RE: [Fault-injection-developer] fith
Status: Alpha
Brought to you by:
rustyl
Menu
▾
▴
RE: [Fault-injection-developer] fith
|
From: Gao, K. <kev...@in...> - 2003-04-10 00:59:01
|
OK, here is the description of most important e100 test codes.
chg_tcb_co_00 //Set both 'c' and 'ok' bit of TCB to 0
//when driver reads SCB status register.
chg_tcb_co_01 //Set 'c' bit to 0, and 'ok' bit to 1
//when driver reads SCB status register.
chg_tcb_co_10 //Set 'c' bit to 1, and 'ok' bit to 0
//when driver reads SCB status register.
chg_tcb_u //Set 'u' bit of TCB to 1, when driver reads
//SCB status register.
link-down.fsml //Simulate plug in and out network cable.
set_tcb_co_01 //Set 'c' bit to 0, and 'ok' bit to 1 when driver=20
//writes CU_RESUME to SCB command register.
set_tcb_co_10 //Set 'c' bit to 1, and 'ok' bit to 0 when driver
//writes CU_RESUME to SCB command register.
set_tcb_co_11 //Set both 'c' and 'ok' bit of TCB to 1 when
//driver writes CU_RESUME to SCB command register.
set_tcb_i //Set 'i' bit of TCB to 1 when driver writes
//CU_RESUME to SCB command register.
set_tcb_u //Set 'u' bit of TCB to 1 when driver writes
//CU_RESUME to SCB command register.
set_rfd_align //Set 'align' bit of RFD to 1 when driver reads
//SCB status register. (A receve package int is received)
set_rfd_co_00 //Set both 'c' and 'ok' bit of RFD to 0 when
//driver reads SCB status register.
set_rfd_co_01 //Set 'c' to 0, and 'ok' to 1 when driver reads
//SCB status register.
set_rfd_co_10 //Set 'c' to 1, and 'ok' to 0 when driver reads
//SCB status register.
set_rfd_crc //Set 'crc' bit of RFD to 1 when driver reads=20
//SCB status register.
set_rfd_fts //Set 'Frame too short' bit of RFD to 1 when
//driver reads SCB status register.
set_rfd_roob //Set 'Ran out of buffer space' bit of RFD to 1
//when driver reads SCB status register.
set_rfd_s //Set 's' bit of RFD to 1 when driver reads SCB
//status register.
Other test suite is only capture and change CSR register, they
are very simple.
Thanks
-Kevin
-----Original Message-----
From: arun [mailto:ar...@np...]=20
Sent: 2003?4?9? 11:21
To: Gao, Kevin
Cc: fau...@li...
Subject: Re: [Fault-injection-developer] fith
Hi Kevin,=20
Trying to identify test cases which can be tested using the fith =
tool... It will be=20
helpful as a reference , if there is a descriptive test cases written =
for e100 driver=20
(for wich the scripts are available in fault_set/e100 dir)... If any =
such test-case=20
document is available, please pass on...=20
Thanks=20
-arun=20
"Gao, Kevin" wrote:=20
Hi Arun, I want to know whether you do the following task before =
using ficl tools. 1. mount the sysfs to directory "/sys" 2. Before =
you run "ficl -i scb_cu_resume.fsml " The scb_cu_resume.ko should be =
inserted which contains the codesegment used by test fsml file.-Kevin =
-----Original Message-----=20
From: Arun Prasad [mailto:ar...@ne...]=20
Sent: 2003?3?20? 15:03=20
To: Gao, Kevin=20
Cc: fau...@li...=20
Subject: Re: [Fault-injection-developer] fith=20
=20
Hi Kevin,=20
I got some problems when I use the FICL command to insert the fsml =
script... Below I have pasted the info about it...=20
=20
Did the following fix to compile the FITH_UTILITY....=20
ficl/Makefile:=20
<<LDFLAGS =3D=20
>> LDFLAGS =3D$(AM_LDFLAGS)=20
Without the above fix, bfd is not linked and was getting link errors=20
ficl_cmd.c:=20
696 // dirp =3D opendir(TRIGGER_PATH); (commented this line)=20
