fault-injection-developer — Developer discussions about creating a fault injection test harness.

Re: [Fault-injection-developer] Couple of issues downloading and building the FITH source code

[Rob R. <err...@li...>]

OK, got the source, and I'm now going through the build steps and
reading the documentation. In short, its way too TLT specific for a
general announcement on the LKML.

I haven't attempted to build it yet, but this is what I see so far that
needs to be fixed before the project is announced (in any form) to the
LKML:

1. Remove all mention & dependencies on TLT, carrierlinux, etc. from
everything. The README file is full of TLT & carrierlinux 'isms. If this
is going to be posted on LKML, the instructions and code had better work
on a stock kernel by applying relevant kernel patches. 

If you rely on the TLT or CGLE kernel no one on the LKML, outside of
OSDL developers, will take a second look at it.

The instructions need to be written so anyone can take a stock kernel
apply the relevant patches and make it work. For example, dprobes is
required, but the instruction don't clearly state that you need to apply
the patch to the stock kernel sources.

2. You need to break-out the required kernel patches from the user-space
components. Separate source tarballs & separate instructions, might be
one way to do it. Right now it's confusing from reading the README what
exactly you need to do in order to get going. At the very least you need
to build a new kernel with the right patches applied, and then make the
user-space components.

3. A lot of the info that's in the README should be in the INSTALL file.
The INSTALL file has good info, but is pretty much generic to running
./configure.

4. To get the best exposure on the LKML, it would be better to also
provide instructions on how to make it work on the 2.5 kernel. I would
think that would be just a matter of getting the patches working on the
2.5 kernel. This could be any area where we ask for community help in
making it work on 2.5. This shouldn't stop us from announcing, but after
listening to Greg K-H yesterday and my past experience this is the
conclusion I come to.

More comments to follow...stay tuned...

-RobR


On Thu, 2002-10-17 at 13:25, Rob Rhoads wrote:
> I'm attempting to download the source to FITH for building, but it
> appears that it is only available via checking out directly from CVS. It
> would be nice if you post a tarball of the source code to both the
> project web page and under the file section on the project page.
> 
> Also make me a member of the project (my SF account errhoads), so I
> don't have to monkey around with trying to get anonymous :pserver to
> work around the our stupid corporate firewall.
> 
> Thanks.
> 
> -- 
> Rob Rhoads                     mailto:err...@li...
> Telecom Software Platforms     office: 503-677-5498
> Intel Communications Group
> 
> <disclaimer value="pointless">
> This email message solely contains my own personal views, 
> and not necessarily those of my employer.
> </disclaimer>
> 
> 
> 
> -------------------------------------------------------
> This sf.net email is sponsored by: viaVerio will pay you up to
> $1,000 for every account that you consolidate with us.
> http://ad.doubleclick.net/clk;4749864;7604308;v?
> http://www.viaverio.com/consolidator/osdn.cfm
> _______________________________________________
> Fault-injection-developer mailing list
> Fau...@li...
> https://lists.sourceforge.net/lists/listinfo/fault-injection-developer

[Fault-injection-developer] Couple of issues downloading and building the FITH source code

[Rob R. <err...@li...>]

I'm attempting to download the source to FITH for building, but it
appears that it is only available via checking out directly from CVS. It
would be nice if you post a tarball of the source code to both the
project web page and under the file section on the project page.

Also make me a member of the project (my SF account errhoads), so I
don't have to monkey around with trying to get anonymous :pserver to
work around the our stupid corporate firewall.

Thanks.

-- 
Rob Rhoads                     mailto:err...@li...
Telecom Software Platforms     office: 503-677-5498
Intel Communications Group

<disclaimer value="pointless">
This email message solely contains my own personal views, 
and not necessarily those of my employer.
</disclaimer>

[Fault-injection-developer] About fsml.dtd

[Wang, C. <che...@in...>]

Dear all:
When I user "dtdpatser" to parse fsml.dtd, the following error encount:
"Duplicated ATTLIST!"
And it worked perproly after I merged all attributes into one ATTLIST.
Is it a bug?

