[U-Boot-Users] BDI vs. Lauterbach

G'Day,
I've been looking at a few of the posts regarding debugging tools, and the standard answer on this list appears to be "Get a BDI2000". I'm presently looking at getting a debugger to use to bring up a new board, get U-Boot going, and eventually do a lot of OS work (a new OS that we're developing - see http://www.ertos.nicta.com.au -- sorry, shameless plug).
Some people that we are working with use the Lauterbach Trace32 tools extensively, and we've had some good experiences with them. I was wondering if anyone on this list had used both (particularly while developing U-Boot), and how the BDI-2000 stacks up against the Lauterbach equivalent. (Apart from being significantly cheaper).
Any insights much appreciated.
Many thanks,
David Snowdon, Research Engineer, National ICT Australia, http://www.ertos.nicta.com.au

On Monday, April 10, 2006, David Snowdon wrote:
G'Day,
I've been looking at a few of the posts regarding debugging tools, and the standard answer on this list appears to be "Get a BDI2000". I'm presently looking at getting a debugger to use to bring up a new board, get U-Boot going, and eventually do a lot of OS work (a new OS that we're developing - see http://www.ertos.nicta.com.au -- sorry, shameless plug).
Some people that we are working with use the Lauterbach Trace32 tools extensively, and we've had some good experiences with them. I was wondering if anyone on this list had used both (particularly while developing U-Boot), and how the BDI-2000 stacks up against the Lauterbach equivalent. (Apart from being significantly cheaper).
Pros: BDI - 10Mbit eth, Lauterbach - 100 Mbit. Lauterbach scalable and simply extendable, BDI - not. Cons: BDI support gnu toolchain natively (in GDB server mode), Lauterbach - not (sometime it parsing elf/dwarf correctly, sometime, usually in critical cases :), not). And you are know, hmm, strange Lauterbach price policy: price of BDI firmware for a new CPU target is approx. 1000 eur, for the Lauterbach - price of new device.
Any insights much appreciated.
Many thanks,
David Snowdon, Research Engineer, National ICT Australia, http://www.ertos.nicta.com.au

In message 427722615.20060410170041@varma-el.com you wrote:
Pros:
meaning: pro for Lauterbach, just to make it clear.
BDI - 10Mbit eth, Lauterbach - 100 Mbit.
This does not mean anything. I haven't seen a single case where the network speed was the limiting factor. Shuffeling the data through the JTAG is usually much slower.
Lauterbach scalable and simply extendable, BDI - not.
What would you want to extend or scale?
Cons: BDI support gnu toolchain natively (in GDB server mode),
In other words: The BDI2000 fits seamlessly into a Linux based development environment. You can use exact the same tools and and user interface for low level stuff (boot loader, OS and drivers - using the BDI) and for application code (using gdbserver).
Lauterbach - not (sometime it parsing elf/dwarf correctly, sometime, usually in critical cases :), not).
Question: does the Lauterbach reliably handle issues like relocating the symbol table as needed für U-Boot?
And you are know, hmm, strange Lauterbach price policy: price of BDI firmware for a new CPU target is approx. 1000 eur, for the Lauterbach - price of new device.
Another pro for the BDI: they have *excellent* support. I know a couple of vendors of hardware and tools etc. Some of them are really good, but Abatron beats them all.
Best regards,
Wolfgang Denk

On Monday, April 10, 2006, Wolfgang Denk wrote:
In message 427722615.20060410170041@varma-el.com you wrote:
Pros:
meaning: pro for Lauterbach, just to make it clear.
Yes
BDI - 10Mbit eth, Lauterbach - 100 Mbit.
This does not mean anything. I haven't seen a single case where the network speed was the limiting factor. Shuffeling the data through the JTAG is usually much slower.
Not always. As ex.: allowable MPC5200 JTAG's clock is ... 25 MHz, clock of 10 Mbit eth is ... + tcp stack overhead. And now we run memory dump command and measure.
Lauterbach scalable and simply extendable, BDI - not.
What would you want to extend or scale?
Lauterbach like Lego, constructing from modules: from dumb LPT<->jtg upto monster-alike in-circuit emulator (I didn't talk about how _much_ smb. will pay for it, I talk about possibility).
Cons: BDI support gnu toolchain natively (in GDB server mode),
In other words: The BDI2000 fits seamlessly into a Linux based development environment. You can use exact the same tools and and user interface for low level stuff (boot loader, OS and drivers - using the BDI) and for application code (using gdbserver).
Yes.
Lauterbach - not (sometime it parsing elf/dwarf correctly, sometime, usually in critical cases :), not).
Question: does the Lauterbach reliably handle issues like relocating the symbol table as needed fur U-Boot?
I'm not sure, since I doesn't work with Lb approx. 3 years, but probably it could.
And you are know, hmm, strange Lauterbach price policy: price of BDI firmware for a new CPU target is approx. 1000 eur, for the Lauterbach - price of new device.
Another pro for the BDI: they have *excellent* support. I know a couple of vendors of hardware and tools etc. Some of them are really good, but Abatron beats them all.
Don't asking them for a help, since my BDI work as predicted, so it may be pros also :).
Best regards, Wolfgang Denk