697 struct dirent *dp;=20
698 int fd =3D 0;=20
699 char cmd[128];=20
700=20
701 dirp =3D opendir(TRIGGER_PATH); (Added this line)=20
********************************************************=20
The Below is the error I get when trying to insert the fsml script using =
ficl command according to your instructions in the previous mail....=20
[root@xcast1hclt-ntl e100]# lsmod=20
Module Size Used by=20
e100 60900 0=20
fi_pf 4672 0=20
fi_core 6176 1 fi_pf=20
eepro100 21964 1=20
mii 2592 1 eepro100=20
autofs 8960 0=20
parport_pc 19620 0 [unsafe]=20
parport 23296 1 parport_pc,[unsafe]=20
[root@xcast1hclt-ntl e100]# uname -a=20
Linux xcast1hclt-ntl.co.in 2.5.63 #5 Thu Mar 20 11:52:09 IST 2003 i686 =
unknown=20
[root@xcast1hclt-ntl e100]# pwd=20
/home/arun/bkt/fith/fith_test/fault_set/e100=20
[root@xcast1hclt-ntl e100]# ficl -i scb_cu_resume.fsml=20
Inserting scb_cu_resume.fsml into fi_dm...=20
Can not open /sys/fi/trigger/ctl.=20
Can not open /sys/fi/code_segment/scb_cu_resume/ctl.=20
=20
=20
Thanks again=20
-arun=20
=20
=20
=20
"Gao, Kevin" wrote:=20
The tree linux-2.5 is out of date.=20
So please dowload the newest tree at=20
http://fau...@fa.../2.5-fi-extern=20
The new tree includes all interceptors.=20
The usage of FITH_TOOL is as follow.=20
We still use tools "ficl"=20
You can use parameters as follow:=20
--attach, -a <modname,file_path>=20
attach a module into dbp.=20
--detach, -d=20
detach all modules that attached to dbp.=20
--insert, -i <file.fsml>=20
insert a faultset written in file.fsml into FITH.=20
--remove, -r {file.fsml|all}=20
remove the faultset written in file.fsml or all faultset.=20
--help, -h=20
help information.=20
thanks=20
-Kevin=20
-----Original Message-----=20
From: Arun Prasad [mailto:ar...@ne...]=20
Sent: 2003?3?18? 17:17=20
To: fau...@li...=20
Subject: [Fault-injection-developer] fith=20
Hi All,=20
To use the fith tool, downloaded the Bitkeeper tool and=20
cloned the=20
below repositories...=20
* fith-tool=20
* linux-2.5=20
both from fault-injection project.=20
Now I have to compile and load the "linux-2.5" (above) and=20
install=20
"fith-tool" on it... Is my assumption right... Because I=20
could see=20
"2.5-kprobes" in the list of repositories which is again a=20
kernel tree....=20
It will be helpful if we know the significance of each=20
repository..=20
Please clarify this....=20
Is there any upto date manual to describe the ficl commands=20
and for=20
general usage of the tool....=20
Thanks=20
-arun=20
-------------------------------------------------------=20
This SF.net email is sponsored by: Does your code think in ink?=20
You could win a Tablet PC. Get a free Tablet PC hat just for playing.=20
What are you waiting for?=20
http://ads.sourceforge.net/cgi-bin/redirect.pl?micr5043en=20
_______________________________________________=20
Fault-injection-developer mailing list=20
Fau...@li...=20
https://lists.sourceforge.net/lists/listinfo/fault-injection-developer=20
-------------------------------------------------------=20
This SF.net email is sponsored by: Does your code think in ink?=20
You could win a Tablet PC. Get a free Tablet PC hat just for playing.=20
What are you waiting for?=20
http://ads.sourceforge.net/cgi-bin/redirect.pl?micr5043en=20
_______________________________________________=20
Fault-injection-developer mailing list=20
Fau...@li...=20
https://lists.sourceforge.net/lists/listinfo/fault-injection-developer
|