BTW:
"KATE" is so wounderful for edit XML file.


Stanley Wang 
SW Engineer, Intel Corporation.
Intel China Software Lab. 
Tel: 021-52574545 ext. 1171 
iNet: 8-752-1171 
 
Opinions expressed are those of the author and do not represent Intel
Corporation

[Fault-injection-developer] Write fsml

[Wang, C. <che...@in...>]

A convenience way to write a fsml file :)

http://www.linuxfocus.org/ChineseGB/May2002/article201.shtml

Stanley Wang 
SW Engineer, Intel Corporation.
Intel China Software Lab. 
Tel: 021-52574545 ext. 1171 
iNet: 8-752-1171 
 
Opinions expressed are those of the author and do not represent Intel
Corporation

[Fault-injection-developer] FITH 1.2 is released!

[Zhuang, L. <lou...@in...>]

Dear all,
          After one and a half month hard working, we are proud to announce
Fault Injection Test Harness (FITH) 1.2. The new release includes several
new features and enhancements and easier-to-use than before. Please see
release notes below:
 
-------------------------------- Release Notes
---------------------------------------------------------
New Features:
1.	Support injecting faultset Anytime
2.	Support arbitrary numbers of faultset
3.	Support injecting writing faults in IO space
 
Enhancements:
1.	FITH setup only need one step.
2.	Several facility scripts make manipulation much easier than before
3.	Revised user manual and README
 
Yours sincerely,
Louis Zhuang
Intel China Software Lab.
 

[Fault-injection-developer] Recall: FITH 1.2 is released!

[Zhuang, L. <lou...@in...>]

Zhuang, Louis would like to recall the message, "FITH 1.2 is released!".

[Fault-injection-developer] FITH 1.2 is released!

[Zhuang, L. <lou...@in...>]

Dear all,
          After one and a half month hard working, we are proud to announce
Fault Injection Test Harness (FITH) 1.2. The new release includes several
new features and enhancements and easier-to-use than before. Please see
release notes below:
 
-------------------------------- Release Notes
---------------------------------------------------------
New Features:
1.	Support injecting faultset Anytime
2.	Support arbitrary numbers of faultset
3.	Support injecting writing faults in IO space
 
Enhancements:
1.	FITH setup only need one step new.
2.	Several facility scripts make manipulation much easier than before
3.	Revised user manual and README
 
Yours sincerely,
Louis Zhuang
Intel China Software Lab.
 

[Fault-injection-developer] Update doc

[Wang, C. <che...@in...>]

fith.texi updated!

Stanley Wang SW Engineer, Intel Corporation.

Opinions expressed are those of the author and do not represent Intel
Corporation

Intel China Software Lab. 
Tel: 021-52574545 ext. 1171 
iNet: 8-752-1171 

 

[Fault-injection-developer] Bug fixed :)

[Wang, C. <che...@in...>]

The bug of can not parse "prefetchnta" in libdisasm fixed :)
I've sent the patch to the developer.
 

Stanley Wang SW Engineer, Intel Corporation.

Opinions expressed are those of the author and do not represent Intel
Corporation

Intel China Software Lab. 
Tel: 021-52574545 ext. 1171 
iNet: 8-752-1171 

 

[Fault-injection-developer] Funny example for e100.o

[Zhuang, L. <lou...@in...>]

The FI will make e100.o crazy to be down and up again and again=A1=AD =
:-) please
take a look in example/E100
=20
Yours sincerely,
Louis Zhuang
Intel China Software Lab.
=20

[Fault-injection-developer] Shell script for clean up fith

[Wang, C. <che...@in...>]

Add "fith_clean" to remove fith's module after completing FI testing.

