Menu
▾
▴
openocd-devel
|
From: Oleksandr T. <go...@bl...> - 2010-08-15 03:59:03
|
Here is initial version of AVR32 support for openocd: http://people.freebsd.org/~gonzo/openocd/avr32.diff So far it supports: - halt/resume - registers inspection - memory inspection/modification I'm still getting up to speed with OpenOCD internals and AVR32 so code is a little bit messy and I'd appreciate any feedback. Config file for my configuration: http://people.freebsd.org/~gonzo/openocd/avr32.cfg |
|
From: David B. <da...@pa...> - 2010-08-15 04:36:21
|
--- On Sat, 8/14/10, Oleksandr Tymoshenko <go...@bl...> wrote: > Here is initial version of AVR32 > support for openocd: Cool! would you happen to know the latest story on mainlining binutils, GCC, and GDB support for that arch? Last I heard, Atmel's code was public but not (for some silly-ass reasons that weren't public) headed any time soon to mainline. That's been true for four or more years now... AVR32 has some virtues, but GCC/etc support has been weak... |
|
From: Øyvind H. <oyv...@zy...> - 2010-08-15 19:49:13
|
Looks like a good start to me. Any objections to committing at this point so it can serve as a starting point for collaboration? Some comments: - I believe error propagation is missing in some places - should there not be a timeout for some of the operations? -- Øyvind Harboe US toll free 1-866-980-3434 / International +47 51 63 25 00 http://www.zylin.com/zy1000.html ARM7 ARM9 ARM11 XScale Cortex JTAG debugger and flash programmer |
|
From: Oleksandr T. <go...@bl...> - 2010-08-15 21:53:53
|
On 2010-08-15, at 10:49 AM, Øyvind Harboe wrote:
> Looks like a good start to me.
>
> Any objections to committing at this point so it can serve
> as a starting point for collaboration?
No objections, quite contrary :)
>
> Some comments:
>
> - I believe error propagation is missing in some places
Yes, I'll fix those places.
>
> - should there not be a timeout for some of the operations?
Yes, busy loops should have timeouts in order to prevent them
from spinning forever. It's on my TODO list.
|
|
From: Øyvind H. <oyv...@zy...> - 2010-08-15 22:00:49
|
On Sun, Aug 15, 2010 at 9:53 PM, Oleksandr Tymoshenko <go...@bl...> wrote: > > On 2010-08-15, at 10:49 AM, Øyvind Harboe wrote: > >> Looks like a good start to me. >> >> Any objections to committing at this point so it can serve >> as a starting point for collaboration? > No objections, quite contrary :) >> >> Some comments: >> >> - I believe error propagation is missing in some places > Yes, I'll fix those places. > >> >> - should there not be a timeout for some of the operations? > Yes, busy loops should have timeouts in order to prevent them > from spinning forever. It's on my TODO list. OK. I've pushed your first patch + a config file from David Brownell. I'll await patches on avr32 as work proceeds then! We find that we get a lot more testing of stuff in the master branch and as this is a new target, there are *very* slight chances of regression on other targets so it's much less work and hazzle just to make it part of the master branch at this point. -- Øyvind Harboe US toll free 1-866-980-3434 / International +47 51 63 25 00 http://www.zylin.com/zy1000.html ARM7 ARM9 ARM11 XScale Cortex JTAG debugger and flash programmer |
|
From: simon q. <sim...@gm...> - 2010-08-16 16:37:23
|
Great!! I have tried to add AVR32 support. Versaloon once can emulate JTAGICE mkII from Atmel, and can support AVR32 both under IAR EWAVR32 and AVR32 Studio. But I have removed these supports due to potential license problem -- I have no license to use JTAGICE mkII protocol. I'll buy a AVR32 demo board and test it soon. 2010/8/16 Øyvind Harboe <oyv...@zy...> > On Sun, Aug 15, 2010 at 9:53 PM, Oleksandr Tymoshenko > <go...@bl...