Dear Andrey,
in message 1618325846.20060410201725@varma-el.com you wrote:
This does not mean anything. I haven't seen a single case where the network speed was the limiting factor. Shuffeling the data through the JTAG is usually much slower.
Not always. As ex.: allowable MPC5200 JTAG's clock is ... 25 MHz, clock of 10 Mbit eth is ... + tcp stack overhead. And now we run memory dump command and measure.
But the 25 MHz also includes a LOT of JTAG communication overhead. I think you won't get even close to a raw 10 Mbps data rate, but I have to admit that I never actually measured it yet. Did you really measure this? For example, what difference do you get when you connect your Lauterbach to a 10 Mbps port vs. a 100 Mbps port?
[This is a serious question; I have never been able to run such a test myself yet.]
Don't asking them for a help, since my BDI work as predicted, so it may be pros also :).
I remember a case when somebody here on the list had the data pins of his flash device swapped and Abatron came up with a special firmware that allowed flashing the devices anyway. And just recently They helped us to get support for a new processor (Intel Monahans) working - without ever seeing such a chip or much of the documentation.
Best regards,
Wolfgang Denk

On Monday, April 10, 2006, Wolfgang Denk wrote:
Dear Andrey,
in message 1618325846.20060410201725@varma-el.com you wrote:
This does not mean anything. I haven't seen a single case where the network speed was the limiting factor. Shuffeling the data through the JTAG is usually much slower.
Not always. As ex.: allowable MPC5200 JTAG's clock is ... 25 MHz, clock of 10 Mbit eth is ... + tcp stack overhead. And now we run memory dump command and measure.
But the 25 MHz also includes a LOT of JTAG communication overhead. I think you won't get even close to a raw 10 Mbps data rate, but I have to admit that I never actually measured it yet. Did you really measure this? For example, what difference do you get when you connect your Lauterbach to a 10 Mbps port vs. a 100 Mbps port? [This is a serious question; I have never been able to run such a test myself yet.]
Unfortunately, but I have not access to a Lauterbach anymore (as I say, I work with it 3 years ago), so I couldn't run this test too :(.
snip..

Wolfgang Denk schrieb:
Another pro for the BDI: they have *excellent* support. I know a couple of vendors of hardware and tools etc. Some of them are really good, but Abatron beats them all.
This is also true for Lauterbach. We are working with Lauterbach for nearly 10 years have experienced always excellent support. Implementing a new RTOS support: no problem.
If you only use a linux based development, Abatron may fit. If you use different development systems, like Windows and Linux, Lauterbach is far better.
-- Schöne Grüße aus Reutlingen
Wolfram Wadepohl Forschung & Entwicklung
E&K AUTOMATION Indumat GmbH & Co. KG

Pros: BDI - 10Mbit eth, Lauterbach - 100 Mbit. Lauterbach scalable and simply extendable, BDI - not.
1) Powerful scripting language 2) Very user friendly 3) Rich commands set that allows to debug hardware problems, too 4) Can "interpret" peripheral registers to provide human readable/changeable values 5) O.S. awarness for several o.s. and RTOSes 6) Modularity (trace module, logic analyzer module)
Cons: BDI support gnu toolchain natively (in GDB server mode), Lauterbach - not (sometime it parsing elf/dwarf correctly, sometime, usually in critical cases :), not). And you are know, hmm, strange Lauterbach price policy: price of BDI firmware for a new CPU target is approx. 1000 eur, for the Lauterbach - price of new device.
1) So far no gdb support 2) USB version doesn't support Linux on host

