EPICS Home

Experimental Physics and Industrial Control System


 
2002  2003  2004  2005  2006  2007  2008  <2009>  2010  2011  2012  2013  2014  2015  2016  2017  2018  2019  2020  2021  2022  2023  2024  Index 2002  2003  2004  2005  2006  2007  2008  <2009>  2010  2011  2012  2013  2014  2015  2016  2017  2018  2019  2020  2021  2022  2023  2024 
<== Date ==> <== Thread ==>
Subject: Re: Failing EPICS Test Harness on RTEMS-mvme5500
From: Kate Feng <[email protected]>
To: Ralph Lange <[email protected]>
Cc: EPICS Core Talk <[email protected]>
Date: Fri, 24 Apr 2009 11:28:38 -0400 
Ralph Lange wrote:

Here's the report from my colleague:

Using RTEMS 4.9.1 (without the patch) yields the original error.

Yes, it needs the patch.  The patch was sent against the 4.9.1.
I did not get a chance to see the difference between 4.9.1 and 4.9.0.
More below.

Using RTEMS 4.9.2 doesn' t work, either.

Applying the patch to RTEMS 4.9.0 or 4.9.1 doesn't compile:

   powerpc-rtems4.9-gcc --pipe -B../../../../../.././lib/
   -B../../../../../.././mvme5500/lib/ -specs bsp_specs -qrtems
-DHAVE_CONFIG_H -isystem ../../../../../.././mvme5500/lib/include -Wall -ansi -fasm -fno-strict-aliasing -mcpu=7450 -Dmpc7455 
   -mno-altivec -mabi=altivec -mvrsave=no -mmultiple -mstring
   -mstrict-align
   -O2 -g -MT irq_init.o -MD -MP -MF .deps/irq_init.Tpo -c -o
   irq_init.o `test -f 'irq/irq_init.c' || echo
'../../../../../../../../rtems-4.9.0/c/src/lib/libbsp/powerpc/mvme5500/'`irq/irq_init.c
../../../../../../../../rtems-4.9.0/c/src/lib/libbsp/powerpc/mvme5500/irq/irq_init.c:43: 
   warning: initialization from incompatible pointer type
../../../../../../../../rtems-4.9.0/c/src/lib/libbsp/powerpc/mvme5500/irq/irq_init.c:43: 
   warning: initialization from incompatible pointer type
../../../../../../../../rtems-4.9.0/c/src/lib/libbsp/powerpc/mvme5500/irq/irq_init.c:43: 
   warning: initialization from incompatible pointer type
../../../../../../../../rtems-4.9.0/c/src/lib/libbsp/powerpc/mvme5500/irq/irq_init.c:43: 
   warning: initialization from incompatible pointer type
   mv -f .deps/irq_init.Tpo .deps/irq_init.Po
make[6]: *** No rule to make target `irq/irq.c', needed by `irq.o'. Stop. 
The patch should have replaced the irq.c file with the BSP_irq.c one.
See below.  Can you check to see if Makefile.am is updated by the patch ?
It is possible that the patched Makefile.am is dated as 2009-03-16.
I would suggest one to remove the "Makefile.in" file,, and "touch Makefile.am". 
Then compile again.  This should solve the problem.  Another way is to
grab the entire tar file from
http://www.nsls.bnl.gov/facility/expsys/software/EPICS/

************************************************************************
--- mvme5500.orig/Makefile.am    2008-09-24 01:38:43.000000000 -0400
+++ mvme5500/Makefile.am    2009-03-16 13:57:56.000000000 -0400
@@ -32,7 +32,9 @@
    ../../powerpc/shared/startup/sbrk.c ../../shared/bootcard.c \
    ../../shared/bsppredriverhook.c startup/bspclean.c \
    ../../shared/bsplibc.c ../../shared/bsppost.c \
-    ../../shared/gnatinstallhandler.c startup/reboot.c
+    ../../shared/gnatinstallhandler.c startup/reboot.c \
+    ../../powerpc/shared/startup/probeMemEnd.c
+
pclock_SOURCES = ../../powerpc/shared/clock/p_clock.c

include_bsp_HEADERS = ../../powerpc/shared/console/uart.h
@@ -45,7 +47,7 @@
    pci/pcifinddevice.c

include_bsp_HEADERS += irq/irq.h
-irq_SOURCES = irq/irq_init.c irq/irq.c
+irq_SOURCES = irq/irq_init.c irq/BSP_irq.c

Cheers,
Kate


(Exactly the same problem with 4.9.1.)