Yours Sincerely, 
Stanley Wang 

Intel China Software Lab. 
Tel: 021-52574545 ext. 1171 
iNet: 8-752-1171 

 

RE: [Fault-injection-developer] fith.info updated.

[Wang, C. <che...@in...>]

fith.texi updated.
-----Original Message-----
From: Zhuang, Louis [mailto:lou...@in...]
Sent: 2002=C4=EA8=D4=C215=C8=D5 10:08
To: 'fau...@li...'
Subject: RE: [Fault-injection-developer] fith.info updated.


User manual updated.
Detail explains for IRQ faultset.
Minor fixes.
=20
  --Louis
=20
-----Original Message-----
From: Wang, Chenguang [mailto:che...@in...]
Sent: Thursday, August 15, 2002 10:03 AM
To: 'fau...@li...'
Subject: [Fault-injection-developer] fith.info updated.
=20
Synchronized fith.texi with fith_user_manual.
Yours Sincerely,=20
Stanley Wang=20
Intel China Software Lab.=20
Tel: 021-52574545 ext. 1171=20
iNet: 8-752-1171=20
=20

[Fault-injection-developer] fith.info updated.

[Wang, C. <che...@in...>]

Synchronized fith.texi with fith_user_manual.

Yours Sincerely, 
Stanley Wang 

Intel China Software Lab. 
Tel: 021-52574545 ext. 1171 
iNet: 8-752-1171 

 

[Fault-injection-developer] updated fith_user_manual.doc

[Zhuang, L. <lou...@in...>]

TODO:
        step-by-step instructions for examples
 <<fith_user_manual.ZIP>> 

[Fault-injection-developer] RE: [Fault_injection] FW: About FAQ of FITH

[Zhuang, L. <lou...@in...>]

1. Why using command "ficl -v", the only info be printed is "verbose"?
we do not give the verbose information now, the verbose information =
will
come just after...
[lyzhuang] verbose is just a trending. It means FITH trends to output =
more
verbose information only if it has. But still now, it hasn't. ?
2. What does the number storage in directory =
"/proc/fith_dm/globalcounter"
mean?
[lyzhuang] it is a message counter.
3. Why a "wp" with a byte IO access may catch a word or double word =
access?
We will intercept all access in the address region. So it is =
uninterested in
access length.
[lyzhuang] FITH will intercept all access aligned with the address. But =
FITH
only corrupts 'len' bytes of data at most.
4. When insert a faultset, does it useful or I need enable it?
As long as the faultset is validated, when insert the faultset, the =
state of
faultset is enabled.
[lyzhuang] FITH will enable the faultset just after the faultset is =
inserted
successfully.
5. When some critical issue happened that terminate the "ABAT", the =
"ABAT"
cannot use again, it will always print an error information:" There is
another test script that is running!\n", how can it be used again?
Remove the hidden file "/tmp/.fithlock.*, it can be used again.
[lyzhuang] ABAT doesn't design for running parallel.  So if you run a =
script
when another script is running, error happens.=20
6.When set the two trigger with same "id", there is no warning =
information.
Is there any issue?
id is not used and not checked at this stage. So this unification check =
will
be done in future.
[lyzhuang] id is not be used. It may never be used. So... don't care =
about
it.
7. When do a mmio_readb operation, why the length information in =
ringbuffer
is still 4?
It is a bug of "gcc 2.96". Redhat released an unofficial GCC named =
"2.96".
It is a root of bugs. The last official GCC in 2.xx series is 2.95.3. =
You
should use this version or else ignore this bug.

8. How can I use parameter "bitmask" in fsml file to define a faulset?
We can use it as: final data=3D(data & ~bitmask), and check if final =
data is
between min and max.

9. Why the IO write operation is corrupted while only insert a io read
operation?
Just because dummy driver will alway do read and write operation no =
matter
we call read or write command.