> wrote: > > > > On 2010-08-15, at 10:49 AM, Øyvind Harboe wrote: > > > >> Looks like a good start to me. > >> > >> Any objections to committing at this point so it can serve > >> as a starting point for collaboration? > > No objections, quite contrary :) > >> > >> Some comments: > >> > >> - I believe error propagation is missing in some places > > Yes, I'll fix those places. > > > >> > >> - should there not be a timeout for some of the operations? > > Yes, busy loops should have timeouts in order to prevent them > > from spinning forever. It's on my TODO list. > > OK. I've pushed your first patch + a config file from David Brownell. > > I'll await patches on avr32 as work proceeds then! > > We find that we get a lot more testing of stuff in the master branch > and as this is a new target, there are *very* slight chances of > regression on other targets so it's much less work and hazzle just > to make it part of the master branch at this point. > > -- > Øyvind Harboe > US toll free 1-866-980-3434 / International +47 51 63 25 00 > http://www.zylin.com/zy1000.html > ARM7 ARM9 ARM11 XScale Cortex > JTAG debugger and flash programmer > _______________________________________________ > Openocd-development mailing list > Ope...@li... > https://lists.berlios.de/mailman/listinfo/openocd-development > -- Best Regards, SimonQian http://www.SimonQian.com |
|
From: Peter S. <pe...@st...> - 2010-08-16 17:01:28
|
simon qian wrote: > I have no license to use JTAGICE mkII protocol. That's a safe option, but on the other hand I don't think anyone can claim that a communications protocol must be licensed. Clean-room reverse engineering might be another safe option. //Peter |
|
From: David B. <da...@pa...> - 2010-08-16 18:44:56
|
> > I have no license to use JTAGICE mkII protocol. Irrelevant. AVR32 uses the Nexus protocols, layered on top of JTAG. Those are public and need no licensing. |
|
From: Michel C. <mic...@gm...> - 2010-08-17 02:24:29
|
On 16/08/2010 12:44, David Brownell wrote: > >>> I have no license to use JTAGICE mkII protocol. >>> > Irrelevant. AVR32 uses the Nexus protocols, > layered on top of JTAG. Those are public and > need no licensing. > > For devices where you need most of the pins Nexus can't be used. Nexus has some real value when you use a large pin count device and don't need most of the pins of the devices. The only AVR32 devices likely to stay on the market have 48 pins to 144 pins. Even with the 144 pins it is often that we need most pins and at times look at perhaps use the single wire debugging to free more pins. Michel -- Tired of Microsoft's rebootive multitasking? then it's time to upgrade to Linux. http://home.comcast.net/~mcatudal |
|
From: David B. <da...@pa...> - 2010-08-17 03:54:49
|
--- On Mon, 8/16/10, Michel Catudal <mic...@gm...> wrote: > > Irrelevant. AVR32 uses the Nexus protocols, > > layered on top of JTAG. Those are public and > > need no licensing. > > > > > > For devices where you need most of the pins Nexus can't be > used. Not wholly true. Trace needs lots of pins, yes. Basic debug only needs JTAG layers. Same with ARM: but Nexus trace has more parallel wiring options. There was some other Nexus option that chewed up pins too. Emphasis: option. |
|
From: Michel C. <mic...@gm...> - 2010-08-17 05:14:13
|
On 16/08/2010 21:54, David Brownell wrote: > > --- On Mon, 8/16/10, Michel Catudal<mic...@gm...> wrote: > > >>> Irrelevant. AVR32 uses the Nexus protocols, >>> layered on top of JTAG. Those are public and >>> need no licensing. >>> >>> >>> >> For devices where you need most of the pins Nexus can't be >> used. >> > Not wholly true. Trace needs lots of pins, yes. > Basic debug only needs JTAG layers. Same with ARM: > but Nexus trace has more parallel wiring options. > > There was some other Nexus option that chewed up > pins too. Emphasis: option. > > > The Nexus connector on my AT32UC3C-EK board uses 16 pins while the JTAG connector uses 6 pins. If I had a device with over 200 pins I may not care but a device with 144 pins or less, no way. Michel -- Tired of Microsoft's rebootive multitasking? then it's time to upgrade to Linux. http://home.comcast.net/~mcatudal |
|
From: simon q. <sim...@gm...> - 2010-08-31 17:18:14
|