Clueless,
Ralph


On 21.04.2009 21:14 Kate Feng wrote:

Hi Ralph,

Ralph Lange wrote:

Hm. We get the attached result when running the test harness on a 5500.
Are we missing something obvious? This is RTEMS 4.9.0 and the 5500 BSP that is shipped with it. 
Yes, the patch is still missing.
It seems that you are using the version, which does not have the patch
I sent on March 18, 2009 for RTEMS 4.9.1.  Please see
http://www.aps.anl.gov/epics/tech-talk/2009/msg00491.php
The patch I sent is not applied to the RTEMS 4.9.0 or 4.9.1 branch
yet.  The staff at  RTEMS are very busy lately (e.g. office relocation
and so on).
Here is what you can do :
1) Download rtems-4.9.1.tar.bz2 from
http://www.rtems.org/ftp/pub/rtems/4.9.1/
2) Unpack the code :
bzcat rtems-4.9.1.tar.bz2 | tar xf -
3) "gunzip" the attached mvme5500-rtems4.9.1-20090401.diff.gz
and  copy the unzipped mvme5500-rtems4.9.1-20090401.diff
patch to the directory of  rtems-4.9.1/c/src/lib/libbsp/powerpc
4) cd mvme5500
5) patch -p1 < ../mvme5500-rtems4.9.1-20090401.diff
6) recompile

Another option is to download the tar file from
http://www.nsls.bnl.gov/facility/expsys/software/EPICS/

Hope this helps.

Cheers,
Kate


Thanks,
Ralph


Copyright Motorola Inc. 1999-2003, All Rights Reserved
MOTLoad RTOS Version 2.0,  PAL Version 1.1 RM02
Tue Jul  1 16:34:30 MST 2003

MPU-Int Clock Speed  =1000MHz
MPU-Ext Clock Speed  =133MHz
MPU-Number/Type = 0/MPC7455
MPU-Int Cache(L2) = 256K, Enabled, L2CR =0xC0000000
MPU-Ext Cache(L3) =  2MB, Enabled, 200MHz, L3CR =0xDF826000

Reset/Boot Vector    =Flash1

Local Memory Found   =20000000 (&536870912)
User Download Buffer =005C3000:007C2FFF
MVME5500> tftpGet -a4000000 -c193.149.12.200 -s193.149.12.29 -g0.0.0.0 -m255.255.255.0 -d/dev/ene 
t1 -f/opt/tftpdir/IOC/rtems/rtems-4.9/rtems5500TestHarness.boot
Network Loading from: /dev/enet1
Loading File: /opt/tftpdir/IOC/rtems/rtems-4.9/rtems5500TestHarness.boot 
Load Address: 04000000

Client IP Address      = 193.149.12.200
Server IP Address      = 193.149.12.29
Gateway IP Address     = 193.149.12.253
Subnet IP Address Mask = 255.255.255.0

Network File Load in Progress...

Bytes Received =&1109192, Bytes Loaded =&1109192
Bytes/Second   =&1109192, Elapsed Time =1 Second(s)
MVME5500> netShut
/dev/enet0                 Disabled
/dev/enet1                 Disabled
MV-----------------------------------------
Welcome to rtems-4.9.0(PowerPC/Generic (classic FPU)/mvme5500) on MVME5500-0163 
-----------------------------------------
Now BSP_mem_size = 0x1FE00000
rtems_configuration_get_work_space_size() = 1F4000
Universe II PCI-VME bridge detected at 0x82000000, IRQ 12
Universe Master Ports:
Port  VME-Addr   Size       PCI-Adrs   Mode:
0:    0x20000000 0x0E000000 0x90000000 A32, D64 [MBLT], Dat, Sup
1:    0x00000000 0x00FF0000 0x9F000000 A24, D64 [MBLT], Dat, Sup
2:    0x00000000 0x00010000 0x9FFF0000 A16, D64, Dat, Sup
Universe Slave Ports:
Port  VME-Addr   Size       PCI-Adrs   Mode:
0: 0x90000000 0x1FE00000 0x00000000 A32, Pgm, Dat, Sup, Usr, PWEN, PREN 
Overriding main IRQ line PCI info with 76
Registering /dev/console as minor 0 (==/dev/ttyS0)

***** Initializing network *****
Startup.

Ethernet driver name wmG unit 1
Copyright (c) 2004,2005 S. Kate Feng <[email protected]> (RTEMS/mvme5500 port) 
Can't set default route: Network is unreachable
***** Initializing NFS *****
RTEMS-RPCIOD $Release$, Till Straumann, Stanford/SLAC/SSRL 2002, See LICENSE file for licensing i 
nfo.
RTEMS-NFS $Release$, Till Straumann, Stanford/SLAC/SSRL 2002, See LICENSE file for licensing info 
.
Trying to mount 193.149.12.29:/srv/IOC on /srv/IOC
***** Starting EPICS application *****