10. What is the max number of faultset, trigger per faultset, op per =
trigger
is validate?
max number of faultset: 16
max number of trigger:  16
max number of op	9

11. How can we set a linear address in faultset?
We use physical address instead of linear address, FITH prefer user to =
set
physical address.[lyzhuang]  Linear address support may be cancelled in
future release.

12. Why must we remove faultset before remove dummy?
[lyzhuang] FITH place some hooks in driver and system for watching =
accesses.
If you remove watched driver before removing faultset, these hooks =
cannot be
clear elegantly. Some unpredicted error may happen.=20


  --Louis

-----Original Message-----
From: Wang, Chenguang [mailto:che...@in...]
Sent: Wednesday, August 14, 2002 8:45 AM
To: SourceforgeFI mail list (E-mail); CarrierlinuxFI mail list (E-mail)
Subject: [Fault_injection] FW: About FAQ of FITH


>  -----Original Message-----
> From:         Gao, Kevin=20
> Sent: 2002=C4=EA8=D4=C213=C8=D5 14:17
> To:   Wang, Frank; Wang, Chenguang; Zhuang, Louis
> Cc:   Duan, Jian Gang; Li, Jack; Pan, Deng; Wang, Chenguang; Wang, =
Frank;
> Yu, Jerry; Zhuang, Louis
> Subject:      About FAQ of FITH
>
> There are the samples I collected. I think some answer I give is not
> clearly, and some I cannot give a good answer. Louis can you help me?
> thanks!!
> Any FAQ samples or comment?
> thank you !^_^
>
> 1. Why using command "ficl =3Dv", the only info be printed is =
"verbose"?
> we do not give the verbose information now, the verbose information =
will
> come just after...
>
> 2. What does the number storaged in directory
> "/proc/fith_dm/globalcounter"  mean?
>
> 3. Why a "wp" with a byte IO access may catch a word or double word
> access?
> We will intercept all access in the address region. So it is =
uninterested
> in access length.
>
> 4. When insert a faultset, does it useful or I need enable it?
> As long as the faultset is validated, when insert the faultset, the =
state
> of faultset is enabled.
>
> 5. When some critical issue happened that termate the "ABAT", the
> "ABAT" cannot use again, it will always print a error information:"
> There is another test script that is running!\n", how can it be used
> again?
> Remove the hidden file "/tmp/.fithlock.*, it can be used again.
>
> 6.When set the two trigger with same "id", there is no warning
> information. Is there any issue?
> id is not used and not checked at this stage. So this unification =
check
> will be done in future.
>
> 7. When do a mmio_readb operation, why the length infomation in =
ringbuffer
> is still 4?
> It is a bug of "gcc 2.96". Redhat released an unoffical GCC named =
"2.96".
> It is a root of bugs. The last offical GCC in 2.xx series is 2.95.3. =
You
> should use this version or else ignore this bug.
>
> 8. How can I use parameter "bitmask" in fsml file to define a =
faulset?
> We can use it as: finaldata=3D(data & ~bitmask), and check if =
finaldata is
> between min and max.
>
> 9. Why the io write operation is corrupted while only insert a io =
read
> operation?
> Just because dummy driver will alway do read and write operation no =
matter
> we call read or write command.
>
> 10. What is the max number of faultset, trigger per faultset, op per
> trigger is validate?
> max faultset: 16
> max trigger:  16
> max op        9
>
> 11. How can we set a liner address in faultset?
> We use physical address instead of liner address, FITH prefer user to =
set
> physical address.
>
> 12. Why must we remove faultset before remove dummy?
>
>
>
>
>
> Yours Sincerely,
> Gao, Xiaohe (Kevin)
>
> Intel China Software Lab.
> Tel: 021-52574545 ext. 1271
> iNet: 8-752-1271
>
>
_______________________________________________
Fault_injection mailing list
Fau...@ca...
http://www.carrierlinux.org/mailman/listinfo/fault_injection

[Fault-injection-developer] FW: About FAQ of FITH

[Wang, C. <che...@in...>]