My AVR32 debugger/programmer will be released next week. I'll get some sample(UC3B0256) from Atmel this week. I think I can do the test after next week. 2010/8/17 Michel Catudal <mic...@gm...> > On 16/08/2010 21:54, David Brownell wrote: > >> >> --- On Mon, 8/16/10, Michel Catudal<mic...@gm...> wrote: >> >> >> >>> Irrelevant. AVR32 uses the Nexus protocols, >>>> layered on top of JTAG. Those are public and >>>> need no licensing. >>>> >>>> >>>> >>>> >>> For devices where you need most of the pins Nexus can't be >>> used. >>> >>> >> Not wholly true. Trace needs lots of pins, yes. >> Basic debug only needs JTAG layers. Same with ARM: >> but Nexus trace has more parallel wiring options. >> >> There was some other Nexus option that chewed up >> pins too. Emphasis: option. >> >> >> >> > The Nexus connector on my AT32UC3C-EK board uses 16 pins while the JTAG > connector uses 6 pins. > If I had a device with over 200 pins I may not care but a device with 144 > pins or less, no way. > > > Michel > > > -- > Tired of Microsoft's rebootive multitasking? > then it's time to upgrade to Linux. > http://home.comcast.net/~mcatudal > > _______________________________________________ > Openocd-development mailing list > Ope...@li... > https://lists.berlios.de/mailman/listinfo/openocd-development > -- Best Regards, SimonQian http://www.SimonQian.com |
|
From: Michel C. <mic...@gm...> - 2010-09-01 01:52:50
|
On 31/08/2010 11:18, simon qian wrote: > My AVR32 debugger/programmer will be released next week. > I'll get some sample(UC3B0256) from Atmel this week. > I think I can do the test after next week. > > Are you planning to support the UC3Cxxxx series as well? They are +5V devices and not +3.3V devices as is the UC3B0256 Michel -- Tired of Microsoft's rebootive multitasking? then it's time to upgrade to Linux. http://home.comcast.net/~mcatudal |
|
From: simon q. <sim...@gm...> - 2010-09-01 08:53:39
|
It should be same for UC3C series, because the port is 5V tolarent. I'm using KiCAD, eagle has a limitation on the size of the PCB. I have released the AVR32 code, but it will be tested later this week. http://code.google.com/p/vsprog/source/browse/trunk/dongle/firmware/STK500v2 http://code.google.com/p/vsprog/source/browse/trunk/dongle/firmware/STK500v2/Target/AVR32_JTAG 2010/9/1 Michel Catudal <mic...@gm...> > On 31/08/2010 11:18, simon qian wrote: > >> My AVR32 debugger/programmer will be released next week. >> I'll get some sample(UC3B0256) from Atmel this week. >> I think I can do the test after next week. >> >> >> > Are you planning to support the UC3Cxxxx series as well? > They are +5V devices and not +3.3V devices as is the UC3B0256 > > > > Michel > > -- > Tired of Microsoft's rebootive multitasking? > then it's time to upgrade to Linux. > http://home.comcast.net/~mcatudal > > -- Best Regards, SimonQian http://www.SimonQian.com |
|
From: Peter S. <pe...@st...> - 2010-08-16 18:46:31
|
David Brownell wrote: > > > I have no license to use JTAGICE mkII protocol. > > Irrelevant. AVR32 uses the Nexus protocols, > layered on top of JTAG. Those are public and > need no licensing. Might matter if making something that also looks identical to the original Atmel product from the PC's viewpoint. //Peter |
|
From: simon q. <sim...@gm...> - 2010-08-16 19:10:16
|
It will only matter if you use the registered mark of Atmel. It should be OK to use protocol of JTAGICE mkII. But I can't find any license for this protocol opened by Atmel. What I say is JTAGICE mkII protocol, not Nexus protocol. Of course, it's no problem to support AVR32 through nexus protocol in OpenOCD or in vsprog. 2010/8/17 Peter Stuge <pe...@st...> > David Brownell wrote: > > > > I have no license to use JTAGICE mkII protocol. > > > > Irrelevant. AVR32 uses the Nexus protocols, > > layered on top of JTAG. Those are public and > > need no licensing. > > Might matter if making something that also looks identical to the > original Atmel product from the PC's viewpoint. > > > //Peter > _______________________________________________ > Openocd-development mailing list > Ope...@li... > https://lists.berlios.de/mailman/listinfo/openocd-development > -- Best Regards, SimonQian http://www.SimonQian.com |