***** epicsEventTest *****
1..11
ok  1 - epicsEventWaitWithTimeout(event, 0.0) = 1
ok  2 - epicsEventWaitWithTimeout(event, 1.0) = 1
ok  3 - epicsEventTryWait(event) = 1
ok  4 - epicsEventWaitWithTimeout(event, 1.0) = 0
ok  5 - epicsEventWaitWithTimeout(event, DBL_MAX) = 0
ok  6 - epicsEventTryWait(event) = 0
# consumer: starting
# producer 0: starting
# producer 1: starting
# producer 2: starting
# setting quit
ok  7 - consumer: errors = 0
ok  8 - producer 2: errors = 0
ok  9 - producer 0: errors = 0
ok 10 - producer 1: errors = 0
# epicsEventWaitWithTimeout(1.000000)  delay error 0.000000 sec
# epicsEventWaitWithTimeout(0.500000)  delay error 0.000000 sec
# epicsEventWaitWithTimeout(0.250000)  delay error 0.010000 sec
# epicsEventWaitWithTimeout(0.125000)  delay error 0.005000 sec
# epicsEventWaitWithTimeout(0.062500)  delay error 0.002500 sec
# epicsEventWaitWithTimeout(0.031250)  delay error 0.011250 sec
# epicsEventWaitWithTimeout(0.015625)  delay error 0.004375 sec
# epicsEventWaitWithTimeout(0.007812)  delay error 0.012188 sec
# epicsEventWaitWithTimeout(0.003906)  delay error 0.016094 sec
# epicsEventWaitWithTimeout(0.001953)  delay error 0.018047 sec
# epicsEventWaitWithTimeout(0.000977)  delay error 0.019023 sec
# epicsEventWaitWithTimeout(0.000488)  delay error 0.019512 sec
# epicsEventWaitWithTimeout(0.000244)  delay error 0.019756 sec
# epicsEventWaitWithTimeout(0.000122)  delay error 0.019878 sec
# epicsEventWaitWithTimeout(0.000061)  delay error 0.019939 sec
# epicsEventWaitWithTimeout(0.000031)  delay error 0.019969 sec
# epicsEventWaitWithTimeout(0.000015)  delay error 0.019985 sec
# epicsEventWaitWithTimeout(0.000008)  delay error 0.019992 sec
# epicsEventWaitWithTimeout(0.000004)  delay error 0.019996 sec
# epicsEventWaitWithTimeout(0.000002)  delay error 0.019998 sec
# epicsEventWaitWithTimeout(0.000000)  delay error 0.000000 sec
ok 11 - Average error 0.013214 sec

   Results
   =======
      Tests: 11
     Passed:  11 = 100.00%
In BSP_exceptionHandler()
Uhuuuh, Exception 10 in unknown task???
        Next PC or Address of fault = C4A08, Mvme5500 Saved MSR = B032
        R0  = 00000AE0 R1  = 1FC27B10 R2  = 00000000 R3  = 001AC778
        R4  = 001075DC R5  = 000A2864 R6  = 00110000 R7  = 000AA6B4
        R8  = 001AC748 R9  = 474E5543 R10 = 432B2B00 R11 = 0010EC60
        R12 = 24400022 R13 = 001164D0 R14 = 00000000 R15 = 00000000
        R16 = 00000000 R17 = 00000000 R18 = 00000000 R19 = 00000000
        R20 = 00000000 R21 = 1FC288C8 R22 = 1FC288F0 R23 = 000EA970
        R24 = 000EA9E4 R25 = 1FC28914 R26 = 00122870 R27 = 0000D45C
        R28 = 00000023 R29 = 001AC778 R30 = 00114B14 R31 = FFFFFFFF
        CR  = 24400022
        CTR = 00000000
        XER = 00000000
        LR  = 000A237C
        DAR = 00000000
Stack Trace:
 IP: 0x000C4A08, LR: 0x000A237C
--^ 0x000A237C--^ 0x000A29F8--^ 0x000A2A18--^ 0x0000D208--^ 0x0000D47C
--^ 0x0001831C--^ 0x000104B0--^ 0x0000330C--^ 0x0002A6AC--^ 0x00067330
--^ 0x000672B0
unrecoverable exception!!! task 0A010001 suspended

***** epicsExceptionTest *****
1..4
References:
Failing EPICS Test Harness on RTEMS-mvme5500 Ralph Lange
Re: Failing EPICS Test Harness on RTEMS-mvme5500 Kate Feng
Re: Failing EPICS Test Harness on RTEMS-mvme5500 Ralph Lange
Navigate by Date:
Prev: Re: Failing EPICS Test Harness on RTEMS-mvme5500 Ralph Lange
Next: Re: some field definitions Andrew Johnson
Index: 2002  2003  2004  2005  2006  2007  2008  <2009>  2010  2011  2012  2013  2014  2015  2016  2017  2018  2019  2020  2021  2022  2023  2024 
Navigate by Thread:
Prev: Re: Failing EPICS Test Harness on RTEMS-mvme5500 Ralph Lange
Next: 3.14.10 on native win32: make problem Ralph Lange
Index: 2002  2003  2004  2005  2006  2007  2008  <2009>  2010  2011  2012  2013  2014  2015  2016  2017  2018  2019  2020  2021  2022  2023  2024