>  -----Original Message-----
> From: 	Gao, Kevin =20
> Sent:	2002=C4=EA8=D4=C213=C8=D5 14:17
> To:	Wang, Frank; Wang, Chenguang; Zhuang, Louis
> Cc:	Duan, Jian Gang; Li, Jack; Pan, Deng; Wang, Chenguang; Wang, =
Frank;
> Yu, Jerry; Zhuang, Louis
> Subject:	About FAQ of FITH
>=20
> There are the samples I collected. I think some answer I give is not
> clearly, and some I cannot give a good answer. Louis can you help me?
> thanks!!
> Any FAQ samples or comment?
> thank you !^_^
>=20
> 1. Why using command "ficl =3Dv", the only info be printed is =
"verbose"?
> we do not give the verbose information now, the verbose information =
will
> come just after...
>=20
> 2. What does the number storaged in directory
> "/proc/fith_dm/globalcounter"  mean?
>=20
> 3. Why a =A1=B0wp=A1=B1 with a byte IO access may catch a word or =
double word
> access?
> We will intercept all access in the address region. So it is =
uninterested
> in access length.
>=20
> 4. When insert a faultset, does it useful or I need enable it?
> As long as the faultset is validated, when insert the faultset, the =
state
> of faultset is enabled.
>=20
> 5. When some critical issue happened that termate the =
=A1=B0ABAT=A1=B1, the
> =A1=B0ABAT=A1=B1 cannot use again, it will always print a error =
information:=A1=B1
> There is another test script that is running!\n=A1=B1, how can it be =
used
> again?
> Remove the hidden file =A1=B0/tmp/.fithlock.*, it can be used again.
>=20
> 6.When set the two trigger with same =A1=B0id=A1=B1, there is no =
warning
> information. Is there any issue?
> id is not used and not checked at this stage. So this unification =
check
> will be done in future.
>=20
> 7. When do a mmio_readb operation, why the length infomation in =
ringbuffer
> is still 4?
> It is a bug of "gcc 2.96". Redhat released an unoffical GCC named =
"2.96".
> It is a root of bugs. The last offical GCC in 2.xx series is 2.95.3. =
You
> should use this version or else ignore this bug.
>=20
> 8. How can I use parameter =A1=B0bitmask=A1=B1 in fsml file to define =
a faulset?
> We can use it as: finaldata=3D(data & ~bitmask), and check if =
finaldata is
> between min and max.
>=20
> 9. Why the io write operation is corrupted while only insert a io =
read
> operation?
> Just because dummy driver will alway do read and write operation no =
matter
> we call read or write command.
>=20
> 10. What is the max number of faultset, trigger per faultset, op per
> trigger is validate?
> max faultset: 16
> max trigger:  16
> max op	9
>=20
> 11. How can we set a liner address in faultset?
> We use physical address instead of liner address, FITH prefer user to =
set
> physical address.
>=20
> 12. Why must we remove faultset before remove dummy?
>=20
>=20
>=20
>=20
>=20
> Yours Sincerely,
> Gao, Xiaohe (Kevin)
>=20
> Intel China Software Lab.
> Tel: 021-52574545 ext. 1271
> iNet: 8-752-1271
>=20
>=20

[Fault-injection-developer] Website is coming online

[Lynch, R. <rus...@in...>]

I have laid down the initial framework for the project web site
(http://fault-injection.sourceforge.net).  The site is database driven, but
I have not setup any of the web based administration tools yet, so editing
content (like the introduction text and news stories) still needs to be done
by hand.

When you have some spare time, take a look at the site and send content
ideas to this mailing list. Some of my ideas (not yet realized) are:

* dedicate a page that list all of the constraints for the project
* dedicate a page that list that brain storms a list off possible
  methods for injecting faults, with pros and cons for each 

(Basically, instead of producing a one document PDF white paper, make the
web site an online white paper.)

    -rusty