On Monday, April 10, 2006, llandre wrote:
Pros: BDI - 10Mbit eth, Lauterbach - 100 Mbit. Lauterbach scalable and simply extendable, BDI - not.
- Powerful scripting language
- Very user friendly
- Rich commands set that allows to debug hardware problems, too
- Can "interpret" peripheral registers to provide human
readable/changeable values 5) O.S. awarness for several o.s. and RTOSes
1-5 common for a both (to a greater or lesser extent), but Lb _demand_ that TRACE32 installed on host and may communicate _only_ with TRACE32. And, if your bumped with problems with TRACE32 (_host software_, not hardware) at the middle of annual subscription, your _must_ prolonger subscription (not for free of charge) since LB released fixed TRACE32 when your current subscription will be closed. We bumped with such situation when worked with Hitachi (Renesas) SH7751 family. TRACE32 know _nothing_ about new revision (7751R), and refuse to work at all. For a defeated of Lb, I could say that I don't know what Abatron offer in this situation, and this situation was 3 years ago. Wolfgang?
- Modularity (trace module, logic analyzer module)
I already point to it (scalable etc)
Cons: BDI support gnu toolchain natively (in GDB server mode), Lauterbach - not (sometime it parsing elf/dwarf correctly, sometime, usually in critical cases :), not). And you are know, hmm, strange Lauterbach price policy: price of BDI firmware for a new CPU target is approx. 1000 eur, for the Lauterbach - price of new device.
- So far no gdb support
- USB version doesn't support Linux on host
Since protocol _closed_ and patented (I think).

In message 1317633204.20060410215844@varma-el.com you wrote:
SH7751 family. TRACE32 know _nothing_ about new revision (7751R), and refuse to work at all. For a defeated of Lb, I could say that I don't know what Abatron offer in this situation, and this situation was 3 years ago. Wolfgang?
This probably depends on what was changed between versions, but I guess the same szenario might happen with Abatron, too. But then, the price for software updates + support is some 142 Euro per year, which makes the decision easy enough.
Best regards,
Wolfgang Denk

Wolfgang Denk wrote:
In message 1317633204.20060410215844@varma-el.com you wrote:
SH7751 family. TRACE32 know _nothing_ about new revision (7751R), and refuse to work at all. For a defeated of Lb, I could say that I don't know what Abatron offer in this situation, and this situation was 3 years ago. Wolfgang?
This probably depends on what was changed between versions, but I guess the same szenario might happen with Abatron, too. But then, the price for software updates + support is some 142 Euro per year, which makes the decision easy enough.
Thanks, you're really calm me, Abatron didn't put such information under public domain.
-- Regards Andrey

Hi all!
I have been using Lauterbach debuggers for several years at different companies and finally bought my own one (Power Debug USB for ARM7/9). I like them very much but unfortunately there doesn't seem to be RDI or GDB stub support. At least on MS Windows there is a DLL available to connect third-party products to the Lauterbach environment.
llandre schrieb:
- USB version doesn't support Linux on host
The USB version works very well on a Linux host, at least with software release ~summer 2005! First I had some problems on a Suse 9.3 Prof. system with the USB hotplug mechanism but this could be solved by a proper chmod.
With best regards Andreas Schweigstill

G'Day all,
Many thanks for all the discussion regarding BDI and Lauterbach. It sounds like the Lauterbach has the edge over the BDI in some ways, but whether its a 2x-the-cost edge is difficult to work out.
I thought I had the powers-that-be convinced on the Lauterbach until someone suggested the Wind-River and Greenhills tools (in front of management). I'm now back to the drawing board: comparing these tools via the information available on the 'net is difficult because they all claim to do the same things. Have any of you had positive or negative experiences with either of those two company's products?
Please ignore if I'm pushing the friendship too far... ;-).
Dave.
-- David Snowdon Research Engineer National ICT Australia (NICTA) Phone: +61-2-8306-0566 Email: David.Snowdon@nicta.com.au Web: http://www.nicta.com.au/ertos
On 11/04/2006, at 8:41 PM, Andreas Schweigstill wrote:
Hi all!
I have been using Lauterbach debuggers for several years at different companies and finally bought my own one (Power Debug USB for ARM7/9). I like them very much but unfortunately there doesn't seem to be RDI or GDB stub support. At least on MS Windows there is a DLL available to connect third-party products to the Lauterbach environment.
llandre schrieb:
- USB version doesn't support Linux on host
The USB version works very well on a Linux host, at least with software release ~summer 2005! First I had some problems on a Suse 9.3 Prof. system with the USB hotplug mechanism but this could be solved by a proper chmod.
With best regards Andreas Schweigstill
-- Dipl.-Phys. Andreas Schweigstill Schweigstill IT | Embedded Systems Schauenburgerstraße 116, D-24118 Kiel, Germany Phone: (+49) 431 5606-435, Fax: (+49) 431 5606-436 Mobile: (+49) 171 6921973, Web: http://www.schweigstill.de/
This SF.Net email is sponsored by xPML, a groundbreaking scripting language that extends applications into web and mobile media. Attend the live webcast and join the prime developer group breaking into this new coding territory! http://sel.as-us.falkag.net/sel? cmd=lnk&kid=110944&bid=241720&dat=121642 _______________________________________________ U-Boot-Users mailing list U-Boot-Users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/u-boot-users

