10701, build #4 Thu Jun 7 02:48:49 GMT 2012

 

And running toolchain test against the built toolchain with this command(target

board is running updated kernel built with the tested toolchain):

cd ~/work/cruise/checkouts/toolchain/buildscript

./toolchain-regtest -s /home/test/work/cruise/checkouts/toolchain -o

~/work/cruise/test_scripts/toolchain/toolchain-build/toolchain-2012_Jun_05_17_22/bfin

-b /home/test/temp/regtest_build1208 -t linux-uclibc -T 10.100.4.50 -a -L

 

This command would produce error log like above.

 

--- Stuart Henderson                                         2012-06-07 12:23:48

Hi,

I'm currently seeing an error when booting uclinux built with gcc 4.5 (even

2011R1_45, which used to work):

...

Calibrating delay loop... 995.32 BogoMIPS (lpj=1990656)

pid_max: default: 32768 minimum: 301

Security Framework initialized

Mount-cache hash table entries: 512

Blackfin Scratchpad data SRAM: 4 KB

Blackfin L1 Data A SRAM: 16 KB (16 KB free)

Blackfin L1 Data B SRAM: 16 KB (16 KB free)

Blackfin L1 Instruction SRAM: 48 KB (35 KB free)

NET: Registered protocol family 16

gpiochip_add: registered GPIOs 0 to 47 on device: BFIN-GPIO

Blackfin DMA Controller

stamp_init(): registering device resources

------------[ cut here ]------------

WARNING: at kernel/resource.c:597

Modules linked in:

 

ADSP-BF537-0.2 500(MHz CCLK) 125(MHz SCLK) (mpu off)

Linux version 3.3.0-ADI-2012R1-pre-00723-ga1bb857 (shender@edin-florence) (gcc

version 4.5.3 (ADI-2011R1-RC4_45) )2

 

SEQUENCER STATUS:               Not tainted

SEQSTAT: 00002021  IPEND: 8008  IMASK: ffff  SYSCFG: 0006

  EXCAUSE   : 0x21

...

 

Is there a specific configuration I should be using for building uclinux at the

moment?  I'm using the default 537 config with strace disabled.

 

--- Mingquan Pan                                             2012-06-07 23:53:26

there is a build kernel script under

toolchain/buildscript/build_toolchain_kernel.exp. It disables some items when

toolchain testing does not need. Run : /build_toolchain_kernel.exp BF537-STAMP

should build a working kernel.

 

I am now running 4.5 toolchain on bf561-ezkit and bf527-stamp. At least the

built kernel can start up and show up shell prompt. It just fails in toolchain

testing especially in gdb.

 

--- Stuart Henderson                                         2012-06-08 09:36:02

bf527-stamp?  i've tried bf537-stamp and i get an "ifconfig: SIOCSIFADDR:

No such device" error.

 

the gdb issues are a problem with GDB being too old a version to fully support

all of gcc 4.5's features.

 

--- Mingquan Pan                                             2012-06-12 05:28:05

sorry, I use bf527-ezkit. and the ethernet ifconfig should work without problem

with the trunk head kernel.

 

root:/> version

kernel:    Linux release 3.3.0-ADI-2012R1-pre-00715-ged5abf2, build #6 Tue Jun

5 23:09:29 CST 2012

toolchain: bfin-uclinux-gcc release gcc version 4.5.3 (ADI-trunk/svn-5845)

user-dist: release svn-10697, build #2 Tue Jun 5 23:08:21 CST 2012

root:/>

root:/> ifconfig

eth0      Link encap:Ethernet  HWaddr 00:E0:22:FE:C0:9A 

          inet addr:10.100.4.50  Bcast:10.100.4.255  Mask:255.255.255.0

          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1

          RX packets:1154 errors:0 dropped:0 overruns:0 frame:0

          TX packets:1034 errors:0 dropped:0 overruns:0 carrier:0

          collisions:0 txqueuelen:1000

          RX bytes:638582 (623.6 KiB)  TX bytes:72779 (71.0 KiB)

 

lo        Link encap:Local Loopback 

          inet addr:127.0.0.1  Mask:255.0.0.0

          UP LOOPBACK RUNNING  MTU:16436  Metric:1

          RX packets:0 errors:0 dropped:0 overruns:0 frame:0

          TX packets:0 errors:0 dropped:0 overruns:0 carrier:0

          collisions:0 txqueuelen:0

          RX bytes:0 (0.0 B)  TX bytes:0 (0.0 B)

 

As you say gdb is too old, is there any plan to upgrade ? and what should we

disable gdb testing for now in regression?

 

 

 

    Files

    Changes

    Commits

    Dependencies

    Duplicates

    Associations

    Tags

 

File Name     File Type     File Size     Posted By

No Files Were Found

Attachments

    Outcomes