In message D4267F77-3E56-4C19-84D9-7A491B07FD7C@nicta.com.au you wrote:
I thought I had the powers-that-be convinced on the Lauterbach until someone suggested the Wind-River and Greenhills tools (in front of management). I'm now back to the drawing board: comparing these tools via the information available on the 'net is difficult because they all claim to do the same things. Have any of you had positive or negative experiences with either of those two company's products?
This has been discussed before on this list. Please search the mailing list archive. Short summary: Trace32 and BDI200 play in a somewhat different league.
Best regards,
Wolfgang Denk

There's no reference to "greenhills" or "green hills" on this mailing list other than the below two posts. <shrug>.
Dave.
-- David Snowdon Research Engineer National ICT Australia (NICTA) Phone: +61-2-8306-0566 Email: David.Snowdon@nicta.com.au Web: http://www.nicta.com.au/ertos
On 12/04/2006, at 8:07 PM, Wolfgang Denk wrote:
In message D4267F77-3E56-4C19-84D9-7A491B07FD7C@nicta.com.au you wrote:
I thought I had the powers-that-be convinced on the Lauterbach until someone suggested the Wind-River and Greenhills tools (in front of management). I'm now back to the drawing board: comparing these tools via the information available on the 'net is difficult because they all claim to do the same things. Have any of you had positive or negative experiences with either of those two company's products?
This has been discussed before on this list. Please search the mailing list archive. Short summary: Trace32 and BDI200 play in a somewhat different league.
Best regards,
Wolfgang Denk
-- Software Engineering: Embedded and Realtime Systems, Embedded Linux Phone: (+49)-8142-66989-10 Fax: (+49)-8142-66989-80 Email: wd@denx.de 2000 pounds of chinese soup = 1 Won Ton

Pros: BDI - 10Mbit eth, Lauterbach - 100 Mbit. Lauterbach scalable and simply extendable, BDI - not.
Lauterbach is the best choiche for both Linux and WinCE targets. Lauterbach have *excellent* support too.
Cons: BDI support gnu toolchain natively (in GDB server mode), Lauterbach - not (sometime it parsing elf/dwarf correctly, sometime, usually in critical cases :), not). And you are know, hmm, strange Lauterbach price policy: price of BDI firmware for a new CPU target is approx. 1000 eur, for the Lauterbach - price of new device.
-- Marco Cavallini ============================================================== Koan s.a.s. - Software Engineering Linux and WinCE solutions for Embedded and Real-Time Software Klinux : the embedded distribution for industrial applications - Atmel AT91 ARM Third Party Consultant - Intel PCA Developer Network Member - Microsoft Windows Embedded Partner Via Pascoli, 3 - 24121 Bergamo - ITALIA http://www.koansoftware.com - http://www.klinux.org ==============================================================

--- Marco Cavallini arm.linux@koansoftware.com wrote:
Pros: BDI - 10Mbit eth, Lauterbach - 100 Mbit. Lauterbach scalable and simply extendable, BDI - not.
Lauterbach is the best choiche for both Linux and WinCE targets. Lauterbach have *excellent* support too.
I have introduced several of my clients to the BDI2000 and have yet to have a single one regret the selection. I have to admit though, if they lowered their price they would probably get more customers.
I don't understand why the have to charge $1500+ for a differnent cpu firmware update...
__________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com

Frank wrote:
--- Marco Cavallini arm.linux@koansoftware.com wrote:
Pros: BDI - 10Mbit eth, Lauterbach - 100 Mbit. Lauterbach scalable and simply extendable, BDI - not.
Lauterbach is the best choiche for both Linux and WinCE targets. Lauterbach have *excellent* support too.
Why best? Because best... :)
I have introduced several of my clients to the BDI2000 and have yet to have a single one regret the selection. I have to admit though, if they lowered their price they would probably get more customers.
I don't understand why the have to charge $1500+ for a differnent cpu firmware update...
Well, nice question, but now, show me modern CPU with onboard JTAG, in public datasheets of which described FULL set of debug commands, not only JTAG, but onchip debug subsystem too (we all know which cost of SoC systems with onboard usb/eth/flash). Any questions after that?
Andrey
participants (9)
-
Andreas Schweigstill
-
Andrey Volkov
-
David Snowdon
-
David Snowdon
-
Frank
-
llandre
-
Marco Cavallini
-
Wolfgang Denk
-
Wolfram Wadepohl